Vegetation change along salinity gradients in the tidal marshes of the upper Savannah River estuary

MISSING IMAGE

Material Information

Title:
Vegetation change along salinity gradients in the tidal marshes of the upper Savannah River estuary
Physical Description:
xviii, 306 leaves : ill. ; 29 cm.
Language:
English
Creator:
Bossart, John M
Publication Date:

Subjects

Subjects / Keywords:
Environmental Engineering Sciences thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Environmental Engineering Sciences -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 2002.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by John M. Bossart.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029640125
oclc - 51947221
System ID:
AA00014260:00001


This item is only available as the following downloads:


Full Text











VEGETATION CHANGE ALONG SALINITY GRADIENTS IN THE TIDAL
MARSHES OF THE UPPER SAVANNAH RIVER ESTUARY















By

JOHN M. BOSSART


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2002















ACKNOWLEDGMENTS

Many friends and colleagues have been involved in the opportunity for me

to undertake and complete this dissertation. I would like to thank the members of

my faculty committee, Dr. Mark Brown, Dr. Thomas Crisman, and Dr. Scot Smith,

for their time and guidance. Especially, I would like to thank my major advisor,

Dr. Clay Montague, for the many hours he spent in the field, in discussion, and in

deciphering the first draft.

Many colleagues at Applied Technology & Management, Inc., in

Gainesville provided invaluable assistance and some have spent literally

thousands of hours in the field, managing data, and conducting analyses. This

group includes George Otto and Kevin Flavin who kept a very stubborn

datalogging network running under extremely harsh conditions. Also, Pete

Wallace graciously shared his many years of experience in plant ecology.

Thanks also go to Chris Ahern, Gary Bazemore, Jenny Carter, Robert Garrren,

Matt Goodrich, Kostas Kalimtgis, Pete Peterson, Michelle Rau, Dan Rich, Bobby

Richardson, and Fran Way for their participation in ways too numerous and

diverse to list. Finally, special thanks go to Thomas Schanze for his patience

and encouragement through what turned out to be a very long process.

The Georgia Ports Authority generously supported this work. Thanks go

to Larry Keegan at Lockwood-Greene, Inc. and Bo Ellis at Applied Technology &

Management, Inc., for facilitating the funding. Thanks go also to Sam Drake and









John Robinette of the U.S. Fish and Wildlife Service, Savannah Coastal Refuges

for allowing access to the Savannah National Wildlife Refuge. Dr. Wiley Kitchens

of the University of Florida Cooperative Fish and Wildlife Research Unit also

provided many valuable hours of collaborative discussion.

Finally, I would like to thank my wife, Jean, for the years of support,

encouragement, and love she has given. She, more than anyone else, allowed

me to be more than I thought I could be. For that I owe her a debt of gratitude

that can never be fully repaid.















TABLE OF CONTENTS

page

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

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

LIST OF FIGURES .................... ................................... viii

A B S T R A C T ........................................................................ ......................... xvi

CHAPTER

1 INTRODUCTION .......................................... ..................... 1

S tudy A rea D description .................................................. .............................. 4
Location ............................................... ........................... 4
Estuarine Extent ...................................................... ................... 5
Defining the Salinity Gradient................................................ .............. 6
History of the Tidewater Rice Industry Land Management Practices ......7
Land-Cover Changes Associated with the Post Tidewater Rice
Industry Era ............................................................ ..................... 12
Tidal Creek Developm ent....................................... .................... ... 14
Marsh Substrate Development........................... .............. 15
Other Anthropogenic Perturbations ...................... ....... ............. 19
Sea-Level R ise ..................................................... ..................... 22
Literature Review............................. ......................... 24
Hypothesis and Approach ........................................... ...................... 29

2 M ETHO DS...................... ...................... ........................... 59

Mapping and Surveying ......................................... 59
Aerial Photography Acquisition........................ ..................... 59
River Channel and Tidal Creek Mapping...............................................60
Aerial Photograph Interpretation of the Tidal Creek Network ................60
Survey Instrumentation and Tidal Creek Cross-Sections ....................61
V egetation Studies ................................................. ....................... 61
Field Surveys............................ ......... ..... .................. 61
Vegetation Analysis ................................... ............................. 64
Hydrologic and Salinity Data Collection......................... ...... .............. 70









CHAPTER page

Water Level Monitoring Instrumentation ................... ................ 71
Salinity Monitoring Instrumentation.................... ........... ............. 72

3 R ES U LTS .................................................... ..................................... 83

Aerial Photography Interpretation of the Tidal Creek Network...................... 83
Marsh Surface Elevation ............................................. ...................... 85
Vegetation Study Results ........................................... .......................87
C luster A nalysis................................ ................. ......................94
Detrended Correspondence Analysis (DCA)........................................95
Tide Stage and Water Level Studies Results ............................................ 101
S alinity Study R esults................................................................................. 106
Integration of Vegetation Data with Environmental Parameters ................ 10
Correlation A nalysis.......................................................................... 110
Detrended Canonical Correspondence Analysis (DCCA).................. 114

4 DISCUSSION ................................................................... 168

R EFER ENC ES ..................................................... ............................... 189

APPENDIX

A VEGETATION DATA ........................................ ...................... 195

B W ATER LEVEL DATA ............................................................. ................. 250

C SALINITY DATA ......................................................... ...................... 265

D SURVEYED CROSS-SECTIONS OF FORMER MARGIN DITCHES AND
MAIN WATER SUPPLY CANALS ...................... .....................293

BIOGRAPHICAL SKETCH ............................................................................. 306















LIST OF TABLES


Table page

1-1 Locations and bottom elevations of four remnant rice trunks...................17

1-2 Summary of dredging projects impacting the Savannah River................20

2-1 Locations of belt transects along their associated river channels............62

2-2 Cover value categories and percent cover ranges for each category ......64

3-1 Tidal creek development on Argyle Island........................................... 83

3-2 Marsh surface elevations at 10-foot intervals along the ten vegetation
belt transects............................ ... ....................86

3-3 Coordinates and elevations at 100-foot points along vegetation
belt transects......................... ... ....... .................. .... ..... 88

3-4 List of all plant species found in the ten permanent vegetation belt
transects during the study period of October 1997 through
O ctober 2001 ..................................................... .. ......................90

3-5 Number of plant species found within each belt transect during each
sam pling event. ....................................... ......................................... 94

3-6 Tidal flooding frequency, depth, and duration at the ten belt transects..102

3-7 Summary statistics for sediment salinity data collected by the
datalogging equipment within marsh sediments ranked by average
salinity ......... .... ................................... .. ..................... 106

3-8 Pearson's and Spearman's correlation coefficients for species versus
average sediment salinity and temporal standard deviation of salinity... 111

3-9 Pearson's and Spearman's correlation coefficients for species versus
average depth of flooding during high tide and temporal standard
deviation of depth......................................................... .................... 112









Table page

3-10 Pearson's and Spearman's correlation coefficients for species versus
average belt transect elevation and temporal standard deviation of
belt transect elevation ............................................... 113

3-11 Summary of the DCCA correlations of the first two species axes with
the environmental variables .................. ...... ..................... 115

4-1 Summary of belt transect parameters.......................................... 170

A-1 Species codes, scientific name, common name, and presence (X) or
absence (-) data for each belt transect by sampling event................... 196

A-2 Frequency and percent cover for each species within
each belt transect for each of the six vegetation-sampling events.........214















LIST OF FIGURES


Figure page

1-1 Regional map of Savannah River drainage basin with study area ...........32

1-2 Project study area .................................................. ...................... 33

1-3 Landsat 7 satellite image showing extent of Savannah River estuary......34

1-4 Savannah River miles .................... .... ........................... 35

1-5 Comparison of gage heights at the USGS gaging stations at Hardeeville,
S C and C lyo G A ........................... ............... .................... ... .... 36

1-6 Flow volumes of the Savannah River recorded at USGS gaging station
near Clyo, GA.................... .................................... 37

1-7 Average daily flow (cubic feet per second) of the Savannah River as
recorded at the USGS gaging station near Clyo, GA...............................38

1-8 Recent average daily flow (cubic feet per second) of the Savannah River
compared to the average daily flow for the period of record (1938 to
present at USGS gaging station near Clyo, GA ......................................39

1-9 Typical growing season salinity probability contour at river flow of 8,200
cfs. Contours represent location of 50th percentile salinity gradient when
tide stage is above 4.5 feet. .............................................................. 40

1-10 Typical dry season salinity probability contour at river flow of 5,900 cfs.
Contours represent location of 50th percentile salinity gradient when tide
stage is above 4.5 feet. ............................................ ................... 41

1-11 Typical wet season salinity probability contour at river flow of 9,500 cfs.
Contours represent Location of 50h percentile salinity gradient when tide
stage is above 4.5 feet. .................................. .. ..................... ... 42

1-12 Conceptual schematic of tidal marsh classification based on salinity.......43

1-13 Historical map (1796) of a portion of Argyle Island drawn by John
M cK inno n ................................ ........................................ 44









Figure page

1-14 Historical map (c. 1840) of a portion of Argyle Island .........................45

1-15 Historical map (1846) of Redknoll Plantation on the northern portion of
Argyle Island drawn by C. de Choiseul ........................................46

1-16 Historical map (1867) of the Gowrie and East Hermitage Plantations on
Argyle Island drawn by either Charles or Louis Manigault .....................47

1-17 Conceptualized cross-section of typical rice field water management
system ............................................................................... ........48

1-18 Infrared aerial photograph (1999) of a portion of Argyle Island ...............49

1-19 Aerial photographs (1952 and 1999) of a portion of Argyle Island with
locations of margin ditch change analyses labeled as A, B, and C..........50

1-20 Location A change analyses of margin ditches .....................................51

1-21 Location B change analyses of margin ditches ..................................52

1-22 Location C change analyses of margin ditches ..............................53

1-23 Channel modification affecting downriver freshwater flow and upriver
salinity transport ................... .....................54

1-24 Tide gage locations on the lower Savannah River................................. 55

1-25 Yearly mean sea-level rise at the Ft. Pulaski, Georgia gage....................56

1-26 Land subsidence since 1933 resulting from municipal and industrial
water withdrawals of groundwater in proximity to the study area. ...........57

1-27 Linkages between river channel salinity and marsh root zone tracked
by salinity monitoring protocol ................... .........................58

2-1 Flight lines and photograph center points for 1:12,000 scale true color
and infrared aerial photography flown August 1999............................... 74

2-2 Flight lines and photograph center points for 1:25,000 scale true color
and infrared aerial photography flown August 1999............................... 75

2-3 Locations and coordinates of aerial targets along the lower Savannah
River used in rectification of August 1999 true color and infrared aerial
photography .................. ............................ ....................... 76









Figure page

2-4 Base map of main river channels and tidal creeks digitized from rectified
1:12,000 and 1:25,000 scale aerial photography acquired August 1999..77

2-5 Permanent vegetation monitoring belt transect locations....................... 78

2-6 Typical 500-foot vegetation sampling belt transect composed of
50 contiguous cover intervals........................................... 79

2-7 Locations of hydrologic and salinity datalogging stations.......................80

2-8 Hydrologic and salinity monitoring equipment setup at low tide and
high tide.................................... ................. ................................ 8 1

2-9 Positioning of salinity probes within marsh sediments and above marsh
su rfa ce .................. ........................................ ...................................... 8 2

3-1 Infrared aerial photograph (1999) with locations of rice-era main water
supply canals on Argyle Island......................... .. ...................124

3-2 Dendritic development of tidal creek networks on Argyle Island
associated with the Middle River and the Little Back River.................. 125

3-3 Aerial photograph (1938) of a portion of Argyle Island and the Little
Back River and an unsupervised classification showing a pool of
open water ................................................................. 126

3-4 False color infrared aerial photograph (1999) of a former rice field
square located on Argyle Island..................................................... 127

3-5 Belt transect Q1 surveyed cross-section ............................................128

3-6 Belt transect Q2 surveyed cross-section ............................................ 129

3-7 Belt transect Q3 surveyed cross-section ......................................... 130

3-8 Belt transect Q4 surveyed cross-section ........................................ 131

3-9 Belt transect Q5 surveyed cross-section ..................... ................... 132

3-10 Belt transect Q6 surveyed cross-section ........................................ 133

3-11 Belt transect Q7 surveyed cross-section ......................................... 134

3-12 Belt transect Q8 surveyed cross-section ......................................... 135









Figure page

3-13 Belt transect Q9 surveyed cross-section ............................................ 136

3-14 Belt transect Q10 surveyed cross-section ........................................ 137

