Downloaded by [University of California, San Diego] at 19:36 04 May 2017

GREEN INFRASTRUCTURE
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

With more than half of the world’s population now living in urban areas, it is vitally
important that towns and cities are healthy places in which to live. The principal aim of this
book is to synthesise the disparate literature on the use of vegetation in the built environment
and its multifunctional benefits to humans. The author reviews issues such as: contact with
wildlife and its immediate and long-term effects on psychological and physical well-being;
the role of vegetation in removing health-damaging pollutants from the air; green roofs and
green walls, which provide insulation, reduce energy use and decrease the carbon footprint
of buildings; and structural vegetation such as street trees, providing shading and air circu-
lation whilst also helping to stop flash floods through surface drainage. Examples are used
throughout to illustrate the practical use of vegetation to improve the urban environment and
deliver ecosystem services.
Whilst the underlying theme is the value of biodiversity, the emphasis is less on existing
high-value green spaces (such as nature reserves, parks and gardens), than on the sealed
surfaces of urban areas (building surfaces, roads, car parks, plazas, etc.). The book shows how
these, and the spaces they encapsulate, can be modified to meet current and future environ-
mental challenges including climate change. The value of existing green space is also covered
to provide a comprehensive textbook of international relevance.

John W. Dover is Professor of Ecology at Staffordshire University, Stoke-on-Trent, UK.

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

“This book comes at exactly the right time. The term ‘Green Infrastructure’ symbolizes
new thinking in relation to planning and constructing cities. Vegetation offers many benefits
to urban dwellers, and now is the time to integrate this knowledge into city planning
procedures. Around the world, examples of GI now exist and national and international
associations support and disseminate such ideas. This textbook highlights many such examples
to introduce the concepts to a wider audience. I wish this book as many readers as possible;
first of all students of related disciplines to take these ideas as seed into their future business
lives, and also urban developers to integrate this new thinking into their daily business to
spread out more green than grey infrastructure around the globe.”

– Manfred Koehler, Professor of Landscape Ecology, University of Applied Sciences Neubrandenburg,

Germany, President of the World Green Infrastructure Network.

“This is a very thoughtful, timely and comprehensive book. If our cities are to be considered
as being truly sustainable then the integrated provision of green spaces will be a vital
component of that sustainability. Green Infrastructure will help us to do just that.”

– Richard Sabin, Director, Living Green City, UK.

 GREEN INFRASTRUCTURE
Incorporating plants and enhancing
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

biodiversity in buildings and urban

environments

John W. Dover
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

First published 2015

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
© 2015 John W. Dover
The right of John W. Dover to be identified as author of this work has been
asserted by him 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
Dover, John W.
Green infrastructure : incorporating plants and enhancing biodiversity in
buildings and urban environments / John W. Dover.
pages cm
Includes bibliographical references and index.
1. House plants in interior decoration. 2. Foliage plants. 3. Interior
landscaping. I. Title.
SB419.D74 2015
635.9’65--dc23
2014048373

ISBN: 978-0-415-52123-9 (hbk)

ISBN: 978-0-415-52124-6 (pbk)
ISBN: 978-0-203-12199-3 (ebk)

Typeset in Bembo
by Fakenham Prepress Solutions, Fakenham, Norfolk NR21 8NN
 CONTENTS
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Preface ix
Acknowledgements xi

1 What is green infrastructure? 1

1.1 Why green infrastructure? 1
1.2 Definitions of green infrastructure 2
1.3 Why should we want to ‘green’ buildings and reduce the amount
of sealed surface in urban areas? 4

2 Benefits of green infrastructure 7

2.1 Introduction 7
2.2 Biodiversity value 9
2.3 Direct value to humans 14
2.4 Food production/urban agriculture 24
2.5 Transport (walking/cycling) 26
2.6 Economics/enhanced building values 28
2.7 Climate control/climate change proofing 30
2.8 Sustainable urban drainage systems (SUDS) 40
2.9 Pollution control 48
2.10 Energy efficiency 62
2.11 Summary 63

3 Indoors 65
3.1 Why vegetation indoors? 65
3.2 Noise reduction 66
3.3 Plants and illness 67
3.4 General health and comfort – and the workplace 68
3.5 Air conditioning 72
 vi Contents

3.6 Microbial load reduction 72

3.7 Air pollution 73
3.8 Indoor green walls 88
3.9 Summary 92

4 Permeable pavements 93
4.1 Introduction to permeable pavements 93
4.2 Permeable pavements and street trees 98
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

4.3 Microbiology of permeable pavements 98

4.4 Permeable pavements and the heat island effect 99
4.5 Front gardens 101
4.6 Alleys 103
4.7 Roads 104
4.8 Swales and strands 104
4.9 Summary 104

5 Green walls 107

5.1 Introduction 107
5.2 Types of green wall 111
5.3 Directly greened walls: rooted directly on the wall 117
5.4 Direct greening: surface climbers 124
5.5 Indirect greening of walls 125
5.6 Specific values of green walls 141
5.7 Summary 162

6 Green roofs 163

6.1 Introduction 163
6.2 Types of green roof 172
6.3 Structure of green roofs 176
6.4 Plants for green roofs 182
6.5 Ecosystem services (or benefits) delivered by green roofs 193
6.6 Whole-life cost analysis of installing green roofs 214
6.7 Summary 217

7 Street trees 219

7.1 Street trees and city greening 219
7.2 Trees and health and safety 221
7.3 The street environment 222
7.4 Street tree inventories 226
7.5 Which tree species is best? Selection criteria 229
7.6 Getting the most out of street trees 235
7.7 Aesthetics and food 235
7.8 Trees and shade 237
7.9 Trees and gaseous pollutants 240
 Contents  vii

7.10 Street trees and particulate pollution 246

7.11 Trees and heavy metals 251
7.12 Street trees and water 253
7.13 Street trees and biodiversity 253
7.14 Societal value 255
7.15 Summary 255

8 Policy, regulation and incentives 257

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

8.1 Public good vs Private good 257

8.2 Creating the environment for implementation 258
8.3 Summary 266

References 267
Index 311
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

7KLVSDJHLQWHQWLRQDOO\OHIWEODQN
 PREFACE
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

This book arose from an Urban Ecology module I developed for a Master’s degree at
Staffordshire University called ‘Greening the Grey’ which explored the value of biodiversity,
mostly in the form of vegetation, to us as humans and also to wildlife. I became deeply
interested in the subject and felt that some kind of summary was needed to help provide
the background evidence to promote the greening of our cities. Writing took longer than
expected, the subject matter is wide and took me down paths I had never before encountered
and gave me diverse experiences. For example, I found beauty in the way that a geotextile
membrane – unseen to us – helps degrade oil leaks from cars; I also found horror in the way
that ultrafine particulates from pollution may enter our bodies. My personal view is that using
biodiversity intelligently in our towns and cities is perhaps the only way we can confront
some of the challenges of urban living facing us now, and in the future.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

7KLVSDJHLQWHQWLRQDOO\OHIWEODQN
 ACKNOWLEDGEMENTS
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

I have a great debt of thanks to pay to the many people who have helped me during the
process of writing this book. The greatest, of course, goes to my wife Christine and to
my children Alice and Charlotte for their tolerance and encouragement. My postgraduate
students on the Master’s module ‘Greening the Grey’ at Staffordshire University and PhD
student (now Dr) Caroline Chiquet have given me a great deal of pleasure and encour-
agement, and kept me on my toes over the years; without you I would not have started the
work at all. Colleagues and friends from academia and the commercial world have been
generous with their time and information, which I have truly valued. Non-academic friends
and relatives (especially Shirley Dixon and Jerry West) also deserve a mention – for putting
up with my constant talk on the subject and waiting patiently whilst I’d take a photograph of
a particularly interesting wall, tree or roof (again!). I’d particularly like to thank Dr Caroline
Chiquet, Christine Dover, Sean Farrell, Richard Sabin and Professor Philip James for reading
and commenting on an earlier version of the manuscript and thus helping improve it; all
remaining errors and inaccuracies are, of course, mine and mine alone. I’d also like to thank
Tim Hardwick at Earthscan for agreeing to take on steering me through the process of the
book’s production, along with Ashley Wright, and not least for reminding me to get on with
it! Thanks to Charlotte Dover for unearthing the information on the Holloway Sanatorium.
Picture credits are given throughout the book, but I’d just like to thank, personally: Christine
Dover and Caroline Chiquet, Staffordshire University, UK; Alan Goodkin, UK; Dr Alex Tan
from the National University of Singapore; Sarah Fisk of Adnams plc, UK; Andrea Fürst from
the Hundertwasser Archive, Vienna, Austria; Marc Grañén of Phytokinetic, Spain; Joel Nash
of ANS Group Europe; Emily Hauth of the City of Portland, Oregon; Edmund Maurer of
the Planning Department, Municipality of Linz, Austria; Dr Tom Smiley of the Bartlett Tree
Research Lab, Charlotte, NC; Sean Farrell of Mobilane, UK; Richard Sabin of Biotecture,
UK; Daniel Taylor and Liam Barrett of Bauder Ltd; David Luukas from Alumasc Exterior
Building Products Ltd; Gary Grant of The Green Roof Consultancy; and Justin Hobson of
Country Life. I am also exceedingly grateful to Rosie Duncan of Staffordshire University’s
Geography Department for redrawing several of the figures and to Christine Dover for
indexing.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

7KLVSDJHLQWHQWLRQDOO\OHIWEODQN
 1
WHAT IS GREEN INFRASTRUCTURE?
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

This chapter will:

• review some existing green infrastructure definitions

• present a modified definition of green infrastructure
• identify the reasoning behind the need to reduce the amount of sealed surface in urban
areas.

1.1  Why green infrastructure?

I have been interested in green infrastructure, on and off, for a long time, since the early 1990s,
even though it wasn’t called green infrastructure then. In the dim and distant past, I worked
for an environmental consultancy company, and one project I handled was to produce the
evidence base to support the green corridors policy which Kirklees Metropolitan Borough
Council (based in Huddersfield) were proposing for their Unitary Development Plan. It was
clear to me then that green spaces and biodiversity in towns provided a number of benefits
to the community (and I mean this in the broadest sense of people, public services and enter-
prises) and I recommended that they recognised that in their policy (Dover, 1992). Moving
back into academia, I subsequently did some rather more esoteric work on wildlife corridors
(Dover & Fry, 2001) and urban ecology (Dover, 2000), amongst other things, and in 2006
worked for the regeneration agency Renew North-Staffordshire on ‘embedding green infra-
structure in housing market renewal’ (Dover, 2006a, 2006b) which ‘renewed’ my interest in
using biodiversity to deliver important ‘ecosystem’ services to humans.
Ecosystem services are generally considered to be of four types:

• Supporting services, e.g. nutrient and water cycling, soil formation, primary production
• Regulating services, e.g. climate control and pollution removal
• Provisioning services, e.g. food, medicines, building materials
• Cultural services, e.g. societal appreciation of nature and environment (EASAC, 2009).
 2  What is green infrastructure?

I passionately believe that green infrastructure is the only way to cost-effectively meet some of
the most important environmental challenges we face in urban areas; challenges that we will
have to meet continuously over the coming decades, and perhaps even over centuries to come.
There are many good reasons for creating green infrastructure, but one which is of singular
importance is coping with the impacts of climate change, and the associated intensification
of other environmental problems such as air pollution.

1.2  Definitions of green infrastructure

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Perhaps I should make it clear from the start: when I use the word ‘green’ I mean using
plants, vegetation or even microbes (even those that are not green). So ‘greening’ buildings
means, literally, growing plants on, around, up, or even in them. I will not be using the term
for other more technological approaches to reducing the environmental impact of operating
buildings such as using ground-source heat pumps (Cho & Choi, 2014) or in the sense of a
‘green’ political movement.
Green infrastructure is often seen as being about the value of ‘open’ green spaces, and
mainly emphasises pre-existing ones. Definitions which exclude the built environment are
unhelpful in that they do not give a policy steer to the greening (literally, with vegetation) of
sealed surfaces (e.g. Benedict & McMahon, 2006; Goode, 2006; TCF, 2011; and NWGITT,
2007). In the UK the Town and Country Planning Association (TCPA) defined green infra-
structure in this way:

Green Infrastructure is the sub-regional network of protected sites, nature reserves,

greenspaces, and greenway linkages. The linkages include river corridors and flood
plains, migration routes and features of the landscape, which are of importance as
wildlife corridors.
Green Infrastructure should provide for multifunctional uses i.e., wildlife, recrea-
tional and cultural experience, as well as delivering ecological services, such as flood
protection and microclimate control. It should also operate at all spatial scales from
urban centres through to open countryside.
(TCPA, 2004)

Yet ironically both the TCPA (2004) and Goode (2006) whilst using a definition that
excludes building-related structures actually include them as examples within their publi-
cations (e.g. green roofs and sustainable urban drainage systems). Whilst many definitions
persist in excluding built-environment components, some newer ones acknowledge the
built-environment features explicitly, but only in respect of a limited set of attributes. For
example, this one concentrates on sustainable urban drainage:

Green infrastructure is management approaches and technologies that utilize, enhance

and/or mimic the natural hydrologic cycle processes of infiltration, evapotranspiration
and reuse. Green infrastructure approaches currently in use include green roofs, trees
and tree boxes, rain gardens, vegetated swales, pocket wetlands, infiltration planters,
porous and permeable pavements, vegetated median strips, reforestation/revegetation,
and protection and enhancement of riparian buffers and floodplains.
(EPA, 2008)
 What is green infrastructure?  3

Others use a vague term that allows inclusion, but does not effectively promote built-
environment features as such. For example:

Green Infrastructure is a strategically planned and delivered network comprising the

broadest range of high quality green spaces and other environmental features.
(NE, 2009a)

The term ‘other environmental features’ was used in the UK government’s White Paper on
the natural environment (Defra, 2011b), although it is unclear exactly what these features
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

are as they are not defined. Nevertheless, the document cites an example that includes green
roofs and walls and refers to the Flood and Water Management Act 2010 that has provision
for encouraging sustainable urban drainage.
So my preference would be for a definition such as (NE, 2009a) but which is more
explicit. For example:

Green Infrastructure is a strategically planned and delivered network comprising the

broadest range of high quality green spaces and other environmental features including
surfaces such as pavements, car parks, driveways, roads and buildings that have been modified to
incorporate biodiversity and promote ecosystem services. (italics = my text)

The definition used by Tzoulas et al. (2007) is pretty good and seems to cover most eventu-
alities and in very few words:

all natural, semi-natural and artificial networks of multifunctional ecological systems

within, around and between urban areas, at all spatial scales.

However, I have two remaining quibbles: 1) the last two definitions suggest a rather
top-down approach and one which discounts individual features that have not been a priori
included in some strategic planning and delivery vehicle – this seems perverse as surely green
infrastructure is a sum of assets whether they fit into a strategic framework/network or not;
2) these definitions stop at the entrance to a building – I do not see why this should be so as
ecosystem services can be delivered by biodiversity inside buildings as well as outside. So the
definition which I will be using is:

Green infrastructure is the sum of an area’s environmental assets, including stand-alone

elements and strategically planned and delivered networks of high-quality green spaces
and other environmental features including surfaces such as pavements, car parks,
driveways, roads and buildings (exterior and interior) that incorporate biodiversity and
promote ecosystem services.

Now, having settled on a definition I am happy with, I am going to blithely ignore a huge
chunk of it. Whilst I want to give an overview of the values of green infrastructure in the first
part of the book, I want then to concentrate on the ‘other environmental features including
surfaces such as pavements, car parks, driveways, roads and buildings (exterior and interior)
that incorporate biodiversity and promote ecosystem services’. Why? Principally because the
‘strategic network’ approach is well documented and a core component of spatial planning.
 4  What is green infrastructure?

The subtitle of the excellent book Green Infrastructure by Benedict and McMahon (2006) is
Linking Landscapes and Communities, and implies – incorrectly to my mind – that GI is the
preserve of the spatial. I do not want to give the impression that spatial planning or the
creation and promotion of networks of green spaces, linkages, etc. is not important: it clearly
is. What I want to do is show how a green infrastructure approach that embraces the modifi-
cation of our existing non-green areas, piecemeal or strategic, can improve our environment.
I want to show how we can ‘green the grey’.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

1.3  Why should we want to ‘green’ buildings and reduce the amount of
sealed surface in urban areas?
Whitford et al. (2001) identified four ways in which the ecology of cities is distinct from the
surrounding countryside: 1) cities are warmer, 2) cities retain less water, 3) cities emit carbon
dioxide, 4) cities have lower biodiversity; all these features are considered undesirable for one
reason or another. These are generalisations, of course, and may not hold true in every case.
Nevertheless, the point is that whilst cities are human habitats, humans are not very good at
creating uniformly good ones for their inhabitants. The four features Whitford et al. (2001)
identify are linked with poor environmental performance, but there are a whole clutch of
others which impact directly and indirectly on human health which, for simplicity, can be
bundled under the headings of ‘pollution’ and ‘green access’. Human health, physical and
mental, has been shown to be better where people have access to good quality green space
(Croucher et al., 2007; GSS, 2008) and especially that which is rich in biodiversity (Fuller
et al., 2007). Sadly, the poorest people in our urban communities are also those with least
access to green space, and tend to have the poorest health. Poor health is correlated with
income (Deaton, 2002; Smith, 1999), and if you do not have much income, you tend to
live in densely populated areas and eat less well than if you are rich (Alaimo, 2001). The
housing stock you live in will be poor too: old, leaky and probably also more expensive to
heat (Burholt & Windle, 2006). The health problems associated with poverty are exacerbated
by lack of access to green spaces (Mitchell & Popham, 2008) and this disparity in access to
environmental goods comes under the general concept of environmental justice (Haughton,
1999). Deprivation is, sadly, nothing new (Newman et al., 1990) and nor is the pollution
that comes with urbanisation (Jacobson, 2002) and resultant health impacts (e.g. PM2.5 related
asthma (Kheirbek et al., 2013)) which we know that plants can help to reduce (see Chapter
2 et seq.).
If you live in a town or city, especially a large one, you will recognise that it generates its
own climate – it is warmer than the surrounding countryside (Changon, 1992). I like snow
and always found it frustrating when I lived in cities: we had miserable rainy winters, but as
soon as I drove out of town, the countryside was covered in beautiful snow – you might not
agree with my preferences, but the reality is that urban climates are different. Back in 1982
a seminal book was published called Urban Ecology (Bornkamm et al., 1982); it contained
papers from the second European Ecological Symposium which was held in Berlin in 1980.
I recall reading it some years later and being struck by the way that much work on vegetation
was looking at how plants took up pollutants such as heavy metals (Greszta, 1982) and could
be selected for tolerance to air pollution such as sulphur dioxide (SO2) (Bell et al., 1982).
In the same volume Horbert et al. (1982) published a paper summarising their research
which identified the relationship between the proportion of sealed surfaces in an urban area
 What is green infrastructure?  5

(buildings, roads, pavements, etc.) and the resulting local climate which was: warmer (on
average 0.5–1.5°C higher, 2–6°C higher on clear days), wetter (5–10% more precipitation),
with higher levels of runoff, more polluted (gaseous pollution 5–25 times greater) and had
lower humidity (winter 2% lower, summer 8–10% lower) and air circulation (windspeed
10–20% lower) than surrounding areas with a higher density of vegetation. In the same paper,
Horbert et al. (1982) identified the ameliorating effect of urban vegetation on temperature
(the central part of the 212 ha Tiergarten park was cited as being as much as 7°C lower in
comparison with adjacent built-up areas) and that it could have localised impacts on the
surrounding areas (depending on the permeability of the built-up areas so as to permit air
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

movement). These urban climate issues are exactly the same challenges that face us now, but
with increased ferocity due to the challenges of climate change, and some of the solutions are
exactly as Horbert et al. (1982) identified: using vegetation.
So, we humans have not only changed our local environments for the worse, through
overcrowding and pollution, but we have also changed the global environment (IPCC, 2007,
2014; Solomon et al., 2007). Very few serious scientists doubt that global climate change is
real and that it is caused by human (anthropogenic) activity, principally through the burning
of fossil fuels (IPCC, 2007). We know from modelling that one of the consequences of
global climate change is that weather is going to become more unpredictable with extreme
events such as heatwaves and severe storms becoming more frequent. Extreme events such
as heatwaves kill people (Golden, 2004; Rey et al., 2007), extreme events like storms cause
flooding (Starke et al., 2010). Towns and cities that have a lot of ‘sealed surface’ (buildings,
roads, car parks) are less well equipped to deal with such, literally, lethal problems than those
with a lot of green space (O’Neill et al., 2009).
In April 2013 the UK government published its National Adaptation Programme aimed
at reducing the impact of climate change (Defra, 2013b) which ‘supports the use of green
infrastructure’. So why is this book concerned largely with the ‘built’ part of the environment
as opposed to large open spaces? It is partly to do with definitions and opportunities. Many
green infrastructure definitions (see above) appear to exclude biodiversity on and in buildings
and the benefits that it brings. This means that when policy is implemented and budgets set
and spent based on such definitions, built-up areas are excluded. The reason for greening
buildings and reducing the proportion of sealed surface is simple: if we don’t, we are going to
have increasingly uncomfortable urban areas with low aesthetic value, high levels of pollution,
poor climate, excessive heat, and flash flooding. In combination, this will have an impact on
physical and mental health.
It is also my perception that most organisations have departments that work in silos and
rarely work together or even appreciate the value of each other’s work. Green infrastructure,
to many officers and political representatives in urban areas, means parks and nature reserves
and the perception is therefore that it is really about aesthetics, recreation and wildlife. Of
course managing such areas costs money, and viewed against the need for other essential
services park management or nature reserve wardening appears a luxury and non-essential
and, as a result, is often the first service to have its budget cut when times are hard. This
view is, of course, to completely misunderstand what green spaces do, and to undervalue the
wide-ranging benefits they provide to the whole community. Unfortunately, many current
definitions (see above) have created new silo thinking: if green infrastructure is perceived as
being only about green open spaces, then this is an important misconception. So this book
aims to demonstrate: 1) the multifunctional values of green infrastructure generally – and
 6  What is green infrastructure?

how this cuts right across most of our concerns as citizens, and 2) to show how some aspects
of the built environment can be modified to deliver many of these benefits in addition to
green spaces. Why? The answer is very simple: few politicians are bold enough to suggest
knocking down valuable buildings to create a new open space in the heart, or even periphery,
of a town or city, but modifying built structures either by modifying existing building and
curtilages (surrounding areas) by ‘retrofitting’ or incorporating green infrastructure elements
in replacement or new builds is far easier and is really a matter of modifying planning policy
and building regulations, and demonstrating economic and community value. It is also worth
remembering that buildings are three-dimensional, whereas many urban green spaces are very
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

flat and essentially two dimensional: because of this it means that you could, potentially, get
more benefit from greening the surfaces of a building than you could from the same area of
open space if the building was not there.
 2
BENEFITS OF GREEN INFRASTRUCTURE
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

This chapter will:

• identify the general benefits of green infrastructure

• review the different ecosystem services delivered by green infrastructure.

2.1 Introduction
Without biodiversity there is no green infrastructure. Unfortunately, as a term ‘biodiversity’ is
really rather vague and can be considered at the level of genetic variability within a species or
population of a species right up to some global measure such as the total number of species on
the planet – which is a scientist’s way of looking at it. Another, very anthropogenic (human-
focused) way of viewing biodiversity is to consider what it can and has done for us. Well,
biodiversity does rather a lot: it provides us with materials for food, clothing, construction,
pharmaceuticals, aesthetic pleasure, oxygen production and carbon dioxide removal, and so
much more – such values have been recognised for a long time (Spellerberg, 1992). What
have become more, or perhaps simply more widely, recognised are the basic ecosystem
services that biodiversity delivers: removing pollutants, acting as air conditioning, insulation,
shading, removal and pre-treatment of rainwater, pollination services, increasing values of
adjacent property, reducing healthcare costs, etc. Of course, in a kind of circular argument,
enhancing green infrastructure can also increase and improve wildlife habitat, which improves
the ecosystem services provided by biodiversity (Figure 2.1).
Vegetation also acts as a carbon store, and whilst the ‘urban forest’ (a term coined to
collectively describe all the trees in urban, as opposed to rural, areas) appears to capture very
little (e.g. less than 1% of the carbon emissions in the USA (Nowak, 1994b) or 6% for the
West Midlands Region of the UK (Stewart et al., undated-a)), this can add-up to a substantial
amount. In addition, Nowak and Heisler (2010) in a report for the US National Recreation
and Park Association estimated that the value of trees in urban parks was in the billions of
dollars annually for air temperature reduction, $500 million per year for air pollution removal,
an unknown but substantial value for shielding from UV radiation in preventing skin cancer
 8  Benefits of green infrastructure

Aspect   A(ributes   Examples  

Visual   Aesthe1cs,  screening   Improved  visual  environment  

Exercise  (walking,  running,  

Reduced  costs  to  health  providers  
green  gym),  pollu1on,  
Human  Health   abatement,  de-­‐stressing,  
through  reduced  admissions;  improved  
mental  health;  faster  recovery  
socialisa1on,  recrea1on  
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Study  and  experience  of   Formal  and  informal  educa1on,  skills  

Educa1on   wildlife,  schools,  ranger   through  volunteering,  hobbies  
services,  volunteering   (photography,  bird  watching,  etc.)  

Allotments,  gardens,  orchards,   Improved  diet,  community  bonding,  

Food  Produc1on   roofs,  walls   educa1on,  biodiversity,  exercise  

Alterna1ve  movement   Reduces  road  conges1on,  safer  routes,  

Transport   corridor  for  cyclists  and   reduced  pollu1on  exposure,  more  
pedestrians   relaxed  seOng  

Property  prices,  inward   Improved  staff  morale,  reduced  sick-­‐

investment,  tourism,   leave,  improved  staff  reten1on,  aQracts  
Economics   improved  business  /  shopping   businesses;  units  let  faster  and  fewer  
environment   voids  

Reduced  heat-­‐related   Provides  shade  against  UV-­‐related  

mortality,  heat  island,  and   cancer,  cardiovascular  mortality,  
Climate  Control   wind;  improved  air  circula1on   heatstroke,  etc.;  reduces  air  and  surface  
and  climate  change  mi1ga1on   temperature,  freshens  air  

Sustainable   Reduces  risk,  economic  losses,  trauma  

Reduced  runoff,  flash  flooding  
Urban  Drainage   and  distress,  processing  costs  

Removes  PM10  and  below  from  air;  

Water,  light,  noise,  air  
absorbs  nitrogen  oxides,  ozone,  vola1le  
Pollu1on  Control   pollu1on  (par1culates,  gases  
organic  chemicals;  acts  as  heavy  metal  
and  aerosols,  odours)  
sink  

Reduces  air  condi1oning  and   Insulates  buildings,  provides  shelter  

Energy  Efficiency   hea1ng  costs   against  draYs,  shades  windows  

Provides  breeding  habitat,  food  and  

Wildlife  habitat,  wildlife  
Biodiversity   corridors  and  stepping  stones  
other  resources;  promotes  dispersal;  
reduces  ex1nc1on  risks  

FIGURE 2.1  Examples of the services provided to humans and wildlife by green infrastructure.
Many of the aspects overlap; for example, human health is improved not just by exercise,
but also by climate control, sustainable urban drainage and pollution control (modified from
Dover, 2000).
 Benefits of green infrastructure  9

and cataracts, $1.6 billion a year for carbon storage and $50 million a year for carbon removal;
the value of the trees themselves was estimated at $300 billion. Specific benefits will depend
on the area under consideration – for the Piedmont region of the USA (from the south of
New Jersey to eastern Texas), the value of trees to the community was dominated by reduc-
tions in stormwater runoff and elevated property prices followed by decreased energy use and
lower levels of air pollutants and CO2 (McPherson et al., 2006).

2.2  Biodiversity value

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

2.2.1  Wildlife and their habitats

In urban areas there are many green spaces, and urban woodland is known to be of value
in European cities to vertebrates (birds, bats) and invertebrates (Tyrväinen et al., 2005). But
not all green spaces will have much in the way of biodiversity value; for example, a closely
cut park lawn. Although such areas have their place as amenity areas, and they will provide
unsealed surfaces helping to reduce runoff following rain. Perhaps ironically, neglected ‘waste’
areas can be valuable for biodiversity; Sukopp (1990) estimated that abandoned urban sites
could have four times as many species as city parks, although about half of the plant species
would be non-native. Residential areas can host substantial communities of invertebrates
as Owen (2010) demonstrated in her 35-year study of her suburban Leicestershire garden
which had: 23 species of butterfly, 375 species of moth, 94 species of hoverfly, honeybees,
8 species of ‘true’ bumble bee, 5 species of ‘cuckoo’ bumble bee, 45 species of solitary bee,
7 species of social wasp, 55 species of solitary wasp, 87 species of sawfly, 183 species of bug
(Hemiptera) and 422 species of beetle. Areas of high plant species richness and heterogeneity
in vegetation structure (from patchy areas with bare ground to full tree canopy cover) are
likely to deliver a wider range of ecosystem services and habitat for fauna than those of low
richness and structure. However, this should not be interpreted as suggesting that every green
space should try and pack everything into it, more as a plea to think more widely about
green space design and functions. The character of a particular space will be reflected by its
principal purpose: play areas, cycleways and walkways, churchyards, parks, sports fields, etc.
But wherever possible it should incorporate native species, and be structurally variable (not
all mown grass) because in this way society will get a bigger benefit whatever the function of the
green space. For example, does a sports field all have to be low mown, hard-wearing, amenity
grass species, or can the non-competitive areas be fringed by more naturalistic vegetation?
Can golf courses become wildlife havens? Can churchyards be managed for wildlife? After
all, Sjöman et al. (2012c) showed that Aarhus had high tree species richness in its cemeteries
(Figure 2.2).
But what is wildlife habitat? There are very technical approaches to this question – for
example, habitat has been described as an ‘n-dimensional hypervolume’ (Begon et al., 1996)
– but we will take a more practical approach. Habitat is often thought of as a single physical
space where a species lives, a place where all its requirements can be found and where the
environmental characteristics are suitable. It might be a place with the right soil type, nutrient
levels, aspect, slope, light and moisture for a particular plant. For some animals that definition
may hold, but in the last decade a wider understanding has developed – that which highlights
the resource needs of a species, and an acknowledgement that these resources may not be
in the same place in space or time (Dennis et al., 2014) – something which is obvious if
 10  Benefits of green infrastructure
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.2  Cemetery managed for wildlife in the village of Kirtlington, UK. It is important to
let visitors know why grass is not cut to demonstrate that it is a deliberate management action
for a specific, positive reason, otherwise it can be interpreted as merely a way to save money or
because of neglect. High vegetative species richness generally leads to high faunal richness. ©
John Dover
 Downloaded by [University of California, San Diego] at 19:36 04 May 2017
TABLE 2.1  Insects recorded as visiting 25 species of native or naturalised plant in the University of Cambridge’s Botanic Garden 1996–1997

Hymenoptera (Ants, Bees, Wasps) Lepidoptera (Butterflies & Moths) Hoverflies

Psithyrus sp.

Ants
Anthidium
manicatum
Apis mellifera
Bombus hortorum
Bombus lapidarius
Bombus pratorum
Bombus pascuorum
Bombus terrestris
†
(Cuckoo species)
‘Solitary bee’
‘Solitary wasp’
Aglais urticae
Celastrina argiolus
Gonepteryx rhamni
Inachis io
Lycaena phlaeas
Maniola jurtina
Pieris brassicae
Pieris napi
Pieris rapae
Polyommatus icarus
Pieris tithonus
Unidentified
Skipper
Vanessa atalanta
‘Moths’
‘Hoverflies’
Episyrphus balteatus
Eristalis intricarius
Eristalis tenax
Heliophilus pendulus
Sphaerophoria scripta
Platycheirus sp.
Syrphus sp.
Melanostoma sp.
Metasyrphus sp.

Common name Scientific name

Cornflower Centaurea cyanus *
Greater knapweed Centaurea scabiosa • • •
Foxglove Digitalis purpurea
Wild teasel Dipsacus fullonum * • • • • • • • •
Hemp agrimony Eupatorium cannabinum • • • •
Field scabious Knautia arvensis * * * • • •
White deadnettle Lamium album *
Toadflax Linaria vulgaris • * •
Bird’s-foot trefoil Lotus corniculatus • • •
Ragged robin Lychnis flos-cuculi * • • • •
Purple loosetrife Lythrum salicaria * * • • * • • • • • • • • • • • • • • •
Musk mallow Malva moschata * * • • • • • • • •
Common mallow Malva sylvestris * * • • • • • •
Marjoram Origanum vulgare • * • • • •
Meadow clary Salvia pratensis * * *
Wild clary Salvia verbenaca *
Soapwort Saponaria officinalis • * • • • • • • • • • • • • • • • • •
Small scabious Scabiosa columbaria * * * * • • •
Red campion Silene dioica • • •
White campion Silene latifolia • • •
Marsh woundwort Stachys palustris • • • * • •
Hedge woundwort Stachys sylvatica • * • • • • •
Red clover Trifolium pratense • * •
White clover Trifolium repens * *
Common vetch Vicia sativa • • * •

Source: Comba et al. (1999b)

†
Now Bombus sp.; * = species most visited; • = species visited. Because of difficulties in identification, the bumblebees were grouped such that records of B. lapidarius may also have included
some B. ruderarius; B. terrestris may include B. lucorum; B. hortorum may include B. ruderatus. Individuals not identified to species level are given either as Genus (e.g. Syrphus sp.) or more
generally as ‘hoverflies’, ‘moths’, etc.
 12  Benefits of green infrastructure

you think of the requirements of migratory species. So, resources may be in several different
places for a butterfly: somewhere to lay eggs and food for caterpillars to eat, somewhere else
to pupate, somewhere else to feed as an adult, somewhere else to roost or find a mate and
the places that are used may be used at different times in the year. This concept is called
complementation, that is, the provision of complementary resources (Dunning et al., 1992).
So even the tiniest flower bed, if it has the right plants, can provide ‘habitat resources’ and
be vitally important as food for adult butterflies. But what are ‘the right plants’? Different
insects have different ways of accessing plants: some are generalist feeders and some are
‘specialists’ that have mouthparts that evolved to access nectar from plants with very specific
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

flower structures (different species of bees have different mouthpart lengths, for example),
so the ‘right plant’ may depend on a particular plant species’ flower structure. However, in
urban parks and gardens there can be a mismatch between plants that have been cultivated
for their looks but which are of little value to wildlife. Comba et al. (1999a) compared a
number of horticultural cultivars of plants that had been modified by breeding for ‘looks’
with the uncultivated ‘wild type’ version of the same species for nectar production and the
ability of invertebrate species to get at the nectar. All the species they studied (nasturtium
Tropaeolum majus, larkspur Consolida sp., snapdragon Antirrhinum majus, viola and pansy Viola
x wittrockiana, French marigold Tagetes patula and hollyhock Alcea rosa) produced good levels
of nectar, but the structural flower changes in the commercial varieties tended to modify the
community of species visiting the flowers and reduce the availability of the nectar (although
this was not always the case). Subsequently Comba et al. (1999b) screened 25 native or
naturalised British plants for their value as nectar flowers for bees, butterflies and hoverflies;
the work was carried out in Cambridge and reflected the species assemblage likely to visit
flowers in urban areas. They identified a number of species that were popular with bees and
which were visited by a wide range of species and could thus be of great value in urban
plantings (Table 2.1).
Nectar for adult invertebrates like butterflies is one thing, but what about breeding
‘habitat’ or ‘resources’? Some butterfly species may require areas of long grass to lay eggs on
– a requirement that rules out many of our green spaces. For some species which do not fly
very far, breeding habitat needs to be close to other resources so they are easily accessible but
species that are very vagile (able to disperse widely), like the peacock (Inachis io), can move
around urban areas pretty well to find what they need (in the case of the peacock: nettles

FIGURE 2.3  Left: peacock nectaring; right: female peacock ovipositing on nettle. The peacock is
a highly mobile butterfly easily able to find nectar sources for food and nettles to lay its eggs on
provided they are in the right situation. © John Dover
 Benefits of green infrastructure  13

Urtica dioica) (Figure 2.3). Some species have very exacting requirements for egg-laying sites;
I have seen the wall butterfly Lassiommata megera in a London park finickly selecting where
to lay its eggs on the interface between bare ground and grass where turf had been removed
around the base of an amenity tree to help it grow.
Native plant species are particularly valuable in the urban setting as they are more useful
to our native fauna. For example, Perre et al. (2011) investigated the value of plants from the
Family Asteraceae (daisies) in the urban environment of the campus of the State University
of Campinas, São Paulo, Brazil, to insect herbivores living in flower heads. They found 30
species of Asteraceae on site and whilst only seven of the species were non-native, their
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

abundance far exceeded that of the native species. Despite this dominance, they supported far
fewer species of herbivore (4) than the native species (26), although there was no difference in
abundance between natives and non-natives in herbivore abundance. However, being purist
may miss opportunities; Angold et al. (2006) found that urban habitat patches developed a
flora that contained both native and non-native species in ‘recombinant communities’ which
provided habitat for a wide range of species.

