spatial variability of coastal organic soil characteristics

Louisiana State UniversityLSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1998

Spatial Variability of Coastal Organic SoilCharacteristics.L. Cecil DharmasriLouisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion inLSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please [email protected].

Recommended CitationDharmasri, L. Cecil, "Spatial Variability of Coastal Organic Soil Characteristics." (1998). LSU Historical Dissertations and Theses. 6824.https://digitalcommons.lsu.edu/gradschool_disstheses/6824

https://digitalcommons.lsu.edu?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6824&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool_disstheses?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6824&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6824&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool_disstheses?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6824&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool_disstheses/6824?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6824&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. U M I

films the text directly from the original or copy submitted. Thus, some

thesis and dissertation copies are in typewriter face, while others may be

from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the

copy submitted. Broken or indistinct print, colored or poor quality

illustrations and photographs, print bleedthrough, substandard margins,

and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send U M I a complete

manuscript and there are missing pages, these will be noted. Also, if

unauthorized copyright material had to be removed, a note will indicate

the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by

sectioning the original, beginning at the upper left-hand comer and

continuing from left to right in equal sections with small overlaps. Each

original is also photographed in one exposure and is included in reduced

form at the back of the book.

Photographs included in the original manuscript have been reproduced

xerographically in this copy. Higher quality 6” x 9” black and white

photographic prints are available for any photographs or illustrations

appearing in this copy for an additional charge. Contact U M I directly to

order.

UMIA Bell & Howell Information Company

300 North Zeeb Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

SPATIAL VARIABILITY O F COASTAL ORGANIC SOIL CHARACTERISTICS

A Dissertation

Submitted to the Graduate Faculty of the Louisiana State University and

Agricultural and Mechanical College in partial fulfillment of the

requirements for the degree of Doctor of Philosophy

in

The Department of Agronomy

byL. Cecil Dharmasri

B.S., University of Peradeniya, 1984 M.S., University of Saskatchewan, 1995

December 1998


UMI Number: 9922073

UMI Microform 9922073 Copyright 1999, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI300 North Zeeh Road Ann Arbor, MI 48103


ACKNOWLEDGEMNTS

I am very much grateful and express my sincere thanks to my major

advisor, Dr. W ayne H. Hud nail for his valuable guidance, advice, and constant

help during my study program. I thank Dr. Nina Lam, Dr. R. D. DeLaune, Dr.

Ray E. Ferrell Jr., Dr. Ramona Travis, Dr. Gary A. Breitenbeck and Dr. Jeff Hoy

for their advice while serving in my graduate committee. The NASA/EPSCoR

program supported the research, under cooperative agreement NCCW -0059

and the Louisiana Board of Regents/LEQSF under the contract NASA/LEQSF

(1994-97)-IM P-03. I am thankful to Dr. Ken Holladay and the members of the

project. I am very much thankful to Joaney Haigler for ICP-AES analysis and

W anda LeBlanc and Xiaogang Xie for mineralogical investigations. My thanks

are also due for my friend Asfaw Bekele who was with me when in need. I am

thankful to the student assistants and the EPSCoR summer intern, Ms. Betty A.

Odom who helped me in the laboratory. I thank my colleagues. Lamine Coly,

Rodney Young, Salena Agra va I, Lois West, Bill Patterson, and Clint Waddell

who shared the resources with me during this period. I express my gratitude

and commitment to the Department of Agriculture, Sri Lanka for releasing me to

pursue higher studies. I am grateful to Anura and Venika, Asoka and Upuli, and

Gumar (Gala) and Usha for their unreserved care and help for our family during

our stay in Baton Rouge. The constant encouragement and patient

understanding of my wife, Ayesha, through out my graduate studies for over six

years, is remembered most affectionately. I dedicate this dissertation to my

dearest sons, Poorna and Chamara, who were unaware of their sacrifice.


TABLE OF CONTENTS

ACKNOWLEDGMENTS..................................................................................... ii

LIST OF TABLES................................................................................................. vi

LIST OF FIGURES............................................................................................. vii

ABSTRACT......................................................................................................... xii

CHAPTER 1: INTRODUCTION........................................................................ 11.1 References...................................................................................... 9

CHAPTER 2: ESTIMATE SALINITY PARAMETERS ANDELEMENTAL RATIOS FOR THE ESTUARINEWATER SYSTEM OF BARATARIA BAY BASIN .................. 13

2.1 Summary and Introduction............................................................. 132.2 Literature Review......................................................................... 15

2.2.1 Barataria Bay Basin Water System .............................. 152.2.2 Seawater Salinity, the Major Elements, and Salinity

Parameters................................................................. 182.3 Materials and M ethods............................................................. 19

2.3.1 Sample Collection and Preparation.............................. 192.3.2 Chemical Analysis......................................................... 212.3.3 Measurement of T D S ..................................................... 21

2.4 Results and Discussion............................................................. 222.4.1 Elemental Composition.................................................. 222.4.2 Electrical Conductivity and Elemental Concentrations.. 242.4.3 Elemental Ratios............................................................. 262.4.4 Electrical Conductivity and Salin ity.............................. 32

2.5 Conclusions................................................................................. 362.6 References................................................................................. 36

CHAPTER 3:USE OF ION-EXCHANGE RESIN STRIPS FORELEMENTAL ANALYSIS OF ESTUARINE WATERAND COASTAL ORGANIC SOILS.......................................... 39

3.1 Summary and Introduction............................................................. 393.2 Literature Review ..................................................................... 43

3.2.1 Multi-Elemental Analysis in Soil Using Ion-Exchange Resins......................................................................... 43

3.2.2 Effect of Soluble Salts on Membrane Extractable Elements......................................................................... 45

3.2.3 Estimation of Membrane Exchangeable S ...................453.2.4 Methodology Adopted in Resin Extraction of S 46

3.3 Materials and M ethods..................................................................... 473.3.1 Salinity and Chloride Effect........................................ 49


3.3.2 Multi-Elemental Analysis of Estuarine W ater.............. 493.3.3 Resin-Extractable S Analysis for Coastal Organic

S o ils ................................................................................. 503.3.3.1 Equilibration and Elution Time Experiment... 503.3.3.2 Soil/Water Ratio Experiment.................. 513.3.3.3 Resin-Extractable and Water Soluble S

Analys is ............................................. 513.3.3.4 Pyritic and Non-pyritic S Analysis 52

3.3.4 Data Analysis................................................................. 533.4 Results and Discussion................................................................ 54

3.4.1 Salinity and Chloride E ffect........................................... 543.4.2 Multi-Elemental Analysis of Estuarine W ater............... 583.4.3 Resin Extractable S in Coastal Organic S o ils ............... 69

3.4.3.1 Equilibration and Elution T im e ........................ 693.4.3.2 Soil/Water R a tio ................................................ 693.4.3.3 Resin Extractable S .......................................... 73

3.5 Conclusions................................................................................. 803.6 References.................................................................................... 81

CHAPTER 4: INFLUENCE OF SALINITY AND LANDSCAPEPOSITION ON ACCUMULATION OF PYRITE AND NON-PYRITIC Fe AND S WITHIN COASTAL ORGANIC SOILS OF BARATARIA BAY BASIN, LOUISIANA............... 86

4.1 Summary and Introduction............................................................. 864.2 Literature Review..................................................................... 89

4.2.1 Marsh Loss around Barataria Bay Basin, Louisiana ... 894.2.2 Seawater Intrusion, Salinity and Marsh Types 904.2.3 Pyrite Formation and Accumulation.............................. 924.2.4 Sulfur Dynamics within the M arsh............................... 954.2.5 Pore Water Dynamics and Mineral Formation............. 994.2.6 Pyrite Crystal Formation................................................. 994.2.7 Spatial Variability of Pedogenic Processes.............. 1004.2.8 Soil Classification and Soil Taxonomy...................... 1024.2.9 Pyrite Determination........................................................ 104

4.3 Materials and Methods................................................................. 1064.3.1 Site Description and Sampling Procedure.................. 1064.3.2 Measurement of EC and T D S ...................................... 1084.3.3 Pyrite and Non-pyritic Fe and S Determination 1084.3.4 Measurement of Soil p H ............................................... 1104.3.5 Microanalysis of Marsh S o ils ........................................ 1104.3.6 Data Analysis................................................................. I l l

4.4 Results and Discussion............................................................. 1124.4.1 Soil Horizon Thickness and Bulk Density.................. 1124.4.2 Soil Salin ity..................................................................... 1184.4.3 Pyrite Accumulation and Pyrite Profiles...................... 1274.4.4 Non-pyritic F e ................................................................. 136

IV


4.4.5 Non-pyritic S ................................................................. 1444.4.6 Soil p H ............................................................................. 1514.4.7 Microanalysis of Soils for Pyrite .................................. 1544.4.8 Soil Classification.......................................................... 162

4.5 Summary and Future Research Needs.................................. 1664.6 References................................................................................. 173

CHAPTER 5: SPATIAL VARIABILITY OF COASTAL ORGANICSOIL CHARACTERISTICS WITHIN THE BARATARIABAY BASIN, LOUISIANA........................................................ 180

5.1 Summary and Introduction......................................................... 1805.2 Literature Review ..................................................................... 184

5.2.1 Barataria Bay Coastal M arsh...................................... 1845.2.2 Soil Morphology and Classification.............................. 1885.2.3 Spatial Variability of Marsh Soil Characteristics............1895.2.4 Spatial Interpolation Methods...................................... 190

5.3 Materials and M ethods............................................................. 1955.3.1 Site Description............................................................. 1955.3.2 Data Collection.................................................................. 1975.3.3 Data Analysis................................................................. 200

5.4 Results and Discussion............................................................. 2025.4.1 Thickness of Organic Subhorizons.............................. 202

5.4.1.1 Exploratory Data Analysis...................... 2025.4.1.2 Spatial Data Analysis.............................. 204

5.4.2 Depth to Mineral Layer (DML) and Soil Subgroup Delineation.......................................................................2105.4.2.1 Exploratory Data Analysis.......................... 2105.4.2.2 Spatial Data Analysis............................... 2135.4.2.3 Spatial Distribution of Soil Subgroups .... 216

5.4.2 pH for Ditferent Subhorizons.................................. 2165.4.2.1 Exploratory Data Analysis.......................... 2165.4.2.2 Spatial Data Analysis.................................. 219

5.4.3 Organic/Mineral Ratio (OMR) for Different Subhorizons................................................................. 2245.4.3.1 Exploratory Data Analysis.......................... 2245.4.3.2 Spatial Data Analysis............................... 227

5.5 Conclusions................................................................................. 2305.7 References................................................................................. 235

CHAPTER 6:C0NCLUSI0NS......................................................... 239

VITA.................................................................................................... 248


LIST OF TABLES

2.1 Sample locations and EC measurements.................................................. 20

2.2 Electrical conductivity, pH and elemental concentrationranges of water samples................................................................................. 23

2.3 Element/Chloride ratios of water samples..................................................... 30

3.1 Electrical conductivity, pH and elemental concentration rangesof water samples used in the experiment................................................... 59

3.2 Regression model for different elements to predict elementsin the water using resin extracted elements.......................................... 67

3.3 Regression models to predict different sulfur fractionsfrom RES for organic and mineral soil layers.......................................... 79

4.1 Sulfur species (ppm) variation in 0-50 cm depth for differentmarshes in the Barataria Bay, Louisiana.................................................. 98

4 .2 Average thickness of different subhorizons of soil profilesat different landscape positions............................................................... 113

5.1 The variables used for spatial analysis for each marsh type 201

5.2 Modal parameters and regression coefficients, for the semivariograms for the thickness of organic subhorizonsat both sites....................................................................................................... 204

5.3 Modal parameters and regression coefficients of thesemivariograms for the DML at both sites.............................................. 213

5.4 Semivariogram model parameters for pH of soil subhorizonsat both s ite s ................................................................................................... 2 2 2

5.5 Modal parameters and regression coefficients for thesemivariograms for OMR for organic soil horizons at both sites ... 227

VI


LIST OF FIGURES

1.1 Schematic representation of the linkage between differentchapters of the dissertation........................................................................ 6

2.1 A map showing different marsh types within the BaratariaBay Basin............................................................................................................ 16

2.2 Electrical conductivity and the major cations of water samples ... 25

2.3 Upper and lower confidence intervals (95% ) for Clconcentration at different EC levels....................................................... 27

2.4 Upper and lower confidence intervals for SO 4

concentration at different EC levels....................................................... 28

2.5 Relationship between C I/S 04 ratio and EC of water samples 31

2.6 Chloride/sulfate ratio data for different marsh types............................. 33

2.7 The relationship between EC and T D S ................................................. 34

2.8 The relationship between EC and T D S ............................................... 35

3.1 Schematic presentation of the general procedurefor resin extraction.......................................................................................... 48

3.2 Effect of salinity and sulfate level on resin extractable sulfate 55

3.3 Effect of chloride concentration on resin extractable sulfateat 600 mg S L" ........................................................................................... 56

3.4 Response surface for resin extractable S as affected bysulfate and chloride levels......................................................................... 57

3.5 Potassium and sulfur concentration as influenced by electricalconductivity in water samples..................................................................... 60

3.6 Resin extracted K at different K levels in water.......................................... 62

3.7 Relationship between electrical conductivity and resin extracted-K... 63

3.8 Resin extracted milliequivalent percentages of major cations............... 64

3.9 Sulfate milliequivalents exchanged onto the resin atdifferent EC levels of water samples..................................................... 65

VII


3.10 Effect of equilibration time on resin extractable S from organic soils.. 70

3.11 Effect of elution time on resin extractable S ............................................... 71

3.12 Effect of soil/water ratio on resin extractable S ....................................... 72

3.13 Boxplots showing data for different S fractions for organicand mineral horizons............................................................................... 74

3.14 Relationship between RES and W SS for organic horizons 76

3.15 Relationship between RES and NFS for organic horizons 77

3.16 Relationship between RES and NFS for mineral horizons 78

4.1 A map showing different marsh types within theBarataria Bay Basin...................................................................................... 107

4 .2 Variation in soil bulk density for saline and brackish m arshes 119

4 .3 Salinity profiles for streamside (SSS) and inland (SIL)for the saline marsh...................................................................................... 1 2 1

4 .4 Salinity profiles for streamside (BSS) and inland (BIL)for the brackish marsh................................................................................ 123

4 .5 Salinity profiles for streamside (ISS) and inland (ML)for the intermediate m arsh ......................................................................... 124

4.6 Salinity profiles for streamside (FSS) and inland (FIL)for the fresh marsh......................................................................................... 125

4.7 Weighted average of salinity for the profile up to 150 cm................. 126

4.8 Fyrite profiles for streamside (SSS) and inland (SIL)for the saline m arsh ............................................................................. 130

4.9 Fyrite profiles for streamside (BSS) and inland (BIL)for the brackish m arsh .................................................................................. 132

4.10 Fyrite profiles for streamside (ISS) and inland (ML)for the intermediate m arsh ......................................................................... 133

4.11 Fyrite profiles for streamside (FSS) and inland (FIL)for the fresh m arsh ...................................................................................... 134

VIII


4.12 Weighted average of pyrite content for the soil profileup to a depth of 150 c m ................................................................. 135

4.13 Non-pyritic Fe profiles for streamside (SSS) andinland (SIL) for the saline m arsh ........................................................ 137

4.14 Non-pyritic Fe profiles for streamside (BSS) andinland (BIL) for the brackish m arsh ........................................................ 138

4.15 Non-pyritic Fe profiles for streamside (ISS) andinland (ML) for the intermediate m arsh ................................................ 139

4.16 Non-pyritic Fe profiles for streamside (FSS) andinland (FIL) for the fresh m arsh ............................................................ 140

4.17 Weighted average non-pyritic Fe for the soil profileto a depth of 150 c m .................................................................................. 142

4.18 Non-pyritic S profiles for streamside (SSS) and inland (SIL)for the saline m arsh .................................................................................. 145

4 .19 Non-pyritic S profiles for streamside (BSS) and inland (BIL)for the brackish m arsh ............................................................................ 146

4.20 Non-pyritic S profiles for streamside (ISS) and inland (ML)for the intermediate m arsh ..................................................................... 147

4.21 Non-pyritic S profiles for streamside (FSS) and inland (FIL)for the fresh m arsh .................................................................................. 149

4.22 Weighted average of non-pyritic S for the soil profileto a depth of 150 c m .................................................................................. 150

4.23 The pH profiles for streamside of all marsh types............................. 152

4.24 The pH profiles for inland of all marsh types..................................... 153

4.25 Scanning electron micrograph showing microstructureof coastal marsh sediments...................................................................... 156

4.26 Scanning electron micrograph showing pyrite framboids within organic material of saline marsh sediments from the O a l horizon... 157

4.27 A close up of pyrite framboids within saline marshsediments from the O a l horizon........................................................... 158

IX


4.28 A single pyrite framboid within saline marshsediments from the Oa1 horizon............................................................. 159

4.29 The XRD patterns of saline marsh sediments (<2 pm)from the O a l horizon.............................................................................. 160

4.30 The XRD patterns of saline marsh sediments (<20 pm)from the O a l horizon..................................................................................... 161

4.31 The XRD patterns of brackish marsh sediments (<2 pm)from the O a l horizon.................................................................................. 163

4.32 The XRD patterns of brackish marsh sediments (<20 pm)from the O a l horizon.................................................................................. 164

4.33 Differential XRD patterns between 2 and 20 pm fractionsfrom the O a l horizons at saline and brackish m arshes.................... 165

5.1 An idealized coastal organic soil profile with subhorizons................. 188

5.2 Major features of an ideal semivariogram............................................... 193

5.3 Site locations for the saline marsh (8 ) and the brackish marsh (B)shown on a satellite image for southern Louisiana............................. 196

5.4 Maps showing dominant open water areas (W) and waterwayswithin the sites at saline (A) and brackish (B) marsh ty p e s .............. 198

5.5 Sampling plan for both sites, showing the sampling pointson the grid......................................................................................................... 199

5.6 Boxplots showing the data for organic subhorizon thicknessat both sites.................................................................................................. 203

5.7 Semivariograms for S0A 1TH , S 0A 2TH , and SOA3TH d a ta 206

5.8 Semivariograms for B0A1TH, B0A 2TH , and BOA3TH d a ta 207

5.9 Spatial variability of O a l , Qa2, and Ga3 horizon thicknessfor the saline marsh..................................................................................... 208

5.10 Spatial variability of O a l , Oa2, and Oa3 horizon thicknessfor the brackish marsh................................................................................. 209


5.11 Boxplots showing the data for depth to mineral layer (DML)for both sites.................................................................................................. 2 1 2

5.12 Semivariograms for the SDML and BDML d a ta ..................................... 214

5.13 Spatial variability of depth to mineral layer for the saline marsh (A)and brackish marsh ( B ) ............................................................................... 215

5.14 Spatial variability of the soils for the saline and brackishmarsh types...................................................................................................... 217

5.15 Boxplots showing the data for pH of soil subhorizons andweighted average pH for the organic layer for both sites................ 218

5.16 Semivariograms for S0A1 PH, S0A 2PH , S 0A 3P H , andSCGPH d a ta .................................................................................................. 220

5.17 Semivariograms for B0A1PH, BOA2PH, BOA3PH, andBCGPH data.................................................................................................... 221

5.18 Spatial variability of soil pH for O a l (A), Ga2 (B), Ga3 (C),and Cg (D) horizons for the saline marsh.............................................. 223

5.19 Spatial variability of soil pH for G a l (A), Ga2 (B), Ga3 (C),and Cg (D) horizons for the brackish marsh.......................................... 225

5.20 Boxplots showing the data for organic/mineral ratio (GMR) for different organic subhorizons and weighted average GMRfor the organic layer for both sites............................................................. 226

5.21 Semivariograms for SGA1GMR, SGA2GMR, and SGA3GMR data 228

5.22 Semivariograms for BGA1GMR, BGA2GMR, and BGA3GMR data 229

5.23 Grganic/mineral ratio for the G a l (A), Ga2 (B), and Ga3 (0 )horizons at the saline marsh....................................................................... 231

5.24 Grganic/mineral ratio for the G a l (A), Ga2 (B), and Ga3 (C) horizons at the brackish marsh................................................................. 232

XI


ABSTRACT

This dissertation is a part of the Louisiana NASA/EPSCoR (Experimental

Program to Stimulate Competitive Research) global change research, which

studied the fate of carbon and sediments within the Barataria Bay Basin,

Louisiana. It studied water composition to assess seawater influence within the

marsh. Ion exchange resin strips were used to study the effect of salinity and

Chloride (Cl) on sulfate (SO 4 ) reduction and their potential for water and soil

analysis. Chloride dominated the water system and the CI/SO 4 ratio can be

used to assess the seawater influence. Resin extractable sulfur (S) predicted

non-pyritic S fraction for the marsh soils. High salinity reduced the affinity of

target ions onto the resin. Limited affinity of SO 4 to resin indicates SO 4

accumulation within root zone, which promotes sulfate reduction and pyrite

formation.

This project mainly studied landscape position and salinity effects on

pyrite accumulation and the spatial variability of soil characteristics within a

saline and a brackish marsh. Salinity, pyrite, and non-pyritic iron (Fe) and 8

varied between streamside and inland. Depressions in mineral layer, accretion

variations and associated hydrology caused the field-scale variability. When

the inland site is landlocked, salinity and pyrite content within the surface

horizon varied. Non-pyritic 8 , pH, and pyrite profiles were different in different

marsh types. Mineralogical evidence also found for presence of pyrite

framboids. These soils should be reclassified to indicate accumulations of

reduced sulfur.

XII


Thickness of subsurface horizons was highly spatially variable. Variation

in depth to mineral layer (DML) can be due to the presence of depressions in

the mineral layer surface. Typic Medisaprists occurred mostly toward inland

areas and away from waterways at the saline marsh. The DML was shallower

for the degrading marsh within the saline marsh type. Typic Medisaprists within

the brackish marsh had thick organic layers due to presence of thick

subhorizons. Spatial variability is evident for pH and organic/mineral ratio

(OMR) within organic subhorizons. The OMR data varied widely for the

brackish marsh compared to saline marsh. Organic soil characteristics vary

spatially due to variations in associated processes, therefore, spatial variability

should be considered for soil sampling schemes.

XIII


CHAPTER 1: INTRODUCTION

A multidisciplinary research team was formed to study the fate of carbon

and sediments within the Barataria Bay Basin, Louisiana under the Louisiana

NASA/EPSCoR (Experimental Program to Stimulate Competitive Research)

global change research cluster (Miller et al., 1995). The research team

comprised of focus groups for hydrology, vegetation, biogeochemistry, soil

science, remote sensing and CIS. This dissertation is part of the soil science

investigations, which also includes studies required by the collaborations with

the hydrology and remote sensing groups.

Relationship between electrical conductivity (EC) and total dissolved

solids (TDS) is required for ground truthing the Airborne Electromagnetic

Profiler (AEM) data. The AEM data can be used to obtain material conductivity

and depth (Bergeron et al., 1989; Travis, 1994) The relationship between EC

and the elemental ratios may indicate the seawater influence and deviation of

the ratios may be used to determine flow direction within marsh for the

hydrology work.

Ion-exchange resin strips are used as a tool to analyze a large number

of samples quickly. Resin strips are very useful, if feasible, to use for water

samples as required by the hydrology studies and soil samples for sulfur (S)

estimations for the spatial variability studies. Currently available S fractionation

methods are complicated, laborious and time consuming (Begheijn, 1978; Lord

III, 1982; Raiswell et al., 1994). Resin extractable S needs to be calibrated for

different sulfur fractions of organic soil samples.

1


Seawater brings in 8 as sulfate that is chemically reduced within the

anaerobic environment within the organic soils. This process is called sulfate

reduction that forms pyrite as an end product (Ponnamperuma, 1972; Patrick

and Jugsujinda, 1992; Rabenhorst and James, 1992). Coastal marsh provides

a suitable environment for pyrite formation because of the presence of

anaerobic environment, continuous S supply by seawater intrusion and the

availability of reduced iron (Pons et al., 1982). During the sulfate reduction

process, microorganisms utilize labile carbon as the energy source (Connell

and Patrick, 1969; Rabenhorst and James, 1992). The amount of pyrite

indirectly estimates the quantity of carbon utilized for the sulfate reduction.

Existence of the favorable conditions for the pyrite formation leads to

pyrite accumulation within the soil profile. However, the availability of easily

decomposable organic matter limits the sulfate reduction (Devai et al., 1996).

Reduction of ferric to ferrous, conversion of sulfate to different forms of reduced

sulfur and pyrite formation, occur in consequent steps. Therefore,

accumulations of pyrite, ferrous and sulfidic material within the soil profile reflect

the biogeochemical status of a marsh, which can be spatially variable. The

accumulation of sulfide is toxic to plants, but the presence of mineral material

buffers sulfide toxicity by soil iron (Koch and Mendelssohn, 1989; Nyman and

DeLaune, 1991). The accumulation of ferrous and sulfide reflects limited

pyritization due to the unavailability of labile carbon.

Previous 8 dynamics studies within these marshes have confined

sampling within close proximity to avoid spatial variability (Feijtel et al., 1988;

2


Krairapanond et al., 1991a, 1991b, 1992). However, DeLaune et ai. (1983)

reported the existence of a sulfide concentration gradient associated with slight

elevation changes that resulted in plant productivity and mortality variations.

The elevation variations are controlled by the accretion rate. The vertical

accretion rates vary between levee areas and backmarshes (Hatton et al.,

1983). As the previous studies indicated the biogeochemical status of the soils

vary between landscape positions. Therefore, a detailed study is needed to

understand the spatial variability for profile distribution of S fractions between

landscape positions within different marsh types.

High net primary production of coastal marsh in terms of the above and

below ground biomass is the main contribution to the fixed carbon pool within

the marsh soil environment. The organic soil profile is comprised of mainly two

layers; the organic layer and the mineral layer. The organic layer can be

divided into different sub horizons based on the color and the composition. The

organic matter accumulation (organic accretion) and sedimentation (inorganic

accretion) are the main processes that form the organic sub horizons (Hatton et

al., 1983; Chmura et al., 1992;Nyman et al., 1993). Thickness of the organic

subhorizons varies due to the variations in the organic and the inorganic

accretions and disturbances to the marsh, which reflects the historical signature

of the marsh soil formation. Organic accretions vary due to the changes in the

productivity of the vegetation and inorganic accretions vary depending on the

changes in the sedimentation rate within the marsh.


Marsh loss to open water is a severe problem in the marshes of northern

Gulf of Mexico (Boesch et al., 1983; Salinas et al., 1986; Turner, 1990).

Degradation and catastrophic events are the main disturbances to the marsh.

Two mechanisms, vegetation die back and subsurface erosion cause marsh

degradation. Vegetation die back is associated with plant stress due to sulfide

toxicity and flooding (Webb et al., 1995). Inability of the marsh to counter the

submergence level has been reported as the main cause of marsh loss (Turner,

1990; DeLaune et al., 1994). Subsurface erosion causes loss of organic

subhorizons and collapse of the surface organic layer (DeLaune et al., 1983;

DeLaune et al., 1994; Nyman et al., 1994). Salinity intrusion also causes

dispersion of sediments within the degrading marshes. In the submerging

marsh the organic matter is either buried within the profile or lost to the open

water areas (Nyman et al., 1995). Catastrophic events influence the Louisiana

marshes frequently. Hurricanes bring salt water and sediments inland and may

also erode lands due to physical action of flooding (Childers et al., 1990;

Turner, 1990; Jackson et al., 1995). Sedimentation due to hurricanes is highly

variably within the marsh (Nyman et al., 1995).

Information on spatial variability of coastal marsh characteristics is

limited at field scale. Soil sampling is very difficult within the marsh and spatial

variability data will help plan soil sampling schemes to be more representative

and minimize the number of sampling points as appropriate. Remotely sensed

data are often used in coastal marsh studies and the spatial variability


information also helps a researcher decide the appropriate pixel size of the

remotely sensed data for a particular study.

The thickness of the organic subhorizons is a direct estimate of fixed

carbon. The organic/mineral ratio and pH reflect the compositional and

biogeochemical status of the marsh. Understanding of the spatial variability of

these characteristics more accurately estimates the carbon pool within the

marsh.

Four studies are reported in this dissertation with the following

objectives. The linkage of the different studies is presented in Figure 1.1.

1. To estimate salinity parameters and elemental ratios for the estuarine water.

2. To study the feasibility of the ion-exchange resin strips for water analysis and

to estimate sulfur status.

3. To determine the effect of landscape position and marsh type on pyrite

accumulation.

4. To study the spatial variability of the coastal organic soil characteristics.

Study 1 establishes the relationships between the EC and the TDS and

the elemental ratios for the water system within the Barataria Bay, which is

mainly collaboration with the remote sensing and the hydrology groups. The

elemental ratios that are characteristic to the water system are identified and

their relationship with the EG is established. Based on this information any

deviation of water samples from their original composition is identified. This

scenario can be used to assess the influence of seawater intrusion and to

determine the flow direction of water within the marsh. The information on

5


:x)CD■oOQ.CgQ.

■oCD

C/)(/)o'=5

8 5c q '

3CD

C

CD■oOQ.CaO3■oO

&

oc■oCD

Chapter 1: Introduction

1

o>

Chapter 6: Conclusions

Chapter 2; Estimate salinity parameters and elemental ratios for the estuarine water (study 1 )

Chapter 5: Study the spatial variability of coastal organic soil characteristics (study 4)

Chapter 3: Study the feasibility of ion-exchange resin strips for water analysis and to estimate soil S status (study 2)

Chapter 4: Determination of the effect of landscape position and marsh type on pyrite accumulation (study 3)

C/)(/)o'3

Figure 1.1. Schematic representation of the linkage between different chapters of the dissertation

salinity and dominant ions is used as a guideline for testing resin strips for

water and soil analysis in study 2 and to select salinity levels for study 3.

Feasibility of resin strips is tested for the estuarine water and coastal

marshes in study 2. Several constraints for soil and water analysis can be

avoided using resin strips. Saline water samples can not be analyzed by

instruments such as Inductively Coupled Plasma (ICP-AES) and Ion

Chromatography (1C) because of the interference of high concentrations of

soluble salts. Dilution of samples reduces the target ions beiow the detection

limits of the instruments. Resin extractions enable one to use these

instruments despite the high salinity of the samples. Multi-elemental analysis

can be performed quickly using resin strips.

Bulk density of the marshes varies widely. Therefore, the elemental

concentration estimates can not be compared unless they are corrected for

bulk density of the soils. Resin strips mimic plant roots and resin extracted

elemental concentrations are independent from bulk density variations. It is

advantageous if we use resins for S analysis of coastal organic soils, especially

for spatial variability studies. Resin strips have never been tested for these

soils. W ater samples collected for study 1 were used to study the resin

performance for water analysis. The effect of salinity and Chloride (01) on the

affinity of target ions on to the resin is presented. Data on resin extracted S are

calibrated with water-soluble S and the pyritic and non-pyritic S data collected

for study 3.


The influence of the landscape position and marsh type on pyrite

accumulation, and non-pyritc Fe and S profiles is presented in the study 3.

The currently available pyrite determination methods are complicated, laborious

and time consuming (Begheijn, 1978; Lord III, 1982; Willett and Beech, 1987;

Leventhal and Taylor, 1990; Raiswell et al., 1994). A modified technique was

used to determine pyrite content indirectly (Schneider and Schneider, 1990). A

quick and easy method is necessary for spatial variability studies because of

the large number of samples needs to be tested. Moreover, the spatial

variability pattern is independent of the analytical technique used. Profile

distribution of pyrite and non-pyritic Fe and S were studied to determine

differences between landscape positions and marsh types.

The spatial variability of selected soil characteristics is presented in

study 4. Two sites representing a saline and a brackish marsh were intensively

sampled on a grid within a one-square mile area. Soil morphological data such

as sub horizon thickness, depth to mineral layer, pH and organic/mineral ratio

for different horizons were collected. Geostatistical analysis was conducted to

identify the spatial variability of the soil variables. The semivariograms are

presented for the soil variables and future marsh sampling schemes can be

planned on the spatial variability data. Data are interpolated within a one-

square mile area on a finer grid using appropriate interpolation methods to

identify associated landscape patterns of these soil characteristics.


1.1 ReferencesBegheijn, L. Th., N. van Breemen, and E. J. Velthorst. 1978. Analysis of sulfur

compounds in acid sulfate soils and other recent marine soils. Commun, in Soil Soi. Plant Anal. 9(9):873-882.

Bergeron, Jr., C. J., J. W. loup, and G. A. Michel, II. 1989. Interpretation of airborneelectromagnetic data using the modified image method. Geophysics. 54(8):1023-1030.

Boesch, D. P., Levin, D., Nummedal, D., and Bowles, K. 1983. Subsidence in coastal Louisiana: Causes, Rates, and Effects on Wetlands. U. S. Fish and Wildlife Service FWS/OBS-83/26, Washington, D. 0., 30pp.

Childers, D. L., J. W. Day, and R. A. Muller. 1990. Relating climatological forcing to coastal water levels in Louisiana estuaries and the potential importance of El Nino-Southern oscillation events. Climate Research 1:31-42.

Chmura, G. L., R. Costanza, and E. Kosters. 1992. Modelling coastal marsh stability in response to sea level rise: a case study in coastal Louisiana, USA. Ecological Modelling, 64:47-64.

Connell, W. E., and W. H. Patrick, Jr. 1969. Reduction of sulfate to sulfide in waterlogged soil. Soil Soi. Soc. Am. Proc. 33:711-715.

DeLaune, R. D., R. H. Baumann, and J. G. Gosselink. 1983. Relationships among vertical accretion, coastal submergence, and erosion in a Louisiana gulf coast marsh. Journal of Sedimentary Petrology 53(1 ): 147- 157.

DeLaune, R. D., C. J. Smith., and W. H. Patrick. 1983. Relationship of marsh elevation, redox potential, and sulfide to Spartina a Item if I ora productivity. Soil Sci. Soc. Am. J. 47:930-935.

DeLaune, R. D., J. A. Nyman, and W. H. Patrick, Jr. 1994. Peat Collapse, ponding and wetland loss in a rapidly submerging coastal marsh. J. Coastal Res. 10:1021-1030.

Devai, I., K. R. Reddy, R. D. DeLaune, and D. A. Graetz. 1996. Sulfatereduction and organic matter decomposition in a wetland soil and lake sediment. Acta Biol. Debr. Oecol. Hung. 6:13-23.


Feijtel, T. C., R. D. Delaune, and W. H. Patrick, JR. 1988. Seasonal pore water dynamics In marshes of Baratarla Basin, Louisiana, Soli Sol. Soc. Am. J. 52:59-67.

Hatton, R. S., R. D. DeLaune, and W. H. Patrick, Jr. 1983. Sedimentation accretion, and subsidence In marshes of Baratarla Basin, Louisiana. Limnology and Oceanography 28:494-502.

Jackson, L. L., A. Lee Foote, and L. S. Ballstrlerl. 1995. Hydrological,geomorphologlcal, and chemical effects of hurricane Andrew on coastal marshes of Louisiana. Journal of Coastal Research SI(21):306-323.

Koch, M. S. and I. A. Mendelssohn. 1989. Sulphide as a soil phytotoxin: Differential response In two marsh species. J. Ecol. 77:565-578.

Kralrapanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1991a. Sulfur dynamics In Louisiana coastal freshwater marsh soils. Soil Sol. 151(4):261-273.

Kralrapanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1991b. Seasonal distribution of sulfur fractions In Louisiana salt marsh soils. Estuaries 14:17-28.

Kralrapanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1992. Distribution of organic and reduced sulfur forms In marsh soils of coastal Louisiana,Org. Geochem. 18:489-500.

Leventhal, J., and C. Taylor. 1990. Comparison of methods to determinedegree of pyrltlzatlon. Geochlmica et Cosmochlmica Acta 54:2621-2625.

Lord III, C. J. 1982. A selective and precise method for pyrite determination In sedimentary materials. Journal of Sedimentary Petrology 52:664-666.

Miller, R. L., M. Glardino, B. A. McKee, J. F. Cruise, G. Booth, R. Rovansek, D. Mulrhead, W . CIbula, K. Holladay, R. E. Pelletier, W . Hudnall, C. Bergeron,J. loup, G. loup, and G. Love. 1995. Processes and fate of sediments and carbon In Baratarla Bay, LA. Proceedings from the Third Thematic Conference on Remote Sensing for Marine and Coastal Environments. Seatle, Washington. Vol. l:233-244pp.

Nyman, J. A. and R. D. DeLaune. 1991. Mineral and organic matteraccumulation rates In deltaic coastal marshes and their Importance to landscape stability. GCSSEPM Foundation Twelfth Annual Research Conference Program and Abstracts, 166-170.

10


Nyman, J. A., M. Carlose, R. D. DeLaune, and W . H. Patrick, Jr. 1994.Erosion rather than plant dieback as the mechanism of marsh loss in an estuarine marsh. Earth Surface Processes and Landforms 19:69-84.

Nyman, J. A., J. C. Callaway, and R. D. DeLaune. 1993. Case study of a rapidly submerging coastal environment: relationships among vertical accretion, carbon cycling and marsh loss in the terrebone basin, Louisiana. Proceedings of the Hilton Head Island South Carolina USA International Coastal Symposium, June 6-9, 1993. vol 2. 452-457p.

Nyman, J. A., R. D. D elaune, S. R. Pezeshki, and W. H. Patrick, Jr. 1995. Organic matter fluxes and marsh stability in a rapidly submerging estuarine marsh. Estuaries 18:207-218.

Patrick, W. H. Jr., and A. Jugsujinda. 1992. Sequential reduction and oxidation of inorganic Nitrogen, Manganese, and Iron in flooded soil. Soil Soi. Soc. Am. J. 56:1071-1073.

