THE FUNCTIONAL ASSESSMENT OF WETLANDS
By
Edwin Martínez Martínez

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Crop and Soil Sciences - Doctor of Philosophy
2014

ABSTRACT
THE FUNCTIONAL ASSESSMENT OF WETLANDS
By
Edwin Martínez Martínez
Wetlands play a diverse and important role in the ecosystem. They provide numerous
environmental, economic, cultural, recreational, aesthetic and ecological benefits to society.
Meanwhile, wetlands are lost at an alarming rate due to human actions such as deforestation,
expansion of agricultural land, pollution, and climate change. Quantifying wetland functionality
is the first step to protect these valuable and biologically diverse ecosystems. However, current
functional assessment techniques only provide a general overview of wetland functions in large
and diverse watersheds. In addition, due to the qualitative nature of these techniques, they cannot
be used to develop future management and restoration plans, which require solid understanding
of hydrological and water quality characteristics of the wetlands. The goal of this research is to
address some of these limitations by examining the impacts of wetland size, depth, and
placement on flow and sediment transport at subbasin and watershed scales. The Soil and Water
Assessment Tool (SWAT), a physically-based watershed model, was used along with the System
for Urban Stormwater Treatment and Analysis Integration (SUSTAIN) to examine flow and
sediment transport in two watersheds in Michigan, the Shiawassee Watershed (southeastern
Michigan) and the River Raisin Watershed (southeastern Michigan and northeastern Ohio). Both
watersheds were selected because they have experienced significant conversion of land from
wetlands to agriculture since European settlement. Wetland area was found to be more influential
in controlling streamflow rate than wetland depth. Meanwhile, wetland implementation has
limited impacts of daily peak flow rates and frequency of peak flow events at the watershed

outlet. In general, rate of streamflow reduction is higher than sediment reduction at the subbasin
level but more comparable at the watershed level. These results reveal the importance of wetland
size, depth and placement as part of restoration efforts. This study introduces an alternative
approach to the functional assessment of wetlands that is more accurate and quantitative.

Copyright by
EDWIN MARTINEZ MARTINEZ
2014

This doctoral thesis is dedicated to my beloved family. To my parents for all their hard work,
sacrifices, support, advices, and for always giving me unconditional love. “Los amo y estoy muy
orgulloso de ustedes”

v

ACKNOWLEDGEMENTS

I want to thank God, first of all for giving me a great family and friends and for being my
guide and my ALL.
I would like to thank my beloved family: Isidora Martínez Santana (mom) and Edwin
Enrique Martínez Martinez (dad), Maria Luisa Martínez Martínez, Tania Martínez Martínez, and
Milagros Martínez Martínez (sisters) for their unconditional support, inspiration, dedication and
love during all my life.
Thanks to my uncle, Monsignor Hector Rivera Ramos (Tío Hector) and the Rivera
Family for all their unconditional support. Thanks to my grandmother Gertrudis Maldonado, who
is in heaven watching over me, for always giving me advice and for her unconditional and true
love.
I want to thank my awesome advisor, Dr. Amir Pouyan Nejadhashemi for being a great
research advisor, for always finding the time to listen to me, and for all the help and advices
provided to me to improve my thesis research. Thanks for making the difference in my
professional and academic career and for your encouragement and motivation since day one of
this journey at Michigan State University. I feel extremely fortunate to have Dr. Nejadhashemi as
my dissertation advisor.
Thank you to my committee members: Dr. Dawn Reinhold, Dr. Tim Harrigan, and Dr.
Sasha Kravchenko for their advice and for their generosity in helping me to improve my research
skills.
I want to thank Sean Woznicki for all his assistance with computer models, programing,
research and most important for his friendship and encouragement.
vi

I want to thank my friends/family: Linda Ortiz, Rafael Vega and Lindsay Martínez,
Dewell Paez and Carolina Solano, Alvin Perez and Becky Baez, Jorge Cintrón and Sheila Luna
for becoming my family and always giving me good advice and encouraging me to complete this
journey. Linda thanks for being an important part of this journey and for the great memories.
I want to thank my great friend Lindsay Cunningham (Graduate Student Toxicology
Department) for his friendship and for all of his advice and support with my writing during these
past few years.
Thank you the Sustainable Ecosystems Modeling Lab graduate students team for their
support, friendship and for making the difference as team players. You guys are the best! Sean,
Brad, Giri, Matt E., Yasseen, Georgina, Melissa, Umesh, Mike, Andy, Fariborz and Irwin.
Thanks go to Dr. Luis Perez Alegria and Dr. Jairo Diaz for their support with my
graduate school application, scholarship applications and research advice.
Thanks go to the Michigan State University faculty members and administrative
members for providing my education and for the excellent work in education that they perform.
Special thanks to Rita House for all her hard work and for providing me with guidance since day
one of this journey.
Thanks to the National Organization of Professional Hispanics Natural Resources
Conservation Service Employees (NOPHNRCSE) for their support through the Career
Enhancement Scholarship during 2009 and 2010.
I want to thank the Natural Resource Conservation Service personnel in the Saint Johns,
Michigan Field Office for their friendship, team efforts and hard work. You guys are a great

vii

staff, friends and family! Thanks go to Eddie Glover (Area 3 Area Conservationist) and Garry
Lee (Michigan State Conservationist) for their support.
The Alliance for Graduate Education and the Professoriate (AGEP) supported materials
for this research and provided me with great educational opportunities and training. Special
thanks to Steven Thomas and the AGEP Steering Committee members for their advices and
friendships.

viii

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................................ xii
LIST OF FIGURES ..................................................................................................................... xiv
KEY TO ABBREVIATIONS .................................................................................................... xviii
1

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

2

LITERATURE REVIEW ..........................................................................................................3
2.1 WETLANDS OVERVIEW ......................................................................................................3
2.2 TYPE OF WETLANDS ..........................................................................................................4
2.3 FUNCTIONS, SERVICES AND VALUES OF WETLANDS ..........................................................6
2.3.1 Wetland Functions ................................................................................................7
2.3.2 Wetland Services ..................................................................................................7
2.3.3 Wetland Values .....................................................................................................8
2.4 DESCRIPTION OF FUNCTIONS, SERVICES AND VALUES OF WETLANDS ...............................9
2.4.1 Water Quality ........................................................................................................9
2.4.2 Water Quantity ....................................................................................................10
2.4.3 Nutrient Transformation .....................................................................................11
2.4.4 Flood Water Storage ...........................................................................................14
2.4.5 Sediment Retention .............................................................................................16
2.4.6 Carbon Sequestration ..........................................................................................17
2.4.7 Wildlife ...............................................................................................................18
2.5 WETLANDS AND AGRICULTURE .......................................................................................18
2.5.1 Crop Production ..................................................................................................20
2.5.2 Livestock Grazing ...............................................................................................20
2.6 WETLANDS AND LANDUSE CHANGE ................................................................................21
2.7 WETLANDS CONSERVATION PROGRAMS ..........................................................................22
2.7.1 Wetland Restoration Program (WRP) ................................................................22
2.7.2 Conservation Reserve Program (CRP) ...............................................................23
2.8 WETLANDS CONSERVATION PRACTICES ..........................................................................23
2.8.1 Wetland Restoration............................................................................................24
2.8.2 Wetland Enhancement ........................................................................................24
2.8.3 Wetland Creation ................................................................................................24
2.8.4 Wetland Wildlife Habitat Management ..............................................................25
2.8.5 Wetland Construction .........................................................................................25
2.9 WETLAND ASSESSMENTS .................................................................................................25
2.10 WETLAND COMPUTER MODELS .......................................................................................29
2.10.1 Computer Models Strengths and Deficiencies ....................................................31
2.10.2 Factors to Consider in Wetland Modeling ..........................................................32
2.11 COMPUTER MODELS FOR WETLANDS AND WATER QUALITY MODELING ........................33
2.11.1 System for Urban Stormwater Treatment and Analysis INtegration ..................35
ix

2.11.2 Soil and Water Assessment Tool ........................................................................36
2.11.3 SWAT Applications on Wetlands .......................................................................38
2.11.4 Hydrologic Simulation Program-Fortran ............................................................42
2.11.5 HSPF Applications on Wetlands ........................................................................42
2.11.6 Storm Water Management Model .......................................................................43
2.11.7 WETLANDS-2D ................................................................................................44
2.11.8 Wetland – DNDC ................................................................................................45
2.11.9 Hydrologic Engineering Center – Hydrologic Modeling System.......................46
2.11.10 Precipitation-Runoff Modeling System .........................................................46
2.11.11 Environmental Fluid Dynamics Code............................................................47
2.11.12 TABS-2 ..........................................................................................................48
2.12 WETLANDS AND GEOGRAPHICAL INFORMATION SYSTEMS (GIS) ....................................48
2.13 WETLANDS AND REMOTE SENSING APPLICATIONS ..........................................................50
3

INTRODUCTION TO METHODOLOGY AND RESULTS .................................................51

4

MODELING THE HYDROLOGICAL SIGNIFICANCE OF WETLAND RESTORATION
SCENARIOS ..................................................................................................................53
4.1 ABSTRACT .......................................................................................................................53
4.2 INTRODUCTION ................................................................................................................53
4.3 MATERIALS AND METHODS .............................................................................................57
4.3.1 Study Area ..........................................................................................................57
4.3.2 SWAT Model Description ..................................................................................61
4.3.3 SWAT Model Wetland Processes .......................................................................61
4.3.4 SWAT Model Setup............................................................................................65
4.3.5 Wetlands Digital Data Assessment .....................................................................66
4.3.6 Wetland Field Data Assessment .........................................................................66
4.3.7 SWAT Calibration and Validation......................................................................69
4.3.8 SWAT Wetland Restoration Scenarios for the Shiawassee Watershed..............73
4.3.8.1 Impact of Wetland Placement by Stream Order on Streamflow ...........74
4.3.8.2 Impact of Wetland Placement on Peak Flow .........................................75
4.4 RESULTS AND DISCUSSION...............................................................................................75
4.4.1 Model Calibration and Validation ......................................................................75
4.4.2 Wetland Restoration Scenarios ...........................................................................78
4.4.2.1 Impact of Wetland Area on Streamflow ................................................78
4.4.2.2 Impact of Wetland Depth on Streamflow ..............................................82
4.4.2.3 Impact of Wetland Placement by Stream Order on Streamflow ...........86
4.4.2.4 Peak Flow Reduction .............................................................................91
4.4.3 High Priority Areas for Wetland Restoration .....................................................93
4.5 CONCLUSION....................................................................................................................96

5

ASSESSING THE SIGNIFICANCE OF WETLAND RESTORATION SCENARIOS ON A
SEDIMENT MITIGATION PLAN ................................................................................98
5.1 ABSTRACT .......................................................................................................................98
5.2 INTRODUCTION ................................................................................................................99
5.3 MATERIALS AND METHODS ...........................................................................................102
5.3.1 Study Area ........................................................................................................102
x

5.4

5.5

5.3.2 Models...............................................................................................................104
5.3.2.1 SUSTAIN Model .................................................................................105
5.3.2.2 SWAT Model .......................................................................................106
5.3.3 SUSTAIN/SWAT Hyrbid Model Approach .....................................................108
5.3.4 Wetland Restoration Scenarios .........................................................................110
5.3.5 Environmental and Economic Aspects of Wetland Implementation ................111
5.3.6 Environmental Aspects of Wetland Implementation ........................................111
5.3.7 Economic Aspects of Wetland Implementation ...............................................112
5.3.8 Wetland Selection Considering Environmental and Economic Factors ...........113
5.3.9 Statistical Analysis ............................................................................................121
RESULTS AND DISCUSSION.............................................................................................122
5.4.1 Model Calibration .............................................................................................122
5.4.2 Wetland Impacts on Streamflow .......................................................................127
5.4.3 Wetland Impacts on Sediment Load .................................................................129
5.4.4 Distance From Watershed Outlet and Sediment Reduction .............................133
5.4.5 Stream Order and Sediment Reduction.............................................................137
5.4.6 Landuse and Sediment Reduction .....................................................................140
5.4.7 Selection of Most Suitable Wetland Considering Environmental and Economic
Factors ...............................................................................................................142
CONCLUSION..................................................................................................................149

6

OVERALL CONCLUSION ..................................................................................................151

7

RECOMMENDATIONS FOR FUTURE RESEARCH........................................................153

APPENDICES .............................................................................................................................154
APPENDIX A ........................................................................................................................155
APPENDIX B ........................................................................................................................184
REFERENCES ............................................................................................................................200

xi

LIST OF TABLES

Table 4-1. Model Calibration parameters ......................................................................... 71
Table 4-2. Model Performance summary ......................................................................... 75
Table 4-3. Statistical comparison of streamflow means for wetland area ........................ 79
Table 4-4. Statistical comparison of streamflow means for wetland depth ...................... 83
Table 4-5. Wetland area/depth combinations for reduction of average annual maximum peak
flow (m3/s) events........................................................................................ 92
Table 4-6. Wetland area/depth combinations selected as optimal for streamflow reduction at the
watershed outlet ........................................................................................... 95
Table 4-7. Wetland area/depth combinations selected as optimal for streamflow reduction at the
individual subbasin outlet ............................................................................ 95
Table 5-1. Pairwise comparison matrix developed for Subbasin 2 based on sediment reduction at
the watershed outlet. .................................................................................. 114
Table 5-2. Priority vector calculation based on sediment reduction at watershed outlet for the
Subbasin 2. ................................................................................................. 115
Table 5-3. Daily streamflow calibration/validation results............................................. 117
Table 5-4. Daily sediment calibration/validation results ................................................ 117
Table 5-5. Statistical comparison of mean flow and sediment reduction for different wetland
surface areas at watershed and subbasin level ........................................... 117
Table 5-6. Distance from outlet and sediment reduction ................................................ 119
Table 5-7. Decision matrix of wetland size alternatives for all criteria developed for watershed
analysis (Subbasin 2). ................................................................................ 120
Table 5-8. Daily streamflow calibration/validation results............................................. 122
Table 5-9. Daily sediment calibration/validation results ................................................ 122
Table 5-10. Statistical comparison of mean flow and sediment reduction for different wetland
sizes at watershed and subbasin level ........................................................ 133
Table 5-11. Distance from outlet and sediment reduction .............................................. 136

xii

Table 5-12. Statistical comparison of the sediment reduction provided by different wetland
surface areas and stream orders at watershed level ................................... 139
Table 5-13. Landuse type and sediment reduction ......................................................... 141
Table 9-1. SUSTAIN model parameters that were used in this study ............................ 185
Table 9-2. Sensitivity analysis of the SUSTAIN model parameters .............................. 187
Table 9-3. SUSTAIN and SWAT model parameters for models set up. ........................ 189
Table 9-4. SWAT model parameters that were adjusted during flow calibration procedure193
Table 9-5. SWAT model parameters that were adjusted during sediment calibration procedure
for station STORET 580046 ...................................................................... 194
Table 9-6. Cost summary for establishment and maintenance of a 0.4-hectare wetland over 10year period (USDA-NRCS, 2013a) ........................................................... 195
Table 9-7. Cost summary for establishment and maintenance of a 0.81-hectare wetland over 10year period (USDA-NRCS, 2013a) ........................................................... 196
Table 9-8. Cost summary for establishment and maintenance of a 2-hectare wetland over 10-year
period (USDA-NRCS, 2013a) ................................................................... 197
Table 9-9. Cost summary for establishment and maintenance of a 4-hectare wetland over 10-year
period (USDA-NRCS, 2013a) ................................................................... 198
Table 9-10. Statistical comparison of the sediment reduction provided by different wetland
surface areas and stream orders at subbasin level ...................................... 199

xiii

LIST OF FIGURES

Figure 4-1. Study Area. ..................................................................................................... 58
Figure 4-2. Shiawassee watershed land use ...................................................................... 59
Figure 4-3. Shiawassee watershed topography. ................................................................ 60
Figure 4-4. Relationship between wetland storage areas and volumes in the Shiawassee
watershed. .................................................................................................... 68
Figure 4-5. Observed versus simulated daily discharge at USGS station 04144500........ 76
Figure 4-6. Observed versus simulated daily discharge at USGS station 04145000........ 77
Figure 4-7. Percent flow change at the watershed outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 91 cm ................ 80
Figure 4-8. Percent flow change at the subbasin outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 91 cm ................ 81
Figure 4-9. Percent flow change at the watershed outlet compared to base scenario – 500 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ........ 84
Figure 4-10. Percent flow change at the subbasin outlet compared to base scenario – 500 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ........ 85
Figure 4-11. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 91 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha 87
Figure 4-12. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 500 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm 88
Figure 4-13. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 91 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha 89
Figure 4-14. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 500 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm 90
Figure 4-15. High, medium, and low impact areas for wetland restoration scenarios ..... 94
Figure 5-1. Study area and location of the monitoring stations ...................................... 104
Figure 5-2. SWAT – SUSTAIN models integration schema......................................... 109
Figure 5-3. River Raisin watershed subbasin 2. ............................................................ 113
xiv

Figure 5-4. Streamflow calibration versus validation for the USGS 04175600 gauging station.
.................................................................................................................... 123
Figure 5-5. Streamflow calibration versus validation for the USGS 04176000 gauging station.
.................................................................................................................... 124
Figure 5-6. Streamflow calibration versus validation for the USGS 04176500 gauging station.
.................................................................................................................... 125
Figure 5-7. Sediment Concentration calibration versus validation for the STORET station
580046........................................................................................................ 126
Figure 5-8. Percentage flow reduction at the watershed outlet. (a) 0.40 ha wetland scenario, (b)
0.81 ha wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha wetland scenario.
.................................................................................................................... 128
Figure 5-9. Percentage flow reduction at the watershed subbasin level. (a) 0.40 ha wetland
scenario, (b) 0.81 ha wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha
wetland scenario......................................................................................... 129
Figure 5-10. Sediment reduction at the watershed outlet. (a) 0.40 ha wetland scenario, (b) 0.81 ha
wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha wetland scenario.131
Figure 5-11. Sediment reduction at the subbasin level. (a) 0.40 ha wetland scenario, (b) 0.81 ha
wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha wetland scenario.132
Figure 5-12. Relationship between distance of the wetland implementation site and the watershed
outlet. (a) individual wetlands and (b) clustered wetlands for 0.4 ha wetlands.
.................................................................................................................... 134
Figure 5-13. Most suitable wetland selected based on different environmental / economic
scenarios (watershed scale). ....................................................................... 145
Figure 5-14. Most suitable wetland selected based on different environmental/economic
scenarios (subbasin level). ......................................................................... 146
Figure 5-15 Most suitable wetland placement considering watershed level
environmental/economic benefits a) 0.1/0.9, b) 0.25/0.75, c) 0.5/0.5, d) 0.75/0.25,
and e) 0.9/0.1.............................................................................................. 147
Figure 5-16. Most suitable wetland placement considering subbasin level
environmental/economic benefits a) 0.1/0.9, b) 0.25/0.75, c) 0.5/0.5, d) 0.75/0.25,
and e) 0.9/0.1.............................................................................................. 148
Figure 8-1. SWAT data frame for wetland modeling processes. .................................... 156
Figure 8-2. SWAT statistical data frame. ....................................................................... 157

xv

Figure 8-3.Percent flow change at the watershed outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 15 cm .............. 158
Figure 8-4.Percent flow change at the watershed outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 30 cm .............. 159
Figure 8-5.Percent flow change at the watershed outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 61 cm .............. 160
Figure 8-6. Percent flow change at the subbasin outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 15 cm .............. 161
Figure 8-7. Percent flow change at the subbasin outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 30 cm .............. 162
Figure 8-8. Percent flow change at the subbasin outlet compared to base scenario –wetlands
areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 61 cm .............. 163
Figure 8-9. Percent flow change at the watershed outlet compared to base scenario – 50 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ...... 164
Figure 8-10.Percent flow change at the watershed outlet compared to base scenario – 100 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ...... 165
Figure 8-11. Percent flow change at the watershed outlet compared to base scenario – 250 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ...... 166
Figure 8-12. Percent flow change at the subbasin outlet compared to base scenario – 50 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ...... 167
Figure 8-13. Percent flow change at the subbasin outlet compared to base scenario – 100 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ...... 168
Figure 8-14.Percent flow change at the subbasin outlet compared to base scenario – 250 ha
wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm ...... 169
Figure 8-15. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 15 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha 170
Figure 8-16. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 30 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha 171
Figure 8-17. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 61 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha 172
Figure 8-18. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 50 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm 173

xvi

Figure 8-19. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 100 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm174
Figure 8-20. Percent streamflow change at the watershed outlet compared to base scenario by
stream order for 250 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm175
Figure 8-21. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 15 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha176
Figure 8-22. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 30 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha177
Figure 8-23. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 61 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha178
Figure 8-24. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 50 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm179
Figure 8-25. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 100 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm180
Figure 8-26. Percent streamflow change at the subbasin outlet compared to base scenario by
stream order for 250 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm181
Figure 8-27. Sampled subbasins ..................................................................................... 182
Figure 8-28. Measuring wetland depth in the Shiawassee watershed ............................ 182
Figure 8-29. Collecting GPS points for surveyed areas in the Shiawassee watershed ... 183
Figure 8-30. Soil hydrology data collection for a farmed wetland ................................. 183

xvii

KEY TO ABBREVIATIONS

ARS – Agricultural Research Service
BASINS – Better Assessment Science Integrating point and Nonpoint Sources
BMP – Best Management Practice
Canmx – Maximum Canopy Storage (mm)
CMSR – Completely Mixed Stirred Reactor
CH_N1 – Manning’s “n” value for the tributary channel
Ch_N2 – Manning’s “n” value for the main channel
CH_K2 – Channel effective hydraulic conductivity (mm/hr)
CN2 – Initial SCS runoff curve number for moisture condition II
CO2 – Carbon Dioxide (ppm)
CWA – Clean Water Act
DEM – Digital Elevation Model
EFTMIX – Tillage mixing efficiency
Epco – Plant uptake compensation factor
Esco – Soil evaporation compensation factor
ET – evapotranspiration (mm)
GIS – Geographic Information Systems
Gw_Delay – Groundwater delay time (days)
HRU – Hydrologic Response Unit
HSPF – Hydrologic Simulation Program-FORTRAN
HUC – Hydrologic Unit Code

xviii

Kg – Kilogram
Km2 – Square Kilometer
MUSLE – Modified Universal Soil Loss Equation
NCDC – National Climatic Data Center
NLCD – National Land Cover Database
NHD – National Hydrography Data
NPS – Non-Point Source
NRCS – Natural Resources Conservation Service
NSE – Nash-Sutcliffe Efficiency
PLOAD – GIS Pollutant Load
PRISM – Parameter-elevation Regressions on Independent Slopes Model
R2 – Coefficient of determination
Rchrg_Dp – Deep aquifer percolation fraction
RCM – Regional Climate Model
RMSE – Root Mean Square Error
RUSLE – Revised Universal Soil Loss Equation
SCS – Soil Conservation Service
SDSM – Statistical Downscaling Model
Sol_Alb – moist soil albedo
SOL_AWC – available water capacity of the soil layer (mm H2O/mm soil)
Sol_K – saturated hydraulic conductivity (mm/hr)
Spcon – Linear re-entrainment parameter for channel sediment routing
Spexp – Exponential re-entrainment parameter for channel sediment routing

xix

SRES – Special Report on Emissions Scenarios
SSURGO – Soil Survey Geographic
STATSGO – State Soil Geographic Database
STORET - STOrage and RETrieval
SUSTAIN – System for Urban Stormwater Treatment and Analysis INtegration
SURLAG – Surface runoff lag coefficient
SWAT – Soil and Water Assessment Tool
TCLW – Tuttle Creek Lake Watershed
Timp – Snow pack temperature lag factor
TKN – Total Kjeldahl Nitrogen
TN – Total nitrogen
TP – Total phosphorus
TSS – Total Suspended Solids
USDA – United States Department of Agriculture
USEPA – United States Environmental Protection Agency
USGS – United States Geological Survey
USLE – Universal Soil Loss Equation
USLE_C – Universal Soil Loss Equation cover factor
USLE_P – Universal Soil Loss Equation support practice factor
WetSpa – Water and Energy Transfer between Soil, Plants and Atmosphere
WSS – Web Soil Survey

xx

1

INTRODUCTION

Wetlands provide important hydrologic, geochemical, and biological functions in a
watershed (De Laney, 1995; Hart, 1995; NRC, 1995). In addition, wetland systems directly
support millions of people throughout the world by providing fertile soils for agriculture
production, food, shelter for wildlife, trees for timber and fuel, recreation areas, and many other
benefits for humans.
Meanwhile, wetlands have been lost at an alarming rate. The impacts of lost wetlands can
affect health and productivity of water bodies such as streams, lakes and rivers downstream
(Meyer et al., 2003). Destruction and degradation of wetlands can also reduce groundwater
levels. For example, it has been estimated that groundwater resources would decline by 45% if
80% of Florida’s cypress swamps were drained to accommodate other land uses (Ewel, 1990).
In order to protect valuable wetland ecosystems, first we need to better understand their
functions and roles, especially within large and diverse areas. However, this is a challenging task
due to the complexity of wetlands’ pollutant removal and transport processes at the watershed
scale. Meanwhile, current functional wetland assessment techniques only provide rough
estimations, which in most cases are site specific and qualitative.
The overall goal of this project is to quantify some water quality and quantity benefits of
wetland implementation scenarios in large and diverse watersheds. The specific objectives of this
project are as follows:
1) Evaluate the impacts of wetland depth, area, and wetland placement in the watershed
on streamflow and peak flow reduction at the watershed scale.
2) Determine the role of wetland placement in watershed sediment dynamics by
considering the distance to the outlet and stream order concept

1

3) Identify the most appropriate sites for wetland implementation by considering the
environmental and economic aspects of restoration scenarios

2

2
2.1

LITERATURE REVIEW

WETLANDS OVERVIEW
Wetlands are diverse environments which are defined by several factors; these include

the soil type, hydrology, topography, climate and vegetation. Natural and human landscape
disturbance activities could potentially negatively affect the above mentioned factors. Abiotic
and biotic characteristics of a wetland are controlled by the hydrology of that wetland (National
Research Council, 1995). Wetlands have a scientific and legal definition. A scientific definition
was provided by Cowardin et al., 1979 who defined wetlands as “lands where saturation with
water is the dominant factor determining the nature of soil development and the types of plant
and animal communities living in the soil and on its surface.” A legal definition of wetlands
under the Clean Water Act is: "those areas that are inundated or saturated by surface or
groundwater at a frequency and duration sufficient to support, and that under normal
circumstances do support, a prevalence of vegetation typically adapted for life in saturated soil
conditions. Wetlands generally include swamps, marshes, bogs and similar areas" (U.S. EPA,
1995). Wetlands can be situated in saline, brackish and freshwater environments along
coastlines, within estuaries, rivers and lakes, on slopes where ground water breaks out as a spring
or seeps, in abandoned ditches or stream channels, as well as other locations.

3

2.2

TYPE OF WETLANDS
We have several major types of wetlands in the United States due to the unique and

diverse climate, land features, vegetation and land management activities. A very important
characteristic of wetland types is the hydrologic conditions of wetlands. This factor will
determine the wetland's geomorphology, habitat quality, water quality, and biodiversity
according to: Mitsch and Gosselink, 1993; Kadlec and Knight, 1996.The EPA described the most
common types of wetlands in the United States and organized them into four general categories:
marshes, swamps, bogs and fens (U.S. EPA, 2001).
The following is a description of each of these types (U.S EPA, 2001; Mitsch and
Gosselink, 2000; Mitsch and Gosselink, 1993):
1. Marshes are periodically saturated, flooded, or ponded with water and
characterized by herbaceous (non-woody) vegetation adapted to wet soil
conditions. Marshes are further characterized as tidal marshes and non-tidal
marshes. Tidal (coastal) marshes occur along coastlines and are influenced by
tides and often by freshwater from runoff, rivers, or ground water. Non-tidal
(inland) marshes are dominated by herbaceous plants and frequently occur in
poorly drained depressions, floodplains, and shallow water areas along the edges
of lakes and rivers. Major regions of the United States that support inland marshes
include the Great Lakes coastal marshes, the prairie pothole region, and the
Florida Everglades.
i. Freshwater marshes are characterized by periodic or permanent
shallow water, little or no peat deposition, and mineral soils.

4

ii. Wet meadows commonly occur in poorly drained areas such as
shallow lake basins, low-lying depressions, and the land between
shallow marshes and upland areas.
iii. Wet prairies are similar to wet meadows but remain saturated
longer.
iv. Prairie potholes develop when snowmelt and rain fill the
pockmarks left on the landscape by glaciers. They provide
excellent habit and breading grounds for migratory birds.
v. Playas are small basins that collect rainfall and runoff from the
surrounding land.
vi. Vernal pools have either bedrock or a hard clay layer in the soil
that helps keep water in the pool.
2. Swamps are fed primarily by surface water inputs and are dominated by trees and
shrubs. Swamps occur in either freshwater or saltwater floodplains. They are
characterized by very wet soils during the growing season and standing water
during certain times of the year. Swamps are classified as forested, shrub, or
mangrove.
i. Forested swamps are found in broad floodplains of the northeast,
southeast, and south-central United States and receive floodwater
from nearby rivers and streams.

5

ii. Shrub swamps are similar to forested swamps except that shrubby
species like buttonbush and swamp rose dominate.
iii. Mangrove swamps are coastal wetlands characterized by salttolerant trees, shrubs, and other plants growing in brackish to
saline tidal waters.
3. Bogs are freshwater wetlands characterized by spongy peat deposits, a growth of
evergreen trees and shrubs, and a floor covered by a thick carpet of sphagnum
moss.
4. Fens are ground water-fed peat forming wetlands covered by grasses, sedges,
reeds, and wildflowers.

2.3

FUNCTIONS, SERVICES AND VALUES OF WETLANDS
Wetlands are valuable, limited, dynamic and unique ecological habitats in the world.

Wetland functions are properties that a wetland naturally provides, services are properties that
are valuable to humans, and values are attributes that humans assigned to wetland services.
These ecological areas have multiple functions, services and values that are defined by the
location, agricultural expansion, urban sprawl, environmental laws and regulations. Some of
these factors (i.e. industrialization, agricultural conversion and timber harvest) have significantly
contributed to the loss and degradation of wetland ecosystems (Mitch and Gosselink, 1993). It is
important to notice that not all wetlands will perform all functions and services nor do they have
the same values. Wetland properties are determined by several factors including: climate and

6

ecological conditions, location and size of the wetland, type of wetland and water availability
among others.
Unfortunately "while wetland functions are natural processes of wetlands that continue
regardless of their perceived value to humans, the value people place on those functions in many
cases is the primary factor determining whether a wetland remains intact or is converted for
some other use" (National Audubon Society, 1993).
2.3.1

Wetland Functions
Wetland functions are defined by Novitzki et al., 1997 as a process or series of processes

that take place within a wetland. Wetland functions are ecological processes which include but
are not limited to surface water storage, subsurface water storage, ground water recharge,
sediment and other particulate retention, shoreline stabilization, stream shading, and other
environmental functions (Novitzki, 1979; Luecke, 1993; Tiner, 1998; Ramsar, 2004; Keddy,
2000; Mitsch and Gosselink, 2000; Hassan et al., 2005; Verhoeven and Setter, 2010). In addition,
wetland functions include transformation of nutrients, growth of living matter, and diversity of
wetland plant. All of these values are for the wetland itself, for surrounding ecosystems, and for
people (USGS, 2004). In many locations, such as the United Kingdom, South Africa, Canada
and the United States, wetlands are part of conservation efforts due to the ecological,
environmental, nutritional and recreational values (Keddy, 2009).
2.3.2

Wetland Services
Wetland ecosystems provide a diversity of services for the well-being of humans, as well

for wildlife species. Some of these services are: food, fiber, biomass production, wildlife habitat
for terrestrial and avian species, retention, removal and transformation of nutrients, flood control,

7

flood water storage and storm buffering, biodiversity, maintenance or improvement of water
quality, carbon sequestration, reduction of soil erosion and sedimentation, discharge and
recharge of ground water. Several studies have been conducted showing the capacity of wetlands
to abate flooding, improve water quality, and support biodiversity (Neely and Baker, 1989;
Crumpton et al., 1993; Richardson and Craft, 1993; Bedford, 1999; Keddy, 2000).
Zedler (2003) attributed the decline of three ecosystem services (flood abatement, water
quality improvement, and biodiversity support) in the Upper Midwestern area to the drainage of
approximately 60% of the regions historical wetlands for agriculture.
2.3.3

Wetland Values
The value of a wetland lies in the benefits that it provides to the environment or to

people, something that is not easily measured Novitzki et al., 1997. Wetlands can have
ecological, social, or economic values. These values include multiple categories such as
population (animals harvested for pelts, waterfowl and other birds, fish and shellfish, timber and
other vegetation harvest, and endangered species), ecosystem (flood mitigation, storm abatement,
aquifer recharge, water quality improvement, aesthetics, and subsistence use) and the biosphere
(nitrogen , sulfur, carbon and phosphorus cycles) as described by Mitsch and Gosselink, 2000.
Many studies have discussed wetland values in an extensive matter (e.g. Wharton, 1970;
Gosselink et al., 1974; Mitsch, 1977; Costanza et al., 1997; Costanza et al., 1989; Turner, 1991;
Barbier, 1994; Gren et al., 1994; Gren, 1994; Bell, 1997; Mitsch and Gosselink, 2000). Values
that are well recognized by humans are productivity of downstream fisheries, recreational
observation and hunting of wildlife, reduced cost of water purification, production of valuable
sources of food and fiber, reduced damage due to flooding, and erosion reduction which is
directly linked to maintenance.
8

Barbier (1989; 1993; 1994) classified the total economic value for tropical wetlands and
divided it between direct use value, indirect use value and existence value. Direct use values are
forest resources, wildlife resources, fisheries, forage resources, agricultural resources and water
supply. Indirect use values are nutrient retention, flood and flow control, storm protection,
ground water recharge, external ecosystem support and micro-climatic stabilization. Existence
values are biodiversity, culture and bequest values. This study shows how important wetland
values are and how extensive these natural system characteristics are.
2.4
2.4.1

DESCRIPTION OF FUNCTIONS, SERVICES AND VALUES OF WETLANDS
Water Quality
Extensive research has shown that wetlands play an important role in improving water

quality (Wolverton et al., 1983; Neely and Baker, 1989; Reed, 1993; Larson et al., 1989;
Crumpton et al., 1993; Richardson and Craft, 1993; Bedford, 1999; Keddy, 2000). Wetland
vegetation will substantially slow the flow of runoff water causing deposition of mineral and
organic particles with adsorbed nutrients (Nitrogen and Phosphorus) (Carter, 1996). The ability
of wetlands to improve water quality has been extensively studied and these systems have been
shown to lower concentrations of water contaminants including Nitrogen, Phosphorus,
suspended solids, biochemical oxygen demand, trace metals, trace organics, and pathogens
(Mitsch, 1994; Hammer 1992, 1993).
Due to the effectiveness of wetlands on nutrients filtration and transformation, artificial
wetlands have been constructed for water quality restoration (e.g., Hammer 1992, Mitsch et al.,
2001). According to these studies, natural and constructed wetlands performing this function

9

will help improve local water quality of streams and other watercourses through the reduction of
nutrients and sediments loads.
Studies have shown that wetlands act as a natural filter that can improve water quality
(Kadlec and Knight, 1996). Wetland capacity to improve water quality has been shown in several
studies. The capacity of wetlands to reduce nutrients (N and P) allows the reduction of
eutrophication (Mitsch et al., 2001) in addition to the capacity to store large amounts of
sediments (Day et al., 2007).
Johnston et al. (1990) developed a method to evaluate the cumulative effect of wetlands
on stream water quality and quantity. A stepwise multiple regression analysis was used to
evaluate relationships between stream water quality variables and wetlands. The results showed
that the proximity of wetlands in relationship with the sampling station was associated with a
decrease of several parameters including: inorganic suspended sediments, Nitrates (NO3-), flowweighted ammonium (NH4), flow-weighed total P (TP). This study clearly shows the efficiency
of the wetlands’ capacity to improve water quality. The method used describes the efficiency of
wetlands removing suspended sediment solids, TP and Ammonia during high flow periods and
the efficiency of wetlands at removing Nitrates during low flow periods.
2.4.2

Water Quantity

Water is a resource that every known form of life on earth depends on and it affects every
aspect of our lives. While water is abundant in many regions, approximately one billion people
around the world don't have clean drinking water, and 2.6 billion still lack basic sanitation. In the
United States we can observe water as an abundant resource in places like Michigan, but at the
same time we can observe water as a limited resource in places like Northern California.