3-15 Belt transect Q1 cover values of the top ten plant species ..................138

3-16 Belt transect Q2 cover values of all plant species ............................ 139

3-17 Belt transect Q3 cover values of the top ten plant species ..................140

3-18 Belt transect Q4 cover values of the top ten plant species ..................141

3-19 Belt transect Q5 cover values of the top ten plant species ..................142

3-20 Belt transect Q6 cover values of the top ten plant species .................. 143

3-21 Belt transect Q7 cover values of the top ten plant species ..................144

3-22 Belt transect Q8 cover values of the top ten plant species ..................145

3-23 Belt transect Q9 cover values of the top ten plant species .................. 146

3-24 Belt transect Q10 cover values of the top ten plant species ................147

3-25 Cluster analysis for the ten most common species occurring for all
sampling events for all belt transects ........................... ................... 148

3-26 Detrended correspondence analysis based on belt transect scores...... 149

3-27 Detrended correspondence analysis based on species scores.............. 150

3-28 Box plots comparing marsh surface elevations at the ten belt
transects to the high tide elevations as recorded in the adjacent tidal
cre e ks ........................................................ .................... 15 1

3-29 Belt transect Q1comparison of water levels between tidal creek and
m arsh interior ............................................................ 152

3-30 Belt transect Q10 comparison of water levels between tidal creek and
m arsh interior ....................................... ..... .................. 153

3-31 Sediment salinities for each of the ten belt transects........................... 154

3-32 Average sediment salinity within each of the ten belt transects during
each sam pling event ........................................................... 155









Figure page

3-33 Q1 comparison of sediment salinity changes and tidal regime
(November 1, 2001 November 21, 2001)........................................... 156

3-34 Q2 comparison of sediment salinity changes and tidal regime
(July 31, 2001 -August 30, 2001) ................... ....................... 157

3-35 Q3 comparison of sediment salinity changes and tidal regime
(July 7, 2001 August 6, 2001) ................................... .................... 158

3-36 Q3 comparison of sediment salinity changes and tidal regime
(November 1, 2001 November 21, 2001)........................................... 159

3-37 Q6 comparison of sediment salinity changes and tidal regime
(November 1, 2001 November 21, 2001)........................................... 160

3-38 Q8 comparison of sediment salinity changes and tidal regime
(July 7, 2001 -August 6, 2001) ...................................................... 161

3-39 Q9 comparison of sediment salinity changes and tidal regime
(November 1, 2001 November 21, 2001).......................................... 162

3-40 Datalogging station E comparison of sediment salinity changes and
tidal regime (July 7, 2001 August 6, 2001)........................................ 163

3-41 Datalogging station W comparison of sediment salinity changes and
tidal regime (July 7, 2001 August 6, 2001)........................................ 164

3-42 Detrended canonical correspondence analysis biplot relating relative
frequency plant data to five environmental variables ........................... 165

3-43 Detrended canonical correspondence analysis biplot relating relative
frequency plant data to ranked elevation and salinity .......................... 166

3-44 Detrended canonical correspondence analyses for the ten belt
tra nsects.................................................................. ................ 167

4-1 Systems diagram relating tidal marsh plant community structure of
upper Savannah River estuary to salinity and hydrologic gradients....... 185

4-2 Current velocities in the Middle River and Little Back River at the
location of the northern main water supply canal................................. 186

4-3 Marsh polygons associated with tidal creek system and connections to
main river channels .................... .... ................. 187









Figure page

4-4 Comparison of average sediment salinity and number of plant species
found at each belt transect .......................................... 188

B-1 Q1 tidal creek stage (November 1,2001 December 10, 2001)...........251

B-2 Q2 tidal creek stage (April 10, 2001 September 5, 2001)..................252

B-3 Q3 tidal creek stage (April 11, 2001 September 5, 2001)....................253

B-4 Q3 tidal creek stage (November 1, 2001 December 10, 2001)...........254

B-5 Q4 tidal creek stage (November 1, 2001 December 10, 2001)...........255

B-6 Q5 tidal creek stage (April 11, 2001 September 5, 2001)....................256

B-7 Q6 tidal creek stage (November 1, 2001 December 2, 2001)............. 257

B-8 Q7 tidal creek stage (November 1,2001 December 10, 2001)...........258

B-9 Q8 tidal creek stage (April 11, 2001 September 5, 2001)..................259

B-10 Q8 tidal creek stage (November 1,2001 December 10, 2001)...........260

B-11 Q9 tidal creek stage (November 1,2001 December 10, 2001)...........261

B-12 Q10 tidal creek stage (November 14, 2001 December 10, 2001).......262

B-13 Datalogging station W tidal creek stage (April 11, 2001 -
Septem ber 5, 2001) ....................................................... .....................263

B-14 Datalogging station E tidal creek stage (April 11, 2001 -
Septem ber 5, 2001) ............... ...................... ... ........................ 264

C-1 Q1 salinity data records for tidal creek (November 1, 2001 -
December 10, 2001) ..................... ........................... 266

C-2 Q1 salinity data records for marsh surface water and marsh
sediments (November 1,2001 December 10, 2001).........................267

C-3 Q2 salinity data records for tidal creek (March 29, 2001 -
September 5, 2001) ....................................... ..................... 268

C-4 Q2 salinity data records for marsh surface water and marsh
sediments (March 29, 2001 September 5, 2001).............................. 269









Figure page

C-5 Q3 salinity data records for tidal creek (March 29, 2001 -
September 5, 2001) ......................................................... 270

C-6 Q3 salinity data records for marsh surface water and marsh
sediments (March 29, 2001 September 5, 2001) .............................. 271

C-7 Q3 salinity data records for tidal creek (November 1, 2001 -
December 10, 2001) ................................................272

C-8 Q3 salinity data records for marsh surface water and marsh
sediments (November 1, 2001 December 10, 2001)...........................273

C-9 Q4 salinity data records for tidal creek (November 1, 2001 -
Decem ber 10, 2001) ......................................... ...274

C-10 Q4 salinity data records for marsh surface water and marsh
sediments (November 1, 2001 December 10, 2001).........................275

C-11 Q5 salinity data records for tidal creek (May 16, 2001 -
September 5, 2001) ........................... ........................ 276

C-12 Q5 salinity data records for marsh surface water and marsh
sediments (November 1,2001 December 10, 2001).........................277

C-13 Q5 salinity data records for marsh surface water and marsh
sediments (March 29, 2001 September 5, 2001).............................. 278

C-14 Q6 salinity data records for tidal creek (November 2, 2001 -
December 1, 2001) ............... ..........................................279

C-15 Q6 salinity data records for marsh surface water and marsh
sediments (November 2, 2001 December 1, 2001)...........................280

C-16 Q7 salinity data records for tidal creek (November 1, 2001 -
December 10, 2001) ....................... ..........................281

C-17 Q7 salinity data records for marsh surface water and marsh
sediments (November 1, 2001 December 10, 2001)............................282

C-18 Q8 salinity data records for tidal creek (March 29, 2001 -
Septem ber 5, 2001) ............................................ .................... ... 283

C-19 Q8 salinity data records for marsh surface water and marsh
sediments (March 29, 2001 September 5, 2001).............................. 284









Figure page

C-20 Q8 salinity data records for marsh surface water and marsh
sediments (November 1, 2001 December 10, 2001).........................285

C-21 Q9 salinity data records for tidal creek (November 1, 2001 -
Decem ber 10, 2001) ........................................... .................... ... 286

C-22 Q9 salinity data records for marsh surface water and marsh
sediments (November 1, 2001 December 10, 2001).........................287

C-23 Q10 salinity data records for marsh surface water and marsh
sediments (November 14, 2001 December 10, 2001).......................288

C-24 Datalogging station E salinity data records for tidal creek
(March 29, 2001 September 5, 2001) ...................... ........... ..........289

C-25 Datalogging station E salinity data records for marsh surface water
and sediments (March 29, 2001 August 25, 2001)............................290

C-26 Datalogging station W salinity data records for tidal creek
(March 29, 2001 September 5, 2001) ........................... .......... 291

C-27 Datalogging station W salinity data records for marsh surface water
and marsh sediments (March 29, 2001 September 5, 2001)............292

D-1 Section A-A' through K-K' surveyed cross-section locations ...............294

D-2 Section A-A' surveyed cross-section ...................... ...................... 295

D-3 Section B-B' surveyed cross-section ............................ ................... 296

D-4 Section C-C' surveyed cross-section ....................... ...................... 297

D-5 Section D-D' surveyed cross-section ......................................298

D-6 Section E-E' surveyed cross-section ...................... ..................... 299

D-7 Section F-F' surveyed cross-section ........................ ..................... 300

D-8 Section G-G' surveyed cross-section ................................................. 301

D-9 Section H-H' surveyed cross-section ......................................302

D-10 Section I-I' surveyed cross-section ................................ ............ 303

D-11 Section J-J' surveyed cross-section ................... ......... ............... 304









Figure page

D-12 Section K-K' surveyed cross-section ........................ .....................305















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

VEGETATION CHANGE ALONG SALINITY GRADIENTS IN THE TIDAL
MARSHES OF THE UPPER SAVANNAH RIVER ESTUARY

By

John M. Bossart

December 2002

Chair: Clay Montague
Major Department: Environmental Engineering Sciences

The tidal freshwater-oligohaline marsh interface was investigated in the

upper Savannah River estuary. Tidal marsh vegetation, tide stages, and salinity

were monitored from October 1997 through November 2001. Permanent belt

transects for vegetation monitoring were established at ten locations chosen to

bracket the salinity gradient between tidal freshwater and subsaline conditions.

Marsh vegetation was monitored six times between November 1997 and October

2001, and data were collected on frequency and percent cover of each species.

Automatic datalogging stations were used to continuously monitor tide stage and

salinity at 12 locations. Tide stages were monitored both within tidal creeks and

within the interiors of adjacent marshes. Salinity was monitored in tidal creeks, in

high-tide surface waters within adjacent marshes, and in marsh sediments. High

tides were shown to flood the marshes between 33.8 and 100% of the time,

depending on location. Mean salinity in marsh sediments ranged from a low of









0.4 + 0.3%o at the site farthest upriver to a high of 8.1 4.0%o at the site farthest

downriver. However, between October 1997 and October 2001, salinity within

marsh sediments rose at all sites, a trend that was attributed to a 3-year drought

in the Savannah River basin. Ordination of vegetation data defined the

vegetation assemblages of each belt transect and separated them along two

major gradients. The primary gradient was salinity; however, the secondary

gradient remained undefined, possibly indicating an influence of sediment

consolidation and differences in elevation. Comparison of belt transects over the

six separate sample periods indicated a subtle shift to more saline vegetation

assemblages at some sites, a result that was again attributed to the drought

conditions. Salinity distribution across the tidal marshes was determined to have

a strong spatial component associated with proximity to an extensive network of

tidal creeks, the remnants of agricultural water management systems constructed

for the tidewater rice industry in the eighteenth and nineteenth centuries. The

influence of the tidal creek system on the salinity distribution was determined to

have potential use in river management actions that could preserve or increase

tidal freshwater marsh habitat.















CHAPTER 1
INTRODUCTION

The upper Savannah River estuary contains a complex mosaic of tidal

wetlands interspersed among braided river channels. This wetland mosaic

includes tidal freshwater marshes intermingled with low-salinity oligohaline

marshes. These marshes are bracketed upriver by tidal freshwater forests and

downriver by extensive coastal salt marshes. Much of this system, in turn, is

confined laterally by upland bluffs in close proximity to the main channels of the

Savannah River. However, the tidal freshwater marshes occupy a unique

landscape position. They are the product of significant tidal range acting over a

flat elevational gradient and against a large volume of freshwater flow. This

combination of environmental factors restricts tidal freshwater marshes to above

the upriver extent of the estuarine salinity gradient (Odum et al. 1984). These

environmental factors also endow tidal freshwater marshes with the potential for

high production and diversity because the marsh receives the benefits of tidal

import and export (i.e., the "tidal subsidy") without the physiologic limitations

imposed by salt stress (Mitsch and Gosselink 1993).

The landscape components that drive marsh production have also

fostered human productivity along the lower Savannah River. The City of

Savannah was established in 1733 along the upland bluffs lining the river. The

river was a source of freshwater and also provided a protected harbor, which has









since developed into one of the largest ports in the United States. The tidal

subsidy that today supports freshwater marshes previously supported a vast

agriculturally managed landscape of rice fields, which were the source of

immense fortunes for their owners in pre-Civil War America.

