Constructed Wetlands and Sustainable Development (Gary Austin Yu Kongjian)

Constructed Wetlands and
Sustainable Development
This book explains how, with careful planning and design, the functions and perform-
ance of constructed wetlands can provide a huge range of benefits to humans and the
environment. It documents the current designs and specifications for free water surface
wetlands, horizontal and vertical subsurface flow wetlands, hybrid wetlands and bio-
retention basins; and explores how to plan, engineer, design and monitor these natural
systems.
Sections address resource management (landscape planning), technical issues
(environmental engineering and botany), recreation and physical design (landscape
architecture) and biological systems (ecology). Site and municipal scale strategies for
flood management, stormwater treatment and green infrastructure are illustrated
with case studies from the USA, Europe and China, which show how these principles
have been put into practice.
Written for upper-level students and practitioners, this highly illustrated book
provides designers with the tools they need to ensure constructed wetlands are
sustainably created and well-managed.
Gary Austin is the author of Green Infrastructure for Landscape Planning (Routledge
2014). He is a landscape architect who studied under John Lyle and taught at the
California State Polytechnic University, Pomona, USA. He has practised in the public
and private sectors and has taught landscape architecture at the University of Wash-
ington, USA, and the University of Idaho, USA. His teaching and research focus on
community revitalization, urban biological diversity and treatment of wastewater and
stormwater for water quality improvement.
Kongjian Yu is co-author of the influential book The Art of Survival (2007). He is

Visiting Professor of Landscape Architecture at Harvard University, USA, and the
principal of Turenscape, a large landscape architecture firm in China. The many con-
structed and monitored wetland projects that his firm has designed appear in this book.
He is also Professor and Dean at Beijing University, China.
This page intentionally left blank
Constructed Wetlands
and Sustainable
Development
Gary Austin and Kongjian Yu

First published 2016
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2016 Gary Austin and Kongjian Yu
The right of Gary Austin and Kongjian Yu to be identified as authors of this
work has been asserted by them in accordance with sections 77 and 78 of
the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or
utilised in any form or by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying and recording, or in
any information storage or retrieval system, without permission in writing
from the publishers.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identification and explanation
without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
Names: Austin, Gary (Gary D.), author. | Yu, Kongjian, author.
Title: Constructed wetlands and sustainable development /
Gary Austin and Kongjian Yu.
Description: Abingdon, Oxon ; New York, NY : Routledge, 2016. |
Includes bibliographical references and index.
Identifiers: LCCN 2016003285| ISBN 9781138908987
(hardback : alk. paper) | ISBN 9781138908994 (pbk. : alk. paper) |
ISBN 9781315694221 (ebook)
Subjects: LCSH: Constructed wetlands. | Water—Pollution. |
Ecological engineering.
Classification: LCC TD756.5 .A97 2016 | DDC 628.3/5—dc23
LC record available at https://lccn.loc.gov/2016003285
ISBN: 978-1-138-90898-7 (hbk)

ISBN: 978-1-138-90899-4 (pbk)
ISBN: 978-1-315-69422-1 (ebk)
Typeset in Sabon
by Florence Production Ltd, Stoodleigh, Devon, UK
Contents
List of figures vi
List of tables xiv
Acknowledgements xvi
1 Water and sustainable urban design 1
2 Characteristics of wastewater 25
3 Free water surface constructed wetlands 31
4 Horizontal subsurface flow treatment wetlands 50
5 Vertical subsurface flow constructed wetlands 70
6 Hybrid constructed wetlands 87
7 Plants in constructed wetlands 98
8 Riparian wetlands 142
9 Stormwater management and sustainable development 175
10 Increasing the sustainability of agriculture 233
11 Treatment of industrial effluent in constructed wetlands 254
Appendix A 269
Appendix B 271
Appendix C 272
Index 279
Figures
1.1 Philadelphia’s campaign to divert stormwater run-off into

infiltration areas 2
1.2 Flooding risk in three categories and outstanding habitat and
cultural features 3
1.3 Flow direction for stormwater through the three corridor types 3
1.4 The use of swales and basins in the narrowest and most urban
corridors 4
1.5 Surface drainage system in a secondary corridor 4
1.6 The widest corridors featuring the best biological diversity 5
1.7 Intersection between one of the widest corridors and an
intermediate one 5
1.8 One of the widest corridors supporting significant habitat 5
1.9 Intermediate corridor at Wulijie 6
1.10 Variety in the visual character and diverse treatment strategies
within the intermediate wetland corridors 6
1.11 The narrowest corridor is the most urban but sufficiently wide to
provide privacy and light 6
1.12 The controlling plan for a small section of the new town of Wulijie 7
1.13 Artist’s rendering of the proposed development of Wulijie 8
1.14 Primary and intermediate corridors meet in Wulijie 8
1.15 The use of spray Irrigation 10
1.16 A stormwater outfall pipe 11
1.17 Map showing precipitation trends in the United States from 1958
to 2007 12
1.18 The principal aquifers in the United States 13
1.19 Fire on the Cuyahoga River in 1952 17
1.20 The city of Roseburg, Oregon’s 350-acre restored wetland and land
application of effluent 18
1.21 Map showing the conditions of rivers and streams in the United
States based on the presence and health of biological indicators 18
1.22 Mine tailing pollution of a small stream 20
1.23 Algae caused by excess nutrients from agricultural and domestic
wastewater 20
1.24 Free water surface wetland treating stormwater in Wulijie, China 22
1.25 Horizontal subsurface flow wetland in Casper, Wyoming 23
1.26 E. coli bacteria colonies in a culture dish 23
2.1 BOD tests of wastewater samples 27
2.2 Conventional wastewater treatment plant using a large pump
powering an aeration device 29
3.1 Blue Heron Water Reclamation Facility, Titusville, Florida 31
3.2 Free water surface wetland mimicking natural wetlands 32
FIGURES vii
3.3 Inflow from the wastewater treatment plant at Blue Heron water
Reclamation Facility, Titusville, Florida 33
3.4 Wastewater inflow being aerated 33
3.5 Design of a free water surface wetland 34
3.6 Differing water depths encouraging a variety of species to flourish 36
3.7 Below ground view of a wetland cell outlet trench 37
3.8 Plan view diagram of the constructed wetland sequence at National
Cheng-Kung University 38
3.9 E L Huie Jr Constructed Treatment Wetlands, Clayton County,
Georgia 40
3.10 Free water surface constructed wetland in Albany, Oregon 41
3.11 Talking Waters Garden, Albany, Oregon 42
3.12 Deep water zones at the Talking Water Gardens, Albany, Oregon 42
3.13 Panoramic view of the treatment wetland at the Talking Water
Gardens, Oregon 43
3.14 The aeration process at the Talking Water Gardens, Albany,
Oregon 44
3.15 Wapato in the Talking Water Gardens, Albany, Oregon 46
3.16 Plant varieties at the Boone Presbyterian Church stormwater
treatment wetland, North Carolina 48
4.1 The horizontal subsurface flow wastewater treatment wetland,
Krovi, Czech Republic 51
4.2 Coarse gravel inlet trench and filter bed for a horizontal subsurface
flow wetland 52
4.3 Example of a horizontal subsurface flow wetland 54
4.4 Horizontal subsurface flow wetland under construction 57
4.5 Comparison of distribution and collection piping 58
4.6 Gravel media in horizontal subsurface flow constructed wetlands 59
4.7 Aerial view of a trench horizontal subsurface flow wetland under
construction 61
4.8 Model of linear horizontal subsurface flow wetland 62
4.9 First wedge of coarse stone 63
4.10 Horizontal subsurface flow trench capped with a layer of fine gravel
and a top layer of sand 63
4.11 Horizontal subsurface flow wetland and elevated effluent storage
tank 64
5.1 Vertical subsurface flow wetland, Silkerode, Germany 71
5.2 Vertical subsurface flow wetland based on Danish standards 72
5.3 Vaults and piping system for a vertical subsurface flow wetland 72
5.4 Vertical flow subsurface wetland during testing and construction 73
5.5 Two-stage French system used for vertical flow treatment wetlands 74
5.6 Achieving an aerobic environment for bacteria 75
5.7 Vertical subsurface flow wetland in Jiangxiang, China 77
5.8 Liner, drainage piping and filter media for a vertical subsurface
flow wetland 78
5.9 Vertical flow subsurface wetland planted with yellow flag
(Iris pseudacorus) 79
5.10 Stand of Papyrus is in a vertical flow treatment wetland in Peru 80
viii FIGURES
5.11 Section of two-stage vertical subsurface flow wetland 81

5.12 Installation of waterproof membrane and drainage system of a
vertical subsurface flow cell 85
5.13 Vertical flow wetland, Shengyang, China 85
6.1 Experimental wastewater facility near Carrión de los Céspedes,
Spain 89
6.2 Vertical subsurface flow wetland stage with Phragmites 90
6.3 Second stage (horizontal subsurface flow) of the hybrid treatment 91
6.4 Free water treatment stage covered by a variety of plant species 92
6.5 Storage tank 93
6.6 Hybrid wetland, Hennersdorf, Germany 94
6.7 Hybrid wetland and central park, Koh Phi Phi, Thailand, 95
6.8 Aesthetics of wetland and ecotourism resort, Lake Yangcheng 96
6.9 Three stages of vertical subsurface flow wetland, Shenyang, China 96
6.10 Treated effluent from the three vertical subsurface flow beds 97
7.1 Example of planting design by Turenscape, Tianjin, China 99
7.2 Free water surface wetland treats polluted river water and
stormwater and creates biodiversity niches 99
7.3 Aggressive plant species dominating wetland 100
7.4 Plant diversity and aesthetic quality, Houtan Park, Shanghai 101
7.5 Full range of wetland plants 101
7.6 Plant zones in a constructed wetland 101
7.7 Plan representing the installed planting design for a stormwater
wetland 103
7.8 Typha latifolia (Broadleaf cattail) 104
7.9 Typha angustifolia 104
7.10 Phragmites australis harvesting 106
7.11 Phragmites australis seed plumes 107
7.12 Example of species attracted to wetlands by Phragmites australis 108
7.13 Phragmites australis in dense single species stands 108
7.14 Juncus effusus 109
7.15 Side view of the Juncus effusus inflorescence 109
7.16 Eleocharis quadrangulata (Squareside Spikerush) 110
7.17 Eleocharis acicularis (Needle Spikerush) 111
7.18 Eleocharis palustris (Common Spikerush) 111
7.19 Eleocharis sphacelata (Tall Spikerush) 112
7.20 Eleocharis dulcis (Chinese Water Chestnut) 113
7.21 Phalaris arundinacea 113
7.22 Phalaris arundinacea as a hybrid 113
7.23 Scirpus californicus 114
7.24 Scirpus validus (River Clubrush) 115
7.25 Scirpus maritimus 116
7.26 Carex riparia 117
7.27 Carex nebrascensis (Nebraska Sedge) 117
7.28 Carex stipata (Awl-Fruit Sedge) 117
7.29 Iris pseudacorus (Yellow Flag Iris) 119
7.30 Sagittaria latifolia (Common Arrowhead) 120
7.31 Sagittaria latifolia 120
FIGURES ix
7.32 Pontederia cordata (Pickerel Weed) 121

7.33 Caltha palustris (Marsh Marigold) 122
7.34 Two Canna hybrid varieties within a shopping area wetland 123
7.35 Canna indica 123
7.36 Canna indica’s long fibrous roots 123
7.37 Canna flaccida 124
7.38 Canna flaccida in Florida 124
7.39 Cyperus papyrus 125
7.40 A Laysan Albatross nestled in Cyperus involucratus on Sand Island,
Midway Atoll 125
7.41 New Cyperus papyrus shoots being removed to be replanted 126
7.42 Arundo donax 126
7.43 Thalia dealbata 127
7.44 Thalia geniculata 128
7.45 Zantedeschia aethiopica 129
7.46 Heliconia psittacorum 129
7.47 Acorus gramineus (Japanese Sweet Flag) 129
7.48 Agapanthus 130
7.49 The exotic flowers of the bird of paradise 130
7.50 Watsonia borbonica 131
7.51 Hymenocallis littoralis 131
7.52 Bacopa Monnieri (Water Hyssop) 132
7.53 Colocasia esculenta 133
7.54 Eichhornia crassipes 134
7.55 Water hyacinth flower and foliage 134
7.56 Pistia stratiotes (Water Lettuce) 135
7.57 Lemna minor (Duckweed) 135
7.58 High plant diversity with species organized in masses and according
to water depth and frequency of inundation 137
8.1 The Sanlihe River restoration project 143
8.2 The Sanlihe River restoration project, restored river 143
8.3 Proposed changes to the Arroyo Secco River Channel 144
8.4 The Isar River, Munich, Germany 145
8.5 Natural riparian vegetation 146
8.6 Riparian wetlands along and within rivers and streams 147
8.7 Wetlands along the Kobuk River, Alaska 147
8.8 Efforts are underway to restore portions of the Los Angeles
River 148
8.9 Constructed wetland along the Tanghe River 149
8.10 Huangpu River upstream from Shanghai 149
8.11 Coal-fired energy and industrial plants along the Huangpu River 150
8.12 Aerial view of Houtan Park along the bank of the Huangpu River
near the centre of Shanghai 151
8.13 Houtan Park along the bank of the Huangpu River near the centre
of Shanghai 152
8.14 The industrial riverbank of the Huangpu River at the proposed
location of Houtan Park 152
8.15 The Chinese Pavilion at the 2010 Shanghai Exposition 153
x FIGURES
8.16 The main axis of the Shanghai Exposition with pools filled with
treated river water 153
8.17 Suspended solids in the water in the first and the last stage of the
wetland 154
8.18 Plan view of the Houtan Park constructed wetland 155
8.19 Broad range of submerged, floating leaf and emergent plant species
that contribute to the treatment of the polluted river water 155
8.20 The original condition of the Houtan park site 157
8.21 Terracing, concrete boardwalk platforms and the division of the
wetland bed into planting strips 157
8.22 The boardwalk and terrace seating alongside wetland vegetation
bands and upland terraces from the opposite direction 158
8.23 Plan showcasing the arrangement of plant species in bands
perpendicular to water flow 158
8.24 The 15 wetland compartments in treatment sequence 159
8.25 Plan view of the initial stage of the wetland 159
8.26 Initial stage of the wetland during construction 160
8.27 Masonry wall with cut stone veneer, section 160
8.28 The established wetland, Houtan park site 161
8.29 Section through wetland, deck and steel sculpture created from
waste material from the existing site 161
8.30 Park terraces just before planting 162
8.31 Section through planted terraces 162
8.32 Wetland valley during construction 162
8.33 Wetland valley completed landscape 163
8.34 Early rendering of the proposal for Houtan Park illustrating the
multifunctional character of the park 163
8.35 Bamboo decking follows a vibrant fiberglass bench between two
community agriculture fields 164
8.36 Compressed spaces between verdant strands of reeds and grasses 164
8.37 Panels punctuate the landscape and frame views of the city 165
8.38 Shade structures supported by panels made from re-purposed steel
create intimate spaces along the river 165
8.39 Sequence of two- and three-dimensional elements fashioned from
the steel plates 165
8.40 Plan view of the steel elements 166
8.41 Steel structure and paving along the boardwalk 166
8.42 Steel paving and bench 166
8.43 The bamboo boardwalk places the user at the water surface 167
8.44 Fishing docks and terraces connected by steps that serve as
seating 168
8.45 This is the last pool before the treated water returns to the river 169
8.46 Plan view of the last pool and the decks below the pergola 169
8.47 The diverse planting at Houtan Park 171
8.48 The terraced fields at Houtan Park 171
8.49 Example of textures and colours 172
8.50 The Sanlihe River Ecological Corridor 172
8.51 The Tanghe River red ribbon bench and boardwalk 173
FIGURES xi
8.52 Aerial perspective of the Tanghe River project 173

8.53 The Isar River restoration near Munich 173
9.1 The Qunli National Urban Wetland Park 175
9.2 The affects of human changes on increased stormwater run-off 177
9.3 The bio-retention swale at Olympic College Parking Garden in
Bremerton, Washington, USA 178
9.4 Stapleton community stormwater management 179
9.5 Map showing and defining the four rainstorm types in the USA 180
9.6 Portion of an isopluvial map showing rainfall depth 181
9.7 Intensity–Duration–Frequency curve for a region in Delaware,
USA 181
9.8 Map showing TR-55 analysis 184
9.9 The urban and suburban watersheds in the study of water quality
in streams near Atlanta, Georgia 189
9.10 Decomposition of algae, as seen in this California urban stream 191
9.11 Suburban context of JEL Wade Park wetland 193
9.12 JEL Wade Park wetland 193
9.13 Plan view of the design for JEL Wade Park and wetland 194
9.14 JEL Wade wetland sedimentation basin and its outlet weirs 195
9.15a Plan view of stormwater management and treatment cells that
operate in parallel 198
9.15b Perspective view of stormwater management and treatment cells 198
9.16 Perspective view of the subsurface flow gravel wetland showing
two treatment cells in series 199
9.17 Three-basin subsurface gravel wetland at the University of
New Hampshire 200
9.18 Sedimentation basin of the subsurface gravel wetland at the
University of New Hampshire 200
9.19 Schematic perspective of the piping system for the gravel wetland 201
9.20 Section through the gravel wetland 201
9.21 Subsurface gravel wetland at Greenland, New Hampshire 202
9.22 Bio-retention basin at Tabor School in Portland, Oregon 205
9.23 Bio-retention basin preceded by a vegetated swale to reduce
sediment and heavy metals 207
9.24 Underground view of a bio-retention basin with a flooded sub-basin
(internal water storage) 208
9.25 Plants being treated formally in bio-retention basin, Portland,
Oregon 210
9.26 Hal Marshall bio-retention basin with spring growth of vegetation 211
9.27 The Street Edge Alternative project in Seattle 214
9.28 Courtyard designed to receive stormwater in a park in Tianjin,
China 216
9.29 Oregon City bio-retention basin 216
9.30 Aerial view of the context for the stormwater park Qunli, China 217
9.31 Aerial view of site Qunli, China 218
9.32 Land-use plan of the Qunli New Town 218
9.33 Plan view of the design for the stormwater park Qunli, China 219
9.34 Transfer process of a bio-retention basin 220
xii FIGURES
9.35 Elevated walkways and pavilions shown in Qunli, China 220

9.36 One of the two viewing towers in the Qunli wetland park 221
9.37 Sedimentation basins at work as well as the mounds forested with
birch trees 222
9.38 Birch groves (Betula pendula) on hillocks 222
9.39 Variety of species planted at Qunli wetland park 222
9.40 Emergency spillways between the sedimentation and treatment
basins 223
9.41 Sediment and velocity control on-site 224
9.42 Sedimentation basins connect to treatment pools and then to the
core wetland 224
9.43 Example of water from the nearby river supplementing the wetland
during the dry season 225
9.44 Siberian iris blooming in May 225
9.45 Texture and colour of the wetland vegetation 225
9.46 The Qunli wetland and park 226
9.47 Restored hydrology supports low maintenance cost and provides
seasonal diversity 227
9.48 Example of spaces within the Wetland for small groups to sit 227
9.49 Pavilion at Qunli wetland park 227
9.50 Pavilion made of bamboo screen 228
9.51 Winter scene of the pavilion made from cord wood 228
9.52 The high production of biomass creates a highly diverse ecosystem 228
9.53 Variety of user amenities as well as a dramatic seasonal display of
perennial flowers 229
10.1 Constructed wetland at a field edge that receives agricultural
drainage water 234
10.2 A field-edge wetland along the Coldwater River 235
10.3 Biologist Charles Bryant and Richard Lizotte measuring water
quality in the linear wetland 236
10.4 The Red River Research Station constructed wetland 238
10.5 Phytoplankton bloom in the Bay of Biscay, France 239
10.6 Distribution and size of dead zones along with ocean carbon
content and human population density 240
10.7 Transparent perspective of a denitrifying wetland 241
10.8 Dairy cattle significantly increase the phosphorus load in the
Town Brook, New York watershed 242
10.9 Stormwater drainage from a feedlot pen overflows into a basin
where sediments can settle 243
10.10 Iowa State University Beef Nutrition Farm 243
10.11 Sheet flow of stormwater across the feedlot pen is collected in a
swale to allow solids to settle out 245
10.12 Diagram of the treatment sequence for a Vermont dairy 245
10.13 Manure composting facility 248
10.14 Will Allen illustrates that tilapia production can be combined with
plant production for a holistic approach 249
10.15 Catfish are inspected in a fish farm at the Warmwater Aquaculture
Research Unit in Stoneville, Mississippi 249
FIGURES xiii
10.16 Trout farm situated to allow water routed from streams to flow
through 252
11.1 Aerial view of a hybrid wetland treating industrial pollution within
the golf course in Casper, Wyoming 255
11.2 This groundwater remediation project in Casper is embedded
within a Robert Trent Jones golf course 256
11.3 The horizontal subsurface flow wetland is the last treatment step
in the hydrocarbon removal project in Casper, Wyoming 256
11.4 Mandan refinery sequence of ponds and swales 257
11.5 Wastewater treatment wetland in Oman 258
11.6 Oman oil recovery and wastewater remediation 258
11.7 Artificial wetland in Oman 259
11.8 The treatment wetland during construction 262
11.9 View toward the northeast of the wetland showing the first stage of
the treatment wetland 262
11.10 Horizontal subsurface wetland featuring Cattail (Typha) 263
11.11 Full-scale wetland treatment of effluent from a coal-burning power
plant in Kansas 265
11.12 Section of the vertical subsurface up-flow wetland 266
A.C.1 Composite CN with connected impervious area 274
A.C.2 Composite CN with unconnected impervious areas and total
impervious area less than 30 per cent 275
A.C.3 Soil texture classification 276
Tables
2.1 Characteristics of raw sewage in the United States 26

2.2 Typical septic tank effluent concentrations 30
3.1 Concentrations and removal of contaminants in a free water surface
treatment wetland 39
3.2 Water quality standards and performance of the E L Huie Treatment
Wetlands 40
3.3 Official visitors in the Talking Water Gardens 44
3.4 Ammonia and nitrogen removal 45
3.5 Bird species photographed at the Talking Waters Gardens 46
4.1 Common substrates 60
4.2 Residential horizontal subsurface flow wetland, Zitenice, Czech
Republic 64
4.3 Treatment efficiency of Horizontal Flow Constructed Wetland at
Spálené Poŕı́čé, Czech Republic 65
5.1 Mean influent and effluent from the two-stage French system
treatment wetland system 82
6.1 Contaminant average concentration by treatment stage 91
6.2 Mean values of emerging organic contaminants in the hybrid
wetland stages 93
8.1 Huangpu River water quality 151
8.2 Project calendar 154
8.3 Chinese water quality standards 156
9.1 Recommended run-off coefficients for the rational method 183
9.2 Runoff curve numbers for urban areas 185
9.3 Stormwater run-off pollution concentration and land use 186
9.4 Levels of parking lot contaminants after a 28-day antecedent dry
period 187
9.5 Mean contaminant concentration for six roads in Texas 188
9.6 Pollutant mean concentrations entering and exiting the wetland 196
9.7 Horizontal subsurface flow gravel wetland pollution removal
performance 204
9.8 Hal Marshall bio-retention basin pollution removal performance 212
9.9 Net mass removal of pollutants from sequential urban run-off
storms 213
9.10 Reduced curve number and depth of rainwater to infiltrate 215
10.1 Sediment, nutrient and pesticide concentrations in the wetland at
100 m from the inlet 237
10.2 Feedlot wetland performance 244
11.1 Summary of tannery effluent treatment performance at the hydraulic
loading rate of 6 cm per day and a hydraulic residency time of
7 days 261
TABLES xv
11.2 Hybrid wetland and the removal of metals 265

A.A.1 Plants by continent 269
A.B.1 Plant species for Houtan Park, Shanghai China 271
A.B.2 Aquatic species for Houtan Park, Shanghai China 271
A.C.1 Run-off curve numbers for cultivated agricultural lands 272
A.C.2 Run-off curve numbers for other agricultural lands 273
A.C.3 Run-off curve numbers for arid and semiarid rangelands 273
A.C.4 Run-off depth for selected CNs and rainfall amounts 274
A.C.5 Criteria for assignment of hydrologic soil groups (water-impermeable
layer less than 100 cm deep) 275
A.C.6 Criteria for assignment of hydrologic soil groups (water-impermeable
layer greater than 100 cm deep) 275
A.C.7 Soil particle sizes 276
A.C.8 Soil infiltration rates 276
A.C.9 Plant species within the JEL Wade Wetland, Wilmington, North
Carolina 277
A.C.10 Bio-retention basin plants – Pacific Northwest, USA 278
Acknowledgements
I’d like to thank the many people that have made this book possible. I’m grateful to
Kongjian Yu for generously providing images, drawings and documentation of his
many amazing landscape architectural projects in China. They are truly courageous
and inspirational. I want to thank him for his kindness during my stay in China. I’m
also grateful for the support of Wei Li at Turenscape. My visits to the many projects
by Kongjian Yu and his firm, Turenscape, were made possible by Zhenyu Liu.
Zhenyu’s efforts in arranging my travel, his companionship and discussions of the
projects on-site enriched my experience and this book. I am also grateful to Zhenyu
for the three dimensional models used to illustrate the book. Stephen Drown and my
wife, Leslie, always provide support and patiently endure the messy and time-
consuming writing process. The reviewers, editors and staff at Routledge improve and
refine my efforts with great skill. I thank them for their understanding and patience
when my health took a turn for the worse and delayed completion of the manuscript.
Many individuals in the Creative Commons have shared images that illustrate the text.
Thank you for your generosity. Other images and technical information that have
made the book better and more comprehensive came from Hans Brix, Jan Vymazal,
Michael Blumberg, Cristina Avila and Joan Garcia. I have relied on their careful and
important research, and I thank them for the images and information that they
generously shared. Finally, I’d like to acknowledge the great research and dis-
semination programmes at the Biological and Agricultural Engineering program of
North Carolina State University and the University of New Hampshire Storm-
water Center (Tom Ballestero and Robert Roseen) and the Centre for New Water
Technologies in Spain.
Gary Austin, 2016

1
Water and sustainable
urban design
Introduction
This book addresses the intersection of development (both rural and urban), and the
supply and quality of water required to sustain it. Sustainable development must include
water quantity and quality as indicators applied to both human and ecosystem health.
Furthermore, these indicators must be considered across the full range of human land
uses, including urban, agricultural and industrial. In the planning for new urban centres,
green infrastructure exhibits its potential as a continuous network of urban and natural
spaces connected by ecological corridors. These linear and nodal green spaces structure
the residential and commercial uses of the city while contributing many ecosystem
services, such as pollution mitigation and provision of recreational open space. When
a green infrastructure network is combined with high-density development, then com-
pact, high-quality human environments are created (Austin, 2014).
Municipal master planning

Establishing sustainable water use is more readily attainable when water supply and
treatment infrastructure are incorporated into master planning. Retrofitting existing
cities to achieve more sustainable use and disposal of water is possible, but requires
intensive, coordinated efforts by government agencies and citizens through years of
consistent effort. The city of Philadelphia, United States, is an example of a city with
an aging infrastructure engaged in this difficult but economically advantageous process
today. The $1.2-billion-dollar plan will require 25 years to complete, but will save
the city $5.6 billion by eliminating the need for new conventional wastewater treatment
plants, upgrades to existing plants and conventional stormwater infrastructure. This
effort includes constructed wetlands and other measures to convert one-third of the
city’s impervious surfaces to pervious conditions to reduce stormwater contributions
to the city’s combined stormwater and sanitary sewer infrastructure (Figure 1.1).
Ultimately, the amount of stormwater entering Philadelphia’s waterways will be
reduced by 85 per cent (Philadelphia Water Department, 2015).
An example of green infrastructure master planning is in the city of Wuhan in
the Hubei province of central China. Wuhan has a rapidly expanding population of
more than 10 million people. The city receives about 47 inches (1,200 mm) of rain-
fall per year, much of it during the hot summer months. In east Wuhan, an area of
2 WATER AND SUSTAINABLE URBAN DESIGN
Figure 1.1
The city of Philadelphia
is engaged in a citywide
campaign to divert stormwater
run-off into infiltration areas
rather than into the combined
stormwater and sanitary sewer
system.
Photo: Philadelphia Water
Department. www.flickr.com/
photos/philadelphiawater.
6.2 square miles (10 km2) has been designated for a new town named Wulijie. The
gently rolling topography of the new town will be the home of 100,000 residents.
The existing valleys and ponds will form the basis of a surface stormwater collection
and treatment network. Retaining and treating the increase in run-off created by urban
buildings and paving are important to preserve the high water quality of Liang Zhi
Hu Lake, which is downstream to the east (Saunders and Yu, 2012).
Figure 1.2 shows two sensitivity maps for the proposed development area. One
studies the hydrology and the risk of flooding, while the second considers the habitat
value and cultural sites. A composite of these plans led the planners to the configura-
tion of the residential and commercial development blocks as well as the transportation
network. Public transportation, pedestrian and bicycle systems were an integral part
of the planning from the outset.
The development plan is structured by three types of corridors. In Figure 1.3, the
three corridor types are distinguished by width and clearly establish the form and scale
of the new town districts. The orange line indicates the major vehicular transportation
routes.
The hierarchy of corridors specified in the master plan will accommodate
increasingly greater stormwater run-off volumes. The tertiary corridors (Figure 1.4)
are 50–100 feet wide (20–30 m); they receive stormwater from the development parcels
and overflow to the larger corridors. The secondary corridors (Figures 1.5 and 1.9)
are 200–300 feet wide (60–90 m); they subdivide the development blocks and deliver
moderate stormwater flows to the major water corridors. The widest corridors (Figures
1.6, 1.7 and 1.8) are 400–500 feet (120–150 m), which define the perimeter and major
development sections.
The width of the corridors (Figure 1.10) planned at Wulijie exceeds the minimum
width recommended by ecological research and green infrastructure planning (Austin,
2014), which suggests minimum widths and lengths for ecological corridors. The
WATER AND SUSTAINABLE URBAN DESIGN 3
Figure 1.2 The plan on the left identifies flooding risk in three categories, while the image on the right
shows outstanding habitat and cultural features.
Image: Kongjian Yu, 2014.
Figure 1.3 This plan shows the flow direction for stormwater through the three corridor types.
It is clear that this is a surface water management approach to stormwater. This approach allows
ancillary benefits to be attached. These include urban habitat, recreation, open space and natural
beauty. The resulting installation and maintenance costs are low compared to a catch basin and pipe
system, but the advantages related to the environment and quality of life are remarkable.
Figure 1.4 The narrowest and the most urban corridors are surrounded by high-rise buildings.
However, planted swales and basins collect and treat stormwater. A – bioswale; B – use node;
C – pedestrian walk; D – infiltration basin; E – stormwater catchment; F – bioswale; G – pedestrian
walk; H – bioswale.
Figure 1.5 This section through a secondary corridor (200–300 feet wide) illustrates the great
attention to landform grading required to implement the surface drainage system and create
treatment areas where non-point source pollution can be mitigated through filtration, sedimentation
and biological processes. A – road; B – bioswale; C – bikeway; D – terrace; E – waterside path;
F – infiltration wetland.
recommendations vary according to the context, but in a low-density residential district,

the minimum recommended width is about 33–66 feet (10–20 m) and their length
should be limited to 3,270–6,335 feet (1,000–2,000 m) before a habitat patch is
provided. At the rural scale, the recommended maximum length of 1,300–6,335 feet
(400–1,000 m) is paired with a minimum corridor width of 66–164 feet (20–50 m).
In order to sustain urban biological diversity, the ecological corridors should connect
to habitat patches having a minimum area of 1.2–12.4 acres (0.5–5 hectares). At the
outskirts of the community or in rural settings, the habitat patches should have a
maximum area of 124 acres (50 hectares) and spaced 6,335 feet apart (2,000 m) (Kubei,
1996).
At Wulijie, the wide corridors are appropriate for high-density development. The
patches of habitat are provided at the corridor intersections in the most highly
developed areas or as nodes attached to the perimeter corridors. The corridors become
narrow as development intensifies, as shown in Figures 1.7 and 1.11.
Both the public landscape and the development parcels have development controls
to assure effective implementation. Figure 1.12 shows a small portion of the controlling
plan for development. These controls include build-to lines and links to the develop-
ment requirements including standards for the public landscape. This plan most clearly
shows the tertiary corridors separating the development parcels.
Figure 1.6 The widest corridors (up to 500 feet wide) feature the best biological diversity, but
recreation and stormwater storage and treatment are equally important. The intersections of the
widest corridors create larger habitat patches.
Figure 1.7
This image shows an
intersection between one of
the widest corridors and an
intermediate one. It is
remarkably consistent
with the character imagined
in the drawing in Figure 1.6.
Photo: Kongjian Yu, 2015.
Figure 1.8
This image of one of the widest
corridors illustrates that they
support significant habitat, as
upland species of shrubs and
trees are adjacent to extensive
wetlands. The presence of
such expansive open space
within high-density urban
development is rare.
Figure 1.9 This is an intermediate corridor at Wulijie. The diversity and beauty of the planting design
is inspiring, but this is also a working landscape. In the foreground, a basin captures sediment in the
first step of water quality improvement.
Figure 1.10
This image illustrates that there
is great variety in the visual
character coupled with diverse
treatment strategies within the
intermediate wetland corridors.
Figure 1.11
The narrowest corridor is the
most urban in character but
sufficiently wide to provide
space between the buildings
for light and privacy. The width
still accommodates the
landscape infrastructure.
Figure 1.12 The controlling plan for a small section of the new town of Wulijie.
Material selection, architecture featuring passive and active solar technology,

public transportation options and other factors must be added to the consideration
of stormwater, open space, recreation and habitat in order to build a city with minimum
adverse environmental impact. Indeed, a high-speed train station and other features
are included in the plans for Wulijie. The first phase of the new town is complete.
As other sections are completed and the project matures, we will have a new model
of sustainable development to monitor and learn from. The images of the planning
documents, conceptual perspectives and, most importantly, the installed landscape
shown here indicate that the result fulfils the promise of the artist’s rendering in Figures
1.13 and 1.14.
Figure 1.13 The artist’s rendering of the proposed development shows the major ecological corridor
in the centre and to the right in this night-time scene. The secondary corridors establish smaller
districts, while the narrowest corridors separate the buildings.
Figure 1.14
This image shows where the
primary and intermediate
corridors meet. Consistent with
the planning and design
intensions, the narrower
corridors capture and treat
stormwater run-off. During
large storms, the narrow
corridors conduct floodwater to
wider corridors in the network.
Boardwalk and other pathways
invite the urban dweller to
explore the naturalistic
network.
Water resource planning

The allocation of water resources often occurs at the scale of the state or province
rather than at the municipal scale, such as in Wulijie. The quantity of water allocated
to users in the United States is regulated by state governments rather than by a national
policy or regulation. However, the federal government participates by way of the
development of water resources and distribution through dam construction and
irrigation supply programmes, for example. Federal mandates, such as the Endangered
Species Act, also establish a minimum allocation for aquatic ecosystem health. With
the major exception of irrigation, most water allocations are non-consumptive. Water
is used and returned to surface water, although at a lower quality, generally. In this
way, freshwater is reused.
Constructed wetlands play a limited role in increasing the quantity or supply of
potable water, but they can have a significant impact on the quality of water available
for reuse, especially reuse for non-potable purposes. The quality of water returned to
streams, rivers and lakes is critical to the healthy function of the ecosystem and the
ecosystem benefits that accrue to humans. Constructed wetlands can play an important
role in achieving water quality that fosters ecosystem and human health.
Water quantity
There are recent positive trends in water consumption, which show a 13 per cent
decrease in water use between 2005 and 2010 in the United States. Self-supplied
industrial withdrawals of water were 15.9 billion gallons per day in 2010, which is
a dramatic decline from the peak use of 47 billion gallons per day in 1970 (Maupin
et al., 2014). The national government assesses water use every 5 years and the next
dissemination of data is scheduled for 2016. It is useful to know how the water use
was distributed in order to devise a plan to efficiently allocate, recycle, treat and reuse
existing water resources. In the United States, legally binding water rights allocate
almost all of the surface water. Agriculture and industry use the great majority of
surface water, while municipalities depend much more on groundwater for potable
supplies or water that can be treated to potable standards.
An estimated 355 billion gallons per day was used in the United States in 2010,
although 48.3 billion gallons per day of saline water was used primarily as industrial
cooling water. Of the 306 billion gallons per day of freshwater used, 230 billion gallons
were from surface water while 76 billion gallons were withdrawn from groundwater
(Maupin et al., 2014).
Thermoelectric power and agricultural industries

While thermoelectric power generators and agricultural irrigators are the two greatest
users of water, both showed a decrease in consumption compared to 2005. Thermo-
electric power generation industry used 161 billion gallons per day and irrigation used
115 billion gallons per day (a 20 per cent and 9 per cent drop, respectively, compared
to 2005). Nevertheless, thermoelectric power and irrigation each used 38 per cent
of the water, totalling to 76 per cent of all freshwater withdrawals. Most of the
irrigation water (65.9 billion gallons per day, 57 per cent) was derived from surface
water while the remainder (49.5 billion gallons per day) came from groundwater
Figure 1.15
Spray irrigation is intermediate
in efficiency between flood and
drip irrigation. Shifting from
flood to spray or from spray to
drip irrigation has the potential
to free significant volumes of
water for other uses.
Photo: Gary Austin, 2015.
(Maupin et al., 2014). The enormous volumes of water used by these sectors and the
high percentage of total water use suggest that explorations of increased water use
efficiencies, recycling and treatment measures in these sectors could yield significant
water resources that could be used elsewhere. For example, flow-through cooling
systems remain the most predominant cooling method for coal-fired power plants,
withdrawing about 55 billion gallons of water per day (Diehl and Harris, 2014). The
conservation potential is clear from an understanding that water use produced from
each kilowatt-hour of energy ranges from 75 gallons per kilowatt-hour to 0.4 gallons
per kilowatt-hour based on flow-through, recirculation or dry-cooling systems.
In 2010, about 62,400,000 acres of land was irrigated, of which 58 per cent was
spray irrigation. The most wasteful irrigation method, flood irrigation, accounted for
26,200,000 acres (42 per cent). Again significant water resources can be conserved
through a shift in irrigation technology. A positive irrigation trend is the decrease in
the depth of agricultural irrigation, from more than 4 feet to about 2.1 feet in 2010.
In contrast, the domestic use of water accounts for only 1 per cent of all water
withdrawals. Therefore, efforts to reduce use or recycle and reuse water are important
where groundwater supplies are limited and in arid areas where water supply is
imported from another region (Maupin et al., 2014).
Water use in other economic sectors also declined with the exception of mining
(5.32 billion gallons per day) and aquaculture (9.42 billion gallons per day). Most of
aquaculture’s surface water withdrawals occurred at facilities that operated flow-
through raceways. Water used for livestock was 2 billion gallons per day, of which
60 per cent was from groundwater (Maupin et al., 2014).
Urban stormwater run-off quantity

A considerable amount of rainwater that would normally infiltrate into the soil to
recharge aquifers or supply streams and lakes during the dry season now flows off
pavement and roofs. The greater volume of run-off also flows away at a greater velocity.
These changes in infiltration and run-off generate great volumes of polluted run-off
by cities (Figure 1.16). For example, Denver, Colorado, covers 153 square miles and
generates over 2.6 million gallons of water from a 1-inch rainstorm. Most of this
amount runs off into streams and rivers through the storm sewer piping system (Sipes,
2010). A certain amount of this stormwater is detained in stormwater basins, but these
are rapidly drained. Very few of the basins are designed to retain significant amounts
of stormwater run-off. Even fewer are designed to infiltrate captured run-off to
recharge groundwater supplies or to maintain base flow (underground discharge) of
water to sustain local streams through the dry season. These circumstances exist today
despite current construction wetland technology to establish landscapes that retain
stormwater run-off and cause it to infiltrate.
Climate change
National governments need to urgently mitigate the already occurring impacts of
climate change. Even more urgently, we must put adaptation plans into effect to
respond to the worsening conditions that we can no longer avoid during the next few
decades. Increased stormwater flows and periods of drought will occur but in an
irregular pattern across the globe (Figure 1.17). We will experience more rainfall during
the most intense storms. This trend is already evident, as the amount of rainfall during
the largest storms (1 per cent) has increased by nearly 20 per cent during the last 50
years. More precipitation is already falling as rain rather than as snow, causing
increased stream flow in the wet months and decreased flow during the dry months
for rivers that have headwaters in the mountains (US Environmental Protection
Agency, 2015). The use of constructed wetlands designed to include infiltration is one
Figure 1.16
The size of stormwater outfall
pipes suggests the tremendous
volumes of, often untreated,
water that discharge to
waterways.
Photo: Darkday. License: CC-BY-
SA-3.0.
Figure 1.17 The impact of climate change on precipitation will have different impacts in regions of
the United States and even within individual states. This map shows the trends from 1958 to 2007.
The hatched areas identify significant trends.
Image: US Global Change Research Program. License: government work product.
www3.epa.gov/climatechange/images/impacts-adaptation/USdrought-large.jpg.
way of adapting to climate change, as this will increase base flow to rivers and streams
during the dry season. The changes in surface water have other economic and
ecosystem impacts that must be mitigated. Increased volumes of stormwater run-off
in urban areas have significant implications for the existing urban stormwater
management infrastructure. For example, many cities are making adjustments to their
stormwater detention and retention standards to prevent future flooding caused by
climate change.
Climate change impact studies for Europe indicate regional alterations in river
flow, with the greatest impacts in the Mediterranean zone (due to reduced precipita-
tion) and in the northern forest areas due to reduced snowpack, faster snowmelt and
subsequent flood risk. In the Mediterranean areas, reduced river flows are anticipated
in all months by 2050. The reduced river flow will expose species to increased extinc-
tion pressure. In addition, lower flows may increase water allocation conflicts, since
existing water withdrawal for irrigation and other human uses may be reduced to
assure species preservation. For example, there will be an increased need to direct
flows to floodplains to sustain existing ecosystems (Schneider et al., 2013).
Groundwater
Some categories of water are particularly precious, such as the pure sources of
drinking water. Groundwater sources of drinking water should be protected from
contamination by disease-causing organisms or toxic substances. Again, constructed

wetlands can be used as a sustainable development measure to capture polluted water
and reduce the concentration of contaminants before the water infiltrates into
groundwater reservoirs.
Figure 1.18 shows a map of the principal aquifers in the United States. Where
these are the sole source of potable water, the quantity and quality of the water
contained in the aquifer are critical to the survival of the cities, towns and people that
depend on them. Generally, municipalities must share groundwater with agriculture
and industrial users.
Aquifers are composed of sands, gravel and porous stone layers that are often
hundreds of feet below the surface but in other instances are near the surface due to
underlying geology. Water is added to aquifers by infiltration of rain and surface water
through permeable soils. The rate at which aquifers are recharged is limited by the
amount of rainfall, soil infiltration rate, and the presence of ponds and lakes. In order
to restore the quantity of water withdrawn from aquifers by humans, the recharge
areas must be identified and preserved. Covering them with impermeable paving and
buildings compromises the amount of water that can enter.
As they are below ground, it is rather difficult to quantify the volume of water
contained within an aquifer. However, it is a simple matter to measure the distance
from the surface to the top of the aquifer using groundwater wells. When the distance
consistently increases, this means that the carrying capacity of the aquifer is being
exceeded. More water is being withdrawn than is being replenished. As the volume
of the aquifer is often unknown, it is difficult to predict the amount of time this
Figure 1.18 Principal aquifers in the United States. The United States Geological Survey (USGS)
provides current and historical information about groundwater at http://waterdata.usgs.gov/nwis/gw.
Image: US Geological Survey, 2003. License: government work product, http://waterdata.usgs.gov/nwis/gw.
unsustainable situation can continue before the aquifer is depleted. It is, however,
certain to occur eventually if there are no changes in the consumption or recharge
rate. Increasing the infiltration of water in the recharge areas or substituting surface
or reclaimed water for non-potable uses to conserve the pure groundwater are two
options that might correct the sustainability balance. In either case, constructed
wetlands have a significant role to play.
Aquifers near the surface are more readily subject to pollution from human
activities. When this occurs, it is possible to pump contaminated groundwater and
treat it in constructed wetlands to remove the pollutants. The chapter on industrial
applications presents a wetland that treats groundwater contaminated with hydro-
carbons. The quality of groundwater wells decreased between 2000 and 2010
compared to the period of 1988–2000. Although the quality was unchanged in
about half of the wells, many wells showed increased levels of chloride, dissolved
solids and nitrate. Nitrate was higher in 23 per cent of the wells tested (US Geological
Survey, 2014).
Water quality
Water resource planning generally focuses on supply and allocation, and less often
on treatment for reuse and recycling as a strategy of expanding water quantity. A
primary benefit of constructed wetlands associated with development is the treatment
of various types of wastewater, so that they can be reused or discharged to the
environment without causing adverse ecosystem impacts. In the case of reuse, treated
water could be applied to consumptive uses, such as irrigation, or non-consumptive
uses, such as toilet flushing or industrial cooling and process water. In both cases, the
reuse of water treated by constructed wetlands reduces the demand for high-quality
water diverted from surface water or groundwater. A second important benefit of
constructed wetlands is the protection or restoration opportunity to return water to
streams, rivers, lakes and coastal areas that is pure enough to foster ecosystem health.
Treating agricultural, industrial and urban effluents and stormwater to this quality is
a major challenge that the United States and other countries are only beginning to
address. This book contains chapters on the application of constructed wetlands for
treating human wastewater, stormwater, agricultural and industrial wastewater.
Whether or not we like to think of it, water from rivers and groundwater is recycled
after treatment for subsequent use by humans and a host of other organisms. Cities
along the lower Mississippi and Yangtze Rivers, for example, withdraw, treat and
consume water that has been previously used, treated and discharged upriver. In this
scenario, it is clear that all water entering the river should be treated to as high a
quality as is technically and economically feasible in the interest of other human and
non-human users. Surprisingly, in the case of urban stormwater, filthy water is
allowed to flow into streams and rivers untreated. Constructed wetlands are effective
and reliable options for the treatment of human wastewater as discussed extensively
in the initial chapters of this book.
About 25 per cent of Americans use septic systems rather than a connection to
sanitary sewer networks that deliver wastewater to centralized wastewater treatment
plants. The estimated lifespan of septic systems ranges from 20 to 27 years. Between
5 per cent and 35 per cent of these septic systems do not function correctly due to
age, poor design and maintenance, or due to construction in unsuitable soils. Failing
septic systems are the source of significant non-point source pollution of groundwater,
streams and lakes (Sipes, 2010). Small-scale applications of constructed wetlands to
replace individual septic systems or those serving groups of homes would greatly reduce
non-point source pollution.
Barriers to sustainable water use

The manner and cost of water allocation among users, the unwillingness to pay for
treatment of contaminated water, centralized wastewater treatment and the lack of
redistribution piping for reclaimed water are conditions that lead to wasteful use
of water.
As pure water is an essential resource, assuring a sufficient and consistent supply
and quality is a basic and primary responsibility of any municipality. Yet we commonly
find that citizens, business groups and even municipal governments thwart efforts to
protect water supplies, especially to treat contaminated water to desirable levels before
discharging it downstream. This avoidance of responsibility for cleaning one’s mess
transfers the problem and cost of full treatment for reuse or discharge to healthy eco-
systems to other agencies or municipalities who are also downstream. Adopting this
subterfuge is unnecessary since low cost technologies can be applied and multiple
benefits can be derived from local treatment and reuse. The economic and multiple func-
tions of treatment wetlands is a consistent theme in the chapters that follow.
A certain amount of water is a free ecosystem service to municipalities and even
property owners. However, the cost of purifying, storing, distributing or pumping from
aquifers and rivers can be high. This cost is often not transferred to the water user,
but is instead subsidized by the municipality or water district and funded by tax revenue
from other sources. This leads to a mistaken assumption by the consumer that water
is inexpensive and abundant when, in fact, it is neither. This situation results in excessive
and wasteful use of water, as this behaviour has no economic or supply consequence
in the short term.
Domestic water use and cost

A very high per capita use of domestic water in places such as Las Vegas, which are
almost entirely dependent on importing water from beyond the region, is the
perplexing result of water pricing and policy. High per capita use is partly the result
of the climate. For example, in arid Las Vegas, the per capita use is about 110 gallons
per day, while in the Boston, where there is ample rain, the per capita use is about
40 gallons per day. However, price is also a factor. In Las Vegas, the monthly cost
for a family of four using 100 gallons per person per day is about $42, while in Boston,
the cost for the same amount is about $78 (Walton, 2010). In the European Union,
nine states have a per capita domestic water use of 30 gallons per day or less. However,
arid Greece has a per capita use of about twice this amount. The per capita use in
the United Kingdom is about 44 gallons (European Union, 2014).
The cost of water in Europe ranges from $0.46 up to $6.58 (€0.40–€5.75) for
264 gallons (1,000 litres) with an average price for 65 nations in Western Europe of
$2.19 (€1.91) for 264 gallons (1,000 litres) per month. The price is based on the use
of 15,000 litres per person per month (132 gallons, 500 litres per person per day).
Higher rates are charged for additional quantities (Conroy, 2013). In comparison to
US prices, the monthly cost of water for four people using 100 gallons per day would
range from $22.50 to $322 (with an average of $107.24).
In the United States, the cost of water also varies widely based on its source, quality
and economic sector. Agricultural customers are often able to purchase an acre-foot
of water for $10, which is less than the cost to provide it. In contrast, urban residential
customers might pay $230 for the same amount (Sipes, 2010).
Centralized wastewater treatment

Centralized versus decentralized application of water treatment efforts is a question
that arises when planning for reuse. Metropolitan areas of industrialized countries
adopted a centralized and industrial approach to the treatment of municipal sewage
long ago. Most municipal wastewater treatment plants depend on gravity flow
wherever possible to deliver raw sewage to the treatment plant. This is to avoid the
cost of pumping water. In practice, this means that the treatment plant is located at
the lowest elevation in the community. Therefore, reuse of the treated effluent would
require pumping the treated water uphill to the prospective customers within the same
community. This system of trading energy for reclaimed water is not an optimal
sustainability strategy. A decentralized approach to wastewater treatment in a city
makes the distribution of treated effluent for non-potable uses, such as irrigation, more
feasible as gravity flow can be utilized. Opportunities for decentralized wastewater
treatment, whether through conventional or constructed wetland systems, are possible
as cities expand through new development. Decentralized treatment encourages
residential water reuse for non-potable applications, such as toilet flushing, or recycling
in the case of industrial and aquaculture applications.
Distribution piping for reclaimed water

Associated with the disadvantages of centralized wastewater treatment at the lowest
elevation in a municipality is the absence of a non-potable water piping network to
redistribute the treated effluent. It is not economically feasible to retrofit existing
developed areas with this piping, except in the most dire circumstances, but it is feasible
for new development. As reclaimed water can be used for any non-potable use within
or outside buildings, the demand on potable water from groundwater or treated surface
water can be significantly reduced.
Condition of wetlands and streams

Since the days of the Cuyahoga River (Figure 1.19) in Cleveland and rivers in other
cities repeatedly catching fire between 1900 and 1969, we have made good progress
in the United States and in other nations in controlling point sources of water
pollution. However, natural systems can be overwhelmed and damaged by high
concentrations of excess nutrients and toxic substances that originate from both point
and non-point sources. The resulting loss of ecosystem health in turn compromises
the potential of human health and economic systems.
An assessment of the conditions of waterways and water bodies reveals an
unsustainable course that can be corrected with little cost but sustained attention.
Wetland ecosystems have greatly diminished in size in every country. Although the
Figure 1.19
Images such as this one of the
Cuyahoga River on fire in 1952,
but published in 1969, shocked
the nation.
Photo: James Thomas, Courtesy
of Cleveland Press Collection at
Cleveland State University
Library. License: public domain.
United States now operates under a ‘no net loss’ policy, more than 50 per cent of the
original wetlands in the country have been destroyed. The situation is even worse in
New Zealand where only 8 per cent of the original wetlands remain. These losses are
ecologically devastating, as wetlands made up only a small portion of the total natural
landscape but are critical to both aquatic and terrestrial species (Sipes, 2010). New
constructed wetlands can replace some of the historic wetlands that were drained to
produce agricultural plots or urban development sites.
A comprehensive study of rivers and streams in the United States revealed that
more than 55 per cent of the length of the nation’s rivers and streams are in poor
condition (Figure 1.21). The assessment is based on biological indicators, such as
aquatic insects, as this is also an indirect measure of the physical and chemical
conditions. The study revealed regional differences with rivers and streams in the eastern
United States having the most impaired water. The poor or fair biological conditions
result, in part, from high levels of nitrogen and phosphorus. High nitrogen levels and
high phosphorous levels were found respectively in 40 per cent and 28 per cent of
the river and stream lengths (US Environmental Protection Agency, 2013). Wastewater
contributes large quantities of both phosphorus and nitrogen to natural streams through
discharge from sewage treatment plants, although agricultural run-off from crop-land
and feedlots, and industries that process food such as milk, cheese, wine, beer, potatoes
and meat, can also contribute great amounts of these excess nutrients. Constructed
wetlands and land applications of treated effluent can reduce phosphorus and other
contaminants in a very economical manner (Figure 1.20). The example of Roseburg,
Oregon, illustrates that the three aspects of sustainability (economic, social and
environmental) can be achieved simultaneously.
It seems obvious that agricultural and domestic applications of herbicides and
insecticides that run-off with stormwater would have a negative impact on natural
aquatic ecosystems. Levels of these chemicals that can kill or disrupt growth and
reproduction of aquatic organisms after several hours of exposure are characterized
as chronic level impacts. During 2002–11, chronic levels of pesticides impacting aquatic
organisms were measured in about two-thirds of streams in agricultural areas and
Figure 1.20 The city of Roseburg, Oregon, spent $10 million to purchase 350 acres of land, restore a
degraded wetland and construct a holding pond, pumping station and irrigation system to spray
treated effluent from their conventional wastewater treatment plant to reduce the amount of
phosphorous reaching the receiving river. This alternative saved $90 million compared to an upgrade
to the wastewater treatment plant that would have achieved the same result. The open space and
wildlife benefits resulting from the land application of the wastewater were free.
Figure 1.21 The conditions of rivers and streams in the United States based on the presence and
health of biological indicators (macroinvertibrates) (US Environmental Protection Agency, 2013).
Image: US Environmental Protection Agency. License: government work product.
about half of the streams in areas with a mix of land uses. Streams in urban areas
contained chronic levels of pesticides in 90 per cent of the streams (Stone, Gilliom,
and Martin, 2014). Constructed wetlands can reduce the concentration of nitrogen
and phosphorus and, surprisingly, pesticides as well.
Physical changes to the rivers and streams also compromise the biological
conditions. The loss of riparian vegetation and the degree of human activities near
the rivers and streams were found to be most detrimental to biological conditions.
The sedimentation of the rivers and streams indicated a 60 per cent likelihood of poor
biological conditions (US Environmental Protection Agency, 2013). Linear constructed
wetlands for the treatment of polluted river water is the topic of one of the chapters
in this book. Another chapter on constructed stormwater wetlands addresses the
problem of sedimentation of natural waterways.
Human health can also be compromised by the poor conditions of rivers and
streams, and the United States’ national assessment found that more than 13,000 miles
of river length was associated with mercury levels in fish tissue that exceeded safe
human health standards. For streams, 9 per cent of their length was associated with
pathogenic bacteria (enterococci) at levels exceeding health standards (US Environ-
mental Protection Agency, 2013). Heavy metals associated with urban stormwater
run-off and industrial effluents are effectively removed by constructed wetlands.
Similarly, high levels of pathogens are dramatically reduced in constructed wetlands
treating domestic and livestock sewage.
Constructed wetlands and sustainable development

Sustainability is an oft-bandied term that has lost some of its meaning through
common use and misapplication. ‘Sustainability’ in this book is an assertion that there
is a link between the health of human systems and ecosystems. The authors recog-
nize that there is capacity in natural systems to adapt to greater or lesser inputs of
nutrients and contaminants from both anthropogenic and natural sources. We acknow-
ledge that there are technological and, more often, economic limits to the removal of
pollutants generated by urban development, agriculture and industry.
Sustainable development proposals are not quests for perfection. Instead, they
illustrate the acquisition of technology, practices and outcomes that lead to continuous
economic growth and ecosystem health rather than levels of contamination that will
result in the collapse of human or natural systems and transfer the cost and suffering
this collapse causes to future generations.
It is necessary to develop a philosophical and conceptual basis for integrated
water management and human development. This includes the agricultural and
industrial enterprises that support human settlements but are located in the hinterland.
Many cities in the United States, Europe and Asia have exceeded the carrying capacity
of their aquifers for potable water. Water returned by agriculture and industry to
groundwater or to natural receiving waters is invariably contaminated to some degree
even when it is in compliance with wastewater discharge regulations. Wineries,
breweries, dairy and other food processing industries generate high levels of organic
pollutants. Heavy metals, phenols and hydrocarbons and solvents are by-products of
mining (Figure 1.22), manufacturing and petrochemical factories, and energy
production. In many nations, these point-source pollutants are inadequately removed
from the effluent discharged to the natural environment.
Figure 1.22
Pollution of a small stream from
mine tailings.
Photo: Admerial Crunch. License:
CC-BY-SA-2.0.
Furthermore, the rainwater that replenishes the surface water collects pollutants
as it flows across urban, agricultural and industrial surfaces before it reaches natural
streams and rivers. This contaminated run-off is non-point source pollution that is
joined by municipal sewage effluent, which contains concentrations of nitrogen,
phosphorus (Figure 1.23), pharmaceuticals and chlorine or other substances even when
it has been treated to the common secondary effluent standards. Wastewater and
contaminated stormwater run-off degrade the quality of water in receiving streams
and rivers, especially those that are effluent-dominated.
The characterization of water in its relationship to human settlement described
above is far from a model for sustainable management of water supply and water
quality. Of course, the consequences accrue most tragically to aquatic organisms that
must survive in the waste created by humans. Impact of an aquatic ecosystem can
occur near the pollution source or hundreds of miles away. Rapidly enlarging dead
zones along the coasts in the temperate zones of the world and especially at the outfall
of polluted rivers, such as the Mississippi and the Yangtze, are dramatic illustrations
of the cumulative impacts of municipal, agricultural and industrial discharges along
hundreds of miles of the watershed. Failing ecosystem health in streams, rivers, lakes
and oceans is indicator and indictment of the unsustainable use of water.
Figure 1.23
Excess nutrients from
agricultural and domestic
wastewater cause algae
blooms that can result in
oxygen levels that are too
depleted to sustain fish and
other aquatic organisms.
Photo: Eric Vance, US EPA.
License: government work
product.
The recommendation to clean our mess as and where we make them is a lesson
learned in kindergarten, which is befitting as a sustainable water quality principle.
However, an unwillingness to adopt sustainable water management by citizens,
municipalities and companies is too simple an explanation for the predicament we
find ourselves in. Citizens have little direct control of the pollution created by their
automobiles or the electrical energy power plants that supply their needs. Stockholders
demand profits but rarely sustainable practices. Our progress towards a sustainable
society is thwarted as our attention is diverted by extremities in weather, floods and
droughts, even though these climate change impacts are caused, in part, by unsustain-
able human practices. In the case of climate change, wilful ignorance is a perplexing
obstacle that curtails creative thinking, research and implementation of solutions. Of
course, constructed wetlands cannot mitigate climate change impacts but they can help
us adapt to our changing circumstances.
Some sustainability questions are how well should we clean the mess we create
and how much should we attempt to recover and reuse? How clean should treated
effluent be? Do we have the technology and can we afford the cost to clean wastewater
as well as we would like? Can we solve the logistical problems and bear the cost of
delivering reclaimed water for reuse? This book offers answers to each of these
questions. How clean depends on which use the reclaimed water is put to. If discharged
into the environment, the answer is simple – clean enough to sustain natural
ecosystems. Unfortunately, defining this standard for the array of contaminants is
more difficult, but possible, as this book will illustrate. If the effluent is to be reused
in the industrial process, the quality can often be much lower than for potable water.
If the reuse is for irrigation, high nitrate levels are actually desirable, but high salt
levels are not, for example. The reader will be required to assess the technology and
the performance of the various constructed wetlands with regard to the context
and intended reuse goals. However, in every case, we should strive to decrease the
vulnerability of our communities and increase their resilience. An interdisciplinary
approach and a search of multiple functions will guide us towards these goals.
Types of constructed wetlands

There are three basic types of constructed wetlands; this book comprises a chapter
on each, which discusses their characteristics, uses, construction and performance. In
these initial chapters, the designs and performance data focus on the treatment of
domestic sewage, for which each type was originally developed. Subsequent chapters
discuss the application of the three types of wetlands, and their combinations, for the
treatment of stormwater and agricultural and industrial wastewater. Another chapter
discusses constructed wetlands for the treatment of polluted river or lake water.
The first type of constructed wetland is called a free water surface wetland (Figure
1.24). This type is similar to a shallow marsh and features a depth of water on the
surface. The free water surface wetland supports emergent vegetation such as common
reed or cattail, and floating vegetation such as duckweed or submerged plants. Often
all types of vegetation are present. Chapter Three explores free water surface wetlands.
The second basic type of constructed wetland is called subsurface flow wetland
(Figure 1.25). In these wetlands, water is below the surface in a bed of filter media
such as sand or gravel. There are two alternate flow options for the subsurface flow
wetland. In the horizontal subsurface flow wetland, water moves laterally through
Figure 1.24 Free water surface wetland treating stormwater in Wulijie, China.
the filter bed, while in the vertical subsurface flow wetland, the water flows vertically
through the filter bed. For the vertical subsurface flow wetland, water is typically
applied to the top of the basin and flows downward, but sometimes the water is added
to the bottom of the filter bed and water is forced upward. Both subsurface types
support emergent vegetation.
Hybrid wetlands are a sequence of two or more of the basic constructed wetland
types. They are designed to achieve the highest water quality or are applied to
wastewater that is particularly difficult to remediate. The common sequences are:
• vertical subsurface flow > horizontal subsurface flow

• horizontal subsurface flow > vertical subsurface flow
• horizontal subsurface flow > free water surface
• free water surface > horizontal subsurface flow
Hybrid wetlands are also designed with multiple stages, where each stage is an
independent cell within the sequence. Common sequences are:
• vertical subsurface flow > vertical subsurface flow > horizontal subsurface flow
• vertical subsurface flow > horizontal subsurface flow > free water surface
• vertical subsurface flow > horizontal subsurface flow > vertical subsurface flow
Finally, there are a number of enhancements that can be applied to the constructed
wetlands, such as artificial aeration, baffles or other elements to modify flow.
Units
Pollutants in water are measured in two ways. The concentration of contaminants is
generally expressed as milligrams present within one litre (about a quart; mg/L). There
are 1,000 milligrams per gram and about 28 grams per ounce. For example, the
Figure 1.25 Figure 1.26

Horizontal subsurface flow wetland in Casper, E. coli bacteria colonies in a culture dish.
Wyoming, treating groundwater polluted with A maximum of 126 colonies per 100mL of water
hydrocarbons. is the standard for primary recreational contact
(swimming) set by the US EPA.
Photo: Madprime. License CC-BY-SA-3.0.
US Environmental Protection Agency (EPA) standard for total suspended solids in

secondary sewage effluent is less than 30 mg/L (this is equivalent to 30 parts per million).
Extremely low concentrations of contaminants are expressed in micrograms per
litre (␮g/L). There are 1,000 micrograms in one milligram. Heavy metal concentrations
in water are defined this way. For example, stormwater from landscaping contains
about 263 ␮g/L of zinc.
The second way of expressing pollutants is as weight (often called the pollutant
load). This is simply calculated as the mg/L times the total volume of water in litres.
In this way, we can define the total weight of phosphorous entering a river, for example.
The US EPA has a programme that limits the pollutant load permitted for com-
promised or sensitive water bodies. The Total Maximum Daily Load (TMDL) is an
example of a pollutant standard based on the total weight of the contaminant. For
example, a wastewater treatment plant might be allowed to discharge no more than
90 kilograms (198 pounds) of phosphorous per year into a receiving river.
Bacteria in water are reported as the number of colonies that form in culture dishes
containing a water sample. The sample is generally 100 millilitres (about 3.3 ounces)
of water. For example, Escherichia coli (E. coli) is a disease-causing bacteria and the
US EPA has established a maximum of 126 colonies per 100 mL for water as the
standard for primary recreational use. New testing methods (Quanti Tray Sampler)
do not actually grow a culture of bacteria (Figure 1.26) and express the test results
as the most probable number (MPN) of bacteria.
The toxicity of pollution is sometimes distinguished as acute and chronic. Both
are harmful, but the first is short-term (hours) exposure while the latter is long-term
(several hours or days). These terms are applied to aquatic ecosystems. For example,
federal and state agencies test the effect of toxic substances, such as pesticides and
heavy metals, on aquatic organisms to determine if death or impairment to growth
or reproduction occurs at various concentrations of the contaminant when the
exposure is short-term and when it is long-term. The acute level of exposure is

sometimes defined as the maximum concentration, and the chronic level of exposure
as the continuous concentration. The acute exposure to lead is 30 ␮g/L for 1 hour
for aquatic organisms according to the standard adopted by the state of Georgia, for
example.
References
Austin, G. Green infrastructure for landscape planning: Integrating human and natural systems.
Abingdon, Oxon: Routledge, 2014.
Conroy, N. “Domestic water charges in Europe,” Public Policy, 1 May 2013. [Online].
Available: www.publicpolicy.ie/domestic-water-charges-in-europe/. [Accessed: 5 October
2015].
Diehl, T. H. and. Harris, M. A. Withdrawal and consumption of water by thermoelectric power
plants in the United States, 2010. Reston, Virginia: US Geological Survey, 2014.
European Union, “Water statistics,” European Union, 2014.
Kadlec, R. H. Treatment wetlands. Boca Raton, FL: CRC Press, 2009.
Maupin, M. A., Kenny, J. F., Hutson, S. S., Lovelace, J. K., Barber, N. L. and Linsey, K. S.
“Estimated use of water in the United States in 2010,” US Geological Survey, Circular 1405,
2014.
Philadelphia Water Department, “Green city, clean waters,” Philadelphia Water, 2015. [Online].
Available: http://phillywatersheds.org/what_were_doing/documents_and_data/cso_long_term_
control_plan. [Accessed: 15 October 2015].
Saunders, W. S. and Yu, K. eds., Designed ecologies: The landscape architecture of Kongjian
Yu. Basel: Birkhäuser, 2012.
Schneider, C., Laizé, C. L. R., Acreman, M. C. and Flörke, M. “How will climate change modify
river flow regimes in Europe?” Hydrology Earth System Science, vol. 17, no. 1, pp. 325–339,
January 2013.
Sipes, J. L. Sustainable solutions for water resources: Policies, planning, design, and imple-
mentation. Hoboken, NJ: John Wiley, 2010.
Stone, W. W., Gilliom, R. J. and Martin, J. D. “An overview comparing results from two decades
of monitoring for pesticides in the Nation’s streams and rivers, 1992–2001 and 2002–2011,”
Scientific Investigations Report US Geological Survey Scientific Investigations Report
2014–5154, 2014.
US Environmental Protection Agency, “National rivers and streams assessment 2008–2009: A
collaborative survey,” US Environmental Protection Agency, Washington DC, DC 20460
EPA/841/D-13/001, 2013.
US Environmental Protection Agency, “Water Resources,” Climate Change, 15 September 2015.
[Online]. Available: http://www3.epa.gov/climatechange/impacts/water.html. [Accessed: 15
October 2015].
US Geological Survey, “A national assessment of changes in Chloride, dissolved solids, and
Nitrate in groundwater.” US Department of the Interior, 2014.
Walton, B. “The price of water: A comparison of water rates, usage in 30 U.S. cities,” Circle
of Blue, 26 April 2010. [Online]. Available: www.circleofblue.org/waternews/2010/world/
the-price-of-water-a-comparison-of-water-rates-usage-in-30-u-s-cities/. [Accessed: 10 October
2015].
2
Characteristics of wastewater
Introduction
This chapter provides a background on the fundamentals of wastewater treatment
including basic chemistry and the characteristics of wastewater, since the next three
chapters of this book focus on the treatment of wastewater using constructed wetlands.
Readers with experience in wastewater treatment or water chemistry may wish to skip
this chapter, but if you are new to topics of water quality improvement, this material
will make the following chapters more meaningful. This chapter begins by defining
the characteristics of domestic wastewater and the regulatory standards that control
the quality of effluent discharges.
Characteristics of raw sewage

Although organic matter is the principal contaminant in domestic wastewater, sewage
is characterized by the concentration of a number of contaminants or by the weight
of the contaminants in a specific volume of water. The strength of domestic sewage
varies depending on the city and country. For example, the per person load in the
United States (for biological oxygen demand [BOD], total suspended solids [TSS], total
nitrogen and total phosphorous) is significantly higher than in Turkey, India, Egypt
or Brazil, but better aligned with Denmark and Germany (Henze and Comeau, 2008).
Furthermore, the strength of wastewater has changed over time with the advent of
the garbage disposal and low-volume toilets and other household appliances. The pH
ranges from slightly acidic to moderately alkaline (US Environmental Protection
Agency, 2002).
Table 2.1 illustrates the range and typical concentrations for a number of
components in the United States. To allow systems to be compared around the world,
a population equivalent rather than a per person standard has been adopted. This is
equal to a volume of 0.2 cubic meters of water and 60 grams of BOD per day.
Water is by far the greatest component of domestic sewage. All substances
contaminating water amount to only 0.1 per cent of the total volume. This fraction
is called total solids; some of these are dissolved (80 per cent) and some are suspended
solids (20 per cent). About 65 per cent of the suspended solids will eventually settle
but the remaining 35 per cent will not settle out of solution. The contaminants can
also be characterized as organic and inorganic. Wastewater from livestock, food
processing industries, breweries and domestic sewage is very high in organic matter
that is easily biodegradable.
26 CHARACTERISTICS OF WASTEWATER
Table 2.1 Characteristics of raw sewage in the United States (US Environmental Protection Agency,
2002)
Component Concentration range Typical concentration Mass loading g/p/d

Total suspended solids 155–330 mg/L 250 mg/L 35–75
Biological oxygen demand 155–286 mg/L 250 mg/L 35–65
Total coliform bacteria 108–1010 CFU/100 mL 109 CFU/100 mL
Faecal coliform bacteria 106–108 CFU/100 mL 107 CFU/100 mL
Ammonium-nitrogen, NH4-N 4–13 mg/L 10 mg/L 1–3
Nitrate-nitrogen, NO3-N Less than 1 mg/L Less than 1 mg/L Less than 1
Total nitrogen 26–75 mg/L 60 mg/L 6–17
Total phosphorus 6–12 mg/L 10 mg/L 1–2
mg/L = milligrams per litre, CFU/100 mL = colony-forming units per 100 millilitres, g/p/d = grams per person
per day.
Treatment standards
Municipal wastewater is treated to achieve standards described as primary, secondary
and tertiary. The compliance standards are established by the US Environmental
Protection Agency (EPA) and other national regulatory agencies. Primary treatment
is simply the removal of solids through screening or settling. Secondary treatment
achieves monthly averages of 30 mg/L or less of BOD and TSS as well as disinfection
to remove virtually all pathogenic bacteria. A wastewater treatment facility is also
required to remove at least 85 per cent of the initial BOD and TSS. Tertiary or advanced
treatment significantly reduces the concentrations of nitrate, phosphorous and
pharmaceutical contaminants, and generally requires lower concentrations of BOD
and TSS than required for secondary effluent. Tertiary treatment might include other
measures, such as temperature, depending on the condition of the receiving waters
and the presence of sensitive or endangered organisms. The effluent water quality is
defined by the discharge permit, which allows some flexibility for the size and type
of treatment technology.
Organic matter and nitrogen

The organic matter in domestic wastewater is heterogeneous but composed primarily
of proteins, carbohydrates and lipids. It contains carbon and is decomposed by
bacteria. Bacteria consume oxygen, as they use organic matter for their metabolic
processes. Therefore, great quantities of organic matter in a water body can result in
a drop in dissolved oxygen (DO). At a DO concentration of 1 mg/L or less, few aquatic
organisms can survive. If DO concentrations fall below 5 mg/L (or parts per million),
fish will be unable to live for very long. Trout require at least 8 mg/L of DO, and
salmon need around 11 mg/L. This is why removing organic material from sewage is
necessary to protect ecosystem and human health. Constructed wetlands and
conventional wastewater treatment systems use bacteria to consume organic material,
and this occurs in either oxygen-rich or low-oxygen environments, although it occurs
more rapidly under aerobic conditions.
The amount of organic material in a sample of water is indirectly measured by
determining how much oxygen is consumed within a given amount of time (Figure
2.1). This characteristic is called biological oxygen demand or BOD. Tests of BOD
CHARACTERISTICS OF WASTEWATER 27
Figure 2.1
BOD tests of wastewater samples.
Photo: SuSanA Secretariat. License: CC-BY-SA-3.0.
are generally based on a five-day time period (BOD5), but seven-day and ultimate (about
20 days) tests are sometimes reported in the literature. Effluent discharge permits
establish maximum and monthly average BOD levels for treated water flowing into
streams and other water bodies.
Understanding the transformational sequence of organic nitrogen (organic matter)
and various types of nitrogen is critical to understanding the purposes and design
features of the various kinds of constructed wetlands used for water quality improve-
ment. Nitrogen, like carbon, is an essential element of organisms, but nitrogen in the
form of ammonia and nitrite is toxic to aquatic organisms at high concentrations.
The removal of organic matter and nitrogen is a primary goal of wastewater
treatment. Organic matter is transformed through bacterial activity into ammonia,
nitrate and then nitrogen gas. This is done in three stages called ammonification,
nitrification and denitrification. Organic nitrogen is transformed into ammonia (there
are two forms: ammonia, NH3 and ammonium, NH4) by microbial decomposition in
either an oxygen-rich or oxygen-depleted environment. After ammonification, popu-
lations of bacteria convert ammonia to nitrite, NO2, and then to nitrate, NO3
(nitrification). This process occurs primarily in an oxygen-rich setting. Finally, other
genera of bacteria convert the nitrate to harmless nitrogen gas (denitrification).
Although a wastewater treatment system might convert almost all of the organic matter
to ammonia or nitrate, the amount of total nitrogen in the wastewater might not change
much. This is because organic matter, ammonia and nitrate are all forms of nitrogen.
Total nitrogen decreases significantly only when the transformations reach the stage
of creating nitrogen gas (denitrification). In addition to total nitrogen, measurements
of total Kjeldahl nitrogen (TKN) are also common. TKN is the combination of organic
nitrogen, ammonia and ammonium but not nitrite or nitrate.
Total suspended solids

The second fundamental indicator of water quality is also an indirect measurement.
TSS are particles that do not settle from a sample of wastewater and cause the water
to appear cloudy. There is a high correlation between TSS and pathogenic bacteria
in wastewater and heavy metals in stormwater.
Phosphorous
In wastewater, phosphorous exists in two forms. Organic phosphorus is combined
with organic matter, while inorganic phosphorous is found as either orthophosphates
or polyphosphates. Furthermore, phosphorous can be in either dissolved or solid states.
Like excess nitrogen, excess phosphorus in natural water bodies causes algae bloom
and precipitous drops in DO that can kill populations of fish and other organisms.
Pathogens
There are many potential disease-causing organisms in domestic sewage including
bacteria, protozoa and viruses. Indicators of pathogens may be used to estimate their
presence rather than testing for specific organisms. For example, coliform bacteria are
a very large group of species, only some of which cause disease in humans. However,
since the correlation of disease-causing organisms is highly correlated with levels of
coliform bacteria, they are used as a guide. Faecal coliform bacteria are a smaller
category of bacteria and are often used as a test. The presence and quantity of specific
species, such as Escherichia coli (E. coli), are sometimes identified, but the information
about individual species is almost always used to indicate the presence or absence of
many other disease causing organisms. The current EPA E. coli recommendation for
primary recreational use (swimming and wading) is fewer than 126 colonies/100 mL.
Enterococci, another genus of bacteria pathogen, has a primary recreation standard
of 30 colonies/100 mL.
Conventional wastewater treatment plants

In metropolitan areas, highly engineered wastewater treatment plants rapidly treat
great volumes of sewage (Figure 2.2). Activated sludge processes are the most common
methods used. This process adds large quantities of bacteria from sewage sludge to
raw wastewater. A sequence of process steps requiring aeration, mixing and moving
water through a series of tanks improves the water quality by removing organic
nitrogen, pathogens and suspended solids. This requires many large pumps to aerate,
mix and move the water. The rapid treatment of large volumes of water in a small
land area is possible through the application of very high capital and operation costs.
Highly skilled personnel and the abundant use of chemicals are other factors necessary
to support conventional wastewater treatment plants.
Decentralized wastewater treatment

However, for small towns and cities, lower capital and operation costs and very little
need for chemicals or trained personnel is an option presented by constructed wetlands.
In fact, tens of thousands of small towns in Europe rely on constructed wetlands to
produce wastewater effluent with similar or better quality compared to their big city
counterparts. Since constructed wetlands can be created as a decentralized technology,
new wetlands can be created for each new neighbourhood just before it is constructed.
CHARACTERISTICS OF WASTEWATER 29
Figure 2.2 Conventional wastewater treatment plants are highly engineered facilities that require
sophisticated operation controls. They require large capital investments as well as substantial annual
energy, operation and maintenance budgets. In this image, a large pump powers an aeration device
to increase the oxygen content of the water and support bacteria that consume organic matter.
Photo: The Chesapeake Bay Program. License: CC-BY-NC 2.0.
For the individual home or group of homes or for institutions, constructed wetlands
create a more effective and reliable alternative to the leach field common in rural
America.
Sewage lagoons
Every constructed wetland, as well as every conventional wastewater treatment plant,
is preceded by a pretreatment step that may be as simple as a series of screens and
grit chambers or more capable as an effective water quality improvement stage, such
as a multi-chamber septic tank. Small towns comprising a few thousand people often
use aerated or facultative lagoons as the pretreatment step (or the only treatment) for
domestic sewage.
Facultative lagoons are 4–8-feet-deep sewage ponds that are not mechanically
mixed or aerated, where raw sewage is held for as long as a month. In very cold winter
areas, sometimes three or four months’ worth of sewage is stored in these ponds until
spring. Three zones are created in the facultative lagoon. The top layer of water is
oxygen rich (aerobic) due to contact with the atmosphere, while the bottom layer
is anaerobic and contains sludge that settles due to gravity. The middle zone (the
facultative zone) has a reduced oxygen concentration (anoxic) and is populated by
organisms that can flourish in a range of oxygen concentrations. These microorganisms
are called facultative, as they have the capacity to survive in various oxygen environ-
ments. Facultative lagoons are characterized by rapid algae growth, objectionable
odours (especially when wind or temperature inversions cause the lagoon layers to
Table 2.2 Typical septic tank effluent concentrations (US Environmental Protection Agency, 2002;
Crites et al., 2014)
Component Concentration range Typical concentration

Total suspended solids, TSS 36–85 mg/L 60 mg/L
Biochemical oxygen demand, BOD 118–189 mg/L 120 mg/L
Faecal coliform bacteria 106–107 CFU/100 mL 106 CFU/100 mL
Ammonium-nitrogen, NH4-N 30–50 mg/L 40 mg/L
Nitrate-nitrogen, NO3-N 0–10 mg/L 0 mg/L
Total nitrogen 29.5–63.4 mg/L 60 mg/L
Total phosphorus 8.1–8.2 mg/L 8.1 mg/L
mg/L = milligrams per litre, CFU/100 mL = colony-forming units per 100 millilitres.
mix) and large land requirements. Therefore, they are invariably fenced to prevent
public access and isolated from residential areas.
Aerated lagoons require pumps to deliver air to the bottom of the ponds. The
oxygenated water becomes the environment for aerobic bacteria that consume organic
matter. The aerobic process is faster but requires energy inputs and performance
declines in cold weather.
Septic tanks
In contrast, septic tanks are airtight and are usually placed below ground. Since there
is no possibility of public contact with contaminated water and a useful landscape
can be installed above the underground tank, they are well suited to public areas and
as a pretreatment stage for subsurface wetlands. Since they are below ground, they
are somewhat insulated from cold weather. They can also be installed with passive
solar windows to raise the temperature inside the tank. Water is held in septic tanks
for 12 hours to a few days, and conditions are anoxic or anaerobic. Modern septic
tanks significantly improve the quality of the effluent (outflow) in comparison to the
raw sewage inflow.
Septic tanks improve water quality significantly as illustrated by a comparison of
the concentrations in Tables 2.1 and 2.2. The pH range for septic tank effluent is
6.4–7.8 with an average of 6.5 (slightly acidic) (US Environmental Protection Agency,
2002).
References
Crites, R. W., Middlebrooks, E. J., Bastian, R. K. and Reed, S. C. Natural wastewater treatment
systems, Second edition. Boca Raton: IWA, CRC Press, Taylor & Francis Group, 2014.
Henze, M. and Comeau, Y. “Wastewater characterization,” in Biological wastewater treatment:
Principles, modelling and design, London: IWA, 2008, pp. 33–52.
US Environmental Protection Agency, “Onsite wastewater treatment systems manual,”
EPA/625/R-00/008, 2002.
3
Free water surface
constructed wetlands
Introduction
This and the next two chapters introduce three types of constructed wetlands. The
designs and treatment performance presented for these wetlands are based on
remediation of domestic or municipal wastewater, because this is the original purpose
for each. Furthermore, this will provide a common basis for comparing the capability
of each wetland type.
Engineered free water surface wetlands

Free water surface wetlands are engineered treatment landscapes, but they are very
similar in appearance to naturally occurring marshes (Figure 3.1). The wetland shown
in Figure 3.2 is a man-made landscape to treat non-point source pollution from
stormwater run-off. The variety of plants and the presence of open water make these
very attractive places of interest to people as recreation and education sites and to
wildlife for shelter, food and water.
Figure 3.1
This wastewater treatment plant uses free water
surface wetlands (six cells of deep and shallow
marsh) to reduce nitrogen and phosphorus levels
to meet very high water quality discharge
standards. Blue Heron Water Reclamation Facility,
Titusville, Florida.
Photo: Google Earth 28°32′25.03″ N 80°51′37.83″ W,
image date: 13 February 2014, image accessed:
23 October 2015.
32 FREE WATER SURFACE CONSTRUCTED WETLANDS
Figure 3.2 Although engineered landscapes, free water surface constructed wetlands, such as this
one to treat stormwater, mimic natural wetlands. Therefore, careful attention to planting local native
species will yield results that reduce maintenance and enhance treatment performance.
Photo: Wendy Patoprsty. License: reproduced with permission.
Experiments with free water surface wetlands in the United States for the treatment
and disposal of municipal sewage began over 100 years ago when partially treated
sewage was discharged into natural wetlands. The first artificial free water surface
wetland designed to treat sewage was built in the Netherlands in 1967. While it was
effective, the design required a considerable amount of space. In the United States,
during the 1970s, experiments yielded engineered constructed wetlands for wastewater
treatment. Although the Europeans focused on the development of the horizontal
subsurface flow wetland as the preferred treatment wetland type, in America the free
water surface type became the most common constructed wetland (Kadlec, 2009a).
In the United States, these wetlands are most often used to treat secondary effluent
from conventional wastewater treatment plants to achieve advanced (tertiary) water
quality (Figure 3.3). This wetland type is also most common for the treatment of
stormwater run-off. Stormwater wetlands are discussed in another chapter.
There are now hundreds of free water surface wetlands for tertiary treatment in
the United States, and some of them are very large (Figure 3.9). For example, a free
water surface treatment wetland in Orlando, Florida, is about 1,235 acres (500
hectares) (Kadlec, 2009b).
Design challenge
From a sustainable development perspective, the primary challenge in the design of
free water surface wetlands for domestic sewage or livestock wastewater treatment is
providing secondary uses simultaneously. Compared to the subsurface flow wetland
types, the potential of free water surface treatment wetlands to provide recreation,
education and wildlife benefits is much greater. However, the public must be protected
FREE WATER SURFACE CONSTRUCTED WETLANDS 33
Figure 3.3
The inflow from the wastewater
treatment plant across the
width of the shallow marsh
wetland at the Blue Heron
Water Reclamation Facility.
Photo: Matt Hixson. License:
reproduced with permission.
Figure 3.4
Water from septic tanks,
sewage lagoons or wastewater
treatment plant secondary
effluent (as in this case) is often
oxygen depleted. Aerating the
inflow with an artificial or
naturalistic cascade adds the
oxygen necessary for rapid
biological removal of
contaminants.
from contacting water that has pathogenic bacteria in excess of US Environmental

Protection Agency (EPA) standards for primary (swimming) or more commonly
secondary (boating and fishing) recreational use. Consequently, free water surface
wetlands are rarely used to treat primary effluent from septic tanks or sewage lagoons
due to odour and contact hazards. Instead, they are employed to raise the quality of
water from secondary to tertiary standards by treating secondary effluent (Figure 3.4).
A second challenge is creating designs where the water flows uniformly through
the wetland to make optimum use of the treatment capacity and maintain a small foot-
print. Mosquito control and wildlife use are also challenges but more manageable ones.
Pretreatment
Extending the longevity of the treatment wetland and reducing the organic load that
is applied are the purposes of pretreatment of domestic sewage. In rural communities,
this is often accomplished in facultative or aerated sewage lagoons, but a septic tank
or Imhoff tank are better pretreatment solutions.
Figure 3.5
The design for a free water
surface wetland includes a
distribution pipe with clean-
outs, a deep inflow trench or
pool, water depth varying
between 12 inches and 30
inches, and dense emergent
vegetation perpendicular to the
flow direction.
Image: Zhenyu Liu and Gary
Austin.
Typical configuration
The general design of a free water surface wetland includes a deep-water pool or trench
at the wetland inlet and outlet (Figure 3.5). Then the wastewater flows on the surface
through dense stands of emergent vegetation. The depth of the water in these emergent
marshes may vary from about 12 inches to about 30 inches to encourage the growth
of a variety of plant species. Sometimes areas of open water 3–6 feet deep are
incorporated into the design. If the wetland is large, it is usually divided into a sequence
of cells in order to collect and redistribute the water uniformly. The wetland is generally
sealed with a high-density polyethylene (HDPE) liner or clay infiltration barrier. Treated
water flows into an outlet structure with an adjustable pipe that controls the water
level within the wetland. This is adjusted to control weeds or respond to extremely
cold weather. If there are high levels of pathogenic bacteria in the water, then the
wetland is fenced to prohibit human access, but where secondary effluent has been
chlorinated or otherwise disinfected before reaching the constructed wetland, then
public access might be permitted (Figure 3.11) (Crites, 2001).
Sizing
Monitoring the performance of existing free water surface wetlands for the treatment
of domestic sewage has established a sizing rule of 48 to 54 square feet (4.5–5 m2)
per person to accomplish secondary treatment standards (Cooper, 2009). This equals
about 1 acre (0.4 hectare) for every 850 people the system serves. For accurate wetland
sizing, matched to target contaminants or nutrient reduction, communities should
consult an environmental engineer.
Loading
In order to achieve 30 mg/L for biological oxygen demand (BOD) and total suspended
solids (TSS) in the effluent of a free water surface treatment wetland, the recommended
aerial loading is 6 grams per m2 and 7 grams per m2, respectively. The standard of
30 mg/L for BOD and TSS is the basis for secondary effluent in the United States. An
aerial load of 1.5 grams per m2 of total Kjeldahl nitrogen (organic nitrogen and
ammonia [TKN]) should yield a 10 mg/L concentration in the wetland effluent.
Phosphorus in domestic wastewater ranges from 4 to 12 mg/L, and little improvement
can be expected if aerial loading is 0.4 grams per m2 or more. If loading of faecal
coliform bacteria is more than 10,000 colony-forming units (CFU) per 100 mL, then
it is unlikely that the effluent will meet secondary recreation standards (200 CFU/100
mL). A hydraulic residency time of 3 days is typically required for a 2 log reduction
in bacteria (Wallace, 2006).
Hydraulic loading rate

For domestic and municipal sewage applications, the amount of water that the
wetland receives is generally determined by the BOD aerial loading criteria noted above.
The amount of water associated with the BOD varies. Water use per person in the
United States is much higher than in Europe and developed countries in Asia. Per person
water use in developing countries is much lower than use in Europe. The average water
use influences the hydraulic loading rate of the wetland (and hydraulic residency time).
In the United States, 4 cm (1.6 inches) to 10 cm (3.9 inches) per day is a common
range for hydraulic loading (Wu et al., 2015). Therefore, if the hydraulic loading is
3.5 inches of water added to a 1-acre wetland, this equals 95,000 gallons of capacity
per day.
If the wetland receives flow from combined stormwater and sanitary sewers, then
the amount of water increases and concentration of BOD decreases, although the BOD
load is the same. This situation results in water moving more rapidly through the
wetland in the high flow condition, and there is less time for bacteria and other
mechanisms to remove BOD and other contaminants. Recycling some of the treated
wastewater to the beginning of the wetland is sometimes recommended where the
concentration of BOD is very high and the volume of the water inflow is low. This
would dilute the BOD concentration to levels compatible with the cultural needs of
the wetland plants.
Cell shape
Initial experiments in the Netherlands led to long narrow wetlands with berms
between them to facilitate harvesting the wetland plants. This configuration doubled
the land area required and was not the optimum shape for treatment effectiveness.
Free water surface wetlands are particularly troubled by the creation of preferential
pathways by the water flowing through them. Therefore, the water flows through more
rapidly and with less contact with the plant stems and bottom surface than intended.
To compensate, the wetland is usually divided into two or more roughly square or
rectangular treatment cells to shorten the flow length and allow the water to be collected
and redistributed evenly across the width of the cells.
Depth
The water in free water surface wetlands generally has high levels of dissolved oxygen
but this declines at the bottom of the treatment pool. A variety of oxygen levels is
generally beneficial, as some contaminants or excess nutrients are removed more
effectively in one condition or another. However, treatment wetlands deeper than
3 feet limit the growth of almost all emergent plants (Figure 3.6), which might be
replaced with floating leaf plants, such as water lilies, or floating plants, such as duck
weed, or open water. In most designs, the water depth is 16–24 inches (15–60 cm).
Figure 3.6
Differing water depths
encourage a variety of species
to flourish, and increase the
ecosystem benefits of the
artificial wetland. Ideally, the
vegetation is planted in bands
perpendicular to the flow
direction of the wastewater.
This improves uniform
distribution, retention time and
treatment performance.
Materials
Waterproof membrane
Preventing infiltration of contaminated water can be accomplished with synthetic liners
or clay. Free water surface wetlands for tertiary treatment of effluent from conventional
wastewater treatment plants are often very large. Therefore, the cost of synthetic liners
is avoided where possible through the use of bentonite or other clay barriers covered
by wetland soil. The compacted depth of a bentonite layer is about 12 inches (30 cm)
and may be an economic alternative to synthetic liners if the bentonite source is near
the proposed wetland. Bentonite is mined in Wyoming in the United States. When
seasonal groundwater is well below the bottom of the wetland, then the US EPA permits
a clay barrier if infiltration can be limited to 10–6 cm per second (Kadlec, 2009b).
Synthetic liners are made of polyvinyl chloride (PVC) or HDPE or other com-
pounds such as polypropylene. PVC, at least 30 mil (0.76 mm), is generally the least
expensive and is suitable if not exposed to sunlight. HDPE should be at least 40 mil
(1 mm) and is about 10 per cent costlier than PVC. Liners with an embedded nylon
or other netting are available where great strength is necessary. When either a clay
or synthetic liner is used, a layer of wetland soil 12 inches (30 cm) deep on top of the
liner will support tall emergent plants (Kadlec, 2009b).
Inlet/outlet
Generally, a deep 6-feet wide (2 m) trench at the inlet of each free water surface
treatment cell is necessary to collect sediment and to uniformly distribute the influent
(Figure 3.5). In practice, many inlet designs are used, including discharge onto a rock
slope above the wetland. However, distribution of water continuously along the entire
face of the wetland through perforated pipes or other methods leads to more uniform
flow and positive treatment results. Similarly, a continuous collection of the effluent
from each cell rather than a single outlet point will encourage uniform flow (Figure
3.7). Two or more small treatment cells rather than one large cell where water is
Figure 3.7
This below ground view shows
the wetland cell outlet trench
with its perforated collection
pipe. The outlet vault includes
a vertical pipe that can be
adjusted to establish the level
of the water in the wetland.
Image: Zhenyu Liu and Gary
Austin, 2016.
distributed and collected as described above will yield the best treatment result. Where
the public has access to the wetland, care must be taken to provide safe conditions.
Therefore, limiting slopes to 3:1 or flatter and water depth to 3 feet, or other measures
to prevent visitors from falling into deep water, are necessary.
Filter media
In free water surface wetlands, the stems of the vegetation serve as the filter media
causing sedimentation. Dense vegetation should be planted in strips or masses
perpendicular to the direction of water flow. Bacteria on the plant stems and in the
top layer of soil and organic material on the bottom of the wetland convert organic
material to ammonia, and ammonia to nitrate.
Cold weather applications

Free water surface wetlands have been implemented successfully for the treatment of
domestic sewage in northern United States and southern Canada, where the average
annual temperature is about 44.6°F (7°C) or higher (Kadlec, 2009b). However,
operation in freezing weather requires active management of the system to create an
air space below a layer of ice or insulation of the wetland with mulch on top of the
ice. The size of the wetland must be larger in cold climates in order to meet treatment
standards in winter.
Clogging
Even very dense stands of emergent vegetation in a free water surface wetland are
very porous, allowing water to move freely through the plants. In fact, uniform flow
is a more important issue than it is with other wetland types, as the water readily
finds the most open route through the wetland. Accumulation of sediment and detritus
in the free water surface wetland diminishes its maximum volume and somewhat
shortens its life; therefore, good pretreatment is necessary.
Site-scale application
An example of the uncommon use of a free water surface wetland for the treatment
of septic tank effluent is a three-cell wetland treating wastewater (toilet and sink) from
300 people in a university building. The wetland was constructed in Taiwan at the
National Cheng-Kung University and began operation in November 2003. The
treatment wetland is composed of two free water treatment cells and an ornamental
pond (Figure 3.8). A filter and a storage tank to facilitate water reuse follow these
basins (Ou et al., 2006).
Local soil 30 cm deep over an impermeable liner supports cattail (Typha orientalis)
(half of cell C, and cell D, Figure 3.8) and water hyacinth (Eichhornia crassipes) in
the second half of the first basin (C). The water in the ornamental pond is between
6 inches (15 cm) and 33.5 inches (85 cm) deep, and contains umbrella papyrus (Cyperus
alternifolius), hayata (Hygrophila pogonocalyx), yellow water lily (Nuphar shimadai),
water primrose (Ludwigia adscendens) and tropical sundew (Drosera burmanni)
(Ou et al., 2006).
The hydraulic loading rate is 4.2 cm per day (which is higher than the typical
rate), which yields a residency time of 6 days. In the septic tank effluent, the level of
organic matter and ammonia was high (9.6–154.7 mg/L TKN) but nitrate was low
(0.05 mg/L). The treatment system resulted in low levels of TSS, BOD and chemical
oxygen demand (COD), and good performance for the reduction of total phosphorus
and pathogenic bacteria (Table 3.1). Initially, 53.1 per cent of nitrogen was organic,
while 46.5 per cent was ammonia. The treatment cells demonstrated effective am-
monification and nitrification as indicated by the increase in nitrate and nitrite to as
much as 25.9 mg/L. The level of nitrite and nitrate dropped from 25.9 to 13.68 mg/L
in the final effluent, indicating that some denitrification occurred, although nitrate in
the final effluent remained high. Since the concentration of nitrate was high in the
effluent, the reduction in total nitrogen was moderate (42.3 per cent). The remaining
concentration of TKN (organic nitrogen and ammonia) in the effluent also indicates
incomplete nitrogen removal. The ornamental basin had a significant impact on the
improvement of water quality especially with regard to BOD, nitrate, TKN, total
nitrogen, total phosphorus and total coliform bacteria (Ou et al., 2006). Recycling a
portion of the treated effluent to the septic tank would probably reduce the nitrate
and total nitrogen in the final effluent.
The first cell (C in Figure 3.8) was fenced to prevent public access, but the
remaining two cells were available. In the effluent of the second cell (D), the E. coli
Figure 3.8
Plan view diagram of the constructed
wetland sequence at the architecture
building at National Cheng-Kung
University. The sequence wraps around
university buildings. The ornamental
pool (E) fills a narrow space between
buildings. A – Septic tank; B – Pump
vault; C – Cell 1; D – Cell 2; E –
Ornamental pond; F – Slag filter;
G – Storage tank (Ou et al., 2006).
Image: Gary Austin adapted from W. S. Ou
et al. (2006).
Table 3.1 Mean concentrations (mg/L) and per cent removal of contaminants in a free water surface
treatment wetland in Taiwan (Ou et al., 2006)
TSS BOD COD TKN NOx TN TP TC EC

Influent 63.1 94 198.5 81.8 0.05 82.1 14.5 8,964 2,258
Effluent 16.4 16.5 60.4 20.4 13.68 35 4.2 947 UD
% removal 74.5 83 65.9 72.3 – 42.3 69.8 78.5 >99.9
TSS = total suspended solids, BOD = biological oxygen demand, COD = chemical oxygen demand, TKN =
total Kjeldahl nitrogen (organic nitrogen and ammonia), NOx = nitrite and nitrate, TN = total nitrogen, TC =
total faecal coliform bacteria, EC = Escherichia coli bacteria.
concentration was 6 CFU/100 mL and not detectable in the water of the ornamental
pond. The US EPA standard for primary recreation contact is 126 CFU/100 mL.
Therefore, public access to the second and third cells is warranted. Mosquitos in the
wetland were controlled by introducing fish (Macropodus opercularis) that prey on
mosquito larva (Ou et al., 2006).
The first two cells of the university free water wetland achieved US EPA secondary
effluent standards. The third cell planted with ornamental wetland plants provided a
campus amenity, while providing some tertiary water quality benefits. The treated water
was reused in the drip irrigation system for the campus landscape (Ou et al., 2006).
Since the water is reused for irrigation, the residual nitrate is a benefit to plant growth
and reduced the need for artificial fertilizer.
Large-scale tertiary treatment applications

The E L Huie Jr Constructed Treatment Wetlands in Georgia are used to treat
secondary effluent from a conventional wastewater treatment plant to a high standard.
The Huie free water surface wetland, shown in Figure 3.9, includes several parallel
sequences of treatment cells within the 263 acres of constructed marsh. The tertiary
treatment allows the water to be delivered to drinking water supply reservoirs. The
amount of water reclaimed (17.4 million gallons per day or MGD) approaches
the amount used (25 MGD) by the customers in the wastewater treatment district
(Jarrin, 2012).
The US EPA established discharge concentration maximums for the Huie wetlands
(Table 3.2). The actual concentration is well below the compliance standards. The
estimated residence time of water through the wetland and reservoirs before reuse is
1 year under drought conditions and 2 years under normal supply and withdrawal
conditions (Jarrin, 2012).
Since the water entering the Huie wetland has been disinfected as a final step at
the conventional wastewater treatment plant, the constructed wetlands are a resource
for wildlife and the public. In fact, there is a visitor centre dedicated to providing
information and tours to view and interpret the wetland ecosystem and wildlife.
Retrofit and rural applications

Rural communities often use sewage lagoons for treatment of wastewater. These are
generally inefficient and consume large land areas. The town of Cle Elum, Washington
(2,300 people), used three aerated lagoons, but these often failed to meet the discharge
Figure 3.9 The E L Huie Jr Constructed Treatment Wetlands in Clayton County, Georgia, USA, have
the capacity to treat more than 17 MGD to tertiary standards.
Photo: Google Earth, 33°29′11.05″ N 84°17′50.82″ W, image date: 5 June 2014, image accessed:
10 August 2015.
Table 3.2 Water quality standards and performance of the E L Huie Treatment Wetlands
Parameter NPDES limit mg/L monthly/weekly average Actual mg/L (2011 average)
BOD 10/15 3
TSS 30/45 5
NH3 1.4/2.1 0.06
TP 0.6 (1) 0.24
1 – Annual average at the lake discharge.
standards for secondary effluent. In order to improve compliance, the third sewage
lagoon was converted into a free water constructed wetland, which was inaccessible
to the public. The new five-acre wetland was divided into three planted marsh sections
by two 3-feet deep (1 m) open water trenches. The water level above the planted beds
was maintained at 18 inches (46 cm) in summer and increased to 24 inches (61 cm)
in winter to accommodate a layer of ice in this cold climate (latitude 47°N) (Crites,
2001).
The original lagoon received 24 inches of fill, a 6 inch depth (15 cm) of 3⁄4 inch
(19 mm) minus, non-angular gravel and an HDPE liner and then another 6 inch depth
of gravel over the liner. This formed the bottom of the open-water zones. The marsh
beds received an additional 12 inches (30.5 cm) of soil fill topped by a 6-inch layer
of planting media (half soil fill and half sewage sludge). The three marsh sections were
planted with hard-stem bulrush (Scripus acutus) and represented 68 per cent of the
wetland area. The deep-water strips increased water retention time and discouraged
short circuit flows by redistributing water evenly across the width of the wetland.
Performance of the system for BOD and TSS was outstanding. Influent BOD (182
mg/L at the first lagoon inlet) was reduced 96 per cent to 6.4 mg/L. Influent TSS
(169 mg/L at the lagoon inlet) were reduced 98 per cent to 3 mg/L at the wetland
outflow. The level of dissolved oxygen at the constructed wetland outfall averaged a
healthy 6.9 mg/L (Crites, 2001) compared to minimum levels of 0.2–0.6 mg/L required
for conversion of organic matter to ammonium and the conversion of ammonium to
nitrate (Zhang, 2010).
The Talking Water Gardens case study

The Albany–Millersburg conventional wastewater treatment plant, in cooperation
with an industrial user, recently constructed a free water surface wetland to achieve
improved water quality (Figure 3.10). The site was formerly a lumber mill. Designed
by CH2M Hill and Kurisu International, the wetland is located in Albany, Oregon
(USA), and discharges to the Willamette River (Figure 3.11). The Talking Water
Gardens project is a public–private partnership where the $13.75 million cost was
shared by the cities, state and industry. An additional $5 million is planned for visitor
amenities.
The Willamette River has a Total Maximum Daily Load (TMDL) for temperature,
bacteria and mercury reduction required by the Oregon Department of Environmental
Quality. Of particular concern is the temperature of the river, where cooler water is
Figure 3.10 This free water surface constructed wetland in Albany, Oregon, like the Huie wetland,
treats secondary effluent to meet tertiary standards. It is also a popular public park and tourist
attraction. The lush wetland vegetation, extensive walking trails and the many bird species draw many
visitors to the Talking Water Gardens.
Figure 3.11
Talking Waters Garden
in Albany, Oregon. The
Willamette River is on
the upper left, and the
wastewater treatment
plant and metal industry
are at the bottom of the
image. The wetland is
divided into 10 basins
and six individual flow
paths. The berms
between the basins form
an extensive pedestrian
network.
Photo: Google Earth,
44°38′47.12″ N
123°04′02.83″ W, image
date: 14 July 2014, image
accessed: 31 August
2015.
Figure 3.12 Deep water zones at the Talking Water Gardens project increase residence time and cool
the wastewater effluent before it is discharged into the Willamette River.
required for certain salmon and trout species for rearing and migration in the stretch
of the river by Albany. The Talking Water Gardens was built to address the tem-
perature and heat load reduction of the treated effluent from the wastewater treatment
plant and the industrial user (Figure 3.12). The option of the constructed wetland,
rather than other mechanical cooling options, such as cooling towers and refrigeration,
was selected partly due to lower cost and other benefits, such as improving habitat
and creating public amenities. In addition to temperature reduction, water quality is
improved with the mixing and aeration of the water in the wetland. Other pollutants
and nitrogen and phosphorous levels are also reduced through biological treatment
and sedimentation (US Environmental Protection Agency, 2012) (Figure 3.13).
The free water surface wetlands constructed in Albany treats a maximum of
9.6 MGD of effluent from the Albany–Millersburg wastewater treatment plant and
an additional 3 MGD of treated effluent from the ATI Wah Chang metal manufactur-
ing plant. The treated effluent from this industry is characterized by high total
dissolved solids and is fully blended with the Albany treated effluent. The wetlands
occupy 39 acres of the 50-acre site. The remaining land is dedicated to public trails
and other user amenities and wildlife habitat. The water averages a detention time of
two days. Several years after construction, the wetland is still growing to achieve the
plant coverage needed for shade and heat load reduction.
Often, free water surface wetlands are used to achieve tertiary quality effluent,
so they can generally be accessible to the public without concern for contact with
pathogenic bacteria. The great advantages of free water surface treatment wetlands
are the ancillary recreation, education, scenic and wildlife benefits that can be achieved
using careful design (Figure 3.14). In the case of the Talking Water Gardens, which
is open every day to the public, thousands of visitors, bird watchers and photographers
come each year to walk the trails and view wildlife and wetland scenery. Organized
activities include tours with professional groups, hosted events including track meets
and volunteer activities. Talking Water Gardens was built as a ‘living laboratory’ to
bring wetland science to life for school grades from kindergarten through high school
as well as for university students. Educational programmes, such as student tours,
field day events and wetland habitat research, occur throughout the year (Neal and
Preston, 2015).
Figure 3.13 Several cells of the treatment wetland are shown in this panoramic view. In the
foreground is a deep pool covered with floating vegetation (Azolla sp.). The vegetation shades the
water and prevents heat gain. Azolla is an effective accumulator of lead, cadmium, nickel and zinc. It
can be harvested and used as a green fertilizer. A heavy mat can be detrimental to submerged plants
and other aquatic organisms.
Table 3.3 Official visitors (Neal and Preston, 2015)
Fiscal year Total visitors

2010–2011 497
2011–2012 986
2012–2013 1,627
2013–2014 1,372
The Talking Water Gardens staff keeps track of numbers of some types of
visitors. The visitors who come to the garden for professional involvement, public
outreach, events, volunteer opportunities or education since construction began in 2010
are shown in Table 3.3. However, the garden staff does not count the numbers of
informal visitors from the public. It is likely that their numbers are far greater than
those visitors that come in groups for events (Neal and Preston, 2015).
The wetland park includes cells 1–5 feet deep (Figure 3.12) and nine waterfalls
(Figure 3.14) and weirs to aerate and mix the water. In addition to meeting the more
stringent TDML standards, the wetlands make other water quality improvements. They
will remove lead and nickel below the most restrictive regulatory limit (81 per cent
and 94 per cent, respectively). Sixty per cent more BOD will be removed than is required
by regulations, and chromium, zinc, mercury, oil and grease levels will also be below
the most restrictive regulatory limits (US Environmental Protection Agency, 2012).
The amount of nitrogen in the sewage and industrial influent is high primarily
due to nitrate (Table 3.4). There are three points where treated wastewater enters the
wetland. The north influent point is associated with industrial wastewater and is
characterized by higher concentrations of ammonia and nitrate. All of the wetland
effluent discharges from a single point. The constructed wetland removal of ammonia
Figure 3.14
The aeration of the wastewater
entering the wetland from the
wastewater treatment plant is
artfully done to enhance the
naturalistic park-like setting.
Table 3.4 Ammonia and nitrogen removal (mg/L) (Frazier, 2012)
November average May average

Ammonia Nitrate Ammonia Nitrate
North influent 3.5 10.1 0.9 14
East influent 2.7 10 0.5 11
South influent 2.1 9.2 0.6 11.3
Final effluent 1.7 8.8 <0.5 7.4
and nitrate is good, particularly in warm weather. However, the nitrate concentration
of 7.4 mg/L is still quite high. Plans to add carbon to a portion of the treatment sequence
is being studied as a way to further reduce nitrate in the outflow (Frazier, 2012).
Denitrifying wetlands that use wood chips as a carbon source are simple subsurface
flow wetlands that were initially created to remove high concentrations of nitrate from
agricultural fields. See the chapter on agricultural applications for a description of
these wetlands.
Reduction of the temperature of wastewater discharges into rivers and streams is
important for the health of fish species that require cold water to sustain healthy popu-
lations. The Snake River, Willamette River and other rivers in the Pacific Northwest
are important salmon habitat, and excessively high temperatures can threaten resident
and migratory fish. For example, Chinook salmon are more prone to disease, injury
and stress when water temperatures exceed 60°F. At a water temperature of 70°F,
these fish become very vulnerable. This was illustrated in July 2015 when water
temperatures in the Willamette River rose steadily, from 70° to 74°F (Friesen, 2015).
The deaths of hundreds of Chinook salmon are attributed to the temperature of river
water in the Willamette, Klamath and Snake rivers. Therefore, regulatory agencies
set temperature discharge targets for wastewater treatment effluent. For example,
the regulatory agency might require that a wastewater treatment plant reduces the
temperature of its effluent from 73°F to 68°F (22.8°C to 20°C). The Talking Water
Gardens wetlands reduce the temperature of the water entering the river.
Birds are quite reliable indicators of urban biological diversity. Table 3.5 lists the
species of birds that have been photographed or observed within the park by invited
photographers and bird census participants from the Audubon Society (Baldwin
and Proebsting, 2015). The extensive list is the result of the site’s proximity to the
Willamette River, which serves as a flyway for migrating birds and the abundance of
water that attracts waterfowl. In addition to documenting the presence of the bird
species with their gorgeous images, photographers capture landscape scenes through-
out the seasons and highlight the varied site features including bridges, water displays
and a waterfall that is formed by a foundation remnant of the former saw mill. See
the great images on the Talking Waters Facebook page (www.facebook.com/talking
watergardens/photos_stream).
The most common plants in the free water surface wetlands at the Talking Water
Gardens are the broadleaf cattail (Typha latifolia) and hardstem bulrush (Scirpus acutus
or Schoenoplectus acutus). Other emergent plants are wapato (Sagittaria latifolia)
(Figure 3.15), softstem bulrush (Scirpus validus or Schoenoplectus tabernaemontani)
and various rushes (Juncus spp.). Species native to the Pacific Northwest are occurring
naturally in the park, and invasive species are removed or managed with integrated
pest management techniques (Neal and Preston, 2015).
Table 3.5 Bird species photographed at the Talking Waters Gardens
Cooper’s Hawk Red-Tailed Hawk

Bald Eagle Cooper’s Hawk
Red-Winged Blackbird Bufflehead Duck
American Coot Killdeer
Green-Winged Teal Cinnamon Teal
Mallard Duck Shoveler Duck
Ring-Necked Duck Northern Pintail
Sora Ruby-Crowned Kinglet
Spotted Sandpiper Great White Egret
Townsend’s Warbler Orange-Crowned Warbler
Yellow-Rumped Warbler Townsend’s Warbler
Great Blue Heron Green Heron
American Wigeon European Starling
Downy Woodpecker Northern Flicker
Song Sparrow Lincoln’s Sparrow
White-Crowned Sparrow Golden-Crowned Sparrow
Canada Goose Cackling Goose
Red-Breasted Sapsucker Dunlin
Marsh Wren Black Phoebe
Lesser Scaup Greater Scaup
Pied-Billed Grebe Double-Crested Cormorant
Western Scrub Jay Cedar Waxwing
Gadwall Dark-Eyed Junco
Dowitcher Wilson’s Snipe
Glaucous-Winged Gull Rock Pigeon
Double-Crested Cormorant American Crow
Red-Breasted Nuthatch White-Breasted Nuthatch
Pacific Wren Bewick’s Wren
Varied Thrush
Figure 3.15 The wapato is a secondary plant at the Talking Water Gardens.
Photo: Talking Water Gardens staff. License: reproduced with permission.
Wastewater performance
Background concentrations
The drainage from healthy, natural, free water surface wetlands contains BOD, TSS,
nutrients, metals and other substances. These naturally occurring concentrations are
called background concentrations. It is not reasonable to expect constructed treat-
ment wetlands to achieve lower concentrations. Background concentrations are
variable depending on the character of the soil, vegetation, climate, etc., but the
following values are reasonable estimates: BOD – 3 mg/L, TSS – 8 mg/L, TKN (organic
nitrogen and ammonia) – 2 mg/L, total phosphorus – 1 mg/L (Wallace, 2006).
BOD and TSS

Free water surface wetlands are quite effective in the removal of BOD and TSS from
wastewater. However, since the wetlands are used by wildlife to a greater extent than
other wetland types, these contaminants are also generated within the wetland. It is
important to understand that especially for this wetland type, but also for the others,
that complete removal of BOD and TSS is not possible. Instead, a concentration
consistent with naturally occurring wetlands is the treatment goal. A removal plateau
for BOD occurs after about 3 days of hydraulic residency time. Wastewater from drain-
age pipes below agricultural fields, conventional wastewater treatment plant secondary
effluent and stormwater run-off may contain very low levels of BOD. This low carbon
condition can limit the capacity of the free water surface treatment wetland to remove
nitrate (Kadlec, 2009b).
Nitrogen
Total nitrogen removal by free water surface wetlands is quite good, as both aerobic
conditions at the water surface and anoxic conditions in the detritus on the bottom
of the wetland encourage ammonification and nitrification at the surface and
denitrification on the bottom. However, organic matter is added by the wetland itself,
as leaves and stems die in the autumn and fall into the water. Wildlife is also often a
significant source of nitrogen contamination. Nevertheless, free water surface treatment
wetlands are economically viable and effective for the achievement of tertiary water
quality with regard to total nitrogen. A total nitrogen decrease of about 40 per cent
is a reasonable expectation and consistent with the Taiwan wetland presented above.
Plants incorporate nitrogen in to the roots, leaves and stems. The amount of nitrogen
removed by plants is less than 10 per cent when treatment is of domestic sewage.
Therefore, plant harvesting to remove nitrogen is generally not highly beneficial.
However, if the wetland has lower inputs of organic matter, ammonia and nitrate,
then harvesting the plants would be more significant. Most plants take up ammonia
preferentially but use nitrate when this is not available (Kadlec, 2009b).
Phosphorus
As with nitrogen, free water surface wetlands are often used to meet increasingly
stringent phosphorus discharge standards. This is often possible at a much lower cost
than is possible using conventional wastewater treatment plant technologies. In the
example of the free water constructed wetland in Taiwan presented above, phosphorus
was reduced from 14.5 to 4.2 mg/L (a 69.8 per cent decrease).
Pathogenic bacteria
There are many mechanisms for the removal of pathogenic bacteria from wastewater
in free water surface wetlands. However, this is not an issue when secondary effluent
is disinfected before it is delivered to the wetland for tertiary treatment. Again, the
presence of wildlife will contribute coliform bacteria to water. Nevertheless, during
secondary treatment, free water surface treatment wetlands can reduce pathogenic
bacteria to meet recreation or irrigation use standards, as demonstrated in the example
of the Taiwanese wetland described earlier. Where free water surface wetlands have
been used in the United States to achieve secondary treatment, the reduction of coliform
bacteria is generally 99–99.9 per cent (US Environmental Protection Agency, 2002).
Metals
Filtration and sedimentation are the primary mechanisms for the removal of heavy
metals. Therefore, free water surface wetlands are quite effective. Although the
concentration of metals is rarely an issue in the treatment of domestic sewage,
municipal sewage including industrial clients can include increased metal contam-
ination. More often, heavy metals are associated with stormwater run-off and are
discussed in detail in the chapter on stormwater treatment.
Figure 3.16
Varying water depth allows for a
great diversity of wetland plants as
illustrated in this image of the Boone
Presbyterian Church stormwater
treatment wetland in North Carolina.
In this image of a free water surface
wetland, the great variety of plants
includes sedges, rushes, spikerush,
arrowhead and cardinal flower.
Photo: Wendy Patoprsty. License:
reproduced with permission.
Conclusion
Free water surface wetlands require fairly large land areas but effectively achieve
secondary quality effluent. In this case, at least the initial treatment cell should be
fenced and isolated from the public. Very much more often, this wetland type is used
to achieve tertiary quality effluent from conventional treatment plant discharge. In
this case, the wetlands can generally be accessible to the public without concern for
contact with pathogenic bacteria. The great advantages of free water surface treatment
wetlands are the ancillary recreation, education, scenic and wildlife benefits, which
can be accomplished using careful design (Figure 3.16).
References
Baldwin M. and Proebsting, W. “Checklist S21283451.” 10 January 2015.
Cooper, P. “What can we learn from old wetlands?” Desalination, vol. 246, pp. 11–26, 2009.
Crites, R. “Design and start-up of constructed wetlands for Cle Elum, Washington,” in Pacific
Northwest Pollution Control Association Conference, Sunriver, Oregon, 2001.
Frazier, D., Wagner, R., Huang, T. and Kelly, C. “Hydraulic characterization of the talking
water gardens wetland,” 2012.
Friesen, T. “Higher Water Temperatures Primary Cause of Early Spring Chinook Mortality,”
Oregon Department of Fish and Wildlife, 18 June 2015. [Online]. Available: www.dfw.
state.or.us/news/2015/june/061815.asp. [Accessed: 10 August 2015].
Jarrin, V., Bays, J. S. and Bowyer, J. “Sustainable water reclamation using constructed wetlands:
The Clayton county water authority success story,” 2012.
Kadlec, R. H. “Comparison of Free Water and Horizontal Subsurface Treatment Wetlands,”
Ecological Engineering, vol. 35, no. 2, pp. 159–174, 2009a.
Kadlec, R. H. Treatment wetlands. Boca Raton, FL: CRC Press, 2009b.
Neal, M. and Preston, K. “Talking waters garden,” 23 October 2015.
Ou, W. S., Lin, Y.-F., Jing, S.-R. and Lin, H.-T. “Performance of a constructed wetland-pond
system for treatment and reuse of wastewater from campus buildings,” Water Environment
Research, vol. 78, no. 12, pp. 2369–2376, November 2006.
US Environmental Protection Agency, “Onsite wastewater treatment systems manual,”
EPA/625/R-00/008, 2002.
US Environmental Protection Agency, “Case study: Albany-Millersburg talking water gardens:
A value-focused approach to improving water quality,” EPA-832-F-12-022, 2012.
Wallace, S. Small-scale constructed wetland treatment systems. London: Water Environment
Research Publishing, 2006.
Wu, H., Zhang, J., Ngo, H. H., Guo, W., Hu, Z., Liang, S., Fan, J. and Liu, H. “A Review on
the Sustainability of Constructed Wetlands for Wastewater Treatment: Design and
Operation,” Bioresource Technology, vol. 175, pp. 594–601, January 2015.
Zhang, L. “Effect of limited artificial aeration on constructed wetland treatment of domestic
wastewater,” Desalination, vol. 250, pp. 915–920, 2010.
4
Horizontal subsurface
flow treatment wetlands
Introduction
German scientist Dr Kathe Seidel was the first researcher to test the capacity of plants
to grow in wastewater. She developed water treatment systems using vegetation
grown in sludge and wastewater from industrial and agricultural sources. Between 1952
and 1956, this German scientist conducted many experiments including filtering
untreated sewage through sandy media planted with macrophytes. In the mid-1960s,
she and collaborator Reinhold Kickuth developed horizontal subsurface flow treat-
ment systems, which initially used soil as the filtration media. Despite this development,
the first horizontal subsurface flow system was not applied to the treatment of
municipal wastewater until it was implemented in Liebenburg–Othfresen, Germany,
in 1974 (Vymazal, 2011). Between 1983 and 1989, an intensive study of two horizontal
subsurface flow treatment wetlands was conducted in Weinitzen and Mannersdorf,
Austria. This transfer of German technology to Denmark led to the construction of
80 horizontal subsurface flow systems shortly after 1987, while in 1987, two systems
based on the German and Danish experience were constructed in the United Kingdom.
These initial demonstrations resulted in more than 20 horizontal subsurface flow
wetlands by 1986. Two of the first horizontal subsurface flow treatment wetlands
constructed in the Czech Republic were constructed in 1991 and 1992 and are still in
operation (Vymazal, 2009). Today, horizontal subsurface flow wetlands treat domestic
wastewater for tens of thousands of small communities in Europe (Vymazal, 2011).
In the United States, facultative and aerated sewage lagoons are often used to
treat municipal wastewater in rural towns. Facultative lagoons can be quite effective
but are single-purpose facilities. A comparative study showed that the land area
requirement for a facultative lagoon was three times greater than for a horizontal
subsurface flow wetland when water was treated to the same standard. This
comparison applied to a tropical climate. Since in a temperate climate, cold weather
causes treatment problems in facultative lagoons, the required area would likely be
even greater (Mburu et al., 2013).
Design
Since the 1970s, horizontal subsurface flow wetlands have been monitored and their
designs refined to produce wastewater treatment facilities that are very simple to
HORIZONTAL SUBSURFACE FLOW TREATMENT WETLANDS 51
construct and operate. Plants and gravity produce the energy to operate the system,
while the subsurface flow virtually eliminates risk of contact with wastewater, the
emergence of mosquitos and unpleasant odours. Similarly, the subsurface flow wet-
lands are insulated from freezing temperatures and not subject to thermal stratification,
which encourages uniform distribution of the water across the wetland (Pozo-Morales
et al., 2014). These wetlands can be aesthetically integrated into the landscape (Figure
4.1). The greatest design challenge is preventing rapid clogging of the pore spaces
between the gravel particles that are used as the filter media. Many of the design prac-
tices discussed below focus on minimizing this problem. As with all treatment
wetlands, effective and reliable pretreatment (primary treatment) of domestic sewage
is critical for the longevity of the horizontal subsurface flow wetlands. Ideally, this
should be accomplished with two or three chamber septic tanks with effluent filters.
However, Imhoff tanks, sewage lagoons and other systems have been successfully
incorporated as the pretreatment step.
The typical purpose of the horizontal subsurface flow constructed wetland is to meet
secondary water quality standards for municipal wastewater. Size and depth
recommendations are intended to provide the time and area for bacteria attached to
the filter media to consume or transform the organic matter, pathogenic bacteria and
other contaminants present in municipal wastewater. Careful specifications for influent
distribution, shape and flow direction of the wetland, and the size of the filter material
reduce the impact of clogging by organic and inorganic material (Figure 4.2).
Figure 4.1 The plot of Common Reed in the middle ground of this image reveals a horizontal flow
constructed wetland. It reinforces the rural landscape aesthetic. This image shows the horizontal
subsurface flow wastewater treatment wetland for Krovi, Czech Republic (500 population equivalent).
Photo: Jan Vymazal, reproduced with permission.
52 HORIZONTAL SUBSURFACE FLOW TREATMENT WETLANDS
Figure 4.2 This section/perspective shows the coarse gravel inlet trench and the filter bed. The
vegetation insulates the bed in the winter and maintains porosity in the filter media. Note that the
earth slopes away from the basin to avoid stormwater inflows. The bottom of the basin is lined to
prevent infiltration.
Image: Zhenyu Liu and Gary Austin, 2016.
Variations or additions to the commonly applied design features are in response

to cold climate settings or efforts to extend the performance of the horizontal
subsurface flow wetland to achieve tertiary water quality standards or treat wastewater
from industrial or stormwater sources.
Design challenge
The greatest design challenge when planning a horizontal subsurface flow wetland
for the treatment of water with high organic content (human domestic wastewater,
food processing wastewater and livestock wastewater) is managing the rate and
extent of media clogging. The existing installations have fulfilled the goal of simple,
reliable wastewater treatment of domestic sewage to secondary standards.
Size
The area required by a proposed horizontal subsurface flow wetland depends on the
rate at which biological and chemical processes transform target organic and inorganic
substances into benign or less problematic substances. Temperature, the amount of
contaminants and the volume of contaminated water all influence the required wetland
size. The amount of pollutants (expressed as weight per area per day) entering the
wetland is one way of estimating the size of horizontal subsurface flow wetlands.
The US Environmental Protection Agency (EPA) recommends a maximum of 6 g/m2
of biological oxygen demand (BOD) per day to achieve an effluent concentration of
30 mg/L, and 20 g/m2 of total suspended solids (TSS) per day to achieve the same
effluent concentration (US Environmental Protection Agency, 2000).
The formula below is commonly used to estimate the required size of horizontal
subsurface flow wetlands treating domestic wastewater. This example focuses on the
inlet and outlet concentrations of BOD5 but in other situations, such as treatment of
stormwater, other contaminants would be targeted (a different rate constant would
also be needed).
Qd(ln Cin – ln Cout)
Ah = –––––––––––––––––––
KBOD
where
Ah is the surface area of the bed (m2),
Qd is the average flow (m3 day),
Cin is the influent BOD5 (mg/L),
Cout is the effluent BOD5 (mg/L), and
KBOD is the rate constant (m day) (Vymazal, 2005).
For municipal wastewater, BOD5 and TSS are the targeted contaminants; the
accepted KBOD rate constant based on field measurements of many horizontal
subsurface flow wetlands is 0.1 m per day (36.5 m per year). Different rate constants
apply to ammonium, nitrate, phosphorus and other substances (Vymazal, 2005).
The following list shows the typical standards applied to estimating the size of
horizontal subsurface flow wetlands.
Hydraulic loading rate = 0.03 m3 wastewater inflow/m2/day

Suspended solids loading rate = 20 grams TSS in wastewater inflow/ m2/day
Organic loading rate = 6.0 grams BOD5 in wastewater inflow/m2/day
Hydraulic retention time = 2.40 days (water volume m3/wastewater flow m3/day)
Velocity of output flow = 0.29 m/second (Kadlec, 2009; Pozo-Morales et al., 2013)
A large number of horizontal subsurface flow wetlands designed to treat municipal

wastewater have been constructed and monitored in various climates using the formula
above and other size estimation methods. This experience indicates that about 54 square
feet (5 m2) of horizontal subsurface flow wetland area for each population equivalent
(PE) is required to meet US EPA standards in summer and winter. This guideline
assumes 60 grams of BOD5 per person and is thus often called a PE (Vymazal, 2011),
as sometimes other sources of BOD are present, as in the case of a combined sewer
system. This standard is useful to planners and landscape architects for the initial size
estimation of horizontal subsurface flow wetlands (Vymazal, 2005) required to meet
a standard secondary effluent with an average concentration of 30 mg/L for BOD and
TSS. More accurate estimates of treatment effectiveness by an environmental engineer
would refine the initial size established by the designer or planner.

The amount of time that water remains in a treatment wetland is called residency time.
The rate of water inflow determines the residency time and the amount of pollutant
reduction. The residency time required to meet standards for various contaminants
differs greatly. Therefore, the inflow rate can be used to estimate the size of horizontal
subsurface flow wetlands for sufficient treatment of different pollutants.
The inflow rate of water is called the hydraulic loading rate, and it generally ranges
from 0.8 to 7.9 inches (2–20 cm) per day (Zhang et al., 2014). This is the depth of
water covering the area of the wetland. For example, at a hydraulic loading rate
of 1.8 inches (4.5 cm) per day for a 1-acre (0.4 hectares) wetland, the equivalent volume
is 48,000 gallons (182 m3) per day. Second only to bed depth, the hydraulic load-
ing rate (increasing the hydraulic loading rate also increases the aerial organic
loading rate) is the most important factor predicting treatment performance. A study
applied three hydraulic loading rates (2, 3.6 and 4.5 cm per day). These rates, in this
study, correlated with the nominal hydraulic retention times of 5.5, 3.0 and 2.4 days.
In the case of domestic sewage, the BOD and TSS load is relatively constant; there-
fore, increasing the inflow rate increases the contaminant load. In general, as the
hydraulic loading rate increases, so does the BOD concentration in the effluent (García
et al., 2005).
Shape
Recommendations for the shape of the horizontal subsurface flow wetland bed are
intended to reduce the risk of clogging the filter media near the inlet of the wetland.
Therefore, rectangular beds where the inflow is distributed along the longest side are
recommended. A 1:2 length-to-width ratio is often recommended for horizontal
subsurface flow treatment cells. The greater width reduces the rate of clogging near
the inlet zone, as the load is distributed over a broader cross section (Figure 4.3). As
clogging is reduced, a more uniform flow through the wetland is likely. When the
filter media is fine (less than 4 mm), the treatment bed should be wider than it is long
(Kadlec, 2009). When coarser filter media (10–20 mm) is used, the shape of the wetland
Figure 4.3 Horizontal subsurface flow wetlands are generally rectangular with cobble inlet and outlet
zones. They are densely vegetated and can be designed to be as beautiful as functional.
Photo: Jan Vymazal, reproduced with permission.
is less critical. In practice, the bed shape varies widely, and there are other measures
that can be implemented to lessen clogging risk. One study of properly functioning
horizontal subsurface flow wetlands found that the bed shape was less important than
bed depth, hydraulic loading rate and media size in predicting treatment effectiveness
(García et al., 2005).
Depth
Water depth is more important than bed shape, media size and hydraulic loading rate
for improved horizontal subsurface flow wetland reduction of BOD, chemical oxygen
demand (COD), ammonia and dissolved reactive phosphorus. Research demonstrates
that horizontal subsurface flow beds in a warm climate that are 10.5 inches deep
(0.27 m) perform better than 20-inch deep (0.5 m) beds. This is thought to be the
result of greater oxygen availability within the media (García et al., 2005). Limiting
the anaerobic zone within the treatment bed may reduce conditions where methane
is produced. This condition seems to be related to clogging problems. Of course,
methane is also a very potent greenhouse gas. Most horizontal subsurface flow
wetlands have been built with a bed depth of 0.6 m or less, but a 0.8 m deep bed is
not uncommon (Kröpfelová and Vymazal, 2008). The 0.6 m depth is based on the
assumption that this is the maximum root penetration of Phragmites australis, the
most commonly planted species in European constructed wetlands. The initial estimate
of the depth of root penetration of Phragmites in the horizontal subsurface flow bed
was inaccurate. Research shows that more than 90 per cent of the root biomass is
located in the upper 12 inches (30 cm) of the bed (Kadlec, 2009).
Pretreatment
Domestic wastewater should be pretreated to reduce the size of the particles reaching
the constructed wetland. Several pretreatment processes are used and their effective-
ness varies. Especially for horizontal subsurface flow treatment wetlands, it is critical
that the pretreatment removes very small particles to extend the life of the wetland
by reducing the rate of clogging the filter media. The simplest pretreatment passes the
raw sewage through a series of screens. Using a series of coarse, fine and micro screens
removes particles more than 0.025 mm. Imhoff and septic tanks typically remove
particles larger than 0.014 mm (Kröpfelová and Vymazal, 2008). Multi-chamber septic
tanks with effluent filters are recommended when septic tanks are the pretreatment
choice.
Organic loading
The section above on size mentioned that the amount, or load, of contaminants is an
important factor to consider in the design of horizontal subsurface flow wetlands. For
domestic wastewater treatment, the loading of organic material is of most concern
when considering the longevity of the wetland and the ability to meet secondary
treatment standards. Although the US EPA recommends a maximum of 6 g/m2 of BOD
per day (US Environmental Protection Agency, 2000), experience with well-designed
horizontal subsurface flow wetlands suggests that rates as high as 11 g/m2 can be
supported while meeting the 30 mg/L secondary effluent standard (Vymazal, 2009).
Clogging
As noted earlier, design measures to prevent rapid clogging of horizontal subsurface
flow wetlands are among the most important efforts to consider. Several design,
operation and maintenance factors can contribute to clogging. Therefore, there are
also many ways to manage this potential problem. A study of 13 horizontal subsurface
flow wetlands in Spain identified eight possible factors responsible for clogging:
1) organic solids on the surface or within the media, 2) accumulation of suspended
solids within the filter media pores, 3) filter media that is too fine, resulting in rapidly
clogged pores, 4) differing sizes of filter media within the treatment bed, 5) non-uniform
distribution of water in the inlet or outlet, 6) growth of microorganisms, 7) precipitation
and deposition of solids caused by chemical interactions between the wastewater and
the filter media, 8) growth of rhizomes and roots (Turon et al., 2009). Many of these
clogging factors can be addressed by the design, careful specification of materials,
inspections during the installation period and routine maintenance.
About 50 per cent of the total organic carbon in domestic wastewater is eliminated
near the inlet (in the first quarter of the wetland length) due to sedimentation and
filtration (García et al., 2005). Biofilm develops quickly in this zone forming a gel-
like coating around the media particles. This gel greatly reduces the size of the pore
spaces near the wetland inlet. Bacteria are part of the biofilm, and these organisms
degrade organic material creating a less degradable sludge as a by-product. Inorganic
material in the wastewater is also deposited near the inlet, exacerbating the loss in
media porosity. Eventually, the media will become clogged with this organic and in-
organic material effectively damming the wetland. Initially, this causes water to flow
to the surface and over the clogged zone. Then the water drops back into the media
until a new clog is created downstream. To delay clogging, a BOD loading of the
cross-sectional area of the wetland inlet with less than 250 g/m2 per day is recom-
mended for filter media, where D10 is greater than 4 mm (Kadlec, 2009). Lower loading
rates are required to avoid clogging when finer grain material media is specified.
The character of the clogging material can vary. A study of the distribution of
clogging material (sludge) was conducted in 2006 to determine its distribution and
constituents in six horizontal subsurface flow wetlands in Spain. The media was 0.35
inches (D60 = 9 mm, coefficient of uniformity = 1.8, initial porosity = 40 per cent),
which is smaller than normally recommended today in the United States. The waste-
water was pretreated by screening and septic tanks. The clogging material in this study
was composed primarily of mineral substances (approximately 90 per cent). Most of
this may have come, not from deposition, but from acid disintegration of the filter
media, which had a high calcium content. Another component of clogging material
is poorly biodegradable substances, such as lignocellulose, left behind after aerobic
or anaerobic digestion (Llorens et al., 2009).
While clogging can be attributed to more than one cause, delaying the rate and
extent of media clogging has generated five design responses. The first is to distribute
the pretreated wastewater over the largest possible area in the inlet zone; this has led
to the general shape recommendation discussed above. The second anti-clogging
recommendation is to increase the size of the media in the inlet zone and in the main
bed to increase the size of the pore spaces between the particles to provide greater
storage of sediment and greater capacity for biofilm without eliminating all porosity.
Therefore, the size and material specification for the inlet zone presented above
will reduce clogging. Gravel 0.6 inches (16 mm) in diameter or a size distribution
between 8 and 16 mm is now commonly recommended as a positive balance between
control of clogging and treatment performance in Europe (Vymazal, 2011; Turon,
Comas and Poch, 2009). The recommended size of the filter media adopted in the
United States is 20–30 mm.
Third, it is clear that subsurface distribution of the wastewater in the inlet zone
results in fewer clogging problems than surface discharge (Figure 4.5). The fourth
measure is not related to the design of the wetland itself. It is the improvement of the
pretreatment of the raw sewage to reduce the organic load and the use of effluent
filters in two or three chamber septic tanks in order to reduce the amount and size of
inorganic and organic particles reaching the horizontal subsurface flow wetland. Two
related measures include the avoidance of construction errors that contaminate the
filter media and diversion of all stormwater surface flow around the wetland. Fifth,
the hydraulic and organic loads that are applied to the wetland influence whether
clogging occurs. These are both a design and operations issue. Designers must measure
or carefully calculate the hydraulic and BOD load with an understanding that the
implementation of low-flow toilets and urinals in recent decades causes changes to
these treatment parameters. Applying loads larger than the wetland was designed to
treat, over a long period, can result in clogging problems.
Materials
Waterproof membrane
Excavating a basin and lining it with a waterproof membrane is the first step
in constructing horizontal subsurface flow wetlands. The liner prevents infiltration of
contaminated water and ranges between 0.5 mm and 1.5 mm. It may be protected
on each side by a geotextile fabric to prevent punctures or tears due to stones or other
material in the soil or from the filter material itself. Bentonite or other dense clay is
also a suitable water barrier; however, its use requires side slopes that are less steep
than is possible with waterproof membranes (Figure 4.4).
Figure 4.4
This horizontal subsurface flow
wetland under construction
shows the creation of a thick clay
liner to waterproof the treatment
cell.
Photo: B. Koelsch SuSanA
Secretariat, license CC-BY-2.0.
Inlet/outlet media
The inlet and outlet zones of horizontal subsurface flow wetlands are usually com-
posed of coarse stone. Various authors recommend a different size for the stones.
However, sizes ranging between 1.5 and 3 inches are common recommendations. As
noted earlier, this coarse aggregate in the inlet zone is critical. A wider inlet zone
also supports the sedimentation and filtration processes. The US EPA recommends
about 6 feet (2 m) in length for the inlet zone and 3 feet (1 m) for the outlet zone
(Figure 4.5) (US Environmental Protection Agency, 2000). However, this guideline
might be adjusted depending on the anticipated load of organic and inorganic matter
in the wastewater.
Filter media
Filter media actually performs several functions. Of course, it is intended to filter the
water causing sedimentation of particles, but it is also the support for the wetland
vegetation. A critical function of media is to serve as a substrate for the growth of a
biofilm including bacteria that performs the biochemical treatment of the wastewater
(Figure 4.6). Finally, the filter media causes the water to flow more or less uniformly
through the wetland. Although not part of the treatment operation, the filter media
includes a layer that is above the water level. This prevents the possibility of contact
with the wastewater, if the treatment area is accessible by the public. This unsaturated
zone also forms an oxygen reservoir.
Figure 4.5 This transparent section/perspective shows that the red distribution piping (often
perforated polyvinyl chloride pipe) is set higher than the yellow collection piping. The inlet, outlet and
filter beds are fully vegetated with strong plants, such as Phragmites, Typha, Scirpus, Carex and
Juncus. As with the free water surface wetland, the outlet structure is a vault buried in the berm and
contains a water level control.
Image: Zhenyu Liu and Gary Austin, 2016.
Figure 4.6
The gravel media has been placed in
the first of a long sequence of cells in
this horizontal subsurface flow
constructed wetland.
Photo: B Koelsch SuSanA Secretariat,
license CC-BY-SA-2.0.
The filter media is often not uniform in size, but the range in size is controlled
by a specification. The media size is expressed by the percentage of particles passing
through screens with different sized openings. For example, gravel designated as
10 mm (0.4 inches), D60 indicates that 60 per cent of the particles fall through a screen
with a 10 mm opening. It is critical that the media is washed well to eliminate clay
and silt particles, which would reduce the porosity of the filter media. In 2000, the
US EPA recommended filter media of 0.75–1 inches (20–30 mm) (US Environmental
Protection Agency, 2000). This is the culmination of a series of filter size recommend-
ations that increased with each iteration.
The size of the media can affect aspects of the performance of the wetland. One
study demonstrated that the biomass (combined above and below ground biomass)
produced by the wetland plants (Phragmites australis) is almost three times greater
when the media size is D60 3.5 mm (0.14 inches) than when it is D60 10 mm (0.4
inches). Similarly, the finer of these two sizes of filter particles resulted in a better
removal of contaminants. This is probably due to the greater surface area on which
bacteria can attach. Also, the pore spaces between the finer media are smaller, creating
better contact between the bacteria and contaminated water. A bed of this fine media
has almost twice the surface area compared to a bed composed of the 10 mm media
(García et al., 2005).
The dichotomy between the desire for fine filter material for best treatment
performance and the need to prevent clogging for as long as possible has led to a wide
range of recommendations for filter media size. The US EPA recommends a gravel
media of 0.75–1 inches (20–30 mm) for horizontal subsurface flow wetlands (US
Environmental Protection Agency, 2000), although evidence from horizontal sub-
surface flow wetlands in Europe demonstrates that gravel ranging between 0.3 and
0.6 inches (8–16 mm) provides excellent long-term performance without clogging
problems (Vymazal, 2011).
Filter media and redox potential

Chemical changes in a biological system are dependent on the availability of electrons.
Molecules that lose electrons become oxidized, while those that gain electrons are
reduced. This capacity for oxidation and reduction is abbreviated as the redox
potential of the wetland system. When oxygen becomes depleted, specialized organisms
are capable of reducing nitrate and other molecules to satisfy their energy needs.
However, the use of organic material is much more rapid when microorganisms are
operating in an oxygen-rich (aerobic) environments than in anaerobic ones. Aerobic
soils usually have a redox potential of +400 to +700 mV, while submerged soils have
a redox potential ranging from –400 to –700 mV (Kröpfelová and Vymazal, 2008).
In the early stage of the development of horizontal subsurface flow wetlands,
environmental engineers focused on achieving a balance among media size, resistance
to clogging and treatment effectiveness. In recent years, this focus has shifted to
experiments with filter media to improve the quality of the effluent for particular
contaminants, such as phosphorus, or to reduce the cost of the system through the
use of media that is also a local industrial waste product. Natural materials have been
investigated to determine those that are most effective for the removal of phosphorus
(those with high calcium content are effective), while other natural materials were
investigated for adsorption capacity for ammonium. One such study found that the
ammonium adsorption capacity of zeolite (11.6 g/kg) was much higher than that of
volcanic rock (0.21 g/kg). Industrial waste materials have shown promise for the
removal of phosphorus, while others (alum sludge, peat, marl, compost and rice husk)
have been proposed for effective removal of organic material and nitrogen. Similarly,
mixtures of materials, such as sand and dolomite, have demonstrated effectiveness
for the removal of phosphorus. Mixtures of materials, such as gravel, vermiculite,
ceramsite and calcium silicate hydrate, have been studied as media for treating surface
water that is low in nutrients (Wu et al., 2015). Table 4.1 lists natural, industrial waste
and artificial materials that were studied between 1997 and 2012.
Case study
Results of long-term monitoring of horizontal subsurface flow wetlands are becoming
more readily available and serve as examples of high performance and enduring designs.
Table 4.1 Common substrates (Wu et al., 2015)
Natural material Industrial by-product Artificial products

Sand Slag Activated carbon
Gravel Fly ash Light-weight aggregates
Clay Coal cinder Compost
Calcite Alum sludge Calcium silicate hydrate
Marble Hollow brick crumbs Ceramsite
Vermiculite Moleanos limestone
Dolomite Wollastonite tailings
Limestone Oil palm shell
Shell
Shale
Peat
Wollastonite
Maerl
Zeolite
A study of a horizontal subsurface flow wetland, constructed in 1993 in a temperate

climate, treating combined sanitary and stormwater sewage for a community of 92
people, continues to demonstrate outstanding performance with no evidence of
clogging problems. In this wetland, the treatment bed is 0.5 m deep, filled with
4–22 mm gravel and planted with common reed (Phragmites). The wetland is 60 feet
wide (20 m) and 85 feet long (26 m). The average BOD5 inflow concentration is
85 mg/L, although the inflow concentration varied widely between 10 and 370 mg/L,
as expected in a combined sewer situation. The average BOD5 outflow is very low at
3.3 mg/L. The organic loading rate is lower than average at 2.78 grams BOD5/m2 per
day (annual average) compared to the US EPA recommended maximum of 6 grams
(Vymazal, 2014).
Innovative design
A design that targets the clogging issue has been built as a full-scale demonstration
at the Experimental Wastewater Treatment Plant of Carrión de los Céspedes in
Seville, Spain. This horizontal subsurface flow wetland is a long narrow trench (length
30 feet (10 m); width 3 feet (1 m); depth 31 inches (0.80 m)), and is built to emphasize
a plug flow movement of water, where the inflow simply pushes before it the water
that is already in the trench. The trench is installed as a series of identical 10-m long
sections. Sections are added until the target water quality is achieved. Figure 4.7 shows
the trench during construction along the right side of the treatment facility (Pozo-
Morales et al., 2014).
Figure 4.7
Aerial view of a trench horizontal subsurface
flow wetland under construction. The long and
narrow wetland is on the edge of the treatment
facility. The very narrow space occupied by the
wetland suggests that it could be incorporated
into the building setback zones often stipulated
for residential development.
Photo: Centre for New Water Technologies
(CENTA) R & D & I, 2008, reproduced with
permission.
Very coarse stone 4–3 inches in size (200–150 mm) is placed in a long wedge
extending from the trench inlet to the outlet (Figure 4.9). The inlet end of the wedge
fills the trench to the top. This wedge is overlaid with a mirrored wedge of somewhat
finer filter media, 3–2 inches (150–50 mm), which is again much larger than commonly
specified for horizontal flow wetlands (Figure 4.8). The purpose of the gravel size and
its placement is to provide larger pore spaces at the inlet, where clogging is more
problematic. Porosity at the inlet is 73 per cent. The smaller pore spaces near the outlet
are intended to optimize contact between the water and bacteria covering the stone
(porosity = 51 per cent). In the typical design, the pore space occlusion is around
66 per cent and it reduces the hydraulic conductivity. Similarly, 28 per cent of the pore
spaces are closed in the outlet of the typical design. However, in this wetland, the inlet
and outlet occlusion are 54 per cent and 12 per cent, respectively. This allows greater
hydraulic loading and greater organic loading than is commonly possible. For example,
the cross-sectional hydraulic capacity of this system is 0.5 m3/m2 per day, which is many
times greater than for typical designs (Pozo-Morales et al., 2014).
This treatment trench was tested with municipal wastewater. Although the trench
is 0.8 m deep, the water depth is limited to 0.45–0.50 m in order to also create an
interior air chamber above the water. The cross-sectional area of the experimental
wetland is small, compared to typical designs that are wider than they are long.
Therefore, the proposed design receives BOD loading of 1,129 grams per m2 of cross-
sectional area per day, rather than the 250 grams of BOD per m2 per day commonly
recommended. Similarly, for the typical design, the recommended amount of total
suspended solids is 2 grams per m2 of cross-sectional area, while for the experimental
design, 10 grams per m2 per day is accommodated. This is possible due to the in-
creased porosity and levels of dissolved oxygen in the wastewater from the design of
a natural forced air circulation technique that is only partially related to the features
associated with clogging prevention, but equally innovative (Pozo-Morales et al., 2014).
Both the configuration of this wetland and the media sizes are contrary to
common design practice. This example suggests that additional research of media size
and distribution may lead to designs with less clogging potential, while enhancing
treatment performance. This treatment trench design did not include plants, which is
a disadvantage if the facility is included in the public landscape, although the
narrowness of the trench would make it easy to conceal it with adjacent plantings
(Figure 4.10). The long narrow trench would be a distinct advantage for a steep site,
as it could follow the contours without the need for extensive cut and fill earthwork.
Figure 4.8
This model of the linear horizontal
subsurface flow wetland illustrates
the overlapping wedges of stone with
the coarse stone filling the inlet (left).
On top of the wedges, a layer of
gravel establishes the oxygen
chamber supported with wells and
chimneys to increase the amount of
dissolved oxygen in the top section of
the trench. The uppermost layer is a
fine sand cap.
Image: Zhenyu Liu and Gary Austin,
2016.

This image shows the first wedge of coarse The horizontal subsurface flow trench is capped
stone. The next step is to overlay the first wedge with a layer of fine gravel and then a top layer of
with a wedge of somewhat finer gravel. sand. The vertical pipes shown in the image are
Photo: Centre for New Water Technologies (CENTA)
test ports, while the coarse stones across the
R & D & I, 2008, reproduced with permission.
trench are the location of the aeration chimneys
and wells that increase the dissolved oxygen
levels of the wastewater. The dissolved oxygen
at the outlet is higher than at the inlet.
Photo: Centre for New Water Technologies (CENTA)
R & D & I, 2008, reproduced with permission.
Horizontal subsurface flow wetlands were initially developed, tested and refined for
high performance in the treatment of domestic wastewater (Figure 4.11). However,
during the last decade, agricultural, industrial and other wastewaters have been
treated using the horizontal subsurface flow wetlands. This section addresses domestic
wastewater in detail, while other chapters discuss horizontal subsurface flow wetlands
in industrial and agricultural settings.
The simple construction and low maintenance operation of horizontal subsurface
flow wetlands make them well suited for wastewater treatment at the scale of the single
residence or residential compound. For example, a 194 square feet (18 m2) system
with a PE of 4 (4.5m2 per PE) was studied after 10 years of operation. The system
included a high-efficiency septic tank with vertical baffles and a 12-hour retention time.
The filter bed was composed of coarse sand (1–4 mm) and planted with yellow flag
(Iris pseudacorus) and Siberian iris (Iris sibirica). Monitored from January 2003
through September 2004, the system demonstrated exemplary treatment performance
(Table 4.2). The septic tank greatly reduced the organic content and TSS, reducing
the load on the wetland and the clogging potential. The removal of BOD, COD and
Figure 4.11
This photo shows an elevated effluent
storage tank in the foreground and a
horizontal subsurface flow wetland
planted with Phragmites karka in
Bayawan City in the Philippines.
The background block of plants is a
slightly elevated vertical subsurface
flow wetland stage.
Photo: Jouke Boorsma, SuSanA
Secretariat, license CC-BY-SA-2.0.
TSS was outstanding (Vymazal, 2005). It should be noted that wastewater discharge
permits for secondary quality effluent focus on TSS and BOD. Secondary treatment
is the typical compliance requirement for small communities. Nevertheless, the data
in Table 4.2 documents a moderate reduction in total nitrogen, and moderate
reduction in ammonia.
BOD and TSS

Horizontal subsurface flow constructed wetlands for treatment of domestic sewage
are very effective for the removal of BOD and TSS in both summer and winter as well
as cold climates. Reliable achievements of secondary effluent water quality standards
make this simple treatment landscape highly suitable for rural residences, groups of
homes and small communities.
Table 4.3 illustrates the performance of one of these community systems after 7
years of operation. Built for a PE of 700 in 1992, this 2,500 m2 wetland treats combined
stormwater and wastewater flows. Therefore, water flow is as high as 52,800 gallons
(200 m3) per day. During these high flows, the contaminant concentrations are low
due to dilution. The wastewater is pretreated in an Imhoff tank, although some
residential septic tanks are also in service. The pretreated water is directed to two
horizontal subsurface flow beds operating in parallel. Each of these is followed by a
second bed forming a two-bed series. The filter bed is gravel (0–16 mm) supporting
Table 4.2 Residential horizontal subsurface flow wetland, Zitenice, Czech Republic (50°33′21″ N
14°9′28″ E) (Vymazal, 2005)
Parameter Inflow After pretreatment Outflow Efficiency

(mg/L) (mg/L) (mg/L) (%)
BOD5 373 73 9.7 97
COD 1,118 182 37 97
TSS 639 44 9.1 99
TP 17.1 10.6 10.6 38
NH4+ 59 62 51 14
NO3 0 0 2.9
N org 24.2 9.1 1.1 95
TN 85 72 55 35
BOD = Biological oxygen demand, COD = Chemical oxygen demand, TSS = Total suspended solids, TP =
Total phosphorus, NH4+ = Ammonium, NO3 = Nitrate, Norg = Organic nitrogen, TN = Total nitrogen
a monoculture of common reed (P. australis), although alternating bands of reed canary
grass (Phalaris arundinacea) were originally planted but have been displaced (Vymazal,
2005).
In this situation, the low concentration of organic material would make conven-
tional activated sludge treatment technologies unfeasible. However, this situation is
well suited to horizontal subsurface flow wetlands. Table 4.3 illustrates that BOD5
and TSS are well below the US EPA discharge standard for secondary effluent
(30 mg/L) (Vymazal, 2005).
Nitrogen
Removal of total nitrogen from domestic sewage is moderately successful using
horizontal subsurface flow constructed wetlands. Removal of nitrates is generally very
high due to the low level of dissolved oxygen in the treatment bed, but conversion of
ammonia to nitrate is quite low for the same reason. Since decreases in ammonia and
nitrate concentrations are not generally required for the discharge of secondary
effluent, the modest capacity of horizontal subsurface flow wetlands to remove these
is often not an issue particularly if the effluent is reused for irrigation. Where higher
quality (tertiary) effluent is required, the horizontal subsurface flow wetland can be
combined with vertical subsurface flow or free water surface treatment landscapes.
Phosphorus
Phosphorus removal is largely dependent on the characteristics of the filter media.
Where gravel with high calcium, magnesium or aluminium content is used as the filter
media, then phosphorus removal is optimized.
Pathogenic bacteria
As with BOD and TSS reductions, pathogenic bacteria removal occurs near the
inlet zone primarily in the horizontal subsurface flow wetland. Removal is due to
sedimentation, filtration, biological processes and natural die-off. A 2-log reduction
(99 per cent) of faecal coliform (an indicator of the presence of pathogenic organisms)
Table 4.3 Treatment efficiency of Horizontal Flow Constructed Wetland (HF CW) at Spálené Poŕíčé,
Czech Republic, November 1992–December 2002 (Vymazal, 2005)
Parameter Inflow Outflow Efficiency

BOD5 23.3 (43) 4.6 (3.4) 80
COD 85 (147) 26.1 (11.5) 69
TSS 91 (228) 9.5 (8.0) 90
NH4 11.6 (5.9) 9.4 (5.0) 19
NO3 3.0 (2.9) 1.79 (2.2) 40
TP 2.25 (1.25) 2.09 (1.52) 7
TCa 6.14 (6.47) 5.01 (5.42) 1.1
FSa 4.47 (4.64) 3.62 (4.03) 0.9
Values in mg/L, bacteria in log10 CFU/100 ml, efficiency in per cent for chemical parameters, in log units for
bacteria. Standard deviations in parentheses.
a TC = total coliforms; FS = faecal streptococci.
is typical of horizontal subsurface flow wetlands, but this is usually insufficient to

meet the US EPA standard for recreational contact of 200 colony-forming units (CFU)
per 100 mL (Wallace, 2006). Improved reduction in the number of pathogens occurs
with increasing time. A residency time of about 3 days is required for a 2-log to
3-log reduction (99–99.9 per cent) of faecal coliform (García et al., 2003). This corre-
sponds to a wetland sized for about 3 m2 per PE, which is smaller than the common
recommendation of 5 m2 set for adequate removal of BOD. A study of horizontal
subsurface flow wetlands in Spain demonstrated that shallow beds (0.27 m) were more
effective than deeper beds (0.5 m) and finer media (D60 3.5 mm/0.14 inches) was
more effective that coarser media (D60 10 mm/0.4 inches) for the removal of pathogenic
bacteria (Morató et al., 2014).
Metals
Domestic sewage typically does not contain much heavy metal, and several researchers
found that the metal concentration in horizontal subsurface flow constructed wetlands
did not differ significantly from sediment in naturally occurring wetlands. However,
if the community operates a combined stormwater and sanitary sewer system, then
the amount of metal reaching the treatment wetland can increase significantly. The
case of a horizontal subsurface flow constructed wetland in Flanders, Belgium (built
in 2001 to serve a population of 350), illustrates that metals are removed and
sequestered in the sediment of the wetland. Located at latitude 50°59″ N, longitude
4°28″ E, this 7,000 square feet (650 m2) wetland is preceded by a 5,800 gallon
(33 m3), three-compartment septic tank. The wetland consists of two 164-feet-long
(50 m) by 43-feet-wide (13 m) treatment cells, 19 inches deep (0.5 m) and filled with
fine gravel (5–10 mm diameter). The inlet and outlet zones contain coarse gravel (5–8
cm diameter), and the wetlands have a 0.7 per cent bottom slope. The two cells are
arranged in parallel and fed alternately with water pumped from the septic tank and
distributed evenly in the inlet zone by vertical pipes (Lesage et al., 2007).
After three years of operation, the effluent was tested for metal concentration.
Copper, lead, chromium and nickel concentrations were below the permitted level of
50 ␮g/L; cadmium was present below 1 ␮g/L and zinc was present below 200 ␮g/L.
All of these concentrations meet the surface water standards. However, the metal
concentration in the wetland sediment increased. The zone within 16 feet (5 m) of
the inlet was significantly contaminated with cadmium, copper, zinc and lead.
Dissolved manganese was exported from the study wetland, as was particulate iron.
The manganese in the effluent probably resulted from reactions in the low-oxygen
environment. Therefore, this study shows that the low-oxygen character of horizontal
subsurface flow treatment wetlands and the filtration and sedimentation functions cause
metals to be captured and held in the wetland. These accumulate over time even when
the metal levels in the incoming wastewater are low. This sink could become a source
of environmental pollution, where metals are flushed into waterways if the wetland
is abandoned in the future and its oxygen concentration changes (Lesage et al., 2007).
Warm and cold climate settings

Treatment processes dependent on biological agents improve in warm climates and
fair less nicely during the winter in cold climates. This is true of all types of treatment
wetlands, but a distinct advantage of horizontal subsurface flow wetlands for water
quality improvement is their reliability under very cold climatic conditions. Second-
ary quality effluent can be achieved due to the inherent insulation of the belowground
filter bed and the insulating effect of the standing vegetation. This benefit can be
extended through a number of operating options. There are four demonstrated
measures to further insulate horizontal subsurface flow wetlands: (1) mulch, straw or
fabric blankets can be added to the standing dead plants, (2) water levels can be lowered
in winter to increase the depth of the layer of dry media above it, (3) a layer of ice
above the filter media can be induced by temporarily raising the water level and then
dropping it to establish a dry gravel zone between the ice and the water level, (4) the
wetland can be designed with an especially deep bed of filter media, so that the water
level can be lowered in winter (Kadlec, 2009). A 6-inch layer (15 cm) of mulch over
a horizontal subsurface flow wetland in Minnesota prevents freezing of the wastewater.
This protection is sufficient during brief periods with temperatures as low as –49°F
(–45°C). In southern Canada and northern United States and other high elevation or
northern latitudes, where the average annual temperature is approximately 44°F (7°C),
horizontal subsurface flow wetlands are an appropriate biological wastewater treat-
ment technology (Kadlec, 2009).
A recent study of the cold climate performance of four horizontal subsurface flow
wetlands in the Czech Republic included one in the town of Zásada, 50°41′57″ N,
15°16′13″ E, at an elevation of 1,706 feet (520 m) above sea level. At Zásada, the
wetland was originally composed of two beds with an area of 21,743 square feet (2,020
m2), but in 2004, an additional bed was added to increase the area to 37,673 square
feet (3,500 m2) to serve a population of 620 people connected to a separated sewer
system. The filter media is gravel with 1⁄2-inch diameter, and the bed is planted with
Phalaris (reed canary grass). The average treated volume is 18,545 gallons (70.2 m3)
per day with a hydraulic loading rate of 2 cm per day. The wetland was monitored
from 1995 through 2013 for this study (Vymazal and Březinová, 2014).
The average concentration of BOD5 in the influent was 247 mg/L and 239 mg/L
in summer and winter, respectively. The effluent concentration was 11 mg/L and 9.8
mg/L in summer and winter, respectively. The standing vegetation is not harvested in
the fall. This along with snow cover usually prevents freezing of the wetland water
surface. This and three other monitored cold climate, horizontal flow wetlands
performed similarly, with a BOD5 removal rate of between 88 per cent and 94 per
cent. Removal of TSS was also high. The TSS influent concentration average of 272
mg/L was reduced to an average of 10.5 mg/L in the Zásada wetland. In all four
wetlands studied, there was no statistical difference between summer and winter
efficiency for BOD and TSS. Ammonium in the effluent was limited by code at Zásada.
The maximum discharge concentration was set at 40 mg/L, which was exceeded only
once during the 18-year study. The annual average ammonia concentration was set
as 20 mg/L maximum. At Zásada, the average annual inflow concentration was 34.5
mg/L, while the effluent concentration annual average was 16.1 mg/L. The removal
efficiency for ammonia is 53.3 per cent. This performance is somewhat higher than
expected from a horizontal subsurface flow wetland. Similarly, total phosphorus
removal at Zásada was high at 59 per cent (Vymazal and Březinová, 2014).
Conclusion
The tens of thousands of horizontal subsurface flow wetlands in a variety of climates
attest to the simplicity and reliability of this constructed landscape for the treatment
of domestic sewage and achievement of secondary effluent quality. We will discuss in
another section the benefits of combining this wetland type with others to achieve
advanced treatment of municipal wastewater and its use to treat industrial pollutants.
References
García, J., Aguirre, P., Barragán, J., Mujeriego, R., Matamoros, V. and Bayona, J. M. “Effect
of key design parameters on the efficiency of horizontal subsurface flow constructed
wetlands,” Ecological Engineering, vol. 25, no. 4, pp. 405–418, 2005.
García, J., Vivar, J., Aromir, M. and Mujeriego, R. “Role of hydraulic retention time and granular
medium in microbial removal in tertiary treatment reed beds,” Water Research, vol. 37,
no. 11, pp. 2645–2653, June 2003.
Kadlec, R. H. “Comparison of free water and horizontal subsurface treatment wetlands,”
Ecological Engineering, vol. 35, no. 2, pp. 159–174, 2009.
Kröpfelová L. and Vymazal, J. Wastewater treatment in constructed wetlands with horizontal
sub-surface flow. Dordrecht: Springer, 2008.
Lesage, E., Rousseau, D. P. L., Meers, E., Tack, F. M. G., and De Pauw, N. “Accumulation
of metals in a horizontal subsurface flow constructed wetland treating domestic wastewater
in Flanders, Belgium,” Science of the Total Environment, vol. 380, no. 1–3, pp. 102–115,
July 2007.
Llorens, E., Puigagut, J. and García, J. “Distribution and biodegradability of sludge accumulated
in a full-scale horizontal subsurface-flow constructed wetland,” Desalination Water
Treatment, vol. 4, no. 1–3, pp. 54–58, April 2009.
Mburu, N., Tebitendwa, S. M., van Bruggen, J. J. A., Rousseau, D. P. L. and Lens, P. N. L.
“Performance comparison and economics analysis of waste stabilization ponds and
horizontal subsurface flow constructed wetlands treating domestic wastewater: A case study
of the Juja sewage treatment works,” Journal of Environmental Management, vol. 128,
pp. 220–225, October 2013.
Morató, J., Codony, F., Sánchez, O., Pérez, L. M., García, J., and Mas, J. “Key design factors
affecting microbial community composition and pathogenic organism removal in horizontal
subsurface flow constructed wetlands,” Science of the Total Environment, vol. 481, pp. 81–89,
May 2014.
Pozo-Morales, L., Franco, M., Garvi, D., and Lebrato, J., “Influence of the stone organization
to avoid clogging in horizontal subsurface-flow treatment wetlands,” Ecological Engineering,
vol. 54, pp. 136–144, May 2013.
Pozo-Morales, L., Franco, M., Garvi, D. and Lebrato, J. “Experimental basis for the design of
horizontal subsurface-flow treatment wetlands in naturally aerated channels with an anti-
clogging stone layout,” Ecological Engineering, vol. 70, pp. 68–81, September 2014.
Turon, C., Comas, J. and Poch, M. “Constructed wetland clogging: A proposal for the
integration and reuse of existing knowledge,” Ecological Engineering, vol. 35, no. 12,
pp. 1710–1718, December 2009.
US Environmental Protection Agency, “Constructed Wetlands Treatment of Municipal
Wastewaters,” Cincinnati Ohio, EPA/625/R-99/010, September 2000.
Vymazal, J. “Horizontal Sub-surface Flow and Hybrid Constructed Wetlands Systems for
Wastewater Treatment,” Ecological Engineering, vol. 25, pp. 478–490, 2005.
Vymazal, J. “Horizontal sub-surface flow constructed wetlands Ondřejov and Spálené Poříčí
in the Czech Republic – 15 years of operation,” Desalination, vol. 246, pp. 226–247, 2009.
Vymazal, J. “Constructed wetlands for wastewater treatment: Five decades of experience,”

Environmental Sciences Technology, vol. 45, no. 1, pp. 61–69, January 2011.
Vymazal, J. “Long-term performance of constructed wetlands with horizontal sub-surface
flow: Ten case studies from the Czech Republic,” Ecological Engineering, vol. 37, no. 1,
pp. 54–63, 2011.
Vymazal, J. “Long-term Treatment Efficiency of a Horizontal Subsurface Flow Constructed
Wetland at Jimlikov, Czech Republic,” Environmental Engineering and Management
Journal, vol. 13, no. 1, pp. 73–80, 2014.
Vymazal J. and Březinová, T. “Long term treatment performance of constructed wetlands for
wastewater treatment in mountain areas: Four case studies from the Czech Republic,”
Ecological Engineering, vol. 71, pp. 578–583, October 2014.
Wallace, S. Small-scale constructed wetland treatment systems. London: Water Environment
Research, 2006.
Wu, H., Zhang, J., Ngo, H. H., Guo, W., Hu, Z., Liang, S., Fan, J. and Liu, H. “A review on
the sustainability of constructed wetlands for wastewater treatment: Design and operation,”
Bioresource Technology, vol. 175, pp. 594–601, January 2015.
Zhang, D. Q., Jinadasa, K. B. S. N., Gersberg, R. M., Liu, Y., Ng, W. J. and Tan, S. K.
“Application of constructed wetlands for wastewater treatment in developing countries –
A review of recent developments (2000–2013),” Journal of Environmental Management,
vol. 141, pp. 116–131, August 2014.
5
Vertical subsurface flow
constructed wetlands
Introduction
The preceding chapter introduced the work of Dr Kathe Seidle, who, in the early 1950s,
conducted experiments in which wetland plants were grown in industrial and dairy
wastewaters. In the 1960s, she introduced a system of draining septic tank effluent
vertically through sandy soils planted with common reed in order to oxygenate the
water (Vymazal, 2011). These experiments led to hybrid wetlands containing vertical
flow and horizontal flow cells in series. It was only in the 1990s that vertical subsurface
flow wetlands were developed on a full scale in France and the United Kingdom as
part of these hybrid systems. During the first few years of the twenty-first century,
full-scale experiments in Denmark established the design and size requirements for
effective and reliable removal of biological oxygen demand (BOD) and phosphorus
by vertical flow constructed wetlands for individual homes and groups of residences
(Figure 5.1). These systems featured the recirculation of some of the treated water to
the septic tank for enhanced nitrate removal (Figure 5.2). The investigations resulted
in design and construction requirements published by the Danish Ministry of Environ-
ment and Energy (Brix and Arias, 2005). Austria (1997), Australia (2000) and the
United States (2000) established less binding guidelines for vertical subsurface flow
constructed wetlands (Vymazal, 2011).
Vertical subsurface flow wetlands are commonly used to treat municipal sewage
for small communities, and require about one-half of the area of horizontal subsurface
or free water surface wetlands. For example, the community of Silkerode, Germany
(51°34′7″ N 10°24′1″ E), treats its domestic sewage with a vertical subsurface flow
wetland constructed in 2007 (Figure 5.1). The sewage from this central German town
of about 425 people is collected in a 30 m3 sedimentation tank and then intermittently
applied to the 0.53-acre wetland (2,160 m2) that is planted with Phragmites. The system
is designed to serve 500 people with a maximum daily flow of 24,832 gallons
(94 m3). The quality of the treated water is very good and meets the national standards
for secondary effluent. Initial chemical oxygen demand (COD) of 654 mg/L is reduced
to 31 mg/L, while initial BOD of 355 mg/L is reduced to 7 mg/L. Although not regu-
lated, the effluent is similarly of high quality with regard to ammonia (3 mg/L), total
nitrogen (4 mg/L) and total phosphorus (0.2 mg/L) (Blumberg Engineers).
VERTICAL SUBSURFACE FLOW CONSTRUCTED WETLANDS 71
Figure 5.1
The vertical subsurface
flow wetlands in Silkerode,
Germany, are consistent with
the agricultural character of the
farming town, and provide low-
cost treatment of domestic
sewage. The parallel treatment
beds are in the upper right
corner.
Photo: Google Earth, GeoBasis
DE, 51°33′38.79″ N 10°23′53.85″
E, image date: 31 December
2007, image accessed:
30 October 2015.
Design
The most common design for a vertical subsurface flow constructed wetland consists
of a temporary pool (Figure 5.4), where water (pretreated in a septic tank or similar
device) (Figure 5.3) is delivered rapidly but periodically to flood the surface of a filter
bed composed of one or more layers of sand or gravel (Figure 5.2). The waste-
water drains through the filter media to be collected and discharged by perforated
drain lines surrounded by coarse gravel at the bottom of the basin (Figure 5.8).
The drain lines are connected to other pipes that extend above the surface of the
wetland. These aeration pipes (Figure 5.6) enhance rapid re-oxygenation of the filter
bed after the water drains out. After a few hours, a new batch of pretreated wastewater
flows into the wetland, but during the hiatus, bacteria decompose organic matter,
convert ammonia to nitrate and prey on pathogenic bacteria. Other biological and
chemical processes also contribute to the decrease of contaminants (Tunçsiper, 2009).
The intermittent loading of wastewater and the aeration pipes makes the vertical
subsurface filter bed an oxygen-rich environment.
As with the horizontal subsurface flow wetland, pretreated water in vertical
subsurface flow wetlands can be held below the surface. Vertical subsurface flow
wetlands are flooded periodically just below a surface layer of gravel (Figure 5.6).
There are several variations to the basic design described above. These include
the absence of the pretreatment step (the French system), recirculating a portion of
the treated effluent, tidal-flow loading, up-flow, saturated down-flow and multi-stage
72 VERTICAL SUBSURFACE FLOW CONSTRUCTED WETLANDS
Figure 5.2 This transparent perspective illustrates a vertical subsurface flow wetland based on Danish
standards and the components of the treatment wetland. The reddish pipes (11⁄2-inch perforated PVC,
spaced maximum 3 feet apart) distribute doses of wastewater from the septic tank (tank on right).
The pipes are covered with wood chips or gravel. The main filter bed consists of uncompacted sand,
125–4 mm with clay and silt < 0.5 per cent. The bottom layer is composed of 1⁄2-inch drain rock (
8 inches thick). The drainage pipes are 4 inches diameter perforated PVC, spaced maximum 3 feet
apart. At one end, they connect to aeration pipes that extend above surface. The basin is lined with
5-mm waterproof membrane between two geotextile layers.
Image: Gary Austin, adapted from Brix, 2005 (Brix and Arias, 2005).
Figure 5.3 This transparent perspective of the Danish system shows the large septic tank and a
smaller tank connected to the wetland drain pipes. Half of the water received from the wetland is
discharged, but another portion is recirculated into the septic tank for additional removal of nitrate.
The pumping tank delivers doses of wastewater to the wetland. The other small tank provides doses
of aluminium polychloride into the third chamber of the septic tank where there is an air-lift pump.
This system causes phosphorus to precipitate for enhanced removal if required by the regulating
agency.
Image: Gary Austin, adapted from Brix, 2005 (Brix and Arias, 2005).
Figure 5.4 This is a vertical flow subsurface wetland during testing and construction. The pretreated
wastewater is being delivered to test for uniform distribution. The entire surface is used as an inlet
area, and the pipes are later covered with gravel after completing the testing.
Photo: Heike Hoffmann, and SuSanA Secretariat. License, CC-BY-2.0.
treatment sequences that combine some of these options. Some of these modifications
include adapting vertical subsurface flow wetlands to treat industrial wastewater.
The French system

Developed in the 1980s, these systems are common for communities of fewer than
2,000 people; over 2,500 systems exist with about 200 added each year. They
generally consist of two or three cells operated in series (Figure 5.5). In France, vertical
subsurface flow constructed wetlands for treatment of domestic wastewater do not
include the pretreatment of the effluent; they only include screening (2–3 cm) of the
solids. Therefore, the cost of a septic or Imhoff tank is avoided and the primary sewage
is applied in periodic doses to the top of the treatment bed. Over the course of a year
or two, a layer of degraded sludge accumulates on the top of the bed and improves
the treatment performance of the system. This is presumably due to a rich micro-biotic
environment and an increased hydraulic retention time (Paing et al., 2015). Of course,
these beds cannot be accessible to the public due to health risks from contact and
highly problematic odours; however, in rural settings where the treatment bed can be
placed away from human activity areas and fenced, this is a cost-effective design.
However, some positive secondary uses, such as adjacent recreation, that might be
valuable in more urban settings would be lost.
The French system – single treatment cell

The newer single cell version of the French design has five layers of media, two of
which are expanded schist, as well as an additional layer of aeration piping. From
the top to the bottom, the media layers are as follows: 30 cm depth of 2–4 mm
expanded schist; 20 cm depth of 4–10 mm gravel; 20 cm depth of 0.5–4 mm expanded
schist (with added aeration piping at the top of this layer); 10 cm depth of 4–10 mm
Figure 5.5 In this view of a two-stage French system, the vertical flow treatment wetlands are
installed on a slope. This situation allows the operation of intermittent feeding of the wetland cells
without pumps or electrical power. The two concrete vaults visible in the model are siphon tanks. On
level ground, they become pump vaults. The two cells on the lower terrace are used alternately. The
cells alternate between a treatment and a regeneration (resting) period.
Image: Gary Austin, 2016.
gravel; and 20 cm depth of 10–20 mm gravel. The size of the treatment cell is a very
efficient 1.2–1.5 m2 per person served. This design meets secondary effluent standards,
although the treatment performance is slightly lower than for the two-stage configura-
tion. The single-cell design coupled with the absence of a pretreatment step provides
a very cost-effective system domestic wastewater treatment for small communities
(Paing et al., 2015).
Up-flow vertical subsurface flow constructed wetland

The basic configuration of the vertical subsurface flow constructed wetland can be
modified to cause the wastewater to be delivered to the bottom of the basin, where
it forces existing water within the filter bed to be discharged from the top of the bed.
This configuration transforms the bed into an oxygen-depleted environment. This
design variation emphasizes the removal of nitrate. This configuration is especially
useful as a second stage cell.
Tidal-flow vertical subsurface flow constructed wetland

This variation on the vertical subsurface flow constructed wetland focuses on
increasing the contact time between the wastewater and the filter media. The purpose
of this modification is to improve the removal of total nitrogen. High nitrogen levels
in dairy, pig and other livestock operations is a particular challenge that this design
can address. Tidal flow basins feature either an up-flow or down-flow distribution of
water. The water is delivered in batches as with the conventional design, but instead
of being allowed to immediately drain through the filter media, it is held for a number
of hours or days. In one promising study, three batches of wastewater per day were
treated. However, each batch was cycled through the wetland and a holding tank thrice
in the 8-hour period. The water was held in the wetland for 1–2 hours, then drained
to the holding tank, while the wetland rested for an hour (Hu, Zhao and Rymszewicz,
2014). This schedule created alternating oxygen-rich and oxygen-depleted environ-
ments, leading to very effective removal of ammonia and total nitrogen.
Enhancements
The variations on the typical design are sometimes augmented with forced aeration
or added substances to respond to special treatment needs. This is most common
when treating industrial wastewater as discussed in another chapter. Designs that create
a permanently saturated layer of gravel at the bottom of the vertical flow basin are a
variation of the original design. This flooded sub-basin improves the decrease of nitrate
and total nitrogen (see the case study below).
Design challenge
The first challenge faced in the development of this treatment technology was to reduce
the amount of area (and cost) required to meet the water quality goals. This is
particularly important in urban areas where land cost is high or the amount is scarce.
Producing water to meet secondary standards for BOD and total suspended solids
(TSS) using an aerobic system, rather than the anaerobic environment (horizontal
subsurface flow), was the second challenge. Once these goals were achieved, improving
removal of nitrogen, pathogenic bacteria, phosphorus and other contaminants to meet
advanced (tertiary) treatment standards was adopted as the topic of research and
application. As with horizontal subsurface flow wetlands, the secondary benefits
associated with treatment systems that could be placed in the public landscape led to
the consideration of habitat, recreation, education and other opportunities.
Pretreatment
Only the French system of vertical flow treatment wetlands requires no treatment
beyond coarse screening. All other designs recognize the great benefit of good pre-
treatment in a septic tank or other device. The advantage of good pretreatment is
Figure 5.6
One goal of the vertical flow
treatment wetland is to create
an aerobic environment in
which bacteria can thrive. The
coarse sand and gravel layers,
intermittent loading of
wastewater and the aeration
pipes achieve this. In this
image of a newly constructed
wetland, the aeration pipe is
shown. It is an extension of a
drain line on the bottom of the
treatment basin.
Photo: SuSanA Secretariat,
license CC-BY-SA-2.0.
reduced organic (BOD) and TSS loads, which contribute to clogging of the filter
media. The pretreatment requirements in the Danish guidelines include a two- or three-
chamber septic tank with the capacity of about 106 gallons (0.4 m3) per person. When
a two-chamber septic tank is used, the first chamber should be 70–90 per cent of
the total tank volume. When a three-chamber septic tank is used, the first chamber
should be 50–70 per cent of the total tank volume with the remaining volume divided
equally in the next chambers (Brix and Arias, 2005). An effluent filter is also beneficial
in reducing the TSS entering the treatment wetland and improving its longevity.
Sludge in the septic tank should be removed annually (Brix and Arias, 2005). Septic
tanks remove up to 40 per cent BOD and up to 70 per cent TSS in raw wastewater
(Stefanakis, 2014).
Large towns might choose to use an up-flow anaerobic sludge blanket (USAB) tank
for pretreatment of municipal sewage. This a more sophisticated and expensive
undertaking, but leads to reliable and excellent decrease in BOD while producing low
amounts of sludge (Stefanakis, 2014). This process also allows for the creation and
capture of methane gas, which can be used as an energy source. Methane is also a potent
greenhouse gas; therefore, preventing its release into the atmosphere is important.
Loading
For design purposes, it is assumed that one population equivalent is equal to 60 grams
per day of BOD5, 13 grams per day of nitrogen and 2.5 grams per day of phosphorus
with a flow of 39.6 gallons (150 litres) per day. This definition may vary somewhat
between countries (Brix and Arias, 2005).
Size and organic loading rate

The recommended size of vertical flow wetlands for treatment of domestic or municipal
sewage is 10.8–32.3 square feet (1–3 m2) per person served. The Danish guidelines
for residential systems serving 1–30 residents require pretreatment in a septic tank
and 34.4 square feet (3.2 m2) of treatment area per person (Brix and Arias, 2005).
The French multi-cell treatment wetland receives a typical organic load of 50 grams
BOD5 m2 per day and a nominal hydraulic load of 38 cm/day (consistent effluent quality
is maintained until the hydraulic loading rate exceeds about 60 cm per day). The initial
cell is sized between 1.2 and 1.5 m2 per person equivalent, depending on whether or
not the wastewater contains stormwater. The cells in the second stage (sized 0.8–1.0
m2 per population equivalent) of the treatment sequence are dosed for about a week
and rest for the same period (Paing et al., 2015). A study of 169 vertical subsurface
flow wastewater treatment wetlands, using the French system (no pretreatment),
shows that they are cost efficient and effective treatment systems. In fact, when a single
cell is used, the sludge accumulation on the surface increases the treatment efficiency.
As little as 1.2 m2, rather than the more common standard of 21.5 square feet (2 m2),
is necessary to achieve secondary quality effluent (BOD – 25 mg/L, COD – 125 mg/L)
in France (Paing et al., 2015).
The wide range of organic loading rates applied to vertical flow wetlands reflects
continuing research and efforts to define an optimum design and treatment efficiency.
Also, loading rates are somewhat lower in temperate climates, as compared to tropical
ones. Loading of up to 30 grams of COD m2 per day is possible in temperate climates,
whereas loading as much as 80 grams of COD is possible in tropical climates. A range

of 20–30 grams of BOD m2 per day is a reasonable expectation (Stefanakis, 2014).

An areal hydraulic loading rate of 2–5 inches (5–10 cm) is recommended for temperate
climates while as much as 6 inches (15 cm) is possible in tropical climates (Stefanakis,
2014). Water is added in doses to the vertical flow wetland. A depth of 2–4 inches
(5–10 cm) in each dose is recommended for coverage of the entire cell. This is
generally drained below the temporary pool in less than 10 minutes. The Danish
guidelines recommend recirculation of 50 per cent of the treated effluent to either the
septic tank or dosing tank to improve nitrogen decrease and add stability to the system
through dilution of the concentration of the septic tank effluent (Brix and Arias, 2005).
The French multi-cell system generally involves the screening equipment, a dosing
tank and an initial treatment cell that is followed by two cells operated in parallel
(Figure 5.5). The first cell receives doses (3–4 cm depth at the rate of more than 0.5
m3/m2 per hour) of wastewater over the course of about a week and then is allowed
to rest for about two weeks before the next dosing period is initiated. The French
system nominal hydraulic load is 38 cm per day. However, consistent effluent quality
is maintained until the hydraulic loading rate exceeds about 60 cm per day.
Shape
The shape of the treatment bed is not important as long as the pretreated wastewater
is distributed uniformly across the top surface (Figure 5.7). This flexibility allows for
creative landscape design when the wetlands are part of the public landscape. This
opportunity has not been arisen much in the past, where the treatment performance
and ease of construction were the only issues of interest.
Figure 5.7
Vertical subsurface flow
wetland in the village of
Jiangxiang, near Changshu,
Jiangsu Province, China, just
after planting. This wetland is
the first treatment stage. A
horizontal subsurface flow
stage follows to improve
decrease in nitrate and total
nitrogen.
Photo: courtesy of Blumberg
Engineering, www.blumberg-
engineers.com.
Depth
The depth of the vertical flow treatment basins fairly varies, but is generally 3 feet
(1 m). The inclusion of a flooded sub-basin adds to the typical depth. Most of the
decrease in organic material, TSS, heavy metals, pathogenic bacteria and hydrocarbon
occurs near the top of the basin. Great depth does not improve treatment performance
and may be problematic where groundwater levels are very high.
Materials
Waterproof membrane and outlet media
Preventing the contamination of groundwater by untreated or partially treated sewage
or other contaminated water is critical. Waterproof liners are most often specified
(Figure 5.8). In areas with rocky soil, the membrane is sometimes placed between two
layers of a geotextile fabric to protect the liner. Where available, a layer of bentonite
clay makes an effective and more natural waterproofing. In this case, the sides of the
treatment basin should be sloped rather than vertical.
Filter media
Clogging problems are rare in vertical flow constructed wetlands due to improvements
in sizing and layering the filter material. There are many variations in the configuration
of the filter layers. However, filter materials specified include sand (0.2–0.6 mm d50),
fine gravel (6–16 mm d50), medium gravel (24–32 mm d50) and coarse stone (60–130
Figure 5.8
Vertical subsurface flow
constructed wetlands like all
others treating contaminated
water are lined with a
waterproof membrane or clay.
The yellow perforated drain
lines are covered with coarse
gravel and then a thick layer of
filter media.
Photo: Heike Hoffmann and
SuSanA Secretariat. License,
CC-BY-2.0.
mm d50). A number of modified or manufactured materials are sometimes specified

as filter media, including expanded clay aggregate, shells, iron ore and zeolite
(Stefanakis, 2014).
The Danish guidelines require only a single layer of filter media 3 feet deep
(1 m). This media is supposed to be washed sand where the diameter of 10 per cent
of the particles (d10) is between 0.25 and 1.2 mm, and where the diameter of 60 per
cent of the particles (d60) is between 1 and 4 mm. The coefficient of uniformity (d60/d10)
should be less than 3.5. Silt and clay particles (less than 0.125 mm in diameter) should
be less than 0.5 per cent (Brix and Arias, 2005). This filter material is much finer than
in the French model.
In the French system, the filter media in the initial cell is organized into three layers:
top – 40–50 cm depth of 2–8 mm gravel; middle – 15–20 cm depth of 10–20 mm
gravel; bottom – 20 cm depth of 20–40 mm gravel. The second treatment cell in the
sequence has a slightly different set of media: top – 40 cm depth of 0–4 mm sand;
middle – 15–20 cm depth of 4–10 mm gravel (or 4–20 mm); bottom – 20 cm depth
of 10–20 mm gravel (or 20–40 mm) (Paing et al., 2015).
Clogging
In vertical subsurface flow wetlands, clogging is generally less problematic than in the
horizontal systems and is confined to the top 10 cm of the filter. Nevertheless, long-
term operation requires close attention to rates of organic and TSS loading.
Plants
Plants are positive contributors to constructed wetlands for the treatment of waste-
water and stormwater (Figure 5.9). By providing shade to the wetland surface, plants
moderate temperature fluctuations within the filter bed. The plants absorb nitrogen,
phosphorus and other substances, especially in spring and summer when growth rates
are high. Fewer varieties of plants have been tested for use in both horizontal and
Figure 5.9 Vertical flow subsurface wetland planted with yellow flag (Iris pseudacorus).
Photo: courtesy of Blumberg Engineering, www.blumberg-engineers.com.
vertical subsurface wetlands compared to free water surface treatment wetlands. This
requires additional research, but plants with deeply penetrating roots, rapid growth
and high biomass production serve the goal of treating contaminated water.
In vertical subsurface flow wastewater treatment wetlands, plants are responsible
for the removal of 4–5 per cent of the nitrogen load (Figure 5.10). To realize the benefit
of the absorption of nitrogen, phosphorus and other substances, the wetland plants
must be harvested and removed in early fall before they move the nutrients from the
leaves to the roots or rhizomes. Some species of plants accumulate greater quantities
of nitrogen than others. For example, Iris pseudacorus is particularly effective for the
accumulation of nitrogen. Ammonia is generally the preferred nitrogen form for plant
absorption; however, nitrate is absorbed when concentrations are high and when
ammonia is limited (Stefanakis, 2014).
As plant roots spread through the filter media, they offer sites for the attachment
of bacteria, thereby indirectly supporting the removal of nitrogen and other substances.
Some plants, such as common reed and bulrush, have deeply penetrating root systems.
These can extend as much as 30–60 inches into the filter media. A shallower roots
zone characterizes most other wetland plant species. About 18 inches is the typical
rooting depth of many wetland plants. As plant roots die and decompose within the
filter bed, they open pathways for the penetration of water and oxygen. Wetland plants
translocate oxygen from the aboveground leaves to the root hairs, creating very narrow
pockets where aerobic processes can take place. The extent of this plant-mediated
oxygen transfer is less important to the effective treatment of wastewater than once
thought. In fact, in horizontal subsurface wetlands, anoxic and anaerobic conditions
Figure 5.10 Vertical subsurface flow constructed wetlands can be harvested to acquire the biomass
or capture the nitrogen, phosphorus, metals and other contaminants moved from the wastewater to
the stems and leaves. This stand of Papyrus is in a vertical flow treatment wetland in Peru.
Photo: Heike Hoffmann and SuSanA Secretariat. License, CC-BY-2.0.
predominate, while in vertical subsurface wetlands, intermittent loading of the filter

bed results in aerobic conditions through most of the bed (Stefanakis, 2014).
Case study
This case study demonstrates the performance of a vertical flow constructed wetland
with a design enhancement that was built in Austria and monitored for 19 months.
The design features two vertical subsurface flow stages operated in series (Figure 5.11).
This system reliably achieved secondary quality effluent in both warm and cold
seasons.
Each vertical subsurface cell was 108 square feet (10 m2) and was planted with
Phragmites australis. The first cell (Figure 5.11, left) included a 20-inch deep (50 cm)
bed of sand ranging from 0.08 to 0.12 inches (2–3.2 mm). Below the sand layer was
a flooded sub-basin of water filled with coarse gravel. The second stage included a
0.002–0.16–inch sand layer (0.06–4 mm) above a coarse aggregate drainage layer with
drain pipes (Guenter Langergraber and Roland Rohrhofer, 2009).
Pretreated wastewater flooded the top of the first basin to a depth of 0.64 inches
(16.2 mm) every 3 hours. The loading rate was 80 grams COD m2 per day. This rate
is very high for a temperate climate where 20–30 grams COD m2 per day is typical.
The average concentration of BOD in the pretreated effluent was very high at 327
mg/L. BOD in the effluent from the second basin was very low at 4 mg/L in summer
and 12 mg/L in winter (Guenter Langergraber and Roland Rohrhofer, 2009).
The Austrian wastewater quality standards require that ammonia in the
effluent is 10 mg/L or less, for water temperatures higher than 53.6°F (12°C). The
average ammonia in the effluent from the second wetland cell was 0.29 mg/L in summer
(99.5 per cent decrease) and 17.5 mg/L in winter (64 per cent decrease). This wet-
land removed 46 per cent more of the ammonia than a single cell vertical subsurface
Figure 5.11 Section of a two-stage vertical subsurface flow treatment wetland similar to the case
study wetland. The red pipes are for influent distribution, the yellow pipes are for effluent drainage
and the vertical pink pipes are for aeration of the filter bed. The cell on the left features the flooded
sub-basin. Notice that the pink drainage pipe between the cells is turned up to keep the sub-basin
filled.
Image: Zhenyu Liu and Gary Austin, adapted from Langergraber (Guenter Langergraber and Roland
Rohrhofer, 2009).
wetland with no flooded sub-basin but with a comparable load (Guenter Langergraber
and Roland Rohrhofer, 2009). The US Environmental Protection Agency (EPA) does
not generally regulate ammonia, but aquatic organisms are sensitive to constant levels
in excess of 1.8 mg/L at pH 8 and 25°C (EPA, 2009).
The concentration of nitrates entering the two-stage wetland averaged 0.37 mg/L
in summer and 0.30 mg/L in winter. Nitrates in the effluent were 30.9 mg/L in summer
and 21.1 mg/L in winter. This large increase indicates complete conversion of
ammonia, but low removal of nitrates (Guenter Langergraber and Roland Rohrhofer,
2009).
Elimination of total nitrogen was 53.2 per cent in summer and 37.1 per cent in
winter. This high performance is attributed to the nitrification of about 80 per cent
of the ammonia in the first stage of the wetland with the flooded sub-basin but with
enough carbon remaining to allow conversion of some of the nitrate to nitrogen gas
in the flooded sub-basin, although not all nitrate was removed (Guenter Langergraber
and Roland Rohrhofer, 2009). This improved system performance is attributed to
alternating high and low oxygen zones supporting increased nitrification (ammonia
conversion) and de-nitrification, respectively.
The two-stage vertical subsurface flow wetland removed about 99 per cent of the
pathogenic bacteria. However, the remaining quantity of E. coli, for example, was
1,585 colony-forming units (CFU) per 100 mL (Guenter Langergraber and Roland
Rohrhofer, 2009), which is still much higher than the 126 CFU per 100 mL standard
for primary contact (swimming) set by the US EPA. Additional treatment in another
wetland stage or ultraviolet light disinfection would be required before this effluent
could be used for primary recreation.
Table 5.1 illustrates the high treatment performance of two-stage French vertical
flow treatment wetland when operated according to the design loading rates. Effluent
quality increased as nitrifying bacteria became established during the first 6 months
of operation. The oldest wetlands monitored had been operating for 12 years. No
clogging or loss in effluent quality could be attributed to the age of the wetland (Paing
et al., 2015).
Nitrogen
As noted in Chapter 2, the first part in the three-phase reduction of nitrogen is ammoni-
fication, followed by nitrification and denitrification. Microorganisms are the primary
agent in the transformation of organic nitrogen to ammonia, which can occur in both
aerobic and anaerobic zones, although the process is much more rapid in an oxygen-
Table 5.1 Mean influent and effluent from the two-stage French system treatment wetland system.
Units are mg/L except pH (Paing et al., 2015)
pH TSS BOD5 COD TKN N-NH4 N-NO3 TP P-PO4

Raw wastewater Mean 7.9 353 360 841 94 70 12 7
2nd stage effluent Mean 6.8 10 6 51 7 5 56 8 8
rich environment. Vertical flow wetlands are quite tolerant of fluctuations in the BOD
(organic) load. In fact, high removal rates occur at a range from about 10 mg/L to
80 mg/L. In vertical subsurface flow wetlands, the rate of ammonification is as high
as 0.53 grams N/m2 per day. Ammonification occurs primarily in the top layer of the
filter bed where the oxygen saturation is highest (Stefanakis, 2014).
Nitrification
During the second step of the nitrogen removal process, nitrogen in the form of
ammonia is oxidized to nitrite by a group of bacteria genus (Nitrosomonas,
Nitrosococcus, Nitrosolobus and Nitrosopira). Then a second population of bacteria
(Nitrospira, Nitrospina, Nitrosococcus and Nitrobacter) convert nitrite to nitrate. The
nitrifying bacteria obtain energy from the oxidation of ammonia and use carbon dioxide
as a carbon source. The conversion of ammonia to nitrate occurs in oxygen-rich
environments, while the conversion of nitrate to nitrogen gas occurs in an anoxic
situation. Vertical flow constructed wetlands feature a high capacity for transforma-
tion of organic material to ammonia and transformation of ammonia to nitrate
(Stefanakis, 2014). Wetlands based on the Danish guidelines result in 100 per cent
nitrification with or without recirculation of treated effluents to the septic tank (Arias,
Brix and Marti, 2005). For very high strength wastewater, artificial aeration of the
filter bed results in higher nitrification and reduction in total nitrogen. In contrast to
the aeration requirements of conventional activated sludge treatment of wastewater
in conventional plants, the aeration of vertical flow subsurface flow wetlands requires
10 per cent of the energy. This is because the aeration is intended to provide small
bubbles in a small area and can be intermittent (Stefanakis, 2014).
Nitrification in vertical flow constructed wetlands treating wastewater can be
increased significantly through the addition of earthworms, according to one micro-
cosm study. The worms increase the height, width and stem diameter of Phargmites
australis, Typha augustifolia and Carex indica. Nitrification was increased 61 per cent,
59 per cent and 65 per cent for the three plants, respectively. Denitrification was also
increased but to a lesser extent (5–15 per cent) (Xu et al., 2013).
Denitrification
The final step in the removal of nitrogen from wastewater is the decrease of nitrite
and nitrate to nitric oxide, nitrous oxide and nitrogen gas by autotrophic bacteria
(Pseudomonas, Bacillus, Microccus, Sprilliums) (Stefanakis, 2014). Since this process
requires anoxic or anaerobic conditions not provided in the typical vertical subsurface
flow wetland, nitrate removal is typically very low. Recycling treated effluent to the
septic tank at the rate to 100–200 per cent of the hydraulic loading rate improves
nitrate (and total nitrogen) removal significantly. A Danish study found that although
a vertical flow constructed wetland reduced the organic nitrogen and fully nitrified
wastewater influent, the reduction in total nitrogen was only about 1 per cent.
However, when the treated effluent was recycled to the septic tank at the rates of 100
per cent, 200 per cent and 300 per cent of the initial hydraulic loading, then the removal
of total nitrogen increased to 52 per cent, 66 per cent and 68 per cent, respectively
(Arias, Brix and Marti, 2005).
Phosphorus
As in horizontal subsurface flow wetlands, phosphorous removal is low (20–40 per
cent) unless the sand or gravel filter media is high in calcium, iron or aluminium to
provide abundant cation exchange sites. Adding additional treatment with materials
such as steel slag, oyster shell, wollastonite or bauxite can more than double the normal
removal rate (Stefanakis, 2014).
Pathogenic bacteria
The Austrian study above illustrated the very high removal of pathogenic bacteria by
vertical flow constructed wetlands. Other studies report as much as 99.9 per cent
removal of pathogens. Finely textured filter media and ample retention time, effluent
recirculation, low hydraulic loading rate and high hydraulic residency time all favour
pathogen reduction (Stefanakis, 2014). The majority of the removal occurs in the top
8 inches (20 cm) of the treatment basin. Although additional treatment will often be
required to meet standards for recreational use, the low concentration and clarity of
the effluent simplifies this operation.
Conclusion
Vertical flow constructed wetlands, partly due to the variety of flow options (up, down
or tidal), are somewhat more complex than horizontal subsurface flow treatment
wetlands. Therefore, they require slightly more maintenance attention and sometimes
the use of small pumps for intermittent feeding. A gently sloping site and the use of
siphon vaults can eliminate the need for pumps and allow the system to be powered
entirely by gravity flow. This is sometimes important in rural areas, especially in
developing nations.
In developed countries, expensive engineering and permitting costs discourage the
widespread adoption of new wastewater treatment technologies such as treatment
wetlands. Once systems have been monitored under a range of conditions, then
standard sizing, materials and construction details can be codified. These standard
design and construction practices encourage rapid and extensive application of
technology. This has been undertaken in Germany, Austria, France and Denmark (Brix
and Arias, 2005), and has resulted in tens of thousands of subsurface flow treatment
wetlands. Large projects or those with special circumstances would continue to
require targeted engineering and construction considerations (Figure 5.12). For
example, the vertical flow constructed wetland in Shenyang, China, was unprecedented
in its size and urban context. Blumberg Engineers designed the system to serve 6,000
residents, and a similar system in Iran serving 4,000 residents. The Shenyang wetland
consists of three vertical subsurface flow beds (380 x 84 feet, 116 m x 56 m) operated
in parallel.
The average removal rates of COD, ammonia and total phosphorous were 90.9
per cent, 98.4 per cent and 89.7 per cent, respectively. The wetland has a dual waste-
water distribution system. A surface flow pipe network delivers the wastewater to the
treatment bed in summer, while a drip irrigation pipe system 30 cm below the surface
delivers the water to the filter bed in the winter. A soil layer (20 cm) and an air space
protect the lower system in freezing winter temperatures (about –28°C, 18°F: the project
Figure 5.12
A vertical subsurface flow cell
of the Shengyang wetland
showing the installation of the
waterproof membrane and
drainage system.
engineers.com.
Figure 5.13
The vertical flow constructed
wetland treats wastewater for
6,000 residents to high water
quality standards, Shengyang,
China.
engineers.com.
is at latitude 41°44′31.92″ N). The wetland occupies two acres (8,000 m2) (Figure
5.13). It cost $9.6 million, and the operating cost is only 20 per cent of the cost to
operate a conventional wastewater treatment plant.
References
Arias, C., Brix, H. and Marti, E. “Recycling of treated effluents enhances removal of total
nitrogen in vertical flow constructed wetlands,” Journal of Environmental Science and Health
Part A, vol. 40, no. 6, pp. 1431–1443, 2005.
Blumberg Engineers, “Reed bed treatment system – Sewage treatment plant, Silkerode,” 2015.
Brix H. and Arias, C. A. “The use of vertical flow constructed wetlands for on-site treatment
of domestic wastewater: New Danish guidelines,” Ecological Engineering, vol. 25, no. 5,
pp. 491–500, 2005.
EPA (US Environmental Protection Agency), “Aquatic life ambient water quality criteria for
ammonia freshwater,” US Government, 822D09001, 2009.
Guenter Langergraber K. L. and Roland Rohrhofer, R. H. “High-rate nitrogen removal in a
two-stage subsurface vertical flow constructed wetland,” Desalination, vol. 246, pp. 55–68,
2009.
Hu, Y., Zhao, Y. and Rymszewicz, A. “Robust biological nitrogen removal by creating multiple
tides in a single bed tidal flow constructed wetland,” Science of the Total Environment, vol.
470–471, pp. 1197–1204, February 2014.
Paing, J., Guilbert, A., Gagnon, V. and Chazarenc, F. “Effect of climate, wastewater
composition, loading rates, system age and design on performances of French vertical flow
constructed wetlands: A survey based on 169 full scale systems,” Ecological Engineering,
February 2015.
Stefanakis, A. Vertical flow constructed wetlands: Eco-engineering systems for wastewater and
sludge treatment, First edition. Amsterdam: Elsevier, 2014.
Tunçsiper, B. “Nitrogen removal in a combined vertical and horizontal subsurface-flow
constructed wetland system,” Desalination, vol. 247, no. 1–3, pp. 466–475, 2009.
Vymazal, J. “Constructed wetlands for wastewater treatment: Five decades of experience,”
Environmental Sciences Technology, vol. 45, no. 1, pp. 61–69, January 2011.
Xu, D., Li, Y., Howard, A. and Guan, Y. “Effect of earthworm Eisenia fetida and wetland
plants on nitrification and denitrification potentials in vertical flow constructed wetland,”
Chemosphere, vol. 92, no. 2, pp. 201–206, June 2013.
6
Hybrid constructed wetlands
Introduction
Hybrid treatment wetlands are a combination of the basic types of constructed
wetlands that are being increasingly applied to particularly difficult wastewater
treatment problems. Industrial and agricultural wastewater is often high strength or
contains an array of contaminants not present in domestic wastewater. Food
processing plants produce high strength biodegradable wastewater, for example.
As documented in previous chapters, after pretreatment, a single stage of either
horizontal, vertical or free water surface wetland is sufficient to achieve secondary
quality effluent for domestic sewage. Each wetland type can achieve this reliably
and economically for individual residences, subdivisions or small communities in all
but the most extreme climates. Since any of the wetland types is effective, the choice
of which type to implement is based on factors not directly related to treatment
performance, such as available land area, climate or proposed secondary benefits.
Unlike conventional wastewater treatment plants, constructed treatment wetlands
are a good choice for a distributed network of small facilities organized for down-
stream reuse of treated water for irrigation or other non-potable uses. Furthermore,
constructed wetlands can be located and designed to provide outstanding secondary
benefits at a low cost. However, treatment wetlands are increasingly designed as multi-
stage facilities where more than one type of constructed wetland is employed to achieve
tertiary effluent quality.
Tertiary water quality

Increasingly, secondary quality effluent is insufficient for the protection or restoration
of streams, rivers, lakes and coastal environments. This is often because a number of
communities and cities are discharging treated effluent to the same receiving waters.
When a stream, for example, becomes effluent dominated because it receives most of
its water from wastewater treatment facilities during, at least, a number of months
each year, then the aquatic ecosystem becomes highly dependent on the quality of the
treated wastewater. In this case, it is important to reconsider the traditional assumption
that polluted water will be diluted by the receiving stream resulting in low concentra-
tions of contaminants. Unregulated concentrations of phosphorus, nitrogen or pharma-
ceutical contaminants discharged in secondary effluent can overwhelm the ecosystem’s
capacity to accommodate them.
88 HYBRID CONSTRUCTED WETLANDS
Alternatively, the receiving waters may be particularly sensitive to the addition

of these substances. In this case, the concentration of nutrients and contaminants typical
of secondary effluent causes ecosystem damage. Fortunately, combining the types of
constructed wetland capitalizes on the capacity of each to remove specific nutrients
or substances and improve the quality of the effluent for protection of the aquatic
ecosystem. This can often be done at a capital and operating cost much lower than
required for conventional wastewater treatment plants.
Tertiary treatment of domestic sewage is increasingly necessary due to emerging
organic contaminants from pharmaceutical and personal care products pesticides
and antiseptics. Conventional wastewater treatment plants often poorly treat these
substances. Many of these contaminants are toxic to aquatic and other organisms (Ávila
et al., 2014). Hybrid wetlands show great promise in eliminating contaminants from
pharmaceutical and personal care products.
In addition to the indirect measures provided by the removal of biological oxygen
demand (BOD) and total suspended solids (TSS) (secondary level of treatment),
wastewater treatment and stormwater discharge permits now often require discharge
maximums for ammonium, nitrates, phosphorus, heavy metals and pathogenic bacteria
into sensitive water bodies and recreation areas. Therefore, more comprehensive or
more complete treatment is necessary.
Hybrid wetlands are designed to treat agricultural wastewaters (see Chapter 8)
and the effluent from a wide range of industries (see Chapter 9), landfill leachate and
contaminated groundwater and river water (see Chapter 7). The combination of aerobic
and anaerobic treatment stages improves the remediation of wastewater with sub-
stances that are not readily biodegradable, such as those in winery or petroleum refinery
effluent. The inclusion of a free water surface wetland as the last stage of a hybrid
wetland provides water quality benefits, but may be most important for the aesthetic,
habitat, recreation and education values that they add to the public landscape.
Wastewater case study

The example of a hybrid constructed wetland presented here demonstrates the out-
standing treatment for a broad range of contaminants in municipal wastewater.
The full-scale wetland (Figure 6.1) is located in Carrión de los Céspedes near Seville
in southern Spain (37°21′34.71″ N 6°20′01.47″ W). There is no industrial land use
in the town; therefore, the wetland performance is an accurate assessment of domestic
sewage treatment. However, the inflow includes stormwater, which is not a recom-
mended practice except in fairly arid zones where stormwater surface run-off is
minimized. Since this is an example of a warm weather region, the performance of
the hybrid wetland would be rather less effective for some contaminants, such as nitrate
removal, during winters in cold climates.
Treatment sequence and configuration

Pretreatment consists of the elimination of solids with 3 cm and 3 mm screens and
the removal of sand and grease. An Imhoff tank serves as the primary treatment step.
It is followed by a vertical subsurface flow wetland of 3,412 square feet (317 m2), a
horizontal subsurface flow wetland of 2,465 square feet (229 m2) and a free water
surface wetland of 2,583 square feet (240 m2) connected in a series. The materials
HYBRID CONSTRUCTED WETLANDS 89
Figure 6.1
Experimental wastewater
facility near Carrión de los
Céspedes, Spain.
Photo: Centre for New Water
Technologies, R & D & I.
License: reproduced with
permission of CENTA.
and other details of each stage are provided below to assess the performance of the
wetland in relation to the installation situation.
Vertical subsurface flow stage

Table 6.1 illustrates that pretreatment is very important for the reduction of TSS, BOD
and chemical oxygen demand (COD) and reduces the risk of clogging the first wetland
stage. The vertical subsurface flow wetland (Figure 6.2) receives about 20 doses of
wastewater each day, amounting to a daily average total of 3,700 gallons (14 m3).
This project and other research indicate that small, but frequent, doses of wastewater
are a feeding strategy leading to high performance of vertical flow wetlands. A modest
organic load of 9 grams of BOD5 per m2 per day is treated. Each dose of influent is
delivered by five 5-inch (125 mm) diameter pipes with 0.4-inch (1 cm) diameter perfor-
ations every 6 feet (1.8 m) (Ávila et al., 2013).
The vertical subsurface flow bed is composed of three layers. The top layer is
2 inches (5 cm) of sand (1–2 mm particle size), followed by a 24-inch layer (60 cm)
of siliceous gravel 0.16–0.5 inches (4–12 mm) and a bottom layer 6 inches (15 cm)
deep of siliceous gravel 1–1.6 inches (25–40 mm). Within the bottom layer of gravel
in the vertical subsurface flow wetland, five perforated drainpipes were installed to
include three 3-feet-tall (1 m) aeration chimneys in order to transfer oxygen into the
filter media. Water is retained in the vertical subsurface flow bed for only a few hours
due to the intermittent loading schedule. The vertical subsurface flow stage, like the
horizontal subsurface flow stage that follows it, was planted with Phragmites australis
that was mature and vigorous by the time the study began (Ávila et al., 2013).
The vertical subsurface flow stage demonstrates excellent treatment perform-
ance for several contaminants. The removal of BOD5 (94 per cent), COD (85 per cent),
TSS (90 per cent) and total nitrogen (66 per cent) is impressive. At the end of the
vertical subsurface flow stage, the concentrations of BOD5 and TSS (Table 6.1)
are already much lower than the secondary treatment standards required by the
US Environmental Protection Agency (EPA). The removal of ammonia was high
Figure 6.2
The vertical subsurface flow
stage with Phragmites occurs
after pretreatment. The vertical
pipes are part of the aeration
system.
Technologies, R & D & I. License:
reproduced with permission of
CENTA.
(74 per cent), but the percentage of total nitrogen removed (66 per cent) by the vertical
subsurface flow stage suggests that both nitrification and some denitrification takes
place within the bed, although it is generally assumed that vertical subsurface flow
wetlands are entirely aerobic (Ávila et al., 2013).
Horizontal subsurface flow stage

The second treatment bed is a horizontal subsurface flow wetland consisting of
16-inch deep (40 cm) siliceous gravel (0.16–0.5 inches, 4–12 mm particle size) (Figure
6.3). The treatment bed includes inlet and outlet zones with stones 11⁄2–3 inch (40–80
mm) in diameter. Effluent from the vertical subsurface flow is distributed throughout
the inlet zone by a polyethylene pipe 2.5 inch (63 mm) in diameter, perforated with
0.4 inch (1 cm) holes every 3 feet (1 m). The horizontal subsurface flow wetland is
drained from the bottom of the outlet zone by two 5-inch (125 mm) diameter drain-
pipes. The outlet pipe is flexible to allow adjustment of the water level at 2 inches
(5 cm) below the top of the gravel bed. Water is retained in the horizontal subsurface
flow wetland for 2.3 days (hydraulic retention time) (Ávila et al., 2013).
Although the vertical subsurface stage is responsible for the dramatic removal of
organic material, BOD, TSS and ammonia, the horizontal subsurface stage made
significant improvements to the water quality as well. This is particularly true for the
removal of nitrate and additional reductions in ammonia, BOD and COD. Nitrate
reduction is an important tertiary water quality achievement. The concentration of
total nitrogen decreased 54 per cent between the end of the vertical flow stage and
the end of the horizontal flow stage (Ávila et al., 2013).
Free water surface stage

The next treatment stage is a 12-inch deep (30 cm) free water surface wetland
with an 8-inch deep (20 cm) bed of siliceous gravel, which is planted with Typha
spp., Scirpus spp., Iris pseudacorus, Carex flacca, Cyperus rotundus and Juncus spp.
The hydraulic retention time was calculated as 5.1 days. As in the other stages, the
Figure 6.3 This image shows the second stage (horizontal subsurface flow) of the hybrid treatment
wetland. The effluent from the vertical subsurface flow stage is distributed into the coarse gravel.
Since there is no public access, the wastewater is discharged onto the gravel surface but this pipe
could be covered. Also, the side slopes of the basins are steeper and taller than required and could
be reduced in a setting where aesthetics is important.
Photo: Centre for New Water Technologies, R & D & I, 2014. License: reproduced with permission of
CENTA.
Table 6.1 Contaminant average concentration by treatment stage (Ávila et al., 2013)
Element Influent Imhoff tank VSSF HSSF FWS

TSS 287 ± 108 98 ± 25 8±6 8±3 6±2
COD 539 ± 200 294 ± 123 46 ± 18 30 ± 16 50 ± 16
BOD5 393 ± 127 204 ± 80 10 ± 5 5±5 7±3
TN 54.6 ± 12.5 52.7 ± 18.1 17.6 ± 6.0 8.2 ± 4.6 7.9 ± 5.1
NH4 42.1 ± 11.2 43.1 ± 15.2 10.4 ± 5.5 5.5 ± 4.8 2.3 ± 2.7
NOx 0.9 ± 0.4 0.8 ± 0.6 3.1 ± 2.6 0.7 ± 0.2 0.9 ± 0.5
TP 8.1 ± 2.5 6.8 ± 2.6 5.7 ± 2.9 4.7 ± 3.8 5.3 ± 2.7
PO4 5.4 ± 2.0 4.8 ± 2.5 4.8 ± 2.3 3.7 ± 3.2 4.7 ± 2.6
All units mg/L. ± = standard deviation.
vegetation was well established. The effluent from the free water surface wetland is
stored in a 5,283 gallon (20 m3) open tank (Figure 6.4) for future use. The combined
water retention time is more than 7.4 days (Ávila et al., 2013).
Since the quality of the water entering the free water surface stage was already
very good, there was a significant reduction in only ammonia. In fact, there was a
small increase in the concentration of most parameters. This is probably due to the
use of the wetland by wildlife, and it indicates that the concentrations of BOD, nitrate
and total nitrogen were already at or below background (naturally occurring) levels.
There was a significant reduction in E. coli by the free water wetland stage (from
3,000 to less than 40 colony-forming units (CFU) per 100 mL) and by each of the
previous stages as well (Ávila et al., 2013).
Figure 6.4
The free water surface
treatment stage is completely
covered by a variety of plant
species. The plants in this
photo are 4 years old.
Technologies, R & D & I, 2014.
The hybrid wetland showed only a 22 per cent reduction in phosphorus. This
poor result could be improved significantly by using filter sands and gravel high in
receptor sites. A second study of this hybrid wetland, one year after the first, yielded
similar results, although it showed a 47 per cent decrease in phosphorus. The levels
of pathogenic bacteria decreased significantly in each of the wetland stages. E. coli
was reduced from 5 million CFU per 100 mL in the Imhoff tank effluent to less than
40 CFU per 100 mL in the free water surface effluent and the water tank. This
represents a 99.999 per cent decrease, and allows the water to be used for groundwater
recharge and irrigation without chlorine or other disinfection measures (Ávila, 2014).
Similar results for pathogenic bacteria removal were observed in the performance
of the Oaklands Park hybrid wetland, serving 65 people, constructed in Britain in
1989. The wetland included vertical and horizontal flow cells and a fish pond. Total
coliform bacteria (an indirect measure of pathogenic bacteria) declined from 2,310,000
CFU per 100 mL in the septic tank effluent to 680 CFU per 100 mL at the pond outlet.
E. coli declined from 500,000 to 0, and faecal Streptococci dropped from 22,000 to
25 CFU per 100 mL (Burka, 1990; Gaboutloeloe et al., 2009). The bacteria level in
the pond stage of the hybrid wetland met the US EPA standards for swimming.
The removal of pharmaceutical and personal care product contaminants by the
hybrid wetland described above was greater than 80 per cent for all compounds,
although the role of each stage of the wetland had a different treatment impact. The
study of the wetland’s performance included three analgesic (pain medicines) and anti-
inflammatory drugs, including Ibuprofen, Diclofenac and Acetaminophen; three
compounds from personal care products, including Tonalide, Oxybenzone and
Triclosan; and two endocrine-disrupting compounds, including Bisphenol A and
Ethinylestradiol. Table 6.2 illustrates that every stage of the hybrid wetland contributed
to the removal of these substances. This suggests that both aerobic and anaerobic
conditions are required for complete or nearly complete removal of the contaminants.
The horizontal subsurface flow stage was particularly effective for the reduction of
Ibuprofen, Diclofenac, Tonalide and Bisphenol A. The free water surface wetland stage
was effective in the continued reduction of the concentration of these same substances
as well as Triclosan (Ávila et al., 2014).
Table 6.2 Mean values of emerging organic contaminants in the hybrid wetland stages (␮g/L)
(Ávila, 2014)
Contaminant Imhoff VF HF FWS Water Overall

tank tank removal (%)
Analgesic/anti-inflammatory drugs
Ibuprofen 14.78 4.01 0.52 0.03 0.03 >99
Diclofenac 0.74 0.50 0.28 0.10 0.10 89
Acetaminophen 3.32 LOD LOD LOD LOD 99
Personal care products

Tonalide 0.33 0.24 0.11 0.05 0.02 90
Oxybenzone LOD LOD LOD LOD LOD –
Triclosan 0.13 0.05 0.05 0.04 0.03 79
Endocrine-disrupting compounds
Bisphenol A 3.90 2.12 1.35 LOD LOD >99
Ethinylestradiol LOD LOD LOD LOD LOD –
LOD = less than the limit of detection.
It is significant that the resulting outstanding water quality was achieved without
artificial aeration, pumping or use of chemicals. When hybrid wetlands are installed
on a gently sloping site, gravity can be used to transfer the water between treatment
stages.
An experimental hybrid wetland with three stages constructed to achieve tertiary
water quality confirms the experience from the wetland presented above. The
experimental wetland was composed of a fully saturated vertical subsurface flow cell
followed by a free-draining, vertical subsurface flow cell. The third cell was a
horizontal subsurface flow bed. Fifty per cent of the effluent from stage two was
pumped to stage one for denitrification (Vymazal and Kröpfelová, 2011).
The organic load was 17.8 grams per day of BOD, compared to the 9 grams per
day in the Spanish hybrid wetland described above. Nevertheless, the performance
was very similar. The average removal rates were: BOD – 94.5 per cent (10 mg/L),
TSS – 88.5 per cent (9.2 mg/L), ammonia – 78.3 per cent (6.5 mg/L) and phosphorus
Figure 6.5
The storage tank that
receives the treated effluent
from the free water surface
stage provided unexpected
water quality improvement.
There was a significant
decrease in the levels of
E. coli due to the penetration
of sunlight to the bottom of
the tank.
Technologies, R & D & I, 2014.
– 65.4 per cent (1.8 mg/L). The concentration of nitrates in the outflow was only
1.1 mg/L. The combined ammonia and nitrate removal was 73.5 per cent (Vymazal
and Kröpfelová, 2011).
Hybrid wetland treating surface run-off

An example of a hybrid treatment wetland applied to the problem of surface run-off
from livestock operations was designed by Blumberg Engineers and constructed in
2013. The hybrid wetland in Hennersdorf, Germany (Figure 6.6), treats stormwater
contaminated with agricultural waste from a cattle operation, including run-off from
silage production. The treatment sequence includes four stages: (1) floating wetlands
over a sedimentation pond 4,445 square feet (413 m2); (2) a buffer pond 4,747 square
feet (441 m2); (3) a vertical subsurface flow wetland 10,900 square feet (1,014 m2);
(4) a horizontal subsurface flow wetland 5,513 square feet (475 m2). The concentration
of organic matter in the wastewater varies but is very high, between 3,400 mg/L and
32,700 mg/L COD with a maximum flow of 7,925 gallons (30 m3) per day. At a median
concentration of 10,100 mg/L COD and a median pH of 4.9, effluent concentrations
for COD and BOD5 are 150 mg/L and 40 mg/L, respectively (Blumberg Engineering).
Secondary benefits
Sustainable development often revolves around the provision of secondary benefits.
Designing for multiple uses frequently prevents the common problem of addressing
and solving a single issue while simultaneously creating other problems.
Figure 6.6 This hybrid wetland in Hennersdorf, Germany, treats surface run-off from a livestock
operation. The newly planted floating mats of emergent vegetation reduce nitrogen and BOD in the
first treatment stage (sedimentation and buffer ponds). After the sedimentation pond and similar
buffer pond, the water is treated in a vertical subsurface stage and a horizontal subsurface flow
wetland. The use of vegetated floating mats is useful in stormwater ponds also.
Photo: Blumberg Engineers, www.blumberg-engineers.com. License: reproduced with permission.
The addition of a free water surface stage to a hybrid treatment wetland

exponentially increases the secondary benefits to the community or the ecosystem or
both. The visual appeal of open water draws visitors, as it adds variety to the
landscape. Water also attracts birds and other wildlife that are of interest to people.
In the image of the central park and municipal wastewater treatment hybrid wetland
shown in Figure 6.7, three wetland types are visible. At the upper left of the image,
there are three vertical subsurface flow panels that are the first treatment stage. In the
middle of the image, two panels of horizontal subsurface flow wetlands are visible.
Adjacent to these is a linear pool, and, at the bottom, the free water surface wetland
completes the treatment sequence. Treated water is collected in a tank for non-potable
reuse. A few of the other park amenities on this intensively developed 1.5-acre site
are also visible in the image. This park is on the island of Koh Phi Phi near the coast
of Thailand, where the tropical climate makes growing wetland plants with large
colourful flowers possible. Canna and Heliconia are planted in the vertical subsurface
wetland, while Canna are planted in horizontal subsurface flow wetlands. The hybrid
wetland performed very well, but illegal sewer connections without septic tanks
caused problems later (Brix, 2011). The design of the free water surface wetland beds
in this project discourages contact with the water. If the water was more accessible,
and if the vertical and horizontal subsurface stages did not result in water quality
appropriate for secondary recreation, then a disinfection step would be required.
Although multifunctional hybrid wetlands are not yet common, the park-like
setting shown in Figure 6.8 is a second example of effective wastewater treatment,
in this case for a restaurant and tourist centre on Lotus Island in Lake Yangcheng
near Suzhou, China. Constructed in 2009, this wetland treats as much as 7,925 cubic
feet (30 m3) of domestic sewage per day. Figure 6.8 shows two round plantings of
Phragmites and Canna. These plant masses indicate vertical subsurface flow wet-
lands (3,767 feet2, 350 m2) and the first stage of treatment after the septic tank. The
second stage is a 1,077 square feet (100 m2) horizontal subsurface flow bed (Blumberg
Engineers).
Free water surface wetlands that are commonly used in the United States as a
tertiary treatment stage after conventional wastewater treatment plants have
demonstrated outstanding ecosystem values, especially when planted with a diversity
of plant species, open water and marsh areas with varying water depth. Free water
surface wetlands are commonly used in the United States as a tertiary treatment stage
Figure 6.7
Koh Phi Phi, Thailand, is a
hybrid wetland for treatment of
municipal wastewater but also
as the central park for the
community.
Photo: Hans Brix, Aarhus
University, Denmark. Reproduced
with permission.
Figure 6.8
Aesthetics were important in
this wetland constructed to
treat of domestic sewage since
it is at an ecotourism resort.
Several similar small-scale
projects along Lake Yangcheng
protect the water quality of the
lake and the tourism industry.
Photo: Blumberg Engineers,
www.blumberg-engineers.com.
permission.
Figure 6.9
The upper-right portion of this
image shows the three vertical
subsurface flow wetland stages
of the Shengyang wetland. The
rest of the wetland receives the
treated wastewater and
stormwater.
permission.
Figure 6.10
The treated effluent from the
three vertical subsurface flow
beds drains into the
stormwater wetland.
permission.
after conventional wastewater treatment plants. When located as a final stage in a

hybrid treatment sequence, free water surface treatment wetlands offer outstanding
ecosystem benefits. These are maximized when the free water surface cell is planted
with a diversity of species and features open water and marsh areas with varying water
depth. Figure 6.9 is an aerial image of the Shenyang (China) project discussed in the
chapter on vertical subsurface flow wetlands. It shows the three panels of the vertical
flow treatment wetlands in the upper right corner. Free water surface wetlands are at
the left and bottom. They receive treated effluent and untreated stormwater. The
aesthetic and wildlife values are remarkable for this densely developing urban centre
(Figure 6.10).
References
Ávila, C., Bayona, J. M., Martín, I., Salas, J. J. and García, J. “Emerging organic contaminant
removal in a full-scale hybrid constructed wetland system for wastewater treatment and
reuse,” Ecological Engineering, vol 80, pp. 108–116, July 2014.
Ávila, C., Salas, J. J., Martín, I., Aragón, C. and García, J. “Integrated treatment of combined
sewer wastewater and stormwater in a hybrid constructed wetland system in southern Spain
and its further reuse,” Ecological Engineering, vol. 50, pp. 13–20, January 2013.
Blumberg Engineering, “Cascade of ponds and constructed wetlands for treatment of
agricultural polluted surface run-off.”
Blumberg Engineers, “Treatment wetlands at Lotus Island on Yangchen Lake, Suzhou, China.”
Brix, H., Koottatep, T., Fryd, O. and Laugesen, C. H. “The flower and the butterfly constructed
wetland system at Koh Phi Phi – System design and lessons learned during implementation
and operation,” Ecological Engineering, vol. 37, no. 5, pp. 729–735, 2011.
Burka, U. “A new community approach to waste treatment with higher water plants,” in
Constructed Wetlands in Water Pollution Control, Oxford: Pergamon Press, 1990.
Gaboutloeloe, G. K., Chen, S., Barber, M. E. and Stöckle, C. O. “Combinations of horizontal
and vertical flow constructed wetlands to improve nitrogen removal,” Water Air Soil
Pollution: Focus, vol. 9, no. 3–4, pp. 279–286, 2009.
Vymazal J. and Kröpfelová, L. “A three-stage experimental constructed wetland for treatment
of domestic sewage: First 2 years of operation,” Ecological Engineering, vol. 37, no. 1, pp.
90–98, 2011.
7
Plants in constructed wetlands
Introduction
Plants in the various types of wetlands and bio-retention basins have multifaceted
roles. They have important engineering functions, but ecological contributions are
a critical concern in sustainable development. There are also aesthetic, educational
and even social dimensions to the plants that are selected (Figure 7.1). Finally,
there are maintenance and even possible economic returns that should be considered.
Since environmental engineers have conducted most of the research leading to the
development of constructed wetlands, many of the roles that plants play have been
poorly explored.
Initial research on plants for constructed wetlands focused on the survivability
and rooting depth of macrophytes planted in domestic wastewater. A few of the most
common species have been investigated according to their capacity to absorb nitrogen,
phosphorus, heavy metals and other substances into their roots, stems and leaves. Less
research has been devoted to the ecological and aesthetic aspects of plants in
constructed wetlands than their engineering properties. Since this book focuses on
sustainability issues, plants are of interest for their full range of contributions.
Since constructed wetlands are engineered systems, the technical performance of
the plants must be balanced with the practical considerations of the effort and cost
of installation and maintenance. The great difference between wastewater treatment
wetlands that are continuously supplied with water and the intermittently supplied
stormwater or agricultural treatment wetlands has an impact on the types of plants
that can be specified. Similarly, differences in climate, particularly precipitation and
cold, limit or expand the palette of plants available to the designer.
From a sustainable development perspective, the opportunity is to increase the
diversity of plants, especially native species that are specified and planted in constructed
wetlands of various types. The challenge is to maintain or improve the engineering
functions that the typical monoculture of plants contributes, while adding ecosystem,
aesthetic and other benefits. This chapter will address the selection of plants to
achieve greater diversity, while managing the more complex installation and main-
tenance issues that diversity might pose.
The following section addresses the physical and cultural characteristics of the
most commonly specified plants for constructed wetlands around the world, but a
section that proposes associations of plants to create additional ecosystem service
benefits and ecosystem functions follows it.
PLANTS IN CONSTRUCTED WETLANDS 99
Commonly selected plants

Although approximately 150 species have been planted in constructed wetlands, by
far the most frequently planted species are Typha latifolia, cattail (common in North
America); Phragmites australis, common reed (very common in Europe and Asia);
Typha angustifolia, narrowleaf cattail; Juncus effusus, common or soft rush; Scirpus
(Schoenoplectus) lacustris, bulrush; Scirpus californicus and Phalaris arundinacea, reed
canary grass. Another common genus is Eleocharis, spikerush. While Phragmites
australis and Typha domingensis, southern cattail, predominate in the Central and
South Americas, and Scirpus validus (S. tabernaemontani), soft-stem bulrush, occurs
broadly in Australia and the islands of the south and central Pacific Ocean. Cyperus
papyrus is commonly used in Africa (Vymazal, 2013).
Most engineered constructed wetlands are planted with a single species that is
selected because it is native, has long roots, spreads assertively and is characterized
by high biomass production. However, a few constructed wetlands have been
studied that are planted with a variety of species. The Houtan Park wetland shown
in Figure 7.2 is an exception that features an expansive plant palette. Multi-species
wetlands seem to perform as well or better in their treatment function, ecological and
Figure 7.1
The planting design for
constructed wetlands should
match the species to the
hydrology, as in this free water
surface stormwater wetland in
Tianjin, China, designed by
Turenscape. The plants occur
in zones that respond to
seasonal soil moisture and the
depth and duration of
inundation.
Figure 7.2
This free water surface wetland
treats polluted river water and
stormwater, but the consistent
water level creates niches for
various species of plants and
aquatic organisms.
100 PLANTS IN CONSTRUCTED WETLANDS
aesthetic roles (Tanner, 1996). A mixed community horizontal subsurface flow

wetland in subtropical China was planted with 20 plants each of Canna indica, Cyperus
flabelliformis, Phragmites australis, Pennisetum purpureum and Hymenocallis
littoralis. These species were equally spaced and randomly distributed. At the end of
the 4-year study, the Phragmites (common reed) had expanded at the expense of all
other species (the Pennisetum disappeared from the plots altogether). Nevertheless,
the mixed community plots generally showed better performance than the monoculture
wetlands after the first year (Liang et al., 2011).
A study in the Czech Republic also demonstrated that sometimes Phragmites
completely replaced Phalaris arundinacea after several years, but in other instances
Phalaris (reed canary grass) maintained its extent (Březinová and Vymazal, 2014;
Vymazal and Kröpfelová, 2005). A study of five subtropical plants (Acorus calamus,
Canna indica, Cyperus flabelliformis, Phragmites australis and Hymenocallis littoralis)
growing in mixed beds in constructed wetlands demonstrated that the mixed plantings
treated the wastewater as well as single specie beds. However, Canna indica, Cyperus
involucratus and Phragmites australis were more aggressive competitors for light and
space than Acorus calamus and Hymenocallis littoralis (Qiu et al., 2011). Similarly,
Typha latifolia often replaces Scirpus cyperinus and Juncus effusus when they are
planted together.
A review of several studies of mixed vegetation highlights the problem of species
competition and change over time. If, for treatment, ecological or aesthetic reasons, a
community of plants is desired, then careful attention to plant selection (and avoidance
of aggressive non-native species) and other measures will be required to meet the design
goal. Changing the installation method and initial maintenance to favour mixed
communities requires additional study. However, rather than mixing the species
randomly, patches of single species and staged planting times may yield more stable
mixed-species constructed wetlands. Similarly, providing root barriers between plants
Figure 7.3
Aggressive species can expand to
dominate the plant community.
Photo: Forest and Kim Starr. License:
CC-BY-SA-3.0.
that spread aggressively via rhizomes or stolons will reduce maintenance and preserve
the distribution of plants. These techniques were successfully employed at Houtan
Park in Shanghai, China, where the landscape architect specified a highly diverse palette
(see Chapter 8 on ‘riparian wetlands’). The expanded palette was possible in part by
the highly variable wetland conditions that ranged from deep to shallow water, oxygen
abundant and oxygen depleted and other variations (Figure 7.4).
Of course, we need not limit our planting proposals to the area of the constructed
wetland. The margins and even upland zones can be fashioned to provide differing
cultural conditions that favour species not normally associated with constructed
wetlands, such as low growing sedges that require a dry season or plants intolerant
of more than a few inches of inundation (Figure 7.1). In summary, there are abundant
opportunities to expand the diversity, aesthetic impact and ecosystem benefit of the
plant palette through creative and knowledgeable selection of plants and organization
of cultural conditions needed to allow them to thrive.
The sections below describe the cultural, aesthetic, ecosystem and treatment
capabilities of some of the plants that are commonly used in constructed wetlands. It
also features a number of plants that are less commonly used but which show promise
according to initial research. The plants have been divided into groups according to
their ecological adaptation to inundation. The groups are emergent, submerged and
floating. The cultural requirements of wetland plants include tolerance for permanent
and temporary inundation depth. Therefore, species can be grouped according to their
position on a gradient from upland to deep water (Figure 7.5).
Figure 7.4
Carefully controlled water
depth, staged planting and
good maintenance during the
first year resulted in plant
diversity and good aesthetic
quality at Houtan Park,
Shanghai.
Figure 7.5
This image shows a full range
of wetland plants. On the left,
submerged plants colonize
much of the open water
channel, while the right side of
the stream features floating
leaved plants and then
emergent vegetation in the
shallow water.
Emergent vegetation
This type of macrophyte is the primary kind of vegetation in all constructed wetland
types with the exception of those depending on free floating plants. The species listed
below may perform somewhat differently in treatment wetlands depending whether
oxygen is abundant (as in vertical subsurface flow wetlands) or limited (as in horizontal
subsurface flow wetlands). In free water surface wetlands treating stormwater, the
length of the dry season, the depth to groundwater, and the depth and duration of
inundation define cultural niches for the plants. In subsurface flow wetlands, the high
strength municipal or industrial wastewater is better tolerated by some species than
others. In free water surface wetlands, plants are often grouped according to the
normal water depth (Figure 7.6), although the depths assigned for each zone vary
somewhat. Common deep marsh species that grow vigorously in water 12–30 inches
deep (30–76 cm) include Phragmites australis, P. karka, Eleocharis dulcis, Typha
latifolia T. angustifolia, Scirpus californicus, S. grossus, S. mucronatus, Lepironia
articulate and Phylidrium lanuginosum (Sim, 2003).
Common marsh species that grow vigorously in water 6–16 inches deep (15–41
cm) include Thalia geniculate, Rhynchospora corymbosa, Eleocharis variegate, Scleria
sumatrana, Fimbristylis globulosa, Polygonum barbatum and Erioucaulon longifolium
(Sim, 2003). In a stormwater wetland, a normal water level of 12 inches for the shallow
marsh should be designed to fluctuate no more than 6 inches above and below this
level.
The shallowest water (2–6 inches; 5–15 cm) is the zone for wet meadow species,
such as Iris pseudacorus and some Carex. For bio-retention and stormwater basins
or other situations where water fluctuates seasonally, species adapted to an extended
dry season should be specified. Some rushes (Juncus), spikesedges (Eleocharis) and
sedges (Carex) tolerate this alternation of inundation and dry land conditions.
For stormwater wetlands, the normal water level must be well controlled to sustain
the intended vegetation. Stormwater wetland zones include deep-water marsh, marsh,
shallow marsh 2–8 inches (5–18 cm) and ephemeral marsh 0–2 inches (0–5 cm) (dries
out) (Greenway, 2012). The expectation is that the water depth varies seasonally but
Figure 7.6 In natural wetlands, water levels fluctuate and create niches for rooted, submerged
and floating wetland plants. In constructed wetlands, the plant zones can be created as benches.
A – Upland terrestrial plants; B – Wet meadow, 2–6 inches water depth with facultative plants;
C – Shallow marsh, 6–16 inches water depth, with emergent plants and aquatic creepers; D – Deep
marsh, 30–36 inches water depth with emergent and floating-leaved rooted plants; E – Open water,
more than 36 inches water depth with submerged and free floating plants.
Image: Gary Austin adapted from Greenway (Greenway, 2012).
that the depth of inundation over the levels listed above are reduced within days through
the use of outlet structures, infiltration rates or evapotranspiration. Furthermore, the
water depth immediately after planting must be regulated to allow about one-third
of the plant’s foliage to be above the water level during a 6–8-week plant establishment
period.
Poor regulation of the water depth or excessive duration of inundation causes
planting designs to fail. Figure 7.7 illustrates a portion of a planting design that was
implemented at a stormwater wetland in Australia. The plants were distributed
based on water depth in several zones: ephemeral marsh, 14 per cent; shallow marsh,
23 per cent; marsh, 24 per cent; deep marsh, 20 per cent; open water, 19 per cent.
However, after 6 months, the plants, particularly in the marsh and deep marsh areas,
were largely replaced by open water. This was due to poor design of the outlet structure
and the presence of short circuit flows. The more direct flow of water through the
wetland during storm events scoured the soil and uprooted the plants. The outlet
established deeper inundation for longer periods than the plants could survive.
Unrelated but also significant was the uprooting of facultative plants by maintenance
workers without the knowledge to distinguish the wetland plants from weeds. To better
preserve the plants and the treatment capacity of the stormwater wetland, a maximum
inundation depth of 12 inches above the normal water level is recommended. The
water level should return to the elevation of the normal pool within 48 hours. These
general recommendations should be modified depending on local conditions (rainfall
intensity, rainfall frequency) and the plants specified. The velocity of flow of
stormwater through a wetland can damage plants with soft foliage, and some are even
uprooted as stormwater carries soil away from the root zone (Greenway, 2012). In
general, designing the treatment wetlands to receive only a water quality storm while
an adjacent pool contains and transmits larger, less frequent storms will yield a better
treatment and planting result than a single basin that detains large storm events, such
Figure 7.7 The depth and duration of inundation must be managed to sustain the planting design and
treatment performance of stormwater wetlands. This plan represents the installed planting design
establishing the following zones: A – Boardwalk; B – Ephemeral marsh, 14 per cent; C – Shallow
marsh, 23 per cent; D – Marsh, 24 per cent; E – Deep marsh, 20 per cent; F – Open water, 19 per
cent. However, the open water zone was greatly enlarged at the expense of the deep marsh and
marsh zones due to hydraulic problems.
Image: Gary Austin adapted from Greenway (Greenway, 2012).
as the 25-year storm. See Chapter 9 on ‘stormwater’ for more details about on-line
and off-line stormwater wetlands.
In general, it is best to specify native species of plants. For large projects, it is
possible to contract nurseries to grow local wetland plants. This often requires
collection and propagation of seed, so it takes several months to get plants ready after
completing the wetland construction. A botanist might be necessary to identify local
species. However, the landscape architect might be able to do this with the support
of local nurserymen and other resources, such as a species list like those published by
the US Army Corps of Engineers for every state and territory in the United States
(http://rsgisias.crrel.usace.army.mil/NWPL/).
The text below provides the cultural requirements of many emergent wetland plants
and also identifies which plants demonstrate accumulation of nutrients, heavy metals
and other contaminants. Similarly, it identifies plants with outstanding visual charac-
teristics and suitability for tropical, subtropical and temperate climates.
Typha latifolia
The common or broad-leaved cattail is a widely distributed North American native
plant that grows 5 to 10 feet tall (1.5–3 m) and in water 2.6 feet deep (0.8 m) or less
(Figure 7.8). Typha angustifolia (Figure 7.9), native to Europe and northern Asia, is
capable of growth in deeper water than T. latifolia. Along a water depth gradient,
broadleaf cattail often grows upslope from bulrush or open water but downslope from

Broadleaf cattail, Typha latifolia, is an attractive Typha angustifolia has slender seed heads and
and vigorous wetland plant. leaves ( 1⁄4–1⁄2 inch wide). It grows 3–6 feet tall.
Photo: Gary Austin, 2015. Photo: Don Pedro28. License: CC-BY-SA-3.0.
common reed (Phragmites australis), reed canary grass (Phalaris arundinacea) and
willow (Salix spp.).
Cattail often forms extensive patches but also grows in mixed stands with bulrush
and other species. Typha stands expand rapidly when there are changes in nutrients,
salinity and hydrology. In this case, they can replace other native species. T. latifolia
is reported to overwhelm Scirpus cyperius and Juncus effuses in a mixed planting
(Calheiros et al., 2009). To reduce an invasive stand as part of a restoration plan,
mowing cattails after the seed heads are well formed but not mature and again a month
later, when new growth is 2 or 3 feet high, will kill at least 75 per cent of the plants
(Stevens and Hoag, 2006).
The best survival of new plants occurs when dormant rhizomes are planted in
October or November in moist soil after frequent rainstorms. Rhizomes should be
planted about 3 feet apart, except in clay soils where smaller spacing is necessary.
Seed germination is high and growth from seed is effective when established in
aerobic, rather than flooded, soil (Stevens and Hoag, 2006).
Typha is quite effective for the treatment of sewage and other effluents high in
organic nitrogen. For example, Typha was able to treat tannery wastewater with high
biological oxygen demand (BOD; 420–1,000 mg/L) and chemical oxygen demand
(COD; 808–2,449 mg/L) concentrations (Calheiros et al., 2009). The species is also
able to accumulate nitrate and ammonium. Although most wetland plants contain
about the same amount of nitrogen in their aboveground stems and leaves, for cattail,
this is lower than in other common species at 0.8 per cent of the dry matter. Cattail
allocates more of its nitrogen to the rhizomes. The rhizomes are about 1 inch thick
and often over 24 inches long. Because the plant is quite vigorous, it produces a
great amount of biomass. Combined, these characteristics result in a substantial accum-
ulation of nitrogen. According to one study, about 22 g/m2 of nitrogen is harvested
annually from the aboveground portions of the plant when the nitrate and ammonium
concentrations in the water are low (86 g/m2) and 107 g/m2 when the nitrogen
concentration is moderate (222 g/m2). These concentrations are similar to those in
stormwater and agricultural run-off, but lower than for domestic wastewater. The
belowground portions of the plants account for an additional 103 g/m2 of nitrogen
(Borin and Salvato, 2012). One study found that T. latifolia was stressed by ammonia
concentrations averaging 160–170 mg/L (Wu et al., 2015). A study of three free water
surface constructed wetlands in Spain found that Typha latifolia generally removed
greater amounts of nitrate and ammonium than Phragmites australis. These wetlands
treated secondary effluent from a wastewater treatment plant. However, the distribution
pattern and amount of coverage of the treatment cells by emergent vegetation had a
great effect on total nitrogen removal (and hydraulic residence time) (García-Lledó
et al., 2011).
T. latifolia demonstrated the capacity to grow in acidic (3.9–4.3 pH) leachate with
average COD of 3,980 mg/L and BOD of 3,565 mg/L from wood waste piles at
a sawmill. These free water surface constructed wetlands were 16 inches deep with a
7-day residency time (Masbough et al., 2005). T. latifolia can also tolerate total
dissolved solids (primarily calcium, phosphates, nitrates, sodium, potassium and
chloride) concentrations as high as 2,500 mg/L and long-term exposure to copper at
30 mg/L (Valipour et al., 2014).
Cattails provide cover and nesting sites for long-billed marsh wrens, red-wing
blackbirds and yellow-headed blackbirds, but are food for few birds, although geese
eat new shoots and a few duck species consume the seeds. The roots are valuable for
muskrat, while moose and elk eat the spring shoots (Stevens and Hoag, 2006).
Typha domingensis and T. orientalis are very similar species also planted in
wastewater treatment wetlands in Australia and elsewhere.
Phragmites spp.
There are several species of common reed, including Phragmites karka (India, Nepal),
Phragmites japonica (Japan) and Phragmites mauritianus (central Africa); however
Phragmites australis (Figure 7.10) has a nearly global distribution, although it is most
common in Europe and Asia. The attractive plants (Figure 7.11) grow between about
3 feet wide and 15 feet high, and spread primarily via rhizomes rather than seed. The
tall plant grows in dense stands and is anchored by deeply penetrating roots (0.6–1.0
m). The stiff stems are hollow between several nodes (Vymazal, 2013).
P. australis is the most commonly specified plant for both subsurface flow and
free water surface constructed wetlands. Therefore, it has been studied extensively for
its capacity to grow in water with a range of nutrient loads and toxic contaminants.
Phragmites tolerates domestic and municipal wastewater that has been simply screened
or pretreated in a septic tank or similar device. It also tolerates industrial wastewater
(Xu et al., 2010). For example, it is used to treat wastewater from tanneries, pulp
and paper mills, and animal wastewater from dairies and piggeries. It grows
successfully in tannery wastewater with BOD concentrations of 420–1,000 mg/L and
COD concentrations of 808–2,449 mg/L (Calheiros et al., 2009).
Because it is prolific and invasive, Phragmites is avoided in parts of the United
States and New Zealand. It can grow in water with salinity as much as 20 parts per
thousand, making it useful for treatment of brackish water or wastewater that has
high salinity due to high evaporation rates, as in arid areas (Vymazal, 2013).
Figure 7.10
Phragmites australis often
grows in dense stands. In
China’s Liaohe Delta, about
197,700 acres (800 km2) of
natural wetland is managed for
maximum production of
Phragmites biomass.
Harvested each winter, this
wetland produces nearly
440,925 tons (400,000 metric
tons) of biomass that is used to
make paper. Unfortunately, this
commercial enterprise
sometimes includes the use of
pesticides and burning to
maximize yields, resulting in
ecosystem shocks (Brix et al.,
2014).
Photo: Botaurus Stellaris.
License: public domain.
Figure 7.11
Phragmites australis features
attractive seed plumes that
persist through the winter.
Photo: Peter aka
anemoneprojectors License:
CC-SA-2.0.
Phragmites australis, due to its high biomass and density, has the capacity to
accumulate 2.5 kg of nitrogen per hectare and 120 grams of phosphorus per hectare
per year. These are more than twice the amount accumulated by Cyperus papyrus
(1.1 kg N hectare per year and 50 g P hectare per year) in the same setting (Brix,
1994).
Like Typha, Phragmites produces a great deal of biomass and accumulates
nitrates and ammonium. Where there is an input of 222 g/m2 of nitrogen in the form
of nitrate and ammonium, Phragmites accumulates 63 and 83 g/m2 of nitrogen in its
above and belowground portions, respectively. This is somewhat less than that
reported for Typha latifolia (Borin and Salvato, 2012). Whether growing in diluted
or full strength domestic wastewater, Phragmites acquires the same amount of
nitrogen. When growing in the presence of 179 ␮M of ammonium (NH4), this species
shows no reduction in growth rate and biomass production was not reduced at 3,700
␮M NH4, whereas these levels were limiting to another common wetland plant,
Glyceria maxima (Reed mannagrass) (Tylová et al., 2008).
Phragmites effectively accumulates heavy metals present in mixed residential and
industrial effluents. This species performed better than Typha latifolia for the removal
of most metals, although a mixed planting of the two species appears to be most
effective for the removal of copper (78.0 per cent), cadmium (60.0 per cent), chrom-
ium (68.1 per cent), nickel (73.8 per cent), iron (80.1 per cent), lead (61.0 per cent)
and zinc (61.0 per cent) when a 14-day water residence time is provided (Kumari and
Tripathi, 2015). The capacity to accumulate copper, zinc, cadmium and nickel was
also evident from an Italian study, where river water was treated in a free water surface
constructed wetland (Bragato et al., 2009).
In its native range, P. australis is subject to damage by the larva of a stem-boring
moth (Archanara geminipuncta), which kills or damages shoots. Although the general
impact on the plant is negative, this insect also encourages a growth response and
fosters greater biological diversity (Tscharntke, 1999) (Figure 7.12).
There is a North American Phragmites named P. australis subspecies americanus
(Figure 7.13) with equal nitrogen removal characteristics and perhaps greater
Figure 7.12
In its native range, Phragmites
australis hosts a diversity of
insects that attract other
species to the wetland.
Photo: Sebastien Bertru. License:
CC-BY-SA-2.0.
Figure 7.13 Phragmites australis often forms dense single species stands.
Photo: USFS. License: CC-BY-SA-2.0.
phosphorus removal capacity. This subspecies should be specified for constructed

wetlands in the United States and Canada to avoid the invasive nature of P. australis
and to create a more authentic plant community (Rodríguez and Brisson, 2015). The
American subspecies may have been introduced from Europe in the 1700s.
Juncus spp.
Several Juncus species are grown in treatment wetlands including J. effuses (Figure
7.14), J. conglomeratus, J. articulatus, J. balticus, J. filliformis, J. ingens, J. kraussii,
J. roemerianus and J. usitatus. They are most commonly specified in Europe and
especially North America.
Juncus (rush) species are leafless, grass-like annual or perennial herbs with
rhizomes. The species vary in height from less than 1 feet to 41⁄2 feet (25 cm–1.5 m).
Juncus species are quite adaptable, tolerating periods of drought and total inundation.
Stands of rushes provide habitat for amphibians, spawning areas for fish and wading
birds find shelter within the dense vegetation. Muskrats feed on the roots while
waterfowl, songbirds and small mammals, such as jackrabbits, cottontail, muskrat,
porcupine and gophers, eat rush seeds. Juncus species are rarely damaged significantly
by insect or diseases problems (Stevens, 2003).
Juncus effusus (L.), soft rush, is the most commonly specified rush species for
constructed wetlands in Europe and in North America (Vymazal, 2013). It occurs in
wet places below 2,500 m and is distributed throughout California to British
Columbia, the eastern United States, Mexico and Eurasia. The sturdy but soft stems
stand erect and reach 2–5 feet (0.4–1.5 m) in height (Figure 7.15). While it grows in
clumps, a mass of the plants exhibits a grass-like appearance due to the green stems,
but brownish, leaf-like sheaths that are present at the base of the plant (Stevens, 2003).
Soft rush is easy to propagate from the division of rhizomes or seed. Planting
rhizomes or seedlings 10–12 inches (25–30 cm) apart encourages full coverage within
a single growing season. Planting should occur in fall at the beginning of the rainy
season in aerobic rather than flooded soil. Seed can be collected in August and seed
germinates in the greenhouse (32–38°C) in about a week. The plants can be installed
in the field within about 3 months, but older, larger plants have better survival.
Established plants are vigorous and can become invasive (Stevens, 2003).
Figure 7.15
The Juncus effusus inflorescence appears to
emerge from the side of the stem.
Photo: Pethan Houten. License: CC-BY-SA-3.0.
Figure 7.14
Juncus effusus.
Photo: Christian Fischer. License: CC-BY-SA-3.0.
In natural wetlands and free water surface constructed wetlands, muskrats will
use Juncus and other species for food and hut construction materials. The animals
typically forage and clear an area around their huts. These open water areas increase
the ecological diversity of the wetland (Stevens, 2003).
Juncus effusus accumulates significant amounts of arsenic within its rhizomes
but not within its stems when exposed to contaminated water (simulated secondary
wastewater effluent) in a subsurface flow tank, according to a laboratory study
(Rahman et al., 2014). Similarly, when J. effusus plants are suspended from floating
rafts in stormwater treatment ponds, the stems and roots accumulate nickel and zinc.
However, the roots accumulated 5.7 and 2.6 times more nickel and zinc, respectively,
than the stems (Ladislas et al., 2014).
Other Juncus species such as Juncus articulatus, jointleaf rush, in Asia, Juncus
conglomeratus (this is a variety of J. effusus) in Europe, or Juncus kraussii in Oceania
have been used only occasionally in constructed wetlands (Vymazal, 2013).
Eleocharis spp.
The genus Eleocharis (spikesedge) contains more than 250 species broadly distributed
across the globe (Figure 7.17). Those reported for use in constructed wetlands are E.
acuta, E. dulcis, E. fallax, E. obtusa, E. palustris (Figure 7.18), E. quadrangulata (Figure
7.16) and E. sphacelata (Figure 7.19). The use of these species is most common in
Asia and especially in North America.
Figure 7.16
Squareside spikerush, Eleocharis
quadrangulata, is native to Mexico and the
United States.
Photo: Mike Ryon. License: government work
product.
Figure 7.17
Eleocharis acicularis, needle
spikerush, has a global
distribution.
Photo: Mike Ryon. License:
government work product.
Figure 7.18
Eleocharis palustris, common spikerush, grows in
USDA hardiness zones 4–8.
Photo: Karelj. License: CC-BY-SA-3.0.
The species range widely in height to about 10 feet tall for Eleocharis sphacelata
(Australia and New Zealand, Figure 7.19) and form patches or mats that are extended
by rhizomes or stolons (Vymazal, 2013). The stems are photosynthesizing while the
leaves are inconspicuously located at the base of the plant. The flowers are simple
and inconspicuous.
Four species are common to constructed wetlands in North America. Eleocharis
palustris, Common spikerush (Figure 7.18), is a widely distributed perennial in the
United States that grows to 4 feet tall and is characterized by a dense root mat that
extends more than 16 inches deep (40 cm). It can grow in water up to 3 feet deep for
3–4 months. It can also grow in drier soil where groundwater is available within
30 inches during the dry season (Ogle, Tilley and John, 2012).
Figure 7.19
Eleocharis sphacelata, tall
spikerush, grows in USDA zone
8, tolerating temperatures from
10 °F to 20 °F (–12 °C to –6.5 °C).
Photo: John Tann. License:
CC-BY-2.0.
Figure 7.19 illustrates tall spikerush, Eleocharis sphacelata. It is frequently planted

in constructed wetlands in Australia and New Zealand where it is native. This is a
perennial species with sturdy rhizomes and stems. It is unusual in its ability to grow
in deep water (up to 6 feet, 2 m).
Eleocharis fallax (creeping spikesedge) is native to southern and eastern United
States; Eleocharis obtusa, blunt spikerush, generally is 1 feet tall, while Eleocharis
quadrangulata (Figure 7.16), squarestem spikerush, grows 1–3 feet tall.
Eleocharis dulcis (Chinese water chestnut), a freshwater rhizomatous perennial
emergent macrophyte (Figure 7.20), is most common in free water surface constructed
wetlands in Asia and Oceania. Stems of this species are hollow, have an average
diameter of 5 mm and grow to 6 feet tall (2 m) (Vymazal, 2013).
Phalaris arundinacea
Phalaris arundinacea, reed canary grass (Figure 7.21), is distributed widely across the
globe. It is cold hardy to –35°F (USDA zone 4) (Stannard and Crowder, 2002).
This plant grows 2–9 feet tall (0.6–2.7 m) and has a medium texture with leaves
3.5–10 inches long (9–25 cm) (Figure 7.22). It sprouts early in the spring and forms
a dense root and rhizome network that dominates the soil mass (Stannard and
Crowder, 2002). In constructed wetlands, the roots grow 8–16 inches long (20–40
cm) (Vymazal and Kröpfelová, 2005). This plant displaces other plants that sprout
later or are deprived of light by the tall dense vegetation. In some temperate climate
horizontal subsurface flow wetlands, Phragmites encroaches on Phalaris, replacing it
after several years but in other instances this has not been observed (Březinová and
Vymazal, 2014; Vymazal and Kröpfelová, 2005).
This grass produces as much as 9 tons per acre (920 pounds of nitrogen per N
acre) of biomass (but less biomass than Phragmites). In fact, it is sometimes cultivated
as a biofuel feedstock in Scandinavia and as a reliable forage crop on poorly drained
soils. Reed canary grass expands rapidly via new shoots along its rhizomes, but the
stems are also capable of forming roots. Reed canary grass also grows from seed (it
produces 30–50 pounds of seed per acre). Moderate germination rates occur after a

Chinese water chestnut, Eleocharis dulcis, grows Phalaris arundinacea.
in USDA zones 9–11, tolerating temperature to Photo: Michael Becker. License: CC-BY-SA-3.0.
25 °F (–3.9 °C).
Photo: Robyn Jay. License: CC-BY-SA2.0.
Figure 7.22
Phalaris arundinacea is also sold as an
ornamental plant as hybrids with variegated
foliage or other characteristics.
Photo: Patrick Standish. License: CC-BY-SA-2.0.
few days of cool temperatures. Seedlings are sensitive to competition, especially when
the soil is dry. When established, the plant survives as much as 9 months of continuous
grazing and responds well to as little as a 2-week grazing hiatus. Spring flooding is
well tolerated, as is burning.
Reed canary grass is often planted in constructed wetlands or used as the target
for irrigation disposal of municipal and industrial wastewater (Stannard and Crowder,
2002). In the Pacific Northwest region of the United States, the plant is aggressive
and invasive, causing problems for riparian restoration as it displaces native species.
Prolonged deep inundation can be used to control the growth of P. arundinacea.
Phalaris biomass can be reduced by nearly 70 per cent through shading from the
planting of willows (Scouler willows, Salix scouleriana, and Pacific willows, Salix
lasiandra) at 2 feet (0.6 m) apart in moist soils. This is less effective where the willows
and reed canary grass are exposed to drought stress (Kim, Ewing and Giblin, 2006).
Phalaris accumulates metals in its stems and leaves when growing in municipal
wastewater. In a temperate climate, harvesting the aboveground biomass in May or
June is optimum for removal of zinc, copper and nickel. Harvest in September is better
for the removal of chromium, in July for removal of cadmium, and in August for
removal of lead (Březinová and Vymazal, 2015). However, this plant is not particularly
tolerant of high concentrations of zinc compared to Phragmites australis, Typha
latifolia, Glyceria fluitans, Eriophorum angustifolium, Festuca rubra and Carex
rostrata (Matthews, Moran and Otte, 2006).
Scirpus (Schoenoplectus) spp.

Bulrushes, Scirpus, are annual and perennial herbaceous plants that grow to 10 feet
tall (3 m) in some species. Root penetration is 27–31 inches deep (70–80 cm). The
plants lack leaves and the stems range from sharply to softly triangular in cross section
(Vymazal, 2013).
In Germany, Dr Kathe Seidel planted Scirpus lacustris (Common clubrush) in her
early wastewater treatment experiments. This species is planted in Europe, while other
species are commonly part of constructed wetlands in North America, including Scirpus
acutus (Hardstem bulrush), S. americanus (Chairmaker’s bulrush), S. californicus (Giant
bulrush) and S. cyperinus (Woolgrass) (Vymazal, 2013).
Scirpus californicus (Figure 7.23) is native to temperate South America, the
southern United States, California and Oregon where it is an important species for
marsh and shore birds. Marshes in California dominated by S. californicus are the
primary habitat for the endangered Santa Cruz long-toed salamander and the rare
giant garter snake. Seedlings should be planted 3 feet apart in aerobic, not flooded,
Figure 7.23
This image of Scirpus
californicus illustrates the
inundation tolerated by this 5–8
feet tall macrophyte. It grows in
USDA hardiness zone 8,
tolerating temperatures of
10 ° to 20 °F (–12.5 ° to 6.7 °C).
Photo: US National Park Service,
Santa Monica Recreation Area.
product.
soil. In a constructed wetland, S. californicus provides good removal of aluminium,

copper, iron, lead, manganese, zinc and total petroleum hydrocarbons.
River clubrush, Scirpus validus (Schoenoplectus tabernaemontani), is the primary
bulrush species planted in constructed wetlands in Australia and New Zealand
(Vymazal, 2013), but is distributed throughout the United States and much of Canada.
It grows naturally in fresh and brackish water, along creeks, lakes and marshes, but
grows best in saline conditions. Planting 3 feet apart yields complete coverage within
one growing season. S. validus grows very rapidly in primary wastewater from dairies,
yielding an aboveground biomass of 2 kg/m2 of dry weight and a belowground biomass
weight of 1.25 kg/m2 in a horizontal subsurface flow constructed wetland (Tanner,
1994). Some studies report a shallow root depth (10–30 cm) for S. validus, which
grows in small groups or large colonies (Figure 7.24).
One microcosm study of S. validus revealed that the plant preferentially acquired
ammonium when both ammonium and nitrate were available. This suggests that it is
well adapted for growth in horizontal subsurface flow wetlands, where ammonium
is generally more abundant than nitrate. S. validus absorbed more ammonium than
Canna indica under the same conditions (growing in simulated secondary effluent).
Similarly, S. validus absorbed slightly more phosphorus (PO4) than Canna (Zhang,
Rengel and Meney, 2009). S. validus also demonstrates the ability to tolerate high
levels of ammonia (160–170 mg/L) (Wu et al., 2015), although a second study
showed biomass reduction when ammonia was in excess of 100 mg/L (Clarke and
Baldwin, 2002). Another study noted root death when ammonia concentrations
reached an average of 222 g/m3 (Tanner, 1994). The highest accumulation of total
nitrogen seems to require a 7-day water residence time in a horizontal subsurface flow
wetland treating primary dairy influent. At this residence time, total nitrogen and total
phosphorus reduction in mass reached as much as 75 per cent and 74 per cent,
respectively (Tanner, Clayton and Upsdell, 1995).
Scirpus maritimus (Bolboschoenus maritimus) is distributed widely around the
world and valuable due to its tolerance to alkalinity and salt (Figure 7.25). It is often
Figure 7.24
River clubrush, Scirpus validus,
grows in USDA zones 3a to 9b,
tolerating temperatures as low
as –40 °F.
Photo: Matt Lavin. License:
CC-BY-SA-2.0.
Figure 7.25
Scirpus maritimus grows in
USDA zone 4 and warmer
tolerating temperatures of –30 °
to –20 °F (–34.4 ° to –28.9 °C).
Photo: Forest and Kim Starr.
License: CC-BY-SA-3.0.
a pioneer species that grows to 4.5 feet tall and tolerates constant water depths of
12 inches and inundation to 3 feet (Tilley, 2012). It naturally colonized portions of
the large constructed wetland near the Venice Lagoon (Italy) where Phragmites
had been planted. While it grew vigorously there, it accumulated less nutrients and
heavy metals (with the exception of sodium) than common reed (Bragato, Brix and
Malagoli, 2006).
Carex
There are a couple of thousand species in this genus, which occur across most of the
globe. Most of these are native to wet environments of various types. With so many
native species to choose from, avoiding the use of invasive Carex species is clearly
possible by collecting plants or purchasing them from regional nurseries. Many species
of Carex are planted in constructed wetlands including C. apressa, C. aquatilis, C.
cristella, C. elata, C. frankii, C. lacustris, C. lurida, C. nebrascensis (Figure 7.27), C.
normalis, C. obnupta, C. rhynchophis, C. riparia, C. rostrata and C. stipata (Figure
7.28). The most commonly planted of these is C. riparia (Figure 7.26). This species
is native to Europe where it is widespread occurring as far south as North Africa and
east to China. It is naturalized in New Zealand and occurs in South America. It grows
to 3–4.5 feet (1–1.5 m) and can grow in water as much as 18 inches deep.
Carex nebrascensis, Nebraska sedge, is an example of a North American plant
that is widely distributed through the Midwest, west and southwest United States,
and central Canada. It grows about 3 feet tall and has inch-wide grass-like leaves.
C. nebrascensis tolerates total inundation for as long as 3 months.
Figure 7.26
Carex riparia, grows 4 feet (130 cm) and is
hardy in USDA hardiness zones 5–9.
Photo: Karelj. License: Public Domain.
Figure 7.27
Carex nebrascensis, Nebraska Sedge grows in
USDA zones 4–8.
Photo: Andrey Zharkikh. License: CC-BY-2.0.
Figure 7.28
Carex stipata (Awl-fruit sedge)
is distributed throughout
Canada and almost all of the
United States. It grows to 36
inches tall and tolerates shade.
It grows in USDA zones 3–9.
Photo: Matt Lavin. License:
CC-SA-2.0.
In its natural setting, this species sometimes forms dense stands, but generally it
is the dominant plant associated with water sedge (C. aquatilis), common spikerush
(Eleocharis palustris), cattail (Typha spp.), bulrush (Scirpus spp.) and other wetland
plants. C. nebrascensis forms a dense mat of roots and rhizomes making it useful for
erosion control and wastewater treatment. The plant requires a season where the soil
dries somewhat, making it valuable for stormwater basins. It should be planted from
seedlings 6 or 24 inches apart for complete coverage in 1 or 3 years, respectively.
Planted seedlings require water depth of 1–2 inches until they are about 12 inches
tall. C. nebrascensis can be propagated from seed in a greenhouse (Boonsaner,
Borrirukwisitsak and Boonsaner, 2011; Calheiros, Rangel and Castro, 2009).
Carex aquatilis, water sedge, is distributed throughout Canada and all but the
southeast quadrant of the United States, although it is threatened or endangered in
Connecticut, New Jersey and Pennsylvania. The plant grows where the water table
is high in bogs, stream and lake margins and wet meadows in monoculture stands or
mixed with other sedges, rushes and bulrushes, willow, birch and cottonwood. It can
withstand as much as two months of inundation, but normally grows in only a few
inches of water. C. aquatilis grows 6–40 inches (15–100 cm) tall and at temperatures
as low as –4°F (–20°C) (Tilley, 2011).
Water sedge is eaten by bison, caribou, mule deer, white-tailed deer and elk as
well as by geese, swans and other waterfowl. The vegetation also provides a source
of cover for birds and, as C. Aquatilis sod overhangs streams, it provides cover for
fish. The sod is also an erosion protection. Propagation is by division of rhizomes
but grows readily from seed in a greenhouse. Seedlings should be planted 6 inches
apart (15 cm) for full coverage in 1 year (Tilley, 2011).
Iris
Iris pseudacorus is native to Europe, western Asia and northwest Africa, and is one
of the best choices for treatment of domestic sewage in constructed wetlands. It grows
39–59 inches (100–150 cm) tall with erect leaves 35 inches tall (90 cm) and 1.2 inches
wide (3 cm). This species can be invasive and difficult to eradicate outside of its native
range. In the United States, its use is restricted in Connecticut, Massachusetts,
Montana, New Hampshire, Oregon and Washington. The plant is very hardy (USDA
zone 4, –30°F). The attractive and long-lasting flowers are a potential economic crop
(Liang et al., 2011; Zhang et al., 2007) (Figure 7.29).
I. pseudacorus accumulates heavy metals and is reported to effectively remove
pathogenic bacteria from wastewater (Tilley, 2011). Significant amounts of cadmium
and zinc were accumulated by I. pseudacorus according to one study, where the metal
concentration in the plant increased as exposure to increasingly polluted water
increased. Cadmium was concentrated in the roots and rhizomes, while zinc also
accumulated in the leaves. These results suggest that this species is valuable for
remediation of soils contaminated with heavy metals (Zhang et al., 2007).
High yield from photosynthesis translates into long root length (18–20 cm). The
root biomass of this Iris is moderate in diluted effluent and low in full strength
wastewater. The roots are very porous, which facilitates transfer of oxygen to the
area around the roots. Accumulation of total nitrogen is high (moderate for ammon-
ium but high for nitrate), for phosphorus accumulation is moderate (Li et al., 2013).
This plant performs better in diluted than in full strength sewage, suggesting that
Figure 7.29
Yellow flag iris, Iris
pseudacorus, is one of the
most highly ornamental of
wetland plants. The flowers are
3–4 inches across.
constructed wetland designs with a portion of recirculated effluent might be most

suitable.
A microcosm study suggests that I. pseudacorus is a good choice for floating
wetlands intended to remove nitrogen and phosphorus from surface water. In the study,
Iris plants removed 277 mg of nitrogen per m2 per day and 9.32 mg of phosphorus
per m2 per day during the growing season. The removal of 74 per cent of the total
nitrogen and 60 per cent of the total phosphorous resulted from plant uptake (Caldelas
et al., 2009).
Iris sibirica, native to Turkey, Russia, and eastern and central Europe, grows 24–36
inches tall (60–90 cm) with highly ornamental blue or purple flowers. I. sibirica appears
to be a viable plant for use in constructed wetlands treating polluted river water
according to a microcosm study (Li et al., 2013). An advantage of this species is its
cold hardiness (USDA zone 4a, –34°F).
Sagittaria
Three species of Sagittaria have been planted in constructed wetlands. Sagittaria
latifolia, Common arrowhead (Figure 7.30), is distributed throughout the United States,
southern and central Canada, Mexico, and is naturalized in Europe (USDA Zone 4a:
(–30°F) to zone 11). Swans, geese, many species of ducks, and king rail eat the plant’s
Figure 7.30
Sagittaria latifolia, Common
arrowhead.
Photo: Mike Ryon License:
seeds and tubers, while muskrat, beaver and porcupine eat the tubers. The plant
can be propagated by transplanting live plants or tubers or from seed, although it
takes two years for the seed to germinate. Plants should be spaced 10–12 inches apart
(25–30 cm) for full coverage within one growing season (Figure 7.31). The transplants
should be covered with 1⁄2 inch of water until establishment. The mature plants are
18–24 inches tall and sustained by water 6–18 inches deep (0.15–0.45 m) where there
is little or no current. S. latifolia was an important food source for indigenous people
in the Americas. Sagittaria cuneata is a similar aquatic species (Stevens, 2003). In a
microcosm study, this species grew well when exposed to nitrate, zinc, copper, lead
and cadmium levels typical of urban stormwater (Kearney and Zhu, 2012). Ammonia
levels below 200 mg/L are tolerated (Clarke and Baldwin, 2002).
Figure 7.31
Sagittaria latifolia.
Sagittaria lancifolia is native to Mexico, Central America and northern South

America, and grows 24–36 inches tall (60–90 cm). The plants are easy to propagate
by dividing rhizomes or by seed and flourishes in USDA Zone 5a (–20°F) to zone 11.
Sagittaria graminea is native to the United States and Canada, and grows in USDA
Zone 5a (–20°F) to zone 9b. The grass-like leaves grow 18–24 inches (45–60 cm) or
more in water 2–4 inches deep (2–10 cm).
Pontederia
Pontederia cordata (P. sagittata), Pickerel weed (Figure 7.32), is native and widely
distributed in North and South America and is hardy in USDA zones 5–11. Pickerel
weed grows 2–3 feet tall with large erect leaves emerging from the base of the plant.
The 6–8 inches long, purplish-blue flower spikes bloom from May to October.
Propagation is by seed or division. To establish by seed, distribute 20–30 seeds per
square foot. The plant forms a mat of rhizomes and roots just below or covering the
surface of the soil. The root mat and sturdy plant make this a valuable erosion control
species. It tolerates flooding to 20 inches deep, but permanent inundation only 6–12
inches deep, and tolerates salinity of only less than three parts per thousand. Birds
feed on the seeds and the young leaves are edible (Miller, 2012).
Pontederia cordata has a demonstrated capacity to grow in a subsurface flow
wetland receiving wastewater from a molasses distillery with high concentrations of
BOD (533 mg/L), COD (1,181 mg/L) and SO4 (75 mg/L). This wetland achieved a
concentration reduction of over 80 per cent for BOD and COD and 69 per cent for
sulphate. This species also accumulates total Kjeldahl nitrogen equalling or exceeding
Phragmites (Common reed) and Scirpus (Bulrush) (Olguín et al., 2008).
Caltha palustris
Caltha palustris, marsh marigold (Figure 7.33), grows 2–18 inches (30–45 cm) in USDA
Zone 3a (–40°F) to zone 11, and is widely distributed throughout the northern
hemisphere. It has attractive yellow flowers 1–2 inches (2–5 cm) in diameter. It grows
Figure 7.32
The coarse texture and the
attractive flower make Pickerel
weed, Pontederia cordata, a
good choice where aesthetics
is important and where water is
shallow.
Photo: Joshua Mayer. License:
CC-BY-SA-2.0.
Figure 7.33
Caltha palustris, Marsh
marigold, creates mats of dark
green foliage with contrasting
yellow flowers.
Photo: Wildfeuer. License:
CC-BY-SA-3.0.
in water 0–6 inches deep and tolerates shade. It is propagated by seed or by dividing
the root ball. There are several hybrids in the nursery trade. C. palustris is much more
sensitive to high sulphide concentrations than J. effusus.
Canna
Canna indica is a highly ornamental wetland plant due to its large leaves (Figure 7.34)
and showy flowers that have been emphasized through the development of many
hybrids. It is native to tropical Central and South America but naturalized across much
of the globe including the south-eastern United States. It is frost tender and appropriate
for USDA hardiness zone 8. Depending on the variety, the plant grows between 1.5
feet and 7.5 feet tal (0.5–2.5 m).
Canna grows along streams and on disturbed land, but is often planted as an
ornamental species in gardens and parks. C. Indica (Figure 7.35) preferentially
accumulates nitrates rather than ammonium (Zhang, Rengel and Meney, 2009).
C. indica is often planted in horizontal subsurface wetlands and performs well, as it
can utilize nitrate as its nitrogen source for rapid growth. However, the plant uses
ammonium when nitrate is less available and is commonly planted in free water surface
and vertical subsurface wetlands. A study of Canna’s growth in the tropics estimated
the accumulation of nitrogen at 0.57 g/m2 per day of which 0.23 g/m2 per day was
translocated to the aboveground portions of the plant. These high rates of nitrogen
accumulation in the leaves and stems allow the removal of an estimated 85 grams of
nitrogen per m2 per year through annual harvest (Wang et al., 2014).
Canna is an excellent plant for treatment of wastewater due to the length of its
roots (18–20 cm), large root biomass and high yield of aboveground leaves and stems.
It grows well in both diluted and undiluted septic tank effluent (Li et al., 2013). The
roots are thin (about 70 per cent are less than 1 mm in diameter) and fibrous but
abundant (Figure 7.36). This results in a large root area (22,832 cm2 per plant after

Two Canna hybrid varieties within a shopping Canna indica is attractive due to its coarse
area wetland. texture and flower colour.
Photo: Gary Austin, 2015. Photo: Forest and Kim Starr. License:
CC-BY-SA-3.0.
Figure 7.36
Canna indica has long fibrous
roots and spreads rapidly. It is
reported to overwhelm Scirpus
validus in a mixed planting
(Calheiros et al., 2015).
Photo: Maja Duamt. License:
CC-BY-2.0.
21 weeks of cultivation in containers). This growth occurred with a 50 per cent dilution
of domestic wastewater (COD – 76.08 ± 11.09 mg/L, total nitrogen – 11.43 ± 1.68
mg/L, total phosphorus – 0.56 ± 0.21 mg/L, phosphorus – 0.47 ± 0.20 mg/L, nitrate
– 1.11 ± 0.01 mg/L, ammonium – 5.14 ± 1.13 mg/L) (Chen et al., 2007).
Canna generalis grows in USDA zones 7b: 5°F (14.9°C) to 11, and is also native
to South and Central America but common as an introduced plant in other tropical
and subtropical areas including southeast United States. It has been investigated
according to its ability to accumulate solvents and petroleum refinery wastes. Benzene,
toluene, ethylbenzene and xylenes are accumulated in the roots and rhizomes, and
translocated to the shoots. At typical contamination levels, C. generalis is able to remove

about 80 per cent of these pollutants from the root zone in 21 days (Boonsaner,
Borrirukwisitsak and Boonsaner, 2011). Canna species have also been shown to remove
heavy metals and pesticides.
In conclusion, Canna is one of the best choices for constructed wetlands treating
domestic sewage (Li et al., 2013), where it thrives in tropical and subtropical climates
(Figure 7.37). In addition, since the plant and its hybrids are ornamental due to their
flowers and large, attractive leaves, Canna is valuable where aesthetics is an important
function (Kearney and Zhu, 2012). Canna indica and other ornamental wetland plants
(Canna flaccida, Zantedeschia aethiopica, Canna indica, Agapanthus africanus and
Watsonia borbonica) were planted in a horizontal subsurface wetland at a resort in
Portugal. In addition to excellent wastewater treatment, the wetland provided
attractive and cut flowers for the facility (Calheiros et al., 2015).
Figure 7.37
Canna flaccida is native to
southeast United States, where
it grows in wetlands and is
hardy to USDA zone 10.
Photo: Peter A Mansfeld.
Figure 7.38
Canna flaccida in Florida.
Cyperus
Cyperus papyrus (Figure 7.39) is characterized by 5 inch long (12 cm) roots, moderate
root biomass but relatively high above ground biomass production. Its removal of
COD is high and its removal of BOD is moderate, compared to other common high
performance plants such as Canna and Iris. The accumulation of nitrogen is high,
but the accumulation of phosphorus is relatively low (Li et al., 2013). One study
demonstrated that the plant grew well when dissolved oxygen was limited (0.53–0.98
mg/L) (Kyambadde et al., 2004).
A number of other species in this genus are planted in constructed wetlands in
tropical climates including C. involucratus, Umbrella papyrus (Figure 7.40), which
grows 4–6 feet (1.2–1.8 m) (USDA Zone 8a: to 10°F, –12.2°C). This attractive plant
Figure 7.39
Cyperus papyrus is a soft
textured, graceful plant that
forms dense stands.
Photo: Hans Hillewaert. License:
CC-BY-SA-4.0.
Figure 7.40
A Laysan Albatross nestled in Cyperus
involucratus on Sand Island, Midway Atoll.
CC-BY-SA-3.0.
is noteworthy for its fine, fibrous root system and good treatment performance. One
free water surface microcosm study in subtropical China found that this plant
developed more biomass in monoculture and mixed stands than Phragmites (Qiu et
al., 2011), although a different microcosm study found its total biomass production
to be lower than Phragmites australis or Canna indica (Li et al., 2013).
Arundo
Arundo donax is native to eastern Asia, although it is globally distributed in tropical
and subtropical climates. It grows 15–20 feet (4.7–6 m) in floodplains, but it also
grows along low gradient streams in Mediterranean climates. Arundo donax grows
in USDA zones 6a: 10°F (–23.3°C) to zone 10b. Giant reed (Figure 7.42) is capable
Figure 7.41
Cyperus papyrus. Growth has
occurred on the surface, but
without enough vertical roots in
this subsurface gravel wetland.
This may cause the plants to
fall over and not contribute to
the treatment process. The new
shoots are removed and
replanted.
Photo: Heike Hoffmann, 2008.
Figure 7.42
Arundo donax growing in
northern California.
Photo: USDA. License:
government work product,
CC-BY-SA-ND-2.0.
of growing in reduced oxygen conditions due to long periods of inundation. It is

potentially valuable as a lignocellulosic feedstock for paper pulp and manufacture of
nylon due to its rapid growth, high biomass and since it is a non-food crop that grows
in areas unsuitable for food production (Pompeiano et al., 2015). For these reasons,
it also has potential as a feedstock for biofuel production. Experiments demon-
strate that biofuel and biochar production is feasible (Saikia et al., 2015). A. donax
performed well when planted in a horizontal subsurface flow wetland in Morocco
(El Hamouri, Nazih and Lahjouj, 2007).
A. donaciformis is an endangered species that could be valuable in stormwater
wetlands, where it is native in southern Europe. This use could mitigate the loss of
wetland habitat in urban areas (Hardion et al., 2015).
Thalia
Two species of Thalia are specified for use in constructed wetlands. The most
ornamental is Thalia dealbata, Powdery Thalia (Figure 7.43). This dramatic tropical
and subtropical plant grows in USDA zones 6a: 10°F (–23.3°C) to zone 10b. It grows
4–6 feet (1.2–1.8 m) tall, has large blue-green leaves and attractive purple flowers.
This species can tolerate water depths of 1.5 feet during the growing season and deeper
water during the dormant season. Propagation is by division of rhizomes and seed.
Seedlings should be planted 24 inches apart for full coverage in one growing season.
New plants should be established with 1–2 inches water depth (water should not cover
more than 2⁄3 of the shoot) (Moss, 2011). The plants should be established before
contaminated water is added. T. dealbata has a demonstrated ability to flourish in
polluted river water and pretreated domestic sewage.
Figure 7.43
Thalia dealbata is a striking and versatile plant
for treatment wetlands in warm climates.
T. geniculata, alligator flag (Figure 7.44), is native to Africa, Central America and
occurs in southeast United States. This species is taller (8–10 feet) and grows in USDA
zones 7a (0°F, –17.7°C) to zone 10b. In one study, this plant grew very rapidly when
transplanted into both horizontal subsurface and free water surface wetlands treating
domestic sewage in El Salvador (Katsenovich et al., 2009).
Zantedeschia aethiopica
Zantedeschia aethiopica, Calla lily (Figure 7.45), is a widely distributed plant native
to southern and eastern Africa. It grows in aquatic as well as in terrestrial settings
and is a common ornamental and market flower in warm climates (USDA zones 8–10).
The famous flowers are white, yellow or light pink, standing slightly above large
leaves. The plants grow rapidly (to 24 inches, 50 cm tall) in pretreated domestic sewage,
producing substantial biomass and roots that are 10 inches long (25 cm) within a few
months (Belmont and Metcalfe, 2003). Calla lily grew more vigorously in a horizontal
subsurface flow wetland than in a vertical subsurface flow wetland setting when
exposed to the same concentration of domestic sewage (Zurita, De Anda and Belmont,
2009). Zantedeschia aethiopica and Canna indica in a mixed planting, in a horizontal
subsurface flow wetland, performed as well as Typha latifolia in treating domestic
sewage after pretreatment.
Heliconia
There are several species of Heliconia that might be suitable for use in treatment
wetlands. Heliconia psittacorum (Figure 7.46) has been investigated and shown to
grow well in a horizontal subsurface flow constructed wetland used to treat domestic
wastewater. While its aesthetic and economic contributions are positive, this species
performs much less well as a treatment plant compared to Canna indica in terms of
root length, biomass, nitrogen accumulation, etc. (Konnerup, Koottatep and Brix,
2009). Therefore, it might be included as a subordinate planting in a constructed
wetland and located where its benefits can be appreciated and managed effectively.
Figure 7.44
Thalia geniculata.
Figure 7.45
Zantedeschia aethiopica.
Photo: Carrie Kellenberger
Hillewaert. License: CC-BY-2.0.

Heliconia psittacorum is a tropical and beautiful Acorus gramineus, Japanese sweet flag, has
wetland plant with economic value. attractive grass-like leaves 6–14 inches long
(15–35.6 cm).
Photo: Pinus Hillewaert. License: CC-BY-SA-3.0.
Photo: Daderot. License: CC-BY-SA-3.0.
Acorus gramineus
Acorus gramineus (Figure 7.47), Japanese sweet flag (USDA Zones 6–9), is a
commonly planted ornamental plant native to eastern Asia. It grows in water 3–6
inches deep (7.6–15 cm) and is not tolerant of drought. It is uncommonly grown in
constructed wetlands but can contribute to treatment wetlands while providing a
commercial horticultural product (rhizomes can be divided in the spring). There are
several varieties that have been developed for the ornamental landscape. A. gramineus
is highly tolerant of domestic wastewater that also contains significant concentrations
of heavy metals (Zhang et al., 2007).
Agapanthus africanus
Native to South Africa, A. africanus (Figure 7.48) grows 24–36 inches (60–90 cm)
and features umbels of fragrant violet flowers that are attractive to bees, butterflies
and birds. It is hardy in USDA Zone 8a: 10°F (–12.2°C) to zone 11.
Agapanthus growing in a vertical subsurface flow wetland treating domestic
sewage produced more leaves and more long-lasting flowers than when growing under
the same conditions in a horizontal subsurface flow wetland (Zurita, De Anda and
Belmont, 2009). In one study, more aggressive wetland plants, such as Phragmites,
quickly overtook the Agapanthus. The plant does not require inundation, so it might
be planted at the margin or even outside the treatment wetland rather than mixed
randomly where its height and slower growth puts it at a disadvantage.
Strelitzia reginae
S. reginae, Bird of paradise (Figure 7.49), is native to South Africa but is widely
distributed in warm climates around the world. It grows in USDA zones 9b: 25°F
(–3.8°C) to zone 11. The plant grows 4–6 feet tall (1.2–1.8 m) and 3 feet (1 m) wide
with bold leaves and flowers.

Agapanthus is a very attractive plant that is The exotic flowers of the bird of paradise make it
commonly grown as an ornamental plant in an attractive prospect for vertical flow wetlands
warm climates. in warm climates.
Photo: Dezidor. License: CC-BY-SA-3.0. Photo: Martina Nolte. License: CC-BY-SA-3.0.
Growing in a vertical subsurface flow wetland treating domestic sewage,

S. reginae produced more and healthier flowers and bigger leaves than when growing
under the same conditions in a horizontal subsurface flow wetland (Zurita, De Anda
and Belmont, 2009). As with Agapanthus, this suggests that the species does not tolerate
low-oxygen environments as well as other wetland species.
Watsonia borbonica
W. borbonica, bugle lily (Figure 7.50), is a long-flowering plant that is native to South
Africa that grows in USDA zones 9a (20°F) to 10b. The plant is 4–5 feet tall with
2.5-inch long fragrant flower spikes. The iris-like leaves are 2.5 feet tall. This species
was grown successfully in a horizontal subsurface wetland treating domestic sewage
(Calheiros et al., 2015). However, in its native range, it is often located in seasonal
wetlands, suggesting that it would be useful in stormwater bio-retention basins.
Figure 7.50
Watsonia borbonica blooms
during the wet season and is
dormant during the dry season
in South Africa.
Figure 7.51
Hymenocallis littoralis is
another tropical and
subtropical plant with good
treatment capacity and
outstanding visual appeal.
Photo: Vicki. License: CC-BY-
2.0.
Hymenocallis littoralis
Spider lily is native to Latin America but has been introduced or naturalized in many
tropical and subtropical regions (Figure 7.51). There are species of Hymenocallis that
are native to southeast United States. The plant grows 18–24 inches (45–60 cm) with
5–7 inches long (14–17 cm) fragrant flowers. Hybrids have been developed for the
ornamental market. Spider lily grows in USDA zones 6a: –10°F (–23.3°C) to zone 11,
and its growth rate and accumulation of root and aboveground biomass make it
competitive with Canna flaccida. Staged planting might allow it to compete with
Phragmites and other fast growing wetland species. One study demonstrated that this
species was effective in removing contaminants from seafood processing wastewater
that was high in organic matter and nitrogen when planted in a free water surface
wetland in Thailand (Sohsalam, Englande and Sirianuntapiboon, 2008).
Bacopa monnieri
Bacopa monnieri, water hyssop, is native to India and Asia, but is globally distributed
in warm climates including the southern tier of states in the United States. It grows
in USDA zones 8a (10°F) to zone 11. It is a creeping perennial (Figure 7.52) but can
be grown hydroponically and is tolerant of brackish water. It attracts bees, butterflies
and birds. Propagation is through stem cuttings.
Colocasia esculenta
Colocasia or Taro grows to 8 feet tall and wide. The heart-shaped leaves are 2–3 feet
long and 1–2 feet wide on 3-feet-long stems that all emerge from a corm that can
weight 1–2 pounds (Figure 7.53). The corms, stems and the young leaves are edible
and nutritious. There are more than 200 cultivars of Taro; some best suited to wetlands
Figure 7.52
Bacopa Monnieri, Water
hyssop.
Figure 7.53
Colocasia esculenta grows in USDA hardiness
zones 8a (10 °F, –12.2 °C) to 11.
CC-BY-SA-3.0.
and others better suited to unflooded fields. Other cultivars highlight ornamental
characteristics, such as foliage colour. In wetlands, Taro is planted 18 inches apart
(45 cm) resulting in 20,400 plants per acre (49,000 per hectare). Taro is propagated
by collecting the small corms with stems that form around the main corm. The small
corms are planted 12 inches deep (30 cm). A number of insects and other pests attack
the plants (Moore and Lawrence, 2003). The corms of Taro are known to accumulate
arsenic, cadmium, lead and mercury from landfill leachate and mining wastewater
(Madera-Parra et al., 2015). Colocasia has been grown successfully in horizontal
subsurface flow wetlands with domestic sewage with COD levels as high as 1650 mg/L.
Floating wetland plants

This group of wetland plants floats freely on the surface of lakes, ponds and slow
moving surface water. They have relatively short roots that take nutrients directly from
the water column.
Eichhornia crassipes
Water hyacinth is native to tropical America but is naturalized and invasive in the
southeast United States and warm climates around the world. The glossy spoon-shaped
leaves are attached to spongy stems with nodules (Figure 7.54) that keep the plant
afloat. The fine roots hang a foot or more into the water. The showy flowers are violet
with yellow spots, and stand on 6-inch spikes above the foliage (attractive as cut
flowers) (Figure 7.55). Water hyacinth grows in full sun and requires at least 50°F to
produce new growth. It survives freezing and grows in USDA zone 9a: –6.6°C (20°F)
to zone 10b. Where it is invasive, it can spread with alarming swiftness, doubling or
Figure 7.54
Eichhornia crassipes has
bladder-like nodules that make
the plant buoyant.
Figure 7.55
The beautiful water hyacinth
flower and attractive foliage
complement the plant’s
wastewater treatment
effectiveness.
Photo: Wouter Hagens. License:
public domain.
quadrupling its population in as little as 2 weeks, creating significant ecosystem impact.

It can create 200 tons of biomass per acre.
Laboratory and pilot scale studies demonstrate that water hyacinth and water
lettuce (Pistia stratiotes) can be used in a recirculating, aerated system to treat
domestic wastewater to secondary standards within 2 to 4.5 days (or 7–8 days
without circulation and aeration) (Zimmels, Kirzhner and Kadmon, 2009). The plant
absorbs nitrogen and phosphorus into its roots and leaves and is a very effective
accumulator of heavy metals. Since E. crassipes increases biomass very rapidly,
frequent harvesting is an effective way to remove nutrients and metals from wastewater
(Marchand et al., 2010).
E. crassipes is also capable of reducing industrial pollutants. The plant is more
tolerant of cyanide than many plants. For example, Salix viminalis (willow) was killed
by an 8 mg/L concentration. Water hyacinth was able to tolerate and entirely remove
a 10 mg/L concentration of cyanide within a 23–32-hour treatment period (Ebel,
Evangelou and Schaeffer, 2007). This capability makes the plant valuable for treating
cyanide effluent from small-scale gold mining effluent.
Naphthalene (a polycyclic aromatic hydrocarbon, PAH) is another industrial

pollutant known to cause cancer. At a concentration of 13 mg/L, naphthalene was
completely removed by water hyacinth during a 9-day treatment period (Nesterenko-
Malkovskaya et al., 2012).
Pistia stratiotes
Pistia stratiotes, Water lettuce (Figure 7.56), is native to Africa but is now distributed
in virtually all tropical and subtropical freshwater habitats. This plant forms 6-inch
wide rosettes of leaves and has 50-cm-long roots. It grows only in warm climates (USDA
Zone 9a: 20°F, –6.6°C). Water lettuce is a particularly effective accumulator of
mercury and chromium from polluted water.
Lemna minor
L. minor is native to the United States and Canada and widely distributed else-
where (Figure 7.57). This tiny plant, just 1⁄4 inch in diameter (2–5 mm), grows in USDA
zone 5a: –20°F (–28.8°C) to zone 11 and reproduces rapidly, although it is short-
Figure 7.56
Pistia stratiotes, water lettuce,
is an attractive free-floating
plant common in warm
climates
Figure 7.57
Lemna minor, duckweed, is a
tiny but prolific floating plant.
Photo: Barbarossa. License:
CC-BY-SA-3.0.
lived (about 30 days). There are even commercial Lemna farms, as it is possible to
produce 10–30 tons of dried duckweed per hectare per year. The crop is high in protein
and is used for cattle feed. The feed contains 25–45 per cent proteins (depending on
the growth conditions), 4.4 per cent fat and 8–10 per cent fibre, measured by dry
weight. In natural and constructed wetlands Lemna is an important food for fish and
waterfowl. From a distance, this plant can appear to be an algae mat, but is actually
indicative of fairly good water quality, while an algae mat often indicates a eutrophic
problem.
A free water surface wetland with L. minor for tertiary treatment of domestic
wastewater was compared in a 3-year Turkish study with a horizontal subsurface flow
wetland planted with Cyperus and surface flow constructed wetland planted with
Cyperus. The Lemna wetland achieved a comparable performance, although the
required hydraulic residence time was more than twice as long (Ayaz, 2008).
Lemna minor demonstrates the ability to grow in ethanol and citric acid
production wastewater with COD as high as 1,400 mg/L, which could not be tolerated
by Eichhornia crassipes (water hyacinth), Azolla filiculoides (water fern), Salvinia
Rotundifolia willd (water fern) and Wolffia columbiana (duckweed) (Valderrama et
al., 2002). It tolerates chromium at 2 mg/L and is an effective accumulator (Uysal,
2013). Lemna minor has also been studied for the treatment of distillery effluent where
COD and BOD concentrations were very high at 17,540 and 25,576 mg/L,
respectively. Lemna provided better treatment than water hyacinth (Ran, Agami and
Oron, 2004). Lemna gibba is very similar to L. minor but is distributed throughout
Europe, Africa and South America.
Planting design
Where treatment wetlands are constructed in a public setting, such as in the Bridged
Gardens, Tianjin, China, or Houtan Park, Shanghai, China, the beauty of the planting
design is nearly as important as the treatment effectiveness. For example, at a guest
house in Portugal, a horizontal subsurface flow wetland to treat domestic sewage was
planted with ornamental flowering plants (Canna flaccida, Zantedeschia aethiopica,
Canna indica, Agapanthus africanus and Watsonia borbonica). The flowers were
harvested for use as decoration inside the buildings. The Agapanthus grew slowly and
was overwhelmed by the more vigorous species, suggesting the need for a staged
installation of plants. The treatment performance was very good, although only the
Canna is a common treatment plant. Removal of BOD and COD were more than 90
per cent, and removal of phosphorus, ammonium and total coliform bacteria were as
much as 92 per cent, 84 per cent and 99 per cent respectively (Calheiros et al., 2015).
Planting design examples are available for free water surface constructed wetlands
(Figure 7.58), but much more research and many demonstration projects are required
to apply effective methods to horizontal and vertical subsurface flow wetlands.
Generally, these have been planted as monocultures with very few of the many
possible plants attempted. Similarly, the development of niches for wetland plants that
are not part of the treatment system but adjacent to the treatment basins is a relatively
unexplored opportunity to increase the diversity of the planting palette and target
ecosystem restoration and enhancement goals. Again, habitat wetlands tangentially
associated with wastewater treatment have been constructed in free water surface
wetland sequences, but have not been created adjacent to the subsurface flow types.
Figure 7.58
High plant diversity with
species organized in masses
and according to water depth
and frequency of inundation
creates a maintainable and
beautiful landscape. Attention
to the texture, blossom colour
and form of the plants
establishes dynamic and
attractive scenes.
Photo: Gary Austin, 2015,
Turenscape planting design.
See the Appendices for plant lists organized by continent and for lists of plants
installed in projects in differing climates.
References
Ayaz, S. Ç. “Post-treatment and reuse of tertiary treated wastewater by constructed wetlands,”
Desalination, vol. 226, no. 1–3, pp. 249–255, June 2008.
Belmont M. A. and Metcalfe, C. D. “Feasibility of using ornamental plants (Zantedeschia
aethiopica) in subsurface flow treatment wetlands to remove nitrogen, chemical oxygen
demand and nonylphenol ethoxylate surfactants – a laboratory-scale study,” Ecological
Engineering, vol. 21, no. 4–5, pp. 233–247, December 2003.
Boonsaner, M., Borrirukwisitsak, S. and Boonsaner, A. “Phytoremediation of BTEX contam-
inated soil by Cannaxgeneralis,” Ecotoxicology and Environment Safety, vol. 74, no. 6,
pp. 1700–1707, September 2011.
Borin, M. and Salvato, M. “Effects of five macrophytes on nitrogen remediation and mass balance
in wetland mesocosms,” Ecological Engineering, vol. 46, pp. 34–42, September 2012.
Bragato, C., Brix, H. and Malagoli, M. “Accumulation of nutrients and heavy metals in
Phragmites australis (Cav.) Trin. ex Steudel and Bolboschoenus maritimus (L.) Palla in a
constructed wetland of the Venice lagoon watershed,” Environmental Pollution, vol. 144,
no. 3, pp. 967–975, December 2006.
Bragato, C., Schiavon, M., Polese, R., Ertani, A., Pittarello, M. and Malagoli, M. “Seasonal
variations of Cu, Zn, Ni and Cr concentration in Phragmites australis (Cav.) Trin ex steudel
in a constructed wetland of North Italy,” Desalination, vol. 246, no. 1–3, pp. 35–44,
September 2009.
Březinová T. and Vymazal, J. “Competition of Phragmites australis and Phalaris arundinacea
in constructed wetlands with horizontal subsurface flow – does it affect BOD5, COD and
TSS removal?,” Ecological Engineering, vol. 73, pp. 53–57, December 2014.
Březinová T. and Vymazal, J. “Evaluation of heavy metals seasonal accumulation in Phalaris
arundinacea in a constructed treatment wetland,” Ecological Engineering, vol. 79, pp. 94–99,
June 2015.
Brix, H. “Functions of macrophytes in constructed wetlands,” Water Science and Technology,
vol. 29, no. 4, pp. 71–78, 1994.
Brix, H., Ye, S., Laws, E. A., Sun, D., Li, G., Ding, X., Yuan, H., Zhao, G., Wang, J. and
Pei, S. “Large-scale management of common reed, Phragmites australis, for paper produc-
tion: A case study from the Liaohe Delta, China,” Ecological Engineering, vol. 73, pp.
760–769, December 2014.
Caldelas, C., Araus, J. L., Febrero, A. and Bort, J. “Ecophysiological effects of Cr and Zn on
Iris pseudacorus grown under semi-controlled conditions,” Comparative Biochemistry
Physiology Part A: Molecular and Integrative Physiology, vol. 153, no. 2, p. S215, June
2009.
Calheiros, C. S. C., Bessa, V. S., Mesquita, R. B. R., Brix, H., Rangel, A. O. S. S. and Castro,
P. M. L. “Constructed wetland with a polyculture of ornamental plants for wastewater
treatment at a rural tourism facility,” Ecological Engineering, vol. 79, pp. 1–7, June 2015.
Calheiros, C. S. C., Rangel, A. O. S. S. and Castro, P. M. L. “Treatment of industrial wastewater
with two-stage constructed wetlands planted with Typha latifolia and Phragmites australis,”
Bioresource Technology, vol. 100, no. 13, pp. 3205–3213, July 2009.
Chen, W., Chen, Z., He, Q., Wang, X., Wang, C., Chen, D. and Lai, Z. “Root growth of wetland
plants with different root types,” Acta Ecological Sinica, vol. 27, no. 2, pp. 450–457, February
2007.
Clarke E. and Baldwin, A. H. “Responses of wetland plants to ammonia and water level,”
Ecological Engineering, vol. 18, no. 3, pp. 257–264, January 2002.
Ebel, M., Evangelou, M. W. H. and Schaeffer, A. “Cyanide phytoremediation by water
hyacinths (Eichhornia crassipes),” Chemosphere, vol. 66, no. 5, pp. 816–823, January 2007.
El Hamouri, B., Nazih, J. and Lahjouj, J. “Subsurface-horizontal flow constructed wetland for
sewage treatment under Moroccan climate conditions,” Desalination, vol. 215, no. 1–3, pp.
153–158, September 2007.
Gao, J., Wang, W., Guo, X., Zhu, S., Chen, S. and Zhang, R. “Nutrient removal capability
and growth characteristics of Iris sibirica in subsurface vertical flow constructed wetlands
in winter,” Ecological Engineering, vol. 70, pp. 351–361, September 2014.
García-Lledó, A., Ruiz-Rueda, O., Vilar-Sanz, A., Sala, L. and Bañeras, L. “Nitrogen removal
efficiencies in a free water surface constructed wetland in relation to plant coverage,”
Ecological Engineering, vol. 37, no. 5, pp. 678–684, May 2011.
Greenway, M. “Macrophyte zonation and sustainability in stormwater wetlands in subtropical
eastern Australia: design and function,” Presented at the International Association for
Ecology, Orlando, FL, 2012.
Hardion, L., Barthélémy, C., Consales, J.-N., Gauthier, P., Thompson, J. D., Verlaque, R. and
Vila, B. “An endangered reed, Arundo donaciformis, in a dynamic urban environment: The
need for interdisciplinary conservation proposals,” Journal of Nature Conservation, vol. 26,
pp. 20–27, July 2015.
Jacobs, J., Graves, M. and Mangold, J. “Plant guide for pale yellow iris (Iris pseudacorus),”
USDA/NCRS, 2010.
Katsenovich, Y. P., Hummel-Batista, A., Ravinet, A. J. and Miller, J. F. “Performance evaluation
of constructed wetlands in a tropical region,” Ecological Engineering, vol. 35, no. 10, pp.
1529–1537, October 2009.
Kearney M. A. and Zhu, W. “Growth of three wetland plant species under single and multi-
pollutant wastewater conditions,” Ecological Engineering, vol. 47, pp. 214–220, October
2012.
Kim, K. D., Ewing, K. and Giblin, D. E. “Controlling Phalaris arundinacea (reed canarygrass)
with live willow stakes: A density-dependent response,” Ecological Engineering, vol. 27,
no. 3, pp. 219–227, October 2006.
Konnerup, D. and Brix, H. “Nitrogen nutrition of Canna indica: Effects of ammonium versus
nitrate on growth, biomass allocation, photosynthesis, nitrate reductase activity and N uptake
rates,” Aquatic Botany, vol. 92, no. 2, pp. 142–148, February 2010.
Konnerup, D., Koottatep, T. and Brix, H. “Treatment of domestic wastewater in tropical,

subsurface flow constructed wetlands planted with Canna and Heliconia,” Ecological
Engineering, vol. 35, no. 2, pp. 248–257, February 2009.
Kumari, M. and Tripathi, B. D. “Efficiency of Phragmites australis and Typha latifolia for heavy
metal removal from wastewater,” Ecotoxicology Environmental Safety, vol. 112, pp. 80–86,
February 2015.
Kyambadde, J., Kansiime, F., Gumaelius, L. and Dalhammar, G. “A comparative study of
Cyperus papyrus and Miscanthidium violaceum-based constructed wetlands for wastewater
treatment in a tropical climate,” Water Research, vol. 38, no. 2, pp. 475–485, January 2004.
Ladislas, S., Gérente, C., Chazarenc, F., Brisson, J. and Andrès, Y. “Floating treatment wetlands
for heavy metal removal in highway stormwater ponds,” Ecological Engineering, October
2014.
Li, L., Yang, Y., Tam, N. F. Y., Yang, L., Mei, X.-Q. and Yang, F.-J. “Growth characteristics
of six wetland plants and their influences on domestic wastewater treatment efficiency,”
Ecological Engineering, vol. 60, pp. 382–392, November 2013.
Liang, M. Q., Zhang, C.-F., Peng, C.-L., Lai, Z.-L., Chen, D.-F. and Chen, Z.-H. “Plant growth,
community structure, and nutrient removal in monoculture and mixed constructed
wetlands,” Ecological Engineering, vol. 37, no. 2, pp. 309–316, February 2011.
Madera-Parra, C. A., Peña-Salamanca, E. J., Peña, M. R., Rousseau, D. P. L. and Lens,
P. N. L. “Phytoremediation of Landfill Leachate with Colocasia esculenta, Gynerum
sagittatum and Heliconia psittacorum in Constructed Wetlands,” International Journal of
Phytoremediation, vol. 17, no. 1, pp. 16–24, January 2015.
Marchand, L., Mench, M., Jacob, D. L. and Otte, M. L. “Metal and metalloid removal in
constructed wetlands, with emphasis on the importance of plants and standardized
measurements: A review,” Environmental Pollution, vol. 158, no. 12, pp. 3447–3461,
December 2010.
Masbough, A., Frankowski, K., Hall, K. J. and Duff, S. J. B. “The effectiveness of constructed
wetland for treatment of woodwaste leachate,” Ecological Engineering, vol. 25, no. 5, pp.
552–566, December 2005.
Matthews, D. J., Moran, B. M. and Otte, M. L. “Screening the wetland plant species Alisma
plantago-aquatica, Carex rostrata and Phalaris arundinacea for innate tolerance to zinc and
comparison with Eriophorum angustifolium and Festuca rubra Merlin,” Environmental
Pollution, vol. 134, no. 2, pp. 343–351, March 2005.
Miller, C. “Plant fact sheet for Pickerelweed Pontederia cordata,” USDA/NCRS, 2012.
Moore, L. and Lawrence, J. “Plant guide for Taro,” USDA/NCRS, 2003.
Moss, T. “Plant fact sheet for powdery thalia (Thalia dealbata),” USDA-Natural Resources
Conservation Service, Jamie L. Whitten Plant Materials Center, Coffeeville, MS 38922., 2011.
Nesterenko-Malkovskaya, A., Kirzhner, F., Zimmels, Y. and Armon, R. “Eichhornia crassipes
capability to remove naphthalene from wastewater in the absence of bacteria,” Chemosphere,
vol. 87, no. 10, pp. 1186–1191, June 2012.
Ogle, D., Tilley, D. and St. John, L. “Plant guide for common spikerush (Eleocharis palustris),”
USDA/NCRS, 2012.
Olguín, E. J., Sánchez-Galván, G., González-Portela, R. E. and López-Vela, M. “Constructed
wetland mesocosms for the treatment of diluted sugarcane molasses stillage from ethanol
production using Pontederia sagittata,” Water Research, vol. 42, no. 14, pp. 3659–3666,
August 2008.
Pompeiano, A., Vita, F., Alpi, A. and Guglielminetti, L. “Arundo donax L. response to low
oxygen stress,” Environmental Experimental Botany, vol. 111, pp. 147–154, March 2015.
Qiu, Z. C., Wang, M., Lai, W.-L., He, F.-H. and Chen, Z.-H. “Plant growth and nutrient removal
in constructed monoculture and mixed wetlands related to stubble attributes,”
Hydrobiologia, vol. 661, no. 1, pp. 251–260, February 2011.
Rahman, K. Z., Wiessner, A., Kuschk, P., van Afferden, M., Mattusch, J. and Müller, R. A.
“Removal and fate of arsenic in the rhizosphere of Juncus effusus treating artificial
wastewater in laboratory-scale constructed wetlands,” Ecological Engineering, vol. 69,
pp. 93–105, August 2014.
Ran, N., Agami, M. and Oron, G. “A pilot study of constructed wetlands using duckweed
(Lemna gibba L.) for treatment of domestic primary effluent in Israel,” Water Research, vol.
38, no. 9, pp. 2241–2248, 2004.
Rodríguez M. and Brisson, J. “Pollutant removal efficiency of native versus exotic common
reed (Phragmites australis) in North American treatment wetlands,” Ecological Engineering,
vol. 74, pp. 364–370, January 2015.
Saikia, R., Chutia, R. S., Kataki, R. and Pant, K. K. “Perennial grass (Arundo donax L.) as a
feedstock for thermo-chemical conversion to energy and materials,” Bioresource Technology,
vol. 188, pp. 265–272, July 2015.
Sim, C. H. The use of constructed wetlands for wastewater treatment. Selangor, Malaysia:
Wetlands International, 2003.
Sohsalam, P., Englande, A. J. and Sirianuntapiboon, S. “Seafood wastewater treatment in
constructed wetland: Tropical case,” Bioresource Technology, vol. 99, no. 5, pp. 1218–1224,
March 2008.
Stannard, M. and Crowder, W. “Plant guide for Reed Canarygrass,” USDA/NCRS, Pullman
Plant Material Center, Pullman, Washington, 2002.
Stevens, M. “Plant guide for Soft Rush (Juncus effusus L.),” 2003.
Stevens, M. “Plant guide for Broadleaf Arrowhead, Sagittaria latifolia Willd.,” USDA/NCRS,
National Plant Data Center, Plant Sciences Department, University of California Davis, 2003.
Stevens, M. and Hoag, C. “Plant guide for broad-leaved cattail (Typha latifolia L.),” USDA/
NCRS, 2006.
Tanner, C. C. “Growth and nutrition of Schoenoplectus validus in agricultural wastewaters,”
Aquatic Botany, vol. 47, no. 2, pp. 131–153, February 1994.
Tanner, C. C. “Plants for constructed wetland treatment systems – A comparison of the
growth and nutrient uptake of eight emergent species,” Ecological Engineering, vol. 7,
no. 1, pp. 59–83, September 1996.
Tanner, C. C., Clayton, J. S. and Upsdell, M. P. “Effect of loading rate and planting on treat-
ment of dairy farm wastewaters in constructed wetlands – II. Removal of nitrogen and
phosphorus,” Water Research, vol. 29, no. 1, pp. 27–34, January 1995.
Tilley, D. “Plant guide for water sedge (Carex aquatilis),” USDA/NCRS, 2011.
Tilley, D. “Plant guide for cosmopolitan bulrush (Schoenoplectus maritimus),” USDA/NCRS,
2012.
Tilley, D. “Plant fact sheet for Nebraska sedge (Carex nebrascensis),” USDA/NCRS, 2013.
Tscharntke, T. “Insects on common reed (Phragmites australis): Community structure and the
impact of herbivory on shoot growth,” Aquatic Botany, vol. 64, no. 3–4, pp. 399–410,
September 1999.
Tylová, E., Steinbachová, L., Votrubová, O., Lorenzen, B. and Brix, H. “Different sensitivity
of Phragmites australis and Glyceria maxima to high availability of ammonium-N,” Aquatic
Botany, vol. 88, no. 2, pp. 93–98, February 2008.
Uysal, Y. “Removal of chromium ions from wastewater by duckweed, Lemna minor L. by using
a pilot system with continuous flow,” Journal of Hazardous Material, vol. 263, pp. 486–492,
December 2013.
Valderrama, L. T., Del Campo, C. M., Rodriguez, C. M., de-Bashan, L. E. and Bashan,
Y. “Treatment of recalcitrant wastewater from ethanol and citric acid production using
the microalga Chlorella vulgaris and the macrophyte Lemna minuscula,” Water Research,
vol. 36, no. 17, pp. 4185–4192, October 2002.
Valipour, A., Hamnabard, N., Woo, K.-S. and Ahn, Y.-H. “Performance of high-rate
constructed phytoremediation process with attached growth for domestic wastewater
treatment: Effect of high TDS and Cu,” Journal of Environmental Management, vol. 145,
pp. 1–8, December 2014.
Vymazal, J. “Emergent plants used in free water surface constructed wetlands: A review,”
Ecological Engineering, vol. 61, pp. 582–592, December 2013.
Vymazal J. and Kröpfelová, L. “Growth of Phragmites australis and Phalaris arundinacea in
constructed wetlands for wastewater treatment in the Czech Republic,” Ecological
Engineering, vol. 25, no. 5, pp. 606–621, December 2005.
Wang, C. Y., Sample, D. J., Day, S. D. and Grizzard, T. J. “Floating treatment wetland nutrient
removal through vegetation harvest and observations from a field study,” Ecological
Engineering, July 2014.
Wu, H., Zhang, J., Ngo, H. H., Guo, W., Hu, Z., Liang, S., Fan, J. and Liu, H. “A review on
the sustainability of constructed wetlands for wastewater treatment: Design and operation,”
Bioresource Technology, vol. 175, pp. 594–601, January 2015.
Xu, J., Zhang, J., Xie, H., Li, C., Bao, N., Zhang, C. and Shi, Q. “Physiological responses of
Phragmites australis to wastewater with different chemical oxygen demands,” Ecological
Engineering, vol. 36, no. 10, pp. 1341–1347, October 2010.
Zhang, X., Liu, P., Yang, Y. and Chen, W. “Phytoremediation of urban wastewater by model
wetlands with ornamental hydrophytes,” Journal of Environmental Science, vol. 19, no. 8,
pp. 902–909, January 2007.
Zhang, Z., Rengel, Z. and Meney, K. “Kinetics of ammonium, nitrate and phosphorus uptake
by Canna indica and Schoenoplectus validus,” Aquatic Botany, vol. 91, no. 2, pp. 71–74,
August 2009.
Zimmels, Y., Kirzhner, F. and Kadmon, A. “Effect of circulation and aeration on wastewater
treatment by floating aquatic plants,” Seperation and Purifiction Technology, vol. 66, no.
3, pp. 570–577, May 2009.
Zurita, F., De Anda, J. and Belmont, M. A. “Treatment of domestic wastewater and production
of commercial flowers in vertical and horizontal subsurface-flow constructed wetlands,”
Ecological Engineering, vol. 35, no. 5, pp. 861–869, May 2009.
8
Riparian wetlands
Introduction
This chapter begins a new section in which we review recently constructed wetlands
intended to solve pollution, urban and environmental problems, in order to assess the
wetlands’ successes and shortcomings. In every case, constructed riparian wetlands
should achieve multiple objectives. This is most apparent when riparian wetlands are
fashioned or restored in the urban environment (Figures 8.1 and 8.2). In these cases,
every category of ecosystem service should be advanced, in addition to ecosystem health
and biological diversity. The Millennium Ecosystem Assessment (Hassan et al., 2005)
divided ecosystem services into provisioning, regulating, supporting and cultural
categories. Ecosystem services are human benefits. The benefits flow to humans only
through viable biotic and abiotic networks. Therefore, planners and designers of
constructed wetlands must generate, restore or sustain wetland ecosystem functions
on-site and off-site in order to maximize human benefit and ecosystem health.
Despite the holistic imperative of the introduction above, the next few paragraphs
will address individual objectives before the case studies illustrate how these can be
integrated and mutually supportive.
Riparian constructed wetlands are often employed to remediate polluted water.
The source of pollution and the intended use of the repaired waters vary. Within
agricultural landscapes, the removal of pesticides and other contamination associated
with the application of artificial fertilizers may be undertaken to reduce impacts
on aquatic ecosystems and to recover habitat previously lost due to filling and
channelization. The Des Plains River wetland restoration in Illinois is a constructed
wetlands project of this type (Kostel et al., 2012). Of course, the scale and length of
rivers means that pollutants can be transported over long distances and can impact
remote urban, natural and even oceanic environments.
Another motivation for the creation of riparian wetlands is to improve the quality
of water in rivers that serve as the drinking water source for downstream towns and
cities. River water is always treated before use as drinking water, but the difficulty
and cost of treatment to potable standards is far more difficult and expensive if
the influent is highly contaminated. The experience of New York City is a well-
documented example demonstrating that preventing pollution of the drinking water
source is far less expensive than treating polluted water to create drinking water (New
York City Department of Environmental Protection, 2012; DePalma, 2006).
Water from mine tailings, uncontrolled industrial discharges, inadequately treated
domestic sewage from wastewater treatment plants, or leaking septic tanks and drain
RIPARIAN WETLANDS 143
Figure 8.1
The Sanlihe River restoration
project removed concrete
channel walls, repaired the
hydrological function by:
routing some water from the
larger Luan River, removing
solid waste dumps, and
preventing the inflow of poorly
treated sewage and treatment
of urban stormwater run-off
before it entered the restored
river. The creation of multiple
channels allowed the retention
of rows of existing trees while
expanding the floodway.
Figure 8.2 Before 2007, this river was a concrete channel. Polluted by industrial discharges and
eventually depleted of water due to industry, agriculture and urban withdrawals, the river had become
a wasteland and civic liability. The Sanlihe River Ecological Corridor Project by Turenscape reversed
the history of pollution and transformed the river into an ecological and community resource. The
restored river is 8.3 miles long (13.4 km), 330–980 feet wide (100–300 m) with a total area of 333
acres (135 hectares).
fields can all contribute polluted water to streams and rivers. The degradation of rivers
from these many discharges can produce pollution levels that challenge the technology
and economics of removal for human use.
While surface water quality standards for swimming, wading, fishing, boating
and other recreational contact are lower than the requirements for drinking water,
rivers often fail to meet these standards. Of most concern are substances in the
water that threaten to make people sick in the short or long term. Pathogenic bacteria
from livestock operations, inadequately treated human sewage and urban storm-
water run-off regularly require the closure of rivers, lakes and coastal beaches for
recreational use.
144 RIPARIAN WETLANDS
Finally, there are several contaminants and excess nutrients that find their way
to rivers through urban stormwater run-off. Many of these are associated with the
pervasive use of the private automobile, which contributes polycyclic aromatic
hydrocarbons (PAH), polychlorinated bi-phenyl hydrocarbons and other cancer-
causing substances. Heavy metals such as cadmium, lead, copper and zinc are also
associated with automobiles and run-off from industrial land uses. Urban and
suburban stormwater run-off also contains surprisingly large quantities of pathogenic
bacteria, nitrogen and phosphorous, which compromise human health and the
functions of aquatic ecosystems.
Is the river margin really the best location for reducing water pollution in rivers?
Absolutely not. A better strategy is to reduce pollution at its source, before it reaches
natural streams and rivers. Unfortunately, rivers have served humans as convenient
waste disposal systems for millenniums. Many energy, chemical, consumer product
and commodity agricultural industries seek the water supply, waste disposal and
transportation opportunities offered by rivers. Therefore, the use of riparian wetlands
to treat river water is a remedial practice, suggesting the absence of comprehensive
treatment policies for industrial discharge, municipal sewage effluent and non-point
source pollution. The broad range of contaminants from these diverse pollution
sources makes the improvement of water quality more challenging than when
engineering for a narrower range of substances associated with homogeneous land
uses. However, the use of constructed riparian wetlands for treatment of point or non-
point sources of contaminated water generated on adjacent parcels, such a stormwater
run-off from a neighbourhood, is a preventative rather than remedial practice. This
is the situation on the margins of the Sanlihe River, as illustrated in Figure 8.2.
Alteration of rivers
Unfortunately, there are relatively few free-flowing rivers in the world. Urbanization,
agriculture and energy demands inspired channelization and damming rivers to control
flooding and provide local economic opportunities (Figure 8.3). During the last few
centuries, these changes have sometimes been engineered without a comprehensive
understanding of regional hydrology, resulting in adverse impacts, as illustrated by
the history of the lower Mississippi River in the United States and the Yangtze River
Figure 8.3 The first image shows the existing condition of the Arroyo Secco River Channel at the
confluence with the Los Angeles River. The second image is a photo simulation of a proposed
revision of the same area.
Photo: US Army Corps of Engineers. License: public domain.
in China. River control efforts have often resulted in unintended consequences, costing
human lives and economic hardship.
Historically, cities were located along rivers for defensive, transportation,
agricultural and industrial purposes. More recently, rivers attract urban dwellers for
the recreation and aesthetic advantages that they offer. Associated with this history
of human use of rivers is their modification for human purposes. The practice of flood
prevention through channelization, levees, dredging, dams and floodgates is a global
activity. Similarly, the filling and hardening of the river edge to serve as development
pads and docking infrastructure is common in all cities built along rivers.
In most cities of this sort, much damage is done, and reversing it will take decades.
For example, 35 million euros were spent to eliminate contaminants and debris left
from World War II, improve flood control and provide recreation and habitat along
the Isar River in Munich, Germany (Figure 8.4) (RESTORE, 2013). We must avail
ourselves of all opportunities to implement palliative and preventive measures if rivers
are to be clean enough for human use and clean enough to support healthy aquatic
ecosystems.
Achievable objectives
Riparian constructed wetlands can significantly improve the quality of water in rivers
if they are frequent and extensive, but the water quality goals should not blind us to
the many other achievable objectives. Flood hazard protection is a common secondary
benefit of constructed wetlands. A natural associated objective is ecosystem health
and enhanced biological diversity. In urban areas, provisions for open space, recreation
opportunities, social spaces and educational resources constitute the long list of
Figure 8.4 The Isar River flows through Munich, Germany. The city removed the walls channelizing
the river. To fortify the riverbanks, sheet piles were installed at the edge of the floodplain. Soil was
filled against these to establish a naturalistic bank while providing flood protection. Habitat by
planting native vegetation and new recreation opportunities resulted from public access points and
new beaches. Water quality is further improved by the more natural river.
Photo: Bbb. License: CC-BY-SA-3.0.
cultural services that could be generated. Important among these cultural services is
the prospect of increased tourism and economic development.
A less obvious objective might be provisioning ecosystem services. Fish, crayfish
and shellfish are examples of foods that could be greatly expanded. In addition, we
could include crops grown on the upland benches of the wetland. Even fibre from
poplar trees, willows or wetland vegetation, such as common reed, could be cultivated
as feedstock for paper, biofuels or other products.
Finally, as suggested earlier, elimination of disease-causing organisms, the cycling
of nutrients, carbon sequestration, soil formation and other regulating and supporting
ecosystem services should not be discounted processes and products of the constructed
wetland.
Natural riparian wetlands

Free-flowing rivers exhibit a range of edge conditions depending on the gradient of
the bed and the bed material. The amount of seasonal change in the water level also
contributes to a range of edge characteristics. In nearly every case, there is a linear
riparian habitat that extends some distance from the edge of the water (Figure 8.5).
Greater humidity, access to groundwater or annual inundation all influence the
vegetation communities, which differ from the adjacent upland vegetation associations.
The vegetation along rivers is not necessarily wetland plants, but low-lying lands
adjacent to a river that floods seasonally or where there is high groundwater develop
into wetlands of various types. These are often associated with a historic route of the
river that has been forsaken for a new main stem route (Figure 8.6).
A natural linear wetland is the model for constructed riparian wetlands presented
in this chapter.
General characteristics of riparian constructed wetlands

In order to repair hydrological damage, restore habitat or treat polluted water,
riparian constructed wetlands are increasingly necessary. The distinction between linear
constructed wetlands and others is somewhat artificial. Broad wetlands, either natural
or engineered, are commonly associated with rivers (Figure 8.7), but the linearity of
constructed riparian wetlands presents particular opportunities and challenges.
Figure 8.5
Natural riparian vegetation.
Photo: Dave Powell. License:
CC-BY-SA-3.0.
Figure 8.6
Riparian wetlands along and
within rivers and streams
develop due to natural
sedimentation and changes in
the course of the river.
Photo: P reIgor. License: CC-BY-
SA-2.0.
Figure 8.7
Wetlands along the Kobuk
River, Alaska.
Photo: Neal Herbert, NPS Alaska
Region. License: CC-BY-SA-2.0.
Riparian constructed wetlands are formed by creating a side channel separated

from the main stem of the river, similar to patterns seen in braided natural rivers. The
main stem feeds the riparian constructed wetland, and the long narrow shape favours
sedimentation and filtration processes.
In the developed urban or agricultural landscape, linear wetlands are often the
most viable spatial option. For example, many urban rivers in the United States were
engineered to flow within concrete channels with a trapezoidal cross section (Figure
8.8). Restoring ecosystem values along these rivers often requires replacing the angled
wall with a vertical wall to generate space along the edge of the water for wetland
vegetation. Even removing the concrete bed often results in vegetation growth and
improved biodiversity. In other cases, railroad corridors or highways following
rivers have been relocated to recover narrow strips along the river for restoration
efforts. Extending the width of the channel is also an important retrofit strategy for
tidal influenced rivers subjected to increasing water level due to global warming.
In fact, this technique is often necessary to adapt to increasing stormwater run-off
volume associated with global warming for streams and rivers away from the coast
(Figure 8.4).
Figure 8.8
Efforts are underway to restore
portions of the Los Angeles
River in this condition to a
more natural configuration.
Photo: Downtowngal. License:
CC-BY-SA-3.0.
Multifunctional riparian constructed wetlands

For the reduction of pollution, the shape of the linear wetland is well suited for removal
of suspended solids, and sequestration of heavy metals and petroleum aromatic
hydrocarbons, but is less well suited for the removal of dissolved metals or nitrates,
for example. However, special design measures can overcome this shortcoming.
Recreation trails and gathering areas for small groups are easily accommodated
along the linear wetland (Figure 8.9), but activities requiring more expansive space
are less well served.
Wetland habitat is particularly rich due to the presence of water and the
heterogeneous mosaic created through the interaction with upland vegetation.
However, the narrow strip will be inadequate for animal species that require large
territories or those that are sensitive to the presence of humans, if the wetland is in
the urban or suburban landscape. Again, urban biological diversity can be maximized
through the application of design measures. Perhaps, most important is the role the
constructed riparian wetland can play as a corridor that connects more expansive and
more natural habitat within and beyond the city (Figure 8.1).
Reducing flood hazard is highly compatible with the other functions of constructed
riparian wetlands, as illustrated in the Isar River (Figure 8.4) and the Houtan Park
case studies.
Houtan Park case study

Huangpu River
The Huangpu River is 70 miles long with its headwaters in an agricultural area
upstream from Shanghai, China. As it flows toward the city, agriculture becomes mixed
with residential and small manufacturing enterprises (Figure 8.10). It is within this
reach that Shanghai diverts five million cubic metres of river water to treatment plants
for purification to drinking water standards. The quality of the water in the river is
declining because of point and non-point source pollution. Agricultural run-off
containing fertilizer, poorly treated sewage from small towns, industrial discharges
and its location downstream from the polluted Lake Taihu all contribute to the poor
water quality at the drinking water intake. At the intake, the concentration of total
Figure 8.9
Public health, ecosystem
restoration and social
opportunities are three of the
many values evident in this
photograph of the constructed
wetland along the Tanghe
River.
Photo: Kongjian Yu.
Figure 8.10 Huangpu River upstream from Shanghai. Notice the high volume of shipping traffic.
Image: Google Earth and DigiGlobe 11/13/2014, 30°8′47.31″ N 121°19′46.01″ E, image accessed:
18 February 2014.
suspended solids ((TSS) an indirect measure of pollution) is between 50 and 150 mg/L.
A desirable level is less than 30 mg/L. Disinfection of the river water requires
increasingly large amounts of chlorine, which interacts with the large amount of
dissolved organic matter in the river water to create harmful compounds that are
delivered with the drinking water to the residents of the city (Xu et al., 2007).
Unfortunately, the quality of the river water continues to worsen as it flows toward
Shanghai. As the river approaches the city, energy and industrial plants burning coal
are common along the banks (Figure 8.11), while highly impervious residential and
commercial districts contribute non-point source pollution from locations further from
the river. Toward the centre of the city, the industrial uses are being replaced by high-
density, mixed-use developments.
Coal-fired power plants discharge phenols, which accumulate in the bottom
sediment of the river and are harmful to fish and people. Phenol is carbolic acid, a
Figure 8.11
Coal-fired energy and industrial
plants along the Huangpu River.
Image: Google Earth and DigitalGlobe
9 February 2015, 31°03′33.58″ N
121°28′15.65″ E, image accessed:
18 February 2014.
common manufacturing chemical used to produce fertilizer and paint. It is a solvent

and a product of coal tar.
PAH have toxic, mutagenic and carcinogenic characteristics. About 40 per cent
of the PAH in the Huangpu River are from coal combustion, while traffic-related
pollution and spills of oil products contribute 36 per cent and 24 per cent, respectively.
Escherichia coli (E. coli), a human pathogen (and a commonly used indicator of the
presence of many other pathogens), is present in high concentration near the centre
of Shanghai. Here, there are between 3,000 to a little less than 30,000 colony-forming
units (CFU) per 100 ml. This level can be compared to the US Environmental
Protection Agency (EPA) standard of 126 CFU per 100 ml of E. coli for primary
recreational contact (swimming).
As contaminant levels in river water increase, a dynamic arises in which high levels
of ammonium and organic material cause explosive growth of bacteria that reduce
dissolved oxygen to levels that threaten fish and other aquatic species. Low oxygen
levels also reduce the conversion of ammonia to nitrate and the transformation
of nitrate into nitrogen gas and harmless by-products. These processes are critical to
effective pollution mitigation (Dong Yue et al., 2013).
Table 8.1 summarizes the compromised water quality of the Huangpu River in
the area of the drinking water intake (upstream) (Figure 8.10) and in the area of Houtan
Park (city centre) (Figure 8.12). The range of values in the table indicates seasonal
variations due to increased dilution of pollutants during the rainy season.
Table 8.1 Huangpu River water quality
Upstream City centre Standard

TSS 50–150 100–350 30 (30)
BOD 1.2–2.3 4.8 -9.8 30 (30)
DO 4–9 1.2–3 6 (4–5)
TN 3.5–5 3–6 0.5 (4)
TP 0.3–0.5 0.23–0.38 0.1
NH4 0.5–2 1.8–4.8 0.5 (0.4)
NO3 1–1.8 0.2–0.6 (10)
E. coli Meets Std 3,000–30,000 (126*)
Phenol 0.001–0.003 0.001–0.006
Oil 0.1–0.16 0.18–1.2
All units Mg/L, except E. coli. Standard – Chinese (US EPA). TSS – total suspended solids; BOD – biological
oxygen demand; DO – dissolved oxygen; TN – total nitrogen; TP – total phosphorous; NH4 – ammonium; NO3
– nitrate. * colony-forming units meeting primary recreational contact.
Figure 8.12
Aerial view of Houtan Park
along the bank of the Huangpu
River near the centre of
Shanghai.
Image: Google Earth and
DigitalGlobe 10 October 2015,
31°11′11.22″ N 121°28′08.93″ E,
image accessed: 18 February
2014
The water quality improvement goals for Houtan Park should have targeted those
contaminants posing the highest risk to human and aquatic health. Therefore, reducing
the high concentrations of E. coli, TSS, total phosphorous (TP) and ammonium is
most important, as is increasing the amount of dissolved oxygen in the river water.
The levels of most of these contaminants are quite high, compared to levels in
European rivers (European Environment Agency, 2015), for example.
The Houtan Park context

Houtan Park is located in the Pudong area of Shanghai on Shibo Avenue and
near the North-South Elevated Road (31°11′10.62″ N 121°28′12.25″ E). Houtan
Park follows the bank of the Huangpu River, which is an average of 1,300 feet wide
(400 m) and 30 feet deep (9 m), dividing Shanghai into two parts. Near the heart of
the city, the area around the park is being rapidly redeveloped. Former industrial and
warehousing operations are being replaced by high-density and mixed-use urban
districts (Figure 8.13). The site of the park was formerly a steel factory and shipyard.
Subsequent use of the parcel as a landfill and industrial staging yard continued a legacy
Figure 8.13
Houtan Park along the bank of the
Huangpu River near the centre of
Shanghai.
Figure 8.14 The industrial riverbank of the Huangpu River at the proposed location of Houtan Park.
Note the vertical floodwall. This was removed when the new park was constructed.
of soil and river contamination (Figure 8.20). A vertical flood control wall separated
the parcel from the river. In fact, the site of the future park and constructed wetland
lay between the 20-year and 1,000-year flood control structures. The Huangpu River
is influenced by tidal flows at this location, causing the water to fluctuate 6 feet (2
m), establishing an inaccessible, muddy and littered river edge (Figure 8.14). The site
is long and narrow, between 100 and 265 feet wide, and features an 18-foot elevation
change between the shoreline and the road. The total area of the park is 34.6 acres
(14 hectares), and the length of the water treatment sequence is 1 mile (1.7 km).
Design programme
There was an extensive programme for the project, as the purpose of the park was
to change in emphasis over time. The initial impetus for the project was the Shang-
hai Exposition hosted by China in 2010. Many international pavilions were planned
(Figure 8.15). Some of the thousands of visitors to the exposition would arrive by
ferry. Therefore, the park includes a dock, embarkation structure and a ceremonial
walkway to the pavilion area. The planners of the exposition followed the inter-
national tradition of highlighting technology and sustainable development, including
a demonstration of the biological treatment of polluted river water (Figure 8.17). The
treated water was to be used in the pools and fountains within the pavilion area (Figure
8.16). The park included structures for displays, restrooms, walking paths and gather-
ing areas. The landscape architect adopted and extended the sustainability theme of
the project to include the reuse of existing structures and materials, the creation
of new habitat for birds, wildlife and aquatic organisms, and the cultivation of crops.
This programme was in addition to providing flood mitigation and reduction of
stormwater run-off. The following sections document the design and performance of
Houtan Park, beginning with a review of water quality and biodiversity improvements.
Figure 8.15
The Chinese Pavilion at the
2010 Shanghai Exposition.
Photo: Brücke Osteuropa, 2010.
License: CC public domain.
Figure 8.16 The main axis of the Shanghai Exposition with pools filled with treated river water.
Photo: Suzuki, 2010. License: CC-BY-2.0.
Figure 8.17
This image illustrates the
suspended solids in the water
in the first and the last stage of
the wetland.
Table 8.2 Project calendar
Activity Calendar
Water introduced 6, 2009
Initial planting 6, 2009
Second planting 10, 2009
First fish, shellfish and other aquatic organisms introduced 10, 2009
Wetland establishment period 6, 2009 to 12, 2009
Exposition open 5, 2010 to 10, 2010
Maintenance assisted period with water quality testing 6, 2009 to 10, 2010
Mature wetland (reduced maintenance) with water quality testing 11, 2010 to 10, 2012
Programme implementation
The implementation of the wetland was conceptualized in three stages with stage one
(Figure 8.18) located nearest to the head of the wetland and featuring plants and other
organisms most resistant to high levels of pollution.
Water was introduced for the first time in June 2009 and followed by a 6-month
start-up and stabilization period (Table 8.2). In October 2009, the first fish, aquatic
snails and shrimp were introduced. Twenty-eight species of wetland plants were planted
in the wetland (Figures 8.19 and 8.22). Those adapted to living in polluted water were
introduced first. Thousands of plants from three species dominated the initial planting.
Twenty-five thousand common reed (Phragmites australis), 15,000 eelgrass (Vallisneria
natans), and 12,000 waterthyme (Hydrilla verticilata) were planted. The plants were
carefully selected so that there were species active in every season with special
attention to plants that would be actively growing during the winter. No invasive plants
were included in the planting palette. Twenty-two species of aquatic organisms were
introduced. The 1,000 Asian clams (Corbicula fluminea) represented the greatest
number of individuals for any species.
Water treatment
This constructed wetland treats a small fraction of the water volume that flows past
the park (Figure 8.24). Nevertheless, riparian treatment wetlands mitigate existing
pollution (Figures 8.19 and 8.22), and the huge volumes of polluted water in the river
call for many more extensive riparian wetlands within the urban area. The daily flow
Figure 8.18 Constructed wetland plan with contours. The river water enters on the left and travels
through a series of cells that target particular contaminants.
Figure 8.19
The planting design for Houtan
Park is in contrast to the
monocultures that characterize
most constructed wetlands
projects. This image shows a
broad range of submerged,
floating leaf and emergent plant
species that contribute to the
treatment of the polluted river
water.
into the wetland is 2,500 m3 (660,430 gallons), while the outflow is 2,400 m3. The
difference represents a 100 m3 per day loss, due to evapotranspiration and infiltration.
A study of the water quality performance of the constructed wetlands in Houtan
Park was conducted between 2009 and 2012. The analysis included testing before
and after the plants were established. The water quality testing was divided into an
establishment phase (June 2009–October 2010) and a mature phase (November 2010–
October 2011). The first phase included more intensive maintenance of the plants,
particularly harvesting of the wetland vegetation in autumn. During both phases, three
water samples were collected every month from 10 different points distributed along
the water route.
The water quality was characterized according to an index that included un-ionized
ammonia (NH3), chemical oxygen demand (COD), nitrite (NO2), nitrate (NO3), total
nitrogen (TN) and TP (Dong Yue et al., 2013). Unfortunately, the index did not include
measurements for biological oxygen demand (BOD), TSS, heavy metals, PAH or
pathogenic bacteria, which are either important indirect or direct measures of water
pollution in the United States.
Using the composite measure of water quality, the constructed wetland results in
a general water quality improvement (Figure 8.17) changing from the Chinese standard
Class 5–4 to Class 3–2. There are five water quality standards defined by the Chinese
government (Table 8.3).
• Class 1: Consistent with protected natural landscapes. Excellent drinking water

quality.
• Class 2: Surface water for drinking purpose (1st grade protection area), water for
aquaculture (precious species).
• Class 3: Surface water for drinking purpose (2nd grade protection area), water
for aquaculture (common species).
• Class 4: Suitable for industrial use.
• Class 5: Suitable for industrial use. Unsuitable for human contact (Wang et al.,
2012).
When reviewing the individual constituents, it’s clear that some contaminants
were more effectively removed than others. Nitrogen improved from Class 5 to 4–3;
TP improved from Class 4 to 2; COD and ammonia met the Class 3 standard, while
DO improved significantly from Class 3 to 1. The 86.4 per cent reduction in nitrate
was dramatic and accounts for the reduction in TN. COD is a water quality indicator
often used for water that contains constituents from industrial land uses. The test for
COD uses a chemical (such as potassium dichromate) to define the oxygen demand
required for degradation of both organic and inorganic matter in the water sample.
Since BOD measures only the oxygen demand to degrade organic matter and COD
is a measure of both organic and inorganic demand, COD values are always higher
than BOD values. When BOD is near 0, COD is approximately 50. If BOD/COD
ratio is 0.5, then the wastewater is easily treatable by biological means.
There was little reduction in the amount of ammonia, which explains the small
reduction of TN (25 per cent). This result is consistent with free water surface
wetlands (Kadlec, 2009). Vertical flow constructed wetlands demonstrate as much as
80 per cent removal of ammonia, relying on bacteria rather than plant uptake,
although this wetland type is also planted. Replacing some of the free water surface
Table 8.3 Chinese water quality standards. All units = mg/L, except faecal coliform – number/litre
Pollutant China
Class 1 Class 2 Class 3 Class 4 Class 5
BOD5 3 3 4 6 10
COD ≤ 15 15 20 30 40
TSS ≤ 20 30
TP ≤b 0.02 0.1 0.2 0.3 0.4
TN ≤ 0.2 0.5 1 1.5 2
TOC 20 30
Ammonium NH4 15 50 100 50 200
DO ≥ 7.5 6 5 3 2
Faecal coliform 200 2,000 10,000 20,000 40,000
Copper (Cu) 0.01 1 1 1 1
Zinc (Zn) 2 5 5
Petroleum 50 50 50 500 1,000
Chlorine <0.5 <6.5a >5a
a – with 1.5 hours of contact time, b – river water (lower for lakes).
BOD – Biological oxygen demand, COD – Chemical oxygen demand, TSS – Total suspended solids, TOC –
Total organic carbon, P – Phosphorus.
Sources: Wang et al., 2012 and http://chinawaterrisk.org.
The two drinking water classes (2 and 3) provide water that can be treated to meet drinking water quality with
only biological processes.
wetlands with vertical subsurface flow cells in the early stages of the treatment
wetland might have improved ammonia removal.
The significance of the biological water treatment system at Houtan Park extends
well beyond its boundaries. Dozens of other treatment wetlands have been created
based on the Houtan Park model.
First stage of the constructed wetland

The wetland is initiated by an interesting, although largely hidden, structure. Water
from the Huangpu River is pumped to the top of a basin. There it drains vertically
through a series of filters before discharging into the first wetland pool. Removal of
sediment and aeration of the polluted water is the primary goal of this stage. Figure
8.25 is a plan view of this area and shows a series of terraces dropping to pools
indicated with contour lines. A graded slope on the west transitions into three or
four levels of planted terraces, which are mirrored on the east side of the wetland.
The permanent pool is at an elevation just below the wall forming the lowest terrace.
Figure 8.20
This image shows the original
condition of the park site. See
Figure 8.21 and 8.22 below for
comparison to the same area
while under construction and
after completion of the park.
Figure 8.21
The same site shown in the
earlier image illustrates the
terracing, concrete boardwalk
platforms and the division of
the wetland bed into planting
strips. The stone dividers,
within the wetland planting
bed, improve the establishment
of the plants and facilitate
maintenance operations. See
Figure 8.22 for the scene in the
completed project and Figure
8.23 for the plant list for this
area.
Figure 8.22 The same area as in the previous figure, but viewed from the opposite direction, shows
the boardwalks and terrace seating along with the wetland vegetation bands and upland terraces.
Figure 8.23 The plan of the area shown in Figures 8.21 and 8.22 identifies the plant species arranged
in bands perpendicular to the water flow. These are bordered by planted terraces, boardwalks and
flights of stairs to gathering areas.
Figure 8.24 Constructed wetland section. There are 15 wetland compartments in the treatment
sequence. The variety in water depth creates a heterogeneous habitat and treatment environment.
Water in each compartment is collected and redistributed in the subsequent cell to optimize contact
of the water with the wetland plants and the soil substrate.
Figure 8.25
GDZQUHGZRRGIRUHVW
Initial stage of the wetland, plan
LQGXVWULDOUHOLF 0 5 10 20 50m view.
ZRRGHQSDWKZD\
Huangpu River
The berm between the wetland and the river provides protection from the 20-year
flood and the daily rise and fall of the river. During larger storms, water flows into
the wetland and floods successively higher terraces.
Figure 8.26 shows the terraces in the first section of the wetland just before
planting. Each terrace is designed to hold a volume of water. This image also makes
it clear that delay and retention of stormwater run-off and detention of flood flows
were dual design objectives. Treatment volumes of water are retained in the terraces
to allow time for contaminant removal. The deep pools provide habitat and a range
of oxygen conditions necessary to address the broad spectrum of pollutants present
in the river water.
Associated with the water inlet is a water wall. The section of the wall (Figure
8.27) clarifies the flood retention capacity of the design. It also illustrates that river
water is aerated as it flows over the surface of the wall veneer. It was important to
oxygenate the water in the initial stage of the wetland, in order to make aerobic
biological treatment of the water possible. Nevertheless, note that at the base of
the wall, the planting soil is underlain by thick strata of sand and gravel to provide
sites for the growth of bacteria in anoxic and anaerobic conditions, below the aerobic
pools.
Figure 8.26
Initial stage of the wetland
during construction, viewing
north.
Photo: Kongjian Yu.

01. slate gray granite top, 250X500mm(50mm thick)
02. planter
03. non-clay brick, 120mm
04. 1:3 cement mortar 120-150 wide bonded slate gray schist
05. 1:3 cement mortar (including 5% water repellent), 20mm
06. waterproof mortar bonded rubble retaining wall, lay1:0.1 tilted
07. slate gray schist, 10mm
08. 1:3 cement mortar (including 5% water repellent), 20mm
09. C30S6 reinforced concrete, 120mm
10. C20 reinforced concrete cushion, 100mm

11. gravel, 300mm
12. prime soil compaction
13. C15 reinforced concrete basis, 150mm
14. sediment planting soil, 400mm
15. rough sand, 400mm
16. gravel, 300mm
17. clay compacted in layers, 600mm
Figure 8.27 Masonry wall with cut stone veneer, section. A reservoir of river water at the top of the
wall feeds a granite weir directing water over the vertical veneer and into the first pool of the
constructed wetland.
Image: Kongjian Yu.
The transformation of the degraded site into a functional and beautiful constructed
wetland is illustrated in Figure 8.28. The plants soften the geometry of the terrace
edges, but the structure remains visible as the plants are grown in single specie masses
in each terrace. Mixing the grass-like species, even if they had similar tolerance to
inundation, would make the design less legible and diminish the visual impact of colour
and texture that are characteristic of the various species.
At right-centre of Figure 8.28, a steel structure is reflected in the slowly moving
water. A section (Figure 8.29) details the structure and its context. The upland soil
is separated from the wetland soil populated with floating leaf plants. The bamboo
decking is attached to a concrete platform that is raised above the wetland on
columns. The steel structures and the decks and platforms will be discussed in some
detail later, but notice in the section that the water elevation is carefully established
Figure 8.28
This image shows the
established wetland from the
same viewpoint as the
construction image in Figure
8.26. The terraces are now
visible as soft panels of
vegetation. The new tree
groves completely screen the
buildings on the right side of
the park. Notice the steel
structure reflected in the water
in the middle ground. A section
through this feature is shown in
the following figure.
Figure 8.29 Section through wetland, deck and steel sculpture created from waste material from the
existing site.
to meet the top of the concrete platform. This imparts the sense that the wood deck
is floating. It also provides the sense that the visitor is within the wetland environment,
establishing a unique sense of place.
Stormwater
Stormwater flows into the Houtan Park treatment wetland where it is treated along
with the river water. However, the slopes are often transformed into level terraces
(Figure 8.30) or are heavily vegetated, which slows run-off, encourages infiltration
and removes many of the contaminants before they reach the wetland. In fact, a portion
of the existing site was 82 per cent impervious and was replaced with parkland that
is only 19 per cent impervious. The design and planting of the park reduced the
stormwater run-off volume by 19.2 per cent (31,500 cubic feet), reversing the trend
of urban development.
Figures 8.30 and 8.31 show that stormwater is captured in level terraces where
sedimentation and biological processes remove contaminants such as lead, zinc,
bacteria and excess nitrogen. Figures 8.32 and 8.33 illustrate that careful grading creates
stormwater retention areas on the bench above the linear pool. These support the
illusion that the boardwalk floats on the water.
Figure 8.30
Park terraces just before
planting. The lower terraces
receive water from the river,
while the upper terraces
capture stormwater from the
neighbourhood.
Figure 8.31 Section through planted terraces.

Figure 8.32 Wetland valley during construction. See Figure 8.33 for the effect after the planting has
matured.
Figure 8.33
Wetland valley completed
landscape.
Cultural ecosystem services

Of course, the primary reason for building Houtan Park was to fulfil a cultural
aspiration. The short-term success of the Shanghai Exposition led to long-term
economic development and a terrific city park with many public amenities (Figure
8.34). There are 3.25 miles of pedestrian walking paths that form a main loop with
perpendicular boardwalks that cross the wetland (Figure 8.35). While this park might
not make people healthy, it affords residents the opportunity to exercise to improve
health and to restore their ability to cope with urban life through a stress-reducing
outing in a naturalistic setting (Figure 8.36).
There is a great deal of public art (Figure 8.37). This is the result of two impulses
– one aesthetic and one environmental. Sustainable development depends on recycling
and reuse of materials and the former industrial use of the site provided a supply of
steel refuse. This material was refashioned into public art that is displayed throughout
the park or incorporated into new structures (Figure 8.38). As illustrated in Figures
8.33 and 8.41, the steel was used to create several folded panels. These punctuate the
Figure 8.34 This early rendering of the proposal for Houtan Park illustrates the multifunctional
character of the park. Note the 10-acre existing wetland at the lower right. This rare wetland along the
Huangpu River was preserved.
landscape and frame views of the Shanghai skyline and highlight the industrial
heritage of the city and this site. The materials were reconfigured to create artful forms,
new paving material for the boardwalk and shelters (Figures 8.39, 8.41 and 8.42).
Groves of bamboo and Dawn redwood trees serve as screens along the pedestrian
paths and create partially enclosed spaces that are used to exhibit the public art and
industrial relics (Figure 8.40).
Figure 8.35
Bamboo decking follows a vibrant fibreglass
bench between two community agriculture fields.
Figure 8.36
There are varied aesthetic
experiences that the visitor
encounters on a walk through
the park. These range from
regional views of the river and
skyscrapers on the opposite
shore to compressed spaces
between verdant stands of
reeds and grasses.
Figure 8.37
The lovely panels punctuate the
landscape and frame views of
the city. Notice in this image
that industrial relics are also re-
purposed as wetland planters
where the intense green of the
grass-like plants contrasts with
the reddish and aging steel.
Figure 8.38
Shade structures supported by
panels made from re-purposed
steel create intimate spaces
along the river.
Figure 8.39 This isometric drawing shows a sequence of two- and three-dimensional elements
fashioned from the steel plates recovered from the former industrial site. See Figures 8.41 and 8.42
for images of the built elements.
Figure 8.40 This plan view of the steel elements shown in the drawing above places them in the
contest of the boardwalk, seating and plantings.

Steel structure and paving along the boardwalk. Steel paving and bench.
Photo: Gary Austin, 2013. Photo: Gary Austin, 2013.
Unlike some constructed wetland projects, Houtan Park is beautiful. The

landscape architecture has achieved the technical outcome (water purification and flood
control) while simultaneously providing scenic diversity. The park is a treat to visit
because of the careful attention to colour, texture and form in the planting design
(Figure 8.33) and the man-made elements and spaces. This dynamic landscape is
bursting with the fecundity of nature expressed in the rapid growth of the wetland
plants, the blossoms of the water lilies and the panels of crops or groves of trees.
Education is a cultural ecosystem service. A display at the entrance to Houtan
Park educates people about the natural treatment of polluted river water. The
constructed wetlands and the associated boardwalk afford opportunities for parents
and other teachers to engage in active learning about the natural world. Parents and
their children wielding nets and studying containers filled with aquatic organisms are
a common sight in Houtan Park (Figure 8.43). Even if education is not the goal of
the visitor, no one can resist the sight and sounds of birds or the masses of grasses
and wetland vegetation. The opportunity to be surrounded by nature is a rare joy for
the urban dweller.
Large and small groups of people can be accommodated in the many terraces
(Figures 8.22 and 8.44) distributed throughout the park. Lovers, families and large
community celebrations can simultaneously find settings that suit them. Having places
to socialize away from the distractions of urban life is very important and improves
the quality of life in the city.
Maintenance
Initially, the maintenance requirement of the park was high in order to establish the
terrestrial and aquatic systems. The high aesthetic expectations associated with an
Figure 8.43
The bamboo boardwalk places
the user at the water surface.
A dip net and bucket are all
that is required to entertain and
educate young and old alike.
Figure 8.44
Fishing docks and terraces
connected by steps that serve
as seating provide sunny
viewpoints or gathering areas
shaded by bosques of trees.
international exposition contributed to an extensive maintenance effort. However, the

landscape architect was aware that the maintenance budget would be reduced when
the exposition ended. The dense planting and ample establishment period reduced the
long-term effort required to sustain the park as an attractive place.
The performance of the constructed wetland was also linked to maintenance efforts.
For a few years, the wetland plants were harvested at the end of the growing season
to remove dead leaves and stems that contribute organic matter to the wetland. The
additional organic matter was undesirable, as it is converted to ammonia and nitrate
as it decomposes. While this is a natural process, the polluted river water contains
these substances at a higher level than desirable. As the constructed wetland became
robust, maintenance was reduced without substantial changes to the performance of
the treatment system. Today, the maintenance requirement of the park is low and the
aquatic ecosystem is stable (Figures 8.45 and 8.46).
Economic impact
The constructed wetland provides a pollution reduction service that can be
economically quantified. The hundreds of thousands of gallons of river water treated
daily in the Houtan Park constructed wetland could have been treated to an equivalent
quality by a conventional wastewater treatment plant. The wetlands save $500,000
per year when compared with the conventional treatment alternative. Furthermore,
the wetland requires almost no energy to operate, as sunlight and biological process
are responsible for the treatment.
Another way that Houtan Park is sustainable is through the recovery and reuse
of materials. The construction of the park reduced waste disposal costs. An estimated
$17,300 was saved by incorporating into the project features 37 tons of steel and
approximately 34,000 post-industrial bricks found on the site.
Most dramatically, Houtan Park and the Shanghai Exposition were the catalysts
for the economic redevelopment of this former industrial district. Blighted, brownfield
sites often lay vacant for decades due to the investment and development risks that
they pose. As in this case, public investment is often necessary to assure lending
institutions and developers that this risk is being addressed and shared. This is
Figure 8.45
This is the last pool before the treated
water returns to the river. The clear water
contains lilies and floating duckweed with a
margin of common reed. The decks that
seem to float on the water are covered with
cloud-like screens. This whimsical space is
a popular destination.
Figure 8.46 Plan view of the last pool and the decks below the pergola.
perhaps even more critical in the uncontrolled markets in western countries than in
China, where the national government plays a leading role in the redevelopment of
derelict urban land.
Houtan Park maintained the flood control function and can contain 136,469 m3
of floodwater. It is important to note that this is a protection of the current and future
economic investment made along the river corridor.
Biological diversity
There is frighteningly little nature in large cities and fewer opportunities to restore
habitat lost to development. Houtan Park restored a place for plants, trees, birds,
wildlife, fish and other organisms. The initial step in this was the removal of toxic
substances that would prevent the emergence of any ecosystem. Followed by creating
the aquatic setting and planting numerous native plants and seeding aquatic organisms,
the park has begun to evolve a more complex community of species within habitat
niches. The continuation of the ecological corridor along the river until it connects
with natural and extensive habitat would extend the biological diversity benefits of
the park. The 28 species of plants and 22 species of aquatic organisms (see Appendices)
that the project placed on the site initiated a recovery. This is neither an entirely natural
ecosystem nor does it accommodate all of the riparian species displaced centuries ago
by urbanization. Nevertheless, this park is an example of the rich urban biodiversity
shared with human users (Figure 8.47). The biodiversity of the site increased
dramatically beyond the introduction of the initial plants and aquatic organisms. Today,
93 species of plants and over 200 species of birds and animals have been observed in
the park.
Climate change adaptation

Mitigation of global warming is an effort that requires changes at every scale. The
massive reforestation and other revegetation necessary to counteract the impacts of
fossil fuel combustion require coordination by national governments. Nevertheless,
each project that increases plant biomass makes a contribution. Wetlands are
characterized by very high primary production. This is augmented in Houtan Park
by the planting of 585 trees, including metasequoia, willow, privet and camphor (Figure
8.47). The park sequesters an estimated 242 tons of carbon annually in the extensive
wetlands, perennial plantings and trees.
The water in the constructed wetland is shaded and returns to the river at a lower
temperature. In an era of rising global temperature, this is an important consideration
for the habitat of aquatic organisms.
Community agriculture
Inspired by the terraced slopes of Chinese agricultural landscape, at Houtan Park
terraces were constructed to create use areas and to negotiate the 15–18 feet (5–6 m)
elevation change from the river edge to the road. These terraces also slow stormwater
run-off flowing toward the constructed wetland. Crops and wetland plants were selected
to create an urban farm on the slopes and terraces of the park. Like biodiversity,
community agriculture has been largely eradicated from the urban landscape, but
Figure 8.47
There is a good diversity of
plants, aquatic and upland
conditions. Combined with
various water depths and
adjacent upland vegetation, the
diverse planting at Houtan Park
creates habitat heterogeneity.
This provides ecological niches
that can be exploited by
different plant and animal
species.
Figure 8.48
There are several terraced
fields in Houtan Park where
food is grown. This is an
extremely rare example of
reintroducing community
agriculture into the public,
urban landscape.
Houtan Park restores an opportunity for locally grown food (Figure 8.48). The
produce is one aspect of the community agriculture, but education for urban dwellers
and the aesthetic benefits should not be overlooked. The agricultural landscape of
Houtan Park features golden blossoms in the spring, splendid sunflowers in the
summer (Figure 8.49), the fragrance of the ripening rice in the fall and a winter blanket
of green clover. The paved walks and the ecologically friendly boardwalks made of
renewable bamboo guide the visitor through the agricultural parcels.
Conclusion
Riparian constructed wetlands at Houtan Park, the Isar River reach through Munich
(Figures 8.4 and 8.53), the Sanlihe River Ecological Corridor Project (Figure 8.50),
and the Tanghe River project (Figure 8.51) reveal important planning and design
lessons. Planned in a multidisciplinary way, a single project can deliver social, eco-
nomic, and ecosystem benefits simultaneously. Restoration of the wasteland that
overwhelmed the Sanlihe River corridor led to exuberant urban development along
Figure 8.49
The textures and colours of the
wetland plants are presented to
the visitor with attention to
displays in every season.
Lotus, water lily, yellow flag iris,
purple loosestrife and this
Thalia are a few of the colourful
blooming plants. The range of
foliage texture and colour
sustains the garden as a
beautiful place to visit, even in
the winter. See the appendices
for a full listing of plants.
Figure 8.50 The Sanlihe River Ecological Corridor where riparian wetlands catalyze urban
development.
Figure 8.51
The Tanghe River red ribbon
bench and boardwalk.
Figure 8.52
Aerial perspective of the
Tanghe River project.
Figure 8.53 The Isar River near Munich. Notice the floodwall is covered with a vegetated
embankment. An example of the former vertical floodwall is in the lower right corner of the image.
Photo: Nikater: License: Creative Commons public domain.
the banks of the new community resource (Figure 8.50). A minor intervention in the
floodplain along the Tanghe River led to a striking and well-loved community
recreation resource (Figure 8.51). Pollution abatement is effective either through a
linear marsh, as at Houtan Park, or within the channel, as in the Tanghe River in
Qinhuangdao City, China (Figure 8.52). Freed from its narrow channel, the Isar River
provides improved flood protection, new recreational beaches, improved water quality,
restored biological diversity and diverse habitat within and near Munich (Figure 8.53).
References
DePalma, A. “New York’s water supply may need filtering,” The New York Times, 20 July
2006.
Dong Yue, Y., Zhang, Y., Liu, X., Li, Y., Jin, J., Duan, T., Li, J. and He, P. “Construction of
ecosystem and water quality control of Houtan Wetlands in Shanghai Expo Garden,” Wetland
Science, vol. 3, pp. 219–226, 2013.
European Environment Agency, “Nutrients in freshwater,” European Union, 2015.
Hassan, R. M., Carpenter, S. R., Chopra, K., Capistrano, D. and Millennium Ecosystem
Assessment, Ecosystems and human well-being. Washington: Island Press, 2005.
Kostel, J., Hey, D. and Wetlands Research, Inc., “Des Plaines River wetland demonstration
project water level fluctuation study,” Army Corps of Engineers, DACW23–01-C-0003, 2012.
New York City Department of Environmental Protection, “New York City 2011 drinking water
supply and quality report,” New York City, 2012.
RESTORE, “River by design: Rethinking development and river restoration,” Environment
Agency, Horizon House, Deanery Road, Bristol, LIT 8146, 2013.
Wang, W., Gao, J., Guo, X., Li, W., Tian, X. and Zhang, R. “Long-term effects and performance
of two-stage baffled surface flow constructed wetland treating polluted river,” Ecological
Engineering, vol. 49, pp. 93–103, December 2012.
Xu, B., Gao, N.-Y., Sun, X.-F., Xia, S.-J., Simonnot, M. O., Causserand, C., Rui, M. and Wu,
H.-H. “Characteristics of organic material in Huangpu River and treatability with the O3-
BAC process,” Seperation and Purifiction Technology, vol. 57, no. 2, pp. 348–355, October
2007.
9
Stormwater management and
sustainable development
Introduction
Constructed wetlands for stormwater comprise a sustainable development practice,
as they can be implemented at various scales, can be distributed or centralized, and
can produce multiple benefits. These benefits include economic efficiency, due to low
capital cost, low operation cost and low cost of recreation, as well as social and
ecosystem benefits. At the neighbourhood or municipal level, stormwater treatment
wetlands can serve as urban design and open space elements to structure and inspire
urban development (Figure 9.1).
Free water surface constructed wetlands are often created to manage stormwater
run-off rates and volumes, and to eliminate contaminants. They are immediately
recognizable as marshes, but bio-retention basins (rain gardens) have standing water
only during and immediately after moderate and severe storms. Although these
wetlands also feature plants, many of them are not commonly found in marshes.
Nevertheless, these treatment landscapes are vertical subsurface or horizontal
subsurface flow wetland designs that have been modified to respond to the periodic
filling and draining caused by storms. The planning, design, performance and
sustainability features of stormwater wetlands and bio-retention basins are the topics
explored in this chapter.
Figure 9.1 The Qunli National Urban Wetland Park receives water from the highly impervious
neighbourhoods that are being developed in this new town adjacent to Harbin, China.
Photo: Kongjian Yu.
176 STORMWATER MANAGEMENT
Management and treatment challenge

Landscape design for the management and treatment of stormwater run-off presents
challenges very different from those associated with the treatment of domestic sewage
or grey water (water from showers, sinks and laundry). Domestic wastewater treatment
is characterized by relatively consistent concentrations of pollutants that are continu-
ously fed to the constructed wetland. In contrast, stormwater flows intermittently to
the treatment wetland. In arid and semi-arid climates, long periods with little rain
stress the wetland plants, while in rainy climates the large volume of rainwater
presents the designer with different challenges. The volume and frequency of
stormwater varies greatly according to the seasonal and regional climate. Furthermore,
the stormwater contaminants and their concentration in the stormwater are determined
by the land use over which the stormwater flows. Therefore, the types and concen-
trations of pollutants are quite variable. The design challenge is to create a treatment
landscape to remove the target pollutants.
There are really two categories of problems associated with stormwater run-off.
One is attenuating flow volume and run-off rate increases caused by development.
The second is mitigating the pollution in non-point source run-off. This chapter
discusses these challenges and demonstrates that management and treatment goals can
be achieved simultaneously by constructed wetlands, while offering additional human
and ecosystem benefits.
Stormwater run-off characteristics
Volume and run-off rate

Typically, after development (75–100 per cent impervious cover), there is a 45 per
cent increase in surface run-off, a 20 per cent decrease in deep infiltration, a 15 per
cent decrease in shallow infiltration and a 10 per cent decrease in evapotranspiration
compared to a natural pre-development landscape. A surface run-off increase of as
little as 10 per cent can result in damage to local streams (US Environmental Protection
Agency, 2010).
Figure 9.2 illustrates that the amount of stormwater that runs off is much greater
after than before development, but it does not show that the run-off rate and velocity
are also greater. The time at which the peak run-off occurs is also reduced after
development, typically. The combination of greater volume and greater velocity often
exceeds the capacity of streams and rivers that receive the run-off, leading to many
adverse impacts. Flooding risks increase as the existing streams and rivers exceed their
capacity of containing the water within their banks. The increased velocity undercuts
banks, causing them to collapse into the channel. This creates channels with vertical
walls and fills the channel with sediment. Streamside vegetation is also undercut and
falls into the water, reducing the amount of shade and increasing water temperature.
For many fish and other aquatic organisms, the higher temperature and a soil, instead
of gravel, riverbed are habitat changes to which they cannot adapt. These observations
led to the first ordinances requiring stormwater management.
STORMWATER MANAGEMENT 177
40% Evapotranspiration 30% Evapotranspiration
10% Runoff
Natural
ground 55% Runoff 75%-100%
25% Shallow
cover impervious
infiltration
cover
10% Shallow
infiltration
25% Deep infiltration 5% Deep infiltration

Figure 9.2 Urban development and most other human changes from natural conditions result in
increased stormwater run-off at the expense of every other water pathway. Each of these deficits can
result in damage to ecological or human systems.
Image: author, adapted from US EPA (US Environmental Protection Agency, 2010).
Planning for stormwater
Parcel vs neighbourhood systems

Both stormwater management and treatment can be accomplished by systems
implemented at the parcel scale or at the neighbourhood scale. For example, the bio-
retention basin shown in Figure 9.3, combined with pervious paving in a parking
area for 550 vehicles, infiltrates more than 91 per cent of the annual run-off. Parcel
scale systems are often selected when retrofitting a failing centralized system. The
City of Philadelphia and many other cities in the United States are representative of
this approach. These aging cities have sewer systems that combine sanitary and storm-
water inputs in a single pipe network. These are called combined sewer systems. These
are problematic since high stormwater contributions overwhelm the capacity of
conventional municipal wastewater treatment plants. This results in the discharge
of inadequately treated sanitary wastewater (not compliant with discharge permits)
to natural receiving streams, rivers, lakes and oceans. In this case, the develop-
ment of a parcel scale programme to disconnect stormwater inputs from the combined
sewer system is much less expensive than construction of more or larger conventional
wastewater treatment plants.
Low-density development or where a conventional system of stormwater sewer
collection and piping is unavailable or prohibitively expensive are other circumstances
that favour a parcel scale approach for separate treatment of both storm and sanitary
flows.
In contrast, constructed wetlands planned and implemented at the subdivision or
neighbourhood scale might be more economically efficient and provide secondary
Figure 9.3
The bio-retention swale at
Olympic College Parking
Garden in Bremerton,
Washington, USA, collects,
manages and treats the
stormwater on-site. This linear
basin consists of a temporary
pool, 18 inches of bio-retention
soil (60 per cent sand, 40 per
cent compost) on top of 24
inches of crushed rock.
Photo: SvR Design Company,
www.svrdesign.com: reproduced
with permission.
neighbourhood benefits, such as recreation space, wildlife habitat, educational

opportunities or other simultaneous advantages. Stapleton, Colorado (population
74,000), is an example of a community that adopted this strategy. The volume of
stormwater is necessary in this neighbourhood to restore wildlife habitat destroyed
by a previous land use (Figure 9.4). Areas adjacent to the stormwater facilities
serve as recreation and open space as well. This multiple use approach takes advantage
of several ecosystem benefits for an economically efficient achievement of several project
goals.
Regardless of whether the parcel or the neighbourhood approach is adopted, the
use of constructed wetlands for stormwater management is an opportunity to
sustainably minimize run-off rate and volume, remove pollutants before they can reach
the natural environment, minimize costs for construction, operation and maintenance,
and increase the aesthetic character of the neighbourhood.
Municipal planning
Comprehensive planning at the municipal or watershed scale is vital for the achieve-
ment of sustainable development. This activity should be initiated by the municipality
in concert with the county through a coordinated public planning process supported
by planners, landscape architects, scientists and engineers. The key goals are to
identify public safety risks associated with flooding and landslides, identify critical
habitat and unique ecosystems, plan a continuous network of corridors and spaces
that connect these, and serve human recreation needs as well. This planning will reveal
opportunities to mitigate existing flooding and pollution problems while avoiding future
pollution discharges. The long-term economic, social and environmental benefits of
this planning and associated development are incalculable. The master planning
process in Wulijie, China, presented in Chapter 1, is an example of this approach.
Incremental development in cities is the most common way that cities expand.
This often happens without visionary planning and most often occurs as planners and
the public react to development proposals and attempt to mitigate some of the
negative outcomes in the presence of pressure from developers and political officials.
Figure 9.4
The community of Stapleton
collects stormwater run-off in a
conventional network of catch
basins and pipes, but directs
the water to recreation and
habitat areas, where high
volumes are detained but lower
volumes are retained to
support the habitat and
recreation goals of the project.
The, often negative, outcome of reactive and incremental development is mitigated

by stormwater regulations.
State, or province, county and municipal government may all establish regulations
to manage and treat stormwater. As development permits are usually awarded at the
municipal or county level, stormwater regulations by these agencies are often most
effective. However, since national agencies often regulate pollution discharges, they
are involved in water pollution mitigation and flood control at the largest scale. When
local regulations and development guidelines are consistent with a comprehensive
municipal or watershed plan, then long-term goals are more likely to be achieved
without future costs for pollution clean-up or environmental restoration that have so
burdened the taxpayer in the past (and current) environment.
Stormwater regulations
There are three stormwater management and water quality goals that have evolved
over the last several decades. The first was the control of flooding and the prevention
of channelization of streams and rivers. The second was the treatment of stormwater
to protect human health and the aquatic ecosystem. The third is matching the amount
of pre-development infiltration in the post-development condition. Various states,
counties and cities require achievement of some or all of these goals to a certain degree.
Flood control and prevention of channelization

State, county and city governments regulate the rate of stormwater run-off from a
site proposed for development. This rate is generally limited to the run-off rate that
exists in the natural pre-developed condition. Therefore, if the pre-development run-
off rate was 10 cubic feet per second (cfs), for example, and the post-development
rate is 15 cfs, then storage of some stormwater on the developed site would be required.
Water may continue to drain from the site (after the storm has passed or abated) at
the pre-development rate (10 cfs) until the temporary storage basin is empty. Therefore,
stormwater is simply detained on-site. The amount or volume of run-off also increases
with development, but the rate and time of peak run-off can be maintained at pre-
development levels in order to prevent channelization of natural streams and rivers.
Later in the chapter, we will see that stormwater treatment and infiltration
requirements are often added to the zero-increase-in-run-off-rate requirement.
As noted above, to maintain a pre-development run-off rate, some stormwater
must be held on the developed site. The amount to be held depends on the regional
climate (Figure 9.5) and land use. Of course, rainstorms occur at different intensities
and durations even within a specific climate context. The US Department of Commerce
through the National Oceanic and Atmospheric Administration provides digital maps
of rainfall depth for storms of different magnitudes for all areas of the nation (Figure
9.6). Stormwater ordinances do not require detention of rainfall from very large storm
events. Therefore, a ‘design storm’ is defined in stormwater regulations. Often, this
is the 10-year and 25-year storm for residential and commercial property, respectively.
This means that the amount of water stored temporarily on-site and the permitted
run-off rate are based on storms of an intensity expected to occur once in 10 years
(a 10 per cent chance of occurring in any year) or once in 25 years (4 per cent chance
every year). Storms larger than these result in temporary flooding. The design storm
is loosely related to the value of the property being protected from flooding. Generally,
only state and federal agencies are charged with protecting property from flooding
by 50- or 100-year storms. Once the design storm is selected, the pre- and post-
development run-off rates and volumes can be estimated for individual parcels or entire
cities. For example, 10- and 25-year 24-hour storms might deliver 2.4 and 3.0 inches
of rainfall to a site on average. Knowing this depth and the area of the site allows for
calculation of the volume of rainfall for the site. The site will capture some of this
volume and the remainder will run-off. Therefore, we can estimate run-off volumes
for both natural and developed conditions of the same property.
Figure 9.5 In different regions of the United States, rainstorms display similar duration, frequency and
intensity characteristics. Four rainstorm types have been defined and mapped. The TR-55 run-off
calculation software requires that the rainfall type be defined.
Image: Gary Austin.
Figure 9.6
This portion of an isopluvial map
shows rainfall depth (in tenths of
an inch) for the 2-year 24-hour
storm.
Image: US National Oceanographic
and Atmospheric Administration.
License: US government work
product (National Oceanic and
Atmospheric Administration, 1973).
Calculation methods
There are several manual and computerized methods of estimating pre- and post-
development volume and run-off rates. Two of the simplest ones are the rational
method and TR-55. The rational method provides the peak run-off rate using the
equation below, but another procedure is necessary to determine volume.
Q = CIA
Where Q is the peak run-off rate in cubic feet per second (cfs)
C = run-off coefficient (from Table 9.1)
I = rainfall intensity (inches per hour), derived from a graph such as the
one in Figure 9.7
A = area (acres)
The rainfall intensity for various design storms is derived from Intensity–Duration–
Frequency curves (Figure 9.7), which are generally available from county engineering
departments or flood control districts or city surface water management departments.
Table 9.1 shows run-off coefficients for common surfaces and development
conditions. The coefficient is the percentage of water that runs off the surface.
To illustrate the use of the rational formula, assume a 10-acre site with loam soil
and light vegetation and an estimated time of concentration of 60 minutes. Figure 9.7
yields a rainfall intensity of about 1.5 inches per hour for a 60-minute time of
concentration and a 2-year storm. The run-off coefficient of 0.10 is acquired from
Table 9.1 for loam soil with light vegetation. With this information, we can calculate
the peak run-off rate for the pre-development situation. This is generally the maximum
run-off rate in the post-development condition.
Q = CIA
Q = 0.10 × 1.5 × 10
Q = 1.5 cfs
Figure 9.7 Intensity–Duration–Frequency curve for a region in Delaware, USA. Enter the graph at the
bottom (the time of concentration is the number of minutes in the storm when all areas are
contributing to the run-off outflow from the site, which is dependent on distance, slope and land
cover). From the TOC, extend a line vertically and pivot on the design storm curve, then continue left
to read the rainfall intensity on the left axis (State of Delaware, 2008).
Photo: State of Delaware, Delaware Department of Transportation. License: government work product.
The post-development rate can also be calculated using the formula and then
compared to the pre-development rate. The temporary storage of water on the
developed site is discharged at the pre-development rate by an opening in the outlet
structure sized to limit the outlet flow rate.
TR-55
TR-55 stands for Technical Report 55. This uninspired title refers to a stormwater
run-off rate and volume calculation methodology developed by the US Government
in 1986, and updated as a public software programme called WinTR-55. It is a
Windows-based programme that can be downloaded without cost, from www.nrcs.
usda.gov/wps/portal/nrcs/detailfull/national/water/?cid=stelprdb1042901. The web
Table 9.1 Recommended run-off coefficients (C) for the rational method, public domain (State of
Delaware, 2008)
Type of drainage Run-off

Area surface coefficient, C
Earth surfaces Sand from uniform grain size with no Bare 0.15
fines to well graded with clay or silt Light vegetation 0.10
Dense vegetation 0.0
Loam, from sandy or gravelly to clayey Bare 0.20
Light vegetation 0.10
Gravel, from clean gravel and gravel Bare 0.25
sand mixtures with no silt or clay to Light vegetation 0.15
high clay or silt contents Dense vegetation 0.10
Clay, from coarse sandy to pure Bare 0.40
colloidal clays Light vegetation 0.30
Lawns Sandy soil Slope, flat to 2% 0.05
Average slope 2% to 7% 0.10
Steep slope over 7% 0.15
Clay soil Slope, flat to 2% 0.13
Average slope 2% to 7% 0.18
Steep slope over 7% 0.25
Pavements Asphalt 0.95
Pavements 0.95
Compacted graded aggregate or gravel 0.70
Railroad yard areas 0.30
Parks, golf courses and cemeteries 0.15
Playgrounds 0.25
Unimproved areas 0.15
Cultivated areas 0.30
Swamp or marsh areas 0.08
Roofs 0.90
Drives and walkways 0.90
City business areas 0.85
City with dense residential areas and varying soil and vegetation conditions 0.70
Residential areas, single family units 0.40
Residential areas, duplexes and twins 0.50
Residential areas, multi-family units 0.70
Suburban residential areas (lots 1⁄2 acre or more) 0.35
Apartment complexes 0.60
Light industrial areas 0.70
Heavy industrial areas 0.80
site includes presentations and user guides with examples of common stormwater
situations.
TR-55 calculates both the peak run-off rate and the run-off volume, and assists
in estimating the size of the wetland basin and discharge piping. The software is
appropriate for small watersheds and sites with one or more catchment areas (Figure
9.8). TR-55 introduces hydrologic soil groups since soil texture, land cover and slope
influence run-off. TR-55 divides soils into four groups (A-D) according to soil texture
and infiltration rate. See the Appendices for hydrologic soil assessment criteria.
Figure 9.8 TR-55 uses local rainfall data, time of concentration, hydrologic soil groups, curve
numbers, drainage network and detention or retention basins in its calculation of stormwater run-off
times, rates and volumes for both the pre- and post-development conditions. The software is simple
to use once the terminology and concepts are mastered.
Image: Natural Resources Conservation Service. License: government work product.
Group A – Soils in this group have low run-off potential when thoroughly wet.
Group A soils typically have less than 10 per cent clay and more than 90 per cent
sand or gravel. Some soils having loamy sand, sandy loam, loam or silt loam textures
may be placed in this group. The saturated hydraulic conductivity of group A exceeds
10 micrometres per second (1.42 inches per hour) (USDA, 2007).
Group B – Soils in this group have moderately low run-off potential when
thoroughly wet. Group B soils typically have between 10 per cent and 20 per cent
clay and 50 per cent to 90 per cent sand, and have loamy sand or sandy loam textures.
Some soils having loam, silt loam, silt or sandy clay loam textures may be placed in
this group. The saturated hydraulic conductivity of group B exceeds 4.0 micrometres
per second (0.57 inches per hour) but is less than 10.0 micrometres per second (1.42
inches per hour) (USDA, 2007).
Group C – Soils in this group have moderately high run-off potential when
thoroughly wet. Group C soils typically have between 20 per cent and 40 per cent
clay and less than 50 per cent sand and have loam, silt loam, sandy clay loam, clay
loam and silty clay loam textures. The saturated hydraulic conductivity of group C
exceeds 0.40 micrometres per second (0.06 inches per hour) but is less than 4.0
micrometres per second (0.57 inches per hour) (USDA, 2007).
Group D – Soils in this group have high run-off potential when thoroughly wet.
Group D soils typically have greater than 40 per cent clay, less than 50 per cent sand,
Table 9.2 Runoff curve numbers for urban areas 1/ (USDA, 1986)
Cover description Curve numbers for

hydrologic soil group
Cover type and hydrologic condition Average A B C D
per cent
impervious
area 2/
Fully developed urban areas (vegetation established)
Open space (lawns, parks, golf courses, cemeteries, etc.) 3/:
Poor condition (grass cover < 50%) 68 79 86 89
Fair condition (grass cover 50% to 75%) 49 69 79 84
Good condition (grass cover > 75%) 39 61 74 80
Impervious areas:
Paved parking lots, roofs, driveways, etc. 98 98 98 98
(excluding right-of-way)
Streets and roads:
Paved; curbs and storm sewers (excluding right-of-way) 98 98 98 98
Paved; open ditches (including right-of-way) 83 89 92 93
Gravel (including right-of-way) 76 85 89 91
Dirt (including right-of-way) 72 82 87 89
Western desert urban areas:
Natural desert landscaping (pervious areas only) 4/ 63 77 85 88
Artificial desert landscaping (impervious weed barrier, 96 96 96 96
desert shrub with 1″–2″ sand or gravel mulch and
basin borders)
Urban districts:
Commercial and business 85 89 92 94 95
Industrial 72 81 88 91 93
Residential districts by average lot size:
1
⁄8 acre or less (town houses) 65 77 85 90 92
1
⁄4 acre 38 61 75 83 87
1
⁄3 acre 30 57 72 81 86
1
⁄2 acre 25 54 70 80 85
1 acre 20 51 68 79 84
2 acres 12 46 65 77 82
Developing urban areas
Newly graded areas (pervious areas only, no vegetation) 5/ 77 86 91 94
Idle lands (CNs are determined using cover types similar to those in the table titled ‘Runoff curve numbers for
other agricultural lands’). For curve numbers for agricultural and range land uses, see the Appendix.
and have clayey textures. The saturated hydraulic conductivity of group D is less than
or equal to 0.40 micrometres per second (0.06 inches per hour) (USDA, 2007).
TR-55 also introduces the run-off curve number, which is very similar to the
coefficient of run-off (C) in the rational calculation method discussed above. Table
9.2 provides the curve numbers for various types of land uses (cover descriptions)
with one of the four hydrologic soil groups. See the Appendices for tables with curve
numbers for additional land uses. We will return to consideration of hydrologic soil
groups and curve numbers in the discussion of infiltration later in this chapter.
Stormwater quality improvement

Stormwater contaminants
While the treatment of domestic sewage concentrates on the reduction of total
suspended solids (TSS), biological oxygen demand (BOD), total nitrogen, ammonia,
nitrate, phosphorus, disease-causing bacteria and pharmaceuticals, stormwater
often contains heavy metals, hydrocarbons, phenols and other toxins, some of which
are known carcinogens. Typical concentrations of pollutants in urban stormwater are
TSS – 54.5 mg/L, total phosphorus – 0.26 mg/L, total nitrogen – 2.0 mg/L, copper –
11.1 ␮g/L, lead – 50.7 ␮g/L, zinc – 129 ␮g/L, faecal coliforms – 1.5 colony-forming
units (CFU)/100 mL (Kadlec, 2009). However, these levels vary considerably depending
on the type of land use that generates the stormwater. Some land uses contribute high
concentrations of particular stormwater contaminants (Table 9.3). Most of the con-
centrations of nutrients, bacteria and metals from the highlighted land uses are far above
the US Environmental Protection Agency (EPA) standards for surface water.
All of the land uses listed in Table 9.3 contribute all the listed pollutants, but the
table displays only pollutant concentrations for the land uses contributing the highest
levels. This helps us select the type of treatment wetland best suited for the removal
of the most problematic contaminants.
Table 9.3 Stormwater run-off pollution concentration and land uses making the greatest
contributions of pollutants (Kadlec, 2009)
Contaminant TSS E. coli TN P Copper Lead Zinc

mg/L CFU mg/L mg/L ␮g/L ␮g/L ␮g/L
Land use
Lawns 602 9.1 2.1
Commercial streets 468
Auto recyclers 335 103 182 520
Industrial parking 228
Landscaping 94,000 94 182 263
Residential streets 37,000 .55
Driveways 2.1 .56
Urban highways 3 .32 54 400 329
Rural highways 22
Industrial roofs 62 43 1,390
Heavy industrial land 148 290 1,600
Water quality standard 30 126 .05 13 65 120
Kadlec, 2009. CFU = colony-forming units per 100 mL; mg/L = milligrams per litre; ␮g/L = micrograms/L; TSS
= total suspended solids; E. coli = pathogenic bacteria; TN = total nitrogen; P = phosphorus.
Species of pathogenic bacteria, in addition to E. coli, noted in Table 9.3, are

also present in high concentrations in storm run-off. Faecal coliform, total coliform
and E. coli bacteria are highly correlated with one another and with turbidity and
suspended solids concentrations (Peters, 2009). Total nitrogen listed in Table 9.3
includes organic nitrogen, ammonium, nitrite and nitrate. While there are no water
quality standards for total nitrogen, concentrations of 0.2 mg/L for nitrite and 1 mg/L
for nitrate are generally accepted maximums (Li and Davies, 2009).
The contaminants in urban and suburban stormwater accumulate on surfaces
through atmospheric deposition during dry periods, through vehicle emissions from
burning fossil fuel, through the wear and disintegration of tyres, brakes, engine and
other components, through animal waste from pets, and through fertilizers, pesticides
and other chemicals applied to the landscape, buildings and paving. However, diesel
oil, concrete, asphalt, undercoating, exhaust and antifreeze contribute contaminates
also. Rain run-off washes these contaminants into surface waters. The concentration
of contaminants is especially high after periods with no rain and during the initial
flow from a storm. About the first 25 per cent of run-off is particularly polluted and
is called the ‘first flush’. Therefore, it is particularly important to capture this portion
of run-off and remove as much of the pollution as possible.
Parking lot contaminants

Table 9.4 shows the levels of metals, suspended solids and 25 types of polycyclic
aromatic hydrocarbon (PAH) from parking lot run-off after simulated rain in a
southern California city (Long Beach). PAHs are formed from burning coal, oil, gas
and garbage, when combustion is incomplete. It also enters the environment through
pavement-sealing coats made from coal tar (common in the eastern United States)
(Mahler et al., 2012). Breathing PAH-polluted air or consuming contaminated water
or food creates human health risks, including cancer. The compounds are long-lasting
and most are insoluble in water. PAHs are associated with high levels of TSS in urban
run-off.
The California study revealed that a rainfall intensity of at least 0.24 inches per
hour (6 mm/hour) is required to wash contaminants from the paving surface. The
Table 9.4 Levels of parking lot contaminants after a 28-day antecedent dry period. Long Beach,
California (Tiefenthaler, Schiff and Bay, 2001)
Parameter Mean (SD) Parameter Mean (SD)

Suspended solids (mg/L) 51.8 (14.1)
Total metals Dissolved metals
Aluminum (␮g/L) 533.3 (119.2) Aluminum (␮g/L) 131.7 (62)
Cadmium (␮g/L) 2.5 (1.4) Cadmium (␮g/L) 1.3 (1.2)
Chromium (␮g/L) 3.6 (0.5) Chromium (␮g/L) 2.3 (1.1)
Copper (␮g/L) 40.3 (7.2) Copper (␮g/L) 28.5 (13.4)
Iron (␮g/L) 810.0 (174.4) Iron (␮g/L) 286.7 (140.0)
Lead (␮g/L) 41.8 (10.6) Lead (␮g/L) 22.8 (14.0)
Mercury (␮g/L) 0 (0) Mercury (␮g/L) 0 (0)
Nickel (␮g/L) 20.7 (2.4) Nickel (␮g/L) 16.2 (7.7)
Silver (␮g/L) 0 (0) Silver (␮g/L) 0 (0)
Zinc (␮g/L) 620.0 (60.4) Zinc (␮g/L) 131.7 (62)
Total PAHs (␮g/L) 82.4 (33.7)
concentration of contaminants increased approximately 600 per cent during a 28-day

antecedent dry period, but was lower than this maximum after 2- and 3-month
antecedent dry periods. All loose particles were washed from the paving surface during
the first 15 minutes of simulated rainfall. The highest concentration of pollutants was
found in samples collected after the first 10 minutes of light simulated rainfall
(6 mm/hour). Higher intensity rainfall (13 and 25 mm/hour) diluted the concentration
of pollutants (Tiefenthaler, Schiff and Bay, 2001).
The toxicity of the stormwater run-off was characterized according to its impact
on sea urchins (Strongylocentro purpura), ocean bacteria (Vibrio fischeri) and mysids
(Americamysis bahia, marine crustaceans). All stormwater samples, whether taken
during the first flush or later in the storms, were toxic to these organisms. The first
flush samples were twice as toxic as other samples. All three organisms died or were
affected in reproduction or growth. The primary factors in this toxicity were heavy
metals, particularly zinc. Street sweeping and the intensity of parking lot use did not
significantly change the contamination levels. Notice in Table 9.4 that the proportion
of total metals which are in the dissolved state is high. This is of concern since this
pollution form is more difficult to remove from stormwater than metals in particulate
form (Tiefenthaler et al., 2001).
Roadway contaminants
Like parking areas, roads are particularly potent sources of pollutants. Annually,
highway stormwater run-off at the concentration of 0.71–1.07 mg/L contributes
4.76–7.64 pounds per acre per year of nitrate to the adjacent landscape and receiving
waters, while total phosphorus at the concentration of 0.10–0.33 mg/L generates
0.72–1.76 pounds per acre per year (Hunt, 2003).
Table 9.5 shows data from a Texas study of highways with three levels of daily
vehicle use (>50,000, >43,000, >35,000) and the pollutants that washed off the
pavement surfaces during storms. This data helps us predict the concentration of
pollutants that will require treatment. This data is confirmed by other studies, but the
values vary considerably according to continent and within continents. Higher
Table 9.5 Mean contaminant concentration for six roads in Texas, with an average daily traffic count
of 35,000 to more than 50,000. Units mg/L (Li et al., 2008)
Parameter >50,000 >50,000 >50,000 >43,000 >35,000 >35,000

TSS mg/L 116 172 124 118 124 173
TKN mg/L 2.131 1.940 1.786 1.130 1.510 1.760
NO3+2 mg/L 0.37 1.064 0.897 0.434 0.340 0.220
Total phosphorus mg/L 0.217 0.238 0.216 0.132 0.130 0.280
Dissolved phosphorus mg/L 0.119 0.112 0.126 0.040 0.050 0.090
Total copper ␮g/L 14 17 16 27 22 30
Dissolved copper ␮g/L 6 6 6 6 6 5
Total lead ␮g/L 7 9 6 13 10 14
Dissolved lead ␮g/L Below detection limits
Total zinc ␮g/L 117 118 112 167 140 175
Dissolved zinc ␮g/L 48 44 45 57 49 50
COD mg/L 73 71 92 64 81 100
TSS = total suspended solids; TKN = total Kjeldahl nitrogen; NO3+2 = nitrite and nitrate; COD = chemical
oxygen demand.
contaminant levels are reported for Europe and Asia than for the United States. There
are strong correlations between TSS, total dissolved solids, total organic carbon and
iron and 13 other constituents and parameters (turbidity, oil and grease, total
petroleum hydrocarbons, total Kjeldahl nitrogen (TKN), electrical conductivity,
chloride, cadmium, copper, chromium, nickel, lead and zinc). Therefore, TSS and the
other three indicators can be used as general assessments of the presence and removal
of the other 13 contaminants (Kayhanian et al., 2012).
General urban contaminants

The negative impact on the ecosystem of stormwater from parking and roads is
confirmed by studies of general urban run-off where heavy metals, bacteria and
excessive nutrients are present and contribute to the impairment of water bodies. For
example, over 7,200 water bodies in California are defined as impaired and about
1,200 of these are impaired due to excessive levels of bacterial pathogens (Kayhanian
et al., 2012).
Although it is diluted as it flows into streams and rivers, the level of pollution in
urban stormwater degrades streams and rivers. A study of the effects of urbanization
on streams was conducted from 2003 to 2007 on watersheds with 69–93 per cent
imperviousness. Twenty stream watersheds (Figure 9.9) near Atlanta, Georgia, were
analyzed for a range of contaminants, including metals that are common in highway
run-off. The first 25 per cent of stormwater run-off from impervious surfaces (the
first flush) resulted in high concentrations of copper, lead and zinc, but high levels
sometimes occurred later in the storms. Copper and zinc in most of the streams
exceeded Georgia’s chronic and acute standards (chronic levels damage aquatic
organisms when exposure exceeds 4 days; acute levels damage aquatic organisms when
exposure exceeds 1 hour). Lead was often detected at chronic levels. Copper, lead
and zinc concentrations are highly correlated with each other and are present at high
levels due to run-off from impervious surfaces (Peters, 2009).
Figure 9.9
The urban and suburban
watersheds in the study of
water quality in streams near
Atlanta, Georgia.
33°44′44.83″ N 84°23′57.21″ W
image date: 8 February 2013,
image accessed: 7 June 2015.
The Atlanta stream water quality study showed that faecal coliform bacteria levels
exceeded the US EPA standard (200 most probable number (MPN)/L) for recreation
uses in more than 90 per cent of the test samples taken from urban watershed streams.
Faecal coliform, E. coli and total coliform bacteria were highly associated with
suspended solids levels in the streams. Alkalinity, sodium, calcium, magnesium
and potassium concentrations in the urban streams were four to ten times higher than
in streams in more natural watersheds. This is probably due to the weathering of
concrete. Generally, nutrient levels in the streams were high compared to streams
in natural areas, but lower than EPA maximum standards for streams. For example,
most of the urban streams had much less nitrate than 10 mg/L (EPA drinking water
standard), and less than 10 per cent of the streams had more than 0.1 mg/L of
phosphate (EPA limit causing algae blooms). High levels of chlorine were associated
with combined storm and wastewater sewer outflow treated with sodium hypochlorite,
as well as drainage from swimming pools and road de-icing salts (CaCl2) (Peters, 2009).
Contaminants from the ornamental landscape

Insecticides, herbicides, pathogenic bacteria and fertilizers containing nitrogen and
phosphorus are stormwater contaminants generated in large quantities by the
ornamental landscape.
Phosphorus
Phosphorus occurs naturally from the weathering of rock, but is present in the urban
environment largely due its distribution as an artificial fertilizer. It is highly reactive,
forming compounds with oxygen, iron, aluminium and calcium. Phosphate is the
most common form and the simplest form of phosphate is orthophosphate (PO4)
(phosphorus and 4 oxygen atoms). Phosphorus occurs in organic, inorganic, soluble
and particulate forms. It is most available to plants and microorganisms in the soluble
form. Phosphorus is added as fertilizer to turf to create about 25 mg/L of soil
phosphorus for optimum growth. However, only 25 ␮g/L (a quantity 1,000 times
smaller) can promote excessive algae growth in lakes; 50–70 ␮g/L of phosphors
cause significant and frequent algae blooms and the eutrophication of lakes. The decom-
position of grass clippings and tree leaves from the ornamental landscape releases
phosphorus. About 0.13 pounds of phosphorus (0.3 pounds P2O5) per 1,000 square
feet is included in grass clippings during the growing season, and a tree canopy of
50 per cent coverage results in about 0.5 mg/L of phosphorus in street stormwater
run-off (Spetzman, 2004). The target for phosphorus is 0.1 mg/L or less to avoid
excessive algae growth, but many streams have this problem even at concentrations
lower than this (US Environmental Protection Agency, 2010). The typical total phos-
phorus concentration in urban stormwater run-off in the United States is 0.26 mg/L.
Nitrogen
Nitrogen is also a nutrient that causes growth of algae in aquatic systems, especially
when phosphorus is also in excess (Figure 9.10). A survey of streams in the United
States revealed that 40 per cent had high levels of phosphorus and 28 per cent had
high levels of nitrogen. These excess nutrients contribute to aquatic conditions
Figure 9.10
The rapid growth and
subsequent death and
decomposition of algae, as
seen in this California urban
stream, have devastating
effects on the ecosystem, as
this process depletes the
oxygen in the water, causing
damage or death to aquatic
organisms.
Photo: University of California,
Davis. License: CC-BY-SA_2.0.
harmful to macro-invertebrates (US Environmental Protection Agency, 2013). Nitro-

gen enters aquatic systems through atmospheric deposition (as a product of burning
fossil fuels), the decomposition of organic matter and the use of artificial fertilizer. In
the urban environment, pet waste and treated sewage are significant sources of excess
nitrogen entering streams, rivers, lakes and oceans. The amount of nitrogen from
fertilizer present in stormwater run-off varies according to the amount and type of
fertilizer (quick or slow release), the soil infiltration rate, the slope of the land, storm
intensity and density of vegetation.
Streams not impacted by human activities have total nitrogen levels ranging from
0.12 to 2.2 mg/L (US Environmental Protection Agency, 2010). The typical total
nitrogen concentration in urban stormwater run-off in the United States is 2.0 mg/L.
Nitrate above approximately 5 mg/L causes algae blooms, loss of dissolved oxygen
and eutrophication. The nitrate drinking water standard for adults is 10 mg/L.
Ammonia (a form of nitrogen) is toxic to aquatic organisms, but toxicity varies by
species, pH and temperature. However, the EPA acute (1 hour) criterion standard for
freshwater is 17 mg total ammonia nitrogen/L and the chronic (30-day average)
criterion is 1.9 mg/L (at pH of 7 and temperature of 20°C). These maximum levels
are intended to protect sensitive species such as salmon and mussels (US Environmental
Protection Agency, 2013).
Nitrogen is present in the form of organic or inorganic compounds. The inorganic

forms (ammonia and nitrate) of nitrogen are readily available to plants.
Stormwater treatment landscapes

The overview of stormwater contaminants above reveals a great diversity in the pollu-
tants generated by transportation, urban and suburban land use. High contaminant
levels, coupled with the episodic nature of storms and run-off makes reducing the
concentration and total amount of pollutants before they enter natural waterways a
challenge. The conventional use of detention basins that temporarily store stormwater
is ineffective for water quality improvement. Conventional wastewater treatment
infrastructure similar to that used to treat municipal sewage could certainly improve
the quality of stormwater, but the cost is prohibitive. Therefore, biological, landscape
and multifunctional solutions have been sought. Many of these have been designed,
constructed and monitored at the field scale. They provide us with reliable models
that we can adjust to respond to particular pollution problems, soil types, rainfall
intensities, etc.
Below, three treatment landscapes are presented with the information that
planners, engineers and landscape architects require for master planning, sizing and
construction detailing. The treatment performance of each system is also presented.
The initial concentration of contaminants is presented for each case study, and this
can be compared to the data in the tables above that characterize non-point source
pollution from stormwater run-off. The treatment of pesticides and run-off from
agricultural and industrial activities are presented in separate chapters. Pesticides used
in the ornamental landscape are of the same classes as pesticides used in agriculture,
and respond to similar treatment measures. This chapter concludes with a case study
of the Qunli National Wetland Park to emphasize the multifunctional advantages and
sustainable development potential of stormwater treatment landscapes.
JEL Wade Park case study

Context
Free water surface wetlands are the most common types of constructed wetlands for
stormwater treatment. The JEL Wade wetland was designed and constructed to
mitigate the impacts of non-point source pollution from suburban development on
Hewletts Creek, a tidal creek, located in the city of Wilmington, North Carolina.
Wilmington receives about 57 inches of rainfall per year. Figure 9.11 shows that the
watershed is 19 per cent impervious due to the mix of urban and residential
development. The wetland in JEL Wade Park has a 589-acre (238 hectares) catchment
area. The size of the wetland (including the upland recreation area of the park)
represents less than 1 per cent of its drainage area, but treats 47 per cent of its potential
run-off. Historically, faecal coliform bacteria and excess nitrogen contaminated the
creek causing algae blooms and the closure of shell fishing areas nearby. The wetland
(Figure 9.12) was designed to treat a 1-inch (2.5 cm) water quality storm, but it can
contain and convey 100-year, 24-hour rainfall storms (Mallin et al., 2012).
Figure 9.11
Approximately 20,200 people
live in the watershed of the JEL
Wade Park wetland.
Photo: Google Earth © Google
2015 © TerraMetrics 2015.
34°10′39.96″ N 77°52′40.43″ W,
image date: 3 January 2015,
image accessed: 25 April 2015.
Figure 9.12
JEL Wade Park was
constructed in 2007. It contains
the 11.5-acre (4.7 hectare)
wetland and 3.4 acres (1.4
hectare) of parkland. The area
around the wetland includes
park buildings, picnicking areas
and a 0.75-mile (1.22 km)
pedestrian loop trail including a
boardwalk across the wetland
(Li and Davies, 2009).
Photo: Google Earth © Google
2015, © TerraMetrics 2015.
34°10′39.96″ N 77°52′40.43″ W,
image date: 3 January 2013,
image accessed: 25 April 2015.
Design
The design of the JEL Wade free water surface wetland illustrates several recommended
design elements. The stormwater enters the wetland at one end through two inlets,
and flows into two sedimentation basins. These are deep basins where the heaviest
particles settle to the bottom. The sedimentation basin or forebay must be cleaned
out after several years for continued water quality performance of the wetland. The
6-feet-deep, 30-feet-long sedimentation basins (item F, Figure 9.13) are graded to
provide a relatively gentle vegetated entering slope for the first two-thirds of their length.
This encourages the growth of vegetation that slows and distributes stormwater, causing
sediments to settle to the bottom of the basin. Capturing the sediment is beneficial
for extending the life of the wetland, easy excavation of the accumulated soil, and
removal of heavy metals and disease-causing bacteria attached to the soil particles
(especially small particles) (Mallin et al., 2012).
Notched weirs at the outlet of each sedimentation basin evenly distribute the
stormwater across the wetland width, while the water drops 6 inches (15 cm) into
the wetland. Two of the weir notches contain flash board risers that slow the flow of
water from the sedimentation basin. Two additional weirs that cross the wetland collect
and redistribute water, as it drops in elevation at these low walls. The wetland contains
two low-flow streams that are fed by groundwater (Mallin et al., 2012).
Figure 9.13 Plan view of the design for JEL Wade Park and wetland. A – play area; B – playground;
C – restrooms; D – park building; E – vegetated area of sedimentation basin; F – sedimentation
basin; G – weirs; H – low-flow streams; I – boardwalk and overlook; J – wetland with emergent and
submerged marsh vegetation; K – outlet structure; L – by-pass channel; M – existing trees;
N – shoreline vegetation.
Image: Gary Austin based on design by Stewart Engineering.
Healthy, dense vegetation is a critical feature of a stormwater wetland. In the

Wilmington case, the groundwater is high enough that it provides sufficient water
to sustain healthy wetland vegetation during the dry periods. The wetland and
the sedimentation areas are dominated by cattail (Typha latifolia), Eastern bureed
(Sparganium americanum), pickerelweed (Pontederia cordata), soft rush (Juncus
effusus), giant cutgrass (Zizaniopsis miliacea), alligatorweed (Alternanthera phil-
oxeroides) and parrot feather (Myriophyllum aquaticum) (Mallin et al., 2012). The
complete plant list for the wetland is available in the Appendices.
The water quality storm flow is shallow and moves slowly through the wetland
to receive maximum contact with the stems of the wetland plants and bottom
sediment. The microorganisms and bacteria growing on the plants, sediment and
decaying vegetation establish an ecosystem that utilizes the organic material, ammon-
ium, nitrate, phosphorus and even complex and toxic organic compounds, such as
petroleum hydrocarbons (Mallin et al., 2012).
Treatment wetlands should be constructed with a flat or minimal slope (0.5–1
per cent), and the JEL Wade wetland maintains the level areas by dropping elevation
at the weirs (Figure 9.14). Water flowing through the wetlands commonly seeks
preferential routes across the wetland, which decreases the treatment effectiveness.
Therefore, redistribution of the water at weirs and a roughly circular or square shape
for the wetland, rather than long and narrow, are recommended. Both of these features
are present in the Wilmington example. The water depth should vary from 6 to 18
inches deep for the water quality storm. In the Wilmington example, the sinuous route
of the low flow channels extends the length of the flow pathway through the wetland
and crosses the storm flow to encourage distributed flow through the vegetation.
Figure 9.14
The JEL Wade wetland
sedimentation basin and its
outlet weirs are shown in
the foreground of this
image. The outlet is the
concrete structure in the
background.
Photo: City of Wilmington,
NC: reproduced with
permission.
Sometimes deeper (4 feet) trenches are placed across the flow path to encourage mixing
and redistribution of water, and to provide cooler water for fish habitat.
There is a small pool adjacent to the concrete outlet structure of the JEL Wade
wetland, where a discharge pipe and overflow weirs control the normal depth of water
in the wetland and the depth for storms greater than the water quality storm. The
entire wetland is contained by a berm that prevents stormwater from flowing into the
basin except at the inlet, and contains the 100-year storm (Mallin et al., 2012).
Water management
Although the JEL Wade wetland is intended primarily for water quality improvement,
the basin has a positive effect on the peak run-off rate and time, and permanently
stores a portion of each storm. Eight storms consistent with the magnitude of the water
quality storm, ranging from 0.4 to 1.4 inches (1 cm to 3.5 cm), were monitored for
6-hour durations. The data shows that the wetland dramatically reduced both the rate
and volume of the stormwater run-off from these storms. The wetland retained
50.3–74.5 per cent (63 per cent average) of the storm inflow. This is significant since
the volume retained in the wetland receives water quality treatment until it is removed
by the subsequent storm, base flow, infiltration or evapotranspiration. The average
inflow of 15,185 cubic feet (430 m3) per hour was reduced by the wetland to an outflow
average of 4,944 cubic feet (140 m3) per hour (Mallin et al., 2012).
Water treatment
The primary water quality deficit in the stormwater was pathogenic bacteria.
Historically, high pathogenic bacteria levels caused the closure of shellfish harvesting
in the downstream tidal reach of Hewletts Creek. Stormwater inflow greatly exceeded
the North Carolina standard for human contact for faecal coliform (200 CFU/100
mL), but the average level in the outflow, after treatment in the wetland, was well
below this standard (Table 9.6). The stormwater wetland decreased an average of 99
Table 9.6 Pollutant mean concentrations entering and exiting the wetland (Mallin et al., 2012)
Pollutant Inflow Inflow Outflow Inflow 1 Inflow 2

1 2 reduction % reduction %
Total suspended solids, mg/L 9.8 15.2 4.1 58.2 73.0
± 10.5 ± 7.2 ± 1.2
Ammonium ␮g/L 229 143 43 81.2 69.9
±128 ± 65 ± 32
Nitrate ␮g/L 123 159 66 46.3 58.5
± 42 ± 39 ±46
TKN, ␮g/L ␮g L 1,088 713 463 57.4 35.1
±1129 ±344 ±130
Total nitrogen, ␮g/L 1,210 871 529 56.3 39.3
± 1109 ± 374 ± 114
Phosphate 20 93 13 35.0 86.0
␮g/L ± 21 ± 25 ±5
Total phosphorus ␮g/L 59 150 43 27.1 71.3
± 44 ± 53 ± 41
Faecal coliform‡ 605 437 42 93.1 90.4
Zinc ␮g/L 14.12 44.12 11.0 22.1 74.0
± 7.562 ± 11.000 ± 6.80
‡ CFU, colony-forming units; CFU per 100 mL, geometric mean. Eight sampling events.
TKN: Total Kjeldahl nitrogen (organic nitrogen and ammonia).
± standard deviation; ␮g/L = micrograms per litre.
per cent of the faecal coliform bacteria. The average concentration of coliform bacteria
in the wetland outflow was 42 CFU/100 mL. Removal of coliform bacteria in the
wetland was highest when the inflow concentration was also high and during the
warmest weather. The warm weather effect is probably due to increased predation of
bacteria by microorganisms (Mallin et al., 2012). Removal of bacteria pathogens in
free water surface and other wetland types is highly dependent on residence time and
internal flow patterns. Rotifers and protozoa are microorganisms that prey
on bacteria. Rotifers are abundant in the outflow of treatment wetlands. They are
commonly about 10 rotifers in each millilitre of water in a treatment wetland. At this
concentration, rotifers can disinfect stormwater detained for 1.2 hours in a marsh
wetland (Kadlec, 2009).
The TSS concentrations flowing into the wetland were generally low. However,
when the TSS concentrations were highest in the inflow, the wetland removed
98.8 per cent of them. The Wilmington wetland was quite effective in the removal of
ammonium and nitrate, which is important for the improvement of the shellfish area
downstream. The wetland removed more that 90 per cent of the ammonium and
orthophosphate, and 89 per cent of the total phosphorus. Nitrate removal was very
high, especially in the warmer months. Nitrate removal was 68–83 per cent when
water temperatures were less than 53.6°F (12°C), but 90–97 per cent when the water
temperature exceeded 59°F (15°C) (Mallin et al., 2012).
In the case of the JEL Wade wetland, it was important to remove nitrate from
the stormwater, as it sustains pathogenic bacteria and causes algae blooms, which
could compromise the shellfishery downstream. Total nitrogen removal ranged from
66 to 96 per cent. Heavy metals concentrations in the inflow were low with the
occasional exception of zinc. The level of zinc in the outflow was always well below
the state standard (Mallin et al., 2012).
The performance of the JEL Wade wetland is confirmed by comparison to the
Dye Branch wetlands located in Mooresville, North Carolina. This wetland watershed
is 60 per cent impervious with a run-off curve number of 87.5. Completed in 2006,
this 0.81-acre stormwater wetland is composed of three treatment cells. The removal
of contaminants is even better than in the JEL Wade wetland, with the exception of
phosphorus, although this was still a very high removal percentage: TSS – 88 per cent,
total phosphorus – 67 per cent, total nitrogen – 58 per cent, organic nitrogen – 36
per cent, total ammonia – 90 per cent, TKN – 48 per cent, nitrite and nitrate – 85
per cent. The removal of organic nitrogen is fairly low. This is probably because free
water wetlands simultaneously remove and add organic nitrogen to the water, so it
may not be possible to improve this significantly unless a different type of wetland is
added to the treatment sequence (Hathaway and Hunt, 2009). In fact, there was only
a small improvement in the quality of the water after the second treatment cell. A
different kind of wetland, such as a horizontal subsurface flow cell, might have made
more significant improvement in some parameters. Other examples of stormwater
treatment wetlands show performance much lower than in the JEL Wade and Dye
Branch examples. This points to the need for careful evaluation of site conditions,
target contaminants and good design.
Other benefits of the JEL Wade wetland

The example of JEL Wade Park shows the multifunctional character of stormwater
wetlands for water quality improvement, open space, recreation and habitat. Although
the park is primarily dedicated to stormwater capture and treatment, there are several
additional uses. A mundane one is the electrical transmission line corridor that shares
the space. There are also public amenities including a restroom, a park building,
playgrounds, turf play areas, a pedestrian/bicycle trail, a boardwalk and even educa-
tional panels (Figure 9.13). In addition, the preservation of many groves of existing
trees and the planting of many other trees, shrubs and wetland vegetation generate
ecosystem and aesthetic benefits. Urban biological diversity is enhanced due to the
size of the park and types of vegetation planted. There is a wide range of ecosystem
elements and recreation and educational amenities that are normally expensive for
the city to provide. The combination of the water management, treatment and these
other ecosystem and human benefits reduces the cost of providing them, compared
to creating parks dedicated to each category (Mallin et al., 2012).
Off-line stormwater treatment wetlands

The JEL Wade stormwater wetland combines the stormwater management and
treatment functions into a single facility. However, this may not be the optimal design,
as during storms larger than the water quality storm, it can result in scouring of the
wetland soil, high water velocity, and deep and lengthy inundation of the vegetation.
Instead, parallel basins provide a better opportunity to control the treatment landscape
that is independent of the stormwater management function. Figure 9.15b illustrates
a design where the water quality storm volume is routed to the treatment cells,
while the larger volume of more intense storms is routed to a pond for detention.
Figure 9.15a
Plan view of stormwater
management and treatment
cells that operate in parallel.
In the plan, the sedimentation
basin is at the upper left. Water
flows to the large stormwater
management basin. A reverse
flow pipe delivers the water
quality storm to the first
treatment basin (lower left).
Image: Austin adapted from
Virginia Department of
Environmental Quality (2013).
Figure 9.15b
Perspective view of stormwater
management and treatment
cells. In the perspective, the
sedimentation basin is at the
lower left and the treatment
basins are at the top. Since
only the water quality volume
enters the treatment cells, the
vegetation is protected from
excessive inundation depth,
duration and other negative
impacts of managing large
storms. The first treatment
basin contains deep and
shallow marsh sections.
Image: Austin adapted from
Virginia Department of
Environmental Quality (2013).
The sedimentation basin (holding 15 per cent of the water quality storm) serves the
treatment cells and an extended detention basin, which is sized to contain the 10- or
25-year storm as required by local ordinances. The extended detention basin water
level might fluctuate several feet, while the inundation level in the treatment cells is
limited to 12 inches above the normal water level. A reverse slope pipe from the
extended detention pond controls the water levels within the treatment cells (Figure
9.15b). A volume of water equal to one half of the water quality storm is permanently
retained in the extended detention basin and is correctly assumed to be part of the
water quality treatment system. Ideally the extended detention basin is well vegetated
with plants tolerant of deep water and deep inundation, such as cattail or common
reed. The limited inundation depths within the treatment cells better preserves
abundance and diversity of plants. The treatment wetland is designed to hold at least
one water quality storm volume below the normal water level and one above it during
storm events (Virginia Department of Environmental Quality, 2013). The treatment
cells can be graded to include benches with water depths consistent with deep, normal
and shallow marsh plant groups as well as ephemeral wet meadows or other zones,
such as habitat mud flats.
Horizontal flow stormwater treatment basin

The treatment landscape presented here combines features of the free water surface
and the horizontal subsurface flow wetland types. This design was created, constructed
and tested by the University of New Hampshire, Stormwater Center, USA. It treats
a 1-inch water quality storm (3,300 cubic feet) from a 1-acre parking lot (Roseen,
Ballestero and Houle, 2009).
Pretreatment
The water from the parking lot is discharged by a pipe into a sedimentation basin
that contains 10 per cent of the water quality storm (Figure 9.18). This basin could
be preceded or replaced by a bio-swale or other technique to remove particulates before
water enters the treatment cells. All berms and basin edges have maximum slopes of
3h:1v (33 per cent) for safety and ease of maintenance. The accumulated sediment
should be removed when 12 inches has been deposited (about every 4 years), and
vegetation should be harvested annually (Roseen, Ballestero and Houle, 2009).
Treatment basins
The treatment basins are rectangular and each is sized to contain 45 per cent of the
water quality storm (Figure 9.16). The basins were excavated and then backfilled with
filter media and a wetland soil layer. The minimum width of the basins is 15 feet (4.6
m) and the length to width ratio is 0.5 or greater (Figure 9.17). From top to bottom,
the basins include a 6-inch freeboard, temporary pool, 8-inch thick (20 cm) wetland
soil layer, 3-inch thick (8 cm) gravel transition layer and 24-inch deep (0.6 m) gravel
filter layer. The top and bottom surfaces of the basins are level (Figure 9.20) (Roseen,
Flow sequence
Water enters the sedimentation basin where it loses velocity before it overflows into
the first stage of the treatment wetland (Figure 9.17). The stormwater floods the surface
of the first basin. The sedimentation basin and temporary pools of the basins are aerobic
treatment zones. The temporary pool drains into the gravel filter on the upstream
end, through vertical perforated pipes (Figure 9.19). The central pipe is 12 inches in
Figure 9.16
This perspective of the
subsurface flow gravel wetland
shows two treatment cells in
series. A sedimentation basin
(Figure 9.17) discharge is on
the left, in red. Image by Austin
based on design by the
University of New Hampshire
Stormwater Center.
Figure 9.17
The three-basin subsurface
gravel wetland at the University
of New Hampshire. The
sedimentation basin is in the
background. The top of the
basin is much higher than the
intermediate berms due to its
location on a slope below the
parking lot. Note that the three
vertical pipes on the upstream
side of the treatment cells
distribute water to the
subsurface gravel filter.
Photo: Tom Ballestero, UNH
Stormwater Center: reproduced
with permission.
Figure 9.18
The sedimentation basin of the
subsurface gravel wetland at
the University of New
Hampshire supports dense
wetland vegetation 4 years
after planting. The inlet into the
first treatment cell is on the
right.
Photo: Tom Ballestero, UNH
Stormwater Center: reproduced
with permission.
diameter; all others are 6-inch diameter perforated PVC pipes (Roseen, Ballestero and
Houle, 2009).
These vertical drains are spaced 15 feet apart and connect to a horizontal 6-inch
diameter perforated PVC pipe that distributes the stormwater evenly across the width
of the wetland (Figure 9.19). Water flows horizontally (15 feet, 4.6 m, minimum) and
slowly through the gravel filter. This is the anoxic (oxygen limited) treatment zone.
At the downstream end of the gravel bed of the first treatment basin, the water is
collected by another horizontal 6-inch diameter perforated PVC pipe (at least 15 feet,
4.6 m, away). This is connected to a horizontal, solid, 6-inch PVC pipe that delivers
the water to the next treatment basin. The piping in the second basin is the same as
in the first stage, until the water drains into the outlet structure. Solid vertical PVC
clean-out pipes are connected to the ends of the horizontal collection pipes (Roseen,
Figure 9.19 This schematic perspective of the piping system for the gravel wetland shows the
sequence from the sedimentation basin on the left. The red colour indicates inlet and overflow pipes,
green indicates perforated pipes, yellow shows the solid clean-out and transfer pipes, and the violet is
the outlet structure. Notice the elevated yellow pipe exiting the outlet structure. It maintains the
permanent saturation of the gravel bed.
Image: Gary Austin.
The outlet elevation is set at 4 inches below the surface of the wetland soil layer
(Figure 9.20). This assures that a volume of water is held in the basin after the storm
ends, and provides anoxic or anaerobic conditions for the removal of nitrates
(denitrification) (Roseen, Ballestero and Houle, 2009).
When the inflow is high enough, the first temporary pool overflows into the second
treatment basin. Although the University of New Hampshire gravel wetland contains
only the water quality storm. The basin can be designed to detain and covey the storms
larger than the water quality storm by increasing the height of the berms and adding
weirs to the outlet structure as required by local stormwater ordinances (Roseen,
Wetland and subgrade soil

The native subsoil below the filter bed should have a hydraulic conductivity less than
0.03 feet per day (0.9 cm per day). If it does not, then a 6–12 inch (15–30 cm) layer
of clay soil, a layer of bentonite or a 30 ml high density polyethylene liner should be
used to prevent infiltration (Roseen, Ballestero and Houle, 2009).
Figure 9.20
This is a section through the
gravel wetland. From top to
bottom, the layers are: the
temporary pool, the 8-inch
deep wetland soil, the pea
gravel transition layer, the
gravel filter bed, the
impermeable liner and the
compacted native soil. The
water is maintained at the
centre of the wetland soil
layer.
Image: Gary Austin.
Materials
The berms of the sedimentation basin and the treatment basins should have a low
infiltration rate or be composed of clay or a fine mesh geotextile fabric, or a
combination of these (Roseen, Ballestero and Houle, 2009).
The wetland soil can be created from loam and 15 per cent mulch, such as shredded
wood, and less than 15 per cent clay and silt. The resulting soil should have a saturated
hydraulic conductivity of 0.1–0.01 feet per day (3.05–0.305 cm/day) (Roseen,
The transition gravel layer should be 1⁄4 inch in diameter (pea gravel) to prevent
migration of silt and clay into the gravel filter. A thin layer of coarse sand could be
added above this transition layer, if desired, to further guard against migration of clay
and silt into the filter media. Geotextile fabric should not be substituted for the gravel
transition layer. The gravel filter bed should be composed of graded and washed gravel
3
⁄4 inch (2 cm) in diameter.
Spillways and swales and pools receiving water from pipes should be covered in
gravel 11⁄2–3 inches (4–8 cm) in diameter, to dissipate energy and prevent erosion.
The plants should be native species. The sedimentation basin may hold enough
water to support emergent macrophytes, such as cattail (USA) (Figure 9.18) or
Figure 9.21 This image is of a subsurface gravel wetland at Greenland, New Hampshire, that is
based on the design developed at the University of New Hampshire Stormwater Center. The receiving
stream is degraded with high levels of TSS – 53 mg/L, total nitrogen – 1.35 mg/L and total
phosphorus – 0.145 mg/L. Water discharged from the gravel wetland has a much lower concentration
of pollutants than stream (TSS – 3 mg/L, total nitrogen – 0.50 mg/L, total phosphorus – 0.005 mg/L).
Photo: Robert Roseen, UNH Stormwater Center: reproduced with permission.
common reed (Europe and Asia). However, wet meadow plants are more likely
to thrive in the treatment basin (Roseen, Ballestero and Houle, 2009) as shown in
Figure 9.21. An application of the University of New Hampshire design to a shopping
centre project shows meadow vegetation in the treatment basin (Figure 9.21).
Outlet structure
The outlet structure can be designed in various ways to comply with local ordinances.
These might require that the 25-year, 24-hour storm be detained and released at pre-
development rates, for example. Similarly, the orifice that controls the release of the
water treated in the gravel bed should be sized to conform to local codes that often
require that the temporary pool be drained within 24–48 hours. This prepares the
basin to receive inflow from subsequent storms. The invert (outlet elevation) of the
outflow pipe should be set 4 inches (10 cm) below the top of the wetland soil to
maintain a saturated gravel bed and make water available between storms to the
wetland plants (Roseen, Ballestero and Houle, 2009). An adjustable riser on the outlet
pipe is recommended to allow adjustment of the water level within the wetland. This
is valuable during plant establishment and subsequent maintenance. A secondary
discharge drain, that is normally closed, is an optional but valuable addition to the
system. This allows rapid draining of the wetland and gravel bed for maintenance
and allows the system to be flushed at a higher water velocity.
Treatment performance
The University of New Hampshire subsurface gravel wetland has been monitored for
a number of years to assess its stormwater management, treatment performance,
maintenance requirements and longevity. It has proved to be highly effective and
reliable. Therefore, the design is being increasingly installed in both the private and
public sectors.
Stormwater management
The stormwater peak flow from the New Hampshire wetland was greatly reduced
and the time of the peak flow was greatly delayed. Both of these characteristics are
significant benefits to the receiving streams within urbanizing watersheds. The peak
flow was reduced by an annual average of 81 per cent (85 per cent in the winter
and 77 per cent in the summer). The time of the peak flow was delayed 315 minutes
on average.

Each treatment basin of the subsurface gravel wetland filters the entire water quality
storm volume for small frequent storms. The water treatment for water quality
improvement is effective because this design takes advantage of sedimentation and
sequestration, physical filtration, biological transformation, plant uptake and even
chemical process that bind pollutants to soil. Hydrocarbons, heavy metals and bacteria
in the polluted inflow are captured in the sedimentation basin and the wetland pools.
Organic material captured in these locations also decomposes to create ammonia and
Table 9.7 Horizontal subsurface flow gravel wetland pollution removal performance (from parking lot
catchment area), 6 years of data
Pollutant In flow Out flow Removal

Total petroleum hydrocarbon mg/L 644.0 6.44 99%
Total nitrogen mg/L 1.1 0.44 56%
Ammonia, nitrite and nitrate mg/L 0.3 0.075 75%
TSS mg/L 57 1.14 98%
Total zinc mg/L 0.04 0.007 83%
Total phosphorus mg/L 0.07 0.031 56%
Ortho phosphorus mg/L 0.02 0.005 75%
Robert Roseen; Thomas Ballestero; James Houle; University of New Hampshire Stormwater Center.
nitrate as this is an oxygen-rich environment. The removal of nitrates depends on the

oxygen-limited environment provided in the subsurface gravel filtration bed. This
combination of pollution reduction pathways results in outstanding water quality
(Table 9.7).
The average annual reduction in the concentration of nitrate in the subsurface
gravel wetland is more than 75 per cent (33 per cent in the winter and 85 per cent in
the summer) with an average concentration of less than 0.05 mg/L (Roseen et al.,
2012). This is a fraction of the 5 mg/L that can cause algae blooms. The average annual
total nitrogen removal is good at 58 per cent. The removal of TSS and petroleum
hydrocarbons is excellent. Phosphorus removal is moderate (56 per cent) and might
be improved through the use of gravel with higher calcium content. Nevertheless, the
0.031 mg/L concentration of total phosphorus in the outflow is below the 0.1 mg/L
target for protection of streams and lakes. The vegetation within the treatment basins
should be harvested at least every 3 years for best treatment performance (Roseen,
Retrofit
Conventional stormwater detention basins have little water quality benefit. However,
they can be modified to perform both their intended stormwater management functions
while incorporating all of the elements of the New Hampshire subsurface gravel
wetland described above for greatly improved treatment performance. Since the
subsurface gravel wetland is intended to have no bottom slope, and no separation
from ground water is required, the system is easily adapted to conventional detention
settings. The hydraulic pressure requirement for the gravel wetland is only 4 inches
(10 cm) (Roseen, Ballestero and Houle, 2009). Converting detention basins to
subsurface gravel wetland is a low cost manner for municipalities to upgrade their
infrastructure and meet more stringent water quality requirements and restore polluted
streams and lakes.
Conclusion
This installation of a subsurface gravel wetland focused on the technology, construc-
tion specifications and testing of the system performance for water management and
treatment only. However, as these systems are incorporated into the public landscape,
additional design considerations are necessary to improve their appearance and their
contribution for wildlife habitat. Municipal master planning to incorporate the
wetlands into a green infrastructure network, good planting design and integration
of user facilities will make the basins positive contributors to the open space networks
as well as essential tools for sustainable development.
Bio-retention basins for stormwater treatment

Stormwater treatment wetlands, such as the marsh landscape and the subsurface gravel
wetland, presented above, are suited to the scale of the neighbourhood or large parcel,
but smaller stormwater treatment landscapes are needed for the scale of residential
or commercial sites. In this setting, the emphasis on aesthetics often increases. The
bio-retention basin is a treatment landscape based on design and experience with
vertical subsurface flow constructed wetlands developed to treat domestic sewage.
However, the design is adapted to treat non-point source pollution with a fraction of
the organic nitrogen content, intermittent stormwater flow and with a different filter
media capable of supporting plants more common to the forest and ornamental
landscape. Rain gardens are similar to bio-retention basins. Rain gardens are simply
shallow depressions in the ornamental landscape intended to hold 6 inches or less of
stormwater from roof downspouts or other impervious surfaces. They are planted with
water-tolerant trees, shrubs and herbs. Rain gardens are valuable low impact
development measures that address stormwater volume, run-off rate and treatment.
Bio-retention basins are more robust versions of rain gardens, sharing the set of
appropriate plants and small-scale application. However, the bio-retention basin is
engineered to improve the quality of stormwater. Like the vertical subsurface flow
constructed wetlands, bio-retention beds are intended to dry (and renew their oxygen
content) between storms (Figure 9.22). Filtration, chemical and biological processes
all contribute to the removal of contaminants in the stormwater. This full set of
treatment processes makes the bio-retention basin much more effective in the removal
of contaminants than detention or retention (wet pond) stormwater basins. In fact,
typical stormwater basins have little, or even a negative, water quality improvement
benefit (Davies, 2001; Mallin, 2002; Hathaway, 2009).
Figure 9.22
Bio-retention basin at
Tabor School in
Portland, Oregon.
Design criteria
Government design standards for bio-retention basins are either unavailable or
outdated in most states. Initially, guidelines were established based on very little
research data and a limited set of goals. Monitoring of installed bio-retention basins
over the last 10 years provides more reliable criteria, installation and performance
expectations.
Pretreatment
When bio-retention basins fail to perform as intended, it is most often due to
construction errors and clogging of the filter media. Therefore, a sedimentation basin,
bioswale or tank is recommended to remove as much sediment and suspended organic
material as possible, before it can enter the treatment basin. When a sedimentation
basin is used, it should be sized to contain 25 per cent of the water quality storm
volume. A wide grass filter strip, upstream of the bio-retention basin, was initially
required by the regulations of some government agencies. This transition area was
intended for sheet flow of stormwater to capture sediment and was related to the single
circumstance of bio-retention basins along roads. This technique is effective, but it is
only one alternative for capturing sediment and TSS.
Treatment basin
The relationship between the drainage area and the area of the bio-retention basins
is typically 5–8 per cent of the catchment area (sometimes this is limited to the size
of the impervious area of the catchment), but the size varies with the stormwater
strength and water quality goals set for the basin. Suggested maximum widths for
bio-retention basins (25 feet) are based on the ability to excavate the basin with heavy
equipment located outside the basin, to avoid compaction of the bottom soil. However,
this does not apply to bio-retention basins, where infiltration of the treated stormwater
is not desired. Even where infiltration is a goal, the basin width could be expanded
with post-excavation measures such as ripping the soil, installing boreholes or infil-
tration trenches. A bucket with teeth should excavate the final 12 inches of a basin
intended for infiltration, in order to limit compaction of the basin bottom. Research
shows that infiltration into loamy sand subsoil is 2.6 inches per hour when the basin
is dug with a rake bucket, as compared to 1.2 inches per hour for a basin dug with
a flat-edged bucket. This translates into a basin drawdown of 12 hours and 27 hours
for the rake and flat edge bucket, respectively. Similarly, infiltration will be higher if
the basin is dry when dug (Brown, 2010).
The basin includes a portion above the soil surface to temporarily contain a pool
equal to 75 per cent of the water quality storm volume (Figure 9.23). The maximum
depth of ponding is a matter of some debate. Most states with regulations set the
maximum surface pond depth between 6 and 18 inches. If the filter media and subsoil
have a high infiltration rate or the basin has an under-drain, the deeper surface pond
is more acceptable. The depth of the surface pond should drop at about 1 inch per
hour due to infiltration or outflow to draw down the temporary pool in several hours
and the entire basin in 24–48 hours.
A 6- to 12-inch freeboard above the maximum water level is a necessary safety
factor. The surface of the basin should be covered with 3–4 inches of wood chips or
Figure 9.23 Bio-retention basin preceded by a vegetated swale to reduce sediment and heavy metals
and delay the storm peak. A – Earth berm with spillway or piped overflow; B – 6–18-inch ponding
depth; C – 4-inch wood chip mulch; D – Vegetated swale; E – Weir to retain water in swale.
other mulch. This appears to be particularly important if hydrocarbons are targeted

for removal, as bacteria in mulch rapidly decompose absorbed hydrocarbons, such
as toluene and naphthalene (Davis et al., 2009). The basin should be deep enough to
contain 24–48 inches of sandy filter media. Underneath the filter media, a 6–16 inch
depth of sand or gravel is sometimes specified to improve infiltration. However, this
is probably not beneficial for a correctly designed and constructed basin. Metals and
suspended solids are reduced significantly in the top 8 inches (20 cm) of the media,
which is where much of the pathogenic bacteria are also removed. However, removal
of hydrocarbons, total nitrogen and phosphorus seems to benefit by depths of at least
30 inches (Davis et al., 2009).
When the subsoil has an infiltration rate below 0.5 inches per hour, an under-
drain in a bed of coarse gravel (1–2-inch diameter) is recommended (Figure 9.24).
As a transition layer to prevent the migration of fine soil particles, the coarse gravel
should be separated from the filter media by a 2–4-inch gravel layer of 1⁄4-inch diameter
(pea gravel). Some installations even include a coarse sand transition between the filter
media and the pea gravel (University of New Hampshire, 2007).
Flow sequence
Stormwater from impervious surfaces or the ornamental landscape is directed through
bioswales, a sedimentation basin or other pretreatment before flooding the treat-
ment basin. The water drains vertically through the sandy filter media and into the
subsoil or into under-drains. Under-drains are used when the infiltration rate of the
native soil is less than 0.5 inches per hour or when the groundwater is seasonally within
3 feet of the bio-retention basin bottom. When there is limited infiltration, perforated
PVC under-drains in a coarse gravel drainage layer discharge the treated stormwater
to the surface waters (Figure 9.24).
Figure 9.24 Underground view of a bio-retention basin with a flooded sub-basin (internal water
storage). Total depth is about 4 feet. A – Sandy loam filter (80 per cent sand, less than 3 per cent silt
and clay); B – 2 inch depth of 1⁄4-inch diameter gravel transition; C – 3-inch depth of 3⁄4-inch diameter
gravel transition; D – 11⁄2-inch diameter gravel sub-basin; E – Perforated drain pipes connected to a
turned up discharge line.
When reduction of nitrate is an important goal, then a permanently saturated 24-

inch-deep layer of gravel is included below the main filter media. This creates an
anaerobic zone that encourages the growth of bacteria that use carbon instead of
oxygen as an energy source. In the process, nitrate is converted to harmless nitrogen
gas that escapes to the atmosphere. In Figure 9.24, the elevated discharge pipe, E, will
cause water to be retained in the gravel beds, C and D. Inflow from the subsequent
storm causes the retained water to be discharged. Therefore, a volume of water is
held, after the storm ends, for longer treatment. Within about 1 hour, oxygen in the
retained water will be depleted by microorganisms, creating anaerobic conditions that
are suitable for denitrification by bacteria. This is the final step in a complex sequence
including organic matter > ammonification > nitrite > nitrate > denitrification > nitrogen
gas.
Subgrade soil
Infiltration of water treated in a bio-retention basin may be desirable for groundwater
recharge, to maintain base flow in the soil or to reduce stormwater run-off volumes.
Where groundwater recharge is implemented, high water quality should be achieved
before infiltration. Nitrates are poorly removed from bio-retention basins constructed
without a water impoundment below the media. Nitrates are also not held in the soil
and, therefore, are likely to drain into groundwater especially if it is within a few feet
of the surface. Therefore, agricultural drainage areas, brownfields or current industrial
land are poor locations for bio-retention basins with infiltration due to elevated levels
of nitrates or toxic chemicals.
If infiltration is desired, then the character of the soil below the bio-retention basin
is important. Generally, a subsoil infiltration rate of 0.5 inches per hour is required
to drain the saturated basin. The saturated hydraulic conductivity of loam is 0.52
inches per hour, while for silty loam and sandy loam it is 0.27 inches per hour and
1.2 inches per hour, respectively. Hydraulic soil groups A and B, as defined by the
US Natural Resources Conservation Service, are most suitable, but silty loam soil
(type C) might be suitable under certain design conditions. A silty loam soil below 13
inches of ponded water in a bio-retention basin will be eliminated in 48 hours, while
in a loam soil this takes only 24 hours. A sandy loam soil will draw down 24 inches
of ponded water in 24 hours. It is important to reduce the standing water in the bio-
retention basin steadily so that capacity is available for storms occurring at short
intervals. However, an excessive infiltration rate of more than 3 inches per hour is not
desirable, as this reduces treatment time and also indicates a soil unsuitable for most
plants. If 0.5–1 inches of water from every storm is infiltrated to recharge groundwater,
this typically meets or exceeds preconstruction infiltration rates (Davis et al., 2009).
Materials
Filter media
Originally the bio-retention basin was planted with a mix of shrubs, ground cover
and trees to resemble a native forest. Concern that using coarse sand or gravel, as in
a vertical subsurface flow wetland, would be too infertile and dry too quickly, led to
initial specifications for loam soil. This caused the bio-retention basins to clog quickly,
as did the use of filter fabrics to separate the media layers in the basin (University of
New Hampshire, 2007). Most specifications today require 80–88 per cent sand for
the main filter layer. Small amounts of shredded bark, mulch and loam soil are generally
specified, but fines (silt and clay) are limited to 7 per cent. However, an effective bio-
retention basin with a sandy clay loam soil (54 per cent sand, 26 per cent silt, 20 per
cent clay) and 12.2 per cent of organic matter performed better in removal of most
pollutants than a basin with sandy loam soil (80 per cent sand, 13 per cent silt, 7 per
cent clay, 5.7 per cent organic matter). In this study, performance testing over 14
months began only 1 month after the basin was completed, so the long-term infiltration
rate is not known. Compost should be used with caution, as it is likely to increase
the amount of nitrogen and phosphorus in the effluent.
Where organic matter, ammonium or nitrates are in high concentrations, such as
in agricultural run-off, solid carbon (wood chips) in a horizontal subsurface flow bed
has proven to be very effective for removal of nitrates from agricultural run-off
(Schipper et al., 2010). A laboratory study demonstrated an 87 per cent reduction of
nitrate when carbon was added to the media of a saturated biofilter (Yang, 2010).
This is a topic requiring additional research.
Specifications should require that all gravel be triple washed and media with a
low phosphorus index be used. A bio-retention basin at Villanova University in Pennsyl-
vania has been in operation for 7 years with no reduction in the infiltration rate. Its
filter media is composed of 50 per cent sand and 50 per cent existing (well drained)
site soil (Council, 2009).
Plants
The high sand content of the filter media necessary to maintain porosity of the filter
bed creates a fairly difficult growing environment for plants. However, there are
ecosystems where sandy soils and high rainfall have resulted in plants’ adaptation to
alternating periods of excess water followed by drought conditions (Figure 9.25).
Specifying these plants for bio-infiltration basins that are native to the region will lead
to sustained porosity of the filter media, better growth, lower maintenance and better
habitat for native species of insects, birds and mammals. In a Seattle, Washington (a
climate with a long, dry period in the summer) study, bio-retention basins with longer
drawdown times (24–72 hours) supported a different set of plants than basins that
drew down in 24 hours. In the basins with more moisture (and longer inundation),
the three most successful plants were Scirpus atrocinctus, Scirpus microcarpus, and
Scirpus acutus, while in the drier basin Carex stipata, Carex pachystachya, Juncus
tenui, Rubus spectabilis and Rosa pisocarpa grew well. Juncus effusus grew well under
both moisture conditions (Reiners, 2008). See the Appendices for regional lists of plants
suitable for bio-retention basins and rain gardens.
Treatment performance
Stormwater management
Assessment of several bio-retention basins confirms that they are effective in reducing
run-off volume, run-off rate and delaying the time of peak run-off. The Hal Marshall
bio-retention basin (Figure 9.26) in Charlotte, North Carolina, was designed to
capture storms of 1 inch or less. However, for rainfall depths of 1.65 inches (42 mm),
the peak storm outflow decreased by 96 per cent even though the entire catchment
Figure 9.25 Plants can be treated formally, as in this Portland, Oregon, bio-infiltration basin, or
more naturally to respond to the project context. This basin receives run-off from the street and
sidewalk areas.
Figure 9.26
Hal Marshall bio-retention
basin with spring growth of
vegetation.
Photo: North Carolina State
University, Biological and
Agricultural Engineering,
William Hunt.
area was impervious. During the monitoring period, the mean storm was 1.08 inches
(Hunt, 2008). This performance is confirmed by a study of a bio-retention basin
constructed at Villanova University in Pennsylvania. Its catchment area is 50,000 square
feet and is 52 per cent impervious. The 4-feet-deep infiltration basin consistently
removes 50–60 per cent of the storm run-off from the surface waters. In fact, for storms
1.95 inches and less, there is rarely any outflow from the basin at all. This is partly
because there is infiltration into the subsoil during the entire storm. Even during a
6-inch storm, the retention basin reduced the storm peak run-off rate (Council, 2009).
A University of New Hampshire bio-retention basin comprised of a 30-inch top layer
of sandy media, and a 16-inch bottom layer of gravel, displayed an 82 per cent
reduction in peak stormwater flow and a 92-minute delay in the storm peak (University
of New Hampshire, 2007).

The example below illustrates the water quality improvement data (Table 9.8) for the
bio-retention basin at the Hal Marshall Municipal Services Building (Figure 9.26). It
is designed to treat 1 inch (25.4 mm) of rainfall (the 2-year, 24-hour storm is 3.36
inches). The basin receives water from a 0.92-acre (0.37 hectare) parking lot. The
surface of the infiltration bed is 2,480 square feet (229 m2), which represents 6.2 per
cent of the catchment area. The bed is composed of 4-feet deep (1.2 m) loamy sand
(silt/clay = 5.7 per cent) with a 6-inch diameter corrugated under-drain. The subsoil
permeability is 0.43 inch per hour, and the basin is planted with a variety of water
tolerant species (Hunt, 2008).
The bio-retention bed reduces contaminants significantly with one exception
(Table 9.8). The low total nitrogen removal was due to low organic matter in the
run-off. The increase in nitrite and nitrate to 0.43 mg/L indicates that the bed provides
aerobic conditions for the conversion of ammonium to nitrite and nitrate. However,
the removal of nitrate requires an oxygen-depleted environment, which is not a
feature of this design. A very similar bio-retention bed was installed in Greensborough,
North Carolina, but the bottom 2 feet was saturated with water to form an anaerobic
zone. This bed performed better with nitrate removal at 75 per cent and total nitrogen
removal of 40 per cent.
The Charlotte bio-retention basin significantly reduced faecal coliform bacteria

(89 per cent) and E. coli (92 per cent). The bio-retention basin outflow met EPA
recommendations for primary recreation contact for E. coli and almost met the
standard for faecal coliform bacteria concentration (Hathaway and Hunt, 2012).
Drying of the soil between storms reduces pathogenic bacteria. As drainage through
the soil media is rapid, even the temporary pool above the soils drops about 1 inch
per hour after a storm. High oxygen levels return to the soil volume quickly, and drying
is due, in part, to evapotranspiration of bio-retention beds (Hunt, 2008).
A University of New Hampshire bio-retention basin (similar in design to the one
in Figure 9.23) featured a 99 per cent removal effectiveness for zinc, which is better
performance than the basin data shown in Table 9.8. The New Hampshire study also
reported total hydrocarbon removal of nearly 60 per cent for a 30-inch-deep filter
bed and 99 per cent for a 48-inch-deep bed (University of New Hampshire, 2007).
Another study of PAH removal in a bio-retention basin reported an annual average
PAH mass load reduction of 87 per cent (DiBlasi et al., 2009).
A laboratory experiment that tested 125 filter media configurations, varied the
plant, filter media, media depth and pollutant concentrations typical of urban storm-
water. The significant difference in the best performing configuration was the presence
of Carex appressa, which is characterized by deep and fine roots. Under various media
and flow conditions, the best filter removed 99 per cent of TSS, 93 per cent of
ammonium, 85–96 per cent of nitrite and nitrate, 71–79 per cent of total nitrogen,
93–96 per cent of total phosphorus and 87–98 per cent of particulate phosphorus.
These are significant improvements in performance compared to the Charlotte bio-
retention basin. The Australian study indicates that using sandy loam filter media,
specifying plants that are known to remove ammonia and nitrate at accelerated rates
(some plants actually increased total nitrogen in the effluent), and avoiding compost
or mulch in the media mix could lead to significant improvements in water quality
in full-scale field applications (Bratieres et al., 2008). Phosphorus removal in bio-
retention basins is generally poor, ranging from 5 to 30 per cent. Mulch in the filter
media and sands or gravel high in phosphorus cause low removal rates, or sometimes
an increase in phosphorus in the outflow. Sedimentation and adsorption are
the primary removal mechanisms for phosphorus. Using media with high levels
of calcium, aluminium or magnesium results in good removal of phosphorus.
Eventually the adsorption sites in the filter media will be filled, and removal rates will
drop. Aluminium-based drinking water treatment residuals (a waste product) as an
Table 9.8 Hal Marshall bio-retention basin (Charlotte, NC) pollution removal performance (Hunt, 2008)
(from parking lot catchment area)
Pollutant Inflow Outflow Removal Pollutant Inflow Outflow Removal

(%) (%)
Total nitrogen 1.68 1.14 32 Faecal coliform 14,700 4,500 69
TKN 1.26 0.70 44 E. coli 938 273 71
Ammonium 0.34 0.10 73 Zinc 72 17 77
Nitrite and nitrate 0.41 0.43 –5 Copper 12.8 5.9 54
TSS 49.5 20.0 60 Lead 4.85 3.33 31
BOD 8.54 4.18 63 Iron 1,110 4,710 330
Total phosphorus 0.19 0.13 31
(Hunt, 2007. City of Charlotte, Pilot BMP Monitoring Program, Hal Marshall Bio-retention)
Units = mg/L, except for coliform and E. coli, which are CFU/100 ml.
amendment to the bio-retention filter media improve adsorption of phosphorus in

laboratory and field studies (Stone, 2013). A laboratory scale study demonstrated
84–100 per cent removal of the common pesticide, atrazine, in bio-retention systems
with and without an anaerobic stage (Yang, 2010).
Hybrid bio-retention basins

When advanced water quality is the project goal, a sequence of treatment techniques
is likely to provide better results than a single type. The alternation of oxygen-rich
and oxygen-depleted environments often treats the widest range of contaminants. Ohio
State University built and tested a three-stage bio-retention basin that demonstrates
outstanding water quality improvement for both urban and agricultural contaminants.
The system is designed to contain 1.76 inches (44.7 mm) of run-off from a 748-square-
feet (69.5 m2) paved surface, resulting in 3.1 m3 of run-off. This is equivalent to the
10-year, 1-hour storm (Yang et al., 2013).
The system consists of three cells operated in series. The first cell is a 1.2-m-deep
vertical subsurface flow basin that is always saturated with water. The second is
a vertical flow, free draining cell and the third is a gravel infiltration bed. The vertical
flow cells have an 8-inch (0.2 m) temporary pool, are planted and have a mulch layer.
The first cell (saturated) contains 2.8-feet-deep (0.85 m) filter media (90.6 per cent
sand, 6.9 per cent silt, and 2.5 per cent clay with 0.7 per cent organic matter, 3.1
meq/100 g of cation exchange capacity, 12.0 cm/h of saturated conductivity). On top
of this filter is a 0.5-foot-deep (0.15 m) layer of fine gravel (3.2–12.7 mm diameter).
With an area of 73 square feet (6.8 m2), the saturated cell can contain 56 cubic feet
(1.58 m3) of water (1-inch rainfall from the paved surface). The area of the saturated
basin is 9.6 per cent of the catchment. The plants in this basin are boneset (Eupatorium
perfoliatum), spiderwort (Tradescantia ohiensis) and culver’s root (Veronicastrum
virginicum) (Yang et al., 2013).
The second cell can contain 1.34 m3 of water. The basin is 20 inches (0.5 m)
deep and composed of filter media (0.35 m deep) above a 0.15 m depth of fine gravel.
The plants in this basin are purple love grass (Eragrostis spectabilis), Indian grass
(Sorghastrum nutans) and purple coneflower (Echinacea purpurea) (Yang et al.,
2013).
Similar to Hal Marshall basin, the hybrid basin has good hydraulic performance.
It reduces peak flow 84 per cent and 88 per cent for medium rainstorms (6–12 mm/24
hours) and heavy rainstorms (more than 12 mm/24 hours). It reduces 59 per cent and
54 per cent of storm run-off volume, respectively, for medium and heavy rainstorms,
and delays the time of peak flow 180 minutes and 80 minutes, respectively, for medium
and high intensity storms (Yang et al., 2013).
Table 9.9 Net mass removal (%) of pollutants from sequential urban run-off storms (Yang et al., 2013)
Pollutant Average influent (mg/L) Maximum effluent (mg/L) Net mass removal (%)
Nitrate (NO3-N) 20 6.9 78–91
Phosphate (PO4-P) 10 2.5 94–99
Dicamba (C8H6Cl2O3) 0.09 0.05 84–92
Glyphosate (C3H8NO5P) 2.0 0.08 98–99
2,4-D(C8H6Cl2O3) 0.3 0.25 81–90
When tested with very high concentrations of common urban pollutants under
simulated rainfall, the hybrid bio-retention basin removed as much as 91 per cent of
the nitrate, 99 per cent of the phosphorus and 90–99 per cent of three pesticides
common in urban run-off (Table 9.9). The tested pesticides were glyphosate
(N-(phosphonomethyl)glycine), dicamba (3,6-dichloro-2-methoxybenzoic acid) and
2,4-D(dichlorophenoxyacetic acid) (Yang et al., 2013).
Base flow and ground water

The traditional goals of flood and channel protection are increasingly expanded with
requirements for water quality improvement, and most recently with stipulations that
post-development infiltration of rainwater shall approach the pre-development
infiltration volume. This third phase of stormwater regulations is intended to preserve
the base flow of local streams and rivers. This is important to preserve minimum water
levels required and ecosystem health. In addition, bio-retention with infiltration can
be used to correct historic damage (channelization) to streams by reducing storm flows
and delaying storm peak flows. Base flow is the discharge of water into receiving
streams by water moving through the soil volume. Declining base flow volumes are
known to reflect the increase of impervious surfaces within the watershed. Sufficient
base flow is critical to the ecology of aquatic environments during the dry seasons.
The volume, water quality and temperature of the base flow are factors important to
the health of streams.
Regulations generally require that a portion of the pre-development infiltration
volume is infiltrated after development. For example, the Wisconsin Department of
Natural Resources requires that 25 per cent or 10 per cent of the 2-year, 24-hour,
pre-development storm is infiltrated by residential and non-residential projects,
respectively. Alternatively, the required infiltration volume may be calculated as
60 per cent (non-residential) or 90 per cent (residential) of the average annual pre-
development infiltration volume. The calculations exclude undisturbed natural areas.
Figure 9.27
The Street Edge Alternative
project in Seattle demonstrated
the effectiveness of bio-
retention for the dual purpose
of mitigating high stormwater
volumes that had damaged a
stream and local flooding of the
basements of homes in a
residential district. The project
reduced annual run-off from
the neighbourhood by 99 per
cent and eliminated basement
flooding. Although some
streets in the neighbourhood
are fairly level, the linear
infiltration basin shown here
demonstrated techniques for
steeper terrain (about 6 per
cent).
Table 9.10 Reduced curve number and depth of rainwater to infiltrate
Hydrologic soil group B

%I RCN PE =1″ 1.2 ″ 1.4 ″ 1.6 ″ 1.8 ″ 2.0 ″ 2.2 ″ 2.4 ″ 2.6 ″
0 61
1 63
10 65
15 67 55
20 68 60 55 55
25 70 64 61 58
30 72 65 62 59 55
35 74 66 63 60 56
40 75 66 63 60 56
45 78 68 66 62 58
50 80 70 67 64 60
55 81 71 68 65 61 55
60 83 73 70 67 63 58
65 85 75 72 69 65 60 55
70 87 77 74 71 67 62 57
75 89 79 76 73 69 65 59
80 91 81 78 75 71 66 61
85 92 82 79 76 72 67 62 55
90 94 84 81 78 74 70 65 59 55
95 96 87 84 81 77 73 69 63 57
100 98 89 86 83 80 76 72 66 59 55
I= Imperviousness.
RCN = Runoff curve number.
RE= Rainfall target.
Exemptions to the rule include areas with existing toxic soil and high seasonal ground
water (Bannerman and Barras, 2004).
Other agencies have similar methods of determining the volume of rainwater that
must be infiltrated. For example, Prince George’s County, Maryland, determines the
required infiltration volume in a more site-specific manner. Table 9.10 shows the values
for areas with hydrologic soil type B. The table is used to determine the rainfall depth
from impervious areas that are to be infiltrated after development, to mimic infiltration
for native forest in good condition. One enters the chart in the left-most column (per
cent imperviousness of the proposed development). From the per cent impervious
cell, move right until the first green cell is encountered, then read the rainfall depth
in inches at the top of the column. For example, 50 per cent imperviousness yields
1.8 inches. This is adjusted by adding a coefficient. The coefficient is calculated with
the formula: 0.05 + (0.009 × per cent impervious cover). For our example with
50 per cent imperviousness, this is 0.05 + (0.009 × 50) = 0.05 + (0.45) = 0.5. This
coefficient is multiplied by the rainfall depth 1.8 inches × 0.5 = 0.9 inches. Therefore,
the infiltration that would be required to comply with the regulation is 0.9 inch of
rainfall depth from the impervious area. If the impervious area of the proposed
development was 10,000 square feet, then 750-cubic-feet volume of water would
require infiltration, 10,000 × (0.9 inches/12 inches) = 750 cubic feet (Maryland
Department of the Environment, 2009).
Other methods of determining the volume of water use the difference between
the TR-55 curve number for pre- and post-development. For example, a wooded site
slated for low-density residential development might typically have a pre-development
Figure 9.28
Stormwater collection and infiltration can be
accommodated in the urban landscape.
Landscape architects are beginning to explore
the multifunctional and aesthetic potential of
stormwater landscapes. This courtyard is one of
several that are designed to receive stormwater
in a park in Tianjin, China. These are beautiful as
well as intriguing places for children and adults.
Figure 9.29
This Oregon City bio-retention
basin receives stormwater run-
off from a highway. Note the
steel grate over the channel
that delivers water from the
gutter to the basin (lower left).
The steel panels in the basin
are sculptural elements that
contrast the plants, which
range from drought tolerant
upland species to water
tolerant plants in the bottom
of the basin. The design is
by Greenworks.
curve number of 55. If after development, the wooded site had a curve number of 70,
then 15 per cent more water is expected to run-off. An infiltration of 0.22 inches of
run-off volume from the developed site would be required to match pre-development
conditions for a storm with a 2-inch rainfall depth (Department of Environmental
Resources, 2007). For a 1⁄2 acre site, this would equal approximately 400 cubic feet
for a 2-inch storm, and could be accommodated in a 20-feet by 20-feet bio-retention
basin.
Conclusion
Bio-retention basins are effective landscapes for the treatment of pollutants,
stormwater management and groundwater recharge. The basin design must be adapted
to site conditions and the primary implementation goal, but are more effective than
stormwater detention and retention basins (Figure 9.29). Lower than desired removal
of nitrates, phosphorus and chloride requires design modification or additional
treatment methods. Use of a saturated volume within the bio-retention basin improves
nitrate removal. Even higher performance might be achieved (as it is in the treatment
of wastewater) if a sequence of treatment stages is implemented (Guenter Langergraber
and Roland Rohrhofer, 2009). This might be configured as a lined aerobic bio-retention
basin followed by a lined anaerobic horizontal subsurface wetland and then an
infiltration bed (Austin, 2010).
Qunli stormwater wetland case study

Case studies, including this one of the National Urban Wetland Park in Qunli, expose
the challenges and realities posed by context, diverse goals, budget, hydrology, soils,
maintenance, construction quality, etc. This masterwork combines stormwater
technology with innovative landscape architectural design, to demonstrate that
multifunctional landscapes can simultaneously be beautiful social places (Figure 9.37),
engineering infrastructure and regenerative ecosystems.
Context
Qunli is a new town district located east of Harbin in northeast China (45°43′46.77″
N 126°33′45.29″ E). Figure 9.30 shows the Qunli area of Harbin in relationship to
the Songhua River and the project site. At the centre of the developing urban district,
the project site was an 84-acre (34 hectares) degraded wetland. The hydrological
isolation created by the new development and declining ground water levels was causing
the wetland to decline. The climate of Harbin is relative dry. The city receives as much
Figure 9.30
Aerial view of the context
for the stormwater park.
The Songhua River is north
of the park site (red
asterisk).
Photo: Google Earth and
CNES Astrium, 45°43′46.77″
N 126°33′45.29″ E, image
date: 15 September 2014,
image accessed: 2 April
2015.
Figure 9.31
Aerial view of site. Longest
dimension about 2,000 feet
(650 m).
Photo: Google Earth and CNES
Astrium, 45°43′32.81″ N
126°32′53.38″ E, image date:
15 September 2014, image
accessed: 2 April 2015.
Figure 9.32 The land-use plan of the Qunli New Town shows the distribution of high-density
residential (yellow), commercial mixed-use (red), residential mixed-use (orange) and open space
(green).
as 70 per cent of its 22 inches (567 mm) of annual precipitation from June to the end
of August, although snow is a significant complication in this temperate climate.
The aerial image of the site (Figure 9.31) and the land-use plan (Figure 9.32)
graphically illustrate the high population density of both workers and residents
around the parcel. The Qunli New Town is an example of an emerging district (6,753
acres, 2,733 hectares) that will be the home and workplace for a population of
approximately 350,000 people. However, the product of high-density urban districts
is the creation of a highly impervious urban environment. At Qunli, nearly 344 million
square feet (32 million m2) of building floor area and impervious infrastructure will
consume all but 16.4 per cent of the permeable landscape (Saunders and Yu, 2012).
Design
The parcel had been originally set aside as future parkland. The simultaneous need
for public open space and the need to preserve and restore the existing wetland was
resolved by the designer, Kongjian Yu. His plan invites the public to participate along
the perimeter, while preserving most of the ecosystem at the centre of the parcel. The
designer established a pattern of birch forest wedges (Figure 9.33) separated by shrub
masses and stormwater ponds and wetland vegetation (Figure 9.34).
Figure 9.33 The plan view of the design for the stormwater park shows the mound (green) and pond
perimeter landscape connected by paths and elevated boardwalks (red line).
Figure 9.34
Water basins in series collect water
from the urban catchment. The transfer
from the sedimentation basins (brown)
to second stage (light blue) and third
stage (dark blue) remove progressively
more sediment and pollutants before
the water is discharged into the
recovering wetland. The light green
wedges represent the upland mounds
planted with tree groves.
Figure 9.35
Elevated walkways and
pavilions create opportunities
for dramatic views and places
where groups can gather in the
sun or shade. The design
includes four pavilions with
unique designs and materials.
Photo: Kongjian Yu.
Figure 9.36
One of the two viewing towers
in the Qunli wetland park differ
greatly in character, but are
striking and provide wonderful
views of the wetland and
human activities.
A network of paths and elevated boardwalks (about 2,500 feet, 762 m) punctuated
by four pavilions (Figure 9.35) and two viewing towers (Figure 9.36) connects the
stormwater and upland landscapes. The plan view of the project (Figure 9.33) shows
an active and diverse perimeter dedicated to human uses, while the interior is dedicated
to ecosystem functions without the disruption of paths and other human activity or
facilities (with the exception of an electrical transmission line).
Stormwater management and treatment

The increase in stormwater run-off volume and rate caused by urban development
could pose significant flooding hazards to the new residents and those downstream,
as saturated soils and flooding were already a historic problem in the flat plain of the
Songhua River where Qunli New Town is being constructed. Furthermore, the infil-
tration of rainwater is required to restore a depleted groundwater volume.
The project is designed to store and treat a water quality year storm in addition
to the normal volume of water in the wetland. Larger storm flows bypass the wetland
through a conventional storm sewer (Figure 9.43) leading to the river. The catchment
area for the wetland is about 111 acres (45 hectares) where the stormwater is collected
in a conventional system of catch basins and stormwater pipes. These distribute the
Figure 9.37
This image shows the
sedimentation basins at work
as well as the mounds forested
with birch trees (Betula
pendula).
Figure 9.38
Birch groves (Betula pendula)
on hillocks are a counterpoint
to the sedimentation and
treatment ponds that prepare
the stormwater for entrance
into the wetland. The
stormwater collection areas
vary in size, depth and
vegetation.
Figure 9.39 Frequency and depth of inundation determine the range of species around or within each
treatment basin and in the core wetland.
Figure 9.40
The stormwater flows
from basin to basin via an
underground pipe in order to
maintain a low peak flow rate.
However, emergency spillways
are provided between the
sedimentation and treatment
basins, and require careful site
grading and sizing of culverts.
stormwater evenly to more than 50 sedimentation ponds along the perimeter of the
park. Water from the sedimentation basins overflow into more than 80 secondary and
tertiary treatment basins. These in turn overflow into the 50-acre (20.2 ha) core wetland.
Figure 9.38 illustrates a second tier sedimentation and treatment pool and a birch grove,
formed through a balanced cut and fill earthwork to create a pool and mound
landscape along the perimeter of the park. The wetland park has proven to be very
effective for both flood management and water quality improvement. It is estimated
that under normal conditions, the wetland contains 58 acre feet (72,000 m3) of water,
while at maximum storage it holds more than 111 acre feet (138,000 m3) of
water (Yan, 2013).
Water flows via gravity from the sedimentation basins to the treatment basins
(Figure 9.40) and to the core wetland. Low velocity inflow and sufficiently sized
sedimentation basins allow particles to settle. A rule of thumb is to accommodate
10–25 per cent of the water quality storm in the sedimentation basins.
Controlling erosion on-site is a critical issue, as the removal of sediment before
it reaches the core wetland and even the sedimentation basins, if possible, will prolong
the life of both. Figure 9.41 shows that the water from slopes is intercepted in slotted
drains covered with coarse gravel. Using erosion control blankets or mulch on slopes
is particularly important until the vegetation on the slope can completely cover the
soil. Similarly, the velocity of water entering the sedimentation basins must be
moderated to allow settling of the heavier sediments before the water overflows into
the next treatment basin.
Figure 9.42 shows ponds after a storm, and water further away that receives the
flow after much of the sediment has been deposited. The sediment in the basins can
Figure 9.41
Sediment and velocity control
on-site.
be periodically removed to avoid filling the wetland and shortening its life. Cleaner
water flows from the sedimentation pools into the wetland, restoring the natural
hydrology and native vegetation (Yan, 2013). However, since the groundwater level
has dropped, restoration of the natural wetland requires supplemental water during
the dry season. This is provided by water pumped from the river (Figure 9.43). This
practice might be suspended if projects similar to this one recharge the aquifer.
The visual character of the wetland changes dramatically with the season.
Specifying plants that bloom (Figures 9.44 and 9.39) in different seasons maintain
colour and detail in the park for many months (Figure 9.46). The enormous increase
in biomass during the growing season conceals the water with tall green growth. In
autumn, the wetland landscape is golden and, in winter, it is ice and snow covered
(Figure 9.45).
The filling and draining of the sedimentation and treatment basins is another
temporal aspect of the design. This makes stormwater (and the monsoon rains) visible
to the user and creates an ever-changing scene. As water in the basins infiltrate or
evaporate, capacity for the subsequent storm is created. The sedimentation and
treatment pools on the perimeter of the wetland capture sediment and the attached
heavy metal and bacteria.
Figure 9.42
Sedimentation basins connect
to treatment pools and then to
the core wetland.
Figure 9.43
Water from the nearby river
supplements the wetland
during the dry season. In the
wet season, this structure
conducts overflow to the river.
Figure 9.44
Siberian iris blooming in May.
Figure 9.45
Texture and colour of the
wetland vegetation provides
terrific aesthetics even after the
first frost of autumn.
Figure 9.46
The Qunli wetland and park
offer great seasonal variety.
Ecosystem services
The designer adopted a strong ecosystem service perspective as the foundation for the
application of design and technology. The wetland park provides ecosystem benefits
to humans. The stormwater and wetland restoration goals are met through the use
of regenerated hydrology and the planting of native plants. The vegetation extends
the efforts of the landscape architect by organizing itself according to tolerance to
saturated soils. This is key to the low maintenance of the park. Critical ecosystem
services are the stormwater storage (flood protection), cleansing of non-point source
pollution and infiltration of the treated water through a recovered wetland to augment
the ground water supply (Figure 9.47). Three methods of providing water to restore
the damaged wetland were investigated. Pumping groundwater into the wetland is
problematic due to the cost, poor water quality and the fact that groundwater
depletion is one of the reasons for the decline of the natural wetland. Pumping water
from the Songhua River also comes with an energy cost. The use of urban stormwater
is the least expensive option, and the quality of the water is intermediate between
water from the river and groundwater (Yan, 2013).
Related human benefits are the stimulation of high-value urban development and
the provision of open space, recreation and social opportunities. These are in great
demand due to the very high population density. Therefore, the design creates a range
of opportunities and experiences for individuals and groups. Although city views are
highlighted at times, the wetland park is a sanctuary for people and a dramatic contrast
to the street, plaza and ornamental landscape environment that surrounds the park.
Literally steps away from towering residential buildings, a family can find a cosy deck
enclosed by native trees (Figure 9.48).
The several pavilions are beautifully and creatively designed with local materials.
The contrast of the cordwood, stone (Figures 9.49, 9.50 and 9.51), metal, brick and
bamboo material makes each pavilion a destination and unique experience. Each
structure is a unique setting that features a shady space and a commanding view. The
two towers provide the most dramatic viewing opportunities (Figures 9.36 and 9.53)
(and good exercise).
Figure 9.47
The restored hydrology
supports the low maintenance
cost of managing the
vegetation and provides
seasonal diversity.

There are many spaces for small groups. This pavilion provides shade to the user with a
novel use of native stone contained in a wire
frame. The extension into the landscape provides
wildlife viewing and views of the surrounding city
buildings.
The landscape and habitat diversity is even greater than the variety of human
settings. Gradients between upland and wetland vegetation, shallow and deep water
create niches for many species. Limiting human activity to the perimeter of the
wetland provides wildlife with a refuge from disturbance that so often limits the variety
of species that might share the urban landscape with people. As the native plant species
respond to the restored hydrology and cleaner water, and grow to provide a vertical
as well as horizontal distribution of niches, the number and variety of species present
will grow. The great production of plant biomass is responsible, in part, for the energy
available for long food chains that begin with herbivores (Figure 9.52).
Figure 9.50
A pavilion made of bamboo
screen.
Figure 9.51 Winter scene of the pavilion made from cord wood.
Figure 9.52
The high production of biomass by the wetland
plants transfers energy to other tropic layers.
This creates a highly diverse ecosystem of
plants, herbivores, predators and even
decomposers.
Conclusion
The urban design team, led by Turenscape, devised a sustainable plan that provides
ecosystem services in every category through careful planning and design. The project
goals include aesthetic, social, recreational, engineering and ecological dimensions,
which the designer conceptualizes as ecosystem services, following the 2005
Millennium Ecosystem Assessment framework of regulating, supporting, provisioning
and cultural categories. The planning and design masterfully create values in each of
these ecosystem service categories. In the regulating category, we find carbon
sequestration in the biomass of the wetland vegetation and tree groves as well as in
the reduction in the urban heat island temperature. In the supporting category, we
find the development of soil and recycling of nutrients and water. In the provisioning
category of ecosystem services, we find water cleaned and returned to the groundwater
or river. The many benefits in the cultural category include economic development,
recreation, open space and aesthetic values (Figure 9.53).
We can correctly describe this as a regenerative project, leading to better urban
sustainability. The multifunctional mission of this project increases the complexity of
planning, programming, design and implementation, but not necessarily increased
complexity of materials or forms. Public open space, recreation and social space
are needed to mitigate population density and the absence of private open space. The
aspects of the design that restore or preserve ecosystem health mitigate the widespread
and sometimes devastating impacts of rapid urbanization on adjacent natural and
rural areas.
Figure 9.53 This image shows a variety of user amenities as well as a dramatic seasonal display of
perennial flowers.
References
Austin, G. “Constructed wetlands for wastewater treatment and as landscape amenities in rural
communities,” Rural Connections, vol. 4, no. 2, pp. 37–40, May 2010.
Bannerman R. and Barras, J. “Infiltration: Fitting into a stormwater management plan.”
Wisconsin Department of Natural Resources, 2004.
Bratieres, K., Fletcher, T. D., Deletic, A. and Zinger, Y. “Nutrient and sediment removal by
stormwater biofilters: A large-scale design optimisation study,” Water Resource, vol. 42,
no. 14, pp. 3930–3940, 2008.
Brown, R. “Impacts of construction activity on bioretention performance,” Joural of Hydrologic
Engineering, vol. 15, pp. 386–394, 2010.
Council, N. R. Urban stormwater management in the United States. Washington DC: National
Academies Press, 2009.
Davies, B. “The fate of stormwater-associated bacteria in constructed wetland and water
pollution control pond systems,” Journal of Applied Microbiology, vol. 89, pp. 349–360,
2001.
Davis, A., Hunt, W., Traver, R. and Clar, M. “Bioretention technology: Overview of current
practice and future needs,” Journal of Environmental Engineering, vol. 135, no. March,
pp. 109–117, 2009.
Department of Environmental Resources, “Bioretention manual,” Prince George’s County,
Maryland, Environmental Services Division, 2007.
DiBlasi, C. J., Li, H., Davis, A. P. and Ghosh, U. “Removal and fate of polycyclic aromatic
hydrocarbon pollutants in an urban stormwater bioretention facility,” Environmental
Science and Technology, vol. 43, no. 2, pp. 494–502, January 2009.
Guenter Langergraber K. L. and Roland Rohrhofer, R. H. “High-rate nitrogen removal in a
two-stage subsurface vertical flow constructed wetland,” Desalination, vol. 246, pp. 55–68,
2009.
Hathaway, J. “Indicator bacteria removal in storm-water best management practices in
Charlotte, North Carolina,” Journal of Environmental Engineering, vol. 135, pp. 1275–
1285, December 2009.
Hathaway J. M. and Hunt, W. F. “An evaluation of the dye branch wetlands,” North Carolina
State University, Biological and Agricultural Engineering, 2009. [Online]. Available:
www.bae.ncsu.edu/stormwater/pubs.html. [Accessed: 12 July 2016].
Hathaway J. M. and Hunt, W. F. “Indicator bacteria sequestration in stormwater wetlands,”
Orlando, FL: International Association for Ecology, 2012. [Online]. Available: www.
conference.ifas.ufl.edu/intecol/presentations/120/1120%20J%20Hathaway.pdf. [Accessed:
4 July 2016].
Hunt, W. “Pollutant and peak flow mitigation by a bio retention cell in Urban Charlotte, NC,”
Journal of Environmental Engineering, vol. 134, pp. 403–408, May 2008.
Hunt, W. F. “Pollutant removal evaluation and hydraulic characterization for bioretention
stormwater treatment devices,” North Carolina State University, Raleigh, NC, 2003.
Kayhanian, M., Fruchtman, B. D., Gulliver, J. S., Montanaro, C., Ranieri, E. and Wuertz, S.
“Review of highway run-off characteristics: Comparative analysis and universal
implications,” Water Resource, vol. 46, no. 20, pp. 6609–6624, December 2012.
Li, H. and Davis, A. P. “Water quality improvement through reductions of pollutant loads
using bioretention,” Journal of Environmental Engineering, vol. 135, pp. 567–576, August
2009.
Li, M.-H., Barrett, M. E., Rammohan, P., Olivera, F. and Landphair, H. C. “Documenting
stormwater quality on Texas highways and adjacent vegetated roadsides,” Journal of
Environmental Engineering, vol. 134, no. 1, pp. 48–59, January 2008.
Mahler, B. J., Metre, P. C. V., Crane, J. L., Watts, A. W., Scoggins, M. and Williams, E. S.
“Coal-tar-based pavement sealcoat and PAHs: Implications for the environment, human
health, and stormwater management,” Environmental Science and Technology, vol. 46,
no. 6, pp. 3039–3045, March 2012.
Mallin, M. “Pollutant removal efficacy of three wet detention ponds,” Journal of Environmental
Quality, vol. 3, pp. 654–660, 2002.
Mallin, M. A., McAuliffe, J. A., McIver, M. R., Mayes, D. and Hanson, M. A. “High pollutant
removal efficacy of a large constructed wetland leads to receiving stream improvements,”
Journal of Environmental Quality, vol. 41, no. 6, p. 2046–2055, 2012.
Maryland Department of the Environment, “2000 Maryland stormwater design manual, vol.
1, 2009 revision.” Maryland Department of the Environment, 2009.
National Oceanic and Atmospheric Administration, “Atlas 2, Volume 5, Precipitation Frequency
Atlas of the Western United States, (1973),” 1973. [Online]. Available: http://hdsc.nws.noaa.
gov/hdsc/pfds/other/id_pfds.html. [Accessed: 12 July 2016].
Peters, N. E. “Effects of urbanization on stream water quality in the city of Atlanta, Georgia,
USA,” Hydrological Processes, vol. 23, no. 20, pp. 2860–2878, 2009.
Reiners, J. “Vegetation growth and success as a function of soil moisture conditions in
bioretention cells,” 2008.
Roseen, R. M., Ballestero, T. P. and Houle, J. J. “UNHSC subsurface gravel wetland design
specifications,” 2009.
Roseen, R., Ballestero, T., Houle, J. and Loram, A. “Subsurface gravel wetlands for stormwater
management,” Stormwater Report, 2012. [Online]. Available: http://stormwater.wef.org/
2012/07/subsurface-gravel-wetlands-for-stormwater-management/. [Accessed: 12 July 2016].
Saunders W. S. and Yu, K. eds., Designed ecologies: The landscape architecture of Kongjian
Yu. Basel: Birkhäuser, 2012.
Schipper, L. A., Robertson, W. D., Gold, A. J., Jaynes, D. B. and Cameron, S. C. “Denitrifying
bioreactors – An approach for reducing nitrate loads to receiving waters,” Ecological
Engineering, vol. 36, no. 11, pp. 1532–1543, 2010.
Spetzman, J. “Phosphorus in lawns, landscapes and lakes,” Minnesota Department of
Agriculture, 2004.
State of Delaware, “DelDOT Road Design Manual.” State of Delaware, 2008.
Stone, R. “Evaluation and optimization of bioretention design for Nitrogen and Phosphorus
removal,” Civil Engineering, University of New Hampshire, 2013.
Tiefenthaler, L. L., Schiff, K. and Bay, S. “Characteristics of parking lot run-off,” Southern
California Coastal Water Research Project, Long Beach, California, SCCWRP Technical
Report #343, 2001.
University of New Hampshire, “Annual report,” CICEET, Durham, NH, 2007.
US Environmental Protection Agency, “Aquatic life ambient water quality criteria for Ammonia
– Freshwater,” Office of Science and Technology, Washington DC, EPA 822-R-13-001,
2013.
US Environmental Protection Agency, “National Pollutant Discharge Elimination System
(NPDES) permit writers’ manual,” EPA-833-K-10-001, 2010.
US Environmental Protection Agency, “National rivers and streams assessment 2008–2009:
A collaborative survey,” US Environmental Protection Agency, Washington DC, DC 20460
EPA/841/D-13/001, 2013.
US Environmental Protection Agency, “Nitrogen and phosphorus in streams in agricultural
watersheds,” February 2010. [Online]. Available: http://cfpub.epa.gov/eroe/index.cfm?
fuseaction=detail.viewInd&lv=list.listbyalpha&r=219683&subtop=200. [Accessed: 8 June
2015].
USDA (US Department of Agriculture), “Part 630 hydrology national engineering handbook,”
US Department of Agriculture, 210–VI–NEH, 2007.
USDA (US Department of Agriculture), “Urban hydrology for small watersheds,” Technical
Release 55, 1986.
Virginia Department of Environmental Quality, “Virginia stormwater design specification

No. 13, Version 2.0,” 2013.
Yan, L. “Evaluation of flood regulation and storage capacity in Harbin Qunli National Urban
Wetland Park,” Modern Landscape Architecture, vol. 10, no. 1, pp. 43–50, 2013.
Yang, H. “Dissolved nutrients and atrazine removal by column-scale monophasic and biphasic
Rain Garden Model Systems,” Chemosphere, vol. 80, pp. 929–934, 2010.
Yang, H., Dick, W. A., McCoy, E. L., Phelan, P. L. and Grewal, P. S. “Field evaluation of a
new biphasic rain garden for stormwater flow management and pollutant removal,”
Ecological Engineering, vol. 54, pp. 22–31, May 2013.
10
Increasing the sustainability
of agriculture
Introduction
Generally, the objective of ‘sustainable development’ is applied to the design and
construction of urban buildings and infrastructure within in cities and towns. Failure
to implement sustainable practices in the enormous land areas that support these urban
centres ignores the impacts with regional or global implications. This is the reason
for this review of agricultural practices and the application of constructed wetlands
to reduce the amount of toxic chemicals and excess nutrients that compromise human
and environmental well-being.
Agriculture consumes more water than any other economic sector. Most of this
comes from surface waters and may be nearly potable, if the source has not been greatly
contaminated by poorly treated sewage, agricultural run-off or industrial discharges.
Some of the water used by agriculture is pumped from aquifers. Therefore, the issue
of water supply and its allocation is the first sustainability concern. The amount of
water withdrawn from either groundwater or surface water must be limited to the
recharge rates of aquifers and the minimum requirements of healthy aquatic ecosystems.
In some cases, such as at dairies and aquaculture operations, process water can be
cleaned by constructed wetlands and reused to conserve potable supplies. Where this
is possible, it is irresponsible to use precious potable water for non-potable uses.
The agricultural sector is diverse. It can be divided conveniently into animal and
fish operations and plant crop operations. Each of these is characterized by different
point and non-point source pollution by-products that can be reduced by treatment
wetlands (Figure 10.1). There are tremendous opportunities to improve the sustain-
ability of agricultural operations due to their diversity and the enormous quantities
of organic material, fertilizers and pesticides that currently mix with process water
or stormwater and drain to natural water bodies.
Application of constructed wetlands in

plant operations
Introduction
The production of commodity crops, such as corn, rice, wheat and cotton, is supported
by large quantities of fossil fuels, insecticides, herbicides and artificial fertilizers.
234 INCREASING THE SUSTAINABILITY OF AGRICULTURE
Figure 10.1
This is a constructed wetland
at a field edge that receives
agricultural drainage water to
capture excess nutrients and
pesticides. When designed in
concert with tile drainage
systems, the water in the
wetland can be used to irrigate
fields during drought years.
Photo: Peggy Greb, USDA,
Agricultural Research Service.
product.
Stormwater run-off, wind and infiltration into groundwater transfer these pesticides
and fertilizers to the aquatic environment. Often, the quantities of these pollutants
overwhelm the capacity of the ecosystem to sequester or transform them into less
harmful substances. In this case, engineered wetlands can restore the balance through
treatment of contaminates near the source. The first example of this is a natural wetland
that was targeted to receive field run-off decades ago. The wetland provides a model
that we can use to design and locate constructed wetlands to make agriculture more
sustainable.
Free water surface wetland for pesticide and nutrient

reduction
The use of a linear riparian wetland to remove total suspended solids (TSS), nitro-
gen, phosphorus and three common agricultural pesticides was extensively studied
in Mississippi, USA. The site is along the Coldwater River (34°40′04.93″ N,
90°13′38.09″ W), where the floodplain has a low gradient resulting in many backwater
channels and wetlands (Figure 10.2). In this case, the linear wetland is naturally severed
from the main stem of the river. However, approximately 865 acres (350 hectares)
of agricultural fields drain first into drainage ditches leading to an intermittently wet
slough, which discharges into the wetland. The riparian wetland was modified through
INCREASING THE SUSTAINABILITY OF AGRICULTURE 235
Figure 10.2
The field run-off enters the
wetland at the left and flows
toward the Coldwater River on
the right.
34°40′04.93″ N 90°13′38.09″ W,
image date: 10/15/2014, image
accessed: 12/5/2015.
the construction of two adjustable crest weirs for research purposes. The upstream
weir serves as a dam to create an upstream pool and the inlet to a linear wetland that
is 1,640 feet long (500 m), 66 feet wide (20 m), with an average depth of 11 inches
(28 cm) (Figure 10.3) (Lizotte et al., 2012).
A simulated storm event equivalent to about a 1 cm rainfall from a 39.5-acre
(16 hectares) field and with a very similar hydrograph to a measured natural storm
hydrograph was routed through the riparian wetland by lowering the upstream weir
and injecting contaminants to match concentrations measured in the slough when water
was present. The peak flow into the wetland was 1,350 gallons per minute (85 litres
per second) and the volume was 161,409 gallons (611 m3). The simulated storm did
not result in outflow from the downstream weir. However, the inflow resulted in a
pulse through the wetland. Results from a chemical tracer study revealed that within
about 5 hours, the pulse reached the outlet (500 m). Outflow occurred 22 days after
the simulated storm due to a local thunderstorm (Lizotte et al., 2012).
Before the simulated storm was applied, the concentrations of contaminants
were measured at six points along the wetland (0, 10, 40, 100, 300 and 500 metres
from the inlet). The concentrations were: TSS – less than 100 mg/L at each location,
total phosphorus – less than 1 to 3.4 mg/L, soluble phosphorus – 0.02 to 0.16 mg/L,
ammonia and nitrite – less than 0.02 mg/L at each location, nitrate – 0.03 to
0.05 mg/L, total nitrogen – 2 to 6.7 mg/L and the target pesticides – below detection
at all locations. During the first 1.3 hours of the simulated storm, the following amounts
of contaminants were introduced into the flow: 270.8 kg sediment, 3.6 kg P2O5,
6.1 kg NH4NO3, 6,600 milligrams of a. i. atrazine + 5,220 milligrams of a. i.
S-metolachlor (Bicep II Magnum®) and 630.4 milligrams of a. i. permethrin (Hi Yield
38®). A tracer of 200 grams NaCl allowed the movement of the contaminants within
the wetland to be followed over time and distance (Lizotte et al., 2012).
The simulated storm added TSS at a concentration of 375 mg/L (higher than the
natural inflow concentration of about 325 mg/L). The water quality parameters were
measured at a point 100 m from the wetland inlet at 1, 1.5, 2, 2.5, 3, 3.5, 4 and
5 hours and at 28 days after the beginning of the simulated storm (Table 10.1). Through
filtration and sedimentation, the wetland removed 90–98 per cent of TSS within
Figure 10.3
Biologist Charles Bryant and
Richard Lizotte measured
water quality in the linear
wetland. Pesticides entering
upstream degraded quickly in
the wetlands before reaching
the downstream areas.
Photo: Richard Lizotte, USDA,
product.
48 hours. Total phosphorus declined from a concentration of 3.94 mg/L at 3 hours,

to 1.33 mg/L at 5 hours and 1.05 mg/L at 28 days. Soluble phosphorus declined from
a peak concentration of 3.44 mg/L at 2 hours, to 0.25 mg/L at 5 hours. Ammonia
peaked at 2 hours at 0.42 mg/L, and declined to 0.06 mg/L at 5 hours. Nitrate levels
were quite low, peaking at only 0.06 mg/L at 1 hour, and declining to 0.03 mg/L
4 hours before rising again to 0.05 mg/L at 5 hours. Total nitrogen peaked at 7.14
mg/L at the 2-hour measurement, and declined to 2.36 mg/L at 4 hours before rising
again slightly to 2.81 mg/L at 5 hours. Overall the treatment effectiveness of the linear
wetland was very high (Lizotte et al., 2012).
The pesticides measured in this study are commonly applied to crops in the United
States. Atrazine is one of the most widely used herbicides in the United States and is
the herbicide most often encountered in drinking water supplies. Atrazine disrupts
the endocrine system and was banned in the European Union in 2004 due to high
levels in drinking water. S-metolachlor is a herbicide that accumulates in fish.
Permethrin is a common synthetic, broad-spectrum insecticide that is used on crops,
such as cotton, alfalfa, maize and wheat and on ornamental turf. It is a likely
carcinogen and is highly toxic to fish and cats.
All of the pesticides reached their peak concentration at the 100-m point 1.5 hours
after the beginning of the simulated storm. Atrazine peaked at 145.66 ␮g/L, and
declined to 19.69 ␮g/L at 5 hours, and was undetectable at 28 days. S-metolachlor
Table 10.1 Sediment (mg/L), nutrient (mg/L) and pesticide (␮g/L) concentrations in the wetland at 100
m from the inlet (Lizotte et al., 2012)
Parameter Time (h)

1 1.5 2 2.5 3 3.5 4 5 28 days
TSS 159 142 76 103 69 93 68 112 160
Total PO4–P 2.77 4.00 3.94 2.86 2.22 2.13 1.46 1.33 1.05
Soluble PO4–P 0.71 3.06 3.44 2.02 1.15 0.51 0.43 0.25 0.43
NH4+–N 0.12 0.40 0.42 0.27 0.16 0.08 0.08 0.06 Ua
NO3–N 0.06 0.06 0.05 0.05 0.04 0.04 0.03 0.05 0.07
NO2–N 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02
Total nitrogen 4.18 6.75 7.14 5.58 2.98 2.52 2.36 2.81 4.07
Atrazine 53.14 145.66 77.95 54.5 36.62 26.80 20.84 19.69 Ub
S-metolachlor 48.0 114.9 65.2 42.0 26.7 18.9 14.5 13.9 Uc
cis-Permethrin 2.6 7.2 3.1 1.7 1.0 0.8 0.6 0.4 0.1
trans-Permethrin 2.1 4.5 2.5 1.3 0.7 0.6 0.4 0.3 0.1
Ua Below detection limit of 0.02 mg/L
Ub Below detection limit of 0.01 ␮g/L
Uc Below detection limit of 0.1 ␮g/L
also peaked at 114.9 ␮g/L, then declined to 13.9 ␮g/L at 5 hours, and to undetectable
levels at 28 days. Cis-Permethrin peaked at 7.3 ␮g/L, and declined to 0.04 ␮g/L at
5 hours. Trans-Permethrin peaked at 4.5 ␮g/L, and declined to 0.3 ␮g/L at 5 hours
(Lizotte et al., 2012).
Field edge free water surface constructed wetland

The example below is a more common situation where a constructed wetland is created
at the edge of a field to protect receiving streams and rivers.
Red River Research Station in Louisiana constructed an agriculture run-off
treatment system including sedimentation basins, a 4.5-acre (1.82 hectares) shallow
wetland cell 18 inches deep (0.53 m) and a 10-acre (4.04 hectare) pond with a maxi-
mum depth of 9 feet (2.3 m) (Figure 10.4). The deep pond can hold about 18,000 m3.
The wetland receives run-off from 400 acres (162 hectares) of farmland (3.6 acres of
wetland per 100 acres of run-off) (Millhollon and Dans, 2010). Water samples taken
at the field edge and in the shallow wetland and deep pond reveal a reduction in the
average concentrations of organic nitrogen and ammonia (65 per cent, to about 1
mg/L), total phosphorus (59 per cent, to about 1mg/L), soluble reactive phosphorus
(45 per cent, to about 0.7 mg/L) and TSS (90 per cent). Nitrate was reduced to less
than 0.5 mg/L. The wetland reduces pollutants entering the nearby, impaired Flat River,
which the US Environmental Protection Agency (EPA) targets with a 30 per cent
reduction in nitrogen and 62 per cent reduction in phosphorus (Jeong, Girouard and
Colyer, 2015).
These examples demonstrate that riparian free water surface wetlands can
sequester or degrade contaminants common in agricultural run-off during small to
moderate rainfall events. These events generally account for the greatest concentrations
of contaminants. Although the example of the wetland along the Coldwater River
was a natural wetland, it had been modified to receive agricultural run-off from a
large watershed. The addition of weirs allowed the hydrology of the wetland to be
controlled seasonally or for contaminant treatment purposes. Constructed wetlands
Figure 10.4 The Red River Research Station constructed wetland features a deep pool (foreground),
shallow wetland and sedimentation basins.
Photo: Google Earth 32°24′57.78″ N 93°37′48.74″ W, image date: 10/29/2104, image accessed: 9/15/2015.
could be designed based on this model, to achieve similar outstanding treatment results.
The Coldwater River study demonstrated that 40–80 per cent of nitrogen and
phosphorus, and 80–98 per cent of the target pesticides were removed within 300 m
of the wetland inlet. Capturing or degrading these contaminants protects the main
stem of the river from agricultural non-point source pollution (Lizotte et al., 2012).
Figure 10.3 shows a highly diverse landscape that is well structured vertically. It is
certainly a regional wildlife refuge. The Red River Research Station example shows
that similar results can be achieved in edge-of-field free water surface wetlands.
Horizontal subsurface flow wetlands for nitrate

reduction in field run-off
Dead zones
In the examples above, the concentration of nitrate entering the wetland was low.
This is often not the case. High nitrogen levels, primarily due to agricultural run-off,
in rivers are largely responsible for the oxygen-depleted (hypoxic) zones in coastal
waters, although concentrated livestock feeding operations and treated municipal
sewage effluent also contribute nitrogen to rivers.
It is quite certain that agricultural sources are the primary cause of the dead zones
due to a natural large-scale experiment. A 15,400 square mile (40,000 km2) dead zone
in the Black Sea developed, as the use of artificial fertilizer and the presence of livestock
operations (especially pig farms) doubled the amount of nitrogen and phosphorus
entering the sea. The result was the collapse of the ecosystem and shellfish populations.
When the Soviet Union collapsed in 1989, artificial fertilizers were too expensive for
farmers to purchase and many of the huge swine farms failed. The abrupt decrease
in nutrient rich run-off resulted in a dramatic reduction in the area of the dead zone
and the eventual recovery of the shellfish ecosystem (Mee, 2006).
The concentration of dissolved oxygen in the water in hypoxic areas is often so

low that aquatic organisms cannot survive. The nitrogen-rich water from the river
causes explosive growth of phytoplankton (Figure 10.5) in the coastal areas. When
these die, bacteria decompose the organic matter and consume almost all of the
dissolved oxygen in the process. The drop in oxygen causes mass deaths in populations
of fish and other organisms.
The number (now more than 400) and size of the dead zones are increasing and
are particularly problematic in the temperate waters of the northern hemisphere (Figure
10.6). The dead zone near the Mississippi River occupies 5,000–6,000 square miles
and is well known, but hypoxic zones are emerging on both coasts of North America,
Europe and China. These zones are clear indications of ecosystems out of balance and
the imbalance is caused by unsustainable human development. Capturing and
removing phosphorus and nitrogen, especially in the form of ammonia and nitrate,
are urgently needed measures to restore ecosystem health and long-term human well-
being.
Denitrifying wetlands
Especially where stormwater and residual water from irrigation are drained from fields
through underground drain-tile networks, nitrates from artificially fertilized soil are
problematic and drain via surface water to pollute rivers and coastal areas. Since nitrate
Figure 10.5
Phytoplankton bloom in the
Bay of Biscay, France.
Photo: Jeff Schmaltz, MODIS
Rapid Response Team,
NASA/GSFC, 2004. License:
Figure 10.6 This image shows the distribution and size of dead zones along with ocean carbon
content and human population density. Dark blue indicates high carbon content, while the brown hue
indicates high human population density. The pink hue indicates dead zones with the largest circles
indicating zones 3,860 square miles (10,000 square km2). The black dots show the location of dead
zones of unknown size.
Image: NASA Earth Observatory, 2008. License: government work product.
is soluble, it can infiltrate through the soil to reach ground water. Edge-of-field
constructed wetlands have been developed to remove nitrate from the drainage system
effluent. These are horizontal subsurface flow wetlands that use carbon (usually in
the form of wood chips) as the filter media, instead of gravel, as in the treatment of
domestic sewage. The principle is the same, however. Nitrate-rich water flows through
the filter media where oxygen is depleted. This environment encourages the growth
of bacteria that uses nitrate as its energy source, resulting in nitrogen gas as a harmless
by-product (Schipper et al., 2010).
Figure 10.7 illustrates that the denitrifying wetlands are simple to construct and
require little or no maintenance. The porosity of the filter material is quite high so
flow-through is rapid. The wetland should be lined with clay or a synthetic liner if
there is a chance that the water can infiltrate into the groundwater. Since the wood
chips are under saturated, oxygen-depleted conditions, they decay slowly, resulting
in a lifespan of as much as 15 years. Overflow spillways or pipes can be added to the
wetlands if necessary. Similarly, large denitrifying wetlands can be divided into cells
to improve uniform distribution of the polluted water and to increase retention time
(Schipper et al., 2010).
Denitrifying wetlands can be located along field edges to limit the displacement
of crops and operate via gravity and very little maintenance. Studies of installed
denitrifying wetlands show a reduction of 2 to 22 grams of nitrogen per m3 of filter
material per day. The higher removal rates are associated with higher concentrations
of nitrate in the inflow and higher seasonal temperatures. At temperatures as low as
Figure 10.7
In this transparent perspective of
a denitrifying wetland, the water
enters from drain tiles at the left
and is distributed across the
wetland. The polluted water
flows through the wood chip
filter and is collected and
discharged to the right. The
wetland can be vegetated (with
reed canary grass, for example)
and harvested. It is best if fully
nitrified run-off enters the
wetland as the last stage in a
treatment sequence, as neither
phosphorus nor ammonia are
removed well by this wetland.
Image: Gary Austin after Schipper
(Schipper et al., 2010).
1–5°C, removal of about 2 grams of nitrogen per m3 per day is still evident. It appears
that a range of high nitrate concentrations (3.1–49 mg/L) results in a consistent nitrate
removal rate by the denitrifying wetland. At lower concentrations, as in urban
stormwater, the removal efficiency may increase with increasing nitrate concentration
(Schipper et al., 2010). At an influent temperature of 19°C and greater and a nitrate
concentration of 66 mg/L, nitrate is reduced to nearly 0 mg/L within 7.5 feet of the
inlet (Haunschild, Leverenz and Darby, 2010).
A denitrifying wall is similar to the horizontal flow wetland discussed above, it
is situated to intercept groundwater along its side rather than top surface to remove
nitrate that has already contaminated the groundwater basin. See the section on dairy
operations for an example of a denitrifying wetland applied to livestock waste.
Application of constructed wetlands in animal

operations
Commercial scale production of livestock generates staggering amounts of organic
waste and contaminated water (Figure 10.8). The quantities vary according to the
type of animal production and whether it is concentrated, as in feedlots and dairies,
or dispersed, as in pasture operations. Constructed wetlands of all types have been
applied to the treatment of wastewater and stormwater from concentrated animal
operations including dairies, feedlots, swine and chicken operations.
Feedlots and dairies

Feedlots
Although feedlots do not produce much wastewater, extremely high levels of nitrogen,
phosphorus and other substances are part of the stormwater run-off from these facilities
(Figure 10.9). Stormwater run-off from concentrated animal feeding operations can
be dramatically cleaner after treatment by constructed wetlands before discharge into
receiving streams and rivers.
Figure 10.8
12 per cent of the phosphorus
load in the Town Brook
watershed in New York is from
dairy cow dung deposited
directly in pasture streams.
This watershed is a major
source of drinking water for
New York City.
Photo: Stephen Ausmus, USDA
product.
The Beef Nutrition Farm at Iowa State University is an example of a four-stage

treatment system (Figure 10.10), which treats stormwater from a farm with about
380 cattle. The first stage is a shallow level concrete trough that receives run-off from
the barn and outdoor feeding area. The feedlot is 56 by 756 feet. Although narrow,
the settling basin width allows equipment to recover and compost the sediment. Liquid
from the settling basin flows through a pipe to a vertical subsurface flow wetland
planted with reed canary grass. Run-off from outdoor pens (34 inches by 110 feet)
and additional farm area is also collected and delivered to the vertical flow wetland.
The third stage is a free water surface wetland, and the fourth stage is a vegetated
swale and wet meadow.
The area of the vertical flow wetland (120 by 350 feet) is 20 per cent of its
catchment area. The basin’s 2-feet high berm contains the 25-year, 24-hour storm.
The detained water infiltrates through 5 feet of porous soil and is collected by three
perforated pipes 4 inches in diameter. The effluent from the subsurface wetland empties
into a 90 by 150-feet free water surface wetland, 18 inches deep, which discharges
into a vegetated swale and wet meadow and finally to Onion Creek (Lorimor, 2003).
Table 10.2 illustrates the outstanding performance of the treatment sequence. The
levels of all contaminants entering the vertical flow wetland (settling basin effluent)
are very high. The vertical flow wetland removes 80 per cent of the organic and
ammonia nitrogen and 77 per cent of the phosphorus (Lorimor, 2003). The large
Figure 10.9
Stormwater drainage from
a feedlot pen overflows into
a basin where sediments
can settle.
Photo: Kongjian Yu.
Figure 10.10 Iowa State University Beef Nutrition Farm. A – Barn and feeding area; B – Concrete
settling basin; C – Cattle pens; D – Vertical subsurface flow wetland; E – Free water wetland;
F – Vegetated swale to stream.
Photo: Google Earth, 42°03′23.70″ N 93°40′57.28″ W, image date: 11 June 2012, image accessed:
16 September 2015.
decrease in ammonia and the simultaneous increase in nitrate indicate very high
nitrification and some denitrification, as the increase in nitrate is modest. The cost of
the vertical flow wetland was very low in this case, because the soil is porous; in other
situations, excavation and replacement of clay soils would be necessary. Although the
vertical flow wetland area is excluded from grazing, the reed canary grass is harvested
and fed to the cattle (Lorimor, 2003).
The free water surface wetland further reduces contaminant concentrations as does
the vegetated swale and meadow, with the exception of a large increase in nitrate.
Table 10.2 Feedlot wetland performance (Lorimor, 2003). Five year averages except for Onion Creek
data, which is a 2-year average. All units are mg/L
Total Kjeldahl Nitrate and Ammonia Total Total

nitrogen nitrite NH3 phosphorus solids
Effluent from settling basin 196.1 0.09 109 46.6 3,294.1
Wetland inflow 39.6 1.7 20.7 10.5 1,149.6
Wetland outflow 30.7 1.3 19.5 8.2 964.8
Edge of Onion Creek 9.7 9.2 3.7 5.4 614.4
Onion Creek upstream 3.3 15.3 0.5 4.1 782.0
This suggests that the remaining ammonia is being converted to nitrate, but that
conditions are inadequate for the conversion of nitrate to nitrogen gas. The addition
of a nitrifying wetland as discussed earlier would eliminate most of this nitrate. The
nitrate levels upstream of the farm indicate even higher levels of nitrate that would
cause eutrophication in lakes. Although the treatment sequence dramatically reduced
the amount of phosphorus, it is also at levels that would encourage eutrophication.
Nevertheless, the overall reductions in Kjeldahl nitrogen (organic nitrogen and
ammonia), ammonia and phosphorus of 95 per cent, 97 per cent and 88 per cent,
respectively, are outstanding (Lorimor, 2003).
Dairies
Dairies are similar to concentrated animal feeding operations (feedlots), as the density
of animals is high, but dairies use a much greater volume of water to keep the animals,
milking floors and equipment clean. At feedlots and dairies, manure accumulates
rapidly. The solids are managed through composting, but the piles are generally open.
Rainstorms cause leachate from the piles and run-off from the pens that is highly
contaminated with sediment, excess nutrients and often with hormones and antibiotics
(Figure 10.9). Dairy wastewater can be polluted with as much as 5,000 mg/L of BOD
compared to the 4 mg/L of unpolluted streams. The BOD concentration of dairy
wastewater is about three times higher than in swine wastewater. Concentrations of
ammonia and nitrate and pathogenic bacteria are also particularly high. The US Clean
Water Act specifies that run-off from these facilities must be collected and disposed
of responsibly. The requirements differ depending on the size of the operation.
Generally, wastewater from livestock operations is collected in sewage lagoons,
which are not very efficient and often emit noxious odours. Land-based disposal
systems, where stormwater run-off from the pens and process water is collected and
then sprayed over a large, vegetated land area, are effective (Figure 10.11), but some-
times not permitted if the water table is high or the soil highly permeable. The depth
of the sprayed effluent is controlled by federal regulations to protect groundwater from
contamination. Of course, if effluent saturated soils receive rainstorms, this could result
in contaminated run-off. If a large land area is not available, then other nutrient removal
measures are necessary.
Horizontal subsurface flow treating dairy wastewater

A Vermont, USA, study demonstrates that horizontal subsurface flow wetlands are
effective for achieving secondary quality effluent from dairy stormwater run-off from
Figure 10.11
Sheet flow of stormwater
across the feedlot pen is
collected in a swale to allow
solids to settle out. In this case,
the water is then sprayed on
adjacent pastureland.
Photo: Stephen Ausmus, USDA.
License: public domain.
Figure 10.12
Diagram of the treatment
sequence for a Vermont dairy.
A – Inflow from feedlot catch
basin; B – Settling pond;
C – Screened outlet; D – Two
settling tanks; E – Inflow from
milking parlour; F – Settling
tank; G – Settling tank;
H – Mixing and dosing tank;
I – Horizontal subsurface flow
wetland; J – Disposal tank.
Image: Gary Austin after
Tunçsiper (Pelissari et al., 2014).
pens and wastewater from the milking parlour. This is significant as Vermont
experiences very cold winters, making treatment in open water wetlands limited. Over
a 4-year period, the dairy tested the performance of four constructed wetland cells:
one vegetated (Scirpus fluviatilis, river bulrush) and aerated, one vegetated and not
aerated, one not vegetated and aerated, and one not vegetated and not aerated. Each
cell was 2,422 square feet (225 m2) and 24 inches deep (60 cm). All cells featured
2-cm diameter gravel for the first one-third of their length and 1-cm diameter gravel
for the remainder. The vegetated and aerated cell produced slightly better water quality
than the other cells. Figure 10.12 illustrates that stormwater run-off from the outdoor
pens was collected in a catch basin and directed to a lagoon for sedimentation. After
the lagoon, the stormwater was screened and directed to two settling tanks in series
before flowing into a tank where it joined wastewater from the milking parlour. The
wastewater from the milking parlour was directed to a settling tank and then to the
common tank (Tunçsiper, Drizo and Twohig, 2015).
The combined wastewater was directed to the wetland. The wetland cells were
designed to contain the 2-year 24-hour storm volume (37.5 m3). The hydraulic
loading rate was between 10.44 and 14.42 cm per day, which is higher than common.
The organic loading rate varied widely but averaged 0.21 kg per m2, which is much
higher than common. Despite the high organic load, the wetland cells produced no
clogging symptoms during the 4-year study. The inflow varied greatly between 132
and 2,642 gallons (500–10,000 litres) per day, depending on the weather and the
milking parlour operation. The hydraulic residence time varied from 3 to 60 days
depending on the inflow rate. The long resting periods may be important for the
removal of organic material and may account for the capacity of the wetland to tolerate
high organic loading rates at other times. The four cells operated in parallel. Aeration
improved the performance of the cells. Performance also improved over time with the
fourth year demonstrating the highest removal of BOD and TSS. The vegetated and
aerated cell removed 86 per cent and 94 per cent of the BOD and TSS, respectively.
The water entering this cell had an extraordinarily high average BOD concentration
of 2,445 mg/L, which was reduced by the wetland to 278 mg/L. The initial TSS
concentration was 829 mg/L, which was reduced to 40 mg/L (Tunçsiper, Drizo and
Twohig, 2015). Although the BOD removal percentage is fairly high, the 278 mg/L
average concentration is far above the 30 mg/L maximum permitted for discharge by
municipal wastewater treatment plants.
When the wetlands were operated in a two-stage sequence, performance also
increased. BOD removal increased by 12 per cent and TSS removal increased by 16
per cent with two cells in series. This experiment demonstrates the outstanding
treatment capability of horizontal flow constructed wetlands in a cold climate. The
performance of the wetland varied only slightly between winter and summer seasons.
The wetland cells were covered with a layer of mulch to insulate them in winter. While
BOD and TSS are basic indirect measures of water quality, excess ammonia, nitrate
and phosphorus might still overwhelm receiving waters. The vegetated cells were much
more effective in capturing phosphorus than cells with no vegetation. The wetland
cell that was planted and aerated removed an average of 48.4 per cent of dissolved
reactive phosphorus. This was 2.4 times more than the cell that was not aerated but
planted. During the first 21⁄2 years of operation, the treatment wetlands showed high
efficiency in dairy effluent treatment. The effluent from the four horizontal flow
wetlands showed a 75 per cent average reduction in organic matter, 90 per cent in
TSS, 90–99 per cent in E. coli and 80 per cent in ammonium (Tunçsiper, Drizo and
Twohig, 2015).
Comparison of horizontal and vertical flow wetlands for

dairy wastewater treatment
As both nitrate and ammonia are generally high in feedlot and dairy wastewater
effluent, constructing a hybrid wetland should lead to better tertiary treatment results.
A study in Brazil collected dairy wastewater from the milking parlour. It received
primary treatment in a pond. A horizontal subsurface flow wetland and a vertical
subsurface flow wetland, operating in parallel, allowed assessment of the treatment

performance of each wetland type. The inflow to the wetland was high in nitrogen
particularly ammonia: organic nitrogen and ammonia – 69 mg/L, ammonia – 55 mg/L
and nitrate – 5 mg/L. The effluent concentrations were consistent with the expected
performance of each wetland type. The vertical flow wetland removed more ammonia
than the horizontal flow wetland. Conversely, the horizontal flow wetland removed
more nitrate than the vertical flow system. This confirms that a two-stage treatment
system with the first stage using the vertical flow distribution to convert ammonia to
nitrate and a second horizontal flow stage to convert the nitrate to nitrogen gas would
be more effective for reducing all species of nitrogen (Pelissari et al., 2014).
In order to reduce nitrate in effluent from feedlot and dairy operations, a
denitrifying wetland can be included in the treatment sequence to form a hybrid
wetland. An example of this is a denitrifying wetland constructed on a dairy farm in
Dargaville, New Zealand. Wash water from the dairy building and stormwater from
outdoor pens flowed to a lagoon and were then treated in a membrane biological
reactor. After this secondary treatment (converting organic nitrogen and ammonia to
nitrate), the water was treated in the denitrifying wetland. The carbon source was 50
per cent wood chips and 50 per cent sawdust from Pinus radiata. The wetland received
an annual load of 68 pounds (31 kg) of nitrogen. The hydraulic flow rate varied greatly
and the total nitrogen was as high as 250 g of nitrogen/m3 of filter material. The amount
of nitrate was as high as 200 g/m3 but was generally below 100 g/m3. The outflow
from the wetland contained less than 1 g/m3 of nitrate. As expected, the denitrifying
wetland did not reduce the amount of ammonia or organic nitrogen (Schipper,
Cameron and Warneke, 2010).
Removal of endocrine disruptors by constructed wetlands

Endocrine-disrupting compounds from dairies are significant pollutants that can
impact ecosystem and human health when discharged into the water supply. The 2.2
million dairy cattle in Britain excrete three times more oestrogen than the human
population of 59 million. Free water surface wetlands treating dairy wastewater
effectively remove endocrine-disrupting pharmaceuticals. One study demonstrated that
95.2 per cent of oestrogen hormones and 92.1 per cent of androgen hormones were
removed by a free water surface treatment wetland at a dairy (Cai et al., 2012). A
test of a hybrid treatment wetland (vertical subsurface flow, horizontal subsurface flow
and free water surface cells operated in series) treating human wastewater revealed
high removal of oestrogen and antimicrobial substances (Ávila et al., 2014). This
sequence might also be applied to agricultural wastewater where animals receive regular
doses of hormones and antibiotics.
Swine wastewater treatment with a vertical flow

constructed wetland
The Iowa Beef Nutrition Farm discussed earlier included a vertical subsurface flow
wetland. This example of a swine facility uses a composting step but also a verti-
cal subsurface flow wetland applied to the treatment of effluent from swine. Vertical
subsurface flow wetlands have also been used to treat wastewater from concentrated
chicken farms producing eggs. The swine application is interesting because a manure
Figure 10.13
A – Manure slurry;
B – Compost of manure and
poplar wood chips (1:2) in two
rows; C – Two leachate tanks,
one for each compost row;
D – Recirculation and dosing
tank; E – Vertical flow wetland;
F – Recirculation; G – Effluent
discharge.
Image: Gary Austin after Vazquez
(Vázquez et al., 2013).
composting facility generates the wastewater and the wetland includes a recirculation
feature in order to dilute the inflow with treated effluent (Figure 10.13) (Vázquez et
al., 2013).
The vertical flow wetland was a fairly standard design planted with Phragmites.
The wetland received 360 litres of water from the recirculation tank every 6 hours
for a hydraulic loading rate of 240 mm per day. The recirculation tank received a
dose (100–300 L) of compost leachate once each week. The concentrations of nitrogen
in the compost leachate were high (459 mg/L of NH3, and 904 mg/L of TKN).
Unfortunately, the concentration of TSS was also very high at more than 3,000 mg/L
during some operating periods. When the areal loading rates in grams per m2 per day
were 9.2 – TSS, 17.9 – total COD, 4.8 – BOD and 1.9 – TN, the concentrations were
reduced by more than 93 per cent by the vertical flow constructed wetland (Vázquez
et al., 2013).
This study lasted 200 days, and the performance of the wetland varied during
that period with respect to ammonia and nitrate removal. Initially, the nitrate
concentration in the wetland effluent was extraordinarily high at 178 mg/L, but after
the wetland matured, the concentration dropped to 10 mg/L. Both nitrification and
denitrification occurred, since the ammonia concentration was 16 mg/L during the
last part of the testing period. This suggests that the wetland developed regions of
anoxic conditions in, at least, portions of the weekly treatment cycle. Generally, vertical
flow constructed wetlands are characterized by high oxygen levels. Since a sedi-
mentation basin or other pretreatment step to remove some TSS and nitrogen did not
precede this wetland, the long-term performance is uncertain. The potential for
clogging or reduction in the areal application rate is likely. A more deliberate design
for TSS removal in pretreatment and for nitrification in the upper portion of the
bed and a denitrification flooded sub-basin, as presented in the chapter on vertical
flow constructed wetlands, might lead to more stable long-term performance (Vázquez
et al., 2013).
Application of constructed wetlands in aquaculture

operations
The commercial production of fish and shrimp generate effluent that is high in excess
nutrients. As the concern for the health of streams and rivers has grown, regulations
requiring treatment have been enacted. Systems are being developed where fish and
plant production contribute to an efficient use of nutrients and water (Figure 10.14).
Treatment of effluent from flow-through systems and treatment to allow recirculation
of production water are options dependent on the species being farmed and availability
of water. Trout and Salmon require very cold, clean water, so flow-through treatment
is most common, although treatment of sludge removed from growing or hatchery
tanks are additional treatment circumstances. Other species are more tolerant of some
contaminants and lower dissolved oxygen.
Low concentrations of contaminants and great volumes of water characterize
aquiculture effluents, whether flow-through or recirculation is adopted. The typical
concentrations are: BOD5 – 15 mg/L, COD – 12 mg/L, ammonia – 1.3 mg/L and total
phosphorus – 0.3 mg/L. Hydraulic loading varies from an aerial loading of only a
few centimetres per day to more than 29 m. Consequently, the hydraulic residence
time also ranges widely from 3 hours to 4 days (Vymazal, 2014). Generally, the highest
hydraulic loading rates result in ineffective removal of nitrate, although removal of
ammonia and nitrite might be acceptable for the culture of fish and shrimp (Figure
10.15).
Figure 10.14
Will Allen illustrates that tilapia
production can be combined
with plant production for a
holistic approach to using
excess nutrients in pond water
to irrigate plants and improve
water quality of the fish.
Photo: Grifray. License: CC-BY-
SA 2.0.
Figure 10.15
Concentrated populations of
fish can create low water
quality in the rearing ponds.
Catfish are somewhat more
tolerant of poor water quality
than many other species. Here,
mature brood fish catfish are
inspected in a fish farm at the
Warmwater Aquaculture
Research Unit in Stoneville,
Mississippi.
Photo: Les Torrans, USDA,
product.
There are generally two treatment goals. The first is producing water quality high
enough to produce optimal growth if the water is re-circulated. The second is
complying with discharge regulations to protect the quality of receiving waters.
Nitrite and ammonia are particularly toxic to fish, whereas they are more tolerant of
higher nitrate and phosphorus concentrations. All of the types of constructed wetlands
have been successfully applied to treatment of aquaculture wastes. Free water surface
wetlands have been used to treat water from channel catfish and salmon farms.
Horizontal subsurface flow wetlands have been used to treat effluent from trout and
tilapia farms, while vertical flow wetlands have treated effluent from tilapia
production. Finally, several combinations of the wetland types have been configured
as hybrid wetlands and applied to a range of species.
Hybrid wetland for tropical fish production

Tropical aquaculture of fish for commercial markets is a potent source of point source
of pollution. This is due to the weekly or daily (or even more frequent) discharge of
the fish pond water to the rivers and streams. Production using this model adds about
47.3 kg of nitrogen to receiving water for every ton of catfish produced. Using
constructed wetlands to clean and recycle the production water has the potential to
avoid the discharge of tens of thousands of tons of nitrogen into every river where
aquaculture is significant, as it is in Vietnam, China and many other countries
(Konnerup, Koottatep and Brix, 2009).
Fish waste, uneaten fish food and the growth of algae and phytoplankton generate
large quantities of organic nitrogen, ammonia and nitrate. The use of hybrid wetlands
to treat these pollutants is necessary to create the aerobic conditions for nitrification
and the anoxic conditions for removal of nitrite and nitrate. Hybrid treatment
wetlands make possible recirculation of pond water with water quality sufficient for
the rapid growth of fish species. In contrast to treating domestic wastewater,
aquaculture effluent has a low concentration of contaminants but a very high volume
of water to be treated. Therefore, devising a treatment system with a high hydraulic
loading rate is important to minimize the size of the treatment wetlands (Konnerup,
Koottatep and Brix, 2009).
A study in Vietnam established a lined fish pond 8 by 8 m and 1 m deep, resulting
in a volume of 40 m3. The pond was stocked with 15 km of Tilapia (125 km were
harvested after 41⁄2 months). Two horizontal subsurface flow wetlands operating in
parallel and two vertical flow subsurface wetlands operating in parallel composed the
treatment system. The horizontal wetlands were 3.7 m long, 0.85 m wide and 0.3 m
deep. They were filled with gravel (30–50 mm diameter) and planted Canna generalis.
The vertical flow wetlands were 1.25 m in diameter and 0.6 m deep, and planted with
Canna generalis. The layers of gravel, from top to bottom, in this wetland were: a
0.2 m depth of 10–20 mm diameter gravel, a 0.4 m depth of 20–50 mm diameter
gravel and a 0.1 m depth of 50–100 mm diameter gravel. The vertical flow wetland
received pulses of water for 30 seconds followed by a 2–10-minute resting period.
The horizontal flow wetland received water continuously (Konnerup, Koottatep and
Brix, 2009).
The treatment system was operated at hydraulic loading rates of 750, 1,500 and
3,000 mm per day. The highest loading rate yielded adequate water quality for fish
production. At the 3,000 mm/day loading rate, dissolved oxygen was greater than
1 mg/L, BOD was less than 30 mg/L, total ammonia was less than 1 mg/L and nitrite
was less than 0.07 mg/L. The wetlands reduced BOD by 50 per cent, but the primary
concern was the concentration of ammonia and nitrite. The toxic limit of nitrite for
Tilapia is 0.5 mg/L compared to 0.07 mg/L produced by the wetland system. Ammonia
is also toxic to Tilapia but the concentration produced by the treatment wetlands was
well below harmful levels (Konnerup, Koottatep and Brix, 2009).
As expected, the horizontal flow wetlands removed more nitrate, while the vertical
flow wetlands removed more ammonia. Performance was the same in the wetlands
for total nitrogen, nitrite, total phosphorus and phosphate, although all of these
increased in concentration with increases in the hydraulic loading rate. Dissolved
oxygen was 2–4 mg/L in the effluent from the horizontal flow wetland and 5.3–
6.1mg/L for the vertical flow wetland (Konnerup, Koottatep and Brix, 2009).
The hydraulic loading rate of this system is many times greater than in previous
aquaculture operations. The high loading rate minimizes the land used and the
construction cost of the treatment wetland. However, the high hydraulic loading rate
did result in a steady increase in nitrogen and phosphorus over time. Most of the
nitrogen increase was organic nitrogen rather than ammonia or nitrate. Nevertheless,
the water quality was sufficient for fish production. The design and performance of
the treatment system would be improved by the inclusion of a settling tank to capture
algae and particles. This would protect the wetland filter media from clogging
problems. Algae grew rapidly in the pond and 45 species were identified. The algae
caused high dissolved oxygen (12–13 mg/L) in the pond during the day but low levels
(as low as 1 mg/L) at night (Konnerup, Koottatep and Brix, 2009).
Significantly, there was no discharge of polluted water during the 41⁄2-month dry
season. Water was discharged only during the heavy rains of the wet season. If widely
adopted, the use of constructed wetland would transform aquaculture from an
unsustainable and damaging agricultural model to a sustainable one (Konnerup,
Koottatep and Brix, 2009).
Hybrid wetland for tropical shrimp production

Shrimp aquaculture is another opportunity for high rate recirculation of production
water to establish a sustainable aquaculture industry. In this example, the hybrid
wetland is composed of a free water surface wetland and a horizontal subsurface flow
wetland. The efficiency of this sequence allowed a much higher hydraulic loading rate
than common, as it did in the hybrid wetland in Vietnam. In this case, the size of the
treatment wetland was 43 per cent of the size of the production pond, compared to
previous designs for shrimp aquaculture where the wetlands occupied 70–270 per cent
of the pond area. A sedimentation cell before the free water surface wetland is an
important component of the treatment system, and the horizontal flow wetland
showed no clogging symptoms. The hydraulic loading rate in the wetlands is 1.95 m
per day, and the hydraulic residence time is 3 days.
The treatment wetland reduced BOD from 7.1 to 2.5 mg/L, total ammonia from
0.18 to 0.06 mg/L and nitrite from 0.25 to 0.06 mg/L. However, nitrate increased to
40.1 mg/L and phosphorus increased to 3.8 mg/L. The concentration of nitrate and
phosphorus was much lower when the hydraulic loading rate was 1.54 m per day
rather than 1.95 m per day (Lin et al., 2005).
Horizontal wetland for trout production

The production of trout or salmon for commercial sale or in hatcheries for release to
natural streams and rivers differs from fish production in tropical climates. Trout and
salmon require very clean, cold water to flourish (Figure 10.16).
An example of a trout farm with a flow-through system in Germany illustrates
the application of a horizontal subsurface flow constructed wetland to treat process
water and effluent from tank cleaning. The wetlands were conversions of six previously
existing concrete sedimentation basins. These were divided into a sedimentation area
and a wetland treatment area. The gravel (4–8 mm grain size) wetland was 1 m deep
with Phragmites communis and Phalaris arundinacea equally dominant.
The hydraulic loading rate during the cleaning operations was very high at 13.6
m per day and the wetland removed 68 per cent of the TSS. During normal operation,
the hydraulic loading rate was 10.6 m per day and the total ammonia removal was
88 per cent. The subsurface wetland proved to be a better treatment solution than
the previous sedimentation system for BOD, TSS, ammonia and nitrite, but nitrate
removal was modestly higher in the sedimentation basin and phosphate concentrations
were unchanged. The high nitrate concentration was due, in part, to their high
concentration in the spring water feeding the farm.
Conclusion
This chapter demonstrated that low-cost and reliable constructed wetland solutions
are available, leading to substantial improvement in effluent water quality from crop,
livestock and aquaculture operations. Although these biological solutions will often
lack the recreation benefits of constructed wetlands in urban areas, they can greatly
improve ecosystem health and availability of wildlife habitat. This is important, given
the history of habitat destruction by agricultural development. Concentrating on the
possible secondary benefits, such as the harvesting of reed canary grass for cattle feed
as in the Red River Station example, reduces the economic cost of creating treatment
Figure 10.16
Trout farms are often situated
to allow water routed from
streams to flow through the
rearing raceways. The
concentrated populations of
fish produce waste and
uneaten food that produce
excess nutrients that can cause
eutrophication downstream.
Photo: Alejandro Linares Garcia.
wetlands. Excess nutrients in agricultural effluent are actually resources that could
be utilized to the benefit of the enterprise if more holistic planning were to be
implemented. Linking fish production to plant production is only the first of these
opportunities.
References
Ávila, C., Matamoros, V., Reyes-Contreras, C., Piña, B., Casado, M., Mita, L., Rivetti, C.,
Barata, C., García, J. and Bayona, J. M. “Attenuation of emerging organic contaminants in
a hybrid constructed wetland system under different hydraulic loading rates and their
associated toxicological effects in wastewater,” Science of the Total Environment, vol.
470–471, pp. 1272–1280, February 2014.
Cai, K., Elliott, C. T., Phillips, D. H., Scippo, M. L., Muller, M. and Connolly, L. “Treatment
of estrogens and androgens in dairy wastewater by a constructed wetland system,” Water
Resources, vol. 46, no. 7, pp. 2333–2343, May 2012.
Haunschild, K., Leverenz, H. L. and Darby, J. Development of design criteria for denitrifying
treatment wetlands WERF Report DEC13U06. Alexandria, VA: WERF, 2010.
Jeong, C., Girouard, E. and Colyer, P. “Tailwater recovery system as a best management practice
to improve irrigation water quantity and quality in Louisiana,” Presented at the 9th
Louisiana Water Conference, Alexandria, Louisiana, 2015.
Konnerup, D., Koottatep, T. and Brix, H. “Treatment of domestic wastewater in tropical,
subsurface flow constructed wetlands planted with Canna and Heliconia,” Ecological
Engineering, vol. 35, no. 2, pp. 248–257, February 2009.
Lin, Y. F., Jing, S. R., Lee, D. Y., Chang, Y.-F., Chen, Y.-M. and Shih, K.-C. “Performance of
a constructed wetland treating intensive shrimp aquaculture wastewater under high hydraulic
loading rate,” Environmental Pollution, vol. 134, no. 3, pp. 411–421, April 2005.
Lizotte, R. E., Shields, F. D., Murdock, J. N., Kröger, R. and Knight, S. S. “Mitigating
agrichemicals from an artificial run-off event using a managed riverine wetland,” Science of
the Total Environment, vol. 427–428, pp. 373–381, June 2012.
Lorimor, J. “An infiltration and wetland system to treat beef feedlot run-off,” Iowa State
University, Department of Agricultural and Biosystems Engineering, Ames Iowa, 2003.
Mee, L. “Reviving dead zones,” Scientific American, vol. 295, no. 5, pp. 78–85, November
2006.
Millhollon, E. P. and Dans, D. R. “Using constructed wetlands to protect Louisiana waterways,”
Louisiana Agriculture, vol. 53, no. 1, pp. 10–11, 30, 2010.
Pelissari, C., Sezerino, P. H., Decezaro, S. T., Wolff, D. B., Bento, A. P., Junior, O. de C. and
Philippi, L. S. “Nitrogen transformation in horizontal and vertical flow constructed wetlands
applied for dairy cattle wastewater treatment in southern Brazil,” Ecological Engineering,
vol. 73, pp. 307–310, December 2014.
Schipper, L. A., Cameron, S. C. and Warneke, S. “Nitrate removal from three different
effluents using large-scale denitrification beds,” Ecological Engineering, vol. 36, no. 11, pp.
1552–1557, November 2010.
Schipper, L. A., Robertson, W. D., Gold, A. J., Jaynes, D. B. and Cameron, S. C. “Denitrifying
bioreactors – An approach for reducing nitrate loads to receiving waters,” Ecological
Engineering, vol. 36, no. 11, pp. 1532–1543, 2010.
Tunçsiper, B., Drizo, A. and Twohig, E. “Constructed wetlands as a potential management
practice for cold climate dairy effluent treatment – VT, USA,” CATENA, vol. 135,
pp. 184–192, December 2015.
Vázquez, M. A., de la Varga, D., Plana, R. and Soto, M. “Vertical flow constructed wetland
treating high strength wastewater from swine slurry composting,” Ecological Engineering,
vol. 50, pp. 37–43, January 2013.
Vymazal, J. “Constructed wetlands for treatment of industrial wastewaters: A review,”
11
Treatment of industrial effluent
in constructed wetlands
Introduction
The concentrations of contaminants in industrial wastewater are generally so high
that some form of pretreatment, such as anaerobic digestion, is necessary. Although
still above acceptable discharge levels, most contaminants can then be reduced in
constructed wetlands to meet regulatory standards. This treatment sequence, or even
one that includes a secondary treatment process, is often an economic alternative to
conventional treatment technologies. Wastewater from beverage and food processing,
manufacturing, pulp and paper mills, petroleum refineries and other industrial
enterprises differ in content and concentration from municipal sewage or stormwater.
Each industrial sector presents a different set of treatment challenges.
A certain group of industries produce highly biodegradable waste. These include
alcohol fermentation (distilleries, wineries and breweries), food processing (potatoes,
milk, cheese and olives,) sugar refineries (cane and beets), fish processing, meat
processing and slaughterhouses. A general measure of biodegradability is the initial
wastewater ratio of biological oxygen demand (BOD) to chemical oxygen demand
(COD). When BOD/COD is more than 0.5, then the wastewater is considered easily
biodegradable. For the industries listed above, the BOD/COD ratio ranges from 0.6
to 0.8. A second group of industries produce wastewater that is less biodegradable.
These include the petrochemical, pulp and paper, and a range of other manufacturing
industries (Vymazal, 2014).
Various constructed wetland types are used to treat all of these effluent types.
Horizontal subsurface flow wetlands are most frequently implemented due to their
capacity to tolerate high organic loading, treatment of effluent colour and where anoxic
conditions are necessary to degrade contaminants. Vertical subsurface flow constructed
wetlands are necessary where aerobic conditions are necessary, such as in the
conversion of ammonia to nitrate. Free water surface wetlands are effective for
effluent high in suspended solids, oil and grease (Wu et al., 2015). The examples of
constructed wetlands offered below illustrate sustainable development, first, because
they treat wastewater on-site rather than deferring this to another time or pushing
the pollution downstream. Second, the wetlands offer secondary benefits not normally
associated with industrial enterprises. Third, they are economical to operate.
TREATMENT OF INDUSTRIAL EFFLUENT 255
Petroleum industry
Crude oil, natural gas and oil sand processing generate wastewater containing organic
compounds, oil and grease, suspended solids, phenols, hydrogen sulphides, salts and
metals. The hydrocarbons present in the effluent include BTEX (benzene, toluene,
ethylbenzene and xylenes), gasoline and diesel organic compounds. Total petroleum
hydrocarbons are a common measure of contaminants from petroleum refining
operations (Vymazal, 2014). The concentration and type of contaminant varies greatly
depending on the industry sector and refining process.
There are many examples of free water surface and horizontal subsurface flow
wetlands used to treat petroleum effluent, but often a hybrid system is employed. For
example, a treatment sequence of an aeration cascade, a free water surface wetland
(1.5 acres, 0.6 hectares) and a horizontal subsurface flow wetland (3.2 acres, 1.3
hectares) has been used in Casper, Wyoming, since 2008 to treat groundwater
contaminated from eight decades of petroleum refining activities (Figure 11.1). The
hybrid wetland treats about 700,000 gallons of groundwater per day and reduces
benzene (170 ␮m/L), BTEX (470 ␮m/L) and gasoline-derived organics (2,020 ␮m/L)
to below detection levels (Wallace, Schmidt and Agency, 2011).
This case is especially noteworthy, as it is located on a golf course and business
park redeveloped on the brownfield of the previous petroleum refinery. The free water
surface wetland is planted with cattail and bulrush, and allows iron to precipitate
after it has been oxidized in the belowground aeration cascade. The cascade is within
a pipe embedded in a hillside slope. The wetland pools with their lush vegetation
become part of the beautifully landscaped golf course (Figure 11.2). The horizontal
subsurface flow wetland includes a forced aeration system and is insulated by a
6-inch layer (15 cm) of wood mulch to prevent the water in the filter media from
freezing in the very cold (to –31°F, –35°C) winters. The horizontal flow wetland is
planted with a variety of wetland species (Figure 11.3) (Wallace, Schmidt and Agency,
2011). It is expected to take at least 50 years to remove the estimated 30 million gallons
of hydrocarbons that leached into the aquifer during the operation of the refinery.
The constructed wetland cost $3.4 million, but compared to other treatment options
it saved $12.5 million in construction cost and will result in an additional $15.7 million
in operating cost savings over the course of the remediation project (Natural Systems
Utilities, 2015).
Figure 11.1
The treatment wetlands are
embedded within the golf
course in Casper, Wyoming.
The free water surface
wetlands are along the bottom
and left of the image, while the
round shapes at the top are the
aerated horizontal subsurface
flow wetlands.
42°50′21.00″ N 106°21′14.90″ W,
image date: 17 August 2012,
image accessed: 15 July 2015.
256 TREATMENT OF INDUSTRIAL EFFLUENT
Figure 11.2
This groundwater remediation
project in Casper is embedded
within a Robert Trent Jones
golf course. The difficulty and
cost of cleaning polluted
groundwater is higher than
treating industrial wastewater
on-site, before it contaminates
the environment. However, this
project allowed sustainable
redevelopment to occur on this
brownfield site.
Photos: Gary Austin, 2015.
Figure 11.3
The horizontal subsurface flow
wetland (with aeration and
mulch insulation) is the last
treatment step in the
hydrocarbon removal project in
Casper, Wyoming.
Photos: Gary Austin, 2015.
An oil refinery in Mandan, North Dakota, uses a lagoon, free water surface
wetlands and vegetated swales to treat its industrial wastewater and to enhance wildlife
habitat (Figure 11.4). The plant uses 1,500,000 gallons (5,700,000 litres) of water
from the Missouri River each day for process purposes. After use, the contaminated
wastewater flows from an oil/water separator to an aerated lagoon and then through
a series of six ponds and vegetated swales. Five additional ponds (47 acres, 19.1
hectares) are used infrequently for storage after heavy rainstorms or snowstorms or
diversions from the primary ponds when a water quality problem is encountered. The
flexibility to bring the additional five ponds into the treatment sequence prevents water
quality permit violations when water is discharged into the river.
The additional ponds are managed for wildlife benefit, primarily, but are needed
as pollution control since stormwater run-off is not separated from process wastewater.
Figure 11.4
Mandan refinery sequence of
ponds and swales assures
water quality compliance while
providing excellent wildlife
habitat through revegetation
and creation of nesting islands.
Photo: Google Earth.
46°51′28.92″ N 100°53′10.04″ W,
image date: 15 September 2014,
image accessed: 14 July 2015.
Nesting islands were constructed within the ponds and 50,000 trees and shrubs were
planted. This resulted in 184 species of plants. Houses and feeders for ducks, swallows
and purple martins were installed. The ponds were stocked with fish, which have been
monitored and reveal no health problems. The state fish and game department has
released Giant Canada Geese and the rearing of goslings has been successful at the
site. Of the 191 bird species identified using the project area, 51 species nest there.
Deer, fox, badger, skunk and raccoons are now common. In fact, the project received
an award from the National Wildlife Federation, demonstrating a sustainable
development configuration (Baker and Schatz, 1989).
The free water surface system at the Mandan refinery effectively reduces BOD
(98 per cent), COD (93 per cent), ammonia (84 per cent), sulphides (100 per cent),
phenols (99 per cent) and oils and grease (99 per cent) in the wastewater. The hydraulic
loading rate is 1.2 cm per day (Vymazal, 2014). A 0.5 mile (0.8 km) vegetated canal
connecting the initial lagoon to the first pond is responsible for most of the reduction
of heavy metals in the wastewater (Baker and Schatz, 1989).
Another example of the effective use of constructed wetlands for the sustainable
development and operation of petroleum industry facilities is in Oman. The free water
surface constructed wetlands in Oman are possibly the largest commercial treatment
wetlands in the world (Figure 11.5). Here 1,433 acres (360 hectares) and an equal
area of evaporation ponds are used to remove hydrocarbon from oil field production
water.
When the Oman environmental protection agency enacted regulations prohibiting
the injection of oil production wastewater into the shallow aquifer, the oil industry
sought alternatives to the energy-intensive option of injection into the deep aquifer.
The constructed wetland alternative requires only 2 per cent of the energy required
by deep injection. Each day, the wetlands receive over 25 million gallons (95,000 m3)
of water after residual oil is removed in an oil/water separator (Figure 11.6). More
than 2 million Phragmites plants were planted in the wetland cells.
The wetlands reduce the hydrocarbon concentration to less than 0.5 mg/L. The
secondary benefits of the project are oil recovery, the avoided cost of deep injection,
generation of a commercially viable salt industry and bird habitat used by over
100 species. The large wetland is now an important resource for migratory birds
Figure 11.5 Wastewater treatment wetland in Oman.