10

Wetlands play an important role in the hydrologic cycle and in some regions a vital role on water
quantity. Wetlands receive, store and release water in numerous ways (e.g. surface water, ground
water, and plant intake). In some regions wetlands will maintain streamflow during dry periods
providing a water source for wildlife. Wetlands possess multiple hydrological functions related
to gross water balance, ground water recharge, base flow and low flow, flood response and river
variability. All of these functions will vary depending on the type of wetland, location,
vegetation, and size, in addition to other important wetland characteristics.

Wetlands have the capacity to catch, retain, filter and release runoff water generated from
heavy rainfall or snowmelt events allowing an increase in ground water infiltration which will
help to reduce river flow downstream and agricultural runoff (Luecke 1993; Comin et al., 1997;
Keddy, 2000). Wetlands will provide short term and long term water storage functions and assist
with the reduction of downstream flood peaks (Ramsar, 2004).
2.4.3

Nutrient Transformation
Nitrogen (N) and phosphorus (P) are the primary macro-nutrients that enrich streams,

lakes and rivers. Phosphorus is the main nutrient controlling productivity and the primary cause
of the excess algal biomass in surface waters (Correll, 1998). The directly available forms of N
and P are mostly inorganic (NO3- and NH4+). Total N and total P include soluble fractions,
particulate and dissolved organic fractions. Total N and total P concentrations are used to predict
algal biomass in lakes and reservoirs. Nutrient concentrations can differ from stream to stream
because of differences in land use, geology, streamflow, point sources and other factors in the
drainage basin. Wetlands have been shown to be very efficient in removal of nutrients from
agricultural runoff (Comin et al., 2001).

11

Current studies show that wetlands have the capacity to significantly reduce nutrients,
sediments and other pollutant concentrations produced from agricultural runoff under different
environmental conditions (Kadlec, 1993; Mander et al., 2000; Trepel and Palmeri, 2002; Jordan
et al., 2003; Archeimer et al., 2004; Skagen et al., 2008). Kelly and Harwell (1985) nutrients
supplied to wetland areas from discharge or runoff are present in the soluble and particulate
forms (i.e. dissolved nitrogen (N) will be introduced as nitrate (NO3-), ammonium (NH4+) or
soluble organic forms and dissolved phosphate (PO4-3 ) or soluble organic Phosphorus).
Wetland vegetation will remove pollutants by slowing runoff and through pollutant plant
uptake. Jordan et al. (2003) studied nutrient and sediment reduction capacity of constructed
wetlands and found a reduction in non-point source pollution (approximately 25% of
Ammonium and 52% of Nitrate were significantly removed from the studied area). A study
conducted in the Houghton Lake wetland system in Michigan, shows the capacity of this wetland
system to remove up to 90% of the Phosphorus load (Kadlec, 1993). Several studies have found
that forested wetlands near rivers and streams are important for nutrient retention and
sedimentation during flood events (Whigham et al., 1988; Yarbro et al., 1984; Simpson et al.,
1983; Peterjohn and Correll, 1982).
Studies show that a combination between buffer strips and wetland areas will effectively
control the nutrient fluxes (Vought and Lacoursiere, 1998). In addition, wetlands can act as
filters removing particulate material, as sinks accumulating nutrients and also as transformers
converting nutrients to different forms (Richardson, 1989).
Healy and Cawley (2002) studied the reduction capacity of a constructed surface-flow
wetland in nutrients P and N from a waste treatment system in Ireland. They found an average

12

percentage reduction of approximately 51% for total N and 13% for total P. In addition to the
reduction in nutrients, a reduction of approximately 87% of suspended sediments was observed
along with a 49% reduction in biological oxygen demand (BOD). This study clearly shows the
nutrient reduction capacity of wetlands.
Several recent studies show the restored wetland nutrient (N and P) removal capacity
(Fleischer et al., 1994; Reinelt and Horner, 1995; Raisin et al., 1997, Hunt et al., 1999, Kovacic
et al., 2000; Braskerud 2002; Jordan et al., 2003). During each of these studies a reduction in
nutrients was found, but the observations are different due to environmental and time factors (i.e.
temporal variability of water inflow, timing required to filter or transform nutrients, hydraulic
loading rates and hydraulic efficiency).
Borin et al. (2001) demonstrated that wetlands reduced Nitrates (NO3-) by 95% and total
dissolved solids by 30% in runoff from cropland. Borin and Tocchetto (2007) found a nitrogen
removal efficiency of approximately 90% during a 5 year study conducted on water and nitrogen
balance for a constructed surface flow wetland treating agricultural drainage waters.
Weller et al. (1996) studied the role of wetlands in reducing P loading in surface waters.
This study showed the calculation of different variables that summarized a variety of
characteristics of wetlands using a geographic information system (GIS) and regression analysis
to measure Phosphorus loading. The wetland variables that were developed for this study were:
quantity of wetland (area number and perimeter), wetland type, land use and the relationship
between wetlands and streams. Significant results were found in this study suggesting that a
hectare of riparian wetland may be many times more important in reducing Phosphorus than an
agricultural hectare is in producing Phosphorus (Weller et al., 1996). This study shows the

13

capacity of wetlands reducing P loading from surface waters which is beneficial to improve and
protect surface water quality.
Recent researches demonstrate the capacity of constructed or restored wetlands to remove
nutrients and sediments from non-point source pollution (i.e. Mitsch, 1994; Raisin and Mitchell,
1995; Whigham, 1995; Jordan et al., 1999). Restoration of wetlands in agricultural watersheds
will provide wildlife habitat as well as improve water quality (Whigham, 1995). This study
shows multiple environmental benefits of wetland restoration. Wetland restoration could have a
domino effect depending on the location (i.e. water quality and wildlife could have an impact in
agricultural production and/or recreation activities (hunting) which could have an economical
effect). Wetland conservation and restoration could improve if all of these valuable
environmental benefits are promoted, recognized and better understood by the general public.
Crumpton et al. (1993) demonstrated the capacity of wetlands to trap nutrients upstream
and downstream. Upstream wetlands trap few nutrients, while down-stream wetlands can
potentially remove up to 80% of inflowing Nitrates.
2.4.4

Flood Water Storage
Numerous studies show that wetlands can decrease flooding, improve water quality, and

support biodiversity (Campbell and Johnson, 1975; Novitski, 1978; Thomas and Hanson, 1981;
Neely and Baker, 1989; Larson et al., 1989; Crumpton et al., 1993; Demissie and Kahn, 1993;
Richardson and Craft, 1993; Bedford, 1999; Keddy, 2000). Wetlands that are located close to
rivers, streams or adjacent water bodies can slow down runoff coming from storm water or snow
melt and to a certain degree contribute to the protection of populations in floodplains areas.
Wetland vegetation (e.g. trees, shrubs, and other wetland plants) will slow down runoff water by

14

slowing the flow which will allow ground water resources to recharge and sediments to be
trapped (Holden et al., 2007). The capacity for wetlands to decrease flooding is limited by
several factors such as: water level fluctuations, plant community and density, habitat elements,
ground water hydrology among other physical factors.
Wetlands will provide short term and long term water storage functions and assist with
the reduction of downstream flood peaks (Ramsar 2004; Mitsch, 1992; Potter, 1994; Hey et al.,
2002). Wetlands will moderate or prevent floods along a watershed area according to their
distribution and size. Ewel (1997) showed that maintaining integrity of wetlands by leaving
vegetation, soils, and natural water regimes intact could potentially reduce the severity and
duration of flooding along rivers. Investigations have shown that small wetlands could
potentially reduce and delay flood peaks serving as storage areas, while larger wetlands could
potentially reduce peak flow levels (Potter, 1994; Hey et al., 2002). Novitzki (1978) studied the
relationship between wetlands and flooding at a watershed scale in Wisconsin and found that
watershed areas with approximately 40 % coverage by lakes and wetlands had significantly
reduced flood flows (80 % less flood flow than similar watersheds having no or few lakes and
wetlands).
Wu and Johnson (2008) performed a hydrologic comparison between a forested and a
wetland/lake dominated watershed in northern Michigan using SWAT as the watershed
modeling tool. The results suggest an important storage function provided by wetlands and lakes,
in which they increase the ability of a watershed to moderate extreme flows and gradually
release water as baseflow.
Acreman and Holden (2013) described how wetlands affect floods from a hydrological
perspective of two wetland types; upland rain-fed wetlands and floodplain wetlands. The study

15

explained how multiple factors such as: landscape location and configuration, soil
characteristics, topography, soil moisture status and management will the ability of these
wetlands to provide flood reduction services.
2.4.5

Sediment Retention
Sediment particles are often vehicles for transporting pollutants such as nutrients (e.g.

nitrogen and phosphorus), pesticides and heavy metals. The accumulation of sediments at the
bottom of water bodies (e.g. reservoirs) could potentially have an impact on fish and aquatic life,
and could also reduce storage capacity in water reservoirs (Soler-Lopez et al., 2001a; 2001b;
2001c).
One of the principal external dynamic agents of sedimentations is that water serves as a
source of pollutant transport. The detachment of particles in the erosion process occurs through
the kinetic energy of raindrop impact or by the forces generated by the flowing water (Vanoni,
1997). Sediments are detached particles carried by rain water into streams, lakes, rivers and bays.
Sedimentation problems are observed in streams, lakes and other important water bodies used by
humans and wildlife. Some of the problems associated with sediment transport and deposition
are: movement of soil particles, loss of soil fertility, reduction of sun light penetration through
the water column, reduction in the reservoirs water storage capacity and reduction of dissolved
oxygen concentration. Sediments can also carry concentrations of pollutants that contaminate
waterways, including nutrients such as phosphorus and nitrogen which promote eutrophication in
surface waters.
The role wetlands play in trapping sediments and preventing them from reaching surface
water bodies is important. When positioned in stream networks, wetlands also mitigate
hydraulically driven variables including sediment, nutrients, temperature, and disturbance
16

(Richards et al., 1996). Wetlands play a unique and important role in improving environmental
quality as discussed previously. However, according to Kuenzler, 1990 wetlands should only be
used to remove sediments and other agricultural pollutants after agricultural best management
practices have been implemented.
Many studies show the capacity of wetlands to trap sediments as a water quality benefit.
Sediment input from agricultural fields has potential to completely fill wetlands and shorten their
effective life-span. Thus, the value placed on wetlands to trap sediments is in conflict with
maximizing the effective topographic life of wetlands (Gleason and Euliss, 1998).
2.4.6

Carbon Sequestration
Wetlands connected to rivers and slope locations are very productive. Their interaction

with streams make them significant sources of dissolved and particulate organic Carbon for
aquatic ecosystems and biogeochemical processes in downstream aquatic habitats (Sedell et al.,
1989 and Vannote et al., 1980).
It has been found that forested wetlands can offer a number of options for reducing Green
Houses Gases (GHGs), particularly carbon (C) emissions. They aid in the removal of CO2 from
the atmosphere into carbon pools (Cui et al., 2005). Several studies (Nieveen et al., 1998;
Schreader et al., 1998; Waddington and Roulet, 2000; Aurela et al., 2001; Lafleur et al., 2001,
2003; Bubier et al., 2004) related to forested wetlands have been developed to observe the
interactions between abiotic and biotic environmental factors and processes (e.g. methane
production and transport from wetland to the atmosphere). In addition, in wetland water level is
the major factor controlling carbon allocation, organic matter decomposition and C fluxes in
wetland (Kettunen et al., 1999; Weltzin et al., 2000).

17

2.4.7

Wildlife
Many studies have been done showing the benefits provided to wildlife by wetlands

(Shaw et al., 1956; Cowardin et al., 1979; Kantrud et al., 1989; Yerkes 2000). Wetlands are a
very significant ecosystem for wildlife. They provide shelter, food, and fish habitat, in addition
to other important factors (e.g. water and diversity). For example, wetlands and surrounding
upland areas provide breeding ducks and other waterfowl with the diverse habitat they need for
feeding, breeding, and nesting (Batt et al., 1989; Kantrud et al., 1989; Yerkes, 2000). These
wetlands can also have a high level of endemism, extensive plant zonation, and high
biodiversity. The high biological productivity of wetlands among other factors has produced a
rich biota associated to these ecological sties (Gibbs, 1995).
Annual flooding in low-gradient rivers and their adjacent flood plain wetlands constitute
a significant subsidy to physical habitat, vegetative communities, and populations of aquatic
organisms (Benke et al., 2000). In addition, seasonal water exchanges between lakes and coastal
wetlands and tidal fluxes between salt marshes, estuaries, and shallow marine areas create and
maintain productive habitats for a variety of plants and animals (Stevens et al., 2006). Most
freshwater and many marine aquatic organisms (birds, fish, insects) utilize wetland environments
at some stage of their development (Mitsch and Gosselink, 1993).

2.5

WETLANDS AND AGRICULTURE
Wetland issues (e.g. conservation, regulation, policies, and degradation) have been an

important part of agricultural and environmental policy debates at Federal, State and Local levels
for more than 25 years, when the Food Security Act regulated/protected wetlands. This has not
stopped the significant loss of these unique and valuable ecosystems in the United States (Dahl,
18

1990, 2007). Society has increased the value it places on the services provided by wetlands (e.g.
water quality improvement, flood control, wildlife habitat, and recreation), however owners of
wetlands often don not gain a profit directly from these services because the benefit is freely
enjoyed by many (Heimlich et al., 1998). These factors and the fact that wetlands are fertile and
productive soils for agricultural production trigger a land use change impact (wetland to
agricultural land) in many locations of the United States (e.g. Midwest Region: Michigan,
Wisconsin, Indiana).
Agriculture is the production of food and goods through farming. Land use changes are
required in order to perform agricultural activities in some locations (e.g. tiling for drainage,
creation of irrigation or drainage ditches), which has been one of the biggest causes of the
degradation of wetlands in the USA (Mitsch and Gosselink, 1993). Due to the high fertility of
wetland soils (e.g. rich content of organic matter) farmers converted over 28 million acres of
wetlands into high-quality cropland in nine Midwestern States since settlement (Heimlich et al.,
1998; Heimlich and Gadsby, 1994). Over 50 % of the area of depressional wetlands, riparian
zones, lake littoral zones and floodplains has been lost, mainly due to land use change as a
conversion to increase agricultural land use, in North America, Europe and Australia
(Millennium Ecosystem Assessment, 2005).
This conversion has not stopped in today’s society, where land use changes are dynamic
and agriculture and urbanization seem to be increasing every day. Increasing population and its
associated increase in demand for food and economic development will continue to create the
pressure to convert wetlands for farm use over the next several decades (Wood and van Halsema,
2008). Maintaining and improving the quality of wetlands is a very important goal because

19

wetlands provide multiple services to society, wildlife and the environment as it has been
discussed in previous sections of this document.
2.5.1

Crop Production
Wetlands are one of the primary sources of crop production, fiber and proteins in some

locations in the world (e.g. Africa and South Brazil). The value given to wetlands for agricultural
activities could be observed in any location of the world at any given time. For example, in the
United States during dry periods farmers can farm wetland areas within their farms without
changing the hydrology. In other locations farming activities are major economic pursuits in and
around wetland areas due to the high agricultural production value of these ecosystems. An
example of this would be areas where crops such as rice, maize, and various vegetables and fruit
are cultivated (Omari, 1993).
The use of chemicals and changes in crop production could potentially affect: soil
erosion, sedimentation in streams and reservoirs, pollution of surface and ground water waters
(NRC, 1989). Wetlands act as filters that can improve water quality (Kadlec and Knight, 1996)
and store sediments (Day et al., 2007) and will potentially have a positive impact in crop
production areas.
2.5.2

Livestock Grazing
Another agricultural activity that wetlands provide is livestock grazing areas. In locations

where seasonal wetlands are part of the ecosystem livestock grazing can continue during dry
seasons facilitated by the large source of biomass associated with these productive areas. These
areas could be grazed directly or used for hay production to feed livestock.

20

2.6

WETLANDS AND LANDUSE CHANGE
Land use/cover changes can alter watershed properties, such as: water infiltration rate,

water velocity, peak flow fluctuations, water storage capacity, and vegetation, as well other
important hydrological factors. This will affect sediment loads which could potentially have an
impact on downstream wetland ecosystems.
Chen et al. (2006) modeled the impacts of land use and land cover change on sediment
loads in wetlands in the Pouyang Lake Basin in Asia. GIS and Remote Sensing technologies
were used for the prediction of annual soil loss in the basin area in addition to a mathematical
model to study the relationship between land use/cover changes and sediment loads. They found
that the land use/cover change in the watershed reduced sediment sources and led to a reduction
in the suspended sediment concentration loads entering the wetland. Significant land use
alterations including the decrease of wetland by 43.55 percent for the study time period was
observed. The authors identified population growth, economy development and urbanization as
the social problems faced by the study area as well as the driving forces that led land use changes
in the studied watershed. Wetlands are also very sensitive ecosystems that are subject to much
stress from human activities (Bergh, 2001).
Previous studies related to wetland ecosystems have found a significant impact of the
land use/cover change on hydrological and fluvial processes. However, there is limited
information available regarding the effect of land use/cover change on water quality, especially
sediment load (Chen et al., 2006). Wetland ecosystems are susceptible to external influences
because of functional relations between hydrological characteristics in the ecosystem and the
surrounding area (Barendregt et al., 1995; De Mars and Garritsen, 1997).

21

2.7

WETLANDS CONSERVATION PROGRAMS
The hydrologic, geochemical and biological functions of wetlands are very important in a

watershed (De Laney, 1995; Hart, 1995; NRC, 1995). For that reason federal and state
governments, in addition to local and regional groups created programs for wetland conservation.
The United States Department of Agriculture (USDA) Natural Resources Conservation Service
(NRCS) has a conservation program to conserve, enhance and create wetlands in the United
States (USDA-NRCS, 2007). The USDA Farm Service Agency also has a conservation program
available to enhance wetlands in the United States.
Implementation of the USDA, CRP and WRP has resulted in the restoration of
approximately 2,200,000 ha (5,436,200 acres) of wetland and grassland habitats in the Prairie
Pothole Region (Gleason et al., 2008).
2.7.1

Wetland Restoration Program (WRP)
The Wetland Restoration Program (WRP) is a voluntary program which offers landowners

the opportunity to protect, restore, and enhance wetlands on their property (USDA-NRCS, 2007).
The WRP provides technical and financial assistance to enable qualified landowners to address
wetland, wildlife habitat, soil, water, and related natural resource concerns on private lands. The
WRP program focuses on: enrolling marginal lands that have a history of crop failures or low
yields, restoring and protecting wetland values on degraded wetlands, maximizing wildlife
benefits, achieving cost-effective restoration with a priority on benefits to migratory birds,
protecting and improving water quality, and reducing the impact of flood events.
In addition to the WRP, the USDA restores, enhances, and protects wetlands through other
conservation programs such as: the Small Watershed/Watershed Rehabilitation Program, the

22

Wildlife Habitat Incentives Program (WHIP), and the Emergency Watershed Program (EWP).
All of these conservation programs are promoted every year nationwide as part of the USDA’s
conservation strategy.
2.7.2

Conservation Reserve Program (CRP)
The Conservation Reserve Program (CRP) is a voluntary program which offers landowners

the opportunity to protect, restore, and enhance wetlands on their property (USDA-NRCS, 2007).
Under this program land is taken out of production and long-term, resource conserving cover
vegetation is established to control soil erosion, improve water and air quality, and enhance
wildlife habitat. The CRP Wetland Restoration Initiative is designed to restore the functions and
values of wetland ecosystems that have been devoted to agricultural use. These wetlands prevent
degradation of the wetland area, increase sediment trapping efficiencies, improve water quality,
prevent erosion and provide vital habitat for waterfowl and other wildlife.
2.8

WETLANDS CONSERVATION PRACTICES
Currently conservation practices are well promoted by government agencies and local

conservation groups (e.g. USDA-NRCS, US Fish and Wildlife Service, EPA, and Conservation
Districts) and accepted by landowners (e.g. dairy, fruits and cash crops farmers). Federal, state
and local agencies with a conservation vision will promote conservation and best management
practices within their target region. All of these conservation practices have something in
common which is addressing resource concerns (e.g. water quality and quantity, air quality, soil
erosion and plant deficiencies). These conservation efforts will assist landowners in improving
air and water quality, controlling/reducing erosion and sedimentation, reducing nutrient and
pesticide pollution; control/reduce impacts to stream ecology, morphology and habitat (USDA-

23

NRCS, 2006). Some of the conservation practices related to wetland ecosystems are discussed
below.
2.8.1

Wetland Restoration
Wetland Restoration is defined as the rehabilitation of a degraded wetland or the re-

establishment of a wetland so that soils, hydrology, vegetative community, and habitat are a
close approximation of the original natural condition that existed prior to modification to a
practical extent. The purpose of this practice is to restore wetland function, value, habitat,
diversity, and capacity to a close approximation of the pre-disturbance condition by: restoring
hydric soil, restoring hydrology and restoring native vegetation (USDA- NRCS, 2003).
2.8.2

Wetland Enhancement
Wetland Enhancement is defined as the rehabilitation or re-establishment of a degraded

wetland, and/or the modification of an existing wetland, which augments specific site conditions
for specific species or purposes, possibly at the expense of other functions and other species. The
purpose of this practice is to provide specific wetland conditions to favor specific wetland
functions and targeted species by: hydrologic enhancement, vegetative enhancement (including
the removal of undesired species, and/or seeding or planting of desired species) (USDA-NRCS,
2003).
2.8.3

Wetland Creation
Wetland creation is defined as the creation of a wetland on a site that was historically

non-wetland. The creation will provide wetland hydrology on a geomorphic setting that was not
originally wetland. Wetland creations usually have the highest cost and management

24

requirements. They are usually done for only one function such as providing wildlife habitat,
educational opportunities, or improving the quality of water from nonpoint source runoff. The
purpose of this practice is to create wetland functions and values (USDA-NRCS, 2003; USDANRCS, 2006).
2.8.4

Wetland Wildlife Habitat Management
Wetland wildlife habitat management refers to retaining, developing, or managing habitat

for wetland wildlife. These practices evaluate several wildlife elements including: food (type and
amount), cover (type, amount and quality), water (quality, quantity and accessibility)
interspersion and distance to crops, grasses or legumes, shrubs and trees and open areas), and
migration (routes, season of use and corridors). The main purpose of this practice is to maintain,
develop, or improve habitat for waterfowl, shorebirds, fur-bearers, or other wetland associated
fauna and flora (USDA-NRCS, 2003).
2.8.5

Wetland Construction
Constructed wetlands have been used for wastewater treatment for nearly 40 years and

have become a widely accepted technology available to deal with both point and non-point
sources of water pollution. They offer a land-intensive, low-energy, and low-operationalrequirements alternative to conventional treatment systems, especially for small communities
and remote locations (Vymazal et al., 1995).
2.9

WETLAND ASSESSMENTS
In today’s world, efficiency and cost are factors that are imperative to consider by private

and government agencies, planners, regulators, and the general public prior to conducting field

25

work in potential areas of concern, including wetlands. These factors have had a significant
impact in the scientific community including an interest in the development of tools to facilitate
environmental evaluations. Proper assessments of these potential areas aids in providing a clear
decision planning process and ensures that the most valuable wetlands are preserved, restored or
enhanced.
Several methods have been developed by the scientific community to better understand
wetland functions (Cowardin et al., 1979; Brinson, 1993; Gilver and McInness, 1994; Hruby,
1999.; Adamus et al., 1987; Leibowitz et al., 1992; Kent, 2001; Amman and Stone, 1991; Hruby
et al., 1995). These wetland assessment methods have been developed for multiple purposes (e.g.
planning and development, conservation projects and restoration practices). The process of
developing wetland assessment tools is not an easy task. However, these methods allow the
scientific community to examine and protect sensitive wetland areas within a short period of
time. These processes are very valuable for Federal and State Agencies that are currently
working together to protect, restore, enhance or create wetlands as part of a national initiative
(e.g. United States Department of Agriculture (USDA) Natural Resources Conservation Service
(NRCS), U.S. Army Corp, U.S. Fish and Wildlife Service, Michigan Department of Natural
Resources and the Environment (MDNRE), among others).
Some examples of methods used to analyze wetlands are: the Cowardin method, the
Washington State rating system, the Oregon method, the EPA synoptic method and the
Evaluation of planned wetlands method (Cowardin et al., 1979; Brinson, 1993; Gilvear and
McInnes 1994; Kent, 2001; Amman and Stone, 1991; Smith et al., 1995).
Hruby (1999) describes several of the above mentioned assessment methods that have
been used to better understand wetland systems. These methods include characterization,

26

classification/categorization, rating, assessment, and evaluation. Each data set obtained from
these methods is analyzed with a “model” or “algorithms” which uses either logic or a
mechanistic approach. Hruby also describes how the lack of model validation negatively affects
the model data output quality. He recommends finding ways to validate models that are more
time efficient and cost effective. He describes how multiple methods that generate a numeric
assessment of performance or value of wetland functions depend on the mechanistic approach of
constructing models. The author also describes how rapid assessment methods based on
mechanistic models do not assess the rates or dynamics of ecological processes taking place in
wetlands. These assessments provide a clear way of organizing our knowledge about wetland
functions.
Other wetland assessment methods that are used by either private and/or government
agencies, planners, regulators, or the general public include the Wetland Evaluation Technique
(WET), the Environmental Monitoring Assessment Program (EMAP-Wetlands) and the Hydrogeomorphic approach (HGM). The WET method assigns values to specific functions of
individual wetlands, was developed for the Federal Highway Administration, and has been used
widely (e.g. highways, restore/create wetlands). The WET method is applicable to all types of
wetlands in the contiguous United States (Adamus et al., 1991). EMAP-Wetlands, developed by
the EPA, focuses on determining the ecological condition of a population of wetlands in a
region. The HGM combines features from the aforementioned methods by measuring the
functions of individual wetlands and also by comparing them to functions performed by other
wetlands, this approach was developed by the U.S. Army Corps of Engineers for assessing
wetland functions.

27

Wheeler and Shaw (2000) used an extensive amount of data from over 80 wetlands in
Eastern England to develop a classification and assessment system called WETMECS. This
system combines landscape situation (location and condition), water supply mechanism, hydrotopographical elements, acidity and fertility to classify the hydrological process on wetlands.
The Army Corp of Engineers has a program dedicated to ecosystem management and
restoration research. Under this program, they have created multiple regional guidebooks for
applying the HGM approach to assess functions in multiple regions of the US. The HGM
approach is a method for developing functional indices and the protocols used to apply these
indices to the assessment of wetland functions at a site-specific scale. This approach was initially
designed by Brinson (1993) to be used in the context of the Clean Water Act, Section 404
Regulatory Program to analyze project alternatives, minimize impacts, assess unavoidable
impacts, determine mitigation requirements, and monitor the success of compensatory
mitigation. Smith et al. (1995) expanded the HGM concept including wetland functional
assessments using Brinson’s HGM classification system. Today the HGM has multiple potential
uses including the design of wetland restoration projects and the management of wetlands. The
hydro-geomorphic classification of wetlands is intended to lay a foundation for, and support
ongoing efforts to develop methods for assessing the physical, chemical, and biological functions
of wetlands. The strengths of this classification approach include its ability to clarify the
relationship between hydrology and geomorphology and wetland function, and its open structure,
which allows adaptation in various types of wetlands and geographic regions of the country.
In 1997 several federal agencies: the U.S. Army Corps of Engineers (USACE); U.S.
Department of Transportation, Federal Highway Administration; USDA, NRCS; U.S.
Department of Interior, U.S. Fish and Wildlife Service (USFWS), and the U.S. Environmental

28

Protection Agency (USEPA) agreed to use the HGM approach as a basis for wetland functional
assessments. Their methodology was to use the HGM classification as a tool for measuring
changes in the functions of wetland ecosystems due to impacts by proposed projects, restoration,
creation, and/or enhancement (Brinson et al., 1997).
2.10 WETLAND COMPUTER MODELS
Hydrological computer models are a very powerful and commonly used tool for
environmental studies. Hydrology-based computer models allow the scientific community to
better understand complex environmental processes such as the hydrological cycle and its
components. To represent these processes, modelers have adopted multiple approaches/systems
such as empirical or theoretical models. The empirical models are functional relationships
defined in terms of statistical data analysis of observed data, while theoretical models are
classified as functional relationships defined from physical laws and relationships (Heatwole,
1998). Empirical models have the limitation of being site specific which makes the applicability
limited or not applicable to other locations. On the other hand, theoretical models are adaptable
to different locations if properly calibrated.
Several computer models have been developed to simulate water quality and quantity at a
watershed scale (e.g. Soil and Water Assessment Tool (SWAT) by Arnold et al., 1998; Neitsch
et al., 2002; Hydrological Simulation Program-Fortran (HSPF) by Donigian et al., 1995;
Agricultural NonPoint Source Pollution Model (AGNPS) by Young et al., 1987,1989, 1994;
Areal Nonpoint Source Watershed Environment Response Simulation (ANSWERS) by Beasley
et al., 1980; ANSWERS-Continuous by Bouraoui and Dillaha, 1996; Bouraoui et al., 2002;
Dynamic Watersheds Simulation Model (DWSM) by Borah et al., 2002 and the PrecipitationRunoff Modeling System (PRMS) by Leavesley et al., 1983; Leavesley and Stannard, 1995).
29

In addition, hydrologic computer models are commonly used to estimate the impact of
Best Management Practices at a watershed scale. These models can potentially simulate the
characteristics of Best Management Practices (BMPs) as close to its physical conditions and
functional design (Renschler, 2007).
Borah and Bera (2003) reviewed some of the above mentioned models (AGNPS,
AnnAGNPS, ANSWERS, ANSWERS-Continuous, DWSM, HSPF, KINEROS, PRMS, and
SWAT) in their publication “Watershed-Scale Hydrologic and Non-point Source Pollution
Models: Review of Mathematical Bases”. AnnAGNPS, HSPF and SWAT are described as longterm, continuous simulation models that contain all of the three major components (hydrology,
sediment, and chemical), which are applicable to watershed-scale catchments (Arnold and Fohrer
2005; Neitsh et al., 2002; Neitsh et al., 2001; Arnold et al., 1998; Young et al., 1987, 1989;
Donigian et al., 1995). AGNPS and DWSM are storm event simulation models that also contain
all three major components.
SWAT is a river-basin or watershed-scale model for continuous simulations in
predominantly agricultural watersheds, and HSPF is suitable for mixed agricultural and urban
watershed conditions (Arnold et al., 1998; Neitsch et al., 2002). AGNPS is similar to SWAT and
it is an event-based model simulating runoff, sediment, and transport of Nitrogen, Phosphorous,
and chemical oxygen demand (COD) resulting from single rainfall events (Young et al., 1987,
1989). Conversely, we have the single-event models like DWSM, which is a potential model for
agricultural and rural watersheds. It simulates distributed surface and subsurface storm water
runoff, propagation of flood waves, upland soil and streambed erosion, sediment transport, and
agrochemical transport in agricultural and rural watersheds during single rainfall events (Borah
et al., 2002).

30

2.10.1 Computer Models Strengths and Deficiencies
The use of computer models in today’s civilization is advanced, extensive and diverse.
This makes finding more than one computer watershed model for addressing any practical
problem much simpler. Computer models for watershed simulation allow researchers and
scientists to evaluate and compare data outputs for validation using more than one available
model, which provides a more broad approach. The variety in watershed computer models is
one of the major strengths we have in today’s technology.
The integration of soils, water, animal, plant, and air (SWAPA) data, ecosystem and
ecology, environmental components, in addition to other factors with hydrology is an additional
strength of watershed computer models. The wide range of computer models that are available
and the applications of these models make them flexible and robust for the water quality study
area.
Even though watershed computer models have become an important part of the scientific
and research community there are many deficiencies that need to be addressed. Some of these
deficiencies are the lack of user-friendliness, extensive data inputs, data management, the
complexity of model calibration procedures and the integration of social, political, economic,
and environmental systems within the models.
The inefficiency of computer models in the prediction of multiple water quality
parameters has been a deficiency according to recent studies (Benaman and Shoemaker, 2004;
Conan et al., 2003). Nasr et al. (2007) compared SWAT, HSPF and SHETRAN/GOPC for
modeling phosphorus export from three catchments in Ireland. The three mathematical models:
SWAT, HSPF and a European model known as “système hydrologique Européen TRANsport”
(SHETRAN)/grid oriented Phosphorus component (GOPC) were used for this study. The results

31

after model calibration of daily flows and total (TP) outputs were compared and assessed. It was
found that the HSPF model was the best at simulating the mean daily discharge while SWAT
gave the best calibration results for daily TP loads. The study showed that no single model is
consistently better in estimating the annual TP export for all three catchments.
In addition, models have some downfalls with temporal resolution. For example, SWAT
operates only at a daily time step compared to HSPF which can simulate at any time step from 1
min up to 1 day. This will also be linked to data availability and management because data is not
always available at intervals equal to or less than the simulation time step.
All of these deficiencies require a solution to obtain better performance of the model. The
ability of users to modify these models for specific climate and locations is essential during this
process.
2.10.2 Factors to Consider in Wetland Modeling
Selecting a hydrological computer model that will fit a specific research area is a process
that takes time and dedication. The ability to understand the hydrology of wetlands and their
characteristics (e.g. wetland type, soils, plants) is very important and can assist during the model
selection process.
According to the Hydrologic Engineering Center 1988, the factors that should be taken
into consideration when modeling a wetland include: (1) the location of the wetland within the
watershed, (2) hydrology, (3) water retention periods of the wetlands (entire year, or just during
wet periods), (4) amount of vegetation in the wetland (this can affect evapotranspiration), (5)
storage and infiltration capabilities of the wetland, and (6) spatial variation of the landscape.
Other factors are also important to consider, such as wetland age.