Ponnamperuma, F. N. 1972. The chemistry of submerged soils. Adv. Agron. 24:29-96.

Pons, L. J., N. van Breemen, and P. M. Driesses. 1982. Physiography ofcoastal sediments and development of potential soil acidity, pp. 1-18. In J. A. Kittrick, D. S. Fanning, and L. R. Hossner (Eds.) Acid sulfate weathering. Soil Sci. Soc. Am. Spec. Pub. No. 10, Soil Sci. Soc. Am., Madison, Wisconsin.

Rabenhorst, M. C., and B. R. James. 1992. Iron sulfidization in tidal marsh soils. In H. C. W. Skinner and R. W. Fitzpatrick (Eds.) Biomineralization processes. Iron, Manganese: modern and ancient environments. Catena supplement 21:203-217.

Raiswell, R., and D. E. Canfield, and R. A. Berner. 1994. A comparison of iron extraction for determination of degree of pyritisation and the recognition of iron-limited pyrite formation. Chemical Geology 111:101-110.

Salinas, L. M., R. D. D elaune, and W . H. Patrick, Jr. 1986. Changes occuring along a rapidly submerging coastal area: Louisiana, USA. Journal of Coastal Research 2(3):269-284.

Schneider, J. W ., and K. Schneider. 1990. Indirect method for thedetermination of pyrite in clays and shales after selective extraction with acid solutions. Ceramic Bulletin 69(1): 107-109.

11


Travis, R. E. P. 1994. Utility of remote sensing technology in wetland soil and gas flux studies. A Dissertation submitted to Louisiana State University, La.

Turner, R. E. 1990. Landscape development and coastal wetland losses in the northern Gulf of Mexico. Amer. Zool. 30:89-105.

Webb, E. C., I. A. Mendelssohn, and B. J. Wilsey. 1995. Causes for vegetation dieback in a Louisiana salt marsh: A bioassay approach. Aquatic Botany 51:281-289.

Willett, I. R., and T. A. Beech. 1987. Determination of organic carbon in pyritic and acid sulfate soils. Commun, in Soil Sci. Plant Anal. 18(7):715-724.

12


CHAPTER 2: ESTIMATE SALINITY PARAMETERS AND ELEMENTAL RATIOS FOR THE ESTUARINE WATER SYSTEM OF BARATARIA BAY

BASIN2.1 Summary and Introduction

This study establishes the relationships between the electrical

conductivity (EC) and the total dissolved solids (TDS) and the elemental ratios

for the water system within the Barataria Bay Basin. Water samples from the

Baratarla Bay Estuary were analyzed for chemical properties. Salinity and the

elemental ratios were studied. Barataria Bay estuary has a chloride dominant

water system with a wide range of salinity. The elemental composition and the

ionic ratios indicate that the water characteristics of this estuary are extensively

under the influence of seawater intrusion. Salinity change is due to the mixing

of seawater with fresh water, because the ratios of major elements are similar

to that of seawater. The chloride/sulfate ratios indicate that seawater intrusion

controls the salinity within the brackish and saline marshes while the fresh

marsh is free from seawater salinity. The deviation in elemental ratios of water

samples from the seawater ratios can be used to assess the influence of

seawater intrusion and to determine the flow direction of water within the

marsh. Chloride/sulfate ratio is proposed to determine the extent of seawater

intrusion within the basin. The information on salinity and dominant ions is

used as a guideline for testing resin strips for water and soil analysis in study 2

in this dissertation.

13


A multidisciplinary research team was formed to study the fate of carbon

and sediments within the Barataria Bay Basin, Louisiana under the Louisiana

NASA/EPSCoR (Experimental Program to Stimulate Competitive Research)

global change research cluster (Miller et al., 1995). The research team

comprised of focus groups for hydrology, vegetation, biogeochemistry, soil

science, remote sensing and GIS. This study is required by the collaborations

with the hydrology and remote sensing groups. Airborne electromagnetic

profiler (AEM) is used to remotely collect data on salinity and organic layer

thickness within the marsh. The AEM data can be used to obtain material

conductivity and depth (Bergeron et al., 1989; Travis, 1994). Relationship

between EC and TDS is required for ground truthing the AEM data. The

relationship between EC and the elemental ratios may indicate the seawater

influence within the marsh.

Characterization of the chemical properties of the estuarine water

system helps to compare the data with other estuarine systems as well as to

use the knowledge generated from other estuaries. Seawater is considered as

the main source of sulfur in to the marsh ecosystem. Sulfate reduction is a

major biogeochemical process within the marsh, which is mainly responsible for

S transformations within the soils and sediments. Biological S assimilation

converts the inorganic S in to organic S forms. Plants respond to high soluble

salt concentrations within the root zone through different mechanisms such as

ion exclusion, secretion and accumulation. Overall salinity controls these

mechanisms within the plants and the species responses are different (Bradley

14


and Morris, 1991; Latham et al., 1994). Ionic ratios may be changed in water

due to these biological and geochemical processes.

Deviation of elemental ratio may indicate lack of seawater influence or

domination of another process. Hydrological studies need flow rate

measurements along the flow path to determine the direction of water flow.

Within the seawater influenced areas the deviation of the elemental ratios may

be used to determine flow direction within marsh for the hydrology work. This

study was planned mainly to provide information for the collaborative work

within the NASA/EPSCoR project. Therefore, it will measure the elemental

composition of the estuarine water at different salinity levels and establish the

relationship between EC and TD S for this water system.

2.2 Literature Review

2.2.1 Barataria Bay Basin Water System

The study was carried out in Barataria Bay Basin, which is a shallow

coastal estuary in Louisiana. The natural levees of the Mississippi River

constitute its northern and eastern boundaries while the abandoned Bayou

Lafourche distributory marks the western boundary. The Gulf of Mexico forms

the southern boundary (Figure 2.1). The basin is triangular in shape, about 110

km long and 50 km wide at the southern boundary (Conner and Day, 1987).

The total area of 628,000 ha is divided into three sub basins; upper freshwater

lake (Lac Des Allemands), middle brackish lakes, and lower saline bays

(Barataria Bay and Caminada Bay), lakes and marshes (Madden et al., 1988).

15


CD■ D

OQ.C

sQ.

■DCD

C/)C/)

8■D

oi«oo' 90 * 30 ' 90 * 00 ' 89 * 30 '

3.3 "CD

CD■DOQ.C

aO3"Oo

CDQ.

■DCD

C /)C /)

30 * 00 ' ____

G)

29 * 30 ' -

BARATARIA BASINLoulsiena

\Barataria Rasln

VegetationNolural lovaet(Urban, Agricultural Areoi

end Bollomlond Hardwood Pareil)

I Foreiled Welland (Swomp)

freih Morih

^ .1 Inlermediola Morih

I \ 1 Brockiih Morih

I Solina Morih

Figure 2.1 A map showing different marsh types within the Barataria Bay Basin (after Kralrapanond et al., 1992)

Salinity varies within the Barataria Bay Basin and this spatial variability has

established a well-defined salinity and vegetation association. Salinity

determines the vegetation type and influences pedogenic processes. Marshes

are delineated into saline, brackish, intermediate, and fresh water marshes

based on the salinity levels. Salinity ranges from below 0.5 to above 20.0 ppt

within the Barataria Bay marshes. Fresh and intermediate marshes occur

below 5 ppt. Brackish marshes occur between 5 and 10 ppt while salt marsh is

found at salinity above 10 ppt (Feijtel et al., 1988; Soil Conservation Service,

1989). Seawater is considered as the main source of salt into the system. The

presence of channels and streams brings seawater into the inland marshes.

Channel construction has changed the natural hydrology and affected the water

characteristics. Even though the magnitude of salinity change is not well

estimated, studies have shown a gradual increase of salinity in this region (U.

S. Army Corp Engineers, 1984). Seasonal variations in salinity and sediment

concentrations were associated with short and long term weather changes.

Fresh water floods lowered salinity within the lower basin. Lake Salvador and

Barataria Waterway transfer sediments and water in both directions with the

influence of weather events (Garrepally, 1994). Hydrology information for this

basin are reported elsewhere (Garrepally, 1994; Rovansek, 1997).

17


2.2.2 Seawater Salinity, the Major Elements, and Salinity Parameters

The major ions of seawater are those that significantly contribute to the

measured salinity. In practice, the elements that are present at concentrations

greater than 1 mg L' are considered as major ions except silicon (Si) which is

conventionally excluded because of its atypical ionic behavior. The major

cations are Na. K, Mg, Ca, and Sr while the major anions include Cl, SO4, Br, F,

and B. The major ions are referred to as conservative because their

concentrations in seawater maintain a constant ratio, one to another. This

constancy of composition of seawater permits the use of salinity to describe the

salt in a seawater sample. This uniqueness is so important that the ratios can

be used to compare compositional differences in sea salt from various parts of

the world (Wilson, 1975). The most important water quality parameter in

relation to salinity is the total salt content that is expressed as TDS. The TDS

can be determined gravimetrically. Measuring TDS is more tedious than

measuring EC, which is the currently preferred measure of salinity (Bresler et

al., 1982). Gravimetric estimates may not be accurate due to the volatility of

some inorganic constituents and possible retention of water of crystallization

(Wilson, 1975; Bresler et al., 1982). However, the gravimetric method is used

when an empirical relationship needs to be defined. The relationship

developed by Forch and others (Wilson, 1975), Salinity (ppt) = 1.805 Cl (ppt) +

0.030, was later replaced by Salinity (ppt) = 1.80655 Cl (ppt) due to

improvements in instrumentation and considering the variability of different

18


seawaters. The EC can be converted to TDS through the approximate

relationship proposed by the U. S. Salinity Laboratory Staff (within the EC

range of 3 to 30 dS m'^) (Bresler at al., 1982); TDS (mg L'^) = 0.64 * 10^ EC

(dS m'^). Seawater salinity is commonly considered as 35 ppt. The ratios of

cations (ppt) to chloride (ppt) were reported by Wilson (1975); Na, 0.5555; K,

0.0206; Mg, 0.06625; Ca, 0.02126; and Sr, 0.00041. The ratio of SO 4 to

chloride is 0.14. Average B content is about 4 .5 mg L' at sea water salinity of

35 ppt and its chloride concentration is about 19.354 g kg‘ (Wilson, 1975).

These chemical ratios can be used as a scheme to characterize the estuarine

water.

2.3 Materials and Methods

2.3.1 Sample Collection and Preparation

W ater samples were collected from different water bodies within brackish

and saline marsh areas of the Barataria Bay Basin during the summers of

1995, 1996, and 1997. Geographic locations were registered using a Global

Positioning System (G PS) for most of the sample locations. Table 2.1 provides

several sample locations with their coordinates and their electrical conductivity.

W ater samples were also collected from small streams or broken marsh areas

within the inland marshes. W ater samples were brought to the laboratory in an

ice-filled cooler and filtered (Whatman No. 42). These samples were used to

measure EC, pH and elemental composition.

19


CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8■D

■DCD

C /)C /)

Table 2.1. Sample locations and EC measurements.

3.3 "CD

^ ëOQ.C

a.O3"OO

CDQ.

Sample Latitude / Longitude EC (dS m" ) Sample Latitude / Longitude EC (dS m-1)

1 29 24.308 / 90 04.024 06.11 25 29 3 1 .2 5 2 /9 0 01.977 05.372 29 2 4 .2 1 0 /9 0 03.962 05.39 26 29 2 9 .6 9 7 /9 0 01.173 07.193 29 23.889 / 90 03.976 05.05 27 29 1 4 .7 4 9 /9 0 07.270 38.614 29 22.245 / 90 03.879 11.28 28 29 1 4 .7 4 9 /9 0 07.270 38.905 29 2 1 .1 3 8 /9 0 04.228 14.77 29 29 1 6 .5 8 7 /9 0 08.182 34.906 29 20.428 / 90 03.968 14.81 30 29 1 6 .5 8 7 /9 0 08.182 38.507 29 1 9 .2 4 4 /9 0 02.020 18.58 31 29 1 6 .5 8 3 /9 0 08.167 26.308 29 1 8 .8 0 2 /9 0 04.381 21.40 32 29 1 6 .5 8 3 /9 0 08.167 27.809 29 2 0 .0 1 8 /9 0 06.282 24.86 33 29 1 6 .5 8 3 /9 0 08.167 29.501 0 29 20.558 / 90 05.272 19.68 34 29 1 6 .5 8 3 /9 0 08.167 27.201 1 29 1 9 .9 8 7 /9 0 07.254 28.02 35 29 24.204 / 90 03.306 19.381 2 29 20.857 / 90 08.370 22.54 36 29 24.204 / 90 03.306 20.6013 29 23.495 / 90 06.628 16.59 37 29 2 4 .2 1 0 /9 0 04.024 19.2414 29 23.245 / 90 01.737 09.87 38 29 2 4 .2 1 0 /9 0 04.024 17.8615 29 24.308 / 90 04.024 06.56 39 29 23.560 / 89 56.500 48.3916 29 24.558 / 90 03.177 06.73 40 29 23.700 / 89 58.700 40.6017 29 2 4 .1 5 2 /9 0 01.684 10.60 41 29 2 4 .1 7 4 /9 0 03.930 42.1018 29 2 3 .7 7 5 /9 0 01.710 10.73 42 29 24.038 / 90 03.382 45.0019 29 23.885 / 90 02.921 06.71 43 29 4 1 .3 0 0 /9 0 12.090 06.822 0 29 23.869 / 90 03.894 05.80 44 29 45.101 /9 0 21.500 0.1932 1 29 2 4 .1 7 4 /9 0 03.930 06.08 45 29 45.101 /9 0 21.500 0.5892 2 29 24.305 / 90 04.067 05.60 46 29 45.101 /9 0 21.500 0.78723 29 24.305 / 90 04.067 09.71 47 29 45.101 /9 0 21.500 0.15624 29 24.204 / 90 03.306 10.08 48 29 45.101 /9 0 21.500 0.144

2.3.2 Chemical Analysis

The EC of the water samples was measured using a Corning M90

conductivity sensor (Corning Inc., Corning New York 14831, USA). The pH

was measured in the laboratory using a combination pH electrode, and chloride

was measured using a chloride ion selective electrode with pH/m V meter (Orion

Research model 701A/digital lONALYZER). Chloride concentration data were

obtained from a linear calibration using 1 0 to 1 0 0 0 0 ppm chloride

concentrations. W ater samples were analyzed using ICP-AES (PERKIN

ELMER model OPTIMA 3000 (Perkin-Elmer Corporation, Norwalk, CT 06859,

USA). Chloride can be estimated from charge balance calculations when

sulfate content is known (Wilson, 1975; Bresler et al., 1982). This uses two

assumptions that are very feasible; total halides are replaced by chloride and all

the sulfur is in an inorganic form.

2.3.3 Measurement of TDS

A volume of 30 - 60 mL from each water sample was weighed and slowly

evaporated in a centrifuge tube at a temperature between 55 - 65 °C in an

oven, cooled in a desiccator and weighed. The TDS was estimated on a dry

weight basis for each sample in triplicate. Total solids in solution were also

estimated based on chemical composition, by adding all the major cations and

anions.

21


2.4 Results and Discussion

2.4.1 Elemental Composition

Concentrations of major cations and anions and their relationship with

electrical conductivity are presented. Elemental concentrations changed

drastically between fresh water and saline water marshes. Descriptive statistics

for elemental composition of water samples are presented in Table 2.2. More

attention is given to the ranges than the mean values because the low

concentrations of fresh water lowered the means. W ater samples were not

collected systematically so that they may or may not have included any

temporal variability in addition to the spatial variability.

EC of water samples ranged from 0.144 to 48.39 dSm'^ (Table 2.2). EC

of water samples from fresh marsh ranged from 0.144 to 0.787 d S m '\ higher

values were obtained for standing water (i.e., 0.787) within the marsh while the

EC of the water in the bayou was less than 0.2 dSm "\ W ater samples from

brackish and saline areas always had greater EC values than water samples

from the fresh marsh. The pH of the water samples ranged between 6.69 and

8.29; pH increased with increasing EC.

Chloride concentrations increased about 300 times from fresh to saline

marshes (Table 2.2). Within the fresh marshes, Cl concentration ranged from

42 to 227 mg L ' \ the bayou water had lower concentrations (42 to 44 mg L'^)

while standing water within the marsh had Cl concentrations ranging from 179

to 227 mg L ' \ Brackish and saline areas had chloride concentrations ranging

22


CD■ D

OQ.C

gQ.

■DCD

(/)Wo"303CD

8■D( O '3"

13CD

3.3 "CD

CD■DOQ.C

aO3

■DO

CDQ.

"DCD

roco

Table 2.2 Electrical conductivity, pH and elemental concentration ranges of water samples.

Descriptive

Statistics

EC pH Cl

Element concentration^

--------------------- (mg L ' ) --------------

SO4 B Na Mg Ca K Sr

Mean 20.39 7.66 6319 888.28 1.72 3501.3 412.25 144.95 126.71 2.271

80+ 12.79 0.385 3867 570.51 1.086 2160.5 252.97 81.1 76.46 1.530

Minimum 0.144 6.69 41.9 3.86 0.0323 7.04 3.83 10.46 1.78 0.051

Median 21.15 7.715 6285.5 884.03 1.5372 3491.6 409.76 143.87 121.34 2.226

Maximum 48.39 8.29 13356.9 2232.1 6.3135 7423.7 878.18 293.5 267.5 5.098

50 samples: Standard Deviation

3(/)o'

from 1919 to 13357 mg L ' \ Sulfate concentration In fresh water ranged from

about 4 to 14 mg L' whiles In saline and brackish areas sulfate concentrations

ranged 223 to 2232 mg L ' \

Sodium was the dominant cation. Its concentration ranged from 797 to

7424 mg L' In brackish and saline waters. The samples contained more Mg

than Ca and K, and Mg concentration ranged from 4 to 878 mg L ' \ with

concentrations less than 16 mg L' for the fresh water. Concentration of Ca

and K were similar but Ca was always higher. Boron concentrations ranged

from 0.7 to 6.3 mg L' while Sr concentrations ranged from 0.7 to 5.1 mg L" In

brackish and saline water (Table 2.2).

2.4.2 Electrical Conductivity and Elemental Concentrations

The relationships between EC and cations are presented In Figure 2.2.

Cation concentrations Increased linearly with Increasing EC. Sodium Is the

dominant cation In water samples. Calcium and K were found In equal

concentrations but Mg concentration was higher than Ca and K. The following

regression equations were obtained between the EC (dSm'^) and the cation

concentrations In mg L '\

Na = 1 6 5 * E C + 136.50 (R^ = 0.954) [1]

Mg = 19.342 * EC + 17.82 (R^ = 0.956) [2]

Ca = 6 . 1 4 * E C + 19.75 (R^ = 0.937) [3]

K = 5.742 * EC + 9.62 (R" = 0.922) [4]

24


CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8■D

3.3 "CD

CD■DOQ.C

aO3"OO

CDQ.

■DCD

N)en

gc01'c8coüco%o

8000

7000

6000

5000

4000

3000

2000

1000

0

ONa

+ Mg

o Ca

A K

OO

° ° ° oo _c%o

+ + + + + +

0 10 20 30 40 50

Electrical Conductivity (dS m’ )

Figure 2.2. Electrical conductivity and the major cations of water samples.

C /)C /)

The relationship for EC and 01 concentration is presented in Figure 2 .3 and EC

and SO4 concentration is presented in Figure 2.4. Both Cl and SO4

concentration increased linearly with increasing EC and can be predicted from

following regression equations.

Cl = 2 9 5 * E C + 303 (R" = 0.952) [5]

SO 4 = 43.35 * EC + 4.21 (R" = 0.944) [6 ]

Chlorinity (Cl (%o)) is highly related with salinity, the relationship produced the

following equation that is very similar to that of seawater.

S (%o) = 1.81057 * Cl (%o) - 0 .045176 (R" = 0.999) [7]

The 95% confidence interval is indicated for equation [5] for Cl (Figure 2.3) and

for equation [6 ] for SO4 (Figure 2.4). At a particular EC value if the elemental

concentration (for Cl or SO4) is out of the confidence inten/al, the elemental

concentration that sample is significantly different from the common sample.

This indicates the water samples is not under the influence of seawater

intrusion. This can happen due to concentration of or depletion of the element

within more saline water system. In less saline areas this deviation can simply

indicate lack of seawater intrusion and influence of freshwater system.

2.4.3 Elemental Ratios

Elemental ratios have been used to characterize seawater. These ratios

may also be appropriate indicators for estuarine water characterization.

Relative concentration of other major elements to chloride concentration is

widely used. Element/chloride ratios for the water samples are presented in

26


CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8■D

3.3 "CD

CD■DOQ.C

aO3"OO

CDQ.

S

16000

14000

12000

g 10000c° 8000 co 6000

8ü 4000

d & o

(D) < 0 0

2000

20 3010 40 50

Electrical Conductivity (dS m )

■DCD

C /)C /)

Figure 2.3. Upper and lower confidence intervals (95%) for Cl concentration at different EClevels.

CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8

ci'

33 "CD

CD■DOQ.C

aO3"OO

CDQ.

■DCD

ro00

3000

g 2000CO

Isc8 1000

ow

10 20 30 40 50

Electrical Conductivity (dS m’ )

Figure 2.4. Upper and lower confidence intervals for SO4 concentration at different EC levels.

C /)C /)

Table 2.3, together with ratios for standard seawater at 35%o (SO 4 /Cl, 0.14;

Na/CI, 0.5555; Mg/Cl, 0.06625; Ca/CI, 0.02126; K/CI, 0.0206). Mean SO 4 /CI

ratio was 0.1342, which is close to the ratio for seawater. This measured ratio

was unchanged within the brackish and saline waters. However the S 04 /C I

ratio decreased within the fresh water to 0.0631, suggesting depletion of sulfate

once the water gets into the fresh marsh. Other elemental ratios were also

found close to those of seawater values; mean ratios for Na, 0.5283; Mg,

0.0664; Ca, 0.0394; and K, 0.0228 (Table 2.3).

Figure 2.5 presents the relationship between EC and CI/SO4 ratio. For

seawater at 35%o this ratio is about 7.0. For the water samples from brackish

and saline areas this ratio ranged between 6 and 8 while for the fresh marsh

this ratio was higher and ranged between 9 and 16. Higher Cl content

associated with standing water samples may be due to the concentration effect.

However, it should be noted that the total salts in the fresh water samples are

very low. Chloride was dominant and contributed 56.0% of salinity compared to

SO4 that contributed 7.5 % and the balance (36.5% ) were major cations.

Considering the salt composition data, it may be inferred that the seawater

influence within the Barataria Basin is very extensive because about 50 water

samples collected randomly over a three-year period had the similar salt

composition.

When CI/SO4 ratio data were separated to represent different marsh

types based on their EC, it is evident that the ratios are always higher for the

29


CD■ D

OQ.C

8Q.

■DCD

C/)Wo"303CD

8■D( O '3"

13CD

3.3 "CD

CD■DOQ.C

aO3

"D

O

CDQ.

■DCD

Table 2.3 Element/chloride ratios of water samples.

03o

Descriptive

Statistics

SO4 /CI B/CI

Elemental Ratio”

Na/CI Mg/Cl Ca/CI K/CI Sr/Cl

Mean 0.1343 0.000281 0.52827 0.06644 0.03940 0.02284 0.000473

SD^ 0.0198 0.000099 0.08251 0.00633 0.05390 0.01090 0.000224

Minimum 0.0631 0.000175 0.1681 0.05046 0.02143 0.00785 0.000377

Median 0.1398 0.000245 0.5524 0.06536 0.02292 0.02032 0.000396

Maximum 0.1804 0.000771 0.5710 0.09138 0.27830 0.07772 0.001330

Seawater^ 0.14 N/A 0.5555 0.06625 0.02126 0.0206

3(/)Wo'

^50 samples; Standard Deviation; ^Seawater at 35%o (Wilson, 1975)

8 )

oLO

o

oco

oCM

coS

■I3TJC0 üs

1 (D

CM 00 O "

O Ü B jj l 'o s /1 0

0Q .Em(/)

10LUT3Cco

1o0s1coC01

lOc\iÿ.3.5)u_

31


fresh marsh (Figure 2.5). The mean CI/SO4 ratio values for different marsh

types are presented in Figure 2.6, the ratio for the fresh marsh is significantly

different from other marsh types. This can simply indicate that seawater mixing

does not extend in to the fresh marsh area because the seawater is highly

diluted. Lower S concentrations may be a result of S assimilation by plants and

S transformations within the soil. Sulfur use may be promoted within the fresh

marsh due to very low ionic competition. Rainwater can deposit considerable

amounts of Cl within the coastal areas (Berner and Berner, 1996). The 01

deposition by the rainwater may considerable change the CI/SO4 ratio at the

lower salinity levels within the fresh marsh. Both these mechanisms can result

higher CI/SO4 ratio. The ratio between SO4 and 01 can be used to determine

the extent of seawater intrusion within the basin and also to identify dominant

processes within the marsh. Therefore, OI/SO4 ratio in pore water and standing

water within the marsh may be different from the bay water.

2.4.4 Electrical Conductivity and Salinity

The relationship between electrical conductivity in dSm'^ and total

dissolved solids in ppt, was constructed based on the gravimetric method

(Figure 2.7) and on the elemental analysis data (Figure 2.8). Chlorine was

estimated based on charge balance calculations. The following relationships

were identified using the regression analysis.

Gravimetric method:

TD S = 0.669 * EC - 0.6208 (R" = 0.960) [8 ]

32


CD■ D

OQ.C

gQ.

■DCD

C/)(/)

8■D

3.3 "CD

CD■DOQ.C

ao3T3O

(DQ .

T3(D

(/)(/)

COCO

a:

oü

14

12

10

8

6

4

2

0FM IM BM SM

Marsh Type

Figure2.6. Chloride/sulfate ratio data for different marsh types (differnet letters indicatesignificant difference at a=.05).

in

CO

COo8 _ O CO I O)

oo to o o toto inCO CO CM CM

bdd) SP!|0S P0AIOSS1Q |B}01

COs

I3■oco0

81 (ULU

COQI -T3CmÜLU

(D

COco%20)jC

r-CM

23O)

34


CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8■D

3.3 "CD

CD■DOQ.C

aO3"Oo

CDQ.

■DCD

Ü1

Q .

(/>;d■5CO"O>8

I

30

Y= 0.524X + 0.664225

[r = 0.9453]

200 0

15

10

5

00 10 20 30 40 50

Electrical Conductivity (dS m" )

Figure 2.8. The relationship between EC and JDS.

C /)C /)

Elemental composition:

TDS = 0.524 * EC + 0.6642 (R== = 0.945) [9]

However, the gravimetric method based relationship seems to be similar to the

relationships reported by the U. S. Salinity Laboratory Staff. It should be

cautioned when these relationships are used for lower EC levels such as below

3 dS m' (B reslereta l., 1982).

2.5 Conclusions

Barataria Bay estuary has a chloride dominant water system with a wide

range of salinity. The elemental composition and the ionic ratios indicate that

the water characteristics of this estuary are extensively under the influence of

seawater intrusion. Salinity changes are due to the dilution effect because the

ratios of major elements and chloride were similar to that of seawater.

Seawater intrusion widely controls the salinity within the brackish and saline

marshes but not in the fresh marsh. Sulfate/Chloride ratio is proposed to

estimate the extent of seawater intrusion within the basin and the water flow

within the marsh.

2.6 ReferencesBerner, E. K. and R. A. Berner. 1996. Global Environment: water, air, and

geochemical cycles. Prentice Hall Press, Upper Saddle River, New Jersey. 68-85p.

Bradley, P. M. and J. T. Morris. 1991. Relative importance of ion exclusion, secretion and accumulation in Spartina altem iflora Loisel. J. Expl. Bot. 42(245): 1525-1532.

36


Bresler, E., B. L. McNeal, and D. L. Carter. 1982. Saline and sodic soils: principles-dynamics-modeling. Adv. Ser. Agric. Sci. 10: 1-27.Spring- Verlag, Ne\w York.

Conner, W . H., and Day, Jr., J. W. 1987. The ecology of Barataria Bay Basin: An estuarine profile. Biological Report 85:1-7.

Feijtel, T. C., R. D. DeLaune, and W. H. Patrick, JR. 1988. Seasonal porewater dynamics in marshes of Barataria Basin, Louisiana, Soil Sci. Soc. Am. J. 52:59-67.

Garrepally, R. R. 1994. Spatial and temporal variability in flow, salinity and sediment concentrations in Barataria Bay basin, Louisiana. A masters project report submitted to Louisiana State University and A and M College. 60pp.

Krairapanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1992. Distribution of organic and reduced sulfur forms in marsh soils of coastal Louisiana, Org. Geochem. 18(4):489-500.

Latham, P. J., L. G. Pearlstine, and Wiley M. Kitchens. 1994. Speciesassociation changes across a gradient of freshwater, oligohaline, and mesohaline tidal marshes along the lower savannah river. Wetlands. 14(3):174-183.

Madden, C. J., J. W. Day Jr., and J. M. Randall. 1988. Freshwater and marine coupling in estuaries of the Mississippi River deltaic plain. Amer. Soc. Limnol. And Oceanogr. 33:982-1004.

Rovansek, R. J. 1997. Waterborne materials exchange between marshes and open water of the Barataria Bay estuary of Louisiana, U. S. A. A dissertation submitted to Louisiana State University and A and M College. 158pp.

Soil Conservation Service (SCS). 1989. East Central Barataria cooperative river basin study, Jefferson, Orleans, Plaquemines and St. Charles Parishes, Louisiana. Rev. May 1989. United States. Soil Conservation Service. National Cartographic Center (U.S.). Ft. Worth, TX : USDA-SCS-National Cartographic Center ; [Alexandria, La. State Conservationist, distributor].

U. S. Army Corp Engineer District, New Orleans. 1984. Louisiana coastal area study. New Orleans, La.

37


Wilson, T. R. S. 1975. Salinity and the major elements of sea water, In J. P. Riley and G. Skirrow (eds.) Chemical Oceanography. 365-408p.

38


CHAPTER 3: USE OF ION-EXCHANGE RESIN STIRPS FOR ELEMENTAL ANALYSIS OF ESTUARINE WATER AND COASTAL ORGANIC SOILS

3.1 Summary and Introduction

The potential of using resin strips for analysis of wetland soils, especially

in coastal wetlands, needs to be investigated because of the high soluble salt

concentrations present within coastal organic soils. Salinity and biogeochemical

status influence sulfur levels within the coastal marshes. Inorganic sulfur

fraction may indicate the spatial variability of pedological processes within the

marsh. Sulfur fractionation for marsh soils involves complicated procedures

therefore, resin extractable 8 may be used when large number of samples

need to be analyzed.

A laboratory experiment was conducted to study the influence of salinity

and chloride on resin extractable S using anion-exchange resin strips. Anion

exchange (AE) and cation exchange (CE) resin strips of 10 X 40 mm, were

equilibrated with ionic solutions having different levels of SO 4 and 0 1

concentrations. Bicarbonate, chloride (Cl), and phosphate were tested for

terminal ions. All three anionic forms equally extracted sulfate. Salinity and Cl

significantly reduced the resin extractable sulfate. Ionic strength and competing

ions can limit the affinity of the target ions for the resin.

The potential of using AE and CE resin strips was tested for water

samples collected from brackish and saline marsh types within the Barataria

Bay Basin, Louisiana. Major cations and sulfate in water samples were well

predicted from resin extractions. Inclusion of Electrical Conductivity (EC) into

39


the regression equations improved R , especially for divalent ions. The 0.5M

HCI can preferably be used for elution of both resins. Salinity and the ionic

competition need to be considered when predicting elements for estuarine

w ater samples using resin strips.

The performance of different eluents, effects of equilibration time and

elution time, and soil/water ratio on resin extractable S were studied. An

equilibration time of 16 h was adequate. A soil/water ratio of 1/10 can be used,

but for high sulfur conditions, either lower ratios or more resin strips may be

used. Over 16 h elution levels off the resin extractable 8 , therefore, A 16-h

elution time period can be used.

Resin extractable 8 was compared with water soluble, pyritic and non-

pyritic 8 in coastal organic soils. Resin extractable 8 predicted water soluble 8

for organic soils and non-pyritic 8 for organic and mineral soils. Adding Cl and

pH of soil suspensions to the regression model improved the predictability.

Mineralization of sulfur may limit the predictability similarly in agricultural soils.

However, using more resin strips for equilibration may alleviate this limitation.

This procedure is proposed as an alternative to the traditional analytical

methods, especially for studying spatial variability within the coastal marshes.

Prediction models presented in this paper are recommended for the Barataria

Bay Basin water system and similar estuarine systems.

Marsh types within the Barataria Bay are affected by seawater intrusion.

The marsh is delineated as saline, brackish, intermediate, and freshwater

marshes depending on the salinity levels (Feijtel et al., 1988; 8 oil Conservation

40


Service, 1989). Seawater provides salts to this environment. Sulfur plays an

important role in this environment. It contributes to salinity, is a major plant

nutrient, and is an electron acceptor for the reduced soil and undergoes sulfate

reduction. Excessive sulfate can cause adverse effects on plants (Carlson and

Forrest, 1982; Mayland and Robbins, 1994; Ewing et al., 1995). Plants adapt

through different mechanisms to tolerate high salt concentrations in their root

zone (Carlson and Forrest, 1982; Bradley and Morris, 1991). Composition of

pore water may be different resulting different ionic ratios, due to differences in

ion uptake by the plants (Bradley and Morris, 1991; Latham et al., 1994). Sulfur

uptake may be limited due to high concentration of Cl in saline marshes. Ionic

competition and Cl effects on affinity of S for roots can be studied using resin

strips.

Nutrient phytoavailability can be evaluated using ion-exchange resin

membranes and resin strips have given promising results for upland soils

(Searle, 1988; Schoenau and Huang, 1991; Qian et al., 1992; Searle, 1992;

Lee and Zheng, 1993; McLaughlin et al., 1993; Qian et al., 1994). Resin strips

need to be tested for marsh soils because use of resins will alleviate problems

of complicated soil sampling and preparation processes. Resins will enable a

direct comparison of nutrient status of soils with varying bulk density.

Hydrology within the Barataria Bay basin was monitored as part of the

Louisiana NASA/EPSCoR global change cluster project (Miller et al., 1995). It

was planned to identify indicator elements to monitor temporal changes in

water flow direction because water flow into or out of the marsh can be

41


determined from water composition data. This requires analysis of a large

number of samples. W ater samples within the project area are highly saline.

Salinity ranges from 5 to above 20 ppt within intermediate, brackish, and saline

marshes of Barataria Bay Basin (Soil Conservation Service, 1989).

Inductively Coupled Plasma (ICR) and the Ion Chromatography (1C) are

mostly used for water analysis and require less sample preparation. Estuarine

waters are highly saline and have very high concentrations of soluble salts.

These excessive ionic concentrations interfere with the target ions when 1C and

ICR are used for elemental analysis. Sample dilution to avoid this problem can

dilute the target elements below the detection limits of the instruments. This

problem could be overcome if element concentrations in the water sample can

be predicted from the extractable elements from resin strips. Another

advantage is the possibility to concentrate the elements that are in trace

concentrations in the sample (Edwards et al., 1993).

Ion-exchange resin strips could also be used in soil testing since they

give a quick assessment of elemental concentrations within the soil (Lee and

Zeng, 1993; McLaughlin et al., 1993). Ionic competition due to high salinity in

coastal wetland soils may limit the affinity of target ions on resin (Amer et al.,

1955). This potential limitation must be identified before using them for these

soils. In this study, resin strips were tested for multi-elemental analysis of

estuarine water and coastal organic soils.

42



3.2.1 Multi-Elemental Analysis in Soil Using Ion-Exchange Resins

Ion exchange resins can extract plant available elements from soil (van

Raij et al., 1986). A mesh bag filled with resin beads was used for multi-

elemental soil extraction (Sibbesen, 1977; Somasiri and Edwards, 1992).

Skogley et al. (1990) proposed a phytoavailability soil test. It is a simplified soil-

extraction methodology based on ion-exchange resin extraction of nutrients

from saturated pastes of soil samples. Somasiri and Edwards (1992)

suggested a multi-elemental analytical technique using a mixed resin system

with ICP for element detection. Use of resins eliminates soil sample

preparations procedures (Sibbesen, 1977; Skogley et al., 1990) and overcomes

several changes caused to the soil samples during conventional soil analysis

(Searle, 1988, 1992). Soil phosphate was tested using anion exchange resins

in the form of beads ( Amer et al., 1955; Vaidyanathan and Talibudeen, 1970;

Barrow and Shaw, 1977) and strips (Schoenau and Huang, 1991).

Resin membrane strips have replaced the use of beads (Qian et al.,

1992) and have overcome problems affiliated with beads, especially cleaning

and separation from the soil (Schoenau and Huang, 1991; Somasiri and

Edwards, 1992). These membranes are available in the form of AE and CE

membranes. The AE strips have been used to estimate plant available

phosphate (Schoenau and Huang, 1991), sulfate (Searle, 1988, 1992), and

nitrate (Qian et al., 1994). Phytoavailable Ca, K, Mg, and Cd have been

satisfactorily estimated using CE strips (Qian et al., 1992; Lee and Zheng,

43


1993; McLaughlin et al., 1993). Resin strips were used in trace metal analysis

(Jing and Logan, 1991, 1992; Liang e ta l., 1994; Tejowulan et al., 1994) and

elemental toxicity assessment in saline soils (Greer and Schoenau, 1994).