Currently, the landscape of the upper Savannah River estuary reflects a

diversity of conflicting land uses and management goals. Taking advantage of

the tidal subsidy, many of the former rice fields are now actively managed for use

by migrating waterfowl. Other fields were simply abandoned and have since

developed into tidal freshwater marsh or reverted to tidal swamp. Some of the

former rice fields have been permanently filled and developed for industrial or

port-related uses. Construction and dredging associated with port expansion and

maintenance continue to alter the tidal and salinity regime of the river, potentially

affecting the interface between the freshwater and saline portions of the tidal

marshes.

Against the backdrop of direct human activity are landscape-level changes

attributed to sea-level rise. Rising sea level and the concomitant upriver

migration of the salinity gradient drive the freshwater-saline marsh interface

landward, promoting a state change in the landscape components (Ricker 1999).

Hackney et al. (1996, pg. 94) stated "vegetation is an indicator of specific

hydrologic and chemical characteristics of established tidal marshes, even if the

mechanism through which this occurs is not clear." Salinity has been shown in

numerous studies to play the key role in the differences between tidal freshwater

and salt marshes (Odum 1988). However, even in water bodies with gradual

salinity gradients, the shift from freshwater to saltwater vegetation along the









gradient "is not a gradual process but occurs in a rather narrow zone of critical

salinity" (Khlebovich 1990, pg. 5). This zone of "critical salinity" is where tidal

freshwater marshes intergrade with tidal oligohaline marshes.

The oligohaline portion of the marsh may be an especially sensitive

indicator of long-term change because of its intermediate position along the

salinity gradient. Activities that can change water levels and salinities within

estuaries include upriver dam construction and reservoir management, waterfowl

management, bridge and causeway construction, navigation improvements such

as jetty construction and channel dredging, and, in the case of the lower

Savannah River, rice field construction and management. It is believed by some

authors that these activities can cause rapid changes (Pearlstine et al. 1990), yet

preliminary observations indicated very few changes after removal of a

controversial tide gate that was constructed in the 1970s. Accordingly, this study

focused on the response of this community to the salinity and water level regime

impinging upon it. The response of the oligohaline community was measured in

relation to that of the freshwater community upgradient and the more brackish

system downriver.

Water levels and salinity in the Savannah River and its associated

channels are subject to dramatic daily fluctuations because of constantly

changing river flows and tide stages. While water levels and salinity are

recognized as major environmental factors in determining vegetation

distributions, constant fluctuation can also be expected in other environmental

factors that may be less easily identified. These undefined factors may play

fundamental, yet unrecognized, roles in driving the self-organization of the









ecosystem. Taken together, these myriad, simultaneous fluctuations can mask

trends that occur over scales from a few years to a decade or more. The

vegetation of the tidal marshes, however, may integrate these diverse

fluctuations and reveal underlying trends. The interface between the tidal

freshwater and oligohaline marshes may potentially provide an especially

sensitive indicator of environmental changes within the river.

Use of the tidal freshwater-oligohaline marsh interface as an indicator of

environmental change is contingent on defining its location and monitoring the

primary influencing factors (i.e., salinity and water levels). Evaluation of the

relationship of the two marsh types in relation to these primary environmental

factors may help to identify additional yet-undefined environmental factors that

also influence the vegetation distributions.

Study Area Description
Location

This study was conducted within the freshwater and low-salinity tidal

marshes of the lower Savannah River (Figure 1-1). Throughout most of its

length, the Savannah River occupies one channel and forms the border between

Georgia and South Carolina. However, approximately 27 miles' upstream of its

mouth, the river becomes an estuarine delta (Day et al. 1989) and divides into

three braided channels named the Front River, Middle River, and the Little Back

River (Figure 1-2). These channels in turn define the large islands Argyle



' The units used in this study are English units and not metric as may generally be expected in an
academic treatise. However, this use of English units reflects the multidisciplinary nature of the
much larger engineering project from which this dissertation is derived. In the overall project,
which concerns deepening more than 40 miles of shipping channel and rerouting flows along
perhaps another 10 miles, English units are the standard.









Island, Ursla Island (also called Isla Island), Onslow Island, and Hutchinson

Island where most of the study area's swamps and marshes are located.

However, additional expanses of swamps and marshes occur on the mainland

margins of both the Georgia and South Carolina sides of the three channels.

The Middle River eventually rejoins the Front River, leaving just two main

channels: the Front River and the Back River. The Back River and Little Back

River form the border between Georgia and South Carolina.

In general, the extent of the study area is defined upstream by the

Interstate 95 bridge across the Savannah River and downstream by the

Savannah River tide gate, a water control structure built across the Back River in

the 1970s by the United States Army Corps of Engineers (USACE) and removed

in 1991-92. The study area is bisected from east to west by the former US-17

(now GA-25 and SC-170), which was constructed in the 1930s. The City of

Savannah is located along the Front River, downstream of the study area.

Estuarine Extent

Day et al. (1989) defined the functional boundaries of a riverine deltaic

estuary as extending from near-shore coastal waters on one end to the upriver

limit of tidal influence on the other. Between these two extremes is the main part

of the estuary, characterized as a mass-mixing zone with strong physical,

chemical, and biological gradients. Figure 1-3 provides a regional overview of

the Savannah River estuary. Figure 1-4 provides a schematic of river miles

along the lower Savannah River. A tidal range of 1 to 3 feet persists at the

United States Geological Survey (USGS) gaging station above Hardeeville,

South Carolina (Figure 1-5), approximately 38 miles upriver (USGS 2001), but no









tidal signal is observed at the gaging station near Clyo, Georgia, approximately

61 miles upriver, indicating that the estuary boundary lies between these two

stations.

River flow volumes within the study area were reported using data from

the Clyo gaging station (Figure 1-6), which has a period of record dating from

1938. Average daily flows computed over the Clyo gage period of record are

shown in Figure 1-7 (USGS 2001). Figure 1-7 also includes the average daily

maximums and minimums, to show the variability on a daily basis. Figure 1-8

compares flows from 1997 through October 2001 to the expected daily average

flow. The daily flows have been lower than the average daily flows over the

period of record (since 1938) because of an extended drought; flows in 1998

were above normal because of the El Nifo weather pattern.

Defining the Salinity Gradient

Within the mixing zone, the salinity gradient is dynamic, and its location at

any given point at any given time is a function of the volume of freshwater flowing

downriver and tide stage (Odum et al. 1984). Salinity will advance upriver with

the incoming tide and retreat downriver with the outgoing tide. Concurrently, high

river flow volumes will push the salinity front downriver, while low flows will allow

the salinity front to advance farther upriver.

Because the estuarine salinity gradient is dynamic, its location is more

conveniently described in statistical rather than absolute terms. In a previous

study of the lower Savannah River (Applied Technology & Management, Inc.

1998. Ecological study of the tidal marshes of the Savannah National Wildlife

Refuge. Prepared for Georgia Ports Authority. 120 pp.), several months of field









data were collected and input to a hydrodynamic modeling to determine the

location of the salinity gradient under different river flow conditions. River flows

of 5900, 8200, and 9500 cubic feet per second (cfs) were considered

representative of average dry season, growing season, and wet season

conditions, respectively. The contours in Figures 1-9, 1-10, and 1-11 represent

50th percentile salinity concentrations under each flow regime (i.e., 50 percent

(%) of the time that salinity value would be located farther upriver, and 50% of

the time farther downriver).

Although the estuarine salinity gradient is a continuum, a tidal marsh

classification based on salinity (Figure 1-12) was developed (Odum et al.1984

and Cowardin et al. 1979). Under this classification system, tidal freshwater

marshes exist in those locations along the salinity gradient where the average

annual salinity is less than 0.5 parts per thousand (%o), except during periods of

extended drought. Oligohaline marshes occupy the zone of 0.5 to 5.0%o, with

mesohaline marshes found in the 5.0 to 18.0%o zone. Using these criteria, the

study area includes tidal freshwater, oligohaline, and mesohaline marshes.

History of the Tidewater Rice Industry Land Management Practices

The tidal freshwater conditions described above were also conducive for

the development of the tidewater rice industry beginning in the mid-1700s

(Richards 1859, Starnes 1886, Rice Association of Savannah 1888, Clifton 1970,

and Clifton 1978, Stewart 1996). Tidewater rice production was established

along restricted portions of some southeastern rivers in areas where both

freshwater conditions and extreme tide range were found. These conditions

were subsequently exploited by the construction of elaborate water management









systems consisting of dikes and levees, distribution canals and ditches, and

water control structures. In an era prior to electric or fossil fuel driven pumps,

these water management systems provided a means to move thousands of acre-

feet of tidally driven freshwater efficiently onto and off of rice fields. The rice

planters were in search of extremely specific conditions:

The rice lands of the Atlantic seaboard occupy the deltas of the
rivers from Pamlico Sound in North Carolina, to the St. Marys River
in Georgia. They are confined in every instance to the fresh
tidewater, the tidal flow being necessary for inundation, and the
water, of course, must be free from salt.

These narrow river strips consequently extend from the extreme
limit of brackish water to the extreme limit of available tidewater, a
distance varying with the volume and location of the rivers. (Starnes
1886, pg. 334)

Historical accounts of development of the tidewater rice industry on the

Savannah and other southeastern rivers indicate that the existing marshes and

swamps of the study area were dominated by tidal forest (see Richards 1859,

Starnes 1886, Rice Association of Savannah 1888, Clifton 1970, and Clifton

1978). For example:

The coasts of Carolina and Georgia afford a stretch of fifty miles
and more of this low swamp land, which, in its primeval condition, is
for the most part occupied by great, dense cypress swamps and
reedy marshes. (Richards 1859, pg. 724)

These descriptions of the tidal forest are augmented by property survey

maps of portions of the study area (Figures 1-13, 1-14, 1-15, and 1-16) that are

clearly marked as tidal forest. In addition, field reconnaissance of the study area

found that the remnant stumps of very large trees, probably cypress, are still

common within the marshes and edges of the tidal creeks.









Beginning in the mid- to late-1700s, the forested swamp system began to

be cleared for development of agricultural fields, which were to be planted and

intensively managed for rice production. Richards (1859) described the initial

step as clearing of the trees in a 50-foot swath around the future field, followed

by the excavation of a ditch (during low tide) in the cleared space. The material

excavated from the ditch was used to make a temporary embankment, or levee,

between the ditch and river, allowing the work area to remain dry during high

tide. The next step consisted of constructing a second and more substantial

embankment within the newly excavated ditch. This placement allowed the

second embankment to have a solid foundation clear of "roots and stumps." This

second embankment, after removal of the temporary embankment, would form

the exterior perimeter of the rice field.

Figure 1-17 provides a conceptualized cross-section of the main

components of a typical rice field including the exterior embankment, margin

ditch, and adjacent main water supply canal. Starnes (1886, pg. 335) provided

dimensions of the exterior embankment (shown in Figure 1-17) as "about five

feet high, with a base of ten feet and a width of four feet." The elevation of the

embankment was "sufficiently high and strong to resist the encroachments of

spring tides and ordinary storms." Richards (1859, pg. 726) described the

dimensions of the embankments as "seven or eight feet" in height, "with base

proportionate."

The area enclosed by the initial exterior embankment was subsequently

cleared of trees by cutting and burning, with some larger trees simply girdled and

left standing (Richards 1859). The enclosed and cleared area was subdivided,









by construction of additional embankments that checked or held the water called

"check banks" (shown in Figure 1-17) into individual fields or "squares" of

manageable size, averaging "seventeen or eighteen acres" (Starnes 1886, pg.

335). These acreage figures are consistent with those obtained through

measurements from rectified aerial photographs (Figure 1-18) of remnant

squares located on Argyle Island. Check banks had the same dimensions as the

main exterior embankments (Richards 1859).

Within its confining embankment, each square was completely surrounded

by a 6-foot wide, 4-foot deep "margin ditch," located 15 to 20 feet inside the

exterior embankment. The rice fields within the square were further ditched with

what were termed "quarter drains, ... one and a half to two feet in depth, usually

seventy-five feet apart," which served to increase the efficiency of moving water

off the planted field (Starnes 1886, pg. 335).

Fields were held dry for planting and harvesting. Water was moved on

and off the fields at various times during the growing season to accommodate

different growth stages of the rice plants, or to control weeds and insects. The

water source for flooding the rice fields was the main river channels. Water was

conveyed from the rivers to each square by a system of main canals excavated

through the former tidal forest (Starnes 1886, pg. 335). Main canals were 20 feet

in width and 5 feet in depth and were fitted with a floodgate at their connection

points with the main river. These floodgates were frequently constructed as

locks to allow boat navigation between the river and the main canals.