32

Hydrologic models often have different modeling capabilities. In order to find the best fit
for a hydrological research interest, one will need the model components. There are multiple
components incorporated within each hydrologic model including the following parameters:
precipitation, snow accumulation and snowmelt, evapo-transpiration, interception, infiltration,
surface drainage and runoff, depression storage and routing, subsurface soil water flow and
channel routing (U.S. Army Corps of Engineers, 1998). Computer models offer multiple options
and capabilities for modelers to select the most appropriate for their specific needs and project
goals. It is up to the modeler/scientist to determine which model meets their criteria. A limitation
during the selection of a computer model could be data availability, quality of data and
quantification of data in addition to other restrictions.
2.11 COMPUTER MODELS FOR WETLANDS AND WATER QUALITY MODELING
Wetland models have been developed to assist the scientific community in understanding
processes that occur in multiple wetland types (e.g. costal, swamps, peatland and ponds). Some
of these studies were conducted for hydrology processes (Hammer and Kadlec, 1986; Walton et
al., 1996), while others studied nutrients (e.g. N and P) (Widener, 1995; 1994; Dorge et al.,
1994). Even when studies have been conducted for these types of ecosystems, modeling wetland
processes is relatively new as compared to other ecosystems (Mitsch et al., 1988).
There are three approaches used to model wetland hydrology: single event models,
stochastic models, and comprehensive water budgets (Koob et al., 1999). These models are
classified into different categories: one-dimensional, two-dimensional or three-dimensional.
These hydrologic computer models have been used in the development and
implementation of total maximum daily load (TMDL) standards and guidelines that are required
by the Clean Water Act (CWA), simulation of floods, and modeling the hydrodynamics of
33

wetlands for flood detention, water quality, water quantity and climate change as well as others
(Neitsch et al., 2002, Borah et al., 2002; Arnold and Fohrer, 2005; Hattermann et al.,
2008;Watson et al., 2008; Liu et al., 2007; Gassman et al., 2005; Conan et al., 2003).
Although some studies have attempted to describe wetland hydrology (Konyha et al.,
1995; Reinelt and Horner, 1995; Hawk et al., 1999; Arnold et al., 2001; Zhang and Mitsch,
2005), there is a limitation in the availability of computer models capable of describing wetland
water flows (e.g. urban) (Drexier et al., 1999; Raisin et al., 1999). This limitation is a critical
factor because most current wetland modeling applications are derived from traditional pond
design engineering (Konyha et al., 1995), which is denoted to be used for wetland modeling.
There are serious limitations to this approach when modeling wetland water fluctuations, which
are characteristically more subtle than water movement captured by pond models (Obropta,
1998). One example of this particular case is the SWAT model which has a pond simulation
capability and lake/wetland algorithms integrated for wetland simulations, but not a specific
application for these types of ecosystems (Wu, 2007). The modification of a hydrological
computer model’s applications related to wetlands or similar areas is an alternative to improve
hydrological simulation process obtain better results.
Some of these computer models (e.g. SWAT, HSPF) have incorporated wetland modules
as part of their applications (Wang et al., 2010; Hattermann et al., 2008; Watson et al 2008; Liu
et al., 2007; Gassman et al., 2005, Conan et al., 2003). For example, the Topographic-based
Nitrogen Transfer (TNT) and transformation model has been used to assess the effect of riparian
wetlands, (Beaujouan et al., 2001). Having these computer models available to assist those in the
scientific community who are interested in wetlands is a great approach for protection,

34

conservation and restoration efforts. Unfortunately, representation of wetland processes in these
computer models is not sufficient and requires more consideration.
Okruszko et al. (2006) used the GIS platform for combining the results of hydrological
models and assessing the ecological consequences of a restoration project on wetlands in Poland.
The author identified three hydrological models: linear optimization model for control of
hydraulic structures, 1-D hydrodynamic model of the surface waters for flood simulation, and a
regional groundwater model (SIMGRO) for predicting the changes in groundwater level. GIS
was used as a data management tool for all of the hydrological models. Their goal was to
simulate the impact of different restoration measures in the area on hydrological regimes of
rivers and riparian wetlands.
The authors found an affirmative relationship between land use changes and potential restoration
projects, which could be used as a tool to demonstrate the benefits of ecological restoration to
farmers.
The following section will describe general modeling approaches and techniques, in
addition to presenting the detailed modeling processes for wetlands used by researchers.
A variety of computer models used to model wetlands or wetlands properties will be
discussed. The following wetland model descriptions are included to present an overview of
approaches taken to model multiple processes related to wetlands.
2.11.1 System for Urban Stormwater Treatment and Analysis INtegration
The System for Urban Stormwater Treatment and Analysis INtegration (SUSTAIN)
model is a decision support system developed by the United States Environmental Protection
Agency (USEPA) to evaluate alternatives for water quality management. The SUSTAIN model
incorporates the best available research that could be practically applied to decision making
35

processes, including the tested algorithms from other watershed scale hydrologic and water
quality simulation computer models such as the Storm Water Management Model (SWMM) and
the Hydrologic Simulation Program – FORTRAN (HSPF) (EPA, 2009). The SUSTAIN model
simulates the ability of individual or a combination of BMP’s in reducing stream peak flow,
nutrients and sediment concentrations taking in consideration watershed characteristics (e.g. soil
type, precipitation, temperature, and landuse). SUSTAIN includes algorithms for simulating
urban hydrology and pollutant loading. The incorporation of these methods provide effectiveness
and balance between computational complexity and practical problem solving (USEPA, 2009,
2010). SUSTAIN also incorporates an advanced approach that allows for cost effectiveness
evaluation of both individual and multiple nested watersheds to address the needs of both localand regional-scale applications (USEPA, 2009, 2010).
2.11.2 Soil and Water Assessment Tool
The Soil and Water Assessment Tool (SWAT) model (Arnold and Fohrer, 2005; Neitsh
et al., 2001, 2002; Arnold et al., 1998) is a river basin, or watershed, scale model developed for
the USDA Agricultural Research Service (ARS). SWAT could be used as part of the USEPA’s
Better Assessment Science Integrating point and Nonpoint Sources (BASINS) (Lahlou et al.,
1998) for TMDL analyses (Di Luzio et al., 2002; Mausbach and Dedrick, 2004). BASINS is a
watershed and water quality−based assessment system that integrates geographical information
system (GIS), national watershed data, and environmental assessment and modeling tools like
SWAT into one package (U.S. EPA, 2001). SWAT has proven to be very efficient in predicting
impacts of management practices on water, sediment, and agricultural chemical yields in large,
un-gauged watersheds (Gassman et al., 2007).This model was developed to predict the impact of
land management practices on water, sediment, and agricultural chemical yields in large,
36

complex watersheds with varying soils, land use, and management conditions over long periods
of time. The model is physically based and computationally efficient; it uses readily available
inputs and allows users to study long-term impacts.
The model features include watershed hydrology, sediment and water quality, pesticide
fate and transport simulation, channel erosion simulation, and rural and agricultural management
practices (e.g. agricultural land planting, tillage, irrigation, fertilization, among others). SWAT
subdivides a watershed into a number of sub-basins. Portions of a sub-basin that possess unique
land use/management/soil attributes are grouped together and defined as one hydrologic response
unit (HRU; Neitsch et al., 2002). Each sub-basin is simulated as a homogenous area in terms of
climatic conditions, and each HRU is assumed to be spatially uniform in terms of soils, land use,
and topography.
The model possesses multiple strengths such as: great documentation, physically based,
GIS interface (BASINS), high-quality land management modules and databases and is suitable
for studying watersheds ranging in scale from small to very large. As any other modeling
application the SWAT model has limitations such as it could not be used for simulating sub-daily
events (e.g. single storm events), it is useful only for simulating conservative metal species from
the point source input, it cannot specify actual areas to apply fertilizers, and it may require
extensive data input and management. Some of the data inputs are: land uses, soils, topography
(Digital Elevation Models (DEM)), sub-watersheds, point sources of pollutant discharge, climate
and weather data, crop data, and long term water quality and flow data. SWAT requires a
significant amount of data and empirical parameters for development and calibration (Benaman
et al., 2001).

37

SWAT has been modified to improve the simulation of specific processes at a watershed
scale for different conditions. Gassman et al., 2005 discussed several examples of these
modifications (e.g. Extended SWAT (ESWAT)), which features enhanced in-stream kinetics and
other modifications (van Griensven and Bauwens, 2001; 2005), the Soil and Water Integrated
Model (SWIM), which is partially based on SWAT (Krysanova et al., 1998, 2005) and
SWATMOD, a version of SWAT that has been linked to MODFLOW to simulate detailed
surface/groundwater interaction (Sophocleus et al., 2000). All of these are examples of how this
powerful watershed modeling tool has been advanced to resolve more complex environmental
evaluations.
Borah and Bera (2002) reviewed several fully developed hydrologic and non-point source
pollution models (e.g. SWAT, HSPF, AGNPS, AnnAGNPS, ANSWERS, ANSWERSContinuous, PRMS, KINEROS, DWSM, and CASC2D among others) and found SWAT to be
one of the most capable models for long-term continuous simulations in predominantly
agricultural watersheds. The model was found suitable for predicting yearly flow volumes,
sediments and nutrient loads. Borah and Bera found that the model daily predictions were
generally poor. In addition, it was found that SWAT should be combined with other single event
models (e.g. DWSM, ANSWERS, KINEROS) for adequately simulating the extreme single
storm events.
2.11.3 SWAT Applications on Wetlands
SWAT applications have been used to study wetland hydrology, functions, and
relationships with water quality and quantity. However, appropriately represent wetlands in
models is challenging, and few SWAT applications reported in the literature have considered
wetlands (Wang et al., 2008). For that reason the incorporation of SWAT into wetland studies is
38

an important issue to be addressed and more research involving the simulation of wetlands
ecosystems should be conducted.
Wang et al. (2010) simulated the effects of wetland conservation and restoration on
water quality and quantity at a watershed scale in Minnesota. Wang used SWAT in addition to
the hydrologic equivalent wetland (HEW) developed by Wang et al., 2008. This combination of
modeling tools can be used to consistently predict effects of wetland conservation/restoration
scenarios and to logically prioritize restoration efforts (Yang et al., 2008). SWAT treats wetlands
as water bodies located within sub-basins (Arnold et al., 2001; Neitsch et al., 2002a) and allows
one wetland at a time per subbasin modeled. The objective of this study was to use the HEW
concept in SWAT to assess effects of wetland restoration within the Broughton's Creek
watershed in addition to wetland conservation within the upper portion of the Otter Tail River
watershed. It was found that the HEW concept allows non-linear functional relations between
watershed processes and wetland characteristics (e.g. morphology) to be accurately represented
in the models. It was found that a reduction (loss) of approximately 10 to 20 percent of the
wetlands in the study area would considerably increase the peak discharge and loadings of
sediment, total Phosphorus, and total Nitrogen. The author compared wetland conservation
versus wetland restoration and described that wetland conservation deserved a higher priority
compared to wetland restoration.
Hattermann et al. (2008) compared two approaches that allow integration of important
wetland processes using the Soil Water Infiltration and Movement (SWIM) model. The SWIM
model allows addition of water to the system as precipitation and removal by runoff, drainage,
evaporation from the soil surface and transpiration by vegetation. These approaches evaluate
water and nutrients fluxes at different levels of complexity allowing modeling results to be

39

improved in terms of seasonal river discharge and nutrient loads in catchments with wetlands. In
addition, these approaches evaluated by Hattermann et al., 2008 are compatible to other models
and can be used for the integration of wetland processes for regional applications.
Wu and Johnson (2008) performed a hydrologic comparison between a forested and a
wetland/lake dominated watershed in northern Michigan using SWAT as the watershed
modeling tool. The specific objectives of this study were to calibrate SWAT to simulate
streamflow and compare the effects of wetland and lake abundance on the magnitude and timing
of streamflow from two watersheds (East branch and Middle branch of the Ontonagon River
basin). SWAT treats wetlands as water bodies located within sub-basins (Arnold et al., 2001;
Neitsch et al., 2002a). The study shows that the watershed containing greater wetland and lake
areas had lower spring peaks and higher sustained flows during summer and fall.
Conan et al. (2003) found that SWAT adequately simulated land use change from
wetlands to dry land in Spain (Upper Guadiana river basin) and evaluated the impact of ground
water withdrawals in the studied basin. The model showed misrepresentation of data for certain
conditions, which could be related to the lack of sufficient data (e.g. rainfall data). Another
limitation was that the model was unable to represent all of the discharge details impacted by
land use changes.
Arnold et al. (2001) studied a hydrologic model for design and constructed wetlands near
Dallas, Texas (Trinity River). For this study the SWAT model was used to assess flow following
a heavy precipitation event and base flow in regards to a wetland ecosystem. The model was run
for 14 years and was compared to observed data from the nearby watershed. The model results
indicate that the wetland should be at or above 85 percent capacity over 60 percent of the time,
showing no signs of dryness during the 14 years of simulation.

40

Wang et al. (2008) used hydrologic-equivalent wetland (HEW) concepts within the
SWAT model to estimate streamflow in a watershed with numerous wetlands in Minnesota. One
of the objectives of this study was to demonstrate how to incorporate wetlands into a SWAT
model using the HEW concept. The HEWs were defined in terms of six calibrated parameters:
the fraction of the sub-basin area that drains into wetlands (WET_FR), the volume of water
stored in the wetlands when filled to their normal water level (WET_NVOL), the volume of
water stored in the wetlands when filled to their maximum water level (WET_MXVOL), the
longest tributary channel length in the sub-basin (CH_L1), Manning's n value for the tributary
channels (CH_N1), and Manning's n value for the main channel (CH_N2). Statistical methods
(the Nash-Sutcliffe coefficient (E j 2), the coefficient of determination (R 2), and the
performance virtue (PVk)) were used to evaluate the model’s performance. The results indicated
that the HEW concept is superior in incorporating wetlands into SWAT for the study area.
Overall, the SWAT model with the HEW assumption had an acceptable or satisfactory
performance in simulating the streamflows at daily, monthly, seasonal, and annual time steps.
Yang et al. (2008) completed a research for Ducks Unlimited in Canada related to water
quantity and quality benefits from wetland conservation and restoration. The SWAT-based
modeling system was applied to examine the effects of wetland conservation and restoration in
the Broughton’s Creek watershed. Multiple scenarios were used for a better simulation; all
scenarios included wetland areas ranging from 2,379 ha in the year of 2005 to 2,998 ha in the
year 1968. The results show that the peak discharge at the watershed outlet was predicted to be
reduced by 1.6% to 23.4%, and the sediment loading was reduced by up to 16.9%. This study
estimated reductions of TP and TN at the watershed outlet as a result of wetland restoration.

41

These estimated reductions each are equivalent to 2.4% to 23.4% of the existing TP or TN export
out of the study watershed.
2.11.4 Hydrologic Simulation Program-Fortran
The Hydrological Simulation Program – Fortran (HSPF) (Donigian et al., 1995) is a U.S.
EPA program for simulation of watershed hydrology and water quality for both conventional and
toxic organic pollutants. HSPF is a comprehensive, continuous, lumped parameter,
watershed−scale model that simulates the movement of water, sediment, pesticides, and nutrients
on pervious and impervious surfaces, in soil profiles, and within streams and well−mixed
impoundments. HSPF allows the user to simulate selected water quality constituents by
specifying their sources, sinks, chemical properties, and transport behavior. (Bicknell et al.,
2000).
HSPF has been incorporated as a non-point source model (NPSM) into the U. S. EPA’s
BASINS. BASINS is use to analyze and develop TMDL standards and guidelines nationwide
(U.S. EPA 2009, Whittemore and Beebe, 2000). HSPF has become a useful tool for water
resource planners, because it is more comprehensive than other modeling systems available
which allows more effective planning.
The model features include: time-series-oriented model for easier data management, the
unified structure makes it simple to operate, easy to modify and extend, and extensive research
and uses of the model are available to use as guidelines.
2.11.5 HSPF Applications on Wetlands
HSPF is one of the two most widely used water quality models for simulation of
hydrology and water quality nationwide (Bicknell et al., 1997). The model has been modified to

42

take into account multiple physical and agricultural databases, which makes it more accessible
and robust.
Schwar et al. (1998) studied the restoration of rivers and wetlands using hydrologic
design. HSPF was used as the continuous hydrologic modeling tool. This computer model offers
the potential to design and evaluate restoration projects based upon the goals of the project (e.g.
magnitude, duration and seasonality of streamflows or water levels). The application of computer
models for restoration goals is very valuable providing data during the decision making process.
Davis (2001) studied the integration of two hydrologic models: HSPF and MODFLOW
in Florida. The results shows the advantages of the integrated hydrologic model, and how the
results of the model can be used to help predict future impacts of groundwater production on
wetland conditions in those areas where wetlands interact with the shallow groundwater
environment.
2.11.6 Storm Water Management Model
Obropta et al. (2008) simulated urban wetland hydrology for the restoration of a forested
riparian wetland ecosystem. The U. S. EPA Storm Water Management Model (SWMM) was
used to characterize water movement through multiple sub-basins in the Teaneck Creek. The
main objectives were to develop a conceptual restoration plan for the site and predict surface
water movement through the re-established wetlands using multiple data inputs (e.g. surface
water flow, ground water flow, vegetation and soils data). The authors developed a methodology
for analyzing water budget in urban wetlands. The results showed that the SWMM model can be
used to analyze water budgets for individual wetland basins. In addition, the model can be used
to analyze each wetland basin, separately or in combination. This will allow the evaluation of the

43

effects of restoration projects (e.g. installing water-control structures). The model was effective
at analyzing nutrient loadings in wetland areas.
Tsihrintzis et al. (1998) simulated wetlands hydrodynamics for flood detention in South
Florida. For this study the SWMM-EXTRAN link-node model (Roesner et al., 1989) was used.
The objectives of this study were: to test the applicability of the SWMM model on wetland
hydrodynamics and to apply the calibrated model, using it with synthetic storms in wetland
design for flood control. Similarities were found between predicted flows and water surface
elevations when compared to observed values. The results showed that the model could
potentially be used as a design tool to optimally size hydraulic structures connecting wetland
areas.
2.11.7 WETLANDS-2D
WETLANDS-2D (flow and transport in variably saturated porous media) is a
mathematical model for one or two dimensional water flow and solute movement in variablysaturated multi-layered porous media featuring optional surface water bodies (ponds) and
multiple root zones. The model supports axially symmetric 2D systems as an approximation to
three-dimensional systems. The model is a modification of the Variably Saturated 2-D Flow and
Transport Model (VS2DT). The VS2DT is a two-dimensional finite difference simulator for
cross-sectional or cylindrical variably saturated flow in porous media. The model allows
consideration of non-linear storage, conductance, sink terms and boundary conditions. Processes
included are infiltration, evaporation and plant root uptake.

44

2.11.8 Wetland – DNDC
Wetland-DNDC is a computer simulation model of water, Carbon and Nitrogen
biogeochemistry in forested wetland ecosystems. The model can be utilized for estimating forest
production, ecosystem C dynamics and emissions of trace gases including methane (CH4),
nitrous oxide (N2O), nitric oxide (NO), Nitrogen gas (N2), and ammonia (NH3).
Wetland-DNDC was constructed by integrating hydrological and forest biogeochemical
processes at site and watershed scales. Wetland-DNDC has been calibrated and validated at site
scale against numerous field data sets measured in forest ecosystems in North America and
Europe (Stange et al., 2000; Butterbach-Bahl et al., 2001; Zhang et al., 2002). In WetlandDNDC, forest growth is simulated by tracking photosynthesis, respiration, C allocation, N uptake
and water demand at a daily time step.
Cui et al., 2005 linked two models together (MIKE SHE to Wetland-DNDC) for carbon
budgeting and anaerobic biogeochemistry simulation in forested wetlands. The wetland-DNDC
model was modified by utilizing parameters for management measures, refining anaerobic
biogeochemical processes, and linked to the hydrological model – MIKE SHE. Simulation
results from this model show that water table changes had a notable effect on Green House
Gases’ (GHGs) fluxes. It was found that anaerobic conditions in forested wetland soils reduce
organic matter decomposition and stimulate CH4 production. Results show that average longterm carbon storage in ecosystem pools increased with increasing rotation length: Soil carbon
showed only minor, long-term responses to harvesting events. It was shown that linking these
two models would allow the assessment of GHGs mitigation strategies, carbon budgeting and
forest management. The basic functions adopted by Wetland-DNDC for simulating forest

45

growth and soil biogeochemistry processes have been well validated in previous publications (Li
et al., 1992; Aber et al., 1996; Li et al., 2000).
2.11.9 Hydrologic Engineering Center – Hydrologic Modeling System
Hydrologic Engineering Center-Hydrologic Modeling System (HEC-HMS) was
developed by the U.S. Army Corps of Engineers (U. S. Army Corps of Engineers 1998). HECHMS essentially replaces HEC-1; it provides numerous options for simulating precipitation
runoff processes. This new program offers the ability to perform continuous hydrograph
simulations over long periods of time. It accomplishes this through the use of a “single-reservoir
soil-moisture representation”. It also computes spatially distributed run-off values using a ‘grid
cell’ depiction of the watershed. HMS can simulate the rainfall-runoff at any point within a
watershed if the physical characteristics of the watershed are utilized. It is a tool for watershed
management in that an HMS model can be developed to account for the human impact and to
determine the effect on the magnitude, quantity, and timing of runoff at points of interest. Results
from an HMS model can be used by a number of other programs to determine impact in areas
such as water quality and flood damage.
2.11.10Precipitation-Runoff Modeling System
PRMS (Precipitation-Runoff Modeling System) is a distributed watershed model that
simulates precipitation and snowmelt driven movement of water through the basin via overland
flow, interflow, and baseflow. The model was developed by the United States Geological Survey
(Leavsley and Stannard, 1995). Watershed response can be simulated at a daily time step or more
frequently over the course of a storm. Kinematic routing of the unidirectional flow and the
transport of sediments through a receiving network of well-mixed channel reaches can be

46

simulated when the model is in "storm mode". Simulation of the energy balance in the snowpack
and the water balance is based on many theoretically and empirically developed relations. The
resulting model is comprehensive and flexible, but also very complex and requires a large
number of parameters. The model contains procedures for parameter optimization and sensitivity
analyses. Some of the model inputs include: daily precipitation, maximum and minimum air
temperature and solar radiation data. This model takes into account snowmelt and uses air
temperature and solar radiation data to compute this process as well as the processes of
evaporation, transpiration and sublimation.
2.11.11Environmental Fluid Dynamics Code
The EFDC (Environmental Fluid Dynamics Code) is a 3-dimensional surface water
model for hydrodynamic and water quality simulations in rivers, lakes, reservoirs, wetland
systems, estuaries, and the coastal ocean. The basic physical process simulation capabilities of
EFDC are similar to the Blumberg-Mellor Model (ECOM3D) and CH3D-WES. Notable
extensions included in EFDC include representation of hydraulic structures for controlled flow
systems, vegetation resistance for wetlands, and high frequency surface wave radiation stresses
in near shore zones. The model solves the hydrostatic, turbulent-averaged equations on an
orthogonal curvilinear horizontal grid and a sigma-stretched vertical grid. EFDC transports
salinity, heat, cohesive or non-cohesive sediments (only one sediment class at a time), and toxic
contaminants that can be described by equilibrium partitioning between the aqueous and solid
phases.

47

2.11.12TABS-2
TABS-2 Horizontal two-dimensional model (Thomas and McAnally, 1990) is a
generalized numerical modeling system for open-channel flows, sedimentation, and constituent
transport. This model is a two-dimensional, depth averaged, finite element hydrodynamic
numerical model developed by the Waterways Experiment Station of U.S. Army Corps of
Engineers. It computes water surface elevation and horizontal velocity components for
subcritical, free-surface flow in two-dimensional flow fields. This model has been applied to a
variety of waterways, including rivers, estuaries, bays, and marshes (wetlands).
Lee and Shih (2004) studied the impacts of vegetation changes (mangrove removal) on
the hydraulic and sediment transport characteristics in a wetland in the Guandu Natural Reserve,
Taiwan. A horizontal two-dimensional model, TABS-2, was applied as part of this study to
simulate the hydraulic and sediment transport characteristics of this wetland. The authors found
that the optimal removal ration was between 10% and 20% according to the variations of the
hydraulic and sediment transport simulation from TABS-2.
NHCP, CODE 600, 2010). Sediment reduction in runoff is achieved through decreasing
length of the hill slope to decrease peak runoff rate, increased settling of sediments in surface
runoff, and interception and retention of water (Arabi et al., 2007; Tuppad et al., 2010).
2.12 WETLANDS AND GEOGRAPHICAL INFORMATION SYSTEMS (GIS)
Many studies related to wetlands and the integration of GIS have been conducted as part
of efforts to better understand these ecosystems (Lyon and Adkins, 1995; Lyon and McCarthy,
1995; Ji and Mitchell 1995; Ramsey and Jensen, 1995; Williams and Lyon 1995; Cedfeldt et al.,

48

2000; Tiner, 2003; Van Lonkhuyzen et al., 2004, White and Fennessy, 2005; Gibson, 2006;
Torbick et al., 2008).
Lyon and Adkins (1995) developed a wetland prediction model for the St. Clair River in
Michigan. The model used GIS and RS technologies to identify wetland locations and then
estimate landuse changes for a period of time (1974 to 1985). The prediction model was based
on a linear equation which combines the data layers and a constant error term to produce a
predicted wetland type. The author proved that GIS/RS technologies could be integrated into
wetland research.
Cedfeldt et al. (2000) developed an automated wetland assessment methodology using
GIS technologies to identify functionally significant wetlands in Vermont. Cedfeldt used
wetlands functions to designate a wetland as functionally important (flood flow alteration,
surface water quality improvement and wildlife habitat). This model showed the ability to
identify potential restoration or mitigation sites.
Gibson (2006) developed a GIS model for potential riparian wetland restoration sites in
Ohio. The model used secondary GIS data to identify and prioritize potential restoration sites
with a restoration category value. The classification for restoration categories showed 68.1% of
the studied sites being classified correctly. According to the author, this model is a potential
planning tool that should be employed in the future in restoration projects.
Torbick et al. (2008) studied the application and assessment of a GIS model for
jurisdictional wetlands identification in Northwestern Ohio. During this study it proved the
effectiveness of the GIS tool’s application to wetlands (accuracy values ranged from 55% to
84%). A wetland classification system was developed to identify the dwindling wetland land
cover types which provided accurate and detailed maps for the county.

49

2.13 WETLANDS AND REMOTE SENSING APPLICATIONS
Remote sensing (RS) technologies in combination with GIS applications allow the
scientific community to better understand environmental processes (e.g. hydrology and climate
change). The utilization of remote sensing technologies for multiple purposes including wetland
inventory and identification has proven to be a useful and commonly used application (i.e.
Lunetta and Balogh, 1999, Townsend and Walsh, 2001).
Remote sensing technologies have been used for years to assess multiple environmental
conditions in the landscape (e.g. sedimentation, loss of vegetation index, and land use change).
An example of RS applications could be to correlated precipitation or runoff data with aerial
photography, commonly U.S. Department of Agriculture (USDA) crop history slides. Using this
tool, it is possible to determinate the number of times that wet signatures (e.g. standing water,
soil saturation, and stressed crops) are visible at a specific site on a series of aerial photographs
taken over a number of years. Using aerial photography and observed data allows RS
classifications to have accuracy levels of (69%-88%). A study conducted by Townsend and
Walsh (2001) showed an accuracy of approximately 90% when classifying thematic mapper
(TM) data for forested wetlands using a series of three season imagery.

50

3

INTRODUCTION TO METHODOLOGY AND RESULTS
This dissertation is in the form of two research papers that have been submitted to

scientific journals. The first paper is entitled “Modeling the hydrologic significant of wetland
restoration scenarios”, and the second paper is entitled “Assessing the significance of wetland
restoration scenarios on sediment mitigation plan”. Both papers are related to the development of
large-scale wetland restoration plans. All of the wetlands simulated in these studies are based on
restoration of natural wetland systems (hydrological restoration).
The first paper utilized a physically-based model, the Soil and Water Assessment Tool
(SWAT), to quantify the impacts of wetland restoration scenarios in the Shiawassee River
watershed located in Michigan. Initially, field surveys were performed in multiple locations in
the watershed to parameterize the wetland model. After model calibration and validation, the
restoration scenarios included 1424 model runs for all possible combinations of wetland depths
(15, 30, 61, and 91 cm), areas (50, 100, 250, 500 ha), and placements (89 subbasins) were
evaluated. Each wetland area and depth combination was implemented one-at-a-time for each
subbasin. Wetland volume in SWAT was calculated on a daily time-step using the water balance
concept over 19-year-period. The reduction in streamflow and peak flow at the watershed outlet
was calculated. A statistical analysis was performed to identify significant differences regarding
the impact of wetland area and depth on streamflow reduction at the watershed outlet. Finally,
high impact areas for wetland restoration were identified for two separate goals: (1) flow
reduction at the watershed outlet and (2) flow reduction at a specific location within the
watershed.
The goal of the second paper was to examine the sediment reduction benefit of wetland
restoration scenarios at the subbasin and watershed levels in the River Raisin watershed of

51

Michigan. To accomplish this goal, two models were utilized. The System for Urban Stormwater
Treatment and Analysis Integration (SUSTAIN) was used to evaluate the local scale benefits of
wetland installation in reducing sediment loadings, while SWAT was used to estimate sediment
loads to and from wetlands and at the watershed outlet. Multiple wetland restoration scenarios
(4480 simulations) were developed to assess the importance of wetland areas (0.4, 0.81, 2 and 4
ha) on sediment reduction. Statistical analyses were performed to explore the effect of stream
order, distance from the outlet, and wetland surface area on mean annual sediment reduction.
Finally, the analytic hierarchy process (AHP) was used to incorporate environmental and
economic aspects of wetland restoration scenarios into the selection of the most cost effective
wetland surface area for implementation in each subbasin.

52

4

MODELING THE HYDROLOGICAL SIGNIFICANCE OF WETLAND
RESTORATION SCENARIOS

4.1

ABSTRACT
The Soil and Water Assessment Tool (SWAT) was used to study the impact of wetland

restoration on streamflow in the Shiawassee watershed of Michigan. Wetland restoration
scenarios were developed based on combinations of wetland area (50, 100, 250, and 500 ha) and
wetland depth (15, 30, 61, and 91 cm). Increasing wetland area, rather than depth, had a greater
impact on annual streamflow. Meanwhile, wetland depth had a limited impact on streamflow.
Wetland implementation resulted in negligible reductions in daily peak flow rates and frequency
at the watershed outlet. In developing high priority areas for wetland restoration, similar
locations were identified for reduction of subbasin and watershed outlet streamflow. However,
the best combinations of area/depth were quite different depending on the goal of the restoration
plan.
4.2

INTRODUCTION
Wetlands play a diverse, unique, and important role in the health and conservation of

vital ecosystems. Wetland systems directly support millions of people throughout the world by
providing such benefits as fertile soils for agricultural production (food and fiber), wildlife
habitat, clean water, trees for timber and fuel, and recreation areas. In addition, wetlands provide
important hydrologic, geochemical, and biological functions in a watershed (De Laney, 1995;
Hart, 1995; NRC, 1995). However, wetlands are an extremely vulnerable environmental system

53

and have significantly vanished in the past century (Nejadhashemi et al., 2012). According to
Dahl (2000), approximately 2,606 km2 of wetlands were lost in the United States between 1986
and 1997 with an estimated loss distribution of: urban development (30%), agriculture (26%),
silviculture (23%) and rural development (21%). Furthermore, in recent years, wetlands in the
U.S. are disappearing at a rapid rate of 243 km2 per year (Dahl, 1990; Dahl 2000). Some
examples of the possible major causes of wetland losses and degradation in the United States are:
artificial drainage, deposition of fill material, diking and damming, conversion to crop
production, construction, induced erosion, changing nutrient levels, increases in urbanization,
and natural causes such as erosion, droughts, hurricanes and climate change (Carter, 1961;
Leopold, 1968; U.S. EPA, 1993, Wray et al., 1995; Burkett and Kusler, 2000; U.S. EPA, 2009).
To protect wetlands, various regulations have been developed. One example is the Clean
Water Act (CWA), administered by the United States Environmental Protection Agency (EPA).
The CWA Section 404 established a program to regulate the discharge of dredged and fill
material into waters of the United States, including wetlands (Copeland, 2006). In addition,
many efforts have been developed to conserve, preserve, and restore wetlands. These efforts
include the development and use of tools to identify wetland restoration and conservation areas,
demonstrate wetland services, and perform wetlands classifications. Although some studies have
attempted to describe wetland functions using watershed models (Konyha et al., 1995; Reinelt
and Horner, 1995; Hawk et al., 1999; Arnold et al., 2001; Kirk et al., 2004; Zhang and Mitsch,
2005, Liu et., al 2008, Wang et al., 2008, Melles et al., 2010, Yang et al., 2010), there are
limitations primarily in over-simplification of wetland processes and understanding flood
mitigation benefits based on wetland placement in a watershed (Drexier et al., 1999; Raisin et al.,

54

1999, Wang et al., 2008, Yang et al., 2008). Efforts to simulate wetlands at the watershed scale
are discussed below.
Conan et al. (2003) found that the Soil and Water Assessment Tool (SWAT) adequately
simulated land use change from wetlands to dry land in Spain (Upper Guadiana river basin). The
model represented the impact of groundwater withdrawals throughout the basin and showed
misrepresentation of certain conditions that could be related to lack of sufficient data (e.g.
rainfall data). Wu and Johnson (2008) performed a hydrologic comparison between a forested
and a wetland/lake dominated watershed in northern Michigan using SWAT. The specific
objective was to compare the effects of wetland and lake abundance on the magnitude and timing
of streamflow from two watersheds (east and middle branches of the Ontonagon River basin).
The study showed that the watershed containing greater wetland and lake areas had lower spring
peaks and higher sustained flows during summer and fall. Wang et al. (2010) simulated the
effects of wetland conservation and restoration on water quality and quantity for a 4506 km2
watershed in Minnesota. In this study, the concept of hydrologic equivalent wetlands (HEWs)
was utilized. A HEW was defined in terms of six calibrated parameters: fraction of the subbasin
area that drains into wetlands, volume of water stored in the wetlands when filled to their normal
water level, volume of water stored in the wetlands when filled to their maximum water level,
longest tributary channel length in the subbasin, Manning's n value for the tributary channels,
and Manning's n value for the main channel (Wang et al., 2008). This study showed that the
HEW concept allows non-linear functional relations between watershed processes and wetland
characteristics (e.g. morphology). A reduction of approximately 10 to 20 percent of the wetlands
in the study area resulted in a considerable increase in peak discharge and loadings of sediment,
total phosphorus, and total nitrogen. They concluded that wetland conservation is a higher

55

priority than wetland restoration (Wang et al., 2010). Yang et al. (2008) studied water quantity
and quality benefits from wetland conservation and restoration scenarios using SWAT in the
Broughton’s Creek watershed (251 km2). Multiple wetland restoration scenarios were examined,
including: 0%, 10%, 25%, 50%, 75%, 90% and 100%. The optimal scenario determined for peak
flow reduction in this study was 90% restoration. However, when compared with cost
effectiveness, scenarios ranging from 50% to 80% were the most cost effective in terms of the
benefit to the wetland acreage ratios. Hattermann et al. (2008) compared two approaches that
allow integration of important wetland processes using the Soil and Water Integrated Model
(SWIM). They compared a simple supply/demand approach versus an advanced hydrotopes
approach and concluded that using the advanced approach significantly improved seasonal river
discharge and nutrient load predictions in catchments with wetlands.
Placement of a wetland for streamflow reduction is an important consideration in the
planning process. Understanding the relationship between stream order and wetland area and
depth allow for targeting stretches of river in a watershed in which restoration will be most
beneficial when project goals involve streamflow reduction. As described above, a number of
studies have explored watershed-scale wetland modeling. However, none of these studies
systematically examined the impact of wetland area, depth, and placement on streamflow and
peak flow reduction in a watershed. This study is also unique in terms of the number of scenarios
and the length of study performed to assess the hydrological function of wetlands. The
Shiawassee watershed was selected for planning of wetland conservation activities because
historically, the majority of the watershed was covered by wetlands (57%). However, vast land
use change has reduced the wetland area to 11% of the watershed. Therefore, this watershed was
considered to be a good candidate for development of wetland conservation and restoration

56

strategies. The hypothesis is that by introducing wetlands onto the landscape, we can
significantly reduce peak flow rate, which ultimately decreases environmental and economic
losses due to flooding.
We utilize SWAT to evaluate the impacts of wetland depth (15, 30, 61, and 91 cm from
normal water surface level to wetland bottom, the average depth of standing water), wetland area
(50, 100, 250, 500 ha), and wetland placement in the watershed on streamflow and peak flow
reduction at the watershed scale. The findings of this study will provide scientific understanding
of wetland functions in controlling and altering the hydrologic cycle of a watershed.
4.3
4.3.1

MATERIALS AND METHODS
Study Area
The Shiawassee watershed (hydrologic unit code 04080203) is located southwest of

Saginaw Bay in the central portion of Michigan’s Lower Peninsula and is part of the Saginaw
watershed (Figure 4-1). It drains approximately 3,000 square kilometers through the Shiawassee
River to the Saginaw River, which ultimately drains to the Saginaw Bay of Lake Huron.