Szmigielska and Schoenau (1994) used resin strips to estimate 2,4-D amine

retention in soil. The exchangeable nutrient levels estimated with resin strips

correlated well with commonly-used analytical extractions (McLaughlin et al.,

1993) and with plant uptake (Schoenau and Huang, 1991) and were

reproducible for a wide range of soils (Qian e ta l., 1992; Searle, 1992). The

tested plant uptake correlations included P uptake by cotton and flooded rice

(Van Raij et al., 1986);N and S uptake by canola and wheat (Qian et al., 1992).

Sulfate extracted from soil using strips of phosphated AE strips were well

correlated with sulfate extracted by monocalciumphosphate (MCP),

(Ca(H2P04)2), therefore, one can avoid several problems associated with the

MCP extraction method (Searle, 1992). Cation and anion exchange

membranes can be used in multi-elemental analysis (Qian et al., 1992;

McLaughlin et al., 1993). Recent studies have shown that field burial of

membrane strips successfully estimated available nutrients in arable soils

(Schoenau et al., 1994).

Resin strips are simple to use and have been used in simulating cation

or anion sinks (Robinson and Syers, 1990), soil testing (Qian et al., 1992),

mineralization studies (Searle, 1992; Parfittet al., 1994), and soil fertility

assessment (McLaughlin et al., 1993).

44


3.2.2 Effect of Soluble Salts on Membrane Extractable Elements

Soluble salts provide competing anions that may decrease the

concentration gradient of phosphate across the solution film surrounding the

resin (Amer, 1955). A concentration of 20 ppm NaCI reduced the resin's P

adsorption by 1%, but this effect was minor for arable soils (Amer at al., 1955).

However, consideration should be given to the soluble salt effects in resin

performance using coastal marsh soils. Resin extractions under saline field

conditions may represent the phytoavailability for field conditions. Absorption of

P was significantly reduced by low pH (pH 4.0 vs. 5.5) due to reduced diffusion

coefficient of P. Adsorption of P by the resin is independent from resin to soil

ratio and the rate and method of continuous shaking (Amer et al., 1955). Ionic

strength and charge balancing for divalent cations improved P adsorption from

soil (Curtin et al., 1987). Charge balancing may not be important in sulfate

adsorption, but ion competition may influence sulfate adsorption from soil.

3.2.3 Estimation of Membrane Exchangeable S

Sulfate can be estimated using AE strips (Searle, 1988, 1991; Qian et al,

1992). An AE strips (BDH cat. no. 55164) of 40 mm X 4 mm, absorbed

approximately 16,000 |ag g"' S from 0.5 M KgSO^ solution (pH 5.8). The

presence of high concentrations of phosphate did not interfere with sulfate

absorption. This indicates the high affinity of the phosphated-membrane for the

sulfate ion (Searle, 1988). Yang et al. (1990) and McLaughlin et ai. (1993)

indicated that mineralization of sulfur could limit sulfate measurements due to

45


high variability. Searle (1988) estimated resin exchangeable sulfate in New

Zealand soils where carbon content ranged from 2.2 to 39.0 %, including a

Histosol. Mineralizable 8 estimates had a higher degree of variation than

membrane extractable sulfate (Searle, 1992). Reported sulfate estimates with

membrane strips covered a wide range; 2 to 108 pg g' in New Zealand soils

(Searle, 1988, 1992) and 2 .2 to 1195 pg g"" in western Canadian soils (Qian et

al., 1992). These estimates indicate that the exchange capacity of AE strips is

well above the sulfate levels for arable soils however, its capacity for coastal

marsh soils must be tested.

3.2.4 Methodology Adopted in Resin Extraction of S

Searle (1988, 1992) used phosphated AE strips while Qian et al. (1992)

used bicarbonated AE strips to estimate phytoavailable S following a similar

procedure. Resin strips are first converted into an ionic form suitable for

exchanging the target ion by shaking in an ionic solution. Next, the membrane

is shaken in a soiliwater suspension overnight to equilibrate. It is rinsed with

deionized water and shaken in eluent solution to extract the target ion from the

membrane. The eluent solution is analyzed for the target element. Sulfate in

the eluent can be analyzed turbidimetrically as barium sulfate (Searle, 1988,

1992), manually using a spectrophotometer (Searle, 1991), or using ICP (Qian

et al., 1992). The membrane can be regenerated for reuse.

There are limited reports that indicate the use of resin membranes for

estuarine water and sediment analysis. Edwards et al. (1993) proposed the

46


ability of using resin strips for monitoring river water composition. Resin strips

can be placed in water for a time period and then exchanged ions can be

estimated. Relationship between resin extractable elements and sample

concentrations need to be established to use this method. Soluble salts

provide competing anions that may decrease the affinity of the target ion to the

resin strip (Amer, 1955).

The primary objective was to study the effect of salinity and chloride

concentration on resin-extractable 8 to indicate possible 8 accumulation within

the root zone. The secondary objective was to study the feasibility of using ion-

exchange resin strips for elemental analysis for estuarine water and 8

estimation for coastal organic soils.


Anion- and cation-exchange resin membranes (BDH cat. no. 55164-2

and 55165-2, respectively) were used. Resin membranes were cut into 10 X

40 mm strips. The AE strips were kept in 0.5 M NaNCO j (pH = 8.5) for 48 h to

convert into bicarbonate form while CE strips were converted into H form by

keeping them in 0.5M HCI for 48 h. Resin strips were rinsed thoroughly and

stored in deionized (Dl) water to avoid drying.

The general procedure for resin extraction is presented in Figure 3.1.

Resin strips are shaken in soil:water suspensions or water samples overnight

(16 h) for ion exchange. The strips are rinsed with Dl water and transferred into

the eluent (25 to 50 mL of 0.5M HCI). The elements exchanged on to the resin

are eluted after shaking in the eluent for 24 h. The eluent solution is analyzed

47


CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8■D

CD

3.3 "CD

CD■DOQ.C

aO3"OO

CDQ.

■DCD

C /)C /)

Resin Strips

H C 03-water sample (50 mL)

or soil suspension

40 m m X 10 mm

(Regeneration)> <

00

— \ (Shake 24 h)

(Rinsed with Dl water)

(Analysis by ICP)

50 m l of eluent 5 M HCI)

(Shake 24 h)

Figure 3.1. Schematic presentation of the general procedure for resin extraction.

for the target element using Inductively Coupled Plasma (ICP-AES). Resin

extractable elements can be expressed in concentration units or based on unit

resin area or unit soil weight, for data analysis. The resin strips are regenerated

for reuse.

3.3.1 Salinity and Chloride Effect

The AE strips were used in three ionic forms, bicarbonate, chloride, and

phosphate separately. Five sulfate levels, 300, 600, 1500, 2400, and 3600 mg

were obtained using KgSO^. Salinity levels were adjusted to approximately

5 ,1 0 , 15, 20, 25, and 30 dS m' by adding KOI to each sulfate level. Each

sulfate level had its lowest salinity level without chloride. Different salinity levels

were obtained based on the relationships reported by the U. 8 . Salinity

Laboratory Staff (Bresler et al., 1982) as given in equation 3.1. Salinity is

expressed as total dissolved solids (TDS).

TDS (mg L'^) = 0.64 * 10^ EC (dS m'"') [3.1]

The general procedure was followed for resin extraction with 24 h equilibration

and 24 h elution times.

3.3.2 Multi-Elemental Analysis of Estuarine Water

W ater samples were collected from different water bodies within the

brackish and saline marsh types of the Barataria Bay Basin, Louisiana. W ater

samples used for the experiment were brought to the laboratory in containers

placed in an ice-filled cooler. W ater samples were filtered with Whatman No.

42 filter paper. The EC of the water samples was measured using the

49


conductivity sensor of a M 90 pH, conductivity and dissolved oxygen meter

(Corning Inc., Corning New York 14831, USA). Water samples were analyzed

using ICP-AES (PERKIN ELMER model OPTIMA 3000 (Perkin-Elmer

Corporation, Norwalk, CT 06859, USA) and pH was measured with a pH meter

(Orion Research model 701A/digital lONALYZER).

Twenty-seven water samples having a wide range of ECs were used for

this experiment. A 50-mL water sample was used for resin extraction following

the general procedure (Figure 3.1) with 24 h for shaking and elution times, in

four replicates. The AE and CE strips were eluted separately. The CE strips

were eluted with 25 mL of 0.5M HCI. Two replicates of AE strips were eluted

with 25 mL of 0.5M HCI while the other two were eluted with 25 mL of 0.5M

NaHCOg to compare the performance of the two eluents, HCI and NaHCO j for

AE strips. The NaHCO^ eluent can interfere with low sulfur detection by the

ICP-AES.

3.3.3 Resin-Extractable 3 Analysis for Coastal Organic Soils

Three preliminary experiments were conducted to study the effects of

equilibration time, elution time and soil/water ratios on resin extractable S for

organic coastal soils.

3.3.3.1 Equilibration and Elution Time Experiment

Time periods, 1, 3, 6 , 10 ,16 , and 24 h were tested to determine the

effect of equilibration time on resin extractable S. Soil collected from a saline

marsh surface organic layer was used for the study. Soil/water ratio of 1 ; 10

was used mixing a 5-mL scoop of soil with 50 mL of Dl water in triplicate.

50


General procedure (Figure 3.1) was followed for resin extractions changing the

shaking time for different samples. In separate sets of three replicates, elution

times of 2, 3, 6 , 10, 16, and 24 h were tested using the same 1/10 soil/water

ratio.

3.3.3.2 Soil/Water Ratio Experiment

Surface organic soil at different ratios of soil/water, 1/5, 1/10, 1/20, and

1/40 were used for resin extractions. Shaking time and elution time was 24 h.

The A E strips were eluted separately with 0.5M HCI.

3.3.3 3 Resin-Extractable and Water Soluble S Analysis

Soil samples were collected from the organic subhorizones and the

mineral horizon of saline and brackish marsh soil profiles of Barataria Bay,

Louisiana. Soil suspensions obtained by mixing 5 mL of soil and 50 mL of Dl

water (1/10, soil/water) were shaken for 24 h following the general procedure

(Figure 3.1) with three replicates for each sample. The AE strips were shaken

for 24 h and the eluted for 24 h in 50 mL of 0.5M HCI. Similar soil/water

suspensions were shaken for 24 h without resins to estimate water soluble S,

EC, chloride concentration and pH of the soil suspension. The EC of the soil

suspensions was measured with the conductivity sensor of M90 pH,

conductivity meter (Corning Inc., Corning New York 14831, USA). The filtrate

was analyzed using ICP-AES (PERKIN ELMER model OPTIMA 3000 (Perkin-

Elmer Corporation, Norwalk, CT 06859, USA), and pH was measured with a pH

meter. Cl was measured using a Cl specific electrode with a pH/mV meter

51


(Orion Research model 701A/digital lONALYZER). Resin extractions were

compared with water soluble S, non-pyritic S and pyritic S for organic soils.

3.3.3.4 Pyritic and Non-Pyritic S Analysis

A method used for pyrite estimation in coal and ceramics was modified

for the analysis of extractions for Fe and S by the ICP-AES. Pyritic 8 was

estimated indirectly from the pyrite content (Schneider and Schneider, 1990).

One scoop (1/8 tsp.) of soil was measured into a digestion tube. Six tubes that

were calibrated to 50 or 75-mL volume, were used for one soil sample. Tubes

were separated into two batches. One was for non-pyritic (NP) and the other

set was for pyritic (P) and non-pyritic (NP) Fe digestion. The 50-mL tubes

were used for non-pyritic Fe, while the 75-mL tubes were used for pyritic and

non-pyritic Fe. A sample was placed at the bottom of the tube. Six mL of

concentrated HCI were added into each tube with soil, under a fume hood and

the tube was covered with a small glass ("reflex") funnel. The soil was stirred

within the acid to make a suspension and left over night, covered with "reflex"

funnels. The next morning, 23 mL of Dl water was added to each tube. The

digestion procedure was different for "NP" and “NP+P" sets, therefore they

were digested separately. To digest the NP set, the sample was placed onto

the digestion block, and the block heater was turned on. It took about 15 to 20

min for the mixture to start boiling. The samples were kept boiling for one hour

and were removed and allowed to cool. The tubes were filled to the 50 mL

level with 1:10 HCI then mixed well and filtered through #5 (or #42) Whatman

52


filter paper. The filtrate was highly acidic; therefore, it was diluted six times to

reduce the acidity and to increase the detection levels by the ICP.

For the “NP+P” set, after adding 23 mL of water, the sample was boiled

for 30 min and then 25 mL of 3N HNO 3 was added to each tube while they were

on the digestion block. The acid was slowly added through the funnel. It took

another 10 to 15 min for the mixture to start boiling. Samples were boiled for

another 30 min and cooled. The tubes were filled to the 75-mL level with 2N

HNO 3 then mixed well and filtered through #5 (or #42) Whatman filter paper.

The filtrate was highly acidic; therefore, it was diluted six times before analysis

by the ICP. The pyritic Fe content was determined based on the difference

between Fe content of “NP+P” and “NP” digestions. The pyritic Fe content was

estimated using equation 3.2.

Pyritic Fe (rng L ') * 1.1487 = Pyritic 8 (mg L ') [3.2]

The S content from the “NP” digestion was the non-pyritic 8 content (Schneider

and Schneider, 1990).

3.3.4 Data Analysis

Regression analysis was performed for the data using the general linear

model procedure (PROC GLM) in Statistical Analysis System (SAS Institute

Inc., 1989) to identify the relationships between the amount of resin-extracted

elements and the elements in water. Single and multiple linear regression

models were fitted for different elements. The EC of water sample was included

as an additional variable. Mean comparison was obtained using the PROC

GLM procedure for different eluents. Regression analysis and analysis of

53


variance (ANOVA) were performed to identify the effects of equilibration time,

elution time, soil/water ratios on resin extractable S. Correlation and regression

analyses were conducted to study the relationship between resin extractable S,

water soluble S and pyritic and non-pyritic S.


3.4.1 Salinity and Chloride Effect

Resin extracted sulfate content was expressed as mg SO 4 cm'^.

Maximum sulfate extracted on the resin strip was 0.59 mg SO 4 cm that is

equivalent to 8287 pg SO 4 g ^soil. Figure 3.2 presents the effect of salinity and

sulfate levels on resin extractable sulfate. Salinity reduced the amount up to

0.007 mg SO 4 cm'^, which is a drop from 100% to 5 %. Statistical analysis

indicated that salinity significantly reduced (p=.0 0 1 ) resin extractable sulfate

content at each sulfate level. Both salinity and chloride concentrations were

highly correlated with resin extractable sulfate. Figure 3.3 presents the

competition by Cl ions with sulfate on resin exchange. The maximum resin

extractable sulfate is limited by the resin’s exchange capacity. The competing

individual ions and the total soluble ionic concentrations affect the affinity of the

target ions to the resin. The interactive effect of 01 and sulfate on resin

extractable sulfate is presented in Figure 3.4. A linear relationship is noted for

both Cl and sulfate levels on resin extractable sulfate.

54


CD■ D

OQ.C

8Q.

■DCD

C/)C/)

8■D

CD

3.3 "CD

CD"OOQ.CQ

■DO

CDQ.

■DCD

enen

0.60

-0-3O O ppm

-B -6 0 0 p p m

-z^ 1500p p m

-* -2 4 0 0 p p m

—h"SBOOppm

0.50

I- 0.40co^ c f "

■G o 0.30 2

SC

I 0.20

0.10

0.0010 200 5 15 25 30

-1 ,Electrical Conductivity (dSm )

Figure 3.2. Effect of salinity and sulfate level on resin extractable sulfate.C /)C /)

CD■ D

OQ .C

gQ .

■DCD

C/)C/)

CD

8■D( O '

3.3"CD

CD■DOQ .C

ao3T3O

(DQ .

Ü1O)

0.50

E 0.40 u

ê0)g 0.303

W0)SI

50.20

SI 0.10CH

0.001000 2000 3000 40000 5000 6000 7000

T3(D

Chloride (mgL’ )Figure 3.3 Effect of chloride concentration on resin extractable sulfate at 600 mg S L-1

(/)(/)

O'

' I AQq.

I.:I

Figure — "

3.4.2 Multi-Elemental Analysis of Estuarine Water

Elemental concentrations, EC, and pH ranges of water samples used in

the experiment are presented in Table 3.1. Descriptive statistics for different

elements, EC and pH were obtained by SAS PROC UNIVARIATE procedure

(SAS Institute Inc., 1989). The EC, pH and major elements were normally

distributed. Sodium, Mg, Ca, and K were the major cations and chloride and

sulfate were the major anions in the water samples. Major cations maintained

the same proportions within the EC range used for the study. For example, out

of the major cations on weight basis, Na was dominant and was 83.5% (SD ±

0.7), with much less proportions of Mg 10 (±0.1)% , Ca 3.6 (±0.5)% , and K 3.0

(±0.1)% . Major cations and sulfur concentrations were highly related to the

electrical conductivity of the water samples; this indicates seawater is the main

source of the elements. Figure 3.5 presents the relationship between electrical

conductivity and concentration for sulfur and potassium. Element

concentrations were linearly related to EC; linear regression accounted for 96%

of the variation for S and 97% of the variation for K. Similar relationship was

observed for the other elements.

Statistical analysis indicated that the use of both resins together did not

make a significant difference in resin extracted elements; this was true for both

anions and cations. Means of different eluents used for anionic resins were

not significantly different. This suggests that 0.5M HCI can be used as the

eluent for anionic resins. This avoids the variability due to dilution of NaHCOg

58


CD■ D

OQ.C

gQ.

■DCD

C/)Wo"303CD

8■D( O '3"

13CD

"nc3.3 "CD

CD Ü 1■D COOQ.C

aO3

■DO

CDQ.

■DCD

(/)(/)

Table 3.1. Electrical conductivity, pH and elemental concentration ranges of water samples used in the experiment.

Descriptive

Statistics

EC pH

Element concentration^

Ba Ca Cu K Mg Na S Sr Zn

Mean 25.04 7.91 0.173 162.72 0.043 139.87 473.22 3996 327.41 2.816 0.097

SD^ 12.97 0.19 0.051 73.49 0.041 70.58 236.98 2038 167.19 1.338 0.078

Minimum 5.37 7.54 0.074 46.93 0.003 31.36 99.24 797 73.14 0.691 0 . 0 0 2

Median 22.80 7.92 0.181 148.15 0.029 124.96 436.03 3623 301.98 2.610 0.072

Maximum 52.70 8.29 0.330 293.51 0.171 267.53 878.18 7423 636.33 5.098 0 . 2 2 1

27 samples: Standard Deviation

CD■ D

OQ.C

gQ.

■DCD

C/)C/)

CD

8■D( O '

3.3"CD

CD■DOQ.Cao3

T3O

CDQ.

O)o

800

A Potassium (K)

Y = 12.66X+ 10.52 [r2 = 0.96]O Sulfur (8 )600 -

O)Eco

5 400 -

§8

E 200 - 0}

111Y = 5.35X + 5.87

[r2 = 0.97]

5 10 150 20 25 30 35 40 45 50

T3CD

(/)(/)

Electrical conductivity (dS m' )

Figure 3.5. Potassium and sulfur concentration as influenced by electrical conductivity in water samples.

eluent for ICP analysis. Therefore, only the data from HCI eluent were used for

regression analysis of resin extracted sulfate.

Resin extracted elements exhibited curvilinear relationships with

elements in water samples, as well as with the EC. Resin extracted-K

increased linearly (r = 0.87) with the increasing amounts of K in water (Figure

3.6) but decreased exponentially (r = 0.94) with increasing EC (Figure 3.7).

Total major cations extracted by the resin at all EC levels were similar

and were about 0.1 meq. However, the percentage of different cations

exchanged on to the resin changed with EC. More Ca and Mg were exchanged

at lower EC levels than Na and K. When the EC increases, monovalents (Na

and K) dominated the system and exchange to the resin was greater. Figure

3.8 shows the resin-extracted proportions of major cations in milliequivalents.

However, the milliequivalent percentages in water were similar at all EC levels;

Na 78% , Mg 17%, Ca 3.5% and K 1.5%. Skogley et al. (1990) proposed a

phytoavailability soil test because resins are known to mimic plant roots. Based

on this information it seems at high salinity, plants may absorb excess Na but at

lower salinity Mg and Ca are in excess. Affinity of sulfate also was dramatically

reduced with the increase in salinity (Figure 3.9). Chloride can be the major

competitor with sulfate, similar to cations at high salinity monovalents dominate

the system.

The following abbreviations are used: amount of element in the water

sample (W(g)) and the quantity eluted by the anionic resin strips (ARE^g,) and

61


CD■ D

OQ.C

gQ.

■DCD

C/)

o'3O

8

ci'

33 "CD

CD■DOQ.C

aO3

■DO

CDQ.

■DCD

C /)C /)

S

0.10

Y = 0.0039X + 0.0334

[r = 0.87]0.08 -

O)

& 0.06 - bdXJd)o2

0.02 -

0.000.00 2.00 4.00 6.00 8.00 10.00

K in water (mg)

Figure 3.6. Resin extracted K at different K levels in water.

12.00 14.00

CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D

3CD

3.3 "CD

CD■DOQ .C

aO3

■DO

CDQ .

■DCD

C /)C /)

G )W

0.025

Y = 0.0565 X-0.574

[r = 0.94]0.020

O)E^ 0.015 - •o 0)c32

Iç 0.010 - tnK

0.005 -

0.00010 20 300 40 50

Electrical Conductivity (dS m’ )Figure 3.7. Relationship between electrical conductivity and resinextracted-K.

m O) mZ Üo □ < O

in

o

o0 0

oCN

o

ooCN

oo

en

.>o=3

X Jcoo

s(U

LU

go

!■§'E

mO )

IS0)Q .

I53O"0)

T3

E

C

'%0 :0000

E3O)

(baui p %) p e u i8 | 3 papejpcg ujsay

64


CD■ D

OQ .C

gQ .

■DCD

C/)

o"3O

8■D

3.3 "CD

CD■DOQ .C

aO3

■DO

CDQ .

■DCD

C /)C /)

O)Ol

0.02

Y = 0.082 X-1.0821

[r = 0.99]0.01 -

cr0)Ed)

3w

0.00 -

0.005 10 15 20 25 30 350 4540 50

-1 ,Electrical conductivity (dS m )Figure 3.9. Sulfate milliequivalents exchanged onto the resin at different EC levels of

water samples

the cationic resin strips (CRE^g)), where e stands for element. Linear and

quadratic regression models were fitted between resin extracted-elements RE^g,

and W(e). The EC was included as a second variable with first, second and third

powers of RE^g,. The best regression models and the statistics for each

element are presented in Table 3.2. Addition of the electrical conductivity of

the water sample (EC) as a second variable improved the R values. Major

cations did not achieve much improvement, but in some cases, addition of EC

accounted for 30 to 70% of additional variability compared to that of ARE^g, or

CRE(g, alone. A linear regression of CREj^g) alone accounted for 91 % of the

variation in W,Na). For Na and K, quadratic relationships barely improved

predictions. Adding EC increased R^ by another 6 % for Wj^a) and 10% for W^^).

Adding EC increased the R for other models considerably. Addition of EC

increased the R by 35% (R^ 0.62 vs. 0.97) for and explained 96% of the

variation in W(ca>- Much larger improvement was observed for Sr when EC was

included in the model (R^; 0.22 vs. 0.96). This points out the importance of

including EC to predict minor elements in the system. The best models only

accounted for about 30 % of the variation in \N^zn) and 25 % of the variation in

W(Ba)- This may be due to the presence of very low concentrations of these

elements. However, prediction of V^cu) was better (R = 0.66) than V^^n) and

W(Ba) and this may be due to the better detection of Cu by ICP-AES. Anionic

66


CD■ D

OO.C

gQ .

■DCD

C/)

o"3O

8■D

( O '3"

i3CD

"nc3.3 "CD

CD■DOQ .C

aO3

■DO

CDQ .

O)

Table 3.2. Regression model for different elements to predict elements in the water using resin extracted elements.

Element Regression Model'*’ R p-value

Ca W(ca) = 1.9284 -1.4331 * CRE(ca) + 0.2676 * EC 0.96 0 . 0 0 0 1

Mg W(Mg) = 4.5162 - 5.0493 * CRE(Wg) + 0.8676 * EC 0.96 0 . 0 0 0 1

Na W(Na) = -26.04 + 91.994 * CRE(Na) + 6.0834 * EC 0.97 0 . 0 0 0 1

K W(K) = -0.7685 + 32.2 * CRE(K) + 0.2321 * EC 0.97 0 . 0 0 0 1

Sr W(sr) = 15.658 - 0.0733 * CRE(Sr) +5.0208 * EC 0.96 0 . 0 0 0 1

SO4 W(so4) = 7.4964 -12.8661 * ARE(so4) + 1 7463 * EC 0.96 0 . 0 0 0 1

Zn W(zn) = -6.9058 + 40.3568 * CRE(zn) - 26.3448 * [CRE<zn)] 0 . 2 2 0 . 0 0 0 1

Ba W(Ba) = 4.49 + 8.82 * CRE(Ba) - 3.91 * [CRE<Ba)f + 0.46 * [CRE(Ba)f 0.29 0 . 0 0 0 1

Cu W(cu) = 2.25 - 5.0523 * CRE(Cu) + 5.4185 * [CRE(Cu)]^ 0 . 6 6 0 . 0 0 0 1

W(e) = element in water; CRE(e) = cationic resin extracted amount;

m ARE(e) = anionic resin extracted amount; weights in mg for Na, Mg, Ca, K and SO4 , g for others)

U)'w EC = Electrical Conductivity of water sample (dS/m)

resin extracted-sulfate explained 65% of the variation in W(so4), and addition of

EC into the model increased the to 0.96. The models are indicative of the

affinity of individual ions for the resin. Improved with additional variable of

EC indicates to what extent their affinity is influenced by ionic competition.

Factors contributing to ionic activity such as ionic strength, and ion pairing may

be responsibly controlling the affinity of individual ions onto the resin. These

models emphasize the role of EC in predicting elements using resin extractions.

Other elements that were not present in adequate amounts in this study need

to be calibrated separately and may be predicted with resin extraction. The

water system in Barataria Bay Basin is dominated by Cl and Na* ions. High Cl

concentration interferes with the detection by Ion Chromatograph (1C) because

it’s peak spreads over the range of other target anions, such as SO 4 , PO 4 and

NO 3 . High Na in the matrix interferes with the target ions during ICP analysis.

This will give proper estimates as well as protect the instrument from damage

due to high salt concentrations. As indicated by the models presented in Table

3.2, the use of resins can be a suitable alternative for estuarine water analysis.

Mean comparison indicated no significant difference between eluents;

therefore, 0.5M HCI can preferably be used for elution of anionic resins. Major

cations and sulfate were well predicted by resin extractions. Inclusion of EC

into the regression equations improved the R . Addition of EC as a second

variable increased the R considerably for the divalent ions Mg, Ca and Sr. The

regression models emphasize the role of EC in predicting elements using resin

extractions in high saline conditions.

68


3.4.3 Resin Extractable S in Coastal Organic Soils

In the following experiments sub samples taken from a bulk soil sample

collected from the saline marsh type surface organic layer were used for

different treatments. Therefore, micro-site variation can be expected within the

soil sample used for the following experiments.

3.4.3.1 Equilibration and Elution Time

Effect of equilibration time on resin extractable S (RES) is presented in

Figure 3.10. The RES increased linearly within the equilibration times tested.

The overall effect of equilibration time on RES was significant (p=.001). The

equilibration periods less than 16 h, were significantly different from 16 h. The

24 h period extracted significantly higher RES than the 16 h equilibration.

However, considering the possible limitation of resin’s capacity, a 16 h

(overnight) equilibration time is recommended for coastal organic soils.

The overall elution time effect was significant on RES (p=.032). This

may be due to the micro-site variation within the soil sample used for the test

because no relationship was observed within the different elution times tested

and RES (Figure 3.11). However, it shows that the RES levels off after 16 h.

Therefore, 16 h can be a suitable time period as elution time.

3.4.3.2 Soil/Water Ratio

Soil/water ratio significantly influenced the amount of RES (p=.002). The

RES at different soil/water ratios is presented in Figure 3.12. Soil/water ratio of

0.2 (1 :5) was significantly lower than other ratios. Dilution of the soil

69


GGDoinCOooo

(O I

in

OOD

OOO

CO

CM

0)Ei -O)cm

CO

ooCM

oin oo oin

( 1 6 lu ) s a iq e p B jp c g u is e y

70

JO’5too'cmg0E

CO

JO§

1Ç

mc o 0)

CO ™

co

*5cr0 )

Ü

Ê

CO23O)


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D

CD

3.3 "CD

CD■DOQ .C

aO3"OO

CDQ .

■DCD

C /)C /)

300

250

|) 200

cooro 150t j2Sç 1 0 0

I

8 12 164 20 240

Elution Time (h)

Figure 3.11. Effect of elution time on resin extractable S.

om

oCMO O

oin

oo

oino

OOO

oo-* o o ooooin

omoooo

(0cr

I

</>Ç)

I2Co01

Ig 8

ü

iCNi

co2=3O)

co CNJ CNJ ^ ^

H 6 lu) s aiqepBjpcg uisay

72


suspensions decreased the EC and that increased RES due to low ionic

competition. At a ratio of 0.025 (1 ;40) the RES was more than two folds

compared to 0.2 ratio. Further dilution may limit the resin strip’s exchange

capacity for the RES. For saline marsh soils, the ratio of 0.1 may be suitable,

but at lower salinity levels, it may be advisable to use lower ratios.

3 4.3.3 Resin Extractable S

There were 112 soil samples used for the analysis, 92 of them were from

organic subhorizons (Oa horizon), and 20 samples were from mineral horizon

(Cg horizon). Resin extractable S (RES), water soluble S (WSS), pyritic S

(PYS), and non-pyritic S (NFS) were measured in triplicate for each sample.

The EC of the 1:10 soil/water suspensions ranged between 0.07 to 1.37 dS m'

which is approximately equivalent to field salinity ranging from 0.77 to 15.07 dS

m '\ The 1:10, soihwater suspensions diluted the salinity effect that limits the

affinity of target ions on to the resin. This may create problems with resins

capacity for exchangeable ions.

Some variables followed a distinct bimodel distribution, therefore, it was

decided to separate data for organic and mineral soil layers for regression

analysis. Figure 3.13 presents the ranges of different S fractions of organic and

mineral soils. The S content is expressed based on soil volume as mg S L' of

soil. The following ranges were observed: RES 69 to 790 mg S L' for the

organic horizons, 64 to 1038 for the mineral horizons, W SS between 142 and

4307 mg S L' for the organic horizons and 813 and 5015 mg S L' for the

73


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D

CD

3.3 "CD

CD■DOQ .C

aO3"OO

CDQ .

S (mg r )

5000

4000

3000

2000

1000

ORES OW SS OPYS GNPS

MRES M W SS MPYS MNPS

S Fraction and Soil Horizon 0 = organic, M= minerai horizon

■DCD Figure 3.13. Boxplots showing data for different S fractions for organic and minerai horizons.

C /)C /)

mineral horizons, PYS between 81 and 5731 mg 8 L' for the organic horizons,

and 121 and 3853 mg 8 L‘ for the mineral horizons, NP 8 between 276 and

1307 mg 8 L' for the organic horizons, and 245 and 969 mg 8 L' for the

mineral horizons. The W 8 8 indicate high variation compared to R E 8 and NP 8 ,

this may be due to 8 mineralization during the measuring procedure. Chloride

concentrations ranged from 5.6 to 865 mg L' in soil/water suspensions, which

is equivalent to 61.6 to 9515 mg L' in the field. The pH ranges differed

between organic and mineral horizons, it ranged from 3.99 to 7.45 for the

organic horizon while for the mineral horizon ranged from 5.67 to 8.51.

Figure 3.14 presents the relationship between RE 8 and W 8 8 for the

organic horizons. The RE 8 increased with increasing W 8 8 . Figure 3.15

presents the relationship between RE 8 and NP8 for the organic horizons.

Figure 3.16 presents the relationship between RE 8 and NP 8 for the mineral

horizons. 8 eparation of mineral and organic horizons improved the regression

coefficients for individual models compared to the models used for all the soil

samples (R^ < 0.5). Different models to predict NP 8 and W 8 8 are presented in

Table 3.3 for organic and mineral horizons. Better prediction can be achieved

using different models for different soil horizons. Introducing pH and 01

concentration of the soil suspensions improved R^ for the organic horizon

models compared to mineral horizons. The W 8 8 had high variation, therefore,

none of the models predicted W 8 8 well except the model with logarithmic

75


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

CD

8■D( O '

3.3"CD

CD■DOQ .C

aO3"OO

CDQ .

•n JO)

4000

3000O )EW _0)

I 2000w

iO o1000 Oo

2000 100 300 400 500 600 700 800

■ DCD

(/)o"3

-1 vResin Extractable S (mg L )

Figure 3.14, Relationship between RES and WSS for organic horizons.

CD■ D

OQ .C

gQ .

■ DCD

C/)

o"3O

8■D

3.3 "CD

CD■DOQ .C

aO3

■DO

CDQ .

1 4 0 0

1200

1000o O

O )

& 8 0 0 wü

O oi , 6 0 0Q.c o Z

o O

4 0 0

200

3 0 0 5 0 0100 200 4 0 0 6 0 00 7 0 0 8 0 0

T3(D

(/)(/)

Resin Extractable S (mg L’ )

Figure 3.15. Relationship between RES and NFS for organic horizons.

CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8

3.3 "CD

CD■DOQ .CaO3"OO

CDQ .

N00

O)Ewü

èz

1200

1000

800

600

400

200

0200 400 600 800 1000

T3(D

(/)(/)

Resin Extractable S (mg L' )

Figure 3.16. Relationship between RES and NFS for mineral horizons.

CD■ D

OQ.C

gQ.

■DCD

C/)Wo"303CD

8■D

( O '3"

13CD

"nc3.3 "CD

CD■DOQ.C

aO3

■DO

CDQ.

■DCD

Table 3.3. Regression models to predict different sulfur fractions'* from RES for organic and mineral soils layers.

CD

Soil Layer Regression Model RZ p-value

Organic WSS = 2 .1 6 9 5 + 0 .8 2 2 2 * Log (RES) + 0 .0 3 5 3 * Log (Cl) 0 .6 2 0 .0 0 0 1

Organic NFS = 3 3 0 + 1 .0 5 6 1 * (RES) 0 .4 8 0 .0 0 0 1

Organic Log (NFS) = 5 .9 2 3 1 + 0 .0 0 1 7 * (RES) 0 .5 5 0 .0 0 0 1

Organic NFS = 1 .0 1 7 5 * (RES) + 9 0 .7 9 4 4 * pH - 1 6 2 .8 3 0 .5 8 0 .0 0 0 1

Organic Log (NFS) = 5 .1 9 5 1 + 0 .0 0 1 5 5 * (RES) + 0 . 1 3 4 0 4 * pH 0 .6 5 0 .0 0 0 1

Organic Log (NFS) = 3 .0 9 6 1 + 0 .4 9 2 3 * Log (RES) + 0 .1 0 3 5 * pH 0 .6 5 0 .0 0 0 1

Organic Log (NFS) = 4 .2 0 4 4 + 0 .3 2 7 9 * Log (RES) + 0 .0 6 7 5 * Log (01) 0 .7 1 0 .0 0 0 1

Organic NFS = 4 .2 8 3 1 + 0 .3 0 2 3 * Log (RES) + 0 .1 0 6 1 8 * Log (01) 0 .71 0 .0 0 0 1

Mineral NFS = 2 7 9 .7 + 0 .6 5 6 5 * (RES) 0 .7 3 0 .0 0 0 1

Mineral Log (NFS) = 3 .9 6 3 3 + 0 .3 9 7 7 * Log (RES) 0 .7 3 0 .0 0 0 1

Mineral Log (NFS) = 4 .5 8 3 7 + 0 .4 6 4 1 * Log (RES) - 0 .1 3 5 4 * pH 0 .7 7 0 .0 0 0 1

* W SS = water soluble S; RES = resin extractable S; NFS = Non-pyritic S.

(/)(/)

conversion of RES and Cl. Sulfur mineralization is a limitation reported in other

studies, for resin use (Searle, 1988). Sulfur mineralization may be the main

cause of low for W SS. Coastal organic soils are rich in other reduced S

compounds, in addition to the organic sulfur fraction that can be mineralized

during shaking for W SS and RES measurements. However, the models are

able to predict over 70 % of the variation despite the variations in marsh types,

salinity, chloride, organic/mineral ratio and pH. Resin extractable S can be

suggested to estimate non-pyritic 8 and water soluble S for spatial variability

studies.

3.5 Conclusions

Resin strips can be used to analyze the cation and anion concentrations

in water accurately, overcoming the high salt concentrations in the saline water

samples. Salinity and the ionic competition need to be considered when

predicting elements for estuarine water samples using resin strips. The models

predicted elemental concentrations in water samples while demonstrating the

ionic competition effect on individual ions. Salinity significantly reduces the

resin extractable sulfate. Either Cl concentration or EC can be added to the

model to predict elements in water from resin extractable elements.

Bicarbonate, Cl or phosphate can be used as a terminal anion for the AE strips.

The procedure and prediction models can be recommended for the Barataria

Bay Basin water system and similar estuarine systems.

80


Both AE and CE strips can be used together for one sample and 0.5 HCI

can be used as the eluent for both resins if the ICP-AES is used for elemental

analysis. This procedure can be proposed as an alternative to the traditional

analytical methods. Equilibration time, elution time and the soil/water ratio

needs to be adjusted depending on the resin strips’ capacity, 8 levels and

salinity of the samples.