The margin ditch on the interior of a square was connected to the main

canal on the exterior of a square via a wooden "trunk" that allowed the water









level in a square to be controlled independently (Starnes 1886, pg. 335). The

trunk, essentially a wooden culvert, was positioned through the levee, connecting

the interior margin ditch to the larger main canals. The trunk was fitted at each

end with height adjustable wooden flap gates and riser boards that provided

control of water flows and levels. When open, the flap gates allowed the

enclosed rice field to be either flooded during high tide or drained during low tide.

When closed, the flap gates allowed the rice field to be kept either dry or flooded

as necessary.

Clearing of the tidal swamp and development of the rice fields occurred

over a number of years, as indicated by a series of historical maps (Figures 1-13

through 1-16). The McKinnon map of 1796 (Figure 1-13) showed that large

tracts on Argyle Island had already been cleared for rice fields, although

extensive tidal forest still remained. Figure 1-14 provides a later (probably c.

1840) map of the portion of Argyle Island contained within the looping meander

on the lower right quadrant of the 1796 map (Figure 1-13). This looping meander

is also prominently depicted along the right side of the 1999 aerial photograph in

Figure 1-18. Note that portions of this area had been cleared and planted in

1839 and 1840, but that a substantial tract remained "In Wood" in 1840. This

same area was also included in the Manigault map of 1867 (Figure 1-16), but

had been entirely cleared of forest by that time, providing a time frame for

clearing of the study area between 1840 and 1867. The C. de Choiseul map of

1846 (Figure 1-15) clearly showed area cleared on northern Argyle Island

accomplished to that date, as well as the remaining tidal forest. The C. de

Choiseul map of 1846 (Figure 1-15) confirmed that clearing of the study area









occurred around mid-1840. Wilms (1972, pg. 55) stated "the period from 1840 to

1860 marked Georgia's 'golden age' of rice production, and it was probably at

this time that the maximum amount of tidewater lands was in rice production."

Land-Cover Changes Associated with the Post Tidewater Rice Industry Era

Historical maps and descriptions of the development of the former rice

fields allow a timeframe to be placed on the clearing of the former forested

swamps. More difficult, however, is placing a timeframe on when the rice fields

were abandoned, and how long the marshes have had to develop into their

current state. Prior to the Civil War, millions of pounds of rice were shipped

annually through the Port of Savannah. However, the war resulted in the

destruction of much of the rice production infrastructure developed over the

preceding century. After the Civil War, the rice industry never regained its pre-

war production capacity and went into a prolonged decline. Although the last rice

harvest occurred in the early 1900s, rice production began decreasing in the late

1800s (Clifton 1970, Stewart 1996).

Several factors have been blamed for the demise of rice production in

Georgia including the number of tropical storms and hurricanes at the turn of the

century, competition from rice plantations in Arkansas, Louisiana, and Texas,

and the inability to use heavy equipment on the unconsolidated soils (Clifton

1970). Plantation records kept by the Manigault family of their two plantations on

Argyle Island document that rice was produced on portions of the island through

at least 1889, at which time plantation records ceased (Clifton 1978). Granger

(1937) noted that rice continued to be planted on the adjacent Ursla Island (Ursla









Island) until approximately 1900. At that time, planting was abandoned on both

Isla Island and Argyle Island because they were

ruined for agricultural purpose when the construction of jetties by the
United States Government in the south channel of the Savannah River
about 1892 so deepened the channel as to render control of the flow over
the rice fields impossible. (Granger 1937, pg. 89)

The inability to control water in the rice fields dealt the final demise of the

remnants of the rice industry. "Dredging of the Savannah River eventually led to

brackish water entering the rice fields and destroying the crop. By 1910, there

were no attempts to cultivate rice" (Wheeler 1998, pg. 118).

Most of the area that constitutes the study area was incorporated into the

Savannah National Wildlife Refuge in the 1930s. Many of the former rice fields

were not completely abandoned but have been managed for migrating waterfowl

since the 1930s when the wildlife refuge was established. However, much of

what had been dense tidal forest in the early 1700s had been cleared,

hydrologically altered through construction of ditches and embankments,

intensively managed for agricultural production for approximately 150 years, and

then abandoned.

Starnes (1886, pg. 335), in his description of tidewater rice plantations,

calculated that a fully developed 640-acre tidewater rice plantation might have

had in excess of 18 miles of embankments. The combined embankments and

ditches would sum to over 118 miles, representing some 317,000 cubic yards of

excavation for a typical 640-acre tidewater rice plantation. Such intensive diking

and ditching, in addition to the loss of the sediments, has served to alter the

environmental gradients that had given rise to the tidal forests completely.









The current vegetation cover is the result of the environmental conditions

that have become established since the time of rice field abandonment. The

remnants of the rice fields' water-management systems persist to this day as

evidenced by the physical presence of trunks, ditches, and canals observed

during field work conducted for this study.

Tidal Creek Development

The abandonment of the rice fields after the demise of the tidewater rice

industry also meant the end of ditch and embankment maintenance, which had

been a constant struggle throughout the rice-growing era. "The ditches are

cleaned out annually, as they foul quite rapidly from abrasion, silt, and water

vegetation (Starnes 1886, pg. 336). Figure 1-19 provides aerial photographs

from 1952 and 1999 of a portion of Argyle Island. The remnants of the main

canals, margin ditches, and embankments are clearly visible and account for the

regular "checker board" pattern.

The extent of the changes that have occurred in the remnant canal and

ditch systems of the former rice field squares is illustrated by comparison of the

1999 aerial photography with historical aerial photographs from 1952 (Figures 1-

20, 1-21, and 1-22). Changes at three locations are compared, indicated as A,

B, C on Figure 1-19.

In 1952, the parallel configuration of the margin ditches was still clearly

discernable at Location A (Figure 1-20). The parallel ditch arrangement reflects

the design that included an embankment between each square, with a margin

ditch constructed within each square near the base of the embankment. As the

images are rectified to state plane coordinates, distances between objects









depicted on them may be measured. The centerlines of the parallel ditches

range from 42 to 47 feet apart. This distance is generally consistent with the

dimensions of embankments and margin ditches provided by Starnes (1886).

At Location B in 1952 (Figure 1-21), one set of parallel margin ditches

(east-west orientation) intersects with a north-south oriented section of a main

canal. By 1999, the north margin ditch had become filled with sediment and

overgrown, its alignment discernable only by the vegetation signature. No

substantial changes in the north-south canal are evident.

In 1952, Location C (Figure 1-22) still had not only remnant margin

ditches, but remnant quarter drains as well. The quarter drains extended

perpendicularly from the margin ditches. The quarter drains were generally gone

by 1999, their former locations detectable via vegetation signatures. Quarter

drains were "one and a half to two feet in depth, usually seventy-five feet apart"

(Starnes 1886, pg. 335).

Marsh Substrate Development

Starnes (1886, pg. 334) described the swamp sediments in which the rice

fields were constructed as

pure alluvium in formation... The soil, in many cases, is ten,
twenty, or even thirty feet in depth to the underlying stratum of
sand. Often the remains of prostrate forests, the result of ancient
hurricanes, with layer of ashes and Indian remains, lie buried in the
alluvium, the logs and stumps frequently so near the surface as to
present a serious obstacle to the ditcher, and greatly enhancing the
cost of reclamation.

Pennington (1913, pgs. 13 and 7, respectively) described the rice field soil as

"moist, dark brown soil, too deep for comfort" with "blue clay which the sun bakes

like a brick." In addition, the sediments of the tidal forest, which then made up









the rice field, seem to have been subject to consolidation and oxidation as a

result of cultivation

the drains imperatively require to be not only thoroughly excavated
in the origin, but to be constantly kept down to their original depth,
and, as the land settles, to be lowered to the same relative depth.
(Starnes 1886, pg. 335)

Heyward (1937, pg. 27) remarked

The fertility of the soil, after years of planting, with little or no fertilizer,
gradually lessened, and the level of the fields sank slightly from year to
year. It has been estimated that through a period of a century and a half
the rice fields of South Carolina and Georgia sank fully a foot, and
perhaps more.

The soil survey for Bryan and Chatham Counties, Georgia (USDA 1974), only

briefly describes the soils of the former rice fields but states that if the marsh is

kept dry for an extended period, the surface will rapidly subside.

Figure 1-17 provided a conceptual cross-section of the embankment and

margin ditch of a typical rice field square as it may have looked during the

tidewater rice era. The horizontal dimensions are based on descriptions of

Starnes (1886) and Richards (1859). The main water supply canals were

approximately 20 feet in width and contained by the 10-foot bottom width

perimeter embankments. The margin ditch is located in the rice field interior

about 15 or 20 feet inside the perimeter embankment. The main water supply

canals are connected to the margin ditches via the rice trunks. Flap gates fitted

at each end of the trunk control water flow through the trunks.

The historical vertical elevations in Figure 1-17 reflect deductions based

on Global Positioning System (GPS) derived elevations of currently existing









conditions. For instance, remnant rice trunks, exposed during low tides, still

protrude from the bases of the former embankments at a number of locations.

During the time of historical operation, these trunks extended beneath the

embankments, connecting the main water supply canals and river channels with

the margin ditches inside the square. The bottom elevation of the trunks was set

so that at low tide the water flooding a square could be completely drained into

the margin ditch and out the trunk. The elevations of four remnant trunks were

determined using GPS survey (Table 1-1).


Table 1-1. Locations and bottom elevations of four remnant rice trunks.
Easting Northing GPS Elevation
Trunk (ft NAD83) (ft NAD83) (ft NGVD29) Comments
1 Little Back River 979800 799414 -0.37 (top) 14 inches thick
2 Little Back River 981286 799102 -0.74 (top) Riverward end of trunk
angled slightly
-2.2 (bottom) downward

3 North tip of 975100 805930 0.0 (top) This trunk is in excellent
Argyle Island at condition with flap
confluence of gates still attached
Middle River and and in working order.
Little Back River It was probably
installed fairly
recently.
4 Front River, near 970909 804442 -0.33 (top) Constructed with hand-
north tip of Ursla forged iron nails.
Island
ft NAD83 = feet North American Datum 1983
ft NGVD29 = feet National Geodetic Vertical Datum 1929
GPS = Global Positioning System

The trunks still had thick (approximately 2-inch thick) wooden planking

attached on both top and bottom. The sides appeared to consist of single pieces

of lumber, set on edge, with what today would be considered a non-standard size

of 2 by 14 inches. The top and bottom planking was attached to the sides by

either wooden pegs or, in one instance, hand-forged iron nails. While some









measurements were of the tops of the trunks, assuming an approximate vertical

dimension of 14 inches, the bottom invert elevations would range from

approximately -1.5 to -2.3 feet (Table 1-1).

Additional information regarding historical topographic elevations within

former rice fields was obtained by survey of former rice fields that were never

fully abandoned and left to the ravages of the tide. As most of the project area is

within the Savannah National Wildlife Refuge, there are a number of the former

rice-field squares that have been maintained for waterfowl management since at

least the 1930s. These "duck-impoundments" have had their dikes maintained

and are regularly drained and planted with forage crops for consumption by

migrating waterfowl (Gordon et al. 1989, Kovacik 1979). Accordingly, these

managed impoundments have not had the sediment accumulation found in the

tidally inundated marshes that now occupy much of the abandoned rice fields

and, therefore, may serve as an indicator of bottom elevations of the rice fields.

GPS survey of these areas found ground elevations within the managed

impoundments of 0.1 to 0.6 feet. These elevations would be consistent with the

elevations determined for the trunks, which would have been set slightly lower

than the surface elevation of the rice field. In addition, the ground elevations

found in the managed impoundments are considerably lower than the surface

elevations of the adjacent abandoned rice fields that have reverted to marsh.

The surface elevations of non-impounded marsh adjacent to the impounded

areas that were surveyed ranged from 4.0 to 4.4 feet, indicating, that in these

areas, some 4 feet of sediments had accumulated within the former rice field

squares.









Other Anthropogenic Perturbations

In addition to the extensive water management systems constructed for

the tidewater rice industry, a number of other alterations to the river system have

occurred over the years since the removal of the tidal forest. On a landscape

scale, beginning with the initial settlement of the colonies of Georgia and South

Carolina, the old-growth upland forests throughout the Savannah River drainage

basin were cleared for forest products and agriculture (Stewart 1996). This

would have had the effect of greatly increasing the sediment load carried

downriver to the estuarine delta and probably increasing the inorganic

composition of the sediment.