2.2.2  Wildlife corridors

Wildlife corridors can be thought of as linkages helping species move through the urban
landscape; discrete patches of habitat, known as stepping stones, provide a similar function
(Diamond, 1975). It has recently become fashionable to suggest that green roofs may act as
stepping stones or be part of wildlife corridor networks.
If the distance between populations of a species is greater than the distance the species is
able to disperse, then populations are isolated and habitat is inaccessible. In the same way the
urban form (buildings, roads) may act as barriers to movement. This can have serious conse-
quences for species as, all things being equal, small populations on small habitat patches are
more vulnerable to extinction than large populations on large habitat patches. The long-term
viability of a species in a particular area may be dependent on many small habitat patches
exchanging individuals in what is known as a ‘metapopulation’ or ‘population of populations’
(Forman, 1995; Levins, 1970). Habitat or resource patches, if they go extinct for some chance
(stochastic) reason such as fire, disease, predation or parasitism, are unlikely to be recolonised
if isolated by distance or barriers. Dawson (1994) identified five main reasons for employing
corridors and stepping stones:

• To allow a species to escape from an extinction event, or to recolonise a vacant patch

after extinction
• To allow movement between resource patches so that all habitat requirements are met
• To facilitate seasonal migrations
• To promote gene flow between populations; to reinforce flagging populations, etc.
• As an escape route from global warming.

In the context of a particular urban area, the global warming escape function is largely irrel-
evant as it is really a very large-scale attribute appropriate to the regional or national scale
(Hill et al., 1993).
Wildlife corridors and stepping stones were theoretical constructs devised at a time when
spatial ecology was in its infancy and questions were being asked such as ‘Is it better to have
 14  Benefits of green infrastructure

one large nature reserve or several small ones?’ (the SLOSS debate: Single Large or Several
Small) (Diamond, 1975; Game, 1980). Following on from this was the idea that if several
habitat patches were available, but their overall area was too small to support a population,
then perhaps linking them together with strips of habitat (corridors) or putting intervening
small patches (stepping stones) to facilitate movement may effectively make a large habitat
from several small ones. These concepts were enthusiastically taken up by planners, including
those responsible for urban areas, but at the time the concept was not supported by much
in the way of data and this caused some problems in the practical application of the concept
and its defence in planning committees, at least in the UK (Dawson, 1994; Spellerberg &
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Gaywood, 1989). You only have to ask a few questions to see the problems:

• Which species benefit from corridors?

• What are the requirements of those species to get them to use corridors?
• What are the optimal dimensions for corridors for a particular species?
• Will corridors actually facilitate movement of disease, parasites and predators between
populations (Dendy, 1987; Simberloff & Cox, 1987)?

Whilst general design advice on many aspects of spatial ecology has been available for some
time (e.g. Dramstad et al., 1996), planners, without advice from ecologists, may not know
what they are dealing with, and there is still too little data on which species use corridors,
but at least the concept is now much more widely accepted. We do know that some species
are strongly affected by even the smallest interruption in habitat – for example, gaps in hedges
affect dormouse Muscardinus avellanarius dispersal (Bright, 1998), but you don’t tend to get
dormice in urban centres. Fortunately studies are appearing that support the use of corridors
in urban areas such as those carried out in the region of Paris, France, by Vergnes et al. (2012)
who reported the positive effect of corridors on spiders, staphylinids and carabids and by
Vergnes et al. (2013) who worked on shrews.
In designating urban corridors the default, in the absence of evidence, has often been to
simply look for contiguous linear open spaces with some greenery and label them ‘wildlife
corridors’ and assume that they are. Within urban areas the banks of rivers, streams and
canals are often labelled corridors as are networks of green spaces even if they are not all
the same kind of biotope (biotope = a generalist description of a general vegetation type
e.g. grassland, woodland). There are problems in this approach as ‘connectedness’, as we
see it, is not necessarily the same as ‘ecological connectivity’ for a given species (Baudry &
Merriam, 1988). In other words, we may see something as connected such as two woods
linked by a stretch of grass, but for a given species there may be some attribute missing
in the grassland, or some subtle barrier attribute present, such that it cannot be used as a
conduit.

2.3  Direct value to humans

The environment has been credited with an incredible range of values: providing food,
building materials, natural drugs and pesticides (which may also act as starting points for
development of more effective synthetic alternatives), inspiration for works of art (poetry,
paintings, music), social engagement (e.g. volunteering), spiritual value, horticultural
products (including plants), a repository of human history (cultural landscapes, artefacts,
 Benefits of green infrastructure  15

archaeological finds), ecosystem services (local and global), promoting physical and psycho-
logical well-being, study of natural history, acting as an educational resource, and much more
(Lees & Evans, 2003). Not all of these aspects, by any means, will be covered here but I will
highlight some provided by urban green infrastructure; coverage of any issue is not intended
to be comprehensive.

2.3.1  Visual amenity

The incorporation of vegetation into urban areas has an immediate and striking effect: it
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

softens the impact of the hard landscape. In purely visual terms it can screen ugly buildings
and structures, enhance beautiful ones, reduce glare, and introduce colour (Figure 2.4). In
a photomontage-based study of street vegetation in Sapporo, Japan, Todorova et al. (2004)
found that trees along the edge of pavements were the most favoured elements of vegetation
in streetscapes but that other elements enhanced the visual amenity. Responses to photo-
montages manipulating the type of vegetation below trees suggested that neat arrangements
of low-growing bright flowers in single species stands were highly favoured, whereas tall
flowering stands with an irregular structure were not. Moving outside the streetscape, larger
areas such as parks, commons, nature reserves, woodland, can be highly structured and
‘horticultural’ or be managed more for wildlife with a more ‘unkempt’ appearance. These
different management approaches, with different values, may attract different segments of
the population depending on their interests and perceptions of beauty. A recent trend in the
UK has been to use wildflower plantings on wide verges and on roundabouts. The aesthetic
improvement is substantial (Figure 2.5), and may have some biodiversity value as well for
invertebrates using them as pollen and nectar sources. But smaller areas can have substantial
aesthetic benefits including balconies, street planters, doorstep plantings and window-boxes
(Figure 2.6).
Access to green space can clearly contribute to the ability of urban dwellers to take
exercise with both physical and mental benefits, but there are significant benefits in terms
of stress reduction which derive purely from the sensory environment. Pretty et al. (2005a)
compiled the evidence base to promote the countryside for health and well-being in the UK;
the review is useful in an urban context as it draws together a substantial body of data on
a wide range of topics. Various ideas have been put forward to try to explain why humans
respond positively to nature, the most famous is probably the ‘Biophilia’ hypothesis which
suggests that we have a built-in response to natural settings and the lack of them creates
a discord with the environment in which we evolved (de Groot & van den Born, 2003;
Grinde & Patil, 2009; Wilson, 1984). Ulrich (1983) proposed a ‘psychophysiological stress
reduction framework’ whereby nature promotes recovery from stress and other psychological
disorders (Croucher et al., 2007; Morris, 2003). The ‘attention restoration theory’ followed,
as promulgated by Kaplan and Kaplan (1989), and centres on the restorative effects of nature
on the results of prolonged concentration and effort (Croucher et al., 2007). Reid and Hunter
(2011) explored different measures of well-being and the philosophy between different camps
of how well-being happens: the ‘hedonic’ approach being the absence of negative factors
(experiences, emotions, feelings) and the ‘eudaimonic’ view whereby well-being is possible
despite the presence of negative factors. Groenewegen et al. (2006) proposed a study on the
effects of green space on health, well-being and social safety in the Netherlands, at a range
of spatial scales; the document is simply a study protocol, but I like the term they used as a
 16  Benefits of green infrastructure
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.4  Top left: climbers on the railings of a building in Oxford, UK, in April bring a
dash of welcome colour to the streetscape; top right: English ivy (Hedera helix) growing over
ugly, rust-splattered corrugated iron fencing in Stoke-on-Trent; bottom left: concrete fencing
being removed from the Staffordshire University campus in Stoke-on-Trent; and bottom right:
immediately after replacement with ivy green screens (extension panels are provided for the ivy
to grow up). © John Dover
 Benefits of green infrastructure  17
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.5  Wildflower plantings on wide verges and roundabouts in Warrington, UK.
© John Dover

catchy ‘sound-bite’ to sum up the value of green space: ‘Vitamin G’. The following sections
explore aspects of how green infrastructure affects our health and well-being.

2.3.2  Human health

The idea that green spaces have health benefits is not a 21st-century concept. A report
by Natural England ‘Our Natural Health Service’ (NE, 2009b) quoted the Registrar for
Births, Marriages and Deaths who wrote in 1839: ‘A park in the East End of London would
probably diminish the annual deaths by several thousands … and add several years to the lives
of the entire population’; Victoria Park, named after Queen Victoria, resulted in 1850. There
are many excellent reports that explore the relationship between the environment and health,
including those by interest groups such as the Royal Society for the Protection of Birds (Bird,
2004), government agencies including forestry (Henwood, 2001), nature conservation (Lees
& Evans, 2003), health (GSS, 2008), recreation (Pretty et al., 2005a), landscape (Croucher et
al., 2007) and academic reviews such as Morris (2003) and Tzoulas et al. (2007), although
not all of these reports relate specifically to the urban environment.
Green infrastructure can impact on human health in a number of ways, the most obvious
are: by increasing the amount of shade available on hot days, through promoting air circu-
lation, cleaning the air and water of pollutants, helping people to ‘de-stress’ and relax tension,
improving recovery from illness or operations, providing food, community integration
 18  Benefits of green infrastructure
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.6  Even small-scale plantings such as balconies, street planters, doorstep plantings and
window-boxes can have a substantial impact on an area’s aesthetic. © John Dover
 Benefits of green infrastructure  19

and opportunities for physical exercise (see section 2.7 for more on heat effects). The link
between city greening and its potential to reduce healthcare costs was already being made in
the late 1970s (Doernach, 1979). Bird (2004) suggests a single park of 8–20 ha could save the
UK economy annually between £1.6 million and £8.7 million depending on the city; this
represents savings of between £0.3 million and £1.6 million to the National Health Service
(NHS), again depending on the population density of the settlement.
Gerlach-Spriggs et al. (1998) report that the earliest hospitals were in monastic institutions
where gardens were part of the healing process. In the design of the Holloway Sanatorium
(lunatic asylum) for women in the UK (Anonymous, 1872), Thomas Holloway explicitly
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

included gardens and countryside views for their therapeutic effect and the retention of
trees for shade. Green spaces have been shown to have positive effects on human health in
a number of primary studies (de Vries et al., 2003; Maas et al., 2006; Mitchell & Popham,
2007; Takano et al., 2002; White et al., 2013) and reports by influential organisations, such as
the Royal Society for the Protection of Birds (RSPB) (Bird, 2004), Natural England (Lees &
Evans, 2003) and Greenspace Scotland (Croucher et al., 2007). Grahn and Stigsdottir (2003)
in Sweden showed that the more often a person visits green spaces, the less likely they are
to report symptoms of stress. Proximity of green space to home was vitally important, as
respondents to their questionnaire survey claimed that they did not actively seek out green
space to visit elsewhere if there was none nearby. There is a well-established link between
health in different income groups, with those on lower income typically experiencing
poorer environments and thus poorer health, as Chaudhuri (1998) graphically described
for child poverty in Canada. Mitchell and Popham (2008) showed that proximity to green
space helped reduce health inequalities in lower income groups. Fuller et al. (2007) in
Sheffield, UK, showed that the psychological benefit of green space to people increases with
increasing levels of biodiversity. White et al. (2013) found lower levels of mental distress
and higher well-being were correlated with higher levels of urban green space. Luck et al.
(2011) examined the impact of vegetation cover, species richness and abundance of birds
and density of plants on well-being and feelings of connectedness to nature in a study of
1,000 individuals from 36 communities in south-eastern Australia. They found that ‘personal
well-being’ was linked to the extent of vegetation cover of an area, but nature connectedness
was only weakly related to the other variables. Once they had controlled for demographic
characteristics of their subjects (e.g. age, level of activity), they found positive relationships
between ‘neighbourhood well-being’, a measure of satisfaction with the locality, and species
richness and abundance of birds and vegetation cover. Van Dillen et al. (2012) found both
quantity and quality was important for higher levels of (self-reported) health; the findings
held for both green spaces and streetscape vegetation. Even images of ‘restorative environ-
ments’ have been shown to relieve mental fatigue (Berto, 2005). Of particular interest from
the study by White et al. (2013), of 10,000 people from the British Household Panel Survey,
was that they highlighted the importance of small improvements of mental well-being at the
individual level which, when aggregated, were amplified at the community level. The studies
of de Vries et al. (2003), Maas et al. (2006) and Takano et al. (2002) indicated that the effect
of urban greening was strongest in groups most likely to be relatively tied to the locality, that
is, housewives, the old, the young and those with poor educational backgrounds. Taylor et
al. (2002) found that self-discipline was higher in inner-city girls whose homes had more
natural views; the relationship was not found for boys, possibly because boys spent less time
playing in and around the home. Kuo and Taylor (2004), in the USA, claimed that outdoor
 20  Benefits of green infrastructure

activities in green spaces reduced attention deficit hyperactivity disorder (ADHD) symptoms
significantly more than activities held outdoors but in a built environment or indoors.
Mitchell and Popham (2007) using census and economic data for the whole of England noted
that though a general effect can be found between green space and health, the relationship
appeared to depend on income. The expected positive relationship between good health
and green space was evident in urban high-income and low-income groups as well as rural
low-income groups, but there was no relationship in suburban high-income areas or rural
high-income areas; for suburban low-income areas, however, there was a negative association
with green space and health. Mitchell and Popham (2007) suggested that this may indicate a
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

disjunct relationship between quality, quantity and accessibility in poor suburban areas. On
a much broader scale, Donovan et al. (2013) correlated tree mortality (100 million trees lost
in 15 American states as a result of the emerald ash borer beetle Agrilus planipennis) with
human mortality between 1990 and 2007 from cardiovascular disease and lower respiratory
tract illness – they estimated there were some 21,193 excess deaths resulting from the reduced
public health benefits afforded by trees.

2.3.3 Exercise
Exercise is well known to have a wide range of health benefits, and Bird (2004) indicated that
a 3 km footpath could contribute, annually, between £0.1 and £1 million to the economy
depending on the population density of the area, and would save the local National Health
Services (NHS) in the UK between £21,000 and £210,000. One estimate by the UK
Department for Culture, Media and Sport in 2002 (DCMS, 2002) indicated that increasing
the physical activity of adults by 10% would save 6,000 lives and be worth £500 million
annually to the economy. The UK National Institute for Clinical Excellence went further and
suggested that lack of exercise costs the UK £8.2 billion in healthcare costs and absence from
work (NICE, 2008a). The Institute issued guidance on how to promote physical activity in
built as well as natural environments; they also identified what kinds of organisations were
responsible for implementation and made links to policy documents (NICE, 2008b).
For older people, the presence of parks and tree-lined streets has the effect of encouraging
walking and has been shown to positively influence longevity in city dwellers (Takano et al.,
2002). Bell et al. (2008) reported that the body mass index of children was lower in neigh-
bourhoods where there was more vegetation and suggested that this was probably because
more time was spent outdoors in physical activity. Coombes et al. (2010) in a study of 6,821
adults in the City of Bristol, UK, found that reduced accessibility (increasing distance) to
green space reduced frequency of visits, and proximity of a park to home decreased the
probability of individuals being obese. The concept of the informal ‘green gym’ has been
popular for some time, but doctors may even refer patients to organised sessions whereby
exercise programmes are carried out in green spaces (Lees & Evans, 2003). However, the
relationship between physical activity and the perception of better health associated with
green areas is probably not simple. For example, Maas et al. (2008), using a sample of 4,899
Dutch people, examined the relationship between physical activity (walking and cycling),
the amount of green space around the home (within 1 km and 3 km radii of the home post
code) and how it related to feelings of health. Their findings were not straightforward; it
appears that living in a ‘green environment’ resulted in less time walking or cycling for leisure
– although they spent more time gardening (in itself an activity known to promote good
 Benefits of green infrastructure  21

health and reduce stress (Milligan et al., 2004; Van Den Berg & Custers, 2011; van den Berg
et al., 2010)). Fewer commuting trips were made with increasing amounts of green space; but
if commuting was undertaken by cycle, trips were longer if the commuter had more green
space around the home. There did not seem to be a link between the additional time spent
gardening and cycling to work and the relationship between green space and health. Part of
the difficulty with identifying links between health perception and green space was probably
because of confounding factors; for example, a comparison of urban with non-urban areas.
Maas et al. (2008) noted that, in their study, subjects with the most green space around their
homes lived in agricultural areas and hence had bigger gardens (requiring more gardening
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

time). As a result, they were further from major infrastructure (such as shops) which would
require the use of a car to access (compared with those living in urban areas). Additionally, if
they did commute to work, the journey would of necessity take longer.
Pretty et al. (2005b) examined the effect of natural views on physical and mental health in
a laboratory setting by projecting a series of ‘green’ and urban images onto walls whilst 100
volunteers (55 female and 45 male, ranging from 18 to 60 years of age) from the University
of Essex in Colchester, UK, exercised on a treadmill. The images comprised ‘rural pleasant’
such as sylvan grassland, ‘rural unpleasant’ such as groups of abandoned rusting cars, ‘urban
pleasant’ such as tree-lined riverside, ‘urban unpleasant’ such as derelict buildings with
smashed windows, or a control where no image was projected. When analysed by ‘scene’ type
(20 volunteers per category), only those viewing rural pleasant scenes were shown to have
significantly reduced blood pressure measures. When analysed by scene group, self-esteem
was still better in every group following exercise compared with before exercise, but the
response was strongest for the pleasant rural and urban scenes. Mood changes were complex
with Anger-Hostility being significantly lower in urban unpleasant and control groups, but
significantly higher in the rural unpleasant group. Confusion-Bewilderment was significantly
reduced only in the urban pleasant and urban unpleasant groups; Depression-Dejection was
significantly reduced in the urban pleasant and control groups; Tension-Anxiety was signifi-
cantly reduced in the control group but the response was stronger in the rural and urban
pleasant groups; Vigour-Activity responses were significant both in the urban and rural
pleasant groups and in the control group. No significant responses were detected for the
Fatigue-Inertia mood measure.

2.3.4  Recreation, relaxation and socialisation

Socialisation
Green spaces are places where communities and individuals can come together to meet,
celebrate or remember. They may be the only places that community events can be held, or
they may be the place of choice by nature of their setting and the lack of special permissions
that may be needed for road closures and the consequent need to pay for extensive policing.
Events can vary from the purely commercial, such as visiting fairgrounds, to non-commercial,
such as arts festivals or produce shows. Such events allow community groups to have stands to
advertise their services and give demonstrations, combining entertainment with recruitment
and fundraising (Figure 2.7).
Kweon et al. (1998) demonstrated that for older people in Chicago, USA, outdoor
green spaces were important for maintaining social ties engendering a sense of community
 22  Benefits of green infrastructure
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.7  Green spaces are useful for community activities. Top left: morris dancing in
a public park (Portsmouth, Hampshire, UK); top right: nature conservation organisations
promoting their activities (Pershore, Worcestershire, UK); bottom left: community event in a
wood (Pershore); bottom right: a ‘family and friends’ visit to a pond (Fleet, Hampshire, UK). ©
Christine Dover, except bottom right © John Dover

and general social integration. These features are considered to be correlated with various
measures of longevity and well-being (see references in Kweon et al., 1998). Open spaces
can generate both positive and negative concerns about personal safety, which may be quite
distinct from the actual risk. The design of such spaces needs to have regard to the perceptions
of danger that some landscape configurations and management practices might evoke as they
may reduce the willingness to use such spaces (Jansson et al., 2013). Kuo (2003) reviewed
research carried out in Chicago, including that with photomontages, which showed that
residents preferred images of their area with, rather than without, trees; and the more trees
the better. The photomontage studies were confirmed by residents’ actual use of areas with
and without trees. Children also prefer to use green spaces for play, and more creative play was
found associated with green spaces. Green cover also increased social interactions between
adults and between adults and children and decreased graffiti, social disorder and crime.

Natural history study

A number of organisations are now beginning to use the enthusiasm of the public for nature
as a way of engaging them more directly and purposefully by creating participatory initiatives
 Benefits of green infrastructure  23
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.8  A university ecology practical in bird surveying using a green space on the edge of a
housing estate in Staffordshire, UK. © John Dover

that provide data for large-scale projects, typically for monitoring purposes. These are collec-
tively termed ‘Citizen Science’ projects and can be local initiatives or national ones (Pocock
et al., 2014a, 2014b). In some cases they are branded as such, but often they are projects that
are linked to membership organisations and are ongoing. The city of Glasgow, in Scotland,
had the ‘BIG’ project: ‘Biodiversity in Glasgow’ (BTO, 2009) which involved the public in
recording birds and butterflies. A regional initiative in the UK was a hedgerow survey run by
the Cheshire Landscape Trust (CLT, 2013). National projects include the ‘Nature’s Calendar
Phenology Network’ run by the Woodland Trust (WT, 2014), which records the first
appearance of a range of wildlife each year and helps track the influence of climate change,
and the Breeding Bird Survey (BTO, 2014) (formerly the Common Birds Census run by the
British Trust for Ornithology) (Lees & Evans, 2003). Not all natural history study in urban
areas needs to be formal, or organised, or part of a major project; probably the majority
of ‘study’ is likely to be something along the lines of watching birds on a pond or lake, or
idly watching squirrels leap from branch to branch – and is none the worse for that. Green
spaces are, however, the obvious places for ecological training to take place (Figure 2.8),
though green space has the potential to be used for almost any part of a curriculum given a
 24  Benefits of green infrastructure

sufficiently inventive teacher. Local spaces could be used for teaching history, mathematics,
biology, chemistry, philosophy, sports, physics, etc.

2.4  Food production/urban agriculture

Most of the food consumed in urban areas is grown elsewhere. In developed countries
much of it is sold by a small number of large supermarket companies; in small shops on
housing estates, fresh fruit and vegetables, which are essential to improve health, are not
always available whilst cheap processed food is (Garnett, 2000). Encouraging local food
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

production can reduce ‘food miles’, improve food security, provide jobs and have substantial
health benefits including having a better balanced diet, exercise and socialisation, and is
a global phenomenon (Ackerman et al., 2014; Brebbia et al., 2002; Cockrall-King, 2012;
Rich, 2012; Specht et al., 2014; Viljoen & Bohn, 2014) (Figure 2.9). Indeed, in 1996 urban
agriculture was estimated to be practised by some 800 million people (Smit et al., 1996 cited
by Lee-Smith, 2010). The majority of this activity is located in the developing world where
many are dependent on it, but interest in urban agriculture has recently been increasing
in the developed world too, primarily in the USA but now also in Europe and elsewhere
(Lee-Smith, 2010; Mitchell et al., 2014). Urban agriculture is not necessarily a ground-level
activity, and Despommier (2011) explores the possibilities of growing upwards with his
book The Vertical Farm. Urban agriculture is not restricted to the vegetable domain; chicken

FIGURE 2.9  Allotment gardens provide opportunities for local food production and in doing so
provide a wide range of benefits including opportunities for exercise, quiet contemplation and
socialisation; biodiversity should benefit too. This example is from the village of Beer, Devon,
UK. © John Dover
 Benefits of green infrastructure  25

keeping, primarily for eggs, has recently become very popular in the UK (Rustin, 2012),
but small animal production including goats, rabbits, guinea pigs, snakes, fish farming and
beekeeping are all practised in urban areas (Iaquinta & Drescher, 2002). Local laws may not
encourage livestock, however, as LaBadie and Scully (2012) found with chicken keeping in
Iowa. Fleury (2002) identifies other features of society enhanced by the introduction of urban
agriculture, including a sense of identity, the strengthening of local heritage values and the
creation of a distinct local landscape. De Lange (2011) found food production was often not
the primary aim of urban agriculture in Amsterdam: it was frequently used as a vehicle for
education and promoting social cohesion.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Food production typically takes place in private gardens, allotments, city farms and
community gardens, school gardens, community orchards, some parks, and even on balconies
and roofs. Edmondson et al. (2014) found soils in gardens and allotments in the City of
Leicester, UK, to be superior to that of farmland surrounding the city which was subject to
intensive, conventional agriculture. Generally speaking, produce grown in gardens has been
shown to have levels of contaminants, such as lead, which are within acceptable levels, but
some have exceeded such ranges and, in a study of vegetables grown in urban gardens by
McBride et al. (2014), this was associated with particles adhering to leaves or from aerosol
deposition. In a test of contaminant levels in the soil of 54 New York City community
gardens, 78% of the 564 samples taken were shown to be within safe limits set for all ten
metal contaminants studied. However, this meant that 70% of the sites had some contami-
nated samples (principally barium: 46% of gardens, and lead: 44%) (Mitchell et al., 2014).
Contaminants were identified less frequently in raised beds, where clean soil and compost
had been imported onto the site, and may be considered a viable ‘risk reduction’ method
along with other practices including careful washing and preparation of food (see references
in Mitchell et al., 2014) and growing fruit rather than vegetables (McBride et al., 2014).
Clearly caution should be exercised when growing food on abandoned land due to the risk
of contamination from old industrial processes, through using contaminated water, and the
inappropriate use of pesticides (Garnett, 2000; Madaleno, 2002; Ranasinghe, 2009; van den
Berg, 2002).
One lovely example of how creative thinking can produce food and have wider benefits
was that reported by Lees and Evans (2003): many allotments become run-down and are
sometimes threatened by closure, due to the small number of active plot holders (EFF, 2006).
A 4.85 ha (12 acre) site in Cowley, Oxfordshire, UK, instead of being closed was regenerated
by planting a small woodland, an orchard and a wildflower meadow. To encourage families
with children to take up plots, the development also included play areas. The changes made
to the site have had a much wider community impact, including being used by local schools
and being part of a visual arts festival. In terms of biodiversity, birds, at least, do benefit from
allotments, though Müller (2007, cited by Strohbach et al., 2009) in his study of bird diversity
in Leipzig’s (Germany) allotments only found those species that are generally associated with
human settlements rather than any rarities.
Whilst urban parks are typically maintained by local authority staff and with visual
amenity in mind, there is a trend to allow food production to take place in defined areas
(Garnett, 2000) and some parks have had fruit and vegetable growing for some time. The
National Urban Park in Stockholm, Sweden, which is biodiversity-rich, has six allotment
areas in Söderbrunn, Kvarnvreten, Frescati, Jakobsdal, Bergshamra and Stora Skuggan.
The first allotment in Stockholm was dug in 1904 at Djurgården and the others were put
 26  Benefits of green infrastructure

in place over the following 50 years. In area terms, the allotments take up a very small
part of the park, but are still considered to add value in ecosystem services (pollination,
seed-dispersal, pest control) (Barthel et al., 2005). Even sophisticated Paris has its urban
agriculture; for example, Montmartre has communal vines, as do some public parks
(Fleury, 2002).
Urban orchards are also undergoing something of a renaissance with the interest
in growing local food. Several reasons for this interest are evident in addition to food
production: their heritage and biodiversity value are perhaps the most obvious additional
benefits, but social and economic capital has also been identified as accruing from community
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

orchards (Countryman, 2009). Traditional orchards, that is, those not under commercial
production and which use older fruit varieties, are known to be valuable for saproxylic inver-
tebrates – those that live on, or in, dead and decaying wood. Species particularly associated
with old orchards in the UK include the beetle the noble chafer Gnorimus nobilis and the
group of specialist species that live on mistletoe Viscum album (which is itself strongly, though
not exclusively, associated with fruit trees) (Biggs, 2003; Williams, 2000). Old orchards
are mainly found in the countryside, but some may have become encapsulated by urban
development. One such is the La Sainte Union School Orchard in Kentish Town, London.
The school moved to the site in 1864 and still has the original convent orchard with apple
Malus domestica, pear Pyrus communis and quince Cydonia oblonga trees (Cruikshank, 2001).
Urban ‘orchards’ may not resemble traditional orchards: the University of East London has
an urban orchard which consists of six fruit trees (two apple, two pear, one plum Prunus sp.,
one cherry Prunus sp.) planted in half-barrels at its Stratford campus (UEL, 2011). Another
example of a small orchard is a collection of 12 crab apple trees Malus sp. in St John’s Gardens,
Lower Byrom Street, Manchester, UK. The trees, along with the rest of the floristically
rich garden, started out as a winning demonstration display, designed by Daniela Coray, at
the Royal Horticultural Society’s flower show at Tatton Park, Knutsford – after the show it
was transferred to St John’s Gardens and has been adopted by local residents who manage it
(CityCo, 2011). In Chicago the ‘Chicago Rarities Orchard Project’ is trying to encourage
the planting up of unused open spaces using rare varieties, using the plantings for educational
and general open space use. Interestingly, historical research has shown one of the targeted
areas on Logan Square was once planted up with trees and one orchard produced some 600
bushels of cherries in 1871 (CROP, 2009, 2012)!

2.5  Transport (walking/cycling)

Green spaces themselves, and traffic-free links between them, provide movement corridors
for wildlife, but also act as ‘zero-carbon’ transport corridors for people either as pedestrians or
cyclists (Chapman, 2007) (Figure 2.10). Such links are expected to promote an active lifestyle
with social benefits and improved physical and mental well-being as well as helping to reduce
traffic congestion (Goode, 2006; Humphreys et al., 2013; Martin et al., in press; Mell, 2007;
Ogilvie et al., 2011; Pucher et al., 2010; Wen & Rissel, 2008). A cost-benefit analysis carried
out by Saelensminde (2004) showed that investment in cycle networks can return benefits of
four to five times the cost of implementation. Green infrastructure can provide more pleasant
and safer routes than roads, factors that have been shown, or suspected, to be important in
encouraging commuters to walk or cycle rather than use cars (Panter et al., 2014; Pucher &
Dijkstra, 2000; Sonkin et al., 2006). In the UK, the Connect2 initiative developed 79 projects
 Benefits of green infrastructure  27
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.10  Zero-carbon transport. Green spaces and the links between them provide
opportunities for pedestrians and cyclists to travel in pleasant surroundings away from polluting
vehicle traffic. © John Dover

across the UK to encourage walking and cycling (Ogilvie et al., 2011) and in Southwark,
London, links between parks have been developed with planting improvements in streets (LS,
undated). The provision of new routes, such as those carried out under Connect2, does appear
to have increased walking and cycling after two years (Brand et al., 2014; Goodman et al., 2014).
Chapman (2007) reviewed the literature on transport and climate change and suggested
that walking and cycling were viable approaches to reducing cars. However, an analysis by
Brand et al. (2014) suggested that the increased walking and cycling activity resulting from
the Connect2 programme had not yet been translated into a substantial reduction in CO2
emissions from transport. Instead, they suggested that the increase in cycling and walking
they recorded was probably not due to more people engaging in the activities, and hence
reducing the number of car journeys, but rather a result of those already using walking and
cycling as a mode of transport changing their patterns of use in response to the new routes.
Brand et al. (2014) considered that provision of opportunities for walking and cycling may
need to be coupled with disincentives relating to car use (e.g. higher parking charges, fuel
taxation, road tolls, etc. (Chapman, 2007)), incentives and other policy measures. Worryingly,
Steinbach et al. (2011), using London, UK, as a case study, found that cycling as a commuting
 28  Benefits of green infrastructure

activity had an ethnic, gender and class bias towards professional white males, and this is
clearly not a desireable situation. Various schemes have been tried to improve the use of
walking and cycling over the use of cars. The evidence from previous work, reviewed by
Ogilvie et al. (2004), is that general schemes such as publicity campaigns have little impact
whereas those targeted specifically at behaviour change of an already motivated group
generally work, though with differing levels of effectiveness. Some approaches including
engineering solutions, financial incentives, and alternative approaches, such as telecom-
muting, have had mixed fortunes, with some having the opposite effect to that desired. Green
infrastructure is potentially an excellent way of improving population health and reducing
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

fossil fuel consumption by encouraging walking and cycling, but the simple provision of
GI without other support programmes may mean that its value is not fully realised; there is
clearly a data lack in this area.

2.6  Economics/enhanced building values

Doernach (1979) predicted that greening urban structures (green roofs, green walls) would
generate new jobs; whether the scale of his prediction of 25,000 in Germany alone has
proved to be accurate is open to question, but new companies have certainly sprung up to
service this new industry.
Does biodiversity make a difference in terms of economic benefits to an area? Benefits
can be both tangible and intangible in financial terms as Abdullah et al. (2002) found in
their study of the building and operation of a new large park in the centre of Kuala Lumpur,
Malaysia. The park was built in 1998 as part of a large mixed development including a
large shopping centre complex, hotel and associated high-rise buildings (the largest being
88-storey twin towers) in an area which was rather low on green space. The park itself is
20  ha in size, designed by the Brazilian landscape designer R.B. Marx to be biodiversity
rich, especially for birds. It has about 1,900 indigenous trees and palms from 74 species,
some of which were transplanted from the old Selangor Turf Club, and also provides a range
of recreational amenities (Abdullah et al., 2002; SKLCC, 2014). Park use is 10,000–20,000
on weekdays rising to 60,000–70,000 at weekends. Interviews and questionnaires revealed
that the shopping centre was used by 42% of visitors before visiting the park and 21% after-
wards. The park is used as a marketing feature, along with the shopping centre and high-rise
twin towers, to attract businesses into the area. The shopping centre’s success was strongly
associated with its location (next to the park) both by the management of the centre and by
visitors who cited the park as being important. Just over half of park visitors gave it as their
main reason for visiting the area, with others in the park primarily intending to visit the
adjacent buildings. The Mandarin Oriental Hotel and local cafés reported high demand for
outdoor seats overlooking the park, preferring them to indoor seats. The park is considered
to have reduced congestion, attracted visitors and contributed to making Kuala Lumpur ‘a
comfortable city’; it is also a low crime and vandalism area. Maintenance costs of such a large
park are equally high and are paid for by the 22 ‘lot holders’; given that these are commercial
organisations and payments come out of their profits, continued support may be vulnerable
to economic slumps.
Studies in the USA have shown that vegetation can enhance the business prospects of an
area (Wolf, 2005). McPherson (1988) indicated that vegetation near residential areas could
increase house prices and also increase feelings of privacy. A report for the UK Commission
 Benefits of green infrastructure  29

for Architecture and the Built Environment (CABE) called ‘Does Money Grow on Trees’
(CABE, 2005) concluded that:

• there is a positive relationship between domestic property prices and proximity to a

high-quality park;
• the enhanced value is affected by local factors including the type of park, the type
of property, the layout of buildings around the park, the location and the local
population;
• a park’s impact is enhanced when it is linked with tree planting along adjacent streets,
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

and where the park’s vegetation can be seen down streets or over rooftops;
• a property that looks onto a park will command a higher premium than one that backs
onto it;
• housing near parks without associations with crime will generally command a premium
compared with those where there is.

Because of the multiplicity of factors and the wide range of park types considered in the
study, there was a wide range in estimated uplift from 0 to 34% for living in the vicinity of a
park and 3 to 34% for overlooking a park. In the latter situation most properties had a 5 to
7% uplift in value. These findings seem to hold generally: Kong et al. (2007) demonstrated
a housing price uplift due to proximity to green space in their case study of Jinan City in
China; Tyrväinen and Miettinen (2000) in the district of Sale, Finland; Melichar and Kaprova
(2013) in Prague, Czech Republic; Pandit et al. (2013) in relation to street trees in Perth,
Western Australia. Mansfield et al. (2005) gives the uplift estimates from a number of studies,
and probes the impact of different definitions of urban forest in relation to the housing market
(e.g. trees on private land compared to large forest parcels and their ecological value) whilst
Gatrell and Jensen (2002) examine how different urban forest greening approaches fulfil
economic local development goals.
Gore et al. (2013) reviewed the literature relating to green infrastructure’s impact on the
economic growth of a locality and concluded that:

• having high-quality parks in an area increases inward investment and local property
prices;
• an area’s attractiveness and the quality of parks has an effect on visitor numbers and their
spending;
• green infrastructure provides ecosystem services such as pollution control, sustainable
urban drainage and temperature moderation, and at a reduced cost compared to more
technological approaches. Reduced costs associated with damage (e.g. from flooding)
allow greater local investment;
• access to green space reduces mental health/stress issues and promotes exercise resulting
in increased productivity;
• local food production makes a (very) small contribution to the local economy;
• creation and servicing of green infrastructure creates employment opportunities.