The RES maintained a linear realtionship with W SS and N FS in both

mineral and organic soils. Separation of mineral soils and organic soils

improved the regression coefficients, for individual models compared to the

overall model. Therefore, better prediction can be achieved using different

models for different soil types. Including pH or Cl concentration and

transforming data helped to obtain better prediction models for N FS and W SS,

especially for organic soils. Sulfur mineralization is a limitation, reported in

other studies, for resin use for S estimates (Searle, 1988). Sulfur mineralization

may be the main cause of low for models predicting W SS. However, the

models predicted over 70 % of the variation in NFS and W SS, despite the

variations in marsh types, salinity. Cl, organic/mineral ratio and pH. Resin

extractable S can be used to estimate non-pyritic S and water soluble S for

spatial variability studies.

3.6 References

Amer, P., Bouldin, D. R., Black, C. A. and Duke, F. R. (1955). Characterization of soil phosphorus by anion exchange resin adsorption and F32- equilibration. Fiant Soil 6:391-408.

81


Barrow, N. J., and T. C. Shaw. 1977. Factors affecting the amount ofphosphate extracted from soil by anion exchange resin. Geoderma 18:309-323.

Bradley, P. M., and J. T. Morris. 1991. Relative importance of ion exclusion, secretion and accumulation in Spartina alterniflora Loisel.. J. Expt. Bot. 42:1525-1532.

Bresler, E., B. L. McNeal, and D. L. Carter. 1982. Saline and sodic soils: principles-dynamics-modeling. Adv. Ser. Agric. Sci. 10:1-27.Spring- Verlag, New York, New York.

Carlson, Jr., P. R., and J. Forrest. 1982. Uptake of dissolved sulfide bySpartina alterniflora: Evidence from natural sulfur isotope abundance ratios. Science 216:633-635.

Curtin, D., J. K. Syers, and G. W. Smillie. 1987. The importance ofexchangeable cations and resin-sink characteristics in the release of soil phosphorus. J. Soil Sci., 38:711-716.

Edwards, T., B. Ferrier, and R. Harriman. 1993. Preliminary investigation on the use of ion-exchange resins for monitoring river water composition. The Science of the Total Environment 135:27-36.

Ewing, K, K. McKee, I. Mendelssohn, and M. Hester. 1995. A comparison of indicators of sublethal salinity stress in the salt marsh grass, Spartina patens (Ait.) Muhl. Aquatic Botany 52:59-74.

Greer, K. J., and J. J. Schoenau. 1994. Salinity and salt contaminationassessment using anion exchange resin membranes. Proceedings of Soils and Crops Workshop, 1994, Univ. of Saskatchewan, Saskatoon, Sask., Canada. p44-48.

Jing, J. and T. J. Logan. 1992. Effects of sewage sludge cadmiumconcentration on chemical extractability and plant uptake. J. Environ. Qual. 21:73-81.

Jing, J., and T. J. Logan. 1991. Chelating resin method for estimation ofsludge-cadmium bioavailability. Commun. Soil Sci. Plant Anal., 22:2029- 2035.


82


Lee, D., and Zeng, H. 1993. Chelating resin membrane method for estimation of soi! cadmium phytoavailability. Commun. Soil Sci. Plant Anal. 24:685- 700.

Liang, J., and J. J. Schoenau. 1994. Determination of heavy metalcontamination in the soil environment using ion exchange membranes. Proceedings of Soils and Crops Workshop,Univ. of Saskatchewan, Saskatoon, Sask., Canada. p268-273.

Mayiand, H. F., and C. W. Robbins. 1994. Sulfate uptake by salinity-tolerant plant species. Commun. Soil Sci. Plant Anal., 25 (13&14):2523-2541.

McLaughlin, M. J., P. A. Lancaster, P. W. G. Salin, N. C. Uren, and K. I.Peverill. 1993. Use of cation/anion exchange membranes for multielement testing of acidic soils. Plant and Soil 155/156:223-226.

Miller, R. L., M. Giardino, B. A. McKee, J. P. Cruise, G. Booth, R. Rovansek, D. Muirhead, W. Cibula, K. Holladay, R. E. Pelletier, W . Hudnall, C. Bergeron,J. loup, G. loup, and G. Love. 1995. Processes and fate of sediments and carbon in Barataria Bay, LA. Proceedings from the Third Thematic Conference on Remote Sensing for Marine and Coastal Environments. Seatle, Washington. Vol. l:233-244pp.

Parfitt, R. L., K. R. Tate, and R. 8 . McKercher. 1994. Measurement of phosphorus mineralization using an anion exchange membrane. Commun. Soil Sci. Plant Anal., 25:3209-3219.

Qian, P., J. J. Schoenau, L. E. Cowell, and L. Dennis. 1 9 9 4 . Assessing nutrient availability variations in landscapes. Proceedings of Soils and Crops Workshop, 1 9 9 4 , Univ. of Saskatchewan, Saskatoon, Sask., Canada. P 2 7 4 -2 7 9 .

Robinson, J. S., and J. K. Syers. 1990. A critical evaluation of the factorsinfluencing the dissolution of Gafsa phosphate rock. J. Soil Sci., 41: 597- 605.

SAS Institute Inc. 1989. SAS/STAT User's Guide, Version 6 , Fourth Edition, Volume 2, Cary, NC: SAS Institute Inc., 846 pp.

Schneider, J. W ., and K. Schneider. 1990. Indirect method for thedetermination of pyrite in clays and shales after selective extraction with acid solutions. Ceramic Bulletin 69(1):107-109.

83


Schoenau J. J., and W . Z. Huang. 1991. Anion-exchange membrane, water, and sodium bicarbonate extractions as soi! tests for phosphorus. Commun. Soi! Sci. Plant Anal., 22:465-492.

Schoenau, J., P. Qian, W. Z. Huang. 1993. Assessing sulphur availability in soil using ion exchange membranes. Sulphur in Agriculture 17:13-17.

Searle, P. L. 1988. The determination of phosphate-extractable sulphate in soil with an anion-exchange membrane. Commun. Soil Sci. Plant Anal. 19:1477-1493.

Searle, P. L. 1992. The extraction of sulphate and mineralisable sulphur from soil with an anion exchange membrane. Commun. Soil Sci. Plant Anal. 23:2087-2095.

Searle. P. L. 1991. A simple manual method for the determination ofphosphate-extractable (ion-exchange membrane) sulphate in soils. Commun. Soil Sci. Plant Anal., 22:1347-1354.

Sibbesen, E. 1977. A simple ion-exchange resin procedure for extracting plant- available elements from soil. Plant Soil 46:665-669.

Skogley, E. O., S. J. Georgitis, J. E. Yang, and B. E. Schaff. 1990. Thephytoavailability soil test - PST. Commun. Soil Sci. Plant Anal., 21:1229- 1243.

Soil Conservation Service (SCS). 1989. East Central Barataria cooperative river basin study, Jefferson, Orleans, Plaquemines and St. Charles Parishes, Louisiana. Rev. May 1989. United States. Soil Conservation Service. National Cartographic Center (U.S.). Ft. Worth, TX : USDA-SCS- National Cartographic Center ; [Alexandria, La. State Conservationist, distributor].

Somasiri, L. L. W., and A. C. Edwards. 1992. An ion exchange resin methodfor nutrient extraction of agricultural advisory soil samples. Commun. Soil Sci. Plant Anal., 23:645-657.

Szmigielska, A. M., and J. Schoenau. 1 9 9 4 . Determination of 2 ,4 -D amine in soils using anion exchange membranes. Proceedings of Soils and Crops Workshop, 1 9 9 4 , Univ. of Saskatchewan, Saskatoon, Sask., Canada. P 2 3 4 -2 4 3 .

8 4


Tejowulan, R. S., J. J. Schoenau, and J. R. Bettany. 1994. Use of exchange resins in soil and plant testing for micronutrient availability. Proceedings of Soils and Crops Workshop, 1994, Univ. of Saskatchewan, Saskatoon, Sask., Canada. p255-267.

Vaidyanathan, L. V., and O. Talibudeen. 1970. Rate processes in thedesorption of phosphate from soils by ion-exchange resins. J. Soil Soi., 21:173-183.

van Raij, B., J. A. Quaggio, and N. M. da Silva. 1986. Extraction ofphosphorus, potassium, calcium, and magnesium from soils by an ion- exchange resin procedure. Commun. Soil Sci. Plant Anal., 17:547-566.

Yang, J. E., E. O. Skogley, and B. E. Schaff. 1990. Microwave radiation and incubation effects on resin-extractable nutrients: 1. Nitrate, Ammonium, and Sulfur. Soil Sci. Am. J. 54:1639-1645.

85


CHAPTER 4: INFLUENCE OF SALINITY AND LANDSCAPE POSITION ON ACCUMULATION OF PYRITE AND NON-PYRITIC Fe AND S WITHIN

COASTAL ORGANIC SOILS OF BARATARIA BAY BASIN, LOUISIANA


Coastal marsh favors pyrite formation because of the highly reduced

environment and the availability of 8 , Fe, and organic 0 within the marsh.

Pyrite accumulation represents the balance between pyrite formation and

oxidation. Pyrite accumulation can be variable due to variations in redox

potential, and availability of reduced 8 , and Fe, and labile 0 within the soil.

Information on pyrite content and non-pyritic Fe and 8 indicates the

biogeochemical status under which a particular soil horizon was formed. This

study examines the influence of landscape position on pyrite accumulation for

different marsh types. 8 alinity, pyrite and non-pyritic Fe and 8 were measured

for soil profiles at a streamside (8 8 ) and inland (IL) landscape position, within

the saline (8 ), brackish (B), intermediate (I) and fresh (F) marshes.

The organic layer was comprised of a variable number of subhorizons

and their thickness varied widely within a degrading marsh. Accretion

differences between 8 8 and IL are the main source of spatial variability, which

forms more subhorizons at 8 8 than at IL. Deposltional differences that

occurred during formation of the mineral horizon may have left depressions

within the marsh. The organic layer is thicker and the mineral layer is found

deeper within these depressions. 8 oil bulk density (BD) was lower in the

brackish marsh compared to the saline marsh. 8 alinity profiles differed widely

between landscape positions within the inland landlocked site. Influence of

86


salinity on pyrite accumulation was variable within the profile. Pyrite content

within the surface horizon was spatially variable. Salinity effect on pyrite

accumulation was dominant for subsurface horizons. More pyrite accumulated

within the profiles of the saline marsh. A thin layer of dark organic material that

occurred immediately above the mineral layer contained the highest pyrite

content within the soil profiles. Non-pyritic Fe contents were significantly

different between the landscape positions within the more saline marshes. The

non-pyritic Fe content could be used as an indicator to assess marsh status.

The landlocked inland sites retained more non-pyritic Fe within sub surface

organic layers due to limited pyritization. Non-pyritic S content within the

organic subhorizons differs significantly between marshes due to the changes

in salinity. The saline marsh contains more non-pyritic S. Soil pH differences

are more pronounced between marshes. The pH for the organic subhorizons

does not vary within the profile. The mineral layer pH was always higher than

that of the organic subhorizons.

Pyrite framboids detected only within the saline marsh organic

sediments, were 6 to 8 pm in diameter. The sediments at both saline and

brackish marshes exhibited a mixed mineralogy with smectite, illite, and

kaolinite clays. Pyrite also was detected from XRD patterns for the sediments

at both marshes, while the coarser fraction ( < 2 0 pm) contains more pyrite.

These organic soils should be reclassified to indicate sulfidic materials within

87


the profile. The pyrite determination technique used in this study can be used

to characterize sulfidic material within the soils.

Coastal marshes provide a favorable condition for pyrite formation

because of their characteristically highly chemically reduced environment.

Sulfate is reduced to sulfide, and with the availability of ferrous ions, pyrite is

formed. Acid sulfate soil conditions can be formed if the soils with pyrite are

drained (Fanning and Fanning, 1989; Dent, 1993), which would be a highly

unlikely event for these lands. However, dredging occurs in this region

associated with the oil and gas industry. This activity can expose the pyritic

material that could be oxidized and eventually produce acidity. Therefore,

information on the amount of pyrite within the soil profile is important when

making land management decisions because of the potential acidity from the

dredged material. Acidity in combination with salinity (Fanning, 1993) can

adversely affect plant productivity and eventually may contribute to marsh loss.

The adverse effects of these soils could extend into the adjacent aquatic

ecosystems and limit the alternate use of these lands. Outbreaks offish

disease and massive fish kills in the water bodies were reported when they

were polluted by water draining from the acid sulfate soils (Lin and Melville,

1994).

Pyrite is the most stable and prevalent sulfide phase occurring within

coastal marsh soils (Berner, 1964; Lord and Church, 1983; Rabenhorst and

James, 1992). Pyrite remains inactive if it is kept submerged and reduced

within the marsh. Pyrite formation can be spatially and temporally variable

88


within the marsh due to variations of sulfate, reactive ferrous, organic matter

content and microbial activity. Sulfate supply can be variable, but closely

related to salinity within the marsh because seawater carries sulfate into the

ecosystem. Marsh types are delineated based on salinity levels (Feijtel et al.,

1988; Soil Conservation Service, 1989). Seawater intrusion, when extended

into the marsh due to sea level rise, enhances the sulfate supply into the

marsh. Sulfur dynamics also are important for the energy flow within this

coastal environment (Whitcomb et al., 1989).

This study evaluated the influence of salinity on pyrite accumulation and

to compare the accumulation of pyrite and non-pyritic Fe and S between two

landscape positions; namely streamside and inland areas within the marsh.

Mineralogical evidence also was sought to detect pyrite within these soils.


4.2.1 Marsh Loss around Baraterie Bay Basin, Louisiana

Marsh loss has occurs in the northern Gulf of Mexico at a high rate

(Turner, 1990). Marshes undergo rapid submergence that is associated with

the relatively high rate of apparent sea level rise. Conversion of brackish

marsh to saline marsh due to seawater intrusion is common in southeast

Louisiana. Marsh longevity depends partly on the deposition of mineral and

organic matter. Mineral matter accumulation is critical and directly proportional

to biomass production. Salinity influences mineral and organic matter

accumulation (Nyman and DeLaune, 1991; Chmura et al., 1992; DeLaune et

al., 1993; DeLaune et al., 1994; Nyman et al., 1994a). The sulfate content

89


increases with increased salinity, and marshes become more liable to sulfide

toxicity (Nyman et al., 1993b).

Two major mechanisms shown to cause marsh loss are peat collapse

and erosion below the living root zone (Nyman et al., 1994a). Peat collapse

occurs following a “hotspot” pattern (DeLaune et al., 1994); a hotspot is an area

within the marsh that experiences rapid loss (Nyman et al., 1993c). Peat

collapses because of the disintegration of the living root network resulting from

root mortality. This is associated with salinity increase and water logging.

Reduced surface elevation of peat forms ponds within the marsh. Hotspots are

common within brackish areas that are converting to saline marsh. Less

frequent occurrence of sedimentation events by winter storms results in

inadequate vertical accretion, and soil bulk density is too low to support

Spartina a ltem iflora. The most common marsh loss pattern, the scattered

broken marsh interior, is less severe than the less common concentrated

hotspot pattern. Marsh loss mechanisms are spatially variable within a

relatively small region (Nyman et al., 1993a, 1994b).

4.2.2 Seawater Intrusion, Salinity and Marsh Types

The isohaline data for the east and central Barataria Bay ranges from

over 20 ppt to less than 0.5 ppt (Soil Conservation Service, 1989). The

marshes around the Barataria Bay have been delineated into saline, brackish,

intermediate, and fresh water marshes based on salinity. Fresh and

intermediate marshes occur below 5 ppt. Brackish marshes occur between 5

and 10 ppt while salt marsh is found at salinity above 10 ppt (Feijtel et al., 1988;

90


Soil Conservation Service, 1989). The species diversity increases with

decreasing salt concentration (Darmody and Foss, 1979). Typha spp. (cattail),

Spartina cynosuro ides (big cord grass), and Pontederia cordata (pickerel weed)

occupies the low salinity areas of the marsh while Scirpus o iney i (three square),

Spartina pa tens and Spartina a ltem iflora are found within high salinity areas

(Darmody and Foss, 1979). Increasing salinity decreases growth of Spartina

a ltem iflo ra and decreases transpiration and plant biomass. Ion exclusion,

secretion, and accumulation are three major salt tolerant mechanisms.

Inhibition of K uptake at high Na concentrations adversely affects plant growth.

Salt glands are an additional mechanism that enabled plants to secrete excess

internal salts. S. a ltem iflora selectively secretes Na relative to K (Bradley and

Morris, 1991). Spartina a ltem iflora grows optimally in the salinity range of 0-20

ppt, but salinity above 40 ppt adversely affects survival (Bradley and Morris,

1991). Increasing salinity above 10 ppt reduces growth of brackish marsh

species, Spartina o ineyi and Spartina patens, while Spartina o iney i is

drastically affected (Broome et al., 1995). The mechanisms adopted are

species dependent. Spartina pa tens tolerates high Na concentrations by ion

secretion while ion accumulation is the dominant salt-tolerance mechanism for

Spartina o iney i (Broome et al., 1995). Presence of toxic levels of elements and

gases is another stress for plants growing within these soils. Resistant plants

contain high concentrations of elements within their roots compared to the

shoots (Rozema et al., 1985). Spartina a ltem iflora tolerates sulfide uptake over

a long growing season through an internal oxidation or detoxification

91


mechanism (Carlson and Forrest, 1982). Sulfide concentration gradient

associated with slight elevation changes results in plant productivity and

mortality variations (DeLaune et al., 1983).

4.2.3 Pyrite Formation and Accumulation

Increased oxygen consumption in submerged organic soil and restricted

oxygen supply due to flooding creates a low oxidation-reduction potential within

the soil. With the disappearance of oxygen in the system, alternate electron

acceptors become active. Sulfate reduction becomes prominent after depletion

of nitrate, manganic and ferric compounds in the system (Ponnamperuma,

1972; Patrick and Jugsujinda, 1992; Rabenhorst and James, 1992). Labile

organic carbon sources are the major regulator of the soil reduction process

(Connell and Patrick, 1969; Rabenhorst and James, 1992). Ferric is reduced to

ferrous, which is readily mobile within the system as it diffuses along the

concentration gradient. Sulfate also diffuses into the reduced zone along the

concentration gradient supported by thermal mixing. This sulfate reduction

process is called sulfidization (Eq. 4.1 to 4.3), and pyrite is a byproduct of

sulfidization (Fanning and Fanning, 1989; Rabenhorst and James, 1992):

2 CH3-CHOH-CHOH + SO4- ===> 2 CH3COOH + HCO3 +HS + H3O [4.1]

F e " +H S +OH ==> FeS + HjO [4.2]

Fez03 + 4 SO4- + 8 CH2O + 1/2 Oz==> 2FeSz + 8HC03' + 4 H2O [4.3]

Pyrite forms by various processes (Rickard, 1970):

92


1.0 Homogeneous ionic reactions

Fe% „) + S"-(3q, ===> FeSz;,) [4.4]

2.0 Solid-Fe reactions

2.1 Mackinawite-polysulfide exchange

Fe%q) + S fp * ===> FeS,„ [4.5]

FeS(c) +S'-(3q) ===> FeS;;,) [4.6]

2.2 Goethite sulfidation

2FeOOH(c, + 3 H 2S ,3q) ===> 2FeS(„ + S“ + OH ^q, + SH^O [4.7]

3. 0 Homogeneous solid-solid reactions

An irreversible conversion of marcasite to pyrite.

4. 0 Heterogeneous solid-vapor reactions

Sulfur oxidation of pyrrohtite can result in the formation of pyrite.

Coastal marshes favor soil pyrite formation because tidal water provides

sulfate to the highly reduced marshes where soluble ferrous is apparently

abundant. Organic matter accumulation, existence of highly reduced

conditions, availability of excess sulfate, 8 reducing bacteria, ferrous ion

concentration, and fluctuation of redox potential (partial reoxidation of S")

promote sulfate reduction and pyrite accumulation (Pons et al., 1982). The rate

of pyrite formation varies depending on the availability of ferrous ions and 8,

and anaerobic conditions. Pyrite may not be formed despite the abundance of

sources if the organic carbon source is limited. Removal of HCOq" from the

system promotes pyrite formation (Fanning and Fanning, 1989). 8 u If ate

reduction produces carbonates that can be precipitated as CaCOg (Ogata and

93


Bower, 1965). Dissolved S concentration of less than 320 ppm limits pyrite

formation (Goldhaberand Kaplan, 1982). Depending on the pH, sulfate

reduction could occur within the Eh range o f +110 and -500 m V (Baas Becking

and Moore, 1961).

Marsh sediments contain both Fe sulfide (FeSg-Fe) and Fe oxide (free

Fe) phases, depending on the redox potential (Griffin et al., 1989).

Noncrystalline Fe oxides rather than crystalline Fe oxides become reduced by

bacterial activity (Munch and Ottow, 1980). Munch and Ottow (1980)

suggested that the Eh drop and change in pH should be regarded as a

consequence, rather than the cause of Fe reduction by bacteria. The degree to

which the reactive Fe has been used for pyrite formation is termed as degree of

pyritization (Py).

Py = FeSg-Fe / (FeSg-Fe + free Fe) [4.8]

Ferrous-Fe precipitates as siderite (FeCOg) unless pyrite is formed (Rickards,

1973). The reactive ferric minerals in the solid phase are important in the

formation of the reactive ferrous portion. Crystalline Ferric minerals are present

at depth within the salt marsh at Lewes, Delaware (Kostka and Luther, 1994).

Mineral precipitation varies within the marsh due to variable ionic concentration

of porewater (Giblin and Howarth, 1984; Feijtel et al., 1988). This produces

micro site variability within the marsh soils (Rabenhorst and James, 1992).

Possible mineral formation can be evaluated from ion activity products of the

porewaters (Giblin and Howarth, 1984). All porewaters are supersaturated with

respect to pyrite (FeSg) and hauerite (MnSg) for the Barataria Bay marshes.

94


Ironsulfide equilibrium indicates super saturation at surface (3 to 5 cm) but

undersaturation at 2 0 cm within brackish and salt marsh sediments of the

Barataria Bay (Feijtel et al., 1988). Rickard (1975) also shows that pyrite

rapidly forms as follows:

Fe^" + S jS"-+ HS- ====> FeSg + S,S^' + H* [4.9]

4.2.4 Sulfur Dynamics within the Marsh

The major processes involved in 8 dynamics are biological assimilation,

mineralization and biogeochemical transformations. The latter is mainly a

redox-controlled process, which includes both reduction and oxidation

reactions. Sulfur occurs in various oxidation states and redox controls solubility

and mineral formation (Lindsay, 1979). Chemical equilibrium in interstitial

water can be indicative of authigenic mineral formation (Giblin and Howarth,

1984; Feijtel et al., 1988; Gaillard et al., 1989). The inorganic form of S

changes with Eh as sulfides at -2 0 0 mV, elemental S at 0 mV, and sulfate at

+600 mV of Eh. There are many other forms of inorganic 8 that occur during

transformations. Sulfate is added continuously by seawater into the coastal

marsh ecosystem. Pore water 8 is at equilibrium with soil matrix 8 . The

gaseous forms of 8 produced by these processes are HgS, CH 3SH, (C H 3)2S,

COS, and CSg (Crozier et al., 1995). Organic and inorganic 8 comprise the

soil 8 pool within the marsh. The organic 8 portion is much larger compared to

the inorganic 8 fraction. The two organic fractions are the 8 directly bonded to

carbon (carbon-bonded 8 ) and ester sulfates that are not directly bonded to

carbon (Freney et al., 1970). The carbon-bonded 8 includes the 8 of cysteine,

95


cystine and methionine. The inorganic S fraction includes four S species: HCI

soluble S, acid volatile S (AVS), elemental S, and pyrite S. The HCI soluble S

represents the soil sulfates and sulfides produced during sulfate reduction.

Sulfide is mobile and can be evolved as H jS and reoxidized forms, or can be

precipitated as Fe monosulfides and pyrite (Whitcomb et al., 1989). The AVS

fraction includes H^S and Fe monosulfide (not pyrite), elemental or organic-S

(Howarth and Merkel, 1984). Sulfide reoxidation may occur at the root surface

or the soil surface. This oxidation reaction can either be biological (Grant and

Bathmann, 1987) or chemical (Whitcomb et al., 1989); both produce elemental

S. Dissolved elemental S can be combined with sulfides to form the polysulfide

ions that are unstable and will transform into elemental S with further oxidation.

If reducing conditions prevail, polysulfide ions will be reduced, or else Fe** or

FeSz will form. The internal S oxidation-reduction cycles for this environment

prevents the release of HgS into the atmosphere (Whitcomb et al., 1989).

However, Crozier et al. (1995) indicates that production of S gas is positively

correlated with the initial sulfate concentration. Soils retain HgS, and the

sorption capacity is influenced by the quantity of Fe oxides present within the

soil (Bohn et al., 1989). Variation of these 8 species can indicate the S

dynamics for the marsh (Krairapanond et al., 1991a, 1991b, and 1992). The

lack of information for gaseous 8 emission and plant uptake can lead to an

erroneous conclusion as to what processes are responsible for the 8 dynamics.

Organic S is considerably higher than the inorganic 8 pool for all three

marshes within the Barataria Bay marshes (0-50 cm depth) (Krairapanond et

96


al., 1991a, 1991b, and 1992). Carbon bonded S is the dominant portion of the

organic S pool. Organic carbon can be highly correlated with total S because

of sulfate retention by organic matter (Haering et al., 1989). The AVS accounts

for <1% of the total 8 while HCI soluble 8 accounts for most of the inorganic 8

pool for all three marshes. Pore water sulfate is significantly correlated with

HCI soluble 8 . Inorganic 8 fraction is 22% for fresh, 13% for brackish, and 24%

for salt marsh. Reduced 8 species, except elemental 8 , increases seaward

along the salinity gradient. Pyrite 8 and elemental 8 account for 8-33% of the

inorganic 8 pool within the brackish marsh. No clear pattern of pyrite 8

formation was observed for the upper 2 0 cm of the soil profile within the fresh

water marsh. At this depth, peak elemental 8 concentration coincides with

peak plant root density.

Pyrite 8 and elemental 8 account for 6 to 21% of the inorganic 8 fraction

within the fresh water marsh (Krairapanond et al., 1991a). Pyrite 8

concentration represents <3% and <2% of total 8 for the fresh and the saline

marshes, respectively (Krairapanond et al., 1991a, 1991b). Pyrite is the major

8 form for most of the streamside organic horizons, while organic 8 was

dominant within inland pedons for Chesapeke Bay marshes (Haering et al.,

1989). Thiosulfate (8 2 0 3 ) and sulfite (8 0 ^ ^ are in equilibrium with elemental

8 (8 °) during H2 8 oxidation in porewaters of Danish salt marsh profiles

(Thamdrup et al., 1994). Redox conditions fluctuate vertically within marsh

soils, producing seasonal 8 species variation (Feijtel, et al., 1988). Therefore,

97


the responsible processes can be assumed based on the knowledge of S

dynamics. Inverse relationship between ester-sulfate S and HCI soluble S

could indicate S mineralization while AVS oxidation can increase HCI soluble S.

High elemental S concentrations can indicate a previous oxidation history of

reduced layers, most importantly the transition of oxidized and reduced layers

within the profile (Krairapanond et al., 1991a, 1991b, 1992). Table 4.1 presents

absolute S estimates for different marsh types from Krairapanond e ta l. (1991a,

1991b, 1992). Sulfur species vary widely between marsh types. Organic S

content is similar between brackish and salt marshes, but is the lowest for the

fresh marsh. High organic S indicates high S assimilation by the marsh

vegetation. Inorganic S species show much higher variation compared to

organic pool for different marsh types. Total soil S content varies considerably

among marsh types. Fresh marsh contains 5.7 ppt S brackish marsh 15.6 and

salt marsh 15.7 ppt S (Krairapanond et al., 1992).

Table 4.1. Sulfur species (ppm) variation in 0-50 cm depth for different marshes, in the Barataria Bay, Louisiana.

Marsh AVS HCI-S S° Pyr-S Ester-S C-S Total-S

Fresh 1 1 1083 91 1 0 0 1180 3249 5700

Brackish 30 1544 84 314 3619 10000 15600

Salt 24 3250 272 245 3313 8588 15700

98


4.2.5 Pore Water Dynamics and Mineral Formation

Mineral precipitation can be variable within the marsh due to variable

ionic concentrations in porewater (Giblin and Howarth, 1984; Feijtel et al.,

1988), and mineral formation can be evaluated from ion activity products of the

porewaters (Giblin and Howarth, 1984). All pore waters are supersaturated

with respect to pyrite (FeSg) and hauerite (MnSg) for the Barataria Bay marshes.

Ironsulfide equilibrium indicates super saturation within the surface (3 to 5 cm)

but undersaturation at 2 0 cm within brackish and salt marsh sediments of the

Barataria Bay (Feijtel et al., 1988). The study by Giblin and Howarth (1984)

includes examination of the formation of FeS (amorphous), FeS (mackinawite),

Fe^S^ (greigite), FeSg (pyrite), MnS (alabandite), FeCO^ (siderite), M nCOj

(rhodocrosite), Fe3(P 0 4 ) 2 (vivianite), and Mn3(P O j 2 (reddingite) based upon

chemical equilibrium. Giblin and Howarth (1984) indicate that pore waters are

undersaturated with respect to all Fe monosulfide minerals, Mn3 (P 0 4 ) 2

(reddingite), and siderite (FeCO j) for the Great Sippewissett marsh,

Massachusetts. But the porewaters are supersaturated with respect to pyrite

(FeS 2), and pyrite is formed as a result of oxidation of polysulfides and

elemental S, rather than the conversion of FeS to FeS 2 (Giblin and Howarth,

1984). Rickard (1975) also shows that pyrite rapidly forms as follows:

Fe"" + S5S"- + HS ====> FeS2 + 8 4 8 "-+ H" [4.10]

4.2.6 Pyrite Crystal Formation

The rapidly formed pyrite occurs as small single crystals (Rabenhorst

and James, 1992). Feijtel et al. (1988) observed single euhedral pyrite

99


particles on external root walls in salt marsh samples. Kosters and Bailey

(1983) reported the occurrence of framboidal pyrite within the heavy fraction of

the marsh soil from Barataria Bay. Framboidal pyrite results from the reaction

of Fe monosulfides and elemental S (Raybould, 1973). Framboids are circular

or slightly oval in cross section, and can be assumed to be spherical or sub-

spherical. Framboids can vary in size (Raybould, 1973; Rickard, 1970). Iron

framboids that do not contain S (i.e., HFeO j) can coexist with pyrite (Giblin and

Howarth, 1984). Raiswell et al. (1993) proposed a 3-D diffusion-with-

precipitation model to predict pyrite precipitation when a spherical mass of

organic matter, producing HgS by sulfate reduction, is surrounded by

porewaters containing dissolved Fe. This model indicates the need of high

dissolved Fe concentrations to react with higher amounts of reduced S from

large organic masses to form pyrite at decay sites. The maximum size of

sphere of organic matter that can be pyritized is about 50 pm in radius that is

close to the radius of framboidal pyrite. However, Raiswell et al. (1993)

indicated that the exposure times are important in this pyritization process.

4.2.7 Spatial Variability of Pedogenic Processes

The salinity gradient along the marsh landscape maintains the S

gradient into the marsh ecosystem of Barataria Bay Basin. Pyrite formation can

be variable in relation to the marsh landscape position (Pons, 1973; Griffin and

Rabenhorst, 1989; Haering et al., 1989; Lin and Melville, 1994). Variation in

available sulfate and ferrous compounds, and oxidation-reduction status within

100


the marsh can vary pyritization (Griffin and Rabenhorst, 1989; Haering et al.,

1989; Lin and Melville, 1994). The pyrite content is not correlated to total S or

carbon content (Haering et al., 1989) but pyrite accumulation is related to Fe

and S supply (Griffin and Rabenhorst, 1989) within coastal marshes in

Maryland. Pyrite accumulation is favored for streamside profiles that received

Fe inputs from tidal flushing (Haering et al., 1989). The sediments provide

ferrous ions. Pyrite formation can be simulated from models that are based on

availability of sulfate. Eh range, ferrous ion supply, and the soil organic matter.

Pyrite content varies across marshes for Barataria Bay, Louisiana; pyrite is

highest for the brackish marsh and lowest for fresh marsh (Brackish > Salt >

Freshwater marshes) (Feijtel et al., 1988). The change of pyrite content is due

to the seasonal balance between pyrite formation and oxidation. Ion activity

products of the porewaters indicate spatially variable mineral precipitation for

Massachusetts’s marshes (Giblin and Howarth, 1984). Pyrite forms at various

depths of the profile and will remain until it oxidizes. Three distinct zones may

be separated within the profile with respect to pyrite accumulation: a frequently

oxidizable surface layer, a conducive second layer, and a third layer with limited

pyritization (Lord and Church, 1983). Within the transition zone between

oxidized and reduced conditions, more elemental S is found than pyrite S.

However, prolonged droughts can lower the water level by which surface soils

can be naturally drained to a considerable depth. A water table drop during the

day as well as night can be due to water uptake by roots. Oxidation reactions

will occur within the drained portion of the profile as air enters into the surface

101


soil (Dacey and Howes, 1984). This promotes pyrite oxidation instead of pyrite

formation. The redox profile fluctuates seasonally within Barataria Bay

marshes (Feijtel et al., 1988). Pyrite formation occurs within the salt marsh

during the summer and fall seasons when Eh values are low. Pyrite can be

completely oxidized during the early spring, which may create acid conditions.

In a brackish marsh, pH and Eh changes govern Fe^^ concentrations, which is

associated with pyrite formation and oxidation. The minimum pH coincides with

increased interstitial Fe" " during the spring (Feijtel et al., 1988). Ion activity

products of the porewaters indicate seasonal variation of mineral precipitation

for Massachusetts’s marshes (Giblin and Howarth, 1984).

4.2.8 Soil Classification and Soil Taxonomy

Organic soiis within the Barataria Bay Basin are mapped as associations

of soil series in the soil survey conducted in 1981 morphology (Soil

Conservation Service, 1984). The associated series within a particular salinity

regime are delineated based on the depth to mineral layer at subgroup level

(i.e., Terric Medisaprists and Typic Medisaprists). Many activities such as

hurricanes, saltwater intrusion, construction of channels, and other man-made

changes, may affect accretion and degradation of organic layers within the

landscape. Reaction classes are considered at family level differentiae of

Histosols in Soil Taxonomy (Soil Survey Staff, 1996). Histosols that have a pH

value of 4 .5 or more in one or more layers of organic soil materials within the

control section for Histosols are differentiated as “Euic” while all other Histosols

102


are called “Dysic”. The two major soil series associations that are dominant

within the brackish and saline marshes differ in salinity but have similar

morphology (Soil Conservation Service, 1984). Timbalier-Bellpass association

(TB) is dominant within the saline marsh. According to Soil Taxonomy, the

Timbalier series is euic, thermic, Typic Medisaprists while the Bellpass series is

clayey, montmorillinitic, euic, thermic, Terric Medisaprists. The Lafitte-Clovelly

association (LA) dominates the brackish marsh. The Lafitte soil series is an

euic, thermic, Typic Medisaprists while the Clovelly soil series is a clayey,

montmorillinitic, euic, thermic, Terric Medisaprists. Terric Medisaprists are the

Medisaprists that have a mineral layer 30 cm or more thick that has its upper

boundary within the control section (130 cm) below the surface tier (60 cm).

Other Medisaprists that have the upper boundary of the mineral layer below

control section (130 cm) are Typic Medisaprists. The main difference between

these soil series is the depth to mineral layer. They are delineated based upon

depth to mineral layer.

Previous soil classification has not considered the presence of S

compounds within these soil profiles. Two major diagnostic horizons that are

directly related to S are the sulfuric horizon and the sulfidic materials. A great

group of Histosols, "Hemists", is differentiated into two subgroups as

Sulfohemists and Sulfihemists. The Sulfohemists has a sulfuric horizon that

occurs within 50 cm of soil surface, while the Sulfihemists contain sulfidic

materials within 100 cm of the soil surface. The sulfuric horizon is a soil horizon

that is 15 cm or more thick and composed of either mineral or organic soil

103


material that has a pH value of 3.5 or less (1:1 soil:water suspension) and

shows evidence that the low pH is caused by sulfuric acid. The evidence can

be jarosite concentrations, directly underlying sufidic material, or 0 .05 % or

more water-soluble sulfate. Sulfidic materials contain oxidizable S compounds

and are mineral or organic soil materials with a pH of more than 3.5 which, if

incubated a layer of 1 -cm thick at field capacity and room temperature, show a

drop in pH of 0.5 or more units to a pH value of 4.0 or less (1:1 soi!:water

suspension) within eight weeks (Soil Survey Staff, 1996). Sulfihemists and

Medisaprists are reported for the organic soils in Florida, where the Sulfihemists

contains more than 0.75% S in the upper 50 cm (Coultas, 1997).

4.2.9 Pyrite Determination

A suitable method is required for pyrite determination to assess pyrite

accumulations. A few different techniques are available for pyrite

determination. The qualitative determination methods include X-ray diffraction

and Scanning Electron Microscopy. The X-ray diffraction technique can also be

used as a semi-quantitative method to estimate pyrite. The quantitative

determination methods are indirect and either Fe based or S based. Empirical

relations also are proposed (Willett and Beech (1987). Fe based sequential

extraction and differential measurement techniques are used to estimate pyrite.

Sulfur based techniques fractionate S mainly into organic and inorganic S. The

inorganic fraction is separated into four groups as HCI-soluble S (mainly sulfate

and H2S), acid volatile S (AVS) (monosulfides and dissolved sulfides),

elemental S and pyritic S (Begheijn et al., 1978; Krairapanond et al., 1991a,

104


1991b, and 1992;). The Fe based estimation involves separation of pyritic and

non-pyritic Fe. This is directly related to the estimation of the rate of pyritization

(Leventhal and Taylor, 1990). More detailed Fe based schemes are used in

acid sulfate soil analysis, where pyrite, as well as jarosite fractions, need to be

estimated separately (Lord, 1982; Raiswell e ta l., 1994). Separation of pyritic

and non-pyritic Fe is sufficient for the scope of this project. The environmental

significance of pyrite accumulation in these ecosystems is limited to the impacts

of dredging. Hence, any pyrite that is accumulated in the soil profile is kept

submerged until the soil is dredged, exposed and allowed to oxidize. There is

no jarosite within the submerged potential in acidsulfate soils. Therefore,

considering the need to estimate pyrite material, a quick and easy method for

separating pyritic and non- pyritic Fe is sufficient. The spatial variability study

assesses the relative variation of pyrite estimates, so the differences are

cancelled. The S-based techniques may contaminate the samples because of

the soluble S from the high salinity of the samples; most of the sea-borne

salinity contains S in the system. Lord (1982) suggested a selective extraction

method for pyrite determination. The HF-H3BO3 solution used for silicate

dissolution dissolved pyrite and required correction factor to adjust for pyrite

content. The indirect estimate of pyrite is based upon the separation of pyritic

and non-pyritic Fe, which is adequate for this study (Schneider and Schneider,

1990). Therefore, their method used for pyrite estimation in coal and ceramics

is appropriate for this experiment.

105



4.3.1 Site Description and Sampling Procedure

Two landscape positions, namely streamside (8 8 ) and inland (IL) were

sampled from each site. Four sites were selected to represent each marsh

type, saline (8 ), brackish (B), intermediate (I) and fresh (F) within the Barataria

Bay Basin (Figure 4.1). 8 ite names will be abbreviated as “8 8 8 ” for streamside

(8 8 ) at the saline marsh (8 ) and “8 IL” for inland at the saline marsh (8 ) and so

forth. Three soil profiles were sampled from each landscape position.

8 treamside profiles were selected 1 to 3 m from the water edge of the marsh.

Inland profiles were selected about 30 m from the streamside. A MacAully

auger (Day et al., 1979; 8 heppard et al., 1993) was used to extract soil cores

from the surface to below the first mineral soil layer or to a depth of 1.50 m.

8 oil horizons were designated according to 8 oil Taxonomy (8 oil 8 urvey 8 taff,

1996), and soil samples were collected from each subhorizon down to the

mineral soil layer. 8 oil samples were stored in an ice-filled cooler in the field

and in a cold room until they were analyzed.

8 oils were sampled separately for bulk density from the different

horizons using centrifuge tubes. Capped centrifuge tubes of 60 mL were

prepared by puncturing the bottom tip and brought to the field. The tubes were

filled completely with soil with minimum possible disturbance. The hole in the

bottom helped avoid air trapping. The soil-filled tube was capped and wrapped

with a polythene bag and brought to the lab in an ice-filled cooler.

106


CD■DOQ .C

gQ .

■DCD

C/)C/)

8■D

91*00' 00*30' 90*00* 80*30*

3.3 "CD

CD■DOQ .C

aO3"Oo

CDQ .

■DCD

C /)C /)

30*00* —

20*30* -

BARATARIA BASINL o u lô in r t a

Dartlarla Baaln

Vagotation

N o iu ro l l«va« i(U /bo n . AnnCulium l A rc o i

ond fio llom laod H ardw ood fo r e t l i

fkp y-: I f o r a i l td Walksnd(Swomp)

M a rih

[J .] In l*(m *d ia t* M orih

I " 1 fiiocliih M o /ik

I I Satin* M oftk

Figure 4.1 A map showing different marsh types within the Barataria Bay Basin (after Krairapanond et al., 1992).

Tubes with soil were weighed and placed in an oven at 55°C. After drying, dry

weight was used to measure the bulk density. Bulk density data were

estimated only for saline and brackish marsh sites.

4.3.2 Measurement of EC and TDS

W ater samples were collected from different water bodies within the

Barataria Bay Basin to cover a wide range of salinity. Samples were brought to

the laboratory in an ice-filled cooler and filtered (Whatman No. 42). These

samples were used to measure electrical conductivity (EC) and total dissolved

solids (TDS). The EC of the water samples was measured using a Corning M90

conductivity sensor (Corning Inc., Corning New York 14831, USA). A volume of

30 to 60 mL from each water sample was weighed and slowly evaporated in a

centrifuge tube at a temperature between 55 to 65 °C in an oven, cooled in a

desiccator and weighed. The TDS were estimated on a dry weight basis for

each sample in triplicate.

4.3.3 Pyrite and Non-pyritic Fe and 3 Determination

The method used for pyrite estimation in coal and ceramics by

Schneider and Schneider (1990) was modified for the analysis of extractions for

Fe by the ICP-AES. Six scoops (1/8 tsp.) of soil were measured into six

digestion tubes that were calibrated to 50 or 75-mL volume. Tubes were

separated into two batches; one was for non-pyritic (NP) and the other set was

for pyritic (P) and non-pyritic (NP) Fe digestion. The 50-mL tubes were used

for non-pyritic Fe, while the 75-mL tubes were used for pyritic and non-pyritic

108


Fe. A sample was placed at the bottom of the tube. Six mL of concentrated

HCI was added to each tube with soil, under a fume hood and the tube was

covered with a small glass (“reflex") funnel. The soil was stirred within the acid

to make a suspension and left over night, covered with "reflex" funnels. Then,

23 mL of deionized water (D l) was added into each tube. The digestion

procedure was different for “NP” and “NP+P” sets, therefore they were digested

separately. To digest the NP set, the sample was placed onto the digestion

block, and the block heater was turned on. It took about 15 to 20 min to start

boiling. The samples were kept boiling for 1 h and were removed and allowed

to cool. The tubes were filled to the 50 mL level with 1:10 HCI then mixed well

and filtered through #5 (or #42) Whatman filter paper. The filtrate was highly

acidic; therefore, it was diluted six times to reduce the acidity and to increase

the detection levels by the ICP.

For the “NP+P” set, after adding 23 mL of water, the sample was boiled

for 30 min and then 25 mL of 3N HNO3 was added to each tube while they

were on the digestion block. Acid was slowly added through the funnel. It took

another 10 to 15 min for the mixture start boiling. Samples were boiled for

another 30 min and cooled. The tubes were filled up to the 75-mL level with 2N

HNO3 then mix well and filtered through #5 (or #42) Whatman filter paper. The

filtrate was highly acidic; therefore, it was diluted six times before analysis by

the ICP. The pyritic Fe content was determined based on the difference

109


between Fe content of “NP+P” and “NP” digestions. The pyrite content was

estimated using the equation 4.11.

Fe (mg L'^) * 2.146 = FeS2 (mg L'^) [4.11]

Fe and S content from the “N P” digestion were the non-pyritic Fe and S

(Schneider and Schneider, 1990).

4.3.4 Measurement of soil pH

The pH was measured directly using a combination pH electrode with

pH/mV meter (Orion Research model 701A/digital lONALYZER). Soil pH was

measured in triplicate in soil water suspensions in the laboratory for each soil

horizon.

4.3.5 Microanalysis of Marsh Soils

Small soil samples were oven dried at 55 °C for a few days to remove

water from the samples and prepared for the SEM analysis. The samples were

gold coated and mounted on the SEM rod mounts for examination (Welton,

1984).

Surface sediments were collected from the streamside of saline (SSS)

and brackish (BSS) marshes. Smears were prepared for X-ray diffraction

analysis for two size fractions, less than 2 0 and less than 2 pm, with minimum

disturbance (treatment) to samples. The samples were dispersed with sodium

phosphate to preserve any Fe and carbonate minerals. The less than 2-pm

and 20-pm size fractions were separated by gravity sedimentation. Clays were

concentrated by centrifuging. XRD patterns were obtained for each sample

110


under air dried, ethylene glycol saturated, and heated (to 300 and 550 C)

treatments (Moore and Reynolds, 1989). The XRD patterns were obtained

from Cu Ka (40 KV, 30 mA) using a Siemens D5000 instrument. Differential

X-ray diffraction (DXRD) patterns were analyzed for Fe minerals, especially

pyrite (Schulze, 1981; Lord, 1982).

4.3.6 Data Analysis

Statistical analysis included mean comparisons, ANOVA, and regression

analysis using the Statistical Analysis System (SAS Institute Inc., 1994).

Random effect analysis of variance was conducted from a completely nested

design with two factors: site, which represented salinity effect, and the

landscape position, including streamside and inland positions. The random

effect model had the following form (SAS Institute Inc., 1994).

Yijr = H + tti + pij + 8 |jr

W here

ijr is the value of the dependent variable observed at the rth replication with

the first factor at its i th level and the second factor at its j th level.

p is the overall (fixed) mean of the sampling population.

tt| Pij 8 |jr are mutually uncorrelated random effects with zero means and

respective variances 0 2 - ^^3 - and (the variance components).

Each variable for every horizon was analyzed separately in a nested design

and mean comparisons were performed only for the pre-planned pairs of

111


means. Any treatment effect is expressed as significant relative to confidence

level (a ) of 0.05, unless the p-value is presented.


4.4.1 Soil Horizon Thickness and Bulk Density

Organic soil layers were comprised of distinct subhorizons designated as

G a l , Ga2, G a3 horizon, and etc. The number of subhorizons varied within the

landscape position, as well as, between the sites. The underlying mineral

horizon was designated as Cg due to gleying; in some cases, a thin mineral

layer was separated based on texture (i.e., C g l and Cg2). A thin layer of dark

organic material occurred immediately above the Cg horizon with a strong

sulfide smell.

Table 4 .2 presents the average thickness of every horizon at the two

landscape positions at each marsh type. Average thickness of the Ga1 horizon

ranged from 10 to 38 cm. Thickness of the G a l horizon was similar between

the landscape positions at all marsh types except at the intermediate marsh

that was degrading. The lowest G a l horizon thickness (10 cm) at ML and the

highest G a l horizon thickness (38.3 cm) at ISS may be indicative of the current

degrading status at this location. Disintegration of the root network may cause

high erosion with further dispersion due to Na that is brought in with seawater.

Gtherwise consolidation of collapsed material may occur within the G a l horizon

at ML. Eroded material from inland may have been deposited at streamside in

addition to the degradation at the streamside. Inland may not experience any

112


CD■DOQ .C

gQ .

■DCD

C/)Wo'303CD

8■D( O '3"

13CD

3.3 "CD

s i■D WoQ .CaO3

■DO

CDQ .

■DCD

I(/)Wo'

Table 4.2. Average thickness of different sub horizons of soil profiles at different landscape positions.

Marsh Type Saline Brackish Intermediate Fresh

Horizon Streamside Inland Streamside inland Streamside inland Streamside inland(SSS) (SIL) (BSS) (BIL) (ISS) (IIL) (FSS) (FIL)

0 a 1 28.3 ± 5 .8 t 26.7 ± 7.6 28.3 ± 7 .6 31.7 ± 7 .6 38.3 ±18 .9 1 0 . 0 ± 0 . 0 33.3 ± 2 .9 36.7 ± 1 1 .5

Oa2 28.3 ± 5.8 36.7 ± 23.6 31.7 ± 12 .6 33.3 ± 5.8 45.0 ± 26.5 53.3 ± 2 .9 38.3 ± 7 .6 40.0 ± 8 .7

Oa3 31.7 + 2.9 38.3 ± 2 .9 30.0 ± 5.0 35.0 ±1 3 .2 35.0 ± 22.9 68.3 ± 1 0 .4 38.3 ± 5 .8 33.3 ± 2 .9

Oa4 26.7 ± 7.6 18.3 ±16.1 15.0 ± 15 .0 21.7 ±2 0 .2 25.0 ± 25.0 1 8 .3 ± 1 1 .5 0 0 . 0 0 0 . 0

OaS 06.7 ± 5 .8 0 0 . 0 06.7 ±1 1 .5 0 0 . 0 06.7 ±11 .5 0 0 . 0 0 0 . 0 0 0 . 0

DML 121.7 ± 5 .8 120.0 ±1 3 .2 111.7 ± 2 .9 121 .7± 2 .9 >150 >150 1 1 0 . 0 ± 0 . 0 1 1 0 . 0 ± 0 . 0

t Mean ± Standard Deviation

re-deposition because the material is eroded away. The remaining material

within the 0 a 1 horizon may be compacted by the standing water at inland sites

when inundated. The thickness of the 0a1 horizon was very uniform at inland

but highly variable at streamside (Table 4.2). The variation in deposition and

sedimentation rates may cause the variability at streamside. The processes,

such as erosion and re-deposition, may create high spatial variability within the

marshes. Studying the spatial variability and identifying the associated patterns

within the landscape may help to understand the mechanisms of marsh loss.

The marsh degradation at IIL may be associated with seawater intrusion, loss

of vegetation, and lack of accretion, and eventually undergo local submergence

(Chmura et al., 1992; DeLaune et al., 1993;DeLaune et al., 1994; Nyman et al.,

1994b). This is following the most common marsh loss pattern, the scattered

broken marsh interior, which is less severe (Nyman et al., 1993a, 1994b). Re

deposition of the material within a close proximity may rate this loss as less

severe damage.

The Oa2 horizon thickness ranged from 28 to 53 cm for all marsh types.

Intermediate and fresh marshes had thicker Oa2 horizon. The Oa3 horizon

thickness ranged from 30 to 6 8 cm for all marsh types. The thickness of the

Oa3 horizon at inland was greater than that of streamside for all marsh types

except for the fresh marsh. The IIL had the thickest Oa2 and Oa3 horizons

despite its thinnest O a l horizon. This may indicate accumulation of organic

material either due to high accretion rates experienced before or concentration

of disintegrated material from degrading surrounding marshes. The fresh

114


marsh site had only three organic subhorizons. The Oa4 horizon thickness

ranged between 15 to 27 cm. The OaS horizon was very thin (about 7 cm) and

occurred only at SSS, BSS and ISS.

Depth to the mineral layer (DML) was thin at the fresh marsh site (110

cm) and it was deepest at the intermediate marsh (deeper than 150 cm). The

DML was equal to the total organic layer thickness. The deepest DML at the

intermediate marsh was associated with thicker Ga2 and Ga3 horizons, which

may be a depression at the surface of the mineral horizon. This can be due to

depositional differences when the mineral horizon was formed. This location

may be within an area that is away from distributories. The area closer to the

distributory experiences more mineral matter deposition can eventually form a

marsh with a shallow DML after organic accretions. Whereas, the depression

that is away from the distributory may receive more organic accretions and can

form a marsh with deeper DML. When marsh degrades these depressions

areas may be more vulnerable and can form the open water patches within the

marsh. Thicker subhorizons at this location may be indicative of the local

submergence within the marsh associated with marsh loss. Therefore, open

water patches with thicker subhorizons within marsh may be characteristic of

“hotspots” that are degrading or vulnerable marshes. This may be considered

as an indicator to assess marsh status. The relationship between occurrence

of hotspots and micro relief variations within the degrading marshes needs to

be further studied to test this hypothesis.

115


Previous investigations mainly studied the surface soil without paying

much attention to the subhorizons and soil morphology of these marshes. The

knowledge of soil characteristics to a particular depth is useful for marsh

management applications therefore, horizon designation from the surface also

seems appropriate. But it should be noted that these organic soils form from

bottom to up, as organic and inorganic accretions are the major soil forming

processes. The present accretions at the streamside can form a distinct Oa1

horizon whereas the present accretions may not be adequate for the inland to

separate the 0a1 horizon that is contemporary to the streamside. The

differences in accretion can be associated with the elevation differences

between the landscape positions (DeLaune et al., 1983). The inland O a l

horizon is more similar to the streamside Ga2 horizon and the inland G a 2

horizon is similar to the streamside Ga3 horizon and so on. Hence, the organic

subhorizon immediately above the Cg horizon at both landscape positions

(either G a4 or Ga5 horizon) can be formed at the same time under similar

environment whereas the G a l horizons at both positions can be different. The

Ga5 horizon was found for streamside but not for inland profiles.

Unequal number of subhorizons may indicate differences in present

accretion and thickness differences between contemporary horizons may

indicate differences in previous accretions, within a healthy marsh. However,

when a marsh degrades an incorporation of two adjacent subhorizons into one

subhorizon may occur. This may form a fewer number of subhorizons and

thicker subhorizons within the profile of a degrading marsh. Understanding the

116


characteristics of different subhorizons within the profile reveals information

about its formation and helps to explain the spatial variability of soil

characteristics at a particular depth within the marsh. Therefore, when

comparing subhorizons at different landscape positions it should be noted that

the horizons with similar designation do not necessarily mean that they were

formed under similar environments. This can be a major cause of spatial

variability within the marsh.

Statistical analysis indicated that the thickness of Oa1 horizon (p=0.03)

and O a3 horizon (p=0.02) was significantly different between landscape

positions within the marsh types. However, this difference was found only in

intermediate marsh; Oa1 horizon, p=0.002; Oa3 horizon, p=0.001. Thickness of

Oa2 and Oa4 horizons indicated significant differences between marsh types.

The O a2 horizon thickness at intermediate marsh was significantly different

(p=0.06) from that of saline and brackish marshes. The significant difference

of O a4 horizon between marsh types is mainly due to the absence of Oa4

horizon at the fresh marsh. Depth to the mineral layer was not significantly

different between landscape positions (p=0.30) or between marsh types

(p=0.17). Higher variability of subhorizon thickness and variable number of

subhorizons within profiles may be indicative of high spatial variability organic

layer thickness within a site. Based on the number of subhorizons, the fresh

marsh seems to be less disturbed and stable. Having a higher number of

subhorizons at SSS, BSS, and ISS may indicate a higher degree of disturbance

experienced by the saline, brackish, and intermediate marshes.

117


Variation in soil BD within the profile is presented in Figure 4.2. The BD

was higher for the saline marsh compared to the brackish marsh for all the

subhorizons. Higher BD may be due to more mineral matter in the sample.

This can be attributed to higher sedimentation rates at the saline marsh

compared to the brackish marsh. Higher sedimentation rates may be expected

due to more sediment supply, as well as, higher flocculation of sediments within

highly saline waters. Lack of sediment feeding to the brackish marshes may

cause low bulk density due to lack of mineral material and eventually low

accretion rates. The mineral layer at site 1 had a BD of 1.08 Mg m'^ compared

to 0.65 Mg m'^ at site 2. BD of the organic layers was very similar for the saline

site (i.e., 0.26 to 0.31 Mg m'^), but at the brackish site, BD varied within the

organic layer (ranged from 0.10 to 0.2 Mg m'^). The BD values are comparable

to those reported by Nyman et al. (1993); they reported mean bulk density for

the top 15 cm of soil as 0.23 Mg m'^ for saline, and 0.12 Mg m'^ for

brackish/saline marshes. The BD variation within the brackish marsh profile

may be due to long-term temporal variations in sedimentation rates.

4.4.2 Soil Salinity

The EC measurements from 1:10, soil:water suspensions were dilution

corrected and converted to salinity as TDS using the relationship between EC

and TDS. The relationship between EC in dS m' and TDS in ppt, was

constructed based on the gravimetric method.

TDS = 0.669 * EC - 0.6208 (r = 0.96) [4.12]

118


rg

sz2(0E(A

2mEl

2

i<DC

1

roO

%O

CO(0O o Ü

ooCD

COo

CD

CVJCD

E05

05Q

3CO

mJC2TOEx:

Üm

"Occc0)c

1 0O)

CO5T3

OCO

co

Î>C \j

2305

uozuoH l!os

1 1 9


Statistical analysis indicated highly significant differences in salinity for

all the subhorizons (p-value ranged from 0.004 to 0.0001) and for the weighted

averages of the profiles (p=0.0001) between marsh types. This confirms that

the sites represented significantly different salinity levels. Mean comparisons

indicated that the salinity for O a l, Ga3, and Cg horizons and weighted average

salinity for the profile were significantly different from each other. The salinity of

the Qa 2 horizon was significantly different except for saline and brackish

marshes. These sites may be under similar salinity regime during the G a2

horizon formation. The saline site had characteristic saline marsh vegetation

and was located within the saline marsh area according to previous maps.

However, the salinity of the SSS and SIL was not within the range of saline

marsh, which is above 10 ppt (Soil Conservation Service, 1989).

Figure 4.3 presents the salinity profiles for the SSS and SIL. Salinity

ranged from 6.36 to 8.40 ppt within the profile at the SSS compared to the

averages of SIL, which ranged from 6.30 to 7.23 ppt. Salinity of the G a l and

Cg were similar at both landscape positions. The SSS had higher salinity

(about one ppt) than the SIL for other subhorizons except the O a l horizon.

This may be attributed to confined or slow water movement between the

landscape positions. Increased salinity of the subhorizons at SIL could be a

result of salt concentration within the root zone when these horizons form.

Plants adapt through different mechanisms to tolerate high salt concentrations

in the root zone (Carlson and Forrest, 1982; Bradley and Morris, 1991). Ion

exclusion mechanism cause concentration of salts within the root zone for salt

120


CO COm □

CO

CO CL&COQI -

coCO

CM

roO

%O

COCDo O

toCDO o

U O ZU O H VPS

sz%E0)c

m

T3CCD

TJCCD

COCO

mSC/)

11Q.

c15COCO

2 3.g)U_

121


tolerant plants (Bradley and Morris, 1991; Latham et al., 1994). Higher salinity

may also be a result of insufficient mixing of pore water within the profile at SIL.

The BSS salinity ranged from 4.70 to 7.01, while at BIL, salinity ranged

from 5.18 to 7.30 ppt. Salinity profiles were very similar at both landscape

positions in the brackish marsh (Figure 4.4). Small streams brought water into

the interior marsh, which maintained similar salinity for both landscapes. This

may be the reason for similar salinity profiles at the BSS and BIL because there

was a small stream running through the BIL. Both the ISS and IIL exhibited

extensive marsh degradation, and the salinity of the ISS ranged from 2.0 to

4.64 ppt compared to 1.47 to 4.95 at IIL. Salinity profiles were very similar for

both landscape positions at the intermediate marsh (Figure 4.5) suggesting free

water movement between landscape positions. More than 2 ppt difference was

observed within the profiles with higher salinity for the subsoil. This may

indicate a local flooding effect as a consequence of marsh loss due to

increased salinity. Figure 4 .6 presents the salinity profiles at the fresh marsh

site, and it’s salinity was extremely low compared to the other sites. Salinity did

not change between profiles at the FSS and FIL, except the FIL Oa2 horizon

had about 1 ppt higher salinity. Weighted average salinity for each soil profile

was estimated to a 150-cm depth (Figure 4.7). Weighted average salinity

ranged between 6 and 8 ppt for brackish and saline marshes. Weighted

averages for ISS and IIL were below 4 ppt. This may be due to the exclusion of

the Cg horizon within these profiles. The FSS and FIL had about 1 ppt

weighted average salinity for their profiles.

122


00 CO

00

CD -KQ .Q .

COQt

ÇCO

CO

eg

21

J 3

u2 XI(U

0 -TOc(0

"Oc(0

COCO000)■g

‘ coEs

1CL

CI dCO

T|-2 3 O)

UOZUOH l!OS

123


m □

00

CO a.Q.

COQ

ç

I

CM

%E

0)

I0)

■acJ5.5■ocm(O(O

ïcoE3

1Q .

ç"mcoiri

23OJ

UO ZU O H n o s

124


m □

00

CO CLsCOQ1-

coco

CM

mE

m(U

. p

T3Cra_c■acCD

COco

Q)2coE3

1

CL

.•S'_c■(ÔcoCD

2 3.rau_

UOZUOH NOS

125


00

CO

CVJ

Q.Q .

COQt

c^ := cn CO

EÜom

a.0)

T3CD

2

1Q.CU

2

c"cDW

(UO)2

IS.5)m

N-

2305

e ; is p u B u o jjjs o d a d E o s p u e i

126


4.4.3 Pyrite Accumulation and Pyrite Profiles

Pyrite content is expressed volumetrically as mg of pyrite (FeS 2 ) per L of

soil enabling direct comparisons between marsh types despite the BD

variations. Statistical analysis indicated that salinity influenced pyrite

accumulation, but its effect was different from horizon to horizon. Pyrite content

of the Oa1 horizon was significantly different between landscape positions

within the site (p=0.02), but was not significant between marsh types (p=0.50).

Pyrite content within the 0a1 horizon is controlled by the balance of pyrite

formation and oxidation (Feijtel et al., 1988). Therefore, the difference between

sites may not reflect the influence of salinity on pyrite formation. It may be

strongly influenced by the hydrological changes, which affect pyrite oxidation.

Tides, rainfall, and groundwater intrusion influence water level fluctuations

within the marshes (Hemond and Fifield, 1982). Plant water uptake could result

in air entry into the surface soil and sediment oxidation (Dacey and Howes,

1984). However, this may not be significant for these marshes because of the

tidal influence. The mechanism of oxygen diffusion through aerenchyma tissue

from aerial parts of the plant to the roots combined with leakage of oxygen into

rhizosphere is insufficient to overcome high dissolved sulfide toxicity by the

spartina. Internal enzymatic oxidation is responsible for tolerance of high

sulfide concentrations by spartina compared to other plants such as rice

(Carlson and Forrest, 1982). Therefore, it is doubtful whether the oxygen

transported through plants is responsible for pyrite oxidation. Any difference of

127


pyrite content within the Oa1 horizon between landscape positions may be due

to the balance of pyrite formation and oxidation. The landscape position

significantly influenced pyrite accumulation at the saline marsh. The mean

value was 4214 mg FeS 2 L" for SSS compared to 846 mg FeS 2 L' for SIL.

This may be due to more oxidation of pyrite within the O a l horizon at the SIL.

Landscape position effect was not significant within other marshes but at the

brackish marsh a lower p-value (0.15) was found when compare the means

between BSS and BIL (2057 vs. 610 mg FeS 2 L'^). This indicates that the

landscape position effect is prominent for highly saline areas. Removal of

HCOg" from the system promotes pyrite formation (Fanning and Fanning, 1989).

High flushing activity may promote pyrite formation at streamside while

accumulation of HCO 3 can inhibit pyrite formation. This can be another reason

for landscape differences of pyrite accumulation. Haering et al. (1989) found

that pyrite accumulation was favored for streamside profiles that received Fe

inputs from tidal flushing. Lord and Church (1983) identified the O a l horizon as

the frequently oxidizable surface layer. Therefore, pyrite oxidation may be

mostly responsible for less pyrite accumulation within the O a l horizon at SIL

compared to SSS.

For the O a2 horizon, neither the marsh type nor the landscape position

significantly influenced pyrite accumulation. This may indicate that pyrite

formation was similar and oxidation due to water table drop did not occur within

the O a2 horizon of the BIL. The landscape position effect on pyrite

128


accumulation dramatically diminished (p=0.80) with increasing depth (Oa3

horizon). However, pyrite accumulation by marsh type was significantly

different (p=0.03). Major differences were found between brackish marsh and

other marsh types. The brackish marsh had the highest mean pyrite

accumulation in the O a3 horizon (4365 mg FeS 2 L'^), while the saline marsh

had 1757 mg FeSg L' , intermediate marsh had 1807 mg FeSg L'"' and the

fresh marsh had 1070 mg FeS 2 L ' \

Within the Cg horizon, pyrite accumulation was independent of both

marsh type and landscape position effects. The weighted mean pyrite content

had no influence from the landscape position or site effects; however, the p-

value was low (p=0.17) for the marsh type effects. This indicates the higher the

salinity of the marsh, the higher the accumulation of pyrite within the profile.

Saline and brackish marshes had means over 2500 mg FeS 2 L"\ while the

means for intermediate and fresh marshes were less than 1600 mg FeS 2 L ' \

As revealed by the statistical analysis, pyrite content was five-fold higher

for the O a l horizon at SSS compared to the SIL (4214 vs. 846 mg FeSg L' )

(Figure 4.8). Other horizons within the profile did not show much difference

between landscape positions. However, it should be noted that the

contemporary subhorizons contain similar amounts of pyrite, the SSS OaS vs.

SIL Oa4 and SSS Oa4 vs. SIL Oa3. Higher pyrite accumulation within the SSS

OaS and SIL Oa4 suggests they represent a subhorizon that was formed under

high saline condition. Higher amounts of pyrite for the Cg horizons (4710 at

129


w wm □

ooooo

ooo

OOOCD

OOOin

ooo

oooCO

ooorg

ooo

O)E,

IoO

ê>.Q.

£2mEmc13wm

CO

T3CCD

•DCCO

COCO52.CDjg

II11CL

•I>%Û.00

23D)

UO ZUO H n o s

130


SSS and 2768 mg FeSg L' at SIL) can be attributed to higher mineral material

within the Cg horizon.

Pyrite accumulation for the brackish marsh indicates a similar pattern of

increased accumulations for the streamside. The Oa3 and OaS horizons

contained more than 5000 mg FeSg L' (Figure 4.9). Pyrite accumulation was

uniform within the organic layer for both landscape positions of the intermediate

marsh and the highest accumulations were found for the bottom organic

horizon (6284 mg FeSg at ISS vs. 5553 mg FeSj at IIL) (Figure 4.10).

Similar to the saline marsh site, ISS OaS and IIL Oa4 seems to represent a

subhorizon that was formed under high saline condition. The Cg horizon was

deeper than 150 cm at both ISS and the IIL and not included for the profile in

Figure 4.10. The streamside data showed more accumulation at all depths.

Pyrite accumulation was minimal for the fresh marsh site except for the

O a2 horizon at the FSS (Figure 4.11). Pyrite content for the O a l horizon was

83 mg FeSg L ’ at the FSS, which was the lowest. Figure 4.12 presents the

weighted average pyrite content to a 150-cm depth for all of the soil profiles.

The highest pyrite content (>3500 mg FeSg L' ) was observed for the BSS while

the lowest (<1500 mg FeSg L^) was for the stable marshes at the FIL.

Most of the soil profiles indicated the highest pyrite accumulation within

the thin layer of dark organic material that occurred immediately above the Cg

horizon. This was characterized with a strong sulfide smell in the field. Except

for the FSS, FIL, and BIL, this horizon contained more than 5000 mg FeSg L \

Its contribution was not dominant when calculating the weighted average for the

131


CO CO

oooCO

ooo

ooo(O

oooin

ooo

ooo0 0

oo

ooo

S2(0E

I(D0)

_ C

CO

J5c

CT (0E "—'03c ^ 5 DOcoÜ

•I>>CL

<U*< /)

E

IJ

Q .

s"C>.CL

05•cf23O)

UOZUOH NOS

132


1

ooooo

ooo

oooco

oooin

oo§

oooco

oooCN

ooo

S I2CDE

IEa

0)

■Dcjgc

O) *o& sc W B Wc C3-o I•■§ E> , CD û. 0)

B1Q .

B'C>.a.

"4"E3O)

CDoCMCDo

coCDo o

lOCDo

U O ZUO H n o s

133


(/) , (/)L L U _

m □

oOO00

OOON

OOOCO

OOOin

ooo

oooCO

oooCNJ

ooo

JZ£2

E

Im

T JcJO^ .5

"gD ) (13

WCO

0;g'inEs

cBcoO

Û-

i1Q .

B'=>.D.

E3.25u_

COO

<N(0O

COCDo o

UO ZUO H n o s

134


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8

( O '

3.3 "CD

CD■DOQ .C

ao3"Oo

CDQ .

■DCD

WÜ1

2 FSS

o BSS

0 1000 2000 3000 4000 5000

Pyrite concentration (mg L'^)

Figure 4.12. Weighted average of pyrite content for the soil profile to a depth of 150 cm.

C /)C /)

profile because of its thickness. The mechanism of this concentration of pyrite

was not clear. However, several hypotheses on its formation can be

developed. This may be formed under highly saline condition with high S and

labile C supply at the initial stage of marsh formation. It may represent bottom

sediments of a lake at that time, which was developed into a marsh later. This

could be the more logical way of formation of this soil layer. Another

hypothesis is the pyrite preferably formed adjacent to the mineral layer because

of high Fe availability. Dissolved organic compounds moved through the pore

water with reduced S and Fe to the bottom of the organic layer and then formed

pyrite. Pyrite formed within the organic layer where more labile carbon may be

present, and eventually moved to the bottom and accumulated at the top of the

mineral layer because the vertical movement was confined. Another hypothesis

is a moving delta with seawater and vegetation that grew for a certain time

period. During that time pyrite accumulated highly within the root zone with

favorably reduced conditions and availability of soluble carbon. With time, the

area would have been isolated from the sea, marsh grew up without much

pyrite formation. However, these scenarios need to be investigated.

4.4.4 Non-pyritic Fe

The HCI digestion during the indirect pyrite determination technique

extracted non-pyritic Fe and S from the soil samples (Schnider and Schnider,

1990). The non-pyritic Fe and S profiles are presented in this section.

Figures 4 .13 through 4.16 present the non-pyritic Fe profiles for the

different sites. The non-pyritic Fe content for the SSS ranged from 1055 to

136


CO CO CO CO 0 □

1

T

oooCO

oooCO

ooo

oooCM

oooo

ooo0 0

oooCO

ooo■M-

oooCM

nmE0).5V)m

JZ

CO

-D

I

Î Îl ï - gÜ M

o

Co

CO

23.5)LL

COO

%O

COCOO O

%O ü

uozuoH NOS

137


CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8■D

3.3 "CD

CD■DOQ.C

aO3"OO

CDQ.

0 a 1

Oa2

g Oa3goX

w Oa4

CO00

Oa5

cg

BBSS

;ÜBIL

2000 4000 6000 8000 10000 12000 14000 16000 18000

T3(D

(/)(/)

Non-pyritic Fe (mg L'^)

Figure 4.14. Non-pyritic Fe profiles for streamside (BSS) and inland (BIL) for the brackishmarsh.

CD■ D

OQ.C

gQ.

■DCD

C/)C/)

8■D

3.3 "CD

CD■DOQ.C

aO3"OO

CDQ.

0 a 1

Oa2

c

I° O a 3

6W

wCO

Oa4

Oa5

EOISS

0 2000 4000 6000 8000 10000 12000 14000 16000 18000

■DCD

C /)C /)

Non-pyritic Fe (mg L )Figure 4.15. Non-pyritic Fe profiles for streamside (ISS) and inland (IIL) for the intermediate

marsh.

% =d u. u_m □

oooCO

oooCD

ooo

oooCM

OOoo

oooCO

oooCO

ooo

oooCM

£2COE

I<DJC.

TDcm

-ac

i lCD

o E

II1CL<D

co

CD

2305

CDoCM(0o

CO

oO)Ü

U O ZUO H I P S

140


8006 mg Fe L ’ within the organic layer compared to 2076 to 8717 mg Fe L ’ for

the SIL. Higher Non-pyritic Fe content at SSS OaS and SIL Oa4 are similar

and also suggests that they represent one subhorizon that was formed with

high Fe supply. This may be formed when the Mississippi River supplied more

sediment to this area. This may also help for high pyrite accumulation within

this subhorizon. Non-pyritic Fe content for the mineral layers (more than 15000

mg Fe L' ) were higher than that of the organic layers at both landscape

positions due to high mineral matter within the Cg horizon. Non-pyritic Fe

profiles followed a similar pattern for the BSS and BIL. However, lower non-

pyritic Fe content was found for the IIL compared to the ISS. For FSS, the non-

pyritic Fe content within the organic layer varied from 1528 to 1722 mg Fe L

as compared to 1764 to 2272 mg Fe L' for the FIL. The variation within the

profiles was very similar for the fresh marsh (Figure 4.16). Figure 4 .17

presents the weighted average non-pyritic Fe contents. For the intermediate

marsh, the weighted averages were very low compared to other sites because

the profiles at ISS and IIL did not contain a Cg horizon within 150 cm. The non-

pyritic Fe content of the Cg horizons is very high because of the high mineral

matter within the Cg horizon.

Statistical analysis indicated that the non-pyritic Fe content was

significantly different between the landscape positions within the marsh types;

differences were detected between SSS and SIL (p=0.0005) and BSS and BIL

(p=0.036). For both marsh types, the streamside samples contained higher

mean non-pyritc Fe content (saline marsh, 4401 vs. 2543 mg Fe L \ brackish

141


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D

CD

3.3 "CD

CD■DOQ .C

ao3T3O

(DQ .

M

FIL

FSS0)w•g IIL m

I ISS■(/)o°- BIL0)Q .

8(0

C

BSS

SIL

SSS

2000 3000 4000 50000 1000 6000

T3(D

-1,Non-pyritic Fe (mg L )

Figure 4.17. Weighted average non-pyritic Fe for the soil profile to a depth of 150 cm.

C /)C /)

marsh, 3630 vs. 2653 mg Fe L' ). This suggests that the inland area may be

suffering from lack of sedimentation or higher sub-surface erosion. This may

be the reason that the O a l horizon in SIL contained less pyrite compared to the

SSS (see section 4.4.3). Haering et al. (1989) also found that pyrite

accumulation was favored for streamside profiles that received Fe inputs from

tidal flushing.

Non-pyritic Fe content for the Oa2 horizon was significantly different

between marsh types; intermediate marsh was significantly different from the

other marshes and was the lowest among them. This may be due to the

disintegration and erosion of material within the intermediate marsh profiles.

The Fe within the Oa2 horizon may be diffused down to the lower horizons of

ISS and IIL, where highest pyrite contents were observed within the Oa3

horizons (see section 4.4.3). This also suggests that lower non-pyritic Fe

content can be a good indicator for a degrading organic soil horizon.

The fresh marsh had the lowest non-pyritic Fe content within the Cg

horizon and was significantly different from all other marsh types. The

weighted mean non-pyritic Fe content was significantly different between marsh

types (p=0.007), as was the landscape positions within the sites (p=0.009).

The lowest mean was measured for the intermediate marsh because the

profiles at ISS and IIL did not contain a Cg horizon. Means for all the marsh

types were significantly different from each other. Landscape position effect

was non-significant for the intermediate marsh, which was extensively broken-

up (less than 50% vegetation cover). Other sites indicated significant

143


differences between landscape positions; inland contained higher amounts

than the streamside for the saline and fresh marshes, but it was reverse for the

brackish marsh. The depth to mineral layer (DML) is similar at both landscape

positions at saline and fresh marshes, therefore, the main differences can be

contributed to organic layers. The higher non-pyritic Fe content within the

organic layers of inland may indicate less pyritization that was not due to lack of

Fe. This may be due to inhibition of pyrite formation or promoted pyrite

oxidation as discussed in section 4.4.3. The DML of the BSS was shallow

hence the Cg horizon was thicker. The non-pyritic Fe content within mineral

horizons was very high. Non-pyritic Fe at BIL may have flushed away through

the small streams that run through the inland marsh. This may be another

reason for low non-pyritic Fe at BIL. Therefore, the non-pyritic Fe content was

lower for the profile at BIL.

4.4.5 Non-pyritic 5

Figure 4 .18 presents the variation in non-pyritic S within the profiles for

the saline marsh. The SSS had higher non-pyritic S for all subhorizons and the

quantity ranged from 842 to 1090 mg S L' ) compared to the SIL, which ranged

from 695 to 950 mg S L '\

The non-pyritic S content for the BSS ranged from 389 to 956 mg S L

and ranged from 584 to 1037 mg S L' for the BIL. The surface organic layers

contained higher values at both landscape positions (Figures 4.19). Profile

variation of non-pyritic S content for the intermediate marsh is presented in

Figure 4.20 in which the Cg horizon was below 150 cm. Non-pyritic S contents

144


œ (/)

ooCM

ooo

OO00

ooCD

O)

CO

S:o

Z

oo

ooCN

JCÿ.mE0)ç"râ</)03sz

•oc_ç5ç

■ocro

COCOw(D■g(/)I0

Q .

COü

Ècoz00

23.g)u_

UOZUOH NOS

145


CO COm □

ooCM

OOO

OOCO

ECO

CO

a>

CO

•OcCO

T3cCO

CO

OOCO

COa

co

oo

<D COE

I"co

1 Q .

CO a

à :

ÉzCJ>

■M"23

L i.

(0O

CM(0O

COCOO O

U5COo

O)Ü

UOZUOH NOS

146


ooo

oooo

ooCD

oo"4"

OOcg

I■si

I<D

-C

T3c

£

clIICO q)u :g•ti CO

1 1o CO

x:£2(0E

1CL

COü

&cozCDCN

■cf230 5

UOZUOH nos

147


were similar within the organic layer and ranged from approximately 500 to

1000 mg S for both ISS and IIL

For the fresh marsh, non-pyritic S content was almost half that of the

other sites and ranged between 300 and 600 mg S L^ (Figure 4.21).

The weighted average S content for each landscape position is presented in

Figure 4.22.

Non-pyritic S content in the O a l horizon was significantly different

between the marsh types. This may be mainly due to the S supply variation

that can be associated with salinity and seawater intrusion. The fresh marsh

had the lowest mean of 439 mg S L ’ for the O a l horizon and was significantly

different from the means of the other sites, which are more than twice that of

the fresh marsh. Landscape position effect was not found to be significant for

the O a l horizon. The SIL may be frequently flooded and receives 8 from the S

enriched water or else 8 forms recycle within the O a l horizon from the

frequently oxidizable surface layer (Lord and Church, 1983). Both marsh type

(p=0.018) and the landscape position (p=0.036) had a significant influence on

the non-pyritic 8 content in the Oa2 horizon. Intermediate and fresh marshes

were significantly different from other marshes. Lower non-pyritic 8 contents at

less saline areas can be attributed to low 8 supply, especially in the fresh

marsh. For the intermediate marsh, sulfate reduction may continue to form

pyrite, which resulted in lower non-pyritic 8 . A higher pyrite content was

determined for the brackish marsh. The BIL was significantly different from the

148


ooCM

wWu_

I i

ooo

jC2(0E

m0)

oooo

TooCO

_L

t O .

êCOü

è :É

oo■M-

T3C_rac

X3Cm

COCO

m;g

I

Io5

Q .(/)ü

OOCM

co

CM

23O)

mO

%O

COmO

enü

U O ZU O H NOS

149


ooCM

OOO

Oo0 0

ooCD

ooM-

OoCM

9 %S pue uoi^isod adeospuB"!

Ig

0

t0)■Dm3

ex

6i/>d)

J Z

II^ 2 0) ^ D)

1T3ÆJZro0)

CMCMM"e3.g)U L

150


BSS (p=0.03). This can be a result of limited pyritization at BIL that contained

lower non-pyritic Fe compared to BSS.

Intermediate and fresh marshes had significantly lower non-pyritic S

contents within the Oa3 horizons compared to other marshes; fresh marsh had

the lowest mean (470 mg S L ’). Mean for the non-pyritic S content for the Cg

horizon for the fresh marsh, was the lowest among marsh types (324 mg S L' ).

Lower horizons indicate pronounced differences of S supply into the less saline

areas. Weighted mean averages of non-pyritic S for the profile indicated

significant differences between marsh types (p=0.03) and the landscape

positions (p=0.01). The fresh marsh had the lowest mean (441 mg S L' ) while

the other marsh types had about twice that amount (saline, 903; brackish, 824;

intermediate, 730 mg S L ’). The landscape positions were significantly

different for the saline and brackish marshes. The SSS had higher amount of

non-pyritic S than the SIL, but this trend was reversed for the brackish marsh.

Differences between the marsh types may be attributed to less S supply, high S

consumption by the plants and the sulfate reduction. Landscape position

differences existed at the more saline areas. Streamside may contain more S

compared to inland when the inland is landlocked. However, when a stream

breaks the hydrological barriers, non-pyritic S may be accumulated within

inland profiles due to limited pyritization.

4.4.6 Soil pH

Variations in soil pH for different marsh profiles are presented in Figure

4.23 for the streamside and Figure 4 .24 for the inland. Differences between

151


CO CO CO CO CO 00 u_

CJ

inmO

O

IoX

oCO

COmO

CMmO

Q.

«iTO

* 0(U•gTOE

I«£

Q.a.CD

COCM■M"23.gU_

TOO

O) 00 CO in CO CM

Hd

152


ü

"%o

CO(0O

O)

c/3 CQ =□ EU m DD

%O

njO

CO in CO CM

C

Io

X

■ 5CO

oQ.

x:E2(0Em'o•Scroc

1Q.XQ.0)iS

2 3 O)

Hd

153


marsh types were more pronounced than differences between the landscape

positions. The pH ranges for the soil profiles were as follows; for SSS and SIL,

5.98 to 7.5; for BSS and BIL, 5.44 to 8.17; for ISS and IIL, 4 .29 to 5.71; for FSS

and FIL, 4 .32 to 6.73. The Cg horizon always had the higher pH. Note that

ISS and IIL did not have a Cg horizon within a depth of 150 cm.

There were statistically significant soil pH differences between marsh

types (p=0.04) and landscape position within marshes (p=0.04) for the Oa1

horizon. The O a l horizon of intermediate and fresh marshes was acid and

significantly different from the more saline marshes. The ISS pH (4.29) was

significantly lower than IIL pH (5.13), showing landscape position differences

only for the intermediate mash. Mean pH of the Qa2 horizon was the lowest for

the fresh marsh (4.41 ) and highest for the saline marsh (6.65). The differences

between marsh types were highly significant for the pH of the Ga2 horizon.

The pH of the Ga3 horizon ranged between 4.86 and 6 . 6 6 among marshes and

the more saline marshes had higher pH. This may be the influence of salinity.

Except for the brackish and intermediate marshes the Ga3 horizon pH was

significantly different between marshes. The pH of the Cg horizons was

significantly different between marshes; brackish marsh had the highest pH of

8.15.

4.4.7 Microanalysis of Soils for Pyrite

Information on clay mineralogy and other associated minerals for coastal

marsh soils indicate the prominent pedogenic processes within the marsh

ecosystem. Coastal marsh environments favor pyrite formation. Pyrite crystals

154


occur in the form of framboids or clusters of single crystals. Framboidal pyrite

forms when Fe monosulfides react with elemental S. Other crystal forms of

pyrite may be formed as a result of polysulfide oxidation. The observations

from XRD and SEM are presented in this section as evidence for pyrite within

these soils. Pyrite framboids were detected within salt marsh sediments.

Figure 4 .25 presents a micrograph showing the microstructure of the organic

sediments. This indicates the micropores and the surface area within the

sediments. Pyrite was detected within the salt marsh sediments as framboids

(spherical aggregates of microcrystals) and clusters of single crystals. Pyrite

framboids were associated with decaying organic material (Figure 4.26). White

arrows indicate the framboids. The framboids are about 8 pm in diameter, and

some of them have disintegrated into single crystals. An enlargement of the

clusters of pyrite crystals is presented in Figure 4.27 and a highly magnified

pyrite framboid is shown in Figure 4.28, which has a diameter of about 6 pm.

The X-ray diffractograms (ZRD) data are presented for air-dried versus

treated smears from the Oa1 horizon to identify pyrite and clay minerals.

Figure 4 .29 presents the XRDs for the < 2 pm fraction from the saline marsh

sediments, while Figure 4.30 presents the < 20 pm fraction. Smectite, illite and

kaolinite peaks were easily detectable for the < 2 pm fraction. For the 20 pm

fraction, quartz peaks are dominant and clay peaks were weak but higher

intensity was observed for the pyrite peaks. The primary peak for pyrite occurs

155


m

1 12 (0

1*o2

%

CO

2ü

O )c

î-CCO

Q .2O)20Ec

11O)c

’ccswiriCM

2 3g)

UL

156


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D

CD

3.3 "CD

CD■DOQ .C

aO3"OO

CDQ .

■DCD

C /)C /)

en

Figure 4.26. Scanning electron micrograph showing pyrite framboids within organic material of saline marsh sediments from the 0a1 horizon. P = Pyrite framboids.

158

IoJZ

(0oa>£EP

SE

1IroEmç

1

(/>■o0SiE

•I' 6CL

1ü

23O)


c

§

CQO0)£EP

c0 )EisCO

fCOE(U_çlôCO

•gEE2

Ià_0)gCO

00CNj

23D )

159


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D( O '

3.3"CD

CD■DOQ .Cao3

T3O

CDQ .

O)o

5000

S=Smectitel=llliteK=KaoliniteQ=Quartz

550 C4000

300 °C3000

C

c2000

EG (Ethylene Glycoiated)

1 0 0 0 i

Air dried

4020 30 50 60100

"OCD

(/)(/)

20

Figure 4.29 The XRD patterns of saline marsh sediments (<2 pm) from the Oal horizon.

CD■oOQ .cgQ .

■oCD

C/Î

o'3

3CD

C3.3 "CD

CD■oOQ .CaO3

■oO

&

oc■oCD

C/5

o'3

2500

2000 ^

O)

(/)cI

1 550 °C Q

1500

1000

500 ^

Q300 X

S IK

S=SmectiteNllliteK=KaoliniteQ=QuartzP=Pyrite

Ethylene Glycoiated

Air dried

10 20 30

2 0

40 50 60

Figure 4.30 The XRD patterns of saline marsh sediments (<20 pm) from the Oal horizon.

at 28° 20, and the other orders occur at 33, 36.5 and 40.3° 20. The X R D ’s for

the brackish marsh sediments are presented in Figure 4.31 (< 2pm) and in

Figure 4 .32 (< 20 pm). Pyrite peaks also were detected from the brackish

marsh sediments. The XRD patterns indicated similar clay mineralogy in the

surface horizon (O a l horizon) of brackish and saline marshes. The XRD data

from both saline and brackish marshes indicated more pyrite within the less

than 20pm size fraction as compared to the < 2pm fraction. Figure 4.33

presents differential XRD pattern of <20pm fraction between saline and

brackish marsh types and between the 2 and 2 0 pm fractions for saline and

brackish marsh types. The main difference between the size fractions was the

higher quartz and pyrite peaks for the <20 pm fraction. The differences

between the saline and brackish marsh < 2 0 pm, indicates higher peaks for

clays and quartz within saline marsh sediments. These data provide

mineralogical evidence for pyrite accumulation within the salt and brackish

marsh sediments.

4.4.8 Soil Classification

There is a considerable amount of 8 containing compounds

accumulated within the soil profile. They occur as pyrite or other reduced 8

forms, in addition to the organic 8 fractions and the sulfates from the seawater.

Most of these compounds are found within the surface or subsurface tire of the

control section of the soil profile. Based on this information, and considering

the need to indicate the possible environmental impact on using these lands, it

162


CD■ D

OÛ .C

gQ.

"DCD

C/)C/)

3000

(D

8

33 "CD

CD■DOQ.C

aO3"OO

CDQ.

■DCD

C/)(/)

O)Ca)

2500

2000

s 1500

1000

500

550 C

S=Smectitel=illiteK=KaoliniteQ=Quartz

300 C

Ethylene Glycoiated

Air dried

I I l _____

10 20 30

20

40 50 60

Figure 4. 31 The XRD patterns of brackish marsh sediments (<2 pm) from the Oal horizon.

CD"D

OQ.C

gQ.

■DCD

C/)C/)

CD

8■D( O '

3.3"CD

CD■DOQ.C

aO3"OO

CDQ.

■DCD

(/)(/)

2000

1600

^ 1200 Bc0)

400

Q

550 C

300 XQ

S I

S=Smectitel=llliteK=KaoliniteQ=QuartzP=Pyrite

EG

Air dried

0 5 10 15 20 25 30 35 40 45 5020

Figure 4.32 The XRD patterns of brackish marsh sediments (<20 mm) from the Oal horizon.

CD■DOQ.C

gQ.

■DCD

C/)C/)

8

ci'

33 "CD

CD■DOQ.C

aO3"OO

CDQ.

■DCD

O)CJI

(/)c0)

cgê

1900

1400

900

400

-100

-600

Q

Saline - Brackish (<20 pm) 'V wiWwHWK'if#*"

I Brackish Marsh

Saline Marsh

10 20 30

20

40 50

S=Smectitei=j|liteK=KaoliniteQ=QuartzP=Pyrite

60

Figure 4.33 Differential XRD patterns between 2 and 20 pm fractions fromthe Oal horizons at saline and brackish marshes.

C/)C/)

is suggested that the sulfuric or sufidic horizons need to be designated within

these soils. The technique used in this study to estimate pyrite content also

provides non-pyritic S content, while the pyritic S content can be estimated from

the pyrite estimates (Schneider and Schneider, 1990). This method has the

potential to be used in soil characterization, especially to identify sulfidic

material (pyritic and non-pyritic S). Otherwise, new differentiae need to be

introduced to classify these soils to imply the potential problem for the use and

m anagement of these soils. A symbol may be introduced to identify pyrite

accumulation, similar to “j” that designates jarosite accumulation (Soil Survey

Staff, 1996). Designation of pyrite accumulating horizon also will be useful in

the development of a scheme for classifying pond soils for aquaculture (Boyd,

1995).

4.5 Summary and Future Research Needs

Organic layers comprise distinct subhorizons within the soil profiles. The

number of subhorizons varies within the landscape and between marsh types.

A thin layer of dark organic material occurs immediately above the mineral layer

with a strong sulfide smell within most of the soil profiles. These organic soils

develop from bottom to up mainly by the organic and inorganic accretions that

will counter local subsidence. Accretions are higher at the SS than the IL,

therefore, a higher number of subhorizons may be found at SS compared to IL.

Eventually contemporary soil layers occur at the bottom of the profiles but not

at the surface between the landscape positions. Therefore the subhorizons

166


that have the same horizon designation may or may not be contemporary, and

this can be a major source of spatial variability within the marsh.

Thickness of the Oa1 and Oa3 horizons is statistically different between

the landscape positions within the intermediate marsh. Incorporation of two

adjacent subhorizons into one subhorizon may occur within a degrading marsh.

The spatial variability of the 0a1 horizon between SS and IL can be an

indicator of the marsh stability at a particular location. Erosion and re

deposition of organic soil material may create wide spatial variability within a

degrading marsh. If the landscape pattern of the variability is identified, the

mechanisms of marsh loss within an area can be revealed. The degrading Oa1

horizon can release mineral material for subsurface horizons while losing its

integrity. Peat collapse may promote consolidation and compaction within the

subsurface horizons because water pressure can be extended deeper into the

profile when flooded. Thickness of the Oa2 horizon at the intermediate marsh

is significantiy different from the other marshes. Therefore, information on the

other organic subhorizons is required to support the data from the 0 a 1 horizon.

The DML equals to the total organic layer thickness unless the marsh surface is

under standing water. Variations in DML can be due to the depositional

variations when the mineral layer was formed. Depressions formed at that time

may have received more accretions and formed marshes with deeper DML.

When marsh degrades these depressions may be more vulnerable and can

form the open water patches within the marsh. The relationship between

occurrence of hotspots and micro relief variations within the degrading marshes

167


needs to be further studied to test this hypothesis. However, the variations in

number of organic subhorizons, and thickness of any subhorizon are indicative

of higher spatial variability within a marsh.

Soil BD was lower for the brackish marsh compared to the saline marsh.

Lack of sediment accumulation into the brackish marsh areas may be the

reason for lower BD. The BD varied highly within the organic layer of the

brackish marsh while it was uniform for the saline marsh. Salinity can increase

sedimentation due to enhanced flocculation.

Salinity for all the subhorizons was significantly different between marsh

types indicating that the sites have well represented different levels of salinity.

Salinity profiles differed between landscape positions when the inland area was

land-locked. This difference may be pronounced in more saline areas.

Evapotranspiration and lack of lateral flow (hydrological barriers) may change

salinity profiles going inland from the streamside. When small streams flow

through the marsh, the hydrological barriers are broken and salinity differences

disappear. Biogeochemical differences may result due to this salinity variation

within the landscape. Therefore, understanding the spatial variability of salinity

is important when monitoring the marsh status within an area.

Salinity influenced pyrite accumulation; however, its influence varied

within the profile. Pyrite content within the Oa1 horizon was significantly

different between landscape positions. There was no evidence of differences

between marsh types because the balance between pyrite formation and

oxidation mainly controls pyrite content within the 0a1 horizon. Pyrite content

168


within the surface organic layers for the inland position was lower compared to

the streamside, especially when the inland is landlocked. The mechanism of

oxygen transport plant to the roots combined with leakage of oxygen into

rhizosphere may be insufficient to oxidize pyrite. The soil aeration due to plant

water uptake can create prolonged oxidized status within the surface soils or

frequent fluctuation of redox status. It is not clear which mechanism is more

significant in pyrite oxidation. For subsurface organic horizons, the landscape

position effect on pyrite accumulation dramatically diminished but the marsh

type effect was prominent. This can be a long-term salinity effect on pyrite

accumulation within the marshes. The higher the marsh salinity, the greater the

pyrite accumulation within the profile. The highest pyrite accumulation was

observed within the thin layer of dark organic material that occurred

immediately above the mineral layer.

The mechanism of formation of the highly pyritic soil layer is not clear.

The most possible scenario is that this soil layer was formed from lake bottom

sediments under high saline condition with enriched supply of labile C and Fe,

later overlain by the organic accretions forming the surface layers. The other

hypotheses is that the pyrite preferably forms top of the mineral layer because

of high Fe availability. Otherwise the dissolved organic compounds can move

through pore water with reduced S and Fe to the bottom of organic layer and

then forms pyrite. Pyrite may form within the organic layer where more labile

carbon is present, and eventually moves to the bottom and accumulates at the

top of the mineral layer where the vertical movement is confined. The

169


hypothesis on movement of organic material can be proven if the underlain

material has a younger age than the overlying organic materials, this is the

test for Humilluvic material within organic soils (Soil Survey Staff, 1996).

Another possibility is a moving delta that grew for a certain time period resulted

in pyrite accumulation within the root zone; later, the area would have been

isolated from the sea resulting in less pyrite accumulation.

Non-pyritic Fe contents are significantly different between the landscape

positions within the more saline marshes (i.e., saline and brackish marshes).

This suggests that the inland areas may be suffering from either lack of

sedimentation if landlocked or subsurface erosion otherwise. Downward

movement through the profile can be another reason, which supports one of

the hypotheses of pyrite concentration within the bottom organic subhorizon.

The soil profiles at ISS and ML had an Oa3 horizon, which contained more

pyrite and underlain the Oa2 horizon with lower non-pyritic Fe and pyrite

contents. This supports the vertical downward movement of Fe compounds

within the profile, as well as, the isolation of an actively growing deltaic marsh.

These results suggest that studying different subhorizons within the profile

(“pedological approach”) can reveal more information than studying only the

surface layer in biogeochemical studies. The intermediate marsh contained

significantly lower non-pyritic Fe within surface organic layers compared to

other marshes. This may be related to the status of the marsh because the

intermediate marsh at this location is extensively broken-up (less than 50%

170


vegetation cover). The non-pyritic Fe content can be a suitable indicator to

assess marsh status. When it is land-locked, the inland position can retain

more non-pyritic Fe within subsurface organic layers due to limited pyritization

(i.e., saline marsh). The non-pyritic Fe can be flushed away when no

hydrological barriers occur between landscape positions.

Non-pyritic S content within the organic sub horizons differs significantly

between marshes due to the salinity changes. As salinity increases from

brackish to saline marsh, the content of non-pyritic S increases. Less saline

marshes are S limited. Unlike the 0a1 horizon, landscape position showed a

significant difference for non-pyritic S within the Oa2 horizon, especially for the

brackish marsh. The BIL contained more non-pyritic S within the profile than

the profile at the BSS. This suggests that when the inland is not land-locked,

non-pyritic S accumulate within the profile. Subsurface organic horizons within

the profiles of less saline areas indicate very limited 8 supply.

Soil pH differences are more pronounced between marshes than the

landscape positions. The pH did not vary significantly within the organic layer.

The mineral layer always has higher pH compared to the organic subhorizons.

The pH of the 0a1 horizon differed significantly between less saline marshes

and more saline marshes. Landscape difference was found only at the

intermediate marsh where marsh was extensively degrading. The ISS soil

profile was more acid than the IIL. The pH of the mineral layers of less saline

marshes was significantly lower compared to the saline marshes.

171


The pyrite and non-pyritic Fe and S data for the soil profiles can be

utilized in making decisions on marsh health and marsh management

strategies. The conditions can be spatially variable within the marsh, therefore,

they need to be evaluated separately. For example the condition of the soil

profiles at the BIL can be considered. The BIL and BSS did not differ in their

salinity, and evidently the BIL was not landlocked. The BIL indicated

accumulation of non-pyritic S within the profile. Low pyrite accumulation may

be due to low non-pyritic Fe. The marsh looked vulnerable and had small

hotspots. This may be due to sulfide toxicity. Improvement of sediment supply

may be one scenario to manage this marsh, which will assure Fe supply.

Pyrite framboids were detected within the saline marsh sediments using

SEM. The framboids were about 6 to 8 pm in diameter and were found in close

association with organic material. The sediments at both saline and brackish

marshes contained mixed silicate mineralogy with smectite, illite, and kaolinite

clays. Pyrite was evident within the sediments at both saline and brackish

marsh types; the coarser fraction ( < 2 0 pm) contains more pyrite.

There is a considerable amount of reduced S compounds accumulated

within these soil profiles, which occurs as pyritic and non-pyritic S. These

organic soils should be reclassified to indicate sulfidic materials within their

profiles. The technique used to estimate pyrite and non-pyritic S can be

adopted as a method to characterize sulfidic material within these soils.

172


4.6 ReferencesBaas Becking, L. G. M., and D. Moore. 1961. Biogenic sulfides. Econ. Geol.

56:227-259.

Begheijn, L. Th., N. van Breemen, and E. J. Velthorst. 1978. Analysis of sulfur compounds in acid sulfate soils and other recent marine soils. Commun. Soil Sci. Plant Anal. 9(9):873-882.

Berner, R. A. 1964. Distribution and diagenesis of sulfur in some sediments from the Gulf of California. Marine Geol. 1:117-140.

Bohn, H. L., F. Young, and H. Chenghe. 1989. Hydrogen sulfide sorption by soils. Soil Sci. Soc. Am. J. 53:1914-1917.

Boyd, C. E. 1995. Bottom soils, sediment and pond aquaculture. Chapman and Hall, New York. 10-66pp.

Bradley, P. M., and J. T. Morris. 1991. Relative importance of ion exclusion, secretion and accumulation in Spartina alterniflora Loisel.. J. Expt. Bot. 42:1525-1532.

Broome, S. W., I. A. Mendelssohn, and K. L. McKee. 1995. Relative growth of Spartina pa tens (Ait.) Muhl. and Scirpus o iney i Gray occurring in a mixed stand as affected by salinity and flooding depth. Wetlands 15:20-30.

Carlson, Jr., P. R., and J. Forrest. 1982. Uptake of dissolved sulfide bySpartina altern iflora: Evidence from natural sulfur isotope abundance ratios. Science 216:633-635.

Chmura, G. L., R. Costanza, and E. C. Kosters. 1992. Modelling coastal marsh stability in response to sea level rise: a case study in coastal Louisiana, USA, Ecological Modelling, 64:47-64.

Connell, W . E., and W. H. Patrick, Jr. 1969. Reduction of sulfate to sulfide in waterlogged soil. Soil Sci. Soc. Am. Proc. 33:711-715.

Coultas, C. L. 1997. Soils. In C. L. Coultas and Y. Hsieh (Eds.) Ecology and Management of tidal marshes: A model from the Gulf of Mexico. St.Lucie Press, Delray Beach, Florida. 53-75pp.

Crozier, C. R., I. Devai, and R. D. DeLaune. 1995. Methane and reduced sulfur gas production by fresh and dried wetland soils. Soil Sci. Soc. Am. J. 59:277-284.

173


Dacey, J. W. H., and B. L. Howes. 1984. W ater uptake by roots controls water table movement and sediment oxidation in short Spartina marsh.Science 224: 487-489.

Day, J. H., P. J. Rennie, W . Stanek, and G. P. Raymond. 1979. Peat testing manual. Technical Memorandum No. 125, Associate Comm, on Geotechnical Res., National Research Council, Ottawa, Canada.

DeLaune, R. D., C. J. Smith., and W. H. Patrick. 1983. Relationship of marsh elevation, redox potential, and sulfide to Spartina alterniflora productivity. Soil Sci. Soc. Am. J. 47:930-935.

DeLaune, R. D., J. A. Nyman, and W. H. Patrick, Jr. 1994. Peat collapse, ponding and wetland loss in a rapidly submerging coastal marsh. J. Coastal Res. 10:1021-1030.

DeLaune, R. D., S. R. Pezeshki, and W. H. Patrick Jr. 1993. Response of coastal vegetation to flooding and salinity: A case study in the rapidly subsiding mississippi river deltaic plain, USA. In M. B. Jackson and C. R. Black (Eds.) Interacting stresses on plants in a changing climate,(NATO ASI Ser. Vol. 116), 211-229pp.

Dent, D. 1993. Bottom-up and top-down development of acid sulphate soils. Catena 20:419-425.

Fanning, D. S. 1993. Salinity problems in acid sulfate coastal soils. In H. Lieth and A. AI Masoom (Eds.) Towards the rational use of high salinity tolerant plants. Vol.1:491-500.

Fanning, D. S., and M. C. B. Fanning. 1989. Soil morphology, genesis, and classification, John Wiley & Sons. New York, New York. 69-80p.

Feijtel, T. C., R. D. Delaune, and W. H. Patrick, JR. 1988. Seasonal pore water dynamics in marshes of Barataria Basin, Louisiana, Soil Sci. Soc. Am. J. 52:59-67.

Freney, J. R., G. E. Melville, and C. H. Williams. 1970. The determination of carbon bonded sulfur in soil. Soil Sci., 109:310-318.

Gaillard, J., H. Pauwels., and G. Michard. 1989. Chemical diagenesis in coastal marine sediments. Oceanologica Acta 12:175-187.

Giblin, A. E., and R. W. Howarth. 1984. Poreater evidence for a synamic sedimentary iron cycle in salt marshes. Limnol. Oceanogr. 29:47-63.

174


Goldhaber, M. B., and I. R. Kaplan. 1982. Controls and consequences ofsulfate reduction rates in recent marine sediments, In J. A. Kittrick, D. S. Fanning, and L. R. Hossner (Eds.) Acid sulfate weathering. Soil Sci. Soc. Am.Spec. Pub. No. 10, Soil Sci. Soc. Am., Madison, Wisconsin. 19-36p.

Griffin, T. M., and M. 0 . Rabenhorst. 1989. Processes and rates ofpedogenesis in some Maryland tidal marsh soils. Soil Sci. Soc. Am. J. 52:59-67.

Haering, K. 0 ., M. 0 . Rabenhorst, and D. S. Fanning. 1989. Sulfur spéciation in some Chesapeake Bay tidal marsh soils. Soil Sci. Soc. Am. J. 53:500- 505.

Howarth R. W ., and S. Merkel. 1984. Pyrite formation and the measurement of sulfate reduction in salt marsh sediments. Limnol. Oceanogr., 29:598- 608.

Kosters, E. C., and A. Bailey. 1983. Characteristics of peat deposits in the Mississippi river delta plain. Transactions Gulf Coast Association of Geological Societies Vol. XXX III:311-325.

Kostka, J. E., and G. W. Luther III. 1994. Partitioning and spéciation of solid phase iron in salt marsh sediments. Geochim. Cosmochim. Acta 58:1701-1710.

Krairapanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1991a. Sulfur dynamics in Louisiana coastal freshwater marsh soils. Soil Sci. 151(4):261-273.

Krairapanond, N., R. D. DeLaune, and W . H. Patrick, JR. 1991b. Seasonal distribution of sulfur fractions in Louisiana salt marsh soils. Estuaries 14:17-28.

Krairapanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1992. Distribution of organic and reduced sulfur forms in marsh soils of coastal Louisiana,Org. Geochem. 18:489-500.


Leventhal, J., and C. Taylor. 1990. Comparison of methods to determinedegree of pyritization. Geochimica et Cosmochimica Acta 54:2621-2625.

175


Lin, C., and M. D. Melville. 1994. Acid sulphate soil-landscape relationships in the Pearl River Delta, southern China, Catena, 22:105-120.

Lindsay, W. L. 1979. Chemical equilibria in soils. John Wiley & Sons, New York. 281-298p.

Lord III, C. J. 1982. A selective and precise method for pyrite determination in sedimentary materials. J. Sediment. Petrol., 52:664-666.

Lord III, C. J., and T. M. Church. 1983. The geochemistry of salt marshes:Sedimentary ion diffusion, sulfate reduction, and pyritization. Geochim. Cosmochim. Acta 47:1381-1391.

Moore, D. M., and R. C. Reynolds, Jr. 1989. X-ray diffraction and theidentification and analysis of clay minerals. Oxford University Press,New York.

Munch, J. C., and J. C. G. Ottow. 1980. Preferential reduction of amorphous to crystalline iron oxides by bacterial activity. Soil Sci. 129:15-21.

Nyman, J. A., J. C. Callaway, and R. D. DeLaune. 1993. Case study of arapidly submerging coastal environment: relationships among vertical accretion, carbon cycling, and marsh loss in terrebonne Basin,Louisiana. In P. Bruun (Ed.) Proceedings of the Hilton Head Island South Carolina U.S.A. International Coastal Symposium. 1993, vol. 2:452-457.


Nyman, J. A., M. Carlose, R. D. DeLaune, and W. H. Patrick, Jr. 1994. Are landscape patterns related to marsh loss processes?. p337-348. In Coastal zone’93, proceedings of 8 th symposium on Coastal and Ocean Management held July 19-23, 1993, New Orleans, Louisiana.

Nyman, J. A., R. H. Chabreck, R. D. DeLaune, and W. H. Patrick, Jr. 1993. Submergence, Salt-water intrusion, and managed Gulf Coast marshes. Coastal Zone’93, p .1690-1704.

Nyman, J. A., and R. D. DeLaune. 1991. Mineral and organic matteraccumulation rates in deltaic coastal marshes and their importance to landscape stability, GCSSEPM Foundation Twelfth Annual Research Conference, Program and Abstracts, December 5, 1991.

176


Nyman, J. A ., R. D. DeLaune, H. H. Roberts, and W . H. Patrick, Jr. 1993. Relationship between vegetation and soil formation in a rapidly submerging coastal marsh. Mar.Ecol. Prog. Ser. 96:269-279.

Ogata, G., and C. A. Bower. 1965. Significance of biological sulfate reduction in soil salinity. Soil Sci. Soc. Am. Proc. 29:23-25.

Patrick, W. H. Jr., and A. Jugsujinda. 1992. Sequential reduction and oxidation of inorganic Nitrogen, Manganese, and Iron in flooded soil. Soil Sci. Soc. Am. J. 56:1071-1073.

Ponnamperuma, F. N. 1972. The chemistry of submerged soils. Adv. Agron. 24:29-96 .

Pons, L. J. 1973. Outline of the genesis, characteristics, classification andimprovement of acid sulphate soils. In H. Dost (Ed.) Proceedings of first International symposium on Acid sulphate soils. International Land Reclamation Institute, Publication No. 18, Vol I. pp 3-27, Wageningen, The Netherlands.

Pons, L. J., N. van Breemen, and P. M. Driesses. 1982. Physiography ofcoastal sediments and development of potential soil acidity, pp. 1-18. In J. A. Kittrick, D. S. Fanning, and L. R. Hossner (Eds.) Acid sulfate weathering. Soil Sci. Soc. Am. Spec. Pub. No. 10, Soil Sci. Soc. Am., Madison, Wisconsin.

Rabenhorst, M. C., and B. R. James. 1992. Iron sulfidization in tidal marshsoils. In H. C. W. Skinner and R. W. Fitzpatrick (Eds.) Biomineralization processes. Iron, Manganese: modern and ancient environments. Catena supplement 21 :203-217.

Raiswell, R., and D. E. Canfield, and R. A. Berner. 1994. A comparison of iron extraction for determination of degree of pyritisation and the recognition of iron-limited pyrite formation. Chemical Geology 111:101-110.

Raiswell, R., K. Whaler, S. Dean, M. L. Coleman, and D. E. G. Briggs. 1993. A simple three-dimentional model of diffusion-with-precipitation applied to localized pyrite formation in framboids, fossils and detrital iron minerals. Marine Geol., 113:89-100.

Raybould, J. G. 1973. Framboidal pyrite associated with lead-zinc mineralisation in mid-Wales. Lithos 6:175-182.

Rickard, D. T. 1970. The origin of framboids. Lithos 3:269-293.

177


Rickard, D. T. 1973. Sedimentary iron sulfide formation. pp28-65. In H. Dost (Ed.) Vol I. Acid sulphate soils. International Land Reclamation Institute, Publication No. 18, Wageningen, The Netherlands.

Rickard, D. T. 1975. Kinetics and mechanism of pyrite formation at low temperatures. Amer. J. Sci., 275:636-652.

Rozema, J., E. Luppes, and R. Broekman. 1985. Differential response of salt marsh species to variation of iron and manganese. Vegetatio 62:293- 301.

SAS Institute Inc., 1994. SAS/STAT User’s guide. Version 6 , fourth edition. Volume 2, Cary, NO: SAS Institute Inc., 846pp.

Schneider, J. W., and K. Schneider. 1990. Indirect method for thedetermination of pyrite in clays and shales after selective extraction with acid solutions. Ceramic Bulletin 69(1): 107-109.

Schulze, 1981. Identification of soil iron oxide minerals by differential X-ray diffraction. Soil Sci. Soc. Am. J. 45:437-440.

Sheppard, M. J., C. Tarnocai, and D. H. Thibault. 1993. Sampling organicsoils. In M. R. Carter (Ed.) Soil sampling and methods of analysis. Lewis Publications. 423-439pp.

Soil Conservation Service (SCS). 1984. Soil survey of Lafourche parish,Louisiana. United States Department of Agriculture, Washington, DC.

Soil Conservation Service (SCS). 1989. East Central Barataria cooperative river basin study, Jefferson, Orleans, Plaquemines and St. Charles Parishes, Louisiana. Rev. May 1989. United States. Soil Conservation Service. National Cartographic Center. Ft. Worth, Texas : USDA-SCS-National Cartographic Center ; [Louisiana State Conservationist, distributor].

Soil Survey Staff. 1996. Keys to soil taxonomy. Seventh Edition. United States Department of Agriculture, Washington, District of Columbia.

Thamdrup, B., K. Finster, H. Fossing, J. W. Hansen, and B. K. Jorgensen.1994. Thiosulfate and sulfite distributions in porewaters of marine sediments related to manganese, iron, and sulfur geochemistry. Geochim. Cosmochim. Acta 58:67-73.

Turner, R. E. 1990. Landscape development and coastal wetland losses in the northern Gulf of Mexico. Amer. Zool., 30:89-105.

178


Welton, J. E. 1984. SEM petrology atlas. The Amer. Assoc. Petrol. Geologists, Tulas, Oklahoma.

Whitcomb J. H., R. D. DeLaune, and W. H. Patrick, Jr. 1989. Chemicaloxidation of sulfide to elemental sulfur; its possible role in marsh energy flow. Marine Chemistry 26:205-214.

Willett, I. R., and T. A. Beech. 1987. Determination of organic carbon in pyritic and acid sulfate soils. Commun, in Soil Sci. Plant Anal. 18(7):715-724.

179


CHAPTER 5: SPATIAL VARIABILITY OF COASTAL ORGANIC SOIL CHARACTERISTICS WITHIN THE BARATARIA BAY BASIN, LOUISIANA


Information on spatial variability of coastal organic soils within Barataria

Bay Basin is limited. Understanding the spatial variability of soil characteristics

helps to find the influence of dominant processes on marsh formation. Spatial

autocorrelation was estimated for several coastal organic soil characteristics,

such as, thickness, pH, and organic/mineral ratio (OM R) for soil subhorizons,

depth to mineral layer (DML), and soil subgroups for saline and brackish marsh

types. Each variable was interpolated for a one square mile area to study the

spatial variability.