Construction of three major dams along the river also affected river flows

and downriver sediment transport (USACE 2002). The J. Strom Thurmond Dam,

completed in 1954, was the first USACE flood control project constructed in the

Savannah River Basin and is located near the City of Augusta at approximately

river mile 240. This dam is "credited with reducing the amount of sediment

carried by the river into Savannah Harbor by 22%" (USACE 2002). Two other

dams are located further up the river, the Richard B. Russell Dam (approximate

river mile 277) completed in 1984 and the Hartwell Dam (approximate river mile

307) completed in 1963 (USACE 2002).

Additional perturbations to the study area are associated with the

continuing development of the Port of Savannah. Table 1-2 provides a summary

of the dredging projects impacting the Savannah River.

The port has been operating since the initial founding of the City of

Savannah during the colonial era. Port development began in earnest though in










the early 1800s with development of the "steam-dredging machine" (Rowland

1987,pg. 132).

Table 1-2. Summary of dredging projects impacting the Savannah River.

Date Description of Project
1733-1850 Various projects, work done when necessary to maintain channel
1873 -90 Channel 22 feet deep at mean high water (MHW) by building a dam at the Cross
Tides
1907-10 Channel 26 feet from the Quarantine Station to the Seaboard Rail Line Bridge
1912 Channel 21 feet from the Seaboard Rail Line Bridge to the foot of Kings Island
1917 Channel 30 feet from the sea to Quarantine Station
1927 Consolidation of projects related to Savannah Harbor, channel 30 feet deep 500
feet wide from the sea to the Quarantine Station, 26 feet deep 400 feet wide to the
Seaboard Rail Line Bridge, 21 feet deep 300 feet wide to Kings Island and
dredging Drakies Cut.
1930 Channel 26 feet deep and 300 feet wide from the Seaboard Rail Line Bridge to
the foot of Kings Island
1945 Deepening the channel and turning basin above the Seaboard Rail Line Bridge
1946 Extending the channel upstream to a point 1500 feet below the Atlantic Coastal
Highway bridge, construct turning basin at upper end
1954 Deepening the channel to 34 feet and widening to 400 feet in the vicinity of the
American Oil Company Refinery wharf to the Savannah Sugar Refinery, with
improvement to the turning basin
1962 Enlargement of the turning basin near Kings Island
1965 Various dredging project including deepening the bar channel and channels by
the wharf and refineries, construct tide gate structure across the Back River,
construct drainage canal across Argyle Island 15 feet deep and 300 feet wide,
control works and canals for supplying freshwater to the Savannah national
Wildlife Refuge
Early-1970s Dredging of McCombs Cut, excavation on New Cut, construction of the tide gate
1976 Modification to turning basins
1984 Construct three new work curve wideners in the inner harbor channel
1986 Under the Water Resources Development Act (WRDA) Savannah Harbor
widening from Fig Island Turning Basin to Kings Island Turning Basin
1991-92 Filling/closure of New Cut, removal of the tide gate
1992-94 Deepening 31 miles of harbor to 42 feet MHW
Source: Applied Technology & Management, Inc. 2001. Tidal amplitude study. Prepared for
Georgia Ports Authority. 37 pp.

The Savannah Harbor project began in 1826, and the steamer Metropolis

arrived in Savannah to begin dredging in 1829 (Rowland 1987). Dredging









facilitated deepening of the navigation channels and manipulation of the course

of the river. The events summarized in this table show that dredging of the

Savannah River occurred almost continuously for more than 150 years since the

mid-1800s. Major dredging projects of the port continued recently with the

deepening of the port to 42 feet below mean high water (MHW) in 1992-94.

Several river oxbows were dredged to facilitate river flows. Drakies Cut

was dredged in the 1927 and McCombs Cut in the 1970s (Figure 1-23). Recent

projects that had great impact were the excavation of New Cut and construction

of the tide gate (1970s), decommissioning of the tide gate (1991), closure/filling

of New Cut (1992), and the deepening of the navigational channel (1992-94).

New Cut connected the Back River with the Middle River and has been opened

and shut more than once.

In the early 1970s, the USAGE constructed the Savannah River tide gate

across the Back River. The purpose of this structure was to reduce the need for

maintenance dredging of the shipping channel within the Front River by

increasing the scour along the river bottom. To accomplish this, the tide gate

was opened during an incoming tide and then closed at high slack tide, just

before the tide began to recede. Closing the tide gate had the effect of

impounding a huge volume of water in the Back River that, to escape as the tide

dropped, was rerouted through New Cut into the shipping channel in the Front

River. The extra water volume in the Front River increased the current velocity

and scour, reducing the need for maintenance dredging.

The tide gate had the unintended consequence of displacing salt water 2

to 6 miles upriver (Pearlstine et al. 1993). During its period of operation from









1974 through 1992, the tide gate is credited with destroying 74% of the tidal

freshwater marshes of the Savannah River (Pearlstine et al. 1990). At the

request of the U.S. Fish and Wildlife Service, the tide gate was removed from

operation in 1991-92 (Latham and Kitchens 1996 and Applied Technology &

Management, Inc. 1998. Ecological study of the tidal marshes of the Savannah

National Wildlife Refuge. Prepared for Georgia Ports Authority. 120 pp.).

Sea-Level Rise

The National Ocean Service maintains an extensive network of tidal

gaging stations including the Ft. Pulaski gage (Station No. 8670870) near the

mouth of the Savannah River (National Ocean Service 2002). This gage has

collected continuous tide stage data since 1 July 1935 and has the longest period

of record for any of the water level gages in the study area (Figure 1-24). A tide

gage must be vertically stable for at least 40 years to be a valid gage for

estimating relative sea-level rise (Dean and Dalrymple 2001).

Figure 1-25 provides the data from the Ft. Pulaski gage plotted as the

annual mean water level. Data from 1973, 1974, and 1990 were not included

because data was recorded less than 50% of the time during those years. The

graph shows a definite increase in relative sea level with the equation of the

trend line indicating an annual increase of 0.0102 feet or 1.02 feet per century.

Based on the Ft. Pulaski data, Hicks et al. (1983) reported a 0.008 feet per

year relative sea-level rise between 1940 and 1980. Hicks et al. (1983) states

this relative rate of sea-level rise includes 0.002 feet of crustal rebound related to

glaciation in the last ice age, 0.003 feet of change in ocean volume due to

warming, and a residual of 0.0001 inches that remains unexplained.









The Savannah area has been subject to substantial land subsidence

resulting from groundwater withdrawals and subsequent decline in hydraulic

head (Davis 1987). This subsidence must be accounted for in order to have a

reliable estimate of relative sea-level rise. However, precise surveys by the

National Ocean Service have shown the Ft. Pulaski gage to be a consistent,

reliable indicator of relative sea-level rise over the period 1940 through 1980

(Davis 1987).

Davis (1987) discussed the extent of land subsidence in Savannah. In his

review, he noted that pumping wells for municipal and industrial water supplies

began in 1887. Water withdrawals were accompanied by artesian head declines

and subsidence. The subsidence was attributed to the settling of fine-grained

sediments in the aquifer, with most of the subsidence occurring after 1933 when

pumping rates were substantially increased. The area of highest pumping and

subsidence identified by Davis (1987) overlaps with the southern portion of the

study area (Figure 1-26).

Based on Davis (1987), potential ground subsidence in the study area

between 1955 and 1975 ranged from approximately 0.049 feet at the northern

end of Argyle and Ursla Islands to more than 0.26 feet near the most downriver

portion of the study area. Subsidence continues at rates of 0.002 to 0.013 feet

per year at these locations (Davis 1987). These rates translate to an additional

0.054 feet of subsidence at the northern end of Argyle and Ursla Islands, and an

additional 0.29 feet of subsidence at the southern end of these islands. In total,

the northern end of the study area has been potentially subjected to 0.10 feet of

subsidence since 1955, with the southern end experiencing a total of 0.55 feet.









Literature Review

Brewer and Grace (1990) characterized oligohaline marsh community

structure in Louisiana. Their study area included distinct vegetative zonation

correlated to distance upriver from the brackish Lake Pontchartrain, with the most

salt tolerant species being found closest to the lake. The vegetative zonation

was not correlated with average soil salinity and was instead attributed to

infrequent, storm-generated salinity pulses that would temporarily raise soil

salinities. The salinity driven upriver by the storm events would attenuate with

distance, resulting in the observed plant zonation. Since the salinity pulses were

temporary, soil salinities would decrease to their former lower levels. Their study

did not address the sediment salinity levels generated by the storm pulses, the

duration of elevated salinity, or the amount of time required for sediment salinities

to drop to their previous levels; however, the salt pulses were characterized as

short-term. Salt tolerant species that temporarily flourished as a result of the salt

pulses would be gradually replaced by less salt tolerant, but more competitive,

species as the time between salt pulses increased; however, the authors

suggested this replacement would occur over a time scale of years or decades,

not seasons.

Howard and Mendelssohn (2000) conducted a greenhouse study of

oligohaline marsh community structure that examined the interaction of salinity

exposure and water depth. A 3-month salinity exposure at 12%o with concurrent

flooding to either 1- or 15-cm resulted in community changes. Changes did not

occur with only 1-month salinity exposure.









Perry and Hershner (1999) studied temporal shifts in vegetative

dominance over a 14-year period in tidal freshwater marshes on Chesapeake

Bay. Average yearly salinity at the site was approximately 0.45%o and ranged

from 0 to 7%o. The study found an increase in oligohaline-associated species,

particularly Spartina cynosuroides. An increase in oligohaline conditions was

attributed to a relative sea-level rise of 4 mm per year. Perry and Hershner

(1999) cited the need for studies on the inundation frequency and salt tolerance

of individual species in order to predict the rate at which community changes

would occur in response to increasing salinity.

Pearlstine et al. (1990), Latham (1990), Latham et al. (1991), Latham et al.

(1994) reported various aspects of a previous vegetation study of the tidal

freshwater and brackish marshes of the Savannah National Wildlife Refuge. This

study, most comprehensively described in Pearlstine et al. (1990) and Latham

(1990), examined the effects of tide gate operation on marsh vegetation

distributions along the salinity gradient on the Little Back River and Back River.

According to this study, tidal freshwater marshes existed downriver to the tide

gate and were replaced by brackish vegetation assemblages as a result of tide

gate operation. The study was initiated in 1985, 8 years after the tide gate began

to operate in 1977, and found the existing vegetation distributions to be

correlated with sediment salinity, distance from river channels and tidal creeks,

and ground elevation. The goal of the study was to predict vegetation changes

that would occur after removal of the tide gate and the subsequent return of

sediment salinity to levels conducive to the reestablishment of tidal freshwater









marsh. The study concluded that reestablishment of tidal freshwater marsh

would occur rapidly after sediment salinities decreased to below 0.5%o.

Latham and Kitchens (1996) reported the successful reestablishment of

freshwater vegetation after the tide gate was removed in 1992. Effects of the tide

gate are discussed in detail in Pearlstine et al. (1993). During the time the tide

gate was operating, Latham et al. (1994) found the change from freshwater to

brackish vegetation along the lower Savannah River to be related strongly to the

salinity gradient. However, their study was not sufficient to explain vegetative

distributions within freshwater areas and suggested interspecies competition to

be a primary factor.

Many studies have been conducted on salt marsh vegetation (see

Montague and Wiegart 1990 for a comprehensive review). The salt tolerance of

plants under laboratory or greenhouse conditions has also been studied

extensively. Broome, et al. (1995) conducted a greenhouse study of Louisiana

marsh plants and Scirpus olneyi to determine the effects of salinity and water

depth on these two species. Based on their results, salinity greater than 10%o

reduced growth of both species, but Scirpus olneyi was more affected than

Spartina patens. Increased flooding depth reduced growth of Spartina patens,

but had little effect on Scirpus olneyi. Baden et al. (1975) cited salinity as the

primary factor in the distribution of vegetation in abandoned rice fields in South

Carolina. Allen et al. (1997) conducted a greenhouse study of baldcypress

seedlings in Louisiana and concluded that increasing salinity reduced leaf

biomass more than root biomass. Flowers et al. (1977) studied mechanisms by

which the naturally occurring halophilic flora survive, including growth, uptake,









and accumulation of salt. The effects of increased water depths (up to 0.5 feet)

on Sagittaria lancifolia were studied with little or no impacts to the plants (Howard

and Mendelssohn 1995). Howard and Mendelssohn (2000) also conducted

greenhouse experiments using pulsing salinities on oligohaline marsh plants and

found that duration of salinity exposure and water depth determined whether

existing vegetation recovered or new species were established.