Photo: Google Earth, 18°39′51.60″ N 55°47′14.08″ E, image date: 8 January 2015, image accessed: 17 July
2015.
Figure 11.6
Oman oil recovery and
wastewater remediation.
Photo: Courtesy of BAUER
Group. www.bauer.de/en/
press/imagearchive/resources.
html and www.bauer.de/en/bre/
(Figure 11.7). The low concentration of nitrogen and phosphorus and the high
concentration of salt and boron content of the wastewater may be stressing the plants
and may require the addition of artificial fertilizers in the future (Wu et al., 2015).
Wineries and breweries

Winery effluent is the result of crushing waste and washing equipment and bottles.
The effluents are characterized by intermittent flow and variable contaminant
Figure 11.7 The artificial wetland has become an important resource for resident and migratory bird
species, Oman.
Photo: Courtesy of BAUER Group, www.bauer.de/en/press/image_archive/resources.html and
www.bauer.de/en/bre/.
concentration (Wu et al., 2015), but the wastewater includes alcohols. There is a large
quantity of readily biodegradable organic matter in the wastewater, including sugars
(glucose and fructose), alcohols (ethanol and glycerol), acids (tartaric, lactic and acetic)
and suspended solids. However, there are also substances that are difficult to degrade
including polyphenols, tannins and lignin (Vymazal, 2014).
Concentrations of COD at 5,000 mg/L or less allow the use of horizontal or vertical
subsurface flow wetlands for the treatment of winery wastewater. This concentration
of COD is often exceeded, especially at large winery operations. Therefore, the raw
wastewater must be diluted by recirculating a portion of the treated effluent to bring
the concentration of COD down from as much as 17,000 mg/L. Alternatively, the
wastewater can be pretreated in various ways including anaerobic digestion before it
enters the constructed wetlands (Wu et al., 2015). Horizontal subsurface flow wetlands
are most common in this industry due to their simple operation, but vertical subsurface
flow wetlands are also effective. Hybrid wetlands are necessary for advanced treatment
and removal of the broadest range of contaminants and excess nutrients.
An example of a constructed wetland for treatment of winery wastes is a system
in Pontevedra, Spain. The treatment system included human wastewater from the
associated agricultural tourism business as well as winery wastewater.
The treatment sequence began with an anaerobic pretreatment step using a 1,585
gallon (6 m3) hydrolytic up-flow sludge bed digester. The anaerobic step was primarily
for the reduction of total suspended solids (TSS). The discharge from the digester went
to a siphon tank and then to a 538 square feet (50 m2) vertical subsurface flow wetland
planted with Phragmites australis. Effluent from the wetland was divided for final
treatment by three 1,076 square feet (100 m2) horizontal flow constructed wetlands
planted with Juncus effusus.
Water entering the vertical flow wetland contained TSS at 72–172 mg/L and BOD
ranging from 216 and 1,379 mg/L and COD ranging between 422 and 2,178 mg/L.
The average hydraulic loading rate was 19.5 mm per day, and the average organic
loading rate was 30.4 grams of COD and 18.4 grams of BOD per m2 per day. Despite
the high organic loading, the wetland performed well. The percentage removal for
the contaminants was: 86.8 per cent TSS, 73.3 per cent COD, 74.2 per cent BOD5,
52.4 per cent total Kjeldhal nitrogen ((TKN) organic nitrogen and ammonia), 55.4
per cent ammonia (NH3) and 17.4 per cent phosphates.
As with winery effluents, brewery wastewater has high concentrations of sugars,
alcohol and suspended solids, but the pH is often very low (Vymazal, 2014).
Distilleries
Spent wash water from distilleries is high in organic content (BOD: 35,000–60,000
mg/L; COD: 60,000–120,000 mg/L) and dark colour (Wu et al., 2015). The
contaminants and unpleasant odour must be reduced before the effluent is discharged
into the environment. Generally, anaerobic reactors are used to treat the waste but
aerobic processes have also been employed (Vymazal, 2014).
Potato processing
Washing, peeling, slicing and precooking potatoes generates large quantities of
wastewater high in starch and suspended solids. Constructed wetlands have been used
to remove the highly biodegradable contaminants. Horizontal subsurface flow
wetlands and hybrid wetlands have been used to treat potato-processing wastewater.
A hybrid wetland consisting of a sequence of free water surface (24.7 acres, 10
hectares), vertical subsurface flow (9.8 acres, 4 hectares) and horizontal subsurface
flow (6 acres, 2 ha) wetland cells was designed to treat the wastewater (Wu et al.,
2015). This approach was very successful in removing organic nitrogen (80 per cent
removal), but ammonia (47 per cent removal) and nitrates remained in the discharged
effluent. However, this met the system goal and allowed the effluent to be used for
irrigation. A hybrid wetland in Japan successfully removed the following contaminants:
COD (94 per cent removal), total nitrogen (90 per cent removal), ammonia (72 per
cent removal) and total phosphorus (93 per cent removal). This system had three
vertical subsurface flow cells, a horizontal subsurface flow cell and a final vertical
subsurface flow cell in series (Vymazal, 2014).
Milking parlours and cheese production

Effluent from milk processing and cleaning contains high concentrations of COD, TSS
and phosphorus, and oil and grease from milk solids (carbohydrates, proteins) and
fat (Wu et al., 2015). Cheese whey is the wastewater that is generated when making
cheese. It is composed of urea, uric, lactic and citric acids, lactose, proteins, lipids and
mineral salts. These are combined with sterilization wash water and typically treated
in an up-flow sludge blanket reactor (Vymazal, 2014). The chapter on agricultural
applications of constructed wetlands reported the performance of constructed wetlands
treating dairy effluent including milking parlour content.
Tanneries
Processing animal skins with either tannins or chromium creates effluents high in salts
(as high as 80 grams of NaCl per litre), suspended solids, ammonia, organic nitrogen
and chromium (up to 50 mg/L) or tannins (up to 3,000 mg/L) (Vymazal, 2014). In a
study of horizontal subsurface flow wetlands for the treatment of tannery wastewater,
Table 11.1 Summary of tannery effluent treatment performance at the hydraulic loading rate of 6 cm
per day and a hydraulic residency time of 7 days (Calheiros, Rangel and Castro, 2009)
Parameter Concentration Removal Loading Mass

(mg/L) (%) (kg ha day) removal
In Out
COD 1,598 252 83 479 404
BOD5 706 159 77 212 164
TSS 80 9 88 24 21
TKN–N 122 55 54 37 20
NH3–N 81 42 48 24 13
TP 0.41 0.27 38 0.09 0.04
influent levels of 420–1,000 mg/L of BOD were reduced as much as 88 per cent and
808–2,449 mg/L of COD were reduced as much as 92 per cent (Table 11.1). The
treatment system began with wastewater collected in an equalization tank followed
by a tank for coagulation, flocculation and sedimentation. Then, the wastewater was
delivered to two horizontal wetland cells operated in series and planted with
Phragmites australis and Typha latifolia. Both species grew well in the wastewater,
although the Phragmites became established and expanded more readily (Calheiros,
Rangel and Castro, 2009).
The study resulted in the recommendation of a maximum organic loading of 1,153
pounds per acre (212 kg per hectare) per day and a 7-day hydraulic retention time in
order to meet tannery discharge standards (Portugal). The wetlands reduced sulphates
by an average of 56 per cent and colour by an average of 60 per cent at a hydraulic
loading rate of 6 cm per day and a hydraulic residence time of 7 days. Nitrate was
reduced about 50 per cent, decreasing from 43 mg/L to 21–24 mg/L (Calheiros, Rangel
and Castro, 2009).
Chemical and textile industry case study

Chemical manufacturing companies produce wastewaters that vary greatly in toxicity
and in the kind of contaminants. A 15-acre (6 hectare) hybrid wetland (Figure 11.8)
to treat effluent from chemical and textile manufacturing enterprises in the Changshu
Advanced Materials Industrial Park in China was completed in 2014. Designed by
Blumberg Engineers, the project is located in Changshu, Jiangsu Province. The wetland
is designed to treat 1.1 million gallons (4,000 m3) of pretreated effluent per day in
order to protect the Yangtze River adjacent to the industrial development (Blumberg
Engineers, 2014).
The first and largest stage of the hybrid treatment wetland is a vertical subsurface
flow wetland (Figure 11.9) (4.7 acres, 18,900 m2) arranged in a semicircle (Figure
11.8–A). The second treatment stage has three cells operated in a series. The first cell
is a deep pond (0.65 acres, 2,652 m2), and the second cell is an artificial rocky creek
(0.3 acres, 1,064 m2) (Figure 11.8–B, C, D). The last cell is a shallow, free water surface
wetland (1 acre, 4,019 m2). Stage three (Figure 11.10) consists of a horizontal
subsurface flow wetland (2.5 acres, 10,062 m2) (Blumberg Engineers, 2014).
Although the inflow is pretreated, it remains highly polluted including high levels
of salinity and fluoride. The vertical subsurface flow wetland transforms organic
compounds and nitrifies the effluent. The pond, stream cascade and shallow marsh
Figure 11.8 The treatment wetland during construction. The hybrid wetland is composed of the arcs
of the vertical flow stage (A), the pond, stream and marsh sequence to the left and bottom of the site.
The horizontal subsurface flow stage (E) is shown on the right side of the image. In this image, the
industrial context is clear, but new moderate density housing is also visible to the south. The Yangtze
River is 1,000 feet (305 m) to the north. A – Vertical subsurface flow wetland; B – Pond; C – Rocky
creek; D – Free water surface wetland; E – Horizontal subsurface flow wetland.
Photo: Google Earth (CNES, Astrium), 31°46′48.30″ N 120°49′04.46″ E, image date: 11 December 2013,
image accessed: 2 August 2015.
Figure 11.9 In this view toward the northeast, the first stage of the treatment wetland is organized as
concentric bands of vertical flow subsurface wetland with a variety of plants. The industrial towers in
the background include pretreatment of the process water, before it is distributed across the vertical
flow wetland.
Photo: Blumberg Engineers, www.blumberg-engineers.com.
Figure 11.10 The horizontal subsurface wetlands feature Cattail (Typha). The visitor centre is visible
in the background. The attractive design, wildlife habitat and visitor centre are valuable secondary
benefits of the project. The reuse of wetland effluent by the chemical plants is another benefit that
reduces the water demand of the factories.
Photo: Blumberg Engineers, www.blumberg-engineers.com.
allow for precipitation and settling of phosphorus. In addition, the shallow marsh
decreases COD levels. The horizontal flow wetland is provided for denitrification and
other anaerobic processes to improve the water quality. The design of the wetlands
allows for recirculation and by-pass depending on the treatment needs (Blumberg
Engineers, 2014).
The advanced treatment of the industrial effluent is necessary to protect the, already
polluted, Yangtze River from further degradation. The proximity of the treatment
wetlands to the river increases its value as replacement habitat for historical riverbank
habitat losses.
Pulp and paper mills

Wood pulping and paper production generates tremendous amounts of wastewater.
Modern mills require about 3,400 gallons (13 m3) of water to produce one ton of
paper (Kamali and Khodaparast, 2015). The wastewater is characterized by high
concentrations of BOD and COD. A mechanical-chemical process can generate more
than 7,000 mg/L of COD and 3,000 mg/L of BOD. Other contaminants include
suspended solids, volatile organic compounds (terpens, phenols, chloroform and
methanol), fatty acids, lignin, and adsorbable organic halogens or resins (Vymazal,
2014). Adsorbable organic compounds include as many as 500 different substances
associated with the use of chlorine to bleach pulp and paper. Many of these
compounds are not readily biodegradable. Which contaminants are contained in the
wastewater is influenced by whether the pulping process is mechanical or mechanical-
chemical. The chemical process uses chemicals such as sodium hydroxide and sodium
sulphide. As many as 700 organic and inorganic contaminants may be present in the
raw wastewater. Although all pulp and paper mills treat their effluent, they remain
one of the major sources of water pollution.
The high concentration of contaminants requires pretreatment before the
wastewater can be directed to constructed wetland for additional treatment. An
anaerobic digester is often used as a pretreatment step, as this effectively removes

organic material with little energy input and results in methane gas that can be used
as an energy resource by the factory. Aerobic treatment, such as activated sludge
technology, is also common. Physiochemical methods are emerging to address some
of the most recalcitrant contaminants.
Free water surface and horizontal subsurface flow wetlands have been used to
provide secondary or tertiary treatment of pulp and paper wastewater. A pilot
horizontal subsurface flow wetland demonstrated a reduction in COD from 1,083 to
98 mg/L with an 8.6-day hydraulic retention time. The colour of the effluent was also
dramatically reduced. The Canna indica planted in the wetland grew prolifically
(Choudhary et al., 2011). There are several process subsystems involved in the pulp
and paper industry, some of which are well suited to wetland wastewater treatment.
Two are the treatment of stormwater and the treatment of leachate from log and chip
piles. Treating these effluents separate from mill process water reduces the volume of
water that must be accommodated by the systems that treat process water.
Thermoelectric power generation

As noted in the first chapter, the thermoelectric power generation industry in the United
States uses more water than any other sector of the economy. Constructed wetlands
allow the reuse of water by this industry after the water is treated. Thermoelectric
power plants use about 90 per cent of their water for cooling, but also use water to
manage and transport waste ash and to operate air pollution scrubbers. This section
presents constructed wetlands that treat wastewater from a coal-fired power plant air
pollution reduction system. According to the US Environmental Protection Agency
(EPA), pollutant discharges from the thermoelectric power industry are increasing in
volume and total mass. Steam electric power plants are responsible for approximately
50–60 per cent of all toxic pollutants discharged into surface waters by all industries
regulated under the Clean Water Act. A rule proposed by the US EPA would limit
the concentration of toxic metals and nitrate in the effluent from these energy plants
(US Environmental Protection Agency, 2013). These metals can be removed from the
wastewater by constructed wetlands.
Jeffrey Energy Centre case study

Removing the air pollutants with the use of scrubbers creates contaminated wastewater
that requires treatment before it can be discharged to the natural environment or reused.
The Jeffrey Energy Centre near St Mary’s Kansas, USA, implemented air pollution
scrubbers for the stacks of its coal-fired power plant. This reduced sulphur dioxide
emissions by 97 per cent (Reitenbach, 2014).
Constructed treatment wetland

The wastewater from the air pollution scrubbers contains mercury, selenium, boron,
fluoride, sulphate and low levels of chlorine. A pilot and then full-scale wetlands (Figure
11.11) were constructed as the lowest cost treatment alternative to reduce these
contaminants. Sulphate was not reduced sufficiently in the pilot wetland. Therefore,
it was removed through a pretreatment precipitation process instead.
Figure 11.11 Full-scale wetland treatment of effluent from a coal-burning power plant in Kansas.
The foreground ponds are equalization ponds, the eight cells in the centre of the image are vertical
subsurface up-flow wetlands and the upper two cells are horizontal subsurface flow wetlands.
Photo: Google Earth 39°16′33.04″ N 96°07′50.87″ W, image accessed: 14 July 2015.
Table 11.2 Hybrid wetland and the removal of metals (Bland, Snider and Haag, 2014)
Constituent Mass load reduction (%) Average influent (ug/L) Average effluent (ug/L)
Mercury 83 1.2 <0.2
Selenium 92 114 12
Boron 42 3,459 2,120
Fluoride 93 11,209 1,551
The 2-acre pilot study (2010) provided positive results. The pilot wetlands
included free surface water with floating mats of wetland vegetation, vertical
subsurface flow with a soil media and horizontal subsurface flow with a gravel filter
below a thin soil layer. Each type of treatment cell was monitored. An engineered soil
in the vertical subsurface flow wetland was key to the removal of many of the target
contaminants. This soil is a mix of topsoil (25 per cent), clay (10 per cent), sand (40
per cent) and compost (25 per cent). Fertilizer, lime and inoculants augment these
constituents. The use of a soil media differs from the sand or gravel media used in
the treatment of municipal wastewater or industrial wastewater high in organic
nitrogen. Use of the soil media was preferable in this case because of the high filtration
capacity. The pilot wetlands resulted in good removal of metals (Table 11.2).
Full-scale hybrid wetland

Based on the pilot study, vertical flow subsurface constructed wetlands were selected
as the primary treatment strategy, with horizontal subsurface flow wetlands as a
polishing step.
The full-scale wetland including a pretreatment stage for sulphur precipitation
was constructed in 2012–13. The wetland includes two equalization ponds (4.5 acres)
to manage the flow rate into the wetland. These are followed by eight treatment cells
(19.2 acres) operated as two parallel sequences of four stages of vertical up-flow
subsurface wetland cells. The four vertical subsurface flow stages precede two
horizontal subsurface flow cells (4.5 acres) that operate in series. The wetland treats
18,000 gallons of effluent from the plant each day, but this is mixed with an equal
amount of freshwater from the river to reduce the concentration of the contaminants.
The hydraulic residence time is more than 30 days (Bland, Snider and Haag, 2014).
The down-flow distribution of water used in the vertical flow wetlands in the pilot
study was replaced with an up-flow option to avoid the concentration metals (selenium
and mercury) on the surface that could impact wildlife (Reitenbach, 2014).
The design of the vertical up-flow wetland begins with a layer of soil formulated
to support wetland plants (Figure 11.12). Below this is a layer of gravel with embedded
perforated pipes to serve as the filter bed outlet for treated effluent. The main filter
bed is composed of a mixture of topsoil clay, sand, compost, fertilizer, lime and
inoculants. The bottom layer is another gravel layer that serves as the inlet to the filter
bed. The wetland is sealed by a 60 mil high-density polyethylene (HDPE) liner over
a compacted clay base.
As with the pilot wetland, the full-scale treatment system effectively treats
the following metals and excess nutrients: mercury, selenium, chromium, iron,
molybdenum, boron, aluminium, fluoride, nitrogen, potassium and barium (Bland,
Snider and Haag, 2014). Data on the performance of the full-scale wetland is still
being collected and analyzed, but the preliminary results are promising. Although not
a process used at the Jeffrey Energy Centre, constructed wetlands have also been used
to treat leachate from bottom ash landfills and impoundments (US Environmental
Protection Agency, 2013).
Figure 11.12 Section of the vertical subsurface up-flow wetland. A – Wetland soil; B – Sand and
gravel layer; C – Perforated drain pipe; D – Filter soil; E – Coarse gravel; F – High-density polyethylene
liner; G – Compacted clay base; H – Native soil.
Image: Gary Austin adapted from Bland (Bland, Snider and Haag, 2014).
The constructed wetlands produce economic, recreation and habitat secondary

benefits. The energy company estimates that over 15 years, the wetlands will save $40
million in capital and operating costs (EEI, 2014). The treated effluent is reused in
the power plant or feeds three free surface water wetlands open to the public for
recreational use. The wetland vegetation in the subsurface wetlands provides some
ecosystem benefits, but the open water wetlands fed by the treated water provide a
more robust habitat.
Conclusion
Wastewater from many other industries including slaughterhouses, seafood processing,
starch production, coffee processing, sugar and molasses refineries, coke production
and mines treating acid drainage have all used constructed wetlands to successfully
treat effluent before reuse or discharge to the environment. As with the examples
presented above, these enterprises have been able to reduce man-made and more natural
but excessive organic and inorganic compounds through the physical and biological
treatment capacity of hybrid wetlands requiring little energy and low operation cost.
While landscape architects and planners do not often have the opportunity to address
pollution mitigation issues within the industrial sector, environmental engineers and
agency regulators can use the examples presented in this and previous chapters to
encourage this accessible technology as a tool toward creating more sustainable
development and ecosystem health.
References
Baker D. and Schatz, D. “Constructed wetlands for wastewater treatment at Amoco Oil
Company’s Mandan, North Dakota Refinery,” in Constructed wetlands for wastewater
treatment: Municipal, industrial and agricultural. Chattanooga, TN: CRC Press, Lewis
Publisher, 1989, pp. 233–237.
Bland, K., Snider, C. and Haag, D. “Selenium reduction in constructed wetland treatment
systems: Naturally attenuating problematic pollutants. Ecological and ecosystem
restoration,” Presented at the CEER, 2014.
Blumberg Engineers, “Treatment wetlands cascade for industrial wastewater at Wetland
Treatment Center of Advanced Materials Changshu Industrial Park, China,” Blumberg
Engineers, 2014.
Calheiros, C. S. C., Rangel, A. O. S. S. and Castro, P. M. L. “Treatment of industrial wastewater
with two-stage constructed wetlands planted with Typha latifolia and Phragmites australis,”
Bioresource Technology, vol. 100, no. 13, pp. 3205–3213, July 2009.
Choudhary, A., Kumar, S., Sharma, C. and Kumar, P. “Performance of constructed wetland
for the treatment of pulp and paper mill wastewater,” Presented at the World Environmental
and Water Resources Congress 2011, Palm Springs, California, 2011.
EEI (Edison Electric Institute), “Westar Energy Awarded EEI’s Edison Award,” 2014. [Online].
Available: www.eei.org/resourcesandmedia/newsroom/Pages/Press%20Releases/Westar%
20Energy%20Awarded%20EEI%E2%80%99s%202014%20Edison%20Award.aspx.
[Accessed: 20 July 2013].
Kamali M. and Khodaparast, Z. “Review on recent developments on pulp and paper mill
wastewater treatment,” Ecotoxicology Environmental Safety, vol. 114, pp. 326–342, April
2015.
Natural Systems Utilities, “British Petroleum, Casper, WY,” 2015. [Online]. Available:
www.naturalsystemsutilities.com/british-petroleum/. [Accessed: 15 July 2015].
Reitenbach, G. “Water Award: Jeffrey Energy Center’s Constructed Wetland Treatment

System.” 2014. [Online]. Available: www.powermag.com/jeffrey-energy-centers-constructed-
wetland-treatment-system. [Accessed: 12 July 2016].
US Environmental Protection Agency, “Effluent limitations guidelines and standards for the
steam electric power generating point source category,” Federal Register, vol. 78, no. 110,
pp. 34432–34543, June 2013.
Vymazal, J. “Constructed wetlands for treatment of industrial wastewaters: A review,”
Wallace, S., Schmidt, M. and Agency, E. P. “Long term hydrocarbon removal using treatment
wetlands,” Society of Petroleum Engineers, vol. 145, no. 797, pp. 1–10, 2011.
Wu, S., Wallace, S., Brix, H., Kuschk, P., Kirui, W. K., Masi, F. and Dong, R. “Treatment of
industrial effluents in constructed wetlands: Challenges, operational strategies and overall
performance,” Environmental Pollution, vol. 201, pp. 107–120, June 2015.
Appendix A: Wetland plants
Lists of plants by continent

The plant species in the lists below have been reported as having been planted in free
water surface constructed wetlands in six regions of the world (Vymazal, 2013) and
in other constructed wetland types. Additional cultural information about these plants
is necessary before finalizing a plant palette for any sub region, as not all plants will
be suitable for every region due to cold hardiness, invasiveness and other factors. The
species listed are not necessarily plants native to the region. Contacting regional native
plant nurseries for availability or, preferably, specifying locally collected seeds and
plants will lead to the best match with the native regional flora.
Table A.A.1 Plants by continent
North America; adapted from Vymazal (2013)
Acorus calamus Eleocharis fallax Sagittaria latifolia

Acorus gramineus Eleocharis obtusa Schoenoplectus mucronatus
Alisma plantago-aquatica Eleocharis palustris Scirpus acutus
Alisma triviale Eleocharis quadrangulata Scirpus americanus
Alternanthera philoxeroides Epilobium brachycarpum Scirpus atrovirens
Bolboschoenus maritimus Equisetum plaustre Scirpus californicus
Butomus umbellatus Festuca angustifolia Scirpus cyperinus
Canna flaccida Glyceria maxima Scirpus fluviatilis
Carex aquatilis Iris versicolor Scirpus pungens
Carex cristella Iris virginica Scirpus tabernaemontani
Carex frankii Juncus balticus Scirpus validus
Carex lacustris Juncus effusus Setaria glauca
Carex lurida Juncus roemerianus Solanum dulcamara
Carex nebrascensis Leerzia oryzoides Sorghum halapense
Carex normalis Ludwigia alternifolia Sparganium americanum
Carex obnupta Ludwigia peploides Sparganium androcladium
Carex rostrata Mentha aquatica Sparganium emersum
Colocasia esculenta Panicum hemitomon Sparganium eurycarpum
Cynodon dactylon Panicum repens Spartina pectinata
Cyperus eragrostis Panicum virgatum Thalia geniculata
Cyperus erythrorhyzos Peltandra virginica Typha angustifolia
Cyperus globulosus Phalaris arundinacea Typha domingensis
Cyperus iria Phragmites australis Typha latifolia
Cyperus retorsus Polygonum lapathifolium Zizania aquatica
Digitaria ischaemum Pontederia cordata Zizaniopsis miliacea
Echinochloa cruss-galli Rumex crispus
Echinochloa walteri Sagittaria graminea
Eleocharis dulcis Sagittaria lancifolia
270 APPENDIX A
Europe; adapted from Vymazal (2013)
Alisma plantago-aquatica Glyceria fluitans Phalaris arundinacea

Alopecurus aequalis Glyceria maxima Phragmites australis
Arundo donax Hymenocallis littoralis Poa sp.
Bidens sp. Iris pseudacorus Scirpus lacustris
Bolboschoenus maritimus Juncus conglomeratus Scirpus palla
Carex aquatilis Juncus effusus Scirpus sylvatiscus
Carex elata Juncus filliformis Scutellaria galerigulata
Carex riparia Lysimachia thyrsiflora Sparganium erectum
Carex sp. Lysimachia vulgaris Typha angustifolia
Eleocharis palustris Lythrum salicaria Typha latifolia
Filipendula ulmaria Mentha aquatica
Asia; adapted from Vymazal (2013)
Acorus calamus Juncus articulatus Scirpus cyperinus

Acorus gramineus Juncus effusus Scirpus grossus
Alisma plantago-aquatica Lepironia articulata Scirpus lacustris
Canna hybrid Miscanthus floridulus Typha angustifolia
Carex aquatilis Miscanthus sacchariflorus Typha domingensis
Carex rhynchophisa Miscanthus sinensis Typha latifolia
Cyperus alternifolius Oenanthe javanica Typha orientalis
Cyperus malaccensis Pennisetum alopecuroides Vetiveria zizanoides
Cyperus papyrus Phalaris arundinacea Zizania aquatica
Digitaria bicornis Phragmites australis Zizania caduciflora
Eleocharis dulcis Phragmites japonica Zizania latifolia
Hygrophila progonocalyx Phragmites karka
Iris pseudacorus Schoenoplectus mucronatus
Australia and South Pacific; adapted from Vymazal (2013)
Alisma plantago-aquatica Eleocharis sphacelata Persicaria sp.

Baumea articulata Glyceria declinata Phragmites australis
Baumea rubiginosa Glyceria maxima Phylidrum lanuginosum
Bolboschoenus fluviatilis Iris pseudacorus Polygonum lapathifolium
Bolboschoenus medianus Juncus articulatus Schoenoplectus mucronatus
Carex apressa Juncus ingens Scirpus tabernaemontani
Cyperus polystachys Juncus kraussii Scirpus validus
Echinochloa cruss-galli Juncus usitatus Typha domingensis
Eleocharis acuta Lepironia articulata Typha latifolia
Eleocharis acuta Paspalum distichum Typha orientalis
Eleocharis dulcis
Africa; adapted from Vymazal (2013)
Canna flaccida Glyceria maxima Pontederia cordata

Carex riparia Iris pseudacorus Scirpus californicus
Cyperus papyrus Phragmites australis Spartina pectinata
Echinochloa pyramidalis Phragmites mauritianus Typha latifolia
Central and South America; adapted from Vymazal (2013)
Agapanthus africanus Scirpus californicus Typha domingensis

Phragmites australis Strelitzia reginae Zantedeschia aethiopica
Pontederia cordata Typha angustifolia
Appendix B: Plants and aquatic
organisms for Houtan Park
Table A.B.1 Plant species for Houtan Park, Shanghai China, USDA Hardiness Zone 10
(Dong Yue et al., 2013)
Acorus calamus Nelumbo nucifera

Arundo donax Nymphaea tetragona
Canna indica Phragmites australis
Ceratophyllum demersum Pistia stratiotes
Colocasia tonoimo Pontederia cordata
Cyperus alternifolius Potamogeton crispus
Elodea nuttallii Potamogeton maackianus
Hydrilla verticillata Potamogeton malaianus
Hydrocotyle vulgaris Potamogeton pectinatus
Iris tectorum Scirpus validus
Lythrum salicaria Thalia dealbata
Miscanthus sacchariflorus Typha angustata
Myriophyllum spicatum Vallisneria natans
Myriophyllum verticillatum Zizania latifolia
Table A.B.2 Aquatic species for Houtan Park, Shanghai China (Dong Yue et al., 2013)
Anodonta sp. Marcrobrachium nipponense

Aristichthys nobilis Megalobrama skolkovii
Bellamya manhungensis Mylopharyngodon piceus
Corbicula fluminea Myxocyprinus asiaticus
Ctenopharyngodon idellus Neocaridina denticulata sinensis
Erythroculter ilishaeformis Ophicephalus argus
Gambusia affinis Oriental weatherfish
Hemibarbus maculatus Pelteobagrus fulvidraco
Hypophthalmichthys molitrix Plagiognathops microlepis
Hypostomus plecostomus Radix auricularia
Siniperca chuatsi
Appendix C: Stormwater and
soils data and bio-retention
plants
Table A.C.1 Run-off curve numbers for cultivated agricultural lands (USDA, 1986)
Cover description Curve numbers for hydrologic soil group

Cover type Treatment 2/ Hydraulic A B C D
condition 3/
Fallow Bare soil — 77 86 91 94
Crop residue cover (CR) Poor 76 85 90 93
Good 74 83 88 90
Row crops Straight row (SR) Poor 72 81 88 91
Good 67 78 85 89
SR + CR Poor 71 80 87 90
Good 64 75 82 85
Contoured (C) Poor 70 79 84 88
Good 65 75 82 86
C + CR Poor 69 78 83 87
Good 64 74 81 85
Contoured & terraced (C&T) Poor 66 74 80 82
Good 62 71 78 81
C&T + CR Poor 65 73 79 81
Good 61 70 77 80
Small grain SR Poor 65 76 84 88
Good 63 75 83 87
SR + CR Poor 64 75 83 86
Good 60 72 80 84
C Poor 63 74 82 85
Good 61 73 81 84
C + CR Poor 62 73 81 84
Good 60 72 80 83
C&T Poor 61 72 79 82
Good 59 70 78 81
C&T + CR Poor 60 71 78 81
Good 58 69 77 80
Close-seeded SR Poor 66 77 85 89
or broadcast Good 58 72 81 85
legumes or C Poor 64 75 83 85
rotation Good 55 69 78 83
meadow C&T Poor 63 73 80 83
Good 51 67 76 80
1 Average run-off condition, and Initial abstraction (Ia) = 0.2S

2 Crop residue cover applies only if residue is on at least 5 per cent of the surface throughout the year.
3 Hydraulic condition is based on combination factors that affect infiltration and run-off, including (a) density
and canopy of vegetative areas, (b) amount of year-round cover, (c) amount of grass or close-seeded legumes,
(d) per cent of residue cover on the land surface (good ł 20 per cent) and (e) degree of surface roughness.
Poor: Factors impair infiltration and tend to increase run-off.
Good: Factors encourage average and better than average infiltration and tend to decrease run-off.
APPENDIX C 273
Table A.C.2 Run-off curve numbers for other agricultural lands (USDA, 1986)
Cover description Curve numbers for hydrologic soil group

Cover type Hydrologic A B C D
condition
Pasture, grassland, or range – continuous Poor 68 79 86 89
forage for grazing 2/ Fair 49 69 79 84
Good 39 61 74 80
Meadow – continuous grass, protected — 30 58 71 78
from grazing and generally mowed for hay.
Brush – brush-weed-grass mixture with Poor 48 67 77 83
brush the major element 3/ Fair 35 56 70 77
Good 30 4/ 48 65 73
Woods – grass combination (orchard or Poor 57 73 82 86
tree farm) 5/ Fair 43 65 76 82
Good 32 58 72 79
Woods 6/ Poor 45 66 77 83
Fair 36 60 73 79
Good 30 4/ 55 70 77
Farmsteads – buildings, lanes, driveways — 59 74 82 86
and surrounding lots
1 Average run-off condition, and Ia = 0.2S.
2 Poor: 75 per cent ground cover and lightly or only occasionally grazed.
3 Poor: 75 per cent ground cover.
4 Actual curve number is less than 30; use CN = 30 for run-off computations.
5 CNs shown were computed for areas with 50 per cent woods and 50 per cent grass (pasture) cover.
Other combinations of conditions may be computed from the CNs for woods and pasture.
6 Poor: Forest litter, small trees and brush are destroyed by heavy grazing or regular burning.
Fair: Woods are grazed but not burned, and some forest litter covers the soil.
Good: Woods are protected from grazing, and litter and brush adequately cover the soil.
Table A.C.3 Run-off curve numbers for arid and semiarid rangelands (USDA, 1986)
Cover description Hydrologic soil group

Cover type Hydrologic A 3/ B C D
condition 2/
Herbaceous – mixture of grass, weeds Poor 80 87 93
and low-growing brush, with brush the Fair 71 81 89
minor element Good 62 74 85
Oak-aspen – mountain brush mixture of Poor 66 74 79
oak brush, aspen, mountain mahogany, Fair 48 57 63
bitter brush, maple and other brush Good 30 41 48
Pinyon-juniper – pinyon, juniper or both; Poor 75 85 89
grass understory Fair 58 73 80
Good 41 61 71
Sagebrush with grass understory Poor 67 80 85
Fair 51 63 70
Good 35 47 55
Desert shrub – major plants include saltbush, Poor 63 77 85 88
greasewood, creosotebush, blackbrush, Fair 55 72 81 86
bursage, palo verde, mesquite and cactus Good 49 68 79 84
1 Average run-off condition, and Ia = 0.2S. For range in humid regions, use the table ‘Run-off curve numbers
for other agricultural lands’.
2 Poor: 70 per cent ground cover.
3 Curve numbers for group A have been developed only for desert shrub.
274 APPENDIX C
Table A.C.4 Run-off depth for selected CNs and rainfall amountsa (USDA, 1986)
Run-off depth for curve number of –

Rainfall 40 45 50 55 60 65 70 75 80 85 90 95 98
1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.08 0.17 0.32 0.56 0.79
1.2 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.07 0.15 0.27 0.46 0.74 0.99
1.4 0.00 0.00 0.00 0.00 0.00 0.02 0.06 0.13 0.24 0.39 0.61 0.92 1.18
1.6 0.00 0.00 0.00 0.00 0.01 0.05 0.11 0.20 0.34 0.52 0.76 1.11 1.38
1.8 0.00 0.00 0.00 0.00 0.03 0.09 0.17 0.29 0.44 0.65 0.93 1.29 1.58
2.0 0.00 0.00 0.00 0.02 0.06 0.14 0.24 0.38 0.56 0.80 1.09 1.48 1.77
2.5 0.00 0.00 0.02 0.08 0.17 0.30 0.46 0.65 0.89 1.18 1.53 1.96 2.27
3.0 0.00 0.02 0.09 0.19 0.33 0.51 0.71 0.96 1.25 1.59 1.98 2.45 2.77
3.5 0.02 0.08 0.20 0.35 0.53 0.75 1.01 1.30 1.64 2.02 2.45 2.94 3.27
4.0 0.06 0.18 0.33 0.53 0.76 1.03 1.33 1.67 2.04 2.46 2.92 3.43 3.77
4.5 0.14 0.30 0.50 0.74 1.02 1.33 1.67 2.05 2.46 2.91 3.40 3.92 4.26
5.0 0.24 0.44 0.69 0.98 1.30 1.65 2.04 2.45 2.89 3.37 3.88 4.42 4.76
6.0 0.50 0.80 1.14 1.52 1.92 2.35 2.81 3.28 3.78 4.30 4.85 5.41 5.76
7.0 0.84 1.24 1.68 2.12 2.60 3.10 3.62 4.15 4.69 5.25 5.82 6.41 6.76
8.0 1.25 1.74 2.25 2.78 3.33 3.89 4.46 5.04 5.63 6.21 6.81 7.40 7.76
9.0 1.71 2.29 2.88 3.49 4.10 4.72 5.33 5.95 6.57 7.18 7.79 8.40 8.76
10.0 2.23 2.89 3.56 4.23 4.90 5.56 6.22 6.88 7.52 8.16 8.78 9.40 9.76
11.0 2.78 3.52 4.26 5.00 5.72 6.43 7.13 7.81 8.48 9.13 9.77 10.39 10.76
12.0 3.38 4.19 5.00 5.79 6.56 7.32 8.05 8.76 9.45 10.11 10.76 11.39 11.76
13.0 4.00 4.89 5.76 6.61 7.42 8.21 8.98 9.71 10.42 11.10 11.76 12.39 12.76
14.0 4.65 5.62 6.55 7.44 8.30 9.12 9.91 10.67 11.39 12.08 12.75 13.39 13.76
15.0 5.33 6.36 7.35 8.29 9.19 10.04 10.85 11.63 12.37 13.07 13.74 14.39 14.76
a – Interpolate the values shown to obtain run-off depths for CNs or rainfall amounts not shown.
Figure A.C.1 Composite CN with connected impervious area (USDA, 1986).

APPENDIX C 275
Figure A.C.2 Composite CN with unconnected impervious areas and total impervious area less than
30 per cent (USDA, 1986).
Table A.C.5 Criteria for assignment of hydrologic soil groups when a water-impermeable layer exists
at a depth less than 100 cm (40 inches) (USDA, 2007)
Soil property Hydrologic soil group

A B C D
Saturated hydraulic >40.0 ␮m/s ≤40.0 to >10.0 ␮m/s ≤10.0 to >1.0 ␮m/s ≤1.0 ␮m/s
conductivity of the least (>5.67 in/h) (≤5.67 to >1.42 in/h) (≤1.42 to >0.14 in/h) (≤0.14 in/h)
transmissive layer
and and and and/or
Depth to water- 50 to 100 cm 50 to 100 cm 50 to 100 cm <50 cm
impermeable layer [20 to 40 in] [20 to 40 in] [20 to 40 in] [<20 in]
and and and and/or
Depth to high water 60 to 100 cm 60 to 100 cm 60 to 100 cm <60 cm
table [24 to 40 in] [24 to 40 in] [24 to 40 in] [<24 in]
Table A.C.6 Criteria for assignment of hydrologic soil groups when a water-impermeable layer exists
at a depth greater than 100 cm (40 inches) (USDA, 2007)
Soil property Hydrologic soil group

A B C D
Saturated hydraulic >10 ␮m/s ≤10.0 to >4.0 ␮m/s ≤4.0 to >0.40 ␮m/s ≤0.40 ␮m/s
conductivity of the least (>1.42 in/h) (≤1.42 to >57 in/h) (≤0.57 to >0.06 in/h) (≤0.06 in/h)
transmissive layer
and and and and/or
Depth to water- >100 cm >100 cm >100 cm >100 cm
impermeable layer [>40 in] [>40 in] [>40 in] [>40 in]
and and and and/or
Depth to high water table >100 cm >100 cm >100 cm >100 cm
[>40 in] [>40 in] [>40 in] [>40 in]
For hydrologic soil groups where an impermeable layer is less than 100 cm (40 inches), see the table above.
276 APPENDIX C
Figure A.C.3 Soil texture classification

Image USDA; license: US government work product.
Table A.C.7 Soil particle sizes
Name Size (mm)

Very coarse sand 2.0–1.0 mm
Coarse sand 1.0–0.5 mm
Medium sand 0.5–0.25 mm
Fine sand 0.25–0.10 mm
Very fine sand 0.10–0.05 mm
Silt 0.05–0.002 mm
Clay < 0.002 mm
Table A.C.8 Soil infiltration rates
Soil texture, type 0–4% Soil texture, type 0–4%

Coarse sand 1.25 Loam 0.54
Medium sand 1.06 Silt loam 0.50
Fine sand 0.94 Silt 0.44
Loamy sand 0.88 Sandy clay 0.31
Sandy loam 0.75 Clay loam 0.25
Fine sandy loam 0.63 Silty clay 0.19
Very fine sandy loam 0.59 Clay 0.13
Note: Rates based on full cover. These figures decrease with time and per cent of cover. Derived from USDA
information. (SCC 0812 § 1, 1990.)
APPENDIX C 277
Table A.C.9 Plant species within the JEL Wade Wetland, Wilmington, North Carolina, USDA
Hardiness Zone 7 (Mallin et al., 2012)
Trees
Carpinus caroliniana – ironwood – deciduous tree
Cornus foemina – swamp dogwood – deciduous tree
Nyssa biflora – swamp tupelo – deciduous tree
Pinus palustris – longleaf pine – evergreen tree
Quercus laurifolia – laurel oak – deciduous tree
Quercus michauxii – swamp chestnut oak – deciduous tree
Quercus virginiana – live oak – evergreen tree
Taxodium distichum – bald cypress – deciduous tree
Shrubs
Alnus serrulata – smooth alder – deciduous shrub
Crategus viridus – green hawthorn – deciduous tree
Aronia arbutifolia – red chokeberry – deciduous shrub
Cephalanthus occidentalis – buttonbush – deciduous shrub
Chamaecyparis thyoides – swamp dogwood – deciduous shrub
Cyrilla racemiflora – titi – evergreen shrub
Ilex glabra – inkberry – evergreen shrub
Ilex verticillata – common winterberry – deciduous shrub
Itea virginica – Virginia sweet-spires – deciduous shrub
Myrica cerifiera – wax myrtle – evergreen shrub
Sambucus canadensis – common elderberry – deciduous shrub
Viburum nudum – southern wild raisin – deciduous shrub
Other
Acorus calamus – sweet flag – grasslike
Carex (species mix) – sedges
Eupatorium – spp. semaphore thoroughwart – fleshy flowering herb
Helianthus angustifolius – swamp sunflower – flowering herb
Hibiscus moscheutus – rose mallow or marsh mallow – flowering herb
Hibiscus coccineus – scarlet rose mallow – flowering herb
Iris fulva – copper iris – flowering herb
Iris virginica – southern blue flag – flowering herb
Juncus coriaceous – leathery rush – rush
Juncus effusus – soft rush – rush
Kosteletskya virginica – seashore mallow – flowering herb
Leersia oryzoides – rice cutgrass – grass
Nuphar luteum – spatterdock – fleshy herb
Peltandra virginica – arrow arum – fleshy herb
Pontederia cordata – pickerelweed – fleshy herb
Schoenoplectus tabernaemontani – softstem or great bulrush – sedge
Sparganium americanum – eastern burr-reed – grass-like
Saggitaria lancifolia – bull-tongue arrowhead – fleshy herb
Saururus cernuus – lizard’s tail – flowering herb
Veronia novaboracensis – New York ironweed – flowering herb
Zizaniopsis miliacea – giant cutgrass – grass
278 APPENDIX C
Table A.C.10 Bio-retention basin plants – Pacific Northwest, USA

The plants below are suitable for the wet zone of bio-retention basins
Trees Shrubs
Botanical name Common name Botanical name Common name
Alnus rubra Red alder Comus sericea Red-twig dogwood
Fraxinus latifolia Oregon ash Comus sericea ‘Kelseyi’ Dwarf dogwood
Malus fusca Pacific crabapple Comus sericea ‘flaviramea’ Yellow dogwood
Salix Iucida Pacific willow Comus sericea ‘Isanti’ Isanti dogwood
Lonicera inuolucrata Black twinberry
Emergent plants Myrica californica Pacific wax myrtle
Carex obnupta Slough sedge Physocarpus capitatus Pacific ninebark
Carex stipata Sawbeak sedge Rosa pisocarpa Clustered Wild rose
Juncus densifolius Common rush Salix purpunea ‘Nona’ Dwarf Arctic willow
Juncus effuses Daggerleaf rush Spiraea douglasii Douglas spirea
Juncus tenuis Slender rush
Scirpus acutus Hardstem bulrush
Scirpus microcarpus Small-fruited bulrush
References
Dong Yue, Y., Zhang, Y., Liu, X., Li, Y., Jin, J., Duan, T., Li, J. and He, P. “Construction of
ecosystem and water quality control of Houtan wetlands in Shanghai Expo Garden,”
Wetland Science, vol. 3, pp. 219–226, 2013.
Mallin, M. A., McAuliffe, J. A., McIver, M. R., Mayes, D. and Hanson, M. A. “High pollutant
removal efficacy of a large constructed wetland leads to receiving stream improvements,”
Journal of Environmental Quality, vol. 41, no. 6, p. 2046, 2012.
Reiners, J. “Vegetation growth and success as a function of soil moisture conditions in
bioretention cells,” 2008. [Online]. Available: www.seattle.gov/util/cs/groups/public/@spu/
@usm/documents/webcontent/spu02_020010.pdf. [Accessed: 12 July 2016].
USDA (US Department of Agriculture), “Part 630 hydrology national engineering handbook,”
US Department of Agriculture, 210–VI–NEH, 2007.
USDA (US Department of Agriculture), “Urban Hydrology for Small Watersheds,” Technical
Release 55, 1986.
Vymazal, J. “Emergent plants used in free water surface constructed wetlands: A review,”
Index
2,4-D 213, 214 Atlanta, Georgia 189

atrazine 235
absorbable organic compounds 263 Austria 81
Acetaminophen 92 Azolla 43
acetic acid 259
Acorus gramineus 130 Bacillus 83
activated sludge 264 background concentrations 47
acute pollution level 17, 23 Bacopa monnieri 132
aerated horizontal subsurface wetland 255 acteria: colonies 23; measurement 23
aerated lagoon 30, 39–41 base flow 11, 214
aerated pipes 72, 75, 90 Bayawan City, Philippines 64
aeration 245 Bay of Biscay, France 239
aeration cascade 255 Beef Nutrition Farm 242
Agapanthus africanus 130 bentonite 36
agriculture: commodities 233; four-stage benzene 255
wetland 242; horizontal subsurface biodegradability 254
wetland 244; hormones 247; land-based biological oxygen demand 25, 245; test 26
disposal 244; phosphorus 236; sewage biomass 80
lagoons 244; TSS 235; water 233 bio-retention: bacteria 212; basins 205;
Albany, Oregon 41–43 design criteria 206; filter media 209;
alcohols 259 flooded sub-basin 208; flow sequence
algae 251 207; hybrid 213; materials 209; PAH
algae bloom 20 212; plants 210, 213, 278; pretreatment
alteration of rivers 144 206; subgrade soil 208; treatment basin
aluminium polychloride 72 206; under-drain 207; water management
ammonia 45, 244, 247, 248, 251 210
anaerobic digestion 254, 264 Bird of paradise 130
androgen hormones 247 Bisphenol A 92
anoxic 29 Black Sea 238
aquaculture 248; goals 250; horizontal Blue Heron Water Reclamation Facility 31,
subsurface wetland 252; pond size 251; 33
water use 250 Blumberg Engineering 77, 94. 96, 261
aquifers, composition 13 BOD 25, 245
aquifers, USA 13 BOD/COD ratio 254
arrowhead 119 Boone Presbyterian Church wetland 47
Arroyo Secco River 144 Boston 15
artificial fertilizer 238 Brazil dairy wetland 246
Arundo spp. 126 broadleaf cattail 45
ATI Wah Chang 43 brownfield 151, 168
280 INDEX
BTEX 255 constructed wetlands: stages 22; stage

bugle lily 131 sequences 22; types 21
bulrush 114 contaminant concentration 22
conventional wastewater treatment plants
cadmium 66 28
Calla lily 128 copper 66
Caltha palustris 121 corridors: ecological 2; stormwater 2, 7;
Canna generalis 250 width 2
Canna indica 264 Cuyahoga River 16
Canna spp. 122–124 Cyperus alternifolius 38
carbolic acid 149 Cyperus spp. 125
cardinal flower 47 Czech Republic 50, 67
Cares spp 116–118
Carex appressa 212 dairies 244; endocrine disruptors 247;
Carrion de los Cespedes 88 wastewater 244
carrying capacity 19 dairy wetland, performance 245
Casper, Wyoming 255 Danish system 72
catfish 249 Dargaville, New Zealand 247
cattail 104 dead zones 20, 238; map 240
Center for New Water Technologies 89 decentralized wastewater treatment 16, 28
centralized wastewater treatment 16 denitrifying wetland 45, 239–241, 247
CH2M Hill 41 Denmark 70
Changshu Advance Materials Industrial Park Denver, Colorado 11
261 Des Plains River wetland 142
Changshu, China 77 dicamba 213, 214
Charlotte, North Carolina 210 Diclofenac 92
cheese production wastewater 260 dissolved Oxygen 26
chemical manufacturing wetlands 261 distilleries 260
Chinese Pavilion 153 distribution piping 72, 73
Chinese water quality: classes 156; standards DO 26
156 Drosera burmanni 38
Chinook salmon 45 duckweed 135
chlorine 263
chromium 260 ecological corridors 2; length 4; see also
chronic pollution level 17, 19, 23 corridors
Clayton County, Georgia 40 ecosystem service, water 15
Clean Water Act 264 Eichhornia crassipes 38, 133–134
Cle Elum, Washington 39–41 Eleocharis dulcis 112
Cleveland 16 Eleocharis fallax 112
climate change 11 Eleocharis palustris 111
climate change, Europe 12 Eleocharis sphacelata 112
climate change, USA 12 Eleocharis spp. 110–112
clogging 251 E L Huie Jr Constructed Treatment
coal-fired power plants 10, 149 Wetlands 39; performance 39
Coldwater River, Mississippi, wetland emergent vegetation 102
234–235; performance 237 Endangered Species Act 9
cold weather wetlands 255 Enterococci 28
coliform bacteria 28 Escherichia coli (E. coli) 23, 38–39, 91, 187;
Colocasia esculenta 132 standard 28
common reed 51, 106 ethanol 259
Composite Curve Numbers 274, 275 Ethinylestradiol 92
concentrated animal operations 241 ethylbenzene 255
INDEX 281
Experimental Wastewater Treatment Plant of 67, 250, 260; cell shape 54; clogging
Carrión de los Céspedes 61 56–57; cold climate 67; configuration 50;
depth 55; design challenge 52; filter media
facultative lagoon 29, 50 58–60; history 50; hydraulic loading 53;
faecal coliform 66 inlet 52, 58; liner 57; loading 52, 55;
feedlots 241 materials 57; metals 66; nitrogen 65;
field edge wetland 234 pathogenic bacteria 65; performance 63,
first flush 187 64, 65; phosphorus 65; planting design
Flanders, Belgium 66 136; plug flow trench 61–63;
floating plants 43 pretreatment 55; residential 63; sizing
flood irrigation 10 52–53
flow through cooling 10 Houtan Park wetland, Shanghai 101;
food processing wastewater 254 aeration 159; aquatic organisms 271;
free water surface wetland 21, 31, 38, 39, bamboo decking 164; biological 153,
41–43; aeration 33, 44; BOD 47; cell 155, 156, 159, 161, 170; case study 148;
shape 35; clogging 37; cold weather 37; climate change adaptation 170; common
configuration 34; depth 35; design aquatic organisms 154; common plants
challenge 32; filter media 37; fish 39; 154; context 151; cultural ecosystem
history 32; hydraulic loading 35; influent services 162; diversity 170; economic
distribution 36; inlet 36; large scale 39; impact 168; ecosystem services 169, 170;
loading 34; materials 36; metals 47; education 167; first stage 157; flood
Netherlands 32; nitrogen 47; Orlando, control 152; maintenance 168; pedestrian
Florida 32; outlet 37; pathogenic bacteria route 163–164; planting design 155;
48; performance 47; phosphorus 47; planting plan 158; program 154; program
plants 38; rural 40; secondary treatment implementation 114; recycled steel 166;
38; site scale 38; sizing 34; slopes 37; site art 165; stormwater 161;
TSS 47; wildlife 47 sustainability 168; terraces 159–162;
fructose 259 treatment performance 155; water
treatment sequence 154; wetland plants
gasoline-derived organics 255 172, 271
geotextile fabric 202 Huangpu River, Shanghai 148; water quality
Giant Canada Geese 257 151
glucose 259 Hubei, China 1
glycerol 259 hybrid bio-retention basins 213
glyphosate 213, 214 hybrid constructed wetlands, 22, 87, 250,
grams 22 251, 260; agriculture 87; case study 88,
Greensborough, North Carolina 211 92, 93, 94; Changshu, China 262; E. coli
Greenworks 216 91; free water surface stage 90; horizontal
groundwater 12; contamination 255; nitrate flow stage 90; loading 89, 93;
14; pollution 14 performance 91; secondary benefits 94;
tertiary treatment 87; three-stage 93;
habitat, patch size 4 vertical flow stage 89
Hal Marshall: bio-retention basin 211; hydraulic loading 245, 248, 250, 252, 261
performance 212 hydraulic residency time 36
Harbin, China 217 hydrocarbon 207; removal 255, 257
hardstem bulrush 40, 45 hydrologic soil group criteria 275
Heliconia spp. 95, 128 Hygrophila pogonocalyx 38
Hennersdorf, Germany 94 Hymenocallis littoralis 132
Hewletts Creek 192, 195 hypoxic Zones 239
high density polyethylene 34
horizontal subsurface flow constructed Ibuprofen 92
wetland 21; case study 61, 63, 64, 66, Imhoff tank 88
282 INDEX
impaired waters 17, 19 mercury 19

industrial wastewater 254 metals: contamination 264; removal 265
industry: free water surface wetland 256, methane 76
257; horizontal subsurface wetland 260; Microccus 83
hybrid wetland 255, 261, 265; liner 266; micrograms 23
metals removal media 266; Phragmites milking parlor 245, 260
257, 259, 261; stormwater 256; Typha Millennium Ecosystem Assessment 142,
261, 263; vegetated swales 256; water 229
use 263; wetlands cost benefit 266; milligrams per litre 22
wildlife 256 mining 19
infiltration 11; requirements 214, 215 Minnesota 67
inundation levels 103 Mississippi River 14, 20, 144, 239
Iowa State University 242 Missouri River 256
Iris pseudacorus 63, 79, 80 mosquitos 39
Iris sibirica 63 most probable number (MPN) 23
Iris spp. 118–119 Munich, Germany 145
iron 66 municipal master planning 1
irrigation water 9; efficiency 10
Isar River 145, 173 NaCl 260
Naphthalene 207
Jeffrey Energy Centre, case study 264 National Cheng-Kung University wetland
JEL Wade Park 192–197 38
JEL Wade wetland plant list 277 National Cheng-Kung University wetland,
Juncus effusus 109, 259 performance 38–39
Juncus spp. 108–110 National Urban Wetland Park 175, 217;
basin sequence 220; context 217; design
Kobuk, Alaska wetlands 147 219; economic development 236;
Krovi, Czech Republic 51 ecosystem services 226; land-use 218;
Kurisu International 41 outlet 225; pavilions 226; site grading
214; water management 221; water
lactic acid 259 treatment 223
Lake Taihu 148 New York City 142
Lake Yangcheng 95 nitrate 27, 45, 240, 244, 247, 248; for
Las Vegas 15 irrigation 260
lead 66 Nitrobacter 83
Lemna minor 135 nitrogen: ammonia 27, 45; in streams and
Liang Zhi Hu Lake 2 rivers 17
Liaohe Delta, China 106 Nitrosococcus 83
Liebenburg-Othfresen, Germany 50 Nitrosolobus 83
liners 36 Nitrosomonas 83
Long Beach, California 187 Nitrosopira 83
Los Angeles River 148 Nitrospina 83
Lotus Island 95 non-point source pollution 20
Ludwigia adscendens 38 Nuphar shimadai 38
Macropodus opercularis 39 Oaklands Park, England 92

Mandan wetland, North Dakota 256; oestrogen hormones 247
performance 257 Ohio State University 213
manure slurry 248 oil recovery 256, 257
marsh marigold 121 oil/water separator 257
measurement units 22 Olympic College Parking Garden 178
mechanical/chemical processing 263 Oman wetland 257
INDEX 283
Onion Creek 242 rain garden 205

Oregon City 216 recirculation 83, 250
organic: loading 245, 261; pollutants 19 reclaimed water, piping 16
organic matter, transformation 27 Red River Research Station, Louisiana 237;
oxidation 60 wetland: stages 237
Oxybenzone 92 redox potential 60
reduction 60
PAH 144, 150, 187 reed canary grass 112
papyrus 80, 125 Reinhold Kickuth 50
pathogens 28 retrofitting 1
permethrin (Hi Yield38) 235 riparian wetlands 142; natural 146;
personal care products 92, 93 vegetation 146
pesticides 236; in streams 17 river: assessment 17, 19; channelization 148;
petrochemical 19 pollution 16; restoration objectives 145
petroleum wastewater wetland 255 Rivers and streams, biological condition 18;
Phalaris arundinacea 65, 102–114, 252 Mediterranean river flow 12
pharmaceuticals 92, 93 Roseburg Oregon 17–18
phenol 149, 263 run-off: curve numbers 272; depth 274;
Philadelphia 1, 177 urban 11; volume 11
phosphorus 28, 212, 236, 244; in steams rush 109
and rivers 17; wastewater 18
Phragmites australis 59, 81, 106 Sagittaria 119–121
Phragmites communis 252 Sagittaria latifolia 45
Phragmites karka 64, 106 St May’s Kansas, USA 264
Phragmites spp. 106–108 salmon 26, 249
phytoplankton 239 Sanlihe River 143, 172
Pickerel weed 121 Schoenoplectus acutus 45
Pistia stratiotes 135 Schoenoplectus tabernaemontani 45
plants 98; benefits 80; common deep marsh Scirpus acutus 40, 45
102; common marsh species 102; nutrient Scirpus fluviatilis 245
uptake 80; listed by continent 269; Scirpus spp. (Schoenoplectus) 114–116
planting design 101, 103, 136; shallowest Scirpus validus 45
water 102 scrubbers 264
point source pollution 16 secondary treatment 26
pollutant load 23 sedimentation 19; cell 251
polycyclic aromatic hydrocarbons (PAH) Seidle, Kathe 50, 70
see PAH septic tanks 76; septic system failure 14;
polyvinyl chloride 36 septic tank effluent 30
Pontederia spp. 121 Seville, Spain 61
Pontevedra, Spain, wetland 259 sewage lagoon 29
population equivalent 25 Shanghai: drinking water 149; Exposition
Portland, Oregon 210 153
potato process wastewater wetland 260 shellfish 192, 238
precipitation 11 Shengyang, China 85, 96
primary treatment 26 shrimp production 251
Prince George’s County, Maryland 215 Silkerode, Germany 70
Pseudomonas 83 S-metolachlor (BicepII Magnum) 235
pulp and paper mill, wastewater 263 softstem bulrush 45
pumping tank 72 soil: infiltration rages 276; particle size 276;
texture classification 276
Qinhuangdao City, China 174 Songhua River 217, 221
Quanti Tray Sampler 23 Spalene Poroco, Czech Republic 65
284 INDEX
spider lily 132 sustainable water use, barriers 15

spikesedge 110 swine 247
Sprilliums 83 synthetic liners 36
standard specifications 84
Stapleton, Colorado 178 Tabor School, Portland Oregon 205
Stoneville, Mississippi 249 Taiwan 38
stormwater: algae 191; basin see stormwater Talking Water Gardens 41–43, 44; birds 46;
basin; calculation methods 181; case performance 45; plants 45
study 217; channelization 179; Tanghe River 149, 173
contaminants and land use 186; design tanneries 260; wetland performance 261
storms 180; development impacts 177; E. Taro 132
coli 190; flood control 179; heavy metals tartaric acid 259
189; hydrologic soil groups 185; temporary pool 206
isopluvial map 181; JEL Wade Park see tertiary treatment 26
stormwater, JEL Wade Park; municipal Thalia spp. 127
planning 178; nitrogen 190; off-line thermoelectric power generation, water use
treatment wetlands 197; parcel 177; 264
parking lot contaminants 187; thermoelectric power generators 9
phosphorus 190; planning 177; rainfall Tianjin, China 99, 216
types 180; rational method 181; roadway tidal-flow 71, 74
contaminants 188; runoff curve numbers Tilapia 251
185; run-off rate 176; sedimentation Titusville, Florida 21
basin 199; stream impact 189; subsurface TKN 27
gravel wetland see stormwater, subsurface TMDL 23, 41
gravel wetland; toxicity 188; TR-55 toluene 255
182–186; treatment 176; treatment Tonalide 92
landscapes 192; typical contaminants total Kjeldahl nitrogen (TKN) 27
186; urban contaminants 189; volume Total Maximum Daily Load 23, 41
176; wetland 47, 97 total suspended solids 27; standard 23
stormwater basin: management 175; Town Brook, New York 242
regulations 179; run-off characteristics TR-55 215
176; scouring 103 Triclosan 92
stormwater, JEL Wade Park 192; design tropical sundew 38
193; performance 196; plan 194; plants trout 26, 249, 252
194; secondary benefits 197; water TSS 245, 248
management 195; water treatment 195; Turenscape I, 229
weirs 194 Typha latifolia 38, 45, 104–106
stormwater, subsurface gravel wetland: filter Typha orientalis 38
media 201; flow sequence 199; materials
199; outlet structure 201; performance umbrella papyrus 38
204; piping 201; retrofit 204; University of New Hampshire: bio-retention
sedimentation basin 199; water basin 212; Stormwater Center 199, 211
management 203; water quality 203 Up-flow: wetland 74; anaerobic sludge
stream assessment 17, 19 blanket (USAB) 76
Street Edge Alternative 214 US Clean Water Act 244
Strelitzia reginae 130 US Natural Resources Conservation Service
Streptococci 92 209
subsurface flow constructed wetland 21 USAB see up-flow: anaerobic sludge blanket
sulphate 264
sulphur dioxide 264 Vermont, USA 244
suspended solids 25 vertical subsurface flow 22, 70, 242, 247,
sustainable development 19 250; wetlands: ammonia 82; bacteria 82;
INDEX 285
case study 70, 81, 85; clogging 79; water use: aquaculture 10; domestic 15;
configuration 71; Danish guidelines 79; livestock 10; mining 10; water use per
denitrification 83; depth 78; filter media kilowatt-hour 10
78; French system 73–74, 77, 79, 82; Watsonia borbonica 104
hydraulic loading 77; intermittent loading wetland: assessment 17, 19; birds 45; images
70; liner 78; loading 76; nitrification 83; 45; insulation 67, 255; loss 17; visitors 44
nitrogen 82; pathogenic bacteria 84; wetland plants 98; common species 99;
phosphorus 84; plants 79; pretreatment floating 133; mixed community 100;
75; shape 77; Shenyang 85; sizing 76; native species 104; water depth 99,
sub-basin 75, 81; tidal-flow 74; two-stage 101–102
81; up-flow 74, 259, 266; wetlands 266 Willamette River 41
Vietnam, wetland 250 Wilmington, North Carolina 192
Villanova University, Pennsylvania 209, winery: wastewater 258; wetland 259
211 Wisconsin Department of Natural Resources
volatile organic compounds 263 214
worms 83
wapato 46 Wuhan, China 1
Warmwater Aquaculture Research 249 Wulijie, China 2; controlling plan 4, 7;
wastewater: characteristics 25; pretreatment flooding 3; habitat 3; intermediate
30, 33; strength 25; treatment plants 17; corridor 6; stormwater flow plan 3;
treatment standards see wastewater, urban corridor 4, 6; widest corridor 5
treatment standards
wastewater, treatment standards 26; nitrogen xylene 123, 255
26; organic matter 26
water: allocation 15; quantity 9; resource Yangtze River 14, 20, 144, 263
planning 9; temperature 41, 43, 45; water yellow flag 79
use yellow water lily 38
water cost: agriculture 16; Europe 15; United
Kingdom 15; USA 15 Zantedeschia aethiopica 128
water hyacinth 38, 133 Zasada, Czech Republic 67
water hyssop 132 zeolite 60
water lettuce 134 zinc 66
water primrose 38 Zitenice, Czech Republic 64

Constructed Wetlands and Sustainable Development (Gary Austin Yu Kongjian)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Constructed Wetlands and Sustainable Development (Gary Austin Yu Kongjian)

Uploaded by

Copyright:

Available Formats

Constructed Wetlands and

Kongjian Yu is co-author of the inﬂuential book The Art of Survival (2007). He is

Gary Austin and Kongjian Yu

ISBN: 978-1-138-90898-7 (hbk)

1 Water and sustainable urban design 1

3 Free water surface constructed wetlands 31

4 Horizontal subsurface ﬂow treatment wetlands 50

5 Vertical subsurface ﬂow constructed wetlands 70

6 Hybrid constructed wetlands 87

7 Plants in constructed wetlands 98

8 Riparian wetlands 142

9 Stormwater management and sustainable development 175

10 Increasing the sustainability of agriculture 233

11 Treatment of industrial efﬂuent in constructed wetlands 254

1.1 Philadelphia’s campaign to divert stormwater run-off into

5.11 Section of two-stage vertical subsurface ﬂow wetland 81

7.32 Pontederia cordata (Pickerel Weed) 121

8.52 Aerial perspective of the Tanghe River project 173

9.35 Elevated walkways and pavilions shown in Qunli, China 220

2.1 Characteristics of raw sewage in the United States 26

11.2 Hybrid wetland and the removal of metals 265

Gary Austin, 2016

Municipal master planning

recommendations vary according to the context, but in a low-density residential district,

Material selection, architecture featuring passive and active solar technology,

Water resource planning

Thermoelectric power and agricultural industries

Urban stormwater run-off quantity

contamination by disease-causing organisms or toxic substances. Again, constructed

Barriers to sustainable water use

Domestic water use and cost

Centralized wastewater treatment

Distribution piping for reclaimed water

Condition of wetlands and streams

Constructed wetlands and sustainable development

Types of constructed wetlands

• vertical subsurface ﬂow > horizontal subsurface ﬂow

Figure 1.25 Figure 1.26

US Environmental Protection Agency (EPA) standard for total suspended solids in

exposure is short-term and when it is long-term. The acute level of exposure is

Characteristics of raw sewage

Component Concentration range Typical concentration Mass loading g/p/d

Organic matter and nitrogen

Total suspended solids

Conventional wastewater treatment plants

Decentralized wastewater treatment

Component Concentration range Typical concentration

Engineered free water surface wetlands

from contacting water that has pathogenic bacteria in excess of US Environmental

Hydraulic loading rate

Cold weather applications

TSS BOD COD TKN NOx TN TP TC EC

Large-scale tertiary treatment applications

Retrofit and rural applications

1 – Annual average at the lake discharge.

The Talking Water Gardens case study

Table 3.3 Official visitors (Neal and Preston, 2015)

Fiscal year Total visitors

Table 3.4 Ammonia and nitrogen removal (mg/L) (Frazier, 2012)

November average May average

Table 3.5 Bird species photographed at the Talking Waters Gardens

Cooper’s Hawk Red-Tailed Hawk

BOD and TSS

Variations or additions to the commonly applied design features are in response