With the exception of the thickness of the Ga1 horizon within the

brackish marsh, organic horizon thickness showed high spatial autocorrelation.

Variation of horizon thickness was greater at depths within the profile. Contour

plots indicated more accretions at edges of the water bodies. This is mainly

due to higher accretions at streamside than inland within the landscape.

Because of this variation in accretion the Ga1 horizons at streamside may be

younger than the inland G a l horizon. This can be a main source of variability

for previous accretion differences, in subsurface horizons.

Spatial variation of the organic layer thickness was associated with

variation of the DML. For the saline marsh, the DML was shallower near water

bodies. The soil subgroups, Terric and Typic Medisaprists were delineated

based on the organic layer thickness. Typic Medisaprists occurred mostly

180


toward inland areas and away from open water at the saline marsh. The DML

was shallower in the northeastern half of the saline marsh. Typic Medisaprists

occurred in the northeastern part of the brackish marsh site, which had thick

organic layers due to presence of thick Oa2 and O a3 horizons.

High spatial autocorrelation was found for pH of organic horizons for

saline marsh except for the Cg horizon. Lower pH was associated with

degrading marsh patches toward the S W and NE corners of the saline marsh.

A smooth pH gradient from NW to SB direction was observed for the Oa3

horizon at the saline marsh. High spatial variability was found for pH for Oa1

horizon within the brackish marsh site, with lower pH values along the N W to

SB diagonal across the site.

The OM R data ranged widely for the brackish marsh compared to saline

marsh for all organic subhorizons. Higher OM R associated with the northern

part and lower O M R at the southern part of the site indicated sedimentation

variations. The high OM R at the NE and NW corners of the saline marsh can

be attributed to marsh degradation. The OM R data were highly spatially

variable at the brackish marsh site. Higher OMR found inland may be a result

of low sedimentation (inorganic accretions).

Spatial variability can be attributed to long-term influence of processes,

such as sedimentation, organic accretions, consolidation of sediments, and

localized damage by hurricane activity. Understanding of spatial variation of

marsh soil characteristics will help to identify the prominent processes within

the marshes and to plan marsh management strategies.

181


High net primary production of coastal marsh in terms of the above- and

below-ground biomass is the main contribution to the fixed carbon pool within

the marsh soil environment. The organic soil profile is comprised of mainly two

layers; the organic layer and the mineral layer. The organic layer can be

divided into different sub horizons based on the color and the composition. The

organic matter accumulation (organic accretion) and sedimentation (inorganic

accretion) are the main processes that contribute to the organic soil formation

(Hatton et al., 1983; Chmura et al., 1992;Nyman et al., 1993). Thickness of the

organic subhorizons varies due to the variations in the organic and the

inorganic accretions and disturbances to the marsh, which reflects the history of

marsh soil formation. Organic accretions vary due to the changes in the

productivity of the vegetation and inorganic accretions vary depending on the

changes in the sedimentation rate within the marsh.

Organic layers degrade in eroding marshes. Organic accretion results in

thick organic layers that help to maintain healthy marshes (DeLaune et al.,

1983). Accretion rates may be variable and the marsh landscape may be

disturbed by hurricane activity, presence of channels, open water areas, and

m an-made changes (Salinas et al., 1986; D e laune et al., 1989). Thin organic

layers may be characteristic of erodible marshes that convert into open water.

Thickness of the surface organic layer is a significant soil morphological feature

that may indicate the current status of the marsh. Depth to mineral layer is

used as soil subgroup differentiae. Understanding spatial variability of organic

layer thickness will enable one to delineate soil subgroups. Critical marsh

182


areas may be identified based on soil variability and marsh management

strategies may be planned.

Organic/mineral ratio of the organic soil layers is a result of both organic

accretion and sediment deposition. Sediment deposition rates can be different

within the marsh depending upon the sediment supply. High OMR ratio may be

either due to lack of sediments or high organic accretions. Organic accretion

results in thick organic layers that help to overcome submergence. However,

lack of sediments does not promote organic accretions. Therefore,

organic/mineral ratio can be a suitable indicator to detect the status of the

marsh. It may be directly related to the marsh status; degrading marshes may

have high organic/mineral ratio. Soil pH indicates the biogeochemical status of

the marsh. The spatial variability of pH within the landscape can be associated

with prominent biogeochemical processes within the area.

The thickness of the organic subhorizons is a direct estimate of fixed

carbon. The organic/mineral ratio and pH reflect the compositional and

biogeochemical status of the marsh. Understanding of the spatial variability of

these characteristics more accurately estimates the carbon pool within the

marsh.

Previous studies carried out within the marsh excluded spatial variability

by sampling surface soils at points within close proximity to each other (Feijtel

et al., 1988; Krairpanond et. al., 1992). The point estimates represented the

entire marsh type while not much attention was paid to understanding the

spatial variability.

183


Practical limitations for sampling restricted detailed sampling within the

landscape. Because of the practical difficulties of a well-planned sampling

scheme within the marsh, understanding the landscape model is helpful to

understanding possible soil variations. Spatial variability data will help plan soil

sampling schemes to be more representative and minimize the number of

sampling points as appropriate. Remotely sensed data are often used in

coastal marsh studies and the spatial variability information also helps a

researcher decide the appropriate pixel size of the remotely sensed data for a

particular study.

This study was planned to assess the spatial variability of organic soil

characteristics at field-scale for brackish and saline marsh types within the

Barataria Bay Basin. The spatial variability of soil morphological data such as

sub horizon thickness, depth to mineral layer, pH and organic/mineral ratio for

different horizons will be studied. Semivariograms will be generated for the soil

variables and associated landscape patterns of these soil characteristics will be

identified.


5.2.1 Barataria Bay Coastal Marsh

Louisiana's coastal wetlands comprise 41% of the U. S. total coastal

wetlands and are significant as a state, national, and international natural

resource. These wetlands directly support fisheries, wildlife, industry and

recreational needs (Turner, 1990). The coastal wetlands of Louisiana covers

about 3.7 million ha of land and water. This area is comprised of 58% open

184


water (2.146 m. ha), 5.6% urban and agricultural lands (0.207 m. ha), 32%

wetlands (1.184 m. ha), and 2 .1% dredged canals and spoil banks (0.078 m.

ha). These wetlands are a combination of six marsh types; salt marsh (15%,

0.1776 m. ha), brackish marsh (47 %, 0.5565 m. ha), intermediate marsh (23% ,

0.2723 m. ha), fresh marsh (12% , 0.1421 m. ha), swamp (<1%), and mangrove

(<1% ). Barataria Bay marshes are a riverine deltaic marsh that extends for an

area of 628,000 ha. The area is divided into three sub basins; upper

freshwater lake (Lac Des Allemands), middle brackish lakes, and lower saline

bays (Barataria Bay and Caminada Bay), lakes and marshes (Madden et al.,

1988).

Marsh loss to open water is a severe problem in the marshes of the

northern Gulf of Mexico. The average annual rate of marsh loss was about

0.86 % from 1955 to 1978 (Turner, 1990). Boesch et al. (1983) reported a loss

of about 49 square miles per year. Inability of the marsh to counter the

subsidence has been reported as the main cause of marsh loss (Turner, 1990;

DeLaune et al., 1994). Marsh is lost to open water when subsidence is greater

than the accretion rates. Because of high Relative Sea Level Rise (RSLR),

deltaic wetlands are affected by an acceleration of eustatic sea-level-rise (Day

et al., 1995). The combination of RSLR and local submergence has been

considered as the apparent sea level rise. Apparent sea level rise is reported

to be around 0.24 cm yr^ for these marshes (Chmura et al., 1992).

Turner (1990) indicated that sediment loading declined drastically in the

mid-1950s after dam and reservoir construction on major tributaries of the

185


Mississippi River, causing an estimated 50% decrease in sediment input to the

marsh system. Sedimentation (inorganic accretion) and organic matter

accumulation control accretion. Vertical accretion is not simply the result of

sediment supply but also of the interaction of plants and the prevailing

hydrologie regime. Salinity intrusion and other biological factors are known to

be responsible for the above and below-ground production of the marsh

vegetation. W ater logging may be more influential than the sediment

deprivation in relation to marsh loss (Turner, 1990). Water logging can make

the vegetation more vulnerable to dieback due to stress.

One method to determine accretion is the use of the ^ ’'Cs isotope. The

^^ Cs maxima are assumed to result from atmospheric deposition of ® Cs that

marks the 1963 surfaces, rather than that from erosional depositions. The ^ ^Cs

profiles in Louisiana marshes represent continuous atmospheric deposition

rather than discontinuous sediment deposition (Nyman et al., 1993). The ® Cs

dating indicated that vertical accretion (0.98 cm y r ’) was extremely rapid in the

Terrebonne Basin marsh relative to other marsh areas. But this accretion rate

was insufficient to counter submergence (1.38 cm y r ’) (Nyman et al., 1993).

Compaction or consolidation of the sediments will govern the effective accretion

rates (Chmura et al., 1992). Erosion below the root zone is another mechanism

for marsh loss (Nyman et al., 1994), which reduces marsh elevation

considerably. The unconsolidated soil below the living root zone at the marsh-

water interface in the broken marsh areas is eroded and the detached marsh

come to rest on the pond bottom (Nyman et al., 1994).

186


Catastrophic events influence Louisiana marshes frequently. Hurricanes

bring salt water and sediments inland and also may erode lands due to physical

action of flooding (Childers et al., 1990; Turner, 1990). Hurricane sediments

are not stable since they can be washed away by the consequent flooding

(Turner, 1990). Hurricanes influence the sedimentation variability within the

marsh. Nyman et al. (1995) reported a wide range of deposition thickness (0 to

9 cm) due to hurricane Andrew in Louisiana marshes along the hurricane track.

This suggests that marsh elevation can be increased due to deposits and at

the same time marsh loss may occur as a result of a hurricane.

Marshes are delineated into saline, brackish, intermediate, and fresh

water marshes based on salinity (Feijtel et al., 1988; Soil Conservation Service,

1989). The salinity map indicates isohaline lines from 0.5 to 20.0 ppt covering

the area in east and central Barataria Bay. Salinity ranges between 5 and 10

ppt within brackish marsh while saline marsh has salinity over 10 ppt (Soil

Conservation Service, 1989). Previous work indicated variations sulfur

dynamics due to differences in biogeochemical environment within different

marsh types (Krairpanond et al., 1992a, 1992b). Griffin and Rabenhorst

(1989), Haering et al. (1989) and Lin and Melville (1994) reports variability in

biogeochemical processes following a landscape pattern within marshes

elsewhere.

Micro-scale landscape variation within the marsh causes variation in

vegetation (DeLaune et al., 1983). This suggests the existence of spatial

variability in the rate and amount of carbon accumulation within the marsh at

187


small scale. Consequently this can be related to a landscape pattern of marsh

accretion rates and soil morphological variability. Marsh loss is attributed to

plant stress due to salinity and prolonged inundation. Since marsh degradation

affects the soil morphology, thickness of an organic horizon is a suitable

indicator to understand the marsh status. Previous studies were limited to point

sampling within this marsh to exclude spatial variability (DeLaune et al., 1983).

Limited studies attempted to understand the spatial variability of marsh

characteristics. Variations in soil morphology would help to understand the

prominent functions within the marsh and to identify hot spots of degrading

marshes.

5.2.2 Soil Morphology and Classification

An idealized soil profile for the marsh area is presented below (Figure

5.1).

Figure 5.1. An idealized coastal organic soil profile with subhorizons.

188


These marshes have a variable organic layer (Oa horizon) over a sandy or

clayey alluvium (Cg horizon). The surface organic layer can be separated into

different sub horizons based on their color differences as Oa1, Oa2, Oa3, and

O a4 horizons. Organic soils within the Barataria Bay Basin are mapped as

associations of soil series in the soil survey conducted in 1981 (Soil

Conservation Service, 1984). The associated series within a particular salinity

regime are delineated based on the depth to mineral layer at subgroup level

(i.e., Terric Medisaprists and Typic Medisaprists). Timbalier-Bellpass

association (TB) is dominant within the saline marsh. According to Soil

Taxonomy, the Timbalier series is euic, thermic, Typic Medisaprists while the

Bellpass series is clayey, montmorillinitic, euic, thermic, Terric Medisaprists.

The Lafitte-Clovelly association (LA) dominates the brackish marsh. The Lafitte

soil series is an euic, thermic, Typic Medisaprists while the Clovelly soil series

is a clayey, montmorillinitic, euic, thermic, Terric Medisaprists. Terric

Medisaprists are the Medisaprists that have a mineral layer 30 cm or more thick

that has its upper boundary within the control section (130 cm) below the

surface tier (60 cm). Other Medisaprists that have the upper boundary of the

mineral layer below control section (130 cm) are Typic Medisaprists. The main

difference between these soil series is the depth to the mineral layer.

5.2.3 Spatial Variability of Marsh Soil Characteristics

Marsh elevation changes from highest at the streamside to lowest in the

inland area. This change is associated with variation in soil and plant

characteristics along this marsh landscape (DeLaune et al., 1983). Marsh loss

189


is associated with soil water logging and salt-water intrusion into non-saline

marshes (Nyman et al., 1993; DeLaune et al., 1994). Total plant mortality

occurred within one year, and hummock elevation decreased almost 15 cm

within two years in a Louisiana marsh (DeLaune et al., 1994). Increased

sediment sources, maintenance of natural hydrologie regime and sediment

distribution, and lower rates of sea level rise contribute to wetland growth.

Considering these factors, different horizonation can be expected within the

organic soil profile. This may lead to spatial variability of relative marsh

elevation within this area. Spatial variability of soils is the integrated outcome

of these various activities happening within the marsh. Identification of spatial

variation of marsh characteristics may enable one to identify the dominant

processes controlling the marsh loss or growth within the area.

5.2.4 Spatial Interpolation Methods

Spatial variability of soils within landscape was identified at the early

stages of soil science when pedologists identified different soils associated with

the landscape position (i.e., Glazovskaya, 1968). In recent soil science

research, spatial variability of soil characteristics at field-scale has been the

interest of many scientists, and the understanding of its impact on soil and crop

management practices has improved.

Application of mapping science techniques such as geostatistics and

Geographic Information Systems (GIS) at field-scale is helpful for the soil

scientist to identify landscape relations of soils. The underlying pedological

processes have been identified within the landscape from spatial patterns of

190


soil properties (Hanna et a!., 1982). Reese and Moorhead (1996) identified

spatial characteristics of soil properties along an elevation gradient. McDaniel

et al. (1992) presented pedochemical indicators to mark field-scale through flow

water movement within the landscape.

Use of data interpolation to identify the surface model to predict the

values for points that have not been measured, is imperative. Lam (1983)

presented a comprehensive review on spatial interpolation methods. With the

rapid expansion of GIS, spatial data analysis and spatial interpolations require

special attention. The following section presents several spatial interpolation

techniques and concerns in soil science.

Performance of different spatial interpolation methods for soil

parameters were presented by W ebster and Oliver (1989), Knotters et al.

(1995), Barry and Ver Hoef (1996), Brus et al. (1996), and G o tw ayeta l. (1996).

The relation of an appropriate interpolation model depends largely on the type

of data, the degree of accuracy desired, and the computational effort afforded.

Every model is based on a hypothesis about the surface, which may or may not

be true (Lam, 1983). Spatial interpolations are either point interpolations (i.e.,

temperature, elevation) used for isometric maps, or areal interpolations (i.e.,

population density) used for isopleth maps. For both interpolations, they can be

either exact or approximate methods depending on whether the method

"preserves or not” the original sampling point values. Depending on the extent

of the data points involved, the point interpolation methods are classified as

global or piecewise methods. The global methods utilize all sample points to

191


determine a value at a new point, while only the nearby points are used by the

piecewise methods.

The interpolation methods also can be categorized as statistical or

analytical. Kriging is considered a statistical technique while other methods are

considered as analytical techniques. Lam (1983) preferred to classify point

interpolation methods as either exact or approximate methods because

preservation of original sampling point values on the inferred surface seems

fundamental in analyzing accuracy and in examining the nature of interpolation

methods. Interpolating polynomials, most distance-weighting methods, kriging,

spline interpolation and finite difference methods are exact methods. The

group of approximate methods includes power-series trend models, distance-

weighted least squares and least square fitting with spline (Lam, 1983).

The principle of distance-weighting methods is to assign more weight to

nearby points than to distant points. The usual expression is

f(x,y) = [ Z w(di)Zj] / [ Z w (d |) ] [5 .1 ]i= 1 i= 1

W here w(d) is the weighting function, Z| is the data value at point I, and d | is

the distance from point i to (x,y). The disadvantages of the weighting methods

are, one is the ambiguity introduced by the weighting function and second is

that the method is affected by uneven distributions. Weighting functions are

essentially smoothing procedures. It is considered as a limitation because the

contours should be able to predict the important features such as maximum

192


and minimum when they are not included within the data set. However, the

simplicity of the principle, the speed of calculation, the ease of programming,

and reasonable results for many types of data have led to wide application of

the weighting methods (Lam, 1983).

Kriging, perhaps the most distinctive of interpolation methods, has

become a major tool in the field of geostatistics. Kriging treats the statistical

surface of interpolated data as a regionalized variable with continuity. More

information on kriging can be found in Isaaks and Srivastava (1989) and

Goovaerts (1997). Two systems of kriging procedures, simple kriging and

universal kriging, can be distinguished based on the assumptions about the

regionalized variables. Mathematically, a semivariogram (y) is defined by:

N

y =%N E [z (X |+ d ) - z (X j) ]^ [5.2]

1=1

d = distance between two samples, and Z is the data value at points.

Parameters for a semivariogram are presented in Figure 5.2.

Sill

8CCO'C

iÎCO

Nugget Range

Lag Distance

Figure 5.2. Major features of an ideal semivariogram.

193


A line can be fitted for the semivariance function with lag distance. The

three properties of the model are the nugget, sill, and range (Figure 5.2). The

nugget is the intercept that is theoretically zero for lag distance of zero.

However, it may deviate from zero depending on the sampling and short

distance variability. The nugget is the amount of variability that is not included

into the spatial correlation model. The leveling off of the semivariogram model

forms the sill, but for a linear model, it may not present. Beyond the sill, the

semivariance on the average is constant or independent from lag distance.

The lag distance value where the semivariance reaches the sill is the range.

The range represents the maximum separation distance within which the

sample values are spatially correlated (Rossi et al., 1992; Smith et al., 1993).

When the semivariance reaches the sill, sample locations are separated

enough to be independent. The range of correlation indicates the relationship

between spatial variability and random variability within the site. Random

variability occur between samples that are farther apart than the range and

spatial variability occur between samples closer together within the range

(Flatman and Yfantis, 1996).

Four common shapes of the semivariance models are linear, spherical,

exponential and guassian. Linear semivariogram with a sill may also be used

(Gamma Design Software, 1994). Simple kriging has more restrictive

assumptions but fewer computational problems, whereas the universal kriging

has more general assumptions but difficult calculation (Lam, 1983). Increasing

nugget effect increases the ordinary kriging variance and makes the estimation

194


procedure become more like simple averaging of data. The ratio of the nugget

to the sill is referred to as the relative nugget effect that is usually quoted in

percentages. The change of the range has relatively minor effect on the

ordinary kriging weights (Isaaks and Srivastava, 1989).

Accuracy of the inverse distance methods increased with the power of

distance for data sets with low coefficient of variation (CV) (< 25% ). High

distance powers in inverse distance methods can give very inaccurate

predictions for data sets with C V higher than 25%. Accuracy of kriging is

generally independent from the CV of the data set (Gotway et al., 1996). Brus

et al. (1996) compared the performance of six spatial interpolation methods,

global mean, moving average, nearest neighbor, inverse squared distance,

Laplacian smoothing splines, and ordinary point kriging. Soil properties

included thickness of the A horizon, soil adsorbed phosphate, and water table

depths. Differences between methods were small. However, Brus et al. (1996)

found kriging was more reliable because it estimated all properties well.

Therefore, Kriging can be used as a interpolation method when the variable has

a well structured spatial variability that can be explained by a semivariogram.

5.3. M aterials and M ethods

5.3.1 Site Description

Two sites were selected representing saline and brackish marsh types

within the Barataria Bay (Figure 5.3). Both sites were in Lafourche Parish. The

saline marsh site was located at N 29° 14' 44", W 90° 07' 16" and the dominant

vegetation was S partina a ltem iflo ra . The brackish marsh site was located at

195


f

im

IIoU5

0)O)CDE

CO03C0 c1CO

S'

03EJZ.SGo2_o03B■ac03

S

ECD

£03C

I d£/)03x:

toc0

101c/)COUO

EZJ03

196


N 29° 16' 35", W 90° 08' 10", and was predominantly covered by S partina

pa tens and D istich lis . Data were collected in 1990 for both sites, each site

covering a one-square mile area.

Figure 5.4 presents site maps for saline and brackish marsh types.

Maps only show the dominant open water areas and waterways within the

sites. The saline marsh site has two major open water areas that are

separated by broken marshes at the southern part of the site. Two other open

water areas are in the NE section of the site. The waterway, which enters from

the north, is divided into two branches and flows into opposite directions; one

westward and the other eastward. The brackish marsh site is comprised with

vast areas of broken marshes that occupy the entire NE and NNW sections of

the site. The major waterways that flow through the site have been blocked at

the center so that only the water from the west and the south mixes. Major

open water areas are found within the SW section and the minor waterway in

the east expanded into a open water area within the SE section.

5.3.2 Data Collection

Soil morphological data and pH were collected from saline and brackish

marsh types. Soil profiles were sampled on a grid formed by transects of 200-

m interval established in both directions across the sites (Figure 5.5). Samples

were collected at 400-m interval on each transect along north-south direction.

Data included thickness of organic subhorizons (Figure 5.1), depth to mineral

soil layer and pH for each soil subhorizon. Depth to mineral soil layer was

similar to the total thickness of organic matter because the samples were

197


1600

1400

1200

1000

I 800

IQ 600

400

200

0

1600

1400

1200

1000

I 800

Q 600

A

W

{ 1 W '

w

w„

w

ww

A. Saline marsh

400

200

0

A

w

w0 200 400 600 800 1000 1200 1400 1600

Distance (m)B. Brackish marsh

Figure 5.4. Maps showing dominant open water areas (W ) and waterways within the sites at saline (A) and brackish (B) marsh types.

198


CD■ D

OQ .C

gQ .

■DCD

C/)Wo'3O

8■D

3.3 "CD

CD■DOQ .C

aO3

■DO

CDQ .

■DCD

C /)C /)

1600

CDCD

€o

8

Ib

1200

800

400

,, * * # a\f w w w #

. i 1 1 1 :

; : M ,

. : : n........ "#

I i Î I : 1 1 ^ F W W W

A ' À 1 À 1 :T T •

.. ! 1 1 1 1 .

' i

11 W , W ; W

1 1 1 I 1 ;

# '!

. . 1

„ ! 1 1 ! L ... I 10 400 800 1200 1600

Distance (m), East - West

Figure 5. 5. Sampling plan for both sites, showing the sampling points on the grid.

collected within vegetative areas. The pH was measured with a pH meter

(Orion Research model 701A/digital lONALYZER). Organic and mineral

contents were estimated on weight loss basis after burning the samples at 425

°C. The variables used for each site are given in Table 5.1.

5.3.3 Data Analysis

The first task in all geostatistical investigations is to conduct exploratory

data analysis (EDA) (Rossi et al., 1992). The EDA was conducted using the

SPSS software (SPSS Inc., 1997). Boxplots were used to present the data

sets and to identify outliers and compare the variables between marsh types.

Spatial data analysis was conducted using the GS^ software (Gam m a Design

Software, 1993). Semivariograms were constructed and appropriate models

were fitted for each variable. The data set was not trimmed to obtain better

shaped semivariograms as is common for some studies (Cahn et al., 1994).

Contours were generated using SURFER software (Golden Software Inc.,

1995), interpolating data for a finer grid with 50-m intervals. Kriging was used

for the variables that have a good spatial structure explained by the

semivariogram models, otherwise the inverse distance squared method was

used to interpolate data.

200


Table 5.1. The variables used for spatial analysis for each marsh type.

Variable Name Code^

Thickness of the O a l horizon = 0A 1TH

Thickness of the Oa2 horizon = GA2TH

Thickness of the Oa3 horizon = OA3TH

Depth to mineral soil layer^ = □M L

pH of the O a l horizon = GA1PH

pH of the Ga2 horizon = 0A 2P H

pH of the Ga3 horizon = 0A 3P H

pH of the Cg horizon = CGPH

Weighted average pH for the organic layer = W GPH

Organic/mineral ratio* for the G a l horizon = GA1GMR

Organic/mineral ratio for the Ga2 horizon = 0A 2G M R

Organic/mineral ratio for the Ga3 horizon = 0 A 3 0 M R

Weighted average “omr" for the organic layer = W GM R

^ S o r B will proceed the code to indicate saline (S) or brackish (B) marsh.

^Mostly similar to the thickness of total organic layer

^Organic matter content / mineral matter content

201



5.4.1 Thickness of Organic Subhorizons

5.4.1.1 Exploratory Data Analysis

The number of subhorizons within the profile varied widely within both

sites. Most of the sampling points had Qa1 and G a2 subhorizons, but the Ga3

horizon was absent at some sampling sites. Some sampling points had an Ga4

horizon, mostly at the brackish marsh site. Very few locations had mineral soils

at the surface.

Boxplots for the data on thickness of organic subhorizons are presented

in Figure 5.6. The components of a box plot are 25^, 50"' (median), and 75*"

percentiles and the lower and upper fences. The upper fence equal to 7S'"

centile + 1 . 5 * interquatile range (IQR), and the lower fence equals to 25'"

centile - 1.5*IQR. Data values above or below the fences are designated as

outliers. The SGA1TH data indicated more variation compared to BGA1TH

data. The range (0 and 50 cm), median (25 cm) and mean (26 cm) for G a l

horizon thickness were similar for both sites. Data for SGA1TH and BGA1TH

were normally distributed. Thickness of the Ga2 horizons was higher

compared to that of the G a l horizons at both sites. The SGA2TH ranged

between 0 and 100 cm while the BGA2TH ranged between 0 and 90 cm,

resulting in a higher mean for the saline marsh (48 cm vs. 37 cm). The

BGA2TH data followed a bimodal distribution. A bimodal distribution may

indicate that the variable is comprised of two populations that are under control

of two different processes. The Ga3 horizon thickness data followed bimodal

202


120

Eü

ïü

S 0A 1TH S 0A 2T H S 0A 3T H

B 0A1TH BOA2TH BOA3TH

Subhorizon

Figure 5.6. Boxplots showing the data for organic subhorizon thickness at both sites.

203


distribution at both sites; for SOA3TH the modes were at zero and 35 cm, for

B 0A 3TH the modes were at zero and 50 cm. Both sites had about 50% of

sample points that did not have the Oa3 horizon, which is indicated by the

boxplot for the S 0A 3T H that has its median at zero. In general, thickness

variation is higher for the lower subhorizons within the profile (Figure 5.6).

5.4.1.2 Spatial Data Analysis

Semivariograms were constructed using horizon thickness data from

both sites, and best models were selected based on the regression coefficients.

The model parameters are given in Table 5.2.

Table 5.2. Modal parameters and regression coefficients, for the semivariograms for the thickness of organic subhorizons at both sites.

Variable^ Shape Nugget (cm^) Sill (cm^) Lag (m) R:

S 0A 1TH spherical 14 83 684 0.993

S 0A 2TH spherical 162 451 609 0.995

S 0A 3TH spherical 331 941 1392 0.973

B0A2TH spherical 168 485 548 0.935

B0A3TH spherical 250 1060 1607 0.983

BOA1TH data did not fit a semivariogram model.

The data for B 0A 1TH did not fit a semivariogram due to lack of spatial

variability within the data set. Data for the thickness of other subhorizons at

both sites indicated significant spatial autocorrelation that was explained by

spherical models (Table 5.2). Semivariograms for SOA1TH, SOA2TH, and

204


S 0A 3TH are presented in Figure 5.7. The sill is higher for the SOA2TH and

SOA3TH compared to S 0A 1TH . The range of spatial correlation is higher for

S 0A 3TH . Figure 5.8 presents the semivariograms for B0A1TH, BOA2TH, and

BOA3TH. Range of spatial correlation was smaller for B0A2TH compared to

B0A 3TH . The B0A1TH data did not show any spatial correlation. The

semivariograms for B0A 3TH had higher sill effect compared to other

semivariograms. The spatial variability of organic horizon thickness could be

attributed to long term influences on processes, such as sedimentation, organic

accretion, and compaction, which can make a significant variation. On the

other hand, this may be indicative of the localized damage to the marsh by

hurricanes or other marsh loss mechanisms.

Kriging was performed and the contour plots were generated for all the

organic horizons that produced good spatial structure. For BOA1TH, the

inverse distance squared method was used to interpolate data. Contour plots

for the Oa1, Oa2, and Oa3 horizon thickness data are presented for the saline

marsh (Figure 5.9) and brackish marsh (Figure 5.10). Thicker O a l horizons

were found at the edges of the water bodies indicating predominantly accretion

within the area at the saline marsh. The peaks were prominent because of the

presence of shallow areas with open water. Thickness of the O a l horizon was

thin towards the open water (Figure 5.9). It may be due to the accretion

differences between streamside and inland within the landscape. The O a l

found in inland may not be contemporary to the O a l at the streamside.

205


ooCD

T 3

OOCM

g(73•acCD

OO00

8

O )CD

iC/D

OC/D

OC/D♦

gCD

iCD

OoM-

O)E2C3)O

îîCDMtri23O)

ooo

ooco

ooCD

oo•M-

OOCM

(^LUO) aO U B U B A ILU aS

206

ReprocJucecJ with permission of the copyright owner. Further reproctuction prohibitect without permission.

I X X1— 1— 1— oT- CM CO o< < <o o oCO CO COO □ O

o

o□

o

o o

□o o

o o□

o o

o □ o

oooCNJ

oolO

ooo

s

IQ

ooIT)

■D

im-ocm

Im

Î2 X

oco

E2O)0"C

1EmWodiri2 3 D)

OOOOg

O§

ooMooco

oolOoo oo

coooCM

oo

( lUO) GOUBUBAjOISS

207


1600

o 1200

1 800

II 400 b

0

1600

1200

800

^ 400I0

1600

g 1200

800

i 400 S

OS

( )

\ \

è

S0A1TH

!

OP’

09

/ &

S0A2TH

/' /

I ?/ /

Ù \ 800

SOA3TH

1600400 800 1200Distance (m), vest - east

Figure 5.9. Spatial variability of 0 a 1 , Oa2, and O a3 horizon thickness for the saline marsh.

208


1 6 0 0

1200

800

f 400 G

0

1600

1200

( )<§>

1 800

I «>s

01600

Q 1200

( ;S

(f. 3o

!ë

9 ) ; s

0 % I .

/

S

8C(D

800

S 400

%

1 0

'è

30

AO

BOA1TH

BOA2TH

B0A3TH

0 ........400 800 1200 1600

Distance (m), west - east Figure 5.10. Spatial variability of Oa1, Oa2, and Oa3 horizon

thickness for the brackish marsh.

209


This may be the main source of spatial variability for organic subhorizon

thickness within a site.

Thickness of the 0a1 horizon was uniform within the brackish marsh

site, except randomly occurring peaks and falls due to the main waterway lying

NW to SE diagonal across the site. Thickness of the Oa2 horizon was highly

variable at both sites. Thicker areas associated with water bodies at the saline

marsh site (Figure 5.9), may be indicative of persistence of accretion processes

at the site. At the brackish marsh site, the Oa2 horizon was thinner near the

streams (Figure 5.10). This may be due to subsurface erosion within the area,

which may be aggravated by the waterways. The contour plot for SOA3TH

shows a uniform gradient from NW towards the eastern sections of the site

(Figure 5.9). Previously the stream at NW corner may be the main source of

accretions. The inland may be converted to open water later and the present

landscape may be developed due to marsh loss or by catastrophic action of

hurricanes. For the brackish marsh type, the Oa3 horizon was thin within the S

to W section, which coincides with the present water body. The stream located

at the western edge of the site may be developed later and separated the water

body (Figure 5.10).

5.4.2 Depth to Mineral Layer (DML) and Soil Subgroup Delineation


Several sampling points had standing water as they occurred in open

water. Except at those points, the DML is equal to the total organic layer

thickness, which is depth to the surface of the mineral soil layer (Cg horizon). A

210


mineral subhorizon (Cg2 horizon) also was found at many sampling points.

Figure 5.11 presents boxplots for the DML for both sites. The BDML had

higher variation compared to SDML, but the mean values were very similar

(104 cm). The SDML ranged from 50 to 150 cm while the BDML had a higher

range for BDML (0 to 165 cm) due to the presence of a surface mineral layer

for some sampling points.

The SDML data followed a bimodal distribution indicating that the data

may represent two populations. Presence or absence of the Oa3 horizon within

the profile was also associated with the difference. The Oa3 horizon thickness

was highly correlated with the SDML. The bimodal variation of data on SDM L

could indicate presence of depressions at the surface of the mineral layer due

to accretion differences within the site.

The BDML data followed a left skewed unimodal distribution and

associated with absence of an Oa2 horizon, Oa3 horizon, or both within the

profile. The Oa2 and Oa3 horizon thickness were highly correlated with BDML.

North to northeastern parts of the brackish marsh site had higher DML as well

as thicker organic horizons (Oa2 and O a3 horizons), indicating accretion

differences at the mineral layer, similar to the saline marsh site.

Exploratory data analysis indicated that the DML varies and correlated

with the presence or absence of organic horizons and their thickness variations.

This may indicate that more accretions occur where the mineral layer is

deeper. The Oa2 and Oa3 horizon thickness may vary due to variations in

accretion rates between depressions and other areas at the mineral layer.

2 1 1


Eo

m

I

I

2 0 0

150

100

SDML BDML

Marsh Type

Figure 5.11. Boxplots showing the data for depth to mineral layer (DML) for both sites.

212


Loss of a subhorizon or thin subhorizon can be due to erosion of lower organic

horizons, however this occurs at marsh-water interface of the degrading marsh.

Subsurface erosion is one of the major causes of marsh loss (Nyman et al.,

1994).

5.4 2.2 Spatial Data Analysis

The semivariograms are presented in Figure 5.12 for SDM L and BDML.

The lag distance of spatial correlation was higher for SDML (1476 m) compared

to the BDML (690 m). The BDML had higher sill effect compared to SDML.

Table 5.3 present model parameters for the semivariograms.

Table 5.3. Modal parameters and regression coefficients of the semivariograms for the DML at both sites.

Variable Shape Nugget (cm^) Sill (cm^) Lag (m) R:

SDML spherical 298 839.1 1476 0.939

BDML spherical 104 2162 690 0.928

Contour plots were generated by kriging using the semivariograms for

SDML and BDML (Figure 5.13). For the saline marsh site, shallow DML was

found close to the water bodies and deeper DML was found at the edges of the

site. Total organic layer thickness decreased towards the main waterway that

flows along the diagonal of the brackish marsh site. The BDML was very

shallow due to presence of surface mineral layer in this area.

213


oolO

omCM

OOo

sr--

oom

omCM

SIQ

T 3

QCO■acCO

QCOm

E5C3)OC. i

CO

c\i

m(D

.g)u_

o o o o o oo o o o oin o in o inCM CM T—

(^Uio) aouBUBAjUjas

214


■coc

ov>

S

Ib

o

3Otfl

S

IQ

1600

1200

800

400

0

1600

1200

800

400

0

%

6o

I

09

00 L

%\

z.

<%? ^Oq

Oq

Ooi

o

! i /\ i

' \ 1,

A. SDML

400 800 1200Distance (m), west - east

B. BDML

1600

Figure 5.13. Spatial variability of depth to mineral layer for the saline marsh (A) and brackish marsh (B).

215


5.4.2 3 Spatial Distribution of Soil Subgroups

The soil subgroups, Terric and Typic Medisaprists, were delineated

based on the organic layer thickness on the contour maps generated for SDML

and BDML. Soils that have the upper boundary of mineral layer within the

control section (130 cm) and below the surface tier (60 cm) are Terric

Medisaprists. If the upper boundary of the mineral layer occurred below control

section (130 cm) then they were delineated as Typic Medisaprists.

The DML was shallower in the north - eastern half of the saline marsh

site where the marsh is being converted into open water. Typic Medisaprists

occurred mostly toward inland areas away from open water and streams. Rest

of the area was delineated as Terric Medisaprists (Figure 5.14). At the saline

marsh site the Timbalier series is the Typic Medisaprists while the Bellpass

series is the Terric Medisaprists. Typic Medisaprists occurred in the north to

northeastern parts of the brackish marsh site, which had deep DML due to thick

organic horizons (Oa2 and Oa3 horizons) (Figure 5.14). The Lafitte soil series

is the Typic Medisaprists while the Clovelly soil series is the Terric Medisaprists

at the brackish marsh site.

5.4.2 pH for Different Subhorizons


The pH data for different soil subhorizons and weighted average pH for

the organic layers are presented in Figure 5.15. Mean pH for horizons for the

saline marsh was always lower than in the brackish marsh. This may be due to

higher accumulation of reduced sulfur material for the saline marsh compared

216


1600 -*---- —

1400

S

IQ

1200

1000

800

600

400

200

0

&Terric Medisaprists

Typic Medisaprists

0 200 400 600 800 1000 1200 1400 1600Saline marsh

1600

1400

1200

E,1000

o>Üc 800SQ 600

400

200

0

‘ i.-’

09

• C’ Î Î ,

<■' '

^ Terric Medisaprists

200 400 800 800 1000 1200 1400 1600Distance (m)

Brackish marsh Figure 5.14. Spatial varibility of the soils for the saline

and brackish marsh types.