In studies conducted by Baldwin and Mendelssohn (1998), oligohaline

plants Spartina patens and Sagittaria lancifolia were not significantly affected by

flooding or salinity unless "disturbance" (clipping of aboveground vegetation)

occurred.

Visser et al. (1999) studied development impacts to oligohaline marsh

associated with the Atchafalaya River of Louisiana using 5 permanent vegetation

stations surveyed over 24 years. Cluster analysis of vegetation data was

conducted using two-way indicator species analysis (TWINSPAN). Johnsson

and Moen (1998) discussed the effects of belt transect size when establishing

vegetation sampling plots.

Kent and Coker (1992) defined a plant community as the collection of

plant species growing together in a particular location that show a definite

association or affinity with each other (i.e., they are found growing together in

certain locations and under certain environmental conditions more frequently

than would be expected by chance). The plant association is the reflection of the

environmental conditions, or collection of environmental factors, that define the

living requirements, or restrictions, under which the association is found. In the

case of the tidal freshwater and oligohaline marshes of this study, these









environmental factors include, but are not limited to, salinity and the hydrologic

parameters of flood depth, duration, and frequency.

As commonly depicted in plant ecology literature (see Whittaker 1975,

Gauch 1995, Kent and Coker 1992, Jongman et al. 1995), the abundance of an

individual plant species along a gradient of a single environmental factor can

hypothetically be plotted as a Gaussian curve. Under the hypothetical curve,

also called a unimodal model, species abundance increase along the gradient to

some peak level and then begins to decline as the intensity of the factor

increases to a point where it induces stress and ultimately intolerance in the

plant. This type of response prevents analyzing for a direct linear correlation of

plant abundance with an environmental factor. A positive correlation may be

found along one portion of the gradient, and a negative correlation along another.

Additionally, the abundance of a single plant species at a particular

location is usually the integration of multiple gradients acting simultaneously

(Kent and Coker 1992). Acting alone, each gradient would induce an abundance

response curve unique to a particular plant species. However, under actual

conditions, the abundance of an individual plant species at a belt transect is

dependent on the belt transect's position in relation to all the gradients; with

some gradients exerting more influence than others. Consequently, the

abundance of a species at a particular location can be thought of as the

intersection of all the individual response curves for that species for all the

environmental factors that define the habitat at that location. For this reason,

plant community data are multivariate and not subject to analysis by linear

regression (Kent and Coker 1992).









As a mulitivariate technique, ordination identifies relationships between

species distributions and the distributions of associated environmental factors

and gradients (Kent and Coker 1992). Indirect ordination techniques, particularly

detrended correspondence analysis (DCA), use only species abundance data

and determine the existence of species association patterns within that data.

Plots produced by DCA group species according to the strength of their

association with one another and separate the groups along one or more

gradients. However, these gradients represent only underlying patterns derived

from the plant species data and are only indirectly associated with an actual

environmental gradient through inference by the researcher.

Detrended canonical correspondence analysis (DCCA) is a direct

ordination technique because it simultaneously considers both species and

environmental data (Kent and Coker 1992, Jongman et al. 1995). DCCA allows

the relative importance of the different environmental variables to be assessed by

determining what combination of the environmental variables best explains the

species variation.

Hypothesis and Approach

Vegetation distributions along the river channels of the upper Savannah

River estuary display distinct zonation that previous studies (Latham 1990,

Pearlstine et al. 1990) have attributed to salinity. The objective of this current

study is to characterize the spatial and temporal differences in the salinity and

tidal gradients between the Front, Middle, and Little Back Rivers, including the

influence of the tidal creek system on distributing salinity across the marsh

surface. The hypothesis to be tested is that substantial differences in salinity are









normally present across the marsh surface and within the marsh sediments,

differences that are manifested in plant species associations present at any

location within the marsh. An unknown factor to be explored is the variability of

sediment salinity at a given location. Salinity levels within the river channels and

high tide flood waters are highly variable, depending on river flow volumes and

tide stage. However, since salinity effects on marsh plant physiology are largely

manifested through the roots, it is the salinity within the sediments surrounding

the plant roots that is the driving factor in plant distributions. Because plant

distributions are generally relatively stable in the absence of major anomalous

disturbances, it is hypothesized that sediment salinity has less variability than the

highly variable salinity found in the adjacent tidal surface waters and that the

sediment salinity represents an integration of the surface water salinity variability.

However, under what circumstances are transient sediment salinity increases

generated, how long do they persist, and what effect may they have on the

overlying vegetation assemblages? Is short-term variability of salinity levels

within the marsh sediments more important in affecting plant species distributions

than long-term trends in sediment salinities?

The project approach involves simultaneous measurements of salinity

and water levels at representative locations across the marsh to determine any

spatial differences that may be expressed as ecological gradients. Figure 1-27

provides a conceptual approach to a sample design that allows salinity in the

river channels to be linked to salinity in marsh sediments, where the physiologic

effects to plants originate. In Figure 1-27, the source of salt within the marshes is

the ocean salinity carried up the river channels by the rising tide. The salinity






31


gradient develops from the interplay of river flow volume and tide stage. While

the river channels overflow their banks during high tide, flooding the marsh, the

tidal creek system greatly increases the interface between the open water and

the vegetated marsh. In addition, the tidal creeks serve as a conduit for saline

waters from a specific reach of a river channel to be transported across a

disproportionate area of marsh surface. Analysis of the gradients for their

influence on vegetation distribution requires seasonal vegetation monitoring at

established locations. Vegetation monitoring allows plant species abundance

and population structure to be described quantitatively and over time.











80o'O"W


360'0"N "\ 360'0"N
NORTH CAROLINA






SOUTH
CAROLINA




S\ GEORGIA ATLANTIC OCEAN

320'O"N Study Area 320'O"N




SSavannah River Drainage Basin


840'0"W 8000O'"W 7600'O"W



Figure 1-1. Regional map of Savannah River drainage basin with study area.


84"0'0"W


76o0'O"W

































































Scale
SMles
2 0 2


Figure 1-2. Project study area.







975000


900000
























800000
























700000


975000 1075000

US State Plane, Georgia East Zone, NAD83, US Survey Feet


Figure 1-3. Landsat 7 satellite image showing extent of Savannah River estuary.


900000
























800000
























700000


1075000






35





60







50
South Carolina
















Georgia o

30



Middle River
25


Front River Little Back River

0
20
0

Back River

5
Legend 15
40 Mile point from river mouth



Figure 1-4. Savannah River miles.
























I
04








2
1-Feb 5-Feb 9-Feb 13-Feb 17-Feb 21-Feb 25-Feb 1-Ma


7
USGS gaging station Savannah River above Hardeeville. SC (02198760)


6








84



3



2
1-Feb 5-Feb 9-Feb 13-Feb 17-Feb 21-Feb 25-Feb 1-M



Figure 1-5. Comparison of gage heights at the USGS gaging stations at
Hardeeville, SC and Clyo, GA during February 2002 showing loss of tidal
signal between the two stations.


tar


r














140000


*I





S


J.. 1i S 1 1
I 1 1 t i i ,


120000


100000


S80000


o 60000


40000


20000


0


1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000


Figure 1-6. Flow volumes of the Savannah River recorded at the USGS gaging station near
Clyo, GA (02198500) for the period of record, 1938 through October 2001.











50000 M a x I I i
50000


45000


40000 --------



I!I
35000 C---L---------------^------ .-
35000
Maximum



| M___ o m 7 \ \ li L y
30000







15000, Meian--.95. Confidence Inierval


10000
Minimum -.
5000



J F M A M J J A S O N D
Month

Figure 1-7. Average daily flow (cubic feet per second) of the Savannah River as recorded at the USGS
gaging station near Clyo, GA (02198500) during the period of record, 1938-1999.












SUOOU q


4UUUU -
Flow 1997 2002




30 Daily Average (+/- 95% Confidence Interval) since 1938


20000




10000





1997 1998 1999 2000 2001 2002
Figure 1-8. Recent average daily flow (cubic feet per second) of the Savannah River
compared to the average daily flow for the period of record (1938 to present) at the USGS
gaging station near Clyo, GA (02198500) from 1997 through October 2001 .













970000 980000

810000 .810000
0.0




00




800000 800000







0.0

0.1
790000 0.2 790000

0.4
0.5
1.0


2.0



780000 780000



0.0 Salinity (ppt)


970000 980000

U.S. State Plane. Georgia East Zone. NAD83/90. U.S. Survey Feet






Figure 1-9. Typical growing season salinity probability contour at river
flow of 8,200 cfs. Contours represent location of 50th percentile
salinity gradient when tide stage is above 4.5 feet.













970000 980000

810000 810000


03







800000 800000





0.1





790000 790000
0.5


1.0

1.5



780000 780000



0.0 Salinity (ppt)


970000 980000

U.S. State Plane. Georgia East Zone. NAD83. U.S. Survey Feet



Figure 1-10. Typical dry season salinity probability contour at river
flow of 5,900 cfs. Contours represent location of 50th percentile
salinity gradient when tide stage is above 4.5 feet.












970000 980000

810000 0.0 810000










800000 800000










0.0
790000 0.2 790000



0.5






780000 1.0 780000


0.0 Salinity (ppt)

970000 980000
U.S. State Plane, Georgia East Zone, NAD83, U S Survey Feet



Figure 1-11. Typical wet season salinity probability contour at river
flow of 9,500 cfs. Contours represent location of 50th percentile
salinity gradient when tide stage is above 4.5 feet.























AVERAGE ANNUAL
SAUDTY


URGINGG
LOW
ROW
CONOWTNS)


EUHALNE
PAFA*)


Figure 1-12. Conceptual schematic of tidal marsh classification based on
salinity (after Odum et al. 1984 and Cowardin et al. 1979).








































Figure 1-13. Historical map (1796) of a portion of Argyle Island drawn by John McKinnon.
(Courtesy of the Georgia Historical Society).























































Figure 1-14. Historical map (c.1840) of a portion of Argyle Island. The map
is oriented with north to the left. The original map is drawn in the margin
of an accounting ledger in the Manigault Plantation Records, Southern
Historical Collection, University of North Carolina at Chapel Hill.





















































Figure 1-15. Historical map (1846) of Redknoll Plantation on the northern
portion of Argyle Island drawn by C. de Choiseul. (Courtesy of the Georgia
Historical Society)









































Figure 1-16. Historical map (1867) of the Gowrie and East Hermitage Plantations on Argyle Island drawn
by either Charles or Louis Manigault. The original map is contained within the Manigault
Plantation Records, Southern Historical Collection, University of North Carolina at Chapel Hill.













EXTERIOR
EMBANKMENT
OR
CHECK
RICk RELD BANK MAIN CANAL
MARGIN
DITCH




-EL = 8.0
---l



EL = 2.0-- ,
EL -2 C _EL =

-10 20'


Figure 1-17. Conceptual cross-section of typical rice field water management system.











976000


795000 g44: 795000








*r I!'r :
J6
1 4-






%












4f. .
$' ; -



.. ""'L:= ,,



*








790000 j# ,~ ~790000










976000 981000

US State Plane, Georgia East Zone, NAD83, US Survey Feet








































Figure 1-18. Infrared aerial photograph (1999) of a portion of Argyle Island.


981000








975000 980000


970000


795000 .1 795000










790000 \ 790000










785000 785000



970000 975000 980000 985000


973000 978000 983000








796000 796000







78600.. 78.60.
*' ..... .



786000 796000










799000 7300097800098


985000













978000 977500


797500


797400


797500












797000


796900


977500 978000


1952


977500 978000
US State ane Coordinates, Georgia East Zone, NAD83. US Survey Feet


1999


Figure 1-20. Location A change analysis of margin ditches (as noted on Figure 1-19).


797000


797400












796900


977500


978000












978600 978100


793100

793000








792600

792500


A- 793100


792600




978100 978600
US State Plane Coordinates, Gerga East Zone, NAD83. US Survey Feet


1999


Figure 1-21. Location B change analysis of margin ditches (as noted on Figure 1-19).


793000










792500


978100



1952


978600


978100


978600












980500 980000


787200


787000


786700


786500


980000 980500


1952


980000 980500
US State Plane Coordinatae Georoia East Zone. NAD83 US Surev Fa.t


1999


Figure 1-22. Location C change analysis of margin ditches (as noted on Figure 1-19).


787000






786500


787200






786700


980000


980500











1.000.000


810,000


760,000


810,000


760.000


970.000 1,000.000
US Sate Plan. Georgia Ea- Zone. NAD83. US Suy mFet




Figure 1-23. Channel modifications affecting downriver freshwater flow
and upriver salinity transport.