The land use in the Shiawassee watershed during pre-settlement was composed by
approximately 57% woody wetlands and approximately 38% of deciduous/mixed forest
(Apfelbaum et al., 2007). Currently land use in the watershed is 57% agricultural (primarily corn,
soybean, wheat, and pasture), 14% deciduous/mixed forest, 11% woody wetlands, 7% grassland,
and 5% urban (Figure 4-2 and Figure 4-3). The main significant land use change in the
watershed was the conversion from marshes, forested bog wetlands and mixed/deciduous forests
into agricultural land by logging, filling and draining (tiling) wetland areas.

57

Figure 4-1. Study Area.

58

Figure 4-2. Shiawassee watershed land use

59

Figure 4-3. Shiawassee watershed topography.

60

4.3.2

SWAT Model Description
The SWAT model is a watershed scale model developed by the US Department of

Agriculture (USDA)-Agricultural Research Service (Arnold et al., 1998). In this study
ArcSWAT2009.93.7a was used. SWAT has proven to be a robust model capable of predicting
impacts of land use change and management practices on water, sediment, and agricultural
chemical yields in large un-gauged watersheds over long periods of time (Gassman et al., 2007).
The model features include watershed hydrology, sediment and water quality modeling, pesticide
fate and transport simulation, channel erosion simulation, and rural and agricultural management
practices (e.g. agricultural land planting, tillage, irrigation, fertilization, among other). SWAT
subdivides a watershed into a number of subbasins based on topography and river network.
Portions of a subbasin that possess unique land use/slope/soil attributes are grouped together and
defined as one hydrologic response unit (HRU) (Neitsch et al., 2005). In this study, the
Shiawassee watershed was delineated into 110 subbasins in SWAT. The average subbasin area is
approximately 2,000 ha, while maximum and minimum subbasin areas are 6,800 ha and 75 ha,
respectively.
4.3.3

SWAT Model Wetland Processes
SWAT has multiple wetland algorithms that simulate water quality and quantity within a

watershed. Below is a description provided by Arnold et al. (2001) and Neitsch et al. (2005)
regarding representation of wetland processes in SWAT.

SWAT models four types of water bodies including reservoirs, ponds, wetlands, and
depressions. Reservoirs are placed on the main channel, although the model does not recognize

61

the difference between naturally occurring and man-made impoundments. Ponds and wetlands
are modeled similarly, but are different in outflow calculation. Ponds model outflow as a
function of flood season and soil water content, while in wetlands outflow occurs when water
volume exceeds normal storage capacity. In contrast to reservoirs, ponds and wetlands are
located off the main channel, receiving flow from a portion of the subbasin in which it is located.
Finally, depressions are simulated at the HRU level, and runoff generated in these areas does not
contribute to the main channel. Water flowing into wetlands must originate from the subbasin in
which the water body is located. SWAT wetland simulations are based on one wetland per
subbasin, which will have a predefined catchment area to capture streamflow discharge within
the subbasin (Neitsch et al., 2005). SWAT employs the water balance for wetlands, presented in
Equation 1.

V =V
+V
−V
+ V pcp −Vevap −Vseep
stored
flowin
flowout

(1)

3
where, V is the volume of water in the impoundment at the end of each day (m ), Vstored is the
3

volume of water stored in the wetland at the beginning of each day (m ), Vflowin is the volume
3

of water entering the wetland during each day (m ), Vflowout is the volume of water flowing out
3

of the wetland during each day (m ), Vpcp is the volume of precipitation falling on the wetland
3

area during each day (m ), Vevap is the volume of water removed from the wetland by
3

evaporation during each day (m ), and Vseep is the volume of water lost from the wetland by
3

seepage (m ).

62

Wetland surface area is used to calculate the amount of precipitation falling on the
wetland as well as the amount of evaporation and seepage. SWAT updates surface area on a
daily basis using Equation 2:

SA = β sa ⋅V exp( sa )

(1)

where, SA is the wetland surface area (ha), β sa is a coefficient, V is the volume of water in the
3

wetland (m ), and exp(sa) is an exponent. The coefficient, β sa , and the exponent exp(sa) are
calculated by solving Equation 3 and Equation 4 using surface area and volume known values.

log (SAmx ) − log (SAnor )
10
exp(sa) = 10
log (Vmx ) − log (Vnor )
10
10

β sa

 SA
=  mx
 Vmx

(3)

exp(sa)






(4)

where, SAmx is the surface area of the wetland when filled to the maximum water level (ha),
SAnor is the surface area of the wetland when filled to the normal water level (ha), Vmx is the
3

volume of water held in the wetland when filled to the maximum water level (m ), and Vnor is
3

the volume of water held in the wetland when filled to the normal water level (m ).
The volume of precipitation falling on the wetland during a given day is calculated in
SWAT using Equation 5:

V pcp = 10 ⋅ R
⋅ SA
day

(5)

63

where, Vpcp is the volume of water added to the wetland by precipitation during the day(m3),
Rday is the amount of precipitation falling on a given day (mm), and SA is the surface area of the
wetland (ha).
The volume of water entering the wetland on a given day is calculated using Equation 6:

V





= frimp ⋅10 ⋅  Q
+ Qgw + Q  ⋅ ( Area − SA)
flowin
surf
lat 


(6)

where frimp is the fraction of the subbasin area draining into the wetland. In this study, the total
wetland area within each subbasin was calculated using the CDL (2009) and subtracted from the
total subbasin area. Then, it was assumed that half of the remaining area drained into the
wetland, similar to the technique used by Wang et al. (2010) for SWAT model calibration. Qsurf
is the surface runoff from the subbas in on a given day (mm), Qgw is the groundwater flow
generated in a subbasin on a given day (mm), and Area is the subbasin area (ha). The volume of
water entering the wetland is subtracted from the surface runoff, lateral flow and groundwater
loadings to the main channel.
The volume of water lost to evaporation on a given day is calculated using Equation 7:

Vevap = 10 ⋅η ⋅ E ⋅ SA
0

(7)

where, η is an evaporation coefficient (0.6) and Eo is the potential evapotranspiration for a given
day (mm).
The volume of water lost by seepage through the bottom of the wetland on a given day is
calculated as shown in Equation 8:

64

Vseep = 240 ⋅ K sat ⋅ SA

(8)

where, Ksat is the effective saturated hydraulic conductivity of the wetland bottom (mm).
Wetlands in SWAT will release water whenever water volume exceeds the normal
storage volume, Vnor. Wetlands have the ability to retain surface floodwaters, releasing the
excess water slowly to downstream areas (Jiang et al., 2007). Wetland soil provides a
considerable amount of floodwater mitigation, holding 3 to 9 times the weight of the soil per unit
volume (Jiang et al. 2007).
Equations 9, 10, and 11 demonstrate this behavior:

=0

V

if V

flowout

V

V

V −Vnor
flowout
10

flowout

4.3.4

=

if

= V − Vmx

if

<Vnor

Vnor ≤V ≤Vmx

V > Vmx

(9)

(10)

(11)

SWAT Model Setup
The SWAT model data input for this project included landuse, topography, soils, wetland

field data, and climate. The 2009 Cropland Data Layer (CDL, 56 m resolution) was acquired
from the USDA - National Agriculture Statistics Service. Elevation and soil datasets were
obtained from the National Elevation Dataset (NED, 30 m resolution) and the USDA State Soil
Geographic dataset (STATSGO), respectively. The STATSGO dataset includes soil physical and

65

chemical properties that are commonly used for hydrologic modeling and (STATSGO2, 2006).
Information obtained from this dataset also can help in wetland classification and placement
within the watershed. Daily climatic data (precipitation and temperature) was obtained from the
National Climatic Data Center (NCDC) for eight precipitation and six temperature stations from
1988 to 2009. A model SWAT data frame for wetland modeling processes is shown on Figure
8-1.
4.3.5

Wetlands Digital Data Assessment
Wetland data was obtained from an extended data analysis of the National Hydrography

Database (NHD), Cropland Data Layer (2009 CDL, 56 meters resolution), and the National
Wetland Inventory (NWI). The NHD is a digital vector dataset used by geographic information
systems (GIS) to map and analyze surface-water systems. The NWI is a United States Fish and
Wildlife Service program that provides an inventory of wetlands primary for scientific purposes.
The combination of these three data sources allowed for development of more accurate wetland
locations and descriptions.
4.3.6

Wetland Field Data Assessment
Multiple wetland locations were assessed and survey data was incorporated into the

model’s wetland parameters for a more accurate representation of wetland conditions in the
study area. Fifteen percent of the watershed area (28 sampling locations) was surveyed to obtain
wetland average area, depth and approximate maximum water storage volume. The locations of
the surveyed locations are presented in Figure 8-27. Standard survey techniques were used to
obtain wetland profiles within the landscape (Figure 8-28, Figure 8-29, and Figure 8-30). During
field data collection, the following wetlands were identified and evaluated: farm wetland,

66

forested wetland, and restored wetland. Multiple parameters were considered while completing
the field data assessment, including wetland hydric vegetation, hydric soils and hydrology.
For the surveyed wetlands, the average depth was measured. This information was used
with the NWI surface area data to calculate wetland volume. From this, a relationship between
wetland volume and wetland area was developed (Figure 4-4) to calculate the volume and depth
of the remaining wetlands that were not surveyed. This approach was initially described and
employed by Yang et al., (2008).

67

Figure 4-4.. Relationship between wetland storage areas and volumes in the Shiawassee watershed.

68

4.3.7

SWAT Calibration and Validation
Model calibration and validation provides credibility to model simulated outputs and is a

fundamental component of hydrological modeling. Calibration is important in hydrological
studies to reduce model simulation uncertainty (Engel et al., 2007). During the calibration
process, sensitive parameters that could potentially have an effect on wetland hydrologic
processes (e.g. infiltration, evaporation and routing) were manipulated within acceptable ranges
to improve model predictions (Table 4-1). Manual calibration on a daily basis was performed for
streamflow followed by multiple statistical analyses for the validation period. This allowed both
a qualitative and quantitative verification of the SWAT model’s accuracy on flow predictions.
Model outputs were compared to observed streamflow data.
Three statistical indices were used to evaluate model performance: Nash–Sutcliffe
coefficient of efficiency (NSE), root mean square error (RMSE)-observations standard deviation
ratio (RSR), and percent bias (PBIAS). NSE ranges from - ∞ to 1 (where 1 is optimal) and is a
normalized statistic that determines the relative magnitude of the residual variance compared to
the measured data variance. NSE is also used to quantitatively describe the accuracy of model
outputs, and is considered satisfactory at values at greater than 0.5 on a monthly basis (Moriasi et
al., 2007). RSR standardizes RMSE using the observations standard deviation, and ranges from
zero to large positive values; with zero being a perfect model simulation (indicating zero residual
variation). Monthly RSR values are considered satisfactory at less than or equal to 0.7 (Moriasi
et al., 2007). PBIAS measures the tendency of simulated data to over- or under-predict the
observed data. The optimal PBIAS is 0, with low positive/negative values indicating accurate
model simulation, and values less than ±25% being acceptable (Moriasi et al., 2007). Ideal model
calibration should consist of three to five years of data that includes average, wet, and dry years
69

to emulate significant hydrological events that will trigger model constituent processes during
calibration (Moriasi et al., 2007). The SWAT model streamflow calibration was performed on a
daily basis from 2002-2005 and validated from 2006-2009 at USGS gauging stations 04144500
(Shiawassee River near Owosso, MI (drainage area of 1246 km2) and 04145000 Shiawassee
River near Fergus, MI (drainage area of 1440 km2).

70

Table 4-1. Model Calibration parameters
USGS
Station

Description

Method

Calibrated
Value

Baseflow alpha factor

Replace

0.85

CH_K(2)

Effective hydraulic conductivity
in tributary channel alluvium

Replace

150

CH_N(2)

Manning's n value for the main
channel

Replace

0.02

ESCO

Soil evaporation compensation
factor

Replace

0.7

GW_DELAY

Ground water delay time

Replace

12

RCHRG_DP

Deep aquifer percolation fraction

Replace

0

SOL_AWC(1) Available water capacity of the
first soil layer

Multiply

0.95

SURLAG

Surface runoff lag time

Replace

1

WET_K

Hydraulic conductivity through
bottom of wetland

Replace

0.01

Baseflow alpha factor

Replace

0.8

Effective hydraulic conductivity
in tributary channel alluvium

Replace

50

Parameter

04144500 ALPHA_BF

04145000 ALPHA_BF
CH_K(2)

71

Table 4-1 (cont’d)

CN2

SCS curve number for soil
moisture condition II

Multiply

0.9

ESCO

Soil evaporation compensation
factor

Replace

0.7

GW_DELAY

Ground water delay time

Replace

5

RCHRG_DP

Deep aquifer percolation fraction

Replace

0

SURLAG

Surface runoff lag time

Replace

1

WET_K

Hydraulic conductivity through
bottom of wetland

Replace

0.01

72

4.3.8

SWAT Wetland Restoration Scenarios for the Shiawassee Watershed
The capacity for wetlands to decrease flooding is determined by several factors such as

water level fluctuations, plant community and density, habitat elements, and ground water
hydrology, among other physical factors such as area and depth (Carter and Novitzki, 1988;
Weller, 1994). Multiple scenarios were developed to assess these physical factors during wetland
restoration processes. These scenarios were formulated after analyzing wetland physical factors
utilizing GIS datasets in addition to collected field data. This allowed for evaluating the effect of
wetland restoration area and depth. Overall, the impacts of four wetland depths and four wetland
areas on streamflow and peak flow reduction were studied.
The evaluation of wetland area as part of restoration scenarios is an important factor to
consider during the planning process to account for potential environmental and ecological
benefits, in addition to financial reasons. Wetland restoration scenarios were represented in
SWAT by two parameters: surface area of wetlands at normal water level (WET_NSA) and
volume of water stored in wetlands when filled to normal water level (WET_NVOL).
Restoration scenarios include sixteen combinations of four areas and four depths. Restoration
area scenarios were 50, 100, 250, and 500 ha, while depth scenarios were 15 cm (6 in), 30 cm
(12 in), 61 cm (24 in), and 91 cm (36 in). The simulated depths were selected to reflect field
assessments. However, wetland depth is not a SWAT parameter, so it was adjusted using the
relationship between area (WET_NSA) and volume (WET_NVOL) (Figure 4-4).
Each wetland area and depth combination was implemented one-at-a-time for each
subbasin in SWAT and the reduction in streamflow and peak flow at the watershed outlet was
determined. In the case that a wetland already existed in a subbasin, the restoration scenario area
and volume were added to the existing wetland area and volume. However, when subbasin area
73

was less than 500 ha, no wetland restoration scenarios were implemented. With four wetland
areas, four wetland depths, and 89 subbasins with area greater than 500 ha in the Shiawassee
watershed, this resulted in 1424 model simulations. Each SWAT model simulation was run from
1990 to 2009, with a two year warm-up period (1988-1989).
A statistical analysis was performed using the SAS 9.2 statistical package (refer to Figure
8-2 for the utilized statistical data frame). Wetland area and depth were compared using Fisher’s
Least Significant Difference (LSD) test (α=0.05), which allowed for examination of statistically
significant differences regarding the impact of wetland area and depth on streamflow reduction
at the watershed outlet.

4.3.8.1 Impact of Wetland Placement by Stream Order on Streamflow
Stream orders within the watershed were classified using the ArcGIS Spatial Analyst
stream order classification tool. The stream order classification is a numerical denomination that
classifies stream segments of a drainage network. This classification used topographical map
features to classify each stream order. The stream order distribution for the Shiawassee
watershed is: 55% of the streams are order one, 20% of streams are order two, 10% of streams
are order three, and 15% of streams are order four. Stream order was used as a reference to
evaluate the impact of wetland location on streamflow reductions. Stream order was considered
for this study because it provides a visual idea of the area and strength of streams in the
watershed (headwater versus main channel).

74

4.3.8.2 Impact of Wetland Placement on Peak Flow
Streamflow data at the watershed outlet was analyzed to observe the impact of wetland
area and depth on peak flow events during the modeling period. Daily peak flow events were
identified as a day where streamflow is greater than the previous and following days. Average
annual maximum peak flow was defined as the 20-year average of the maximum peak flow event
occurring in each year of the study period.

4.4
4.4.1

RESULTS AND DISCUSSION
Model Calibration and Validation
Statistical indices for the SWAT model calibration (2002-2005) and validation (2006-

2009) are presented in Table 4-2. According to the model evaluation criteria described by
Moriasi et al. (2007) was satisfactory in simulating streamflow. Time-series plots of observed
versus simulated streamflow are presented for USGS station 04144500 and 04145000 in Figure
4-5 and Figure 4-6, respectively.
Table 4-2. Model Performance summary
USGS Station 04144500

USGS Station 04145000

Statistical
Measures

Calibration

Validation

Calibration

Validation

NSE

0.69

0.67

0.71

0.58

PBIAS

7.18%

-8.40%

17.77%

0.69%

RSR

0.56

0.57

0.54

0.65

75

Figure 4-5.. Observed versus simulated daily discharge at USGS station 04144500

76

Figure 4-6.. Observed versus simulated daily discharge at USGS station 04145000

77

4.4.2

Wetland Restoration Scenarios

4.4.2.1 Impact of Wetland Area on Streamflow
Streamflow changes from the base scenario at the outlet of the Shiawassee watershed
were estimated based on different wetland areas (50 ha, 100 ha, 250 ha and 500 ha) with constant
depth. For example, a comparison of streamflow change based on wetland area with a constant
depth of 91 cm is presented in Figure 4-7 (Figure 8-3, Figure 8-4, and Figure 8-5 in the
appendix). Overall, average annual streamflow changes ranged from 0 to 0.35 percent. As
wetland area in a subbasin increased, the streamflow change at the watershed outlet increased.
For instance, subbasins in the 50 ha scenario had negligible impact on streamflow, while 500 ha
wetland implementation resulted in the greatest changes in streamflow for all depths, generally in
the form of streamflow reductions. In addition, as wetland area increases the number of
subbasins contributing to streamflow reduction at the watershed outlet increases. These results
are similar to those found by Yang et al. (2008), which demonstrated that larger flow reductions
require more wetlands to be restored.
Comparing average daily streamflow across the 20-year study period resulted in
negligible changes at the watershed outlet. However, the results of the Fisher’s LSD indicate
statistically significant, but not necessarily relevant (from a stakeholder standpoint), differences
between the wetland area scenarios (Table 4-3).

78

Table 4-3. Statistical comparison of streamflow means for wetland area
Wetland Area (ha)
Streamflow at watershed outlet (m3/s)
50

20.320 a

100

20.318 b

250

20.309 c

500

20.306 d

*Streamflow means followed by the same letters are not significantly different (α=0.05).

Streamflow changes from the base scenario at individual subbasin outlets on which
wetlands were implemented were estimated based on different wetland areas (50 ha, 100 ha, 250
ha and 500 ha) with constant depth. With a constant wetland depth, area is varied to evaluate its
impact on streamflow (e.g. Figure 4-8 represents 91 cm depth across all wetland area scenarios;
additional areas shown in Figure 8-6, Figure 8-7 and Figure 8-8 in the appendix). Streamflow
changes at the subbasin level range from 0 to 25 percent depending on the area of wetland
restoration. Based on visual observation, regardless of depth, the 250 ha and 500 ha wetland area
restoration scenarios result in a majority subbasins with greater than 5% streamflow change at
the subbasin outlet. Finally, restoring 50 ha of wetland has negligible impact on subbasin level
streamflow.

79

Figure 4-7. Percent flow change at the watershed outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 91 cm

80

Figure 4-8. Percent flow change at the subbasin outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 91 cm

81

4.4.2.2 Impact of Wetland Depth on Streamflow
The impact of wetland depth in each subbasin on streamflow change at the watershed
outlet was evaluated using a constant wetland restoration area for each wetland depth. For
example, all depths at a constant area of 500 ha are presented in Figure 4-9 (Figure 8-9, Figure
8-10, and Figure 8-11 in the appendix). Overall, streamflow at the watershed outlet is minimally
impacted among all depths when area is constant. Wetland depth scenarios showed streamflow
reduction for wetland areas greater than 100 ha, while no impacts on streamflow reduction were
observed in any of the 50 ha scenarios when examining the depth effect. Under the 500 ha
wetland restoration scenario, most subbasins influence streamflow at the watershed outlet
regardless of depth. Therefore, in developing wetland restoration scenarios, wetland area is more
important than depth in affecting streamflow at the watershed outlet.
The statistical analysis comparing treatment means (wetland depth) across all studied
wetland areas using Fisher’s LSD (α=0.05) is presented in Table 4-4. All wetland depths were
found to be statistically similar. This supports the conclusion that selection of wetland depth does
not have as much impact on streamflow at the watershed outlet as selecting a suitable wetland
area.
At the subbasin level, the trend among wetland depths in streamflow change was similar
to the outlet, although the percent changes were greater. For example, as demonstrated in Figure
4-10 (Figure 8-12, Figure 8-13, and Figure 8-14in the appendix), all depths have a similar impact
on subbasin level streamflow. This was the case for depth across all wetland areas.

82

Table 4-4. Statistical comparison of streamflow means for wetland depth

Wetland Depth (cm)

Streamflow at watershed outlet (m3/s)

15

20.314 a

30

20.314 a

61

20.313 a

91

20.313 a

*Streamflow means followed by the same letters are not significantly different (α=0.05).

83

Figure 4-9. Percent flow change at the watershed outlet compared to base scenario –
500 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

84

Figure 4-10. Percent flow change at the subbasin outlet compared to base scenario –
500 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

85

4.4.2.3 Impact of Wetland Placement by Stream Order on Streamflow
The wetland performance in percent streamflow change at the watershed outlet was
classified based on the stream order in which the wetland restoration scenario was implemented.
For example, Figure 4-11presents percent flow change at the watershed outlet with a constant
depth of 91 cm (Figure 8-15, Figure 8-16, and Figure 8-17 in the appendix), while Figure 4-12
presents percent flow change at the watershed outlet with a constant area of 500 ha (Figure 8-18,
Figure 8-19, and Figure 8-20 in the appendix). Average streamflow changes were insensitive to
stream order at restoration areas up to 250 ha, while at 500 ha wetland placement on stream order
one and four has a greater impact on streamflow (Figure 4-11). Under constant area a similar
trend was observed in which wetland placement in stream orders one and four resulted in greater
streamflow changes at all depths (Figure 4-12). Therefore, it can be concluded that wetland
restoration on first and fourth order streams may result in higher streamflow reduction when
wetland size is greater than 250 ha.
In terms of the impact of wetland restoration scenarios on streamflow change at the
subbasin level, wetlands placed on subbasins with stream order one are more effective than other
orders. As stream order increases moving from one to four, effectiveness of wetland restoration
scenario decreases among all depths and areas as shown in Figure 4-13 and Figure 4-14 (Figure
8-21 through Figure 8-26 in the appendix). This can be explained by assuming that the flow rate
of first order streams are generally less than those of higher stream order. Therefore, the same
wetland size is more effective in reducing flow in subbasins with first order streams than those
with orders two, three, or four. This indicates that specific subbasin location in which the
wetland was implemented plays an important role in local streamflow reduction.

86

Figure 4-11. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 91 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

87

Figure 4-12. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 500 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

88

Figure 4-13. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 91 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

89

Figure 4-14. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 500 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

90

4.4.2.4 Peak Flow Reduction
Peak flow is defined peak as the maximum water level reached during a runoff event.
This is identified as a day in which the streamflow rate is greater than both the preceding and
following days. Two criteria for examining peak flow were used (daily and average annual
maximum peak flow). Peak flow reduction was compared between the case in which no wetlands
are restored (base scenario) and the scenarios containing wetlands (all combinations of
area/depth one-at-a-time in each subbasin). The base scenario produced 553 daily peak flow
events with a 20-year average peak flow rate of 40 m3s-1 at the watershed outlet. Negligible
changes in peak flow rate were observed, while the number of peak flow events was reduced by
one to eight events, depending on the wetland restoration scenario and location. There was a
greater reduction in peak flow events with increasing wetland depth and area.
The base scenario average annual maximum peak flow was 157 m3s-1. Overall,
implementation of wetland scenarios resulted in reductions of 0.2 m3s-1 to 1.9 m3s-1, depending
on the combination of depth, area, and location (Table 4-5). Similar to the previous observation
concerning wetland area and depth, increasing area results in a greater decrease in annual
maximum peak flow. However, increasing depth does exhibit decreases in annual maximum
peak flow, although they are not as large as increasing area.

91

Table 4-5. Wetland area/depth combinations for reduction of average annual maximum peak flow (m3/s) events

Area/depth

15 cm

30 cm

61 cm

91 cm

50 ha

0.21

0.25

0.32

0.43

100 ha

0.34

0.41

0.55

0.67

250 ha

0.87

0.85

1.12

1.33

500 ha

1.17

1.35

1.71

1.85

92

4.4.3

High Priority Areas for Wetland Restoration
High priority areas for wetland restoration were developed with two implementation

goals: flow reduction at the watershed outlet or flow reduction at a specific location within the
watershed. These areas were identified by selecting the combination of wetland area and depth
for each subbasin that resulted in the greatest streamflow reduction at the watershed outlet
(Figure 4-15a) and individual subbasin level (Figure 4-15b). This ultimately allows for strategic
identification of locations for wetland restoration and the specific area and depth of wetland that
should be implemented. High, medium, and low impact areas for wetland restoration were
identified using the Jenks natural breaks classification method, which arranges values into
classes based on natural groupings inherent in the data, while maximizing differences between
classes (Jenks, 1967). In this specific watershed, high priority locations for wetland restoration
were similar when targeting streamflow reduction at the watershed outlet and individual subbasin
outlets. However, this may not be the case when applied to different geographic regions.
The number of subbasins selected for each wetland area/depth combination resulting in
the greatest streamflow reduction is presented for the watershed outlet (Table 6) and individual
subbasin outlets (Table 7). Wetlands of 500 ha were selected most often for both the watershed
outlet and subbasin outlets (45 and 55 subbasins, respectively). Meanwhile, the depths selected
most frequently were 15 cm for the watershed outlet and 91 cm for subbasin outlets. The most
frequently selected combination was a 500 ha wetland of 15 cm depth for flow reduction at the
watershed outlet, while for subbasin outlets 500 ha with 15 cm or 91 cm were selected most
often. Despite the identification of similar high priority locations for the watershed and subbasin
outlets, the depth and area of wetland implementation can be different to achieve maximum flow
reduction. Finally, there is not a definitive optimal wetland area and depth combination, but it is
93

apparent that wetlands of larger area and smaller depth were selected most frequently as being
effective in streamflow reduction.

Figure 4-15. High, medium, and low impact areas for wetland restoration scenarios

94

Table 4-6. Wetland area/depth combinations selected as optimal for streamflow reduction at the watershed outlet
Area/depth
15 cm
30 cm
61 cm
91 cm
Total
50 ha

14

0

0

0

14

100 ha

0

0

2

0

2

250 ha

17

2

7

2

28

500 ha

33

3

5

4

45

Total

64

5

14

6

89

Table 4-7. Wetland area/depth combinations selected as optimal for streamflow reduction at the individual subbasin
outlet
Area/depth

15 cm

30 cm

61 cm

91 cm

Total

50 ha

2

0

1

0

3

100 ha

1

0

1

6

8

250 ha

3

1

4

15

23

500 ha

26

3

3

23

55

Total

32

4

9

44

89

95

4.5

CONCLUSION
This study examined the impact of wetland area, depth, and placement on streamflow and

peak flow reduction at the watershed scale. Using the SWAT model, the impacts of various
wetland depths, areas, and location placements were examined. The specific objectives of this
study were to evaluate the impacts of wetland areas (50, 100, 250, and 500 ha), depths (15, 30,
61, and 91 cm), and (3) wetland placement on streamflow characteristics at the watershed and
individual subbasin outlets.
Wetland area was determined to be an important factor in reducing streamflow at the
watershed and subbasin level. Greater wetland areas (250 and 500 ha) resulted in greater
streamflow reductions, likely because of increased storage volume, while the smallest area (50
ha) had negligible impacts on streamflow reduction. In addition, there were statistically
significant differences between all wetland area scenarios at the watershed outlet. However,
wetland depth was found to have minimal impact on reducing streamflow. In the statistical
analysis, it was found that all depths resulted in no significant difference in streamflow at the
watershed outlet.
Wetland placement based on stream order demonstrated that average streamflow changes
at the watershed outlet were insensitive for wetland areas less than 500 ha. However, at 500 ha
wetland restoration scenarios on first and fourth order streams had greater streamflow reductions.
At the subbasin level, wetland implementation on first order streams was most effective for all
depths and areas. Therefore, wetland restoration in headwater streams and locations near the
watershed outlet were determined to be most beneficial for streamflow reduction. However, due
to the comparatively large variation in subbasin level streamflow reduction on first order
streams, caution should be exercised when implementing wetlands in these areas.
96

Average daily peak flows and events at the watershed outlet were negligibly impacted by
wetland implementation. However, average annual maximum peak flow was slightly reduced
due to wetland restoration. In addition, both increasing area and depth were somewhat effective
in reducing average annual maximum peak flow magnitude.
Finally, high priority areas for wetland restoration were determined based selection of the
smallest area and depth that resulted in the greatest streamflow reduction at the watershed and
individual subbasin outlets. Despite the similarity between watershed and subbasin level high
priority maps, the area and depth combinations to achieve to maximum streamflow reductions
were often different. This distinction is important for the decision making process, as wetland
implementation size will change based on project goals and resources, although location may
not.
The findings of this study will provide an understanding of wetland functions in
controlling streamflow. In addition, it explores the manner in which optimal wetland area and
depth can be selected for maximum benefit at the watershed and subbasin scale. Although
optimal wetland area and depth were identified, their impacts on streamflow and peak reduction
at the watershed outlet are generally minimal. However, these results may vary regionally based
on watershed physiographical characteristics. Meanwhile, streamflow reductions are just one
reason for wetland restoration; water quality improvements are an even more widespread
motivation. Therefore, future studies should incorporate water quality parameters to compare
restoration scenarios in achieving watershed-scale management goals.

97

5

ASSESSING THE SIGNIFICANCE OF WETLAND RESTORATION
SCENARIOS ON A SEDIMENT MITIGATION PLAN

5.1

ABSTRACT
Wetlands have many environmental, social, and economic values. However, due to

accelerated land use change and lack of understanding of the functions of wetland ecosystems,
they have deteriorated, if not lost in many areas worldwide. Meanwhile, current functional
wetland assessment techniques only provide rough estimation, where in most cases are site
specific and qualitative. The overall goal of this project is to examine the sediment reduction
benefit of wetland implementation scenarios both at subbasin and watershed scales. Two set of
models were used to accomplish this goal. First, a watershed model – the Soil and Water
Assessment Tool (SWAT), was employed to estimate sediment load at the subbasin scale.
However, due to limitations of wetland functions of SWAT model, a second model – the System
for Urban Stormwater Treatment and Analysis Integration (SUSTAIN) was used. The sediment
load generated for each subbasin was incorporated in the SUSTAIN model. This allows for
evaluating sediment reduction capability of wetlands at subbasin level. Next, a portion of
sediment not treated by a wetland was fed back to the SWAT model and routed to the watershed
outlet. The impacts of four different wetland surface areas (0.40, 0.81, 2, and 4 ha) on sediment
load mitigation were examined one-at-the-time for all subbasins within the River Raisin
watershed located in southeastern Michigan and northeastern Ohio. Comparison of the sediment
reductions due to different wetland restoration scenarios reveals the importance of wetland
placement in a watershed. In general, rate of streamflow reduction resulting from wetland
implementation is higher than sediment reduction at the subbasin level but more comparable at
the watershed level. In addition, clusters of wetlands installed at the distance of 150 - 200 stream
98

km from the outlet outperformed other clustered wetlands at closer and farther distances.
Wetlands associated with 1st order streams performed better at the subbasin level, while
wetlands located at 4th order streams performed better at the watershed level. Considering
environmental and economic issues of wetland restoration scenarios revealed that the 0.4 ha
wetland was the most suitable for subbasin and watershed levels outcomes due to the
significantly lower cost of 0.4 ha wetland installation and maintenance than other scenarios.
5.2

INTRODUCTION
Wetlands perform essential hydrological, geochemical, and biological functions at the

watershed level (De Laney, 1995; Hart, 1995). They have the capacity to significantly reduce
nutrients, sediments, and other pollutant concentrations produced from runoff under different
environmental conditions at the watershed scale (Kadlec, 1993; Mander et al., 2000; Trepel and
Palmeri, 2002; Jordan et al., 2003; Arheimer et al., 2004; Skagen et al., 2008, Wang et al., 2010,
Fan et al. 2012). Processes that contribute to pollutant removal in wetlands are chemical,
physical, and biological in nature (Kadlec et al., 2008). Chemical processes include precipitation
of phosphorus by iron, aluminum or calcium, and precipitation of heavy metals (Nilsson et al.,
2011). Additionally, wetlands facilitate chemical transformation of nitrogen, which leads to the
release of nitrogen to the atmosphere (Vymazal, 2007). Physically, wetland vegetation
substantially slows runoff, leading to deposition of mineral and organic particles and adsorbed
contaminants (Carter, 1996). Wetland microbial activity is a biological process that results in
decomposition of organic matter removal of nitrogen through microbial transformation
(nitrification-denitrification) (Brix, 1993). Plant uptake of organic chemicals into plant tissue is
another biological process attributed to wetlands’ ability to treat pollutants (Ryan et al., 1988).
Furthermore, wetlands provides many ecosystem services, especially related to water quantity
99

(Luecke 1993; Comin et al., 1997; Keddy, 2000, Ramsar, 2004). Wetlands are effective in
catching, retaining, and filtering runoff water generated from heavy rainfall or snowmelt events
and promoting groundwater infiltration, which helps reduce river peak flow (Luecke 1993;
Comin et al., 1997; Keddy, 2000; Ramsar, 2004).
Wetland restoration and construction technologies for the treatment of pollutants is an
emerging field (EPA, 2000; Schröder et al., 2007). A robust understanding of pollutant removal
processes and wetland environmental characteristics is needed for conservation, restoration,
planning, and design purposes. For this reason wetland water quality improvement capabilities
have been studied for different types of wetlands in specific settings, as described above.
However, the challenge to optimize ideal restoration conditions on a larger scale persists due to
the complexity of wetlands and pollution transport processes at the watershed scale.
Researchers use numerous models to simulate pollutant transport at catchment and
watershed scales. The Soil and Water Assessment Tool (SWAT) is one of the most widely used
model in watershed and river basin simulation (Gassman, 2014). Arnold et al. (2001) used
SWAT to simulate the water budget in a constructed wetland in Texas, where the model was
modified to include the interaction between ponded water in the wetland and the soil profile and
shallow aquifer. Wang et al. (2008) developed the hydrologic equivalent wetland (HEW) method
to represent wetlands in SWAT model and applied the method to successfully simulate
streamflow in a watershed located in Minnesota. Liu et al. (2008) developed a SWAT extension
to simulate flow and sediment in a riparian wetland, but did not validate the model due to the
limitation of observed data. Wu and Johnston (2008) compared SWAT performance between
forested and a wetland/lake dominated watershed in Michigan, and reported satisfactory model
calibration but discrepancies in summer streamflow prediction. Wang et al. (2010) applied the

100

HEW method to estimate streamflow, sediment, total nitrogen (TN), and total phosphorus (TP)
loads under wetland restoration and conservation scenarios in Manitoba, Canada. However, the
authors only calibrated the model for streamflow. Therefore, evaluation of wetland performance
for sediment, TN, and TP is highly uncertain. Feng et al. (2013) incorporated a wetland module
into SWAT to simulate wetland hydrology in northeast China, where the method performed well
in reconstructing wetland hydrological processes. Martinez-Martinez et al. (2014) used SWAT to
simulate streamflow rates and peaks under wetland restoration scenarios and reported that
average streamflow fluctuation at the watershed outlet is more sensitive to wetland area than
depth. Numerous other researchers have used different modeling approaches to incorporate
wetlands in their simulations such as the Hydrological Simulation Program-Fortran (HSPF)
(Schwar et al., 1998; Zhang et al., 2009), MIKE-SHE (Thompson et al., 2004; Zacharias et al.,
2005; Dai et al., 2010), DRAINMOD (Caldwell et al., 2007; Skaggs et al., 1995; Jia and Luo,
2009), and SWMM (Obropta et al., 2008; Tsihrintzis et al., 1998; Koo et al., 2013). Overall,
most studies have only considered wetland hydrology in watershed scale modeling and have
either ignored or not calibrated the model for sediment and nutrients to simulate the impact of
wetland restoration scenarios on pollutant treatment, as such a task is still a challenge to
scientists (Wang et al., 2008).
Among physically-based watershed/water quality models, SWAT is a comprehensive
model that combines spatial and temporal analysis, is open source, and has strong model support,
making it one of the most widely used water quality models in watershed and river basin
modeling (Gassman, 2014; Srinivasan et al., 1998). However, a major drawback of using SWAT
for watershed scale wetland modeling is that SWAT assumes a completely mixed wetland
system in pollutant routing. In addition, SWAT ignores nutrient transformation in simulating

101

nutrient removal in wetlands, ponds, and reservoirs and considers settling as the sole method of
nutrient removal (Neitsch et al., 2011). In order to solve this problem, we proposed to couple
SWAT with a second model capable of addressing these issues. The System for Urban
Stormwater Treatment and Analysis INtegration (SUSTAIN) model (USEPA, 2009) allow users
to simulate wetlands as either a plug flow reactor or a continuously stirred tank reactor (CSTR)
in series with a user-defined number of CSTRs. In addition, SUSTAIN models pollutant removal
by either first-order order decay or a modified kinetic model (K-C*) (Shoemaker et al., 2009).
This research addresses the challenges of using a hybrid of two water quality models to examine
the sediment reduction benefits of wetland implementation scenarios at subbasin and watershed
scales. The specific objectives of this project are to: 1) assess the impacts of wetland restorations
scenarios on flow and sediment, 2) determine the role of wetland placement in watershed
sediment dynamics by considering the distance to the outlet and stream order concept, and 3)
evaluate the environmental and economic aspects of wetland restoration scenarios at the subbasin
and watershed scale.
5.3
5.3.1

MATERIALS AND METHODS
Study Area
The River Raisin watershed (Hydrologic Unit Code 04100002) is located primarily in

southeastern Michigan, with a small portion located in northern Ohio (Figure 5-1). The River
Raisin watershed drains approximately 2681 km2 into Lake Erie. The watershed is
predominantly agricultural, covering approximately 66% of the total watershed area (CDL,
2012). The primary crops grown in the watershed are corn, soybeans, and wheat. The remaining

102

land cover is 13% forest, 12% urban, 7% wetlands, 1% range grasses, and 1% water (CDL,
2012).
The River Raisin watershed is characterized by hilly to moderately rolling topography in
the western and northwestern regions and by relatively flat terrain in the southeast. Soils are
characterized as having slopes of 0-5 percent. (Knutilla and Allen, 1975). Sandy loams, loams,
and clay loam soils with moderate to high infiltration rates dominate the upstream northwestern
portion of the River Raisin watershed. The streams in this portion of the watershed have more
stable flows and consistent groundwater recharge. Meanwhile, the southeastern portion of the
watershed is dominated by primarily clays, clay loams, and silty clays with low to very low
permeability and slow infiltration rates (Dodge, 1998).
The River Raisin watershed was selected due to its significant variation in soil types, land
use patterns, topography and geology. Historically, this watershed was a swamp (wetland) with
flat topography and muck and clay soils (Dodge, 1998). Comparison of the current land use map
(NLCD, 2001) with prehistoric land use (MNFI, 2014) reveals a loss of 59% of the woody
wetlands and 91% of emergent herbaceous wetlands since the mid-1800s. Due to land use
changes (tiling and drainage) the watershed is now highly agricultural.

103

Figure 5-1. Study area and location of the monitoring stations

5.3.2

Models
Wetlands are complex, diverse, and dynamic ecosystems, and watershed-scale wetland

assessment is a challenge currently faced by watershed managers and conservationists. Due to
limitations of watershed models in quantifying the benefits of wetland restoration strategies,
SUSTAIN was used to evaluate the local scale benefit of wetland installation in reducing runoff
and sediment loadings, while SWAT was used to estimate sediment loads and runoff to and from
wetlands all the way to the watershed outlet. Linking SUSTAIN and SWAT to model water
quantity and quality is a unique approach to evaluate wetland behavior at multiple spatial scales.
104

5.3.2.1 SUSTAIN Model
The SUSTAIN model (version 1.2) is a decision support system developed by the Tetra
Tech for the United States Environmental Protection Agency (USEPA) to evaluate water quality
management alternatives (USEPA, 2009). SUSTAIN simulates the ability of individual or a
combination of best management practices to reduce peak streamflows and sediment and nutrient
concentrations. Data required by SUSTAIN includes: climate, soils, landuse, topography,
delineated subbasins, stream network, stream geometry, streamflow, water quality, and best
management practices.

SUSTAIN provides users with multiple BMP process simulation methods. Flow routing
processes can be simulated either by a stage-outflow storage routing using weir and/or orifice
equations or by kinematic routing through solving the Coupled Continuity equation and
Manning’s equation (Linsley et al., 1992). Infiltration processes can be simulated by either the
Holtan-Lopez equation or the Green-Ampt method (Maidment, 1993; Huber and Dickison,
1988). Evapotranspiration (ET) processes can either be simulated using a constant ET rate or
through calculating potential ET and actual ET (Bicknell et al., 1997; Maidment, 1993).
Pollutant routing processes are simulated using completely mixed, continuous stirred tank
reactors in series (CSTRs) (Wong et al., 2002), or plug flow reactors (Persson et al., 1999). The
first-order decay, Kadlec and Knight’s (K'-C*) method (Kadlec and Knight, 1996), or sediment
settling under quiescent and turbulent conditions (James et al., 2002) can be used to simulate
pollutant removal.

Wetland simulation processes were performed according to the SUSTAIN user’s manual
recommendations (Alvi et al. 2009). Wetlands were simulated as wet ponds and flow and
105

sediment output from SWAT were used as wetland inputs. Wetlands in SUSTAIN are defined by
size specification, pollutant removal method, and removal capacity. In this study all wetland
alternatives were simulated with a relatively shallow depth (0.3048 m); and therefore, it is
reasonable to assume vegetation is present. The Holtan infiltration method accounts for
vegetation effects and therefore was selected for infiltration simulation. The K-C* model,
proposed by Kadled and Knight (1996), is a widely used pollutant reduction model in wetlands
(Park and Roesner, 2009) and was used to estimate sediment removal. In this model, the total
suspended solids removal rate (K) in wetlands was considered as 5,000 m/yr and the background
concentration value (C*) was adjusted to 6 mg/L (eWater, 2012). Sediment routing was modeled
using 1.5 CSTR in series as recommended by the SUSTAIN model manual (USEPA, 2009).
Depth of the wetland was defined by the height of the weir. As suggested by USDA-NRCS
(2009), length-to-width ratio of the wetlands was maintained at 2:1. Percentage of sand, silt and
clay in the inflow were obtained from the Soil Survey Geographic (SSURGO) dataset for the
upper layer soils (USDA-NRCS, 2013b). Additional parameters and their sources are listed in
Table 9-1. A sensitivity analysis of key SUSTAIN parameters was performed and the results are
presented in Table 9-2.
5.3.2.2 SWAT Model
The SWAT model was developed by the US Department of Agriculture (USDA)Agricultural Research Service (Arnold et al., 1998) and is able to simulate watershed hydrology,
sediment and water quality, pesticide fate and transport, channel erosion, and agricultural
management practices. SWAT delineates a watershed into subbasins based on a digital elevation
model and river network. Following delineation, the subbasins are further divided into
hydrologic response units (HRUs). An HRU is an area of homogeneous land use, soil type,
106

slope, and management practices. The SWAT model has multiple algorithms that simulate water
quality and quantity within the watershed, as described in Arnold et al. (2012). Land-based
calculations in SWAT (hydrology and pollution generation) are completed at the HRU level and
aggregated to the subbasin level. Pollutants are then transported to the nearest reach and routed
through the river network.
In this study SWAT 2012 (v. 591) was used. The SWAT model data inputs include:
precipitation and temperature data obtained from the National Climate Data Center (NCDC),
soils (SSURGO), landuse (Cropland Data Layer (CDL), 30 m resolution), DEM (US Geological
Survey (USGS) National Elevation Dataset (NED), 30 m resolution), watershed and subbasins,
stream networks from the Michigan Institute of Fisheries Research based on the USGS National
Hydrography Dataset), streamflow measurements (US Geological Survey (USGS) gauging
stations), and water quality grab samples (Environmental Protection Agency STOrage and
RETrieval (STORET) data warehouse).
The River Raisin SWAT model was calibrated for streamflow at three locations (USGS
stations 04175600, 04176000, and 04176500) and sediment concentration at the STORET site
580046 (figure 1). During the calibration process model parameters were adjusted to improve
model predictions of observed streamflow and sediment concentration data (Table 5-8 and Table
5-9). The SWAT model streamflow calibration was performed on a daily basis from 1996-2000
and validated from 2001-2005. The sediment calibration and validation was performed on grab
sample data from 2000-2002 and 2003-2005, respectively. This period was selected based on
limited availability of sediment data.
Three goodness-of-fit criteria were used to evaluate SWAT model performance:
Nash-Sutcliffe coefficient of efficiency (NSE), root mean square error-observations standard

107

deviation ratio (RSR), and percent bias (PBIAS). NSE ranges from - ∞ to 1 (where 1 is optimal)
and is a normalized statistic that determines the relative magnitude of the residual variance
compared to the measured data variance (Moriasi et al., 2007). NSE is considered satisfactory
when greater than 0.50 for a monthly time step. RSR standardizes RMSE using the observations
standard deviation, and ranges from zero to large positive values, with zero representing perfect
model simulation (indicating zero residual variation). RSR values are considered satisfactory at
less than 0.70 for a monthly time step (Moriasi et al., 2007). PBIAS measures the tendency of
simulated data to over- or under-predict compared to observed data. The optimal PBIAS is 0,
with low values indicating accurate model simulation. Values between ±25% are satisfactory for
streamflow and between ±55% for sediment at a monthly time step (Moriasi et al., 2007).
5.3.3

SUSTAIN/SWAT Hyrbid Model Approach
This study integrates SWAT and SUSTAIN to evaluate the impacts of wetland

restoration on sediment loads at subbasin and watershed scales. Figure 5-2 shows the SWAT and
SUSTAIN model integration schema. First, the calibrated/validated SWAT model was used to
estimate surface runoff and sediment yield for all 1,235 subbasins in the River Raisin watershed.
Surface runoff and sediment yield from the SWAT model were input into the SUSTAIN model
wetlands.

108

Figure 5-2. SWAT – SUSTAIN models integration schema

109

Sediment removal and surface runoff attenuation were simulated using SUSTAIN.
Finally, the SUSTAIN outputs (surface runoff and sediment load) were input back into the
calibrated SWAT model as a point source for each subbasin and routed to the watershed outlet.
However, it is challenging to seamlessly integrate the results of these two models. As SWAT
runs on daily time step and SUSTAIN requires hourly input data, SWAT outputs for each
subbasin were converted to average hourly data prior to use in SUSTAIN. Finally, after
completing the simulation, SUSTAIN outputs were aggregated to daily loads and used as point
source inputs in SWAT. This process was repeated one-at-a time for all wetland sizes for each
subbasin. The SWAT and SUSTAIN models were run for 10 years from 1996 to 2005.
5.3.4

Wetland Restoration Scenarios
Wetland surface area, location, and depth are critical components of wetland restoration

planning to abate sediment loads at the watershed scale. Multiple wetland restoration scenarios
were developed to assess importance of wetland physical factors under restoration conditions.
Restoration scenarios were formulated by using a combination of wetland areas and placement
within the River Raisin watershed. This approach was based on 0.4, 0.81, 2 and 4 ha (1, 2, 5 and
10 acres) of wetland restoration with an average wetland depth of 30.48 cm (12 in). Wetland
surface area selection was based on wetland restoration and management literature resources
obtained from government entities and guidance from the USDA-NRCS National Engineering
Handbook (USDA-NRCS, 2003, USDA-NRCS, 2009). Different wetland restoration depths
were not explored based on findings from Martinez- Martinez, et al. (2014), where wetland
depths were found to be less important than surface area with respect to surface runoff reduction
at the subbasin and watershed scales. Each wetland surface area was implemented one at a time
for each subbasin in SUSTAIN and a sediment reduction capacity was generated per individual
110

scenario. Wetlands were not implemented in subbasins with a total area less than 4 ha (the
maximum wetland restoration size determined for the study), which resulted in 4480 simulations.
SUSTAIN simulation of the wetlands were performed at an hourly time step and were based on
the assumption that the restored wetlands were placed at the outlet of each subbasin.
5.3.5

Environmental and Economic Aspects of Wetland Implementation
Environmental and economic factors were considered during the selection of the most

effective wetland surface area for each subbasin based on sediment reduction at the watershed
outlet using analytical hierarchy process (AHP). AHP is an optimization algorithm that can solve
complex decision-making problems (Linkov and Steevens, 2013) by using evaluation criteria
based on expert knowledge and pairwise comparison (Young et al., 2009). The AHP algorithm
was developed by Saaty (1980) and has been widely used in diverse fields such storm water
BMP selection (Young et al., 2011), environmental vulnerability assessment of river basins
(Chang and Chao, 2011), land suitability assessment for irrigation (Chen et al., 2010), integrated
watershed management (Biswas et al., 2012), landfill site selection (Sener et al., 2010), and
solvent selection in pharmaceutical processes (Perez-Vega et., 2011). However, to the best of our
knowledge, using AHP for wetland placement in large and complex watersheds such as the River
Raisin Watershed has not been explored.
5.3.6

Environmental Aspects of Wetland Implementation
Four different wetland surface areas (0.40, 0.81, 2 and 4 ha) were implemented in each

subbasin one at a time, and the resulting sediment load was estimated at the watershed outlet.
The sediment reduction at the watershed outlet following wetland implementation in each

111

subbasin was calculated by subtracting the sediment load after implementation from the base
sediment load where wetlands were not implemented in the watershed.
5.3.7

Economic Aspects of Wetland Implementation
The implementation costs of the four wetland scenarios were obtained from the United

States Department of Agriculture, Natural Resources Conservation Service (USDA-NRCS) Field
Office Technical Guide (USDA-NRCS, 2013a). The USDA-NRCS implementation cost
includes, but is not limited to, site preparation, excavation, backfilling, grading and finishing,
vegetation planting, trees and shrubs establishment, and operations and maintenance
(Biebighauser, 2007; Biebighauser, 2011; USDA-NRCS, 2013). The cost and needs of each
implement component is site specific and the costs can be lower or higher than listed. The
SUSTAIN model includes a wetland restoration cost database; however, it was developed using
national sources and is not site specific (Alvi et al. 2009). In addition, SUSTAIN does not
account for costs associated with wetland installation and maintenance. Therefore, economic
analysis was performed independently (Table 9-6, Table 9-7, Table 9-8, and Table 9-9). The
tables include three different cost estimates (low, average, and high) based on the 2013 statewide
survey. The cost differences are due to site locations, existing site conditions, labor cost, and
other factors. The average wetland implementation costs over a 10-year lifespan of wetland
operation for 0.4, 0.81, 2 and 4 ha with a depth of 30.48 cm are $1,490, $2,874, $6,947, and
$13,656, respectively.

112

5.3.8

Wetland Selection Considering
onsidering Environmental and Economic Factors
actors
AHP was used to integrate two different factors (environmental benefits and economic

cost) with different units in the selection of most suitable wetlands for sediment reduction at the
watershed outlet. The AHP is a stru
structured
ctured technique consisting of four steps to convert the
complex decision making processes into an algorithm: 1) constructing a pairwise matrix, 2)
computing the priority vector, 3) calculating the consistency ratio, and 4) ranking the alternatives
(Saaty, 1980; Young et al., 2009; Giri and Nejadhashemi, 2014
2014).
). In order to better describe the
AHP process, subbasin 2 (Figure
Figure 5-3) was selected as an example.

Figure 5-3.. River Raisin watershed subbasin 2.

113

Step 1) Constructing the pairwise comparison matrix: A pairwise comparison matrix was
constructed for each subbasin based on environmental benefits of sediment reduction at the
watershed outlet by four wetland surface areas. Sediment reduction was calculated by subtracting
sediment loads at the outlet before and after wetland implementation. The process was repeated
for all subbasins. In the pairwise matrix all rows and all columns were compared. This allows
that relative importance of each alternative to be compared to other alternatives on a scale of 1/9
to nine. In this matrix, one means that two alternatives are equally important. Nine indicates that
one alternative is absolutely more important than the other, while the reverse is true for 1/9. The
pairwise comparison matrix for subbasin 2 is presented in Table 5-1.

Table 5-1. Pairwise comparison matrix developed for Subbasin 2 based on sediment
reduction at the watershed outlet.
Wetland Size
(ha)
0.40

0.4

0.81

2

4

1.0000

0.9735

0.9168

0.8475

0.81

1.0272

1.0000

0.9417

0.8705

2.00

1.0907

1.0619

1.0000

0.9243

4.00

1.1800

1.1488

1.0819

1.0000

Step 2) Computation of priority vector: Before calculating the priority vector, the
principal eigenvector is constructed. To construct the principal eigenvector, first each column
within the pairwise comparison matrix developed in the step 1 was summed. Then each element
in the matrix is divided by the sum of the value for the corresponding column by dividing each
column in the pairwise comparison matrix by sum of its values. Therefore, the total sum of each

114

column should be one (Table 5-2). The average of each row/column in the matrix was calculated,
giving the priority vectors (Table 5-2). The priority vectors were calculated as 0.2327, 0.2390,
0.2538, and 0.2746 for 0.4 ha, 0.81 ha, 2 ha and 4 ha wetlands, respectively. The remaining
priority vectors were calculated using the AHP extension developed for ArcGIS by Marinoni
(2009).

Table 5-2. Priority vector calculation based on sediment reduction at watershed
outlet for the Subbasin 2.
Wetland Size
(ha)
0.40
0.81
2.00
4.00
Total

0.4

0.81

2

0.2327
0.2390
0.2538
0.2746
1.000

0.2327
0.2390
0.2538
0.2746
1.000

0.2327
0.2390
0.2538
0.2746
1.000

4
0.2327
0.2390
0.2538
0.2746
1.000

Row Average
(Priority Vector)
0.2327
0.2390
0.2538
0.2746
1.000

Step 3) Computation of consistency ratio: The consistency ratio is a measure of the
consistency of importance between different elements of a priority matrix. This step verifies
whether the priory vectors obtained in the previous step are acceptable. If the consistency ratio is
less than 0.1, then the pairwise comparison matrix developed in step 1 is consistent and the
priority vectors are acceptable (Saaty, 1980). If it is greater than 0.1, the pairwise comparison
matrix should be adjusted and checked for consistency and logical relationships. To calculate the
consistency ratio, the original pairwise matrix constructed in step 1 was multiplied with the
priority vector, resulting in a new matrix (Matrix A, shown below). This new matrix was divided
by the priority vectors to create Matrix B, shown below.

115

1.0000
1.0272
൮
1.0907
1.1800

0.2327
0.9307
0.9735 0.9168 0.8475
0.2390
0.9560
1.0000 0.9417 0.8705
൲ൈ൮
൲ൌ൮
൲
0.2538
1.0151
1.0619 1.0000 0.9243
0.2746
1.0982
1.1488 1.0819 1.0000
Priority Vector

0.9307
0.2327
4.0000
0.9560
0.2390
4.0000
൮
൲ൊ൮
൲ൌ൮
൲
1.0151
0.2538
4.0000
1.0982
0.2746
4.0000

Pairwise Comparison Matrix

Matrix A

Matrix A

Priority Matrix Matrix B

The maximum eigenvalue (λmax) was calculated by averaging all elements in Matrix B.
Then, the consistency index (CI) was calculated based on equation (1) (Saaty, 1980):
‫ ܫܥ‬ൌ

ሺఒ೘ೌೣ ି௡ሻ
ሺ௡ିଵሻ

(1)

Where n is to the total number of alternative solutions that are examined (number of
rows/columns in the pairwise comparison matrix). In this study, there are four alternatives for
wetland surface area. After obtaining the CI value, which in this case is zero, the consistency
ratio (CR) can be calculated using equation 2 (Saaty, 1980):
‫ ܴܥ‬ൌ

஼ூ

(2)

Random index

The random index value for four wetland alternatives was 0.9 from the Saaty scale
(Forman, 1990). Therefore, the consistency ratio for subbasin 2 was zero, which indicates that
priority vector for subbasin 2 is acceptable.
116

This procedure was repeated to develop and test the priority vectors for each subbasin
based on the average implementation costs for different surface areas of wetlands Table 5-3 to
Table 5-6 ). The priority vector for the wetland implementation costs were calculated as 0.5428,
0.2815, 0.1164, and 0.0592 for wetland surface areas of 0.4, 0.81, 2 and 4 ha, respectively. In
addition, the consistency ratio for economic criteria was also zero, which indicates that priority
vector for subbasin 2 is acceptable.
Table 5-3. Daily streamflow calibration/validation results

Statistical
Measures

USGS Station
4175600

USGS Station
4176000

USGS Station
4176500

Calibration Validation Calibration Validation Calibration Validation
NSE
PBIAS
RSR

0.48
23.07%
0.72

0.47
8.63%
0.73

0.61
16.35%
0.62

0.47
13.35%
0.73

0.72
8.50%
0.53

0.57
0.13%
0.66

Table 5-4. Daily sediment calibration/validation results

Statistical
Measures
NSE
PBIAS
RSR

USGS Station
4175600
Calibration Validation
0.65
17.49%
0.59

0.46
-2.57%
0.74

Table 5-5. Statistical comparison of mean flow and sediment reduction for different
wetland surface areas at watershed and subbasin level

Wetland area
(ha)

Mean reduction at
watershed level
Streamflow* Sediment*
(cms)
(tons)
117

Mean reduction at
subbasin level
Streamflow* Sediment*
(cms)
(tons)

0.4
0.81
2
4
*

0.010a
0.010ab
0.010ab
0.011b

17.1a
17.4a
18.3ab
19.6b

21.5a
40.4b
90.2c
159.2d

118.9a
131.2a
139.9a
141.9a

Streamflow and sediment means followed by the same letters are not significantly different (α = 0.05).

118

Table 5-6. Distance from outlet and sediment reduction
Distance
from outlet
(km)

Number Average
of
subbasin
subbasins area (ha)

0-50

159

197.3a

Average
annual
sediment
yield
(ton)
55.6a

Average
slope (%)

1.0a

Average sediment reduction
Watershed level
(tons)
0.4
0.81
2
4
(ha)
(ha)
(ha)
(ha)
13.3a
13.6a
14.5a 15.9a

Average sediment reduction
Subbasin level
(tons)
0.4
0.81
2
4
(ha)
(ha)
(ha)
(ha)
52.0a
54.5a
55.3a
55.4a

50-100

320

237.8ac

166.1b

1.6b

14.9ab

15.2ab

16.2a

17.8a

143.3b

157.0b

164.9b

166.1b

100-150

229

273.9c

204.1b

1.9b

16.2ab

16.5ab

17.5a

19.0a

165.4b

184.3b

199.0b

203.3b

150-200

278

274.6c

176.8b

3.9c

21.9c

22.3c

23.1b

24.3b

143.1b

159.7b

173.0b

176.2b

200-250

134

214.3a

10.7a

5.0d

18.5bc

18.7bc

19.1ab

19.6a

10.3a

10.5a

10.6a

10.6a

*

Values followed by the same letters are not significantly different (α = 0.05).

119

Step 4) Ranking of alternatives: The priority vectors for different criteria were entered
into a decision matrix (Table 5-7), in which the importance weights of the criteria are
determined.

Table 5-7. Decision matrix of wetland size alternatives for all criteria developed for
watershed analysis (Subbasin 2).
Wetland size
( ha)

Sediment
reduction

Implementation
Cost

0.4

0.2327

0.5428

0.81

0.2390

0.2815

2

0.2538

0.1164

4

0.2746

0.0592

For this example, we assumed a weight of 0.9 for environmental factor (sediment reduction
at the outlet) and a weight of 0.1 for economic factor (wetland implementation cost). The final
rank of the wetlands were calculated by multiplying the decision matrix by the weights of
importance of the criteria.

0.2327
0.2390
൮
0.2538
0.2746

0.2637
0.5428
0.9
0.2432
0.2815
൲ൈቀ ቁൌ ൮
൲
0.1
0.2400
0.1164
0.2530
0.0592

From the above result, it can be concluded that 0.4 ha wetland is the most desirable
(because it has the largest weight) followed by 4 ha, 0.81 ha, and 2 ha as the final weights are
0.2637, 0.2530, 0.2432, and 0.2400, respectively.
120

Five scenarios, or combinations, of different environmental and economic weighting
factors were developed. The five different scenarios were environmental/economic (0.1/0.9),
environmental/economic (0.25/0.75), environmental/economic (0.5/0.5),
environmental/economic scenario (0.75/0.25), and environmental/economic (0.9/0.1). For
example, the environmental/economic scenario (0.1/0.9) indicates that the environmental factor
is given less preference (0.1), while the economic factor is given more importance (0.9). This
indicates that when selecting a wetland surface area for sediment reduction the cost of
implementation is more important than environmental benefit in the context of watershed scale
restoration strategy. The AHP algorithm was developed for all scenarios and the most desirable
wetland were selected.
5.3.9

Statistical Analysis
Annual sediment load generated under different wetland restoration scenarios were

compared to the base annual sediment load (before wetland implementation) to calculate
sediment load reduction. The sediment load reductions obtained by four wetland restoration
scenarios were then averaged over the simulation period to obtain the mean annual sediment
reduction at both subbasin and watershed level. Statistical analyses were performed to explore
the effect of stream order, distance from the outlet, and wetland surface area on mean annual
sediment reduction. Mean input sediment loads across different stream orders were also
compared. In order to compare the mean differences between sediment loads, the Fisher’s Least
Significant Difference test was used to perform a multiple pairwise comparison at 5%
significance level (α=0.05) using MATLAB 7.12 (R2011a).

121

5.4

RESULTS AND DISCUSSION

5.4.1

Model Calibration
The results for SWAT streamflow and sediment concentration calibration are presented

in Table 5-8 and Table 5-9, respectively. The calibration and validation was evaluated based on
the criteria established by Moriasi et al. (2007) and a visual comparison (Figure 5-4,Figure 5-5,
Figure 5-6, and Figure 5-7 ).

Table 5-8. Daily streamflow calibration/validation results

Statistical
Measures

USGS Station
4175600

USGS Station
4176000

USGS Station
4176500

Calibration Validation Calibration Validation Calibration Validation
NSE
PBIAS
RSR

0.48
23.07%
0.72

0.47
8.63%
0.73

0.61
16.35%
0.62

0.47
13.35%
0.73

Table 5-9. Daily sediment calibration/validation results

Statistical
Measures
NSE
PBIAS
RSR

USGS Station
4175600
Calibration Validation
0.65
17.49%
0.59

122

0.46
-2.57%
0.74

0.72
8.50%
0.53

0.57
0.13%
0.66

Daily Streamflow (m3/s)

14.00

USGS observed

12.00

SWAT predicted

10.00
8.00
6.00
4.00
2.00

Date
Figure 5-4. Streamflow calibration versus validation for the USGS 04175600 gauging station.

123

Jul-05

Jan-05

Jul-04

Jan-04

Jul-03

Jan-03

Jul-02

Jan-02

Jul-01

Jan-01

Jul-00

Jan-00

Jul-99

Jan-99

Jul-98

Jan-98

Jul-97

Jan-97

Jul-96

Jan-96

0.00

USGS observed

140

SWAT predicted

Daily Streamflow (m3/s)

120
100
80
60
40
20

Date
Figure 5-5. Streamflow calibration versus validation for the USGS 04176000 gauging station.

124

Jul-05

Jan-05

Jul-04

Jan-04

Jul-03

Jan-03

Jul-02

Jan-02

Jul-01

Jan-01

Jul-00

Jan-00

Jul-99

Jan-99

Jul-98

Jan-98

Jul-97

Jan-97

Jul-96

Jan-96

0

125

Date

Figure 5-6. Streamflow calibration versus validation for the USGS 04176500 gauging station.

Jul-05

Jan-05

Jul-04

Jan-04

Jul-03

Jan-03

Jul-02

Jan-02

Jul-01

Jan-01

Jul-00

Jan-00

Jul-99

Jan-99

Jul-98

Jan-98

Jul-97

Jan-97

Jul-96

Jan-96

Daily Streamflow (m3/s)
250
USGS observed

SWAT predicted

200

150

100

50

0

Daily Sediment Concentration (mg/L)

140

SWAT simulated
EPA STORET observed

120
100
80
60
40
20

Date
Figure 5-7. Sediment Concentration calibration versus validation for the STORET station 580046.

126

Jul-05

Jan-05

Jul-04

Jan-04

Jul-03

Jan-03

Jul-02

Jan-02

Jul-01

Jan-01

Jul-00

Jan-00

0

5.4.2

Wetland Impacts on Streamflow
Overland and streamflows are the main drivers of sediment transport in a watershed.

Therefore, we first discuss the overall impacts of wetland restorations scenarios on streamflow at
both watershed and subbasin levels.
In general, runoff water is captured and retained in a wetland for a short or long time
period of time based on the wetland capacity. This will ultimately reduce downstream flood
peaks and flow magnitude. However, based on the results of this study wetland installation
resulted in minimal streamflow reductions at the watershed level. Figure 5-8 shows the percent
average annual streamflow reduction at the watershed outlet was below 1.2% under all wetland
restoration scenarios. Minimal flow reduction might be due to the small surface area of the
wetlands (0.4 to 4 ha) when compared to the average subbasin area of 233 ha. These results are
consistent with the findings by Martinez-Martinez et al. (2014), where wetlands with an area less
than 50 ha did not significantly reduced streamflow at the outlet of a watershed of similar surface
area. In addition, no visual difference was observed between wetland restoration scenarios
although it was obvious that wetlands located in the headwaters of the watershed (first order
streams) perform better than ones in the mid-reaches of the watershed (Figure 5-8), which is in
line with other wetland studies (Mitsch, 1993; Martinez-Martinez et al, 2014).
Meanwhile, the impacts of wetland restoration scenarios on runoff at the subbasin level
were significantly different from the watershed level. Streamflow was substantially reduced at
the subbasin level but the range was very large, varying from 0% to 100% at different locations
Figure 5-9). There was a substantial increase in streamflow reduction at the subbasin level with
increasing wetland surface area. The average annual streamflow reductions for the four modeled
wetland restoration scenarios (0.4, 0.81, 2 and 4 hectares) at the subbasin level were: 21.45 cms,
127

40.42 cms, 90.22 cms and 159.17 cms, respectively. Only 2% of the 0.4 ha wetlands reduced
greater than 50% streamflow, which increased to 33% when the wetland surface area increased
to 4 ha. Likewise, none of the 0.4 ha wetlands reduced more than 90% of the streamflow, but
12% of the 4 ha wetlands accomplished this.

Figure 5-8. Percentage flow reduction at the watershed outlet. (a) 0.40 ha wetland
scenario, (b) 0.81 ha wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha wetland
scenario.

128

Figure 5-9. Percentage flow reduction at the watershed subbasin level. (a) 0.40 ha
wetland scenario, (b) 0.81 ha wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha
wetland scenario.
5.4.3

Wetland Impacts on Sediment Load
The SWAT/SUSTAIN hybrid model approach captured sediment reduction at the outlet

and subbasin levels. In general, wetland restoration scenarios were determined to be more
effective in reducing sediment at the subbasin level than at the watershed outlet. Watershed level
sediment reduction ranged from 0 to 189 tons/year for all wetland surface areas during the
129

simulation period. However at the subbasin level, sediment reduction ranged from 0 to 4015
tons/year. Figure 5-10 shows the impact of wetland restoration on sediment reduction at the
watershed level averaged over the simulation period while Figure 5-11 shows the sediment
reduction for the same period at the subbasin level. At the watershed level no visual difference
was observed among wetland scenarios (Figure 5-10). In addition, the subbasins with highest
impact on sediment reduction at the watershed outlet (Figure 5-11) did not align with the
subbasins that resulted in the most flow reduction (Figure 5-9) at watershed level (Figure 5-9).
Similar results were observed for the flow and sediment reductions at the subbasin level (Figure
5-9 and Figure 5-11, respectively). For example, installation of a 4 ha wetland resulted in
significant sediment reduction (>80%) for most subbasins. However, the highest flow reduction
(>80%) was only observed in less than 20% of the watersheds. Significant sediment reduction
can be explained by relatively small subbasin size to wetland area. Therefore, it can be
concluded that the streamflow reduction was not highly correlated with sediment reduction at the
subbasin level (R = - 0.25 to - 0.33) or the watershed level (R = 0.30) for all wetland restoration
scenarios.
At the subbasin level, streamflow reduction increased at a faster rate than the sediment
reduction with the increase in wetland area. However, at the watershed level, streamflow and
sediment reduction increased at a comparable rate with the increase in wetland area. At subbasin
level, a 100% increase in wetland area (0.81 ha scenario) from 0.4 ha resulted in a 70% increase
in streamflow reduction and a 10% increase in sediment reduction, whereas a 900% increase in
wetland area (4 ha scenario) resulted in 357% increase in streamflow reduction and 19% increase
in sediment reduction. At the watershed level, a 100% increase in wetland area from 0.4 ha
resulted in a 2% increase in both streamflow and sediment reduction, while a 900% increase in

130

wetland area resulted in a 17% increase in streamflow reduction and 15% increase in sediment
reduction.

Figure 5-10. Sediment reduction at the watershed outlet. (a) 0.40 ha wetland
scenario, (b) 0.81 ha wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha wetland
scenario.

131

Figure 5-11. Sediment reduction at the subbasin level. (a) 0.40 ha wetland scenario,
(b) 0.81 ha wetland scenario, (c) 2 ha wetland scenario, and (d) 4 ha wetland scenario.

Table 5-10 compares the average annual streamflow and sediment reductions based on
different wetland surface areas at the watershed and subbasin levels. At the watershed level, only
the mean annual sediment reduction provided by 4 ha wetland was significantly higher than the
sediment reduction by 0.4 ha and 0.81 ha wetlands. As sediment transport is inextricably tied to
streamflow, significant differences in sediment reduction at the watershed level might be due to
the larger flow attenuation caused by the 4 ha wetland. As shown in Table 5-10, both flow and
sediment were significantly reduced between 0.4 ha and 4 ha wetlands at the watershed level,
132

confirming the above hypothesis. However mean sediment reductions at the subbasin level were
not significantly different despite that flow rates were significantly different for each wetland
scenario.
Table 5-10. Statistical comparison of mean flow and sediment reduction for different
wetland sizes at watershed and subbasin level

Wetland area

*

Mean reduction at
watershed level

Mean reduction at
subbasin level

(ha)

Streamflow*
(cms)

Sediment*
(tons)

Streamflow*
(cms)

Sediment*
(tons)

0.4

0.010a

17.1a

21.5a

118.9a

0.81

0.010ab

17.4a

40.4b

131.2a

2

0.010ab

18.3ab

90.2c

139.9a

4

0.011b

19.6b

159.2d

141.9a

Streamflow and sediment means followed by the same letters are not significantly different (α = 0.05).

In the next two sections we explore whether characteristics of a wetland implementation
site can be used to better explain sediment dynamics in a watershed. The two characteristics
under considerations are the river distance of the wetland implementation site to the watershed
outlet and the order of the stream in which the wetland was installed.
5.4.4

Distance From Watershed Outlet and Sediment Reduction
Figure 5-12 (a) shows the relationship between sediment reduction and the distance from

the implementation site to the watershed outlet for the 0.4 ha wetland scenario. No recognizable
trend was observed. Similarly, no significant trend was observed for the remaining wetland
restoration scenarios.
133

Figure 5-12. Relationship between distance of the wetland implementation site and
the watershed outlet. (a) individual wetlands and (b) clustered wetlands for 0.4 ha
wetlands.

Further, the subbasins were grouped into five categories (distance from outlet) as shown
in Figure 5-12 (b) and pairwise comparison between the categories was performed on the
difference in sediment reduction at the subbasin and watershed levels. Table 5-11 compares
sediment reduction with average subbasin area, annual sediment yield, and slope. Average
annual sediment yield was the greatest for subbasins located between 100-150 km from the
watershed outlet and was significantly different from the subbasins between 0-50 or 200-250 km
from the outlet. However, the 100-150 km distance was on par with the subbasins at distances of
50-100 or 150-200 km from the outlet. Average annual sediment reduction for different wetland
surface areas at the subbasin level showed similar trend to the total sediment yield. Unlike the
sediment reduction at subbasin level, the wetlands at the subbasins located at a distance of 150200 km from the outlet showed the highest sediment reduction at the watershed outlet in all

134

wetland restoration scenarios. In addition, the outlet sediment reductions were significantly
higher than all shorter distances (0-50 km, 50-100 km, 100-150 km). Further exploration of the
subbasins with 150-200 km distance revealed that most subbasins belonged to stream order 1
(153 of 178), dominant soil type C (151 of 178), and forested land cover (103 of 178). Also,
mean slope of the subbasins associated with greatest sediment reduction at the watershed level
(150-200 km from the watershed outlet) had significantly greater slope than the subbasins that
were situated 0-50, 50-100 and 100-150 km but significantly less than the subbasins situated at
200-250 km from the watershed outlet.

135

Table 5-11. Distance from outlet and sediment reduction
Distance
from outlet
(km)

Number Average
of
subbasin
subbasins area (ha)

Average
annual
sediment
yield

Average
slope (%)

(ton)

Average sediment reduction
Watershed level

Average sediment reduction
Subbasin level

(tons)

(tons)

0.4

0.81

2

4

0.4

0.81

2

4

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

0-50

159

197.3a

55.6a

1.0a

13.3a

13.6a

14.5a

15.9a

52.0a

54.5a

55.3a

55.4a

50-100

320

237.8ac

166.1b

1.6b

14.9ab

15.2ab

16.2a

17.8a

143.3b

157.0b

164.9b

166.1b

100-150

229

273.9c

204.1b

1.9b

16.2ab

16.5ab

17.5a

19.0a

165.4b

184.3b

199.0b

203.3b

150-200

278

274.6c

176.8b

3.9c

21.9c

22.3c

23.1b

24.3b

143.1b

159.7b

173.0b

176.2b

200-250

134

214.3a

10.7a

5.0d

18.5bc

18.7bc

19.1ab

19.6a

10.3a

10.5a

10.6a

10.6a

*

Values followed by the same letters are not significantly different (α = 0.05).

136

5.4.5

Stream Order and Sediment Reduction
Two-way analysis of variance (ANOVA) showed that the interaction of stream order and

wetland surface area was not significant (p=1) in sediment reduction at the subbasin and
watershed levels. Hence, the interaction term was dropped and pairwise comparison was
performed using Fisher’s LSD separately for stream order and wetland surface area.
At the watershed level, wetlands at 4th order streams provided the highest annual average
sediment reduction and the wetlands associated with 2nd order streams provided the lowest
sediment reduction for all four wetland restoration scenarios. As shown in table 8, only sediment
reduction by wetlands associated with 4th order streams was significantly greater than the
sediment reduction by wetlands associated with 2nd order streams. Other than this case, stream
order did not play a significant role in sediment reduction. Table 5-12 also showed that in
general, wetland surface area does not play a major role in sediment reduction for a specific
stream order except stream orders 1 and 5, in which 4 ha wetlands reduce significantly more
sediment than the 0.4 ha wetland at the watershed level.
At the subbasin level, wetlands associated with 1st order streams produced the highest
mean sediment reduction in all wetland restoration scenarios. Nonetheless, there was no
significant difference in annual sediment reduction between the wetlands located in stream
orders except for the wetlands at 4th and 1st order streams. Wetlands located at the subbasins
associated with 1st order streams reduced significantly more sediment than wetlands located on
4th order streams, which is completely opposite to wetland performance on these streams at the
watershed level. As shown in Table 5-12, mean sediment yield in subbasins associated with 1st
order streams was the highest and was significantly higher than sediment yield in the subbasins
associated with 4th order streams. As shown in Table 5-12, wetland surface area did not
137

significantly impact sediment reduction at subbasin level in all of the wetland restoration
scenarios at all stream orders.

138

Table 5-12. Statistical comparison of the sediment reduction provided by different wetland surface areas and stream
orders at watershed level
Stream
order

No. of
subbasins

Average
annual
sediment load
yield

Average annual sediment reduction
(tons) for different stream order based

Average annual sediment reduction
(tons) for different wetland sizes
based on stream orders

on wetland sizes

(horizontal comparison)

(tons)

(vertical comparison)
0.4

0.81

2

4

0.4

0.81

2

4

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

1

571

162.7a

18.3a

18.6a

19.5a

20.9a

18.3a

18.6ab

19.5ab

20.9b

2

273

130.1ab

13.4b

13.8b

14.6b

15.8b

13.4a

13.8a

14.6a

15.8a

3

128

136.9ab

15.7ab

16.0ab

16.8ab

18.2ab

15.7a

16.0a

16.8a

18.2a

4

104

80.3b

21.1a

21.4a

22.3a

23.6a

21.1a

21.4a

22.3a

23.6a

5

31

135.8ab

19.1ab

19.5ab

20.5ab

22.1ab

19.1a

19.5a

20.5ab

22.1b

6

13

59.1ab

20.0ab

20.4ab

21.0ab

22.1ab

20.0a

20.4a

21.0a

22.1a

*Mean values followed by the same letters are not significantly different (α = 0.05).

139

5.4.6

Landuse and Sediment Reduction
Landuse type significantly impacted sediment generation per unit area, total sediment

yield, and ultimately sediment reduction at subbasin and watershed levels as shown in Table
5-13. The subbasins with dominant agricultural landuse (> 50% of the subbasin area) produced
significantly higher sediment per unit area and subbasin (sediment yield) compared to the other
landuses. Urban landuse produced higher average sediment per unit area and total sediment
compared to forest and water, but the three landuses were statistically similar for sediment
generated per unit area and total sediment yield.
The wetlands implemented in agricultural subbasins also were more effective and
reduced significantly higher sediment at the subbasin level than the wetlands in subbasins with
forest, urban or water. Wetlands within agricultural subbasins reduced significantly higher
sediment at the subbasin level, but no significant difference was observed at the watershed level
between agricultural and forested subbasins. Wetlands implemented in urban and water
dominated subbasins were not significantly different in sediment reduction at the subbasin and
watershed level. The results show that wetlands are more effective in sediment reduction when
implemented in subbasins with higher sediment yield, which, in this study were those dominated
by agriculture. As demonstrated in Table 5-14, watershed level sediment reduction within
forested subbasins is more than 20 times that of the subbasin level. The sediment produced in all
the subbasins were routed to the wetlands as point sources, and SWAT model may not have
adequately handled the point source input within forest dominated subbasins. Hence, caution
should be exercised in interpreting the sediment reduction results from the wetlands within
forested subbasins.

140

Table 5-13. Landuse type and sediment reduction
Landuse

Sediment
generation
per unit area
(ton/ha)

Average
sediment
yield per
landuse
type and
subbasin*
(tons)

Annual average sediment reduction at
watershed level

Annual average sediment reduction at
subbasin level

(tons)

(tons)

0.4

0.81

2

4

0.4

0.81

2

4

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

(ha)

Agriculture

0.72a

187.9a

18.3a

18.7a

19.6a

21.1a

157.0a

173.3a

184.7a

187.3a

Forest

0.01b

0.8b

15.9a

16.2a

16.7a

17.5a

0.7b

0.8b

0.8b

0.8b

Urban

0.09b

13.8b

8.2b

8.6b

9.5b

11.1b

11.3b

12.4b

13.4b

13.7b

Water

0.00b

0.0b

0.3b

0.3b

0.3b

0.3b

0.0b

0.0b

0.0b

0.0b

* Each subbasin contains only one landuse type

141

5.4.7

Selection of Most Suitable Wetland Considering Environmental and Economic Factors
As described earlier, five combinations of environmental and economic scenarios were

studied using AHP to evaluate the importance of considering these factors when developing
watershed-wide restoration scenarios. Figure 5-13 presents the number and surface area of
wetland selected as most effective (primary y-axis) and cost of sediment reduction per ton
(secondary y-axis) based on different environmental/economic scenarios.
Overall, the 0.4 ha wetland was selected as most suitable for almost all subbasins for three
environmental/economic scenarios (0.1/0.9, 0.25/0.75, and 0.5/0.5). This is due the much lower
cost associated with 0.4 ha wetland installation ($1490) versus $13,656 for the 4 ha wetland.
Meanwhile, the environmental benefits associated with installation of larger wetland surface area
is minimal (less than 2% between 0.4 ha and 4 ha wetlands). The breakeven point was found to
be the scenario that gave equal importance to environment as well as to economy. The breakeven
point is the location where the response variable changes sharply with respect to the predictor.
This point provides the maximum sediment reduction with an associated minimum cost. In the
environment/economic scenario (0.5/0.5), after 0.4 ha wetland, the 4 ha wetland was selected as
the most effective in a small number of subbasins due to slightly higher sediment reduction
efficiency compared to other wetland alternatives (0.8 and 2 ha). From environmental/economic
scenario (0.5/0.5) up to environmental/economic scenario (0.9/0.1), the number of subbasins in
which 4 ha was selected as optimal increased. In the 0.75/0.25 environmental/economic scenario
the 0.4 ha wetland was most favorable for a majority of the subbasins while 4 ha wetland was the
second most favorable and the two other wetland surface areas (0.8 and 2 ha) were rarely
selected. For environmental/economic scenario (0.9/0.1), the 0.4 ha wetland was widely replaced
by the 4 ha wetland as most suitable because the environmental factor was given much more
142

importance than the economic factor. The 0.8 ha and 2 ha wetlands were selected as most
effective in a limited number of subbasins only for environmental/economic scenarios 0.75/0.25
and 0.9/0.1. Overall, it can be concluded that smaller wetland surface area should be selected
reduce sediment at the watershed outlet at lowest cost. However, if cost is not a limiting factor in
a wetland restoration strategy, larger wetlands can be implemented.
The above analysis was repeated at the subbasin level (Figure 5-14). Contrary to the
watershed scale analysis, the 0.4 ha wetland was selected as the most suitable option in all
scenarios. And in four scenarios the 0.4 ha wetland was the only recommended wetland. The
breakeven point was found to be the scenario that gave more weight to the environmental
component than the economic component (0.75/0.25 environmental /economic scenario). For
0.9/0.1 environmental/economic scenario, the 0.4 ha wetland was the most suitable wetland,
while 0.8 ha, 2 ha, and 4 ha wetlands were selected as most effective in a limited number of
subbasins.
The spatial distribution of wetland placement based on different environmental /economic
benefits (watershed versus subbasin levels) are presented in Figure 5-15 and Figure 5-16,
respectively. As shown in these figures, the differences between 0.1/0.9 and 0.25/0.75, and
0.5/0.5 environmental/economic benefits at the watershed and subbasin levels were minimal and
the 0.4 ha wetland was the preferred choice. However, the differences were magnified for
0.75/0.25 and 0.9/0.1 environmental/economic scenarios. Especially, for the 0.9/0.1
environmental/economic scenario, larger wetlands were selected throughout the watershed in the
watershed scale analysis (Figure 5-15), while selection of larger wetland areas was limited to the
middle part of the watershed for the subbasin level analysis. This shows the importance of level

143

of analysis (watershed versus subbasin) for the appropriate placement of wetlands in a
watershed.
These results provide a solution to policymakers based on the importance of environmental
and economic factors. This procedure can be applied in selecting different wetland surface areas
and using different weighting factors depending on the region or goals of a specific wetland
restoration project.

144

0.4 ha

0.8 ha

2 ha

4 ha

$/ton

2000

1000

No. of wetland selected

1600

800
1200
600
800
400

400
200

0

Total cost per ton of sediment reduction ($/ton)

1200

0
0.1/0.9

0.25/0.75

0.5/0.5

0.75/0.25

0.9/0.1

Environmental/Economic Scenario

Figure 5-13. Most suitable wetland selected based on different environmental / economic scenarios (watershed scale).

145

1ac

1200

2 ac

5 ac

10ac

$/ton

103

No. of wetland selected

101
800
100
99

600

98
400
97
200

Total cost per ton of sediment reduction ($/ton)

102
1000

96
95

0
0.1/0.9

0.25/0.75
0.5/0.5
0.75/0.25
Environmental/Economic Scenario

0.9/0.1

Figure 5-14. Most suitable wetland selected based on different environmental/economic scenarios (subbasin level).

146

Figure 5-15 Most suitable wetland placement considering watershed level
environmental/economic benefits a) 0.1/0.9, b) 0.25/0.75, c) 0.5/0.5, d) 0.75/0.25, and e)
0.9/0.1.

147

Figure 5-16. Most suitable wetland placement considering subbasin level
environmental/economic benefits a) 0.1/0.9, b) 0.25/0.75, c) 0.5/0.5, d) 0.75/0.25, and e)
0.9/0.1.

148

5.5

CONCLUSION
Due to limitations of current modeling practices, functional assessment of wetland is widely

used, which in most cases is site specific and qualitative. The goal of this study is overcome
these disadvantages by coupling watershed (SWAT) and site scale (SUSTAIN) models for
evaluating the sediment reduction benefits of wetland restoration scenarios. The SWAT model
simulates land surface and river routing processes while SUSTAIN handles the implementation
of restoration wetlands in subbasins. Finally, the analytic hierarchy process was used to
incorporate environmental and economic considerations when selecting wetland restoration
scenarios.
The results of this study showed that the impacts of wetland restorations scenarios on flow
and sediment reduction at the watershed outlet was minimal (<1.5%), which can be attributed to
small surface area of wetland restoration scenarios (0.4 to 4 ha) compared to the average subbasin
size in the watershed. In addition, streamflow and sediment reduction rates were consistent with
increases in wetland area. Conversely, wetland installation resulted in significant flow and sediment
reductions at the subbasin level up to 100%.
In the next step the wetland restoration sites were examined to better explain sediment dynamics
in the watershed. Two specific site characters (order of the stream and distance to the watershed
outlet) were used for this analysis. The subbasins were grouped into five clusters based on their
distance to the outlet and six classes based on the stream order number for the implementation site.
Average annual sediment yield generated at subbasins located at 100-150 km from the watershed
outlet were the highest. For the watershed level analysis, wetlands located at 150-200 km from the
outlet performed the best while wetlands located at 100-150 km from the outlet performed the best at
subbasin level. Wetlands associated with 1st order streams performed better at subbasin level.
Conversely, wetlands located in subbasins with the 4th order streams resulted in the highest sediment

149

reduction at the outlet. Overall, larger wetlands performed better in sediment removal at both
watershed and subbasin levels, though the reductions were not statistically different.
Finally, both economic and environmental impacts of wetlands restoration scenarios were
considered for final placement of wetlands in the study area. Overall, smaller wetlands should be
selected to control the sediment reduction at the watershed outlet because of their low cost but
acceptable performance. However, if cost is not a limiting factor in a wetland restoration strategy,
larger wetlands can be implemented. Contrary to the watershed scale analysis, the 0.4 ha wetland was
selected as the most suitable option in all scenarios for the subbasin level sediment reduction goal.

This study introduced an alternative approach to functional assessment of wetlands that is
more accurate and quantitative. The approach can be used and adapted by local, state, and
national level agencies for wetland restoration initiatives to identify priority areas for the lowest
cost and the greatest environmental impact resulting from sediment control at subbasin and
watershed scales.

150

6

OVERALL CONCLUSION

The model integration exercise in this study demonstrated the impacts of wetland
restoration scenarios at the watershed scale. In addition, it quantified the benefits of
implementation based on location, size, and, depth of wetlands into a more understandable
concept compared to functional assessments. This will facilitate conversation among watershed
stakeholders that may include farmers, conservation groups, and federal and local agencies. This
study also assists watershed managers in making informed decisions and allocating conservation
dollars based on environmental and economic benefits.
The following conclusions are based on the results of the first study, “Modeling the
hydrologic significance of wetland restoration scenarios”:

•

Wetland size was determined to be an important factor in reducing streamflow at the
watershed and subbasin level. Larger wetland area resulted in more streamflow reduction due
to increased storage volume and evaporation. While differences in streamflow reduction
among wetland sizes were statistically significant, they may not be relevant to
implementation plans. Under these scenarios, a higher level of statistical significance should
be considered.

•

Long-term daily average streamflow reductions were insensitive across stream orders for
wetlands with areas less than 500 ha. However, wetland restoration of 500 ha on first and
third order streams were most effective in streamflow reduction. Wetland implementation on
first order streams was most effective at the subbasin level regardless of depth and area, but
reduction was minimal at the watershed outlet.

151

•

While the high impact maps for the watershed (smallest area and depth combination resulting
in greatest streamflow reduction) were generally similar, optimal area and depth
combinations for streamflow reduction were often different. This is useful in the decision
making process, where targeted location may not change but implementation size will change
based on project goals and resources.
The following conclusions are based on the results of the second study, “Assessing the

significance of wetland restoration scenarios on sediment mitigation plan”:

•

Overall, larger wetlands removed more sediment at the watershed and subbasin levels.
However, the impacts of wetland restoration scenarios on sediment reduction at the
watershed outlet was minimal (<1.5%) compared to the subbasin level (up to 100%).

•

The subbasin clustering based on the distance from the watershed outlet showed that the
optimum wetland restorations sites (clusters) are be different based on the goal of the study
(sediment reduction at the subbasin versus outlet)

•

Wetland implementation in sites associated with 1st order streams performed better at the
subbasin level. Conversely, wetlands located in subbasins with 4th order streams reduced the
most sediment at the watershed outlet.

•

After considering environmental benefits and economic costs of wetland restoration
scenarios, the smallest wetland size was selected for both subbasin and watershed level
analysis. The smallest wetland size was the best in terms of cost of sediment reduction in the
study area.

152

7

RECOMMENDATIONS FOR FUTURE RESEARCH

This study examined the impact of wetland area, depth, and placement on streamflow, peak flow
reduction, and sediment reduction at subbasin and watershed scales. However, there are
significant gaps in key areas of knowledge concerning wetland ecosystems functions, services
and values. Based on the results of this study areas of further research include the following:

•

Many modeling tools exist to evaluate environmental and hydrologic benefits of wetlands,
but due to their limitations (e.g. input intensive, evaluate a single or few water
quality/quantity parameters, operate only at field scales, etc.) they cannot be used for
development of watershed scale restoration plans. Integrating existing models or developing
new models is necessary to account for the wide range of wetland functions that include
additional water quality and quantity benefits.

•

Evaluate the impacts of restoration scenarios beyond water quality and quantity to include
other wetland functions such as wildlife habitat conservation, groundwater recharge,
greenhouse gas sequestration, etc.

•

Perform additional studies in unique physiographic regions to better understand wetlands role
in achieving watershed-scale management goals in different environmental settings.

•

Study the social challenges and benefits of wetland restoration in addition to the economic
and environmental aspects of sustainable watershed management.

153

APPENDICES

154

APPENDIX A
ADDITIONAL MATERIAL TO SECTION 4 TITLED: “MODELING THE
HYDROLOGICAL SIGNIFICANCE OF WETLAND RESTORATION SCENARIOS”

155

Figure 8-1. SWAT data frame for wetland modeling processes.

156

Figure 8-2. SWAT statistical data frame.

157

Figure 8-3.Percent flow change at the watershed outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 15 cm

158

Figure 8-4.Percent flow change at the watershed outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 30 cm

159

Figure 8-5.Percent flow change at the watershed outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 61 cm

160

Figure 8-6. Percent flow change at the subbasin outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 15 cm

161

Figure 8-7. Percent flow change at the subbasin outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 30 cm

162

Figure 8-8. Percent flow change at the subbasin outlet compared to base scenario –
wetlands areas: (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha at 61 cm

163

Figure 8-9. Percent flow change at the watershed outlet compared to base scenario –
50 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

164

Figure 8-10.Percent flow change at the watershed outlet compared to base scenario
– 100 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

165

Figure 8-11. Percent flow change at the watershed outlet compared to base scenario
– 250 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

166

Figure 8-12. Percent flow change at the subbasin outlet compared to base scenario –
50 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

167

Figure 8-13. Percent flow change at the subbasin outlet compared to base scenario –
100 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

168

Figure 8-14.Percent flow change at the subbasin outlet compared to base scenario –
250 ha wetlands at depths of (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

169

Figure 8-15. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 15 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

170

Figure 8-16. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 30 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

171

Figure 8-17. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 61 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

172

Figure 8-18. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 50 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

173

Figure 8-19. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 100 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

174

Figure 8-20. Percent streamflow change at the watershed outlet compared to base
scenario by stream order for 250 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

175

Figure 8-21. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 15 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

176

Figure 8-22. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 30 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

177

Figure 8-23. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 61 cm and (a) 50 ha, (b) 100 ha, (c) 250 ha, and (d) 500 ha

178

Figure 8-24. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 50 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

179

Figure 8-25. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 100 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

180

Figure 8-26. Percent streamflow change at the subbasin outlet compared to base
scenario by stream order for 250 ha and (a) 15 cm, (b) 30 cm, (c) 61 cm, and (d) 91 cm

181

Figure 8-27. Sampled subbasins

Figure 8-28. Measuring wetland depth in the Shiawassee watershed

182

Figure 8-29. Collecting GPS points for surveyed areas in the Shiawassee watershed

Figure 8-30. Soil hydrology data collection for a farmed wetland

183

APPENDIX B
ADDITIONAL MATERIAL TO SECTION 5 TITLED: “Assessing the Significance of
Wetland Restoration Scenarios on Sediment Mitigation Plan”

184

Table 9-1. SUSTAIN model parameters that were used in this study
Parameter

Value

Reference

Comments

Substrate properties
Depth of soil

0.23 m

22.8 cm of media

Soil porosity

0.398

USEPA (2006)

Sandy clay loam soil

Soil field capacity

0.244

USEPA (2006)

Sandy clay loam soil

Soil wilting point

0.136

USEPA (2006)

Sandy clay loam soil

Initial surface water depth

0.09 m

Assumed

Initial moisture content

1

Saturated soil

Saturated soil infiltration

2.54 mm/hr

ET multiplier

1

USEPA (2006)

Clay lining below the soil
media
Assumed

Infiltration parameters
Vegetative parameter A

0.8

Monthly growth index

0.1-1

USEPA (2006)

Recommended value
Assumption based on
vegetative growth in
Michigan

Water quality
parameters
K

5000 m/year

eWater (2012)

Recommended value

C*

6 mg/l

eWater (2012)

Recommended value

Porosity

0.5

USEPA (2006)

Recommended value

Sand fraction

0.530

USDA-NRCS (2013b)

Silt fraction

0.251

USDA NRCS (2013b)

Clay fraction

0.219

USDA NRCS (2013b)

Sediment

185

Table 9-1 (cont’d)
Sand effective diameter

0.1 cm

USEPA (2006)

Average of typical value

Sand velocity

5.3 cm/sec

USEPA (2006)

Average of typical value

Sand density

2.45 g/cm3

USEPA (2006)

Average of typical value

Sand coefficient

0.255

USEPA (2006)

Average of typical value

Sand exponent

2.5

USEPA (2006)

Average of typical value

Silt effective diameter

0.0034 cm

USEPA (2006)

Average of typical value

Silt velocity

0.013 cm/sec

USEPA (2006)

Average of typical value

Silt density

2.25 g/cm3

USEPA (2006)

Average of typical value

Silt deposition stress

7.42 Pa

USEPA (2006)

Average of typical value

Silt scour stress

13.17 Pa

USEPA (2006)

Average of typical value

Silt erodibility

48.12 Pa/day

USEPA (2006)

Average of typical value

Clay effective diameter

0.002 cm

USEPA (2006)

Average of typical value

Clay velocity

0.0127 cm/sec

USEPA (2006)

Average of typical value

Clay density

2.25 g/cm3

USEPA (2006)

Average of typical value

Clay deposition stress

7.42 Pa

USEPA (2006)

Average of typical value

Clay scour stress

13.17 Pa

USEPA (2006)

Average of typical value

Clay erodibility

48.12 Pa/day

USEPA (2006)

Average of typical value

186

Table 9-2. Sensitivity analysis of the SUSTAIN model parameters
Run

Parameter

Initial
Value

New value

Conditions

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

WIDTH
LENGTH
SAND_FRAC
SAND_FRAC
SAND_FRAC
SILT_FRAC
SILT_FRAC
SILT_FRAC
CLAY_FRAC
CLAY_FRAC
CLAY_FRAC
WEIRH
EXITYPE
RELEASETYPE
POROSITY
POROSITY
POROSITY
POROSITY
AVEG

466.69
933.38
0.53
0.25
0.22
0.53
0.25
0.22
0.53
0.25
0.22
0.33
1
3
0.396
0.396
0.396
0.396
0.8

Not changed
Not changed
0.22
0.53
0.25
0.22
0.53
0.25
0.22
0.53
0.25
Not changed
Not changed
Not changed
0
0.25
0.5
1
0

Depends on planned wetland size
Depends on planned wetland size
Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Depends on planned wetland size
Best represent restored wetland
Best represent restored wetland
Depends on Soil Type

20
21
22
23
24
25
26
27

AVEG
AVEG
AVEG
FCAPACITY
FCAPACITY
FCAPACITY
FCAPACITY
WPOINT

1.8
2.8
3.8
0.244
1.244
2.244
0.244
0.136

0.5
0.75
1
0,0.5, 0.75, 1
0.5
0.75
1
Not changed

Empirical value (Holtan
infiltration Method)

Depends on Soil Type

Depends on Soil Type
Depends on Soil Type
187

Flow
Change in outflow
(%)
-42.3 -0
-42.3 -0
-42.3 -0
-42.3 -0
-42.3 -0
-42.3 -0
-42.3 -0
-42.3 -0
-42.3 -0
0
0
0
0
-

Sediments
Change in
outflow (%)
-17.50 -0
-17.50 -0
-17.50 -0
-17.50 -0
-17.50 -0
-17.50 -0
-17.50 -0
-17.50 -0
-17.50 -0
0
0
0
0

0
0
0
0-0.5
0-0.5
0-0.5
0-0.5
-

0
0
0
0
0-.9
0
0
-

Table 9-2 (cont’d)
28
29
30
31
32
33

INFILTM
INFILTM
INFILTM
POLROTM
POLREMM
SDEPTH

2
2
2
3
1
0.75

0
1
2
1>
0,1
Not changed

Depends on Soil Type
Depends on Soil Type
Depends on Soil Type
Best represent restored wetland
Best represent restored wetland
Depends on Soil Type

188

0
0
0
0
0
-

0
0
0
0
0
-

Table 9-3. SUSTAIN and SWAT model parameters for models set up.
Parameter

Land Use data

NA

Streamflow data

NA

Source
National Elevation Dataset (NED) at
30 meter resolution
Cropland Data Layer (CDL) 56 meter
resolution
USGS

Stream Network

NA

National Hydrography Dataset (NHD)

Precipitation

NA

National Climatic Data Center
(NCDC). NCDC Summary

Required for internal land simulation and for estimating storm
sizes for the post-processor

Evapotranspiration

NA

National Climate Data Center
(NCDC)
GIS
SWAT
NA
NA
Web soil survey

Calculated using daily temperature

DEM

Latitudes
Land simulation option
Pollutants
Time series for land use
Sediment fraction:
Sand
Silt
Clay
BMP template
Infiltration method
Pollutant removal
Pollutant routing method
Define BMP parameters
Wetland area
Aspect ratio

Value
NA

43.361955
External
Sediment
0 input
0.530
0.251
0.219
Wet pond
Holtran
K-C* method
CSTR in series:
1.5 CSTR
1, 2, 5, 10 acres
2:1

NA
NA
SUSTAIN manual (pp. 3-58), Type B

Determinate based on watershed
characteristics
Crites, R. W., Middlebrooks, E. J., &
Reed, S. C. (2010). Natural
wastewater treatment systems. CRC
Press.
USEPA, 2000. Constructed wetlands
treatment of municipal wastewaters.