Whilst ecosystem service valuation has started to become influential in policy terms (ten
Brink et al., 2009), valuation of specific services in specific urban areas (other than for
pollution – see below) are infrequent. Zhang et al. (2012) calculated the reduced runoff
 30  Benefits of green infrastructure

in Beijing, China, due to its 61,695 ha of green space to be 154 million m3.y-1 (or 2,494
m3.ha-1), which equated to an economic value of RMB 1.34 billion in 2009 in runoff
reduction. As might be expected, the reduced runoff was uneven across Beijing, varying with
locality due to different proportions of green space.

2.7  Climate control/climate change proofing

2.7.1  Health effects of heat

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Too much heat can be lethal, with the old being particularly susceptible; other vulnerable
groups include children, those in hot-environment occupations, the physically or mentally
ill, alcoholics, tranquiliser users and those living in the upper stories of tall buildings
(Hoffmann et al., 2008; Ishigami et al., 2008; Kilbourne et al., 1982; McMichael et al.,
2006). However, Rey et al. (2007), in their study of heatwaves and ‘ordinary’ summer
temperatures over the period 1971–2003 in France, concluded that no sector of the
population was immune to heat effects. Mortality is typically highest in the most urbanised
and deprived areas (Rey et al., 2009) as Johnson and Wilson (2009) found for heat-related
deaths resulting from the 1993 heatwave in Philadelphia. The heatwave of 2003 is thought
to have caused 2,000 excess deaths in the UK (Defra, 2013b) and 15,000 in northern
France (Defra, 2013a). In Paris alone the 2003 heatwave resulted in over 2,600 extra
emergency hospital visits, 1,900 extra admissions and 475 excess deaths (Dhainaut et al.,
2004). Reducing the impact of the heat island effect (see below), which will be exacerbated
by climate change and the increased frequency of predicted extreme events, is clearly of
importance for human health (Golden, 2004). The Heatwave Plan for England was initially
developed in 2004 and explicitly mentions green infrastructure as a way of reducing heat-
related mortality (Defra, 2013a) and O’Neill et al. (2009) include trees and vegetation as
ways of ‘preventing heat-related morbidity and mortality’ (see also Kilbourne et al., 1982).
Climate change has potentially considerably more impact on health than mortality directly
caused by heat. Predicted impacts include an increase in: cold-related illness and deaths (as
a result of extreme events); flooding, resulting in drowning and sewer overflows (leading to
disease and psychological trauma); algal blooms releasing toxins into reservoirs; incidence
of food poisoning and water-borne disease; infections resulting from invertebrate vectors
such as ticks and mosquitoes; skin cancer; pollution-related health impacts (Kovats, 2008;
McMichael et al., 2006).

2.7.2  The urban heat island effect

The city climate differs from that of the countryside around it (Oke et al., 1999; Stülpnagel
et al., 1990; Taha, 1997; Villiger, 1986), with northern hemisphere towns and cities having
lower relative humidity, less sun, more cloud, more snowfall and more rain than rural areas
(Kusaka, 2008; Shepherd et al., 2002; Taha, 1997). Perhaps the most well-known facet of
city climates is the ‘urban heat island effect’ whereby cities are warmer, by several degrees
Celsius, than the surrounding countryside, and the larger the city the larger the temperature
difference (Kusaka, 2008; Oke, 1973; Sieghardt et al., 2005). Whilst a warmer urban area
might be considered a benefit during winter (McPherson, 1994) (‘the winter heat island’
(Whitford et al., 2001)), as experienced in Tokyo (Ichinose et al., 1999), it is most definitely
 Benefits of green infrastructure  31

a negative impact in the summer, making urban areas unpleasantly warm. The impact is felt
most at night, and especially when the weather is calm and skies are clear, potentially resulting
in ‘heat stress, circulation problems and insomnia’, especially if high temperatures last for
some days (Stülpnagel et al., 1990). The heat island also exacerbates other urban problems
such as increasing levels of pollution (Rosenthal et al., 2008). The general pattern of temper-
ature difference between built areas and vegetation becomes pronounced mid-afternoon and
continues building during the day, eventually becoming greatest at night as buildings, roads,
etc. release their stored heat energy more slowly than vegetation (Gedzelman et al., 2003)
and heat release is impeded by the reduced ‘sky view’ available as a result of tall buildings
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

(Kusaka, 2008; Oke, 1981). This ‘sky view factor’ is important as it is the fraction of the sky
to which the urban surface can radiate heat to, and this is lower in densely built-up areas
with many tall buildings than (say) in parks and the countryside (Graves et al., 2001). The
heat island then dissipates rapidly towards daybreak and, because built-up areas warm more
slowly than vegetation, pockets of cool air can develop in the early morning soon after dawn
(Gedzelman et al., 2003; McPherson, 1994). The development of a heat island and its dissi-
pation also needs to be seen in the context of urban versus rural interactions. At night, as
heat rises above the urban areas, cool air from the rural perimeter replaces it, and a system
known as the Urban Heat Island Circulation can develop (Haeger-Eugensson & Holmer,
1999). Oke (1979) made a distinction between the ‘canopy-layer heat island’ (CLHI) from

TABLE 2.2  Factors influencing the development of the urban heat island (Nitis et al., 2005)

Radiation balance

↑ anthropogenic heating
↑ absorption of short-wave radiation
↓ outgoing long-wave radiation
↓ long-wave radiation loss (reduced sky)
↓ evapotranspiration
↓ heat transport

Physical factors
topography
relief
urban structure
building density
building materials

Urban–rural differences in:

albedo moisture
roughness length
thermal capacities
 32  Benefits of green infrastructure

about roof height to the ground and the ‘boundary layer heat island’ (BLHI) above building
levels. The CLHI is most affected by local urban processes, whilst the BLHI is more affected
by larger-scale processes (Golden, 2004). However, there is clearly an interaction between
the boundary-layer heat island and mesoscale processes, as the urban heat island has been
demonstrated to influence the wider (mesoscale) climate, with increased rainfall downwind
of cities (Shepherd et al., 2002). Nitis et al. (2005) document the factors which contribute to
the urban heat island effect (Table 2.2).
The heat island phenomenon is not new, Golden (2004) tracks the earliest publication on
it back to Howard (1833). New York may well have had an urban heat island since the late
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

1800s, increasing in intensity as the city has grown (Rosenthal et al., 2008). Recent estimates
give New York’s heat island effect as approximately 4°C in summer and autumn, with a
winter heat island (including spring) of about 3°C (Gedzelman et al., 2003). London’s heat
island was estimated to be 3°C higher than average peak temperatures in the surrounding
countryside in the summer of 1999 (Graves et al., 2001). A number of studies have identified
long-term trends in increasing urban temperatures (Brazel et al., 2000; Wilby, 2003). The
additional heat comes from darker surfaces (buildings, roads, pavements) absorbing more solar
heat, energy use for industrial processes, transport, air conditioning and reduced amounts of
vegetation resulting in less shade (e.g. from trees) and evapotranspirative cooling (Akbari et
al., 2001) and, because rainwater is diverted to drains, there is also less water for evapotran-
spiration (Taha, 1997).

2.7.3  Reducing the heat island effect with vegetation

Buyantuyev and Wu (2010) considered that vegetation was most efficient at reducing the
heat island effect during the day as most plants close their stomata at night reducing transpi-
ration. Stülpnagel et al. (1990) noted that on calm nights there could be a 9°C difference in
temperature between the centre of Berlin and the surrounding countryside. Buyantuyev and
Wu (2010) found the same difference when comparing nearby vegetated and non-vegetated
parts of Phoenix, Arizona, and Gedzelman et al. (2003) estimated New York’s at 8°C. The
heat island is subject to modification according to local, seasonal and periodic weather condi-
tions, with winds typically reducing it or shifting it downwind from its ‘normal’ location
(Gedzelman et al., 2003). In his German thesis, Stülpnagel (1987) (given in English in
Stülpnagel et al., 1990) concluded that the impact of a vegetated area on the local climate:

• related to area (the bigger the green area the better)

• extended further downwind than upwind
• extended further at higher windspeeds
• extended further in urban areas that are more open and have additional green spaces
• showed vegetated areas that are situated in a dip, or are enclosed (by walls, tall vegetation)
have less impact
• showed internal fragmentation of green areas by sealed surfaces (e.g. roads) reduces their
value.

Increased summer urban temperatures result in increased electricity consumption for air
conditioning (Akbari et al., 2001) and hence increased CO2 emissions that contribute to
global warming (Akbari, 2002). As a result, the carbon reduction value of a shading tree in
 Benefits of green infrastructure  33

urban Los Angeles was considered by Akbari (2002) to have a greater value than an equiv-
alent forest tree by three to five times in terms of carbon reduction (urban tree saves 18 kg
carbon, of which 4.5–11 kg is direct sequestration in tree tissues and the rest from reduced
electricity use).
Most vegetation is green because plant leaves contain pigments that absorb red and blue
visible light; carotenoids take in blue light and chlorophyll takes in red and blue light, other
pigments (flavonoids) absorb ultra-violet (Yoshimura et al., 2010). Contrast the relatively
soothing colour of plants with that of the two main urban ground surfaces: a typical concrete
plaza or walkway/pavement which is pretty close to white and asphalt/tarmac roadways
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

which are black and age to grey. Light surfaces (i.e. those with a high albedo (Taha, 1997))
reflect light (and heat) and black surfaces do not (Akbari et al., 2009). Light surfaces, in
reflecting heat, have their place in reducing the urban heat island and light-coloured paving
and roofing is recommended for that reason (Akbari et al., 1992, 2009). However, on sunny
days many people walking in paved areas with white or light surfacing (such as concrete)
would find that they would need sunglasses to avoid squinting or risk UV damage to their
eyes (Heisler & Grant, 2000), whereas light surfacing on roofs would not impact on human
vision. At the opposite end of the spectrum are tarmac road surfaces which, being black,
reduce the need for pedestrians, drivers, etc. to wear sunglasses but, by absorbing heat and
later re-radiating it, create uncomfortably warm conditions and contribute to the urban heat
island (Pomerantz et al., 2000). Asphalt surfaces do lighten with age though: when freshly
laid they have a very low albedo (0.04–0.05) after five years this increases to 0.12±0.03 until
replacement after about 10–15 years (in California) (Pomerantz et al., 2000). Pomerantz
et al. (2000) have suggested that asphalt surfaces could practicably be given higher albedos
by surfacing with lighter coloured chippings achieving (say) an albedo of 0.35 by using
chippings with an albedo of 0.5. Of course albedo is subject to seasonal variation especially
with snowfall, loss of leaves from deciduous trees, etc. and urban structure (Brest, 1987)
(Table 2.3). Even under snow conditions albedo is variable, with flatter surfaces having
higher albedo than those with substantial vertical structure like trees; urban areas retain even
snow cover (and thus high albedos) for a shorter time than rural areas as roads are cleared for
traffic, pavements for pedestrians, and heat from poorly insulated buildings causes snow melt
on roofs (Brest, 1987).
Shade cast by tall vegetation, such as trees, can substantially reduce the surface temperature
of sealed surfaces (Fintikakis et al., 2011) (Figure 2.11). Plants also, by transpiring, cool the
ground (on a warm day try touching a road or path surface then some grass). But, taller
vegetation is cooler than shorter vegetation by some considerable margin. Jeremy Thomas,
whilst trying to unravel the ecology of the Adonis blue butterfly (Lysandra bellargus), showed
that there was an 8°C difference between the ground temperature when grass was 1 cm high
(24°C) compared with grass 10 cm or more high (16°C) (Thomas, 1983). Studies carried
out in municipalities such as Stuttgart and Munich in Germany show that large parks,
and especially those with woodland and trees, are some of the coolest parts of the urban
environment being 2–3°C cooler than surrounding areas (Tyrväinen et al., 2005). Peters and
McFadden (2010) demonstrated that soil and surface temperatures were negatively correlated
with increasing number of trees and leaf area index. In a study of Greater Manchester, UK,
28 ‘urban morphology’ types were recognised and their proportion of sealed surface:green
space estimated (Handley & Gill, 2009). Maximum surface temperatures were clearly
positively correlated with proportion of sealed surface. Highest temperatures (>30°C) were
 34 Benefits of green infrastructure

TABLE 2.3 Albedo of different land-uses derived from Landsat imagery of Hartford, Connecticut, on
snow-free days (Brest, 1987)

Location Albedo (%)

City centre ‘downtown’ 08–12
High density residential 08–12
City outlying 11–15
Suburban 11–17
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Trees 07–19
Wetland 05–20
Agriculture 11–19
Range 11–19
Non-forest park 17–23
Lake 00.2–5

found in the town centre with about 20% vegetation; the coolest (<20°C) were categories
with over 90% of their surface vegetated (or, as they say, evaporative) such as informal open
space, woodland, or rivers and canals (Handley & Gill, 2009). Studies in a Manchester park
showed air temperatures to be, on average, 0.8°C cooler in the park compared with the
surroundings. Grass was shown to be substantially cooler than sealed surfaces and that tree
shade had a cooling effect on surface and air temperatures (Armson et al., 2012). Grass and
trees (because of shade and evaporative cooling) were considered by Armson et al. (2012) to
be better in combination than alone, and capable of producing surface temperatures between
4 and 7°C below that of the air. The urban forest in Beijing was estimated to cover 16.4%
of the city area and reduce air temperature by 1.6°C which, because of a reduced load on air
conditioning systems, also reduced electricity use by 0.238 GWh (Yang et al., 2005). Parks
in Singapore have been shown to have lower temperatures than built-up areas with the effect
extending into the adjacent areas (Chen & Wong, 2006).
The cooling effect of urban vegetation extending into the surrounding areas is clearly
an area that excites interest (e.g. Chen & Wong, 2006) but the distance within which the
effect is evident is relatively short, only 100–500 m at best (Chen & Wong, 2006; Tyrväinen
et al., 2005; Wilmers, 1988, 1991), suggesting that urban areas need an extensive network
of green spaces and street vegetation for their effect to be maximised. Interestingly, Natural
England’s minimum set of standards for accessibility to green space (ANGSt: Accessible
Natural Green space in Towns and Cities) includes one criterion which is: ‘that no person
should live more than 300m from their nearest area of natural green space of at least 2ha
in size’ (Handley et al., 2003), a figure that dovetails rather well with the estimate of how
far a cooling effect extends from green space! Buyantuyev and Wu (2010) contributed
an increased understanding of the heterogeneity of urban heat islands in their study of
Phoenix, Arizona; they concluded that heat islands should not be considered a uniform
entity, but a mix of hot and cold spots with temperature differences of such areas sometimes
exceeding the difference between the urban centre and the surrounding countryside. They
coined the term ‘urban heat archipelago’ to describe the scatter of hot and cold patches
 Benefits of green infrastructure 35

35
Surface temperature reduction by

30

25
shading (°C)

20

15
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

10

0
White walls Yellow walls Asphalt Light Dark Grass
pavements pavements
Surface type and colour

FIGURE 2.11 Reduction in the surface temperature (°C) of various urban surfaces in the city of
Tirana, Albania, when shaded. Data from a microclimatic survey carried out in 2008, source
Fintikakis et al. (2011).

within urban areas (Buyantuyev & Wu, 2010). Giannopoulou et al. (2011) used a network
of 25 fixed locations to examine the urban heat island in five zones of Athens, Greece, and
demonstrated substantial differences (5°C) within the city with the highest temperatures in
the city centre and in the western industrialised zone. City parks are now being considered
as ‘cooling islands’ and sophisticated approaches are being used to model their impact on
the urban heat island (Declet-Barreto et al., 2013; Vidrih & Medved, 2013). Arnfield
(2003) pointed out that urban climate should be considered at different scales from the
individual components to the whole city (Table 2.4). Recent work by Connors et al. (2013)
suggests that very-fine-grain modifications to spatial patterns, such as complexity of grass
patch shape, use of high albedo roof surfaces, shade cast by vegetation and tall buildings
(Lindberg et al., 2003), etc., may have an impact on local temperature. Rosenthal et al.
(2008) have produced a report on mitigation approaches for New York’s heat island which
includes a range of measures including low albedo surfacing, green roofs and increased tree
planting. Onishi et al. (2010) examined the potential of tree planting in car parks as a way
of mitigating the urban heat island in Nagoya, Japan. By greening a car park with 30% tree
cover and 70% grass their simulations indicated that summer temperatures could be reduced
by 7.26°C (Onishi et al., 2010); whilst not perhaps being practically achievable in area
greened, the study certainly suggests that greening car parks could reduce small hot spots
(Figure 2.12). For Atlanta, modelling studies indicated that a 100% increase in vegetation
cover or a tripling of its albedo could reduce the UHI; on a practical basis increasing the
amount of vegetation was considered to be the more feasible option (Zhou & Shepherd,
2010).
 36  Benefits of green infrastructure

TABLE 2.4  Hierarchical nature of urban settlements based on Arnfield (2003)

Scale Type Components

0 Individual surfaces Walls, roofs, gardens, lawns, parks, road, etc.
1 Building Aggregation of walls + roof
1 Ground-level Mix of irrigated and non-irrigated green space components (including
gardens) and sealed surfaces
2 Urban canyon Buildings and space between (e.g. roads)
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

3 City blocks Urban canyons + building roofs

4 Neighbourhoods Collection of city blocks
5 Land-use zones Urban centre, industrial, residential, etc.
6 Entire city Aggregation of land-use zones

FIGURE 2.12  Trees shading car parking places in France help keep local temperatures down and
potentially reduce fuel evaporation. © John Dover
 Benefits of green infrastructure  37

2.7.4  Plants and wind

Vegetation that has a lot of structure, e.g. trees, shrubs and hedges compared with cut lawns
and flower borders, will also have the beneficial effect of reducing windspeed. Farmers
have always known this, planting lines of trees (windbreaks) or tall hedges to slow it down
around sensitive vegetation, or shorter hedges to protect animals (Pollard et al., 1974). If it
is very windy and warm, plants can wilt as they try to combat the heat through opening
their stomata and having the water evaporated off faster than it can be replenished via the
roots; plants can also be killed by wind-chill on cold days as the wind increases the exposure
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

factor. It always surprises me how many architects and urban designers still create draughty
urban plazas and roads without any apparent thought of the need to stop the wind tunnel
effect (Figure 2.13). Klaus Tham designed the Bo01 development in the Western Harbour
area of Malmö with the need to create shelter in mind – tall buildings were placed along
the water’s edge to shield lower-level buildings behind them; the street layout was more
sinusoidal than straight and the streets narrow (Kruuse, 2011). A rough rule of thumb is
that the lee-side shelter footprint of an object (hedge, tree, building) extends to 12 times
its height (Lewis, 1965); so a hedge 2  m high would throw a shelter footprint of 24  m.
Of course the shelter effect is greatest near the structure, within 2 times and 3 times of its
height (Lewis, 1965), and the effect is greatest with windbreaks that have a dense structure
(i.e. those that do not let a lot of wind through) (Lewis & Stephenson, 1966). However,

FIGURE 2.13  On a summer’s day this walkway in Liverpool, UK, is cooled by a strip of
mini-fountains, though the surface reflectance (albedo) is high and sunglasses are needed. In the
winter it is likely to be a cold wind-tunnel. © John Dover
 38  Benefits of green infrastructure

the optimum permeability is considered to be 40% (Pollard et al., 1974) with very dense
structures resulting in vortices on the lee side and a rapid return to the windspeeds found
on the upwind side. Heisler (1990) compared windspeeds at 2 m above ground at a local
airport with those in residential areas with no trees and those with the same building
density and different levels of tree cover (24–77%). Results indicated that the buildings
reduced windspeed by 22%, but in the area with 77% tree cover average windspeeds could
be reduced by 70% in summer and 65% in winter.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

2.7.5  Climate change mitigation

Towns and cities are wetter than surrounding areas because the hot air from sealed surfaces
and heat-sources rises promoting cloud formation (and hence rain) (Stülpnagel et al.,
1990). Climate change predictions indicate drier summers and wetter winters with more
extreme events in both (Solomon et al., 2007). Green infrastructure can help cope with
increased storm events by reducing the likelihood of flooding that would result (see section
2.8 Sustainable urban drainage below) and with heatwaves by evapotranspiration and
shading. Climate change is likely to exacerbate urban heat island effects (Wilby, 2003).
The importance of green infrastructure was underlined by climate change modelling of the
Greater Manchester data (Handley & Gill, 2009) described above. This work suggested that
increasing the area of vegetation in the town centre by 10% would keep temperatures in
the range experienced between 1961 and 1990 under both high and low emission scenarios
for 2020 and the low scenario for 2080; for the high 2080s scenario the 1961–1990 range
would just be exceeded using 10% vegetation (Handley & Gill, 2009). Unfortunately, to
increase vegetative cover requires unused open space if conventional ground-level planting
is used. Solecki et al. (2005) note, from their study of the urban heat island effect in New
Jersey, that the more deprived, high-density residential areas of inner cities are less likely
to have capacity for such solutions as tree planting than more affluent lower density areas.
Greening solutions in such situations will require modifications to building surfaces (green
roofs, green walls) to achieve thermal benefits; an alternative is to use high albedo roofing
materials (Solecki et al., 2005).
Climate scenarios that predict more extreme temperatures are likely to impact on all
sectors of society, but most severely on the elderly and frail (Ishigami et al., 2008). With
elevated temperatures, people may become effectively trapped inside their homes and
offices, reducing socialisation and opportunities for exercise. In such circumstances it is not
unlikely that air conditioning will be increasingly relied upon to moderate internal building
temperatures: increasing costs and contributing to greater emission levels of greenhouse gases
(which then feed into increased climate change) – of course if there are also extreme lower
temperatures in winter heating costs will increase as well. If the likely impacts of climate
change and the role of vegetation in mitigating those effects are clearly understood by local
authorities, businesses, and industry, relatively low-cost action can be taken now to meet the
threat in future years when it will cost more to combat using more technological approaches.
For example, it costs less to plant a small tree than a large tree (and survival and growth
may well be superior). Consider the case of a care home for the older members of society
where residents have access to a patio or grassed area for exercise, or simply to take the air:
planting with suitable shade trees will allow the area to continue to be used in periods with
elevated temperatures whilst having relatively low start-up and maintenance costs. The more
 Benefits of green infrastructure  39

ambulatory residents of the care home may have access to local shops, libraries or green
spaces, but if the streets do not provide sufficient shade, they may be effectively prevented
from accessing them: the planting of street trees and the use of hedges may create conditions
that facilitate access and the strategic siting of seating underneath trees or in strip parks may
actually improve, as opposed to simply maintain, quality of life. The correct siting and species
use/mix is important to ensure that shading in summer and wind-shielding in autumn/winter
does not mean that solar heating is compromised in winter (e.g. deciduous trees shade in
summer but allow most sunlight through in winter). Whilst the scenario described is for a
care home, it is equally applicable to domestic residences, businesses and schools, and for all
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

age groups. Vegetation can also be used to shade the windows of conservatories, windows or
verandas to moderate indoor temperatures.
A number of authors have demonstrated the energy-saving value of trees (Akbari,
2002; Akbari et al., 1992, 2001; Heisler, 1986a, 1986b; Huang et al., 1987; McPherson &
Rowntree, 1993; McPherson & Simpson, 2003; Parker, 1983; Simpson, 1998). For example,
McPherson and Simpson (2003) estimated that the 177 million trees in California, at the
time of their study, saved 2.5% (6,408  GWh) on air-conditioning costs ($486 million at
wholesale prices) – to customers this was collectively worth about $970 million. There can
be a downside to shading as it can increase the need for winter heating (even after taking into
account the windspeed-reducing effects on drafts); nevertheless, the net economic benefit of
trees taking into account cooling and heating was $458 million at wholesale prices. The value
of trees in energy saving was greatest during times of peak demand – peak load reductions
were estimated at 10%. Even trees that do not cast direct shade on building surfaces can have
benefits (estimated at 25–50% of shade trees). McPherson and Simpson (2003) considered
that there were 242 million unused planting sites, and if 50 million were planted up to shade
the east and west walls of homes, savings would increase by 1.1%.
The value of trees in reducing summer air-conditioning costs varies due to aspect
(McPherson & Simpson, 2003). In a detailed study of 460 single-family homes in Sacramento,
California, Donovan and Butry (2009) demonstrated that trees on the east-facing side of a
building appeared to have no effect on electricity use – this was probably because trees on
the east cast morning shade before air temperatures become uncomfortable. South-facing
trees saved energy on air conditioning, but only if they were within 12.2 m of the house –
primarily because the sun is higher in the sky when they would be effective, and trees further
from the house would not cast long shadows and hence effective shade. West-facing trees cast
shade later in the day, and at the time of peak temperatures and with lower sun-angles; as a
result they could be further away than south-facing trees and still have an effect on energy
consumption. Trees on the north-side did not cast shade on the study houses, yet curiously
trees planted close to the north side of a house (within 6.1  m) actually increased energy
consumption; factors suggested to explain this included close proximity reducing the cooling
effect of wind, promoting the retention of heat at night, or simply because electric lights
might be required more frequently because of light-blocking. Donovan and Butry (2009)
considered that, combined, south- and west-facing trees saved 5.2% electricity and north-
facing ones increased it by 1.5%.
 40  Benefits of green infrastructure

2.8  Sustainable urban drainage systems (SUDS)

2.8.1 Introduction
Increasing urbanisation results in increasing areas of sealed surface and the requirement for
large capacity stormwater systems to rapidly route water away to prevent flooding; the higher
the proportion of sealed surface, the higher the volume of water to be dealt with (Tyrväinen
et al., 2005). Where the sewer system combines both rainwater collection and sewage, heavy
rainfall events can result in sewer overflows, creating flash flooding contaminated by faecal
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

matter. Such events are an issue globally; for example, in San Francisco, USA, (Bandy, 2003)
and in the UK as the floods in 2007 (Coulthard & Frostick, 2010; Pitt, 2008) and February
2014 demonstrated (Helm, 2014; Helm & Doward, 2014). Higher rainwater volumes are
something that it looks like we are inevitably going to have to contend with because of
climate change and the implications, as predicted by modelling, are frightening. In Europe
heavy rain is predicted to increase in the central and northern areas and decrease in the south
(Beniston et al., 2007). Whitehead et al. (2009) reviewed the impacts on aquatic systems
caused by a combination of increased temperatures and extreme events. Whilst there may
be droughts, there will also be more extreme rainfall events and heavy flushing after drought
may have negative consequences in terms of mobilisation of concentrated physical, chemical
and also microbiological entities (Bandy, 2003; Whitehead et al., 2009). Estimated effects on
streams, rivers and lakes include:

• changes in river morphology as a result of erosive power of increased flows and sediment
loads (and increased siltation of lakes);
• changes in mobility and dilution of contaminants (low flows increasing concentrations);
• decreased water quality leading to reduced habitat quality resulting from:
-- increased frequency of flash floods
-- increased stormwater runoff from urban areas
-- increased nutrient loads
-- changes in pH levels
-- increased temperatures increasing chemical reaction speed
-- toxic algal blooms
-- reduced dissolved oxygen levels;
• increased colonisation and movement of alien species (Whitehead et al., 2009).

Extreme rainfall events are more important than overall averages because they result in
substantial economic losses (including damage to property) and have public health implica-
tions (including direct mortality) (Easterling et al., 2000; Katz & Brown, 1992; Sillmann &
Roeckner, 2008). Effects are primarily due to the overloading of sewer systems. In the EU
the Water Framework Directive (2000/60/EC) (WFD) (EU, 2000) was devised to (amongst
other things):

• protect and enhance aquatic ecosystems (and terrestrial systems and wetlands that depend
on the aquatic ecosystems);
• promote sustainable water use and protect water resources;
• enhance protection and improvement of the aquatic environment;
 Benefits of green infrastructure  41

• progressively reduce pollution of groundwater;

• contribute to mitigating the effects of floods and droughts.

In the UK, responses to the WFD and flooding evident in 2007 (Pitt, 2008) included: the
Flood and Water Management Act 2010 (Defra, 2010) for England and Wales, the Water
Environment and Water Services (Scotland) Act 2003 (SG, 2003) and the Flood Risk
Management (Scotland) Act 2009 (SG, 2009).
The central purpose of sustainable urban drainage systems is: to reduce, or slow down, the
flow of (rain)water to sewers. As such, the use of SUDS is clearly relevant to the implemen-
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

tation of several of the primary WFD aims (EU, 2000).

But what is a sustainable urban drainage system? There is not a simple answer as there
are so many components that can be considered as being part of SUD; perhaps the easiest
approach is to start by breaking SUDS down into their basic functions and then thinking
about what can be used to supply those functions later (Figure 2.14).
While rain is falling two things can immediately happen depending on the SUD
component: 1) rainfall can be intercepted before it reaches the ground and is slowed down,
or 2), on reaching the ground, it is immediately infiltrated (‘soaked-up’), so that it does not
runoff to the sewer system. If the rain has been intercepted, it then follows three routes
depending on circumstances: it can continue to the ground to be infiltrated, if not soaked up
it can be transferred to a containment area or, if it has stopped raining, it can be evaporated off
the interception surface. If the rain was infiltrated, there are a number of possible subsequent
fates: it can be re-evaporated from the infiltration surface, it can be taken up by plant roots

Time  from  start  of  rain  

Evaporate  

Intercept   Infiltrate  
RAINFALL  

Transfer   Contain   Evaporate  

Infiltrate   Transpire   Transpire  

Evaporate  

Transfer  

FIGURE 2.14  Generalised functions of a sustainable urban drainage system (SUDS).

 42  Benefits of green infrastructure

and transpired, it can be transferred to a containment area, or it can remain in the infiltration
medium and be transferred to the surrounding soil. The transfer to the containment area
can be via conventional pipes or, preferably, via ‘swales’ (a nice name for a shallow vegetated
ditch system). The containment area could be a pond, soakaway or ‘rain-garden’ from where
the water can infiltrate, evaporate or be transpired off by marginal or aquatic plants (this can
also happen during the transfer process). The containment area may also act as a temporary
processing area (to remove pollutants, suspended solids, etc.) prior to disgorging into a stream
or river. This means that planning for SUDS needs to be considered at a strategic level, and at
a number of scales from the very local to regional (EPA, 2008). In Malaysia, SUDS are seen
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

as a key way to prevent flash flooding, improve a low water supply and treat water pollution;
implementation is via swales for transport and subsurface detention and various kinds of
pondage for biofiltration and water storage (Zakaria et al., 2003).
In the UK a major barrier to the introduction and use of SUDS was the wordage in
legislation: SUDS simply were not covered by the definition of a sewer. As EU directives
have to be translated into the laws of member states, the need for legislation to incorporate
the WFD was been taken as an opportunity to modify those definitions, as in Scotland (SG,
2003; SW, 2011). The Water Environment (Controlled Activities) (Scotland) Regulations
2005 (SG, 2005) required runoff from new developments to be treated using SUDS (Stewart
& Hytiris, 2008). In 2007 the various organisations concerned with environmental protection
in the USA joined together to collaborate on the use of green infrastructure (GI) as a means
of reducing stormwater runoff and published an Action Strategy (EPA, 2008) composed of
the following elements:

• Research – on the performance and effectiveness of GI.

• Outreach and communication – dissemination of information including the need to examine
and modify the way things are done – including the administrative realm.
• Tools – the development and provision of predictive and design models to aid in
implementation.
• Regulatory support – identification of the provisions that already exist for use of GI in
water management but that may not be recognised as such due to custom and practice.
• Economic viability and funding – assisting in making the financial case for use of GI
including highlighting financial support mechanisms.
• Demonstration and recognition – promoting take-up by provision of case studies and existing
projects, etc.
• Partnerships – expanding the number of organisations that signed up to the inital collabo-
rative effort.

Following the publication of the Action Strategy, the US Environmental Protection Agency
published a points-based ‘Water Quality Scorecard’ in 2009 and noted that regulatory changes
were, after all, necessary to facilitate the use of SUDS, but that these were in what might
have been considered ancillary activities such as landscaping, car park and street design (Hall
et al., 2009). The aim of the Scorecard was thus to help identify where changes needed to be
made and to provide information on policy approaches and materials, and to give examples
of successful implementation and how SUDS incorporation could be incentivised (Hall et al.,
2009). An example of the latter being a stormwater tax discount the ‘Clean River Rewards
Program’ as operated in Portland, Oregon (CoP, 2014). In completing the Scorecard, a
 Benefits of green infrastructure  43

multi-agency or department approach was urged to maximise opportunities – poor commu-

nication within and between large bodies is an obvious problem when multidimensional
problems need addressing. The Scorecard was divided into five sections which reflect some
of the major responsibilities found in different municipal departments (or agencies):

1. Protect natural resources (including trees) and open space.

2. Promote efficient, compact development patterns and infill.
3. Design complete, smart streets that reduce overall imperviousness.
4. Encourage efficient provision of parking.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

5. Adopt green infrastructure stormwater management provisions (Hall et al., 2009).

The Scorecard asks basic questions linked directly to stormwater management outcomes and
as such has a wider applicability than in the purely US context.
SUDS components have a range of different functions and most incorporate vegetation as
an important part of the solution rather than as an incidental inclusion. Pratt (2004) reviewed
the literature on the following SUDS components:

• Constructed wetlands
• Detention basins
• Detention ponds
• Filter drains
• Infiltration pits and trenches
• Pervious surfaces
• Retention/infiltration basins
• Swales and filter strips.

A nice set of examples is given in SUDSWP (2010):

• Collection – infiltration drains and ditches and permeable materials may collect water
directly from roadsides (sometimes with pre-treatment by sediment-removing gullies).
Typically these are filled with some kind of graded stone media which allows runoff to
fill air gaps between the stones before running into a perforated drainage pipe at the
base of the trench for transport to a storage area. Another approach is to use shallow,
wide vegetated ditches (swales) to transport water. Swales may also have gullies catching
sediment before runoff enters them. They may also have small ‘check dams’ incorporated
so that they can store water in the deeper part of their profile. So as well as transporting
water, swales can store and also infiltrate – reducing the amount of water reaching the
storage ponds or basins.
• Storage – typically in shallow basins and ponds where water is captured, held and slowly
released. Such structures can be dry most of the time or have a deeper, permanently
water-filled pool incorporated. This ‘attenuation storage’ captures runoff and prevents
the stormwater surges that overwhelm sewers. The water may have a (narrow) outflow
to traditional sewer systems, or run to infiltration areas. Some designs may include a
sediment trap by having a ‘forebay’ which is divided off from the rest of the pond
area by a permeable material and typically takes up about a third of the pond length.
Water entering the forebay slows and deposits its silt, with the permeable material
 44  Benefits of green infrastructure

also removing suspended solids. Such structures offer pre-treatment and attenuation
storage.
• Infiltration – infiltration basins may have outflows as for storage basins, but their primary
role is to hold water so that it can slowly infiltrate into the ground. This will be deter-
mined both by the nature of the ground (clay-based soils would not be suitable) and by
the level of the water table.
• Wetlands – these combine the concept of ponding with marshes and have a combination
of deeper and shallower areas. They have inflows and outflows at either end of the
wetland and are intended to allow sedimentation (by slowing water down) and to remove
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

pollutants and nutrients through uptake by plants.