217


CD"OOQ .C

gQ .

"OCD

C/)Wo"3O

8"O

CD

3.3 "CD

CD"OOQ .C

ao3

T3O

(DQ .

M0 0

g

8

7

Q .

6

5

4S0A1PH S0A2PH S0A3PH SCGPH SWTORPH

B0A1PH B0A2PH B0A3PH BCGFH BWTORPH

Subhorizon

T3(D

C/)c/)

Figure 5.15. Boxplots showing the data for pH of soil subhorizons and weighted average pH for the organic layer for both sites.

to the brackish marsh. The pH of the Oa1 horizon, saline marsh (SOA1PH)

and brackish marsh (BOA1PH) indicate similar variation. The S 0 A 1 P H and the

BOA1PH followed normal distribution with means at 6.3 and 7.0, respectively.

The pH of the Oa2 horizon also follows approximately normal distribution for

both sites with means of 6.5 for S 0A 2P H and 7.1 for BOA2PH. The S 0A 3P H

followed a left skewed distribution with a mean of 6.5 while the B 0A 3P H

followed approximately normal distribution with a mean of 6.9. The distribution

of the SCGPH was bimodal with a range of 6.4 to 7.7. The BCGPH was

normally distributed with a wide variation, ranging from 5.2 to 7.2. Weighted

average pH for the organic layers at both marshes are comparable.

Distribution for SW OPH was bimodal while BWOPH followed a left skewed

distribution.

5.4.2 2 Spatial Data Analysis

Semivariograms for pH of different subhorizons are presented in Figure

5.16 for the saline marsh, and Figure 5.17 for the brackish marsh. Table 5.4

presents the model parameters of the semivariograms for pH for different

subhorizons. High spatial autocorrelation was found for pH of organic

subhorizons for the saline marsh. However, the pH for the Cg horizon did not

show any spatial autocorrelation. Contour plots showing pH variation within

different subhorizons of the saline marsh is presented in Figure 5.18. The

SOA1 PH was higher and uniform within the western half of the site. Lower pH

was found towards the S W and NE corners of the site where degrading

marshes occur. Similar spatial pattern was observed for the SOA2PH at the

219


on

ooo

o so

tnoCMd

lOCM

OCOinCOd o

ooR

oolO

ooo

8cCO

O)CO

OOin

T 3

XQ.0 ü CO■OcCO

Xo.

gCO

XCL

gCO

1

gCO

12 C3)0 "C1E0)COdd2 3 C3J

aouBUBAjujag

220


lO

cmooo

oc n

ooCO

ddlOd

COd

IT3XCLÜÜCOT3CCO

Q.

iCO

^ X

& ^

8

IQ

oCO

Q .

5 gCO

(/)E2O)gCD>

ÎCO

lO23LL

90UBUBAILU9S

221


Table 5.4. Semivarlogram model parameters for pH of soil subhorizons at both sites.Variable" Shape Nugget Sill Lag (m)

SOA1PH exponential 0.0933 0.1973 445 0.989

SOA2PH linear 0.0632 0.3171 2263 0.836

SOA3PH linear 0.0474 0.1895 2263 0.900

SW O PH linear 0.0427 0.2153 2263 0.845

BOA1PH spherical 0.0085 0.2165 480 0.920

BOA2PH spherical 0.1371 0.2141 856 0.732

BW OPH spherical 0.0406 0.1726 672 0.921

^ SCG PH, B 0A 3PH , and BCGPH data did not fit a semivarlogram.

222


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D

3.3 "CD

CD■DOQ .C

aO3"OO

CDQ .

roroCa)

1600-

g 1 2 0 0 !I

§ o800!

§ i<5 400, (0

0 -1600-

tc 1200i

3

8

S

80CH

S 400i5 I

p>Oï m<D

\0.tO

A. S0A1PH

Ia>If

C. S0A3PH

6.5-lO(d

3S

1600-

12001

800

iS>

8ra 400h□ i

o;160Ck-

rc 1200,I i

800i

s 400i.<aa

6.5

B. S0A2PH

lOtd

D. SCGPH

T3(D

(/)(/)

400 800 1200

Distance (m), west - east

1600Or 400 800 1200

Distance (m), west - east1600

Figure 5.18. Spatial variability of soil pH for O a l (A), Oa2 (B), Oa3 (C), and Cg (D)horizons for the saline marsh

saline marsh site with reduced pH in western portion of the site. A smooth pH

gradient from NW to SE direction was observed for the S 0A 3P H . The SCGPH

was quite uniform within the site with few exceptions. Figure 5.19 presents the

spatial variability of soil pH for different subhorizons at the brackish marsh. The

BOA1PH indicated high spatial variability, with lower pH values along the NW

to SE diagonal across the site. Lower pH values were observed for samples

close to open water areas. Similar spatial pattern was observed for the

BOA2PH but towards the NE corner of the site, pH increased then leveled off

and decreased. The decrease may be associated with open water areas. The

BOASpH and BCGPH did not show any spatial autocorrelation. Spatial pattern

for the BOASPH was similar to that of Qa2 horizon. The BCGPH was uniform

except around the SW section of the site where the pH was highly variable.

This may be due to closer proximity of the open water areas that promotes

leaching of acidic material into the Cg horizon.

5.4.3 Organic/Mineral Ratio (OMR) for Different Subhorizons


Organic/mineral ratio data ranged widely for the brackish marsh site

compared to saline marsh site for all organic subhorizons. Figure 5.20

presents the boxplots for the O M R data for different subhorizons and for the

weighted-average of the organic layer. Organic/Mineral ratios were higher for

the brackish marsh site indicating more organic material compared to the saline

marsh site. The S O A 10M R data were normally distributed with a mean of

0.529. Right skewed B 0 A 1 0 M R data had a higher mean (1.637) with two

224


CD■ D

OQ .C

gQ .

■DCD

C/)C/)

8■D

CD

3O(/)

S

I

1600

1200,

800

400

A. B0A1PH

-j’o

O )bi - 1

'oo.r-

7.0

■co

1600

1200Î

800i

§I 400]â i

B. B0A2PH

3.3"CD

CD■DOQ .C

ao3T3O

(DQ .

T3(D

(/)(/)

rototn

Ol

1600-

I 1200'

C. BOASPH

38

S

800

OZm 400' (0

0-1600r

400 800 1200

Distance (m), west - east

1600

c 1200]

§ :S 400]3 ;

o.r-

D. BCGPH

800] O'L

s>.T-

0 1600400 800 1200

Distance (m), west - east Figure 5.19. Spatial variability of soil pH for Oal (A), Qa2 (B), Ga3 (C), and Cg (D)

horizons for the brackish marsh.

CD■ D

OQ .C

gQ .

■DCD

C /)C /)

8

ci'

33 "CD

CD■DOQ .C

aO3"OO

CDQ .

ma:

Ic

ü'c§,o

5

4

3

2

0S0A 3 SW TD

B0A3

S0A1 S0A 2

B0A1 B0A2 BWTD

Subhorizon

■DCD

Figure 5.20. Boxplots showing the data for organic/mineral ratio (OMR) for different organic subhorizons and weighted average OMR for the organic layer for both sites.

(/)(/)

outliers. The S O A 20M R data showed a bimodal distribution with a mean of

0.790. The B 0 A 2 0 M R data were normally distributed with a wider range and

its mean was 2.034. The B 0A 20M R data were highly variable. The

S 0 A 3 0 M R data were normally distributed and very similar to the O a l horizon

data. The B 0 A 3 0 M R data were bimodal with the lowest mean (1.202) at the

brackish marsh. The weighted OM R data for the saline marsh (SW OM R)

followed a bimodal distribution with a mean of 0.640. The BW OMR data were

normaily distributed with a mean of 1.506.

5.4.3 2 Spatial Data Analysis

Well-structured semivarlogram models were fitted only for O a l horizon

for the saline marsh (Figure 5.21 ), and for Oa2, and Oa3 horizons for the

brackish marsh (Figure 5.22). The semivarlogram model parameters are

presented in Table 5.5.

Table 5.5. Model parameters and regression coefficients for the semivariograms for OMR for organic soil subhorizons at both sites.

Variable^ Shape Nugget Sill Lag (m) R2

S 0 A 1 0 M R linear 0.0271 0.0456 2263 0.871

SW OM R linear 0.0373 0.0615 2263 0.493

B 0 A 2 0 M R spherical 0.143 0.501 561 0.934

B O A 30M R linear 0.145 1.124 2263 0.787

BW OMR spherical 0.173 0.356 799 0.683

S 0 A 2 0 M R , S 0 A 3 0 M R , and B 0A 10M R data were not autocorrelated.

227


□

Qi a: ce

o oi s

œ (O c/3 O □ <g

□

O

□

□

□ < o

□ < o

0

1

oom

ooo

oCM

LOO

od

ood

8

IQ

Ooin

T3

O

§COT3Cm

0

1CO

ce<2 O

gO)CDJC

CO

IO)o*wffl>E(D

CO

CMlOÿ.3.g)LL

aouBUBA^uag

228


oc a: a:

8 §o o

00 00 00o □ <g

<

□

i:

□

<□

oooCM

OOm

ooo

oocsi

oCO

oCM

ooo

oMfCO

oo

8cCD

O)CD

oolO

T 3

OCso

g00TDCCD

OC

o

g00

oc

g00eu

c/>iO)0"C1Ec%c\iCMcri2 o c:)

90UBUBAIUJ8S

229


The inverse distance squared method was used for spatial interpolation for the

variables that do not fit a semivariogram.

Figure 5.23 presents contour plots for OMR for soil subhorizons at the

saline marsh site. Higher OM R associated with northern part of the site. This

indicates more sedimentation for the open water areas within the southern part

of the site. The S 0 A 2 0 M R data were highly variable but did not show any

spatial autocorrelation. The S O A 30M R was uniform within the site with few

exceptions and the NE and N W corners had the highest organic matter content.

This may be attributed to marsh degradation at these locations within the site.

High spatial variability was evident at the brackish marsh site (Figure 5.24).

The B O A 10M R data were uniform with an exception of a high organic patch at

the western part of the site. Higher organic/mineral ratios were associated with

inland of the landscape indicating lack of sedimentation in Oa2 and O a3

horizons. This may also be due to subsurface erosion close to open water

areas or waterways. Mineral sediments from the degrading surface organic

layers may be washed away or moved into subsurface organic horizons.

Weighted average of the organic/mineral ratio was used as a separate variable

to study the composition of the organic layer at both sites. Spatial variability at

the brackish marsh site indicates the increased mineral content towards open

water areas.

5.5 Conclusions

Thickness of the Oa2 and Oa3 horizons were highly variable compared

to the Oa1 horizon. Variation of horizon thickness was greater at depths within

230


1600

g 1 2 0 0

3

& 800

II 400b

01600

o 1200

0.6

0.5

3O(n

8

800

I 400 b

01600

I 1200

) / ) (

I ).o

00o/

3o0) 800

8I 400 b

(\f>.o

0.5

400

■C>o,

(A)

(B)

(C)

800 1200 1600

Distance (m), west - east Figure 5.23. Organic/mineral ratio for the O a l (A), Ga2 (B), and OaS (C)

horizons at the saline marsh.

231


1600,

1 1200

800

I 400

01600

€g 1200

38

800

Iè 400ë

01600

Ic 1200

I 800

IJS 400 wB

2.0

O f

la

^■0

2.0

roo\

400 800

■O

1200

(A)

(B)

(C)

1600Distance (m), west - east

Figure 5.24. Organic/mineral ratio for the O a l (A). Q a2 (B), and G a3 (C ) horizons at the brackish marsh.

232


the profile. Except for the thickness of the 0a1 horizon within the brackish

marsh, organic horizon thickness showed high spatial autocorrelation, and the

semivariograms represented spherical models. Contour plots indicated more

accretions near water bodies. Lower thickness towards open water areas may

indicative of recent accretions at the streamside of the landscapes. The thicker

0 a 1 horizon at the inland areas may not be contemporary to the streamside

0a1 horizons. Contour plots for subsurface organic horizons may also carry

the variation in soil formation between landscape positions within the site.

Spatial variation of depth to mineral layer (DML) is a result variation in

deposition when mineral layer was formed. Variations in organic subhorizons

are indicative of previous marsh status, and variations in organic and inorganic

accretions. Sometimes it may be indicative of recent changes in marsh status,

especially for a degrading marsh. Thicker organic layer is indicative of healthy

marsh that can keep up with the local subsidence because of higher accretion

rates. Thin organic layer can indicate degrading marsh experiencing

subsurface erosion at marsh-water interfaces.

At the saline marsh, the DML was shallower near water bodies. Terric

and Typic Medisaprists were delineated based on the organic layer thickness.

Typic Medisaprists occurred mostly toward inland areas and away from open

water at the saline marsh. The DML was shallower in the northeastern half of

the saline marsh where the marsh is degrading. Typic Medisaprists occurred in

the northeastern part of the brackish marsh site, which had thick organic layers

due to presence of thick Oa2 and Oa3 horizons.

233


High spatial autocorrelation was found for pH of organic horizons for

saline marsh. However, the pH for the Cg horizon did not show any spatial

autocorrelation. Lower pH was associated with degrading marsh patches

toward the S W and NE corners of the saline marsh. A smooth pH gradient

from NW to SB direction was observed for the Oa3 horizon at the saline marsh.

High spatial variability was found for pH for 0a1 horizon within the brackish

marsh site, with lower pH values along the N W to SB diagonal across the site.

The pH was lower close to open water areas and pH for the Oa3 and Cg

subhorizons did not show any spatial autocorrelation.

Organic/mineral ratio (OM R) data ranged widely for the brackish marsh

compared to saline marsh for all organic subhorizons. More organic matter at

the brackish marsh gave higher OM R compared to the saline marsh. Higher

OMR associated with northern part and lower OMR at southern part of the site

indicated more inorganic accretions (sedimentation) from open water areas.

The high OM R at NE and N W corners of the saline marsh can be attributed to

marsh degradation within this site. The OMR data were highly spatially variable

at the brackish marsh site. Higher O M R was associated with inland indicating

lack of sedimentation.

Spatial variability can be attributed to long-term influence of processes

such as sedimentation, organic accretions and compaction and consolidation.

Localized damages to marshes by hurricane activity and due to marsh

degradation also can cause spatial variability within the marsh. Degradation of

a surface organic layer may result in frequent flooding due to local subsidence.

234


The standing water can cause consolidation of sediments within the surface

layer. The consolidation and compaction may be extended into the subsurface

horizons with water activity. Frequent flooding also may provide more

sediment. Understanding of spatial variation of marsh soil characteristics will

help to identify the prominent processes within the marshes and to plan marsh

management strategies.

5.7 References

Barry, R. P., and J. M. Ver Hoef. 1996. Blackbox kriging: Spatial predictionwithout specifying variogram models. J. Agric. Biol, Envtl. Stat. 1(3):297- 322.

Boesch, D. P., Levin, D., Nummedal, D., and Bowles, K. 1983. Subsidence in coastal Louisiana: Causes, Rates, and Effects on Wetlands. U. S. Fish and Wildlife Sen/ice FW S/OBS-83/26, Washington, D. 0 ., 30pp.

Brus, D. J., J. J. De Gruijter, B. A. Marsman, R. Visschers, A. K. Bregt, A. Breeuwsma, and J. Bouma. The performance of spatial interpolation methods and choropleth maps to estimate properties at points: A soil survey case study. Envirometrics 7:1-16.

Cahn, M. D., J. W. Hummel, and B. H. Brouer. 1994. Spatial analysis of soilfertility for site-specific crop management. Soil Sci. Soc. Am. J. 58:1240- 1248.

Childers, D. L., J. W . Day, and R. A. Muller. 1990. Relating climatological forcing to coastal water levels in Louisiana estuaries and the potential importance of El Nino-Southern oscillation events. Climate Research 1:31-42.

Chmura, G. L., R. Costanza, and E. Kosters. 1992. Modelling coastal marsh stability in response to sea level rise: a case study in coastal Louisiana, USA. Ecological Modelling, 64:47-64.

Day, J. W., D. Pont, P. F. Hensel, and C. Ibanez. 1995. Impacts of sea-level rise on deltas in the Gulf of Mexico and the Mediterranean: The importance of Pulsing Events to sustainability. Estuaries 18(4):636-647.

235


DeLaune, R. D., C. J. Smith, and W . H. Patrick, Jr. 1983. Relationship of marsh elevation, redox potential, and sulfide to Spartina alterniflora productivity. Soil Sci. Soc. Am. J. 47:930-935.

DeLaune, R. D., J. H. Whitcomb, W. H. Patrick, Jr., J. H. Pardue, and S. R. Pezeshki. 1989. Accretion and canal impacts in a rapidly subsiding wetland. I. ^^^Cs and ^^°Pb techniques. Estuaries 12(4):247-259.

DeLaune, R. D., J. A. Nyman, and W. H. Patrick, Jr. 1994. Peat Collapse, ponding and wetland loss in a rapidly submerging coastal marsh. J. Coastal Res. 10:1021-1030.

Feijtel, T. C., R. D. DeLaune, and W. H. Patrick, JR. 1988. Seasonal pore water dynamics in marshes o f Barataria Basin, Louisiana, Soil Sci. Soc. Am. J. 52:59-67.

Flatman, G. T., and A. A. Yfantis. Geostatistical sampling designs forhazardous waste sites. P.779-801. In L. H. Keith (ed.) Principles of environmental sampling. Secon Edition. Am. Chem. Soc., Washington, DC.

Gamma Design Software. 1994. GS+ Geostatistics for the EnvironmentalSciences Version 2.3 User's Guide. Gamma Design Software, Plainwell, Michigan 49080, USA.

Glazovskaya, M. A. 1968. Geochemical landscapes and types of geochemical soil sequences. Transactions of 9'’’ Int. Congr. Soil Sci. Adelaide. 4:303- 312.

Golden Software Inc. 1995. SURFER for Windows, Version 6 User’s Guide. Golden Software Inc., Golden, Colorado, 80401-1866, USA.

Goovaerts, P. 1997. Geostatistics for natural resources evaluation. Oxford University Press Inc., New York, New York. 483pp.

Gotway, C. A., R. B. Ferguson, G. W. Hergert, and T. A. Peterson. 1996.Comparison of kriging and inverse-distance methods for mapping soil parameters. Soil Sci. Soc. Am. J. 60:1237-1247.

Griffin, T. M., and M. C. Rabenhorst. 1989. Processes and rates ofpedogenesis in some Maryland tidal marsh soils. Soil Sci. Soc. Am. J. 52:59-67.

236


Haering, K. C., M. C. Rabnhorst, and D. S. Fanning. 1989. Sulfur spéciation in some Chesapeake Bay tidal marsh soils, Soil Sci. Soc. Am. J. 53:862- 870.

Hanna, A. Y ., P. W . Harlan, and D. T. Lewis. 1982. Soil available water as influenced by landscape position and aspect. Agron. J. 74:999-1004.

Hatton, R. S., R. D. DeLaune, and W. H. Patrick, Jr. 1983. Sedimentation accretion, and subsidence in marshes of Barataria Basin, Louisiana. Limnology and Oceanography 28:494-502.

Isaaks, E. H., and R. M. Srivastava. 1989. An introduction to applied geostatistics. Oxford University Press, New York. 561p.

Knotters, M., D. J. Brus, J. H. Oude Voshaar. 1995. A comparison of kriging, co-kriging and kriging combined with regression for spatial interpolation of horizon depth with censored observations. Geoderma 67:227-246.

Krairpanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1992a. Distribution of organic and reduced sulfur forms in marsh soils of coastal Louisiana, Org. Geochem. 18:489-500.

Krairpanond, N., R. D. DeLaune, and W. H. Patrick, JR. 1992b. Sulfur dynamics in Louisiana coastal freshwater marsh soils. Soil Sci. 151(4):261-273.

Lam, N. S. 1983. Spatial interpolation Methods: A review. The American Cartographer 10(2): 129-149.

Lin, C., and M. D. Melville. 1994. Acid sulphate soil-landscape relationships in the Pearl River Delta, southern China, Catena 22:105-120.

Madden, C. J., J. W. Day Jr., and J. M. Randall. 1988. Freshwater and marine coupling in estuaries of the Mississippi River deltaic plain. Amer. Soc. Limnoi. And Oceanogr. 33:982-1004.

McDaniel, P. A., G. R. Bathke, S. W. Buol, D. K. Cassel, and A. L. Falen. 1992. Secondary manganese/iron ratios as pedochemical indicators of field- scale throughflow water movement. Soil Sci. Soc. Am. J. 56:1211-1217.

Nyman, J. A., J. C. Callaway, and R. D. DeLaune. 1993. Case study of a rapidly submerging coastal environment: relationships among vertical accretion, carbon cycling and marsh loss in the terrebone basin, Louisiana. Proceedings of the Hilton Head Island South Carolina USA International Coastal Symposium, June 6-9, 1993. vol 2. 452-457p.

237



Nyman, J. A., C. R. Crozier, and R. D. DeLaune. 1995. Roles and patterns of hurricane sedimentation in an estuarine marsh landscape. Estuarine Coastal and Shelf Science 40:665-679.

Reese, R. E., and K. K. Moorhead. 1996. Spatial characteristics of soilproperties along an elevational gradient in aCcarolina Bay wetland. Soil Sci. Soc. Am. J. 60:1273-1277.

Rossi, R. E., D. J. Mulla, A. G. Journel, and E. H. Franz. 1992. Geostatistical tools for modeling and interpreting ecological spatial dependence. Ecological Mongraphs 62(2):277-314.

Smith, J. L., J. J. Halvorson, and R. I. Papendick. 1993. Using multiple-variable indicator kriging for evaluating soil quality. Soil Sci. Soc. Am. J. 57:743- 749.

Salinas, L. M., R. D. DeLaune, and W . H. Patrick, Jr. 1986. Changes occuring along a rapidly submerging coastal area: Louisiana, USA. Journal of Coastal Research 2(3):269-284.

Soil Conservation Service (SCS). 1984. Soil survey of Lafourche Parish,Louisiana. United States Department of Agriculture, Soil Conservation Service. Washington, DC.

Soil Conservation Service (SCS). 1989. East Central Barataria cooperative river basin study, Jefferson, Orleans, Plaquemines and St. Charles Parishes, Louisiana. Rev. May 1989. United States. Soil Conservation Service. National Cartographic Center (U.S.). Ft. Worth, TX : USDA-SCS-National Cartographic Center ; [Alexandria, La. State Conservationist, distributor].

SPSS Inc. 1997. SPSS for Windows Release 8.0.0.

Turner, R. E. 1990. Landscape development and coastal wetland losses in the northern Gulf of Mexico. Amer. Zool. 30:89-105.

Webster, R., and M. A. Oliver. 1989. Optimal interpolation and isarithmic mapping of soil properties. VI. Disjunctive kriging and mapping the conditional probability. J. Soil Sci. 40:497-512.

238


CHAPTER 6: CONCLUSIONS

This dissertation presents the soil science investigations for the

multidisciplinary research under the Louisiana NASA/EPSCoR (Experimental

Program to Stimulate Competitive Research) global change research cluster,

which study the fate of carbon and sediments within the Barataria Bay Basin,

Louisiana. This project presents the spatial variability of coastal organic soil

characteristics at different scales. It estimates the spatial variability of pyrite

accumulation to represent variability in the labile carbon pool and the

biogeochemical status of the marsh.

First study presents the ability of assessing seawater influence within the

marsh based on water composition data. The effect of salinity and ionic

competition on affinity of S to roots, and the feasibility of ion exchange resin

strips for water and soil analysis for spatial variability studies are presented in

the second study. Third study presents the pyrite accumulation within the soil

profiles between landscape positions for different marsh types. Spatial

variability is identified and spatial autocorrelation of selected marsh soil

characteristics is presented in the fourth study.

The main objective of the first study is to establish the relationship

between electrical conductivity (EC) and total dissolved solids (TDS) for ground

truthing the Airborne Electromagnetic Profiler data. The relationship, TDS (ppt)

= 0.669 * EC (dS/m) - 0.6208 is recommended for conversion of EC to TDS.

The relationship between EC and the elemental ratios indicate the

seawater influence and deviation of the ratios may be used to determine flow

239


direction within marsh for the hydrology work. Barataria Bay estuary has a

chloride dominant water system with a wide range of salinity. The elemental

ratios are similar to that of seawater, which indicates that the water

characteristics of this estuary are extensively under the influence of seawater

intrusion.

Salinity change is mainly due to the dilution effect by mixing with the

fresh water flow. Chloride and SO 4 maintain a linear relationship with EC; Cl

(mg L'1) = 303 + 295 * EC (dS m'^), and SO 4 (mg L'^) = 4.21 + 43.35 * EC (dS

m'^). A significant deviation from this relationship may indicate that the other

processes such as plant uptake or assimilation has changed the water

composition. The CI/SO 4 ratio is about 7.0 for seawater. W ater samples

collected from saline, brackish, and intermediate marsh types maintain a ratio

similar to seawater but fresh marsh had a significantly higher CI/SO 4 ratio

indicating it is independent from seawater influence. Therefore, CI/SO 4 ratio is

proposed to assess the extent of seawater intrusion within the basin.

It is evident from water composition data that Cl dominates the water

system and SO 4 has to compete with Cl as a major anion. Therefore, it is

important to study the influence of Cl and ionic competition on sulfate reduction.

Ion-exchange resins are helpful in understanding the processes by simulating

root uptake. Salinity and Cl significantly reduced the resin extractable sulfate.

This explains a possible reason for accumulating SO 4 within the root zone.

Plants exclude excess soluble salts from uptake as a tolerant mechanism.

240


Because of high Cl content, SO 4 may have limited affinity to root. This may

accumulate of SO 4 that undergo reduction and form sulfide and pyrite.

Spatial variability studies require an analysis of many samples.

Therefore, a quick method would be very useful to study the spatial variability of

coastal organic soils. Resin strips have been used to measure soil parameters.

However, resins have not been tested for estuarine conditions. This study also

examines the potential use of resins for multi-elemental analysis of estuarine

water and to estimate the sulfur status of coastal organic soils.

Bicarbonate, chloride or phosphate can be used as the terminal anion for

the anion exchange strips. Either 01 concentration or electrical conductivity can

be added to the models to predict elements in water from resin extractable

elements. These procedures and prediction models are recommended for the

Barataria Bay Basin water system and similar chloride dominant estuarine

systems.

Both anion and cation exchange resin strips can be used together for

one sample and 0.5 HOI can be used as the eluent for both resins if the

ICP-AES is used for elemental analysis. This procedure will provide an

alternative to the traditional analytical methods. Equilibration time, elution time

and soil/water ratio needs to be adjusted depending on the S levels and salinity.

The resin extractable S (RES) maintained a linear relationship with water

soluble 8 (W SS) and non-pyritic 8 (NP 8 ) for both mineral and organic soils.

Including pH and Cl concentration as additional variables helped to obtain

better models to predict NP 8 and W 8 8 , especially for organic soils. Separation

241


of mineral horizons and organic horizons improved the regression coefficients

for individual models when compared to the overall model using for both soil

horizons. Therefore, better prediction can be achieved using different models

for different soil horizons. Sulfur mineralization is a limitation reported in other

studies for resin use in 8 analysis, which may be the main cause of low for

prediction models. However, the models are able to predict over 70 % of the

variation despite the variations in marsh types, salinity, chloride, organic/mineral

ratio and pH. Resin extractable 8 can be suggested to estimate non-pyritic 8

and water soluble 8 for spatial variability studies.

Each marsh type occurs within a certain range of salinity depending on

the salt tolerance ability of its vegetation. Pyrite accumulation and sulfate

reduction may be different between the marsh types because of the salinity

differences and associated changes in 8 O4 content and the competition

between 8 O4 and Cl. As indicated by the resin study, salinity and 01 influence

sulfate reduction because of the ionic competition between 8 O4 and 01,

however, this effect may be lowered at low salinity levels. Effect of landscape

position and salinity level on pyrite accumulation is studied because of the

reported accretion differences between streamside and inland within the marsh

landscape.

Organic layers comprise distinctly different subhorizons within the soil

profiles and the number of subhorizons is variable. The surface horizons at

streamside represent the recent accretions, which may be younger than the

surface horizons inland. Erosion and re-deposition of organic soil material may

242


create wide spatial variability within a degrading marsh. Depth to mineral layer

indicates considerable spatial variation, especially at the brackish marsh site.

Deeper DML may represent depressions formed due to differences in

deposition when the mineral layer was formed.

Soil BD is lower for the brackish marsh compared to the saline marsh.

Lack of sediment feeding into the brackish marsh may be the reason for the

iower BD. More mineral amounts may stabilize the BD in saline areas. The BD

varies highly within the organic layer of the brackish marsh while it is uniform at

the saline marsh. High salinity can increase higher sedimentation due to

enhanced flocculation.

Salinity profiles differ between landscape positions when the inland

position is landlocked. This difference may be pronounced in more saline

areas. Evapotranspiration and lack of lateral flow (due to hydrological barriers)

may change the salinity profiles inland as compared to streamside. When small

streams flow through the marsh, the hydrological barriers are broken and

salinity differences disappear. Biogeochemical differences may result from this

salinity variation within the landscape. Therefore, understanding the spatial

variability of salinity is important when monitoring the marsh status within an

area.

Salinity influences pyrite accumulation, however, its influence varies

within the profile. Pyrite content within the surface horizon is significantly

different between landscape positions. There is no evidence of differences

between marsh types because the balance between pyrite formation and

243


oxidation mainly controls pyrite content within the surface horizons. Pyrite

content within the inland surface organic layers is lower compared to the

streamside, especially when the inland is landlocked. The landscape position

effect on pyrite accumulation dramatically diminishes deeper in the soil profile

but marsh type effect is prominent. This may indicate the salinity effect on

pyrite accumulation along the salinity gradient. The higher the marsh salinity

the higher the pyrite accumulation within a soil horizon. The highest pyrite

accumulation is observed within the thin layer of dark organic material that

occurred immediately above the mineral layer. This may represent the marine

environment when it was formed.

The non-pyritic Fe contents are significantly different between the

landscape positions within the more saline marshes (i.e., saline and brackish

marshes), suggesting that inland lacks sedimentation. The intermediate marsh

contains significantly lower non-pyritic Fe within surface organic layers

compared to other marshes. This may not be related to the salinity but to the

status of the marsh. The intermediate marsh at this location is extensively

broken-up (less than 50% vegetation cover). The non-pyritic Fe content can be

a suitable indicator to assess marsh status. When landlocked, inland positions

can retain more non-pyritic Fe within sub surface organic layers due to limited

pyritization, otherwise the non-pyritic Fe can be flushed away.

Non-pyritic S content within the organic sub horizons differs significantly

between marsh types due to salinity differences. More saline marshes contain

more non-pyritic S while less saline marshes are limited by the S supply. The

244


brackish marsh inland contains more non-pyritic S within the profile compared

to the streamside. This suggests that when the inland is not landlocked, non-

pyritic S accumulate within the profile. This is similar to the reason suggested

by the resin study. Subsurface organic horizons within the profiles of less

saline areas are S limited as also indicated by the water composition data.

Soil pH differences are more pronounced between marshes than the

landscape positions. The pH does not vary within the organic layer. The

mineral layer always has higher pH compared to the organic subhorizons,

which may represent the saline (marine) environment under which it was

formed. The surface horizon pH differs significantly between less saline

marshes and more saline marshes. Landscape difference is found only at the

intermediate marsh where the marsh is extensively degrading. The pH of the

mineral layers of less saline marshes is significantly lower compared to the

saline marshes.

Pyrite framboids are detected within the saline marsh sediments using

scanning electron microscopy. The framboids are 6 to 8 pm in diameter and

are found in close association with the organic material. Both the saline and

brackish marsh sediments contain mixed mineralogy with smectite, illite, and

kaolinite clays. Pyrite is evident within the sediments for both marshes with

coarser fractions ( < 2 0 pm) containing more pyrite.

There is a considerable amount of reduced S compounds accumulated

within these soil profiles, which occur as pyritic and non-pyritic S. These

organic soils should be reclassified to indicate sulfidic materials within their

245


profiles. The technique used to estimate pyrite and non-pyritic S can be

adopted as a method to characterize sulfidic material within these soils.

Soil morphology and biogeochemical status is variable within the

landscape as shown in the previous study. This study estimates the spatial

autocorrelation for several coastal organic soil characteristics within a saline

and a brackish marsh at a larger scale. This information is required for planning

future marsh sampling schemes as well as understanding the spatial variability

pattern of marsh soils within the landscape.

Variation of subhorizon thickness is greater at depths within the profile

and shows high spatial autocorrelation that follows semivariograms represented

by spherical models. Accretion differences within the landscape may be the

major source of variability of soil morphology within these sites. Contour plots

indicated more accretion near water bodies. Lower thickness within open water

bodies may be due to consolidation of sediments by the standing water.

Contour plots for subsurface organic horizons may indicate a historical

signature of marsh status in the past. Spatial variation of DML is a result of

variations in differences in deposition at mineral layer formation, which is

associated with organic subhorizon thickness variation. Thicker DML is

indicative of healthy marsh that could counter the local subsidence because of

higher accretion rates. Thin DML may indicate degrading marsh that also may

be experiencing subsurface erosion, or consolidation of sediments.

Terric and Typic Medisaprists are delineated based on the organic layer

thickness. Typic Medisaprists occur mostly toward inland areas and away from

246


open water and streams for the saline marsh. The DML is shallower in the

northeastern half of the saline marsh where the marsh is degrading. Typic

Medisaprists occur in the northeastern part of the brackish marsh site, which

had thick organic subhorizons.

High spatial autocorrelation is found for pH within organic subhorizons of

the saline marsh. However, the pH of the Cg horizon does not show any spatial

autocorrelation. Lower pH is associated with degrading marsh patches toward

the SW and NE corners of the saline marsh. A smooth pH gradient from NW to

SE direction is observed for the subhorizons within the saline marsh. High

spatial variability is found for pH for the surface organic subhorizons within the

brackish marsh site, with lower pH values along the N W to SE diagonal across

the site. The pH is lower close to open water areas and pH for the organic

subhorizons and mineral horizon does not show any spatial autocorrelation.

Organic/mineral ratio (OMR) data ranges widely for the brackish marsh

compared to saline marsh for all organic subhorizons. More organic matter for

the brackish marsh give higher OMR compared to the saline marsh. Higher

O M R associated with the northern part and lower OM R at the southern part of

the site indicates more inorganic accretions (sedimentation) from the open

water. The high OM R at the NE and N W corners of the saline marsh can be

attributed to marsh degradation within this site. The OMR data are highly

spatially variable for the brackish marsh site. Higher OM R is associated with

the inland marsh indicating lack of sedimentation.

247


VITA

Langappulige Cecil Dharmasri was born on August 8, 1961, in Matara,

Sri Lanka. He was raised in Matara, where he graduated from St. Servatius

College in 1980. He received his bachelor of science degree in Agriculture from

Faculty of Agriculture, University of Peradeniya, Sri Lanka, in 1984 specializing

in Soil Science as undergraduate major.

He worked for the Department of Agriculture as a Research Officer from

1985 to 1992. He was stationed at the Regional Agricultural Research Center,

Angunakolapelessa, Sri Lanka. His work involved applied and adaptive

research in Agronomy, Soil Science and Cropping Systems. Cecil received a

scholarship from International Development Research Center, Canada, to study

for his masters in 1992. He obtained his master of science degree in Soil

Science from the Department of Soil Science, University of Saskatchewan,

Canada, in 1995. His research focussed soil-plant-water relations for semiarid

conditions in Canada and Sri Lanka and worked under the supervision of Dr. E.

deJong.

Cecil started his doctoral program in the Department of Agronomy at

Louisiana State University in 1995, under the supervision of Dr. W ayne H.

Hudnall. Cecil's dissertation research is part of the Louisiana NASA/EPSCoR

Global Change research program. He majored in Soil Science with emphasis in

mapping sciences. Cecil is a member of Sigma Xi, Gamma Sigma Delta, Soil

Science Society of America, American Society of Agronomy, Soil and W ater

Conservation Society, Clays and Clay Minerals Society, Soil Science Society of

248


Sri Lanka, Sri Lanka Association for the Advancement of Sciences,

International Erosion Control Association, and the International Soil Science

Congress.

Cecil met Ayesha when he was working at Angunakolapelessa and they

got married in 1990. His elder son, Poorna, was born in 1991 and the younger

son, Chamara, was born in 1996. He lived with his wife and two sons in Baton

Rouge. Cecil accepted a position as a Post Doctoral Fellow at Zeneca Ag

Products, Western Research Center, California.

249


DOCTORAL EXAMINATION AND DISSERTATION REPORT

Candidate: L. Cecil Dharmasri

Major Field; Agronomy

Title of Dissertation: Spatial Variability of Coastal Organic SoilCharacteristics

Approved:

Date of Examination:

May 4. 1998

Major'Professer and Chairman

Dean of the Graduate School

EXAMINING COMMITTEE:


IMAGE EVALUATION TEST TARGET (Q A -3 )

y/l

/

1.0

l . l

[ f i a IIIIM

: l~ ilia I lia

mil 1.8

1.25 1.4 1.6

15 0 m m

V

>■*

V

O /

' / y

x 4 P P l_ IE D A I IV W G E , In c_ = 1653 East Main Street

' Rochester, N Y 14609 USA Phone: 716/482-0300

- = - - ^ = Fax: 716/288-5989

O 1993. Applied Image, Inc.. Ail Rights Reserved


spatial variability of coastal organic soil characteristics

Documents