189

Comments
Required for automatic delineation of drainage areas
Required for defining land use distribution
Required for calibration of internal modeling of runoff;
recommended for system testing
Required for automatic delineation of drainage areas and for
placing on-stream management practices

NA
NA
Sediment Flag: Sediment
Modeled as point source data
NA

To incorporate the effect of vegetation
Value in manual is 1.4, rounded to 1.5

0.25:1—4:1

>1:1

Table 9-3 (cont’d)

Length
Width
Wetland depth
Weir height
Weir width

NA
NA 2087.1 ft
0.33 ft
Depth of
wetland
Width of
wetland

Substrate properties
Depth of soil
Soil porosity

0.75 ft
0.398

Soil field capacity

0.244

Soil wilting point

0.136

Initial surface water depth
Initial moisture content
Saturated soil infiltration

0.3 ft
1
0.01 in/hr

ET multiplier
Infiltration parameters
Vegetative parameter A

1

Monthly growth index
Water quality parameters
K
C*
Sediment
Bed width
Bed depth
Porosity

0.8
0.1-1

16400 ft/year
6 mg/l
Width of
wetland
Depth of
wetland
0.5

EPA/625/R-99/010. Cincinnati,
Ohio.
Based on area and aspect ratio
Based on area and aspect ratio
Based on field data collection

NA
NA
Discussed on Martinez at al., 2013

NA

NA

NA

NA

6-9 inch of media
Sandy clay loam soil
(Table 3-8 Sustain manual)
Sandy clay loam soil
(Table 3-8 Sustain manual)
Sandy clay loam soil
(Table 3-8 Sustain manual)
Assumed
Saturated
Clay lining
(Table 3-8 Sustain manual)
NA

Soil type found in the watershed
Soil type found in the watershed
Soil type found in the watershed
NA
NA
Clay lining below the soil media
NA

Recommended in Sustain manual (pp
3-54)
Assumed based on vegetative growth
in Michigan

NA
NA

Sustain manual (Table 3-17, pp 3-59)
Sustain manual (Table 3-17, pp 3-59)

NA
NA
NA

NA

NA

NA

NA

HSPF manual recommended value

NA

190

Table 9-3 (cont’d)
Sand fraction
Silt fraction
Clay fraction
Sand effective diameter
Sand velocity

0.530
0.251
0.219
0.05 in
2.1 in/sec

Web soil survey
Web soil survey
Web soil survey
HSPF manual recommended value
HSPF manual recommended value

NA
NA
NA
NA
NA

Sand density
Sand coefficient
Sand exponent
Silt effective diameter
Silt velocity
Silt density
Silt deposition stress
Silt scour stress
Silt erodibility
Clay effective diameter
Clay velocity
Clay density
Clay deposition stress
Clay scour stress
Clay erodibility

165.434 lb/ft3
0.255
2.5
0.00135 in
0.005 in/sec
140.4629 lb/ft3
0.155 lb/ft2
0.275 lb/ft2
1.005 lb/ft2
0.0008 in
0.005 in/sec
140.4629 lb/ft3
0.155 lb/ft2
0.275 lb/ft2
1.005 lb/ft2

HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value
HSPF manual recommended value

Wet pond placement

On stream BMP

Watershed delineation

Import subbasin

HSPF manual recommended
procedure
HSPF manual recommended
procedure
HSPF manual recommended
procedure
HSPF manual recommended
procedure

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

Evaluation factor
Flow

Flow
TSS
TSS

Average annual
flow volume
ft3/year
Peak discharge
cfs
Average annual
load (lb/year)
Average annual
concentration
(mg/l)

HSPF manual recommended
procedure
HSPF manual recommended
procedure
HSPF manual recommended
procedure

191

NA
NA
NA

NA
NA
NA

Table 9-3 (cont’d)
Point source
Create input file
Simulation time period
Land time series time
BMP simulation time step
CRRAT
Output time step

SWAT output
data
01/01/199612/31/2005
60 min
15 min
1.5
hourly

Martinez-Martinez, 2013

NA

Data availability

NA

NA
NA
Default value
NA

NA
NA
NA
NA

192

Table 9-4. SWAT model parameters that were adjusted during flow calibration procedure
USGS Station 04175600
Parameter
Value
Method
CN2
0.75
Multiply

USGS Station 04176500
Parameter
Value
Method
CN2
0.92
Multiply

ESCO
ALPHA_BF
SOL_AWC
OV_N
GW_DELAY
CH_N2
REVAPMN
EPCO
CH_K2
SLOPE
SURLAG
CH_N1
CH_K1

ALPHA_BF
GW_DELAY
ESCO
CH_N2
CH_K2
OV_N
GW_REVAP
REVAPMN
SOL_AWC
SLSUBBSN
SLOPE

0.85
1.00
0.80
0.15
35.00
0.10
5.00
0.01
20.00
1.30
0.30
0.04
0.01

Replace
Replace
Multiply
Add
Replace
Replace
Replace
Replace
Replace
Multiply
Replace
Replace
Replace

0.90
7.50
0.85
0.01
40.00
1.50
0.08
0.01
0.90
1.25
0.90

Replace
Replace
Replace
Replace
Replace
Multiply
Replace
Replace
Multiply
Multiply
Multiply

193

Parameter
CN2

USGS Station 04176000
Value
Method
0.88
Multiply

ALPHA_BF
GW_DELAY
ESCO
CH_N2
OV_N
RCHRG_DP
GW_REVAP
Sol_AWC
SLSUBBSN
CH_K1

0.99
20.00
0.90
0.25
1.50
0.15
0.05
0.75
1.40
0.50

Replace
Replace
Replace
Replace
Multiply
Replace
Replace
Multiply
Multiply
Replace

Table 9-5. SWAT model parameters that were adjusted during sediment calibration procedure for station STORET
580046
Parameter
CHCOV1
CHCOV2
USLE_P
ADJ_PKR
PRF
SPCON
SPEXP

Value
0.10
0.65
0.82
0.98
0.05
0.01
1.47

194

Method
Replace
Replace
Replace
Replace
Replace
Replace
Replace

Table 9-6. Cost summary for establishment and maintenance of a 0.4-hectare wetland over 10-year period (USDANRCS, 2013a)
Cost Range (US $ in 2013)
Wetland Restoration Components

Unit

Required Amount
Low

Average

High

Site Preparation

ha

0.40

175

181

250

Excavation

m3

2.76

73

109

146

Backfilling/Grading and Finishing

m3

2.76

49

73

97

Vegetation Planting

ha

0.40

190

383

575

Trees & Shrubs Site Preparation

ha

0.40

387

554

700

Operation and Maintenance

ha

0.40

190

190

190

Total Cost

$1,064

$1,490

$1,958

195

Table 9-7. Cost summary for establishment and maintenance of a 0.81-hectare wetland over 10-year period (USDANRCS, 2013a)
Cost Range (US $ in 2013)
Wetland Restoration Components

Unit

Required Amount
Low

Average

High

Site Preparation

ha

0.81

350

362

500

Excavation

m3

3.90

103

154

207

Backfilling/Grading and Finishing

m3

3.90

69

103

138

Vegetation Planting

ha

0.81

380

766

1,150

Trees & Shrubs Site Preparation

ha

0.81

774

1,108

1,400

Operation and Maintenance

ha

0.81

380

380

380

Total Cost

$2,056

$2,874

$3,774

196

Table 9-8. Cost summary for establishment and maintenance of a 2-hectare wetland over 10-year period (USDA-NRCS,
2013a)
Cost Range (US $ in 2013)
Wetland Restoration Components

Unit

Required Amount
Low

Average

High

Site Preparation

ha

2.00

875

905

1,250

Excavation

m3

6.17

163

244

327

Backfilling/Grading and Finishing

m3

6.17

109

163

218

Vegetation Planting

ha

2.00

950

1,915

2,875

Trees & Shrubs Site Preparation

ha

2.00

1,935

2,770

3,500

Operation and Maintenance

ha

2.00

950

950

950

Total Cost

$4,982

$6,947

$ 9,120

197

Table 9-9. Cost summary for establishment and maintenance of a 4-hectare wetland over 10-year period (USDA-NRCS,
2013a)
Cost Range (US $ in 2013)
Wetland Restoration Components

Unit

Required Amount
Low

Average

High

Site Preparation

ha

4.00

1,750

1,810

2,500

Excavation

m3

8.72

231

345

462

Backfilling/Grading and Finishing

m3

8.72

154

231

308

Vegetation Planting

ha

4.00

1,900

3,830

5,750

Trees & Shrubs Site Preparation

ha

4.00

3,870

5,540

7,000

Operation and Maintenance

ha

4.00

1,900

1,900

1,900

Total Cost

$9,805

$13,656

$17,920

198

Table 9-10. Statistical comparison of the sediment reduction provided by different wetland surface areas and stream
orders at subbasin level
Stream
order

No. of
subbasins

Average
annual
sediment yield
(tons)

1

571

162.7a

135.1a

149.7a

160.0a

162.1a

135.1a

149.7a

160.0a

162.1 a

2

273

130.1ab

108.3ab

119.4ab

127.8ab

129.8ab

108.3a

119.4a

127.8a

129.8a

3

128

136.9ab

113.7ab

125.1ab

134.1ab

137.0ab

113.7a

125.1a

134.1a

137.0a

4

104

80.3b

68.7b

76.0b

78.0b

80.4b

68.7a

76.0a

78.0a

80.4a

5

31

135.8ab

129.8ab

135.0ab

135.7ab

135.8ab

129.8a

135.0a

135.7a

135.8a

6

13

59.1ab

56.0ab

58.4ab

59.1ab

59.1ab

56.0a

58.4a

59.1a

59.1a

Average annual sediment reduction
(tons) for different stream order based
on wetland sizes
(vertical comparison)
0.4
0.81
2
4
(ha)
(ha)
(ha)
(ha)

*Mean values followed by the same letters are not significantly different (α = 0.05).

199

Average annual sediment reduction
(tons) for different wetland sizes based
on stream orders
(horizontal comparison)
0.4
0.81
2
4
(ha)
(ha)
(ha)
(ha)

REFERENCES

200

REFERENCES
Acreman M., J. Holden. 2013. How wetlands affect floods. Wetlands 33:773-786.
Adamus, P. R., Stockwell, L. T., Clairain Jr, E. J., Morrow, M. E., & Rozas, L. P.
1991. Wetlands Evaluation Technique (WET). Volume 1: Literature Review and
Evaluation Rationale (No. WES/TR/WRP-DE-2).
Adamus, P.R. 1987. Wetland evaluation technique (WET) volume II: Methodology. US Army
Corp of Engineers, Vicksburg, Mississippi State.
Alvi, K., Zhen, J. X., Paul, S., Rafi, T. 2009. SUSTAIN a framework for placement of best
management practices in urban watersheds to protect water quality. EPA/600/R-09/095,
United States Environmental Protection Agency, Office of Research and Development,
National Risk Management Research Laboratory- Water Supply and Water Resources
Division, Cincinnati, OH.
Amman, A.P., and A.L. Stone. 1991. Method for the Comparative Evaluation of Nontidal
Wetlands in New Hampshire, New Hampshire Department of Environmental Services,
Concord, New Hampshire.
Apfelbaum S, Bell J, Roland R, Mackey S, Khoury M, Hinz L. 2007. Identifying and valuing
restoration opportunities and resource improvements at watershed and subwatershed
scales. Final Report GLPF Grant 758. [Online] Available:
http://www.glpf.org/sites/default/files/project_files/758%201%20Exec%20Summ.pdf.
Accessed: 1 May 2012.
Arheimer, B., G. Torstensson, and H. B. Wittgren. 2004. Landscape planning to reduce coastal
eutrophication: agricultural practices and constructed wetlands. Landscape and Urban
Planning 67:205–15.
Arnold, J. G., & Fohrer, N. 2005. SWAT2000: current capabilities and research opportunities in
applied watershed modelling. Hydrological processes,19(3), 563-572.
Arnold, J.G., Allen, P.M., Morgan, D.S. 2001. Hydrologic model for design and constructed
wetlands. Wetlands 21(2): 167–178.
Arnold, J.G., Kiniry, J.R., Srinivasan, R., Williams, J.R., Haney, E.B., Neitsch, S.L. 2011. Soil –
and Water Assessment Tool Input/Output File Documentation, Version 2009. Agrilife
Blackland Research Center, Temple, TX.
Arnold, J.G., R. Srinivasan, R.S. Muttiah, and J.R. Williams. 1998. Large area hydrologic
modeling and assessment Part I: Model development. Journal of American Water
Resources Association. 34(1):73-89.

201

Aurela M., Laurila T. and Tuoviner J.P. 2001. Seasonal CO2 balances of a subarctic mire. J.
Geophys. Res. 106: 1623–1638. Available at
http://www.umimra.org/action.alerts/action_alerts5.htm.
Barbier, E.B. 1989. The Economic Value of Ecosystems: 1 – Tropical Wetlands. LEEC
Gatekeeper Series 89-02. London Environmental Economics Centre, London.
Barbier, E.B. 1993. Valuing Tropical Wetland Benefits: Economic Methodologies and
Applications. Geographical Journal. Part 1, 59: 22-32.
Barbier, E.B. 1994. Valuing Environmental Functions: Tropical Wetlands. Land Economics. 70
(2): 155-73.
Barendregt, A., Wassen, M. J., and Schot, P. P. 1995. Hydrological systems beyond a nature
reserve, the major problem in wetland conservation of the Naardermeer (The
Netherlands). Biol. Cons. 72: 393–405.
Batt, B. D. J., M. G. Anderson, C. D. Anderson, and F. D. Caswell. 1989. The use of prairie
potholes by North American ducks. p. 204–227. In A. van der Valk (ed.) Northern Prairie
Wetlands. Iowa State University Press, Ames, IA, USA.
Beasley, D. B., L. F. Huggins, and E. J. Monke. 1980. ANSWERS: A model for watershed
planning. Trans. ASAE 23(4): 938-944.
Bedford BL. 1999. Cumulative effects on wetland landscapes: links to wetland restoration in the
United States and southern Canada. Wetlands 19: 775–88.
Bell, F.W. 1997. The economic value of saltwater marsh supporting marine recreational fishing
in the southeastern United States. Ecol. Econ. 21, 243–254.
Benaman, J. and C. A. Shoemaker. 2004. Methodology for analyzing ranges of uncertain model
parameters and their impact on total maximum daily load processes. J. Environ. Engr.
130(6):648- 656.
Benke, A.C., I. Chaubey, G.M. Ward, and E.L. Dunn. 2000. Flood pulse dynamics of an
unregulated river floodplain in the Southeastern U.S. Coastal Plain. Ecology. 81(10)
2730–2741.
Bergh V. D., J., A. Bartender, A. Gilbert, M. van Herwijnen, P. van Horssen, P. Kandelaars and
C. Lorenz. 2001. Spatial economic–hydroecological modelling and evaluation of land use
impacts in the Vecht wetlands area. Environmental Modeling and Assessment 6: 87-100.
Bhaduri, B., Minner, M., Tatalovich, S., & Harbor, J. 2001. Long-term hydrologic impact of
urbanization: a tale of two models. Journal of Water Resources Planning and
Management, 127(1), 13-19.

202

Bicknell, B. R., J. C. Imhoff, J. L. Kittle, Jr., T. H. Jobes, and A. S. Donigian, Jr. 2000.
Hydrological Simulation Program − Fortran (HSPF): User’s Manual for Release 12.
Athens, Ga.: USEPA.
Bicknell, B.R., Imhoff, J.C., Kittle, J.L., Jr., Donigian, A.S., Jr., and Johanson, R.C. 1997.
Hydrological Simulation Program--Fortran, Users manual for version 11: U.S.
Environmental Protection Agency, National Exposure Research Laboratory, Athens, Ga.,
EPA/600/R-97/080, 755 p.
Biebighauser, T. R. 2007. Wetland drainage, restoration, and repair. University Press of
Kentucky, Lexington, KY.
Biebighauser, T. R. 2011. Wetland Restoration and Construction: A Technical Guide. Upper
Susquehanna Coalition, Owego, NY.
Biswas, S., Vacik, H., Swanson, M. E., Haque, S. S. 2012. Evaluating integrated watershed
management using multiple criteria analysis—a case study at Chittagong Hill Tracts in
Bangladesh. Environ. Monit. Assess. 184(5), 2741-2761.
Borah, D. K. 2002. Watershed-scale nonpoint-source pollution models: Mathematical bases.
ASAE Paper No. 022091. Presented at the 2002 ASAE Annual International
Meeting/CIGR World Congress. St. Joseph, Mich.: ASAE.
Borin M, Bonaiti G, Santamaria G, Giardini L. 2001. A constructed surface flow wetland for
treating agricultural waste waters. Water Sci Technol;44:523–30.
Borin, M., and D., Tocchetto. 2007. "Five year water and nitrogen balance for a constructed
surface flow wetland treating agricultural drainage waters." Science of the Total
Environment 380.1: 38-47.
Bouraoui, F., and T. A. Dillaha. 1996. ANSWERS-2000: Runoff and sediment transport model.
J. Environ. Eng. 122(6): 493-502.
Bouraoui, F., I. Braud, and T. A. Dillaha. 2002. ANSWERS: A nonpoint-source pollution model
for water, sediment, and nutrient losses. Chapter 22 in Mathematical Models of Small
Watershed Hydrology and Applications, 833-882. V. P. Singh and D. K. Frevert, eds.
Highlands Ranch, Colo.: Water Resources Publications.
Braskerud, B.C. 2002. Factors affecting nitrogen retention in small constructed wetlands treating
agricultural non-point source pollution. Ecol. Eng. 18:351–370.
Brinson, M. M. 1993. "A hydrogeomorphic classification for wetlands," Technical Report WRPDE-4, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. NTIS No.
AD A270 053.

203

Brix, H. 1993. Wastewater treatment in constructed wetlands: system design, removal processes,
and treatment performance, in: Moshiri, G.A. (Ed.), Constructed Wetlands for Water
Quality Improvement. CRC Press Inc.., Boca Raton, FL, pp 9-22.
Bubier J.L., Bhatia G., Moore T.R., Roulet N.T. and Lafleur P.M. 2004. Between year and site
variability in growing season net ecosystem CO2 exchange at a large peatland. Ontario,
Canada, Ecosystems, in press.
Burian, S.J., G.E. Streit, T.N. McPherson, M.J. Brown, and H.J. Turin. 2001. Modeling the
atmospheric deposition and stormwater washoff of nitrogen compounds. Environmental
Modelling & Software 16(5): 467–479.
Burkett, V., Kusler, J. 2000. Climate change: Potential impacts and interactions in wetlands of
the United States. J. Am. Water Resour. Assoc. 36:313-320.
Caldwell, P. V., Vepraskas, M. J., Skaggs, R. W., Gregory, J. D. 2007. Simulating the water
budgets of natural Carolina bay wetlands. Wetlands, 27(4), 1112-1123.
Campbell, K.L., and H.P. Johnson. 1975. Hydrologic Simulation of Watersheds with Artificial
Drainage. Water Resources Research 11(1): 120-126.
Carter, R. W. 1961. Magnitude and frequency of floods in suburban areas. U.S. Geological
Survey Paper 424-B, B9-B11.Washington, DC: U.S. Geological Survey.
Carter, V. 1996. Wetland hydrology, water quality, and associated functions. In: J.D. Fretwell,
J.S. Williams, and P.J. Redman (compilers). National Water Summary on Wetland
Resources. U.S. Geological Survey, Reston, VA. Water-Supply Paper 2425. pp. 35-48.
Carter, V., Novitzki, R., Hook, D.D. 1988. Some comments on the relation between groundwater
and wetlands. Ecology Manage.Wetlands 1: 68-86.
CDL. 2012. USDA national agricultural statistics service cropland data layer. Published cropspecific data layer. Available online at: http://nassgeodata.gmu.edu/CropScape//.
Assessed 20 July 2013.
Chang, C. L., Chao, Y. C. 2012. Using the analytical hierarchy process to assess the
environmental vulnerabilities of basins in Taiwan. Environ. Monit. Assess. 184(5), 29392945.
Chen X., S. Bao, H. Li, Z. Wu, W. Fu, X. Cai, H. Zhao, Y. Liu, G. Peng. 2006. Modeling The
Impacts of Land Use/Cover Change on Sediment Loads in Wetlands of the Pouyang Lake
Basin. Hydrology and Management of Forested Wetlands Proceedings of the
International Conference. 701P0406.
Chen, Y., Yu, J., Khan, S. 2010. Spatial sensitivity analysis of multi-criteria weights in GISbased land suitability evaluation. Environ. Modell. Softw., 25(12), 1582-1591.
204

Choi, K.-S., and J.E. Ball. 2002. Parameter estimation for urban runoff modelling. Urban Water
4(1): 31–41.
Comin, F. A., J. A. Romero, V. Astorga & C. Garcia. 1997. Nitrogen removal and cycling in
restored wetlands used as filters of nutrients for agricultural runoff. In Wetland systems
for water pollution control 1996. Proceedings of the 5th International Conference,
Vienna, Austria, 15-19 September 1996., 255-261.
Comin, F.A., Romero, J.A., Hern´andez, O., Men´endez, M. 2001. Restoration of wetlands from
abandoned rice fields for nutrient removal, and biological community and landscape
diversity. Restoration Ecol. 9 (2), 201–208.
Conan, C., de Marsily, G., Bouraoui, F., Bidoglio, G. 2003. A long-term hydrological modelling
of the Upper Guadiana river basin (Spain). Phys. Chem. Earth, Parts A/B/C 28:193-200.
Copeland, C. 2000. Water Quality: Implementing the Clean Water Act. Washington, DC:
Congressional Research Service. The National Council for Science and the Environment.
Copeland, C. 2006. Water quality: Implementing the clean water act. Washington, DC:
Congressional Research Service. Retrieved from:
http://www.nationalaglawcenter.org/assets/crs/RL33466.pdf.
Correll, D. 1998. The role of phosphorus in the eutrophication of receiving waters: a review.
Journal of Environmental Quality 27:261–266.
Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B.1997. The value of the
world’s ecosystem services and natural capital. Nature 387, 253–260.
Costanza, R., Farber, S.C., Maxwell, J. 1989. Valuation and management of wetland ecosystems.
Ecol. Econ. 1, 335– 361.
Cowardin, L.M., Carter, V., Golet, F.C., La Roe, E.T. 1979. Classification of Wetlands and
Deepwater Habitats in the United States. Fish and Wildlife Service Report FWS OBS79/31. Department of the Interior, Washington D.C.
Crumpton WG and Baker JL. 1993. Integrating wetlands into agricultural drainage systems:
predictions of nitrate loading and loss in wetlands receiving agricultural subsurface
draininage. In: Mitchell J (Ed). Constructed wetlands for water quality improvement. St.
Joseph, MI: American Society of Agri-cultural Engineers. p 118-26.
Crumpton WG, Isenhart TM, and Fisher SW. 1993. Fate of non-point source nitrate loads in
freshwater wetlands: results from experimental wetland mesocosms. In: Moshiri GA
(Ed). Integrated resource management & landscape modification for environmental
protection. Boca Raton, FL: CRC Press, Inc/Lewis Publishers. p 283-91.

205

Cui, J., Li, C., Sun, G., Trettin, C. 2005. Linkage of MIKE SHE to Wetland-DNDC for carbon
budgeting and anaerobic biogeochemistry simulation. Biogeochemistry 72, 147–167,
doi:10.1007/s10533-004-0367-8.
Dahl, T.E. 2000. Status and Trends of Wetlands in the Conterminous United States 1986–1997.
U.S. Fish and Wildlife Service. Washington, D.C. 82p.
Dahl, T. E., Allord, G. J. 1997. Technical aspects of wetlands: History of wetlands in the
conterminous United States. USGS Water Supply Paper 2425. Washington, D.C.: United
States Geological Survey. Available at:
http://water.usgs.gov/nwsum/WSP2425/history.html Accessed 1 June 2012.
Dahl, T.E. 1990. Wetland losses in the United States: 1780's to 1980's. U.S. Department of the
Interior, Fish and Wildlife Service, Washington, D.C. Jamestown, ND: Northern Prairie
Wildlife Research Center (Version 16JUL97). Retrieved from:
http://www.npwrc.usgs.gov/resource/wetlands/wetloss/index.htm.
Dahl, T.E. 2000. Status and trends of wetlands in the conterminous United States 1986–1997.
U.S. Fish and Wildlife Service. Retrieved from: http://library.fws.gov/Pubs9/wetlands8697_lowres.pdf.
Dai, Z., Li, C., Trettin, C., Sun, G., Amatya, D., Li, H. 2010. Bi-criteria evaluation of MIKE
SHE model for a forested watershed on South Carolina coastal plain. Hydrology and
Earth System Sciences Discussions, 7(1), 179-219.
Day J, et al. 2007. Restoration of the Mississippi Delta: Lessons from hurricanes Katrina and
Rita. Science 315: 1679 -1684.
De Laney, T. A. 1995. Benefits to downstream flood attenuation and water quality as a result of
constructed wetlands in agricultural landscapes. J. Soil Water Cons. 50 (6), 620-626.
De Mars, H. and Garritsen, A. C. 1997. Interrelationship between water quality and groundwater
flow dynamics in a small wetland system along a sandy hill ridge. Hydrol. Processes 11:
353–351.
Demissie, M., and Kahn, A. 1993. “Influence of wetlands on streamflow in Illinois,” Contract
Report 561, Illinois State Water Survey, Champaign, IL.
Department of Environmental Quality (DEQ). 2010. The Status and Trends of Michigan’s
Wetlands. Online source, Accessed on 1/29/2014. Available:
http://www.michigan.gov/documents/deq/wrd-waterwords-2013-10-08_436561_7.pdf
Development Assistance Committee Organization for Economic Co-operation and Development
(OECD). 1996. Guidelines for aid agencies for improved conservation and sustainable
use of tropical and sub-tropical wetlands. OECD.

206

Dodge, K. E. 1998. River Raisin assessment, in: Fisheries Special Report No. 23. Michigan
Department of Natural Resources, Fisheries Division, Ann Arbor, MI.
Donigian, A. S., Jr., B. R. Bicknell, and J. C. Imhoff. 1995. Hydrological simulation program Fortran (HSPF). Chapter 12 in Computer Models of Watershed Hydrology , 395-442. V.
P. Singh, ed. Highlands Ranch, Colo.: Water Resources Publications.
Dorge, J. 1994. Modeling nitrogen transformations in freshwater wetlands. Estimating nitrogen
retention and removal in natural wetlands in relation to their hydrology and nutrient
loadings.
Drexier, J.Z., Bedford, B.L., DeGaetano, A.T., Siegel, D.I. 1999. Quantification of the water
budget and nutrient loading in a small peatland. J. Am. Water Resour. Assoc. 35(4):
753–769.
Drexier, J.Z., Bedford, B.L., DeGaetano, A.T., Siegel, D.I. 1999. Quantification of the water
budget and nutrient loading in a small peatland. J. Am. Water Resour. Assoc. 35(4):
753–769.
Ehrenfeld, J.G., H.B. Cutway, R. Hamilton IV, and E. Stander. 2003. Hydrologic description of
forested wetlands in northeastern New Jersey, USA—an urban/suburban region.
Wetlands 23(4): 685–700.
Engel, B., Storm, D., White, M., Arnold, J.G. 2007. A hydrologic/water quality model
application protocol. J. Am. Water Resour. Assoc. 43(5): 1223-1236.
eWater. 2012. Appendix G: Selecting Appropriate k and C* Values. eWater, Australian Capital
Territory, Australia. Available online at:
http://training.ewater.com.au/mod/resource/view.php?id=41
Ewel, K.C. 1990. Multiple demands on wetlands. BioScience 40: 660-666.
Ewel, K. 1997. Water quality improvement: evaluation of an ecosystem service. Pages 329-344
in G. Daily, editor. Nature's Services: Societal Dependence on Natural Ecosystems.
Island Press, Washington, D.C.
Fan, X., Cui, B., Zhao, H., Zhang, Z. 2012. The changes of wetland network pattern associated
with water quality in the Pearl River Delta, China. Clean–Soil Air Water 40(10): 10641075.
Feng, X.Q., Zhang, G.X., Xu, Y.J. 2013. Simulation of hydrological processes in the Zhalong
wetland within a river basin, Northeast China. Hydrology and Earth System Sciences. 17:
2797-2807.
Fleischer, S., A. Gustafson, A. Joelsson, J. Pansar, and L. Stibe. 1994. Nitrogen removal in
created ponds. Ambio 23:349–357.
207

Forman, E. H. 1990. Random indices for incomplete pairwise comparison matrices. Eur. J. Oper.
Res. 48(1), 153-155.
Gassman, P.W., Reyes, M. R., Green, C.H., and Arnold, J.G. 2005. SWAT peer-reviewed
literature: A review. In Proceedings of the 3rd International SWAT Conference, 12-15
July, Zurich, Switzerland: Swiss Federal Institute for Environmental Science and
Technology.
Gassman, P.W., Reyes, M.R., Green, C.H., Arnold, J.G. 2007. The soil and water assessment
tool: Historical development, applications, and future research directions. T ASABE
50(4): 1211-1250.
Gassman, P.W., Sadeghi, A.M., Srinivasan, R. 2014. Applications of the SWAT model special
section: overview and insights. J. Environ. Qual. 43(1), 1-8.
Gibbs, J.P. 1995. Hydrologic needs of wetland animals. Pages 267-276 in W. A, Nierenberg,
editor. Encyclopedia of environmental biology. Volume 2. Academic Press, New York.
Gilvear, D. J., and R. J. McInnes. 1994. "Wetland hydrological vulnerability and the use of
classification procedures: a Scottish case study." Journal of environmental
management 42.4: 403-414.
Giri, S., Nejadhashemi, A. P. 2014. Application of analytical hierarchy process for effective
selection of agricultural best management practices. J. Environ. Manage. 132: 165-177.
Gleason, R. A. and Euliss, N. H Jr. 1998. Sedimentation of prairie wetlands. Great Plains
Research 8: 97–112.
Gleason, R.A., Laubhan, M.K., and Euliss, N.H., Jr. 2008. Ecosystem services derived from
wetland conservation practices in the United States Prairie Pothole Region with an
emphasis on the U.S. Department of Agriculture Conservation Reserve and Wetlands
Reserve Programs: U.S. Geological Professional Paper 1745, 58 p.
Gosselink, J.G., Odum, E.P., Pope, R.M., 1974. The Value of the Tidal Marsh.’ Center for
Wetland Resources Publ. LSU-SG-74-03, Louisiana State University, Baton Rouge, 30
pp.
Gren, I.-M. 1994. Costs and benefits of restoring wetlands: two Swedish case studies. Ecol. Eng.
4, 153–162.
Gren, I.-M., Folke, C., Turner, K., Batemen, I. 1994. Primary and secondary values of wetland
ecosystems. Environ. Res. Econ. 4, 55–74.
Hammer, D.A. 1992. Creating Freshwater Wetlands. Lewis Publishers, Inc., Chelsea, MI.

208

Hammer, D.A. 1993. Designing constructed wetlands systems to treat agricultural nonpoint
source pollution. In: R. K Olson (ed). Created and natural wetlands for controlling
nonpoint source pollution. U. S. Environmental Protection Agency (EPA), Office of
Wetlands, Oceans and Watersheds. Boca Raton: CRC Press LLC.
Hart, R.A. 1995. Mussel (bivalvia: unionidae) habitat suitability criteria for the Otter Tail River,
Minnesota. MS thesis. North Dakota State University.
Hassan R, Scholes R, Ash N. 2005. Ecosystems and human well-being. Current state and trends.
Washington, DC: Island Press.
Hattermann, F. F., Krysanova, V. & Hesse, C. 2008. Modelling wetland processes in regional
applications. Hydrol. Sci. J. 53(5), 1001–1012 (this issue).
Hawk, L.L., A.P. Lipstein, and H.T. Solanki. 1999. Wetland hydroperiod analysis. Proceedings
from the Sixth Biennial Stormwater Research and Watershed Management Conference at
Tampa, Florida, 14–17 September 1999.
Healy, M. & A. M. Cawley. 2002 Nutrient processing capacity of a constructed wetland in
western Ireland. Journal of Environmental Quality, 31, 1739-1747.
Heatwole, C. D. 1988. Class Notes from BSE 5354: NPS modeling. Virginia Polytechnic
Institute and State University, BSE Department.
Heimlich R.E., Weibe K.D., Claassen R., Gadsy D., House R.M. 1998. Wetlands and agriculture:
private interests and public benefits, Resource Economics Division, E.R.S., USDA,
Agricultural Economic Report 765.10.
Heimlich, R.E., and D.M. Gadsby. 1994. "Strategies for Wetlands Protection & Restoration,"
Agricultural Outlook. U.S. Dept. Agr., Econ. Res. Serv., Sept. pp.32-37.
Hey DL, McGuiness D, Beorkrem MN. 2002. Flood damage reduction in the Upper Mississippi
River Basin: an ecological means. Minneapolis, MN: The McKnight Foundation.
Holden, J., M. Gascoign, N.R. Bosanko. 2007. Erosion and natural revegetation associated with
surface land drains in upland peatlands. Earth Surface Processes and Landforms.
32:1547-1557.
Hruby, T. 1999. "Assessments of wetland functions: What they are and what they are
not." Environmental Management 23.1: 75-85.
Huber, W.C., Dickinson, R. E. 1988. Storm water management model: User’s manual version
IV. US Environmental Protection Agency, National Risk Management Research
Laboratory, Office of Research and Development, Cincinnati, OH.