970,000


























































Station No. Station Name 0 Mies 5

US Geological Survey Stations
02198500 Savannah River near Clyo. GA
02198760 Savannah River above Hardeeville, SC
02198840 Savannah River (1-95) near Pt. Wentworth, GA
02198979 Little Back River (Lucknow) near Limehouse, SC
02198920 Savannah (Front) River (US-17) near Pt. Wentworth. GA

National Ocean Service Station
8670870 Ft. Pulaski. GA

Figure 1-24. Tide gage locations on the lower Savannah River.










1.00

0.90

0.80

* 0.70

0.60

S0.50
Sy = 0.0102x -19.72
0.40

2 0.30

0.20

0.10

0.00
1930 1940 1950 1960 1970 1980 1990 2000 2010
Year

Figure 1-25. Yearly mean sea-level at the Ft. Pulaski, Georgia gage (Station No. 8670870) on the
Savannah River. Data from NOAA, National Ocean Service 2002.











970.000 1.000,000

810,000 I )810,000










Area of greater than 0.066 feet of
S subsidence between 1933 and 1975











I ,'
,,i, _, d: h. .




















970,000 1,000,000
US State ane Go -, N 83. US Suy



Figure 1-26. Land subsidence since 1933 resulting from municipal
and industrial water withdrawals of groundwater in proximity to the
study area (after Davis 1987).














Long-term
Vegetation
Changes


Tidal Marsh
Creeks Flood Water
River Channels



Marsh Root Zone


Figure 1-27. Linkages between river channel salinity and marsh root zone
tracked by salinity monitoring protocol.















CHAPTER 2
METHODS

Mapping and Surveying

Aerial Photography Acquisition

Growing-season aerial photography was acquired during August 1999 by

contract with a commercial aerial survey company (Aerial Cartographics of

America, Inc., Orlando, Florida). Both true color and false-color infrared

photographs were acquired at scales of 1:12,000 and 1:25,000; Figures 2-1 and

2-2, respectively, provide the flight lines and center locations for each scale.

Photograph overlap along the flightlines is 80%.

To rectify the aerial photographs, targets were placed across the project

area at ten locations specified by the contractor (Figure 2-3). Each target

consisted of a white cross, 20-feet in width, made from 48-inch wide white plastic

aerial flagging. Coordinates of each target center were determined using a

differentially corrected GPS (Trimble Pro-XR GPS, Trimble Navigation, Inc.,

Sunnyvale, California). This model GPS receives a U.S. Coast Guard correction

signal that is used to provide real time coordinates accurate to within 3.28 feet.

The coordinate system used for the entire project was U.S. State Plane, Georgia

East Zone, North American Datum (NAD) 1983 (with the 1990 correction), with

units in U.S. Survey Feet.








Target coordinates were provided to the aerial survey contractor for

subsequent preparation of rectified photography. The contractor supplied

rectified images prepared by scanning 9- by 9-inch positive transparencies

developed from the original 9- by 9-inch negative. Using the triangulated target

coordinates, the scanned 1:25,000 scale infrared aerial photographs were

rectified and digitally joined together to form a single photomosaic that covered

the entire project area. The mosaic of the 1:25,000 aerial photographs was used

as the reference image for rectification of selected 1:12,000 scale true color and

infrared images.

River Channel and Tidal Creek Mapping

A detailed base map of river channels and tidal creeks (Figure 2-4) was

prepared by on-screen digitizing from the rectified 1:12,000 and 1:25,000 scale

aerial photographs. The digitizing was done in AutoCAD Map 2000, release 4

(AutoDesk 1999). This base-map was compiled into a shape file using ArcView

3.2 (Environmental Systems Research Institute, Inc., Redlands, California). For

information on the use of ArcView and its terminology see ArcView GIS Exercise

Book, Second Edition (Hohl and Mayo 1997).

Aerial Photograph Interpretation of the Tidal Creek Network

Photointerpretation and measurements of the existing tidal creek network

were conducted using two sets of aerial photography of Argyle and Ursia Islands.

One set was historical aerial photographs dated from 1952 that were obtained

through the U.S. National Archives and Records Administration in College Park,

Maryland. The other set was the 12,000 scale 1999 false-color infrared aerial

photographs described above. Both sets of photographs were digitally scanned









by a commercial service (Aerial Cartographics of America, Inc., Orlando, Florida).

The scanned photographs were rectified against the 1:25,000 scale base image

using ERDAS Imagine. Tidal creek lengths as existing in 1952 were measured

for comparison to the length of the same creeks as photographed in 1999.

Length measurements were made using Imagine's measurement tools. The

cumulative length, in linear feet, of each tidal creek system was measured as a

comparative index of the extent of tidal creek development. These

measurements included all channels and ditch remnants that comprised a

particular tidal creek system.

Survey Instrumentation and Tidal Creek Cross-Sections

Cross-section elevations of tidal creeks were surveyed at 11 locations

using a survey grade global positioning system (GPS) (Trimble Model 4800 RTK

GPS, Trimble Navigation, Inc., Sunnyvale, California). This instrument provided

vertical accuracy to 0.066 feet. All elevation data for the project were referenced

to National Geodetic Vertical Datum (NGVD) 1929. Care was taken to ensure

that the instrument was resting on the marsh sediment surface and not on raised

areas such as root clumps or rhizomes. In addition, sediments of marsh interiors

can be soft and depress under weight. Consequently, care was taken to not to

disturb the marsh surface prior to obtaining an elevation reading.

Vegetation Studies

Field Surveys

To monitor vegetation, permanent belt transects were established at ten

locations, labeled Q1 throughQ10, across the study area (Figure 2-5) in fall 1997.

All belt transects were 2 feet in width. Nine belt transects were 500 feet in length









and one (Q3) was 600 feet in length. Q3 was extended an additional 100 feet to

incorporate a Spartina alterniflora dominated area.

Transect locations were chosen to bracket the salinity gradient from

freshwater to mesohaline, with an emphasis on the oligohaline-freshwater

interface. Table 2-1 summarizes the general locations of the belt transects in

relation to their associated river channels. River miles are measured from the

mouth of the Savannah River. To highlight the position of each belt transect in

relation to the salinity gradient, the transects are arranged from upriver to

downriver.

Table 2-1. Locations of belt transects along their associated river channels.
Savannah River Mile Salinity Regime
Front River:
Q1 23.5 freshwater
Q7 22.0 oligohaline
Middle River:
Q9 24.0 freshwater
Q6 23.5 freshwater
Q5 22.5 oligohaline
Q10 21.5 oligohaline
Little Back River:
Q8' 24.5 freshwater
Q4b 21.5 oligohaline
Q30 20.5 oligohaline
Back River:
Q2d 17.0 mesohaline
aFormer Study Area 1 of Pearlstine et al. (1990) and Latham (1990)
bFormer Study Area 2 of Pearlstine et al. (1990) and Latham (1990)
cFormer Study Area 3 of Pearlstine et al. (1990) and Latham (1990)
dFormer Study Area 4 of Pearlstine et al. (1990) and Latham (1990)

However, these four sampling locations are all located along the Little

Back River and Back River. To include marshes associated with the Middle

River and Front River, six additional belt transect locations were selected in fall

1997 based on a preliminary field reconnaissance of the Front River and Middle

River, where a number of salinity measurements were obtained during high tide









using a hand held conductivity meter (YSI Model 30, Yellow Springs Instruments,

Inc., Yellow Springs, Ohio). This preliminary work provided an estimate of the

upriver extent of the salinity gradient, as it existed at that time on the two

additional river channels, and the six additional belt transects were located to

bracket the salinity gradient from freshwater upriver to brackish downriver.

For each belt transect, both ends and the intermittent 100-foot points were

permanently marked with 10-foot long iron rebar stakes driven into the marsh

sediments and, for visibility, covered with a stave of white PVC pipe. The x and y

locations of all stakes marking the belt transects were determined by GPS.

Ground elevations along the entire length of each belt transect were also

determined by GPS. Mean marsh surface elevations were calculated based on

these belt transect elevation surveys.

Herbaceous vegetation was quantified along the entire length of the belt

transect using a line-intercept method modified from Phillips (1959) and

described in Wallace (1996) (Wallace, P. M., R. A. Garren, and D. R. Rich.

1996. Ecology of natural wetland communities in the Orange County eastern

service area reclaimed water wetland system. Final report to the Orange County

Public Utilities Division. Ecosystem Research Corporation, Gainesville, Florida.

358 pp.). Each belt transect was divided into contiguous 10-foot intervals along

its length as illustrated in the top of Figure 2-6. Vegetative cover was assigned

by species within each 10- by 2-foot cover interval using the scale given in Table

2-2. Cover is defined as "the area of ground within a belt transect which is

occupied by the aboveground parts of each species when viewed from above"

(Kent and Coker 1992). Although cover is usually estimated by visual observation









as a percent, total cover values can exceed 100% due to stratification or multiple

layering of vegetation (Kent and Coker 1992).

Table 2-2. Cover value categories and percent cover ranges for each category.
Cover Category Percent Cover Range
0 0
1 <1
2 1 -10
3 10-30
4 30-50
5 50-70
6 70-90
7 90-100

Species frequency was determined as presence or absence along each

foot of the belt transect, so a maximum frequency of 10 was possible for each

10-foot interval. For example, if a given species was present in any three 1-foot

segments of a given 10-foot interval, the frequency for that species for that

interval would be 3.

Vegetation Analysis

Raw data for a belt transect were compiled in a rectangular data table with

one row of data for each species found in the belt transect. Columns in the table

corresponded to each of the 10-foot intervals along the belt transect and were

further subdivided into sub columns in which the frequency and cover data for

each species was entered. The data were summarized using the following

statistical methods:


Total Frequency = the total number of 1-foot segments in the entire belt
transect where the species occurred. The maximum value is 500 for a 2
by 500-foot belt transect and 600 for the 2 by 600-foot belt transect.









Relative Frequency = the total number of 1-foot segments where the
species occurred divided by the sum of the total frequencies of all species
x100.

Percent Cover = sum of cover assignment for each species within each
10-foot interval divided by the number of 10-foot intervals by 100.

Relative Cover = the percent cover of a species divided by the sum of
percent cover of all species x 100.

Importance Value = the sum of relative frequency and relative cover.
The maximum value possible is 200 for an individual species.

Frequency Rank = the numerical rank of the species within the plot based
on the relative frequency of each species. A rank of 1 indicates the
species occurred more frequently than any other.

Cover Rank = the numerical rank of the species within the plot based on
the percent cover of each species. A rank of 1 indicates the plant covered
more area than any other plant.

During the study, vegetation within the ten belt transects were sampled six

times: October 1997, October 1999, May 2000, October 2000, June 2001, and

October 2001.

Concurrent with the collection of vegetative data, qualitative observations

were recorded regarding the stability of marsh sediments. At some locations,

sediments were well consolidated and easily supported weight. However, at

other locations, sediments were unconsolidated and unable to support weight.

The ability to walk in these areas depended on whether or not a vegetative root

mat was present.

To facilitate interpretation of the data in relation to the numerous

environmental gradients that exist, data analyses consisted of cluster analysis,

detrended correspondence analysis (DCA), correlation analysis, and detrended

canonical correspondence analysis (DCCA).









Cluster analysis was used to classify belt transects according to

vegetation similarity and to detect change in floristic composition of the belt

transects over time. The analysis was conducted using a commercial statistical

ecology software package (Pisces Conservation, Ltd. 2000. Community analysis

package, version 1.1. United Kingdom.). Some preliminary runs indicated that

data analysis was more manageable if only those species with importance values

greater than 10 were included. This reduced the number of species included in

the analysis from 150 to 29. While rare species were eliminated, reducing the

size of the data set made only small differences in the absolute similarity

between each belt transect, and no difference in the final analysis. Cluster

analysis is performed only on vegetation species data and does not include

environmental data. However, the cluster results were interpreted with respect to

general gradients that exist among belt transect locations.

DCA, an indirect ordination technique, was used as the first step in a

process to quantify the underlying environmental gradients within the existing

vegetation data and was conducted using the same software package as cluster

analysis. As with cluster analysis, only floristic data were used in DCA and, in

this case, consisted of importance values calculated from relative frequency and

cover statistics. For the same reason as cluster analysis, only species with an

importance value greater than 10 were included in the DCA. With DCA, a set of

belt transect and species similarity scores were generated based on the floristic

data. These data were then plotted on a set of axes in which the primary axis (x-

axis) represents the strongest separation of plant data, presumably according to

an underlying environmental gradient.