All the above use plants in their operation, but there are opportunities with most of them
to increase the biodiversity component at the design stage. However, in concentrating on
engineered solutions (whether or not they include plants), it is possible to overlook the
obvious: that plants on their own are SUDS components, something that San Francisco has
understood. Bandy (2003) explored options to decrease rainwater-caused sewer overflows in
San Francisco and examined three options: permeable pavements, harvesting of rainwater
(water can be captured in tanks for direct reuse, e.g. in toilets, car washing, garden watering,
and not simply for attenuation storage (see EA (2008)) and trees. In Portland, Oregon, a
stormwater tax discount is available via the ‘Clean River Rewards Program’ if properties
have four or more trees over 4.57 m (15 feet) high – or can demonstrate a canopy cover of
18.58 m2 (200 square feet) (CoP, 2014). Schmidt (2009), in terms of mitigating the urban
heat island effect, considers rainwater harvesting for cooling and irrigation to be of higher
priority than simple trough and trench infiltration, or harvesting for toilet flushing etc., but
lower than the use of unpaved green areas (including street trees), green roofs and green
façades, artificial lakes and open waters.
Using a SUDS approach therefore uses (plant) biodiversity to achieve specific storm
drainage aims; they can also increase faunal biodiversity. Jackson and Boutle (2008)
reported findings from the SUDS system incorporated in a new development at Upton
in Northampton, UK. They found that swales and retention ponds were colonised by 34
aquatic macroinvertebrate families from the River Nene Valley as well as newts, frogs and
toads. Birds typical of wetland habitats had also been observed using the SUDS elements,
including the reed bunting Emberiza schoeniculus. In New Jersey, USA, McCarthy and Lathrop
(2011) found that 92% of stormwater basins with continuous water (i.e. they never run
dry) contained fish and thus successful reproduction in such water bodies depended on the
juvenile stages of anurans (frogs and toads) being resistant to fish predation. Basins that dry out
in late summer, however, make potentially good anuran habitat (Brand & Snodgrass, 2010).
In Melbourne, Australia, Kazemi et al. (2011) compared the above-ground invertebrate fauna
of roadside ‘bioretention’ swales with two types of roadside verge composed of either grass
with occasional trees or a more ornamental verge which the authors termed ‘gardenbed-
type’ – typically small patches of vegetation below a tree with the patches separated by bare
ground. The swales were designed to hold water which, after infiltration through a filter
medium, was collected by perforated pipes and transferred to a storage area. Invertebrates
were sampled by sweep netting vegetation and the study showed that species richness and
abundance were both higher in the swale vegetation. Swales were probably better because
of the improved food and shelter provided by the increased structural vegetation cover,
 Benefits of green infrastructure  45

abundance of flowering plants, lower pH, and sloping profile of the swales compared with
conventional verges.

2.8.2  Plants and rain

If we look at some of the fundamentals of higher plants, we see that they have roots, a stem
and (usually) green leaves. Features will vary between species with, for example, some plants
appearing to have no, or very short, stems whilst others have huge ones (e.g. carline thistle
Carlina vulgaris compared with giant redwood trees Sequoiadendron giganteum). These plants
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

also share an internal water transport system with roots taking up water from the soil, trans-
porting it round the plant in tubes of ‘woody’ xylem, and the leaves letting it out through
holes called stomata. Releasing water at the leaf surface can be thought of as being similar to
humans sweating, and has a similar end result: it cools leaves down. Of course if it gets too hot,
the plant can lose too much water and wilt, so stomata can be opened and closed by the plant
to increase or slow down the amount of water that flows through it. Plants that live in very
hot conditions have fewer stomata, to reduce water loss, and have various other adaptations
to their environment such as waxy cuticles, leaves reduced to spines, and swollen stems to act
as water storage organs. This process of giving-off water at the leaves is called transpiration.
So why is the structure of plants and their water transport mechanism important in green
infrastructure? If we first start with rain; it is pretty obvious that anything underneath it is
going to get wet. It is also obvious that, as long as the rain is not part of an electrical storm
(thunder and lightning), standing under a tree with leaves on is a good idea as it keeps
you dry, or rather dryer. Eventually water seeps through the canopy until you start to get
wet, when it stops raining it is a good idea to get out from under the tree because it takes
some considerable time for it to stop dripping water. This is the first reason why trees,
and vegetation in general, are useful in green infrastructure: they slow rain down. One of
the predictions of climate change is increased variability in weather (Murphy et al., 2009),
including heavier bouts of rainfall. Whether or not the flooding events that have happened
in 2007 in the UK and elsewhere (Coulthard & Frostick, 2010) are specifically climate
change events (Whitehead et al., 2009), one of the things that can help reduce flash floods
is green infrastructure because vegetation slows rainfall down, reduces overall rain volumes
that reach the ground, and in consequence results in lower peak flows (the largest volume of
water flowing to the drains at any given time). It is obvious, given the above, how structural
vegetation slows rainfall down – but how does it reduce the amount that reaches the ground?
Well hopefully the sun starts to shine after the rain, and the water still on the surface of the
leaves evaporates and so never gets to the ground. ‘Steam’ rising from trees after rainfall is
always a delight to see! The second way plants help to reduce flash floods is simply because
they take water up from the soil and ‘transpire’ it off, reducing the water held in the ground
so that new rain has somewhere to go. Having green areas allows the rain to percolate into the
soil instead of running over pavements, roads, plazas and parking lots to the drains (Bolund
& Hunhammar, 1999; Guo et al., 2000; Niemelä et al., 2010). Now, it is also pretty obvious
that at different times of the year different kinds of plants will have different values: deciduous
species are going to be better at slowing down rain when they have leaves, for instance (i.e.
in spring, summer and early autumn), whilst evergreens have leaves all year round. But even
evergreen species are not going to do much transpiring in the winter – though they will be
better at catching rain and snow.
 46  Benefits of green infrastructure

2.8.3  Implementing SUDS

It is clear that plants are central to the operation of sustainable urban drainage, and green
roofs, green walls, grassland, street trees, hedges, shrubs, allotments, etc. can all be considered
elements of SUDS. However, it does not have to be green to be part of a SUDS – permeable
pavements (where water drains through permeable concrete, asphalt, gravel gaps between
paving blocks, etc.) often do not have any vegetation, though some incorporate spaces
for grass to grow. Nevertheless, even in such situations where there is no obvious surface
component, at the subsurface level there is likely to be a microbial community that is actively
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

degrading pollutants such as oil (see section 4.3.1). Ponds could be considered ‘blue’ infra-
structure. Whatever the individual components of a particular implementation in a particular
place, the general guiding principles are the same, but the multiplicity of elements which
can be used as SUDS components means they are extremely flexible and can be adapted to
almost any situation.
The City of Portland, Oregon, USA has been very active in introducing SUDS (Liptan,
2006) and there are two particularly nice examples. The first comes from NE Siskiyou Green
Street where a residential road was narrowed at one end through having two SUDS installed.
The net result was to improve the visual amenity, control traffic and, most importantly from
the initial design viewpoint, to reduce and slow down water flow into the drains (Figure
2.15). In a test using a fire hydrant to provide water for a 25-year storm event, the two
elements reduced the peak flow by 81%, delayed the peak flow by 16 minutes and reduced
the volume going to the drains by 82% (Liptan, 2006).

FIGURE 2.15  Before(left) and after (right) installation of two SUDS elements at the end of NE
Siskiyou Green Street, Portland, Oregon, USA (the SUDS were installed on both sides of the
road). © City of Portland (see Liptan, 2006).
 Benefits of green infrastructure  47
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 2.16  The installation of stormwater planters in SW 12th Avenue Green Street, Portland,
Oregon, USA. Top left: before installation, right: old plantings removed and soil excavated;
middle left: wooden shuttering in place for concrete, right: after planting up; bottom: the
planters during rain showing collection of kerbside runoff. © City of Portland (see Liptan, 2006).
 48  Benefits of green infrastructure

Another project was the conversion of conventional surface plantings along SW 12th Avenue
Green Street to SUDS. In this situation, existing rather routine planting beds were excavated,
SUDS planters were concreted in in situ, using shuttering, and finally planted up with a
mixture of trees and plants tolerant to periodic inundation (Figure 2.16). The result is visually
more pleasing than the initial plantings and reduces runoff.

2.9  Pollution control

Pollution comes in many forms: as light, noise, air, gases and other vapours, soil and water
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

contamination, smells and particulates; vegetation has a role in reducing all these.

2.9.1 Water
The implications for removing soil and water contaminants from the human environment
are obvious. Apart from the primary reason for preventing flash floods, SUDS (see
section 2.8) are particularly useful in preventing ‘non-point source pollution’ from being
discharged into streams and rivers as part of those flood events. This type of pollution
has no single obvious source that can be controlled; for example, oils and particulates
containing toxic chemicals that accumulate on roads (Bandy, 2003). Wetland plants such
as reeds (Phragmites australis) and cattails (reed mace, or bulrushes) (Typha latifolia) often
found in SUDS are known for their ability to sequester heavy metals (Grisey et al., 2011).
Cunningham et al. (2010) looked at the conductivity (primarily from de-icing salt) and
total inorganic nitrogen loading along the 18 km-long Casperkill stream in New York state,
USA and also measured the quality of the stream using a biotic index (of aquatic macroin-
vertebrates). They showed convincingly that nitrogen loads in the water decreased when
the stream went through open green space (presumably because runoff from such areas had
lower nitrogen levels) with a consequent improvement in the biotic index. Conductivity
levels did not follow this pattern, however, and they suggested that management of this
pollutant needed to be tackled at the watershed level. Even some of the most potent
chemicals can be taken in by plants and detoxified. Whilst only a laboratory demonstration,
it is impressive that some Chinese plant species (Chinese elder Sambucus chinensis, upright
hedge-parsley Torilis japonica, hybrid willow Salix matssudana, water lily Nymphea teragona
and poplar Populus deltoides) were all shown to be capable of taking in cyanide from an
aqueous solution and presumably converting it, via a well-known metabolic pathway, into
the amino acid asparagine. Of the species tested S. chinensis was the most effective (Yu et
al., 2005).

2.9.2 Light
Vegetation’s ability to improve the ‘look’ of an area by acting as a visual screen (McPherson,
1988) extends to helping reduce unwanted light from industrial buildings, shops, sports
pitches, roads, etc. Light pollution is implicated in impacts on human health by affecting
production of the hormone melatonin by the pineal gland and hence disrupting operation
of the natural sleep cycle (circadian rhythm, body clock) with potential impacts on heart
disease, diabetes and obesity, and reducing the protection it affords in reducing cancer
(Falchi et al., 2011). For protection year-round, clearly using some form of evergreen
 Benefits of green infrastructure  49

shrub, tree or climber, perhaps coupled with an earth bund to increase the height of
the structure, will be important (rather than using deciduous species). Whilst the use of
vegetation to reduce obvious nuisance light pollution is hardly a novel concept, there
appears to be little research on the positive or negative impact of light pollution on
nocturnal ecosystem services in the scientific literature and Lyytimäki (2013) attributes
this to shifting baseline syndrome. This syndrome is typically seen as a generational issue
whereby older generations (e.g. grandparents) have seen more change over their lifetimes
and (typically) deprecate the changes they have seen, whilst the youngest generation,
having seen no change, accepts the current situation as normal (Leather, 2010; Papworth et
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

al., 2009) (see RCEP (2009) and Longcore and Rich (2004) for impacts of light pollution
on wildlife).

2.9.3 Noise
Apart from its nuisance value, noise has human health implications including cardiovascular
disease (Lercher et al., 2013) and diabetes (Sorensen et al., 2013); its role in such diseases
probably results from promoting stress and especially sleep disturbance (Sorensen et al., 2013).
Diabetes has also been linked with nitrogen dioxide levels in the environment (Raaschou-
Nielsen et al., 2013b) but Sorensen et al. (2013) were able to control for NO2 levels and
were able to demonstrate a plausible cause-and-effect model based on noise-induced stress
increasing cortisol levels which in turn depresses insulin production by pancreatic cells.
Noise also affects wildlife, including birds (Fontana et al., 2011; Herrera-Montes & Aide,
2011).
Vegetation can act as a sound barrier, reducing noise from adjacent areas, especially if the
vegetation is on top of an earth bund (Cook & Van Haverbeke, 1974, 1977; Herrington,
1974; McPherson, 1988). Cook and Van Haverbeke (1977) demonstrated that one to two
rows of dense shrubs in front of one to two rows of trees (either the shrubs or trees, or
both, should be evergreen) should reduce noise from cars and light commercial vehicles
travelling at 35 mph (56 kph) to acceptable levels in suburban areas. However, this assumed
that the building being studied was 25 m from the road centreline; vehicles closer to the
road edge would probably require some other acoustic shielding. Trees have been recom-
mended as noise barriers in towns (Leonard & Parr, 1970) and evaluated for their resistance
to air pollution for the purpose (Pathak et al., 2011). In Uttar Pradesh, India, the species
that Pathak et al. (2011) found to be most valuable as noise barriers (graded as ‘excellent’),
and also due to their tolerance of air pollution as air cleansers, from a test list of 35 species
were: the white-fruited wavy fig tree Ficus infectoria, mango Mangifera indica, and the sacred
fig Ficus religiosa. Three other species graded as ‘very good’ were: the Indian banyan Ficus
benghalensis, Indian rosewood Dalbergia sissoo and the false ashoka or Buddha tree Polyanthia
longifolia. Fang and Ling (2003) examined the characteristics of 35 evergreen tree plantations
in Taiwan for their ability to reduce traffic noise. The experimental approach was ingenious
in that urban traffic noise was recorded and then used on-site to produce noise conditions
at 48  2dB A (this is a measure weighted towards the medium-range frequencies that the
human ear is most sensitive to). The noise was then sampled in transects through the planta-
tions at 5 m intervals until the rear of the plantation under study was reached. Along with
the usual parameters that one might expect to be recorded (e.g. height of trees, length and
width of plantation), a measure of the density of the plantation ‘visibility’ was used – this
 50  Benefits of green infrastructure

is the distance into the plantation that an observer standing outside the wood can see an
object before it disappears from view. In this study the ‘object’ was one of the researchers;
although estimates were made in three of the plantations as their visibility exceeded their
width. To determine the noise reduction effect of the trees, the same sound measured at
known intervals over open ground had the sound measured at the same distance through
the plantations subtracted to give the ‘relative noise attenuation’. The strongest relationship
identified was a negative logarithmic relationship between visibility and relative noise
attenuation (i.e. the lower the visibility the greater the noise reduction). Relative attenu-
ation was also shown to have a positive logarithmic relationship with the width, length
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

and height of the plantations. Clearly the plantations and arrangements of trees within
plantations will be related to geographic region, growing conditions and management
practices, but the results of Fang and Ling (2003) show that ‘visibility’ appears to be a good
surrogate for estimation of noise reduction. The other parameters, especially width, were
also important, but given that sound does not propagate along a single transect line but
effectively in a cone (from a point source) the length and height of a shelter belt can also
be influential in noise attenuation. Ten of the 35 plantations were shown to give effective
relative attenuation values in excess of 6  dB  A; such tree belts were composed of large
shrubs with less than 5 m visibility; 14 plantations gave moderate reductions of 3–5.9 dB A
and had visibilities between 6 and 19 m. Both groups had lateral branches at the height
of the sound receiver, the best had substantial foliage and much forking of the branches;
the poorer performing plantations (<2.9 dB A reduction) were composed of species with
branches above the height of the receiver. The clear message is that an understory of shrubs
is likely to increase the performance of tree belts planted for noise reduction. Subsequently,
Fang and Ling (2005) developed a map of noise reduction by trees to help in belt design
using three dimensionless parameters. The three components were: h’ – receiver and noise
source height/tree height; d’ – distance between noise source and receiver/tree height; and
m’ – belt width/visibility.
The perception of noise, and the ability of vegetation to reduce it, does not seem to be
straightforward, but appears to be a visual and auditory interaction (Herrington, 1974) influ-
enced by the landscape setting. The perception of traffic noise by residents in their homes
(in Hong Kong) was reduced more by views of wetland and garden parks than by views of
grassy hills (Li et al., 2010).

2.9.4  Air pollution

‘Air pollution’ groups together the effects of a wide range of gases and aerosols – including
very fine particulate matter. Some have human health impacts such a nitrogen dioxide and
particulates on asthma (Anderson et al., 2013; Gehring et al., 2010; Kheirbek et al., 2013) and
increased cancer risk from particulates (Raaschou-Nielsen et al., 2013a); others have impacts
on ecosystems such as sulphur and nitrogen oxides (Bignal et al., 2004; Lovett, 1994); some
have global effects such as carbon dioxide (climate change) and chlorofluorocarbons (CFCs)
(ozone hole formation) (Solomon, 2004). Substantial mortality from air pollution in 1952,
the ‘London Smog’, lead to the UK Clean Air Act (1956) (Harrison, 1992) and air quality
has improved since with successively revised and new legislation. In the UK, between 1990
and 2001 improved air quality was estimated to have saved 4,200 lives and 3,500 hospital
admissions/year (Defra, 2007a). In the USA, Fann and Risley (2013) estimate that reductions
 Benefits of green infrastructure  51

in PM2.5 and ozone over a seven-year period (2000–2007) have reduced premature deaths
by between 22,000 and 60,000 for the former and between 880 and 4,100 for the latter.
Whilst the estimates have rather broad ranges, as might be expected working with such a
large country and with constantly varying air quality and weather data, the message is clear:
improving air quality saves lives.
The 1995 UK Environment Act introduced a requirement for the development of an Air
Quality Strategy: the first was produced in 1997, revised in 2000, and the current one (at
the time of writing) introduced in 2007 (Defra, 2007a). The strategy contains the current air
quality objectives and the EU directive limit and target values for pollutants. The executive
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

summary makes a bald and sobering statement: ‘Air pollution is currently estimated to reduce
the life expectancy of every person in the UK by an average of 7–8 months.’ Of course,
seven to eight months is an average – with some people (those exposed to the highest levels
of pollution, or the most susceptible (O’Neill et al., 2012)) having considerably shorter life
expectancy than someone living in an unpolluted area (all other things being equal). In
Estonia, Orru et al. (2011) estimated that reduced life expectancy could be in excess of 12
months in polluted city centres. The new UK Air Quality Strategy aims to lower reduced life
expectancy to five months by 2020 (Defra, 2007a), though many air pollutants are proving
difficult to control (particulates, nitrogen dioxide, ozone (Defra, 2007b)). There are places in
this world with far worse air pollution than the UK.
Vegetation can help reduce a range of air pollutants either by direct deposition of
particulates to surfaces, with lipid-soluble volatile organic chemicals being taken in
directly through the cuticle (e.g. PCBs – polychlorinated biphenyls) or, as with gases,
by the stomata (Barber et al., 2002) as are some ultrafine (<0.1  μm) particles (Fowler,
2002). For some compounds, such as PCBs, the compound may be, at least in part,
subsequently re-released into the air whilst other fractions are held in the leaf epidermis
or mesophyll (Barber et al., 2003; Wild et al., 2006). In terms of their pollutant removal
capacity, different types of vegetation, species, ecotypes, and varieties can all have different
attributes and values and it is becoming increasingly clear that deposition to vegetation
is not a simple process (Lovett, 1994). Inevitably, vegetation size is a factor: large trees
remove more air pollutants than small trees; Nowak (1994a) estimated that trees with a
diameter at breast height (dbh) of 76 cm removed 70 times the amount of a tree of less
than 8 cm dbh (1.4 kg vs 0.02 kg). Trees are also better than grassland at removing air
pollutants. Jonas et al. (1985) compared the capture efficiency of a range of tree species
with grass in removing particulates from the air (size range 0.4–17 μm) and found trees
to be substantially better. Trees that were well spaced from other trees captured more
particulates, on a tree-for-tree basis, than those in close groups because of improved air
circulation (Figure 2.17), but groups of trees were very effective overall (Stülpnagel et
al., 1990). The soil under trees has also been shown to have higher levels of lead, for
example, than grassland (Fowler et al., 2004). In 1991 Nowak (1994a) estimated urban
trees removed 591 tonnes (t) of air pollution in Chicago (which had 11% tree cover) =
15 t CO carbon monoxide, 84 t SO2 sulphur dioxide, 89 t NO2 nitrogen dioxide, 191 t
O3 ozone and 212  t of particulates PM10 and below. A recent project to evaluate the
ecosystem services provided by Torbay’s urban forest in the UK showed that 11.8% of
the ground was covered in 818,000 trees and that they removed 50 t of air pollutants per
year (0.0005 t CO, 7.9 t NO2, 22.0 t O3, 18 t PM10, 1.3 t SO2) equivalent in value to
£1.5 million in externality costs and social damage (Treeconomics, 2011). The trees have
 52 Benefits of green infrastructure

45
40
35
Capture efficiency

30
25
20
15
10
5
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

0
ba

is

a
bu
de

lu

ie

tri
ul

ic

pe

br
ep
nu
al

tu

at

ab
nd

es

ro

ru
oi

ro

ol
ta

be

lv
s

an

lv
pe

eu

pt

us

us
ie

a
as

sy

si
ce
Ab

le
at

rc

rc
oc

la

rix
nu

us

s
Pi
pl

rix

ue

ue
nu
tu
pp

La
pi

g
er

Be

La

Q
Fa

Pi
ar
hi
Ac

C
us
ul
sc
Ae

Species

FIGURE 2.17 Relative particulate capture efficiency of trees compared with grass (Deposition
on trees/Deposition on grass). Dark green columns: data for single tree well spaced from other
trees; light green columns: data for trees with reduced air circulation (e.g. planted in groups).
Data from Jonas et al. (1985).

a replacement value of £280 million and store about 98,000  t of carbon a year. Luley
and Bond (2002) produced a report aimed at integrating the management of urban trees
into the air quality management plan of New York. Across the USA removal of ozone,
PM10, NO2, SO2 and CO by trees and shrubs was estimated at 711,000 tonnes per year
worth $3.8 billion by Nowak et al. (2006). In Beijing, Yang et al. (2005) estimated that
there were 2.4 million trees in the metropolitan area which removed about 772 tonnes
of PM10 per year and stored 0.2 million tonnes of CO2. In addition to the ‘normal’
gaseous pollutants such as nitrogen dioxide, a wide range of volatile chemicals have also
been shown to be removed from the air by plants (Fries & Morrow, 1981; Ockenden
et al., 1998; Smith & Jones, 2000; Topp et al., 1986) including PCBs (polychlorinated
biphenyls).

2.9.5 Particulates
Air pollution in urban centres is a worldwide problem (e.g. in China (Jim & Chen, 2008),
Europe (Raaschou-Nielsen et al., 2013a), India (Deshmukh et al., 2013; Rai & Kulshreshtha,
2006), Thailand (Phoothiwut & Junyapoon, 2013), Greece (Vlachokostas et al., 2012)) and
particulates are an important component of it. Plants filter out dust (Rentao et al., 2008) and
small particles from the air (Beckett et al., 1998; Dochinger, 1980; Hosker & Lindberg, 1982;
Smith & Jones, 2000; Stülpnagel et al., 1990; Varshney & Mitra, 1993) and deposition rates
are higher than for buildings and sealed ground surfaces (Pugh et al., 2012). Vegetation is of
particular value to urban dwellers; low air circulation in cities results in high concentrations of
small particles given off by motor vehicles, especially those fuelled by diesel, in their exhaust
(Chaloulakou et al., 2003; Paoletti et al., 1989). In addition to carbon particles from diesel
exhaust, road dust also contains other particles that originate from engine wear and tear:
 Benefits of green infrastructure  53

typically composed of iron, chromium and nickel. Espinosa et al. (2001) found cadmium,
calcium, cobalt, copper, iron, magnesium, manganese, lead, nickel, titanium and vanadium
in urban aerosols in Seville, Spain. In countries, such as Finland, where winter snow cover
is inevitable and snow-tyres with studs are used, particles with a high proportion of tungsten
are also likely to be found (Bućko et al., 2011). Airborne particles are removed by three
main processes: sedimentation (via gravity), deposition (via precipitation), and, particularly
important with respect to the role of vegetation, impaction via wind – the smaller the particle
the more effective is the process of impaction (Smith, 1977).
The particle ranges most often quoted in the literature are those known as PM10 and
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

PM2.5. PM stands for ‘Particulate Matter’ and the number (e.g. 10) is the diameter of the
particle in ‘microns’ or ‘micrometres’ (mm); a micrometre is 1 x 10-6 (or 0.000001) of a
metre. The distinction between the two size ranges is because particles of 10 mm or below
can get into the lower respiratory system (and are called ‘thoracic’ particulates) whilst
particles of 2.5 mm and below can penetrate further, into the narrower spaces in the lungs,
and are therefore called ‘respirable’ particles (Brunkeef & Holgate, 2002). These particles
are easily breathed in and have serious health effects such as increased likelihood of death
from respiratory and cardiovascular disease (Brunkeef & Holgate, 2002; Pope III et al.,
2011; Sunyer et al., 2000), atherosclerosis (Araujo, 2011), lung cancer (Raaschou-Nielsen
et al., 2013a), and increased incidence of asthma (Anderson et al., 2013). Ultrafine particles
which are  0.1 μm in diameter can cross into the body through the lungs and other parts
of the respiratory tract and can thus be transported throughout the body; they have been
found in the liver, spleen and kidneys and can move rapidly (<1 h) from the nose, via the
olfactory nerves, to the brain (Solomon et al., 2012). The level of mortality caused by
particulates is not trivial, with one study estimating an annual death toll of 2,100 deaths
per year in the Netherlands alone (Brunkeef & Holgate, 2002). Particulates are not simply
inert carbon particles but can carry some quite nasty organic chemicals including cancer-
causing polyaromatic hydrocarbons (PAHs), the highest concentrations of which are carried
on the respirable PM0.35–1.2 size class (Burkart et al., 2013; Phoothiwut & Junyapoon, 2013).
Sram et al. (2011) in a study of 950 policemen and bus drivers estimated that non-smokers
exposed to more than 1  ng/m-3 of benzo [a]pyrene bound to particulates could cause
damage to DNA. Stewart et al. (undated-a) indicate that for every 10  μg.m-3 increase in
PM10 concentration in the air there is an associated increase in human mortality of 1%.
This was based on a 24-hour average of PM10 concentrations and the authors claimed that
doubling the number of trees in the West Midlands area of the UK could save 140 lives
per year.
Taking the UK as an example, the European Union Directive 2008/50/EC on ambient
air quality and cleaner air for Europe (EU, 2008) laid down particulate thresholds for
human health. This has resulted in a set of UK air quality objectives for PM10 such that
from 2010:

• The daily threshold concentration must not exceed 50  μg.m-3 on more than 35
occasions/year.
• The annual average must not exceed 40 μg.m-3.

Scotland, with its devolved administration, has tighter thresholds because its background level
of pollution is lower than in the rest of the UK:
 54  Benefits of green infrastructure

• Daily: 50 μg.m-3 on seven occasions.

• Annually: 18 μg.m-3.

PM2.5 annual mean values were set at 25  μg.m-3 for the whole of the UK (excluding
Scotland); for 2020 more stringent limits (12  μg.m-3) have been set for Scotland. For
urban areas, specifically, PM2.5 reduction targets between 2010 and 2020 were set at 15%
reduction compared to the urban background (Defra, 2007a). Local authorities in the
UK have also been given duties under the Environment Act (1995) and the Environment
(Northern Ireland) Order 2002 to establish local air quality management systems, including
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

the introduction of ‘air quality management areas’ (AQMAs) for PM10s and six other air
pollutants if areas are unlikely to achieve compliance with national standards (Defra, 2007a).
In 2012, 98 AQMAs for particulates had been declared by the UK’s 405 local authorities
(LA) (Defra, 2012) – although it is worth noting that an individual LA may have more
than one action plan in force within its area. Defra (2012) indicates good compliance
with particulate pollution standards in 2012 (compared with when monitoring started in
1992) with only London exceeding the daily PM10 limit, and all areas complying with the
annual limits for PM10 and PM2.5. Compliance with standards does not mean the absence
of pollution, as comparison of the standards for Scotland and the rest of the UK demon-
strate – and the standards also allow the subtraction of any natural particulates in the area
before testing against threshold limits – so they are not about absolute pollution levels, but
those only due to anthropogenic sources. It is also worth remembering that a 24-hour mean
could equate to quite high levels of pollution during the period of maximum commuting
during the day (and especially in the morning), when an urban area’s population is most
likely to be exposed, compared with perhaps half that peak level at night when far fewer
people are about (Muir, 1998). Whilst Freer-Smith et al. (2005) did not present any data,
they noted two peaks in particulate concentration (Monday–Friday) during the day (8–9am
and 3–6pm) in their study of particulate deposition on trees. Pedestrians are exposed to
higher levels of particulates than those driving a car and at much higher levels than static
monitoring devices (which are the basis for air quality monitoring) indicate (70% higher
for pedestrians and 25% for people in cars) (Briggs et al., 2008; Gulliver & Briggs, 2007).
However, Kaur et al. (2007), reviewing personal exposure studies, considered that pedes-
trians and cyclists experienced lower particulate and carbon monoxide concentrations than
those inside vehicles. It is also interesting that PM10 and PM2.5 levels were forecast to rise in
the UK after 2015 (Defra, 2007b). Baxter et al. (2013) in a study of PM2.5 and ‘respiratory
and other non-accidental mortality’ in 27 US communities found an unexpected negative
relationship with distance travelled in a vehicle; in other words, the longer your commute,
the less likely you are to have an air-pollution-related health impact. This rather counterin-
tuitive finding was suggested to arise as a result of those living in polluted urban settlements
having shorter travel distances and thus being exposed to (in this case) PM2.5 for longer than
those living in suburban or out-of-town areas with cleaner air. An alternative interpretation
offered was that the data was identifying the difference between the younger, working-age
population with long commutes compared with the older/retired segment of the population
with short or no commutes – the group which would be more likely to suffer air-pollution-
mediated illnesses. Kaur et al. (2007) recommended that pedestrians use ‘back street’ routes
that take them away from major traffic areas, and hence pollutant levels, as a way of reducing
exposure.
 Benefits of green infrastructure  55

Zhu et al. (2002b) examined the deposition of ultrafine particulates (in the 6–220
nanometre range; 1 nm = 1 x 10-3 μm) perpendicular to Freeway 405 along Constitution
Avenue in Los Angeles’s National Cemetery and demonstrated that loads declined exponen-
tially with distance from the road; loads were heaviest within the first 30 m and were half
that at about 100  m, background levels were achieved at 300  m from the road primarily
because of coagulation (smaller particles clumping together) and atmospheric dilution. Zhu
et al. (2002b) concluded that people living, working, and commuting 0–100 m from traffic
will have enhanced exposure to PM10 and below. See also Zhu et al. (2002a) for results from
Interstate 710, which had more diesel vehicles than Freeway 405.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Options for reduction include removing particulates at source (including more efficient
engines, fuels and catalytic particulate filters), removing them from the air, or moving
particulate emission sources from areas where high concentrations build up. In the latter
case, this is primarily by trying to influence road travel by having ‘green travel plans’,
routing traffic away from hotspots, establishing low emissions zones, congestion charging
and workplace charging for parking (Defra, 2007a). Given that much particulate pollution
comes from motor vehicle exhaust (Ayala et al., 2012), it is a tricky task to tackle without
efficient diesel particulate filters that work well in both urban and long-haul situations
and this remains a problem (Chilton, 2011; Fino, 2007; Fino et al., 2003; van Setten et
al., 2001). Encouraging use of public transport is also often mooted as a way of reducing
emissions by reducing car use; however, those waiting at bus stops in polluted areas may
be subject to increased risk due to their proximity to traffic (Godoi et al., 2013) and Kaur
et al. (2005) found pedestrians in London were exposed to greater levels of particulates if
they walked next to the kerb (road edge) compared with next to buildings (furthest from
the kerb).
The ability of vegetation to remove particulates has been known for some time (e.g.
Zulfacar, 1975) and especially in a rural context by the use of windbreaks to reduce
pesticide drift (e.g. by 70–90% in the 3  m zone downwind; see references in Hewitt,
2001). Dochinger (1980) studied the entrapment of particulates by trees in urban areas of
Steubenville, Ohio, USA from March to October using simple dustfall traps. He demon-
strated that the canopies of deciduous trees (black locusts Robinia pseudoacacia, sugar maples
Acer saccharum and red oaks Quercus rubra) could reduce dustfall by 27% and conifers (red
pines Pinus resinosa and blue spruce Picea pungens) by 38% compared with mown grassy
areas. The deciduous trees also had an understory of elms (Ulmus sp.), sassafras (Sassafras
albidum) and dogwood (Cornus florida). In the absence of leaves (March, April and October)
the deciduous species did not differ from the open grassy areas in the level of dust,
although some weather conditions may improve capture rates (e.g. with snow covering
bare branches) (Hosker & Lindberg, 1982). Over the five-month period when deciduous
species had leaves (May–September) they were very similar to the coniferous species
with particulate extraction of 37 and 38% respectively; there were monthly variations in
collection efficiency. Dochinger (1980) also used a second method to sample particulates, in
this case not dustfall, but the number of suspended particles in the air. The method involved
sucking air at a constant rate through a sampling device with the particles being collected
on filter paper. Using this method, over the eight months, deciduous species removed 9%
of the suspended particles and conifers 13%; bare deciduous trees did not remove suspended
particulates. Hagler et al. (2012), working in North Carolina, studied the effect of trees
as particulate barriers alongside roads and found particulate loads varied, sometimes being
 56  Benefits of green infrastructure

TABLE 2.5  Methods of dry deposition of aerosols to vegetation

Deposition method
Brownian diffusion Random motion of small particles; a 1 μm particle can knock into
others at the rate of about 1016 a second. Deposition is limited by the
ability of the particle to penetrate the boundary (air) layer surrounding the
surface
Interception Affects particles with low inertia which follow air currents and encounter
a surface and are retained – typically by surface roughness or hairs
Affects particles of high inertia which cannot follow the air currents round
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Impaction
a surface and hit it
Sedimentation Settling out of particles from the air due to gravity
Rebound Mainly affects particles of 5 μm and above: the kinetic energy of the
particle causes it to rebound from the surface it hits – but affected by the
nature of the surface adhesion.

Sources: UoC (UoC, undated); Petroff et al. (2008)

more, sometimes less than in an area without tree cover. They suspected that the variability
related to the nature of their study system, which was highly porous with rather thin, gappy
tree stands with modest leaf area.
Smith and Jones (2000) noted that particulates deposited on plants are generally smaller
than 100  mm and described the various ways that they were trapped by vegetation: wet
or dry deposition; if dry, by interception, impaction or sedimentation (although there are
other processes that operate i.e. Brownian diffusion and rebound (Lovett, 1994; Petroff
et al., 2008)) (Table 2.5). Wet deposition is more important for particles present at great
heights, whilst dry deposition is more important for particles in the air nearer the ground
(the atmospheric surface layer) (Sehmel, 1980). Another ‘wet’ deposition method is
via occult deposition – from ‘wind-driven cloud water’ (Unsworth & Wilshaw, 1989).
Hosker and Lindberg (1982) indicate that wet and dry deposition remove roughly equal
levels of pollutants, with wet deposition removing them by two methods: 1) within-
cloud scavenging (rainout) and 2) below-cloud scavenging (washout). Smith and Jones
(2000) highlighted uncertainties in the literature relating to the capture of particulates by
vegetation; for example, grass appeared to have higher particulate loads than more complex
vegetation such as herbs, shrubs and trees, but they pointed out that this may be because
grasses have longer to accumulate particulates through retaining dead leaves from the
previous growing season. The implications of this observation are substantial, as vegetation
management and seasonality may strongly impact the efficiency of particulate trapping.
Hosker and Lindberg (1982) list different vegetated and inert surfaces and the atmospheric
features that influence deposition.
Despite the uncertainty in the literature, the findings of some studies reviewed by Smith
and Jones (2000) may be taken as generalities: plants which have smooth, water-repellent
leaves being less likely to collect or retain particulate matter compared with those that are
sticky, hairy or have rugose surfaces; smaller-sized particles (1-5 μm) are likely to be trapped
more efficiently by ‘hairy’ plants. Wang et al. (2011) examined 14 species commonly used
in urban greening in Xi’an, China for their particulate-capturing ability. The species were
 Benefits of green infrastructure  57

predominantly trees and shrubs, but also included white clover (Trifolium repens) – the
results showed substantial differences with T. repens capturing only 0.23 g of particulates m-2
compared with 4.51g.m-2 by the evergreen shrub Japanese mock orange Pittosporium tobira.
Scanning electron microscopy revealed that cell surfaces of the most poorly performing
species were covered in a dense layer of wax tubules, and that the species which performed
best in particulate capture had rough surfaces with some having hairs and pits. The leaves of
evergreens are exposed to air pollutants for longer than deciduous species, and as a result the
surfaces of such species (e.g. Chinese red pine Pinus tabulaeformis) appeared much modified,
and were considered to have enhanced capture abilities partially due to secretions from leaves
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

and microbial colonisation. Sehmel (1980) gives a comprehensive run-down of the factors
that influence the deposition of particulates, including vegetative characteristics. Particulates
that are collected may remain on plant parts, be re-mobilised (e.g. by wind), or be washed
off onto other plants below them, onto the soil surface, or transferred via vegetated swales to
ponds. Whether in situ, or following immobilisation or transfer, some particulate contami-
nants such as nitrogen and phosphorus may be utilised directly by plants (Abe & Ozaki, 1999)
or bound to organic material and immobilised or move through the soil horizon (Kalbitz et
al., 1997; Smith & Jones, 2000). As leaves age, and ultimately die, their collection efficiencies
will change – in some cases more particulates will be collected, in others less – and leaf fall
will ultimately bring collected particulates to the soil surface (though they may be remobi-
lised in the process – via strong winds). More dense vegetation is likely to remove more
particulates, whilst grazing may reduce collection efficiency (by removing plant material).
This latter is more likely to have a substantial effect in rural rather than urban areas, although
grass cutting and hedge trimming is likely to have an equivalent effect. It is also obvious
that weather conditions will have a substantial impact on the amount of particulates that are
removed by vegetation (Sehmel, 1980).
Whilst plants are capable of capturing particulates and removing them from the air, they
are not immune to damage from them, with particulates clogging stomata, affecting photo-
synthetic efficiency, promoting disease attack, etc. (Prajapati & Tripathi, 2008; Prusty et al.,
2005; Rai & Kulshreshtha, 2006). Particulates can contain materials that are toxic to some
plants, and even the physical action of particles impacting on leaves can result in abrasive
damage (see Beckett et al., 1998).
McDonald et al. (2007) modelled the impact of tree planting on PM10 concentrations
in the West Midlands and Glasgow, UK, and considered that tree planting, on its own,
could make measurable improvements to PM10 levels. Increasing tree density from 3.7%
to a realistic target of 16.5% in the West Midlands was estimated to result in a reduction
in PM10 by 10% (equivalent to 110 tonnes/year), for Glasgow an increase in tree densities
from 3.6 to 8.0% would result in a 2% reduction in PM10 (= 4 tonnes/year). These modest
‘global’ levels may mask considerably higher particulate removal rates at the local level.
Pugh et al. (2012) estimated reductions in urban canyons (streets with buildings either side)
could be as high as 60%! Deshmukh et al. (2013) studied particulates in the industrial city
of Raipur, India, and concentrations exceeded the annual average air quality standards of
60  μg.m-3 for PM10 and 40  μg.m-3 for PM2.5 by about 4 times and 3 times respectively;
the 24-hour standards (100  μg.m-3 for PM10 and 60  μg.m-3 for PM2.5) were exceeded
all the time for PM10 and almost all the time (94%) for PM2.5. The air quality was rated
as ‘moderate to hazardous’ (Deshmukh et al., 2013). As a comparison, Chaturvedi et al.
(2013) attribute the low air pollution in Nagpur (24 h PM10 concentration = 53 μg.m-3),
 58  Benefits of green infrastructure

to it being one of the greenest cities in the country with 18% tree cover (including parks)
and 17% in cultivation.
Convincing local authorities of the merits of vegetation planting for particulate control
is difficult and especially so in terms of quantifying expected health benefits especially given
the differing capture efficiency of different types of vegetation (grasses, trees, shrubs) and
their different morphological characteristics (e.g. leaf size, shape, surface characteristics)
(Tiwary et al., 2009). Modelling approaches have therefore attempted to bridge the gap. An
interesting example was based on a 10 x 10 km section of the Boroughs of Newham and
Greenwich in East London; different planting scenarios were modelled for the 5.5% of the
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

area planned as green space (proposed and existing) using two tree species – sycamore (Acer
pseudoplatanus), which has relatively poor PM10 capture efficiency (similar to grassland), and
Douglas fir (Pseudotsuga menziesii), which is much better – and grassland. Although several
modelling scenarios were used from all grassland to all conifer, the ‘realistic’ scenario of green
space planting with 75% grass, 20% sycamore and 5% fir was estimated to remove 0.17 t of
PM10 per ha.y-1. Modelling of the health benefits suggested that modifying this 10 x 10 km
section of London would reduce mortality by two deaths and two hospital admissions per
year (Tiwary et al., 2009). Health benefits appear modest, but over ten years in a 10 x 10 km
area the ‘savings’ are 20 premature deaths. The ‘savings’ are also likely to be an underestimate,
as the study did not include street trees or other vegetation in the analysis. The tree compo-
sition could also be tweaked for better capture efficiency – it is unlikely, for example, that
sycamore would be the only deciduous tree planted.

2.9.6  Gases and aerosols

Burnett et al. (1998) examined the impact of gaseous air pollutants on mortality in 11
Canadian cities from 1980 to 1991 and estimated that nitrogen dioxide pollution resulted
in a 4.1% increased risk of death, ozone 1.8%, sulphur dioxide 1.4% and carbon monoxide
0.9%. As this class of compounds is wide, this section will be limited to briefly touching
on the global effects of carbon dioxide, a global pollution issue due to its climate change
impacts, but will mainly concentrate on the major gases implicated in human health impacts
and the nuisance value of odours. Volatile organic chemicals as indoor pollutants are covered
in Chapter 3.

Carbon dioxide
Whilst carbon dioxide is a major polluter (resulting from anthropogenic activity such as
clearing land, burning biomass and fossil fuels) and climate change gas, it is also a natural
product of respiration by organic life. In terms of climate change, the additional CO2 in the
atmosphere reduces long-wave radiation (heat) emissions from our planet, resulting in global
warming and resultant climate change as weather systems respond to an altered planetary heat
balance (IPCC, 2014). Vegetation (and, come to that, animal life) can act as a carbon sink,
incorporating it into biomass, some of which will, in turn, be incorporated into soils and
some retained in longer-term structures such as trees and ultimately timber (though that will
probably end up being burned or decay releasing the stored CO2). As a result carbon seques-
tration by vegetation has become a hot topic with such actions as tree planting and biomass
 Benefits of green infrastructure  59

burning frequently being promoted (on the assumption that the latter is carbon neutral or at
least better than fossil fuel burning).
Increasing vegetation density in urban areas will obviously make some contribution to
reducing atmospheric carbon through sequestration, although it is likely to be slight in
global terms and reducing carbon emissions in the first place is likely to be more effective.
Nevertheless, collectively, trees can act as substantial carbon offsets. Nowak and Crane (2002)
using data from ten cities and national data on tree cover estimated that urban trees in the
USA stored about 700 million tonnes of carbon with an annual increment of 22.8 million
tonnes of carbon a year. More recently, Nowak et al. (2013) expanded this data set to cover
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

28 cities and reported that each m2 of tree cover stored 7.69 kg of carbon (C); the annual
sequestration rate was estimated at 0.28  kgC.m-2. In total they estimated that urban trees
across the USA stored 643 million tonnes of carbon (data for 2005) with annual sequestration
of 25.6 million tonnes, which makes their earlier estimates pretty good. Peper et al. (2008)
calculated that in 2005 Indianapolis’s street trees collectively absorbed 5,489 kg of CO2 per
year – the equivalent of that produced by 2,338 vehicles.
The age structure of trees in urban areas also affects the carbon storage and uptake figures;
Nowak (1994b) indicated that annually a small tree (<8 cm diameter at breast height (dbh))
only sequesters the amount of carbon emitted by driving one car 16  km whereas a large
tree (>76  dbh) takes-up the carbon emitted by driving a car 1,460  km. Growth rates also
impact on carbon storage and Stewart et al. (undated-a) ranked 30 species growing in the UK
by their growth rates – the fastest growing being Lawson cypress Chamaecyparis lawsoniana,
Leyland cypress Cupressocyparis leylandii, larch Larix decidua, poplar (Populus sp.), silver birch
Betula pendula, and willows (Salix spp.). However, slow growing, long-lived trees will tend
to store more carbon; Stewart et al. (undated-a) cite English oak Quercus robur as storing the
most carbon of any tree in the West Midlands for this reason. In Chicago, Nowak (1994b)
estimated that the urban trees stored about 0.9 million tonnes of carbon and sequester some
40,100 tonnes/year.
The value of planting additional urban vegetation in relation to climate change is likely to
be rather more immediately felt in the area of mitigating its effects, such as through reducing
flooding risk and mortality from heatwaves, etc. However, urban vegetation also has a role
to play in reducing primary carbon emissions by reducing energy consumption in buildings;
for example, by reducing the need for air conditioning by shading buildings from direct sun
(deciduous trees can shade windows in the summer but let light in in the winter) and also by
reducing winter heating costs (e.g. evergreen wind shields). Also, paths and cycleways that are
pleasant visually are more likely to be used (reducing fossil fuel consumption in the form of
petrol and diesel); the aesthetics of paths and cycleways should be an important component
during their design and in subsequent maintenance programmes. Getting people out of cars
(and even buses) is important not only to promote exercise, but also because air quality inside
vehicles may not always be optimal (Kadiyala & Kumar, 2012, 2013). However, as many
(but not all) of the contaminants in vehicles arise from outside, getting people onto paths/
cycleways as a positive benefit in air-quality terms does assume that the outdoor air quality is
better (see above, under ‘particulates’).
 60  Benefits of green infrastructure

Ozone
Ironically, whilst ozone in the stratosphere (the upper layer of our atmosphere) protects our
planet from dangerous UV radiation and is not a pollutant, low level ozone produced in
the ‘boundary layer’ where we live (below the troposphere) (Clarke, 1992) is dangerous to
humans and vegetation (Davidson & Barnes, 1998). Ozone is known as a secondary pollutant
as it is a result of reactions between other ‘primary pollutants’ (Fowler, 2002). Volatile
organic compounds in the air combine with nitrogen oxides (collectively known as NOx),
the most important of which are nitrogen monoxide NO and nitrogen dioxide NO2, in the
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

presence of sunlight (a photochemical reaction) to form ozone. Impacts on humans are as

varied as cardiovascular and respiratory diseases, including increased mortality rates, rashes,
dermatitis/eczema, urticaria, conjunctivitis and cellulitis (a bacterial inflammation of the skin
exacerbated by ozone damage) (Anderson et al., 2012; Berman et al., 2012; Hunova et al.,
2013; Szyszkowicz et al., 2012a; Szyszkowicz et al., 2010; Szyszkowicz et al., 2012b). Chen
and Schwartz (2009) in a study of 1,764 US adults showed that increased ozone levels had
measureable, negative impacts on cognitive performance. For every 10 ppb increase in annual
ozone exposure, the decrease in coding ability (measured by a symbol-digit substitution test)
was equivalent to ageing 3.5 years, and attention and short-term memory by 5.3 years. The
EU have a number of threshold reporting values for ozone: the ‘Long-term objective for the
protection of human health’ (LTO) is a maximum daily eight-hour mean of 120 μg.m-3; the
‘information threshold’ is a 1 h mean concentration of 180 μg.m-3; the ‘alert threshold’ is a
1 h mean concentration of 240 μg.m-3. The LTO is used to assess whether a member state has
met the ‘target value for the protection of human health’ (TV). The TV (as in EU Directive
2008/50/EC) is achieved if the LTO is not exceeded on more than 25 days/year (based on
a three-year mean). In 2012 the TV was exceeded in 17 member states (Černikovský et al.,
2013).
Ozone (O3), along with components of vehicle exhaust, is removed by vegetation (Manes
et al., 2012). However, ozone may also damage the plant (Chauhan, 2010a; Davidson &
Barnes, 1998; Joshi & Bora, 2011). Paoletti et al. (2009) suggest that such damage can be
turned to good use by using particularly susceptible species, such as the rose mallow Hibiscus
syriacus, which is used as an ornamental hedge in Italy, as ozone bioindicators. Ozone
produced characteristic damage to the rose mallow foliage, with extent (of damage) related
to the O3 concentration. Chauhan (2010b) also suggested that trees could be used as bioin-
dicators of vehicle pollution and Ockenden et al. (1998) suggested pines (Pinus sylvestris) and
lichen (Hypogymnia physodes) could be used similarly to monitor PCBs.

Nitrogen oxides
Nitrogen dioxide (NO2) is a common urban air pollutant and can have impacts on human
health including asthma (Anderson et al., 2013) and also by acting as a neurotoxin affecting
cognitive development in children (Freire et al., 2010). This latter study, carried out in the
Granada region of southern Spain, used nitrogen dioxide as a proxy for air pollution from
traffic in general. Children exposed to the highest NO2 levels showed decreases in general
cognitive scores, quantitative and working memory, and gross motor areas as measured by
the McCarthy Scales of Children’s Abilities. Whilst only the gross motor area scores were
shown to be significantly lower in high-exposure children, the results were suggestive of a
 Benefits of green infrastructure  61

general effect of traffic pollution on children’s neurodevelopment. Air pollution is also well
known to have impacts on biodiversity with lichens being particularly sensitive and used as
biomonitors (Jovan, 2008). Gombert et al. (2003) investigated the nitrogen concentration of
two epiphytic lichens: Physcia adscendens a known nitrophyte (species reacting positively to
nitrogen in the environment) and H. physodes an acidophyte (responding positively to a more
acidic environment) with the traffic density in the cities of Meylan and Saint Martin d’Hères
in Grenoble, France. They found a relationship between traffic density, proximity to roads
and nitrogen levels in P. adscendens, but not H. physodes. Honour et al. (2009) demonstrated,
in an experimental ‘solardome’ experiment, that nitrogen oxides at realistic concentrations
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

for the urban street environment (NOx concentrations of 77–98 nl.l-1 and NO:NO2 ratios of
1.4–2.2) reduced growth, produced early senescence, delayed flowering and induced changes
in the leaf wax characteristics of 12 species of ground flora (broadleaved plantain Plantago major,
black knapweed Centaurea nigra, common sorrel Rumex acetosa, autumn hawkbit Leontodon
autumnalis, bird’s-foot trefoil Lotus corniculatus, white clover, sowthistle Sonchus oleraceus, fat
hen Chenopodium album, Oxford ragwort Senecio squalidus and the grasses perennial rye-grass
Lolium perenne, annual meadow grass Poa annua and timothy Phleum pratense). Responses were,
however, species-specific. Further work by Bell et al. (2011) confirmed that diesel exhaust,
and specifically the nitrogen oxides emitted, affected urban plant performance – stimulating
some aspects and reducing others – and called for air-quality laws to reflect impacts on urban
vegetation.
Trees are known to be nitrogen dioxide sinks, with Scott et al. (1998) estimating that
Sacramento’s urban trees removed 1–2% of the area’s daily emissions, an annual equivalent of
13.9 kg/ha. Pugh et al. (2012) suggest that greening urban canyons could reduce NO2 levels
by up to 40%.

Sulphur dioxide
As with other air pollutants, sulphur dioxide exposure is implicated in an increased mortality
risk as Katsouyanni et al. (1997) found in a study of 12 European cities. An increase in
50  μg.m-3 of sulphur dioxide was associated with a 3% increase in daily mortality in five
western European cities (with 4% increase in cardiovascular and 5% increase in respiratory
mortality). For settlements in central eastern Europe, the increase in daily mortality was 0.8%.
Impacts were greatest in the summer months. As with nitrogen oxides, SO2 can be taken up
by plants (Bell et al., 1982; Peper et al., 2008; Tan & Sia, 2005).

Odour
Some animal husbandry and waste treatment processes generate unpleasant odours, and it
is largely from this body of literature which the value of outdoor vegetation to outdoor
odour removal can be seen. Whilst the examples are rural, the principles are likely to be
similar in urban situations where industry and local authority facilities generate odours.
Agriculturally generated odours can be particularly troublesome, and research on poultry
units and on pig production units has demonstrated that vegetated barriers such as shelter-
belts can help control odours (Parker et al., 2012; Tyndall & Colletti, 2007; Tyndall &
Larsen, undated). The vegetation seems to act primarily by removing particulates that
contain the odiferous volatile organic chemicals, and by promoting dilution through
 62  Benefits of green infrastructure

turbulent air mixing (Parker et al., 2012). Chapter 3 covers the value of indoor vegetation
in removing odours.

2.10  Energy efficiency

Dunnett and Kingsbury (2004) suggest that reducing the internal temperature of a building
by 0.5°C can reduce energy use of air conditioning by as much as 8%; Bass et al. (2003)
state that above a threshold temperature (not elaborated in their report) every 1°C increase
results in 5% increased electricity use. Peck et al. (1999) noted that about 33% of the heating
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

requirement of a building is caused by wind-chill and that wind protection by vegetation

could reduce the effect by 75%, equivalent to a reduced heating requirement of 25%. Pérez
et al. (2011a) summarise the mechanisms by which vegetation helps save energy either from
avoiding or reducing the use of cooling or heating systems:

• shade cast by vegetation intercepting sunlight

• thermal insulation provided by the plants and the growing medium (and associated
structures)
• cooling effect of evapotranspiration
• moderation of windspeed.

These features will naturally be affected by the general climate of the area, the season,
specific weather on any given day, the type of plants used, their seasonal nature (evergreen/
deciduous); the structure of the vegetation (i.e. just a single layer or several different layers
of plants of different height), density and depth of the plant layer, the proportion of the
wall covered, the nature of the growing medium and its water retention capacity, and any
additional infrastructure such as drainage layers, waterproof membrane, additional insulation,
as well as the type of building material and wall orientation (Kontoleon & Eumorfopoulou,
2010; Perez et al., 2011a, 2011b).
In general terms, the energy-saving aspects of vegetation through tree shading require
trees to be in the right place and in leaf at the right time. So providing shade in the summer
to sunlit surfaces will save energy used in air conditioning, but the same trees will not be
helpful if they are in leaf in the winter as they would stop sunlight warming the house. So
deciduous trees would be a good choice for aspects of the house that need shade in summer
but sun in winter; house aspects that need insulation from cold weather, especially wind, in
the winter (the north in the northern hemisphere) would be better off with evergreen species
(McPherson et al., 2006). Trees are most effective in shading, and reducing air-conditioning
costs, on the west-facing side of walls (in the northern hemisphere) and windows; and, for
shade, the density of tree crowns is less important than shape (McPherson, 1994) (see section
2.7.4 for trees and wind and 2.7.5 for value in reducing climate change impacts including
through improving energy efficiency). Chapter 5 on green walls and Chapter 6 on green
roofs include sections on how cladding buildings with vegetation can improve insulation and
hence reduce heating and air-conditioning costs.
 Benefits of green infrastructure  63

2.11 Summary
Vegetation in urban areas has multifunctional value.

• Human health and comfort are directly enhanced by vegetation (e.g. by pollution
control, shade).
• Mental health is improved by exposure to vegetation.
• Green spaces provide opportunities for exercise, recreation, socialisation, education and
food production.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

• Climate change and urban heat island impacts can be reduced by sustainable urban
drainage, shading, and the climate moderation effects of vegetation.
• Areas can experience economic benefits from intelligent use of vegetation.
• Energy efficiency of buildings can be improved by the use of vegetation.
• Wildlife habitat can be created and enhanced by a green infrastructure approach which
in turn can improve the quality of life for humans.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

7KLVSDJHLQWHQWLRQDOO\OHIWEODQN
 3
INDOORS
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

This chapter examines the value of vegetation inside buildings including:

• improvement of the work environment

• improved physical and mental health
• reduction of pollutants including volatile organic chemicals.

3.1  Why vegetation indoors?

Green infrastructure is not just about the outside of buildings (Figure 3.1). This chapter will
demonstrate the value that incorporating green elements inside buildings, and even views of
green spaces from windows, has for mood and mental and physical health. Vegetation can
also moderate the microclimate of buildings and remove air pollutants. Examples will include
a range of situations such as the home, computer rooms, hospitals, offices and department
stores. Green elements in such situations can vary from potted plants (Burchett et al., 2010)
through to internal green walls adjacent to stairways, walkways and foyers (Butkovich et al.,
2008; Cosgrave, 2009; Knowles et al., 2003).
Grinde and Patil (2009) reviewed 50 studies for evidence to support the ‘Biophilia’
hypothesis – that a lack of visual contact with nature creates a discord that impacts negatively
on health and well-being – and concluded that the concept had some merit. Bringslimark et
al. (2009) critically examined a number of studies on the psychological value of indoor plants
and found that they were very heterogeneous in nature and recommended more rigour in
future experimental designs. EPA (2009) indicated that Americans spend 90% of their time
indoors and Jenkins et al. (1992) that Californians over the age of 11, in 1987 and 1988
when their survey was carried out, spent on average 87% of their time in buildings and 7%
in enclosed vehicles. Hitchings (2010) suggested there is now a preference for being indoors
among many in the Western/developed world. Some 50% of the office workers he studied did
not leave their building during working hours, the remainder doing so only briefly – and did
not consider this unusual behaviour despite their indoor environment being ‘unremarkably
bland’. As vegetation is well known for having beneficial effects on the mental and physical
 66 Indoors
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 3.1  Indoor plants create atmosphere, improve aesthetics and deliver pollution-reduction
services: left: in the home © John Dover; right: at work, this example is from Centrica’s offices in
Oxford, 600 people can see this wall, containing 9,000 plants, from their desks. © Biotecture Ltd

state of people (see below), as well as acting as air conditioners and cleaners, the presence of
plants in the workplace and at home becomes even more important with the reluctance of
people to leave buildings. But which plants are best? Wolverton (1997) tested 50 plants for
a range of attributes and the top ten were given as: areca palm Chrysalidocarpus lutescens, lady
palm Rhapis excelsa, bamboo palm Chamaedorea seifrizii, rubber plant, the dracaena Dracaena
deremensis ‘Janet Craig’, English ivy Hedera helix, dwarf date palm Phoenix roebelenii, the Ficus
Ficus macleilandii ‘alii’, Boston fern Nephrolepis exaltata and peace lily Spathiphyllum sp. The
following sections examine the values that plants bring to the indoor environment.

3.2  Noise reduction

Costa and James (1995) suggest that plants are unlikely to operate in the same way as walls
in terms of sound reduction. The sound insulation properties of a brick wall of 450 kg.m-2
being given as 48 dB reduction and a plasterboard wall of 7.5 kg.m-2 = 22 dB reduction; a
(1906) estimate of sound insulation by plants was quoted at 11 dB. Plants were considered
to reduce noise inside buildings through absorption, diffraction and reflection of sound
– properties which would vary with species, size, plant architecture, potting media and
containers, moisture content of soil, planting density, etc. Effects would be most noticeable
and valuable in rooms with many hard surfaces and Costa and James (1995) considered that
plants would perform best at reducing the reverberation time for noises of higher pitch with
lower frequencies being diffracted or reflected and absorbed by other materials. However,
Wong et al. (2010b) tested a green wall system, using the Boston fern Nephrolepsis exaltata, in
a reverberation chamber and showed that it had good sound-absorption qualities at a range
of frequencies and that performance increased with the density of the planting. Praag (2011)
cited in Rutgers (2012) compared the sound-reduction qualities of a living wall built on a
wooden framework with one based on an open framework supporting plants in aluminium
containers and with a glass façade. The wooden structure performed best, probably because
the gaps between the aluminium containers allowed sound propagation. At 1  kHz the
 Indoors  67
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 3.2  Breakout space at Q-Free’s headquarters in Trondheim, Norway. The wall was
installed for both its aesthetic and noise-reduction properties. © Biotecture Ltd

wooden-framework living wall absorbed almost 80% of the sound, the best the glass façade
could do was 10% at the 125 Hz band. Natural soil also appeared to be better than artificial
media as a sound absorber (Rutgers, 2012). Green walls may be particularly useful in reducing
noise in rooms, as Costa and James (1995) suggest that walls and corners of rooms are better
places to site plants than in the centre of such spaces (Figure 3.2). Costa and James (1995)
include a table that shows sound reduction by Dracaena fragrans, Dracaena marginata, Ficus
benjamina, Howea forsteriana, Philodendron sp., Schefflera arboricola and Spathiphyllum sp. at a
range of frequencies from 125 Hz to 4 kHz.

3.3  Plants and illness

Ulrich (1984) in a study based in a hospital in Pennsylvania, USA, showed how even visual
access to vegetation could have positive outcomes on health. He showed that post-operative
recovery periods were shorter in patients who had access to a window that looked out onto
trees compared with those who had windows looking out onto a brick wall. These patients
were also more likely to receive fewer doses of strong painkillers, being more likely to need
asprin and acetaminiophen rather than potent narcotics. In Korea, Park and Mattson (2009)
looked at post-operative recovery of 80 female thyroidectomy patients who either had flower
and foliage plants in their rooms or no plants at all. All ‘plant’ rooms had an arrowhead
vine Syngonium podophyllum ‘Albolineatum’, Cretan brake fern Pteris cretica ‘Albolineata’,
variegated vinca Vinca minor ‘Illumination’ and yellow star jasmine Trachelospermum asiaticum
‘Ougonnishiki’ and two plants each of dendrobium Dendrobium phalaenopsis, peace lily
 68 Indoors

Spathiphyllum aureum, golden pothos Epipremnum aureum and kentia palm H. forsteriana.
As with the study of Ulrich (1984), post-operative recovery periods were shorter and the
use of analgesics lower for patients in rooms with plants; in addition, perceptions of pain,
anxiety and fatigue were also lower and a more positive attitude evident in association with
plants. Preventing fatigue and restoring the attention of patients undergoing serious illness
through environmental stimulus was considered to be important in cancer patient recovery
programmes (Cimprich, 1993). Breast cancer patients exposed to the natural environment
for two hours per week during early therapy were better able to direct their attention than
those from a control group not so exposed (Cimprich, 2003).
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

3.4  General health and comfort – and the workplace

Staff working in the Institute of Physics building in Berlin, which has a façade covered in
wisteria, like seeing the plants from their windows and prefer it to having sunshades (Steffan
et al., 2010). The presence of plants in the work environment or at least views of nature is
a recurring theme in the literature, some from in situ experiments and some lab-based. An
example of the latter was that of Chang and Chen (2005) in Taiwan who reported responses
to computer-manipulated images of various office environments in their laboratory and
monitoring responses to the images using electromyographic, electroencephalographic, blood
volume pulse and psychological state questionnaire techniques. The most positive office
configuration (least stressful environment) was a window view of a nature scene coupled
with indoor plants. Joye et al. (2010) suggested that incorporating greenery in retail outlets
should have benefits to retailers, and Brengman et al. (2012) demonstrated, in an online
photographic experiment in the Netherlands with 4,293 participants drawn from a 100,000-
strong consumer panel, that ‘in-store’ greenery gave shoppers a degree of pleasure and stress
reduction in complex store environments, but the effect was not significant in ‘lean’, or more
sparse, settings. The boutique Anthropologie managed to get free full-colour coverage of the
opening of its new store in Regent Street, London by a national newspaper partly on the
basis of its stairwell being fitted with a 160 m2, 200 m high interior green wall composed of
18,000 plants from 14 species (Cosgrave, 2009) (Figure 3.3).
In an early study Lohr et al. (1996) demonstrated the positive values that plants could have
on people working in enclosed office environments without access to windows. Her team
gave a computer-based task to 98 students and staff at Washington State University, USA, in
a bland computer room with and without plants. Students working in an environment with
plants demonstrated higher productivity as measured by faster reaction times, less stress (as
recorded by lower systolic blood pressure) and, at the end of the task, higher attentiveness.
Bringslimark et al. (2007) also found that plants had a positive effect on productivity. Dravigne
et al. (2008) examined the responses of 449 self-selected full-time office workers in Texas
and the Midwest, USA, to their working environment using an 80 question survey. Subjects
were selected from a larger initial sample by matching for work schedule, ethnicity, gender
and salary. Four unequal groupings were used (proportion of respondents in parentheses):

• no plants in the office or windows with views of green spaces (50.6%)

• plants in the office, but no views of green space (18.2%)
• no plants in the office, but views of green space (13%)
• plants in the office and views of green space (18.2%).
 Indoors  69
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 3.3  Green Wall installed in the stairwell of Anthropologie’s store in London.
© Biotecture Ltd. Staff report that they go home with fewer headaches than when working in
other retail spaces because of the ‘pleasant atmosphere created by the living wall’ (Richard Sabin,
pers. comm.).
 70 Indoors
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 3.4  The author’s old office before moving to a modern building with no external view.
The author preferred his old office.

The results showed that office workers with plants and green views and just plants in their
office had higher job satisfaction and quality of life scores than those without plants (Figure
3.4). Respondents with natural views in the absence of plants also had an overall more
positive view of their quality of life than those with no plants or natural views. When mean
job satisfaction scores were compared for gender, males responded positively to plants and
plants+natural views, but females did not. Other than commenting that earlier studies had
not found such a gender difference, no explanation was offered. Shibata and Suzuki (2002)
also found a gender effect with males performing better when plants were present for
‘creative’ tasks.
Smith et al. (2011a) examined the impact of introducing plants into the open-plan
offices of two floors of a large multinational financial services company based in Edinburgh:
one floor had plants, the other did not. Initially (for 3.5 months) a low level of plants was
introduced into the ‘planted’ office, followed by a further 2.5 months with a higher density.
Staff sickness data were compared between experimental groups and with the same time
period in the year prior to the study; perception data was gathered via a questionnaire survey
carried out before and after plant installation. Results indicated that plants reduced feelings
that the work environment contributed to work pressure and that health-related concerns
were lower where there were plants. Workers in the control office also indicated that they
would like plants. Morale was not improved in the office with plants but this was attributed
to the nature of the work rather than to the presence of plants; this view might be corrobo-
rated by the very high sickness levels prior to plant installation. During January to June, on
 Indoors  71

the ‘planted’ floor in 2007 (the year before plants were introduced) 1,351 days were lost
to sickness compared with 340.5 days on the control floor. The figures for the equivalent
period in 2008, when plants had been installed, was: ‘planted’ floor 701 days and control
floor 425 days. The results suggest a very strong influence of plants in reducing sickness
levels. Shibata and Suzuki (2002) and Hesselink et al. (undated) make a distinction between
the impact of plants on ‘creative’ workers compared with ‘production’ workers engaged
in routine tasks whereby plants enhance ‘creative’ activities but distract from ‘production’
activities – though noting that production workers would still benefit from the restorative
presence of plants in areas away from their workstations. Hesselink et al. (undated) did not
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

report on gender effects but did find that the subjects who reported themselves as being
physically exhausted or with high work-related stress benefitted most from the presence of
plants. Nieuwenhuis et al. (2014) in a series of three studies, which sequentially refined their
experimental approach, concluded that plants in offices compared to offices without plants
(or any other distractions: ‘lean’ offices) had beneficial effects on workers in terms of both
perception of the work environment (air quality, concentration, workplace satisfaction) and
their productivity. They considered that the improvement in the perception of the work
environment persisted over time, rather than being just a brief response to a change in condi-
tions. The productivity change was quantified, in their final study using tasks designed to test
‘vigilance’ and ‘information processing’, as a 15% increase in offices with plants compared
to those without plants.
In Norway, Fjeld et al. (1998) investigated the effect of plants in the workplace on the
health of 51 office workers. The work was carried out during three months in the springs
of 1995 and 1996, with half the offices having plants and half without for one spring and
then the situations being reversed the following spring. Staff filled in questionnaires on
12 symptoms (five neurophysiological: fatigue, feeling heavy-headed, headache, nausea/
drowsiness, concentration problems; four related to mucous membranes: ‘itching, burning,
irritation of eyes’, ‘irritated, stuffy or running nose’, ‘hoarse, dry throat’ and coughing; three
skin-related: ‘dry or flushed’ facial skin, ‘scaling/itching’ of scalp or ears, hands with ‘dry,
itching, or red’ skin). The plants used were Chinese, Philippine or golden evergreen Aglaonema
commutatum, striped dragonpalm dracaena D. deremensis, golden pothos Epipremnum aureum
and heart-leaf philodendron Philodendron scandens on the window sill and the ‘Janet Craig’
dracaena and E. aureum by the door. There was a significant, 23%, reduction in neuropsycho-
logical symptoms overall, principally due to a 30% reduction in perceived fatigue. In addition,
mucous-related scores were significantly reduced by 24% overall, primarily due to a reduction
in coughing by 37% and symptoms of dry/hoarse throat by 25%. Skin-related conditions
were not significantly reduced when taken as a group, but ‘dry or flushed facial skin’ was
significantly reduced by 23%. The causes of the improvements in health and reduction of
discomfort were considered to probably be due to an improvement in air quality, and an
increase in well-being mediated by a more congenial working environment. A questionnaire
from this study was later reported by Fjeld (2000) with 82% of the participants in the study
agreeing to the statement ‘I feel more comfortable if I have plants in my office’ and 82% to
‘I would like to have plants in my office in the future’. These findings were confirmed by a
later study in the diverse environments of a hospital X-ray department and a high school, the
latter with students aged 14–16 (Fjeld, 2000). These studies differed from Fjeld et al. (1998) in
that full-spectrum lights were installed as well as plants. In the X-ray department the benefits
were assessed in relation to exposure time and were found to have increased with the amount
 72 Indoors

of time spent during the day in the radiology room (decrease in symptoms by 34% working
all day; 21% present half a day; 17% less than half a day); the benefits were confirmed to be
still operating 11 months after the end of the study (with the plants still in place). The school
classrooms were modified by having plants in three rooms (air was also passed through the
growing medium/root zone), the control situation was a further three rooms without plants.
Significant impacts included a reduction in reported headaches by 37%, dry, itching eyes by
30%, dry or hoarse throat by 36%. Perceptions of the working environment by the students
were also improved by the presence of plants. Bringslimark et al. (2007) demonstrated, in a
survey of 364 office workers, that having plants in view or in their own work area signifi-
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

cantly reduced the amount of sick leave they took. A contrary finding was that plants in the
area very near the workstation had a positive relationship with sick leave, though why such
a result should be generated was not satisfactorily explained, although the distinction made
between ‘creative’ and ‘production’ values of plants by Hesselink et al. (undated) might offer
some insight.

3.5  Air conditioning

Cañero et al. (2012) cite the Spanish building regulations (MITTMH, 2009) which give
acceptable ranges for both temperature (20–24°C) and humidity (30–70%) and note that
in warm climates air conditioning, requiring significant energy consumption, is needed
to maintain such regimes. Even in temperate countries such as the UK, peak summer
temperatures can make office buildings uncomfortable, especially with the reliance on
heat-emitting electronic equipment such as computers for everyday business activities. Lohr
and Pearson-Mims (1996) found that interior plants improved the humidity levels of both
computer labs and offices at Washington State University. Smith and Pitt (2011) found
plants provided more comfortable levels of humidity when introduced into an open-plan
office; installation of plants raised levels from below 30% to 40–60% relative humidity.
Costa and James (1995) cite references detailing the evaporative cooling value of plants in
buildings.

3.6  Microbial load reduction

One of the fears about plants in buildings is that they encourage mould organisms
because of elevated humidity (Wolverton & Wolverton, 1996). However, Wolverton and
Wolverton (1996) in a simple study found that rooms with plants could have 50% lower
airborne microbes compared with plant-free rooms. Petri dishes with nutrient agar plates
were exposed to the air in rooms with or without plants and cultured in an incubator
and the resulting microbial colonies counted. Organisms identified included bacteria,
actinomycetes (bacteria with fungal-like filamentous colonies) and moulds. Wolverton
and Wolverton (1996) concluded that plants actively suppressed microorganisms and
suggested that plant defence chemicals (allelochemicals) (Ehrlich & Raven, 1967; Vollmer
& Gordon, 1975) such as terpenes and phenolic compounds on leaf surfaces and exudates
from roots (Bakker et al., 2012) were the primary source of control. Fjeld (2000) reported
that including plants in a hospital radiology department did not result in an increase (or
decrease) in the fungi/fungal spore content of the air. Burchett et al. (2010) in a study of
offices in Sydney with and without plants showed no significant increase in fungal spores
 Indoors  73

in offices with plants – and all offices had almost a 20-fold lower concentration of spores
than outdoor air.
Curiously, although describing the chemical-degrading abilities of microorganisms
associated with plants, Wolverton and Wolverton (1996) did not consider that such
organisms could have a role in suppressing the microbial load in rooms. However, the
surfaces of plant leaves and stems (the phylloplane) and the root zone (the rhizosphere)
maintain communities of microorganisms potentially useful in removing airborne microbes.
The rhizosphere, for example, acts as host to a wide range of organisms including bacteria,
some of which help protect plants against disease not only by stimulating resistance, but
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

also by the production of suppressive antibiotics and enzymes which kill other microbes,
including fungi, by directly attacking cell walls of potential competitors (van Loon et al.,
1998; Whittaker & Feeny, 1971). Plants actively promote the growth of the root zone
microbial community and tailor it to their needs via root exudates (Bais et al., 2006; Bertin
et al., 2003). Likewise, the phylloplane hosts microorganisms that can, either by competition
or direct antagonism, protect plants from disease (Blakeman & Fokkema, 1982) and direct
inoculation with ‘beneficial’ organisms has been demonstrated in agricultural contexts (Li et
al., 2006) and also in laboratory studies on the degradation of volatile organic compounds
(VOCs) (Sandhu et al., 2007). The protective action of the microbial community of
both phylloplane and rhizosphere may be disrupted by the use of fungicides (Blakeman
& Fokkema, 1982), which probably should be avoided, if possible, during routine plant
maintenance programmes.

3.7  Air pollution

3.7.1  Carbon monoxide, carbon dioxide and nitrogen dioxide

Wolverton et al. (1985) showed that the spider plant Chlorophytum elatum var. vittatum and
golden pothos E. aureus could reduce levels of carbon monoxide; in the same study the
spider plant was also shown to reduce nitrogen dioxide levels. Carbon monoxide and VOC
levels were reduced in an open-plan office after introduction of plants compared with an
‘unplanted’ office (Smith & Pitt, 2011). Tarran et al. (2007) showed that carbon monoxide
levels could be significantly reduced in offices by three D. deremensis ‘Janet Craig’ plants (in
300  mm pots). The scale of the reduction was impressive: 92% reduction (from 0.225 to
0.017  ppm) in an air-conditioned building, 86% (from 0.071 to 0.010  ppm) in naturally
ventilated offices; carbon dioxide levels were also reduced by 10 and 25% respectively. Carbon
dioxide removal by indoor plants is not a subject that has received extensive coverage, but
Torpy et al. (2014) assessed eight indoor plant species for their ability to remove CO2 and
found, on a leaf area basis, that bamboo palm Dypsis lutescens was the most effective followed
by dracaena ‘compacta’ D. deremensis, weeping/Benjamin’s fig Ficus benjamina and kentia palm
H. forsteriana. The first three are plants adapted to high-light conditions and the last adapted
to low-light conditions. Torpy et al. also found that light levels affected uptake of CO2,
with low-light species such as dwarf mountain palm Chamaedorea elegans, Chinese evergreen
Aglaonema commutatum and H. forsteriana having reduced uptake in high-intensity light condi-
tions. The other two species tested were aspidistra Aspidistra elatior and the Moreton Bay
chestnut Castanospermum australe.
 74 Indoors

3.7.2 Odours
Unpleasant odours in buildings, such as hydrogen sulphide, ammonia and methyl mercaptan,
are clearly undesirable and some are characteristic of the specific environment; for example,
ammonia is a typical odour present in hospitals, nursing homes, lavatories and refuse stores
(Oyabu et al., 2003). Oyabu et al. (2003) investigated the potential of plants for the removal
of odours using golden pothos (E. aureum) in three different soils and also used a ‘snake
plant’ and a ‘rubber plant’ as comparisons. Formaldehyde, acetone and ammonia were used
as test odours. Whilst the authors noted the harmful nature of the chemicals, it was the
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

subjective nature of the smell rather than the toxicity they were interested in. The plants were
considered effective in odour removal, with time needed for removal being longer for higher
concentrations of odours and odours of larger molecular weight. Soil type influenced odour
removal; unfortunately no controls (using pots with soil but no plants) were used in the study.

3.7.3  Volatile organic chemicals

Volatile organic compounds (VOCs) are a group of about 18 million organic (carbon-
based) chemicals (Golding & Christensen, 2000). The term ‘volatile’ relates to their ability
to vaporise at typical room temperatures (for more precise definitions see Guieysse et al.,
2008). Many hundreds of VOCs have been identified from indoor air (Posudin, 2010) and
are usually found at far higher concentrations indoors than out (Brown, 2002). Wolkoff
(2003) points out that as building products constitute the largest indoor surface area, they are
therefore also the source of most indoor VOCs. Wolkoff (2003) also documents the various
EU and Nordic standards to minimise such emissions and gives examples of trade labelling
schemes such as the Gemeinschaft Umwelt Teppisch Ordnung (Society for Testing Carpets
in Europe). VOC sources include the resins used in chipboard wood products (Chiappini et
al., 2011) and are emitted from a wide range of building materials, electrical goods, furniture,
paint, household products and print media (Posudin, 2010; Wang et al., 2007; Wolverton &
Wolverton, 1993). Many VOCs have known biological activity and can be found at concen-
trations potentially damaging to human health in indoor air (Wang et al., 2007). VOCs
have been implicated in ‘sick building syndrome’ (ALA, undated) and exposure can result in
nausea, vomiting, abdominal pain and diarrhoea (Xu et al., 2010). Whilst newly constructed
buildings/products will have the highest concentrations of VOCs, the chemicals may be
continually emitted ‘outgased’ at lower levels for years (Wolverton & Wolverton, 1993).
The US National Aeronautics and Space Administration (NASA) initiated work on plants
in the early 1980s to identify ways of purifying air as part of their research into independent
structures such as space stations and also energy-efficient homes, which, by definition, have
reduced ventilation and are sealed against unwanted draughts (Dimitroulopoulou, 2012;
Wolverton et al., 1984; 1989b).

3.7.4  Plants and mechanisms of VOC removal from air

A number of studies have indicated that plants are capable of removing VOCs from indoor
air via a variety of mechanisms (Figure 3.5). Of the hundreds of VOCs contaminating
indoor air, perhaps most work using plants has been carried out on formaldehyde and will
be examined below – studies on other VOCs are briefly given in Table 3.1 and Table 3.2.
 Indoors  75

Plants are probably able to detoxify many compounds directly (Schaffner et al., 2002), but
Wolverton et al. (1989a), in a pioneering experiment with benzene and whole plants and
plant roots of the dragon tree D. marginata, showed that the major site for removal of the
compound by D. marginata was via the root zone. Wolverton and Wolverton (1993) inves-
tigated the differential proportions of formaldehyde and xylene removed by different plant
parts and estimated that 33–49.5% was via the plant leaves and stems and the balance by the
roots – or rather by some of the microorganisms growing in association with plant roots.
Tests with the plants F. benjamina or Spathiphyllum growing in potting soil for five months
showed the root zone removed 660 and 659  μg.h-1 respectively more formaldehyde than
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

potting soil left for five months with no plants; the equivalent figures for S. trifasciata and
Kalanchöe were worse than potting soil on its own at -91 and -92 μg.h-1 respectively. Not
all plant species tested appeared to have VOC-degrading microorganisms which, in their
tests, were shown to be gram-negative rod bacteria: Kalanchöe soil contained gram-positive
bacteria whilst Spathiphyllum contained gram negative bacteria. Wood et al. (2002) demon-
strated benzene and hexane removal by rhizosphere microorganisms with Spathiphyllum
wallisii ‘Petite’ peace lily, D. deremensis ‘Janet Craig’ and Kentia palm H. forsteriana in both
potted and hydroponic media. Chun et al. (2010) confirmed activity of root microorganisms
by growing bacterial material removed from the rhizosphere and testing with the VOCs
benzene and toluene. Inoculation of the cultured isolates into pots with living plants further
confirmed the VOC-removing potential of the rhizosphere bacteria when compared with
un-inoculated plants. Wolverton and Wolverton (1993) hypothesised that VOCs taken in
by leaves and stems might be translocated by the plant to the root zone for removal by
microorganisms, but recent work by Sandhu et al. (2007) has demonstrated degradation of
phenol by microbial communities on leaf surfaces and Schaffner et al. (2002) demonstrated
formaldehyde degradation directly by the plant. Tani and Hewitt (2009) isolated peace lily
Spathiphyllum clevelandii and golden pothos E. aureum leaves from the rest of the plant and
demonstrated uptake of 13 of 14 VOCs by the former and five of six VOCs by the latter (the
exception in each case being acetone) (Table 3.1). The removal mechanisms postulated were
assumed to be metabolism in the leaf or translocation to other parts of the plant following
stomatal uptake. Wolverton and Wolverton (1993) also felt that the degradation rate may
have improved over time during their experiments, an observation confirmed by Orwell
et al. (2004) who coined the term ‘induction’ for the initial, slower, take-up period (over a
2–4 day period) and confirmed the activity of root-zone microorganisms. A lag in uptake,
given the rapid ability of microorganisms to replicate, could have simply been through an
increased population of microorganisms capable of breaking down the VOCs, but whilst
Wolverton et al. (1989b) demonstrated increased bacterial populations after six weeks of
periodic exposure to benzene (from 3.1 x 104 at the start rising to 5.1 x 104 cfu.g-1 after six
weeks) they felt that this correlation was not an adequate explanation of their observations.
Wolverton and Wolverton (1993) suggested that the accelerated uptake after a lag-phase
was via microbial adaptation, a suggestion repeated by Orwell et al. (2004) who considered
that it was actuated by the induction of a specific biochemical pathway in the microbes or
plant (or both). Kim et al. (2012) considered that the induction reflected a change in gene
expression by the plants, microorganisms, or both, rather than simple population increase
by the microorganisms. However, Wood et al. (2002) demonstrated that whilst the overall
population size of the microbial community did not change following exposure to VOCs (in
their study of benzene) the community structure did change with increases in population size
 76 Indoors

TABLE 3.1  Studies on volatile organic chemicals other than formaldehyde in relation to plant uptake

VOC studied Study Summary

α-Pinene Yang et al. See Table 3.2 for tests of plants for this VOC.
(2009)
Acetone Tani & Hewitt Acetone exposed to peace lily Spathiphyllum clevelandii and
(2009) golden pothos Epipremnum aureum leaf attached to the plant
but contained within a teflon bag into which acetone was
continuously introduced via a flow-through mechanism. The
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

VOC uptake was estimated using proton transfer reaction

mass spectrometry. The plants did not take acetone up
consistently and only when concentrations were very low.
Ammonia Wolverton Eleven species of plant assayed for ammonia removal, rates
and Wolverton varied from 7356 μg.h-1 to 984 μg.h-1 (see Table 3.3 for
(1993) details).
Cornejo et Spider plant Chlorophytum comosum was shown to remove
al.(1999) trichloroethylene.
Benzaldehyde Tani & Hewitt Benzaldehyde taken up by peace lily S. clevelandii leaf (see
(2009) Acetone above for more detail).
Benzene Wolverton et al. C. morifolium, D. deremensis ‘Warneckei’, D.massangeana,
(1989a,b) F. benjamina, G. jamesonii could all reduce benzene
concentrations – albeit with differing levels of efficiency. See
further Table 3.2 and also, for rhizosphere effects, Chun et al.
(2010) under Toluene.
Cornejo et al. Eight plant species (flaming Katy Kalanchoe blossfeldiana,
(1999) spiderwort Tradescantia fluminensis, dracaena D. deremensis,
Chinese primrose Primula sinensis, spider plant Chlorophytum
comosum, regal geranium Pelargonium x domesticum, strawberry
saxifrage Saxifraga stolonifera, and Magnesia sp.) were tested
for removal of benzene in a sealed chamber with 108 μg
benzene. On a whole plant basis the best were P. domesticum
(95% removal), K. blossfeldiana (85% removal) and S. stolonifera
(41% removal); the remaining five species had between
10 and 25% uptake. Expressed as μg of benzene removed
per gram of foliage over 24 h, the best species were P.
domesticum (8.5 μg), K. blossfeldiana (3.4 μg) and S. stolonifera
and Magnesia sp. (both 2 μg). Later tests with K. blossfeldiana
and rubber plant Ficus elastica confirmed the activity of the
former, but also the good removal efficiency of the latter.
Orwell et al. Seven plant species tested for benzene removal: kentia palm
(2004) Howea forsterianai, peace lily Spathiphyllum floribundum var.
Petite, S. floribundum var. Sensation, D. deremensis var. ‘Janet
Craig’, D. marginata, devil’s ivy E. aureum, and Queensland
umbrella tree Schefflera actinophylla var. Amate. The results
showed that most of the activity was via microbes in the
root zone, but some species also removed benzene via the
foliage. On a per plant basis, D. deremensis var. ‘Janet Craig’
performed best immediately after ‘induction’ and H. fosteriana
worst but data suggests that removal rates may vary with time
and plant species from induction. Data also presented on
removal rate related to leaf area, dry weight of shoots, roots
and pot mix.
 Indoors  77

VOC studied Study Summary

Liu et al. (2007) Seventy-three species of plant were screened for their ability
to remove benzene from a 150 ppb benzene-in-air feedstock
over a 2 h period: 20 species removed no benzene, 13
removed 0.1–9.99% benzene, 17 removed 10–20% and 17
removed 20–40%. Ten species were then chosen for further
evaluation from the latter group. Plants were fumigated for
8 h/day for 2 days using the same feedstock concentration.
On the basis of benzene absorption expressed as μg.m-2
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

leaf area.day-1 by far the most effective species was the jade
plant Crassula portulacea (syn. C. ovata) at 724.9 μg.m-2.d-1
the next best being the hydrangea Hydrangea macrophylla at
293.7 μg.m-2.d-1. The worst two species of the ten were the
Boston fern N. exaltata cv. ‘Bostoniensis’ and D. deremensis cv.
variegata with 73.5 and 59 μg.m-2.d-1 respectively. The other
plants were intermediate between these values (higher to
lower: Cymbidium ‘Golden Elf ’, Ficus microcarpa var. fuyuensis,
Dendranthema morifolium, Citrus medica var. sarcodactylis,
Dieffenbachia amoena cv. Tropic Snow, Spathiphyllum
‘Supreme’).
Crotonaldehyde Tani & Hewitt Taken up by peace lily S. clevelandii leaf (see Acetone above
(2009) for more detail).
Diethyl ketone Tani & Hewitt Diethyl ketone taken up by peace lily S. clevelandii and golden
(2009) pothos Epipremnum aureum leaves (see Acetone above for
more detail).
Methacrolein Tani & Hewitt Methacrolein taken up by peace lily S. clevelandii leaf (see
(2009) Acetone above for more detail).
Methyl ethyl Tani & Hewitt Methyl ethyl ketone taken up by peace lily S. clevelandii and
ketone (2009) golden pothos E. aureum leaves (see Acetone above for more
detail).
Methyl isobutyl Tani & Hewitt Methyl isobutyl ketone taken up by peace lily S. clevelandii
ketone (2009) and golden pothos E. aureum leaves (see Acetone above for
more detail).
Methyl iso-propyl Tani & Hewitt Methyl iso-propyl ketone taken up by peace lily S. clevelandii
ketone (2009) and golden pothos E. aureum leaves (see Acetone above for
more detail).
Methyl n-propyl Tani & Hewitt Methyl n-propyl ketone taken up by peace lily S. clevelandii
ketone (2009) and golden pothos E. aureum leaves (see Acetone above for
more detail).
Iso-butyraldehyde Tani & Hewitt Iso-butyraldehyde taken up by peace lily S. clevelandii leaf (see
(2009) Acetone above for more detail).
Iso-valeraldehyde Tani & Hewitt Iso-valeraldehyde taken up by peace lily S. clevelandii leaf (see
(2009) Acetone above for more detail).
N-butyraldehyde Tani & Hewitt N-butyraldehyde taken up by peace lily S. clevelandii leaf (see
(2009) Acetone above for more detail).
N-valeraldehyde Tani & Hewitt N-valeraldehyde taken up by peace lily S. clevelandii leaf (see
(2009) Acetone above for more detail).
 78 Indoors

VOC studied Study Summary

Octane Yang et al. See Table 3.2 for plants tested for removal of this VOC.
(2009)
Pentane Cornejo et Pentane showed to be removed by spider plant C. comosum,
al.(1999) and rubber plant Ficus elastica.
Proprionaldehyde Tani & Hewitt Proprionaldehyde taken up by peace lily S. clevelandii leaf (see
(2009) Acetone above for more detail).
Toluene Kim et al. (2012) Initial low levels of toluene shown to induce a stronger
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

response in Begonia maculata (x3.85), Ardisia crenata (x3.18)

and A. japonica (x2.52) after exposure. The enhanced response
was short-lived requiring continual exposure to toluene or
re-exposure.
Chun et al. Nine plant species tested for the ability of microorganisms
(2010) in their rhizosphere for removal of VOCs; rhizosphere
community of Pachira aquatica performed best and inoculated
into media with other plant species. Demonstrated that
rhizosphere microorganisms could remove removed toluene,
benzene and xylenes from the air. See further Table 3.2 for
plants tested for removal of this VOC.
Orwell et al. Toluene and m-xylene exposed to D. deremensis ‘Janet
(2006) Craig’ and Spathiphyllum peace lilies ‘Sweet Chico’ at 0.2,
1.0, 10 and 100 ppm and in mixtures. VOC removal
induced at 0.2 ppm and reduced below 20 ppb over 24 h.
The presence of toluene also increased the rate of m-xylene
removal.
Trichloroethylene Wolverton et al. Chrysanthemum morifolium, D. deremensis ‘Warneckei’, D.
(1989a,b) massangeana, F. benjamina, Gerbera jamesonii could all reduce
trichloroethylene concentrations – albeit with differing levels
of efficiency. See also Table 3.2 for plants tested for removal
of this VOC.
Xylene Wolverton Thirty plant species were tested for their ability to remove
and Wolverton xylene, rates varied from 610 μg.h-1 to 47 μg.h-1 (see Table
(1993) 3.3 for details). As with formaldehyde (see text) two Boston
fern plants or three ‘Janet Craig’ dracaenas would be capable
of removing the typical new office loading of 0.22 μg.l-1 of
xylene from a 22.32 m3 office (493 μg xylene), although a
single pygmy date palm or two dumb canes should achieve
the same effect. See further Chun et al. (2010) under
‘Toluene’ for rhizosphere effects, and Orwell et al. (2006) for
removal by D. deremensis and Spathiphyllum peace lilies and
synergistic effects.

Sources: as indicated.
 Indoors  79

TABLE 3.2  Performance of plants examined by Yang et al. (2009) for their ability to remove five
VOCs from four classes of compound

Poor removal performance Intermediate removal Excellent removal

performance performance
Peperomia clusiifolia Ficus benjamina Hemigraphis alternata
Chlorophytum comosum Polyscias fruticosa Hedera helix
Howea belmoreana Fittonia argyroneura Tradescantia pallida
Spathiphyllum wallisii Sansevieria trifasciata Asparagus densiflorus
Schefflera arboricola Guzmania sp. Hoya carnosa
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Codiaeum variegatum Anthurium andreanum

Calathea roseopicta Schefflera (Dizygotheca)
elegantissima
Aspidistra elatior
Maranta leuconeura
Dracaena fragrans
Ficus elastica
Dieffenbachia seguine syn.
amoena
Philodendron scandens ssp.
oxycardium
Syngonium podophyllum
Epipremnum aureum
Scindapsus aureus
Pelargonium graveolens
Range of removal efficiency by VOC (mg.m–3. m–2.h–1)
Benzene 0.03 to 1.20 0.66 to 2.74 2.21 to 5.54
Toluene 1.54 to 3.18 3.60 to 5.09 5.81 to 9.63
Octane 0.00 to 2.03 0.65 to 3.98 2.76 to 5.58
Trichloroethylene 1.48 to 2.86 3.58 to 6.15 5.79 to 11.08
α-Pinene 2.33 to 4.61 4.30 to 8.68 8.48 to 13.28
Total VOCs 5.55 to 12.98 16.78 to 24.13 26.08 to 44.04

of some species, reductions in others and the appearance of some species in cultures only after
benzene had been administered. Orwell et al. (2004) presented data on benzene removal by
a range of plants (see Table 3.1) using a range of parameters: standard pot plant (12 months
old, 0.3–0.4 m in height in 150 mm diameter pots with a defined potting medium), leaf area,
of shoot dry weight, root dry weight, and dry weight of the potting mix and demonstrated
that the ranking of best/worst in terms of removal efficiency varied depending on parameter
and time since induction. Liu et al. (2007) screened 73 plant species for benzene removal
(Table 3.1) but their initial screen was of very short duration (2 h) and they may have missed
effective plants that needed an induction period. Kumar et al. (2011) suggest that simple
compounds of low molecular weight and high solubility are those most easily degraded by
microorganisms given that more complex compounds require more energy to break them
down, a limitation that may not apply to degradation by the host plants themselves. Kim et
al. (2008) using Ficus benjamina and Fatsia japonica plants convincingly demonstrated that the
 80 Indoors
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Phylloplane

Stomata

Rhizosphere

FIGURE 3.5  Routes and modes of uptake of volatile organic chemicals (VOCs) by pot plants.
VOCs may be degraded by the microbial community living on the Phylloplane (leaf surface);
VOCs may also adhere to the cuticle (the waxy outer surface of leaves) or penetrate directly
through it. The Stomata are used for taking in carbon dioxide and giving off oxygen. VOCs
can also enter by this route and be translocated throughout the plant to be metabolised. The
Rhizosphere is the root zone and supports a microbial community which contains species which
can breakdown VOCs. © John Dover
 Indoors  81

TABLE 3.3  Plants surveyed by Wolverton and Wolverton (1993) for their ability to remove the volatile
organic chemicals (VOC) formaldehyde, xylene, and ammonia from the air*

Removal rate of VOC μg.h-1 and rank (x)

Species type Scientific name Common name Formaldehyde Xylene Ammonia

Bromeliad
Aechmea fasciata Urn plant 234 (31)
Guzmania cherry Cherry guzmania 146 (26)
Neoregelia cv. 47 (30)
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Bulb
Tulip ‘Yellow Present’ Tulip ‘Yellow Present’ 717 (16) 229 (14) 2815 (7)
Climber
Cissus rhombifolia Venezuela treebine 376 (24)
Hedera helix English ivy 1120 (6) 131 (27)
Syngonium podophyllum Arrowhead plant 341 (25) 220 (16)
Ferns
Nephrolepis exaltata Boston fern 1863 (1) 208 (18)
‘Bostoniensis’
Nephrolepis obliterata Kimberley queen fern 1328 (5) 323 (6)
Herbaceous foliage plants
Aglaonema sp. ‘Silver Chinese evergreen 564 (21)
Queen’ ‘Silver Queen’
Anthurium andraeanum Flamingo flower/Tail 336 (26) 276 (8) 4119 (4)
flower
Calathea ornata Prayer plant 334 (27)
Calathea vittata (elliptica) 3100 (6)
Chlorophytum comosum Spider ivy ‘Vittatum’ 560 (22) 247 (12)
‘Vittatum’
Cyclamen persicum Cyclamen 295 (29) 173 (21)
Dieffenbachia camille Dumb cane 469 (23) 341 (2)
Dieffenbachia maculata Dumb cane 325 (4)
Dieffenbachia sp. Dumb cane 754 (15)
‘Exotica compacta’
Dracaena deremensis Janet Craig dracaena 1361 (4) 154 (25)
‘Janet Craig’
Dracaena deremensis Striped dracaena 760 (12) 295 (7)
‘Warneckei’
Dracaena marginata Dragon tree 772 (11) 338 (3)
Dracaena fragrans Cornstalk dracaena 938 (9) 274 (9)
(deremensis)
Euphorbia pulcherrima Poinsettia 309 (28) 116 (28)
Ficus benjamina Benjamin tree 940 (7) 271 (10) 1480 (9)
Ficus sabre Saber Ficus 692 (17)
Homalomena sp. Homalomena 668 (18) 325 (5) 5208 (2)
Kalanchöe Kalanchoe 170 (22)
Liriope spicata Creeping lilyturf 758 (13) 230 (13) 4308 (3)
Rhododendron indicum Azalea 617 (20) 168 (23) 984 (11)
Sansevieria trifasciata Snake plant or mother- 189 (32) 157 (24)
in-law’s tongue
 82 Indoors

Removal rate of VOC μg.h-1 and rank (x)

Species type Scientific name Common name Formaldehyde Xylene Ammonia

Senecio cruentu (Pericallis Cineraria 115 (29)
cruenta)
Spathiphyllum sp. Peace lily 939 (8) 268 (11) 1269 (10)
‘Clevelandii’
Herbs
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Chrysanthemum Hardy garden mum 1450 (2) 201 (19) 3641 (5)
morifolium
Orchids
Dendrobium sp. Orchids 756 (14) 200 (20)
Phalaenposis sp. Moth orchid 240 (30)
Succulent
Aloe barbadensi (vera) Barbados aloe 188 (33)
Woody foliage plants
Chamaedorea elegans Parlour palm 660 (19) 223 (15) 2453 (8)
Phoenix roebelenii Pygmy date palm 1385 (3) 610 (1)
Rhapis excelsa Lady palm 876 (10) 217 (17) 7356 (1)
Notes
The rank of the plants for efficiency is given in parentheses and the top five in removal efficiency are given in
bold for each pollutant.
The temperatures when measurements were made varied between 21.9 and 26.9°C and pot sizes varied from as
little as 8.0 cm and 12.7 cm with all the rest in the range 15.2 to 35.6 cm, so the ranking is only indicative – for
exact details see the original paper.
*Common names have been taken from the general literature and may not be accurate; likely synonyms are given
in brackets after the scientific name. Plants have been grouped in a similar way to that of Kim et al. (2010).

leaves and root zone took in similar amounts of formaldehyde during the day when the plant
was actively taking in air via the stomata, but at night when the stomata were closed the
majority (ratio about 1:11) of the formaldehyde was removed by the root zone. Kim et al.
(2008) suggested that formaldehyde removed by the leaves at night was via the cuticle – they
did not consider phylloplane effects. Yoo et al. (2006) also showed differential uptake between
day and night for the above-ground parts of English ivy Hedera helix, peace lily Spathiphyllum
wallisii, S. podophyllum and grape ivy Cissus rhombifolia when exposed to benzene or toluene;
Cissus rhombifolia was the least effective of the four. The uptake was greater during the day for
all species with the exception of H. helix with benzene which appeared to remove compa-
rable amounts of benzene whether during the day or night. Differences in day:night removal
were less pronounced than implied by the findings of Kim et al. (2008). Yoo et al. (2006)
commented that the ability of H. helix to remove similar amounts of VOCs during the day
or night indicated that the major removal route was mainly by adhesion to the cuticle or
cuticular absorption, but did not consider microbial degradation at the phylloplane.
Wolverton and Wolverton (1993) demonstrated that a particular plant species may not
be uniformly good at removing all pollutants; their best plant at removing formaldehyde,
the Boston fern, was 18th at removing xylene from substantially the same group of plants
 Indoors  83

tested. Likewise, the best plant at removing ammonia in their assays, the lady palm Rhapis
excelsa, was 10th out of 33 for formaldehyde removal and 17th out of 30 for xylene removal
(Table 3.3). Cornejo et al. (1999) found selectivity in uptake of VOCs when presenting
a mixture of benzene and toluene to flaming Katy Kalanchöe blossfeldiana; whilst benzene
was removed by the plant, toluene was not. In other tests with the spider plant C. comosum
and mixtures of trichloroethane, benzene and pentane the uptake of benzene and pentane
appeared to be depressed by the presence of trichloroethane. Yoo et al. (2006) also found
that mixtures of VOCs (benzene and toluene) affected uptake by H. helix, S. walisii, S.
podophyllum and C. rhombifolia – interestingly H. helix outperformed the other three species
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

in the removal of both benzene and toluene from a mixture of the two VOCs. Differential
uptake was also found by Yang et al. (2009) who screened 28 plant species for their ability
to remove five VOCs from four classes of chemical: benzene and toluene to represent
the ‘aromatic’ hydrocarbons (chemicals based on a benzene ring), octane to represent the
straight-chain organic hydrocarbons, trichloroethylene to represent organic compounds
that have a halogen group (in this case chlorine) and α-pinene, a terpenoid compound
found in many essential oils emitted by plants in scents and also used as defence chemicals.
Plants were compared on an uptake/leaf area basis. Of the 28 plants, 16 were considered
poor at VOC uptake, seven ‘intermediate’ and five ‘superior’ (Table 3.2). Four plants
removed all VOC classes effectively: red flame ivy Hemigraphis alternata, English ivy H.
helix, the wax plant Hoya carnosa, and plume asparagus/asparagus fern Asparagus densiflorus
removed all five compounds well and purple spiderwort Tradescantia pallida was effective
against all but octane. Other plants had more selective responses: the nerve plant Fittonia
argyroneura removed the two aromatic hydrocarbons and the halogenated hydrocarbon; the
Benjamin tree F. benjamina removed octane and the terpene; the Ming tree Polyscias fruticosa
only removed octane.
Plants and their associated microbial communities are clearly capable of making a valuable
contribution to indoor air quality, but the removal methods, detoxification mechanisms,
time dependency of response and involvement of unquantified microbial communities of
potentially plant species-specific nature make for a complex system that has not yet been
fully elucidated, but holds considerable promise for improvement in indoor air quality and
significant opportunities for development of novel systems such as biofiltration (Darlington
et al., 2000; Guieysse et al., 2008; Kumar et al., 2011) and indoor living walls (see below).
As yet there appears to be no information on the impact of controlled-release fertilisers,
slow-release fertilisers or liquid applications of fertiliser on the microbial communities of
the VOC-degrading rhizosphere of indoor potted plants or green walls (Šrámek & Dubský,
2007; Tan et al., 2012), although benzene uptake by the arrowhead vine S. podophyllum
grown in hydroculture has been shown to be almost as good as that with conventionally
grown plants (Irga et al., 2013). Whilst it is known that plants expend considerable energy
by diverting up to 45% of photosynthetic output to foster rhizosphere microorganisms,
hypothesised signalling by the plant to elicit a tailored VOC removal has not yet been
confirmed or rejected (see further, and references, in Wood et al., 2002). Recent work has
indicated that ‘plant growth-promoting rhizobacteria’ emit VOCs themselves (including
ethylene) which appear to induce systematic resistance in plants to a range of negative
factors including insect pests, disease, heavy metal contamination and drought (Farag et al.,
2013).
 84 Indoors

Formaldehyde
Formaldehyde (CH2O) is a common indoor pollutant causing a range of symptoms
including eye conditions, skin and respiratory distress, nausea, vomiting and diarrhoea
and is a potential carcinogen (Dingle et al., 2000), although recent work (Golden, 2011)
suggests that the causal link with cancer is not proven and that indoor limits should be set
at 0.1 ppm. Pollution sources include chipboard and plywood glues, foam house insulation,
soft furnishings, paper, cosmetics, smoking and gas fires (see references in Dingle et al., 2000).
Wolverton et al. (1984) tested the ability of golden pothos (Scindapsus aureus), the arrowhead
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

plant (nephthytis) S. podophyllum and spider plants Chlorophytum elatum var. vittatum to
remove formaldehyde from sealed plexiglass chambers; experimental controls using pots
with soil, but without plants, were used to examine the impact of soil microorganisms on
formaldehyde removal. In these experiments the controls and experimental plants removed
formaldehyde. The soil-only control reduced formaldehyde from a starting concentration
of 15 to 10  ppm over a 24  h period whilst golden pothos and nephthytis plants reduced

TABLE 3.4  Indoor plants graded as ‘Excellent’ (removing 1.2 μg.m-3.cm-2 of leaf area over a
5 h period) by Kim et al. (2010) in removing formaldehyde

Type Scientific name Common name μg.m-3.cm-2

Woody foliage plants
Psidium guajava Guava 2.39
Rhapis excelsa Lady palm 1.67
Zamia pumila Coontie 1.32
Herbaceous foliage plants
Chlorophytum bichetii Airplane plant/Siam lily 1.25
Dieffenbachia amoena Dumbcane 1.24
Tilandsia cyanea Pink quill 1.23
Anthurium andraeanum Flamingo flower/Tail flower 1.22
Korean native plants
Nandina domestica Heavenly/sacred bamboo 1.58
Dendropanax morbifera Korean dendropanax 1.50
Ardisia crenata Coralberry 1.46
Laurus nobilis Sweet bay 1.40
Ferns
Osmunda japonica Japanese royal fern 6.64
Selaginella tamariscina Resurrection fern 4.84
Davallia mariesii Squirrel’s foot fern 4.15
Polypodium formosanum Caterpillar/grub fern 3.62
Pteris dispar Amakusa fern 1.95
Pteris multifida Spider brake 1.92
Microlepia strigosa Lace fern 1.49
Botrychium ternatum Winter flowering warabi 1.42
Herbs
Lavendula spp. Lavender 2.12
Pelargonium spp. Geranium 1.87
 Indoors  85

it from 18 to 6 ppm over the same period. In initial tests with a starting concentration of
14 ppm, the spider plant was so effective that a second set of experiments was conducted
with a higher starting concentration of 37 ppm; after 24  h the concentration was below
2 ppm. Whilst these values are impressive, Wolverton et al. (1984) estimated that a typical
418 m3 energy-efficient home would need some 70 spider plants to purify the air of the
7,000–8,000 μg of formaldehyde a day generated by the building (a figure that includes an
estimate of the formaldehyde released by cooking). Wolverton et al. (1989a) subsequently
showed that corn/mass cane Dracaena massangeana, ficus F. benjamina, gerbera daisy Gerbera
jamesonii, pot mums Chrysanthemum morifolium and the striped dracaena Dracaena deremensis
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

‘Warneckei’ could all reduce formaldehyde concentrations in the air of sealed chambers.
Wolverton and Wolverton (1993) subsequently published data on 33 plant species for their
ability to remove formaldehyde (Table 3.3); the best, the fern N. exaltata ‘Bostoniensis’, was
shown to remove the pollutant at the rate of 1,863 μg.h-1 and the worst, the succulent Aloe
barbadensi (Aloe vera), ten times less at 188 μg.h-1.
Kim et al. (2010) examined 86 species of plant for their ability to remove formaldehyde
and found a range of plants which were ‘excellent’ (Table 3.4) – defined as removing at least
1.2 μg.m-3 per cm2 of leaf area over a 5 h period. As a group, ferns were best at formal-
dehyde removal, but herbaceous foliage plants were also effective. As different researchers
tend to use different methodologies, it is difficult to make comparisons between them,
there tend to be few species in common, and very few assay large numbers of species. It
is almost certainly coincidental that the 10th ‘best’ species identified by Kim et al. (2010),
R. excelsia, is also the 10th ‘best’ species of 33 identified by Wolverton and Wolverton
(1993) and Anthurium andraeanum 21st in the list of Kim et al. (2010) and 26th in that of
Wolverton and Wolverton (1993). However, it is interesting that, as with Kim et al. (2010),
Wolverton and Wolverton (1993) found their most effective formaldehyde remover was the
Boston fern N. exaltata. Using data from the US Environmental Protection Agency (EPA)
for formaldehyde emissions in new office buildings (0.173 μg.l-1) Wolverton and Wolverton
(1993) estimated that a 22.32 m3 office with a floor area of 9.3 m2 would contain about
3,916  μg of formaldehyde which could be removed by two Boston fern plants or three
‘Janet Craig’ dracaenas.
Portable buildings are particularly susceptible to formaldehyde pollution and Dingle et al.
(2000) examined the potential of plants to remove the pollutant in five such structures in
Perth, Australia. Five plants (one from each of the following species: spider plant C. comosum,
fig tree Ficus sp., cast-iron plant Aspidistra elatior, dumb cane Dieffenbachia amoena and arum
ivy Epipremnum aureum) were placed as a group in each building. An additional group of five
plants was added every second day until 20 plants had been added. Formaldehyde levels were
monitored throughout the experiment and showed that plants did not reduce levels signifi-
cantly from the 856  ppb (parts per billion) starting concentration without plants until the
maximum of 20 plants had been added, when an 11% reduction was recorded. At this time
the average concentration of formaldehyde was 761 ppb, which was still 661 ppb over the
occupational guideline exposure level for non-industrial buildings. The number of plants to
achieve this reduction was equivalent to one per 1 m3 and reduction below this level would
clearly require an impractical number of plants on horizontal surfaces, although indoor
green walls may be of considerable value in such contexts. Dingle et al. (2000) concluded
that plants were not an effective way of reducing formaldehyde in the air of portable office
buildings. The experiment and interpretation suffered from a number of limitations: 1) the
 86 Indoors

short duration of the experiment whereby it was assumed that a two-day period would be
sufficient for a group of plants to exert a noticeable impact on a large volume of air with
continual outgasing; 2) the assumption that all plants were equal in their formaldehyde
extraction efficiency; and 3) that in a non-portable office building or home situation a
reduction of 95  ppb might be considered a useful impact, especially as their study of 18
conventional offices showed background formaldehyde concentrations in the range 10 to
78 ppb (mean = 22 ppb).
Xu et al. (2010) examined the effect of a biofilter on formaldehyde removal from air.
The filter was composed of compost, vermiculite powder and ceramic particles and formal-
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

dehyde-laden air entered from below the cylindrical filter and exited at the top. The filter
was operated with and without a spider plant grown in the filter medium. The presence of
the spider plant changed the profile of accumulation of formaldehyde in the biofilter and
Xu et al. (2010) suggested that this was because the spider plant was actively removing the
chemical via its root system and also that the microbial community surrounding the roots
may be involved in formaldehyde removal. The activity was later confirmed by Xu et al.
(2011). Aydogan and Montoya (2011) examined four species of plant for their ability to
remove formaldehyde and showed that the root zone was more effective than the leaves.
Kim et al. (2008) showed that the ratio of formaldehyde removal in percentage terms by the
aerial parts compared with the root zone for F. japonica was 61:39 during the day but 2:98
at night, and for F. benjamina the ratios were 43:57 during the day and 6:94 at night. Very
little (about 10%) of the formaldehyde removed by the root zone was considered to have
been removed by the potting medium – the majority was via microorganisms. Aydogan and
Montoya (2011) also demonstrated that some plants were better than others in their VOC
removal efficiency. Schaffner et al. (2002) showed that plants can detoxify formaldehyde
directly via enzymatic action (in this case via a ‘glutathione-dependent formaldehyde
dehydrogenase’).

3.7.5  The real world and VOC removal

The majority of work on VOCs has been in laboratory situations; studies such as Dingle
et al. (2000) using buildings are relatively rare and even their study was atypical as most
people do not work or live in portable buildings. Wood et al. (2006) examined total VOC
(TVOC) removal by two species of plant – D. deremensis ‘Janet Craig’ and the peace lily
Spathiphyllum ’Sweet Chico’ – in 30–50 m3 staff offices in three buildings at the University
of Technology in Sydney, Australia. The VOCs present in buildings used in the study
were identified as: acetone, dodecane, ethanol, ethylbenzene, limonene, methylbenzenes,
methylbutane, methylecyclopentane, 2-methylpantane, n-decane, n-hexane, n-pentane,
toluene and xylenes. In their first experiment, air-conditioned and non-air-conditioned
offices were provided with either three or six plants (plants were individually potted in
300 mm diameter pots) of D. deremensis and compared with offices without plants; weekly
measurements were taken of TVOC concentrations for nine weeks. The treatments (0, 3,
6 plants/office) were then randomly reallocated and measurements taken for a further nine
weeks. The results showed that indoor TVOC levels were far higher than outdoors. When
all offices were analysed together there was an obvious trend in TVOC reduction with
increasing number of plants, although the difference was not statistically significant. When
only offices with initially very high concentrations (>100 ppb) were compared, it was clear
 Indoors  87

TABLE 3.5  Impact of Dracaena deremensis plants on total VOC levels in Australian university staff
offices

Treatment All offices High VOC offices

TVOC (ppb) TVOC (ppb)
Outside 150±15 150±15
Inside 0 plants 110 ±15 190±40
Inside 3 plants 180±7 105±15
Inside 6 plants 180±7 100±10
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Source: Wood et al., 2006.

that substantial reductions in VOCs had occurred in offices with plants (Table 3.5) and that
there appears to be a ‘trigger’ concentration. The scale of TVOC removal depended on
whether the offices were in an air-conditioned building or in a naturally ventilated building.
TVOC reductions were evident in the air-conditioned building when plants were present,
but the reductions were not statistically significant. In the naturally ventilated building, the
impact of plants on TVOC levels was substantial, and statistically significant, with reductions
of the order of 75% whether three or six plants were used (decline from 280 ± 120 ppb to
65 ± 10 ppb).
In their second experiment, Wood et al. (2006) used smaller Dracaena plants and a
mixture of the latter with peace lilies. The design was simpler with either six plants (in
individual 200 mm diameter pots: five peace lilies + one Dracaena) per office or no plants.
Subsequent air-quality monitoring showed some interesting deviations from the first
experiment. The air-conditioned building had higher indoor VOCs than outside, but the
naturally ventilated one did not have levels significantly different from outside. When all
offices were compared in the air-conditioned building, TVOC levels did not differ with
or without plants – probably because the peace lilies were producing their own VOCs as
they were flowering. However, when TVOC concentrations exceeded the threshold of
100 ppb in control offices, offices with plants had 70% lower TVOC levels. In the naturally
ventilated building TVOC levels never exceeded the threshold for TVOC reduction. Both
experiments showed that plants could remove VOCs under realistic office conditions and
that the VOC removal was triggered by exposure to a threshold level of VOCs which then
reduced to below the trigger threshold; the latter was later confirmed in laboratory studies
by Orwell et al. (2006).

3.7.6  Dust and particulates

The negative impact of very fine dust particles from diesel vehicles and other combustion
sources is well known (see Chapter 2) but exposure to particulate matter indoors results
from external sources as well as internal ones such as cooking and fires (Hänninen et al.,
2011). The infiltration of particulates from outside to inside varies seasonally, with summer
having the highest loads with close to twice the concentration of winter, which experiences
the lowest levels; spring/autumn have particulate infiltration in between the two extremes
– probably as a result of the temperature-related incidence of window opening in naturally
ventilated buildings (Hänninen et al., 2011). Lohr and Pearson-Mims (1996) found that dust
 88 Indoors

loads on horizontal surfaces were reduced when plants were used in indoor environments.
House dust has also been shown to contain high molecular weight phthalates used as plasti-
cisers; diisononyl phthalate and diisodecyl phthalate have been tentatively linked with asthma
in a study of Norwegian children (Bertelsen et al., 2013). House dust has been shown by
Rudel et al. (2003) to carry at least 66 endocrine disrupting chemicals (EDCs) including
penta- and tetrabrominated diphenyl ethers used as fire retardants and a banned carcinogen
fire-retardant precursor 2,3-dibromo-1-propanol, 27 pesticidal compounds and synergists.
Fulong and Espino (2013) showed that decabromodiphenyl ether – a fire retardant used in a
range of electrical and electronic devices, building materials and fabrics – was present in dust
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

taken from various indoor sites at the Diliman Campus of the University of the Philippines.
The compound has been found in indoor dust in studies in Belgium, Canada, China, Hong
Kong, Japan, the Philippines, Portugal, the UK, USA and Thailand (see references in Fulong
and Espino, 2013). EDCs are compounds that affect hormonal functions, including sexual
behaviour, and the development of the male and female reproductive systems; they are also
implicated in the initiation of various cancers (Crisp et al., 1998; Diamanti-Kandarakis et al.,
2009). Particulate sampling by Lohr and Pearson-Mims (1996) was fairly crude, using 60 mm
diameter aluminium weighing dishes to capture air-deposited dust. Each collecting dish was
placed inside an open container 70 mm in diameter by 45 mm deep to reduce dust remobi-
lisation and accidental spillage and left for one week before dust loads were estimated. Plants
used were primarily smooth-leaved and included Chinese evergreen Aglaonema sp., bamboo
palm Chamaedorea seifrizii, dragon tree D. marginata, golden pothos E. aureum and peace lily
Spathiphyllum sp. In a computer lab, with 2% of the room filled with plants, dust load was
significantly reduced by 15% by the presence of plants. In a second experiment, an office
environment, plants were located around the edge of a room and took up 5% of the space;
dust was collected as before (except the surrounding cans were omitted) and was significantly
reduced by 21%. Whilst Lohr and Pearson-Mims (1996) did not characterise the particulate
size ranges captured in their study, it has been shown by Hänninen et al. (2011) that indoor
air can include the dangerous PM10 and PM2.5 particulate fractions and indoor plants are
likely to assist in their removal.

3.8  Indoor green walls

Green walls inside buildings can be considered as part of a building’s air-conditioning
system, helping to cool the air and reduce the energy requirement for conventional
air-conditioning systems (Franco et al., 2012), remove air pollutants, and reduce noise and
potentially dust and microbial loads (see above). Indoor green walls can be constructed
using any of the approaches used for external green walls (Chapter 5), but most are of the
‘living wall’ type using a felt mat or modular approaches that hold plants and rooting media;
water and nutrients are typically delivered using an irrigation system (see Chapter 5). The
latter requires electrical and water connections. Variants have been developed that are free-
standing with water troughs to capture and recycle irrigation water to permit more flexible
placing, but still require an electrical connection (Figure 3.6). However, new designs are
constantly emerging such as self-contained free-standing room dividers and wall modules
that incorporate water tanks and use capillary action, avoiding the need for irrigation pumps
or permanent piped water (Figure 3.6). Small systems such as LivePicture need re-watering
only once every four to six weeks. Such passive systems have obvious aesthetic value as well
 Indoors  89

as providing some air conditioning and pollutant removal services (section 3.7). More active
systems have also been developed which use the walls as biofilters to remove pollutants
and to cool and condition the air; such systems promote air flow through the green walls
(Franco et al., 2012).
Passive living wall systems are typically attached to structural and internal walls, and damp
penetration from the green wall to the building walls is prevented through the use of water-
proof membranes. In the case of active green walls, this waterproofing is offset to create a
route for the recirculation of the processed air. Typically such internal green walls have the
warmer air from the building forced through the rear of the growing media (through the root
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

zone first, then the leaf zone), so conditioned air emerges from the wall into the building
envelope (Figure 3.7) (Franco et al., 2012). External green walls can also be used as active
air-conditioning systems. Franco et al. (2012) showed that the choice of growing medium
in living walls had an impact on the amount of water retained, with polyurethane retaining
most in their comparison test with polyamide-polypropylene and polyester; the latter held
least water and had a more hydrophobic character. As might be expected, resistance to airflow
through the substrate was increased by the presence of plants, by the level of saturation
with water and by higher windspeeds. The polyurethane substrate had the highest airflow
– attributed to its more porous nature; with vegetation the overall airflow rates were (highest–
lowest): polyurethane>polyester>polyamide-polypropylene. The crucial feature of the system
is, obviously, how saturated the air is when it leaves the green wall: fast-moving air picks up
less water, so the relationship between saturation of the growing medium and its porosity
combined with the airflow-reducing impact of vegetation are all important contributing
factors to the eventual air saturation. In their tests, Franco et al. (2012) considered air speeds
of 0.25–0.5 m.s-1 to be optimum, and in that range found that polyamide-polypropylene was
the best at promoting air saturation (whilst also being the most frugal in water consumption)
followed by polyurethane and finally polyester. Whilst the work of Franco et al. (2012) is
helpful, the results may not reflect reality after the walls have been established for some time.
Their test sections of substrate were planted up with only one species of plant (Scindapus
aureus), covering only 50% of the test area, and no details were given of the length of time
the plants had been left to grow in situ. Green walls are typically grown with a mixture of
plant species to add a range of visual textures (growth form, leaf morphology) and colours;
this variability, combined with additional leaf and root-zone growth, may strongly impact on
water retention and pressure drop.
In a later study, Cañero et al. (2012) examined the performance of an approximately
8 m2 living wall installed in the hall of a building at the University of Seville (195.36 m3).
Four rooting substrates (two organic: coconut fibre and ‘Xaxim’ composed of fern root;
two synthetic: ‘Epiweb’ (polyetylentereftalat) and a geotextile (acrylic textile fibre mix on a
polypropylene base)) were used with the plants inserted into growing ‘pockets’. The wall was
planted-up with 24 indoor-tolerant species in 2008 and evaluated in 2009 when the plants
had established. Plants were maintained using built-in irrigation and additional lighting was
provided. Over a 15-week study period, the wall moderated internal temperatures, reducing
them, on average, by 4°C and increasing humidity by 15%. The thermal performance of the
substrates was similar, with the geotextile giving the best result. The mean room temperature
at the far end of the hall, some 11 m away, was the control (27.1°C); the average air tempera-
tures 30  cm from substrates was 23.8°C for the geotextile, 24.1°C for both Xaxim and
coconut fibre and 24.2°C for the Epiweb. The estimated energy saving on air conditioning
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

90 Indoors
 Indoors  91
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 3.6  Indoor green walls. Top left and right: before and after installation of Mobilane
LivePicture wall modules at Staffordshire University’s Science Centre in Stoke-on-Trent, UK.
Middle row left: installation requires only three bolts; middle right: installing a pre-planted
cassette into LivePicture; bottom left: indoor living wall showing water reservoir at base (Bin
Fen system); bottom right: free-standing room divider (Mobilane); above: LivePicture planted
up with arums for colour. © John Dover

was 20%, although it was not clear how far the cooling effect extended from the living
wall. The organic media appeared better for plant growth than the synthetics and Xaxim,
being composed of fern root, released spores which germinated on its surface. However, the
durability of most of the media appeared similar with the exception of coconut fibre, which,
despite being reinforced, was clearly degrading. The synthetic media retained less water and
dried faster than the organics, leading to higher water consumption and irrigation require-
ments. The authors note that although Xaxim performed well, it is not recommended due
to the endangered nature of the species used in its construction.
 92 Indoors
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

(a) (b)

FIGURE 3.7  Airflow schematics for active living walls where a) walls are located inside the
building and indoor air is recirculated through the living wall, and b) walls are located outside
the building and air is drawn from the outside of the building, through the living wall and into
the building. Redrawn from Franco et al. (2012).

3.9 Summary
• Vegetation in buildings has a positive benefit on psychological well-being and has a
positive effect on work activities.
• Vegetation in buildings has the potential to remove harmful air pollutants – in particular
VOCs that contribute to ‘sick-building syndrome’.
• There are indications that dust and microbial loads in indoor air may be reduced by
plants – but more work is needed.
• Plants, typically in the form of living walls, can be incorporated into the air-conditioning
systems of buildings, actively modifying temperature, humidity and pollutant levels.
• Novel wall systems have been developed that allow substantial numbers of plants to be
grown in offices and workplaces which would otherwise take up large areas of desk or
floorspace.
 4
PERMEABLE PAVEMENTS
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

This chapter will:

• introduce the concept of permeable pavements

• describe the impacts that permeability has on stormwater runoff
• examine the effects of the subsurface microbiology of permeable pavements on pollutants
• examine the way permeable pavements are being used to modify various common urban
elements including alleys, parking areas and roads.

4.1  Introduction to permeable pavements

If a ground surface is made of impermeable material such as concrete or asphalt, the amount
of rainwater deposited on it that is diverted to drains can be around 95% (Schuler, 1994) with
the remainder presumably evaporating from the wet surface. In contrast, with a vegetated
surface, runoff can be very low, around 10%, with a high proportion being evapotranspired
by plants, and the remaining water taken up by the soil (Arnold & Gibbons, 1996). Schuler
(1994) gave a specific example of a meadow considered to be 1% impervious with a runoff
of 6%. Different proportions of vegetation cover and impermeable surface result in different
levels of infiltration (Arnold & Gibbons, 1996) (Table 4.1).

TABLE 4.1  Effect of different proportions of sealed surface on runoff, evapotranspiration and infiltration

Source Land cover Runoff (%) Evapotranspiration (%) Infiltration (%)

1 Natural vegetation 10 40 50
2 10–20% sealed surface 20 38 42
3 35–50% sealed surface 30 35 35
4 75–100% sealed surface 55 30 15
5 100% 95 not given not given

Sources: (1–4) data from a figure in EPA (1993) cited in Arnold and Gibbons (1996), and (5) Schuler (1994).
 94  Permeable pavements

TABLE 4.2  Proportion of surface covered by impervious material depending on land use and area
(data from USA)

Area of plot Sealed surface (%)

Land use (ha) (acre)

Residential 0.809 2.00 12
0.405 1.00 20
0.202 0.50 25
0.135 0.33 30
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

0.101 0.25 38
0.051 0.13 65
Industrial 72
Commercial and business 85
*Shopping centres 95

Sources: Cronshey (1986) and *Arnold & Gibbons (1996)

The quality of a running water body is related to the proportion of its catchment that is
composed of impermeable surfacing. Arnold and Gibbons (1996) indicate that it takes as little
as 10% sealed surface for stream health to become ‘impacted’, with ‘degraded’ status after
only 30% of a watershed has become sealed. So the use of a permeable material to replace
impermeable surfaces should not only improve infiltration, and reduce runoff to sewers and
drains, but also improve urban water quality including reducing impacts on aquatic macroin-
vertebrates and fish (Schuler, 1994). Contaminants, such as heavy metals (e.g. copper and
zinc) and polyaromatic hydrocarbons (PAHs), have been shown to be substantially reduced
in water percolating through permeable pavements compared to runoff from sealed surfaces
(Boving et al., 2008; Brattebo & Booth, 2003). Of course, the degree to which an area is
sealed depends on its function (Table 4.2), with the more intensive non-residential uses
(industrial, commercial, business and shopping centres) having the highest levels of sealed
surface and residential areas becoming increasingly sealed with decreasing curtilage (plot)
area.
Whilst the basic concept of increasing surface permeability by substituting sealed
surfaces with permeable ones (to promote infiltration of rain in situ) is a simple one, there
is a range of approaches that can be employed (Scholtz & Grabowiecki, 2007). These
include:

• loose gravel: cheap and simple, but with a tendency to scatter and develop ridges –
especially if the base underneath the gravel surface is not adequate to bear loads;
• permeable asphalt: which looks very similar to ordinary tarmac, except when water is
thrown onto it – with one you get puddles, with the other you don’t as the water drains
through pores in the surface;
• porous concrete: no fine aggregates are used so gaps between stones bound by cement
provide drainage pathways;
• permeable blocks: these look like normal blocks used in hard landscaping but rain can drain
through pores in the material and/or around the block edges;
• wheel tracks: where only the area driven on is a hard surface and the rest is planted-up;
• open concrete blocks or plastic blocks: here the load-bearing surface is covered with open
 Permeable pavements  95
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 4.1  Top: open concrete blocks seeded with grass in the car park of the cable car station
at Fuente De, Picos de Europa, Spain; bottom left: interlocking plastic grid system; bottom
right different grades and styles of turf reinforcement mesh. © John Dover

blocking into which gravel or a plant-growing medium is inserted – such areas are often
planted up with grass (Figure 4.1);
• reinforcing mesh: a variation on the open concrete/plastic blocks approach used on grassed
surfaces or to retain gravel. A wide range of semi-rigid plastic mesh is available. Such
materials when used with grass are often coloured green so that if exposed by hard wear
they do not look unsightly (Figure 4.1).

The range of colours, types and shapes of surfacing materials is very wide, so there is no real
reason why the majority of hard landscaped surfaces should not now be constructed with
a permeable action. Of course the surface type, colour, etc. will be chosen to complement
the context, but the longevity and success of the installation will depend on the design of
the sub-base, which will in turn depend partly on the surface type, the soil type (clays cause
problems as they are impermeable), the load to be supported and the frequency of use. The
intensity of rainfall may also determine which system, if any, is most appropriate; Brattebo and
Booth (2003) demonstrated extremely good performance of permeable pavement systems in
the Pacific Northwest, where their peak rainfall intensity was 7.4 mm.h-1, but noted that their
experience might not be directly applicable to areas with higher rainfall intensities.
 96  Permeable pavements

The basic concept of the permeable pavement is capable of further refinement; for example,
Tota-Maharaj et al. (2010) have experimented with combining permeable pavements with
geothermal heat pumps – an approach which seeks to gain additional benefits by recycling
the water for non-potable use (e.g. toilet flushing, car washing, garden watering).
Scholtz and Grabowiecki (2007) reviewed the literature on permeable pavement systems
and noted that they have the following attributes, which are superior to using roadside
drains/gully systems:

• better collection and infiltration of rainfall and stormwater runoff

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

• evaporation of water stored in pores

• recharge of groundwater
• prevention of pollution (removal of heavy metals and hydrocarbons)
• water recycling
• reduction of Biochemical Oxygen Demand (BOD), Chemical Oxygen Demand
(COD), and ammonia levels, in runoff
• reduction in suspended solids
• low maintenance.

As permeable pavements are essentially a modification of surfaces that would normally be

impermeable, they can be incorporated in areas where space concerns would mean that
additional structures such as swales, used to transfer water to storage basins, could not be
used (Starke et al., 2010).
The sub-base of a permeable pavement is typically made up of graded stones, and the air
space between them allows water to freely drain (and also acts as a water store during rainfall
(Fassman & Blackbourn, 2010)). Sub-bases are usually at least 200 mm deep and are placed
on top of a geotextile membrane (such as Terram® www.terram.com/) that is laid on the
soil surface to help spread the load (Figure 4.2).
This basic design (DCLG, 2008) will need to be modified depending on the properties of
the installation site; for example, if placed on clay subsurface, drainage pipes may need to be

Porous  surface  

Graded  sub-­‐base  

Geotex1le  

Soil  

FIGURE 4.2  Schematic cross-section of a permeable driveway, pavement or parking area; there
may also be a geotextile above the sub-base (DCLG, 2008; Fassman & Blackbourn, 2010).
 Permeable pavements  97

incorporated to take the water away to a soakaway, rain garden or pond. In more industrial
situations, where there is the potential for pollution, the water may need to be isolated from
the soil by an impermeable membrane and drainage pipes laid to direct the water to a water
treatment area, which may be, for example, a reed bed, before final infiltration or ponding.
Technical specifications have been drawn up for performance of permeable systems; for
example, German regulations specify the hydraulic conductivity of such pavements (Starke
et al., 2010).
There are some reports of maintenance and construction issues with respect to permeable
pavement systems (see below). However, a study of four different systems, and a conventional
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

sealed asphalt surface, installed at the car park at the King County Public Works in Renton,
Washington (USA) demonstrated continued permeable function after six years of constant
use (Brattebo & Booth, 2003). The four systems used were as follows (primary source:
Brattebo and Booth, 2003):

• Two plastic grid systems – one with sand in the grid cells on which grass was grown
(Grasspave2®), the other composed of the same cell type but filled with gravel
(Gravelpave2®). The plastic grid provides good load-spreading action with very low
impermeability (see ISI, 2014).
• A concrete ‘honeycomb’ system with topsoil and grass in the void (Turfstone®, see
Tobermore, 2014), with about 60% impermeability.
• A block paving system (UNI Eco-stone®), 88% impermeable, with the between-block
gaps filled with gravel (UNI-Group, 2014).

Visual comparison of the surface of the permeable pavement systems and the asphalt control
after six years showed minimal wear and tear, with only a couple of incidences where the
Grass-/Gravelpave2® systems had shifted. Water runoff from the asphalt control was rapid
and followed the rainfall patterns; in the permeable systems almost all water was infiltrated
from the same rainfall events. The only water runoff from the permeable systems that could
not be attributed to leaks in water conduits probably came from localised saturation of the
Grasspave2® system where heavy rain ran off the parked cars and an incidence with the same
system where 121 mm of rain fell in a 72 h period resulting in 4 mm of runoff. This storm
was the most extreme event of the six-year study and runoff, as a percentage of the total
rainfall event, was a mere 3%. The permeable pavement systems also delayed the peak flow
by about 1 h (Brattebo & Booth, 2003).
Permeable pavements which are composed of porous blocks or porous asphalts have been
reported as having surface clogging (including at block joints) by sediments and as a result
of mechanical action breaking up the pavement surface (this latter where there is frequent
heavy traffic). Clogging occurs typically around three years after installation (Scholtz &
Grabowiecki, 2007), which can reduce efficiency and result in additional servicing costs
to open the surface up again. Grass-based systems may get clogging of the geotextile
membrane. Brattebo and Booth (2003) noted that Grasspave2® and Gravelpave2® systems
might require some maintenance if exposed to more frequent vehicle movements than
typical in their six-year study of a works car park. Boving et al. (2008) found that the
geotextile under the porous pavement they were studying appeared to impede infiltration.
The need for maintenance of permeable pavements may affect the decision to install such a
system in the first place; factors promoting clogging appear to be the proximity to very fine
 98  Permeable pavements

soil, catchment characteristics (including leaf fall), and fine particles from adjacent bitumen
asphalt (Fassman & Blackbourn, 2010). Boving et al. (2008) found that material brought
into a permeably paved car park from the road system on vehicles contributed to clogging,
especially sand.

4.2  Permeable pavements and street trees

Permeable pavements might be expected to have benefits for street and other trees over
impermeable surfaces because the latter are known to provide poor growth conditions.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

However, recent work by Viswanathan et al. (2011) comparing permeable and imper-
meable concrete systems on root growth by the American sweetgum (Liquidambar styraciflua)
found no benefit. Both systems resulted in less root growth than a soil-surfaced control,
probably due to high CO2 soil concentrations. However, as the experiment was performed
on a low permeability, clay-based soil, the response may not be generaliseable to all soils.
Given that the soil surface around street trees is frequently capped with a resin-bonded
aggregate, information on the impact of such systems is essential (for more information see
section 7.11).

4.3  Microbiology of permeable pavements

4.3.1  Hydrocarbon management (especially in car parks)

One of the major issues in dealing with permeable pavements in areas such as car parks is
the potential for pollution by hydrocarbons from lubricating oil reaching the soil beneath the
pavement and potentially contaminating aquatic systems. Fortuitously, it appears that a hydro-
carbon-utilising microbial community can develop on the geotextile mat below the sub-base
(Pratt et al., 1999). The hydrocarbons are metabolised into sugars and act as a carbon source
for the community which, in composition, can be extremely diverse consisting of bacteria,
fungi and bacteria-eating protozoa (Coupe et al., 2003). Oil conversion efficiency can be
extremely high; Coupe et al. (2003) suggest up to 98.7%. The degrading community does
require aerobic conditions however (Newman et al., 2006; Pratt et al., 1999), so maintaining
good air spaces in the sub-base is probably critical and may be facilitated by the protozoa
in the oil degrading community (see refs in Coupe et al., 2003). Whilst new permeable
pavements can be inoculated with a commercial starter culture, over time the microbial
community changes from that in the inoculum and it appears that after about 39 weeks
an un-inoculated and inoculated pavement are equally good at degrading hydrocarbons (at
least under laboratory conditions) (Newman et al., 2006, 2002). After five weeks protozoan
communities, whilst initially richer for the first four weeks in inoculated systems, became
richer in un-inoculated systems (Coupe et al., 2003). The oil-degrading community may
need inputs of nitrogen and phosphorus to function (Newman et al., 2006), and fertilisers are
often added, but Rushton (2001) in her study of a car park at the Florida Aquarium in Tampa
showed that rainfall was a good source of nitrogen (but not phosphorus). Tota-Maharaj and
Scholz (2010, 2013) pointed out that dog faecal material would be a likely component of
‘real-world’ water entering permeable pavements and that this would carry with it a range of
nutrients. Their experimental permeable pavement systems substantially reduced Biochemical
Oxygen Demand by 99%, ammonium-nitrogen by 97% and orthophosphate-phosphate by
 Permeable pavements  99

95% (Tota-Maharaj et al., 2010). Infiltrated water from permeable systems compared with
runoff from a sealed asphalt surface showed significantly lower levels of lubricating oil in a
six-year car park study (Brattebo & Booth, 2003). Newman et al. (2004) noted that different
geotextiles perform differently in their efficiency of oil retention, and also that membrane
type becomes less important over time.

4.3.2  Microbiology of permeable pavements: pathogens

Whilst the ability to host a hydrocarbon-degrading microbial community is an extremely
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

positive aspect of permeable pavements, Scholtz and Grabowiecki (2009) articulated fears
that the real-world contaminants in water which infiltrated into permeable pavements, such
as dog faeces, might lead to contamination and multiplication of bacteria associated with
public health risks – especially if the water was recycled for general non-potable use. The
risk of microbial contamination in permeable pavements when combined with a geothermal
heat pump system might be considered more risky than conventional permeable pavements
as the former would have warmer subsurface conditions than the latter, promoting microbial
growth. Their experiments (using plastic ‘wheelie bins’) thus used material sourced from
gully pots (the ‘trap’ part of a drain which holds water from, for example, road runoff),
which would contain a range of contaminants, mixed with de-chlorinated tap water and
fresh dog faeces. Whilst potentially pathogenic organisms were recovered from the permeable
pavements, the conclusions were not particularly helpful in determining real risk as organisms
were not identified to species. The experiment was continued and refined, however, using
more advanced techniques of microbial identification using DNA analysis (Tota-Maharaj et
al., 2010). Standard bacterial analyses from culturing inflow (‘runoff’) and outflow (infiltrate)
on specialist agars (for details, see Tota-Maharaj et al., 2010) showed very high reductions in
bacterial contamination, especially for Escherichia coli. The more exact DNA analyses carried
out on some bacterial colonies derived from infiltrate did not contain the classic human
health-related species that might have been expected with dog faecal contamination of water
(Salmonella, E. coli, faecal Streptococci) and Legionella – this latter was assayed due to the warm
nature of the system Tota-Maharaj et al. were investigating which combined a permeable
pavement with a geothermal heat pump (Tota-Maharaj et al., 2010). Tota-Maharaj and Scholz
(2010) showed that, on average, the permeable pavement system removed 98.6% of total
coliforms, E. coli and faecal Streptococci. Despite these impressive results, Tota-Maharaj and
Scholz (2013) noted that potential adopters of permeable pavement systems used to process
and recycle water for non-potable domestic use might be sufficiently risk averse to consider
the reductions they had previously demonstrated insufficient. They therefore experimented
with a disinfection phase whereby remaining microbial contamination could be removed and
demonstrated that a combination of suspended titanium oxide powder (TiO2) and UV light
could completely remove total coliforms, E. coli and faecal Streptococci after 80–100 minutes
of exposure.

4.4  Permeable pavements and the heat island effect

As Starke et al. (2010) note, the evapotranspiration component of vegetation in urban areas
not only contributes to the reduction in runoff, but also has a cooling effect. The opportu-
nities for evaporation of water from permeable pavements (and transpiration from any plants
 100 Permeable pavements

Surface Temperature (°C) 60

50

40

30 grass
ceramic
20 block

10 asphalt
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

0
00:00 06:00 12:00 18:00
Time of day

FIGURE 4.3 Surface temperatures in °C of different ground cover materials: grass, porous
ceramic, porous block, impermeable asphalt on a warm summer’s day in Tokyo with air
temperature of 35°C at 12:00. Re-drawn from data in Figure 5 of Asaeda and Ca (2000).

incorporated in some designs) has the potential to reduce their surface temperature compared
to similar sealed surfaces and hence reduce the amount of heat stored during the day and
subsequently released at night. As a result, permeable pavements have the potential to reduce
the heat island effect.
Asaeda and Ca (2000) explored the heat-reducing potential of permeable pavements in
Tokyo in 1994 and 1995 using ten 2 x 2 m plots composed of a number of materials including
a porous block pavement, normal (impermeable) asphalt, a permeable ceramic and grass. The
porous blocking had large pores that water drained through rapidly, whilst the ceramic had
a range of pore sizes, some of which were very fine, and retained water within its structure
for long periods (days). They recorded surface and below-ground temperatures (and other
parameters) but here we examine just the surface temperature data. The data presented were
from a 30 h period on 9–10 August 1994 (Figure 4.3). The peak air temperature was 35°C at
12:00. At the same time-point the porous block was over 48°C, similar to that of the normal
asphalt which was just in excess of 50°C; the ceramic and grass were nearly indistinguishable
with a surface temperature of about 43°C (Figure 4.3).
The take-home message from this study was that grass was always the coolest surface, but
that a porous structure that retains some water for evaporative cooling comes a pretty close
second and is sometimes as cool as grass. Impermeable surfaces and porous structures that
do not retain water have similar thermal properties and are warmer. Asaeda and Ca (2000)
suggested that the good performance of the ceramic material might also be, in part, because
some of the pores were fine enough to act as capillaries taking up water from the substrate
below it and enhancing its evaporative potential.
Starke et al. (2010) compared evaporation from two concrete block systems – one
permeable, the other impermeable – and found that overall evaporation was 16% greater
using the permeable system, and that the evaporation took place over several days compared
with the relatively fast evaporation of the small amount of water on the impermeable surface.
In the former, the local air quality was moderated over a longer period of time than the
latter.
 Permeable pavements  101

At the other end of the temperature scale, a study in Lulea, northern Sweden, demon-
strated that permeable pavements are less prone to freezing, and thaw more quickly, than
impervious pavements (Bäckström, 2000). However, the capillary action that seems to work
so well in surface temperature reduction work by Asaeda and Ca (2000) may not be as helpful
as it may seem when used under cold conditions, as capillarity is deliberately avoided in road
surfacing to prevent damage through freeze–thaw action (Starke et al., 2010).

4.5  Front gardens

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Perry and Nawaz (2008) used aerial photographs to track the change in surface sealing of
domestic gardens in a 1.16 km2 area of Leeds in the UK. They found a 13% increase over a
33-year study period (1971–2004), three-quarters of which was attributable to the paving-
over of front gardens, presumably for off-road car parking. Smith et al. (2011b) showed that
front gardens in Greater London made up 25% (9,400 ha) of the area of garden surrounding
houses compared with 63% for back gardens from a total estimated area of 37,900 ha (24% of
Greater London). Of the front garden area, 63% (5,900 ha) was estimated to be ‘hard surfaced’
in 2006–2008 and is probably primarily impermeable (sealed) surfacing; this contrasts with
an estimated 4,200 ha (45%) in 1998–1999.
On 1 October 2008 a new amendment to the Town and Country Planning (General
Permitted Development) Order 1995 came into force in the UK (DCLG, 2008), as a result
of the trend of converting the front gardens of houses into parking areas for cars and the effect
such sealed surfaces had in increasing the pressure on sewer systems – with increased potential
for flooding and the transport of car-related pollutants.
Before this change property owners could convert gardens to parking areas without needing
planning permission under ‘permitted development rights’; following the change this was still
possible, but required planning permission unless the parking area was permeable to water, or
was less than 5 m2 in area, or that water from the sealed surface was directed to some form of
soak-away such as part of the garden (DCLG, 2008). This regulation change was a response
to the substantial flooding that many areas of the UK had experienced in 2007, and a recog-
nition that further climate change would exacerbate the situation. Interestingly, the regulation
also affects replacement driveways (DCLG, 2008). At the watershed level, increasing levels of
sealed surface also impact on the ecology of freshwater systems with negative effects apparent
with as little as 10% sealed surface and severe effects being apparent at 30% (Stone, 2004).
Before the change in regulations in the UK, the concept of sustainable urban drainage
would probably have had a very limited ‘circulation’ in the country, restricted to those
charged with designing, implementing and building urban infrastructure; after the change,
knowledge of the concept (if not the terminology) potentially expanded to anyone with a
front garden who wanted to park their car on it! The Royal Horticultural Society in the UK
has even produced a guidance note explaining the issues and giving example layouts whereby
car parking, drainage and greenery can be integrated in front gardens (RHS, undated). Stone
(2004) noted that in Madison, Wisconsin, planning regulations for residential areas at the
time had the unfortunate effect of resulting in larger areas of sealed surface than necessary
compared with alternatives. He also noted that typical driveways in Madison made up 20%
of the sealed surface of a residential plot and suggested the approach that was implemented
in the UK four years later. Interestingly, Stone (2004) suggested a tax on sealed surfaces,
which would have impacted on new driveways and also promoted retrofitting of permeable
 Downloaded by [University of California, San Diego] at 19:36 04 May 2017

TABLE 4.3  Attributes of common approaches to creating permeable driveways to avoid need for planning permission in the UK (DCLG, 2008)

Planning Maintenance/ Wheelchair

permission Installation Environmental features Management issues accessibility
needed?
Cost Construction Use on slopes Visual Pollution control Local
microclimate
Loose gravel None Inexpensive Simple No Integrates well Gravel scatters and Unsuitable
with garden ruts form, will
design need sweeping
Wheel tracks None, if less Inexpensive Simple Integrates well Simple but regular
than 5 m2 with garden
design
Reinforced grass None Moderate Depends on Integrates well Absorbs Reduces local Regular mowing.
system with garden pollutants, dust, temperature Grass will die if
design reduces noise vehicles left in
same place for
long periods.
Reinforced gravel None Moderate Depends on OK, but see Integrates well Will need
system maintenance with garden sweeping to avoid
design gravel spreading
especially on slopes
Impermeable None Expensive Experienced/ Wide variety of Low maintenance,
block paving with specialist block types and long life
gaps between, contractor colours
permeable blocks, needed
permeable asphalt
 Permeable pavements  103

surfaces. In some ways the UK approach is more egalitarian, as a tax has the potential to
impact low-waged families more than the well off. However, a tax might be expected to
have an immediate stimulatory impact on reducing the levels of existing residential sealed
surface whereas the UK approach will take much longer as existing sealed surfaces wear out
and are replaced with permeable ones (the regulation requires replacement driveways to be
permeable).
The nature of the sub-base is quite critical to the success of a permeable pavement and in
the context of front gardens in the UK the Environment Agency has made specific recom-
mendations (Table 4.3). It is important that the sub-base does not contain fine material that
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

will be compacted and fill the gaps between the spaces of the larger stones. In the UK the
‘old’ type of hardcore used in driveways was called ‘MOT Type 1’ and this should not be
used – two suitable alternatives being ‘4/20’ and ‘Type 3’ (DCLG, 2008).

4.6 Alleys
Chicago has a large network of narrow accessible alleyways (3,058 km or 1,900 miles) and has
an active programme to convert them from sealed surfaces to permeable. The Chicago Green
Alley Handbook (CDT, undated) has been created to explain and promote the concept. For
Chicago this is particularly important as many of the alleys were not designed with connec-
tions to the city’s sewers and frequently flood. Making the alleys permeable to water is seen as
a less expensive flood control solution than upgrading with standard drainage solutions whilst
also providing significant environmental benefits including:

• recharging groundwater;
• reducing local temperatures (new surfaces are designed to have a higher albedo);
• using recycled material in the sub-base reducing pressure on landfills;
• reduction of light pollution.

Because lighting systems are upgraded at the same time with modern energy-efficient,
low-glare, ‘dark sky compliant’ systems, light pollution is reduced with the tantalising
suggestion that it will be possible ‘to see the stars’ again. Citizens are also encouraged, as part
of the programme, to:

• create rain gardens to prevent runoff from gardens into the alleys;
• carry out recycling including composting organic material;
• plant garden trees to shade the alleys;
• use native plants and trees in their gardens that are adapted to local (northern Illinois)
conditions to cut down on the need for garden watering;
• collect roof water in barrels to water the garden;
• use permeable paving materials for paths and parking areas;
• install green roofs on buildings and especially garages;
• create water retention ponds for stormwater control;
• create vegetated swales (which they also call ‘bioswales’) (CDT, undated).
 104  Permeable pavements

4.7 Roads
Fassman and Blackbourn (2010) monitored a 395 m2 experimental road section (including
200 m2 of permeable pavement and 195 m2 of path, driveway and grass) and an adjacent
conventional asphalt area of the road (850 m2 including associated features) along Birkdale
Road in Auckland, New Zealand between 2006 and 2008. This time-period included
81 storms amounting to 1,128.1  mm of rain. The permeable pavement was constructed
of impermeable blocks with 10 mm joints between them filled with aggregate chippings.
The design specification was for a ten-year life and capability of coping with a two-year
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

24 h ‘annual recurrence interval’ peak flow. Downslope of the permeable section, a heavy-
duty perforated drain pipe was positioned to remove underdrain water (i.e. water that
passed through the permeable pavement and sub-base). The results showed that surface
runoff, over 24  h periods, from the conventional road section mirrored the peak rainfall
events, but in the case of the permeable pavement the feature measured (the underdrain
discharge) was considerably reduced in intensity with a substantial lag in the case of small
(25 mm) and two-year storm events (63 mm) progressively becoming shorter with more
intense storms (five-year events (98 mm) and ten-year events (152.3 mm)). Some rainfall
was probably held in the sub-base air pockets, some evaporated off, and some infiltrated
into the relatively impermeable (clay) soils, all contributing to a reduction in the volume
of water in the underdrain discharge. Essentially the permeable pavement was able to
replicate pre-development conditions – an aim of the local Department of Environmental
Resources Programs and Planning Divisions to reduce impacts on stream integrity and
habitats (Fassman & Blackbourn, 2010).

4.8  Swales and strands

Swales are typically shallow vegetated channels; strands are larger channels that collect water
from a number of swales. As with permeable pavements, they are often included as compo-
nents of sustainable urban drainage systems frequently leading to retention ponds. They not
only convey excess water away from sealed surfaces, but, as they are vegetated, they also
act as infiltration structures and can remove pollutants (Rushton, 2001). Rushton (2001)
demonstrated that a permeable-surfaced area of a car park, coupled with a swale, could
reduce runoff by 50% compared with an asphalt-surfaced area with no swale; a swale reduced
runoff by 32% even when coupled with a sealed parking area. However, runoff reduction
was best during ‘small storms’; for large rain events there was less of a benefit. A number of
small structures have been used or suggested to help reduce runoff; an early example was
the use of small sunken vegetated islands in car parks, whereby water would run off the car
park surface into the basins for infiltration and treatment (Bitter and Bowers, 1994 cited in
Arnold and Gibbons, 1996).

4.9 Summary
• Permeable pavements can reduce the overall volume of water reaching drains, and
increase the time between rainfall and peak loads reaching those drains.
• Permeable pavements can be made out of a range of materials which incorporate
vegetation or inorganic materials which allow infiltration.
 Permeable pavements  105

• Because of the variety of materials that are available, the majority of ground surfaces
could be made permeable, though there may be maintenance issues.
• The biological component of permeable pavements is not restricted to vegetation, but
includes a subsurface community of microorganisms that can degrade pollutants.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

7KLVSDJHLQWHQWLRQDOO\OHIWEODQN
 5
GREEN WALLS
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

This chapter will:

• define the different kinds of green wall

• identify the structural issues that relate to different green walls
• describe green wall components/media
• cover planting and maintenance
• identify biodiversity and other ecosystem services
• consider the issue of sustainability.

5.1 Introduction
The green wall concept is essentially very simple: instead of growing plants on roofs or
on the ground where they take up space, grow them up unused space on walls. Just in
terms of aesthetics, green walls are far superior to green roofs as they are visible to a much
greater proportion of the population. As for wider environmental benefits such as reducing
the heat island effect and surface temperatures (Wilmers, 1988), there is a far greater
opportunity for delivering some environmental services with green walls than with green
roofs. Köhler (2008) estimated that walls represented double the ground footprint of a
building, in urban centres, although this seems likely to be on the conservative side, with
Peck et al. (1999) estimating it at between four and twenty times the roof area depending
on the height of the building, and much of the space on flat roofs is taken up with other
infrastructure, reducing growing opportunities (Cheng et al., 2010). On the negative side,
environmental conditions become increasingly harsh for wall vegetation at higher eleva-
tions (Peck et al., 1999) and at lower elevations salt spray during winter may cause damage
(Whittinghill & Rowe, 2011). Green walls have been known and grown for thousands of
years; Köhler (2008) identifies vines grown in palace gardens in the Mediterranean some
2,000 years ago as the earliest deliberately planted examples. Whilst many green walls
establish naturally, there is increasingly global interest in incorporating them in building
and city design because of their multifunctional benefits (e.g. Köhler, 2008; Loh, 2008;
 108  Green walls
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 5.1  In this student accommodation development (The Minories, near Tower Bridge,
London) the 191 m2 Biotecture Living Wall has been used as the external rain cladding of the
building allowing the costs of the green wall to be offset against using conventional cladding.
© Biotecture Ltd
 Green walls  109

Sheweka & Magdy, 2011) and because retrofitting green walls is relatively straightforward
compared with green roofs.
Most early published work on green walls was in German with ecological, rather than
purely botanical, publications only dating from the end of the 1970s (Köhler, 2008).
Köhler (2008) tracked green wall publication numbers from the 1850s to 2005 and
showed an explosion in publications in the 1980s and 1990s; since that paper a flurry
of studies devoted to green walls has been published in English (Chiquet et al., 2013;
Francis & Lorimer, 2011; Franco et al., 2012; Lindberg & Grimmond, 2011; Melzer
et al., 2011; Perini et al., 2011b; Sternberg et al., 2011b; Viles et al., 2011; Wang et al.,
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

2011; Whittinghill & Rowe, 2011; Xia et al., 2011). It is easy to get swept away with
enthusiasm for green walls, but it is perhaps worth remembering that they are not cost
free. Establishment costs will vary with the different types of wall (see below for different
types) and can be completely free if natural colonisation of existing structural walls is used,
or can be in the many hundreds of € per square metre for the most sophisticated living
wall systems. Perini et al. (2011b) give estimated costs for a range of systems (Table 5.1).
Much of the technology for the more complex systems is relatively new and low volume;
it is likely that costs will decrease over time. However, substantial cost savings are possible

TABLE 5.1  Example costs for different types of green wall system (mostly after Perini et al., 2011b)

System type Approach Materials €.m-2

Direct Plants growing on wall surface, planted Climbing plants 130–45
in ground
Indirect Plants growing up support materials, Climbing plants + support 140–75
(façade) planted in ground e.g. stainless steel wire
Indirect Plants growing up support materials, Climbing plants +
(façade) planted in planter boxes filled with soil. • galvanized steel support 600–800
Planters can be at different heights up a • coated steel 400–500
building to increase coverage • HDPE 100–150
Living wall Plants growing in a module subdivided Plants + HDPE planter 400–600
into smaller planters. Plants rooted in module + support structure
a soil or similar substrate (Greenwave
system)
Living wall Plants growing into panels containing Plants + panels + support 500–1000
hydroponic stonewool substrate structure
(Biotecture system)
Living wall Plants growing rooted directly in a Plants + modules using a 750–1200
panel which provides support with foam-based substrate in a
nutrients and water supplied wholly steel meshwork + support
hydroponically (Fytowall system) structure
Living wall Plants growing in pockets between Plants + felt layers + support 350–750
two-layers of a felt mat medium; structure
hydroponic

NB: living wall systems have more flexibility in the range of plants used than direct and façade systems, but will
require complex automatic irrigation and in-line fertiliser control. Planter box systems will also need irrigation
and fertiliser such as used for the more complex living wall systems if mounted above ground level, but could be
manually serviced if at ground level only.
 110  Green walls
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 5.2  Types of green wall: a) plants growing directly on the wall or rooted into cracks; b)
climbing plants adhering directly to the wall surface; c) climbing plants growing up a support
structure not directly on the wall; d) climbers growing in planters typically on support structures
but could also use plants that adhere directly to the wall (irrigated); e) plants cascading over
the top of a wall from a planter (irrigated); f) modular ‘living wall’ where modules contain a
growing medium (irrigated); g) hydroponic ‘vertical garden’: plants grown in felt pockets either
pre-formed or pockets cut into felt layers in situ (irrigated); h) free-standing green screen –
plants grown up a metal mesh support; i) hedge (after Ottelé et al., 2011).
 Green walls  111

if the green wall is designed as part of the external cladding of the building rather than as
an ‘add-on’ (Figure 5.1).
Maintenance costs also need to be considered: depending on the wall type it could be a
quick annual survey to check that undesirable plants have not established on the wall, or that
climbing plants have not obscured windows, got into the eaves or blocked gutters (Köhler,
2008), perhaps nothing more than a simple matter of a bit of cutting back. As wall sophis-
tication increases, so will management costs; for example, for irrigation and fertilisation of
living wall systems, as well as checking and managing plant growth, integrity of substrate, etc.
Set against this will be the many positive values of the walls, but it is best to know exactly
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

why you are creating green walls, what attributes you value in them (or why you are installing
them), and their potential establishment and maintenance costs.

5.2  Types of green wall

5.2.1  The basics

There are several different types of green wall (or ‘vertical greenery system’ (Wong et al.,
2010a)); the simplest is a building or curtilage wall that plants have colonised naturally; the
most complex is one which is composed of a pre-planted modular framework of subdivided
plastic partitions, fed water and nutrients through a network of pipes, and problems with
irrigation detected by sensors and transmitted to the installation/maintenance company so
engineers can be dispatched to repair the fault (Figure 5.2). Naturally the continuum between
these extremes also reflects a continuum in establishment and maintenance costs and also the
complexity and range of plants that can be incorporated in designs. Perez et al. (2011b), like
others (e.g. Loh, 2008), have attempted to bring some order to the classification of green walls
to enable like-for-like comparisons of, for example, energy conservation values. Unfortunately,
they have omitted some categories and their approach is not entirely satisfactory as they class
vegetation growing up a range of support systems, which they call ‘double-skin’ green walls,
as ‘extensive’, grouping them with climbers which adhere directly to built walls.
Ottelé (2011) developed a classification for ‘vertical green’ which made a distinction
between vegetation growing directly on the wall (direct greening) and that separated from
the wall surface (indirect greening) – a useful distinction (see below). However, Ottelé
(2011) identifies indirect greening by having an air gap between the wall and the vegetation
– given that this air gap can be filled with insulation, this distinction is unduly restrictive
and I will use the convention that the vegetation simply does not contact the wall surface
directly. ‘Direct’ and ‘indirect’ can be useful terms, but unfortunately terminology overlaps
so ‘green façades’ can include both direct and indirect approaches: compare Figure 5.2b,
a direct greening approach, with Figures 5.2c and d, indirect approaches, and with 5.2e,
somewhat intermediate as the vegetation may touch the wall but will not necessarily adhere
to it.
 112  Green walls

5.2.2  Direct greening

Structural walls made of stone, brick or other material where flora are:

• either rooted directly in or on the wall (Figure 5.2a, Figure 5.3)

• or surface climbers: rooted in the ground, but adhere to the vertical surfaces directly
through adhesive pads (Figure 5.2b, Figure 5.3) – a form of green façade.

These two forms can develop naturally, or can be deliberately planted. Whilst flora will
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

naturally colonise the surfaces of buildings and free-standing walls such as boundary walls,
given suitable conditions and enough time, there is little indication that walls intended to
be structural components of buildings which have been deliberately created with pockets
for plants to be grown in them has been taken up as a concept. An exception to this is the
‘Growcrete’ concept which facilitates growth of plants on exterior surfaces of buildings – see
below. Free-standing walls, such as boundary walls and retaining walls, may well be designed
with plant ‘pockets’.

FIGURE 5.3  Direct greening of walls. Left: Red valerian Centranthus ruber plants directly rooted
in a wall in Slapton, Devon, UK; right: Boston ivy Parthenocissus tricuspidata adhering to a
building in Durham, UK. © John Dover
 Green walls  113

5.2.3  Indirect greening

These are typically structural walls, composed of the same material as for directly greened
walls; here the flora are grown on support infrastructure of various kinds which are fixed to
the wall but where the vegetation does not touch the wall surface:

• Green façades: rooted in the ground or in planters. Plants twine around support materials
fixed to the vertical surfaces but are offset from them (Figure 5.2c, d; Figure 5.4), these
may also have roof planters to ‘cascade’ vegetation down the walls from above (Figure
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

5.2e).
• Living walls – modular type: rooted into soil or other substantial substrate and water
delivered through an irrigation system, along with additional nutrients (Figure 5.2f;
Figure 5.4).
• Living walls – mat type: rooted into a thin, synthetic, ‘felt-like’ blanket and maintained
wholly hydroponically (Figure 5.2g; Figure 5.4).

In all cases the structures are separated from the supporting wall (by an air space or insulation)
and, in the case of the latter two, by a waterproof membrane.

5.2.4  Free-standing ‘curtain’ walls and hedges

Another class of green wall can be identified: where the vegetation is supported by a
framework, temporary or permanent, which is designed to be free-standing. An example
of this approach is the green wall in Shepherd’s Bush, London, which is used to screen the
Westfield Shopping Centre (Figure 5.5) and uses the ANS modular system. Such walls can
use directly adhering plants, support structures for twiners, or living wall approaches as the
underlying support structure can be custom designed. Free-standing green walls have a wide
range of applications and some designs have the advantage of taking up very little ground
space (Figure 5.2h and Figure 5.6).
There is, of course, another type of green wall: a hedge, which provides its own support
mechanism (Figure 5.2i; Figure 5.7). A hedge can also be grown against a wall as well as
being completely free-standing. There is surprisingly little known about the value of hedges
in the urban environment (Chiquet et al., 2012), which contrasts strongly with the wealth of
knowledge about their country cousins (e.g. Barr and Petit, 2001; Dover, 2012).
 114  Green walls
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 5.4  Indirect greening approaches. Top left: Green façade but with support infrastructure
for plants so they are held away from the wall © John Dover; top right: plastic modular unit
holding organic substrate © John Dover; bottom: Patrick Blanc felt-based system (on a road
bridge in Aix-en-Provence, France) © Caroline Chiquet.
 Green walls  115
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 5.5  Two faces of the same green wall © John Dover. The free-standing ‘curtain’ green
wall at the Westfield Centre, Shepherd’s Bush, London, has different flora either side reflecting
the requirements of the different aspects, but also the need for high visual quality on the
‘shopping side’ of the structure (top). The plants for this particular wall are grown in the ANS
modular system. The ridged stonework at the base of the wall on the shopper’s side is actually
a water feature designed to stop passers-by picking vegetation, but when this photograph was
taken the water had been turned off and the structure was being used as welcome seating and
picnicking space by shoppers.
 116  Green walls
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 5.6  The installation of a

Mobilane ‘Green Screen’ at the
Staffordshire University campus in
Stoke-on-Trent, UK. The structure
is free-standing, requires little space,
being extremely narrow, and the root
system is contained in a ‘U’-shaped
bend in the supporting meshwork. ©
John Dover

FIGURE 5.7  Hedges can be completely free-standing to mark boundaries or divide areas into
intimate plots or can be grown against a wall – evergreen species can be very effective in hiding
ugly structures. © John Dover
 Green walls  117

5.3  Directly greened walls: rooted directly on the wall

5.3.1 Introduction
These walls are effectively direct analogues of natural vertical features such as cliff faces
and ravines; they are also subject to the same limitations as these natural features (Larson et
al., 2000, 2004). A number of publications examine the severe environmental constraints
imposed on plants that colonise and root into walls (e.g. Woodell, 1979; Gilbert, 1996;
Francis, 2010) and the ‘urban cliff hypothesis’ compares the conditions in extreme natural
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

conditions with that pertaining in urban areas (Lundholm, 2006). Essentially these constraints
break down into structural, environmental and biotic considerations.

5.3.2  Structural issues

Unbroken, sheer cliff faces are inhospitable to plants because gravity is a great motivator:
without any cracks or ledges plants cannot find a purchase. In the same way, a new wall
made of natural stone, brick, concrete or plastic material with a fine surface and well
mortared presents little opportunity for colonisation. The acidity–alkalinity of the building
material will also influence plant colonisation, especially by lichens and mosses, although
Gilbert (1996) considers this to be of minor importance for most other species. In the case
of both cliff and wall, time and erosive processes (e.g. alternating heat/cold cycles, frost, rain
(with dissolved acids), wind-blasting with sand or other particles, and mechanical damage)
eventually cause cracks to appear, small ledges to develop, and aggregations of dust or eroded
material to accumulate (Woodell, 1979), providing plants with both colonisation niches and
substrate. Lisci et al. (2003) summarise the process of plant succession once erosive processes
have provided colonisation niches (Figure 5.8).

FIGURE 5.8  The stages of wall colonisation following initial disruption of the wall surface: (a)
a moss spore falls on porous or damaged stone; (b) it develops into a plant which accumulates
atmospheric dust; (c) the moss-captured dust acts as growing medium for a plant seed which
germinates and grows; and (d) the moss does not damage the substrate, but the roots of the
growing plant penetrate it. Redrawn from Lisci et al. (2003).
 118  Green walls
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

FIGURE 5.9  Walls in the village of Slapton, Devon, UK. Clockwise from top left: boundary
(curtain) wall showing plants colonised in aggregated material at the base of the wall, specialised
wall flora such as navelwort on the main surface of the wall and ivy and fern on the top of the
wall; capstone colonised by lichen, ferns and mosses; top of wall colonised by brome grass;
soil-retaining wall completely covered in a diverse flora and especially red valerian. © John Dover
 Green walls  119

With walls, the mortared joints are typically the weak points where erosive processes act
first, unless the construction materials are naturally friable or of poor quality. It is general
practice to make the mortars of a similar strength as the building materials (Busby, 1960) or
slightly weaker. Of course, not all walls have mortared joints, some being fitted ‘dry’. From
about 1870 lime-based mortar, which is susceptible to weathering, started to decline in use
(Gilbert, 1996) and since 1930 most new buildings in the UK have been made with much
stronger non-lime mortars (BDA, 2001). This change in bonding material means that plant
colonisation is less likely on new walls, or at least will take considerably longer (Woodell,
1979): of the order of 40–80 years (Gilbert, 1996). There are three main classes of mortar:
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

lime as the bonding agent plus an aggregate such as sand; cement mortars typically using
ordinary Portland cement and sand; and mortars containing both cement and lime as bonding
agents with sand as aggregate (Busby, 1960). In London, it appears that some species are
now restricted to very old walls because of the change to non-lime mortars (Gilbert, 1996).
Repairs to soft, old or easily eroded structures, if done with inappropriate mortars (e.g.
cement rather than lime), can result in the mortar eventually protruding from the eroding
face, creating colonisation ledges. Walls are often constructed with unintentional built-in
colonisation ledges, typically at the tops of boundary or retaining walls where there are
capstones (Figure 5.9), but often near the ground where there are sloping ‘knees’. Vertical
walls are more challenging for plants to colonise than sloping walls, and the flora of vertical

FIGURE 5.10  The main microsites for plant growth. Gaps in stone/brickwork at: (a) ground
level, (b) on a slope, (c) where two different materials meet, (d) on a vertical surface, (e) on
a horizontal surface, (f) at the junction of vertical and horizontal faces, (g) at the junction of
two vertical faces. Additionally, growth may take place on: (h) a porous horizontal surface
(see further Figure 5.8) and (i) on broken or damaged surfaces. Also, rhizomes may penetrate
retaining walls (j), and (k) soil accumulations at the base of a wall may provide growing media.
Adapted from Lisci et al. (2003).
 120  Green walls

walls is much more distinctive (Gilbert, 1996). Aggregations of soil particles can also build
up at the base of walls where a wider variety of plants can colonise (Figure 5.9). Lisci et al.
(2003) produced a nice diagram of potential wall colonisation sites, redrawn and modified
here as Figure 5.10.

5.3.3  Environmental issues

If structural issues were the only factors involved in plant colonisation of walls, then almost all
older walls would be covered in plants! The best place I know for plant colonisation of stone
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

walls in England is the pretty coastal village of Slapton in south Devon (Figure 5.9) and even
Slapton hardly approaches that ideal! The local abiotic environment of the wall is crucial,
and in particular that which relates to moisture (Lisci & Pacini, 1993). There are several ways
that plants on walls can obtain moisture: directly from rain, from highly humid environments
(e.g. Francis & Hoggart, 2008, 2009), from seepages or water spilling over wall surfaces, from
capillary action from the wall foundations, and from lateral transmission through the wall if
it acts as a soil-retaining wall (e.g. as part of a garden) (Lisci & Pacini, 1993; Woodell, 1979).
Different parts of the wall will experience different moisture regimes in any given situation.
If water comes from capillary action, it will not penetrate very far up the wall, and not at
all if a damp-course is installed. Walls are unlikely to be uniformly inundated by spills or
seepages, sheltered parts of walls may not receive as much moisture as more exposed areas,
etc. A soil-retaining wall (Figures 5.9 and 5.10) is perhaps that which has most consistent
access to moisture (Woodell, 1979), and even then the upper part of the wall will have less
as it is likely to be at least at soil surface level if not somewhat higher. Whatever the source
of moisture, it will be modified by the thickness of the wall, the aspect (with north-facing
walls the dampest and south-facing the driest in the northern hemisphere) and exposure
to the prevailing winds. Some plants are resistant to desiccation and may appear dead and
shrivelled during the dry season, especially on more exposed parts of walls, but return to
good condition as the seasons change (Woodell, 1979) (Figures 5.9 and 5.11). Geographical
location also plays a part. Within the UK the eastern half of England is less suitable for wall
flora than the western half, Wales and Scotland (Gilbert, 1996). In continental Europe, the
suitability of walls as plant habitat declines with increasing distance eastward from the coast
(Gilbert, 1996). Nutrients for plants on walls can come from material in solution, depending
on the water source, or from dust, soil, guano, etc. Essentially, there is a tension between
moisture and elements that reduce moisture; in principle this means that sheltered and shady
places should produce a good flora. However, plants need light to grow and so another axis
enters the equation: with dark, moist walls having a limited range of shade-tolerant plants
capable of utilising the moisture (Woodell, 1979) just as bright, exposed walls have few plants
capable of colonising them (unless there is a constant water source).

5.3.4  Biotic issues

The specific make-up of the flora on any particular wall or part of a wall will be influenced
by a) the structural and environmental issues already discussed and b) the adaptations of
individual plants to those environmental and structural issues (Lisci et al., 2003); because
of this, wall ecologists recognise discrete zones which have distinctive floral communities
(Figure 5.11).
 Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Top: sometimes with a

capstones and ledges.
Plenty of gaps for
establishment

b
Upper and middle: little
water, highly stressed
environment, distinctive
vegetation
c

Lower: damper (capillary

action), splashes from d
passing vehicles

Base: soil build-up from

dust, fertile
FIGURE 5.11  The main areas identified by wall ecologists and the different environmental conditions and flora they support: a) the wall top typified by a
range of species including sedums and woody plants (e.g. red valerian), lichens, ferns, grasses, and flowers (e.g. Mexican fleabane Erigeron karvinskianus);
b) the upper and middle zones which have the most specialist wall flora, e.g. plants such as ivy-leaved toadflax, navelwort, yellow corydalis; c) the lower
zone which is typically damper and can thus support plants which are less tolerant of stressful conditions; and d) the base of the wall where a wide range of
plants can grow, though they will become subject to drought stress in warm weather. Adapted from a diagram in Gilbert (1996). Images © John Dover
 122  Green walls

TABLE 5.2  Examples of papers with data on naturally colonised wall flora

Town/City Country Authors

Pavia, Rome, Siena Italy Lisci & Pacini (1993); Lisci et al. (2003)
Srinagar, Kashmir India Ahmed & Durani (1970)
London England Francis & Hoggart (2009); Woodell (1979)
Berwick-on-Tweed, Chester, Chichester, England Gilbert (1996)
Norwich, Southampton, York
Caernarvon, Chepstow, Conway, Denbigh, Wales Gilbert (1996)
Tenby
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Concarno, Brittany France Gilbert (1996)

There is a third dimension which is also important: a source of propagules. As is well

understood from restoration work in grasslands (Bekker et al., 2000), if there is no nearby
source of colonists, then the flora will remain very simple – even so, because of the stressed
nature of the wall environment, only a subset of plants would be able to colonise. So dispersal
mode of plants and dispersal distances becomes an important consideration, especially in very
urban areas. Woodell (1979) notes that more than 1,200 species of plant have been recorded
from walls and also that the majority of plants probably colonise because they have unspe-
cialised dispersal mechanisms, with wind being the most likely dispersal mode (e.g. Ahmed
& Durrani, 1970; Lisci & Pacini, 1993). Ants are also known to be important in the dispersal
of some wall flora (Ahmed & Durrani, 1970), which is facilitated by the presence of a food
‘reward’ for the ants on the seeds: the elaiosome (Gilbert, 1996). Similarly, bird guano may be
an important ‘carrier’ of the seeds of fleshy fruits (Ahmed & Durrani, 1970; Lisci et al., 2003).
It should be obvious from Chapter 2 that the city environment is different from that of the
countryside, and it can be inimical to plants, with some (e.g. lichens) extremely sensitive to
pollution whilst others (e.g. hartstongue fern Phyllitis scolopendrium) are remarkably resilient
(Woodell, 1979). Whilst the flora of most walls is likely to be composed largely of relatively
common and widespread species, albeit ones capable of tolerating tough conditions, walls are
known to harbour some rarities (Woodell, 1979); for example, the plants yellow whitlow-
grass Draba aizoides, wall bedstraw Galium parisiense, perfoliate penny-cress Thlaspi perfoliatum,
London rocket Sisymbrium irio, the moss Trichostomopis umbrosa and the lichens Calicium
corynellum and Roccella phycopsis (see further Gilbert, 1996). In the UK, about 20 species
characteristic of cliffs and rocky outcrops (half native, half introduced) are mainly found on
walls (Gilbert, 1996). Table 5.2 gives a number of references to wall flora studies. There
appears to be little information on the value to fauna of naturally colonised green walls, but
see the work of Attrill et al. (1997) on river walls along the Thames, UK (Table 5.8).

5.3.5  New walls

Naturally colonised walls can be beautiful or nondescript, be vertical carpets or simply a
varied patchwork of bare wall and many or few plants. What is undoubted is that natural
colonisation by flora takes a considerable time. In terms of green infrastructure, walls with
existing flora should be cherished and maintained as appropriate (see Gilbert, 1996), but,
generally speaking, developing new walls and waiting for natural colonisation will not
address the needs of the present. We know from references in Woodell (1979) that deliberate
 Green walls  123

inclusion of plants in walls has been going on for some considerable time; my favourite is
the charming instance of a man planting tulip, iris, fritillary and daffodil bulbs in a mud wall
in Harwan, Srinigar, in Kashmir (Ahmed & Durrani, 1970). There is no reason why new
boundary, garden and retaining walls should not be designed to incorporate vegetation that
is deliberately planted and Kontoleon and Eumorfopoulou (2010) use the term ‘Landscape
Walls’ to describe such structures. For suitable plants and design ideas, Jekyll (1982) (initially
published in 1901 and revised eight times) is as good a place as any to start. Whilst her book
is concerned with stone walls in ornamental garden settings, many of the principles will
be relevant to green infrastructure plantings whether of native or more horticultural flora.
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

Clearly the issues that impact on existing naturally colonised walls in terms of structure,
environment and plant suitability need to be addressed, but designing a wall with planting
specifically in mind should allow a wider range of species to be used. It may even be possible
to design walls specifically for the conservation of rare flora. Soil or other rooting media
can be incorporated in the core of new walls and cavities included for direct planting –
low boundary walls should pose no particular safety issues, but the integrity of larger walls
would probably need the input of a structural engineer at the design stage (see also Woodell,
1979; Francis, 2010). We know that soil-retaining walls are likely to provide good oppor-
tunities for lateral water movement into the wall, so free-standing walls are likely to pose
the greatest challenge and may need to incorporate piped irrigation for use during very dry
periods – incorporation of sensors for automatic watering would be an obvious refinement.
For river walls, irrigation is unlikely to be a significant problem; but the smooth materials
used in modern construction are not plant-friendly and Francis and Hoggart (2008) describe
approaches that can encourage colonisation.
Rock gabions, metal mesh cubes filled with rocks, are often used to reinforce steep banks
or to prevent erosion, and plants can be grown on soil used to cover the horizontal surfaces.
A different approach is to use strips of high density polyethylene (HDPE) which are linked
together to form a honeycomb and which are filled with soil on site. The ‘Webwall’ system
produced by ABG Ltd (ABG, undated) can be stacked in an offset to form a stepped gradient
to a bank. Substantial-sized banks can be stabilised in this way using a geotextile behind the
walls to provide extra stability. Webwall panels come in standard heights of 250 or 500 mm,
in 4  m lengths, and widths between 1 and 2  m. The advantage of this type of system is
that on-site material can be used to fill the honeycomb, reducing labour, fuel and transport
costs, and the open upper surface of the steps can be planted up with vegetation to hide the
HDPE. The deeper soil profile means a wider range of species (grass to trees) can be planted
compared with rock gabions, and with better soil moisture retention.
Another new approach to growing wall flora is the use of the Growcrete system inves-
tigated by Ottelé et al. (2010a) at Delft University in the Netherlands. These are essentially
panels of concrete with an upper (or outer, when mounted) surface with a very open
granular structure created by using 32 mm diameter lava stones. This open structure (depth:
80 mm, on an 80 mm structural base) is used to hold a growing medium. The pH of the
soil medium is influenced by the highly alkaline nature of the concrete/lava substrate and in
tests increased from a starting level of pH 7.2 to pH 9.2 over a three-month period. As with
non-retaining walls, watering is via precipitation and, because of the extreme environment,
they are likely to support the classic wall vegetation of standing or building walls. Ottelé et
al. (2010a) found, as might be expected, that of their seven test plantings (wild or creeping
thyme Thymus praecox, sedum Sedum spp., grey cranesbill Geranium cinereum subcaulescens,
 124  Green walls

maidenhair spleenwort Asplenium trichomanes, hart’s tongue fern A. scolopendrium, ivy-leaved

toadflax Cymbalaria muralis and aubretia Aubretia sp.), ivy-leaved toadflax and the sedums did
best and aubretia, grey cranesbill and the spleenwort were not suitable. Because this wall is
designed with a rugose surface and with potentially high plant coverage, it should be quite
effective in particulate removal compared with naturally colonised walls that tend to have
sparse vegetation, but data are not yet available.

5.4  Direct greening: surface climbers

Downloaded by [University of California, San Diego] at 19:36 04 May 2017

This section deals with plants that are rooted in the ground but climb up the wall surface.

5.4.1  Plants that adhere to the vertical surfaces directly through

adhesive pads
Whilst many plants can be grown against vertical surfaces, only those, like ivy, that are capable
of adhering to building surfaces through aerial rootlets which exude an adhesive substance
(Endress & Thomson, 1976; Steinbrecher et al., 2010; Xia et al., 2011) can do so without
additional support. As well as the visual and environmental improvement of urban walls,
climbing vegetation is also useful in the stabilisation of slopes (see Wang et al., 2009). There
are about 16 species of ivy (Ackerfield & Wen, 2003), of which Hedera helix (with over 400
varieties) is the most dominant the UK (Sternberg et al., 2011a). A single ivy plant can reach
30 m in height and cover 600 m2 of wall (Dunnett & Kingsbury, 2004). Other self-adhering
species commonly found on buildings and walls include: Boston ivy Parthenocissus tricus-
pidata, trumpet creeper Campsis radicans, Chinese Virginia creeper Parthenocissus henryana and
climbing hydrangea vine Schizophragma hydrangeoides (Dunnett & Kingsbury, 2004; Sternberg,
2010; Wang et al., 2009). Species that can attach themselves to walls but which are better
for having some support include: Virginia creeper Parthenocissus quinquefolia, winter creeper
Euonymus fortunei and climbing hydrangea Hydrangea petiolaris (Dunnett & Kingsbury, 2004).
Ahmed and Durrani (1970) indicate that species of Lonicera are self-supporting on the walls
of Srinigar. Dunnett and Kingsbury (2004) note that most climbers have a strong tendency
to grow towards light (positive phototropism) but that this tendency is quite weak in P. tricus-
pidata – useful for ensuring the whole wall surface is covered rather than just the middle and
upper parts.
Witter (1986) described a project initiated by the Kassel City Council in Germany to
encourage the greening of walls in the city. The project ‘went live’ in 1984 with much
publicity including an A2 poster, planting tips, access to further council advice services and
even a procedure for requesting planting trenches to be dug by the council along footpaths.
Despite the fanfare, and interest, nothing spectacular in the way of uptake resulted. Whilst the
public were being wooed, plantings were made at the base of particularly ugly civic buildings
and it appears that in 1985, when the project was re-launched, the combination of publicity
and examples worked. When Witter (1986) reported on the project, there were almost 40
public buildings with climbing plants. At pretty much the same time, Berlin had developed
an incentive scheme for green walls which resulted in 245,584 m2 being grown over the 15
years of the programme (1983–1997) (Köhler, 2008).
 Green walls  125

5.4.2  Do climbing plants damage walls?

The answer to this is ‘yes and no’. Woodell (1979) points out that plants cannot establish
until some, typically abiotic, structural change has been made in a wall to allow colonisation;
this appears to include lichens, once regarded as primary colonisers (Lisci et al., 2003).
Warscheid & Braams (2000), however, note that microbial biofilms establish quickly when
stone surfaces are exposed to the environment with equally speedy surface-deteriorating
effects including providing moisture for freeze-thaw cycles and facilitating the acquisition of
atmospheric pollutants. Ironically, Warscheid and Braams (2000) also note the deliberate use
Downloaded by [University of California, San Diego] at 19:36 04 May 2017

of some microbial agents in creating protective surfaces on stone and even in the removal of
some surface pollutants. Once in, plants can accelerate degradation through acidic exudates
(Lisci et al., 2003) and mechanical effects via root growth forcing gaps to expand (Lisci et al.,
2003; Viles et al., 2011). Colonisation niches may also allow ingress of pests into buildings.
Fortunately, after this catalogue of horrors, Woodell (1979) pulls things back into perspective
by noting that any structural damage by plants is likely to be mostly through tree and shrub
growth rather than by herbaceous species, and is likely to happen extremely slowly, allowing
remedial action to be taken before real damage occurs. Viles et al. (2011) convincingly
demonstrated that ivy (Hedera helix) rooted directly in cracks in walls, as opposed to growing
as a climber up walls, can have serious structural impacts – and that failed attempts to control
ivy may actually promote damage (see also Lisci et al., 2003). Some ancient monuments may
be susceptible to degradation by physical and chemical (acid) attack from lichens. In such
situations, elimination of some species of lichen may be advisable, but for others it may not,
as removal can accelerate deterioration (Lisci et al., 2003). For non-heritage walls, lichen
damage is likely to be so slow as to be irrelevant.

5.5  Indirect greening of walls

5.5.1  Types of indirect greening systems

‘Indirect’ green walls are composed of three classes of wall: green façades and two types of
living wall (mat and modular) (section 5.2.3).
The advantages of these ‘designed’ systems are obvious:

• they allow a greater degree of control over which part of a building or wall the plants
grow over (over and above simple pruning);
• they allow the use of a much wider range of plant species;
• they have the great advantage over green roofs in that they are very easy to retrofit and
can, potentially, cover a greater area than roofs.

The living wall systems allow greater artistic expression in the planting designs, with Patrick
Blanc’s creations being quite startling and dramatic in their exposition (Blanc, 2008).

5.5.2  Plants that twine around support materials fixed to vertical surfaces
Plants that adhere directly to wall surfaces (see section 5.3.1) have an advantage in estab-
lishment as they require no additional infrastructure: simply prepare the ground appropriately,
 126  Green walls

plant them and off they go; in many cases they do not even seem to need planting but occur
without human intervention! They do have downsides: man