209

Hunt, P.G., K.C. Stone, F.J. Humenik, T.A. Matheny, and M.H. John-son. 1999. In-stream
wetland mitigation of nitrogen contamination in a USA coastal plain stream. J. Environ.
Qual. 28:249–256.
James, W., Huber, W.C., Pitt, R.E., Dickinson, R.E., Rosener, L.A., Aldrich, J.A., James, W.R.
2002. SWMM 4 User’s Manual (User’s Guide to the EPA Stormwater management
Model and to PCSWMM). University of Guelph and CHAI, Guelph, Ontario, Canada.
Jenks, G.F. 1967. The data model concept in statistical mapping. International Yearbook of
Cartography 7, 186-190.
Jia, Z., & Luo, W. 2009. A modified climate diagram displaying net water requirements of
wetlands in arid and semi-arid regions. Agricultural water management, 96(9), 13391343.
Jiang, C.L., Fan, X.Q., Cui, G.B., Zhang, Y.B. 2007. Removal of agricultural non-point source
pollutants by ditch wetlands: implications for lake eutrophication control. Hydrobiologia
581, 319–327.
Johnston, C. A., N. E. Detenbeck, and G. J. Niemi. 1990. The cumulative effect of wetlands on
stream water quality and quantity. A landscape approach. Biogeochemistry 10:105-141.
Jordan, T.E., D.F. Whigham, K. Hofmockel, and N. Gerber. 1999. Restored wetlands in crop
fields control nutrient runoff. p. 49–60. In J. Vymazal (ed.) Nutrient cycling and retention
in natural and constructed wetlands. Proc. of a Conf. in Trebon, Czech Republic.
September 1997. Backhuys Publ., Leiden, the Netherlands.
Jordan, T.E., D.F. Whigham, K.H. Hofmockel, and M.A. Pittek. 2003. Nutrient and sediment
removal by a restored wetland receiving agricultural runoff. Journal of Environmental
Quality 32:1534-1547.
Kadlec, R.H. 1993. Natural wetland treatment at Houghton Lake Michigan: The first fifteen
years. AC93-023-003. P. 73-84. In 66th Annual Conference and Exposition. Anaheim,
CA. 3-7 Oct. 1993. Water Environ. Fed., Alexandria, VA.
Kadlec, R. H. 2008. The effects of wetland vegetation and morphology on nitrogen processing.
Ecol. Eng. 33(2), 126-141.
Kadlec, R.H., Knight, R.L. 1996. Treatment Wetlands. Lewis Publishers, Boca Raton, FL.
Kantrud, H. A., G. L. Krapu, and G. A. Swanson. 1989. Prairie basin wetlands of the Dakotas: a
community profile. U.S. Fish and Wildlife Service, Washington, DC, USA. Biological
Report 85(7.28).
Keddy PA. 2000. Wetland ecology: principles and conservation. Cambridge, UK: Cambridge
University Press.
210

Kelly, J. R., and M.A. Harwell. 1985. Comparisons of processing of elements by ecosystems. I.
Nutrients. Pages 137-157 in P.J. Godfrey, E. R. Kaynor, S. Pelczarski, and J. Benforado
(eds.), Ecological considerations in wetlands treatment of municipal wastewater. Van
Nostrand Reinhold, New York.
Kent, D. M. 2001. Evaluating wetland functions and values. Applied wetlands science and
technology, 2, 55-80.
Kettunen A., Kaitala V., Lehtinen A., Lohila A., Alm J., Silvola J. and Martikainen P.J. 1999.
Methane production and oxidation potentials in relation to water table fluctuations in two
boreal mires. Soil Biol. Biochem. 31: 1741–1749.
Kirk, J.A., W.R. Wise, and J.J. Delfino. 2004. Water budget and cost-effectiveness analysis of
wetland restoration alternatives: a case study of Levy Prairie, Alachua County, Florida.
Ecological Engineering 22(1): 43–60.
Knutilla, R.L., Allen, W.B. 1975. Water resources of the River Raisin basin, southeastern
Michigan. United States Geological Survey, Reston, VA.
Konyha, K.D., D.T. Shaw, and K.W. Weiler. 1995. Hydrologic design of a wetland: advantages
of continuous modeling. Ecological Engineering 4(2): 99–116.
Koob, T., Barber, M. E., & Hathhorn, W. E. 1999. HYDROLOGIC DESIGN
CONSIDERATIONS OF CONSTRUCTED WETLANDS FOR URBAN
STORMWATER RUNOFF1. JAWRA Journal of the American Water Resources
Association, 35(2), 323-331.
Kovacic, D.A., M.B. David, L.E. Gentry, K.M. Starks, and R.A. Cooke. 2000. Effectiveness of
constructed wetlands in reducing nitrogen and phosphorus export from agricultural tile
drainage. J.Environ. Qual. 29:1262–1274.
Kuenzler, E. J. 1990. Wetlands as sediment and nutrient traps for lakes. In Enhancing the States'
Lake and Wetland Management Programs, ed. J. Taggart, 105-12. Chicago: U.S.
Environmental Protection Agency, North American Lake Management Society,
Northeastern Illinois Planning Commission, Proceedings of a National Conference.
Lafleur P.M., Roulet N.T. and Admiral S.W. 2001. Annual cycle of CO2 exchange at a bog
peatland. J. Geophys. Res. 106: 3071–3081.
Lai, F. H., Dai, T., Zhen, J., Riverson, J., Alvi, K., & Shoemaker, L. 2007. SUSTAIN-AN EPA
BMP process and placement tool for urban watersheds. Proceedings of the Water
Environment Federation, (5), 946-968.
Lai, F. H., Shoemaker, L., Alvi, K., Riverson, J., & Zhen, J. 2010. Current Capabilities and
Planned Enhancements of SUSTAIN. In World Environmental and Water Resources
Congress 2010@ sChallenges of Change (pp. 3271-3280). ASCE.
211

Larson, J.S., Adamus, P.R. & Clairian, E.J. Jnr. 1989. Functional assessment of freshwater
wetlands: A manual and training outline. WWF and Environmental Institute, University
of Massachusetts, Amherst, U.S.A.
Leavesley, G. H., and L. G. Stannard. 1995. The precipitation-runoff modeling system - PRMS.
Chapter 9 in Computer Models of Watershed Hydrology, 281-310. V. P. Singh, ed.
Highlands Ranch, Colo.: Water Resources Publications.
Leavesley, G.H., Lichty, R.W., Troutman, B.M., and Saindon, L.G. 1983. Precipitation-Runoff
Modeling System: Users Manual: U.S. Geological Survey Water-Resources
Investigations Report 83-4238, 207 p.
Lee, J. G., Selvakumar, A., Alvi, K., Riverson, J., Zhen, J. X., Shoemaker, L., & Lai, F. H. 2012.
A watershed-scale design optimization model for stormwater best management practices.
Environmental Modelling & Software, 37, 6-18.
Leibowitz, S. G., E. M. Preston, L. Y. Arnaut, N. E. Detenback, C. A. Hagley, M. E. Kentula, R.
K. Olson, W. D. Sanville, and R. R. Sumner. 1992. Wetland research plan FY92-96: An
integrated risk-based approach. EPA/600/R-92/060, US Environmental Protection
Agency, Environmental Research Laboratory, Corvallis, Oregon.
Leopold, Luna B. 1968. Hydrology for urban planning, a guidebook on the hydrologic effects of
urban land use. U.S. Geological Survey Circular 554. Washington, DC: U.S. Department
of the Interior. Retrieved from:
http://enviro.lclark.edu/resources/Tryon/Water/Hydrology.pdf.
Li, S. N., G. X. Wang, W. Deng, Y. M. Hu, W. W. Hu. 2009. Influence of Hydrology Process on
Wetland Landscape Pattern: A Case Study in the Yellow River Delta, Ecol. Eng. 2009,
39, 1719–1726.
Lin, S.-S., S.-H. Hsieh, J.-T. Kuo, Y.-P. Liao, and Y.- C. Chen. 2006. Integrating legacy
components into a software system for storm sewer simulation. Environmental Modelling
& Software 21(8): 1129–1140.
Linkov, I., Steevens, J. 2013. A detailed analysis of the theoretical foundation of different
MCDA (Chapter 35). Appendix A: Multi-Criteria Decision Analysis (Online)
http://www.epa.gov/cyano_habs_symposium/monograph/Ch35_AppA.pdf.
Linsley, R.K., Franzini, J.B., Freyberg, D.L., Tchobanoglous, G. 1992. Water-resources
Engineering, fourth ed. McGraw-Hill, New York, NY.
Liu, Y., Yang, W., & Wang, X. 2008. Development of a SWAT extension module to simulate
riparian wetland hydrologic processes at a watershed scale. Hydrol. Processes, 22(16),
2901-2915.
Luecke, D. F. 1993. Commentary on the Mississippi Flood. Environment 35:4.
212

Maidment, D. 1993. Handbook of Hydrology. McGraw-Hill, New York, NY.
Mander, U., A. Kull, V. Kuusemets, and T. Tamm. 2000. Nutrient runoff dynamics in a rural
catchment: Influence of land–use changes, climatic fluctuations and ecotechnological
measures. Ecological Engineering 14:405–17.
Marinoni, O. 2009. AHP 1.1 e Decision support tool for ArcGIS (Online)
http://arcscripts.esri.com/details.asp?dbid¼13764.
Martinez-Martinez, E., Nejadhashemi, A. P., Woznicki, S. A., Love, B. J. 2014. Modeling the
hydrological significance of wetland restoration scenarios. J. Environ. Manage. 133, 121134.
Melles, S.J., Benoy, G., Booty, B., Leon, L., Vanrobaeys, J., Wong, I. 2010. Scenarios to
investigate the effect of wetland position in a watershed on nutrient loadings. 2010
International Congress on Environmental Modelling and Software. Retrieved from:
http://www.iemss.org/iemss2010/index.php?n=Main.Proceedings.
Millennium Ecosystem Assessment. 2005. Ecosystem and human well-being: wetlands and
water. Synthesis. World Resouces Institude, Washington, DC.
Mitsch, W. J., Day J.W. Jr., Gilliam J.W., Groffman P.M., Hey D.L., Randall G.W., Wang N.
2001. Reducing nitrogen loading to the Gulf of Mexico from the Mississippi River Basin:
Strategies to counter a persistent ecological problem. BioScience 51: 373–388.
Mitsch, W. J., Gosselink, J. G. 2000. The value of wetlands: importance of scale and landscape
setting.Ecological Economics. 35:25–33.
Mitsch, W.J. 1994. The nonpoint source pollution control function of natural and constructed
riparian wetlands. p. 351–361. In W.J. Mitsch (ed.) Global wetlands: Old World and
New. Elsevier Sci., New York.
Mitsch, W.J., Gosselink, J.G. 1993. Wetlands, 2nd ed. John Wiley, New York.
Mitsch, W.J., J.K. Cronk, X. Wu, R.W. Nairn, and D.L. Hey. 1995. Phosphorus retention in
constructed freshwater riparian marshes. Ecol. Appl. 5:830–845.
MNFI. 2014. Vegetation circa 1800. Michigan Natural Features Inventory, Lansing, Michigan.
Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Blingner, R.L., Harmel, R.D., Veith, T.L. 2007.
Model evaluation guidelines for systematic quantification of accuracy in watershed
simulations. T. ASABE 50(3): 885-900.
Nasr, A., Bruen, M., Jordan, P., Moles, R., Kiely, G., & Byrne, P. 2007. A comparison of
SWAT, HSPF and SHETRAN/GOPC for modelling phosphorus export from three
catchments in Ireland. Water Research, 41(5), 1065-1073.
213

National Research Council. 1995. Wetlands, Characteristics and Boundaries. Washington D.C.:
National Academy Press. 307p.
Neely RK and Baker JL. 1989. Nitrogen and phosphorus dynamics and the fate of agricultural
runoff. In: van der Valk A (Ed). Northern Prairie Wetlands. Ames, IA: Iowa State
University Press. p 92–131.
Neitsch S.L., J.G. Arnold, J.R. Kiniry, and J.R. Williams. 2001. Soil and Water Assessment Tool
User's Manual, Version 2000. U.S. Department of Agriculture, Agricultural Research
Service, Temple, TX. Available at:
http://www.epa.gov/waterscience/basins/basinsv3.htm
Neitsch, S. L., J. G. Arnold, J. R. Kiniry, J. R. Srinivasan, and J. R. Williams. 2002. Soil and
Water Assessment Tool: User's Manual (version 2000) . Temple, Tex.: USDA-ARS
Grassland, Soil and Water Research Laboratory, Blackland Research Center, Texas
Agricultural Experiment Station.
Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Williams, J.R. 2005. Soil and Water Assessment Tool
theoretical documentation Version 2005. Retrieved from:
http://swatmodel.tamu.edu/media/1292/swat2005theory.pdf.
Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Williams, J.R. 2011. Soil and Water Assessment Tool
Theoretical Documentation Version 2009. Texas Water Resources Institute Technical
Report No. 406. Retrieved from: http://twri.tamu.edu/reports/2011/tr406.pdf.
Nejadhashemi, A.P., Wardynski, B.J., Munoz, J.D. 2012. Large-scale hydrologic modeling of the
Michigan and Wisconsin agricultural regions to study impacts of land use changes. T.
ASABE 55(3): 821-838.
Nieveen J.P., Jacobs C.M.J. and Jacobs A.F.G. 1998. Diurnal and seasonal variation of carbon
dioxide exchange from a former true raised bog. Global Change Biol. 4: 823–850.
Nilsson, E., Sha, L., Qian, W., Leedo, M. 2011. Constructed Wetland: Wastewater treatment.
VVAN01 Decentralized Water and Wastewater Treatment, 1-13.
NLCD. 2001. National land cover database. U.S. Geological Survey, Multi-Resolution Land
Characteristics Consortium (MRLC), Sioux Falls, SD. Available at:
www.mrlc.gov/nlcd.php. Accessed: 10 April 2014.
Novitski, R. P. 1978. “Hydrogeological characteristics of Wisconsin’s wetlands and their
influences on floods, stream flow, and sediment.” Wetland functions and values: The
state of our understanding. P. E. Greeson, J. R. C. Clark, and J. E. Clark, ed., American
Water Resources Association, Minneapolis, MN, 377-88.
Novitzki R.P., R.D. Smith, and J.D. Fretwell. 1997. Restoration, Creation, and Recovery of
Wetlands: Wetland Functions, Values, and Assessment. United States Geological Survey
214

Water Supply Paper 2425. Available on-line at
http://water.usgs.gov/nwsum/WSP2425/functions.html.
NRC. 1995. Wetlands, characteristics, and boundaries . Committee on Characterization of
Wetlands, National Academic Press, Washington D.C.
Obropta, C., Kallin, P., Mak, M., Ravit, B. 2008. Modeling Urban Wetland Hydrology for the
Restoration of a Forested Riparian Wetland Ecosystem. 5 (1), 183-198.
Omari, C.K. 1993. Social and cultural values of wetlands in Tanzania. In G.L. Kamukala & S.A.
Crafter, eds. Wetlands of Tanzania, Proceedings of a Seminar on the Wetlands of
Tanzania. Morogoro, Tanzania, 20–27 November 1991, pp. 95–102. Gland, Switzerland,
IUCN.
Park, D., Roesner, L. A. 2009. Modeling performances of detention basins with uncertainty
analysis, in: Proceedings of the World Environment and Water Resources Congress, May
17–21, 2009, Kansas City, Missouri.
Perez-Vega, S., Salmeron-Ochoa, I., Nieva-de la Hidalga, A., Sharratt, P. N. 2011. Analytical
hierarchy processes (AHP) for the selection of solvents in early stages of pharmaceutical
process development. Process Saf. Environ. 89(4), 261-267.
Persson, J., Somes, N.L.G., Wong, T.H.F. 1999. Hydraulics efficiency of constructed wetlands
and ponds. Water Sci. Technol. 40 (3), 291–300.

Peterjohn, W.T. and D.L. Correll. 1982. Nutrient dynamics in an agricultural watershed:
observations on the role of a riparian forest. Ecology 65: 1466-1475.
Potter KW. 1994. Estimating potential reduction flood benefits of restored wetlands. Water
Resource Update, 97: 34–38.
Raisin, G., Bartley, J., Croome, R. 1999. Groundwater influence on the water balance and
nutrient budget of a small natural wetland in Northeastern Victoria. Australia. Ecol. Eng.
12(1-2): 133-147.
Raisin, G.W., and D.S. Mitchell. 1995. The use of wetlands for the control of non-point source
pollution. Water Sci. Technol. 32:177–186.
Raisin, G.W., D.S. Mitchell, and R.L. Croome. 1997. The effectiveness of a small constructed
wetland in ameliorating diffuse nutrient loadings from an Australian rural catchment.
Ecol. Eng. 9:19–35.
Ramsar. 2004. Wetland: Values and functions fact sheet [online source: accessed 7/18/2012].

215

Reed, S.C. 1993. Subsurface flow constructed wetlands for wastewater treatment: A technology
assessment.
Reinelt, L.E., and R.R. Horner. 1995. Pollutant removal from stormwater runoff by palustrine
wetlands based on comprehensive budgets. Ecological Engineering 4(2): 77–97.
Renschler, C. S., & Lee, T. 2005. Spatially distributed assessment of short-and long-term
impacts of multiple best management practices in agricultural watersheds. Journal of Soil
and Water Conservation, 60(6), 446-456.
Richards, Carl, Lucinda B. Johnson, and George E. Host. 1996. Landscape scale influences on
stream habitats and biota. Canadian Journal of Fisheries and Aquatic Science 53 (suppl.
1): 295-311.
Richardson, C. and Craft C. 1993. Efficient phosphorus retention in wetlands: fact or fiction? In:
Moshiri GA (Ed). Integrated resource management & landscape modification for
environmental protection. Boca Raton, FL: CRC Press, Inc/Lewis Publishers. p 271–82.
Richardson, C.J. 1989. Freshwater wetlands: Transformers, filters, or sinks? p. 25–46. In R.R.
Sharitz and J.W. Gibbons (ed.) Freshwater wetlands and wildlife. U.S. Dep. of Energy
Office of Sci. and Tech. Info., Oak Ridge, TN.
Rossman L.A. 2005. Stormwater Management Model User’s Manual, Version
5.0.EPA/600/R‐05/040. U.S. Environmental Protection Agency, Water Supply and
Water Resources Division, National risk Management Research
laboratory, Cincinnati, OH.
Ryan, J.A., Bell, R.M., Davidson, J.M., O’Connor, G.A. 1988. Plant uptake of non-inorganic
chemicals from spills. Chemosphere. 17, 2299-2323.
Saaty, T. L. 1980. The Analytic Hierarchy Process, McGraw-Hill, New York.
Schaafsma, J.A., A.H., Baldwin, and C.A., Streb. 2000. An evaluation of constructed wetland to
treat wastewater from a dairy farm in Maryland, USA. Eco. Eng. 14:199-206.
Schreader C.P., Rouse W.R., Griffis T.J., Boudreau L.D. and Blanken P.D. 1998. Carbon dioxide
fluxes in a northern fen during a hot, dry summer. Global Biogeochem. Cycles 12: 729–
740.
Schröder, P., Navarro-Avino, J., Azaizeh, H., Goldhirsh, A.G., DiGregorio, S., Komives, T.,
Langergraber, G., Lenz, A., Maestri, E., Memon, A.R., Ranallill, A., Sebastiani, L.,
Smrcek, S., Vanek, T., Vuilleumier, S., Wissing, F. 2007. Using phytoremediation
technologies to upgrade waste water treatment in Europe. Environ. Sci. Pollut. Res. 14,
490-497.

216

Schwar, M.T. 1998. Restoration of rivers and wetlands: Advancing the State of Hydrologic
Design. Proceedings of the ASCE Wetlands Engineering River Restoration Conference,
New Jersey, p.6.
Sedell, J. R., Richey, J. E., and Swanson, F. J. 1989. ‘The River Continuum Concept: a basis for
the expected ecosystem behaviour of very large rivers?’ Can. Spec. Publ. Fish. Aquat.
Sci., 106, 49-55.
Şener, Ş., Şener, E., Nas, B., & Karagüzel, R. 2010. Combining AHP with GIS for landfill site
selection: a case study in the Lake Beyşehir catchment area (Konya, Turkey). Waste
Management, 30 (11), 2037-2046.
Shaw, S. P., and C. G. Fredine. 1956. Wetlands of the United States. U.S. Fish Wildl. Serv., Circ.
39. 67 pp. Kantrud, H. A., G. L. Krapu, and G. A. Swanson. 1989. Prairie basin wetlands
of the Dakotas: a community profile. U.S. Fish and Wildlife Service, Washington, DC,
USA. Biological Report 85(7.28).
Shoemaker L., Riverson J., Alvi K., Zhen J., Paul s., and Rafi Teresa. 2009. SUSTAIN –
SUSTAIN A Framework for Placement of Best Management Practices in
UrbanWatersheds to Protect Water Quality. National Risk Management
ResearchLaboratory. United States Environmental Protection Agency. Report:
EPA/600/R09/095.
Shoemaker, L., Riverson, J., Alvi, K., Zhen, J. X., Paul, S., Rafi, T. 2009. SUSTAIN: a
framework for placement of best management practices in urban watersheds to protect
water quality. National Risk Management Research Laboratory, Office of Research and
Development, US Environmental Protection Agency.
Available at: http://www.epa.gov/nrmrl/wswrd/wq/models/sustain/
Simpson, R.L., R.E. Good, R. Walker, and B.R. Frasco. 1983. The role of Delaware River
freshwater tidal wetlands in the retention of nutrients and heavy metals. J. Environ. Qual.
12: 41-48.
Skagen, S. K., Melcher, C. P., Haukos, D. A. 2008. Reducing sedimentation of depressional
wetlands in agricultural landscapes. Wetlands. 28, 594–604.
Skaggs, R. W., Breve, M. A., Mohammad, A. T., Parsons, J. E., Gilliam, J. W. 1995. Simulation
of drainage water quality with DRAINMOD. Irrigation and Drainage Systems, 9(3), 259277.
Smith, D. R., Ammann, A., Bartoldus, C., and Brinson, M. M. 1995. "An approach for assessing
wetland functions using hydrogeomorphic classification, reference wetlands, and
functional indices," Technical Report WRP-DE-9, U.S. Army Engineer Waterways
Experiment Station, Vicksburg, MS. NTIS No. AD A307 121.

217

Smith, D., J. Li, and D. Banting. 2005. A PCSWMM/GIS-based water balance model for the
Reesor Creek watershed. Atmospheric Research 77(1–4): 388–406.
Soler-López, L.R. 2001a. Sedimentation survey of Lago Dos Bocas, Puerto Rico,
Soler-López, L.R. 2001b. Sedimentation survey of Lago Caonillas, Puerto Rico, February 2000.
U.S. Geological Survey Water-Resources Investigations, Report 01-4043, 25 p.
Soler-López, L.R. 2001c. "Sedimentation survey results of the principal water supply reservoirs
of Puerto Rico"; in W.F. Sylva (ed.), Proceedings of the Sixth Caribbean Mayagüez,
Puerto Rico, February 22 and 23, 2001, unpaginated CD
Srinivasan, R., Ramanarayanan, T. S., Arnold, J. G., Bednarz, S. T., 1998. Large area hydrologic
modeling and assessment part II: Model application1. 91-101.
STATSGO2. 2006. U.S. General Soil Map (STATSGO2). Retrieved from:
http://soils.usda.gov/survey/geography/statsgo/.
Stevens, P.W., C.L. Montague, and K.J. Sulak. 2006. Patterns of fish use and piscivore
abundance within a reconnected saltmarsh impoundment in the northern Indian River
Lagoon. Florida. Wetlands Ecology and Management. 14:147–166.
Tetra Tech. 2009. System for Urban Stormwater Treatment and Analysis Integration (SUSTAIN)
Step by step application guide, Version 1.0. Prepared by: Tetra Tech, October 15, 2009.
Thomas, D. M., and Hanson, M. A. 1981. “Generalization of streamflow characteristics from
drainage-basin characteristics,” Water Supply Paper 1-55, U.S. Geological Survey,
Washington, DC.
Thompson, J. R., SORENSON, H. R., Gavin, H., Refsgaard, A. 2004. Application of the coupled
MIKE SHE/MIKE 11 modelling system to a lowland wet grassland in southeast England.
J. Hydrol., 293(1), 151-179.
Tiner, R.W. 1998. In Search of Swampland: A Wetland Sourcebook and Field Guide. Rutgers
University Press, New Brunswick, NJ.
Trepel, M. and L. Palmeri. 2002. Quantifying nitrogen retention in surface flow wetlands for
environmental planning at the landscape–scale. Ecological Engineering 19:127–40.
Tsihrintzis, V.A., and R. Hamid. 1998. Runoff quality prediction from small urban catchments
using SWMM. Hydrological Processes 12(2): 311–329.
Turner, K. 1991. Economics and wetland management. Ambio 20, 59–63.
U. S. Army Corps of Engineers. 1998. HEC-HMS Hydrologic Modeling System User’s Manual,
Version 1.0.
218

U.S. Environmental Protection Agency (USEPA). 1995. America's wetlands: Our vital link
between land and water. Office of Water, Office of Wetlands, Oceans and Watersheds.
EPA843-K-95-001.
U.S. Environmental Protection Agency (USEPA). 2009. SUSTAIN A Framework for Placement
of Best Management Practices in Urban Watersheds to Protect Water Quality.
PA/600/R-09/095. Office of Research and Development, Cincinnati,
OH.http://www.epa.gov/nrmrl/pubs/600r09095/600r09095.pdf
U.S. Environmental Protection Agency (USEPA). 2010. Stormwater Management Model User’s
Manual, Version 5.0. EPA/600/R-05/040. Office of Research and Development,
Cincinnati, OH. http://www.epa.gov/nrmrl/pubs/600r05040/600r05040.pdf
U.S. Environmental Protection Agency. 1993. Constructed wetlands for wastewater treatment
and wildlife habitat. Retrieved from:
http://water.epa.gov/type/wetlands/upload/2004_10_25_wetlands_ConstructedWetlandsComplete.pdf.
U.S. Environmental Protection Agency. 2009. National water quality inventory. 2004 Report to
Congress. EPA 841-R-94-001. Retrieved from:
http://water.epa.gov/lawsregs/guidance/cwa/305b/upload/2009_01_22_305b_2004report_
2004_305Breport.pdf.
U.S. Geological Service (USGS). 2004. Restoration, Creation, and Recovery of Wetlands
Wetland Functions, Values, and Assessment. Available at:
http://water.usgs.gov/nwsum/WSP2425/functions.html. Accessed 18 September 2011
USDA-NRCS. 2003. ECS 190-15 Wetland Restoration, Enhancement, Management &
Monitoring. 425 pp.
USDA-NRCS. 2003. Wetland restoration and enhancement techniques. Wetland restoration,
enhancement and management. Indiana Biology Technical Note No. 1. Available online
at: http://www.parcplace.org/images/stories/pdfs/indianatechnote1.pdf
USDA-NRCS. 2009. Chapter 3: Constructed Wetlands. National Engineering Handbook. 210VI-NEH, amend.33.Available online at:
ftp://ftp.wcc.nrcs.usda.gov/wntsc/H&H/NEHhydrology/ch3.pdf
USDA-NRCS. 2013a. Field Office Technical Guide- Statewide Conservation Practice: Typical
Installation Cost Information. Michigan USDA/NRCS State Office, East Lansing, MI.
Available at: http://efotg.sc.egov.usda.gov/treemenuFS.aspx, Assessed April 10, 2013.
USDA-NRCS. 2013b. Web soil survey. United States Department of Agriculture, Natural
Resources Conservation Service. Available online at: http://websoilsurvey.nrcs.usda.gov
(Assessed February 15, 2013).
219

USDA-NRCS. 2006. Wetland Restoration, Enhancement, or Creation, Engineering Field
Handbook Chapter 13, Part 650, pp. 3, 24, 77, 78.
USDA-NRCS. 2007. Wetland Reserve Program. Washington, D.C.: USDA Natural Resources
Conservation Service. Available at: http://policy.nrcs.usda.gov
USEPA. 2000. Constructed wetlands treatment of municipal wastewaters. U.S. Environmental
Protection Agency, Cincinnati, OH, USA, pp. 152.
USEPA. 2009. SUSTAINeA Framework for Placement of Best Management Practices in Urban
Watersheds to Protect Water Quality. Available at:
http://www.epa.gov/nrmrl/pubs/600r09095/600r09095.pdf
USEPA. 2010. Stormwater Management Model User’s Manual, Version 5.0. Available at:
http://nepis.epa.gov/Adobe/PDF/P10011XQ.pdf
Vannote, R. L., Minshall, G. W., Cummins, K. W., Sedell, J. R., and Cushing, C. E. 1980. ‘The
river continuum concept’, Can. J. Fish Aquat. Sci., 37, 130- 137.
Verhoeven, J. T. A. & T. L. Setter (2010) Agricultural use of wetlands: opportunities and
limitations. Annals of Botany, 105, 155-163.
Vought, L.B.-M., Lacoursie`re, J. 1998. Constructed wetlands for treatment of polluted waters
from agricultural fields: Swedish experiences. In: Ecological and Socio-Economic
Consequences of Land Use Changes, VII International Congress of Ecology, INTECOL
‘98, 19–25 July 1998, Florence, Italy. Invited paper at the Symposium 13.2.
Vymazal J. 1995. Constructed wetlands for wastewater treatment in the Czech Republic: state of
the art. Water Sci. Technol. 32(3), 357±365.
Vymazal, J. 2007. Removal of nutrients in various types of constructed wetlands. Sci. Total
Environ. 380(1), 48-65.
Waddington J.M. and Roulet N.T. 2000. Carbon balance of a boreal patterned peatland. Global
Change Biol. 6: 87–97.
Walton, R., R. S. Chapman, and J. E. Davis. 1996. Development and application of the wetlands
dynamic water budget model. Wetlands, 16(3): 347-357.
Wang, X., S. Shang, Q. Zhongy, L. Tingzi, A.M. Melesse, and Y. Wanhong. 2010. Simulated
Wetland Conservation–Restoration Effects on Water Quantity and Quality at Watershed
Scale, J. Environ. Manage. 91, 1511–1525.
Wang, X., Yang, W., Melesse, A.M. 2008. Using hydrologic equivalent wetland concept within
SWAT to estimate streamflow in watersheds with numerous wetlands. Trans. ASABE 51
(1), 55e72.
220

Watson, B.M., R.A. McKeown, G. Putz, and J. D. MacDonald. 2008. Modification of SWAT for
modeling streamflow from forested watersheds on the Canadian Boreal Plain. J. Environ.
Eng. Sci. 7(S1): 145– 159.
Weller, C. M., Watzin, M. C., & Wang, D. 1996. Role of wetlands in reducing phosphorus
loading to surface water in eight watersheds in the lake Champlain basin. Environmental
Management, 20(5), 731–739. doi:10.1007/BF01204144.
Weller, M.W. 1994. Freshwater marshes: Ecology and Wildlife Management. University of
Minnesota Press, Minneapolis/London.
Weltzin J.F., Pastor J., Harth C., Bridgham S.D., Updegraff K. and Chapin C.T. 2000. Response
of bog and fen plant communities to warming and water-table manipulations. Ecology
81: 3464–3478.
Wharton, C.H. 1970. The Southern River Swamp, A Multiple- Use Environment. Bureau of
Business and Economic Research, Georgia State University, Atlanta, GA.
Wheeler, B.D. and S.C. Shaw. 2000. A Wetland Framework for Impact Assessment at Statutory
Sites in Eastern England. Environment Agency Research and Development Technical
Report. W6-068/TR1.
Whigham, D.F., C. Chitterling, and B. Palmer. 1988. Impacts of freshwater wetlands on water
quality: a landscape perspective. Environ. Management. 12: 663-671.
Whigham, D.F. 1995. The role of wetlands, ponds, and shallow lakes in improving water quality.
p. 163–172. In K. Steele (ed.) Animal waste and the land–water interface. Lewis Publ.,
Boca Raton, FL.
Widener, A. S. 1995. A mathematical model of the nitrogen cycle in a constructed wetland. MS
Wolverton, B.C., R.C. McDonald, and W.R. Duffer. 1983. Microorganisms and higher plants for
waste water treatment. J. Environ. Qual. 12:236–242.
Wong, T. H., Fletcher, T. D., Duncan, H. P., Coleman, J. R., Jenkins, G. A. 2002. A model for
urban stormwater improvement conceptualization. Global Solutions for Urban Drainage,
pp.8-13.
Wong, T.H.F., Duncan, H.P., Fletcher, T.D., Jenkins, G.A. 2001. A unified approach to
modelling urban stormwater treatment. In: Proceedings of the Second South Pacific
Stormwater Conference, Auckland, New Zealand, June 2001.
Wong, T.H.F., Duncan, H.P., Fletcher, T.D., Jenkins, G.A. 2001. A unified approach to
modelling urban stormwater treatment. In: Proceedings of the Second South Pacific
Stormwater Conference, Auckland, New Zealand, June 2001.

221

Wood A. P., and G.E van Helsema. 2008. Scoping Agriculture – Wetland Interactions Towards a
Sustainable multiple-response strategy. FAO Water Reports. ISBN 978-92-5-106059-9.
Rome, Italy. 33. P 153.
Wray, R.D., Leatherman, S.P., Nicholls, R.J. 1995. Historic and future land loss for upland and
marsh islands in the Chesapeake Bay, Maryland, U.S.A. J. Coastal Res. 11(4): 1195–
1203.
Wu, K. S., and Johnston, C. A. 2008. Hydrologic comparison between a forested and a
wetland/lake dominated watershed using SWAT, Hydrol. Processes, 22, 1431– 1442,
doi:10.1002/hyp.6695.
Yang, W., Wang, X., Gabor, T.S., Boychuk, L., Badiou, P. 2008. Water quantity and quality
benefits from wetland conservation and restoration in the Broughton's Creek watershed.
Ducks Unlimited Canada. Retrieved from:
http://www.ducks.ca/conserve/research/projects/broughtons/pdf/broughtons.pdf.
Yang, W., Wang, X., Liu, Y., Gabor, S., Boychuk, L., Badiou, P. 2010. Simulated environmental
effects of wetland restoration scenarios in a typical Canadian prairie watershed. Wetlands
Ecology Manage. 18: 269–279.
Yarbro, L.A., E.J. Kuenzler, P.J. Mulholland, and R.P. Sniffen. 1984. Effects of stream
channelization on exports of nitrogen and phosphorus from North Carolina Coastal Plain
watersheds. Environ. Manag. 8: 151-160.
Yerkes, T. 2000. Nest-site characteristics and brood-habitat selection of redheads: an association
between wetland characteristics and success. Wetlands 20: 575–580.
Young, K. D., Dymond, R. L., Kibler, D. F. 2011. Development of an improved approach for
selecting storm-water best management practices. J Water Res. Plan. Manag., 137(3),
268-275.
Young, R. A., C. A. Onstad, D. D. Bosch, and W. P. Anderson. 1987. AGNPS, Agricultural
nonpoint-source pollution model: A watershed analytical tool. Conservation Research
Report No. 35. Washington, D.C.: USDA.
Young, R. A., C. A. Onstad, D. D. Bosch, and W. P. Anderson. 1989. AGNPS: A nonpointsource pollution model for evaluating agricultural watersheds. J. Soil and Water
Conservation 44(2): 168-173.
Young, R. A., C. A. Onstad, D. D. Bosch, and W. P. Anderson. 1994. Agricultural Non-Point
Source Pollution Model, Version 4.03: AGNPS User’s Guide. Morris, Minn.: USDAARS North Central Soil Conservation Research Lab.
Zedler, J. B. 2003. Wetlands at your service: reducing impacts of agriculture at the watershed
scale. Frontiers in Ecology and the Environment, 1(2), 65-72.
222