In contrast to the indirect approach of DCA, DCCA is referred to as a

direct ordination technique since it directly regresses gradients in community

species composition to environmental factors such as salinity or elevation.

DCCA was conducted using a commercial statistical ecology software package

(CANOCO 4, Microcomputer Power, Inc., Ithaca, New York).

Like DCA, DCCA calculates species scores and sample scores and

calculates one or more ordination axes that explain the variation in the species

data. The first axis explains the greatest percentage of the variance. The

second axis explains the next greatest percentage of the variance while being

uncorrelated with the first axis. Additional axes may be calculated as well.

Unlike DCA, in DCCA the ordination axes are constrained to be linear

combinations of environmental variables (Kent and Ballard 1988). This allows

the environmental factors to be regressed against the ordination axes. This is

accomplished during the calculation of the species ordination axes by

simultaneously calculating multiple linear regressions of the available

environmental variables to find a combination that is most highly correlated with

the ordination axes. The multiple linear regression of the environmental data that

is most highly correlated with the first species axis is designated as the first

environmental variable. Additional environmental variables are derived for the

remaining species axes as necessary, with the constraint that the environmental

axes are uncorrelated with one another.

The significance of the regressions derived by the DCCA is tested by a

Monte-Carlo permututaion. When the p-value is less than 0.05 the relation









between the species axis and the environmental variable is significant at the 5%

level.

Kent and Coker (1992) described the components of the DCCA diagram,

also called a species-environment biplot. Superimposed on the plant community

gradient ordination plot, an environmental variable is represented by an arrow

pointing in the direction of maximum change of that variable. The more parallel

the arrow is to either axis, the more highly correlated the environmental variable

is to that axis. A longer arrow represents a greater magnitude of change and is

therefore more important in influencing community variation.

The relationship of a species or a sample to the environmental variable

can be determined by projecting a perpendicular line between the arrow and the

plotted point representing the sample. Samples that have their perpendicularly

projected points falling near or beyond the tip of the arrow are strongly correlated

with the environmental variable. The farther away the projected points fall from

the tip of the arrow, the less the samples represented by the points are

influenced by the environmental variable. Arrows oriented orthogonal, or

perpendicular, to one another are highly uncorrelated. The environmental

variables represented by orthogonal arrows can therefore be highly influential in

separating the samples along the first and second axes (CANOCO reference

manual and user's guide: software for canonical community ordination, version 4.

Microcomputer Power, Ithaca, New York. 352 pp.).

DCCA was used to compare belt transects to one another to determine

environmental factors significant in affecting plant distributions from upriver to

downriver locations. The environmental factors used in these analyses were









average salinity within the belt transect, average ground elevation of the belt

transect, and the depth, frequency, and duration of tidal flooding. A second belt

transect analysis was conducted using ranked data for salinity, where the belt

transects were ranked from 1 to 10 based on average salinity, as well as ranked

data for elevation.

DCCA was also used to compare species distributions within each belt

transect. For these analyses, the data set reflected the species abundance and

environmental factors at each 10-foot interval along the belt transect.

Environmental factors comprising the data set were location (distance) of each

10-foot interval along the belt transect, elevation of the interval, average salinity

within the 10-foot interval, and temporal standard deviation of the average

salinity. Elevation data files used in the within belt transect DCCA were

developed from the GPS surveys of each belt transect by extracting elevations at

10-foot intervals from cross-section drawings.

The salinity data files used in the within belt transect DCCA were

developed from sediment salinity measurements taken along the belt transects

during the six vegetation monitoring events. During each sampling event, salinity

measurements were taken at 50 or 100-foot intervals along the belt transects.

These data were averaged for the multiple sample times and linear interpolation

was used to create data files with average salinity values at 10-foot intervals.

Correlation analysis is used to determine the "strength of relationships

between variables" (Kent and Coker 1992, pg. 134). Correlation analysis

attempts to relate the vegetation species variation with different environmental









variables, such as salinity and elevation. The degree of relationship between two

variables lies between -1.0 through 0.0 to +1.0.

Two methods of correlation analysis were used to relate the data:

Pearson's product-moment correlation coefficient parametricc) and Spearman's

rank correlation coefficient (non-parametric). The significance of the correlation

(r) using both methods is reflected in the results of a "t" test, which expresses

how strongly the vegetation responds to the environmental data. "The

significance test is designed to calculate the probability that for the given sample

size, the correlation coefficient could have been derived by chance" (Kent and

Coker 1992, pg. 137). The test is based on the use of "t" tables available in

general statistics books (Steel and Torrie 1980). The measure of how much the

variation in the vegetation data is explained by the environmental variables is

determined by squaring the correlation coefficient (r2).

Hydrologic and Salinity Data Collection

Salinity values were measured every 10 minutes at each of the ten

vegetation belt transect locations as well as two additional locations, referred to

as datalogging stations E and W (Figure 2-7). These two stations were located

relatively near one another (approximately 1,200 feet) but were associated with

two different tidal creek systems. Datalogging station E was associated with a

tidal creek connected to a freshwater reach of the Little Back River.

Consequently, freshwater was delivered to the marsh at datalogging station E

during high tide. Conversely, datalogging station W received more saline waters

via a tidal creek connected to the Middle River. In addition to salinity, water

levels were measured in tidal creeks adjacent to each of the belt transects, as









well as datalogging stations E and W. Marsh surface elevations of datalogging

stations E and W were determined using the GPS.

The monitoring stations and their associated sensors were placed to

provide simultaneous monitoring of salinity in tidal creeks and in the adjacent

marshes. The sensors monitored salinity within the tidal creeks, within the

waters that flooded the marsh during high tide, and within the marsh sediments.

Figure 2-8 provides a conceptual configuration of a typical monitoring station,

depicted at both high tide and low tide. The marsh is flooded only during high

tide. The specifics of water level and salinity monitoring instrumentation are

discussed below.

Water Level Monitoring Instrumentation

Each monitoring station was configured around a 2-megabyte (MB)

datalogger (Model CR10X, Campbell Scientific, Inc., Logan, Utah) housed in a

weatherproof enclosure. Tide stages in tidal creeks were monitored using 10-

foot lengths of Aquatape (Consilium US, Inc., Littleton, Massachusetts). This

device consists of two thin, flexible metal ribbons attached along their edges to

form a narrow double-sided tape. The electrical resistances of the tape changes

as the two sides of the tape are pressed together by rising or falling tidal

floodwater. The Aquatape was installed in a tidal creek within a stilling well

constructed of 2-inch diameter PVC pipe, which provided both structural support

for the Aquatape and protection from floating debris carried by the tide. The

Aquatape resistance was calibrated to the water level by determining the water

level elevation with the GPS.









The frequency, depth, and duration of tidal flooding were calculated from

the automatically logged data. Frequency was calculated by counting the

number of tide events whose stage exceeded the marsh elevation. Depth was

determined by measuring the height of these tide events. Duration was

determined by counting the number of data points (i.e., 10 minute intervals)

collected when tide stage was above marsh elevation.

Salinity Monitoring Instrumentation

Concurrent with the installation of water level monitoring equipment,

salinity-monitoring equipment was installed in both tidal creeks and marsh

locations, corresponding to the 12-datalogging stations (Figure 2-9). Salinity was

measured at each of the 12-datalogging stations and stored in one of two types

of dataloggers. In tidal creeks, salinity was monitored using a YSI Model 6600

submersible datalogger equipped with a conductivity/temperature probe (Yellow

Springs Instruments, Inc., Yellow Springs, Ohio). These instruments were

suspended from buoys to maintain a constant depth below the water surface of

approximately 1.64 feet. The datalogger was programmed to record conductivity,

temperature, and salinity at 10-minute intervals. The recorded data were

downloaded approximately every 30-days.

In marsh areas, salinity was measured with CSI Model 547 conductivity/

temperature probes (Campbell Scientific, Inc., Logan, Utah) wired to the same

Campbell Scientific CR10X dataloggers to which the water level sensors were

attached. Salinity probes were mounted within each marsh in two configurations.

One probe was mounted to a fixed post and positioned approximately 0.066 feet

above the marsh surface to monitor the high-tide floodwater salinity, and another









was installed just below the dense root mat, at the top of the unconsolidated

sediments. To prevent plugging of the probe by fine clay, the latter probes were

encased in a 1-foot long, gravel filled section of 4-inch PVC pipe capped on each

end (Figure 2-9). The possibility that installation of the probe below the root mat

would create an unnatural exchange pathway was of concern. To minimize this

effect, the marsh root mat was neatly sliced with a sharp saw, carefully lifting the

root mat and inserting the gravel packed sensor beneath it, and then replacing

the root mat and leveling the area.

In addition to salinity data collected by the automatic dataloggers,

additional field measurements of the sediment salinity within the plant root zones

were collected concurrently with the vegetation monitoring. These salinity data

were collected using a hand held salinity-conductivity-temperature meter (Model

30 SCT meter, Yellow Springs Instruments, Yellow Springs, Ohio). Readings

were taken at 50-foot or 100-foot intervals along the belt transects by punching a

1-foot deep hole through the root mat with a 1.5-inch diameter piece of PVC well

screen fitted with a sharp point. To prevent mixing of marsh surface water and

the underlying interstitial water and to ensure integrity of the salinity

measurement, field readings were only collected at low tide when there was no

surface water on the marsh.
















980000 1000000




820000 820000




*





800000 800000










780000 S 780000













*
760000 760000




980000 1000000
US State Plane. Georgia East Zone. NAD83. US Survey Feet




Figure 2-1. Flight lines and photograph center points for 1:12,000 scale
true color and infrared aerial photography flown August 1999.
















980000 1000000



820000 820000


*







800000 800000










780000 ( 780000










760000 760000




980000 1000000
US State Plane. Georgia East Zone, NAD83. US Survey Feet




Figure 2-2. Flight lines and photograph center points for 1:25,000 scale
true color and infrared aerial photography flown August 1999.

















980000 1000000




820000 Target (A) Coordinates 820000
7 TARGET EATING NOTHING
1 992870 759870
2 971792 770556
3 966936 779067
6A 4 968449 793566
5 973573 789510
6 966317 811558
7 974214 817983
8 986902 796030
9 992354 778754
10 998939 769918

800000 800000

8
4







780000 A3 9 780000











760000 760000




980000 1000000
US State Plane. Georgia East Zone. NAD83. US Survey Feet



Figure 2-3. Locations and coordinates of aerial targets along
the lower Savannah River used in rectification of August 1999
true color and infrared aerial photography.














970000 990000 1010000





820000 820000









800000 800000
Ins Scale )
2000 0







780000 780000









760000 760000


970000 990000 1010000
US Sl.e Ptne,. G.oa ast Z, NAD83. US Su- e Faet







Figure 2-4. Base map of main river channels and tidal creeks digitized
from rectified 1:12,000 and 1:25.000 scale aerial photography acquired
August 1999. The detail to which the tidal creek system was digitized is
highlighted in the inset.











970000 980000 990000

810000 810000




Q9
800000 800000

1 6


5
790000 7 790000

10




780000 i780000


Q2


770000 770000





970000 980000 990000
US State Plane, Georgia East Zone, NAD83. US Survey Feet



Figure 2-5. Permanent vegetation monitoring belt transect locations.











0 100 200 300 400 500ft.

I1 1 1 1 1 1 1 I I i


4I Frequency Interval







Cover Interval I


2 ft.


Figure 2-6. Typical 500-foot (ft.) vegetation sampling belt transect composed of 50 contiguous cover intervals.
Enlargement of single cover interval details placement of 10 contiguous 1 by 2-foot frequency intervals within each cover
interval. A total of 500 frequency intervals are present for each 500-foot quadrat.










970000 980000


IpI~ ssfi I~'d~


800000











790000











780000


970000 980000
US State Plane. Geogia East Zone. NAD83, US Survey Feet


Figure 2-7. Locations of hydrologic and salinity datalogging stations.


800000











790000











780000














i--- 2 Ft


AQUATAPE
/ Waler Level Sensor


Pressure Transducer


LOW TIDE


- 200 Ft -


AQUATAPE
/ Waer Le-e Sensor


HIGH TIDE




Figure 2-8. Hydrologic and salinity monitoring equipment setup at low tide and
high tide.


Datalogger Enclosure


Datalogge eosure










Salinity Sensors
-7-


High Tide
Flood Level


Figure 2-9. Positioning of salinity probes within marsh sediments and above
marsh surface.


I


C~t




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ETA37Q4UG_54YN93 INGEST_TIME 2014-04-18T23:29:16Z PACKAGE AA00014260_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES