Ecological Research Monographs

Futoshi Nakamura Editor

Green
Infrastructure and
Climate Change
Adaptation
Function, Implementation
and Governance
Ecological Research Monographs

Series Editor
Yoh Iwasa, Department of Biology, Kyushu University, Fukuoka, Japan
The book series Ecological Research Monographs publishes refereed volumes on all
aspects of ecology, including Animal ecology Population ecology Theoretical
ecology Plant ecology Community ecology Statistical ecology Marine ecology
Ecosystems Biodiversity Microbial ecology Landscape ecology Conservation
Molecular ecology Behavioral ecology Urban ecology Physiological ecology Evo-
lutionary ecology The series comprise books and edited collections by international
experts in their fields.

More information about this series at https://link.springer.com/bookseries/8852

Futoshi Nakamura
Editor

Green Infrastructure and

Climate Change Adaptation
Function, Implementation and Governance
Editor
Futoshi Nakamura
Laboratory of Ecosystem Management,
Graduate School of Agriculture
Hokkaido University
Sapporo, Japan

ISSN 2191-0707 ISSN 2191-0715 (electronic)

Ecological Research Monographs
ISBN 978-981-16-6790-9 ISBN 978-981-16-6791-6 (eBook)
https://doi.org/10.1007/978-981-16-6791-6

© The Editor(s) (if applicable) and The Author(s) 2022. This book is an open access publication.
Open Access This book is licensed under the terms of the Creative Commons Attribution 4.0 Inter-
national License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation,
distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons license and indicate if changes
were made.
The images or other third party material in this book are included in the book’s Creative Commons
license, unless indicated otherwise in a credit line to the material. If material is not included in the book’s
Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copyright holder.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

Cover illustration: A flood control basin as a green infrastructure and a pair of Red-crowned cranes
nesting there (Photos provided by Hokkaido Development Bureau and Naganuma Town)

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface

This volume consists mainly of studies conducted during two projects, titled “Green
Infrastructure with a Declining Population and Changing Climate: Assessment
of Biodiversity, Disaster Prevention, and Social Values (2015–2018)” and “Com-
plementary Role of Green and Gray Infrastructures: Evaluation from Disaster
Prevention, Environment, and Social and Economic Benefit (2018–2021),” that
were supported by the Environment Research and Technology Development Funds
(4-1504 and 4-1805) of the Ministry of the Environment of Japan. The leader of the
two projects was Dr. Futoshi Nakamura. Before these projects, Dr. Nakamura and
other members from various scientific fields organized the Japan Green Infrastruc-
ture Association, which aims to promote the establishment of a scientific basis of
green infrastructure (GI), to apply GI for biodiversity conservation and disaster risk
reduction under a climate change-affected environment, and to exchange scientific
knowledge of GI with other countries. Thus, this volume includes articles provided
by the members of the GI association and foreign scientists that exchange academic
information with the association.
The aims of this volume are (1) to introduce the progress of the conception,
evaluation, implementation, and governance of GI in Japan and other countries; (2)
to provide basic information regarding the structure, function, and maintenance of
GI for scientists, university students, government officers, and practitioners; and
(3) to accelerate the transformative changes from gray-based strategies to green- or
hybrid-based strategies to adapt to climate change.
The ecosystem functions of forests and wetlands have long been recognized
and analyzed quantitatively; thus, the idea of GI is not new. However, developed
countries such as Japan have lost natural and seminatural GI due to the historical
overuse of natural resources and intensive land use development. To compensate
for the loss of GI functions, many engineering gray infrastructures have been built.
For example, Japan has been building continuous artificial levees and large dams
to protect human lives and assets from floods in the last century. However, the
construction and maintenance costs of gray infrastructure are enormous, and they
have various negative impacts on biodiversity and ecosystem services. According
to the WWF Living Planet Report 2014 Living Planet Index (LPI: a measure

v
vi Preface

of the state of the world’s biological diversity based on population trends of

vertebrate species), the freshwater index has shown the greatest decline of any of
the biome-based indices. The main threats are habitat loss and degradation, such as
through direct impacts from dams and unsustainable water extractions, followed by
overexploitation.
Moreover, climate change adds another constraint on the use of gray infras-
tructure because extraordinary events caused by global warming may frequently
exceed the planning level of gray infrastructure, resulting in a great deal of damage
to human lives and properties. Japan experienced frequent occurrences of mega-
flood disasters, such as torrential rain disasters in 2018 that caused landslides and
flooding, which killed 200 people, and Typhoon Hagibis in 2019, which caused 140
breaks of artificial levees, 25,000 ha of inundation, and 64 deaths.
Thus, the strategy of dependence on gray infrastructure may not be acceptable,
specifically in a depopulating society such as Japan. The drastic aging and depop-
ulation in Japanese society, like other developed countries, will likely have impacts
on social security (such as medical services and care), pensions, tax revenues,
and the maintenance of existing infrastructures and will lead to farmland and
forest abandonment. The Ministry of Agriculture, Forestry and Fisheries of Japan
identified an increase in abandoned farmlands in Japan from 130,000 ha in the
late 1980s to 400,000 ha in 2011. This outlook appears pessimistic, but it may
provide other opportunities that we did not have in the past. If we can implement
the best management of land use changes, these areas will become GIs that play
simultaneous roles in disaster prevention and biodiversity conservation.
The purposes of implementing GI differ among countries. The United States
strongly focuses on stormwater management in urban areas by GI, which reduces
the amount of flooding and the polluted runoff that reaches sewers, streams, and
lakes. In contrast, GI refers to the ecological network in the European Union, which
delivers a wide range of ecosystem services and functions, such as biodiversity
conservation, water purification, air quality, space for recreation, and climate
mitigation and adaptation. The purpose of introducing GI in Japan is relatively
closer to that of EU countries. In the EU, in May 2013, the Commission adopted
an EU-wide strategy promoting investments in green infrastructure to restore the
health of ecosystems, ensure that natural areas remain connected together, and allow
species to thrive across their entire natural habitat. In Japan, GI was promoted in the
National Spatial Planning and Priority Plan for Social Infrastructure Development
in 2015. However, scientific research on GI in Japan is limited and is still in the
beginning stage; there are many uncertainties and knowledge gaps in dealing with
GI, especially the structure and function of individuals and networks of GI and
management strategies, along with the stewardship of local communities.
The current major concern for people worldwide is COVID-19. As indicated
by many scientists, the spread of viruses, loss of biodiversity, and climate change
are interrelated with each other, and all are generated from the overuse of natural
resources by humans. Many plants and animals have become extinct due to
intensive exploitation of forest, river, wetland, and underground resources. Global
warming proceeds with carbon dioxide emissions associated with the consumption
Preface vii

of fossil fuels and the destruction of forests and wetlands. Once natural ecosystems
are destroyed by human actions, opportunities for pathogens to pass between
animals and people will increase. Green infrastructure will improve the situation by
preserving natural and seminatural ecosystems and by restoring healthy ecosystems
and will function as an adaptation strategy for climate change.
As a result of the pandemic, individuals may have to refrain from traveling and
may have to telework full-time, isolated from coworkers, friends, and family, which
is very stressful and makes it difficult to maintain physical and mental health. GI
in the hometowns of these individuals may provide opportunities for walking and
jogging and other recreation activities, which alleviate anxiety and stress resulting
from the COVID-19 pandemic. I think the ultimate goal of the introduction of GI
into society is the improvement of the quality of life, health, and well-being of
individuals and communities.
I am grateful to the Ministry of Environment, Japan, for providing financial
support for the two projects. Additionally, I would like to thank all the authors for
submitting their chapters, all the reviewers who were willing to review the original
manuscripts, and Ms. Fumiko Yamaguchi and other staff at the publisher, Springer
Japan, for their encouragement and assistance.

Sapporo, Japan Futoshi Nakamura

Contents

1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 1
Futoshi Nakamura

Part I Concept and Synthesis

2 Concept and Application of Green and Hybrid Infrastructure . . . . . . . 11
Futoshi Nakamura
3 An Economic Analysis of Optimal Hybrid Infrastructure: A
Theoretical Approach in a Hydro-Economic Model .. . . . . . . . . . . . . . . . . . . 31
Ayumi Onuma
4 Flood Management Policy in Shiga Prefecture, Japan:
Implementation Approach of a Risk-Based Flood
Management System at Catchment Scale . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 43
Kentaro Taki
5 Toward Social Infrastructure: Typological Idea for
Evaluating Implementation Potential of Green Infrastructure . . . . . . . 61
Takeshi Osawa and Takaaki Nishida

Part II Forest Ecosystem

6 Riparian Forests and Climate Change: Interactive Zone
of Green and Blue Infrastructure .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 73
Futoshi Nakamura
7 Improvement of the Flood-Reduction Function of Forests
Based on Their Interception Evaporation and Surface
Storage Capacities .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 93
Takao Tamura

ix
x Contents

8 Forest Green Infrastructure to Protect Water Quality: A

Step-by-Step Guide for Payment Schemes . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 105
Thomas R. Nisbet, Maria-Beatrice Andreucci, Rik De Vreese,
Lars Högbom, Sonja Kay, Mary Kelly-Quinn, Alessandro Leonardi,
Mariyana I. Lyubenova, Paola Ovando Pol, Paula Quinteiro,
Ignacio Pérez Silos, and Gregory Valatin

Part III River and Floodplain Ecosystem (Including Paddy Field

and Other Farmlands)
9 Wetland Paddy Fields as Green Infrastructure Against Flood .. . . . . . . 135
Yasunori Muto and Ryo Yokokawa
10 Change in Floodwater Retention Function of a Paddy Field
Due to Cultivation Abandonment in a Depopulating Rural
Region in Japan .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 161
Yota Imai, Yasunori Muto, and Mahito Kamada
11 Paddy Fields as Green Infrastructure: Their Ecosystem
Services and Threatening Drivers . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 175
Takeshi Osawa, Takaaki Nishida, and Takashi Oka

Part IV Wetland Ecosystem (Including Flood-Control Pond)

12 Flood-Control Basins as Green Infrastructures: Flood-Risk
Reduction, Biodiversity Conservation, and Sustainable
Management in Japan.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 189
Nobuo Ishiyama, Satoshi Yamanaka, Keita Ooue, Masayuki Senzaki,
Munehiro Kitazawa, Junko Morimoto, and Futoshi Nakamura
13 Natural Succession of Wetland Vegetation in a Flood-Control
Pond Constructed on Abandoned Farmland . . . . . . . .. . . . . . . . . . . . . . . . . . . . 209
Junko Morimoto, Susumu Goto, Akito Kuroyanagi,
Motoko Toyoshima, and Yuichiro Shida
14 Biodiversity Conservation through Various Citizen Activities
in a Flood Control Basin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 225
Jun Nishihiro, Shohei Koike, and Atsushi Ono

Part V Urban and City Ecosystem

15 Toward Holistic Urban Green Infrastructure Implementation . . . . . . . 239
Takanori Fukuoka
16 Changes in the Use of Green Spaces by Citizens Before
and During the First COVID-19 Pandemic: A Big Data
Analysis Using Mobile-Tracking GPS Data in Kanazawa, Japan .. . . . 257
Yusuke Ueno, Sadahisa Kato, Tomoka Mase, Yoji Funamoto,
and Keiichi Hasegawa
Contents xi

17 Land Use Planning as a Green Infrastructure in a Rural

Japanese Depopulated Town . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 271
K. Watanabe and K. Ishida
18 Toward an Equitable Distribution of Urban Green Spaces
for People and Landscapes: An Opportunity for Portland’s
Green Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 289
Vivek Shandas and Dana Hellman

Part VI Coast and Estuary Ecosystem

19 Effectiveness and Sustainability of Coastal Hybrid
Infrastructures for Low-Frequency Large-Scale Disasters:
A Case Study of Coastal Disaster Assessment for a
Complex Disaster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 305
Ryoichi Yamanaka and Kosuke Nakagawa
20 Challenging a Hybrid Between Green and Gray
Infrastructure: Coastal Sand-Covered Embankments . . . . . . . . . . . . . . . . . 323
Hajime Matsushima and Xiangmei Zhong
21 Green Infrastructures in Megacity Jakarta: Current Status
and Possibilities of Mangroves for Flood Damage Mitigation . . . . . . . . . 335
Yukichika Kawata
22 Implementation of Japanese Blue Carbon Offset Crediting
Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 353
Tomohiro Kuwae, Satoru Yoshihara, Fujiyo Suehiro,
and Yoshihisa Sugimura

Part VII Economic Evaluation

23 Understanding Preference Differences Among Individuals
for the Reduction in Flood Risk by Green Infrastructure . . . . . . . . . . . . . 381
Yasushi Shoji, Takahiro Tsuge, and Ayumi Onuma
24 Assessing Public Preference for Construction of Giant
Seawalls Using the Best –Worst Scaling Approach .. . . . . . . . . . . . . . . . . . . . 401
Takahiro Tsuge, Yasushi Shoji, and Ayumi Onuma
25 Coastal Community Preferences of Gray, Green, and Hybrid
Infrastructure Against Tsunamis: A Case Study of Japan . . . . . . . . . . . . 415
Yui Omori, Koichi Kuriyama, Takahiro Tsuge, Ayumi Onuma,
and Yasushi Shoji
26 Forest Green Infrastructure and the Carbon Storage
and Substitution Benefits of Harvested Wood Products .. . . . . . . . . . . . . . . 443
Gregory Valatin
xii Contents

Part VIII Governance

27 Social System in Collaborative Activities for Conserving
Coastal Pine Forest in Karatsu City, Kyusyu, Japan . . . . . . . . . . . . . . . . . . . 459
Fumika Asanami and Mahito Kamada
28 Governance for Realizing Multifunctional Floodplain:
Flood Control, Agriculture, and Biodiversity in Yolo Bypass
Wildlife Area, California, USA . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 471
Mahito Kamada, Jun Nishihiro, and Futoshi Nakamura
29 Analysis of the Description of the Multifunctionality
of Farmland in the Administrative Plans of Local Municipalities . . . . 487
Yosuke Masuda, Takashi Oka, Erika Yoshinari, Takaaki Nishida,
and Tadashi Ikeda

Index . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 503
Chapter 1
Introduction

Futoshi Nakamura

Abstract Worldwide, Green Infrastructure (GI) has mainly been discussed from an
adaptation strategy perspective in cities and urban areas. However, we believe that
GI can also function in rural and suburban areas where depopulation is prominent.
From 2015 to 2021, my colleagues and I have launched two projects, titled “Green
Infrastructure with a Declining Population and Changing Climate: Assessment of
Biodiversity, Disaster Prevention, and Social Values” and “Complementary Role of
Green and Gray Infrastructures: Evaluation from Disaster Prevention, Environment,
and Social and Economic Benefit,” which were supported by the Environment
Research and Technology Development Funds (4-1504 and 4-1805) of the Ministry
of the Environment of Japan. This volume introduces some of our achievements
in the projects. Additionally, I invited active foreign scientists from the United
Kingdom and the United States to contribute their experiences and knowledge to
this volume. As suggested by the studies, one of the important characteristics of
GI is multifunctionality, which maintains biodiversity and traditional landscapes.
Using a natural and seminatural GI network in a watershed, we are able to adapt to
elevated disaster risks in a changing climate while sustaining traditional land use and
restoring natural ecosystems that provide a suite of ecosystem services and human
welfare.

Keywords Hybrid infrastructure · Flood control · Climate change · Ecosystem

services · Biodiversity · Multifunctionality

The global average air temperature has been increasing over the long term; since the
1890s, it has risen at a rate of 0.72 ◦ C per 100 years (Ministry of the Environment
et al. 2018). The IPCC (2013) showed that precipitation is different from air
temperature, revealing an increasing trend across the Earth and increases in North

F. Nakamura ()
Laboratory of Ecosystem Management, Graduate School of Agriculture, Hokkaido University,
Sapporo, Japan
e-mail: [email protected]

© The Author(s) 2022 1

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_1
2 F. Nakamura

America and Europe at the midlatitudes in the Northern Hemisphere since the
1900s. In Japan, the fluctuation of yearly precipitation has increased since the 1970s,
and the frequency of hourly heavy rains of 50 mm or more has also increased
(Ministry of the Environment et al. 2018). The climate projections for the twenty-
first century in Japan indicate that the mean precipitation may increase by more than
10% (Kimoto et al. 2005), and other projections predict an increase in the frequency
of high-magnitude floods and a reduced discharge from snowmelt floods.
The Japanese archipelago frequently suffers from not only climate disasters but
also geological disasters, such as earthquakes and volcanic eruptions. In 2011,
a large-magnitude tsunami hit Tohoku district following the Great East Japan
Earthquake in 2011 and caused more than 30,000 people to die or go missing. In
2018, an earthquake struck the Iburi district in southern Hokkaido and triggered
numerous landslides, reaching 44 km2 in total area, killing 43 people and injuring
782 people. After the Great East Japan Earthquake, Japan enacted a basic law
for National Resilience Contributing to Preventing and Mitigating Disasters for
Developing Resilience in the Lives of the Citizenry in 2013. Since then, fundamental
plans and action programs for national resilience have been established at the
national, prefectural, and local government levels.
The world’s population growth rate is slowly falling, and the population is
projected to level off or decrease before the end of this century (United Nations
2019). In particular, in at least 55 out of 235 countries or areas, including Japan,
populations are predicted to decline between 2019 and 2050. The United Nations
(2014) has provided estimates showing that population decline has already been
occurring in Germany since 2005 and in Italy and Japan since 2010. In a drastically
depopulating society such as that in Japan, it will be more difficult to maintain
engineered gray infrastructure with limited tax income.
Considering the above-described circumstantial background together with the
importance of biodiversity conservation and the United Nations Sustainable Devel-
opment Goals in Japanese society, national policies for disaster risk reduction
have started to change from gray measures using artificially structured facilities
such as dams and dikes to more environmentally friendly measures such as green
infrastructure (GI) and nature-based solutions. In 2014, the former Prime Minister
Shinzo Abe announced that the “application of the concept of green infrastructure
using our rich natural environment is socially and economically effective and
of great importance. We should preserve the natural environment and ecosystem
services for future generations and use their functions for disaster risk reduction.”
Japan has been developing traditional measures for disaster prevention since the
sixteenth century. These are nature-friendly technologies and are recognized as GI
from the present perspective (Nakamura Chap. 2). Unfortunately, the significance
and necessity of these measures have been forgotten since modern technologies
were introduced from European countries in the Meiji era (1868–1912). However,
these modern technologies are vulnerable to extraordinary events such as mega-
tsunamis and floods. Thus, we need to learn more about the wisdom and philosophy
of traditional knowledge and technology. This volume includes those contents.
1 Introduction 3

Worldwide, GI has mainly been discussed from an adaptation strategy perspec-

tive in cities and urban areas (e.g., Gill et al. 2007; Keeley et al. 2013; Netusil
et al. 2014). However, we believe that GI can also function in rural and suburban
areas where depopulation is prominent (Nakamura et al. 2020). Moreover, to protect
cities, which are generally situated at lower, downstream elevations, we should
explore the preservation and restoration of forest GI in headwater basins and
floodplain wetland GI along rivers from a catchment perspective. Additionally,
disaster risk reduction by a hybrid of green and gray infrastructure has been
examined for stormwater, floods, and coastal flooding (Keeley et al. 2013; Sutton-
Grier et al. 2015; Zelner et al. 2016), but very few studies have quantitatively
examined flood risk, biodiversity, and socioeconomic benefit by defining existing
GI (e.g., forest and wetland in a catchment) and additional layered GI (e.g., flood
control basin along a river).
Moreover, farmlands, especially paddy fields, are one of the prevalent land
uses in Southeast Asian countries, including Japan, which has an Asian monsoon
climate, and play vital roles in providing various ecosystem services, such as
biodiversity conservation and rain and floodwater retention, in addition to rice
production (Natuhara 2013). We interpreted this type of seminatural environment
as GI and evaluated its functions in this volume. Recently, some of these farmlands
have been abandoned in the depopulating society. These abandoned farmlands may
lose ecological and hydrological functions or may succeed in quasi-original natural
environments after abandonment where various ecosystem services are provided
(biodiversity, water retention, water quality, and recreation) (Queiroz et al. 2014,
Koshida and Katayama 2018, Hanioka et al. 2018).
From 2015 to 2021, my colleagues and I have launched two projects, titled
“Green Infrastructure with a Declining Population and Changing Climate: Assess-
ment of Biodiversity, Disaster Prevention, and Social Values” and “Complementary
Role of Green and Gray Infrastructures: Evaluation from Disaster Prevention,
Environment, and Social and Economic Benefit,” which were supported by the
Environment Research and Technology Development Funds (4-1504 and 4-1805)
of the Ministry of the Environment of Japan. This volume introduces some of
our achievements in the projects. Additionally, I invited active foreign scientists
from the United Kingdom and the United States to contribute their experiences and
knowledge to this volume. The chapters are summarized below.
In Part I, the concept, history, theoretical approach, and practical model of green
and hybrid infrastructure (i.e., the combination of green and gray infrastructure)
are introduced. Nakamura (Chap. 2) presented a conceptual model of GI and gray
infrastructure based on the model introduced by Onuma and Tsuge (2018) and then
developed a hybrid model by combining these two models. He also introduced
historical GI for floodwater management in Japan, which still provides important
insights for current river management in a changing climate. Onuma (Chap. 3)
developed an optimal hybrid model by economically maximizing social net benefits.
Taki (Chap. 4) introduced one of the most advanced flood management policies
in Shiga Prefecture, Japan, implementing GI at the watershed scale. Osawa and
Nishida (Chap. 5) classified the types of GI (natural, seminatural, and artificial) and
presented the principle to evaluate the implementation potential of GI.
4 F. Nakamura

In Part II, the forest ecosystem as a GI is the focus. Forests cover 67% of the
land area in Japan, and 40% of the forests are plantations. There is a long history
of forest studies examining the effects of forests on hydrological cycles, including
rainwater storage and floodwater discharge attenuation. One of the current major
concerns in Japan is the abandonment of artificial forest management, which results
in high-density and thin-diameter trees. Unmanaged plantations are very vulnerable
to windthrow and landslides; therefore, they may not be able to sustain forest GI
functions. Nakamura (Chap. 6) focused on riparian forests as an interactive zone of
green and blue infrastructure and discussed adaptation strategies to climate change
using riparian forest GI. Tamura (Chap. 7) examined the effect of forest management
on water discharge using a runoff model, focusing on the evaporation of intercepted
rainfall and water storage in forest soil. Nisbet et al. (Chap. 8) provided guidance on
designing appropriate and cost-effective forests for water payment schemes in the
United Kingdom that support tree planting and forest management to protect and
improve water quality.
In Part III, river and floodplain GI, including paddy fields and other farmlands,
are the focus. Muto and Yokokawa (Chap. 9) built a hydraulic simulation model
that is able to calculate both surface water flooding and river flooding, and functions
for reducing flood risks by paddy fields with proper land-use management were
evaluated. Imai et al. (Chap. 10) examined the negative impact of floodwater
retention function after the abandonment of a paddy field. Osawa et al. (Chap. 11)
reviewed the effects of the consolidation and abandonment of paddy fields in recent
years on ecosystem services represented by habitat provision and regulating services
(i.e., flood control) besides rice production.
In Part IV, the flood control basin (FCB) in an agricultural landscape that provides
habitats for wetland flora and fauna was studied with special reference to biological
conservation. In the past, flood control dams and diversion channels were dominant
measures to prevent flood disasters. However, these engineered gray infrastructures
have a detrimental influence on river and floodplain biota and have recently tended
to be avoided by managers and practitioners. In contrast, FCB can be regarded as
GI, functioning by attenuating peak discharge during a flood and providing wetland
environment for a wide array of plants and animals and recreational opportunities
for people during ordinary times. Ishiyama et al. (Chap. 12) introduced a case study
in the Chitose River, Japan, which features a network of FCBs, regarding how FCBs
and their networks contribute to the regional species biodiversity of various taxa.
Morimoto et al. (Chap. 13) studied the succession of wetland vegetation in the FCB
and proposed management practices to enhance the species diversity of wetland
vegetation. Nishihiro et al. (Chap. 14) highlighted the need for human intervention
and activities to maintain the FCB function of biodiversity.
In Part V, GI adaptation strategies in cities and urban areas are the focus. Cities
and urban areas are vulnerable to heavy rains associated with climate change.
Roads, buildings, and parking lots are paved and covered by impermeable surfaces,
which increases surface runoff, leading to poor water quality and elevated peak
discharge in urban streams. Additionally, most urban residents are eager to relax
in green space after their hard official work. In this regard, Fukuoka (Chap. 15)
1 Introduction 5

introduced GI projects in various countries from the site scale to the urban land-use
scale and discussed GI visions and frameworks that are needed to provide broader
perspectives, ranging from urban heat mitigation and water disaster reduction to
healthy and walkable cities. Interestingly, Ueno et al. (Chap. 16) paid attention
to the use of GI before and during the COVID-19 pandemic in Japan and found
that GI plays a role in maintaining health and refreshment during the pandemic.
Watanabe and Ishida (Chap. 17) proposed a comprehensive land-use plan combined
with GI, considering flood risk reduction in a depopulated local town on Shikoku
Island, Japan. Finally, a case study of Portland, Oregon, which has a long history
of implementing GI, was introduced by Shandas and Hellman (Chap. 18). They
discussed the potential for activating a “green grid” in Portland that may help
alleviate ongoing socioeconomic disparities.
In Part VI, GI and hybrid infrastructures in coastal and estuary ecosystems are
studied. Since the Great East Japan Earthquake in 2011 and the subsequent tsunami
disaster, disaster risk reduction in coastal zones has been a major nationwide
concern for Japanese people (Suppasri et al. 2013). In addition to tsunamis triggered
by earthquakes, high tidal waves associated with the rising sea level in the changing
climate and their combinations are also anticipated disasters that Japan will certainly
face in the future. Yamanaka and Nakagawa (Chap. 19) examined the effects of
hybrid infrastructure consisting of a seashore, coastal embankment, coastal forest,
and dunes on the spread of tsunamis and/or tidal waves in combination with the
rising sea level. Matsushima and Zhong (Chap. 20) examined the effects of sand
coverage on seawall slopes with the aid of local citizens on vegetation establishment.
Kawata (Chap. 21) recognized mangrove forests in Jakarta as a GI mitigating
flood damage and noticed floating garbage problems that may hinder the growth
and regeneration of mangrove forests. Kuwae et al. (Chap. 22) summarized the
current status of “blue carbon (carbon captured by marine organisms)” initiatives
and reviewed three carbon offset projects in Japan.
In Part VII, public preference and willingness to pay (WTP) regarding GI and
hybrid infrastructure are investigated. Shoji et al. (Chap. 23) found heterogeneous
responses of the general public depending on their background knowledge of GI.
Tsuge et al. (Chap. 24) conducted a survey in areas where giant seawalls were
constructed after the Great East Japan Earthquake and found that citizens were
strongly concerned about the negative impact of higher seawalls on the natural
environment. Omori et al. (Chap. 25) quantified the economic value of coastal
ecosystem services, including species richness, landscape, recreational services,
and disaster risk reduction. They found that a hybrid infrastructure (seawalls +
coastal forests) received higher positive responses. Valatin (Chap. 26) evaluated the
climate change mitigation benefit of carbon storage in wood products and of carbon
substitution associated with the use of wood instead of more fossil energy-intensive
materials in the United Kingdom.
In Part VIII, governance systems to maintain and manage GI by various sectors
and their collaborations are introduced. Asanami and Kamada (Chap. 27) studied
the key role of the NPO group in maintaining collaborative activities for restoring
and conserving pine forests in Kyushu, Japan. Kamada et al. (Chap. 28) introduced
6 F. Nakamura

the Yolo Bypass Wildlife Area in the United States as a good example of GI
governance through the collaboration of various sectors, such as federal, state,
and local governments, landowners, NGOs, and citizens. Masuda et al. (Chap. 29)
investigated the administrative plans of local governments across Japan to determine
whether the plans contain the multifunctional features of farmland GI. They
suggested effective strategies to implement GI, considering population, financial
strength, extent of farmland and abandoned farmland, and flood risks.
As suggested by the above studies, one of the important characteristics of
GI is multifunctionality, which maintains biodiversity and traditional landscapes.
Using a natural and seminatural GI network in a watershed, we are able to adapt
to elevated disaster risks in a changing climate while sustaining traditional land
use and restoring natural ecosystems that provide a suite of ecosystem services
and human welfare (Nakamura et al. 2020). In contrast, if we heavily depend on
engineered gray infrastructure for disaster risk reduction, it may ruin biodiversity
and traditional landscapes, as experienced with the seawall construction after the
Great East Japan Earthquake. Even if the seawall guarantees safety against tsunamis
in a return period of once every several decades to centuries, the local people who
lost their houses to the tsunami would not return to live in a hometown where the
original landscapes were destroyed by seawalls. The populations of Iwate, Miyagi,
and Fukushima prefectures, where the tsunami disasters had prevailed in 2011,
have rapidly decreased by approximately 15% between 2011 and 2020. Without
considering an appropriate balance between gray and green infrastructures, we may
create unacceptable land-use recovery plans for residents after such disasters. A
land-use plan for adapting to climate change should be devised to allocate GI and
gray infrastructure to function complementarily and comprehensively and to nurture
local and regional landscapes in which GI contributes to an improved quality of life.

References

Gill SE, Handley JF, Ennos AR, Pauleit S (2007) Adapting cities for climate change: the role of
the green infrastructure. Built Environ 33:115–133
Hanioka M, Yamaura Y, Senzaki M, Yamanaka S, Kawamura K, Nakamura F (2018) Assessing the
landscape-dependent restoration potential of abandoned farmland using a hierarchical model of
bird communities. Agric Ecosyst Environ 265:217–225
IPCC (2013) Climate change 2013: the physical science basis. In: Stocker TF, Qin D, Plattner
GK, Tignor M, Allen SK, Boschung J et al (eds) Contribution of working group I to the fifth
assessment report of the intergovernmental panel on climate change. Cambridge University
Press, Cambridge, p 1535
Keeley M, Koburger A, Dolowitz DP, Medearis D, Nickel D, Shuster W (2013) Perspectives on the
use of green infrastructure for stormwater management in Cleveland and Milwaukee. Environ
Manag 51:1093–1108
Kimoto M, Yasutomi N, Yokoyama C, Emori S (2005) Projected changes in precipitation
characteristics around Japan under the global warming. SOLA 1:085–088. https://doi.org/
10.2151/sola.2005-023
1 Introduction 7

Koshida C, Katayama N (2018) Meta-analysis of the effects of rice-field abandonment on

biodiversity in Japan. Conserv Biol 32(6):1392–1402
Ministry of the Environment, Ministry of Education, Culture, Sports, Science and Technology,
Ministry of Agriculture, Forestry and Fisheries, Ministry of Land, Infrastructure, Transport and
Tourism, Meteorological Agency of Japan (2018) Kikou henndou no kannsoku yosoku eikyou
hyouka nikannsuru tougou report 2018 nihonn no kikouhenndou to sono eikyou [Consolidated
report on observations, projections and impact assessments of climate change: climate change
and its impacts in Japan]. http://www.env.go.jp/earth/tekiou/report2018_full.pdf
Nakamura F, Ishiyama N, Yamanaka S, Higa M, Akasaka T, Kobayashi Y, Ono S, Fuke N,
Kitazawa M, Morimoto J, Shoji Y (2020) Adaptation to climate change and conservation of
biodiversity using green infrastructure. River Res Appl 36:921–933
Natuhara Y (2013) Ecosystem services by paddy fields as substitutes of natural wetlands in Japan.
Ecol Eng 56:97–106
Netusil NR, Levin Z, Shandas V, Hart T (2014) Valuing green infrastructure in Portland, Oregon.
Landsc Urban Plan 124:14–21
Onuma A, Tsuge T (2018) Comparing green infrastructure as ecosystem-based disaster risk
reduction with gray infrastructure in terms of costs and benefits under uncertainty: a theoretical
approach. Int J Disaster Risk Reduct 32:22–28
Queiroz C, Beilin R, Folke C, Lindborg R (2014) Farmland abandonment: threat or opportunity
for biodiversity conservation? A global review. Front Ecol Environ 12(5):288–296
Suppasri A, Shuto N, Imamura F, Koshimura S, Mas E, Yalciner AC (2013) Lessons learned from
the 2011 Great East Japan tsunami: performance of tsunami countermeasures, coastal buildings,
and tsunami evacuation in Japan. Pure Appl Geophys 170(6):993–1018
Sutton-Grier AE, Wowk K, Bamford H (2015) Future of our coasts: the potential for natural and
hybrid infrastructure to enhance the resilience of our coastal communities, economies and
ecosystems. Environ Sci Pol 51:137–148
United Nations (2014) World population prospects: the 2012 revision. ESA/P/WP.235. United
Nations, New York, NY
United Nations (2019) World population prospects 2019: data booket. ST/ESA/SER.A/424. United
Nations, New York, NY
Zelner M, Massey D, Minor E, Gonzales-Meler M (2016) Exploring the effects of green
infrastructure placement on neighborhood-level flooding via spatially explicit simulations.
Comput Environ Urban Syst 59:116–128

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
 Part I
Concept and Synthesis
Chapter 2
Concept and Application of Green
and Hybrid Infrastructure

Futoshi Nakamura

Abstract Recently, Japan has suffered extraordinary damage from typhoons, heavy
rains, and megafloods, each of which has exceeded the upper limit of control by
managed infrastructure for flood mitigation. First, I present a conceptual frame-
work of hybrid infrastructure at the watershed scale, combining (1) fundamental
green infrastructure (GI), composed of forests and wetlands in the watershed; (2)
additional multilevel GI, such as flood control basins; and (3) existing engineered
disaster prevention infrastructure, such as dams and artificial levees. Second, I
introduce the disaster risk reduction function of natural forests and wetlands
and three representative traditional flood control measures: discontinuous levees,
overflow embankments, and flood protection forests. This GI should be properly
allocated and maintained at the watershed scale to reduce damage by megafloods.
The multiple types and functions of GI may provide essential habitats for wildlife
and recreational opportunities for local residents and others. Finally, I address key
points for planning, implementation, and governance of GI at the watershed scale.

Keywords Hybrid infrastructure · Flood control · Climate change · Ecosystem

services · Biodiversity · Ecological network

2.1 Introduction

Flood protection strategies using current gray infrastructure, such as dams and
artificial levees, face limits and are not able to protect human lives and assets from
megafloods (e.g., exceeding 100-year return interval) exacerbated under climate
change. Here, I define gray infrastructure (GYI) as engineering infrastructure
associated with water resources, such as water and wastewater treatment plants,
pipelines, and reservoirs. Recently, Japan has been experiencing large flood disasters

F. Nakamura ()
Laboratory of Ecosystem Management, Graduate School of Agriculture, Hokkaido University,
Sapporo, Japan
e-mail: [email protected]

© The Author(s) 2022 11

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_2
12 F. Nakamura

Fig. 2.1 Flooding of the Abukuma River by Typhoon Hagibis, Sukagawa city, central Japan
(Ministry of Land, Infrastructure, Transport and Tourism, Japan)

every year associated with typhoons and seasonal rains, which have exceeded the
river discharge that the infrastructure was designed to deal with (hereafter I call it
“design flood discharge”). Therefore, floodwater spilled over the artificial levees and
inundated farmlands and residential areas (Fig. 2.1).
There are two alternatives to respond to the increasing floodwater discharge: one
is increasing the design flood discharge by conventional measures, such as construc-
tion of dams, excavation of channel sediment, setback of levees, and enhancement of
levee height, and the other is keeping the design flood discharge at the current level
while providing green infrastructure (GI) outside of the levees including forests
on the hillside slopes and wetlands on floodplains to store rainwater and retain
spilled floodwater there. Here, GI can be defined as “a strategically planned network
of natural and semi-natural areas with other environmental features designed and
managed to deliver a wide range of ecosystem services such as water purification,
air quality, space for recreation and climate mitigation and adaptation” (European
Commission 2016) (Fig. 2.2).
The revision of design flood discharge to a higher level is not easy. Reservoir
dams have been built without considering future climate changes in design flood dis-
charge. Floodwater control by levees and channelization started from downstream
reaches (urban areas) where densely populated cities and towns are located and
progressed to upper reaches (rural areas) where farmlands are developed. However,
there are many rural areas in Japan where flood control measures have not yet been
completed. Thus, if we raise the design flood discharge in response to climate
change, we must restart flood control management again from the downstream
reaches and leave upstream rural areas unprotected for a long time. Even if we
can fulfill the flood control management corresponding to a certain higher level,
2 Concept and Application of Green and Hybrid Infrastructure 13

㻲㼛㼞㼑㼟㼠 㻳㻵
䠄㻯㼛㼚㼟㼑㼞㼢㼍㼠㼕㼛㼚 㼛㼒 㼔㼑㼍㼐㼣㼍㼠㼑㼞 㼎㼍㼟㼕㼚㼟䠅

㻲㼍㼞㼙㼘㼍㼚㼐 㻳㻵
㻔㻼㼞㼛㼢㼕㼟㼕㼛㼚 㼛㼒 㼙㼡㼘㼠㼕㼒㼡㼚㼏㼠㼕㼛㼚㼟䠅

㻾㼕㼢㼑㼞 㼍㼚㼐 㼞㼕㼜㼍㼞㼕㼍㼚 㻳㻵 Urban GI

䠄㻾㼑㼠㼑㼚㼠㼕㼛㼚 㼛㼒 㼒㼘㼛㼛㼐㼣㼍㼠㼑㼞䠅 䠄Rainwater infiltration,
Attractive urban life䠅

Coast GI
䠄Protection from tidal wave and tsunami,
and opportunity for tourism䠅

Fig. 2.2 Multi-functionality of GI and linkages at the watershed scale

intense flooding exceeding the revised management level is likely to occur under
climate change. It is not realistic to revise the design flood discharge continuously
and repeatedly.
When we explain green infrastructure (GI), we tend to situate GI as an opposing
and conflicting concept against gray infrastructure (GYI). In some sense, this is
true, but exaggerating the differences between GI and GYI may hinder productive
discussions. Rather, under climate change, these two infrastructure types should
complement each other and contribute to a comprehensive land-use plan that simul-
taneously provides amenities, convenience, and safety to local towns, cities, and
communities (Green Infrastructure Association (Japan) 2017, 2020). For example,
we consider that forest ecosystems are one of the most important types of GI and
function to regulate water discharge from a basin. In Japan, someone who is opposed
to dam construction may argue that forests are a surrogate for large dams (GYI);
therefore, dam construction is not necessary if we leave forests in upper basins. This
type of emotional argument is not productive; rather, we should manage plantation
forests to keep appropriate tree densities and soil infiltration rates to reduce the
dependence on water regulations by dams.
A suite of GI measures provide various ecosystem services across a watershed
(Fig. 2.2). Our future land-use plan should be built from the viewpoint of how
we can create a best mix of advantages provided by GI and GYI at the watershed
scale. The objective of this chapter is to devise climate change adaptation strategies
for disaster prevention and biodiversity conservation using GI in Japan. More
specifically, I will (1) introduce a conceptual framework for hybrid infrastructure
at the watershed level with concrete evidence of previous research results, (2)
14 F. Nakamura

reevaluate Japanese traditional measures from the viewpoint of climate change

adaptation, and (3) propose a best mix of GI and GYI at the watershed scale mainly
focusing on flood protection and biodiversity conservation.

2.2 Conceptual Framework for Hybrid Infrastructure

First, I would like to define gray, green, and hybrid infrastructure at local and
landscape (or watershed) scale. GYI at the local scale is represented by artificial
elements, such as dams and channelization, whereas an individual natural ecosys-
tem, such as a patch of wetland and forest, is at opposite end of the naturalness
spectrum (Fig. 2.3). A mixture of natural and artificial elements, such as gardens and
parks, is regarded as hybrid infrastructure. At the landscape (or watershed) scale,
urban areas are built by various GYI, whereas wilderness areas consist of diverse
natural ecosystems (Fig. 2.3). In this spectrum, a mosaic distribution of natural (or

Limited Multi-
functions functions

Element (local scale)

Land use (watershed or landscape scale)

Hybrid infrastructure (Gray and Green)

Fig. 2.3 Definition of green, gray, and hybrid infrastructure at local and watershed scale (adapted
from Green Infrastructure Association (Japan) ed. (2020))
2 Concept and Application of Green and Hybrid Infrastructure 15

Degree of safety
Magnitude
of disaster
Degree of safety

Magnitude
Upper limit of plan of disaster

Fig. 2.4 Conceptual framework of gray (a), green (b), and hybrid infrastructure (c). Shaded and
white areas denote the safety zones created by gray and green infrastructure (GI), respectively. The
area denoted by GI-1 is the fundamental GI, while GI-2 is an additional multilevel GI. (a) and (b)
are modified from Onuma and Tsuge (2018) (from Nakamura et al. 2020)

seminatural) and artificial ecosystems in a suburban and rural watershed can be

referred to as hybrid infrastructure.
GYI elements, such as dams and artificial levees, usually assure 100% disaster
protection until the magnitude of the disaster reaches the upper limit of the
management plan, though unexpected risks associated with structural flaws and
human errors still exist. However, once the magnitude exceeds the upper limit, the
GYI will have no additional flood mitigation function; for example, floodwater will
spill into residential areas where artificial levees are breached (see Fig. 2.1). Thus,
the safety-magnitude curve for GYI has a rectangular shape (Fig. 2.4a). In contrast,
I expect the response of GI to show a gradually decreasing trend. In addition, the
disaster prevention function of GI may be sustained longer than that of GYI (Onuma
and Tsuge 2018). However, the response curve of risk reduction against disaster
magnitude by GI at the local scale is not well studied and can vary depending on the
kind of GI (Fig. 2.4b). The uncertainty of the function is high for GI.
In the past, we have discussed the advantages and disadvantages of GYI and
GI and compared them to determine which approach was better. Sometimes, such
16 F. Nakamura

debates are not productive and promote the polarization of opinions between gray
and green approaches. Here, I discuss a combination of GYI and GI with the aim of
applying GI for disaster control in a society at high risk of various natural disasters,
as is the case in Japan. I present a hybrid, combining the two types of infrastructure
at watershed scale in Fig. 2.4c. In this conceptual diagram, GI-1 represents the
fundamental GI composed of forests and wetlands in the watershed, while GI-2 is
additional multilevel GI, such as flood control basins that function when floodwater
exceeds the design flood discharge determined by artificial levees. In this diagram,
an increase in the combined area of GYI and GI guarantees safety, even at a very
high magnitude of flooding.
How much should we expand or reduce the areas of GYI or GI to meet present
and future needs as we face climate change? We have to evaluate the effectiveness
of hybrid infrastructure in terms of disaster prevention, biodiversity protection,
and social and economic values to determine which combination is best in a
given natural and social condition. If the area of GYI is expanded, high levels of
disaster control may be achieved, but there may be losses in biodiversity and the
hometown landscape and increases in maintenance costs. Historically, Japan has
been losing forest and wetland GI-1 through overharvesting of forest resources and
land conversion from wetland to agricultural fields (Nakamura et al. 2017). As a
result, the peak flood discharge has gradually elevated, and Japanese government has
had to compensate for the water retention ability that natural ecosystems provided in
the past with GYI, such as dams and artificial levees. The combination of GYI and
GI may change depending on future land use. We may preserve or rebuild GI using
abandoned farmlands at lower cost in rural areas. In contrast, it may be difficult to
restore natural ecosystems in highly populated urban areas, and GYI therefore still
plays an important role in disaster risk reduction with a limited introduction of GI
represented by gardens and city parks.

2.3 Attenuation of Flood Peaks by Forest and Wetland

Ecosystems (Examples of GI-1 in Fig. 2.4c)

Natural forests and wetlands have the ability to reduce the disaster risk by
attenuating and delaying flood peaks. The mechanism and detailed information are
introduced in other chapters, and I will briefly explain their functions as GI-1 in Fig.
2.4c. These natural and seminatural ecosystems function to keep flood risks low by
storing rainwater on hillslopes and floodplains at a watershed scale.
In one such example, Tsukamoto (1985) investigated how water discharge
changed in association with urbanization and loss of forest cover (Fig. 2.5). From
this figure, we understand that the peak discharge increased during floods, while
the low-flow discharge decreased in the urbanized watershed. On the other hand,
during the beginning of a series of rainfall events, the peak discharges from the
forested watershed were greatly attenuated, and the low-flow discharge was slightly
2 Concept and Application of Green and Hybrid Infrastructure 17

Precipitation (mm/hr)

25/08/1974 26/08
Water discharge (mm/hr)

Urbanized basin (13 km2)

Forested basin (26 km2 )

25/08/1974 26/08
Fig. 2.5 Comparison of hydrographs between urbanized and forested basins (from Tsukamoto
1985)

increased. However, these effects were not clearly observed at the end of the
sequence of rainfall events; the degree of peak discharge attenuation was reduced,
and the low-flow discharge was greatly increased. The timing of flood peaks is
almost simultaneous with rainfall peaks in the urbanized watershed throughout the
investigated period in part because of their extensive paved surfaces. In contrast,
delays in flood peaks can be observed in forested watersheds, especially at the end
of the sequence of rainfall events. This is because most of the precipitation was
absorbed by the forest soil and retained in the soil pores at the beginning of a series
18 F. Nakamura

of rainfall events, and thereafter, precipitation added to the forest soils at the end of
rainfall events was rapidly released from macropores and transported by saturated
subsurface and surface flow. Thus, the flood risk reduction provided by forests may
vary over time during a rainfall event. However, forest ecosystems certainly function
as GI-1 and can stabilize variations in stream discharge.
Japan has traditionally recognized the functions of forest ecosystems, including
the abovementioned hydrological effects, and developed forest protection laws
starting in 1897. In contrast, in the past, wetlands have not been acknowledged
as GI, but rather considered unproductive lands. Due to the lack of knowledge of
wetland ecosystem services, wetlands have been converted to agricultural lands.
Approximately 60% of the original wetlands in Hokkaido, a northern island in
Japan, have disappeared over the last 50 years (Nakamura et al. 2005). However,
people have gradually recognized the importance of the ecosystem services that
wetlands provide, including biodiversity conservation, flood control, and improve-
ment of water quality. In general, wetlands develop in relatively flat areas where
streamwater and rainfall are retained, whereas forests become established not only
in flat areas, but also on steep terrain. However, currently remaining forests are
mainly on relatively steep land because most flat areas have been developed for other
land uses. Thus, the remaining wetlands in flat areas are highly valued and should
be retained and restored to capitalize on their capability of stormwater retention. In
contrast, forests on steep slopes, where the hydraulic gradient is high, have lower
water retention capability.
Nakamura et al. (2020) used the hydrological model GETFLOWS to simulate the
effects of the Kushiro Mire on the water discharge of the Kushiro River in Hokkaido.
The Kushiro Mire is the largest wetland complex in Japan, extending across
approximately 26,000 ha. Three typhoons accompanied by heavy rains arrived in
the Kushiro region from August to September 2016, producing three major flood
peaks. The authors used the model to calculate water discharge for the case in which
approximately 55% of wetlands were converted to residential lands. The simulated
hydrograph of the modified wetlands showed a higher peak discharge, sharper
rising and descending limbs, a 2-day earlier arrival of the peak, and a lower low-
flow discharge than those of the present situation in which wetlands exist without
55% area-based land conversions (Fig. 2.6). The peak discharges of the simulation
cases for the present and for the partial loss of wetlands were 390 and 580 m3 /s,
respectively. The results of the model simulation clearly demonstrated that the
Kushiro Mire acts as GI-1 (a large natural reservoir) that attenuates the peak
discharge during floods.
As shown above, forests and wetlands in a watershed are fundamental GI that
stabilize water discharge and substantially attenuate flood peaks. Thus, if we lose
these GI-1 features, we will face a risk of flooding and have to compensate with
more GYI. However, these GYI features lead to high maintenance costs and may
not be economically feasible in depopulating societies, such as Japan and other
countries.
2 Concept and Application of Green and Hybrid Infrastructure 19

700
(a) ほPresent
್ situation (observation)
600
Present situation
෌⌧ゎᯒィ⟬್ (simulation)
500

400

300
Water discharge (m3/sec)

200

100

0
July 1 Aug. 1 Sep. 1 Oct. 1 Nov. 1 Dec. 1 , 2016

700
(b) Partial loss
‵ཎ䛜䛺䛔ሙྜ of wetlands (simulation)
600 Present situation
‵ཎ䛜䛒䜛ሙྜ (simulation)
500

400

300

200

100

0
July 1 Aug. 1 Sep. 1 Oct. 1 Nov. 1 Dec. 1 , 2016

Fig. 2.6 Attenuation of flood peaks by the Kushiro Wetland. The fit of the simulation results to the
observational hydrograph (a) and comparison of the simulation results with and without wetland
green infrastructure (b) (from Nakamura et al. 2020)

2.4 Traditional Measures Against Large Floods (Examples

of GI-2)

Japan has developed traditional flood control measures since the sixteenth cen-
tury. These are nature-friendly technologies and are recognized as GI from the
present perspective. Unfortunately, however, the significance and necessity of
those measures for flood control have been forgotten since modern technologies
were introduced from European countries in the Meiji era (1868–1912), which
promoted channelization and artificial levee construction to convey floodwaters
rapidly from headwater basins to river mouths. However, these modern technologies
are vulnerable to megafloods, which easily exceed the design flood discharge. The
Typhoon Hagibis disaster in 2019, for example, brought a huge amount of rainfall,
and the floodwater level exceeded the upper limit of the artificial levees. As a result,
artificial levees breached at 140 locations, and the floodwater broadly spilled into
residential areas (see Fig. 2.1). The estimated flood damages were enormous.
20 F. Nakamura

Overflow
embankment

Discontinuous
Flood control levees
Receiving basin
levee floodgate

Flood protection forest

Fig. 2.7 Traditional flood control measures in Japan

Modern technologies for flood control tend to constrain floodwater in river

channels by artificial levees, while traditional flood control measures do not confront
flood power directly, but rather reduce it by the spread of the floodwaters over
surrounding land-use areas, mainly agricultural lands. Thus, some of the floodwaters
will be stored and retarded in land-use areas and slowly released when the
floodwater level gradually recedes. This idea contrasts with modern techniques that
focus on how rapidly floodwater can be discharged from basins. In the modern
flood control framework, local and national governments tend to increase the heights
of artificial levees to sustain safety in adjacent land-use areas once flood disasters
occur. However, this policy encourages people to live closer to rivers and develop
surrounding land-use areas more intensively. Under this situation, once artificial
levees are breached by a large flood, the damage is enormous.
Here, I introduce three representative traditional flood control measures: dis-
continuous levees (Kasumitei in Japanese), overflow embankments (Etsuryutei),
and flood protection forests (Suiborin) (Fig. 2.7). These measures still exist along
Japanese rivers, but are greatly reduced through replacement with modern tech-
nologies. When the Japanese government now applies traditional measures to flood
control, the intentional spillage of the floodwater into designated areas is required
for some extraordinary floods exceeding the design flood discharge. The traditional
measures including the designated areas should function as GI. In contrast, in
modern techniques, we cannot predict where continuous artificial levees will break
once the floodwater level exceeds the levee height. This may lead to serious disasters
when levee breaches occur in intensively developed and populated areas. We should
avoid this situation by learning the advantages of traditional measures and adopting
GI.
2 Concept and Application of Green and Hybrid Infrastructure 21

2.4.1 Discontinuous Levees

The discontinuous levee was devised and developed by Shingen Takeda (Fig. 2.7),
who was a military commander from 1521 to 1573. Discontinuous levees still
existed along 54 rivers among the 109 major large rivers in 1987 in Japan (Okuma
1987). The roles of discontinuous levees differ among the locations where those
measures were set. According to Okuma (1987), discontinuous levees built on high
gradient (>1/500) alluvial, gravel-bed rivers aim to return floodwaters that spill over
upstream levees and/or are generated from tributaries and inlands to the mainstem of
the river. For example, many discontinuous, overlapping levees may be constructed
over 1–2 km reaches, such as on the Tedori River, Ishikawa Prefecture (Fig. 2.8). In
this case, even when the first levee closest to the river channel is broken by a flood,
other levees situated behind the first levee will function to return the floodwater to
the main channel repeatedly. This is a form of multiple protection measures.
In contrast, discontinuous levees built on low gradient (<1/1000) alluvial, sand-
bed rivers aim to store floodwater by the backwater effect between discontinuous
levees during periods of high discharge. Due to the gentle gradient of the riverbed,
the storage capacity and retention effect of floodwater are expected to be large, and,
therefore, substantially reduce the flood peak in the mainstem river.

0 4km

Fig. 2.8 Discontinuous levees (denoted by red lines) in the Tedori River, Ishikawa Prefecture
22 F. Nakamura

In both cases, natural confluences of mainstem and tributaries are maintained

with discontinuous levees, and, thereby, the natural connectivity of the mainstem
and tributaries can be sustained; otherwise, we have to build GYI, such as floodgate
and sluice gate, to manage the discharge of tributaries into the mainstem. In general,
there is an elevational gap between the water level of the mainstem and floodgate,
and, therefore, it is difficult for fish to migrate into tributaries through the sluiceway.
The open spaces created by the discontinuous levees function as GI and have been
used for flood protection forests or agricultural lands. Moreover, the measure allows
floodwater to inundate GI repeatedly, and, therefore, provides habitats for wetland
plants and animals.
Currently, few local people and engineers understand the multiple functions of
discontinuous levees; thus, many were changed to continuous levees following the
approach of dam construction and flood control channels. In this way, functions
originally provided by GI were replaced by those of GYI. However, Japan has
recently experienced breaches of continuous levees, resulting in large flood disasters
influenced by climate change.
The use of discontinuous levees represents the needed shift in Japan from a
focus on constraining floodwater to the river channel to accepting more flow over
floodplains. This is similar to a shift in thinking evident in Europe where the
objectives are to cope with climate change and biodiversity conservation (Buijse
et al. 2002). River managers in Japan and other developed countries should learn
more about the advantages of traditional GI measures and keep them in our river
landscape. Rather than the replacement of GI by GYI, we should consider the best
mix of the two measures.

2.4.2 Overflow Embankment

The overflow embankment was devised and developed by Kiyomasa Kato, who
was a military commander from 1562 to 1611. This measure determines in
advance where the floodwater overflows the artificial levees (Fig. 2.7). Together
with an overflow embankment, a secondary levee was constructed to receive the
floodwaters, and the areas surrounded by the overflow embankment and receiving
levees can function as GI-2.
Unfortunately, these traditional measures also tended to be removed after
the introduction of modern technologies, such as strong embankment and road
construction, and areas of GI have been developed as farmland and/or residential
areas (Tanabe and Okuma 2001). However, most of the current artificial levees in
Japan have not been equipped with overflow embankment system; therefore, we do
not know where the floodwater will overflow the levees, which results in increased
damage and delayed evacuation of residents. Thus, once the levees are partially
breached by overbank flow, the exposure risk to flood damage increases. According
2 Concept and Application of Green and Hybrid Infrastructure 23

to Nakajima et al. (2013), who conducted hydraulic simulations to evaluate the

flood control effects of overflow embankments and discontinuous levees in the
Jobarugawa River in Kyushu, the flood level in the mainstem did not change with
increasing water discharge when floodwater overflowed these facilities.

2.4.3 Flood Protection Forest

Flood protection forests are one of 17 kinds of protection forests prescribed by the
Japan Forest Law established in 1897. Flood protection forest is defined as riparian
forests (including bamboo forest) located within, on, or outside of the levees (Fig.
2.7). Forests function to reduce damage to residential areas and farmlands behind
levees during floods (Matsuura et al. 1988).
The root systems of trees and bamboo hold soils on levees and reduce flood
velocity, thereby protecting riverbanks and levees from scouring and facilitating
sediment deposition and driftwood retention. The floodwater running through pro-
tection forests is filtered and slowed, and, therefore, even if it inundates residential
areas, the damage to houses and farmlands is greatly reduced. On the other hand,
there is an argument that flood protection forests may add roughness elements to
water flow, which may increase floodwater depth and increase the flooding risk.
Moreover, they may not be able to withstand the flood power and may become
a source of driftwood, which may increase the risks of wood-laden flooding. Also,
floated large wood may damage riparian forest. In this situation, river managers may
decide to cut flood protection forests. There are many uncertainties regarding the
physical functions of flood protection forests and trade-offs to balance, and therefore
further research is required.
To maintain the traditional GI, the understanding and cooperation of local
communities and river managers are essential. Local residents who inhabit areas
of GI may suffer flood damage repeatedly, and some watershed communities and
government agencies should support them and compensate for this damage. In
general, these GI features are situated in rural areas in the upstream and midstream
reaches of the mainstream and tributaries. Thus, inundation by floodwater in
upstream and midstream reaches certainly slows the arrival time of flood peaks
and reduces the flood discharge in downstream reaches where cities and towns are
generally situated. In this situation, people in urban areas should financially offset
the damages in rural areas, and GI should be managed and maintained through the
cooperative efforts of communities over the whole watershed. The education of river
managers is also important. Young river managers may not know the functions and
advantages of traditional GI in river management. We should publicly explain how
these measures are applicable as an adaptation strategy for climate change.
24 F. Nakamura

2.5 Hybrid Infrastructure at Watershed Scale

The Japanese government has introduced new flood control management policies
twice using hybrid (green/gray) infrastructure. There are some differences between
these two management policies and the hybrid infrastructure concept in the present
paper; the former strongly focuses on flood control measures by GYI, while the
latter stresses GI having multi-functionalities including biodiversity conservation at
a watershed scale. The first is called “Sogo-chisusui,” which means comprehensive
flood control measures. The measures focus on flood control in urbanized areas
where the flood peaks were elevated as a result of land-use development in the
1970s, as shown in Fig. 2.5. During rapid economic growth in the 1960s and 1970s,
asphalt pavement covered roads in urban areas and buildings were constructed.
Streams in urban areas were channelized, and sewer systems were installed. Most of
these facilities are impermeable and drained rainwater into sewer systems without
allowing the water to penetrate into soils. The water flowing through drainage pipes
reaches the straightened stream channel. Thus, the current water management policy
aimed to flush out rainwater as fast as possible. As a result, flood peaks quickly
increased, even if the amount of precipitation did not change (Fig. 2.5).
To control elevated streamflows in urban areas, the government had to change the
policy from flushing rainwaters rapidly to retaining rainwaters in GI, such as woods,
parks, gardens, and ponds, and in GYI, such as stormwater retention facilities
underneath buildings. Then, the water is released slowly into the groundwater
system. Additionally, the stream channels were partially widened, and sidewalk and
parking pavement was changed to permeable materials. However, some of the recent
stormwater retention facilities were built as GYI, which aims only at flood control
without considering other functions, such as the provision of habitats for wildlife
and recreational opportunities. Parks and schoolyards were built to function as
stormwater retention facilities whose grounds were excavated and lowered relative
to the surrounding areas.
After the introduction of the above flood control policy called “Sogo-chisusui,”
natural restoration or rehabilitation projects began in the 1990s, reflecting the world-
wide movement to consider the protection of biodiversity and natural landscapes.
These projects replaced concrete disaster prevention measures in streams, such
as groundsill and revetment, with measures composed of stones which are more
natural materials than concrete. The stream channel was partially broadened, and
the vertical side slopes of the streambank were reformed to gentle slopes with some
tree plantations. In the 1960s and 1970s, Japanese rivers and streams have been
channelized and straightened for urban and agricultural land-use development and
flood control. However, in the 1990s to 2000s, some of the straighten rivers are
restored to original meandering rivers and wetlands by restoration projects. The
river restoration projects function as an adaptation strategy to spread floodwater
over floodplains and reduce the peak discharge of floods (Nakamura et al. 2014).
In recent years, Japan has experienced megaflood disasters due to climate change.
A megaflood can easily exceed the design flood discharge of GYI (artificial levees
2 Concept and Application of Green and Hybrid Infrastructure 25

and dams). We experienced serious flood damage in broad regions of eastern Japan
during Typhoon Hagibis in 2019. In 2020, the Japanese government announced the
need for “Ryuiki chisui,” which is a new flood control concept at the watershed
scale. This policy focuses on entire watershed areas from headwater basins covered
by forests to midstream areas covered by farmlands to downstream areas where
cities are situated. The policy encouraged active use of GI measures at the
watershed scale. Natural forests in headwater basins (GI-1) should be preserved, and
abandoned artificial forests should be managed by periodic thinning and pruning.
Discontinuous levees, overflow embankments, and flood protection forests should
be preserved, and land-use development in flood-prone areas should be restricted in
midstream reaches (GI-2). Stormwater retention facilities and permeable pavement
should be installed to compensate for the increase in water discharge volume caused
by new developments.
This new policy matches the hybrid infrastructure concept presented in this
chapter, but focuses on flood control with little emphasis on the multi-functionality
and ecological network of GI as shown in Fig. 2.2. The Japanese government
is still planning to build and/or reinforce GYI to adapt to climate change, and,
therefore, it is unlikely that river managers will actively introduce GI at a watershed
scale. One of the reasons why GI is not easily introduced is sectionalism between
ministries and agencies. For example, the collaborative and coordinated actions
between the Ministry of Land, Infrastructure, Transport and Tourism and the
Ministry of Agriculture, Forestry and Fisheries of Japan is essential for flood
control at a watershed scale, but it is hard to achieve due to strong divisions and
cultural differences between administrative agencies. Another problem is the greater
uncertainty and ambiguity of GI function for rainwater retention and floodwater
control as compared to GYI. Thus, development of simulation models evaluating
GI function quantitatively is required to find a best mix of GI and GYI for future
river management.

2.6 Green Infrastructure Functioning as Ecological

Networks

One of the important characteristics of GI is an interconnected network of green

spaces (Maes et al. 2015). Dispersal and breeding success of red-crowned crane
(Grus japonensis) using GI habitat provides an excellent example (Fig. 2.9). The
red-crowned crane was distributed widely on the Japanese islands of Hokkaido,
Honshu, and Kyushu at the beginning of the twentieth century. The population
significantly declined to approximately 40 individuals with extensive farmland
development of wetlands and overhunting (Masatomi and Masatomi 2018) and
then recovered to approximately 1500 individuals with the help of artificial feeding
during the winter season by local residents. The breeding and nesting sites of the
cranes are concentrated in the Kushiro Mire (an example of GI-1 in Sect. 2.3), where
26 F. Nakamura

Fig. 2.9 Pair of red-crowned cranes that arrived in the Maizuru flood control basin in August
2012. The image in the upper right corner is an aerial photo of the entire basin

food and habitat limit the carrying capacity. The genetic diversity of the current
population is low owing to the bottleneck effect (Masatomi and Masatomi 2018;
Miura et al. 2013). Thus, the Ministry of the Environment plans to expand their
nesting and breeding sites outside of the Kushiro region.
In the Chitose River watershed, six large-scale, flood control basins with areas
of 150–280 ha were constructed to control extraordinary floods (acting as GI-2).
These basins have also created a high-quality wetland landscape where many swans,
greater white-fronted geese, and bean geese gather in the early spring. I, together
with the Ecosystem Conservation Society of Japan, proposed restoring one of the
flood control basins to wetlands dominated by a reed (Phragmites australis) that is
preferred by the crane for nesting. A study that compared the abundance of different
organisms (fish, aquatic insects, birds, and aquatic plants) among various water
bodies (natural ponds, artificial channels, etc.) revealed that their abundance was
highest in the flood control basins (see Chap. 12, Yamanaka et al. 2020). Thus, flood
control basins have the potential to act as ecological networks for the conservation
of the red-crowned crane on a large scale, employing the Kushiro Wetland as
a hub wetland (GI-1) and flood control basins as satellite wetlands (GI-2). The
GI networks promote migrations of cranes to extended areas, and add population
stability in case of disaster, which may damage some habitats. In 2012, a pair of
red-crowned cranes appeared in the Maizuru flood control basin, which is one of the
six basins in the Chitose River watershed (Fig. 2.9). Since then, one or two pairs of
2 Concept and Application of Green and Hybrid Infrastructure 27

Fig. 2.10 Juvenile bird born in the Maizuru flood control basin in 2020

cranes have appeared in the Maizuru flood control basin and neighboring farmlands
every year and finally overwintered, and one chick hatched in the Maizuru flood
control basin in 2020 (Fig. 2.10).

2.7 Important Points for Planning, Implementation,

and Governance of GI

Here are three important points for the planning and implementation of GI at the
watershed scale.
1. Forests in headwater basins, the stream network (main channel and tributaries)
and farmlands in midstream reaches, and green spaces in cities should be
properly allocated to exploit the multiple functions of GI (e.g., flood control,
biodiversity conservation, and recreation) together with existing GYI measures
(Fig. 2.2).
We should preserve natural and quasi-natural forests and wetlands still
remaining in a watershed. The present natural ecosystems already perform
ecosystem-based disaster risk reduction (Eco-DRR) (Figs. 2.5 and 2.6). Once
we lose them, it takes enormous resources and time to restore them, and the
restored ecosystems may never perform the multiple functions provided by the
original ecosystems. The existing green parks and gardens should be preserved
to maintain the retention capability for stormwater. In developed countries,
abandoned tree plantations and farmlands will increase with depopulation in
28 F. Nakamura

the future (Kobayashi et al. 2020). The passive and active restoration of natural
forests, grasslands, and wetlands after abandonment is one of the challenging
options to maintain the functions of GI (Morimoto et al. 2017).
The confluences of the mainstream and tributary rivers are key areas for
the installation of GI. Approximately 80% of levee breaches by the 2019
flood disaster associated with Typhoon Hagibis occurred within 1 km of the
confluences of the mainstream and tributaries. Levee breach sites at confluences
accounted for 112 (62 rivers) out of 140 (71 rivers) breach sites in total (Asahi
Newspaper, November 8 in 2019). Confluences are vulnerable to floodwaters
and provide diverse habitats for various plants and animals (Benda et al. 2004).
Thus, the riverbed should be wide, and traditional flood control measures, such
as discontinuous levees, overflow embankments, and flood protection forests,
should be used at appropriate locations.
2. One of the most important characteristics of GI is multi-functionality. When GI
is used for Eco-DRR, the ecological functions of GI that create a network system
that provides habitats and migration routes for wildlife should be considered,
taking into account surrounding green areas, such as riparian forests, street
trees, and windbreak forests. This ecological network will offer recreational
opportunities for walking, jogging, and cycling and will contribute to creating
attractive urban and rural landscapes. Moreover, consumers may selectively
purchase agricultural goods produced from local communities with many GI
measures contributing to the preservation of biodiversity, reduction of flood
damage, and provision of recreational opportunities.
3. Collaborative management of GI.
Natural ecosystems functioning as GI can be maintained by letting ecosystems
undergo natural succession. However, parks, farmlands, and other seminatural GI
may require management activities. Currently, local and prefectural governments
may conduct maintenance of those GI, but collaborative management by local
residents with the help of the government is recommended to obtain long-term
stewardship of GI.
Disaster risks are determined by hazard, exposure, vulnerability, and social
capacity building (United Nations International Strategy for Disaster Reduction
2017). Collaborative management contributes to capacity building. GI may
provide common lands (commons) for local people and can provide vari-
ous ecosystem services, strengthen community solidarity through collaboration
activities, and nurture the power of community resilience against natural dis-
asters. Although local Japanese villages used to have commons in the form of
forests for firewood and grasslands, those lands were subdivided into individual
ownerships, and the commons system sustained by collaborative management
was abolished. However, as indicated in Fig. 2.2, we should build a new
commons system at the watershed scale that connects people living in up- and
midstream forestry and/or agricultural villages with people living in downstream
urban cities through the management of GI. The system will facilitate the
2 Concept and Application of Green and Hybrid Infrastructure 29

exchange and cooperation of rural and urban people, which will promote
social solidarity and economic benefits in local communities under normal
circumstances and will provide evacuation shelters for urban people when they
encounter serious disasters.

Acknowledgments The author sincerely appreciates all thoughtful comments and suggestions
made by Dr. Frederick J. Swanson that significantly improve the quality of the original manuscript.
This research was funded by the Environment Research and Technology Development Funds (4-
1504 and 4-1805) of the Ministry of the Environment of Japan.

References

Benda L, Poff LR, Miller D, Dunne T, Reeves G, Pollock M, Pess G (2004) Network dynamics
hypothesis: spatial and temporal organization of physical heterogeneity in rivers. Bioscience
54:413–427
Buijse AD, Coops H, Staras M, Jans LH, Van Geest GJ, Grift RE et al (2002) Restoration strategies
for river floodplains along large lowland rivers in Europe. Freshw Biol 47(4):889–907
European Commission (2016) Supporting the Implementation of Green Infrastructure. European
Commission, Rotterdam
Green Infrastructure Association (Japan) (ed) (2017) Authorized edition. (Ketteiban). Nikkei BP.
392p
Green Infrastructure Association (Japan) (ed) (2020) Practical edition. (Jissenban). Nikkei BP.
520p
Kobayashi Y, Higa M, Higashiyama K, Nakamura F (2020) Drivers of land-use changes in societies
with decreasing populations: a comparison of the factors affecting farmland abandonment in a
food production area in Japan. PLoS One 7:e0235846
Maes J, Barbosa A, Baranzelli C, Zulian G, e Silva FB, Vandecasteele I et al (2015) More green
infrastructure is required to maintain ecosystem services under current trends in land use change
in Europe. Landsc Ecol 30:517–534
Masatomi H, Masatomi Y (2018) Ecology of the red-crowned crane and conservation activities
in Japan. In: Nakamura F (ed) Biodiversity conservation using umbrella species. Springer,
Singapore, pp 83–105
Matsuura S, Yamamoto K, Hamaguchi T, Homma H (1988) Historical review of groves against
flood hazards. In: Proceedings of the 8th historical study of Japanese civil engineering, pp
193–204 (in Japanese)
Miura Y, Shiraishi J, Shiomi A, Kitazawa T, Hiraga T, Matsumoto F et al (2013) Origin of three
red-crowned cranes (Grus japonensis) found in Northeast Honshu and West Hokkaido, Japan,
from 2008 to 2012. J Vet Med Sci 75:1241–1244
Morimoto J, Shibata M, Shida Y, Nakamura F (2017) Wetland restoration by natural succession in
abandoned pastures with a degraded soil seed bank. Restor Ecol 25:1005–1014
Nakajima H, Ohgushi K, Hino T (2013) Numerical simulations of flood and inundation for
evaluating the effects of Nokoshi and open levee in Jobaru River. Ann J Hydr Eng Japan Soc
Civil Eng 69(4):I_1537–I_1542. (in Japanese with English abstract)
Nakamura F, Inahara S, Kaneko M (2005) A hierarchical approach to ecosystem assessment of
restoration planning at regional, catchment and local scales in Japan. Landsc Ecol Eng 1:43–52
Nakamura F, Ishiyama N, Sueyoshi M, Negishi J, Akasaka T (2014) The significance of meander
restoration for the hydrogeomorphology and recovery of wetland organisms in the Kushiro
River, a lowland river in Japan. Restor Ecol 22:544–554
30 F. Nakamura

Nakamura F, Seo J Il, Akasaka T, Swanson FJ (2017) Large wood, sediment, and flow regimes:
Their interactions and temporal changes caused by human impacts in Japan. Geomorphology
279:176–187
Nakamura F, Ishiyama N, Yamanaka S, Higa M, Akasaka T, Kobayashi Y, Ono S, Fuke N,
Kitazawa M, Morimoto J, Shoji Y (2020) Adaptation to climate change and conservation of
biodiversity using green infrastructure. River Res Appl 36:921–933
Okuma T (1987) A study on the function and etymology of open levee. In: Proceedings of the 7th
historical study of Japanese civil engineering, pp 259–266. (in Japanese)
Onuma A, Tsuge T (2018) Comparing green infrastructure as ecosystem-based disaster risk
reduction with gray infrastructure in terms of costs and benefits under uncertainty: a theoretical
approach. Int J Disast Risk Reduct 32:22–28
Tanabe T, Okuma T (2001) A study on the roles and effects of “NOKOSHI” on the Jobaru-River
Basin. Example of flood control that allows overflowing and its future possibility. Historic Stud
Civil Eng 21(2):147–158. (in Japanese with English abstract)
Tsukamoto (1985) Effects of forests on flood and base flow discharge (Shinrin no Suigen Kanyou
Kinou). Chisan 30-5:116–121. (in Japanese)
United Nations International Strategy for Disaster Reduction (2017) National disaster risk
assessment: governance system, methodologies, and use of results. UNISDR. 77p
Yamanaka S, Ishiyama N, Senzaki M, Morimoto J, Kitazawa M, Fuke N, Nakamura F (2020) Role
of flood-control basins as summer habitat for wetland species—a multiple-taxon approach. Ecol
Eng 142:105617

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 3
An Economic Analysis of Optimal
Hybrid Infrastructure: A Theoretical
Approach in a Hydro-Economic Model

Ayumi Onuma

Abstract This chapter features the functions of green infrastructure and gray
infrastructure in a hydro-economic model and integrates them into a hybrid infras-
tructure. We show that adding green infrastructure to an existing gray infrastructure
generates a new benefit, which is referred to as absorption effects. Using the
model, we investigate the optimal hybrid infrastructure by maximizing the social
net benefits, in which disaster risk reduction is defined. We then discuss how
green infrastructure can be used to augment existing gray infrastructure from
the perspective of economic optimality. We derive some conditions to determine
whether both or either of gray and green infrastructures should be introduced to
implement disaster risk reduction.

Keywords Green infrastructure · Gray infrastructure · Hybrid infrastructure ·

Disaster risk reduction · Hydro-economic model

3.1 Introduction

The purpose of this chapter is to model the functions of green infrastructure (GNI)
and gray infrastructure (GYI) to integrate them into a hybrid infrastructure (HBI)
and to express it in a simple hydro-economic model for deriving optimal HBI.
Although ecosystem-based disaster risk reduction has been widely attracting
attention, there are few theoretical analyses in economics; Barbier (2012) and
Barbier and Enchelmeyer (2014) provide theoretical frameworks in the context of
GNI to prevent storm surges in coastal areas, but they only focus on GNI.
Meanwhile, Onuma and Tsuge (2018) present a hydro-economic framework
to compare GYI and GNI in terms of cost–benefit by considering the size of
the population and formulating the features of their respective disaster prevention

A. Onuma ()
Keio University, Minato-ku, Tokyo, Japan
e-mail: [email protected]

© The Author(s) 2022 31

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_3
32 A. Onuma

functions. In addition, disaster risk and disaster risk reduction are clearly defined
in the above model. While this direction is suitable for discussing optimal disaster
risk reduction, the model is based on a binary choice between GNI and GYI. In
reality, even if a preventing flood infrastructure is called a GNI, it is often an HBI,
i.e., a mixture of GNI and GYI. Examples include the “Slowing the Flow” project
implemented in Pickering in the UK, which is known as a nature-based solution to
flood management and uses one large storage bund to effectively reduce flooding
(Nisbet et al., 2015). Therefore, providing a framework for the analysis of HBI is
significant for the economic study of disaster prevention infrastructures, which are
becoming increasingly important in the days of frequent natural disasters.
In this chapter, we formulate the features of HBI, which combine the features
of GNI and GYI, and construct a hydro-economic model expressing HBI. We then
show that adding GNI to an existing GYI to form an HBI generates a new benefit,
which is referred to as absorption effects. We then discuss how GNI can be used to
augment existing GYI from the perspective of economic optimality.

3.2 Green, Gray, and Hybrid Infrastructures

Let us consider a hazard or rainfall that leads to a flooding river. In this chapter, we
use “hazard” and “rainfall” exchangeably and denote it by H . Residents directly face
H if there is no preventing infrastructure, whereas H is mitigated if we introduce
GNI and/or GYI.

3.2.1 Green Infrastructure

We consider a GNI, which is often exemplified by forests located in the upstream

of the river. GNI has the ability to save water. In the case of forests, some of the
water that flows in does not flow out because it penetrates the ground or stays on the
leaves, while the capacity of water saving is limited.
Let Xf denote the direct outflow from the GNI or the forest. Xf is the level of
H net of the level of flow absorbed by the GNI, which is denoted by A. Therefore,
Xf = H − A. The level of rainfall at which the capacity of absorption is used up is
expressed by Ĥ . Ĥ is supposed to increase with the scale of the forest, Kf , so we
define Ĥ as

Ĥ = Ĥ (Kf ) (3.1)

with Ĥ  (Kf ) > 0. The level of absorption depends on Kf and H . A larger forest
flows out smaller Xf , that is, higher A. Thus, we write A as A(H, Kf ). We assume
that A increases with H up to Ĥ , whereas A is constant with H beyond Ĥ . On the
3 Optimal Hybrid Infrastructure 33

Fig. 3.1 Outflow Xf and

absorption A(H, Kf ) of GNI

contrary, Xf also increases with H . That is, for H < Ĥ , 0 < AH (≡ ∂A/∂H ) < 1,
and AH (≡ ∂ 2 A/∂H 2) < 0. Moreover, AH = 0 for H ≥ Ĥ and AKf > 0.
The maximum absorption of the forest, Â, is defined as

 = A(Ĥ , Kf ) ≡ Â(Kf ) (3.2)

By definition, we have  > 0 and A(H, Kf ) < Â(Kf ) for H < Ĥ . In summary,
the resulting outflow from GNI, Xf , is determined as

H − A(H, Kf ) for H < Ĥ
Xf = (3.3)
H − Â(Kf ) for H ≥ Ĥ

Xf is described in Fig. 3.1.

3.2.2 Gray Infrastructure

Let Kc represent the GYI. Let Hc be the hazard or flow arriving to GYI. Following
Onuma and Tsuge (2018), GYI protects the flow completely up to a threshold Ĥc ,
but it does not protect beyond Ĥc . We call Ĥc threshold hazard, which increases
with Kc ; that is, the threshold hazard increases as the scale of the GYI increases.
We assume that the threshold hazard is proportional to the scale of Kc :

Ĥc = θ Kc (3.4)
34 A. Onuma

Therefore, the resulting hazard that the residents face, Xc , is expressed as


0 (Hc ≤ Ĥc )
Xc = (3.5)
Hc (Hc > Ĥc )

As noted by Onuma and Tsuge (2018), the most striking difference between GNI
and GYI is their hazard prevention rates: in GYI, the rate is 1 up to the threshold but
zero beyond that, while the rate never becomes 1 but is always positive in the case
of GNI.

3.2.3 Hybrid Infrastructure

In our model, an HBI is defined as a mixture of GNI and GYI. A GNI is located
upstream of the river, while a GYI is built downstream to protect nearby residents.
Therefore, Xf in (3.3) flows to the GYI. That is,

Hc = Xf = H − A(H, Kf ) < H (3.6)

In this way, GNI mitigates the hazards coming to GYI. Xh denotes the hazard
resulting from the HBI. GYI itself protects against the hazard completely up to a
threshold of Ĥc , but it fails beyond Ĥc ; that is,

0 (Xf ≤ Ĥc )
Xh = (3.7)
Xf (Xf > Ĥc )

This is a function of the HBI (Kc , Kf ). Under the HBI, we derive the level of hazard
H ∗ such that

Xf∗ ≡ H ∗ − A(H ∗ , Kf ) = Ĥc ⇒ H ∗ = Ĥc + A(H ∗ , Kf ) (3.8)

That is, H ∗ is the level of rainfall in a forest where flow to GYI attains the threshold
hazard Ĥc . This shows that the threshold hazard increases from Ĥc under (Kc , 0) to
A(H ∗ , Kf ) + Ĥc under (Kc , Kf ). Let H ∗ be Ĥc + ω, which is the new threshold
under HBI. Then, it holds that Ĥc + ω = Ĥc + A(Ĥc + ω, Kf ), so

ω = A(Ĥc + ω, Kf ) (3.9)

ω is an important variable when considering the HBI, in that the threshold hazard of
GYI increases as a result of setting the GNI. We may call ω a substantial increment
in threshold hazard. In the following lemma, we prove that ω is determined
uniquely, given (Kc , Kf ).
3 Optimal Hybrid Infrastructure 35

Lemma 1 For any Ĥc > 0, ω satisfying (3.9) exists uniquely.

Proof From the definition of A, it holds that 0 < A(Ĥc + ω, Kf ) ≤ Â(Kf ). This
means that there always exists a ω belonging to (0, Ĥ ) that satisfies (3.9). Next,
suppose that there exist ω1 and ω2 with ω1 > ω2 that satisfy (3.9) for some Ĥc > 0.
However, this implies

ω1 − ω2 = A(Ĥc + ω1 , Kf ) − A(Ĥc + ω2 , Kf ) (3.10)

However, this is not feasible because we assume that AH < 1. Thus, ω that
satisfies (3.9) must be unique.
The next lemma shows that ω depends on GYI as well as GNI when GYI is not
large enough.
Lemma 2 For Ĥc > 0 and ω satisfying (3.9),

dω AH
= ≥0
d Ĥc 1 − AH

where equality holds only if A(H, Kf ) = Â(Kf ) or Ĥc + ω ≥ Ĥ .

Proof Differentiating (3.9) leads to

dω = AH (d Ĥc + dω) (3.11)

AH
From here, we obtain dω
= 1−AH
, where AH = 0 if Ĥc + ω ≥ Ĥ .
d Ĥc
Then, we get the following result, which shows the magnitude of the substantial
increment in the threshold hazard.
Proposition 1 The HBI increases the threshold Ĥc by ω where (i) ω = Â(Kf ) if
Ĥc ≥ Ĥ − Â(Kf ). (ii)ω < Â(Kf ) if Ĥc < Ĥ − Â(Kf ).
Proof For Ĥc = Ĥ − Â(Kf ),

A(Ĥc + ω, Kf ) = A(Ĥ − Â(Kf ) + ω, Kf ) (3.12)

It is obvious that if ω = Â(Kf ), (3.9) is satisfied as

ω = A(Ĥc + ω, Kf ) = A(Ĥ , Kf ) = Â(Kf ) (3.13)

Then, from Lemma 1, ω = Â(Kf ) is the solution of (3.9). Moreover, from

Lemma 2, AH = 0 at Ĥc , so ω = Â(Kf ) is also the solution for Ĥc > Ĥ − Â(Kf ).
Furthermore, for Ĥc < Ĥ − Â(Kf ), Lemma 2 says ω < Â(Kf ).
36 A. Onuma

Fig. 3.2 Prevention of hazard by HBI and GYI

In summary, by introducing (Kc , Kf ), the resulting hazard Hh is


0 (H ≤ Ĥc + ω)
Hh = (3.14)
H −ω (H > Ĥc + ω)

Equation (3.14) is depicted in Fig. 3.2, where Ĥc ≥ Ĥ − Â(Kf ) so that ω = Â(Kf ).
In the figure, the benefits of HBI can be graphically compared with GYI. HBI, in
which GNI is added to the existing GYI, provides two benefits. First, even though
the hazard is higher than Ĥc +ω, which means that the hazard cannot be prevented, it
is always mitigated. It is worth noting that this benefit cannot be obtained by setting
GYI solely; any hazard beyond the new threshold hazard is mitigated in HBI, due
to the function of GNI.
Second, the threshold Ĥc of GYI increases by ω, despite the fact that GYI is
kept the same as before. It might be interesting that the magnitude of ω sometimes
depends on Kc as well as Kf , as demonstrated by Lemma 2. This occurs when
Kc is not sufficiently large enough so that Ĥc < Ĥ − Â(Kf ) by proposition 1.
This property might be important when we consider the replacement of a part of the
preventing function of GYI with GNI. In the replacement, suppose that we keep the
threshold hazard as before. It might be intuitive that ω depends only on Kf and is
determined independently of Kc . However, if Kc is not large enough, if there is a
decrease in Kc , it will reduce ω. Hence, without careful planning, such replacement
could result in the reduction of the threshold hazard.
These features are depicted in Fig. 3.3, which expresses A(H, Kf ) as a green
line. Suppose that the initial threshold hazard is Ĥc0 , in which case the substantial
increment in the threshold hazard ω, created by Kf , is ω0 . The figure shows that if
3 Optimal Hybrid Infrastructure 37

Fig. 3.3 Change of ω as a result of reducing Ĥc

Ĥc0 is reduced to Ĥc1 , ω0 decreases to ω . On the contrary, if we suppose that the

initial threshold hazard is sufficiently large, Ĥc2 , where ω = ω2 . In this case, even if
Ĥc2 is reduced to Ĥc3 , which changes ω2 to ω , the level of ω is equivalent.

3.2.4 Ratios of Prevented Hazard by GNI and GYI

As shown by (3.14), the HBI completely prevents any hazard H below Ĥc + ω
and partly H beyond it. Now that we have a mixture of two preventing hazard
infrastructures, GNI and GYI, in HBI, seek each ratio of hazard that is prevented in
each infrastructure. Given (Kc , Kf ), we define rf = (H −Xf )/H = A(H, Kf )/H
and rc = (Xf −Xh )/H , which formally shows the ratio of hazard prevention. Thus,
in terms of (3.14), both rf and rc are positive with rf + rc = 1 for H (< Ĥc + ω),
whereas rf > 0, but rc = 0 for H (> Ĥ + ω), where rf + rc < 1.
Note that rf (H ) decreases with H because AH > 0 for H < Ĥ and AH = 0 for
H ≥ Ĥ . Note also that

A(H, Kf )
rf (0) = lim = AH (0, Kf ) (3.15)
H ↓0 H

by l’Hôpital’s rule. Thus, we can describe the behavior of rf and rc as follows.

Proposition 2 In HBI expressed by (3.14), the ratio of prevention by GNI, rf =
A(H, Kf )/H , decreases with H from rf (0) = AH (0, Kf ). Meanwhile, the ratio
of GYI, rc = (Xh − Xf )/H , is equal to 1 − rf for H (< Ĥc + ω) and to zero for
H (≥ Ĥc + ω).
38 A. Onuma

Fig. 3.4 rc and rf

Figure 3.4 depicts rc and rf .1 GNI reduces the flow considerably when the
rainfall is small.

3.3 Optimal Hybrid Infrastructure

In this section, we incorporate the function of HBI into a simple economic model,
that is, a hydro-economic model, and analyze some properties of the optimal HBI.
We define disaster risk reduction (DRR) in this model. Denoting the considerable
maximum hazard with H m , let f (H ) : [0, H m ] → [0, 1] be the function expressing
 Hm
the probability that a hazard H occurs, with 0 f (H )dH = 1. The hazard risk is
 Hm
defined as H R = 0 f (H )H dH .
In contrast, D(H ) shows the disaster caused by hazard H , where D(0) =
0, D  > 0, D  > 0. That is, the marginal disaster of hazard is positive and

1 Although AH < 0, the signs of rf and rc are ambiguous. Figure 3.4 shows that AH is small
enough (|AH | is large enough), so that rf < 0 and rc > 0 for H (< Ĥ ). It is possible that the
opposite case occurs, depending on the value of AH . However, rf > 0 for H (≥ Ĥ ).
3 Optimal Hybrid Infrastructure 39

increases with H . Disaster risk (DR) without preventing infrastructure, that is,
(Kc , Kf ) = (0, 0), is defined as
 Hm
DR(0, 0) = n f (H )D(H )dH (3.16)
0

where n is the population exposed to hazard H . On the contrary, DR under GYI,

that is, (Kc , Kf ) = (Kc , 0) from (3.5) with its threshold hazard, Ĥc , is expressed as
 Hm
DR(Kc , 0) = n f (H )D(H )dH (3.17)
Ĥc

Disaster risk reduction by installing GYI at the initial state without preventing
infrastructure, DRR from (0, 0) to (Kc , 0) is, therefore, defined as
 Ĥc
DRR((Kc , 0)|(0, 0)) = DR(0, 0) − DR(Kc , 0) = n f (H )D(H )dH
0
(3.18)

DRR((Kc , 0)|(0, 0)) is considered to be the benefit of setting GYI to the site
without preventing infrastructures.
Now, let us suppose that the government strengthens an existing GYI to increase
the threshold hazard from Ĥc by H to Ĥc∗ (= Ĥc + H ) by investing Ic into GYI
and/or by introducing GNI. That is, we introduce (Ic , Kf ) to satisfy

θ Ic + ω = H (3.19)

Here, we assume that Ĥc is large enough in view of proposition 1 to satisfy ω =

Â(Kf ). This increment can be implemented only with Ic = θ −1 H , where GYI
increases to Kc + θ −1 H ; or only with Kf = Â−1 ( H ); or with a mixture of Ic
and Kf satisfying (3.19).
Disaster risk under this improvement is
 Hm
DR(Kc + Ic , Kf ) = n f (H )D(H − Â(Kf ))dH (3.20)
Ĥc∗

Thus, DRR gained by this improvement is defined as

 Ĥc∗
DRR((Kc + Ic , Kf )|(Kc , 0)) = n f (H )D(H )dH
Ĥc
 Hm
+n f (H )(D(H ) − D(H − Â(Kf )))dH
Ĥc∗
(3.21)
40 A. Onuma

The first term is the benefit from enhancing the threshold hazard, which we call the
threshold effect. The second term, however, is the benefit from absorption by GNI,
which we call the absorption effect.
Let us define the costs of implementing Ĥc + H . We denote the cost of
expanding GYI as Cc (Ic ), assuming Cc > 0 and Cc ≥ 0. On the contrary, the
cost of GNI is expressed as μCf (Kf ), where μ(> 0) reflects the opportunity costs
of setting GNI or the prices of land acquired for setting GNI because GNI requires
a wider range of geographic spaces. We assume Cf > 0 and Cf > 0.
Meanwhile, environmental benefits and costs are included as Ef = αf Kf (αf >
0) for GNI and Ec = −αc Kc (αc > 0) for GYI, where αf > 0 and αc > 0. Then,
social net benefit, W , is expressed by

W = DRR((Kc + Ic , Kf )|(Kc , 0))

− Cc (Ic ) − μCf (Kf ) + αf Kf − αc Ic (3.22)

The optimal HBI is defined as maximizing (3.22) with (Ic , Kf ) subject to (3.19).
Assume that

Â(Kf ) = γ Kf , γ > 0 (3.23)

The optimum must satisfy if it is interior (both Ic and Kf are positive)

 Hm γ 
n f (H )D  (H − γ Kf )γ dH = (μCf − αf ) − (C + αc ). (3.24)
Ĥc∗ θ c

γ /θ shows how much Ic is replaced by one unit, say, one hectare increase in
Kf . Here, the left-hand side is the marginal absorption effect of this (marginal)
replacement, while the right-hand side is the marginal net cost of replacement,
including environmental benefits and costs. Equation (3.24) states that the marginal
absorption effect must be equal to the net cost of replacement.
To make the argument clearer, we specify the function Cc as Cc (Ic ) = qIc (q >
0), where q denotes the unit cost of investing in GYI. Moreover, the l.h.s. of (3.24)
is denoted by V (Kf ) for convenience, where it holds that V  < 0. Let us denote the
maximum feasible GNI as K̄f ≡ H /γ . Note that V (0) > V (Kf ) > V (K̄f ) for
Kf ∈ (0, K̄f ). Then, Kf = 0 becomes optimal if and only if

γ
V (0) + αf + (q + αc ) ≤ μCf (0) (3.25)
θ
where Ic = H /θ . The l.h.s. of (3.25) includes the marginal benefit from launching
GNI, marginal absorption benefit, and marginal environmental benefit, added by the
saved marginal costs of GYI. Roughly, this occurs when μCf (0) is sufficiently high.
3 Optimal Hybrid Infrastructure 41

Developing even the first hectare of GNI is very costly, relative to the unit investment
cost q. However, Ic = 0 is optimal if and only if
γ
V (K̄f ) + αf + (q + αc ) ≥ μCf (K̄f ) (3.26)
θ

This condition is satisfied when μCf (K̄f ) is still low enough, even though the
targeted increase in threshold hazard is achieved only by GNI. We state these
properties in the following proposition.
Proposition 3 To enhance the threshold hazard from Ĥc by H , no use of GNI (i.e.,
Kf = 0) is optimal if μCf (0) is sufficiently high to satisfy (3.25). On the contrary,
zero investment in GYI, that is, Ic = 0, is optimal if μCf (K̄f ) is sufficiently low to
satisfy (3.26). Otherwise, it holds that (Ic , Kf ) >> (0, 0), where (3.24) is satisfied.
Let us suppose that the optimum is interior; that is, (3.24) is satisfied at the
optimum. Now, we assume that μ depends on n, which can be justified by the fact
that the opportunity cost of setting GNI increases with the size of the population.
Thus, μ = μ(n), with μ > 0. Let  μ express the population elasticity of
opportunity cost, that is,  μ = nμ /μ, which expresses how much the percentage of
opportunity cost increases when population increases by 1%. We observe the sign
of dKf /dn to see how the use of GNI changes with population.
Proposition 4 Suppose that both Ic and Kf are positive under the optimum HBI.
Then, it holds

dKf αf + γ θ −1 (q + αc )
 0 ⇔ μ  1 −
dn μCf

Proof Differentiating Kf with n under (3.24) leads to

 Hm
dKf Ĥc∗
f (H )D  (H − γ Kf )γ dH − μ Cf
=  Hm
dn μCf + n Ĥ ∗ f (H )D  γ 2 dH
c

μCf (1 −  μ ) − (αf + γ θ −1 (q + αc ))
=  Hm (3.27)
n(μCf + n Ĥ ∗ f (H )D  γ 2 dH )
c

 Hm
using Ĥ ∗ f (H )D  (H −γ Kf )γ dH = ((μCf −αf )− γθ (Cc +αc ))/n from (3.24).
c
Then, since the denominator is positive, the claim is straightforward from the
numerator.
It should be noted that 0 < (αf + γ θ −1 (q + αc ))/μCf < 1 holds because
the r.h.s. of (3.24) must be positive. Therefore, a paradoxical case dKf /dn > 0
occurs if  μ is sufficiently small, as in the claim. In other cases, the share of GNI
use decreases when implementing H as population increases.
42 A. Onuma

3.4 Concluding Remarks

This chapter formalizes HBI and incorporates it into economics. Based on the
model, we discuss the optimal mixture of GNI and GYI, that is, the optimal HBI.
To augment the disaster prevention function of the existing GYI, we show the
nature of the optimal HBI. In particular, the logic for not using GNI at all is justified
only when the marginal cost, including the opportunity cost when deployment is still
zero, is sufficiently high, i.e., (3.25) holds. As the opportunity costs are expected to
be lower for smaller populations, it is likely that the augmentation of disaster risk
reduction infrastructure without the use of GNI would be inefficient, especially in
sparsely populated areas.
In this chapter, GNI is assumed to be located upstream of GYI, but it is also
possible that GYI is the first to receive the hazard and GNI is located behind it, as
in the case of coastal seawalls. An HBI with a different arrangement of GNI and
GYI could also be characterized in the same way as described in this chapter. This
attempt is left for future research.

References

Barbier EB (2012) A spatial model of coastal ecosystem services. Ecol Econ 78:70–79
Barbier EB, Enchelmeyer BS (2014) Valuing the storm surge protection service of US Gulf Coast
wetlands. J Environ Econ Policy 3(2):167–185
Nisbet T, Roe P, Marrington S, Thomas H, Broadmeadow S, Valatin G (2015) Project RMP5455:
Slowing the Flow at Pickering Final Report: Phase II, Department for environment, food and
rural affairs, UK
Onuma A, Tsuge T (2018) Comparing green infrastructure as ecosystem-based disaster risk
reduction with gray infrastructure in terms of costs and benefits under uncertainty: A theoretical
approach. Int J Disaster Risk Reduct 32:22–28

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons licence and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons licence and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder
Chapter 4
Flood Management Policy in Shiga
Prefecture, Japan: Implementation
Approach of a Risk-Based Flood
Management System at Catchment Scale

Kentaro Taki

Abstract Shiga Prefectural Government uses “site safety level” to assess the flood
risk of major floodplains in the prefecture in order to implement an integrated
floodplain management system in society. Site safety level is determined based on
a risk matrix expressing the relationship between the frequency of inundation and
the degree of damage at sites surrounding a river channel complex. A department
has been set up for floodplain management separate from river management. The
Department of Floodplain Management promotes disaster mitigation measures,
such as land use and construction regulations, and conservation of traditional flood
control facilities, including flood prevention forests, open levees, and ring levees.
Such traditional facilities are highly functional as green infrastructure because they
are focused not only on disasters but also on a daily basis in order to wisely utilize
local ecosystem service.

Keywords Shiga Prefecture · Risk-based flood management · Risk assessment ·

Land use · Building regulation

4.1 Changes in Flood Risk Management in Japan

“Those who control the water, control the country.” Flood control has been
considered to be the basis of national administration since long ago, and the system
for doing so has gradually changed over time to support the socio-economy of Japan.
Recently, the situation surrounding flood control has changed dramatically, and
the risk of flooding is higher than ever before. Several challenges are emerging,
including a decline in investment capacity owing to the decline in population,

K. Taki ()
School of Environmental Science, Lake Governance Research Center, University of Shiga
Prefecture, Hikon, Shiga, Japan
e-mail: [email protected]

© The Author(s) 2022 43

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_4
44 K. Taki

changes in social structure and lifestyles, decline in literacy regarding water-related

disaster mechanisms, and an increase in extreme events associated with climate
change. Against this background, it is important to understand how we can eliminate
human casualties and avoid catastrophic damage that would make it difficult to
rebuild livelihoods. We are currently confronted with the issues that contribute to
the initiation of flood control.
The general measures that have been taken to minimize the risk of flood
inundation so far, including basin and in-channel measures such as rainwater storage
and infiltration facilities, river improvement, and flood control facilities, are not
adequate. In addition to these, there is a need for a multilayered approach for
flood disaster risk reduction measures in floodplains including the maintenance
and development of flood flow control facilities such as flood prevention forests,
open levee systems (combination of discontinuous levees and secondary levees),
and ring levees as shown in Fig. 4.1, land use and building regulations, and further
enhancement of flood prevention activities and evacuation guidance. Floodplain in
this section refers to the area of a plain that is inundated by the overflow of water
from a river channel or other source during a flood. In other words, floodplain refers
to the entire area of a valley floor plain, alluvial fan, alluvial plain, or delta that is
inundated by floodwater.
Measures in the floodplain are not a new approach to disaster mitigation.
Flood control during the Sengoku period (Warring States period, late fifteenth–
late sixteenth century) was based on defensive measures to prevent flooding as
exemplified by open levees; flood mitigation measures based on the premise of
inundation rather than continuous embankments were the mainstream measures.
In the Edo period (1603–1867), as the development of new rice fields and the
intensification of land use progressed, flood control by continuous levees gradually
became the mainstream owing to improvements in civil engineering techniques. The
River Law was enacted in 1898 against the background of the modernization of civil

Fig. 4.1 Flood prevention forests, open levee system, and ring levees
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 45

Table 4.1 Changes in modern flood control planning (based on Hori et al. 2008)
First stage The largest floods to date can be handled by the channels and
(largest recorded flooding) dam reservoirs without causing flooding
(Late 1890s)
Second stage The designed external force for flood control facilities is
(probable flooding) assessed according to the probability of exceeding the annual
(Late 1950s) maximum rainfall; rainfall of a certain probability scale is
used as the planned rainfall amount, and the main type of
hydrographs produced by this rainfall is handled by the river
channels and dam reservoirs without causing inundation
Third stage In addition to the measure of handling rainwater after it enters
(reduce inflow from the the channels, alternative plans include measures in the
watershed) watershed to reduce the amount of rainwater that enters the
(Late 1970s) channels itself
Fourth stage Flood inundation is considered a premise, and alternative
(flood risk management) plans take into account floodplain damage reduction measures
(2000s) as well as watershed and channel facilities

engineering technology, and flood control systems using continuous levees became
the basis for flood control in Japan. Currently, with the exception of a few nonurban
rivers (e.g., Kita River in the Gokase River System (Miyazaki Prefecture), Kita
River in the Kita River System (Fukui Prefecture), etc.), there are few flood control
riparian forests that explicitly incorporate floodplain disaster mitigation measures.
Here, I would like to summarize the changes in flood control systems since the
Meiji era. Hori et al. (2008) organized the changes in modern flood control planning
in Japan into four stages (Table 4.1) and argued for the need to transition to the
fourth stage. Flood control in the fourth stage is an expansion of flood control
theory from the planning of watersheds (catchment areas) and river channels to
encompass the entire basin including the floodplain, which is the affected area. This
stage incorporates the concept of base defense into the continuous levee system.
Flood control in the fourth stage is based on the premise that flood inundation is
a natural phenomenon; mitigation measures such as floodplain land use and housing
practices can be considered as approaches to reduce damage by avoiding exposure
and reducing vulnerability (Fig. 4.2).

4.2 Flood Management Policy in Shiga Prefecture: Policy

Formulation

4.2.1 The History of Wise Land Use and Urban Development:

Hints from History

The Amano River, which flows through Maibara City, Shiga Prefecture, was
severely damaged by the Isewan Typhoon in 1959. Extensive riverbed excavation
46 K. Taki

Fig. 4.2 Approaches to

damage factors and
mitigation, and mitigation
measures in the floodplain
reduce damage by avoiding
exposure and reducing
vulnerability. (Sources:
Ministry of the Environment,
Natural Environment Bureau
2016; ADRC 2015)

and revetment work was conducted in the recovery, but the prefectural public
works office decided to keep the open levees. River administrators at that time well
understood inundation characteristics and were also aware of excess flooding. The
bullet train, Tokaido Shinkansen, crosses the Amano River. A railroad bridge was
built across the river. Furthermore, in the sections of farmland, such as paddy fields,
the viaduct extends instead of embankments, as shown in Fig. 4.3. Continuous
embankments can interfere with the flow of floodwater and cause inundation.
Therefore, bridges were built over non-riverine sections to allow floodwater to flow
down through the section. The original track was planned with an embankment
structure, but there was opposition from local residents against the embankment
system owing to concerns about the retention of floodwater and inland water (Photo
4.1). In addition, the town of Kazuramaki in Higashiomi City, located near the
confluence of the Hino River and the Hokyoji River, has autonomously raised
the level of its residential land in conjunction with the height of its levees after
restoration and repair based on its past experience with typhoons such as the Ise Bay
typhoon. In areas where retention-type flooding occurs, such as at the confluence of
a banked river, the frequency of outwash flooding is reduced by the embankments,
but in the unlikely event that flooding does occur, floodwater and inland water may
raise the depth of inundation to the height of the levee.
Inundation analysis throughout the prefecture showed that only approximately
1800 households would be submerged (inundated by more than 3.0 m) even with
a rainfall probability of once in 200 years (Taki et al. 2010). In addition, many of
these houses were developed after the late 1940s. In contrast, many old settlements
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 47

Fig. 4.3 Extension of the viaduct onto the paddy field so as to not block the flow of floods

Photo 4.1 A sign opposing the embankment-type Shinkansen. (Source: Shiga Prefectural Gov-
ernment)

are located in slightly elevated areas such as natural levee belts (avoidance of
exposure in Fig. 4.2) or have had measures implemented such as raising housing
land (reduction of vulnerability in Fig. 4.2).
Several approaches described here were adopted in various parts of the prefecture
and are still implemented in certain areas. However, with passage of time, many
of these innovative measures have lost their importance. Open levees have also
been gradually disappearing without their disaster mitigation effects being fully
evaluated.
Until a few decades ago, river modification and urban development were linked.
In the case of Lake Biwa, paddy fields are located on the lowlands around the river,
48 K. Taki

and villages are located on very highlands such as natural levees. The result is
a beautiful and rational landscape, harmonized with the providence of nature, in
which the flow of the rivers and the positioning of farmlands and towns are all
planned integrally. Such designs certainly made the best use of local resources
or watershed ecosystem services given the size of the population and level of
technology at the time.

4.2.2 Policy Features: Process Management and Framework

Design

Shiga Prefecture’s flood management policy began with the establishment of

the Floodplain Management Office within the Public Works and Transportation
Department in July 2006. This is a separate organization from the River Works
Division.
In March 2012, approximately 6 years after its conception, the “Flood Risk
Management Basic Policy in Shiga Prefecture” (hereinafter, “Basic Policy”) was
approved, and 2 years later, in March 2014, the “Flood Management Ordinance in
Shiga Prefecture” was approved after rigorous discussions. The Flood Management
Ordinance provides a legal basis for the Basic Policy and ensures the effectiveness
of flood risk management measures conducted by Shiga Prefectural Government.
The Basic Policy defines the objectives and measures of flood management in Shiga
Prefecture as follows.

Flood risk management integrates self-help, mutual aid, and public assistance
to comprehensively implement both in-river and off-river measures for flood
risk management to avoid (1) loss of life (the highest priority) and (2) damage
that makes it difficult to rebuild daily life after any kind of flood.

The responsibility for river management is stipulated in the River Law. The
obligatory responsibility for flood control is understood to be “setting up design
floods and ensuring that design floods flow safely through river channels.” The
obligatory responsibility for flood control is understood to be “setting up design
floods and ensuring that design floods flow safely in the river channels.” The
Basic Policy includes such river management systems according to the River Law.
The Basic Policy positions conventional river management as a core project and
specifies that “basin retention measures,” “floodplain disaster mitigation measures,”
and “measures to improve local disaster preparedness” are to be implemented to
achieve the flood control goals set forth (Table 4.1). Beyond the framework of the
River Law, the City Planning Act, Building Standards Act, and Flood Fighting Act
were included in the scope of protection. The framework is to achieve the goal by
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 49

covering the areas that cannot be addressed by river maintenance with the other three
measures. Of the three measures, the “floodplain disaster mitigation measures,”
in particular, are intended to “avoid exposure” and “reduce vulnerability” to flood
inundation.

4.2.3 Site Safety Level: Basic Information for Policy Decisions

When considering flood control measures, the scope of defensive measures should
be expanded to include floodplain land use and housing practices in order to move
into the fourth stage of flood control. In addition, it is necessary to directly weigh the
risks of the floodplains newly subject to planning. In other words, floodplain risk is
the frequency at which a certain kind of damage occurs at each point surrounded by
a river/channel complex. For example, if you buy the house shown in Fig. 4.4, you
may be more concerned about how often and to what degree your home would be
flooded rather than the individual safety of the rivers and waterways that surround it.
The risk at each site surrounded by a river/channel complex in this way is referred
to as the “site safety level” by SPG.
Site safety level can be calculated from hydraulic parameters (inundation depth
and speed and fluid force) obtained from flood hydraulic analysis. For example, it
is possible to express the site safety level as a risk matrix, as shown in Fig. 4.5,
according to the frequency of damage and degree of damage at each point. The
magnitude of damage can be classified into five levels: level 1 (h ≤ 0.1 m) indicates
no damage; level 2 (0.1 ≤ h ≤ 0.5 m) indicates damage from inundation below the
ground floor of a building; level 3 (0.5 ≤ h < 3.0 m) indicates severe damage from
inundation above the ground floor; level 4 (3.0 ≤ h) indicates fully submerged;
and level 5 (2.5 m3 /s2 ≤ u2 h) indicates completely destroyed. The probability of
occurrence of each flood event was evaluated by means of its return period, namely,
2, 10, 30, 50, 100, 200, 500, and 1000 years, where h is depth (m) of the inundation
and u is velocity (m/s). The term u2 h means the “fluid force.”

Observation Point
What kinds of risk is here?

A River Channel Swege line Agricultural channel B River

Safty level of flood control 1/30 Safety level of flood control 1/2 Safty level of flood control safty level of flood control 1/5 safty level of flood control
1/10 1/10

Fig 4.4 Site safety level. This is the level of safety of a point surrounded by a river/channel
complex and for the setting of daily life (Source: Taki et al. 2010)
50 K. Taki

Fig 4.5 Site safety level at a point expressed using a risk matrix (Taki et al. 2010; Shiga Prefectural
Government 2012)

SPG has developed and published a numerical analysis model for evaluating site
safety level (Taki et al. 2013). The basic requirements of this numerical analysis
model are as follows:
Requirement 1: Consider combined flooding from the river/channel complex sur-
rounding a site on the floodplain (living sphere).
Requirement 2: Consider high- to low-frequency inundations (multiple occurrence
probabilities).
Requirement 3: Be able to evaluate risk/safety level equally by policy units (in this
case the entire prefecture).
Requirement 1 indicates that the model evaluates the risk/safety level at each
location on the floodplain, rather than performance of each river levee or other
disaster prevention facility. The model evaluates the performance of the system as
a whole group of disaster prevention facilities rather than individual facilities. SPG
incorporates (rainwater) sewage systems and agricultural drainage channels into the
model besides the 240 major rivers and analyzes them in an integrated manner.
Requirement 2 is to assess damage based on the frequency (or magnitude) of floods.
Since administrative measures must be impartial within a jurisdiction, requirement 3
allows assessments by administrative units, such as prefectures and municipalities,
in the same way.
The structure of the numerical analysis model and the flow of calculations are
shown in Fig. 4.6 (Taki et al. 2019) and Fig. 4.7 (Taki et al. 2010). Although the
calculation algorithm is orthodox, the most distinctive feature of the model is that it
integrates data from not only major rivers managed by the national and prefectural
governments but also from small rivers, (rainwater) sewage systems with different
management categories, and agricultural drainage.
Modeled rainfall events with a probability of occurrence once in 2, 10, 30, 50,
100, 200, 500, and 1000 years estimated from the Shiga Prefecture rainfall intensity
equation are used as the external force for the calculation.
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 51

ph

ph
Discharge basin
gra

ra
River channels (Distribution model)
dro

og
㸦1D-flow㸧 Hy

dr
Hy
graph

Hy dro
Hy dro
Floodplain

graph
㸦2D-Flow㸧
Hy drograph

Catchment boundary
La k e Biwa , Water channels,
etc. networks

• River channels (approx. 240 rivers) are single dimensional and flood plain is two-dimensional
• Small rivers and large water channels are treated as channels of uniform flow.
• The scope of area for agricultural land readjustment and rainwater drainage development is obtained
by deducting rainfall equivalent to discharge capacity and totaling in the downstream area.

Fig 4.6 Composition of the numerical analysis model (Source: Taki et al. 2019)

Fig 4.7 Flow of calculations (Taki et al. 2010)

Some of the calculation results are shown in Figs. 4.8 and 4.9. For official
prefectural site safety level data, see the Shiga Prefecture disaster preparedness
information map (https://shiga-bousai.jp/dmap). GIS data (shapefiles) are available
along with maps of areas of possible inundation for each river.
The flood hazard maps published by local governments across the country
usually show the depth of inundation for each river in the event of a design flood
as an assumed external force. Recent amendments to the Flood Fighting Act in
2015 mandated the designation of areas with a high probability of inundation in the
52 K. Taki

Fig 4.8 Maximum inundation depth (1000-year flood) (Source: Taki et al. 2019)
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 53

Fig 4.9 Frequency of inundation above floor level (≥50 cm) (Source: Taki et al. 2019)
54 K. Taki

maximum expected scale of flooding and led to an urgent need to review inundation
assumptions and hazard maps across the country. The assumed maximum size is
often calculated as the equivalent of an event that has a probability of occurring once
in 1000 years. Despite this increased awareness of excessive flooding in inundation
assumptions, it is not yet common to express flood risk in a matrix of “frequency of
inundation” and “degree of damage” as in the case of “site safety level.”
Based on this “site safety level,” SPG has developed disaster mitigation measures
for floodplains.

4.2.4 Floodplain Disaster Mitigation Measures

One of the floodplain disaster mitigation measures in the Basic Policy was to
regulate land use and building in high-risk locations. As shown in Fig. 4.10, Area A
is “in principle not included in the urbanization promotion area” in order to avoid
“severe property damage.” In Area A, inundation on the floor occurs frequently,
causing economic damage that makes it difficult to rebuild lives. Area B requires “a
structure with a floor higher than the expected inundation level to which evacuation
is possible” or “a strong structure that will not be washed out by the expected fluid
force” for building permits to avoid house washout and submersion that directly lead
to human suffering (regulations based on fluid force are not yet in operation (as of
July 2021)). In Area B, even if the frequency is rather low, human damage such as
submersion or loss of houses will occur.
However, with the enactment of the Omnibus Decentralization Act in 2000,
both of the above notifications were redesignated as legally nonbinding technical
advice. Therefore, SPG enacted a Flood Management Ordinance (March 31, 2014,
SPG Ordinance 55) to clarify the legal basis for regulation. In addition, Flood-
Resistant Building Guidelines (2015) were established to provide building permit
requirements under the Flood Management Ordinance.

Fig 4.10 Scope of risk covered by land use (Area a) and building regulations (Area b) (Source:
Taki et al. 2010; Shiga Prefectural Government 2012)
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 55

Land use and building regulations are nothing more than restrictions on the
private property rights guaranteed by the Constitution. In other words, the strength
of the regulations must be kept to a minimum in light of the socially accepted
norms at the time and after due consideration of the public interest to be secured,
such as the protection of life and property. Policy and legal considerations are
critical when applying a regulatory approach. Simply discussing the need without
fully identifying institutional common ground does not lead to social implementa-
tion.
Furthermore, floodplain disaster reduction measures impose a duty of care on
developers to ensure that the installation of continuous embankment structures,
such as road projects, does not cause significant changes in flood zones, for
example, by making bridges over floodplain. The “Guidelines for installation of
embankment structures pursuant to Article 25 of the Flood Management Ordinance
(2015)” have been released as technical standards. The preservation of open
levees, which serve as flood flow control facilities, and the development of armor
levees are included in the statutory river improvement plans for each area in the
prefecture.

4.3 Floodplain Management and Green Infrastructure

4.3.1 Establishing a “Floodplain Manager”: Social

Implementation and Points of Practice

Among the four measures for flood risk management (Table 4.2), river improve-
ments (river works) and basin storage measures (retention measures) primarily aim
to reduce the frequency of flood inundation through structural measures (levees and
flood control facilities) in the basin and river channel area. For river improvements
in particular, the River Law and related laws and regulations, as well as major
court cases, such as the Daito flood lawsuit and the Tama River flood lawsuit, have
clarified the scope of obligatory liability for facility management. Most basin stor-
age measures are installed by developers in accordance with prefectural ordinances
and guidelines, but some are positioned in river plains within the framework of
comprehensive flood control projects and installed by river administrators such as
public works offices of the Ministry of Land, Infrastructure, Transport and Tourism
or prefectural governments.
Measures to improve local disaster preparedness (preparation measures) have
been developed mainly by local governments with the support of river administra-
tors under various systems related to crisis management, such as the Flood Fighting
Law and the Disaster Countermeasure Basic Law, as a complementary measure to
river maintenance and basin storage measures.
In contrast, floodplain disaster mitigation measures (containment measures)
such as open levees, two-line levees, forest protection, and land use and building
56 K. Taki

Table 4.2 Classification of flood management measures (Shiga Prefectural Government 2012)
River works Measures to improve rivers and waterways in order to prevent
Measures to ensure safe flow floods from river overflowing. These include artificial levees,
of design floods in river reservoirs, and flood control dams
channels
Watershed works Measures to mitigate rapid flood runoff into rivers and
Basin retention measures waterways. These include rainwater retention in ponds,
grounds, forest soils, paddy fields, and reservoirs
Floodplain works Urban planning measures to minimize damage in the event of
Flood disaster reduction flooding that exceeds the capacity of river and waterway
measures facilities. These include ring levees, secondary levees, flood
prevention forests, land use regulations, and building flood
proofing
Community works Measures to support evacuation and flood prevention
Disaster resilience activities. These include disaster drills and the dissemination
improvement of disaster preparedness information

regulations have long been adopted as risk reduction measures but have not been
fully taken into account in modern flood control systems. The open levee, which is
said to have been developed by Shingen Takeda (1521–1657, a famous territorial
lord), was seen in many places in the past, but many of them have now disappeared
without finding a place in the current institutionalization. With regard to land use,
the risk of flooding in the area has not been treated as a major determinant during
subsequent annexation of an existing residential zone into an urbanized zone or
during the conversion of agricultural land.
Under the current legal system, the primary responsibility of river management
is to handle design floods within river channels. Therefore, when floodplain disaster
mitigation measures are implemented by river administrators, they have to be
implemented as an extension of river management for the purpose of handling
design floods. This means that part of the floodplain is considered a river channel
(river management facility) and used to handle design floods. For example, ring
levees and secondary levees that are constructed as part of a land use integrated
water disaster prevention project fall under this category. That is, the current
position is different from the idea that disaster mitigation is accomplished by
responding throughout the floodplain (floodplain management) rather than by using
river management facilities.
However, the current flood control system has been deeply adapted to the
socioeconomic needs of Japan since the enactment of the River Law in 1898,
with modifications added as necessary. Dramatically altering the system would
cause social friction and confusion. A practical and appropriate approach to
floodplain disaster mitigation practices, therefore, is to respect the various plans
implemented by river administrators as a given condition and add a separate system
of administration (floodplain managers) to complement them.
To enable the establishment of a complementary relationship, floodplain man-
agers should be given the objective of minimizing damage from external forces
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 57

beyond the capacity of facilities, regardless of the stage of river maintenance. A

separate system of administrative procedures (legal, organizational, and budgetary
measures) of river management should also be added.
Based on this idea, SPG established the goals and system of flood risk man-
agement as mentioned in Sect. 4.2. Specifically, an independent department in
charge of river basin-wide flood management (Floodplain Management Office) was
established, which is separate from the department in charge of rivers (River and
Port Division). This was successful, and floodplain disaster mitigation measures
can now be promoted in parallel with river maintenance, rather than as a binary
choice.

4.3.2 Artificial Structures and Green Infrastructure

The Science Council of Japan (2014) describes the benefits of artificial structures
as “achieving a single function that contributes to a specific and distinct purpose
with a high degree of precision (accurately providing the performance required by
the society),” and the benefits of green infrastructure as “maintaining and creating
diverse spaces that can provide a variety of ecosystem services and contribute to
the conservation of biodiversity.” Green infrastructure should be understood as
complementing rather than replacing the function of artificial structures.
In terms of the sole function of flood protection, green infrastructure is nowhere
near the standard of artificial structures. Nonetheless, although green infrastructure
may be less effective in a specific function, it exhibits diverse functions in normal
times. For example, there are still some flood prevention forests along the Azumi
River, which runs through the western part of Shiga Prefecture. Many are bamboo
groves, which not only mitigate flooding but also support rich riverine biota, provide
food, and have been used as materials to make the frames of traditional Japanese
fans and the like. In this way, green infrastructure, which is closely related to
lifestyle and culture, is itself a local resource and forms the character of the
region and its original landscape. This is a role that could not easily be played by
functionally oriented concrete embankments.
In advancing flood control policy, it is essential to find the best combination of
artificial structures and green infrastructure in accordance with the legal system and
regional characteristics, considering the characteristics of each.

4.3.3 Green Infrastructure: The Trump Card in Floodplain

Management

To ensure that administrative responsibilities are fulfilled under the river manage-
ment system based on River Law, artificial structures such as levees and dams
58 K. Taki

that reliably perform the required functions are essential. In contrast, attempts to
build artificial structures to prevent excess floods are met with irrefutable policy
arguments from the fiscal authority that the responsibility of dealing with design
floods should first be executed before engaging in excess flood control measures.
Therefore, in this author’s view, floodplain management must focus on a response
with green infrastructure that can perform a variety of functions during normal
times. From the above, the scope of defense by artificial structures and green
infrastructure can be summarized as follows.
River management: continues to play a role in ensuring the safe flow of design floods
within river channels, based on reliably functioning artificial structures to reduce
the frequency of flood inundation. Promote nature-oriented river development to
the extent that it does not impair flood control functions with respect to design
floods.
Floodplain management: addresses floods that exceed the capacity of artificial
structure complexes through Eco-DRR-oriented green infrastructure, including
land use, with a focus on “avoiding exposure” and “reducing vulnerability”.
As we have seen, Eco-DRR-oriented green infrastructure has a strong affinity
with floodplain disaster mitigation measures, but we cannot expect to see a
large amount of public investment on the so-called “extra” measures when river
management is still in its infancy. In addition, many of the current laws, such as
the Rivers Act, the Cropland Act, and the Forest Act, are organized according
to purpose. Hence, it is inherently difficult to develop something as multifaceted
as green infrastructure within a single legal framework. It is an issue of vertical
compartmentalization.
However, in the face of a declining population and climate change, we must
establish methods to solve problems by going beyond the vertical division of the
legal system and the administrative structure. For example, the disaster mitigation
effects and multifaceted functions of Eco-DRR-oriented green infrastructure should
be evaluated, and the ways to raise funds directly from financial markets should
be considered. I expect research for social implementation to rapidly advance from
now on.

References

ADRC (2015), Total disaster risk management - good practices

Hori T, Furukawa S, Fujita A, Inazu K, Ikebuchi S (2008) An optimal design framework of a flood
control system including in-floodplain countermeasures based on distributed risk assessment
- basic concepts and methodology. Jpn Soc Civil Eng 64(1):1–12. [in Japanese with English
abstract]
Natural Environment Bureau, Ministry of the Environment (2016) Approaches to disaster reduction
and mitigation using ecosystems [in Japanese]
Science Council of Japan (2014) Proposal: recommendations for the use of ecosystem infrastruc-
ture in reconstruction and national resilience [in Japanese]
4 Flood Management Policy in Shiga Prefecture, Japan: Implementation. . . 59

Shiga Prefectural Government (2012) Flood management basic policy in Shiga prefecture [in
Japanese]
Shiga Prefectural Government (2014) Flood management ordinance in Shiga prefecture. Shiga
Prefecture Ordinance No 55, March 31, 2014 [in Japanese]
Taki K, Matsuda T, Ukai E, Ogasawara Y, Nishijima T, Nakatani K (2010) Design for flood
disaster-reduction system in a floodplain of small-medium sized rivers. Adv River Eng 64:477–
482. [in Japanese with English abstract]
Taki K, Matsuda T, Ukai E, Nishijima T, Egashira S (2013) Method for evaluating flood disaster
reduction measures in alluvial plains. J Flood Risk Manag 6:210–218
Taki K, Yamashita K, Hirayama N, Takanishi S (2019) A study of regional disaster preparedness
improvement strategy based on hydraulic analysis of a small and medium-sized river system.
Adv River Eng 25:79–84. [in Japanese]

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 5
Toward Social Infrastructure:
Typological Idea for Evaluating
Implementation Potential of Green
Infrastructure

Takeshi Osawa and Takaaki Nishida

Abstract Green infrastructure (GI) comprises widely distributed objects in human

residential communities. However, because of the variety of certain objects, it is
sometimes difficult to improve public awareness and enhance social implementation
of GI. To expand the idea of GI and apply it widely in our society, we should
understand clearly what exactly GI is and where and how it can be applied. In
this article, we classify the types of GI and present a basic approach to evaluate
their implementation potential as the first step for expanding the application of
GI in human society. First, based on the definition of GI, we classified it as
the infrastructure involving the natural ecosystem, seminatural ecosystem, and
artificial ecosystem in each. The essential differences among these types arise
from their installability depending on human activities. Then, we considered the
principle of evaluation of the implementation potential of GI for the three types
GI based on three dimensions, natural condition, top-down regulation, and bottom-
up regulation, in human society. Additionally, appropriate ideas for the evaluation
of each dimension were considered. Relative importance of the natural condition,
top-down regulation, and bottom-up regulation differs among the types of GI. We
believe that the findings of this work will be useful for decision makers dealing with
the application of GI in their administrative areas.

Keywords Artificial ecosystem · Decision support · Driving force pressure state

impact response (DPSIR) framework · Installability · Seminatural ecosystem

T. Osawa ()
Graduate School of Urban Environmental Sciences, Tokyo Metropolitan University, Tokyo, Japan
T. Nishida
Kyoto Sangyo University, Kyoto, Japan
Mitsubishi UFJ Research and Consulting Co. Ltd., Osaka, Japan

© The Author(s) 2022 61

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_5
62 T. Osawa and T. Nishida

5.1 Introduction

Green infrastructure (GI) is defined as the natural, seminatural, and artificial

networks of multifunctional ecological systems within, around, and between human
residential communities on all spatial scales (Tzoulas et al. 2007; Pearlmutter
et al. 2017). This concept can cover a wide range of elements such as intact
forests (Svensson et al. 2019), wetland (Nakamura et al. 2019), agricultural land
(Osawa et al. 2020a), and urbanized green spaces (Tzoulas et al. 2007; Matsunaga
and Osawa 2021). Thus, GI can exist widely in human residential communities.
However, such a broad definition sometimes makes it difficult for decision makers,
such as government personnel who are the stakeholders for the development of
infrastructures, to perceive GI. For example, a policy maker who is responsible
for the mitigation of climate change tends to focus on the GI that is useful for
carbon fixation (e.g., Chen 2015). Health-care professionals, such as public health
department personnel and hospital operators, tend to focus on the GI that can
benefit the human health, both physical and mental (e.g., Tzoulas et al. 2007).
Although many categories of people have a potential interest in GI, integrating their
interests to link various GI implementation practices is a challenge. In addition,
local residents often cannot perceive the existing GI itself, i.e., the ecosystem
with ecosystem services existing around them (Zhang et al. 2010; Cortinovis
and Geneletti 2018). These challenges hamper the application of GI as social
infrastructure for the human society. To expand the idea and application of GI and
to contribute the human society, a clear understanding of what GI comprises and
where and how it can be applied is necessary.
Proposing a typology for conceptual term is one of the effective ways to perceive
future challenges for research and management (Eggermont et al. 2015). Thus, we
can consider that clarifying and classifying the concept of GI are important steps to
discuss the feasibility of the application of GI as social infrastructure. In this work,
we expanded the basic concept of GI to evaluate its implementation potential for the
human society and to improve its applicability as social infrastructure, especially for
land managers and decision makers. First, based on the basic definition of GI and the
concept of human society–environment interactions, we classified the GI for three
types of ecosystems, i.e., the natural, seminatural, and artificial ecosystems. Second,
we proposed the underlying concept of the components of GI implementation
potential for GI and considered basic methods to evaluate it. Finally, we discussed
the next challenges and perspectives for the application of GI.

5.2 Classification Basis for GIs

One of the challenges in expanding the idea and implementing GI in the human
society is to link with concept of that already widely accepted in other fields
(Romano et al. 2015; Spanò et al. 2017). Spanò et al. (2017) indicated the use-
5 Toward Social Infrastructure: Typological Idea for Evaluating. . . 63

fulness of the driving force–pressure–state–impact–response (DPSIR) framework

to integrate knowledge between diverse disciplines and GI (Svarstad et al. 2008).
This framework considers the driving forces (D) (e.g., human activity) that exert
pressure (P) (e.g., land-use change), leading to changes in the state (S) (e.g.,
ecological processes) and thereby cause impacts (I) and the human health and
society that may elicit a societal response (R) (Spanò et al. 2017). At the heart of
this framework is a simple incorporation of the interaction between human society
and environment. The GI should support the human society, and thus, GI should
interact with human activities. That is, for an ecosystem to be classified as GI,
humans should take appropriate efforts (driving force and pressure) to regulate
that ecosystem as required by humans. Interaction between the human society and
environment is an essential factor in determining the GI, and hence, this factor
should be considered when classifying GI.

5.3 Classification of GI

Based on the conceptual definition of GI considering the interactions between the

human society and environment, we can specify at least three types of elements,
namely, the natural ecosystem, seminatural ecosystem, and artificial ecosystem
as GI (Fig. 5.1). Examples of natural ecosystems as GI are coral reefs, which
can reduce the impact of ocean waves (Martin and Watson 2016), and natural
wetland, which can reduce flooding occurrences (Nakamura et al. 2019). Examples
of seminatural ecosystems as GI are paddy fields, which have a variety of functions
such as food production, wetland habitat support (Natuhara 2013; Osawa et al.
2020b), and disaster prevention (Osawa et al. 2020a, b). Artificial ecosystems as GI
include rain gardens, which are urban green spaces that store rain water and prevent
urban flooding (Ishimatsu et al. 2017). These three types have clear differences in
terms of their installability depending on human activities, i.e., the driving force
and pressure in the DPSIR framework (Fig. 5.1). To be specific, we cannot create a
natural ecosystem, and it is difficult to create a seminatural ecosystem by ourselves;
nevertheless, we can support these ecosystems and create an artificial ecosystem in
any area in theory. Thus, the implementation potentials for natural and seminatural
ecosystem-based GI are strongly regulated by the location of the target area, whereas
there are few such limitations on GI based on artificial ecosystems—this is an
important aspect related to the implementation potential of GI. Of course, it may be
possible to combine these types of GI. However, for simplicity, we have classified
them clearly in this work.
This classification is similar to the classification of nature-based solution (NbS),
proposed by Eggermont et al. (2015). They proposed that NbS could be applied
along with two evaluation axes. Those are (1) the extent of involvement of
biodiversity and ecosystems’ engineering in NbS and (2) the number of ecosystem
services and stakeholder groups that are targeted by a given NbS. Moreover, they
discussed that low level of axis 1 combined with high level of axis 2 constitutes
64 T. Osawa and T. Nishida

Fig. 5.1 Essential differences among the three types of green infrastructure (GI)

ecosystem with no or minimal intervention; both mid-levels are connected with

human-mediated ecosystem, namely, seminatural; and high level of axis 1 combined
with low level of axis 2 constitutes artificial ecosystem. Although they categorized
GI as the artificial ecosystem, which comprises high level of axis 1 combined with
low level of axis 2 (Eggermont et al. 2015), the basic idea established by interaction
between the human society and environment is common.

5.4 Three Dimensions for the Implementation Potential of GI

We propose three dimensions for evaluating implementation potential of GI (Fig.

5.2). The first is natural condition. Implementation potential of natural and seminat-
ural GI is basically regulated by their geographical setting. We cannot introduce a
coral reef as GI in a mountainous or hilly area. We cannot introduce a forest as GI
for the sea. Thus, implementation potential of a GI type should be defined based on
the geographical setting, i.e., the natural condition in the target area. This is one of
essential dimensions for evaluating the implementation potential of GI.
Second is top-down regulation (Fig. 5.2). Artificial ecosystem such as urban
green spaces can be created by humans. In other words, the implementation
potential of this type of GI is not as strongly regulated by the geographical setting
as in the case of the other two types. That is, in theory, artificial GI can be
introduced anywhere. However, such GIs are generally introduced at a large scale
by public work departments, not by individuals. For example, roadside trees are
often established and maintained by local governments. Thus, to introduce such
GI, the public administration should take interest. Therefore, the implementation
potential of this type of GI can be evaluated based on the situation of human society,
mainly the top-down regulation such as that by the government. This administrative
regulation is the second essential dimension for evaluating the implementation
potential of GI.
5 Toward Social Infrastructure: Typological Idea for Evaluating. . . 65

Natural condition
(Ecosystem in the target area)

Introduction Potential of GI

Top-down regulation Bottom-up regulation

(Regulatory rules) (Residents demands)

Fig. 5.2 Three essential dimensions to estimate the implementation potential for GI

Third is bottom-up regulation (Fig. 5.2). To decide the specific public work,
public administrations should respect the will of the residents. Thus, to introduce
GI in residential areas, the residents too should take interest in the GI. This
demand/requirement, i.e., bottom-up regulation, is the third essential dimension for
estimating the implementation potential of GI. Thus, the implementation potential
of GI is based on three essential dimensions that are often related to each other
(Fig. 5.2). We can clearly define and evaluate these aspects independently. Next, we
discuss these three dimensions more in detail.

5.4.1 Natural Condition

The first dimension, i.e., the natural conditions, is the easiest dimension to perceive
and evaluate because it is essentially the ecosystem existing in the target area. Thus,
the expected benefits of the GI are almost the same as the existing ecosystem
services. The aim of introduction of the GI should be to employ and/or enhance
their existing ecosystem services in the target area. For example, the Ministry
of Agriculture, Forestry and Fisheries, Japan, has adopted a GI-related policy
to enhance the ecosystem services produced from agricultural areas using direct
66 T. Osawa and T. Nishida

payment systems (MAFF web site https://www.maff.go.jp/j/nousin/kanri/tamen_

siharai.html, accessed on February 10, 2021). The Forestry Agency, Japan, has
made efforts to promote GI with efforts to maintain forests (Maeda 2017). Both
agricultural and forest ecosystems are already known to provide several ecosystem
services (Matsuno et al. 2006; Natuhara 2013; Osawa et al. 2020a, b); hence, these
GI strategies aim to use and maintain these existing ecosystem services. Therefore,
if a decision maker plans to apply natural and seminatural ecosystem GI in a specific
area, they must consider the local ecosystem and the type of ecosystem services that
can be obtained in this target area as the basis of GI. This is a very simple condition.
Additionally, the practitioner should focus on the benefits they expect from that
GI. For example, a paddy field is an extremely useful GI because of its multiple
functions besides food production: ecosystem services such as regulating service,
cultural service, and supporting service (Natuhara 2013). However, the relative
values of the services strongly depend on their local conditions. A previous study
reported a trade-off relationship between the provisioning service and supporting
service in a paddy field (Osawa et al. 2016). Further, the regulating service, in
specific, flood prevention, strongly depends on the location (Osawa et al. 2020a).
Therefore, to introduce a natural and seminatural ecosystem-based GI, practitioners
should consider the types of the existing ecosystems and the services that can be
extracted from them.

5.4.2 Top-Down Regulation

The second dimension, i.e., top-down regulation, refers to the social regulatory
rules, such as laws, administrative programs, and related individual rights. We
cannot consider any ecosystem as GI without considering such top-down regulation
even when the focusing ecosystem has high potential. For example, a protected area
such as a national park that is strongly regulated for any development work can be a
potential GI (Benedict and McMahon 2006), but its uses are restricted according
to the regulatory roles. In other words, within the protected area, the expected
ecosystem services, i.e., benefits to humans, may be regulated. For example, it is
difficult to receive any provisioning services from a closed zone in a protected
area. The decision maker should not consider these unavailable ecosystem services
in the focusing area as the benefit of GI. However, the regulatory roles could
promote the application of the area as GI. For example, according to the Urban
Park Act, Japan, urban green spaces are expected to provide the temperature
reduction effects in summer (https://www.kantei.go.jp/jp/singi/tiiki/kokusentoc_
wg/hearing_s/150123siryou03_1.pdf, in Japanese, accessed on February 10, 2021).
This type of ecosystem services is easy to use as the function of GI for urban
green space (Matsunaga and Osawa 2021). Both the introduction advisability and
expected ecosystem services are basically decided by the role of the human society.
Therefore, one effective way to evaluate the implementation potential of GI is to
review administrative documents. Administrative documents include laws, local
5 Toward Social Infrastructure: Typological Idea for Evaluating. . . 67

codes, and administrative programs such as city development, environment, and

water usage. One specific example of administrative document in the environment is
the establishment of the biodiversity strategy by the national government and some
local governments based on the recommendation by the Convention on Biological
Diversity. Any mentions of both the protected area and ecosystem services expected
from the current ecosystems in the document can be a key factor for introducing
GI in the area. These documents provide us hints regarding the implementation
potential of GI.
The top-down regulation is not only administrative but can also include individ-
ual rights such as property rights. To simplify, if a practitioner wants to apply GI
for a given forest land, the permission of the landowners is essential. Of course,
this type of regulation can exist for intact ecosystem. Such top-down regulations are
essential for applying GI because almost of all land, sea, and rivers, i.e., ecosystems,
are owned and/or managed by someone.

5.4.3 Bottom-Up Regulation: Residential Demands

and Requirements

The third dimension is also a part of social regulation, but this arises from bottom-up
regulation, i.e., the demand/requirements of local residents. If a government wants
to introduce GI in an administrative area, they should first estimate and visualize the
merits of the GI for residents because public works must be undertaken only if it
has public benefits. Unfortunately, estimating the benefits of GI in detail is difficult
because of the uncertainty of the multiple functions of GI, at least partly owing to
the limited knowledge regarding them at present. Thus, residential acceptability, i.e.,
the appreciation and understanding of ecosystem functions and/or services, is one
of the key factors determining the implementation potential of GI.
Residential appreciation and understanding of ecosystem function and/or ser-
vices are difficult to estimate because they are qualitative factors. Recently, these
qualitative factors were estimated quantitatively via interview and questionnaire
surveys (Cheng et al. 2019). People who have interest and/or affinity toward
ecosystem or natural environment may have a relatively high acceptability for the
introduction of GI as a part of the ecological system.

5.5 Evaluation of the Potential for Introduction of GI

To summarize, the three essential dimensions of evaluating the implementation

potential of GI are the natural condition, top-down, and bottom-up social regulations
(Fig. 5.2). Notably, relative importance of three dimensions among GI types
could differ (Fig. 5.3). For example, the natural ecosystem-based GI is strongly
68 T. Osawa and T. Nishida

(a)

Importance of natural condition

Natural
ecosystem

Seminatural
ecosystem

Artificial
ecosystem

Installability by humans

(b) (c)
Impoortance of top-down regulation

Importance of bottom-up regulation

Seminatural Artificial
ecosystem ecosystem

Natural Artificial
ecosystem ecosystem

Natural Seminatural
ecosystem ecosystem

Installability by humans Installability by humans

Fig. 5.3 Relative importance of three dimensions: (a) natural condition, (b) top-down regulation,
and (c) bottom-up regulation for each GI type

Table 5.1 Relative importance of each dimension in each type of GI

GI type Natural condition Top-down regulation Bottom-up regulation
Natural ecosystem High Moderate Low
Seminatural ecosystem Moderate High Low
Artificial ecosystem Low Moderate High

regulated by the natural condition, i.e., the first dimension (Fig. 5.3a). A seminatural
ecosystem-based GI is strongly regulated by top-down regulation, i.e., the second
dimension (Fig. 5.3b). An artificial ecosystem is strongly regulated by bottom-up
social regulation, i.e., the third dimension, but rarely regulated by the first dimension
(Fig. 5.3a, c). A decision maker should consider all the three dimensions when
introducing GI in the planning stage. Table 5.1 summarizes the relative importance
of each dimension for each type of GI, which can be used by the practitioners as
a checklist in the local planning of GI; at least the items marked “high” should be
carefully considered to introduce GI suitably for the target area.
5 Toward Social Infrastructure: Typological Idea for Evaluating. . . 69

5.6 Conclusion and Perspective

We classified the GI into three types on the basis of ecosystems and considered the
basic concept for evaluating the implementation potential of each type of GI in three
dimensions. Essentially the ubiquitous objects, i.e., ecosystems in our society, can
function as GI, but there exists a large variety in such elements. Some types of GI can
cover a wide area and provide several benefits, while others may exist locally and
provide fewer benefits. GI is a comprehensive concept of the use of an ecosystem for
the human society. Thus, the basic idea of GI itself is not new. However, the word
“infrastructure” refers to the basic systems and services for human society that are
managed by a country or an organization. Thus, the idea of GI may include the
concept that the human society should use natural environment more systematically
and effectively. Expanding the idea of GI is almost the same as establishing a society
that is in harmony with nature. Estimating the implementation potential of GI is an
important step to expand the GI for our society.

Acknowledgments This work was supported by the Environment Research and Technology
Development Fund (4-1805) of the Ministry of the Environment, Japan, and partially supported
by the JSPS KAKENHI Grant Number 20 K06096. Profs. F. Nakamura and N. Furuta provide us
many useful suggestions.

References

Benedict MA, McMahon ET (2006) Green infrastructure. Island, Washington, DC

Chen WY (2015) The role of urban green infrastructure in offsetting carbon emissions in 35 major
Chinese cities: a nationwide estimate. Cities 44:112–120
Cheng X, Van Damme S, Li L, Uyttenhove P (2019) Evaluation of cultural ecosystem services: a
review of methods. Ecosyst Serv 37:100925
Cortinovis C, Geneletti D (2018) Ecosystem services in urban plans: what is there, and what is still
needed for better decisions. Land Use Policy 70:298–312
Eggermont H, Balian E, Azevedo JMN et al (2015) Nature-based solutions: new influence for
environmental management and research in Europe. GAIA Ecol Perspect Sci Soc 24:243–248
Ishimatsu K, Ito K, Mitani Y et al (2017) Use of rain gardens for stormwater management in urban
design and planning. Landsc Ecol Eng 13:205–212. https://doi.org/10.1007/s11355-016-0309-
3
Maeda S (2017) Green infrastructure: forestry and trends in securing new Funcing sources. Q J
Public Policy Manag 2017:47–57. (in Japanese)
Martin TG, Watson JEM (2016) Intact ecosystems provide best defence against climate change.
Nat Clim Chang 6:122–124. https://doi.org/10.1038/nclimate2918
Matsunaga K, Osawa T (2021) A method for evaluating temperature reduction effect in urban
green space: case study in the Central Tokyo metropolitan. Jpn J Conserv Ecol. (in Japanese)
Matsuno Y, Nakamura K, Masumoto T et al (2006) Prospects for multifunctionality of paddy rice
cultivation in Japan and other countries in monsoon Asia. Paddy Water Environ 4:189–197.
https://doi.org/10.1007/s10333-006-0048-4
Nakamura F, Akasaka T, Ishiyama N et al (2019) Adaptation to climate change and conservation of
biodiversity using green infrastructure. River Res Appl:1–13. https://doi.org/10.1002/rra.3576
70 T. Osawa and T. Nishida

Natuhara Y (2013) Ecosystem services by paddy fields as substitutes of natural wetlands in Japan.
Ecol Eng 56:97–106
Osawa T, Kohyama K, Mitsuhashi H (2016) Trade-off relationship between modern agriculture
and biodiversity: heavy consolidation work has a long-term negative impact on plant species
diversity. Land Use Policy 54. https://doi.org/10.1016/j.landusepol.2016.02.001
Osawa T, Nishida T, Oka T (2020a) High tolerance land use against flood disasters: how paddy
fields as previously natural wetland inhibit the occurrence of floods. Ecol Indic 114:106306
Osawa T, Nishida T, Oka T (2020b) Paddy fields located in water storage zones could take over
the wetland plant community. Sci Rep 10:1–8
Pearlmutter D, Calfapietra C, Samson R et al (2017) The urban forest: cultivating green
infrastructure for people and the environment. Springer, Cham. https://doi.org/10.1007/978-
3-319-50280-9
Romano G, Dal Sasso P, Trisorio Liuzzi G, Gentile F (2015) Multi-criteria decision analysis for
land suitability mapping in a rural area of Southern Italy. Land Use Policy 48:131–143. https:/
/doi.org/10.1016/j.landusepol.2015.05.013
Spanò M, Gentile F, Davies C, Lafortezza R (2017) The DPSIR framework in support of green
infrastructure planning: a case study in Southern Italy. Land Use Policy 61:242–250. https://
doi.org/10.1016/j.landusepol.2016.10.051
Svarstad H, Petersen LK, Rothman D et al (2008) Discursive biases of the environmen-
tal research framework DPSIR. Land Use Policy 25:116–125. https://doi.org/10.1016/
j.landusepol.2007.03.005
Svensson J, Andersson J, Sandström P et al (2019) Landscape trajectory of natural boreal forest
loss as an impediment to green infrastructure. Conserv Biol 33:152–163
Tzoulas K, Korpela K, Venn S et al (2007) Promoting ecosystem and human health in urban areas
using green infrastructure: a literature review. Landsc Urban Plan 81:167–178
Zhang B, Li W, Xie G (2010) Ecosystem services research in China: progress and perspective.
Ecol Econ 69:1389–1395

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
 Part II
Forest Ecosystem
Chapter 6
Riparian Forests and Climate Change:
Interactive Zone of Green and Blue
Infrastructure

Futoshi Nakamura

Abstract Riparian forests are recognized as green/blue infrastructure that provides

various ecosystem services, including water transport, shade from sunlight energy,
supply of leaf litter, input of terrestrial insect, delivery of large wood, maintenance
of water quality, and corridors for wildlife. However, these forests have already
experienced negative impacts from various anthropogenic stressors, such as forest
cutting, agricultural development, river regulation, and dam construction. Moreover,
the climate projections for the twenty-first century in Japan indicate that mean
precipitation may increase by more than 10%, and other projections predict an
increase in the frequency of high-magnitude floods and a reduction in the discharge
of snowmelt floods. In this chapter, we describe probable changes in the structure
and function of riparian forests that might result and describe adaption strategies to
reduce the potential impacts of climate change on stream and riparian ecosystems
exacerbated by anthropogenic stressors.

Keywords Riparian zone · Climate change adaptation · Green/blue

infrastructure · Ecological function · Ecosystem services · Restoration

6.1 Introduction

A “riparian forest” refers to a forest that is transversely distributed from a riverbank

to its floodplain at the foot of adjacent hillslopes (Fig. 6.1). Riparian forests include
not only forests associated with stream/river courses but also those around wetlands
and lakeshores. Riparian forest communities can survive intensive and frequent
disturbance regimes through physiological and structural adaptations while utilizing
the materials (sediment, nutrients, and organic matter) trapped by these disturbance
events in complex river geomorphic surfaces (i.e., secondary channels, point bars,

F. Nakamura ()
Laboratory of Ecosystem Management, Graduate School of Agriculture, Hokkaido University,
Sapporo, Japan
e-mail: [email protected]

© The Author(s) 2022 73

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_6
74 F. Nakamura

Fig. 6.1 Riparian forest of Satsunai River, Japan

and backswamps) as habitats for their growth and maintenance (Nakamura and
Inahara 2007).
A riparian forest longitudinally extends from headwater streams to lowland rivers
at the catchment scale (Niiyama 1987). The disturbance regime also changes from
headwaters to low-gradient alluvial rivers (Nakamura et al. 2000; Nakamura and
Swanson 2003) (Fig. 6.2). A dominant disturbance in headwater streams is mass
movement, such as landslides and debris flows, characterized by their intensity
but rare occurrence. In downstream, braided reaches, seasonal flood disturbances
generate frequent high flows that form extensively developed geomorphic surfaces
on floodplains. Low-gradient alluvial rivers in developing peat marshes rarely
experience disturbances that result in landform alterations, although prolonged
inundation in a marsh can cause substantial physiological stress to trees. Thus, a
variety of disturbance types, frequencies, and intensities can be found at a catchment
scale and may have diverse ecological consequences for riparian communities. In
general, the structure and composition of riparian forests are determined by numer-
ous environmental gradients, such as longitudinal and cross-sectional elevation,
substrate (or soil) conditions, and disturbance frequency (Nakamura et al. 1997;
Shin and Nakamura 2005; Sakio 1997).
6 Riparian Forests and Climate Change 75

Headwater basin
¾ Landslide and earthflow
¾ Debris flow

Alluvial fan
¾ Flood
¾ Channel avulsion and migration

Natural levee and backmarsh

¾ Flood, Waterlogging
¾ Channel migration

Fig. 6.2 Riparian forests and dominant disturbance regime from headwater basins to lowland
alluvial rivers

Moreover, riparian forests and ponds in floodplains provide essential habitats

for various terrestrial and aquatic organisms in all or specific stages of their life
cycles. Thus, biodiversity in a riparian forest is generally higher than that in upland
ecosystems, and riparian habitat specialists increase overall regional biodiversity
(gamma diversity) at the catchment scale. Species richness, especially that of plants,
is commonly higher in riparian forests than in upland forests, mainly because
complex and diverse mosaics of habitat patches are created through land-water
interactions (Amoros and Bornette 2002).
The species richness and diversity of macroinvertebrates and fishes in floodplain
waterbodies vary along the gradients of hydrological connectivity (Gallardo et
al. 2014; Schomaker and Wolter 2011; Ishiyama et al. 2014). Some waterbodies
close to the main channel in a floodplain experience frequent floods, while other
waterbodies more distant from the main channel may be disturbed only during
high flood events or extremely large floods. Some waterbodies may be connected
by drainage channels and/or small tributaries with other ponds or mainstems.
These differences in connectivity alter the potential frequency of immigration
and emigration of individuals and the physical and chemical characteristics of
76 F. Nakamura

waterbodies. Consequently, taxon richness and diversity as well as species traits vary
with connectivity gradients (Ishiyama et al. 2014). Riparian forests are recognized
as a key habitat for the conservation of bird communities. Bird assemblages in
riparian zones are distinctively different from those in neighboring areas, owing
to differences in the plant communities (Yabuhara et al. 2019). Additionally, the
species richness and abundance of birds are generally higher in riparian zones than
in adjacent upland habitats due to the structural complexity of the riparian habitat.
Riparian forests provide essential ecological functions through a variety of mech-
anisms, including reducing sunlight and influencing in-stream primary production
and water temperature, supplying food resources for aquatic invertebrates (e.g.,
fine organic matter, leaves, wood), supplying large wood to streams, filtering fine
sediment and nutrients before their entry into streams, adding large roughness ele-
ments that reduce water velocity and protect stream banks, and providing migratory
corridors for animals (Gregory et al. 1991; Naiman and Decamps 1997) (Fig. 6.3).
A riparian forest is a unique landscape element that consists of diverse plant species,
provides essential habitats for various terrestrial and aquatic organisms, and plays
a pivotal role in the functioning of ecological processes between terrestrial and
aquatic ecosystems. Thus, we can refer to a riparian forest as an “interactive zone
of green and blue infrastructure” that should be preserved and restored for the
twenty-first century. Here, green/blue infrastructure is defined as “a strategically
planned network of natural and seminatural areas with other environmental features
designed and managed to deliver a wide range of ecosystem services such as water
purification, air quality, space for recreation and climate mitigation and adaptation”
(European Commission 2016). Green infrastructure refers to terrestrial elements like
trees, parks, hedgerows, and riparian vegetation, while blue infrastructure refers to
aquatic elements such as rivers, wetlands, ponds, estuaries, and coasts.
However, in lowland areas of river basins, riparian forests occur on wet, fertile
soil with flat topography. As a result, they have been exploited historically to
harvest timber, construct forestry roads, cultivate farmlands, and develop urban

Fig. 6.3 Structures and functions of the riparian zone (from Nakamura and Yamada 2005)
6 Riparian Forests and Climate Change 77

Fig. 6.4 Fragmentation of the riparian forests along the Toikanbetsu River, Japan (from Nagasaka
and Nakamura 1999)

land uses (Sedell and Froggatt 1984; Nagasaka and Nakamura 1999) (Fig. 6.4).
Dams greatly alter flow, sediment, and large wood regimes and thereby riparian
vegetation (Nakamura 2017). The loss and degradation of riparian forests cause
serious environmental problems; therefore, scientists and land managers have
developed management approaches to conserve and restore riparian zones (e.g.,
riparian buffers, reforestation, channel remeandering, floodplain reconnection).
Another key driver that may greatly alter riparian forest ecosystems and their
functions is climate change (Capon et al. 2013). In this chapter, we explore new
roles of riparian forests under climate change scenarios and discuss a process
for incorporating green/blue infrastructure for adaptation strategies to conserve
biodiversity and address effects of climate change. Although this chapter focuses
on riparian forests in Japanese landscapes, the derived adaptation strategies and
principles are applicable to other parts of the world.

6.2 Riparian Forest as an Interactive Zone of Green

and Blue Infrastructure

River and floodplain ecosystems are among the most sensitive of all ecosystems to
climate change because they are tightly linked to atmospheric thermal regimes and
global hydrological cycles, and they are affected by interactions between climate
change and non-climatic stressors (e.g., farmland development, river regulation,
78 F. Nakamura

impoundments, water pollution, urbanization, hydropower development) (Ormerod

2009; Thomas et al. 2016). According to the World Wide Fund for Nature (WWF)
Living Planet Report 2014 Living Planet Index (LPI: a measure of the state of the
world’s biological diversity based on population trends in vertebrate species), the
freshwater index has shown the greatest decline of any of the biome-based indices
(WWF 2014).
Maintenance and enhancement of ecosystem services provided by riparian
forests (Fig. 6.3) are important issues in biodiversity conservation and adaptation
to climate change (Palmer et al. 2009). Stream temperature is the primary water
quality variable that affects animal species distribution, phenology, and ecosystem
processes and is expected to increase with rising air temperature under climate
change (Thomas et al. 2016). Riparian forest canopies significantly reduce the
amount of solar radiation reaching a stream surface. The removal of a riparian
forest greatly increases insolation and thereby increases stream temperatures. A field
study in northern Hokkaido streams suggested that the abundance of masu salmon
is primarily limited by water temperature (Inoue et al. 1997). Masu salmon density
in lowland grassland reaches, where the summer maximum water temperature is
generally higher than 20 ◦ C, was lower than density in nearby forest reaches.
Effects of riparian canopies on stream water temperature have been examined
in streams flowing through deciduous broad-leaved forests (Nakamura and Dokai
1989; Sugimoto et al. 1997). Nakamura and Dokai (1989) developed a heat budget
to analyze the effect of riparian canopies and concluded that 86% of the daily total
input of solar radiation is intercepted by the riparian canopy and its removal would
result in a 4 ◦ C increase in water temperature. Thus, the use of riparian forest
shade to mitigate potential increases in stream temperature with climate change is a
fundamental action in adaptation strategies for protecting cold-water species (Seavy
et al. 2009; Wilby et al. 2010; Suzuki et al. 2021).
A riparian forest supplies various types of litterfall to streams (Benfield 1997).
Litterfall is defined as allochthonous organic matter that enters streams from riparian
vegetation. It includes leaves and leaf fragments, floral parts, bark, wood (stems,
branches, twigs), cones and nuts, fruits, and other plant parts. Among these, the
food quality of leaf litter is substantially greater than the others. Litter may reach a
stream by directly falling into the stream or by laterally moving into the stream.
Plant litter supplied from riparian forests is consumed by stream invertebrates.
Terrestrial invertebrates fall into streams from riparian forests, providing important
food resources for stream-dwelling fishes (Kawaguchi and Nakano 2001). In
addition to the supply of leaf litter and terrestrial invertebrates into streams, large
wood and twig accumulation, gravel bars, pools, and complex stream edges retain
transported food resources for aquatic invertebrates and fishes. However, we expect
more frequent, high flood peaks with climate change. In Japan, the fluctuation in
yearly precipitation has increased since the 1970s, and the frequency of hourly
heavy rains of 50 mm or more has increased (Ministry of the Environment et
al. 2018). Rivers in developed countries have been regulated for over a century,
and naturally meandering rivers have been altered to form straighten channelized
rivers (Nagayama and Nakamura 2018). Thus, organic matter supplied from riparian
6 Riparian Forests and Climate Change 79

forests into streams may be flushed downstream without retention on riverbeds,

making it less available to stream biota.
Large wood is supplied from riparian forests into streams and rivers by various
processes, such as landslides, debris flow, floods, windthrow, and bank erosion
(Harmon et al. 2004). Large wood alters the structural and functional characteristics
of stream channels, thereby influencing stream biota. In particular, the formation of
pools by large wood has been investigated with reference to fish habitats (Fausch
and Northcote 1992). Another important function of in-stream wood is modifying
the movement and transformation of energy, nutrients, and food for stream-dwelling
organisms. Step-pool sequences created by large wood provide storage sites for
sediment, organic matter, and nutrients (Nakamura and Swanson 1993). It is widely
recognized that trees growing along streambanks protect against bank erosion
through their root system and create complex habitat along river margins (Gregory
et al. 2019). Trees, however, will eventually fall into streams when flood power
exceeds their tolerances. In this situation, riparian forests may be eroded and
uprooted and transported downstream by floods. Wood-laden floods may destroy
artificial levees, bridges, and houses along the river course, and the intensity and
possibility of such disasters will increase with climate change. Thus, river managers
have to balance the ecological benefits and geomorphic risks, both of which are
important issues in a changing environment. The risks generated by floated wood
should be avoided in urban areas, whereas the ecological functions of large wood
should be sustained in rural and seminatural areas.
The effects of a riparian zone on water quality are well-known functions,
particularly the reduction of sediment and nutrients before their entry into streams
(Nakamura and Yamada 2005). Fine sediment eroded in uplands is efficiently
removed by riparian vegetation that is 10–30 m in width, while nutrients such
as nitrogen and phosphorus are efficiently removed by riparian vegetation that is
10–50 m in width. However, their effects vary with a number of local conditions,
such as sedimentation rates, surface and subsurface drainage characteristics, soil
characteristics (e.g., particle size composition and groundwater-level regulating
redox potential), organic matter content, temperature, vegetation type (e.g., grass or
forest, forest age, and density), nutrient loading from uplands, and slope steepness.
Soil erosion and nutrient leaching from uplands are predicted to increase with
increasing rainfall intensity under climate change (Jeppesen et al. 2009). Thus, the
filtering function of riparian forests is a critical process in maintaining the health of
river and wetland ecosystems.
Riparian forests are corridors in fragmented landscapes that connect healthy
intact habitats and allow animals to move between them. Corridor use by bird
and mammal species has been well investigated relative to that of other types of
wildlife (Machtans et al. 1996; Hilty and Merenlender 2004). Terrestrial wildlife
moves preferentially along or inhabits riparian forests. Wide, well-preserved cor-
ridors maintain a nearly complete species assemblage of the region. The home
range of animals may change with climate change, driving animals to migrate
more appropriately to higher-elevation areas. Thus, the current connectivity of
riparian and stream corridors should be protected to facilitate climate-induced
80 F. Nakamura

movement and restored when land use and dam construction impede their connec-
tivity.

6.3 Adaptation Strategies to Climate Change Using Riparian

Green/Blue Infrastructure

Riparian forests have been acknowledged as green/blue infrastructure that provides

the various ecosystem services explained above. However, these forests have already
experienced negative impacts from various anthropogenic stressors, such as forest
cutting, agricultural development, river regulation, and dam construction. These
stressors generally have detrimental effects on stream and floodplain biota, and
climate change may magnify negative impacts (Palmer et al. 2009). Thus, strategies
to conserve biodiversity at the landscape level must reduce or eliminate negative
impacts of existing anthropogenic stressors on riparian ecosystems and anticipate
the possible interactions of past impacts with climate change are key management
needs for the future.

6.3.1 Maintain River Dynamics Using Artificial Floods

The structure and composition of riparian vegetation have been maintained by

frequent migration of river channels and the active movement of sediment (Naka-
mura and Inahara 2007). Plant and animal species in streams and riparian zones
are adapted to the dynamic features of rivers and floodplains (Poff et al. 1997).
Additionally, the large wood produced by forests plays a critical role in providing
a wide array of habitats for plants, invertebrates, and fishes (Nakamura et al.
2012; Nakano et al. 2018; Fausch and Northcote 1992). Thus, the dynamics of
water, sediment, and large wood are three key components that maintain the
regeneration of riparian plant species and thereby riparian ecosystems (Nakamura
et al. 2017). However, water, sediment, and large wood regimes are greatly
altered by dams and land use in Japan and other developed countries which
results in drastic changes in the dynamic and interactive features of rivers and
floodplains. These regime shifts of physical conditions convert diverse river habitats
to stable, single-thread channels and increased forest establishment on gravel
bars and floodplains (Nadler and Schumm 1981; Johnson 1994; Bejarano et al.
2011). Riparian forests play a pivotal role in maintaining stream and floodplain
ecosystem, but they should be sustained by dynamic feature of the above three
components. In natural rivers, pioneer tree species dominate gravel-bed rivers,
and unvegetated gravel bars are essential sites for these species to germinate. In
contrast, mid- and late-successional tree species start to colonize under pioneer
trees and on gravel bars in dam-regulated rivers. The spread of late-successional
6 Riparian Forests and Climate Change 81

Fig. 6.5 Forest expansion over the gravel bars and floodplains of the Satsunai River

trees over gravel bars potentially threatens the survival of native species that are
dependent on unvegetated habitat (Tiedemann and Rood 2015; Nakamura and Shin
2001).
One of the main causes of forest expansion in gravel rivers is dam construction
and reservoir management (Takahashi and Nakamura 2011). Dams regulate water
discharge by storing floodwater in their reservoirs and releasing water resources
for power generation and agricultural use, which generally results in leveling the
fluctuation in water discharge. The stabilization of water discharge and substantial
reduction in floodwater discharge promote colonization of tree species on gravel
bars and forest expansion over valley floors (Fig. 6.5). Dams and other human
activities, such as channelization and revetment or spur-dike construction, promote
channel incision and forest colonization over gravel bars and floodplains. Forest
expansion causes two major problems for river management. One issue is an
increase in channel roughness, which reduces its floodwater transport capacity,
and the other issue is a reduction in gravel-bed habitat on which native, rare
plant, and animal are dependent species (Nakamura et al. 2020; Yabuhara et al.
2015).
The climate projections for the twenty-first century in Japan indicate that mean
precipitation may increase by more than 10% (Kimoto et al. 2005), and other
projections predict an increase in the frequency of high-magnitude floods and
a reduced discharge from snowmelt floods. With forest expansion and climate
change, we expect more frequent inundation with floodwaters due to the increase
in roughness elements and an increase in log-laden floods as riparian forest trees
82 F. Nakamura

Fig. 6.6 Wood-laden flood caused by heavy rain in northern Kyushu in 2017

are removed by extreme floods (Fig. 6.6). Additionally, reduced discharge from
snowmelt floods may cause recruitment failure of pioneer tree species such as
Populus and Salix spp., which time their seed dispersal periods to the descending
period of snowmelt floods (Nakamura and Inahara 2007).
Environmental flow regimes, including artificial floods, can be used to mitigate
magnified effects of dams and climate change. In the Satsunai River, the Japanese
government launched a restoration project in 2012 to partially restore its riparian
ecosystems, releasing a maximum water volume of 120 m3 /s, with a 2-year
return-period flood before dam construction (Nakamura et al. 2020). This was a
large-scale experiment developed jointly with an interdisciplinary science team
and river managers, who conducted monitoring and evaluations under an adaptive
management scheme. Artificial floods have been initiated once a year at the end of
June since 2012, and they have been synchronized with the seed dispersal period of
Salix arbutifolia, which is endangered species and a high conservation priority for
the project (Fig. 6.7). The project is successful so far in restoring a shifting mosaic of
floodplain habitat patches and gravel bars and thereby in regenerating S. arbutifolia
(Nakamura et al. 2020). With this project, an increase in dense forest cover has been
minimized within the flood-disturbed area and regulated by the current flow regime,
including artificial floods.
6 Riparian Forests and Climate Change 83

150

Natural Controlled Seed dispersal period

of Salix arbutifolia
Water discharge(m3/s)

100

Artificial flood

50

0
April 1 May 1 June 1 July 1 Aug. 1

Hydrographs (water discharge) of natural flows (flows into the dam from the
upper basin) and controlled flows (flows out of the dam)

Fig. 6.7 Hydrographs (water discharge) of natural flows (flows into the dam from the upper basin)
and controlled flows with artificial floods (flows out of the dam) in the Satsunai River (from
Nakamura et al. 2020)

6.3.2 Maintain Riparian Forests Continuously

from Headwater Streams to Lowland Rivers

In Japan, considerable areas of hills and plains that were originally covered by
forests have been converted to farmland, with streams and rivers being channelized
to develop and conserve farmlands (Nagasaka and Nakamura 1999). One of the crit-
ical roles of riparian forests is the maintenance of cold water through their canopies;
therefore, cold-water species, such as salmonids, are able to inhabit streams in
Hokkaido, northern Japan. A laboratory experiment and field study indicated that
feeding and growth of masu salmon decline when water temperatures reached
24 ◦ C (Sato et al. 2001). Summer maximum water temperature is significantly
related to the longitudinal extent of canopy removal (Sugimoto et al. 1997). This
study concluded that removal of 1 km of canopy would be associated with a 2–
4 ◦ C increase in water temperature, and a 6-km canopy removal would result in
a 5–8 ◦ C increase. Based on these data, they estimated a historical change in the
summer maximum water temperature in a small stream and determined that stream
temperature increased by 6 ◦ C from 1947 (22 ◦ C) to 1990 (28 ◦ C), with a sharp
increase during the 1950s as a result of rapid agricultural development. The results
suggest that suitable habitat for masu salmon decreased historically as a result of
these land use changes.
84 F. Nakamura

Water temperature is expected to increase with increasing air temperature with

climate change. However, studies that predict an increase in water temperature
are scarce and limited by uncertainties associated with catchment geohydrology.
In particular, porous volcanic rock is extensive across the Japanese archipelago,
enhancing the contribution of groundwater discharge. This scenario insulates areas
from exposure to sunlight energy and exchanges of latent and sensible heats and
creates spring-fed streams with water temperatures that are cool in summer and
warm in winter compared with those of runoff-fed streams. Spring-fed streams
may provide refuges for cold-water fishes in hot summers under climate change
conditions. Riparian forests will reduce the sunlight energy reaching the stream
surface; therefore, maintaining continuous belts of riparian forests from headwater
streams to lowland rivers is an adaptation strategy for climate change. In particular,
the shading effect of a riparian forest canopy dominates in headwater streams
where the stream width is relatively smaller than the size of the riparian forest
canopy. In the case of Japanese streams, shading by riparian canopies is effective
in streams less than 30 m in width (Nakamura and Dokai 1989). Thus, riparian
cover should be protected along headwater streams as well as spring-fed streams
to mitigate increases in air temperature. Such continuous belts of riparian forests
will provide corridors for terrestrial birds and mammals to migrate to cooler
environments.

6.3.3 Remove or Improve Check Dams for Material Transport

and Fish Migration

Management agencies have constructed many check dams in headwater streams in

Japan to prevent human lives and infrastructure from being affected by sediment-
related disasters (e.g., landslides, debris flows, and hyperconcentrated flows) and to
control erosion. Some check dams are not equipped with fish ladders, and prevent
or reduce upstream migration of socially and economically important fish. Even
if check dams have fish ladders, all ladders reduce passage to some degree, the
continuity and dynamics of sediment transport are interrupted, and some organic
matter produced from riparian forests is trapped by dams and buried by sediment.
Check dams substantially alter the natural sediment regime, which sustains riparian
forests in headwater basins (Nakamura et al. 2017).
Check dams greatly alter channel morphology and degrade salmonid habitats
(Kishi and Maekawa 2009). In general, the channel above check dams becomes
wider and shallower due to the wide dam face, and riparian trees become established
farther away from the wide stream channel. Thus, the riparian canopy cover over the
stream surface is reduced, and greater amounts of sunlight reach the stream surface
and increase water temperature. On the Shiretoko Peninsula, currently designated
as a World Natural Heritage Site, forest canopy cover over the stream surface is less
than 10% in dam-installed reaches in contrast to 80–90% in natural reaches, and the
6 Riparian Forests and Climate Change 85

maximum water temperature is 15–21 ◦ C in dam-installed reaches versus 12–17 ◦ C

in natural reaches (Kishi and Maekawa 2009).
Check dams may exacerbate the thermal effects of global warming, resulting
in greater heat energy exchange between air and waterbodies and even greater
increase in stream temperature. To reduce the potential impacts of climate change
on stream ecosystems exacerbated by anthropogenic stressors, check dams should
be improved and modified to allow fish migration, restore the natural regime of
sediment and organic matter, and ensure stream temperatures remain cool. Partial
removal of check dams is a modification alternative that recovers quasi-natural
systems and maintains disaster prevention functions. The center of the dam is cut
to create a rectangular opening or slit (Nakamura and Komiyama 2010; Nagayama
et al. 2020). The width of the opening is set to equal the low-flow channel width,
thereby maintaining the natural transport of water, sediment, and organic matter
under ordinary conditions and allowing fish to migrate upstream and downstream
(Fig. 6.8). When debris flows arrive at the dam site, all sediment and turbid water
cannot pass through the slit simultaneously, depositing sediment behind the dam
due to the backwater effect. Subsequently, the trapped sediment behind the dam is
gradually released from the slit during the descending limb of hydrographs and/or
subsequent floodwaters. Riparian forests establish along the low-flow channel and
their canopies cover the stream surface to a greater extent.

Fig. 6.8 Dam modification and recovery of up- and downstream linkages at the Shiretoko World
National Heritage Site
86 F. Nakamura

6.3.4 Broaden Riparian Forests and Remeander Rivers Where

Possible

Based on the climate change projection for 2040 under the RCP2.6 scenario in
Japan, the average air temperature will increase by 2 ◦ C from the current level,
rainfall with a 100- to 200-year recurrence interval will increase by 10%, and
floodwater discharge will increase by 20%. Thus, we must prepare for megafloods
and associated environmental problems. We can expect an increase in the discharge
of sediment, organic matter, and nutrients from upper basins. A study conducted in
the Kushiro River basin, northern Japan, under the harshest climate change scenario
(RCP 8.5), predicted that suspended sediment, total nitrogen, and total phosphorus
will increase by 4.3–8.3-, 2.3–3.3-, and 2.6–4.2-fold, respectively, by the end of the
twenty-first century (Ministry of the Environment 2020).
Adaptation actions to expand riparian vegetation cover on floodplains is a
potential strategy to increase floodwater retention capacity and reduce floodwa-
ter velocity, which will reduce peak discharge of megafloods and increase the
retention of transported materials (e.g., sediment, organic matter, and nutrients)
in floodplains (Dybala et al. 2019). Remeandering artificially straightened rivers
associated with the extension of floodplain areas will enhance the efficacy of
this management strategy by increasing inundation and retention of carbon and
nutrients in floodplains (Dahm et al. 1995; Nakamura et al. 2014) (Fig. 6.9).
Movement of large wood into streams will increase with bank erosion associated
with gradual movement of meandering river courses. Furthermore, large wood
pieces that are transported by floodwater from upstream to downstream and that
may destroy downstream infrastructure and houses during megafloods can be
efficiently trapped on the outside of meandering rivers due to the inertial force of
floodwater.
The adaptation strategies of extending floodplains and remeandering rivers also
create various habitats for aquatic and terrestrial plants and animals. This approach
is similar to the “Room for Rivers” concept in the EU (Buijse et al. 2002).
The historical development of agricultural and urban land use converts naturally
meandering rivers to straight river channels with revetments, disconnecting stream
from their floodplain ecosystems. Adaptation measures will restore connectivity
between rivers, ponds, and floodplain wetlands with increasing floodwater inun-
dation. Ishiyama et al. (2014) investigated the effects of wetland connectivity on
the species richness of aquatic invertebrates and fishes in agricultural landscapes.
They found that the species richness of highly mobile groups of insects and fishes
increased with increasing spatial connectivity. Thus, the frequent inundation of
floodwater over riparian forests increases temporal and spatial connectivity among
various waterbodies in floodplains and rivers.
6 Riparian Forests and Climate Change 87

Flow direction

During the restoration

Before the restoration (June 12, 1999) (February 23, 2010)

Restored reach

After the restoration

(August 3, 2011)

Fig. 6.9 Meander restoration in the Kushiro wetlands. The top left-hand photograph shows a 1.6-
km stretch of the main channel before it was restored by reconnecting the remaining former cutoff
channel and backfilling the straightened reach (from Nakamura et al. 2014)

6.4 Conclusion

A riparian forest is regarded as an “interactive zone of green and blue infrastruc-

ture,” and should be preserved and restored for the twenty-first century because
it provides essential habitats for plant and animals and precious ecological func-
tions in a changing environment. However, forests have already experienced
negative impacts from various anthropogenic stressors, such as forest cutting,
agricultural development, and river regulation due to its high productivity and
geomorphic setting close to rivers. However, predictions of future climate change
and associated impacts on biodiversity and ecosystem services project that the
interaction of anthropogenic stressors and climate change will magnify those
impacts. If those stressors are not reduced in the future, biodiversity and riparian
ecosystems would not be sustained and important ecosystem services would
be lost. Thus, active management and conservation strategies should be imple-
mented to lessen non-climate stressors. Ecological restoration projects conducted
throughout the world to protect and restore biodiversity and ecosystem services
will be essential components of adaptation strategies for climate change in the
future.
88 F. Nakamura

Acknowledgments The author deeply appreciates all thoughtful comments and suggestions made
by Dr. Stanley Gregory that significantly improve the quality of the original manuscript. This
research was funded by the Environment Research and Technology Development Funds (4-1504
and 4-1805) of the Ministry of the Environment of Japan.

References

Amoros C, Bornette G (2002) Connectivity and biocomplexity in waterbodies of riverine flood-

plains. Freshw Biol 47:761–776
Bejarano DM, Nilsson C, del Tanago GM, Marchamalo M (2011) Responses of riparian trees and
shrubs to flow regulation along a boreal stream in northern Sweden. Freshw Biol 56:853–866
Benfield EF (1997) Comparison of litterfall input to streams. J N Am Benthol Soc 16:104–108
Buijse AD, Coops H, Staras M, Jans LH, Van Geest GJ, Grift RE et al (2002) Restoration strategies
for river floodplains along large lowland rivers in Europe. Freshw Biol 47(4):889–907
Capon SJ, Chambers LE, Mac Nally R, Naiman RJ, Davies P, Marshall N et al (2013) Riparian
ecosystems in the 21st century: hotspots for climate change adaptation? Ecosystems 16(3):359–
381
Dahm CN, Cummins KW, Valett HM, Coleman RL (1995) An ecosystem view of the restoration
of the Kissimmee River. Restor Ecol 3:225–238
Dybala KE, Matzek V, Gardali T, Seavy NE (2019) Carbon sequestration in riparian forests: a
global synthesis and meta-analysis. Glob Chang Biol 25(1):57–67
European Commission (2016) Supporting the implementation of green infrastructure, Rotterdam
Fausch KD, Northcote TG (1992) Large woody debris and salmonid habitat in a small coastal
British Columbia stream. Can J Fishery Aquat Sci 49:682–693
Gallardo B, Doledec S, Pallex A et al (2014) Response of benthic macroinvertebrates to gradients in
hydrological connectivity: a comparison of temperate, subtropical, Mediterranean and semiarid
river floodplains. Freshw Biol 59:630–648
Gregory SV, Swanson FJ, McKee WA, Cummins KW (1991) An ecosystem perspective of riparian
zones. Bioscience 41:540–551
Gregory SV, Wildman R, Hulse D, Ashkenas L, Boyer K (2019) Historical changes in hydrology,
geomorphology, and floodplain vegetation of the Willamette River, Oregon. River Res Appl
35(8):1279–1290
Harmon ME, Franklin JF, Swanson FJ, Sollins P, Gregory SV, Lattin JD et al (2004) Ecology of
coarse woody debris in temperate ecosystems. Adv Ecol Res 34:59–234
Hilty JA, Merenlender AM (2004) Use of riparian corridors and vineyards by mammalian predators
in northern California. Conserv Biol 18:126–135
Inoue M, Nakano S, Nakamura F (1997) Juvenile masu salmon (Oncorhynchus masou) abundance
and stream habitat relationships in northern Japan. Can J Fish Aquat Sci 54:1331–1341
Ishiyama N, Akasaka T, Nakamura F (2014) Mobility-dependent response of aquatic animal
species richness to a wetland network in an agricultural landscape. Aquat Sci 76:437–449
Jeppesen E, Kronvang B, Meerhoff M, Søndergaard M, Hansen KM, Andersen HE et al (2009)
Climate change effects on runoff, catchment phosphorus loading and lake ecological state, and
potential adaptations. J Environ Qual 38(5):1930–1941
Johnson WC (1994) Woodland expansion in the Platte River, Nebraska: patterns and causes. Ecol
Monogr 64:45–84
Kawaguchi Y, Nakano S (2001) Contribution of terrestrial invertebrates to the annual resource
budget for salmonids in forest and grassland reaches of a headwater stream. Freshw Biol
46:303–316
Kimoto M, Yasutomi N, Yokoyama C, Emori S (2005) Projected changes in precipitation
characteristics around Japan under the global warming. SOLA 1:085–088. https://doi.org/
10.2151/sola.2005-023
6 Riparian Forests and Climate Change 89

Kishi D, Maekawa K (2009) Stream-dwelling Dolly Varden (Salvelinus malma) density and habitat
characteristics in stream sections installed with low-head dams in the Shiretoko Peninsula,
Hokkaido, Japan. Ecol Res 24(4):873–880
Machtans CS, Villard M, Hannon SJ (1996) Use of riparian buffer strips as movement corridors by
forest birds. Conserv Biol 10:1366–1379
Ministry of the Environment, Ministry of Education, Culture, Sports, Science and Technology,
Ministry of Agriculture, Forestry and Fisheries, Ministry of Land, Infrastructure, Transport
and Tourism and Meteorological Agency of Japan (2018) Kikou henndou no kannsoku yosoku
eikyou hyouka nikannsuru tougou report 2018 nihonn no kikouhenndou to sono eikyou [Con-
solidated report on observations, projections and impact assessments of climate change: climate
change and its impacts in Japan]. http://www.env.go.jp/earth/tekiou/report2018_full.pdf
Ministry of the Environment of Japan (2020) Final reports of regional adaptation consortium
project. 1-6
Nadler CT, Schumm SA (1981) Metamorphosis of South Platte and Arkansas Rivers, Eastern
Colorado. Phys Geogr 2:95–115
Nagasaka A, Nakamura F (1999) The influences of land-use changes on hydrology and riparian
environment in a northern Japanese landscape. Landsc Ecol 14:543–556
Nagayama S, Nakamura F (2018) The significance of meandering channel to habitat diversity and
fish assemblage: a case study in the Shibetsu River, northern Japan. Limnology 19:7–20
Nagayama S, Ishiyama N, Seno T, Kawai H, Kawaguchi Y, Nakano D, Nakamura F (2020) Time
series changes in fish assemblages and habitat structures caused by partial check dam removal.
Water 12(12):3357
Naiman RJ, Decamps H (1997) The ecology of interfaces: riparian zones. Annu Rev Ecol Syst
28:621–658
Nakamura F, Dokai T (1989) Estimation of the effect of riparian forest on stream temperature
based on heat budget. J Jpn For Soc 71:387–394. (in Japanese with English summary)
Nakamura F, Swanson FJ (1993) Effects of coarse woody debris on morphology and sediment
storage of a mountain stream system in western Oregon. Earth Surf Process Landf 18:43–61
Nakamura F, Yajima T, Kikuchi S (1997) Structure and composition of riparian forests with special
reference to geomorphic site conditions along the Tokachi River, northern Japan. Plant Ecol
133:209–219
Nakamura F, Swanson FJ, Wondzell SM (2000) Disturbance regimes of stream and riparian
systems - a disturbance-cascade perspective. Hydrol Process 14:2849–2860
Nakamura F, Shin N (2001) The downstream effects of dams on the regeneration of riparian
tree species in northern Japan. Geomorphic processes and riverine habitat (Dorava J M,
Montgomery D R, Palcsak B and Fitzpatrick F A eds.). AGU Water Sci Appl 4:173–181. ISBN:
0-87590-353-3
Nakamura F, Swanson FJ (2003) Dynamics of wood in rivers in the context of ecological
disturbance. The ecology and management of wood in world rivers (Gregory S V, Boyer K
L and Gurnell AM eds.). Symp 37:279–297. ISBN: 1-888569-56-5
Nakamura F, Yamada H (2005) The effects of pasture development on the ecological functions of
riparian forests in Hokkaido in northern Japan. Ecol Eng 24:539–550
Nakamura F, Inahara S (2007) Fluvial geomorphic disturbances and life history traits of riparian
tree species. In: Johnson EA, Miyanishi K (eds) Plant disturbance ecology: the process and the
response. Academic Press, pp 283–310. ISBN: 978-0-12-088778-1
Nakamura F, Komiyama E (2010) A challenge to dam improvement for the protection of both
salmon and human livelihood in Shiretoko, Japan’s third natural heritage site. Landsc Ecol Eng
6:143–152
Nakamura F, Fuke N, Kubo M (2012) Contributions of large wood to the initial establishment and
diversity of riparian vegetation in a bar-braided temperate river. Plant Ecol 213:735–747
Nakamura F, Ishiyama N, Sueyoshi M, Negishi J, Akasaka T (2014) The significance of meander
restoration for the hydrogeomorphology and recovery of wetland organisms in the Kushiro
River, a lowland river in Japan. Restor Ecol 22:544–554
90 F. Nakamura

Nakamura F, Seo JIL, Akasaka T, Swanson FJ (2017) Large wood, sediment, and flow regimes:
their interactions and temporal changes caused by human impacts in Japan. Geomorphology
279:176–187
Nakamura F, Watanabe Y, Negishi J, Akasaka T, Yabuhara Y, Terui A, Yamanaka S, Konno M
(2020) Restoration of the shifting mosaic of floodplain forests under a flow regime altered by a
dam. Ecol Eng 157:105974
Nakano D, Nagayama S, Kawaguchi Y, Nakamura F (2018) Significance of the stable foundations
provided and created by large wood for benthic fauna in the Shibetsu River, Japan. Ecol Eng
120:249–259
Niiyama K (1987) Distribution of Salicaceae species and soil texture of habitats along the Ishikari
River. Jpn J Ecol 37:163–174. (in Japanese with English summary)
Ormerod SJ (2009) Climate change, river conservation and the adaptation challenge. Aquat
Conserv 19:609–613
Palmer MA, Lettenmaier DP, Poff NL, Postel SL, Richter B, Warner R (2009) Climate change and
river ecosystems: protection and adaptation options. Environ Manag 44(6):1053–1068
Poff LR, Allan JD, Bain MB, Karr JR, Prestegaard KL, Richter BD, Sparks RE, Stromberg JC
(1997) The natural flow regime. Bioscience 47:769–784
Sakio H (1997) Effects of natural disturbance on the regeneration of riparian forests in a Chichibu
Mountains, Central Japan. Plant Ecol 132:181–195
Sato H, Nagata M, Takami T, Yanai S (2001) Shade effect of riparian forest in controlling
summer stream temperature: impact on growth of masu salmon juveniles (Oncorhynchus
masou Brevoort). J Jpn For Soc 83:22–29. (in Japanese with English Summary)
Schomaker C, Wolter C (2011) The contribution of long-term isolated water bodies to floodplain
fish diversity. Freshw Biol 56:1469–1480
Seavy NE, Gardali T, Golet GH, Griggs FT, Howell CA, Kelsey R et al (2009) Why climate
change makes riparian restoration more important than ever: recommendations for practice
and research. Ecol Restor 27(3):330–338
Sedell JR, Froggatt JL (1984) Importance of streamside forests to large rivers: the isolation of
the Willamette River, Oregon, U. S. A., from its floodplain by snagging and streamside forest
removal. Verh Int Ver Limnol 22:1828–1834
Shin N, Nakamura F (2005) Effects of fluvial geomorphology on riparian tree species in Rekifune
River, northern Japan. Plant Ecol 178:15–28
Sugimoto S, Nakamura F, Ito A (1997) Heat budget and statistical analysis of the relationship
between stream temperature and riparian forest in the Toikanbetsu River basin, northern Japan.
J For Res 2:103–107
Suzuki K, Ishiyama N, Koizumi I, Nakamura F (2021) Combined effects of summer water
temperature and current velocity on the distribution of a cold-water-adapted sculpin (Cottus
nozawae). Water 13:975
Takahashi M, Nakamura F (2011) Impacts of dam-regulated flows on channel morphology and
riparian vegetation: a longitudinal analysis of Satsunai River. Jpn Landsc Ecol Eng 7:65–77
Thomas SM, Griffiths SW, Ormerod SJ (2016) Beyond cool: adapting upland streams for climate
change using riparian woodlands. Glob Chang Biol 22(1):310–324
Tiedemann RB, Rood SB (2015) Flood flow attenuation diminishes cotton-wood colonization
sites: an experimental test along the Boise River, USA Ecohydrology. https://doi.org/10.1002/
eco.1619
Wilby RL, Orr H, Watts G, Battarbee RW, Berry PM, Chadd R et al (2010) Evidence needed
to manage freshwater ecosystems in a changing climate: turning adaptation principles into
practice. Sci Total Environ 408(19):4150–4164
WWF (2014) Living planet report 2014. https://www.wwf.or.jp/activities/data/
WWF_LPR_2014.pdf
Yabuhara Y, Yamaura Y, Akasaka T, Nakamura F (2015) Predicting long-term changes in riparian
bird communities in floodplain landscapes. River Res Appl 31:109–119
6 Riparian Forests and Climate Change 91

Yabuhara Y, Yamaura Y, Akasaka T, Yamanaka S, Nakamura F (2019) Seasonal variation in patch

and landscape effects on forest bird communities in a lowland fragmented landscape. For Ecol
Manag 454:117–140

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 7
Improvement of the Flood-Reduction
Function of Forests Based on Their
Interception Evaporation and Surface
Storage Capacities

Takao Tamura

Abstract Forests have a flood-reduction function that reduces flood peak flow and
delays the flood peak time. In the mountains of Japan, artificial forests planted
between the 1950s and 1970s are widespread; however, many of these forests are
not well managed. The effective use of the flood-reduction function of forests as a
remarkable approach for river basin management has been discussed for several
years. In this study, two aspects of the water cycle in forests were explored:
the interception evaporation process in the forest canopy and the groundwater
storage process on the forest slope. A runoff model was applied to the hydrological
data obtained in several forest basins with different characteristics to evaluate the
effects of the processes. In the case of the Japanese cedar plantations studied,
it was suggested that the improvement of interception evaporation capacity and
surface storage capacity by conversion to mixed forests and selective logging would
significantly reduce the peak flood discharge on a timescale of approximately 20–
30 years.

Keywords Forest · Flood-reduction function · Interception evaporation ·

Surface runoff · Selective cutting

7.1 Introduction

The influence of climate change is conspicuous in Japan. Severe flood disasters

occur almost annually in various regions of Japan. Forests in Japan account for
approximately 70% of the land, while 40% of the forests are artificial forests
comprising coniferous trees. These forests are distributed at mountainous areas,

T. Tamura ()
Department of Civil and Environmental Engineering, Graduate School of Technology, Industrial
and Social Science, Tokushima University, Tokushima, Japan
e-mail: [email protected]

© The Author(s) 2022 93

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_7
94 T. Tamura

which are the source areas of streams. Therefore, heavy rain in mountainous forests
has become a major cause of flood disasters. Forests have a flood control function
for storing a part of the rainfall and delaying runoff discharge from their basins.
This contributes toward decreasing the quantity of flood flow. As forests are part of
the green infrastructure, an effective use of the flood-reduction function for disaster
prevention has been previously discussed.
Generally, it is recognized that forest soil plays a principal role in the flood-
reduction function. Since forest soil has a large void structure, forests can store a
significant amount of rainwater as compared with other land cover types. Therefore,
methods used to strengthen the flood-reduction function, such as thinning of forests
and conversion of vegetation structure, have been discussed. However, it has been
pointed out that the flood-reduction function of forests is not sufficiently effective
for large floods (Laurance 2007). It is also extremely difficult to increase the
thickness of the soil layer, which determines the water storage capacity. The soil
formation speed has been reported in the range of 0.05–0.2 mm/year (Amundson et
al. 2015); it is estimated that a significant amount of soil is washed away by heavy
rainfall in the steep Japanese mountains. Therefore, the increase in the thickness of
the surface soil layer will not be able to adapt to the large-scale heavy rainfall that
is predicted to accompany the rapid progression of climate change. In addition, it is
difficult to incorporate this method in flood control plans, since the plans in Japan are
developed for the subsequent 20–30 years. Two methods that can improve the flood-
reduction function of forests in a relatively short period of time include enhancement
of interception evaporation capacity and enhancement of ground surface storage
capacity. These two processes are examined and their effects are discussed using a
rainfall-runoff model.

7.2 Rainwater Runoff Mechanisms in Mountainous Forests

and Measures to Improve the Flood-Reduction Functions
of Forests

Figure 7.1 shows the rainwater runoff processes in a mountainous forest. First,
the interception evaporation on the forest crown or canopy affects the rainfall.
Some raindrops are intercepted by the branches and leaves, wherein they become
microscopic particles and drift through the air. The interception evaporation process
was strongly affected by the multilayered structure around the crown. The rainwater
lost through interception evaporation reduces the flood amount and its peak flow.
This forest property can be effectively improved by forest management through
planting, thinning, and felling of trees.
Thereafter, the rainwater that reaches the ground surface will then infiltrate
the soil layer if its intensity is below the infiltration and storage capacity of the
soil. Rainwater that infiltrates the soil becomes groundwater and contributes to
subsurface runoff components, flowing out relatively slowly as compared to the
7 Improvement of the Flood-Reduction Function of Forests Based on Their. . . 95

Evaporation

Interception

Throughfall
Rainfall

Stored Not stored

Interception
Throughfall evaporation

Surface
runoff

Surface storage
Infiltration
Storage

Subsurface
runoff
River
Groundwater
runoff

Fig. 7.1 Rainwater runoff processes in mountainous forest

surface runoff components. This process is generally recognized as the key role of
the flood-reduction function of forests. However, rainwater becomes a surface runoff
component when the rainfall intensity increases beyond the infiltration and storage
capacity of the soil layer. The surface runoff component immediately flows down a
slope, resulting in a flood. The flood-reduction function of the forest then reaches
its limit. The infiltration and storage capacity are determined by the thickness of the
forest soil layer. Therefore, it would be extremely difficult to increase the storage
capacity of the soil because a period greater than 100 years is necessary for recovery
of soil layer thickness (Ogawa et al. 2011).
Surface flows account for most of the peak flood discharge during heavy rainfall
events that can cause flood disasters. It flows down forest slopes under the influence
of the amount of understory vegetation and the roughness of the ground surface.
Therefore, a forest with significant ground surface roughness can decrease the speed
of surface runoff flow. Consequently, the surface runoff component is stored on
the ground surface for a short time. The amount of water storage on the ground
contributes to the reduction and delay of the flood peak flow. For instance, the
management of understory vegetation is important for increasing the ground surface
roughness. This can be actualized by daily forest management, and the effects are
expected to manifest rapidly.
96 T. Tamura

In this study, two aspects were explored: the interception evaporation process in
the forest canopy and the water storage process on the ground surface. Based on
hydrological observations and analysis using a rainfall-runoff model for two forest
areas with different forest characteristics, the possibility of improving the flood-
reduction function of forests by enhancing the operation of these two hydrological
processes is discussed.

7.3 Runoff Model for Evaluating Flood-Reduction Function

The tank model is used as the runoff model in this study, as shown in Fig. 7.2. The
author refers to it as the two-stage tank model with separated surface flow (Tamura
et al. 2006). The model is broadly divided into a surface tank (soil layer) and a
groundwater tank (bedrock layer) to represent the infiltration, storage, and runoff
processes of rainwater in forest soils, as well as the various runoff components:
surface runoff, subsurface runoff, and groundwater runoff. Rainfall input to the
model is either the amount of throughfall that reaches the ground surface or the
amount of rainfall multiplied by the percentage of throughfall (the reciprocal of
the interception evaporation rate). A characteristic feature of this model is that
the surface runoff coefficient is calculated from the average slope, average slope

Fig. 7.2 Two-stage tank model with separated surface flow

7 Improvement of the Flood-Reduction Function of Forests Based on Their. . . 97

N
Tomisato (Besshi-Dozan)

Tokushima
Japan Dozan river
Naka
Tokyo river
Sameura
Tokushima
Shirakawatani
Hashimoto

400 km 50 km
Shikoku island

Fig. 7.3 Location of the study sites

length, and roughness of the forest slope (Eq. 7.1). The coefficients of the tank
model, including the surface roughness, are identified as a set of parameters that can
uniformly reproduce multiple flood hydrographs obtained from field observations;
these are used in flood simulations to compare the flood-reduction functions of forest
basins.
 
λo = 2.52 × 10−3 · Is / rmax 0.8 (N · Ls )1.8 (7.1)

where λo is the surface runoff coefficient (/h), Ls is the slope length, Is is the average
slope (m), N is the surface roughness (m-1/3 s), and rmax is the observed maximum
rainfall intensity considered as the maximum surface runoff intensity (mm/h).

7.4 Enhancing Interception Evaporation Capacity Through

Afforestation

In the upper basin of the Dozan River, which is one of the prevailing tributaries
of the Yoshino River on Shikoku Island (Fig. 7.3), a large copper mine (Besshi-
dozan) had been operated from 1691 to 1973. A significant number of trees had
been felled for copper refinement business, and a substantial quantity of soil had
been washed away from the mountains. A severe flood was caused by a heavy rain-
related typhoon in August 1899, and several people lost their lives. Thereafter, a
large-scale tree plantation was created, and the vegetation was restored (Sumitomo
group Public Affairs Committee 2021).
The flood-reduction function of the Tomisato dam basin, including the Besshi
Copper Mine area, was evaluated at the beginning of the twenty-first century when
the vegetation had recovered for approximately 100 years after the beginning of
98 T. Tamura

large-scale planting. A runoff model was used in the evaluation to describe the
water cycle processes in a forested basin (Tamura et al. 2008). In this study, the
water storage capacity of the soil layers and the interception evaporation effect were
estimated; Figs. 7.4 and 7.5 show the calculation results. The characteristics of the
flood-reduction function of the basin were examined by comparing its runoff results
with those of the Sameura dam basin, because it is located next to the Tomisato dam
basin and has similar geology and vegetation. The role of interception evaporation
is very important during the forest recovery stage (Fig. 7.4). The simulation results
demonstrate that the flood peak flow would increase by 50%, compared to the
current condition, if clear-cutting was performed in the Tomisato dam basin. The
rate of increase of the flood peak flow is higher than that of the Sameura dam.
This is because the runoff model estimated that the interception evaporation rate
of the current Tomisato dam basin is larger than that of the Sameura dam basin,
and the thickness of the surface soil layer of the current Tomisato dam basin is
much smaller than that of the Sameura dam. In other words, if clear-cutting is
implemented in the Tomisato dam basin, the large amount of rainfall lost due to
interception evaporation will reach the ground surface, but the thin soil layer will
quickly become saturated, resulting in an increase in peak flood discharge. Figure
7.5 shows the maximum groundwater storage volume for the two basins under the
clear-cut conditions (the situation is shown in Fig. 7.4b). It shows the incremental
storage from the start of rainfall to the time of peak flood flow in the surface and
groundwater tanks shown in Fig. 7.2. The possibility that the thickness of the soil
layer at the Tomisato dam basin has not recovered sufficiently since the start of the
planting program is shown in Fig. 7.5. Nevertheless, the fact that the peak flood
discharge in the Tomisato dam basin is lower than that in the Sameura dam basin
(Fig. 7.4a) suggests that the interception evaporation in the flood-reduction function
of the current Tomisato dam basin is quite effective. This can be considered as the
greatest effect of the large-scale afforestation project in terms of flood reduction.
As a result, it can be concluded that the enhancement of interception evaporation
capacity is more effective than the enhancement of soil layer thickness as a means
of improving the flood-reduction function of forests in a short period of time.
In addition, it is important to avoid excessive thinning and clear-cutting in forest
management when applying this method to basins with poor forest soil layer
thickness, such as the Tomisato dam basin.

7.5 Enhancing the Interception Evaporation Capacity

and Surface Storage Capacity by Vegetation Conversion

The upper basin of the Naka River is notable for the production of Kito-sugi
Japanese cedar. This basin is known to be a high precipitation area, with an annual
precipitation of approximately 3000 mm. The Hashimoto forest is located in the
upper district of the Naka River (Fig. 7.3), wherein selection cutting forestry was
7 Improvement of the Flood-Reduction Function of Forests Based on Their. . . 99

0 0
Rainfall
(mm/h)

Total 486mm Total 486mm

100 100
70 70

Tomisato
60 .9 (Besshi)
60 60

50 .7
50
Sameura 50
Sameura
Runoff discharge (mm/h)

46 .9

Tomisato
40 .4
40 (Besshi) 40

30 30

20 20

10 10

0 0
0 6 12 18 24 0 6 12 18 24

(a) When interception (b) When interception

evaporation is effective evaporation is ineffective
(current condition) (clear-cutting simulation)

Fig. 7.4 Comparison of flood hydrographs under current and clear-cut conditions to evaluate the
impact of interception evaporation considering Typhoon No. 23 in 2004 (summarized from Tamura
et al. 2008)

performed from the 1980s and broadleaf trees were preserved. Its forest type is
a mixed forest consisting of conifers and broadleaf trees, with a multilayered
crown structure. The standard thinning rate is approximately 30% in common cedar
artificial forests in Japan, but the rate of the Hashimoto forest ranges from 15% to
20%. However, due to the mixture of various aged trees and species, the canopy is
not dense and the understory vegetation is rich (Fig. 7.6a).
The flood-reduction function of the Hashimoto forest was evaluated using a
runoff model (Tamura et al. 2020). The effects of the interception evaporation
capacity and the ground surface storage capacity in the Hashimoto forest, a mixed
conifer and broadleaf forest, were discussed and compared with those in the
100 T. Tamura

250
Total rainfall 486mm

Maximum water storage(mm) 200

150
126 Soil Layer

100

60 114
50
42 Bedrock
Layer
18 12
0
Tomisato Sameura
(Besshi)
Fig. 7.5 Estimation of maximum groundwater storage in forested basins using the two-stage tank
model with separated surface flow considering Typhoon No. 23 in 2004 (summarized from Tamura
et al. 2008)

Fig. 7.6 Trees and understory vegetation at the observation sites

Shirakawatani experimental forested basin. Shirakawatani is a Japanese cedar forest

where general Japanese forestry (large-scale planting and felling) is performed
(Fig. 7.6b), and its geological features and topography are similar to those of
the Hashimoto forest. The model parameters of the interception evaporation rate
and surface roughness obtained from the runoff analysis of the Hashimoto forest
7 Improvement of the Flood-Reduction Function of Forests Based on Their. . . 101

Table 7.1 Main parameters of the in-line two-stage tank model for simulating the conversion of
Japanese cedar artificial forest to mixed needle and broadleaf forests
Parameter Hashimoto Shirakawatani
Mixed needle
Mixed needle Artificial cedar and broadleaf
and broadleaf forest (current forests
forests condition) (simulation)
Interception evaporation rate 25 14 25
I (%)
Ground surface roughness 1.8 1.2 1.8
N (m−1/3 s)
Apparent surface thickness 137 100 100
γ D (mm)
Groundwater supply coefficient
λg (/h) 0.05 0.01 0.01
Final infiltration rate 6.45 1.45 1.45
f* (/h)
Soil saturation capacity 168 40 40
h1 (mm)

were applied to the Shirakawatani experimental forested basin model to study the
improvement of flood-reduction function of a general cedar artificial forest by
converting it into a mixed needle-hardwood with only selective logging. The model
parameters for the two basins, as well as the parameters when the Shirakawa valley
is changed to a mixed needle-hardwood forest, are listed in Table 7.1.
The flood-reduction function of the Hashimoto forest is greater than that of the
Shirakawatani. Runoff simulation using the same rainfall pattern as shown in Fig.
7.7 for both runoff models showed that the peak flow rate in Hashimoto forest
area was 22% lower than that in Shirakawatani (Fig. 7.8). When the interception
evaporation capacity rate of the Hashimoto forest model was applied to the
Shirakawatani model, the peak flood discharge in the Shirakawatani decreased from
66.8 mm/h to 56.3 mm/h. In addition, when the surface roughness parameter of the
Hashimoto forest was applied to the Shirakawatani model, it was further reduced
to 50.9 mm/h. As a result, the peak flood discharge was estimated to be 24% lower
than that for the current Shirakawatani model.
The flood-reduction function of the Hashimoto forest is higher than that of the
Shirakawatani owing to several factors. In the Hashimoto forest, cedar trees of
different ages and broad-leaved trees of various species are mixed; the canopy is
multilayered. The density of the canopy of the multistoried forest seems to be
sparse in the vertical space. The thinning rate in the Hashimoto forest is low, but
sunlight reaches the forest floor easily. This is a favorable condition for understory
vegetation. Therefore, the interception evaporation rate and the surface roughness
are high, and it is assumed that a high flood-reduction function is demonstrated in
the Hashimoto forest.
102 T. Tamura

100
870mm / 24hr
90
80
70
Rainfall (mm/h)

60
50
40
30
20
10
0
1 3 5 7 9 11 13 15 17 19 21 23
Time (h)
Fig. 7.7 Rainfall patterns used to evaluate the Hashimoto forest and Shirakawatani experimental
forested basin (Tamura et al. 2020)

70

60 Shirakawatani 66.8mm/hr
Discharge(mm/h)

50
Hashinoto 51.9mm/hr
40

30

20

10

0
0 24 48
Time(h)

Fig. 7.8 Comparison of flood hydrographs between Hashimoto forest and Shirakawatani experi-
mental forested basin under current conditions (summarized from Tamura et al. 2020)
7 Improvement of the Flood-Reduction Function of Forests Based on Their. . . 103

70
Shirakawatani (Current condition)
Interception evaporation rate 0.14
60 Ground surface roughness 1.2
66.8mm/h
Discharge(mm/h)

50
Shirakawatani
Interception evaporation rate 0.25
Ground surface roughness 1.2
40 56.3mm/h

30 Shirakawatani
Interception evaporation rate 0.25
Ground surface roughness 1.8
20 50.9mm/h

10

0
0 24 48

Time(h)

Fig. 7.9 Flood hydrographs of the Shirakawatani experimental forested basin under the condition
that the interception evaporation rate and surface roughness are changed by the conversion to mixed
needle and broadleaf forest (summarized from Tamura et al. 2020)

The results suggest that the flood-reduction function of cedar forests with poor
understory vegetation, such as those in the Shirakawatani experimental forested
basin, can be improved on a timescale of approximately 20–30 years by conversion
to mixed forests with only selective logging (Fig. 7.9).

7.6 Methods and Limitations of Early Enhancement of Flood

Mitigation Functions of Forests

In this study, the reinforcement measure for the flood-reduction function of forests
and the expected effects were discussed in terms of interception evaporation in the
forest canopy and water storage on the forest ground surface. These methods have
two advantages compared to the restoration of soil layer thickness. First, they are
achieved easily via daily forest operations. Second, the effects can be expected to
be realized in 20–30 years. However, the expected effect is still restrictive (10–20%
reduction of the flood peak runoff) for heavy rainfall events that are targeted by the
flood control plan. In addition, because the interception evaporation rate observed
in forests has changed owing to a rainfall event (Tanaka et al. 2005), it is believed
that the flood-reduction function of forests fluctuates due to forest site and structure.
For utilizing the flood-reduction function of a forest is in river basin management, a
thorough understanding the characteristics of this function is required.
104 T. Tamura

References

Amundson R, Berhe AA, Hopmans JW, Olson C, Sztein AE, Sparks DL (2015) Soil science.
Soil and human security in the 21st century. Science 348:1261071. https://doi.org/10.1126/
science.1261071
Laurance WF (2007) Environmental science: forests and floods. Nature 449:409–410
Ogawa K, Tamura T, Takigawa N, Kuwahara M, Takanishi S (2011) Flood runoff characteristic
change over 50 years by tree species conversion in upper Dozangawa River Basin. Ann J
Hydraul Eng JSCE 55:571–576
Sumitomo group Public Affairs Committee (2021). https://www.sumitomo.gr.jp/history/
besshidouzan/index04.html
Tamura T, Hashino M, Tachibana D (2006) Parameter identification of runoff model using rainfall
and water-level data. Ann J Hydraul Eng JSCE 50:355–360
Tamura T, Hashino M, Anamizu H, Araki T (2008) Evaluation of flood reduction function of the
forested basins located upstream in the Yoshino river. Ann J Hydraul Eng JSCE 52:379–384
Tamura T, Ueda S, Muto Y, Kamada M (2020) Improvement of flood reduction function of forested
basin by reinforcement of evaporation rate and ground roughness. J Jpn Soc Civ Eng Ser
B1(Hydraul Eng) 76(2):127–132
Tanaka N, Kuraji K, Shiraki K, Masanori S, Masakazu S, Ohta T, Makoto S (2005) Throughfall,
stemflow and rainfall interception at mature Cryptomeria japonica and Chamaecyparis obtusa
stands in Fukuroyamasawa watershed. Bull Tokyo Univ For 113:197–240

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 8
Forest Green Infrastructure to Protect
Water Quality: A Step-by-Step Guide for
Payment Schemes

Thomas R. Nisbet, Maria-Beatrice Andreucci, Rik De Vreese, Lars Högbom,

Sonja Kay, Mary Kelly-Quinn, Alessandro Leonardi, Mariyana I. Lyubenova,
Paola Ovando Pol, Paula Quinteiro, Ignacio Pérez Silos, and Gregory Valatin

T. R. Nisbet () · G. Valatin

Forest Research, Alice Holt Lodge, Farnham, UK
e-mail: [email protected]
M.-B. Andreucci
Department of Planning, Design, Technology of Architecture, Faculty of Architecture, Sapienza
University of Rome, Rome, Italy
R. De Vreese
Independent Researcher and Consultant in Urban Forestry and Ecosystem Services, Ghent,
Belgium
L. Högbom
Forestry Research Institute of Sweden (Skogforsk), Uppsala, Sweden
S. Kay
Agroscope, Research Division Agroecology and Environment, Research Group Agricultural
Landscapes and Biodiversity, Zurich, Switzerland
M. Kelly-Quinn
School of Biology and Environmental Science, University College Dublin, Dublin 4, Ireland
A. Leonardi
ETIFOR | Valuing Nature, Padova University Spin-off, Padova, Italy
M. I. Lyubenova
Department of Ecology and Environmental Protection, University of Sofia “St. Kliment
Ohridski”, Sofia, Bulgaria
P. O. Pol
The James Hutton Institute, Aberdeen, UK
P. Quinteiro
Centre for Environmental and Marine Studies (CESAM), Department of Environment and
Planning, University of Aveiro, Aveiro, Portugal
I. P. Silos
Environmental Hydraulics Institute, Universidad de Cantabria, Santander, Spain

© Crown 2022 105

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_8
106 T. R. Nisbet et al.

Abstract This chapter describes how to design appropriate and cost-effective forest
green infrastructure for water payment schemes to protect and improve water
quality. It is structured by the main steps involved in establishing a payment
scheme, starting with identifying the water issues and how tree planting and
forest management can help, managing potential disbenefits and exploring multiple
benefits, followed by scheme design, monitoring and communication. The approach
is relevant to all actors involved in sustainable water management, farming and
forestry, from policy makers, catchment planners and land managers to private
investors, practitioners and local communities. We provide a common language
and framework to help ensure schemes are successful in delivering water and other
ecosystem services while minimising possible trade-offs (such as the potential for
tree planting to reduce water resources).

Keywords Forest green infrastructure · Forests for water · Diffuse pollution ·

Payments for ecosystem services · Sustainable water management

8.1 Introduction

The main aim of European Union (EU) water policy is to ensure that a sufficient
quantity of good-quality water is available for both people’s needs and the environ-
ment. Despite ongoing efforts by Member States to improve water status, only 40%
of surface waters (rivers and lakes) are in good ecological status or potential (EEA
2018). Diffuse pollution is a major pressure, dominated by agricultural sources in
the form of excessive emissions of nutrients (nitrates and phosphates), pesticides,
sediment and faecal indicator organisms (FIO) (Fig. 8.1). Agriculture is estimated
to contribute to 25% of surface water bodies failing good ecological status, and is
the main cause of groundwater bodies failing to achieve good chemical status (EEA
2018).
Member States continue to develop and invest in best practice farming measures
to reduce diffuse pollution from agriculture, such as farm-level nutrient planning,
reduced tillage and the use of catch crops. These have improved water quality but
in most cases by an insufficient margin to meet environmental quality standards.
Notably, there has been limited improvement in the proportion of water bodies
achieving good ecological status since the first River Basin Management Plans were
published in 2009 (EEA 2018). While there is scope for further improvement in the
effectiveness and uptake of farm measures, there is a growing recognition that wider
delivery of good ecological status will only be achieved by a significant degree of
land-use change (Stutter et al. 2012). This chapter focuses on the water benefits of
forest green infrastructure (FGI), although it is recognised that the use of tree-based
measures such as tree planting and management should be part of a wider framework
of integrated catchment management (e.g. including opportunities for peatland and
wetland restoration) and associated strategies and plans.
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 107

Fig. 8.1 Catchment sources and pathways of agricultural diffuse pollution impacting on the water
environment and water users

The multiple benefits provided by FGI in the form of trees, woodlands and forests
(the terms woodlands and forests are used throughout the document and describe
land predominantly covered by trees; woodland means a relatively small area of
trees, while forest refers to a large tract of trees) are increasingly recognised and
valued for society. Benefits for the water environment include the ability to protect
aquatic habitats and species from disturbance, preserve the quality of drinking water,
alleviate flooding and guard against erosion, landslides and the loss of soil (Nisbet
et al. 2011). Tree planting provides a very effective and relatively secure measure
for tackling agricultural diffuse pollution, in addition to helping with carbon storage
and providing other environmental benefits. Small-scale, targeted use of FGI such as
planting of woodlands on or around pollutant sources, or along pollutant pathways
in the form of ‘woodland buffers’, offers a smart way of attenuating or eliminating
pollutant delivery to surface waters and groundwaters while minimising land take
and impacts on food security.
108 T. R. Nisbet et al.

Although the benefits of FGI for water are well known (Creed and Noordvijk
2018), progress is highly constrained by the significant cost to landowners and
managers in terms of reduction in land value and agricultural income resulting from
land-use change. This is especially the case for tree planting on the more productive
and intensively managed agricultural land that represents the greatest source of
diffuse pollutants. Achieving enough tree planting to make a difference for water
body status will require better incentives in the form of payments for the water and
other ecosystem services provided. Maintaining and protecting the water benefits
provided by existing FGI may also require funding support, especially if changes
to forest design and management are needed to address threats posed by climate
change.
There are many different types of payment for ecosystem services (PES)
schemes, and the approach and definitions continue to evolve (Forest Europe 2019).
The main purpose of PES is to protect and enhance the provision of ecosystem
services for environmental gain and better management of natural resources by
incentives (Gatto et al. 2009). Ideally, five conditions should be met in a PES
scheme, which are: (1) the identification of a well-defined ecosystem service to
be exchanged, in this case principally targeted tree planting and the appropriate
management of new or existing woodlands and forests to improve water quality;
(2) the presence of at least one service buyer and (3) at least one seller; (4)
the voluntary nature of the marketing of the ecosystem service; and finally, (5)
the conditionality of the payment, requiring the seller/provider to ensure that the
expected benefit is delivered and sustained over time. Often these conditions are
not met, especially condition 4, such as where schemes are implemented within a
compulsory regulatory framework. Such cases are often referred to as ‘PES-like’
schemes.
We adopt a broad definition of a FGI for water payment scheme based upon
three criteria: (1) a transfer of resources between at least two stakeholders; (2) a
transaction explicitly targeted at obtaining water-related services; and (3) a payment
for actions related to trees, either primarily for water services or for bundled
(including water) ecosystem services.

8.2 Identifying the Problem

As a consequence of the introduction in 2000 of the EU Water Framework Directive

(WFD), there is now far greater knowledge about the water environment across
Member States. Regular monitoring and assessment of surface- and groundwaters
at more than 130,000 sites has generated a detailed understanding of the condition
of Europe’s water bodies, as well as of the pressures that are preventing the majority
achieving the targets of good ecological status or potential (Fig. 8.2).
National water regulators compile and regularly update datasets and maps
showing which water bodies are at less than good status, the causal activities and
progress made with introducing programmes of measures to achieve target status.
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 109

Fig. 8.2 Proportion of surface water bodies by Member State at less than good ecological status.
Reproduced from EEA (2016) © European Environment Agency, 2016

This includes knowledge of which water bodies are failing good status due to diffuse
pollutants such as nutrients, sediment and pesticides derived from agriculture, as
well as impacted by other anthropic pressures on water (Fig. 8.3). Water regulators
also have information on the location and condition of particularly sensitive waters,
such as drinking water protected areas and high-status waters supporting priority
habitats and species. These data are submitted to the European Environment Agency
(EEA) at regular intervals to allow an assessment of the status and pressures acting
on European waters, with the last assessment published in 2018 (EEA 2018).
Addressing water pressures and achieving water targets requires co-ordinated
and long-term actions at the level of the catchment or sub-catchment of the water
body. This is particularly the case with managing diffuse pollution, which often
has a variety of sources spread across the landscape and land ownerships. In some
regions and countries, pollutant models have already identified pollutant sources
and pathways to aid targeting of measures (Collins et al. 2018; Mockler and
Bruen 2018). Catchment partnerships have been formed in many Member States
to adopt an integrated, catchment-based approach to tackling polluting activities
and delivering improvements. Partnerships are often led by trusted intermediaries
who are better able to achieve change on the ground supported by co-ordinated
funding bids (ribblelifetogether.org/improve/woodlands, wrt.org.uk/project/3rivers-
project, www.woodlandsofireland.com, www.etifor.com/en/studies-and-research).
110 T. R. Nisbet et al.

Fig. 8.3 Distribution of groundwater bodies across Member States failing for nitrate

8.3 Role of Opportunity Mapping

Opportunity mapping can help identify and prioritise water bodies and component
areas of land for targeting FGI to reduce water pressures (Broadmeadow and Nisbet
2012). Evidence-based planning supports integrated catchment management, and
the maps guide and underpin the development of FGI for water PES (PESFOR-W)
schemes. This approach is based on using geographic information systems (GIS)
and integrates a wide range of spatial datasets to determine the most effective
locations for changing land use and management, to meet WFD targets and generate
multiple benefits for society (Box 8.1).
The key steps involved in evidence-based planning are summarised below:
1. Use WFD datasets to identify boundaries of surface and groundwater bodies
failing good ecological or chemical status due to diffuse pollution from agri-
culture; determine which and how many diffuse pollutants are causing failure,
either through WFD measurements in relation to water chemical or biological
standards or an assessment of risk.
2. Draw on any available statistics (e.g. from agricultural inventories, or surveys of
fertiliser, or pesticide use), site surveys and measured or modelled pollutant data
to identify and rank the spatial sources and pathways of each diffuse pollutant
draining to the contributing catchments.
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 111

Box 8.1 Opportunity Mapping to Reduce Diffuse Pollution and Flood

Risk in England and Wales
Opportunity mapping was applied to England and Wales in 2014 to better
target grant aid and private investment for FGI to help deliver positive
outcomes for water quality and flood risk management. The mapping used
national datasets of modelled pollutant loads and pressures at a 1 km2 scale
for each of phosphate, sediment, nitrate, total pesticides and faecal indicator
organisms. These were overlaid with datasets on the risk of flooding from
rivers, including on the propensity of soils to generate rapid runoff. Target
areas for woodland creation were identified based on the scope to reduce
one or more diffuse pollutants and contribute to flood risk management (the
colours on the map and the associated values in the key refer to the number
of diffuse pollutants that tree planting could benefit in a given location).The
maps were subsequently used to score water benefits to inform planting
applications and Rural Development Programme grant support.
For further information, see www.forestresearch.gov.uk/research/forest-
hydrology/opportunity-mapping.

1 2 3 4 5
112 T. R. Nisbet et al.

3. Map any spatial constraints (e.g. designated open habitats or archaeological

features) and sensitivities (e.g. landscape views) to FGI within the catchments;
overlay spatial datasets to identify pollutant hotspots free of constraints to FGI
where there are opportunities for woodland creation to reduce one or more diffuse
pollutants in failing water bodies.
4. Consider and map any other water issues that could benefit from FGI (e.g. local
downstream communities or assets at flood risk) and overlay these to determine
scope to deliver multiple benefits through tree planting where there is greatest
need.
5. Map any potential water trade-offs associated with FGI (e.g. water bodies failing
due to poor quantitative status or inadequate flows) and use mapped sensitivities
to guide tree planting design and management to minimise disbenefits (e.g. by
changing tree type or species to reduce tree water use).
6. For existing FGI, map data on tree type, species and age to determine oppor-
tunities for forest redesign and management, to reduce future risks to forest
water protection functions (e.g. from climate change and related effects on the
incidence of fires, storms and drought, as well as pest and disease outbreaks).
7. Use findings to amend and integrate FGI, water, flood and related strategies and
plans to deliver a more effective, catchment-based approach to tackling diffuse
pollution and achieving WFD objectives.

8.4 How Can FGI Help?

Forests are widely recognised as the preferred land cover for protecting water
supplies. This reflects a range of attributes, including the ability of forest canopies to
moderate rainfall inputs due to wet canopy evaporation; the well-structured nature
of forest soils resulting from sustained organic matter inputs, tree rooting and lack
of soil disturbance, reducing erodibility and promoting slope stability; active uptake
and tight canopy recycling of nutrients; and the generally very low level of chemical
inputs to forests such as fertilisers or pesticides (Nisbet et al. 2011; Creed and
Noordvijk 2018). Consequently, waters draining forests are typically of high quality
and good ecological condition, requiring little or no treatment for public water
supply.
Historic clearance of forests for agriculture has resulted in the widespread loss
of these water benefits and a shift to a more intensive land use often associated with
frequent soil disturbance, soil damage, increased erosion and high inputs of nutrients
and chemicals. Despite recent improvements to farming practice, many agricultural
activities typically generate significant losses of sediment, nitrate, phosphate,
pesticides and/or FIO to the water environment. These result in diffuse pollution
and cause a large number of water bodies to fail to achieve good ecological status.
Food scarcity may prevent large-scale forest replanting to tackle the issue, but there
is significant scope for FGI in the form of targeted, small-scale, woodland planting
on agricultural land to make a difference. This includes the use of agroforestry,
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 113

shelterbelts and tree-lined hedges to help capture and remove diffuse pollutants from
adjoining arable crops or livestock pasture.
Targeted planting works because the sources of pollutants, the pathways by
which they move to watercourses and the vulnerability of downstream water users
are spatially variable (Fig. 8.4). For example, soils vary in their vulnerability to
damage, ability to retain nutrients and chemicals, propensity to generate rapid
surface runoff and degree of connectivity to watercourses. Once pollutants are
mobilised in water or the air, they tend to move along preferred pathways such as
surface channels, drains/ditches and the prevailing wind direction. Water receptors
such as groundwater boreholes draw water from distinct areas and depths of ground.
Tree planting on, around, across or along these key pollutant sources, pathways

Fig. 8.4 Preferred locations in a farmed landscape for tree planting to reduce diffuse pollution
114 T. R. Nisbet et al.

and receptors can potentially be very effective at reducing pollutant delivery to

watercourses and water supplies, thereby markedly improving water quality for a
limited land take.
Planting across or along pollutant pathways in the form of buffer areas or strips
offers a dual water quality benefit (Fig. 8.4). Firstly, the pollutant input associated
with the previous agricultural activity on this sensitive area of ground will be
removed. Secondly, there is a significant opportunity for the planted trees to act
as a barrier to the movement of pollutants from upslope or upwind (Ucar and Hall
2001). Pollutants can be retained or removed by runoff being encouraged to infiltrate
into the better structured soil of the buffer; by filtration and surface deposition as
surface runoff passes through the leaf litter layer or is held in surface depressions
created by tree roots; by root uptake and incorporation into growing trees; or by
interception and capture as the polluted airflow passes through the tree canopy.
Riparian woodland buffers have the added benefit of removing pollutant inputs and
reducing damage to this very vulnerable and connected area of land, as well as
providing scope for planted trees to remove pollutants carried downstream within
the main watercourse during out-of-bank flows.
WFD monitoring data show a marked difference in water quality between forest
and agricultural land uses, with the magnitude of difference depending on the
intensity and quality of land management. The quantity of pollutant inputs in the
form of fertiliser, organic amendments and chemicals and typical exports in surface
runoff or leaching to groundwater are well known for each land use and can be used
by pollutant models to estimate the impact and effectiveness of a given area of land-
use change (Table 8.1). In contrast, it is more difficult to predict the barrier effect
of buffer areas since this is influenced by many design and management factors, as
well as by the nature and type of pollutant and the scale of intervention. However,
studies have shown that with good design and appropriate management, tree buffer
areas can be highly effective at reducing pollutant delivery in surface runoff from
upslope land, with efficiencies of up to 100% possible for certain pollutants (Perez-
Silos 2017).
A review of 65 studies found buffer width to be a dominant factor, with pollutant
removal generally decreasing with declining buffer width (Perez-Silos 2017). There
are a number of important factors that act to reduce the efficiency by which tree
buffers can remove diffuse pollutants from upslope land. These include increasing
volume of runoff, increasing pollutant load (especially if the quantity of pollutant
draining from upslope land exceeds the capacity of the trees and soil to remove
or process it), the presence of newly established/very young or old trees, poor
tree condition or weak tree growth, wider tree spacing and the presence of any
bypass channels such as drains. Great care is therefore required in the design and
management of buffer areas to cope with local pollutant loads and to efficiently
achieve and maintain high levels of pollutant removal. This may require productive
woodland management and regular harvesting to sustain nutrient uptake (see Sect.
8.5).
Landowner pressure to minimise land take from agriculture acts as a major
constraint on buffer width that can result in suboptimal performance. As a rough
Table 8.1 Nutrient loads and modelled export coefficients to water for different crops vs woodland in Great Britain. Nutrient loads taken from the British
Survey of Fertiliser Practice for 2000–2011 (BSFP 2013) and export coefficients based on the same data modelled for the UK National Ecosystem Assessment
Follow-on Report (Bateman et al. 2014)
Permanent grassland Rough pasture Wheat Barley Maize Oil seed rape Woodland
Nitrogen input (kg/ha/year) 94–135 10 131–167 120–132 46–62 155–189 20
Nitrate-N export (kg/ha/year) 0.86–10.58 0.02–0.05 1.54–19.72 1.54–19.72 1.52–19.72 3.29–17.4 0.02–0.1
Phosphate input (kg/ha/year) 6–16 0 13–35 18–41 27–43 15–37 0
Phosphate export (kg/ha/year) 0.012–0.169 0.008 0.038–0.458 0.038–0.458 0.038–0.458 0.15–1.834 0.008
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . .
115
116 T. R. Nisbet et al.

Table 8.2 Per cent reduction in diffuse pollutant concentration in surface runoff from upslope
land to watercourses achievable from a well-designed and managed woodland buffer of variable
width. Interpolated from relationships derived from review by Perez-Silos (2017)
Buffer width 5m 10 m 20 m 50 m 100 m
Nitrate-N 20% 30% 40% 80% 90+%
Phosphate-P 10% 20% 30% 60% 90+%
Suspended sediment 80% 90+% 90+% 90+% 90+%

guide, Table 8.2 shows what could be expected in terms of reduction in different
diffuse pollutants in surface runoff from upslope land by an increasing width of a
well-designed and managed woodland buffer, based on a recent review (Perez-Silos
2017). Data are lacking for FIO and many pesticides, with removal efficiency for
the latter greatly depending on the type of pesticide and its pathway of movement.
While the above tables can be used to explore the environmental effectiveness
of woodland planting compared to other measures and help underpin rough cost-
effectiveness estimates, the design of a scheme and especially one that seeks
to make a difference at the catchment level is best informed by hydrological
modelling. Understanding soil and hydrological processes is crucial to identifying
pollutant sources and pathways for correct targeting of measures, as well as for
quantifying and upscaling their environmental effectiveness. It is important to use a
spatially distributed model that can be incorporated within GIS to determine the
most effective placement and integration of measures, accounting for dynamics
in pollutant movement and the evolution of measures. The physically based Soil
and Water Assessment Tool (SWAT) model is often preferred for such applications,
although it is highly data demanding, is unsuited to very small catchments (<150 ha)
and requires great care to ensure correct model parameterisation, especially in
relation to forest processes (Baksic 2018). A groundwater model would be needed
to predict the movement of pollutants to deep boreholes. Where resources and data
are lacking to allow process modelling, more simple tools can be applied such as
the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) suite of
models (Kareiva et al. 2011).

8.5 Importance of Design and Management of FGI

The benefits of tree planting for water can evolve over a number of years due to
the time taken for trees to grow, a forest canopy to form and the associated effects
on water use and soil improvements to become fully established. There can also be
a significant lag before soil stocks of nutrients or contaminants such as pesticides
from the previous land-use breakdown, or are washed out of the soil-rock system.
The delay and lag time will vary with type of pollutant, the depth of water pathways
(being much slower for deeper groundwater) and the nature of tree planting (e.g. tree
type, species and stocking density). For example, reductions in suspended sediment,
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 117

adsorbed/total phosphate and FIO can be rapidly delivered (within 1–3 years) by
the cessation of soil disturbance, removal of livestock and improvement in soil
infiltration with tree rooting following planting. In contrast, while there will be a
rapid step-change in nitrate and pesticide inputs following land-use change, soil
and groundwater stocks of these chemicals can take decades to be removed from
groundwater supplies.
A particular advantage of tree planting is the semi-permanent nature of the land-
use change such that water benefits can be secured in the long term. However,
this relies on planted trees and established woodlands and forests being sustainably
managed and replanted if felled or lost to fire, storms or pest and disease. Productive
management can provide economic benefits to landowners and managers in the form
of timber and wood fuel but poses risks of pollution due to forestry operations such
as timber harvesting and extraction, as well as the temporary loss of water benefits
until trees regrow after felling. These risks can be minimised through good forestry
practice but require great care in forest planning, design and management, especially
involving sensitive locations such as riparian buffers (Forestry Commission 2019).
The targeted planting of woodland buffers for intercepting diffuse pollutants from
adjacent agricultural land requires more detailed design and active management
to sustain and promote pollutant removal. The technical specification will vary
between sites and needs to be determined on a case-by-case basis, with woodland
design tailored to site type, the nature of the pollutant and its pathway of movement.
For example, sediment trapping is enhanced by creating a vegetated and rough
ground surface that benefits from less canopy shade and wider tree spacing. In
contrast, nitrate removal either requires wet woodland with waterlogged soils
to promote nitrate loss by denitrification (note that this increases nitrous oxide
emissions, a potent greenhouse gas (GHG)) or planting close-spaced, faster growing
tree species to maximise nitrate uptake. Where nitrate inputs in surface runoff from
upslope land are very high, regular harvesting of timber or wood fuel may be
necessary to avoid nitrate saturation and overloading buffers. In such cases, wider
buffers would facilitate phased harvesting of strips to maintain some degree of
nitrate removal, although particular care is required to avoid ground damage during
operations.
Other pollutants such as ammonia and pesticides that are dispersed via aerial
pathways require special attention to the design of the structure of the woodland
canopy to maximise aerial deposition and pollutant trapping. A typical example is
the design of woodland buffers around animal housing to reduce ammonia emissions
(Bealey et al. 2016). Another issue can be the presence of drains or soil fissures
that allow pollutants in drainage waters to bypass the soil and the potential for soil
retention and root uptake. This may require physical interventions to disrupt these
pathways such as by drain blocking, although blockage can also occur naturally over
time by tree rooting or soil shrinkage and swelling (Stutter et al. 2020).
Some view leaving FGI unmanaged to be a more attractive option for securing
water benefits, but this is increasingly challenged by climate change and associated
risks of storm, fire and disease outbreaks. Managing these risks is driving the need
for greater intervention to increase tree species and age diversity or install fire
118 T. R. Nisbet et al.

breaks to enhance forest resilience, especially in forests where historic management

has left relatively even-aged or single species stands. Examples include extensive
forest conversion from conifer to broadleaves in areas of Germany as spruce and
pine stands are increasingly impacted by pests and disease (Schuller et al. 2011),
and the installation of forest-fire prevention measures in parts of France. However,
the absence of productive management in some forests makes such interventions
uneconomic for forest owners necessitating economic support.

8.6 Managing Potential Disbenefits

While FGI is generally very good for protecting water quality, there is one common
potential disbenefit. This relates to the ability of trees to use more water than
shorter types of vegetation, resulting in less water runoff or recharge (e.g. due to
interception/wet canopy evaporation and/or potentially higher transpiration rates
sustained by deeper rooting) (Nisbet 2005). The subject is complex, is widely
researched and still attracts debate. Much depends on a wide range of site factors,
especially geographical scale, climate, altitude, geology, soil type, forest type, tree
species, tree age and the counterfactual land cover. In general: conifers reduce water
yield more than broadleaves; differences between individual species tend to be small
(although with a few exceptions); reductions are much less for very young and
old trees; and the impact on catchment water yield is relatively limited (difficult to
measure) when less than 20% of a catchment is planted or cleared of forest (Creed
and Noordvijk 2018).
In some locations, forests can have the opposite effect and increase water yield.
Notable examples include high altitude forests that are effective in trapping cloud
water, the planting of broadleaved forest on grassland overlying chalk geology and
where forests replace irrigated agriculture or crops with a high water use (Creed and
Noordvijk 2018; Roberts and Rosier 2005). The water use of an existing forest can
be reduced by changing forest type from conifers to broadleaves, diversifying forest
age and introducing more open space, although these represent a significant cost to
forest owners.
The water-use issue becomes further complicated concerning the impact of
forests on dry weather flows, when water supplies are most limiting. The generally
greater water use by trees can be expected to reduce low flows but much depends
on the nature of local soils and geology. Permeable geologies are the most
vulnerable, while reductions can be expected to be small or even reversed on
impermeable geologies with poorly structured soils. Here, tree planting can improve
soil infiltration leading to a greater proportion of net rainfall draining to depth
and supplementing low flows. Another exception concerns the presence of riparian
and floodplain woodland, which can enhance the storage of flood waters and their
subsequent release, helping maintain dry season flows. A further complicating factor
is that the higher water use and potential water yield reduction by forests can
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 119

be beneficial for reducing flood flows in flood risk catchments and tackling soil
salinisation issues in dryland environments.
Models (e.g. SWAT, WaSSI-C, Hydro-JULES) are available to estimate the
effect of FGI on water yield, but these vary in their ability to represent and
handle key forest hydrological processes (e.g. wet canopy evaporation), as well as
forest design and management factors. Great care is therefore required in selecting
appropriate models and parameter values, as well as making a serious effort to test
and validate model predictions. Modelling impacts on extreme flows is the most
difficult.
Other potential water disbenefits tend to be more localised in extent and
include the risk of increased water acidification and mobilisation of metals such
as aluminium due to forest canopy scavenging of acid deposition. This issue
is largely limited to acid-sensitive, upland geologies and rapidly declining due
to EU success in emission control (Nisbet and Evans 2014). Another issue is
excessive canopy shade and poor channel morphology resulting from conifer
plantations located too close to watercourses. Care is also required to avoid the
build-up of captured pollutants in tree biomass or in the soil, which could be
released back to the water environment. Lastly, while tree planting and forest
cover generally act to reduce flood risk, there are some localised exceptions,
such as the backing-up of floodwaters upstream of floodplain woodland and the
blocking of culverts and bridges by the washout of woody material (Nisbet et al.
2011).
All above potential disbenefits can be effectively managed by good design and
management of FGI, especially by planting the most appropriate type and species
of tree in the right place based on site factors and ecological requirements.

8.7 Identifying and Assessing Multiple Benefits

While the emphasis of this chapter is on how FGI can protect and improve water
quality, it is important to recognise that this strategy will also increase natural
capital and deliver multiple benefits (from the so-called ecosystem services) for
other policy agendas. There is particular scope for FGI to reduce downstream flood
risk as well as ameliorate rising water temperatures through shade provision, which
are topics of growing concern in the context of climate change (Burgess-Gamble
et al. 2017). FGI will also directly contribute to climate change mitigation through
carbon sequestration and building soil carbon, thereby helping to offset agricultural
GHG emissions (Morison et al. 2012). Other notable benefits include improving
biodiversity and tackling the decline in woodland birds by increasing woodland
habitat and linking-up fragmented woodlands within agricultural landscapes; the
provision of timber and wood fuel to diversify agricultural businesses; and the
120 T. R. Nisbet et al.

potential to improve open landscapes and provide increased access for recreation
(Bateman et al. 2014).

8.8 How to Design a PES Scheme

There are eight operational steps involved in designing a successful PES scheme,
which are described below in the context of using FGI to improve or protect water
quality. Schemes can be simple or complex and large or small, depending on the
location of interest and nature of the water issue. Three case studies summarising
successful schemes are referenced in accompanying text boxes (Boxes 8.2, 8.3, and
8.4), and more information on these can be found via the PESFOR-W web site
(www.forestresearch.gov.uk/research/pesforw/case-studies).
1. Defining the water quality issue: This can take a number of forms. For an
agricultural area, the water issue could typically be an excessive level of diffuse
pollutants generated by farming activities resulting in a failure to achieve water
quality standards and good water status. For an existing forest, it could be that its
inherent water protection function is threatened by an environmental pressure
such as an increased risk of storm damage or fire due to climate change, or
spread of a pest or disease. Alternatively, the threat could be posed by agricultural
intensification or urban expansion. Whatever the issue, the starting point is to
clearly define its nature, including its spatial extent and temporal dimension. In
the case of diffuse pollution, there is a need to determine which pollutants (e.g.
nitrate, phosphate, sediment, pesticides or FIO) are involved, their sources (e.g.
which fields, areas or soils) and pathways of movement (e.g. surface runoff or
groundwater). The water regulatory authority will be a key partner in defining
the water quality issue.
2. Identifying local actors: This involves identifying all stakeholders linked to and
affected by the water issue. In theory, PES schemes can be limited to single
buyers and sellers but are more likely to involve a broad range of actors, espe-
cially for more extensive water issues. Local actors can be categorised into five
main groups: regulatory bodies, suppliers and sellers, beneficiaries and buyers,
intermediaries and designers (Fig. 8.5). Regulatory bodies or beneficiaries and
buyers are more likely to take the lead in developing a PES scheme, while work
will be required to raise awareness of the water issue amongst some actors and
to persuade them to fully engage.
3. Assessing the feasibility of a PES scheme: Bring together stakeholders to
explore the water quality issue and consider potential solutions and opportunities,
drawing on wider experience and examples of different PES schemes. Assess
the existing water quality baseline and margin of improvement needed to meet
a water quality target or standard, or in the case of the loss of an existing
forest protection function, the degree of damage likely to be caused. Check
that the issue cannot be addressed by good management practices, or regulatory
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 121

Fig. 8.5 Main actors who may be involved in designing a PES scheme

mechanisms. Examine the alternative measures that could be adopted, exploring

how costs, benefits and avoided risks differ between them and according to the
extent to which they are used, and identify the least cost and most acceptable
option. Establish whether there are willing buyers and sellers to implement and
finance the preferred measures, a desire to collaborate amongst actors, and trusted
122 T. R. Nisbet et al.

intermediaries to assist with organising and managing a PES scheme. If so, assess
transaction, management and monitoring costs.
4. Exploring potential win-win solutions: Consider whether the identified option(s)
will deliver additional benefits from ecosystem services (e.g. for carbon seques-
tration, flood risk management, recreation and biodiversity [see CICES (Haines-
Young and Potschin 2017)] and if so, whether there is a market for these.
Where there is a willing buyer, try to quantify the potential benefits and assess
the scope to develop an integrated scheme that also considers these other
ecosystem services and impacts on natural capital. To help design the PES
scheme, underpin investment decisions and increase public support, a compar-
ative cost-effectiveness analysis of the per unit environmental improvement can
be undertaken by intermediaries or designers where data are available (Box 8.5).
5. Defining roles and responsibilities: Providing there is local support for develop-
ing a PES scheme, define roles and responsibilities of key actors. This should
include setting clear spatial boundaries for the scheme and agreeing measures,
associated costs, payments and timelines. Seek and draft agreements.
6. Resolving or minimising potential legal issues: Consider legal, fiscal and regu-
latory issues for key actors, such as implications for taxes, property rights and
pollution control, especially for those making or receiving payments. Where
necessary, legal advice should be sought to aid decision making.
7. Drawing-up technical specifications: Technical specifications should be devel-
oped and agreed for the design and management of the selected measure(s) to
address the water quality issue. Regulatory bodies and designers are usually best
placed to advise on the required specification. For tree planting, this will vary
from site to site and include consideration of location (e.g. local climate, geology,
soils and topography), extent/area or width of planting (see Table 8.2), tree
type, species mix, planting density, timing and required management practices
to ensure effective tree establishment and growth (e.g. ground preparation, weed
control and fencing). For an existing forest, it could involve elements of forest
redesign to reduce risks or improve resilience, such as changing forest type or
species mix, altering age structure or introducing fire breaks. The specification
should build-in some leeway (e.g. extra tree planting) to allow for uncertainty in
the effectiveness of a given measure.
8. Formalising scheme contract: A formal contract should be drawn up between
buyers and sellers, covering the technical specification for measures to be
implemented, timelines for delivery, baseline water quality conditions, success
criteria, monitoring needs, staged payments and scheduled reviews. However, it
is best to incorporate a degree of flexibility in the Terms and Conditions to allow
for future adjustments informed by monitoring and evaluation. Care is required
to avoid a high level of bureaucracy and transaction costs, ensuring that scheme
management and monitoring are fit for purpose.
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 123

Box 8.2 Case Study: Tree Planting to Secure Water Benefits

Globally, many water utilities are increasingly recognising the growing threats
to water supplies and rising cost of water treatment. Consequently, attention
is shifting away from grey to green infrastructure to better secure future water
quality and quantity in source catchments. PES schemes are emerging as a
more sustainable approach to water management whereby landowners such as
farmers and forest owners are incentivised to change land use or management
to better protect water supplies. Targeted measures such as tree planting within
safeguard zones potentially offer a more cost-effective way of addressing
diffuse water pollution issues compared to water treatment. At the European
Level, Article 9 of the WFD stipulates that ‘Member States shall take
account of the principle of recovery of the costs of water services, including
Environmental and Resource Costs’. This has led to some water utilities in the
UK (South West Water and United Utilities), Germany (Saxony and Hanover)
and Italy (ETRA and Romagna Acque) to charge consumers for the costs
incurred in source area protection, with payments passed onto landowners
and managers. These schemes are sometimes characterised as ‘PES-like’ as
consumers are not voluntarily paying for the benefit/bill; however, they have
proven to be the most effective systems at EU level for improving water
quality at the catchment level (UNECE 2018).
124 T. R. Nisbet et al.

Box 8.3 Case Study: Tree Planting to Protect Groundwater Quality

There are three notable woodlands for water PES schemes in Denmark. Two
of these are located near Odense on Funen and the third near Aalborg on
Jutland. All are designed to tackle the growing issue of groundwater pollution
by agricultural practices, especially the contamination of drinking water by
rising nitrate and/or pesticide levels. The scheme at Aalborg is one of the
oldest and established in 1991 with funding from EU LIFE and the Aalborg
Municipality to purchase land from farmers within vulnerable groundwater
recharge zones. 900 ha of intensive farmland were converted into 500 ha
of broadleaved woodland and 400 ha of low-input pasture, primarily to
reduce nitrate levels. The drinking water benefit was estimated at a minimum
of AC489/ha/year and the net social benefit (excluding drinking water) at
A
C189/ha/year, which included the provision of local recreation and carbon
gain. The two schemes near Odense are Elmelund Skov and Brylle Water,
both of which involve woodland creation to reduce pesticide pollution of
local groundwater supplies. This is achieved by a voluntary process of land
consolidation whereby agricultural land is purchased in low vulnerable areas
and used to encourage land swaps with farmers for land within vulnerable
groundwater recharge zones. The land is transferred to public or private
partners at a reduced price for woodland planting and management, with a
permanent change from farmland to forest legally guaranteed. At Elmelund
Skov, 380 ha of farmland have been converted to woodland since 2001
under a partnership agreement between the local water utility, the Odense
Municipality and the state forestry agency. The Brylle Water scheme is the
most recent and commenced in 2014. 156 ha of farmland were purchased and
planted with woodland by a private foundation, who met 40% of the cost, with
the other 60% funded by the local water utility. The land consolidation process
involved a significant transaction cost in negotiating agreements with farmers
and building trust. Public access for recreation was a strong component of the
schemes and underpinned municipality funding and support.
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 125

Box 8.4 Case Study: Tree Planting to Improve Groundwater Infiltration

This PES scheme is located on the edge of the town of Carmignano di
Brenta, near Padua in northern Italy. It was established in 2012 as a ‘Forested
Infiltration Area’ (FIA) to help replenish and improve groundwater resources
in the area. Overexploitation of the aquifer had led to the disappearance
of local springs and streams, while agricultural activities had degraded
groundwater quality. A 2.5 ha broadleaved woodland was planted on arable
(maize) land and a system of trenches dug to channel surface water (at a
rate of ~1 million cubic metres per hectare per year) onto the site during
periods of excess flow in the nearby River Brenta. The establishment of the
woodland helped to facilitate water infiltration into the aquifer and enhanced
phyto-purification, removing nutrients and other contaminants. The woodland
also provided a carbon gain to the landowner and woodland products such as
firewood, biomass and timber in the longer term, as well as benefiting the
local community as a valuable habitat and opportunities for recreation and
education. A group of local and regional stakeholders, including municipali-
ties and local companies, formed a partnership to bid for supporting funds that
were used to design and set up the PES scheme on private farmland. Around
80% of implementation costs were financed by LIFE+ and RDP funds. The
loss of income to the landowner from the change in use from maize cropping
to woodland was exceeded by payments from the Brenta Land Reclamation
Board for the infiltration water service (A
C1200/ha/year), the municipality for
community access and related recreation and education events (A C1500/year),
plus the value of generated wood products and carbon gain.
126 T. R. Nisbet et al.

Box 8.5 Cost-Effectiveness Analysis (CEA)

Cost-effectiveness analysis (CEA) is a technique that can be used to compare
the cost of FGI with that of alternative measures for improving water quality.
It can be important both in making a case to adopt woodland planting and in
selecting which woodland (and/or other) measures to implement. The main
challenge is to estimate how much planting and/or the extent of an alternative
measure is required to achieve a given water quality target. It is easiest to
focus on a specific diffuse pollutant or set of pollutants such as reducing the
concentration of nitrate in a watercourse by X% or to below an environmental
standard, rather than adopting a broader water quality index like ‘water status’
or using a biological metric. The assessment is likely to require the use of a
spatially distributed model such as SWAT to estimate the effect of different
levels and locations of planting on water quality, as described in Sect. 8.5. An
alternative, simpler approach would be to limit consideration to the effect of
alternative measures on pollutant loads, with values for the latter relatively
well known for different land covers and crop types (Table 8.1). In cases
where measures are proposed in order to reduce the risk of infringing water
quality targets that are currently met, the improvement would need to be
considered in terms of the reduced risk of exceeding this target.
Cost-effectiveness analysis (CEA) involves dividing the cost of the wood-
land planting or other measure by the improvement/gain made in water
quality. Depending on whether a reduction in concentration or load is the
focus, the cost-effectiveness ratio (RCE) computed may be expressed in units
such as AC/mg/l or A C/kg/ha of a given pollutant. The aggregate cost should
be computed by discounting costs in future years over the lifespan of the FGI.
The costs need to include the revenue forgone from the change in land use and
any transaction costs. Where a private sector perspective is adopted, the costs
should also include any changes in financial incentives involved. However,
where CEA is undertaken from a societal perspective, changes in subsidy
payments (e.g. associated with EU agri-environmental schemes) should be
excluded because these are considered transfer payments. Although the focus
here is on water quality, the value of other benefits generated by the woodland
can also be included in the calculation and used to offset the cost, so reducing
the RCE. This can be a challenging task for some benefits such as habitat gain
but easier for others like carbon sequestration.
Key steps involved in CEA are:
• Identify the water quality issue and level of required improvement, such
as in the concentration or load of a specific diffuse pollutant to meet or
contribute to achieving a given water quality standard.
• Estimate the amount of FGI and other alternative measures needed to
deliver the desired reduction in pollutant concentration or load using mod-

(continued)
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 127

elling (e.g. SWAT) or pollutant budgeting. Calculate the aggregate cost of

implementing the measures, including revenue forgone, transaction costs
and (for a private sector perspective) incentives, discounting management
and other costs expected to arise in the future over the lifespan of the
measures.
• In cases where the expected changes in measure effectiveness over time
differ between measures, one of two approaches should be adopted: either
divide the aggregate costs by an environmental benefit index that weights
future environmental improvements according to when they are expected to
arise (e.g. by applying a discount rate to improvements in future years) or
compare measures using a benchmark cost-effectiveness cost-comparator
that takes account of when improvements arise (an approach sometimes
used in comparing options for climate change mitigation—e.g. DBEIS
2019).
• Consider other benefits and any disbenefits provided by the measures or by
the original land use and where feasible, estimate the value or costs of these
(similarly by applying a discount rate to those arising in future years), to
compute the net cost of each measure.
• Taking into account the above net costs and environmental effectiveness
of potential measures, including an allowance for risk and uncertainty,
calculate the average and range of the RCE for each measure and compare
performance. Marginal Abatement Cost Curves can be used to explore the
relationship between the cost-effectiveness of different measures and the
total amount of diffuse pollution abated at specific points in the future.
• Use results to design an appropriate payment scheme to deliver the
least-cost measure or mix of measures to meet and secure water quality
target(s) within a given timeframe, taking account of local suitability and
acceptability of measure(s).
• Where attracting a range of investors for FGI is important, consider
computing cost-effectiveness from other perspectives too (e.g. climate
change mitigation and/or flood risk attenuation).
• It is important to keep all relevant documentation detailing calculations to
inform future review and learning.

8.9 Monitoring, Evaluation and Review

Monitoring can take many forms that vary greatly in cost. Firstly, implemented
measures will require a certain level of monitoring to ensure that they are designed
as planned and managed appropriately to deliver and maintain their effectiveness
in pollutant reduction. For FGI, this includes checking that the trees become fully
128 T. R. Nisbet et al.

established (e.g. replacing any losses and managing grazing and potential weed and
pest issues) and that the established woodland is sustainably managed and, where
appropriate, the woodland replanted to secure long-term effectiveness. In the case
of more targeted planting such as in the form of riparian woodland buffer areas,
there is a need to check that these are designed and managed appropriately to deal
with the diffuse pollution issue (and to deliver any co-benefits) over the expected
lifespan of the buffer. This includes checking that tree growth rates are sufficient
and sustained to deal with nutrient runoff from adjacent land, the buffer does not get
overloaded/saturated by the pollutant(s) and any management interventions do not
damage the site.
Secondly, in many cases, there will be a need to monitor the water quality
response to check that the FGI is having the desired effect. It may be possible to
rely on existing monitoring networks (e.g. for WFD assessments) operated by water
regulators for this purpose, but these will usually be undertaken at the water body
scale and so may need to be supplemented by local measurements. Monitoring needs
to be tailored to the nature of the implemented measure, how it will affect diffuse
pollutant sources and pathways and the specific pollutant(s) involved. For example,
small-scale woodland planting is less likely to justify river water quality monitoring
and better suited to conducting plot-based measurements such as changes to soil
conditions. There should be agreement on the location, type, frequency and cost of
monitoring; data handling, storage and ownership; and on data analysis, reporting
and publication of results.
Tackling water body scale diffuse pollution issues is likely to require extended
areas of targeted FGI involving multiple landowners and taking multiple years to
deliver. This will necessitate strategic planning and integrated catchment manage-
ment, with regular review of progress and actions. Progress made in achieving
water quality targets can be directly linked to contracted payments, although the
high temporal variability of water quality parameters (e.g. due to variability in
weather conditions and river levels) can make this very challenging, especially
over short-medium timescales. Consideration should be given to undertaking a
periodic evaluation of the cost-effectiveness of the scheme to inform the need for
any changes, share lessons and ultimately ensure long-term success.

8.10 Spreading the Word

Successful communication, dissemination and marketing of PES actions are depen-

dent on messages being appropriately targeted to the relevant actors. Levels of
awareness and understanding of the benefits of FGI for water will vary greatly
between actors, and therefore different approaches and stories may be required.
Be aware of different learning styles and tailor information accordingly. Academic
approaches that apply in the classroom may not work out in the field. Be sensitive
to local conditions and relationships.
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 129

Sources and pathways of diffuse pollution are likely to be disputed, especially

amongst landowners and managers, necessitating an open discussion of the evidence
and drawing on expert opinion. A number of alternative options for tackling a
given water issue will often be available, and those involving land-use change and
particularly woodland creation will face inherent resistance by some. Focus on
careful targeting of FGI to enhance effectiveness and minimise impact on existing
land use.
Bringing different groups together, discussing contrasting viewpoints and resolv-
ing differences are critical. Consideration should be given to using a facilitator
such as a trusted intermediary, especially where there are deep-seated differences
between actors. The use of maps and model outputs can be very helpful but
must be handled sensitively and not used to apportion blame. Many will not trust
these products and question their ability to reflect reality. Consequently, it is very
important to ground truth such information, which is best done by key actors
meeting and discussing the issues and proposed countermeasures in the field.
Use should be made of regional and wider case studies to develop options and
build confidence and consensus, including by visiting existing demonstration sites
where possible (check relevant case studies at www.forestresearch.gov.uk/research/
pesforw/case-studies). These studies can be a great way of showcasing the water
and wider benefits of FGI, as well as the protection function of existing forests and
threats faced.

Acknowledgements This chapter is based upon work from COST Action CA15206, supported by
COST (European Cooperation in Science and Technology). COST is a funding agency for research
and innovation networks. Our Actions help connect research initiatives across Europe and enable
scientists to grow their ideas by sharing them with their peers. This boosts their research, career
and innovation. See www.cost.eu

References

Baksic N (2018) Reviewing available pollutant models and decision support tools for informing
the design and management of woodland creation measures for reducing agricultural diffuse
pollution. STSM Scientific Report. COST Action 15206. Forest Research, Surrey. 12 pp
Bateman I, Day B, Agarwala M, Bacon P, Bad’ura T, Binner A, De-Gol A, Ditchburn B, Dugdale
S, Emmett B, Ferrini S, Carlo Fezzi C, Harwood A, Hillier J, Hiscock K, Hulme M, Jackson B,
Lovett A, Mackie E, Matthews R, Sen A, Siriwardena G, Smith P, Snowdon P, Sünnenberg G,
Vetter S, Vinjili S (2014) UK National Ecosystem Assessment Follow-on. Work package report
3: economic value of ecosystem services. UNEP-WCMC, LWEC, UK, 246 pp
Bealey WJ, Dore AJ, Dragosits U, Reis S, Reay DS, Sutton MA (2016) The potential for tree
planting strategies to reduce local and regional ecosystem impacts of agricultural ammonia
emissions. J Environ Manag 165:106–116
130 T. R. Nisbet et al.

Broadmeadow S, Nisbet T (2012) National map of woodland creation opportunities: targeting

eWGS to help meet the objectives of the WFD and reduce flood risk in England. Final Report
to Forestry Commission England. Forest Research, Alice Holt Lodge, Farnham, Surrey. 29 pp
BSFP (2013) British survey of fertiliser practice dataset. From Department for Environment, Food
& Rural Affairs, London
Burgess-Gamble L, Ngai R, Wilkinson M, Nisbet T, Pontee N, Harvey R, Kipling K, Addy S,
Rose S, Maslen S, Jay H, Nicholson A, Page T, Jonczyk J, Quinn P (2017) Working with
natural processes—evidence directory. Environment Agency, Bristol. 298 pp
Collins AL, Newell Price JP, Zhang Y, Godday R, Naden PS, Skirvin D (2018) Assessing the
potential impacts of a revised set of on-farm nutrient and sediment ‘basic’ control measures for
reducing agricultural diffuse pollution across England. Sci Total Environ 621:1499–1511
Creed IF, Noordvijk Mv (eds) (2018) Forest and water on a changing planet: vulnerability,
adaptation and governance opportunities. A global assessment report, IUFRO World Series,
vol 38. IUFRO, Vienna. 192 pp
DBEIS (2019) Valuation of energy use and greenhouse gas. Supplementary guidance to the HM
Treasury Green Book on Appraisal and Evaluation in Central Government. Department of
Business, Energy and Industrial Strategy, London. https://assets.publishing.service.gov.uk/
government/uploads/system/uploads/attachment_data/file/794737/valuation-of-energy-use-
and-greenhouse-gas-emissions-for-appraisal-2018.pdf
EEA (2016) Annual indicator report series (AIRES): surface waters. EEA Environmental Indicator
Report No 30/2016. EEA, Copenhagen
EEA (2018) European waters—assessment of status and pressures 2018. European Environment
Agency EEA Report 7/2018. EEA, Copenhagen. https://www.eea.europa.eu/publications/state-
of-water
Forest Europe (2019) Valuation and payments for forest ecosystem services in the pan-European
region. Final report of the forest Europe Expert Group on valuation and payments for forest
ecosystem services. Forest Europe, Bratislava. https://foresteurope.org/publications/
Forestry Commission (2019) Managing forest operations to protect the water environment.
Forestry Commission Practice Guide, Forestry Commission, Edinburgh, 48 pp
Gatto P, Pettenella D, Secco L (2009) Payments for forest environmental services: organisational
models and related experiences in Italy. iForest-Biogeosci Forest 2(4):133–139
Haines-Young R, Potschin MB (2017) Common International Classification of Ecosystem Services
(CICES) V5.1 and guidance on the application of the revised structure. www.cices.eu
Kareiva P, Tallis H, Ricketts TH, Daily GC, Polasky S (2011) Natural capital: theory and practice
of mapping ecosystem services. Oxford University Press, Oxford
Mockler AEM, Bruen M (2018) Support tools for characterisation and evaluation of programmes
of measures. Report no. 249. Environment Protection Agency, Dublin
Morison J, Matthews R, Miller G, Perks M, Randle T, Vanguelova E, White M, Yamulki S (2012)
Understanding the carbon and greenhouse gas balance of forests in Great Britain. Forestry
Commission research report. Forestry Commission, Edinburgh. 149 pp
Nisbet TR (2005) Water use by trees. Forestry Commission Information Note 65. Forestry
Commission, Edinburgh
Nisbet TR, Evans CD (2014) Forestry and surface water acidification. Forestry Commission
research note 16. Forestry Commission, Edinburgh
Nisbet T, Silgram M, Morrow K, Broadmeadow S (2011) Woodland for water: Woodland measures
for meeting Water Framework Directive objectives. Forest Research Monograph 4. Forest
Research, Surrey. 156 pp
Perez-Silos I (2017) Assessing the effectiveness of woodland creation for reducing agricultural
diffuse pollution—developing value ranges to create look-up tables. STSM scientific report.
COST action 15206. Forest Research, Surrey. 11 pp
Roberts J, Rosier P (2005) The impact of broadleaved woodland on water resources in lowland
UK: III. The results from Black Wood and Bridgets Farm compared with those from other
woodland and grassland sites. Hydrol Earth Syst Sci 9:614–620
8 Forest Green Infrastructure to Protect Water Quality: A Step-by-Step Guide. . . 131

Schuller G, Pfister L, Vohland M, Seeling S, Hill J (2011) Large scale approaches to forest and
water interactions. In: Bredemeier M, Cohen S, Godbold D, Lode E, Pichler V, Schleppi P (eds)
Forest management and the water cycle: an ecosystem-based approach, Ecological Studies, vol
212. Springer, New York, pp 435–452
Stutter MI, Chardon WJ, Kronvang B (2012) Riparian buffers as a multifunctional management
tool in agricultural landscapes: introduction. J Environ Qual 41:297–303
Stutter M, Wilkinson M, Nisbet TR (2020) Improving the benefits from watercourse field margins
using 3-D buffers. Environment Agency report. Environment Agency, Bristol
Ucar T, Hall FR (2001) Windbreaks as a pesticide drift mitigation strategy: a review. Pest Manag
Sci 57:663–675
UNECE (2018) Forests and water—valuation and payments for forest ecosystem services.
UNECE, Geneva. 108 pp

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
 Part III
River and Floodplain Ecosystem (Including
Paddy Field and Other Farmlands)
Chapter 9
Wetland Paddy Fields as Green
Infrastructure Against Flood

Yasunori Muto and Ryo Yokokawa

Abstract Inundation simulation was carried out for an inland depression mainly
covered by paddy fields to estimate its flood retention volume. In addition to surface
water flooding, river flooding combined with it was also explored in a wide range
of precipitation including a 1000-year return period. The results showed that the
paddy fields in the studied area can retain approximately 60% of the total inundation
volume. Judging from the relationship between the flood retention volume in the
paddy fields and the number of inundated buildings above floor level, we could
estimate two reference values regarding the potential volume of floodwater retention
in paddy fields: one at which damage to the buildings starts to appear and the
other where damage seriously expands in the main residential area. By using
these reference values, we demonstrated the importance of land-use strategies in
flood alleviation. In other words, by transferring buildings inundated between these
levels to safer places, the potential volume of floodwater retention in paddy fields
can be increased to a higher level, which well agrees with the concept of green
infrastructure. The idea of flood alleviation deduced from this study can be applied
to similar near-shore depressions cut off by sand dunes, which are common in Japan.

Keywords Wetland paddy fields · Flood retention · Surface water flooding ·

River flooding · Combined flooding

Y. Muto ()
Research Center for Management of Disaster and Environment, Tokushima University,
Tokushima, Japan
Department of Civil and Environmental Engineering, Graduate School of Technology, Industrial
and Social Sciences, Tokushima University, Tokushima, Japan
e-mail: [email protected]
R. Yokokawa
Water and Land Sector, CTI Engineering Co., Ltd., Osaka, Japan
e-mail: [email protected]

© The Author(s) 2022 135

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_9
136 Y. Muto and R. Yokokawa

9.1 Introduction

Paddy fields are well known to have many functions as green infrastructure, includ-
ing flood risk reduction, water resources conservation, food production for both
humans and animals, landscape improvement, community maintenance through
agricultural activities, etc. (Ministry of the Environment, Japan 2016). As for flood
risk reduction, although this has been pointed out qualitatively in early days (e.g.
Masumoto 1998), there are few studies that quantitatively evaluate the flood control
capacity of paddy fields. Recently in Japan, the so-called tambo-dam, using an
orifice plate in a drain to improve the flood control volume of paddy fields, has
been widely explored and has been adopted in some places (e.g. Yoshikawa et al.
2010). However, in contrast to such a micro device, exploring the macro function of
paddy fields to floods, in connection with land-use planning, is still not popular.
From another aspect, owing to recent advances in numerical simulation tech-
niques, inundation simulation models have been widely applied to urban flooding
and have been proven to reproduce well flooding processes of surface water flooding
(e.g. Thang et al. 2004). Such models can also be applied to wetlands to evaluate
their flood control capacity through inundated volume and its time variation (e.g.
Miyazu et al. 2012). However, most of the simulations considered surface water
flooding and river flooding separately. Since a drainage network system consists of
channels of various sizes, from small streams to large rivers, irrespective of either
urban or rural areas, their design flood levels are different according to the channel
size. In other words, particularly in a low and depressed area, flooding is likely to
start from a small stream due to surface water flooding, and then inundation by river
flooding from a larger river is combined.
Based on these exposed points, we conducted inundation simulations for an
inland depression. The studied area is a small plain near the sea, cut off by sand
dunes and presumably used to be a lagoon. Lower areas are mainly used as paddy
fields and residential areas are located in higher places. Such topography and land
use are typically seen along the shore in Japan (Kusaka 2020); thus, it is valuable to
highlight the inundation characteristics of the area. The outputs will be anticipated
to show a standard for flood alleviation measures in such areas.

9.2 Studied Area

The studied area is located in the southern part of Tokushima Prefecture, Japan
(see Fig. 9.1). It is surrounded by the Osato beach (East), Kaifu River (South) and a
mountainous area (North and West), and forms a depression of 2.86 km2 in an inland
area. The elevation of the studied area is shown in Fig. 9.2 in terms of the altitude
distribution drawn from the 5-m mesh DEM data. As shown in the figure, the eastern
part along the beach is relatively higher because of geological sand dunes; thus, the
inland part, even away from the beach, is lower than the dunes. The lowest parts are
9 Wetland Paddy Fields as Green Infrastructure Against Flood 137

Fig. 9.1 The studied area: the Zenzo, Osato and Nishinosawa rivers basin, Tokushima, Japan

located along the three small streams, the Zenzo, Osato and Nishinosawa rivers. The
land-use classification for the same mesh size is shown in Fig. 9.3. Nearly 35% of
the area is utilised as paddy fields, mainly in the lower western and southern parts.
The three streams together with irrigation channels that spread over the wetland
are eventually connected to the Kaifu River, and build a network drainage system
for the area. In this study, the downstream boundary of the simulation area was
set at the confluence point of the Kaifu River. Although floodgates are set at the
confluence, they are open in normal time; thus, the water level in the streams is
affected by the tide level through the Kaifu River and the confluence, particularly in
their downstream reaches. During floods in the Kaifu River, floodgates are closed
and a pump of 5 m3 /s is operated for drainage when the water level in the Zenzo
River exceeds a certain value.

9.3 Inundation Simulation

9.3.1 Simulation Model

A commercial inundation simulation software, AFREL (© NITA Consultant Co.,

Ltd. n.d.), was employed in this study. As shown in Fig. 9.4, AFREL can treat not
only the surface flow but channel and pipe networks for sewerage overlapped on
the surface. Moreover, flow exchanges among these three forms (surface, channel
and sewerage pipe flows) can also be simulated; therefore, both inundation into the
drainage and flooding from a river into the ground surface can be considered. In the
138 Y. Muto and R. Yokokawa

Fig. 9.2 The elevation of the studied area

studied area, a sewerage pipe network is not installed; thus, it was omitted in the
simulation.
The governing equations for the 2-D surface flow based on Iwasa and Inoue
(1982) are as follows:

∂h ∂M ∂N
+ + = r(t) + qCH + qSW (9.1)
∂t ∂x ∂y
∂M ∂ ∂ ∂H τbx
+ uM + vM = −gh − (9.2)
∂t ∂x ∂y ∂x ρ
∂N ∂ ∂ ∂H τby
+ uN + vN = −gh − (9.3)
∂t ∂x ∂y ∂y ρ
√ √
τbx gn2 M u2 + v 2 τby gn2 N u2 + v 2
= , = (9.4)
ρ h4/3 ρ h4/3

The governing equations for the channel flow are basically the same as those for
the surface flow but omitting the advection terms in the momentum equations, as
9 Wetland Paddy Fields as Green Infrastructure Against Flood 139

Fig. 9.3 The land-use classification of the studied area

Fig. 9.4 Structure of the inundation simulation model, AFREL (©NITA Consultant Co., Ltd. n.d.)

follows:
∂h ∂qx ∂qy
+ + = qGR + qSW (9.5)
∂t ∂x ∂y
∂qx ∂H τx
= −gh − (9.6)
∂t ∂x ρ
140 Y. Muto and R. Yokokawa

Table 9.1 Roughness coefficients and infiltration coefficients adopted in this study
Land-use type Roughness coefficient Infiltration coefficient
Paddy fields 0.060 0.70
Other farmlands 0.050 0.60
Forests 0.060 0.60
Uncultivated 0.050 0.20
Buildings 0.050 0.80
Roads/railways 0.047 0.85
Others 0.050 0.80
Water areas 0.030 1.00

∂qy ∂H τxy
= −gh − (9.7)
∂t ∂y ρ
 
τx gn2 Rx 1/3 |qx | qx τxy gn2 Ry 1/3 qy  qy
= , = (9.8)
ρ h2 ρ h2

where qCH , qSW and qGR are the interacting discharges with the channel, sewerage
pipe and ground surface, respectively. qx and qy are defined at the same grid point
as those for M, N and h.
These equations are numerically solved explicitly on the staggered grid in
space, and by the leapfrog method in time. Runoff discharge from the surrounding
mountainous areas of 2.78 km2 is estimated by the rational method and given at
the upstream end of a channel connected to the relevant mountain stream. Table
9.1 shows the roughness coefficients and infiltration coefficients adopted here in
accordance with land-use classification.

9.3.2 Precipitation and River Discharge Setting

The model hyetograph was generated based on the records at the nearest observation
point to the studied area during Typhoon Nakri 1412, August 2014, as shown in
Fig. 9.5. At this event, the total rainfall was approximately 640 mm within 72 h;
however, for the studied area, 48 h is more suitable as a design rainfall duration,
because the main part of consecutive rainfall in this event was observed from 6:00
on August 2 to 5:00 on August 4. Consequently, the annual maximum rainfall in
48 h of 590.5 mm recorded during that time was selected. In addition, probability
analysis was conducted using 45 annual maximum 48-h rainfall data recorded at the
same observation point. The results of the probability analysis (LogP3) are shown in
Table 9.2. As an input for the simulation, the shape of the hyetograph was identical
9 Wetland Paddy Fields as Green Infrastructure Against Flood 141

Fig. 9.5 The original hyetograph (recorded in Typhoon 1412, August 2014)

Table 9.2 Probability analysis of 48 h precipitation (LogP3)

Return period (year) Total precipitation (mm) Magnification ratio
2 332.0 0.56
5 443.8 0.75
10 508.5 0.86
20 564.5 0.96
25 581.1 0.98
27 (original) 590.5 1.00
30 594.4 1.01
50 629.7 1.07
80 660.3 1.12
100 674.3 1.14
150 698.8 1.18
200 715.5 1.21
400 754.0 1.28
500 765.9 1.30
1000 801.3 1.36

(see Fig. 9.5), but the unit precipitation (for 10 min) was uniformly enlarged in
accordance with the calculated ratio shown in the table.
Before estimating the flood discharge from the Kaifu River into the studied
area, the model hydrograph of river discharge was calculated as follows: first, the
peak discharge at a certain probability in the river was estimated using the rational
method, with the corresponding probability precipitation as an input. Next, using the
stage-discharge curve and the recorded stage hydrograph during Typhoon Nakri,
the discharge hydrograph was worked out at the upstream boundary of the 2-D
river flow simulation model, 12.4 km upstream from the river mouth. Then, a 2-D
142 Y. Muto and R. Yokokawa

Table 9.3 River discharge conditions at the upstream boundary

Return period (year) River discharge (m3 /s) Magnification ratio
27 (original) 2230 1.00
50 2490 1.12
80 2600 1.17
100 2680 1.20
1000 4480 2.01

unsteady flow simulation was performed to obtain a stage hydrograph at an arbitrary

levee breaching point. The peak discharge at the upstream boundary was changed
in accordance with the required probability precipitation; however, the shape of the
hydrograph was fixed, just enlarging the unit discharge. Table 9.3 summarises the
peak discharges at the upstream boundary. Since the studied area is located close
to the sea, the water stages in both the Kaifu and Zenzo rivers are affected by tidal
levels. Thus, the water stage recorded at the river mouth of the Kaifu River during
Typhoon Nakri was given as the downstream boundary condition.
Levee breaching points were set at 0.4 km and 1.8 km from the river mouth,
as those were indicated as weak points against floods in the simulation in terms
of relation between the levee height and the peak stage during floods. The levee
breaching level was set at the record maximum water level (RMWL) as the primary
condition. We assumed that river water starts to flood into the adjacent plains when
the water stage in the river reaches the given breach level. In addition, two more
breaching levels, the design high water level (HWL) and HWL + 0.6 m, were
also considered as cases of levee reinforcement. The relation between water stage
hydrograph and the three levee breaching levels considered here is schematised in
Fig. 9.6. Since RMWL was lower than HWL, breaching at RMWL provides more
flooding volume into the studied area than at HWL. The river flooding discharge as
an input into the studied area in the inundation simulation was calculated using a
modified Homma’s formula for overflow (JSCE 1999).
Two inundation scenarios were explored: one considering only surface water
flooding and the other river flooding in addition to surface water flooding. In
the latter, the probability of precipitation for river discharge estimated by the
aforementioned method and that for surface water flooding was set identically.

9.4 Results and Discussions

9.4.1 Model Verification

In order to check whether the model could adequately reproduce an actual inun-
dation event, model verification was performed with the inundation records in
Typhoon Nakri. The given precipitation hyetograph is shown in Fig. 9.5. Table 9.4
9 Wetland Paddy Fields as Green Infrastructure Against Flood 143

Fig. 9.6 A schematic image of water stage hydrograph and levee breaching levels

Table 9.4 Model Point no. Maximum inundation depth (m) Errors (m)
verification: error analyses for
Observation Simulation
the point depths
1 0.2 0.21 +0.01
2 0.7 0.69 −0.01
3 0.3 0.25 −0.05
4 0.3 0.21 −0.09
5 0.2 0.20 ±0
6 0.8 0.72 −0.08
7 0.8 0.88 +0.08
8 0.4 0.32 −0.08
9 0.3 0.37 +0.07
10 0.4 0.36 −0.04
11 0.4 0.42 +0.02
12 0.6 0.62 +0.02
13 0.2 0.22 +0.02

shows a comparison between the water depths read from the flood marks at 13 points
in the field (see Fig. 9.7) and the simulation results. The table shows that the errors
were less than 10 cm at all points, and their average was 1 cm, which indicates that
the model shows quite high reproducibility.
144 Y. Muto and R. Yokokawa

Fig. 9.7 The locations of inundation depth verification

9.4.2 Surface Water Flooding

Figure 9.8 shows the relationship between the return period of total precipitation
and the peak inundation volume for surface water flooding. The dotted line shows
their regression. The relation considering the logarithm of the return period shows
almost linear.
Figure 9.9 shows inundation depth distributions at the peak volume for various
return periods. As shown in the figure, the depth gradually increases as the
return period increases; however, the inundated area does not considerably spread.
Referring to the land-use classification in Fig. 9.3, the inundated area consists
mostly of paddy fields. In particular, the southern part along the downstream
of Zenzo River and the western part around the Nishinosawa River start to be
inundated even in a high frequent flood like a 2-year return period. Most of the
flooding water remains within paddy fields in these parts even in low frequent
floods.
9 Wetland Paddy Fields as Green Infrastructure Against Flood 145

Fig. 9.8 The relationship between the return period of total precipitation and the peak inundation
volume for surface water flooding

Figure 9.10 shows the relation between the return period of total precipitation and
the area of buildings inundated above floor level. Here inundation above floor level
is defined as the water depth on a building mesh over 45 cm. Similar to the relation
between the return period and the inundation volume shown in Fig. 9.8, a linear
function can be drawn for the inundated buildings. Nevertheless, its ratio to the
total building area, 48.2 ha, is utmost 3.1% among the studied cases. This can also
be recognised in Fig. 9.11, where locations of the inundated buildings are shown.
The figure shows that most of the inundated buildings suffer from 2-year return
period flooding, and the number of buildings dose not considerably increase as the
return period increases. In other words, only a quite limited amount of buildings is
damaged by surface water flooding in the studied area.
Figures 9.12 and 9.13 show the distributions and ratio of damaged paddy fields
in response to the precipitation return period, respectively. Here, the threshold of
damaged/non-damaged is defined as the water depth 30 cm. Although the number
of damaged paddy fields becomes larger as the return period increases, from 25% to
50%, they are located within a limited area, mainly around the Nishinosawa River
and the upstream of Osato River. Referring to the land elevation in Fig. 9.2, those
areas are identical to the lower area. In such areas, lots of endangered hygrophytes
can be found (Muto et al. 2020). It should be stressed that 25% of the paddy fields
are damaged even in a high frequent flood like 2-year return period.
Figure 9.14 shows the estimated inundation volumes stored only in the paddy
fields, plotted as a function of the total inundation volume at its peak. The graph
shows that the paddy fields retain 64% of the inundation volume, and this value is
nearly constant over the tested return period.
146 Y. Muto and R. Yokokawa

Fig. 9.9 The depth distributions at the peak inundation volume for surface water flooding

9.4.3 River Flooding Combined with Surface Water Flooding

Figure 9.15 shows the relation between the return period of total precipitation and
the peak inundation volume for the river flooding combined with surface water
flooding. Here the breaching level was set at RMWL, and the results for the surface
water flooding are also plotted. The dotted lines in the figure indicate tendencies.
By adding the river flooding, the inundation volume drastically increases compared
with the surface water flooding only. In addition, the inundation volume is different
as the breaching point changes. In this study, flooding from 1.8 km supplies more
inundation volume than from 0.4 km. Thus, in sequence, the results shown are
mainly for breaching at 1.8 km.
Figure 9.16 shows the inundation depth distributions at the peak volume for
various return periods. In the figure, the breaching point is denoted by red X.
Considering the depth distribution only for surface water flooding shown in Fig.
9.9, the combined flooding in a 50-year return period matches as the surface water
flooding in a 1000-year return period regarding inundated area and depths. This
9 Wetland Paddy Fields as Green Infrastructure Against Flood 147

Fig. 9.10 The relation between the return period of total precipitation and the area of buildings
inundated above floor level for surface water flooding

can also be seen in Fig. 9.15, where the peak inundation volumes are more or less
similar, 700 × 103 m3 . As the precipitation return period increases, the inundation
depth initially increases gradually, and then both depth and inundated area largely
increase from 100-year to 1000-year return period. Such expansion of inundation is
influenced by the breaching point location. When comparing the same return period,
breaching at 1.8 km causes more severe flooding damage than at 0.4 km. This is
probably due to not only larger flooding volume (see Fig. 9.6) but wider spreading
flooding water because breaching occurs upstream.
In order to estimate the depth increment due to river flooding only, the inundation
depth due to surface water flooding, shown in Fig. 9.9, was subtracted from the depth
by combined flooding, as shown in Fig. 9.16, in each mesh, that is:

Hrf = Hcf − Hsf (9.9)

where Hrf is the depth increment attributed solely to river flooding and Hcf and
Hsf are the maximum water depths in each mesh in combined and surface water
flooding, respectively. The distributions of Hrf for various return periods are shown
in Fig. 9.17. Not surprisingly, the area near the breaching point shows large
increments. By using Hrf , the expanding process of the inundated area in accordance
with the return period becomes clearer. For example, the downstream of the Zenzo
River is also extensively affected, although apart from the breaching point to some
distance. This indicates that the flooding water from 1.8 km point runs easterly
between the Kaifu and Zenzo rivers. Another interesting feature is that paddy
fields in the western and northern parts take charge of floodwater through the three
streams. At least up to 100-year return period, floodwater remains within the paddy
fields, except for the area near the breaching point. However, in the 1000-year return
148 Y. Muto and R. Yokokawa

Fig. 9.11 The locations of buildings inundated above floor level in response to the precipitation
return period for surface water flooding

period, the inundated area spreads devastatingly, and the maximum depth increment
reaches up to 3–4 m near the downstream end of the studied area.
Figure 9.18 shows the relationship between the return period of total precipitation
and the area of buildings inundated above the floor level. Similar to the peak
inundation volume in Fig. 9.15, the number of inundated buildings increases
drastically in the combined flooding. Figure 9.19 shows the locations of the
inundated buildings. As shown in the figure, the number of inundated buildings
increases gradually as the return period increases. In addition, there are inundated
buildings even in the central part including the town office and railway station. In
the worst case, the ratio of inundated buildings reaches 46.3%. Such an expanding
process of the damaged area is well coincident with the expansion of the inundated
area mainly attributed to river flooding, as shown in Fig. 9.17.
Figures 9.20 and 9.21 show the distributions and ratio of damaged paddy fields in
response to the precipitation return period, respectively. As shown in these figures,
the ratio of damaged paddy fields in combined flooding is fairly higher than that in
9 Wetland Paddy Fields as Green Infrastructure Against Flood 149

Fig. 9.12 The distributions of damaged paddy fields in response to the precipitation return period
for surface water flooding

Fig. 9.13 The ratio of damaged/non-damaged paddy fields for surface water flooding
150 Y. Muto and R. Yokokawa

Fig. 9.14 The estimated inundation volume stored in the paddy fields for surface water flooding

Fig. 9.15 The relation between the return period of total precipitation and the peak inundation
volume for combined flooding

surface water flooding. The ratio in the 1000-year return period becomes as high
as 94%. Comparing Figs. 9.20 and 9.12, in the western and northern parts, the
locations of damaged paddy fields are almost the same under a 100-year return
period. Consequently, this increment of damaged paddy fields is mainly brought
9 Wetland Paddy Fields as Green Infrastructure Against Flood 151

Fig. 9.16 The depth distributions at the peak inundation volume for combined flooding (breaching
at 1.8 km denoted by red X, breaching level at RMWL)

by inundation close to the breach point by river flooding. In the 1000-year return
period, almost all the western and northern paddy fields are damaged.
Figure 9.22 shows the estimated inundation volumes only in the paddy fields,
plotted in the same manner as in Fig. 9.14. The graph shows that the paddy fields
retain 60% of the inundation volume for combined flooding, which slightly smaller
than that for surface water flooding shown in Fig. 9.14. Again, this value is nearly
constant regardless of the return period and breaching point.

9.4.4 Discussions

As described in Sect. 9.4.2, for surface water flooding, damaged (inundated above
floor level) buildings are quite rare, representing only 3.1% of the total buildings
even in the 1000-year return period. The total inundation volume in the 1000-year
152 Y. Muto and R. Yokokawa

Fig. 9.17 The inundation depth increment due to river flooding onto surface water flooding
(breaching at 1.8 km denoted by red X, breaching level at RMWL)

Fig. 9.18 The relationship between the return period of total precipitation and the area of buildings
inundated above floor level for combined flooding
9 Wetland Paddy Fields as Green Infrastructure Against Flood 153

Fig. 9.19 The locations of buildings inundated above floor level for combined flooding (breaching
at 1.8 km denoted by red X, breaching level at RMWL)

return period is approximately 764 × 103 m3 , and 64% of the volume is allocated
to the paddy fields. Although some damaged paddy fields become apparent, disaster
risk directly affecting mankind is still low. In other words, paddy fields secure
human lives and properties by retaining most of the flooding volume. In the studied
area, as a result, the current land-use classification can make sense in terms of
disaster risk reduction against surface water flooding.
For river flooding combined with surface water flooding, as shown in Sect.
9.4.3, flooding from the river increases the inundated area and depth. However,
the floodwater is primarily deposited in the vicinity of the breaching point, and
in the western and northern paddy fields through the streams, surprisingly a bit far
from the breach point. The studied area again can retain the floodwater volume to
remain human risk relatively low, up to 80- to 100-year return period for combined
flooding. In addition, the area between the Kaifu and Zenzo rivers is mainly used as
paddy fields, directly receiving the floodwater from the river; thus, it can store the
floodwater without increasing the threat to human lives to some extent. This should
also be appreciated as a wise example of land use against floods.
154 Y. Muto and R. Yokokawa

Fig. 9.20 The distributions of damaged paddy fields for combined flooding (breaching at 1.8 km
denoted by red X, breaching level at RMWL)

Under a quite low-frequency event such as a 1000-year return period, damage

to housing areas in the eastern part, mainly on high sand dunes along the beach,
is inevitable. As mentioned above, inundation in the studied area starts from the
southern (near the river), western and northern (paddy fields) parts, then it spreads
from these three directions toward the central part as the precipitation becomes
larger, and finally the relatively high eastern part gets inundated. Here, we can
consider two reference values regarding damages: one is when a building in the
central part near the town hall gets inundated and the other does other buildings
on the higher sand dune. The former is recognised as damage to buildings starts to
expand, and the latter is considered as seriously enlarge damage by inundating the
residential area on high and relatively safer parts. Figure 9.23 shows the calculated
inundation volumes only in the paddy fields plotted as a function of the return
period, together with the locations of these two reference values. Judging from the
simulation results, the former lies within 80- (1.8 km breaching) and 100- (0.4 km
breaching) year return periods, indicated with a blue band in the figure, whereas the
latter lies over a 100-year return period, shown as a green band. By doing this, we
can estimate the potential volume of floodwater that can be retained by the paddy
9 Wetland Paddy Fields as Green Infrastructure Against Flood 155

Fig. 9.21 The ratio of damaged/non-damaged paddy fields for combined flooding (breaching at
1.8 km, breaching level at RMWL)

Fig. 9.22 The estimated inundation volume stored in the paddy fields for combined flooding

fields until damage to buildings and houses starts to occur. From the figure, the
paddy fields in the studied area can retain 590 to 640 × 103 m3 of floodwater
before damage starts to expand in the central part, and of 950 to 1150 × 103 m3
of floodwater before huge damage occurs in the eastern part. The figure can also be
used to explain why damage to buildings is small for surface water flooding.
156 Y. Muto and R. Yokokawa

Fig. 9.23 Two reference values regarding damage to buildings with reference to the estimated
inundation volumes in the paddy fields

Considering the expanding process of floodwater together with the two reference
values described above, it is important to secure the buildings in the central
part, which start to be inundated between two reference levels. If these buildings
are protected in some ways, a critical inundation volume in which damage to
buildings starts can be enlarged. One of traditional methods is levee reinforcement.
Figure 9.24 shows how much levee reinforcement reduces damaged buildings. Here
three additional breaching scenarios, changing the breaching level at (1) HWL,
(2) HWL +0.6 m and (3) at the same breaching level (RMWL) but overflow
starting from the hydrograph peak, were explored for a 1000-year return period.
The third condition is supposed to reinforce the levee in some ways, to avoid
breaching immediately after reaching the setting level and to endure against flow
until the hydrograph peak. The figure indicates that these levee reinforcements are
all effective since they can reduce the total flooding volume. From the viewpoint of
hybrid infrastructure, that is, the best matching of grey and green infrastructures,
such considerations will be useful to seek how much risk the grey should take.
On the other hand, because the levee reinforcement does not change alluvial
plains and their drainage system at all, once flooding exceeds a certain level and
pours into the plains, the same expansion process of flood flow as without levee
reinforcement will be reproduced. Rather, in order to make full use of the flood
control function of paddy fields, it would be better to consider land-use alteration,
including building transference and wetland restoration. By doing this, the potential
9 Wetland Paddy Fields as Green Infrastructure Against Flood 157

Fig. 9.24 The effect of the breaching level on locations of damaged buildings for combined
flooding (1000-year return period, breaching at 1.8 km denoted by red X)

volume of floodwater retention in paddy fields can be extended at least up to the

green band level, which well meets the concept of green infrastructure.

9.5 Concluding Remarks

By performing the inundation simulation, not only the potential storage volume but
the expanding process of inundation in the studied area can be clarified. This also
provides insights into the advantage of current land-use towards flooding, and leads
to present a land-use strategy to be in the future against low-frequency floods.
Our recommendation for future land use is to transfer the buildings from the
vulnerable central part to floods and to turn it back to the wetland. This may sound to
require considerable efforts; however, with reference to a past aerial photo shown in
Fig. 9.25, this part used to be mainly utilised as paddy fields. The potential volume
of floodwater retention in paddy fields in the studied area should presumably be
larger in the past. This can be a good reference for considering land-use strategies.
158 Y. Muto and R. Yokokawa

Fig. 9.25 An aerial photo of the studied area in the mid-1960s

Recently, Watanabe and Muto (2021) also confirmed the advantage of building
transfer from the central part, from the viewpoint of town planning.
We do not take a position criticising the current land use, since it has been
realised as a result of various factors, such as population growth, high economic
growth, technological innovation against floods and so on. These factors sometimes
allowed, and sometimes pushed, to move residential areas onto low and vulnerable
parts. Nevertheless, the tendencies have now changed, particularly regarding the
population. We can now do manage to live outside the vulnerable area. It must be a
good time to consider withdrawing from the risk exposure.
In this study, we do not quantify damage to inundated paddy fields, but we intend
to maximally utilise their flood retention capacity to reduce damage in residential
areas as the most important aspect. However, this concept will make controversy
among the concerned persons. To realise our proposal, not only explaining its
economic advantages but consensus building is crucially important.

References

Iwasa Y, Inoue K (1982) Mathematical simulation of channel and overland flood flows in view of
flood disaster engineering. J Nat Dis Sci 4(1):1–30
Japan Society of Civil Engineers (edited and published) (1999) Formulary of Hydraulics. 713 pp.
(in Japanese)
Kusaka M (2020) Plains and Japanese history. Kadokawa. 209 pp. (in Japanese)
9 Wetland Paddy Fields as Green Infrastructure Against Flood 159

Masumoto T (1998) Paradigm shift in the evaluation of water storage function of paddies and in
watershed management. J Japan Soc Hydrol Water Res 11(7):711–722. (in Japanese)
Ministry of the Environment, Japan (2016) Ecosystem-based disaster risk reduction in Japan. In:
A handbook for practitioners. 24 pp
Miyazu S, Yoshikawa N, Abe S, Misawa S, Yasuda H (2012) Development of application of
evaluation model for interior flood damage mitigation effect of a paddy field dam. IDRE J
282:15–24. (in Japanese)
Muto Y, Imai Y, Kamada M (2020) Ecosystem services of paddy fields and basinwide compre-
hensive flood alleviation. In: Green infrastructure—its practices. Nikkei BP, pp 308–317. (in
Japanese)
NITA Consultant Co., Ltd (n.d.) Introduction to AFREL. http://www.nita.co.jp/index.php/
software/afrel. (in Japanese)
Thang NT, Inoue K, Toda K, Kawaike K (2004) A model for flood inundation analysis in urban
area: verification and application. Ann Disast Prev Res Inst Kyoto Univ 47B:13 pp
Watanabe K, Muto Y (2021) A study on future land use considering flood disaster in Osato district
in Kaiyo Town, Tokushima. In: Proceedings of Awagakkai, no 63, pp 145–153 (in Japanese)
Yoshikawa N, Miyazu S, Koide H, Misawa S, Yasuda H (2010) Evaluation of flood mitigation
effect of a paddy field dam in an unimproved paddy field area. Adv River Eng 16:507–512. (in
Japanese)

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 10
Change in Floodwater Retention
Function of a Paddy Field Due
to Cultivation Abandonment
in a Depopulating Rural Region in Japan

Yota Imai, Yasunori Muto, and Mahito Kamada

Abstract Flood control concept at watershed scale is used as an adaptation strategy

for climate change; rainwater and floodwater should be stored in forests and
agricultural lands in the watershed, rather than relying on only river systems; and
paddy fields are important ecosystems for floodwater retention. The abandonment
of paddy cultivation, however, is increasing year after year in Japan due to depop-
ulation in rural regions. Through two-dimensional numerical inundation analysis,
we evaluated the negative effects of topographical changes after abandonment on
the floodwater retention function. The bipolarization of the inundation depth is
apparent in the current topography. The areas with deepwater inundation are wider,
particularly in the most downstream areas, whereas in the past, areas with shallow
water were widely distributed. The peak water level in the current land use is 0.4 m
higher than in the past. After the peak level is achieved, the water level decreases in
a shorter time in the present land use than in the past. Owing to hydraulic changes
after the abandonment of rice cultivation, the risk of flooding disasters for people
living downstream has increased. To maintain floodwater retention function, one
of the regulating services, by artificial wetlands, such as paddy fields, continuous
management is essential to prevent vegetation succession. Thus, it is important to
provide incentives for local people to perform the daily maintenance activities.

Keywords Abandoned paddy field · Flood management · Rural area ·

Depopulation · Green infrastructure

Y. Imai ()
Department of Civil and Environmental Engineering, Graduate School of Advanced Technology
and Science, Tokushima University, Tokushima, Japan
Department of Civil Engineering, Kobe City College of Technology, Kobe, Hyogo, Japan
Y. Muto · M. Kamada
Research Center for Management of Disaster and Environment, Tokushima University,
Tokushima, Japan
Department of Civil and Environmental Engineering, Graduate School of Technology, Industrial
and Social Sciences, Tokushima University, Tokushima, Japan

© The Author(s) 2022 161

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_10
162 Y. Imai et al.

10.1 Introduction

Owing to climate change, rainfall is increasing every year along with the magnitude
and frequency of flooding (http://www.env.go.jp/earth/tekiou/report2018_full.pdf).
Although floods have been controlled by strengthening river embankments and dam
construction, solving this issue solely by adopting structural measures is becoming
difficult from a financial viewpoint due to changes in the social condition of
Japan: decreasing tax income associated with a rapid decrease in population and
increasing maintenance costs for existing infrastructures (https://www.mlit.go.jp/
sogoseisaku/maintenance/_pdf/research01_02_pdf02.pdf). Therefore, the Ministry
of Land, Infrastructure, Transport and Tourism steered toward “Ryuiki-chisui”
which refers to a new flood control concept at the watershed scale as an adaptation
strategy for climate change (https://www.mlit.go.jp/river/kasen/ryuiki_pro/index.
html): the rainwater and floodwater should be stored in forests and agricultural lands
in the watershed, rather than relying on only river systems.
The use of paddy fields to reduce flood damage has been proposed as one
of the flood control methods at the watershed scale (https://www.mlit.go.jp/river/
kasen/suisin/pdf/renkei_siryou02/siryou04.pdf). Paddy fields are a land use for rice
production in Japan and provide several ecosystem functions (Natuhara 2013). For
example, paddy fields provide a water storage function as well as a habitat for
several wetland species (Katayama et al. 2015; Osawa et al. 2020), and the function
has been utilized for flood disaster risk reductions (Teramura and Shimatani 2021).
The integrated management of paddies, ditches, and ridges by farmers has
contributed to maintaining the water storage function of paddy fields (Yoshikawa
2014). Previous studies have been using various approaches to evaluate the water
storage function of paddy fields in different regions (Matsumoto et al. 2013;
Yoshikawa 2014; Muto et al. 2018; Teramura and Shimatani 2021).
In Japan, the abandonment of paddy cultivation is increasing every year due to
depopulation (Osawa et al. 2013, 2016); this may have a negative effect on water
storage functions due to the collapse of the ditches-ridges structure (Yoshida et
al. 2013). Although several studies have evaluated changes in runoff volume from
paddy fields after abandonment of cultivation, few have described the hydraulic
characteristics of abandoned paddy fields. Hydraulic characteristics can be linked
to sediment deposition, followed by the development of woody communities
through vegetation succession. If the hydraulic characteristics of abandoned paddy
fields were described, a technique which control floodwater may be developed for
improvement of water storage functions. Such a technique can only be realized with
a lower cost of introduction and maintenance, particularly in a depopulated rural
area.
On the other hand, abandoned paddy fields have often restored to wetland/biotope
and are maintained as a wildlife habitat. In wetland restoration, construction of
ridges and removing shrub are conducted for keeping topographical condition flat,
for retention of floodwater and rainwater. Thus, the water storage function is likely
to be improved at restored abandoned paddy fields.
10 Change in Floodwater Retention Function of a Paddy Field Due. . . 163

In this paper, first, we clarify the negative effects of the abandonment of

rice cultivation on water storage function in Shitaru District, Tsushima Island in
Nagasaki Prefecture, Japan. The hydraulic characteristics of paddy fields before and
after abandonment are described, compared, and discussed in terms of the process of
functional decline. Second, a method of restoring the function is discussed based on
the case of Tai District, Toyooka City in Hyogo Prefecture, where abandoned paddy
fields have been used as wildlife habitat wetlands (Imai et al. 2020). Finally, we
referred to the changes in water storage function with structural change and wetland
restoration at paddy fields and to the management of abandoned paddy fields based
on the comparison of two districts.

10.2 Material and Methods

10.2.1 Study Area

The study was conducted in Shitaru District in Kamiagata, Tsushima Island, in

Nagasaki Prefecture (34◦34 N, 129◦18 E; Fig. 10.1). The watershed comprises
forests, paddy fields, and villages (Fig. 10.2b). The highest monthly average
temperature is 26 ◦ C in August, while the lowest is 6 ◦ C in January. The monthly
average precipitation is the highest in July (350 mm), followed by June (320 mm;
(http://www.city.tsushima.nagasaki.jp/deta/post.html). The population of Shitaru
District in 2020 was 73, and is estimated to decrease to 21 by 2050.
The area used to be under paddy cultivation until the 1960s (Shigehara and
Shibata 2018); however, almost all paddies have been abandoned and transformed
to wetland vegetation (Fig. 10.2c). The rainwater in the watershed spreads across
the floodplain and eventually accumulates downstream (Fig. 10.2d).
Figure 10.3a shows the past distribution of paddy fields in the watershed.
Sediments that flowed into the channels and paddy fields were removed by local
farmers to maintain cultivation after flooding (Imai et al. 2019). Figure 10.3b
shows the current vegetation map. Almost all paddy fields in the watershed have
transformed to natural vegetation. The vegetation was classified into four types
of plant communities. According to an inundation analysis, the spatial patterns of
plant communities corresponded to the velocity of water and inundated depth during
flooding (Imai et al. 2019).

10.2.2 Two-Dimensional Inundation Analysis

To obtain the current topographical conditions, aerial and field surveys were
conducted from December 5 to 8, 2017. For aerial surveys, unmanned aerial vehicles
(UAVs; DJI Phantom3 Pro and Phantom4 Pro) were used. To minimize errors in
164 Y. Imai et al.

Fig. 10.1 (a) Location of Tsushima Island, Japan. (b) Map of Tsushima Island. (c) The study
floodplain (hatched area in c), in Shitaru District, Tsushima Island, Japan. The black borders shown
in (c) are sub-watersheds (Imai et al. 2019)

the digital terrain model, flight routes were planned with 80% overlap and 60%
sidelap. A digital surface model (DSM) with a 0.8 m grid resolution was generated
from 3447 images using the Agisoft Photoscan Pro v1.0.0 and ArcGIS v10.2.2.
The parameters of Align Photos and Build Dense Cloud, structure from motion-
multi view stereo (SfM-MVS) photogrammetry using Agisoft Photoscan Pro v1.
0. 0 were high, respectively. Errors in position coordinates were corrected using
the coordinates from 19 ground control points (GCPs: Fig. 10.4), which were set
in the watershed and measured using RTK-GNSS (TOPCON GR-2100) and Total
Station. Because almost all areas of the watershed were covered by vegetation, the
representative height was measured by using a handy measure for each vegetation
type by field surveys on December 9, 2017; then, the digital terrain model (DTM)
was obtained by subtracting the vegetation heights from the DSM.
10 Change in Floodwater Retention Function of a Paddy Field Due. . . 165

Fig. 10.2 Shitaru District. (a) The study area, (b) distribution of village and floodplain, (c) woody
area, and (d) downstream area

When rice was cultivated, the farmers also maintained the ridges, and therefore,
rainwater and floodwater were stored in every paddy field before overflowing
into the lower fields in a step-by-step manner. Thus, the past topography was
reconstructed as follows: the mean current elevation in each polygon of the paddy
field was calculated from the DTM using zonal statistics (ArcGIS 10.2.2, Spatial
Analyst). Then, 0.3 m, representative height of ridges remained at study floodplain
was added to the elevation of cells surrounding each paddy field to include the height
of the ridge.
Differences in hydraulic characteristics before and after abandonment of culti-
vation were clarified by two-dimensional inundation analyses in conditions with
annual maximum rainfall for past and current topography using iRIC Nays2DFlood
Solver (https://i-ric.org/en/solvers/nays2dflood/). Nays2DFlood Solver enables the
user to easily set the inflow conditions and topographical conditions for flood flow
analysis and obtains hydraulic characteristics such as water velocity and depth at
river and floodplain.
166 Y. Imai et al.

Fig. 10.3 (a) Past distribution of paddy fields and (b) current distribution of plant communities
Abbreviations: Aim-Ic, Artemisia indica var. maximowiczii – Imperata cylindrica community;
Bp, bamboo plantation; Bnn, Boehmeria nivea var. nipononivea community; Cd, Carex dis-
palata community; Dhs, dried herbaceous community (short); Jed, Juncus effusus var. decipiens
community; Ms., Miscanthus sacchariflorus community; Msi, Miscanthus sinensis community;
Oj-Pa, Oenanthe javanica – Phalaris arundinacea community; Ow, open water; Pf, paddy field;
Pa, Phragmites australis community; Ps, pioneer shrub community; Pt, Polygonum thunbergii
community; Pu, Pueraria montana community; Sps, Salix pierotii community (short); Spt, Salix
pierotii community (tall); Ta, Typha angustifolia community; Ups, Ulmus parvifolia community
(short); Upt, Ulmus parvifolia community (tall); Uf, upland field; Zl, Zizania latifolia community
(Imai et al. 2019)

For inundation analyses, the model hyetograph was generated based on records
from the nearest observation points (34◦33 N, 129◦ 18 E; http://www.kasen-sabo.
pref.nagasaki.jp/nagasaki/main/index.php) on Tsushima Island: rainfall of 14 h,
from 12:00 on June 30 to 2:00 on July 1, 2018. The total rainfall was 73 mm, and
10 Change in Floodwater Retention Function of a Paddy Field Due. . . 167

Fig. 10.4 Flight routes of

UAV (yellow circles) were
planned for 80% overlap and
60% sidelap and 19 GCPs
were set (white circles). Black
line shows study floodplain

it was estimated to be 1/1-year probability based on 20-year records. The model

hyetograph was produced by uniformly enlarging the annual maximum rainfall
of 342 mm (Fig. 10.5). For the model hydrograph, discharge flowing into the
target watershed from the northern and northwestern sub-watersheds was considered
(Fig. 10.5).
168 Y. Imai et al.

Fig. 10.5 Model hyetograph and hydrograph. Main inflow and sub-inflow show inflows from the
northern and north-western sub-watersheds (Fig. 10.1(c)), respectively

10.3 Results

Figure 10.6 shows the hydraulic characteristics in different topographies—in the

present (Fig. 10.6a) and before the abandonment of cultivation (Fig. 10.6b). The
bipolarization of inundation depth is apparent in the current topography; areas
without inundation and with deepwater depths (red) are wider, particularly in the
most downstream area, whereas in the past, areas with shallow water (blue) were
widely distributed. These results indicate that floodwater now accumulates in the
most downstream area connecting to the residential area.
Figure 10.7 shows the water level at the most downstream point (see Fig. 10.2d).
The peak water level in the current land use is 0.4 m higher than in the past. After
the peak level is achieved, the water level decreases in a shorter time in the present
land use than in the past. These results indicate that the water storage function has
been decreased by the abandonment of cultivation.

10.4 Discussion

The peak water level at the downstream in the current land use is 0.4 m higher than
in the past topographical condition (Fig. 10.7). This result shows that the hydraulic
characteristics have changed, and the risk of flooding has been increased through
abandonment of paddy cultivation.
The abandonment of cultivation leads to halting of farmer activities to maintain
paddy field structures, and ridges collapse due to erosion (Yoshikawa et al. 2009).
10 Change in Floodwater Retention Function of a Paddy Field Due. . . 169

Fig. 10.6 Depth of inundation immediately after peak discharge for maximum rainfall: (a) present
and (b) past land use

Fig. 10.7 Water level change at downstream area

170 Y. Imai et al.

Once the ridges collapse, a water channel is formed as a natural process (Yoshida
et al. 2012, 2013), and rainwater and floodwater flow into the channel as runoff.
Thus, the water storage function is reduced, and the moisture condition of the area
changes (Ohkuro et al. 1996).
Because of hydraulic changes after the abandonment of rice cultivation, natural
vegetation has established (Imai et al. 2019), but the risk of flooding disasters
for people living in downstream areas has increased. The same situation has
occurred in many rural regions in Japan undergoing depopulation. It is necessary
to adopt appropriate measures against flood disasters in these regions, where little
administrative support can be obtained.

10.5 Mitigating Flood Risk Using Abandoned Paddy Field

Through Wetland Restoration

At Tai District in Toyooka City, Hyogo Prefecture, Japan (35◦38 37.9N

134◦50 41.5E; Figs. 10.7 and 10.8), where the population is estimated to decrease
from 136 in 2020 to 73 by 2050, abandoned paddy fields have restored to

Fig. 10.8 Tai District. (a) Study site, (b) distribution of village and floodplain, and (c, d) weirs
and flashboards maintained by local people and volunteers for wetland improvement
10 Change in Floodwater Retention Function of a Paddy Field Due. . . 171

wetland/biotope and are maintained as feeding sites for the Oriental white stork
(Ciconia boyciana). This is an endangered bird protected by the government and
local people, through collaborative activities by local people and volunteers from
outside the district.
The wetlands have been used as a floodwater retention area. These structures
reduce the flow velocity during flooding. In addition to voluntary work, overflow
levees and small water reservoirs were constructed in the uppermost stream of the
wetland with the support of Toyooka City.
The results of two-dimensional flood flow analysis for 1/30-year rainfall showed
that 22% of peak discharge was stored in abandoned paddy fields (Fig. 10.9; Imai
et al. 2020) and 25% of the total discharge was stored in the wetland (Fig. 10.10;
Imai et al. 2020). Thus, managed abandoned paddy fields can be used as floodwater
retention areas, and water storage function can be improved by wetland restoration
for flood disaster reduction.

Fig. 10.9 Temporal changes

in discharge (Imai et al.
2020). Outflow shows the
discharge at the downstream
boundary and inflow shows
the discharge at the upstream
boundary

Fig. 10.10 Comparison of

floodwater volume between
the calculated area (total) and
abandoned paddy fields on its
area during flooding (Imai et
al. 2020)
172 Y. Imai et al.

10.6 Changes in Water Storage Function with Structural

Change and Their Management at Abandoned Paddy
Fields

Paddy fields provide provisioning services for rice cultivation as well as regulating
services via water storage functions through maintenance of ditches and ridges
structure. However, the water storage function of paddy fields is decreased by
structural changes after abandonment of cultivation; the quality of regulating
services worsens after relinquishing provisioning services, as in the case of Shitaru.
In the case of Tai, the water storage function was provided at the restored abandoned
paddy fields with weirs and flashboards introduced by the local people due to habitat
improvement for supporting Oriental white storks. Thus, the water storage function
decreases from cultivated paddy fields to abandoned paddy fields, but its function
of abandoned paddy fields could be improved through wetland restoration.
At the Tai District, supporting Oriental white storks was an incentive for local
people and volunteers from outside the district, rather than flood control. To
maintain regulating services of artificial wetlands, such as abandoned paddy fields,
continuous management to prevent vegetation succession is essential. Hence, it
is important to provide monetary compensation and to emerge incentives such as
recovery of biodiversity for local people to contribute to this management (Satake
2014). A technique that emerges both water storage function and biodiversity
conservation should be developed with adaptive management approach.

Acknowledgments We are grateful to the local people of the Shitaru and Tai districts, who
allowed us to carry out our survey and provided useful information in this regard. This study was
supported by the Environment Research and Technology Development Fund (4-1504, 4-1805) of
the Ministry of the Environment, Japan.

References

Imai Y, Shigehara N, Shibata S, Muto Y, Kamada M (2019) Spatial pattern of plant communities
in relation to hydraulic conditions at wetland emerged at abandoned paddy fields. In: E-
proceedings of the 38th IAHR World Congress, 1–6 September, 2019, Panama City, Panama,
pp 3311–3320
Imai Y, Mitsuhashi H, Kamada M, Muto Y (2020) Fundamental study on water storage function
of abandoned paddy field on small river floodplain. Annu J Hydraulic Eng JSCE 65:793–798;
[In Japanese with English Abstract]
Katayama N, Baba YG, Kusumoto Y, Tanaka K (2015) A review of post-war changes in rice
farming and biodiversity in Japan. Agric Syst 132:73–84
Matsumoto T, Fukuoka S, Sumi T (2013) Estimation of flood storage volume in the three retarding
basins along the tone river. Annu J Hydraulic Eng JSCE 69(4):793–798; [In Japanese with
English Abstract]
10 Change in Floodwater Retention Function of a Paddy Field Due. . . 173

Muto Y, Kotani S, Miyoshi M, Kamada M, Tamura T (2018) Retarding capacity change of wetland
paddy fields due to house land development -utilizing paddy fields as green infrastructure
against flood. In: Proceedings of the 21th IAHR-APD congress 2018, 8pp
Natuhara Y (2013) Ecosystem services by paddy fields as substitutes of natural wetlands in Japan.
Ecol Eng 56:97–106
Ohkuro T, Matsuo K, Nemoto M (1996) Vegetation dynamics of abandoned paddy fields and their
levee slopes in mountainous regions of Central Japan. Jpn J Ecol 46:245–256; [In Japanese
with English Abstract]
Osawa T, Kohyama K, Mitsuhashi H (2013) Areas of increasing agricultural abandonment
overlap the distribution of previously common, currently threatened plant species. PLoS One
8(11):e79978
Osawa T, Kohyama K, Mitsuhashi H (2016) Multiple factors drive regional agricultural abandon-
ment. Sci Total Environ 542:478–483
Osawa T, Nishida T, Oka T (2020) Paddy fields located in water storage zones could take over the
wetland plant community. Sci Rep 10(1):1–8
Satake S (2014) Habitat creation and restoration in collaboration with local residents. Reintroduc-
tion 3:25–27; [In Japanese with English Abstract]
Shigehara N, Shibata S (2018) The transition of biological resources use and land use in the
1950s and later in Shitaru, Tsushima city, Nagasaki prefecture. Jpn Inst Landscape Architect
81(5):699–702; [In Japanese with English Abstract]
Teramura J, Shimatani Y (2021) Advantages of the open levee (Kasumi-Tei), a traditional Japanese
river technology on the Matsuura river, from an ecosystem-based disaster risk reduction
perspective. Water 13(4):480
Yoshida T, Masumoto T, Horikawa N (2012) Changes in rainfall-runoff characteristics of small
watersheds due to cultivation conditions of hilly paddies. Irrigation Drainage Rural Eng J
80(2):117–124; [In Japanese with English Abstract]
Yoshida T, Masumoto T, Horikawa N, Minakawa H (2013) Modelling of short-term runoff pro-
cesses in watersheds dominated by hilly paddies. Irrigation Drainage Rural Eng J 81(3):235–
244; [In Japanese with English Abstract]
Yoshikawa N (2014) Can paddy fields mitigate flood disaster? Possible use and technical aspects of
the paddy field dam. In: Social-ecological restoration in Paddy-dominated landscapes. Springer,
Tokyo, pp 197–207
Yoshikawa N, Nagao N, Misawa S (2009) Watershed scale evaluation of flood mitigation function
of paddy fields installed with runoff control devices. Irrigation Drainage Rural Eng J 261(3):44–
48; [In Japanese with English Abstract]

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 11
Paddy Fields as Green Infrastructure:
Their Ecosystem Services
and Threatening Drivers

Takeshi Osawa, Takaaki Nishida, and Takashi Oka

Abstract The paddy field is a typical ecosystem in monsoon Asia. The main
purpose of a paddy field is food production (i.e., a provisioning ecosystem service),
but it also has several other ecosystem services. Therefore, we can consider paddy
fields as components of green infrastructure (GI) that are already widely distributed.
To maintain and improve the value of paddy fields as GI, we should strive to
maintain and improve other ecosystem services besides provisioning services.
However, issues that reduce the value of paddy fields as GI remain, and these are
related to the consolidation and abandonment of paddy fields in recent years. In
this paper, we focus on two ecosystem services of paddy fields, other than the
provisioning service: (1) the habitat service and (2) the regulating service. Both
these services are relatively well-evaluated ecosystem services of paddy fields in
Japan. We discuss the effects of both consolidation and abandonment on these
ecosystem services as well as their corresponding countermeasures. Based on this,
we also discuss the challenges in applying and expanding the idea of paddy fields
as GI.

Keywords Agricultural abandonment · Consolidation · Ecosystem-based

disaster risk reduction · Ecosystem service · Habitat service · Wetlands

T. Osawa ()
Graduate school of Urban environmental Sciences, Tokyo Metropolitan University, Tokyo, Japan
T. Nishida
Mitsubishi UFJ Research and Consulting Co. Ltd., Osaka, Japan
Kyoto Sangyo University, Kyoto, Japan
T. Oka
Kyoto Sangyo University, Kyoto, Japan

© The Author(s) 2022 175

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_11
176 T. Osawa et al.

11.1 Introduction

Paddy fields are a typical agricultural land type in monsoon Asia, mainly for rice
crops. In Japan, paddy fields occupy much of the country’s coastal plain (Natuhara
2013), with an area of approximately 24,000 km2 , which represents 6% of the total
land area (Ministry of Agriculture, Forestry and Fishery [MAFF], Japan; https://
www.maff.go.jp/j/tokei/sihyo/data/10.html; accessed on March 10, 2021). Paddy
fields are the second largest land cover type in Japan after forests and are distributed
throughout the country (Osawa et al. 2013, 2016a, b). Thus, paddy fields represent
one of the most common seminatural ecosystems in Japan.
The essential purpose of maintaining paddy fields is for agricultural food
production, i.e., for a provisioning ecosystem service. It is well known that paddy
fields can provide several ecosystem services other than their provisioning service.
For instance, paddy fields provide a supporting (habitat) service for several wetland
species (Yoon 2009; Natuhara 2013; Katayama et al. 2015) in addition to providing
a regulating service, such as flood control (Matsuno et al. 2006; Natuhara 2013;
Osawa et al. 2020a), greenhouse gas emission control (Ahmad et al. 2009; Yang
et al. 2012), landslide prevention (Natuhara 2013), and decreasing pests (insects)
(Jincai et al. 1994). Japan’s MAFF has claimed that paddy fields provide not only
food production, i.e., a provisioning service, but also diverse multiple services,
referred to as “multiple functions of agricultural lands.” Thus, the MAFF has
supported some of these functions through a grant system (https://www.maff.go.
jp/j/nousin/kanri/tamen_siharai.html; accessed on March 10, 2021).
Green infrastructure (GI) has been defined as natural, seminatural, and artificial
networks of multiple functional ecological systems within, around, and between
human residential communities at all spatial scales (Tzoulas et al. 2007). At the
same time, GI should have multiple ecosystem services based on its ecosystem
functions (Sandstrom 2002; Tzoulas et al. 2007). Thus, based on these definitions,
we can also consider paddy fields as GI (Fig. 11.1). However, in recent years, the
situations of many agricultural lands, including paddy fields, have changed. Changes
in agricultural land mainly consist of either modernization or abandonment (Osawa
et al. 2013; Katayama et al. 2015). Both these changes in land use can exert negative
effects on the ecosystem functions of the land (Benayas et al. 2007; Yoshida et al.
2012; Osawa et al. 2013). Consequently, such changes in paddy fields might cause
their inherent ecosystem functions and services to collapse (Osawa et al. 2020a, b).
The concept of GI, particularly in monsoon Asia (including Japan), should
be expanded. Paddy fields represent one of the most important ecosystems with
potential as GI because they are very widespread and could thus offer multiple
ecosystem services. Therefore, it is important to understand the threats and current
status of multiple ecosystem services of paddy fields for sustainable use. In this
paper, we focus on two ecosystem services—habitat and regulating services—of
paddy fields that have been relatively well-evaluated in Japan recently and discuss
their current status and threats as well as the countermeasures against these threats.
11 Paddy Fields as GI 177

Green Infrastructure
Ecosystem services A

Ecosystem services B

Ecosystem services C

Green Infrastructure should have multiple ecosystem services

Paddy field
Provisioning services
(Food production)

Regulating services
(Flood control)

Habitat services
(Wetland habitat)

Paddy fields have several ecosystem services other than food production

Fig. 11.1 The relationship between the definition of green infrastructure (GI) and paddy fields

Then, we discuss the challenges in applying and expanding the idea of paddy fields
as GI.

11.2 Habitat Service: Biodiversity Conservation in Paddy

Fields

The idea that paddy fields can provide habitats for several wetland species is already
well-established (Toral et al. 2012; Katayama et al. 2015; Osawa 2017; Osawa et al.
2020b). The review of the biota of Japanese paddy fields by Katayama et al. (2015)
revealed that paddy fields can harbor several small aquatic organisms (Yamazaki
et al. 2001), at least 184 weed (i.e., non-crop) species, 27 aquatic invertebrate
species (not including rice pests), 15 species and 2 subspecies of amphibians, 11
178 T. Osawa et al.

fish species, and 49 bird species, even considering underestimation due to data
availability (Katayama et al. 2015). The value of paddy fields as habitats for several
wetland species and in contributing to regional biodiversity is clear. However, in
recent years, a drastic decline in species has been observed in Japanese paddy fields
(Katayama et al. 2015; Osawa 2017). This trend indicates that paddy fields have
been recently losing their ecosystem services, namely, their habitat service. Losing
an ecosystem service is linked to losing GI value. Thus, maintaining the habitat
service is an important issue when considering paddy fields as GI.
Agricultural modernization is one of the most severe drivers of habitat degra-
dation in paddy fields (Katayama et al. 2015; Osawa 2017). After World War II,
Japanese agriculture increased both chemical usage and land consolidation, both
related to mechanization (Katayama et al. 2015; Osawa 2017). These practices have
been clearly successful in increasing the effectiveness of agricultural activities such
as time efficiency and yield per area (Katayama et al. 2015). In other words, this
strategy specifically improved the provisioning service of paddy fields. However,
there is often a trade-off between such agricultural modernization and biodiversity
in paddy fields (Katayama et al. 2015; Osawa et al. 2016b; Osawa 2017; Osawa
and Mitsuhashi 2017). In particular, the effect of land consolidation, which changes
the ecosystem structure from a wetland habitat to a dryland habitat, is severe
because paddy fields act as alternative wetland habitats (Osawa et al. 2016b, 2020b).
Wetlands are one of the habitats that are on the decline worldwide (Denny 1994;
Lougheed et al. 2008). This is one of the reasons that could affect the value of
paddy fields as GI. However, paddy fields that have experienced land consolidation
provide little or no habitat service for wetland species (Osawa et al. 2016b, 2020b).
Thus, because such paddy fields lack multiple ecosystem services, their GI value
might be relatively low.
Agricultural modernization can be accompanied by agricultural abandonment,
mainly in areas that are unsuitable for modernization, such as hilly and mountainous
areas (Katayama et al. 2015). Agricultural abandonment signifies losing the provi-
sioning service in that area, thus reducing the value of paddy fields as GI. Moreover,
previous studies have suggested that agricultural abandonment could degrade the
habitat service of paddy fields (Osawa et al. 2013; Uchida and Ushimaru 2014).
Abandoned paddy fields have relatively little value as GI because they do not have
any ecosystem services, i.e., at least provisioning and/or habitat services. Therefore,
changes in agricultural land use, either through consolidation or abandonment,
might reduce the land’s value as GI. In particular, these could reduce the variety
of the ecosystem service, because abandonment is a severe negative driver that
causes the loss of not only habitat services but also provisioning services, which
are essential ecosystem services in paddy fields.
11 Paddy Fields as GI 179

11.3 Regulating Service: Disaster Risk Reduction (DRR)

by Paddy Fields

In recent years, interest from both researchers and practitioners has grown in ecosys-
tem services that reduce the frequency and severity of natural disasters (Dudley et al.
2015; Renaud et al. 2016; Ministry of Environment, Japan; https://www.env.go.jp/
nature/biodic/eco-drr.html; accessed on April 28, 2021). These types of ecosystem
services are often known as Ecosystem-based DRR (Eco-DRR) (Martin and Watson
2016; Furuta and Shimatani 2017; Scarano 2017; Nakamura et al. 2019; Osawa et
al. 2020a). The use of GI for Eco-DRR has gained much research interest because of
the low costs of introduction and maintenance and its provision of other ecosystem
services (Sudmeier-Rieux et al. 2013; Furuta and Shimatani 2017; Onuma and
Tsuge 2018; Osawa et al. 2020a). Paddy fields have been expected to function
mainly for flood control, i.e., by increasing the water storage capacity of river basins,
reducing the peak flow of rivers, and increasing groundwater recharge (Matsuno et
al. 2006), all of which can help mitigate the risk of flood disaster. As an ecosystem
service, Eco-DRR is a regulating service (Bennett et al. 2009), and thus we should
also consider Eco-DRR to be relevant when considering paddy fields as GI.
Recently, studies on Eco-DRR for flooding in paddy fields have progressed. For
instance, we have shown that paddy fields located in water storage zones, i.e., paddy
fields that are located in previous wetland habitats, are very effective in preventing
flooding (Osawa et al. 2020a). A simulation study revealed that paddy fields with
runoff control devices installed in their drainage boxes effectively draw down flood
by the storing rainwater (Yoshikawa et al. 2009, 2010). Thus, there is some evidence
that paddy fields could have Eco-DRR potential against floods.
However, a previous study indicated that abandoned paddy fields could reduce
the Eco-DRR function against flooding, both due to decreased penetration capability
and water runoff from degraded ridges (Yoshida et al. 2012). Thus, abandonment
could result in the loss of not only the provisioning service but also the regulating
service of a paddy field. Moreover, as we have already shown in the previous
section, abandonment could also result in loss of the habitat service (Osawa et
al. 2013; Uchida and Ushimaru 2014). Conversely, to the best of our knowledge,
there are no studies that indicate a negative effect of agricultural modernization, i.e.,
consolidation, on Eco-DRR. Consolidation might actually have a positive effect on
Eco-DRR because concrete ridges could prevent water runoff from paddy fields,
which could increase the capacity of water storage during heavy rain (Yoshikawa
et al. 2010). Although both abandonment and consolidation are major changes in
agricultural land use, their effects on paddy fields (in relation to GI) differ.
180 T. Osawa et al.

11.4 A Strategy to Use and Maintain Paddy Fields as GI

GI is expected to provide multiple ecosystem services (Sandstrom 2002; Tzoulas

et al. 2007). Therefore, to use an ecosystem as GI, land managers and decision
makers should focus on certain expected ecosystem services and then make the
appropriate efforts to maintain and enhance these services. However, managing the
target ecosystem services can be difficult, as ecosystem services can be related
to each other (Bennett et al. 2009). Moreover, Bennett et al. (2009) reported that
some ecosystem services could be positively or negatively influenced by one driver:
human activity (Fig. 11.2). Thus, if land managers and decision makers put some
effort into maintaining and enhancing only one ecosystem service, it might result
in a reduction in other ecosystem services. Agricultural modernization of paddy
fields is a typical example of this because it improves the provisioning service but
reduces the habitat service (Osawa et al. 2016b, 2020b). Thus, if land managers and
decision makers expect the habitat service from paddy fields when they use them
as GI, it will be difficult to achieve the high yield and work efficiency provided by
modernization. Conversely, previous studies indicate that organic farming, which
is an environmentally friendly farming system, could contribute to the biodiversity
of several taxa (Fuller et al. 2005; Hole et al. 2005; Katayama et al. 2019). Thus,
organic farming could even be compatible with both provisioning and habitat
services that have relatively low yields or low work efficiencies compared to
fully modernized agricultural areas. This suggests that certain related ecosystem
services that have a trade-off relationship could be compatible at moderate levels.

Fig. 11.2 Positive and negative drivers for ecosystem services in paddy fields
11 Paddy Fields as GI 181

To achieve multiple ecosystem services from a target ecosystem to be considered as

GI, land managers and decision makers should carefully manage these ecosystem
services by considering the relationship between focused ecosystem services and
the management method, i.e., the driver for ecosystem services.
However, abandoned paddy fields have already lost their essential role, namely,
their provisioning service. Thus, abandoned paddy fields cannot act as agricultural
GI land that should provide multiple ecosystem services including the provisioning
service. In addition, we believe that abandonment not only results in the loss of the
provisioning service but also reduces the regulating and habitat services (Yoshida
et al. 2012; Osawa et al. 2013). Nevertheless, recent studies suggested that some
abandoned farmlands, including paddy fields, can harbor the biodiversity of several
taxa, i.e., provide habitat services (Morimoto et al. 2017; Yamanaka et al. 2017;
Hanioka et al. 2018a, b), depending on the location, condition, and historical land
use of that area (Osawa and Mitsuhashi 2017). As habitat services (i.e., supporting
services) are prerequisites for all ecosystem services (Sarukhán et al. 2005), we
could expect that areas with high habitat services could provide several ecosystem
services. Therefore, for the use of abandoned paddy fields as GI, land managers and
decision makers should change their strategy to focus on ecosystem services other
than provisioning services, such as using areas like non-farmland. If an abandoned
paddy has a high level of biodiversity, converting it into a conservation area could
represent a practical option.

11.5 Conclusion and Perspective

In this paper, we have focused on two ecosystem services and have discussed the
potential of paddy fields as GI as well as changing land uses as a driver threatening
this potential. The most important point in this paper is that the significance of
a paddy field as GI might be different from that of a paddy field as a specific
agricultural area, i.e., an intense food production zone. As paddy fields that represent
GI should have multiple ecosystem services, concentrating on the provisioning
service alone through modernization is not an effective way to promote GI. If
land managers and decision makers want to apply the idea of GI to paddy fields,
they should adopt a different strategy that focuses on combining paddy fields
with provisioning services and other ecosystem services. Traditional agricultural
promotion basically concentrates on food production; however, this does not suit a
strategy that considers a paddy field as GI. Moreover, many countries, including
Japan, are currently experiencing a phase of population decline (Osawa 2017;
https://population.un.org/wpp/; accessed on April 29, 2021). Given this situation,
a strategy that concentrates on food production might not be an effective strategy
for maintaining paddy fields because of decreasing food demands. Other than the
provisioning service (i.e., food production), we assessed two ecosystem services,
providing wetland habitat and Eco-DRR, that are appropriate for paddy fields as
GI; however, it remains unclear which service could coexist with the provisioning
182 T. Osawa et al.

service. If a trade-off exists between food production and other services, not
focusing on food production might be one option for land managers and decision
makers to promote paddy fields as GI, considering the situation of population
decline. Fortunately, previous studies have suggested that flood control (i.e., Eco-
DRR) and food production could coexist (Kramer et al. 1997; Bennett et al. 2009),
whereas there might be a trade-off between the habitat service and agricultural
modernization, which contributes to food production (Katayama et al. 2015; Osawa
et al. 2016b; Osawa and Mitsuhashi 2017) (Fig. 11.2). To balance food production
and the habitat service, these ecosystem services should be maintained at a moderate
level.
Nevertheless, one driver could enhance multiple ecosystem services (Bennett
et al. 2009). Thus, focusing on compatible ecosystem services could be another
effective option to promote GI. Flood control and food production is one potential
combination. However, in our opinion, enhancing human activities such as consoli-
dation is not a smart way to expand on the concept of GI because ecosystem services
should depend on inherent ecosystem functions. Ideas of ecological intensification
already exist and are defined as realizing the intensive and smart use of ecosystem
functions for multiple ecological services (Bommarco et al. 2013; Tittonell 2014).
This should represent a nature-based solution, which is defined as a transition
toward ecosystem services with decreased input of nonrenewable natural capital
and increased investment in renewable natural processes (Maes and Jacobs 2017).
Ecosystem- or nature-based approaches should be one of the most important ideas
to be applied to the idea of GI.
In addition, we should emphasize that abandonment is a driver of the loss
of essential ecosystem services, namely, the provisioning service in paddy fields
(Fig. 11.2). Therefore, land managers and decision makers should strive to pre-
vent agricultural abandonment when they want to promote paddy fields as GI.
However, when abandonment does occur, land managers and decision makers have
different options, including recovering the paddy field (i.e., provisioning service) or
recovering its natural ecosystem. Regardless of the option selected for restoration,
efforts should be made to support the recovery of any ecosystem service (Osawa
et al. 2016b). In abandoned paddy fields, food production is probably not the most
important ecosystem service. Thus, returning the paddy fields to an intact ecosystem
is one of the options. Given the importance of GI, human society should invest in
maintaining such infrastructure. Moreover, promoting GI is a cost-effective strategy
because of its multiple benefits to people. Overall, paddy fields have good potential
as GI because of their multiple ecosystem services and widespread distribution.

Acknowledgments This work was supported by the Environment Research and Technology
Development Fund (4-1805) of the Ministry of the Environment, Japan, and partially supported
by the JSPS KAKENHI Grant Number 20K06096. Drs. F. Nakamura and N. Katayama provide us
many useful suggestions.
11 Paddy Fields as GI 183

References

Ahmad S, Li C, Dai G et al (2009) Greenhouse gas emission from direct seeding paddy field under
different rice tillage systems in Central China. Soil Tillage Res 106:54–61
Benayas JMR, Martins A, Nicolau JM, Schulz JJ (2007) Abandonment of agricultural land: an
overview of drivers and consequences. CAB Rev Perspect Agric Vet Sci Nutr Nat Resour 2:1–
14
Bennett EM, Peterson GD, Gordon LJ (2009) Understanding relationships among multiple
ecosystem services. Ecol Lett 12:1394–1404
Bommarco R, Kleijn D, Potts SG (2013) Ecological intensification: harnessing ecosystem services
for food security. Trends Ecol Evol 28:230–238. https://doi.org/10.1016/j.tree.2012.10.012
Denny P (1994) Biodiversity and wetlands. Wetl Ecol Manag 3:55–611. https://doi.org/10.1007/
BF00177296
Dudley N, Buyck C, Furuta N et al (2015) Protected areas as tools for disaster risk reduction. A
handbook for practitioners. IUCN, Gland
Fuller RJ, Norton LR, Feber RE et al (2005) Benefits of organic farming to biodiversity vary among
taxa. Biol Lett 1:431–434
Furuta N, Shimatani Y (2017) Integrating ecological perspectives into engineering practices -
perspectives and lessons from Japan. Int J Disaster Risk Reduct 32:87–94. https://doi.org/
10.1016/j.ijdrr.2017.12.003
Hanioka M, Yamaura Y, Senzaki M et al (2018a) Assessing the landscape-dependent restoration
potential of abandoned farmland using a hierarchical model of bird communities. Agric Ecosyst
Environ 265:217–225. https://doi.org/10.1016/j.agee.2018.06.014
Hanioka M, Yamaura Y, Yamanaka S et al (2018b) How much abandoned farmland is required
to harbor comparable species richness and abundance of bird communities in wetland?
Hierarchical community model suggests the importance of habitat structure and landscape
context. Biodivers Conserv 27:1831–1848. https://doi.org/10.1007/s10531-018-1510-5
Hole DG, Perkins AJ, Wilson JD et al (2005) Does organic farming benefit biodiversity? Biol
Conserv 122:113–130
Jincai W, Guowen H, Jian T et al (1994) Studies on the regulation effect of natural insect on the
community food web in paddy field. Acta Ecol Sin 14:381–386
Katayama N, Baba YG, Kusumoto Y, Tanaka K (2015) A review of post-war changes in rice
farming and biodiversity in Japan. Agric Syst 132:73–84
Katayama N, Osada Y, Mashiko M et al (2019) Organic farming and associated management
practices benefit multiple wildlife taxa: a large-scale field study in rice paddy landscapes. J
Appl Ecol 56:1970–1981
Kramer RA, Richter DD, Pattanayak S, Sharma NP (1997) Ecological and economic analysis of
watershed protection in Eastern Madagascar. J Environ Manag 49:277–295
Lougheed VL, Mcintosh MD, Parker CA, Stevenson RJ (2008) Wetland degradation leads to
homogenization of the biota at local and landscape scales. Freshw Biol 53:2402–2413. https://
doi.org/10.1111/j.1365-2427.2008.02064.x
Maes J, Jacobs S (2017) Nature-based solutions for Europe’s sustainable development. Conserv
Lett 10:121–124
Martin TG, Watson JEM (2016) Intact ecosystems provide best defence against climate change.
Nat Clim Chang 6:122–124. https://doi.org/10.1038/nclimate2918
Matsuno Y, Nakamura K, Masumoto T et al (2006) Prospects for multifunctionality of paddy rice
cultivation in Japan and other countries in monsoon Asia. Paddy Water Environ 4:189–197.
https://doi.org/10.1007/s10333-006-0048-4
Morimoto J, Shibata M, Shida Y, Nakamura F (2017) Wetland restoration by natural succession in
abandoned pastures with a degraded soil seed bank. Restor Ecol 25:1005–1014. https://doi.org/
10.1111/rec.12516
Nakamura F, Akasaka T, Ishiyama N et al (2019) Adaptation to climate change and conservation of
biodiversity using green infrastructure. River Res Appl 1–13. https://doi.org/10.1002/rra.3576
184 T. Osawa et al.

Natuhara Y (2013) Ecosystem services by paddy fields as substitutes of natural wetlands in Japan.
Ecol Eng 56:97–106
Onuma A, Tsuge T (2018) Comparing green infrastructure as ecosystem-based disaster risk
reduction with gray infrastructure in terms of costs and benefits under uncertainty: a theoretical
approach. Int J Disaster Risk Reduct 32:22–28. https://doi.org/10.1016/j.ijdrr.2018.01.025
Osawa T (2017) Agricultural land use policy in Japan in an era of declining population. Wildlife
Hum Soc 5:17–27; (in Japanese)
Osawa T, Mitsuhashi H (2017) Trial for area zoning in Japanese agricultural area based on
ecological function. Ecol Civ Eng 19:211–220
Osawa T, Kohyama K, Mitsuhashi H (2013) Areas of increasing agricultural abandonment overlap
the distribution of previously common, currently threatened plant species. PLoS One 8:e79978.
https://doi.org/10.1371/journal.pone.0079978
Osawa T, Kohyama K, Mitsuhashi H (2016a) Multiple factors drive regional agricultural abandon-
ment. Sci Total Environ 542:478–483. https://doi.org/10.1016/j.scitotenv.2015.10.067
Osawa T, Kohyama K, Mitsuhashi H (2016b) Trade-off relationship between modern agriculture
and biodiversity: heavy consolidation work has a long-term negative impact on plant species
diversity. Land Use Policy 54:78–84. https://doi.org/10.1016/j.landusepol.2016.02.001
Osawa T, Nishida T, Oka T (2020a) High tolerance land use against flood disasters: how paddy
fields as previously natural wetland inhibit the occurrence of floods. Ecol Indic 114:106306
Osawa T, Nishida T, Oka T (2020b) Paddy fields located in water storage zones could take over
the wetland plant community. Sci Rep 10:1–8
Renaud FG, Sudmeier-Rieux K, Estrella M, Nehren U (2016) Ecosystem-based disaster risk
reduction and adaptation in practice. Springer, Cham. https://doi.org/10.1007/978-3-319-
43633-3
Sandstrom U (2002) Green infrastructure planning in urban Sweden. Plan Pract Res 17:373–385
Sarukhán J, Whyte A, Hassan R et al (2005) Millennium ecosystem assessment: ecosystems and
human well-being
Scarano FR (2017) Ecosystem-based adaptation to climate change: concept, scalability and
a role for conservation science. Perspect Ecol Conserv 15:65–73. https://doi.org/10.1016/
j.pecon.2017.05.003
Sudmeier-Rieux K, Ash N, Murti R (2013) Environmental guidance note for disaster risk
reduction. IUCN, Gland, Switzerland. https://www.iucn.org/sites/dev/files/content/documents/
2013_iucn_bookv2.pdf
Tittonell P (2014) Ecological intensification of agriculture-sustainable by nature. Curr Opin
Environ Sustain 8:53–61. https://doi.org/10.1016/j.cosust.2014.08.006
Toral GM, Stillman RA, Santoro S, Figuerola J (2012) The importance of rice fields for glossy ibis
(Plegadis falcinellus): management recommendations derived from an individual-based model.
Biol Conserv 148:19–27
Tzoulas K, Korpela K, Venn S et al (2007) Promoting ecosystem and human health in urban areas
using green infrastructure: a literature review. Landsc Urban Plan 81:167–178
Uchida K, Ushimaru A (2014) Biodiversity declines due to abandonment and intensification
of agricultural lands: patterns and mechanisms. Ecol Monogr 84:637–658. https://doi.org/
10.1890/13-2170.1
Yamanaka S, Akasaka T, Yabuhara Y, Nakamura F (2017) Influence of farmland abandonment on
the species composition of wetland ground beetles in Kushiro, Japan. Agric Ecosyst Environ
249:31–37. https://doi.org/10.1016/j.agee.2017.07.027
Yamazaki M, Hamada Y, Ibuka T et al (2001) Seasonal variations in the community structure of
aquatic organisms in a paddy field under a long-term fertilizer trial. Soil Sci Plant Nutr 47:587–
599
Yang S, Peng S, Xu J et al (2012) Methane and nitrous oxide emissions from paddy field as affected
by water-saving irrigation. Phys Chem Earth Parts A/B/C 53:30–37
Yoon CG (2009) Wise use of paddy rice fields to partially compensate for the loss of natural
wetlands. Paddy Water Environ 7:357–366. https://doi.org/10.1007/s10333-009-0178-6
11 Paddy Fields as GI 185

Yoshida T, Masumoto T, HOrikawa N (2012) Changes in rainfall-runoff characteristics of small

watersheds due to cultivation conditions of hilly paddies. Trans Jpn Soc Irrig Drain Rural Eng
80:117–124; (in Japanese)
Yoshikawa N, Nagao N, Misawa S (2009) Watershed scale evaluation of flood mitigation function
of Paddy fields installed with runoff control devices. Trans Jpn Soc Irrig Drain Rural Eng
77:273–280. (in Japanese)
Yoshikawa N, Nagao N, Misawa S (2010) Evaluation of the flood mitigation effect of a paddy field
dam project. Agric Water Manag 97:259–270

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
 Part IV
Wetland Ecosystem (Including
Flood-Control Pond)
Chapter 12
Flood-Control Basins as Green
Infrastructures: Flood-Risk Reduction,
Biodiversity Conservation,
and Sustainable Management in Japan

Nobuo Ishiyama, Satoshi Yamanaka, Keita Ooue, Masayuki Senzaki,

Munehiro Kitazawa, Junko Morimoto, and Futoshi Nakamura

Abstract Green infrastructure (GI) is a strategic planning infrastructure that

uses the functions of ecosystems. Under an increased river flood risk, flood-
risk management utilizing GI is gaining attention from managers and ecologists
in Japan. Flood-control basins are facilities that temporarily store river water in
adjacent reservoirs to mitigate flood peaks and gradually drain the water back to the
main channels after a flood. GI is expected to provide multiple functions, such as
flood-risk reduction and habitat provisions. However, there are limited studies on
the ecological functions of flood-control basins. In this article, we first introduce
the characteristics of flood-control basins constructed in Japan. Next, we show
the ecological importance of flood-control basins in terms of wetland organism
biodiversity conservation. Finally, to aid the integration of GI into conventional
flood-control measures, we highlight ecological and social issues about introducing
and managing flood-control basins.

Keywords Climate change · Biodiversity conservation · Ecosystem-based

adaptation · Ecosystem services · Flooding · Habitat connectivity · Sustainable
development

N. Ishiyama ()
Forest Research Institute, Hokkaido Research Organization, Bibai, Hokkaido, Japan
S. Yamanaka
Hokkaido Research Center, Forestry and Forest Products Research Institute, Sapporo, Japan
K. Ooue
Environmental Policy Division, Ministry of Land, Infrastructure, Transport and Tourism, Tokyo,
Japan
M. Senzaki
Faculty of Environmental Earth Science, Hokkaido University, Sapporo, Hokkaido, Japan
M. Kitazawa · J. Morimoto · F. Nakamura
Graduate School of Agriculture, Hokkaido University, Sapporo, Hokkaido, Japan

© The Author(s) 2022 189

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_12
190 N. Ishiyama et al.

12.1 Introduction

Among disasters triggered by natural hazards, flood disasters have been most
frequently reported worldwide. Under a changing climate, an increased flood risk
is predicted to affect human and economic losses globally (Dottori et al. 2018).
Historical records from 1962 to 2011 in the central United States demonstrated an
increase in flooding frequency (Mallakpour and Villarini 2015). In Europe, peaks
of 1/100 river floods are projected to double in frequency within the next three
decades (Alfieri et al. 2015). Jongman et al. (2012) reported that the amount of the
global population exposed to a 1/100 river flood reached 800 million by 2010, of
which 73% was living in Asia. Moreover, Dottori et al. (2018) estimated that the
population exposed to flooding will increase as a result of anthropogenic warming;
an average increase of more than 120% is expected in a 3 ◦ C-warming scenario.
Considering the increased disaster risk, adaptation efforts for flood-risk man-
agement are urgently needed. In Japan, heavy rainfall events (e.g., those above
50 mm/h) have increased in the last half a century (Fig. 12.1), and a flood disaster
with the most severe economic damage occurred in 2019 (Fig. 12.2). Typhoon
Hagibis in 2019 bore down on central Japan and caused 19 billion USD in economic
damage. The heavy rainfall event caused 142 levee collapses and overflowed along
rivers managed by the Japanese government, resulting in ca. 25,000 ha of inundated
land. Flood-control measures in Japan have focused on the construction of dams and
artificial levees for the last century. These conventional gray infrastructures usually

Fig. 12.1 Historical occurrence of heavy rainfall events in Japan. The number of monitoring sites
is 1300. In the first decade (1976–1985) and last decade (2010–2019), the mean annual occurrences
were 226 and 327, respectively. (Data are provided by the Japan Meteorological Agency (https://
www.data.jma.go.jp/cpdinfo/extreme/extreme_p.html))
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 191

Fig. 12.2 Historical economic damage caused by flood disasters in Japan. (Data are provided
by the Japanese Ministry of Land, Infrastructure, Transport and Tourism (https://www.mlit.go.jp/
report/press/content/001359046.pdf))

assure 100% disaster protection until the magnitude of the disaster reaches an upper
limit determined by the prevention plan, but the function will completely fail once
the magnitude exceeds the upper limits (Nakamura et al. 2020). Against the back-
drop of global warming, it has been widely recognized that conventional measures
that depend highly on gray infrastructures are no longer adequate for flood-risk
management. The Ministry of Land, Infrastructure, Transport and Tourism (MLIT)
has started to shift conventional measures to basin-wide flood-risk management that
focuses on both river and floodplain management using green infrastructure (GI).
GIs are conceptually classified into fundamental GI (GI-1) and multilevel GI (GI-
2) (Nakamura et al. 2020; see Chap. 2 for details). In flood-risk management, GI-1
are natural ecosystems, and GI-2 are seminatural basins that reduce or delay river
flooding. In 2016, Hokkaido Island was hit by three typhoons in one summer, which
was the first time this occurred since records began in 1952. Nakamura et al. (2020)
analyzed Kushiro wetland’s water retention function (GI-1), the largest remnant
wetland in Japan. Their study reported that a hydrograph of the simulation case of a
partial loss of wetlands (ca. 55% loss) showed 1.5 times higher peak discharge and
a 2-day-faster peak arrival. However, natural wetlands have been globally reduced
by human activities (Davidson 2014), and the importance of GI-2 would increase
in regions where the function of natural ecosystems is degraded. Considering the
effectiveness of GI, MLIT tries to integrate both gray and green infrastructures (GI-
1 and GI-2) into basin-wide flood-risk management to adapt to the increasing hazard
risk. This combination of different GIs is referred to as “hybrid infrastructures” in
Nakamura et al. (2020) (see Chap. 2).
192 N. Ishiyama et al.

In Japan, flood-control basins are rapidly gaining attention as a practical mul-

tilevel GI (GI-2). Multifunctionality is a fundamental property of sustainable GI
(Lovell and Taylor 2013; Wang and Banzhaf 2018). Biodiversity conservation in
a changing climate and landscape is another key challenge, and GI can be a tool
for mitigating or restoring declined biodiversity (Nakamura et al. 2020). In this
article, we first introduce the characteristics of flood-control basins constructed in
Japan. Next, we show the ecological importance of flood-control basins in terms of
wetland organism biodiversity conservation. Finally, to aid the integration of GI into
conventional flood-control measures in Japan, we would like to raise ecological and
social issues about introducing and managing flood-control basins. We believe that
this chapter provides insight into the future management of flood-control basins in
other Asian countries experiencing rapid urbanization and increased flooding risk.

12.2 Flood-Control Basins in Japan

There are two types of flood-risk management using retention/detention ponds. The
first is to store rainwater in floodplain areas to reduce runoff into the main channels
and mitigate the flood peak in farmland and urban areas. Urban retaining ponds
are examples, and rice paddies also may serve a similar function. Another is to
temporarily store river water in adjacent reservoirs to mitigate the flood peak and
gradually drain the water back into the main channels after the flood. The latter type
is called “Yusuichi” in Japanese. In this chapter, we focus on Yusuichi and represent
it as a flood-control basin.
Flood-control basins generally consist of reservoirs surrounded by artificial
levees and an overflow embankment or sluice gate adjacent to rivers (Fig. 12.3).
Reservoirs of flood-control basins are used for various purposes, such as sports

Fig. 12.3 Illustration of a flood-control basin. (Adapted from materials provided by the Hokkaido
Regional Development Bureau (https://www.hkd.mlit.go.jp/sp/kasen_keikaku/kluhh40000001qfy.
html)). (a) Normal flow. (b) High flow
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 193

grounds, farmland, urban parks, and wildlife habitat, and the presence or absence
of permanent water varies depending on the region and the type of usage. In Japan,
flood-control basins have been constructed nationwide (Suwa and Nishihiro 2020),
and their construction is also planned in several regions to reduce flood risks (e.g.,
Hitachi River and National Highway Office 2020; Kumamoto Prefecture 2020).
These flood-control basins are reported to have mitigated the disaster risk level.
For example, the four flood-control basins located in the Tone and Watarase rivers
in central Japan are estimated to have stored 250 million cubic meters of river water
during Typhoon Hagibis in 2019, resulting in mitigating flood damage downstream
(Tone River Upstream Office 2020).
Flood-control basins are also assumed to provide alternative habitats for wetland
species. Suwa and Nishihiro (2020) demonstrated that most flood-control basins in
Japan are located in floodplains and that 88% of them have natural observation
areas, paddy fields, reed marshes, or water surfaces, suggesting that they can
potentially provide wetland environments. Therefore, there is a possibility that
flood-control basins can be used not only for disaster risk reduction but also for
regional biodiversity conservation.

12.3 Case Study: Biodiversity Conservation in Flood-Control

Basins

Flood-control basins are assumed to contribute to maintaining regional biodiversity

and thus can work as GI. For example, the Watarase flood-control basin in central
Japan was designated as a Ramsar site having the largest reed bed and provides
wetland habitats for many species, including endangered species (Ministry of
Environment of Japan 2020b). However, there are limited studies assessing the
importance of flood-control basins for regional biodiversity conservation. Therefore,
we examined wetland species in flood-control basins recently implemented in the
Chitose River basin in Hokkaido, northern Japan (Yamanaka et al. 2020).
In this region, river flooding often occurred because of the gentle riverbed
slope (1/7000 on average). In particular, a heavy rain event in August 1981 led
to flooding and caused severe damage to urban areas and farmlands in this region
(Hokkaido Regional Development Bureau 2018; Segawa et al. 2008). To mitigate
flood risk, MLIT drove the construction of six flood-control basins near the main
stream and tributary of the Chitose River (Fig. 12.4). Construction started in 2008
and finished in 2020. The reservoirs of the basins comprise a total of 1150 ha,
and they have wetland environments where river water accumulates, except in
areas used for pasture and other purposes (Fig. 12.5). In the Chitose River basin,
agricultural land use, such as rice paddies and cropland, dominates, and there
are many watercourses for irrigation. The expansion of agricultural land in this
area began approximately 100 years ago, and most of the natural wetlands have
been converted to farmland (GSI 2000), resulting in a massive decrease in habitats
194 N. Ishiyama et al.

Fig. 12.4 Location of the Chitose flood-control basins

for wetland species. Therefore, flood-control basins consisting of new wetland

environments are expected to provide alternative habitats for wetland biota.
We examined the species composition of four wetland taxa, fish, aquatic insects,
birds, and plants, in five flood-control basins in summer 2016 (Yamanaka et al.
2020). In 2016, the construction of the basins was not completed except for the
Maizuru basin (the construction of all basins was completed and operations started
in 2020). Thus, we surveyed the above taxa in a part of the reservoir of each flood-
control basin. We also examined the species composition of three other waterbodies
in this area (channelized watercourses, drainage pumping stations, and remnant
floodplain ponds) to compare the species compositions with those of flood-control
basins.
We found that flood-control basins have a comparable or higher species richness
and abundance of wetland species than other waterbodies (e.g., remnant floodplain
ponds and drainage pumping stations) (Fig. 12.6). We also found that flood-control
basins were characterized by some pioneer species that preferred shallow water
or adapted to fluctuations in water levels (e.g., herbivorous insects, shorebirds,
and hygrophyte plants). However, channelized watercourses, which are widely
distributed in the study region, have lower species richness and abundance (Fig.
12.6). This result could be because of their simplified habitats led by channelization.
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 195

Fig. 12.5 Chitose flood-control basins

We also found some red list species of each taxon in flood-control basins although
there was lower abundance and richness than those in drainage pump stations
and remnant floodplain ponds. These results suggest that flood-control basins
provide alternative habitats for wetland species, including endangered species.
Nevertheless, for fishes, we observed a high abundance of nonnative species, such as
Pseudorasbora parva and Rhodeus ocellatus ocellatus, in some flood-control basins
(Fig. 12.6b).
In the Maizuru flood-control basin, whose construction finished in 2016, the
breeding of the red-crowned crane, Grus japonensis, was observed in 2020 (Min-
istry of Environment of Japan 2020a). The red-crowned crane is an endangered
species whose population in Japan experienced a significant decrease approximately
100 years ago, and is one of the flagship wetland species in Japan. The Ministry of
the Environment has implemented conservation measures for this species, such as
promoting breeding, and the distribution area is expanding from the east to the west
of Hokkaido Island. Therefore, flood-control basins are also expected to contribute
to the dispersal and recolonization of this species into uncolonized areas.
Our findings suggest that newly created environments in the Chitose flood-
control basins provide suitable habitats for wetland species. However, there is
room for future research to evaluate the ecological function of flood-control basins.
196 N. Ishiyama et al.

Fig. 12.6 (continued)

12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 197

Fig. 12.6 (continued)

198 N. Ishiyama et al.

Fig. 12.6 Estimated species richness and abundance of four taxa; (a) native fishes, (b) nonnative
fishes, (c) aquatic insects, (d) birds and (e) plants. CW channelized watercourse, DPS drainage
pumping station, POND remnant pond, FCB flood-control basin. Black circles denote values
estimated by generalized linear models (GLMs). The whiskers indicate the 95% confidence interval
(CI). Gray circles denote each observed value. Different letters indicate significant differences
in the multiple comparison analysis (p < 0.05). The values for species richness and coverage of
vegetation indicate values per quadrat (2 × 2 m). (See Yamanaka et al. (2020) for details)

First, Yamanaka et al. (2020) limited the study season to summer. The importance
of flood-control basins for biodiversity conservation can vary with the studied
season because wetland species use different environments in different seasons.
For example, the importance of flood-control basins as spawning sites for some
fishes in spring and as stopover sites for immigrant birds in spring and autumn
was not examined in a previous study. Second, Yamanaka et al. (2020) did not
consider wetland-plant succession. Accumulated sediment delivered with flooding
in reservoirs will change the water level, which could change the vegetation types
from hydrophytic to terrestrial plants. Such changes in vegetation will affect habitat
qualities for other wetland organisms. For effective biological conservation using
flood-control basins, further studies are needed to assess the species composition of
flood-control basins with different seasons, and the succession should be monitored
over a long-term period.
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 199

12.4 Future Issues for the Construction and Management

of Flood-Control Basins

12.4.1 Social Issues for Construction

To adapt to increased disaster risks caused by climate change, the planning and
construction of flood-control basins are urgently needed in Japan. Nevertheless,
the construction of flood-control basins in floodplain areas tends to be costly in
terms of both time and money. The construction of flood-control basins requires a
large space around rivers; however, most areas of floodplains in Japan are used for
residential areas and farmland. Managers therefore spend much time negotiating
with landowners to acquire land for construction. For the Chitose flood-control
basins, lands for each basin were owned by approximately 50 people, and it took
3 years for the land acquisition process. Managers can leave private lands, such
as farmland, within reservoirs, but the approach could be costly because they need
to establish easements for private lands and provide momentary compensation to
landowners. Therefore, a pre-investigation of information on landowners is essential
for construction to start quickly. It is not always possible to construct a large flood-
control basin due to many landowners and stakeholders (e.g., nature conservation
groups), while a large storage capacity is needed for flood-risk management. In
such cases, managers need to consider more feasible plans, such as selecting sites
for multiple small basins to achieve the total desired storage capacity.
Another solution for rapid construction is utilizing unused land, such as aban-
doned farmland. Farmland was abandoned after the period of economic growth in
the 1950s and 1960s in developed countries, including Japan (e.g., Kobayashi et al.
2020). The use of such degraded lands with depopulation is increasingly recognized
as a tool for biodiversity conservation (Ishiyama et al. 2020a; Nakamura et al.
2020). For example, Nakamura et al. (2020) proposed a prioritization technique
for construction sites by estimating the distribution of abandoned farmland and the
biodiversity of wetland organisms (e.g., birds and plants). In their study, created
maps overlapped with a flood hazard map for selection and revealed a financially
and ecologically high-priority area.

12.4.2 Ecological Issues for Constructions

GI is defined as “a strategically planned network of natural and seminatural

areas with other environmental features designed and managed to deliver a wide
range of ecosystem services” (European Commission 2013). In human-modified
floodplains, the contemporary migration of wetland organisms has been spatially
restricted (e.g., Ishiyama et al. 2015b). However, artificial watercourses and remnant
wetlands create a seminatural wetland network in current landscapes, and landscape
connectivity supports the high biodiversity of wetland organisms (Ishiyama et al.
200 N. Ishiyama et al.

2014, 2015a). Flood-control basins provide large open water spaces compared to
other lentic waterbodies in human-modified floodplains (Yamanaka et al. 2020),
suggesting that the construction of flood-control basins potentially contributes to
enhancing a habitat network of wetland organisms by interacting with the existing
lentic waterbodies (Ishiyama et al. 2017). Spatial network analyses can be a solution
for creating “a strategically planned network” (Hermoso et al. 2020). One of the
simple methods is site-scale measures that consider direct connections between
the focal habitat patch and surrounding patches. The nearest neighbor connectivity
measure (e.g., the distance between the focal habitat patch and the nearest patch) and
the buffer measure (e.g., the summed area of the habitat patch within a circle around
the focal habitat patch) are typical examples (Moilanen and Nieminen 2002). The
site-scale measures require little data because the method is simple, as mentioned
above, which is a positive aspect of the method for practitioners in the spatial
planning of flood-control basins. However, the measures potentially evaluate the
connectivity of candidate sites with low precision, especially for highly mobile
animals such as insects and waterfowls. This is because the dispersal ranges of
such animals are not restricted to the area around the focal habitat patch; they
seasonally use multiple patches in the landscape by using stepping-stone patches.
That is to say, both direct and indirect connections among habitat patches should
be assessed for such animals. Alternatively, regional-scale measures can prioritize
candidate sites for flood-control basin construction by considering the importance
of stepping-stone habitats (e.g., Saura and Rubio 2010). For instance, a graph-
theoretical approach can calculate overall connectivity (i.e., regional-scale habitat
availability) by considering the spatial position and size of habitat patches (Saura
and Torné 2009). By using the regional-scale measure, practitioners can assess the
importance of individual patches in maintaining the entire habitat network (e.g.,
Ishiyama et al. 2014, 2015a). However, the regional-scale measures may be more
strenuous for practitioners compared to buffer measures due to more complicated
operations for creating a landscape graph and matrix calculations. More effective
sites for improving existing wetland networks can be selected by using the network
measures according to the species traits of conservation targets and available
conservation resources.
However, improving a habitat network can have adverse effects on indigenous
species. A key reason for the species decline is expanding introduced species
(Rudnick et al. 2012). We therefore should carefully select the construction sites
of flood-control basins and the connections with other waterbodies in consideration
of the potential negative impacts. In northern Japan, many floodplain waterbodies,
such as oxbow lakes and backswamps, are invaded by a small invaded cyprinid,
Pseudorasbora parva. Cyprid abundance is higher in waterbodies with high hydro-
logic connectivity (Ishiyama et al. 2020b). Ishiyama et al. (2020b) showed that
populations of the endangered minnow Rhynchocypris percnurus sachalinensis,
which has an ecological niche similar to that of P. parva, decline with the invasion
of P. parva, suggesting that small isolated waterbodies function as refuges for the
endangered minnow (Fig. 12.7). Nevertheless, small isolated habitats are easily
diminished and degraded by human activities. Considering the importance of small-
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 201

Fig. 12.7 Biological invasion in a wetland network and decline in endangered species. Example
of P. parva and R. p. sachaliensis in the Tokachi floodplain, northern Japan. (See Ishiyama et al.
(2020b) for details)

isolated waterbodies, such remnants should be preserved when managers plan the
construction of flood-control basins (Fig. 12.8). As previously mentioned in the
third section, we should also remember that the creation of flood-control basins
can support the establishment of nonnative species and create new sources for the
secondary spread of these species to connected waterbodies in wetland networks.

12.4.3 Sustainable Management of Flood-Control Basins

Various management practices are needed to fulfill both functions of flood-control

basins (i.e., disaster risk reduction and habitat provision). In the Maizuru and
Watarase flood-control basins, managers have taken measures to conserve endan-
gered species, such as making a breeding site and eliminating invasive species
(Chino and Mizuno 2019; MLIT Kantou Regional Development Bureau 2020a). In
addition, the Maizuru flood-control basin is legally protected as a wildlife reserve
area by the Hokkaido government. Long-term maintenance of wetland environments
is also a vital issue for sustaining functions. The accumulation of dead reed plants
202 N. Ishiyama et al.

Fig. 12.8 Network thinking in biodiversity conservation using flood-control basins. Improvement
of existing wetland networks and preservation of small isolated wetlands should be strategically
conducted

and the establishment of pioneer tree species can change the setting in flood-control
basins from fen to swamp dominated by Salix spp., which will decrease the flood-
control capacity and amount of habitat for wetland organisms. For example, wetland
environments in the Asahata flood-control basins, Shizuoka prefecture, have dried,
and the habitat used by endangered wetland species is shrinking (Shizuoka pre-
fecture 2017). Control burning is one of the historical measures for preventing
vegetation succession in Japan. In some Japanese flood-control basins, from winter
to early spring, control burning is conducted by local communities, including the
government, nonprofit organizations (NPOs), and residents, to maintain the function
(e.g., MLIT Kantou Regional Development Bureau 2020c). Creating open water and
an extended hydroperiod by dredging accumulated sediment and organic material is
another option for wetland management (e.g., Stevens et al. 2003). In the Watarase
flood-control basin, dredging works were conducted to restore a heterogenous
wetland landscape (MLIT Kantou Regional Development Bureau 2020d). Wetlands
with different flood frequencies and multiple open waters were created by this work,
which contributed to restoring floodplain vegetation (Ishii et al. 2011) and foraging
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 203

sites for a top avian predator, the eastern marsh harrier Circus spilonotus (Hirano
2015).
As mentioned above, various types of management are necessary for maintaining
the functions of flood-control basins. However, the running cost for deteriorated
gray infrastructure is increasing (Council for Social Infrastructure 2020), and it
is difficult for the government alone to sustain GI functions. Cooperative man-
agement of flood-control basins among the government, researchers, the private
sector, NPOs, and residents will be essential to maintain the multiple functions
of flood-control basins. For the Watarase flood-control basins, the government has
established a council for management that aims to conserve and sustainably use
wetland environments. The council is composed of multiple organizations, such
as NPOs and local governments, and conducts biological monitoring and environ-
mental education to help local residents to understand the importance of wetland
environments and participate in maintenance (MLIT Kantou Regional Development
Bureau 2020b). The attraction of private investment to flood-control basins has
received growing attention as a useful tool for sustainable management. The council
of the Watarase flood-control basin established the fund “Watarase-Mirai-Kikin” in
2001 and raises funds from private companies that are interested in environmental,
social, and governance (ESG) investment and corporate social responsibility (CSR).
This fund is used for sustainable management, such as restoration of wetland
environments and human resource development. In March 2020, MLIT launched
the GI Public-Private Partnership Platform to manage GI, including flood-control
basins, more sustainably. The platform is organized by multiple stakeholders, such
as national and local governments, private companies, research institutes, NPOs, and
citizens. Working toward sustainable and attractive city development using flood-
control basins, the platform is working on promoting GI, developing construction
and management technologies, and financing techniques.
The gap between conservation science and real-world action is a genuine phe-
nomenon (Knight et al. 2008; Osawa and Ueno 2017). The research-implementation
gap will impede the success of the sustainable management of flood-control
basins regardless of how many academic papers are published. Knight et al.
(2008) observed that “conservation planners must facilitate a solution to a specific
practitioner’s need; it is generally not effective to conduct a conservation assessment
and then attempt to promote it post hoc to a practitioner.” For the Watarase flood-
control basins, MLIT established a committee for conserving wetland environments,
composed of both practitioners and researchers, that held multiple meetings to
develop the restoration plan. One of the key benefits of establishing the committee is
facilitating interactions between researchers and practitioners. In such a committee,
researchers can formulate research questions collaboratively with stakeholders and
understand specific practitioner’s needs and implementation constraints. Indeed,
the information on potential impacts of topsoil dredging on the distribution of
threatened plant species assessed by one of the committee members (Obata et
al. 2012) delivered a management plan for the flood-control basin; practitioners
selected dredging sites by referring to the academic evidence.
204 N. Ishiyama et al.

12.4.4 Importance of Multifunctionality

In this chapter, we introduced that flood-control basins can have environments

similar to those of natural wetlands, providing important functions for humans
and living organisms (i.e., flood-risk reduction and habitat provision). However,
natural wetlands (GI-1) are multiple-value systems (Mitsch and Gosselink 2000)
and provide more functions for humans, such as water quality improvement and
recreational activities. In future flood-risk management, we should also focus on
such functions and wisely utilize flood-control basins (GI-2) to increase residents’
quality of life.
In Japan, some flood-control basins located around cities are used for public
parks with green spaces (e.g., the Tsurumi River flood-control basins). A growing
body of literature has emphasized the importance of people’s green space use for
mental health during the COVID-19 pandemic (e.g., Soga et al. 2020). Considering
lifestyle changes after the global pandemic, flood-control basins around cities can
serve as good health and physical resources for citizens. As previously mentioned in
the third and fourth sections, spatial planning of flood-control basins would benefit
wetland organisms in terms of improving habitat networks. Humans will also profit
from the strategic restoration of habitat networks through recreational activities such
as fisheries and birdwatching.
Flood-control basins with wetland environments can be a mitigation tool for
improving water quality in river networks. Recent work in the USA has shown that
a spatially targeted increase in wetland area by 10% (i.e., wetlands are preferentially
placed in areas with the highest nitrate surplus) would double wetland nitrate
removal (Cheng et al. 2020). For example, farmland expansions in northern Japan
cause a decline in juvenile Oncorhynchus masou, which is one of the key fishery
resources, through nutrient enrichment (Ishiyama et al. 2020a). Strategic spatial
planning of flood-control basins can mitigate such degradation in river ecosystems
and their delivered services.
As mentioned above, flood-control basins can bring various benefits inside
and outside of the GI and possibly contribute to improving residents’ quality of
life. However, ecosystem services can have trade-off relationships with each other
(Bennett et al. 2009). To maximize the multiple functions of flood-control basins
and manage them sustainably, scientists and managers should further understand
the relationships among the provided functions and the mechanisms behind the
relationships.

References

Alfieri L, Burek P, Feyen L, Forzieri G (2015) Global warming increases the frequency of river
floods in Europe. Hydrol Earth Syst Sci 19(5):2247–2260
Bennett EM, Peterson GD, Gordon LJ (2009) Understanding relationships among multiple
ecosystem services. Ecol Lett 12(12):1394–1404
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 205

Cheng F, Van Meter K, Byrnes D, Basu N (2020) Maximizing US nitrate removal through wetland
protection and restoration. Nature:1–6
Chino SK, Mizuno H (2019) Environmental monitoring in the Maiduru flood-control
basin (in Japanese). https://www.hkd.mlit.go.jp/ky/jg/gijyutu/splaat000001t2er-att/
splaat000001t2op.pdf. Accessed 15 Jan 2021
Council for Social Infrastructure (2020) 44th Planning committee: socioeconomic conditions
surrounding social infrastructure development (in Japanese). https://www.mlit.go.jp/policy/
shingikai/content/001342849.pdf. Accessed 15 Jan 2021
Davidson NC (2014) How much wetland has the world lost? Long-term and recent trends in global
wetland area. Mar Freshw Res 65(10):934–941
Dottori F, Szewczyk W, Ciscar J-C, Zhao F, Alfieri L, Hirabayashi Y, Bianchi A, Mongelli I,
Frieler K, Betts RA (2018) Increased human and economic losses from river flooding with
anthropogenic warming. Nat Clim Chang 8(9):781–786
European Commission (2013) Green Infrastructure (GI) — Enhancing Europe’s Natural Capi-
tal. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52013DC0249. Accessed 15
Jan 2021
GSI (2000) National survey of lakes and wetlands. Geospatial Information Authority of Japan
(GSI). http://www.gsi.go.jp/kankyochiri/shicchimenseki2.html (in Japanese). Accessed 25
April 2018
Hermoso V, Morán-Ordóñez A, Lanzas M, Brotons L (2020) Designing a network of green
infrastructure for the EU. Landscape Urban Plann 196:103732
Hirano T (2015) The effects of newly created ponds on hunting flights of eastern marsh harriers
during the wintering season (in Japanese with English abstract). Bird Res 11:a21–a30
Hitachi River and National Highway Office (2020) Emergency flood control project in the Kuji
River and the Naka River (in Japanese). https://www.ktr.mlit.go.jp/hitachi/hitachi00814.html.
Accessed 26 Dec 2020
Hokkaido Regional Development Bureau (2018) Initiatives for river development: flood con-
trol measures in the Chitose River Basin. https://www.hkd.mlit.go.jp/sp/kasen_keikaku/
kluhh40000001qfy.html. Accessed 25 April 2018
Ishii J, Hashimoto L, Washitani I (2011) Early vegetation growth in an experimental restoration
site in the Watarase wetland. Jpn J Conserv Ecol 16:69–84
Ishiyama N, Akasaka T, Nakamura F (2014) Mobility-dependent response of aquatic animal
species richness to a wetland network in an agricultural landscape (in Japanese with English
abstract). Aquat Sci 76(3):437–449
Ishiyama N, Koizumi I, Yuta T, Nakamura F (2015a) Differential effects of spatial network
structure and scale on population size and genetic diversity of the ninespine stickleback in a
remnant wetland system. Freshwat Biol 60(4):733–744
Ishiyama N, Sueyoshi M, Nakamura F (2015b) To what extent do human-altered landscapes retain
population connectivity? Historical changes in gene flow of wetland fish Pungitius pungitius.
R Soc Open Sci 2(7):150033
Ishiyama N, Nagayama S, Iwase H, Nakamura F (2017) Restoration techniques for riverine aquatic
connectivity: current trends and future challenges in Japan (in Japanese with English abstract).
Ecol Civil Eng 19:143–164
Ishiyama N, Miura K, Inoue T, Sueyoshi M, Nakamura F (2020a) Geology-dependent impacts of
forest conversion on stream fish diversity. Conserv Biol
Ishiyama N, Miura K, Yamanaka S, Negishi JN, Nakamura F (2020b) Contribution of small
isolated habitats in creating refuges from biological invasions along a geomorphological
gradient of floodplain waterbodies. J Appl Ecol 57(3):548–558
Jongman B, Ward PJ, Aerts JC (2012) Global exposure to river and coastal flooding: long term
trends and changes. Glob Environ Chang 22(4):823–835
Knight AT, Cowling RM, Rouget M, Balmford A, Lombard AT, Campbell BM (2008) Knowing but
not doing: selecting priority conservation areas and the research–implementation gap. Conserv
Biol 22:610–617
206 N. Ishiyama et al.

Kobayashi Y, Higa M, Higashiyama K, Nakamura F (2020) Drivers of land-use changes in societies

with decreasing populations: a comparison of the factors affecting farmland abandonment in a
food production area in Japan. PLoS One 15(7):e0235846
Kumamoto Prefecture (2020) Pamphlet for special emergency project as the countermeasures
against terrible disasters in the Shira river and the Kuro River (in Japanese). https://
www.pref.kumamoto.jp/soshiki/105/5723.html. Accessed 26 Dec 2020
Lovell ST, Taylor JR (2013) Supplying urban ecosystem services through multifunctional green
infrastructure in the United States. Landsc Ecol 28(8):1447–1463
Mallakpour I, Villarini G (2015) The changing nature of flooding across the Central United States.
Nat Clim Chang 5(3):250–254
Ministry of Environment of Japan (2020a) The birth of a red-crowned crane chick at the
Maizuru flood-control basin in Naganuma (in Japanese). http://hokkaido.env.go.jp/pre_2020/
post_123.html. Accessed 26 Dec 2020
Ministry of Environment of Japan (2020b) Ramsar sites in Japan. https://www.env.go.jp/en/nature/
npr/ramsar_wetland/pamph/index.html. Accessed 26 Dec 2020
Mitsch WJ, Gosselink JG (2000) The value of wetlands: importance of scale and landscape setting.
Ecol Econ 35(1):25–33
MLIT Kantou Regional Development Bureau (2020a) The birth of a stork chick at
the Watarase flood-control basin (in Japanese). https://www.ktr.mlit.go.jp/river/news/
river_news00000009.pdf. Accessed 15 Jan 2021
MLIT Kantou Regional Development Bureau (2020b) Council for conservation and utiliza-
tion of the Watarase flood-control basin (in Japanese). https://www.ktr.mlit.go.jp/tonejo/
tonejo_index028.html. Accessed 15 Jan 2021
MLIT Kantou Regional Development Bureau (2020c) Information of the Watarase flood-control
basin (in Japanese). https://www.ktr.mlit.go.jp/tonejo/tonejo00081.html. Accessed 15 Jan 2021
MLIT Kantou Regional Development Bureau (2020d) Wetland restoration plan in the Watarase
flood-control basin (in Japanese). https://www.ktr.mlit.go.jp/tonejo/tonejo00661.html.
Accessed 11 May 2021
Moilanen A, Nieminen M (2002) Simple connectivity measures in spatial ecology. Ecology
83:1131–1145
Nakamura F, Ishiyama N, Yamanaka S, Higa M, Akasaka T, Kobayashi Y, Ono S, Fuke N,
Kitazawa M, Morimoto J (2020) Adaptation to climate change and conservation of biodiversity
using green infrastructure. River Res Appl 36(6):921–933
Obata T, Ishii J, Kadoya T, Washitani I (2012) Effect of past topsoil removal on the current
distribution of threatened plant species in a moist tall grassland of the Watarase wetland,
Japan: mapping of selected sites for wetland restoration by topsoil removal. Jpn J Conserv
Ecol 17:221–233
Osawa T, Ueno Y (2017) Research-implementation gap in Japanese ecology field. Jpn J Ecol
67:257–265
Rudnick D, Ryan SJ, Beier P, Cushman SA, Dieffenbach F, Epps C, Gerber LR, Hartter JN,
Jenness JS, Kintsch J (2012) The role of landscape connectivity in planning and implementing
conservation and restoration priorities. Issues Ecol 16:1–20
Saura S, Rubio L (2010) A common currency for the different ways in which patches and links can
contribute to habitat availability and connectivity in the landscape. Ecography 33:523–537
Saura S, Torné J (2009) Conefor Sensinode 2.2: a software package for quantifying the importance
of habitat patches for landscape connectivity. Environ Model Softw 24(1):135–139
Segawa A, Minato T, Yoshikawa K (2008) The history of flood plain development and the levee
construction of the Ishikari River down stream (in Japanese with English abstract). J Construct
Manage JSCE 15:429–440. https://doi.org/10.2208/procm.15.429
Shizuoka Prefecture (2017) Action plan for the conservation and utilization of the
Asahata flood-control basin (in Japanese). http://asabata.org/wp-content/uploads/2017/04/
4fad93fec4e8c2ca38736fb06cf4bb38.pdf. Accessed 15 Jan 2021
Soga M, Evans MJ, Tsuchiya K, Fukano Y (2020) A room with a green view: the importance of
nearby nature for mental health during the COVID-19 pandemic. Ecol Appl e2248
12 Flood-Control Basins as Green Infrastructures: Flood-Risk Reduction,. . . 207

Stevens CE, Gabor TS, Diamond AW (2003) Use of restored small wetlands by breeding waterfowl
in Prince Edward Island, Canada. Restor Ecol 11:3–12
Suwa Y, Nishihiro J (2020) Distribution and location characteristics of flood-control reservoir in
Japan (in Japanese with English abstract). Ecol Civ Eng 23:85–97
Tone River Upstream Office (2020) Report on the outflow of water in Typhoon No. 19 (in
Japanese). https://www.ktr.mlit.go.jp/tonejo/tonejo00367.html. Accessed 26 Dec 2020
Wang J, Banzhaf E (2018) Towards a better understanding of green infrastructure: a critical review.
Ecol Indicators 85:758–772
Yamanaka S, Ishiyama N, Senzaki M, Morimoto J, Kitazawa M, Fuke N, Nakamura F (2020) Role
of flood-control basins as summer habitat for wetland species-a multiple-taxon approach. Ecol
Eng 142:105617

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 13
Natural Succession of Wetland Vegetation
in a Flood-Control Pond Constructed
on Abandoned Farmland

Junko Morimoto, Susumu Goto, Akito Kuroyanagi, Motoko Toyoshima,

and Yuichiro Shida

Abstract Can a flood-control basin be considered a component of green infras-

tructure (GI) with not only disaster prevention functions but also biodiversity
conservation functions? We studied the succession of wetland vegetation in a flood-
control basin constructed in a floodplain. The number of species composing the
buried seeds depended on the depth of the soil layer, with shallower (recent) layers
having more species. Plants germinated from buried seeds in deep (ancient) soil
layers were small and difficult to identify based on morphology alone, but DNA-
based analyses made it possible to identify these species. The occurrence of three
species of Juncus and Cyperus in the peat layer that developed between 840 and
1215 years ago indicates that it is possible to regenerate wetland plants from the past
if the buried seeds are well preserved and in good condition. The excavated section
of the experimental site, which was set in part of the flood-control basin, received
dispersed seeds from the surrounding agricultural land and contained more species
than did the soil layer itself. The shallower the water depth was (the shallower the
excavation depth), the greater the number of species that occurred. In addition,
species with different life forms occurred depending on the water depth. More
species occurred in the entire flood-control basin than in the experimental site. The
factors that determined the type of plant community were years since excavation,
water depth, and water quality. Finally, we discussed the management practices
concerning the supply of propagules and the management of suitable habitats as
essential factors for flood-control basins constructed on fallow land to function as
GI.

J. Morimoto () · A. Kuroyanagi · M. Toyoshima

Graduate School of Agriculture, Hokkaido University, Sapporo, Hokkaido, Japan
e-mail: [email protected]
S. Goto
Graduate School of Agricultural and Life Sciences, The University of Tokyo, The University of
Tokyo Forests, Tokyo, Japan
Y. Shida
Wildlife Research Institute Co. Ltd., Sapporo, Hokkaido, Japan

© The Author(s) 2022 209

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_13
210 J. Morimoto et al.

Keywords Buried seeds · Seed dispersal · Water depth · Succession · DNA

barcoding · Field experiment

13.1 Introduction

The recent development of agricultural land has led to a global decline in wetland
biodiversity (World Water Assessment Programme 2009; Zedler and Kercher 2005).
This has led to concerns about the decline in biodiversity and ecosystem services
in wetlands. On the other hand, in developed regions such as North America and
the United Kingdom, the emergence of abandoned farmland has been reported
since the beginning of the nineteenth century, and in Eastern Europe, Russia, and
Latin America, this has occurred since the beginning of the twentieth century due
to socioeconomic deterioration such as urbanization and industrialization and a
decrease in the rural population (Baumann et al. 2011; Benayas and Bullock 2012;
Kuemmerle et al. 2008). Agricultural land converted from wetlands by artificial
draining is particularly prone to abandonment due to the high maintenance costs
(Benjamin et al. 2005; Middleton 2003). Abandoned farmland that was once a
wetland and adjacent to a river can be expected to serve as a flood-control basin
and to restore wetland ecosystems (Morimoto and Shibata 2018). If these goals can
be realized at the same time, it will be useful not only for disaster prevention but
also for biodiversity conservation, making such sites beneficial green infrastructure
(GI) from the perspective of improving the overall ecosystem services (Nakamura
et al. 2020). In this chapter, we introduce a series of studies on the process of natural
succession of wetland vegetation at a project site where a flood-control basin was
established on abandoned farmland that had been a floodplain until it was developed
into farmland in the twentieth century.

13.2 Materials and Methods

A flood-control basin (1150 ha) was constructed in a tributary of the Chitose River in
northern Japan, Hokkaido (42◦ 55 N, 141◦41 E, annual mean temperature of 7.2 ◦ C,
annual precipitation of 1005.6 mm), which was farmland that had been used as
paddy and dryland fields for approximately 10 years and then left fallow for 7 years
(refer to Ishiyama et al. of Chap. 12 for the positioning of the flood-control basin
and overall flood management). We chose the flood-control basin named “Maizuru
Yusuichi” in Japanese as our study site (hereafter referred to as Crane Pond). The
oldest topographic map of 1920 and later maps confirm that Crane Pond had been
at the bottom of a lake until 1960. Carbon dating analysis indicated that the humus
layer from 0 m to 0.3 m below the surface was deposited after 1739; the clay layer
from 0.3 m to 1.2 m below the surface, after between 1190 and 1275; and the peat
layer from 1.2 m to 1.5 m below the surface, after between 775 and 790 or 800
13 Natural Succession of Wetland Vegetation in a Flood-Control Pond. . . 211

(a) (b)

Fig. 13.1 Soil layers in Crane Pond (a) and assessment of the seed bank (b) Ta: volcanic deposit
from Mt. Tarumae

and 980 (Fig. 13.1a). Therefore, it is assumed that this area, including Crane Pond,
was a floodplain of a network of meandering rivers from the eighth or ninth century
until, at the latest, the twentieth century, when the land was developed for farming.
In this study area, before the construction of Crane Pond, an experiment was started
to clarify the composition of the seed bank in the fallow farmland (2.2), and after the
construction of Crane Pond, a survey of the naturally regenerated vegetation (2.3)
was conducted.

13.2.1 Experiment Started Before the Construction of Crane

Pond: Exploring the Seed Bank Species in Fallow
Farmland

Soils from the humus, clay, and peat layers were collected in December 2014 and
June 2015 from fallow farmland where Crane Pond was to later be constructed. The
soil blocks were mixed by layer with 0.1 times the volume of water and loosened,
and gravel and branches were removed. Two-centimeter-deep trays (0.0576 m2 )
were filled with soil that was sterilized by drying, and each layer of soil was
spread out at a thickness of 1 cm (Fig. 13.1b). The soil was submerged in water
so that the water surface coincided with the soil surface. The soil was covered with
mesh to prevent seed contamination from outside. As a control, we also prepared
212 J. Morimoto et al.

trays sown with sterilized soil at a depth of 1 cm. The germinating plants were
observed until the development of new plants stopped (until the end of August
for the soils collected in 2014 and until October for the soils collected in 2015).
The plants were then grown until October 2016, and flowering and fruiting were
confirmed. We identified the species by their morphological features except for
three plants. These plants germinated from the peat layer as old as ca. 1000 years
(hereafter, the ancient group). The size of the three plants was so small that we
could not identify the species by morphological features alone, so chloroplast DNA
information was used for species identification. First, we sampled collected leaves
from the present group of Juncus decipiens (Buchenau) Nakai, Juncus tenuis Willd.,
Juncus bufonius L., and Juncus wallichianus Laharpe (which were candidate species
based on their morphological features). Recently, DNA barcoding (barcode of life:
BOL) has proven useful for species identification. For plants, a part of the sequence
of the rbcL gene is effective for species identification. In this study, we extracted
DNA from the leaves of ancient and present plants, amplified the BOL region of the
collected samples, and identified species based on sequence data. After aligning the
sequences using MEGA ver. 7, we checked the corresponding sequences against the
NCBI database.

13.2.2 Experiment and Survey Started After the Construction

of Crane Pond: Exploring the Vegetation that
Regenerated Naturally

Crane Pond was constructed between 2009 and 2014 and was put into service
in April 2015. In July 2017, we randomly settled 51 quadrats (2 m × 2 m) in
Crane Pond, and the vegetation and aquatic environment was surveyed. Aquatic
environmental parameters, including electrical conductivity (EC, μS/cm), dissolved
oxygen (DO, %), pH, and depth (cm), of surface water were measured. Plant
communities were identified by TWINSPAN (two-way indicator species analysis)
from the vegetation dataset. The relationships between plant communities and the
water environment were analyzed by canonical correspondence analysis (CCA).
We built an experimental site (0.17 ha) to observe the relationships between
the regenerated species and water depth at the corner of Crane Pond (Fig. 13.2a).
Excavation was carried out at three different depths (Fig. 13.2b, c). The water
surface was at approximately the same level as the topmost ground surface (Fig.
13.2d). On the basis of the inaugural year of Crane Pond in 2015, we recorded the
plant cover ratio of each species regenerated from 2015 to 2019. The succession
over 5 years at three sections with different depths was monitored and recorded.
13 Natural Succession of Wetland Vegetation in a Flood-Control Pond. . . 213

(a) Crane Pond, flood-control basin (ca. 200 ha)

Field experiment site

(FES, 0.17 ha)

(b) Overhead view of FES (c) Cross-sectional view of FES

66 m
water level U
8m Upper section: U M 0.3 m alt. 4.1 m:
0.3 m average
L
8m Middle section: M altitude of
the pond
8m Lower section: L
8m 8m 8m
(d) Transition of the FES landscape over time
May, 2015 July, 2017

September, 2015 July, 2018

Fig. 13.2 Field experimental site built in a corner of a flood-control basin

214 J. Morimoto et al.

13.3 Species Composition of Buried Seeds in Crane Pond

13.3.1 Species Identification by a Combination

of Morphological and DNA Features

Three plants that germinated from the peat layer were too small to identify, but the
first one (P1) was assumed to be J. decipiens based on its morphological features.
rbcL DNA is used as a DNA barcoding region in plants (CBOL Plant Working
Group 2009) because the diversity of rbcL is conserved within a species but is
likely to vary between species. The rbcL DNA of many plant species is registered,
which is advantageous in that a large amount of information is available but a
disadvantage in that rbcL genes are often shared among closely related species, as
shown by the results of our analysis. According to a phylogenetic tree based on the
rbcL region (Tsubota et al. 2014), Juncus growing in wetlands in Hokkaido can be
roughly classified according to the nucleotide sequence of this region. Therefore,
we extracted DNA from leaves of P1 (Sample No. 4) and present plants of J.
decipiens (Sample Nos. 1–3) and J. tenuis (Sample Nos. 5–6). After comparing the
BOL sequences of the extracted DNA samples, at least five single-base substitutions
occurred among J. decipiens and J. tenuis, and P1 had the same sequence as two J.
tenuis plants collected in the southern part of Japan (Fig. 13.3). Then, we compared
their sequences with those registered in the database of related species. We found
that the sequences of Sample Nos. 1, 2, and 3 were identical to the sequence of J.
decipiens (database ID number: KT695563) and that the sequences of Sample Nos.
4 (P1), 5, and 6 were identical to those of J. tenuis (KJ593488) and J. bufonius L.
(KJ204357).
Since J. decipiens is abundant near the study site, we expected that P1 emerging
from the peat layer might be this species before we performed the DNA analysis.
However, the chloroplast DNA sequence clearly showed that P1 was not J. decipiens
but rather J. tenuis or J. bufonius. J. tenuis and J. bufonius can be identified by
the morphological features of their leaves (presence or absence of a membranous
auricle in the leaf sheath). Finally, P1 was identified as J. tenuis because there
was an auricle in the leaf sheath. Later, P2 and P3, which germinated from the

No.1 J decipiens
No.2 J decipiens
No.3 J decipiens
No.4 (P1)
No.5 J. tenuis
No.6 J. tenuis

Fig. 13.3 Sequences of the rbcL (BOL) region of the chloroplast genome for six plants. P1 (No.
4) germinated from the peat layer and was initially assumed to be Juncus decipiens. However, the
sequence was different from that of this species (Nos. 1, 2, and 3) at several sites (yellow arrows).
We found that the sequence of P1 (No. 4) was completely consistent with that of J. tenuis (Nos. 5
and 6)
13 Natural Succession of Wetland Vegetation in a Flood-Control Pond. . . 215

peat layer, were analyzed in the same way as P1, and the results were 100%
consistent with J. wallichianus (registered as AB985730.1) collected in Hiroshima
Prefecture. However, some of the sequences were different from those of the native
J. wallichianus in Crane Pond, and further study is needed. Thus, the combination
of morphological and DNA analyses showed the possibility of reliable species
identification of small plants, which are difficult to identify based on morphological
information alone.

13.3.2 Species Composition in Each Soil Layer from Different

Depths

All plants except the three plants germinated from the peat layer were identified
by morphological features alone. All the germinated species were present, and one
of them, Sparganium erectum L. var. coreanum (H. Lév.) H. Hara, was a near-
threatened species (Table 13.1). The number of species and plants was the highest in
the latest layer of humus and was lower in older layers of clay and peat (Fig. 13.4).
However, each layer produced unique species (Table 13.1). The planned elevation
of Crane Pond (4.1 m above sea level) is comparable to the height of the clay layer,
meaning that the humus layer containing the largest number of species has been
stripped off and has spread out into the depression inside Crane Pond. Thus, the
main contribution of buried seeds to natural succession in Crane Pond was assumed
to be buried seeds from the humus and clay layers. In addition, notably, three species
of Juncus and Cyperus germinated from the peat layer as old as ca. 1000 years old.
The result that the latest layer produced the most abundant species and plants
coincides with the view that the germination ratio of seeds declines over time
after production and that the seed longevity of most species is less than 5 years

Table 13.1 Number of plants from each soil layer

(m−2 )
Humus (N = 16) Clay (N = 14) Peat (N = 32)
Mean SD Mean SD Mean SD
Cyperus difformis 0.0 0.00 0.0 0.00 0.5 3.02
Juncus bufonius 1.1 4.1 0.0 0.00 0.0 0.00
Juncus decipiens 316.8 509.5 0.0 0.00 0.0 0.00
Juncus prismatocarpus 9.8 25.2 0.0 0.00 1.6 6.7
Juncus tenuis 0.0 0.00 0.0 0.00 0.5 3.02
Juncus wallichianus 10.9 20.5 0.0 0.00 0.0 0.00
Schoenoplectiella hondoensis 5.4 11.5 0.0 0.00 0.0 0.00
Schoenoplectiella triangulata 0.0 0.00 1.2 4.47 0.0 0.00
Schoenoplectus tabernaemontani 55.3 78.6 1.2 4.5 0.0 0.00
Sparganium erectum 2.2 8.2 0.0 0.00 0.0 0.00
216 J. Morimoto et al.

Number of species

Humus Clay Peat

Fig. 13.4 Number of species (/0.025 m2 ) from each soil layer. Open circles indicate outliers; the
top, center, and bottom lines of each box indicate the third, second, and first quartiles, respectively.
The horizontal lines above and below the box indicate the maximum and minimum values,
respectively (refer to Fig. 13.1 for the position of each soil layer)

(Baskin and Baskin 2014). Long-lived seeds of more than 5 years have been reported
occasionally (e.g., 100 years in Malvaceae, 80 years in Leguminosae, and 77 years in
Lamiaceae), but reports of long-lived seeds exceeding 1000 years are very rare (Oga
1951). The germination of three species of Juncus and Cyperus from the ancient peat
layer indicates their uniqueness. The conditions under which seeds can be stored for
long periods of time are said to be cool, mildly alternating temperatures, and low
oxygen (Baskin and Baskin 2014). These conditions might be easily possible in
Crane Pond, which was the bottom of a lake until land reclamation for agriculture
in the 1960s.

13.4 Natural Succession in Crane Pond

13.4.1 Features of Species Generated from Different Water

Depths

Plant succession at the field experimental site is shown in Table 13.2. The number of
species was higher than that contained in the soil layers (Table 13.1). This indicates
that the number of samples was limited for the survey of buried seeds and that
propagules dispersed by wind, water, and birds from surrounding agricultural land
13 Natural Succession of Wetland Vegetation in a Flood-Control Pond. . . 217

contributed to plant establishment at the experimental site. In addition, it was found

that plants were first established in the shallow water stage and that plants with
different life forms were established depending on the water depth (Table 13.2).
Annual plants that emerged in the first year did not appear after the second year.
The occurrence of alien species was negligible.
Hygrophytes and emergent macrophytes mainly emerged in the upper and
middle sections half a year after the excavation; however, it took 3 years for
plant establishment in the lower section, and floating-leaved macrophytes and
submerged macrophytes emerged. Hygrophytes and emergent macrophytes are
mainly dispersed by wind, whereas floating-leaved macrophytes and submerged
macrophytes are mainly dispersed by water or birds (Kameyama and Ohara 2007;
Mikulyuk and Nault 2009; Phartyal et al. 2018). It can be considered that the
vegetation recovery in the lower section was slower than that in the upper and
middle sections because the opportunities for dispersal by water inflow and bird
visits to Crane Pond were more limited than those by wind.
Hygrophyte species that occurred in the first year were annual plants dependent
upon disturbance (Euliss et al. 2004). Many of them were presumed to have not
occurred after the second year because flood disturbance did not occur in the field
experimental site. Human or natural disturbance is necessary to sustain disturbance-
dependent species.
Mesophytes, including alien species, disappeared by the second year of the
project, which means that there is no need to worry about alien species of
mesophytic plants invading the water body as long as the wet environment is
maintained.

13.4.2 Years from Excavation, Water Depth, and Water Quality

Determine the Plant Community

From the field experiments, it was found that the native species occur in order from
shallow to deep water, and their life forms depend on the water depth. Next, when
we looked at plant establishment throughout the entire Crane Pond, 26 native species
(including five rare species) appeared, which is more than that in the experimental
site, and no exotic species appeared (Kuroyanagi et al. 2019). The 13 unique species
that did not appear in the experimental site and in the seed bank appeared in Crane
Pond (Fig. 13.5). Crane Pond is vast and includes areas where the water was deeper
than that at the experimental site, and because of the large amount of incoming
water and waterfowl flights, it was suitable for the establishment of floating-leaved,
free-floating aquatic, and submerged macrophytes. Among the species that have
contributed to vegetation recovery in Crane Pond, there are three species that are
the same as those found in the buried seeds. Wetland vegetation in Crane Pond
regenerated mainly from propagules supplied from outside Crane Pond, but buried
218 J. Morimoto et al.

Table 13.2 Changes in plant species observed at the field experimental site
Life Life Red list or
Sitea formb historyc Blue listd Species The year of presence
2015 2016 2017 2018 2019
U m A Persicaria longiseta •
A,B  Gnaphalium uliginosum •
m-wet P  Trifolium spp. •
P  Rudbeckia laciniata •
P Juncus tenuis • • •
h A Persicaria sagittata var. • • • • •
sibirica
A Echinochloa crus-galli • • •
var. crus-galli
A Echinochloa crus-galli • •
var. aristata
A Cyperus difformis • •
A *** Persicaria foliosa var. • •
paludicola
A Persicaria lapathifolia •
A Juncus bufonius •
A Persicaria muricata •
P Lythrum salicaria • • •
P  Phalaris arundinacea • • •
P *** Carex capricornis • • •
T Salix udensis • • • • •
T Salix schwerinii • •
T Salix triandra subsp. • • •
nipponica
T Salix integra •
e,h A ** Monochoria korsakowii •
P Alisma • • • • •
plantago-aquatica var.
orientale
P Juncus decipiens • • • • •
P Scirpus radicans • • • • •
P Juncus wallichianus • • • • •
P Schoenoplectiella • • • •
triangulata
P Juncus prismatocarpus • •
subsp. leschenaultii
P Alisma canaliculatum • •
P Cicuta virosa • •
e A Schoenoplectus hotarui • •
P Schoenoplectus • • • • •
tabernaemontani
(continued)
13 Natural Succession of Wetland Vegetation in a Flood-Control Pond. . . 219

Table 13.2 (continued)

Life Life Red list or
Sitea formb historyc Blue listd Species The year of presence
2015 2016 2017 2018 2019
P Typha latifolia • • • • •
P Zizania latifolia • • • •
P Bolboschoenus fluviatilis • • •
subsp. yagara
P * Sparganium erectum • •
P Schoenoplectus triqueter • • •
P Eleocharis mamillata • • • •
P Bolboschoenus •
koshevnikovii
fr A Lemna aoukikusa subsp. •
aoukikusa
fl A Trapa japonica •
s, fl P Potamogeton octandrus •
var. octandrus
M e P Typha latifolia • •
P Zizania latifolia •
fl A Trapa japonica • •
s, fl P Potamogeton octandrus • •
var. octandrus
fr,s P * Utricularia japonica •
L fl A Trapa japonica • •
fr A Lemna aoukikusa subsp. •
aoukikusa
s, fl P Potamogeton octandrus • •
var. octandrus
fr,s P * Utricularia japonica •
aU upper section, M middle section, L lower section
b Life form is identified for wetland plants and others based on the description from picture books.
h hygrophyte, e emergent macrophyte, fl floating-leaved macrophyte, fr free-floating aquatic
macrophyte, s submerged macrophyte, m mesophyte, m-wet methophyte that can live in a wet
environment
c A annual plant, B biennial plant, P perennial plant, T tree
d We denoted the species categories of national (Japan) and regional (Hokkaido Prefecture) red

lists according to Japanese Red lists (Ministry of Environment of Japan, 2017) and Hokkaido
Red lists (Hokkaido Prefecture, 2001, 2017, 2018), respectively. The categories of Japanese Red
lists (2017) are EN (Endangered), VU (Vulnerable), and NT (Near Threatened). The categories
of Hokkaido Red lists (2001) are EN (Endangered), VU (Vulnerable), and R (Rare). We classified
species as alien species according to Hokkaido Blue lists (Hokkaido Prefecture, 2010). ***, VU/R;
**, NT/VU; *, NT/R; , blue list
 220

P.arundinacea
C.capricornis
S.udensis
Experimental site (21) Crane pond (26) A.plantago-aquatica var. Orientale
T.latifolia
Z.latifolia
B.fluviatilis subsp. yagara
E.mamillata
L.aoukikusa subsp. Aoukikusa

P.longiseta
P.foliosa var. paludicola
P.lapathifolia
S.triangulata S.miyabeana
P.sagittata var. Sibirica S.tabernaemontani M.korsakowii
E.crus-galli var. Aristata C.virosa
S.triandra subsp. Nipponica S.trifolia
A.canaliculatum P.australis
S.erectum
S.triqueter T.japonica
U.japonica
P.berchtoldii
P.octandrus var. octandrus
Mesophyte
H.verticillata
Hygrophyte P.Longiseta
Hygrophyte, C.Difformis J.bufonius
emergent J.decipiens J.prismatocarpus subsp. leschenaultii
J.wallichianus S.triangulatus
Emergent
Floating-leaved,
Free-floating
Submerged
Seedbank (10)
Fig. 13.5 Species in Crane Pond, the experimental site, and the seed bank. The numbers in the figure are the numbers of species
J. Morimoto et al.
13 Natural Succession of Wetland Vegetation in a Flood-Control Pond. . . 221

䕿 Trapa japonica community

CCA2

䕧 Schoenoplectus tabernaemontani -Typha latifolia

WD 䠇 S. tabernaemontani community
㽢 Zizania latifolia community
䕻 Z latifolia -Bolboschoenus fluviatilis community
pH
YR 䕱 B. fluviatilis community
EC
䖶 Phragmites australis community

CCA1

Fig. 13.6 Relationships between plant communities and environmental factors. WD, water depth;
YR, years since excavation; EC, electrical conductivity

seeds also played an important role in maintaining the species diversity of wetland
vegetation in Crane Pond.
Six types of plant species communities were present: the Schoenoplectus
tabernaemontani-Typha latifolia community, Schoenoplectus tabernaemontani
community, Zizania latifolia community, Z. latifolia-Bolboschoenus fluviatilis
community, B. fluviatilis community, and Phragmites australis community. The
factors characterizing the habitat of plant communities were the time since
excavation, water depth, and water quality (Fig. 13.6). In Crane Pond, where 3–
7 years have passed since excavation, floating-leaved Trapa japonica has developed
in deeper water, and emergent macrophytes such as Zizania latifolia and Phragmites
australis have developed in shallower water. These results coincide with those of the
experimental site. Eutrophication, expressed as increasing EC, tended to increase
with years since excavation, and the transition to a reed (Trapa japonica) community
was observed (Fig. 13.6). These results suggest that creating various depths of water
inside a flood-controlled basin and controlling the progress of eutrophication are
effective in maintaining the various wetland species and plant communities.

13.5 Implications for Management of the Flood-Control

Basin as GI

There are many cases in which past wetland vegetation did not recover on fallow
land (e.g., González et al. 2017; Sheldon et al. 2016). The initial requirements
for the recovery of wetland vegetation are the supply of propagules of wetland
plants (Bischoff et al. 2009; Vécrin et al. 2002) and a suitable habitat for their
222 J. Morimoto et al.

growth and reproduction (Morimoto et al. 2017; Weiher and Keddy 1995). What
kind of management is necessary for a flood-control basin built on fallow land to
function as GI, conserve wetland ecosystems, and control floods? We will discuss
the management of the supply of propagules and the management of suitable
habitats.

13.5.1 Management of the Supply of Propagules

We can expect two types of propagules in the flood-control basin, buried and
dispersed propagules. Our study showed that buried seeds were preserved even in
the deep ancient soil layers that had accumulated at the time of the floodplain, and
these seeds can be utilized for wetland restoration. A persistent seed bank functions
as a gene pool (Bossuyt and Honnay 2008). Effective utilization of soil containing
ancient seeds will enable wetland restoration considering genetic diversity. On the
other hand, propagule dispersal from the surroundings is an essential requirement
for the restoration of wetland vegetation with high diversity because the species that
can be supplied from buried seeds are limited. Waterways running in an agricultural
landscape function as refuges for wetland plants (Toyoshima et al. 2020). It is
important to conserve these waterways as propagule sources for flood-control basins
established in agricultural landscapes by organic farming or farming with low levels
of agricultural chemicals.

13.5.2 Management for Suitable Habitat

Excavation is a general practice for the construction of flood-control basins, and

the removal of surface soil is a general practice for the restoration of wetlands.
Both practices have a similar effect on suitable habitat creation. Such practices
increase the number of vacant sites, which permit the establishment of new plants
(Klimkowska et al. 2010), decrease soil nutrients (Verhagen et al. 2001), increase
the water level (Galatowitsch and van der Valk 1996; Klimkowska et al. 2010), and
increase flood frequency (Hölzel and Otte 2004). However, both excavation and the
removal of the surface soil will decrease the availability of seed banks in the soil,
making these practices not a good idea if the propagule source is limited in the
vicinity. Instead, it is recommended that drainage facilities be removed to raise the
groundwater level and create a waterlogged environment to promote the effective
use of seed banks (Morimoto and Shibata 2018). The reuse of topsoil removed
during excavation can be an option, although keeping the waterlogged environment
is a requirement to prevent the domination of alien species. In addition, a design
that creates a variety of depths within the flood-control basin would contribute to
the species diversity of plants, as it would promote the establishment of plants with
different life forms suitable for different depths. After the construction of a flood-
control basin, controlling the eutrophication of water from the surrounding farmland
13 Natural Succession of Wetland Vegetation in a Flood-Control Pond. . . 223

and removing the sediment soils to maintain the water depth should be effective for
the conservation of wetland vegetation. Ideally, river floods should be large enough
to occasionally wash away the accumulated sediment in a flood-control basin.
One of the important attractions of GI is that it does not require or reduces
maintenance costs. Even before the construction of the flood-control basin, it is
essential to select a site and design a structure that aims to regenerate and maintain
the wetland ecosystem by natural processes for the flood-control basin to function
as GI.

Acknowledgments We appreciate Ms. Mitsuko Goto (The Asian Natural Environmental Science
Center, The University of Tokyo) for carrying out the DNA analyses and Dr. Kazuo Yabe (Sapporo
City University) for providing advice on the succession of wetland vegetation.

References

Baskin CC, Baskin JM (2014) Seeds: ecology, biogeography, and, evolution of dormancy and
germination. Academic Press, San Diego, CA
Baumann M, Kuemmerle T, Elbakidze M, Ozdogan M, Radeloff VC, Keuler NS, Prishchepov AV,
Kruhlov I, Hostert P (2011) Patterns and drivers of post-socialist farmland abandonment in
Western Ukraine. Land Use Policy 28:552–562
Benayas JMR, Bullock JM (2012) Restoration of biodiversity and ecosystem services on agricul-
tural land. Ecosystems 15:883–899
Benjamin K, Domon G, Bouchard A (2005) Vegetation composition and succession of abandoned
farmland: effects of ecological, historical and spatial factors. Landsc Ecol 20:627–647
Bischoff A, Warthemann G, Klotz S (2009) Succession of floodplain grasslands following
reduction in land use intensity: the importance of environmental conditions, management and
dispersal. J Appl Ecol 46:241–249
Bossuyt B, Honnay O (2008) Can the seed bank be used for ecological restoration? An overview
of seed bank characteristics in European communities. J Veg Sci 19:875–884
CBOL Plant Working Group (2009) A DNA barcode for land plants. Proc Natl Acad Sci U S A
106(31):12794–12797
Euliss NH, LaBaugh JW, Fredrickson LH, Mushet DM, Laubhan MK, Swanson GA, Winter
TC, Rosenberry DO, Nelson RD (2004) The wetland continuum: a conceptual framework for
interpreting biological studies. Wetlands 24:448–458
Galatowitsch SM, van der Valk AG (1996) The vegetation of restored and natural prairie wetlands.
Ecol Appl 6:102–112
González E, Masip A, Tabacchi E, Poulin M (2017) Strategies to restore floodplain vegetation after
abandonment of human activities. Restor Ecol 25:82–91
Hölzel N, Otte A (2004) Assessing soil seed bank persistence in flood-meadows: the search for
reliable traits. J Veg Sci 15:93–100
Kameyama Y, Ohara M (2007) Population genetic consequences of predominant clonal reproduc-
tion in free-floating aquatic bladderworts. Jpn J Ecol 57:245–250
Klimkowska A, Kotowski W, Van Diggelen R, Grootjans AP, Dzierz̈a P, Brzezińska K (2010)
Vegetation re-development after fen meadow restoration by topsoil removal and hay transfer.
Restor Ecol 18:924–933
Kuemmerle T, Hostert P, Radeloff VC, van der Linden S, Perzanowski K, Kruhlov I (2008) Cross-
border comparison of post-socialist farmland abandonment in the carpathians. Ecosystems
11:614
Kuroyanagi A, Morimoto J, Shida Y, Shinsho H, Yabe K, Nakamura F (2019) Recovery of wetland
vegetation in the Maizuru flood-retarding basin in the Chitose River watershed. J Jpn Soc Reveg
Technol 45:45–50
224 J. Morimoto et al.

Middleton BA (2003) Soil seed banks and the potential restoration of forested wetlands after
farming. J Appl Ecol 40:1025–1034
Mikulyuk A, Nault ME (2009) Water chestnut (Trapa natans): a technical review of distribution,
ecology, impacts, and management. Wisconsin Department of Natural Resources Bureau of
Science Services, Madison, WI
Morimoto J, Shibata M (2018) Vegetation succession on fallow land. In: Nakamura F (ed)
Biodiversity conservation using umbrella species: Blakiston’s fish owl and the red-crowned
crane. Springer Singapore, Singapore, pp 197–206
Morimoto J, Shibata M, Shida Y, Nakamura F (2017) Wetland restoration by natural succession in
abandoned pastures with a degraded soil seed bank. Restor Ecol 25:1005–1014
Nakamura F, Ishiyama N, Yamanaka S, Higa M, Akasaka T, Kobayashi Y, Ono S, Fuke N,
Kitazawa M, Morimoto J, Shoji Y (2020) Adaptation to climate change and conservation of
biodiversity using green infrastructure. River Res Appl 36:921–933
Oga I (1951) The longevity of seeds – my dream – my longing – excavation of old lotus seeds in
the Kemigawa Peatland, Chiba Prefecture. Collect Breed 13:206–207
Phartyal SS, Rosbakh S, Poschlod P (2018) Seed germination ecology in Trapa natans L., a widely
distributed freshwater macrophyte. Aquat Bot 147:18–23
Sheldon CJ, Ficklin RL, Fawley KP, Fawley MW, Bataineh M, Nelson AS, Wilson S (2016)
Vegetation diversity in natural and restored forested wetland sites in Southeast Arkansas. J
Ark Acad Sci 70:37
Toyoshima G, Morimoto J, Nakamura F (2020) The role of waterways in agricultural land
landscapes in preserving the diversity of aquatic and epiphytes. J Jpn Green Eng Soc 46:15–20
Tsubota H, Inoue Y, Nakahara-Tsubota M, Shimamoto T, Matsuda I, Uchida S, Mukai S (2014)
Combination of DNA barcoding and molecular phylogenetic analysis as an identification tool
for plant species: a case study using imperfect herbarium specimens. Hikobia 16:475–490
Vécrin MP, Diggelen R, Grévilliot F, Muller S (2002) Restoration of species-rich flood-plain
meadows from abandoned arable fields in NE France. Appl Veg Sci 5:263–270
Verhagen R, Klooker J, Bakker JP, Diggelen R (2001) Restoration success of low-production plant
communities on former agricultural soils after top-soil removal. Appl Veg Sci 4:75–82
Weiher E, Keddy PA (1995) The assembly of experimental wetland plant communities. Oikos
73:323–335
World Water Assessment Programme (2009) The United Nations World Water Development
Report 3: Water in a Changing World. UNESCO, Paris; London: Earthscan
Zedler JB, Kercher S (2005) WETLAND RESOURCES: status, trends, ecosystem services, and
restorability. Annu Rev Environ Resour 30:39–74

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 14
Biodiversity Conservation through
Various Citizen Activities in a Flood
Control Basin

Jun Nishihiro, Shohei Koike, and Atsushi Ono

Abstract Flood control basins (FCBs) are artificially constructed facilities with
the potential to conserve the biodiversity of floodplain ecosystems. However, the
intensity of disturbance in FCBs is generally lower than that in natural floodplains;
thus, habitats for plants with disturbance-dependent life histories and for animals
that use temporary waters or mudflats can be lost in the years following FCB
construction. Here, the Asahata FCB in Shizuoka city, Japan, was studied as an
example, where the species diversity of wetland plants has been conserved as a result
of diverse activities. Although most activities had objectives other than biodiversity
conservation, they contributed to conserving plant diversity and providing habitats
for endangered plants. The FCB is a green infrastructure that not only enables flood
control but also supports activities for various purposes. Additionally, its proper use
contributes to biodiversity conservation.

Keywords Biodiversity · Citizen participation · Education · Endangered plants ·

Welfare

14.1 Introduction: Floodplain and Flood Control Basin

14.1.1 Floodplain

In lowland plains, rivers that have not been artificially modified meander and change
the channel morphology during flooding events (Bridge 2003). Such dynamics of

J. Nishihiro ()
National Institute for Environmental Studies, Ibaraki, Japan
e-mail: [email protected]
S. Koike
Toho University, Tokyo, Japan
A. Ono
Showa Sekkei Co., Ltd., Osaka, Japan

© The Author(s) 2022 225

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_14
226 J. Nishihiro et al.

the river channel result in a floodplain. Because the intensity of flood disturbances
varies among locations within a floodplain, a mosaiclike ecosystem composed of
heterogeneous elements is formed, providing a habitat for diverse flora and fauna
(Junk et al. 1989; Tockner et al. 1998; Ward et al. 1999; Robinson et al. 2002;
Tockner and Stanford 2002).
In recent years, the area and biodiversity of floodplains have decreased world-
wide (Tockner and Stanford 2002; Ishiyama et al. 2017). In Japan, which is
mostly mountainous, floodplain areas have been targeted for land use development.
Many major Japanese cities are located on alluvial plains at low elevations, which
constituted the seabed during the Holocene glacial retreat. Furthermore, these
alluvial plains are suitable for modern rice cultivation; therefore, large areas of the
plain have been converted into modernized agricultural land. In addition to land
conversion, flow control by dams in the upper reaches of rivers causes biodiversity
loss in floodplains. The reduction in flood disturbance in turn reduces the habitat
areas of plants with life history characteristics that depend on disturbance and of
animals that exploit temporary water (Nakamura et al. 2006).

14.1.2 Potential of Flood Control Basins

To conserve floodplain biodiversity, it is important to preserve areas affected by

flood disturbances. In flood control basins (FCBs), the biodiversity of floodplain
ecosystems can be conserved, even in Japan, where the population density is high
and lowland plains are often subject to development. An FCB is an artificially
constructed facility intended to reduce river flow due to rainfall occurring during
peak floods (Fig. 14.1). Although FCBs are designed to reduce flood risk, they have
similar hydrological characteristics to natural floodplains, i.e., they are temporarily
inundated only during floods. Therefore, they can provide habitats for species with
life history characteristics that depend on disturbances caused by floods.
There are 143 FCBs in Japan, with a total area of 15,104 ha (Suwa and Nishihiro
2020). However, not all of these factors may contribute to the conservation of
floodplain ecosystems. There are many differences between FCBs and natural
floodplains, a major difference being the intensity of the disturbance. In many
FCBs, only water that exceeds the overflow dike, which is built as a relatively
low part of the dike between the river channel and FCB, can enter the basin.
Thus, there is insufficient energy to disturb vegetation and soil. In some areas
of natural floodplains, the flood energy is sufficiently high to remove preexisting
vegetation; however, identification of such areas in FCBs is difficult. In stable
wetland environments, the biomass of aboveground vegetation commonly increases
over time, and a transition to woodland occurs through vegetation succession (Salo
et al. 1986). This change can be rapid because the nutrient input is high in FCBs as a
consequence of sedimentation. Thus, habitats for plants with disturbance-dependent
life histories and for animals that use temporary waters or mudflats can be lost within
only a few years of FCB construction.
14 Biodiversity Conservation through Various Citizen Activities in a Flood. . . 227

Fig. 14.1 The functioning

Asahata flood control basin.
(a) Floodwaters overflowing
part of the levee (overflow
dike), entering the FCB. (b)
FCB during floods. (c) FCB
under normal circumstances
228 J. Nishihiro et al.

14.1.3 Utilization of Land in a Flood Control Basin

FCB functions to reduce flood risk but can also be used for sports, recreational
activities, environmental education, and paddy cultivation; additionally, their vege-
tation can be used as a source of materials for traditional thatched roofs. Appropriate
planning of these activities is expected to contribute to biodiversity conservation. If
so, disturbances caused by floods can be replaced by disturbances caused by human
activities.
However, not all disturbances have a positive effect on biodiversity. Local
biodiversity is high only under moderate disturbances (“intermediate disturbance
hypothesis,” Connell 1978). Excessive anthropogenic disturbances can lead to
a decline in biodiversity. Moreover, the intermediate hypothesis is that species
diversity within a spatial unit can be considered a homogeneous environment (α-
diversity). Higher β-diversity, defined as a greater heterogeneity between different
locations, is another characteristic of floodplain biodiversity. The effects of distur-
bance on FCB biodiversity should be evaluated in terms of both α- and β-diversity.
The Asahata FCB in Shizuoka city, Japan, was constructed to control floods in the
Tomoe River. Many groups and individuals, including nature conservation groups,
private companies, special schools for students with disabilities, and hospitals,
participate in activities in the FCB (Nishihiro 2018). We studied the effects of these
activities on the vegetation in the FCB. First, we provide background information
and discuss the research results regarding the relationship between these activities
and vegetation.

14.2 Asahata Flood Control Basin

14.2.1 Geomorphological Features and Changes in Land Use

The Tomoe River is 17.98 km long, and its watershed covers an area of 104.8 km2 .
The watershed comprises approximately 7.6% of the total area of Shizuoka city but
contains 47% of the total population (approximately 340,000 people). The lower
parts of the river basin are urbanized areas.
The Tomoe River Basin was a bay during the Holocene glacial retreat, and most
of the basin was flat. The FCB was built at the inflection point of the riverbed slope,
where the river leaves the mountains and reaches the low land (Fig. 14.2). Until
about 50 years ago, many swamps were scattered in this area (Yasumoto 1979). The
paddy fields of this swampy area were inundated throughout the seasons, and rice
productivity was low. Traditional fisheries for eels, loaches, shrimp, and carp were
established in the Edo era (the late 1800s; Yasumoto 1979). From 1959 to 1973,
a project was carried out to improve agricultural productivity, including wetland
drainage and paddy field enlargement.
14 Biodiversity Conservation through Various Citizen Activities in a Flood. . . 229

Asahata FCB

Asahata area

Altitude
~5m
5 ~ 10 m
10 ~ 50 m
50 ~ 100 m
1 km 100 ~ 500 m
500 m ~

Central Shizuoka

Fig. 14.2 Location of the Tomoe River (black line of upper panel) and location of the Asahata
flood control basin. The lower panel shows Google Earth Image (March 2021 image)
230 J. Nishihiro et al.

Although this project greatly increased the rice productivity of agricultural land,
the discharge capacity of the Tomoe River itself was low, and the risk of flooding
in the watershed remained high. On July 7 and 8, 1974, Typhoon #8 caused heavy
rain, leading to 41 deaths, the destruction of 32 houses, and above- and underfloor
inundation of 11,981 and 14,143 houses, respectively. Because of this damage, a
hydraulic control plan was established for the Tomoe River Basin in 1978, which
included the construction of the Asahata FCB.

14.2.2 Activities in the Asahata Flood Control Basin

In the early 2000s, during the construction of the FCB, various plants, including
the endangered disturbance-dependent species Monochoria korsakowii, grew exten-
sively in the FCB due to disturbances caused by the construction work. The Nature
Restoration Council, a local council for ecosystem management established under
the Act to Promote Nature Restoration, was launched in 2003. The council initially
focused on activities aimed at conserving biodiversity, of which the primary focus
was the maintenance of the habitat of endangered plants. However, the range of
activities was limited, and the number of participants did not increase over time.
In 2014, several new groups started to use the Asahata FCB, including people
with disabilities attending special schools located near the FCB and staff from two
local hospitals. New activities were encouraged by executives of the prefectural
department and hospitals. Paddy fields were created to provide opportunities of job
training to students with disabilities. With the provision of activities aimed mainly at
welfare and education, local private companies and nongovernmental organizations
also participated; thus, the total number of organizations participating in the Nature
Restoration Council increased rapidly (Fig. 14.3). The number of participants in the

80 Group
஄ର Individual
‫ݺ‬ਕ

60
# Participants

40

20

0
2004 2006 2008 2010 2012 2014 2016 2018

Fig. 14.3 Number of group and private participants of the Nature Restoration Council, Asahata
flood control basin
14 Biodiversity Conservation through Various Citizen Activities in a Flood. . . 231

council was about four times larger in the period when groups whose main purposes
were welfare and education were included in the council than that in the period when
it was composed of groups whose main purpose was biodiversity conservation.

14.3 Effect of Activities on Plant Diversity

14.3.1 Activities

We identified nine types of activities that caused anthropogenic disturbances in

the Asahata FCB (Table 14.1, Fig. 14.4). Only one activity, management of M.
korsakowii habitat, had biodiversity conservation as its primary purpose (Table
14.1). Other activities included the management of paddy fields for students with
disabilities, as mentioned above, as well as conducting classes for students with
disabilities; focusing on the study of soil seed banks; transplanting plants to other
parts of the retarding basin; excavating a pond to increase the storage capacity of the
FCB, collecting Carex dispalata and Zizania latifolia to make traditional grass hats
and grass mats, respectively; controlled burning to prevent vegetation succession;
and mowing to make firebreaks to facilitate controlled burning.
Some activities had biodiversity-related objectives. For example, an educational
program for studying soil seed banks was conducted as follows:
1. The students attended a lecture on plant diversity and soil seed banks.
2. The students collected soil samples from the FCB and spread them in containers
placed in the school garden.
3. The plants that emerged from the soil samples were checked and identified
throughout the year, and the flora of the seed bank was recorded.
4. After the experiment, the soil samples containing high-density seed banks (newly
dispersed seeds were added from the plants that emerged in the soil) were moved
and spread in the FCB.
The main objective of this program was to educate students, but the content of the
training program was closely related to the conservation of plant diversity. In another
case, while the main purpose of excavating the pond was to increase the capacity
for water retarding, biodiversity was also considered. In contrast, the activities
involving collection of plant materials did not consider biodiversity conservation
as an objective. Overall, there was a high diversity among activity objects.

14.3.2 Vegetation

The activities listed in Table 14.1 can have different impacts on the ecosystems of
the FCB. Of the nine activities, four involved only mowing aboveground vegetation,
 232

Table 14.1 Activities and their location in the Asahata flood control basin. Stars and circles represent the main actor and participants, respectively
Stakeholder
Nature School for
conservation Welfare people with Private Local Disturbance
Activities Main objective NGO NGO disabilities company Researchers government Citizens regime
(a) Spreading Environmental ◦  ◦ Management of
soil containing education aboveground
seeds vegetation and soil
disturbance
(b) Conservation Biodiversity  ◦
of endangered conservation
plants
(c) Transplant of Public 
Zizania construction
work
(d) Paddy fields Welfare and   ◦ ◦
education
(e) Pond Flood control 
construction
(f) Carex Traditional hat   Management of
collection making aboveground
vegetation
(g) Zizania Traditional mat 
collection making
(h) Control Vegetation ◦ ◦  ◦
burning management
(i) Mowing Fire stopping ◦ ◦  ◦
J. Nishihiro et al.
14 Biodiversity Conservation through Various Citizen Activities in a Flood. . . 233

Fig. 14.4 Various activities in Asahata flood control basin. (a) Rice planting in a paddy field by
students from a special-needs school. (b) Soil with high density of seed banks is being returned
to the FCB by students from a special-needs school. (c) Collection of Zizania to make traditional
rush mat. (d) Winter burning with the participation of many citizens

while five included both mowing and disturbing the soil. We conducted a vegetation
survey to assess the impact of these activities on vegetation in the Asahata FCB.
A 10 m2 plot was set up in each area where the activities listed in Table 14.1 were
conducted, and all vascular plants were recorded. A similar survey was conducted in
a control plot in an area adjacent to each activity site. The control plots were similar
in terms of altitude and soil origin.
In total, 97 plant species were recorded, of which 77 were native species
(including 53 wetlands or aquatic species). Of these 97 species, 58 (60%) were
observed only in plots that underwent activities (Fig. 14.5). In addition, five species
on the Japanese Red List were recorded, all of which were found only in plots with
activities. There were more species in the activity than in the control plots in areas
where the soil was disturbed, with one exception (Fig. 14.6). Although the number
of species was small in this exceptional plot, submerged plants were found only
234 J. Nishihiro et al.

Fig. 14.5 Endangered species observed in the sites with activities: Monochoria korsakowii,
Penthorum chinense, and Persicaria conspicua (from left to right). The numbers of species (sp.) in
sites with activities and/or without activities (control) are also shown

40
Aquatic plants Wetland plants

Terrestrial plants Unknown

30
Alien plants

20

10

0
C A C A C A C A C A C A C A C A C A C A C A
(d) (c) (d’) (b) (b’) (a) (e) (f) (g) (h) (i)
Management of above ground vegetation Management of above
& Soil disturbance ground vegetation

Fig. 14.6 Number of species observed in each site with activity (A) and control (C). The activities
(a) to (i) correspond with these in Table 14.1. (b’) and (d’) represent the ridge of the paddy (b) and
(d), respectively

in the plot. In other words, highly specific vegetation, which contributes to the β-
diversity of flora in Asahata FCB, was established in this plot.
Activities in the Asahata FCB contributed to the maintenance of plant diversity.
Most of the observed plants were wetlands and aquatic plants, suggesting that
activities in the Asahata FCB facilitated the conservation of floodplain vegetation.
Anthropogenic disturbances due to activity in the FCB might partly replace natural
14 Biodiversity Conservation through Various Citizen Activities in a Flood. . . 235

disturbances, although it will be necessary to compare the vegetation of FCB to that

of a natural floodplain to confirm this hypothesis.
In many cases, the number of alien plants was also higher in the activity plots than
that in the control plots, suggesting that disturbance can increase the opportunity
for the invasion of many plants. Controlling alien species can be an issue in the
management of biodiversity conservation.

14.4 Generality of the Results

This study examined activities in the Asahata FCB from the viewpoint of biodi-
versity conservation. However, these results cannot be applied to all FCBs. The
floodplain-like vegetation observed here is thought to have been derived from the
soil seed banks of plants that had grown in paddy fields and swamps before the
FCB was built. The topsoil is removed in most FCBs to maintain the capacity for
retarding floodwater. However, in the construction of the Asahata FCB, the topsoil
from the paddy fields was first secured in another location and then returned to
the surface of the FCB after construction. This was thought to have enabled the
establishment of vegetation with a high species diversity. The use of topsoil is
particularly important for the conservation of plant diversity in FCBs.
Our study showed that the utilization of wetlands usually increases α-diversity.
However, not all activities in FCBs facilitate conservation, and some might signifi-
cantly reduce biodiversity. For instance, under conditions of high invasion pressure
from invasive alien species, disturbance can lead to a decrease in biodiversity. To
prevent such situations, it is important to assess their impact before implementing
activities. We consider that the Asahata FCB is being used appropriately, at least for
the moment, owing to the consensus built among various stakeholders, including
ecology experts.

14.5 Activity Redundancy and Conservation Sustainability

A shortage of participants may lead to discontinuation of the activities in the future,

namely, for the “Use of Zizania” and “Use of Carex” activities listed in Table 14.1.
In addition, the nature conservation group considers the aging of its members as
problematic. Societal changes can make it difficult to continue an activity, the scale
of which may have to be reduced. If an activity cannot be continued, the environment
that depends on it will disappear, thus decreasing the biodiversity of the FCB. To
maintain biodiversity under such uncertain circumstances, “activity redundancy,”
which refers to a variety of actors playing a similar role, is thought to be an important
factor in long-term natural resource management. It will be important to allow
activities with various purposes, not limited to biodiversity conservation, and to
scientifically monitor the effects of these activities. It will also be important to have
236 J. Nishihiro et al.

a mechanism to facilitate communication among participating groups and to help

each other when necessary.

Acknowledgments We sincerely thank Ms. Kuriyama (Shizuoka city) for her cooperation. This
study was partly supported by the Environment Research and Technology Development Fund
JPMEERF20202001 of the ERCA of Japan.

References

Bridge JS (2003) Rivers and floodplains: forms, processes, and sedimentary record. Blackwell
publishing, Oxford
Connell JH (1978) Diversity in tropical rain forests and coral reefs. Science 4335:1302–1310
Ishiyama N, Nagayama S, Iwase H, Akasaka T, Nakamura F (2017) Restoration techniques for
riverine aquatic connectivity: current trends and future challenges in Japan. Ecol Civ Eng
19:143–164. (in Japanese)
Junk W, Bayley PB, Sparks RE (1989) The flood pulse concept in river-floodplain systems. In
Dodge DP (ed). Proceedings of the international large river symposium (LARS). Canadian
special publication of fisheries and aquatic sciences 106, pp. 110–127
Nakamura K, Tockner K, Amano K (2006) River and wetland restoration: lessons from Japan.
Bioscience 56:19–429
Nishihiro J (2018) Efforts to balance the usage and conservation of biodiversity in Asahata
retarding basin, Shizuoka. National Parks 763:16–19. (in Japanese)
Robinson CT, Tockner K, Ward JV (2002) The fauna of dynamic riverine landscapes. Freshw Biol
47:661–677
Salo J, Kalliola R, Hakkinen I, Makinen Y, Niemela P, Puhakka M, Coley PD (1986) River
dynamics and the diversity of Amazon lowland forests. Nature 322:254–258
Suwa Y, Nishihiro J (2020) Distribution and location characteristics of flood-control reservoir in
Japan. Ecol Civ Eng 23:85–97. (in Japanese)
Tockner K, Stanford JA (2002) Riverine flood plains: present state and future trends. Environmental
Conservation 29:308–330
Tockner K, Shiemer F, Ward JV (1998) Conservation by restoration: the management concept for
a river-floodplain system on the Danube River in Austria. Aquatic Conservation: Mar Freshw
Ecosyst 8:71–86
Ward JV, Tockner K, Schiemer F (1999) Biodiversity of floodplain river ecosystems: ecotones and
connectivity. Regul Rivers: Res Manage 15:125–139
Yasumoto H (1979) History of Asahata. Shizuoka city office. (in Japanese)

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
 Part V
Urban and City Ecosystem
Chapter 15
Toward Holistic Urban Green
Infrastructure Implementation

Takanori Fukuoka

Abstract This article aims to reexamine green infrastructure (GI) visions and
frameworks and site-scale GI implementation toward holistic urban GI implementa-
tion. For GI visions and frameworks, both City of Philadelphia and Singapore have
demonstrated progressive approaches to weave different divisions and planning,
accomplishing holistic frameworks as well as enhancing GI implementation at site-
scale level. Through three site-scale GI implementation case studies, the creative
solutions have been demonstrated how streets, retention ponds, and the existing
urban fabric were to transform into GI open spaces. This article suggests the way
how GI could be interwoven into visions, planning, and design to create places for
people.

Keywords Urban green infrastructure · Visions and framework · GI

implementation · Open space

15.1 Toward Holistic Urban Green Infrastructure

Implementation

Green infrastructure (GI) implementation needs to integrate visions and frameworks

of the city scale with the ones of the site scale. In recent years, many practices have
demonstrated good GI implementation; however, there are gaps between city-scale
GI visions and site-scale GI implementation. Like many other concepts, the term
GI often creates the confusion of its definition. As per Environmental Protection
Agency (EPA) in the USA, GI is defined as “systems and practices that use mimic
natural process such as infiltration and evapotranspiration.”

T. Fukuoka ()
Faculty of Regional Environment Science, Department of Landscape Architecture Science,
Laboratory of Landscape Design and Geoinformatics, Tokyo University of Agriculture, Tokyo,
Japan
e-mail: [email protected]

© The Author(s) 2022 239

F. Nakamura (ed.), Green Infrastructure and Climate Change Adaptation,
Ecological Research Monographs, https://doi.org/10.1007/978-981-16-6791-6_15
240 T. Fukuoka

Section 502 of the Clean Water Act defines green infrastructure as “the range of
measures that use plant or soil systems, permeable pavement or other permeable
surfaces or substrates, stormwater harvest and reuse, or landscaping to store,
infiltrate, or evapotranspirate stormwater and reduce flows to sewer systems or to
surface waters” (United States Environmental Protection Agency n.d.).
This sustainable stormwater management system has become a big trend since
the early 2000s, and this system is introduced in the form of bioswales, green roofs,
rain gardens, and such small implementation. On the other hand, the European
Commission defines GI as “a strategically planned network of natural and semi-
natural areas with other environmental features that was designed and managed to
deliver a wide range of ecosystem services. It incorporates open spaces and other
physical features in terrestrial and marine areas. On land, GI is present in rural
and urban settings” (European Commission 2013). Though these two perspectives
indicate principles of GI, it is not clear how both concepts can be applied for
GI implementation. In other words, EPA expects GI put focus on solution, and
performance of urban ecological system, especially for stormwater. EU covers
much broader open spaces and natural systems as its framework. How can we
implement these perspectives on the ground? This section aims to reexamine
“visions and approaches” toward holistic urban green infrastructure implementation
with reference to some project case studies.

15.2 GI Visions and Frameworks: “Green City, Clean

Water” Citywide Green Infrastructure Implementation
Frameworks in the City of Philadelphia

City of Philadelphia, USA, challenged on creating a new type of the holistic

framework toward. strategic GI planning and implementation. Through a literature
review and the interviews with the City of Philadelphia Water Department GI
group, and by consulting “Green City, Clean Waters (GCCW)” to understand the
development and framework of GI planning, three phases of GI planning and
implementation were identified through the analysis of selected GI planning and
policy (Fukuoka et al. (2020)). The first phase (1990s–2008) can be described as
“Water quality control period.” Combined sewer overflow (CSO) was a big problem
in Philadelphia. Polluted water would be released directly to the river system after
extensive storm event. CSO Control Policy was implemented in 1994, and the long
term Control Plan by Water Department was released in 1997. In this phase, GI was
implemented as a site-scale project within Watershed Planning (Office of Water-
shed). The other approach was a grassroot West Philadelphia project directed by
Anne Spirn, professor at the University of Pennsylvania. West Philadelphia project
demonstrates very progressive approaches which cover stormwater management for
social issues and food production. Prof. Spirn collaborated with PWD (Philadelphia
15 Toward Holistic Urban Green Infrastructure Implementation 241

Water Department) and advocated GI at early stage (West Philadelphia Landscape

Project n.d.).
The second phase (2008–2012) can be described as “GI planning development
period.” Mayor Nutter (2008–2016) started it in 2008, and orchestrated very
powerful leadership in creating sustainable cities. Multiple Policies and Guidelines
such as Green Works (Sustainability Vision by Office of Sustainability), GCCW (GI
Plan by PWD), and Green Philadelphia (Open Space Plan by Park and Recreation)
are enforced from one after another.
The third phase (2012–) can be described as “GI implementation acceleration
period.” City of Philadelphia joined a partnership with EPA in 2012 to implement
GI further. Contents of GI partnership include GI model project implementation,
GI engineering development, and water quality research and communication. For
organizational structure, GI Planning group was created in 2017.
City of Philadelphia shifted GI implementation unit from watershed-based
implementation to district-scale implementation which is more effective to create
synergy between city planning and GI implementation. In addition, Green Street
Design Manual (PWD and Transportation), Stormwater Design Manual (PWD), and
Planning and Design Manual (PWD) had been issued in 2018. Now, cities challenge
to accelerate GI implementation further. GCCW performed key roles in setting GI
goals, visions, and frameworks.
As a summary, Philadelphia developed the holistic framework toward strategic
GI implementation. GI visions and frameworks set the big pictures and the essential
structures for diverse governmental divisions in order to move forward with GI
concepts. Those frameworks also help build up and activate a wide range of
site-scale GI projects from green streets, open spaces to urban redevelopment.
Especially, shifting GI framework scale from watershed-based one to district scale
helped to create a synergetic effort with city planning work (Fig. 15.1). It was rather
solo frame work before. The next phase for challenges aims to make this vision and
framework much broader so that it will include new types of GI such as vacant lots
or brownfields that are significantly increasing around the city periphery areas.

15.3 National-Scale Holistic GI Visions and Approaches:

“ABC Water Design Guidelines in Singapore”

In contrast to GI implementation in Philadelphia, Singapore challenged to create

national-scale holistic GI visions and frameworks with Design Guidelines. This case
study is an excellent example of sustainable stormwater management at a national
scale and may be applicable in other cities and nations in the similar monsoon
climate. Singapore is located at the heart of Southeast Asia, and is comprised of
63 islands. Singapore Island is 42 km long from east to west and 23 km from north
242 T. Fukuoka

Fig. 15.1 GI implementation in the City of Philadelphia in 2019. *GSI indicates green stormwater
infrastructure (Fukuoka et al. (2020) created it by using open Data PHL map Green Stormwater
Infrastructure Public Projects Points/Street data set)

to south. Historically, Singapore has relied on Malaysia for 40% of its drinking
water resource coming through pipelines. However, due to uncertain future with
water problems, Singapore government made a decision to become self-sufficient of
water by applying watershed-based management system. Especially, annual rainfall
(around 2340 mm) is targeted as the major water resource to be wisely used together
with gray water and the water stored in reservoirs. Public Utilities Board (PUB),
Singapore’s national water agency, has embarked on water management using the
15 Toward Holistic Urban Green Infrastructure Implementation 243

3P (People, Public, and Private) approach to take joint ownership of Singapore’s

water resource management.
This is embodied in PUB’s tagline – Water for All: Conserve, Value, Enjoy
(Public Utilities Board 2016a).
The Active, Beautiful, Clean Waters (ABC Waters) Program, launched by PUB
in 2006, is the cornerstone of the 3P approach. The program intends to transform
Singapore’s extensive network of reservoirs and water bodies into beautiful and
clean streams, rivers, and lakes, creating a vibrant City of Gardens and Water. More
than 100 potential locations will be identified as the site for the implementation of
the program by 2030 (Public Utilities Board 2016b). As of June 2014, 23 projects
had been completed (Public Utilities Board 2014).
The first edition of the ABC Waters Design Guidelines was launched in 2009
and the second edition in 2011. In June 2014, PUB upgraded the guidelines for
locally built examples to the ones for the showcase developers, architects, and
engineers who have incorporated the ABC Waters concept in their developments
(Public Utilities Board 2016c). ABC Waters Program helps implement ABC Waters
Design Guidelines and function based on GI as a hinge between visions and site-
scale projects. This new set of reference material aims to meet the industry’s
needs better and continue building up technical expertise in the industry (Kato and
Fukuoka 2016). First, the ABC Waters Design Guidelines provide actual stormwater
design tools for three different stages which are applicable to various types of
green infrastructure project scale. These stormwater design tools are categorized
into (1) catchment elements, (2) treatment elements, and (3) conveyance and
storage elements. First, catchment elements aim to collect water based on land
use typologies such as road, canal, water bodies, pedestrian walkways, public
open spaces, plazas, and buildings. For different surface conditions, appropriate
stormwater planning and design methodologies are shown with a clear illustration.
At the stage of project planning, research and concept were made based on the ABC
Waters, and location and volume of buildings were to meet ABC Waters’ goals,
and applicable green infrastructure implementation methodologies were introduced
at catchment stage. For example, building was divided into various catchment
elements such as green roofs, terraced green balconies at multi levels, and ground
level elements such as planted areas and water features. The ABC Waters Design
Guidelines provided engineering and design procedures so that the basic knowledge
and methodologies could be easily integrated into projects. Second, methodologies
how to implement treatment elements such as swales, bioretention pond, detention
pond, stormwater planter, rain garden, and cleaning biotope are explained through
visual information. Water treatment part covers a wide range of methods, and
this provides concept, benefits, and design methodologies for management issues.
Third, conveyance and storage elements focus on large water bodies and provide
methodologies such as bioengineering, erosion control, and water quality control.
244 T. Fukuoka

Second, in order to realize ABC Waters concept, ABC Waters Certification

and ABC Waters Professional Program were created to make loop of projects be
recognized and adopted by people (Kato and Fukuoka 2016). The ABC Waters
Certification launched by PUB on 1 July 2010 is a scheme to provide a recognition to
public agencies and private developers who embrace the ABC Waters concept and
incorporate ABC Waters Design Features in their developments (Public Utilities
Board 2016d). Besides providing the recognition, the scheme also aims to ensure
that the design features incorporated within the developments meet the minimum
design standard. Since 2010, 62 projects have been certified by the ABC Waters
Certification. Certification is judged in four categories: Active, Beautiful, Clean,
and Innovative. Out of total 110 possible points, for a project to be certified, it needs
to receive a minimum of 45 points with at least 5 points in each of the first three
categories. ABC Waters Professional Program is supported by multiple institutions
such as Institution of Engineers Singapore (IES) with the support of the Singapore
Institute of Architects (SIA) and Singapore Institute of Landscape Architects (SILA)
(Public Utilities Board 2016e).
As a summary, Singapore put force on setting nationwide Water Design Guide-
lines and provided both urban green infrastructure visions and frameworks at a
national scale. Singapore’s case study demonstrates “top-down” approaches and
sets effective schemes to translate GI visions to real projects by creating multiple
programs. Both Certificate and ABC Waters Professional Program help accelerate
GI implementation. As described above, two cases demonstrated differently, but
both are powerful in GI visions and frameworks toward implementation. In the next
section, three cases of site-scale GI implementation in urban settings have been
examined.

15.4 GI Approaches: Site-Scale GI Implementation

In this section, site-scale GI implementation project cases are introduced. GI visions

and frameworks often express broader ideas but remain vague. At the same time,
site-scale GI implementation tends to split into single bits of spaces such as
bioswales, green roof, and rain garden. As shown in Fig. 15.2, toward urban GI
implementation, each project still needs to contain a big picture about how each
piece fits to create the whole GI. Site-scale GI’s function is also difficult in setting a
good balance among multiple choices such as stormwater management, biodiversity,
heat mitigation, and others. Figure 15.3 illustrates GI goals with an envision. It
varies from “disaster reduction,” “healthy city” to “soft infrastructure, community.”
How can we set appropriate GI goals and clarify the function of GI required? In
this section, three site-scale GI projects are introduced to describe its character and
functionality.
15 Toward Holistic Urban Green Infrastructure Implementation 245

Park / Open Space

River

Vacant Land, Urban Farm

Urban Greening

Green Urban Development

Garden

Green Road / Street

Fig. 15.2 GI implementation in city scale

15.5 Street as GI: “City of Copenhagen’s New GI Street

Approaches”

City of Copenhagen, Denmark, illustrates an excellent GI street project adapting

to the climate change. In July 2011, cloudburst, heavy stormwater event, caused
a significant damage to the city. Extensive 150 mm of rainfall in 2 h left major
part of the old city area under 1 m of water, and the damage caused by that was
approximately 1 billion Euro worth (Fig. 15.4). Based on the research, the expected
flood damage caused by both cloudburst and expected sea level rise forced City
of Copenhagen to work on cloudburst master plan with blue-green infrastructure
in order to reduce the future risk of disaster (ASLA 2016). Existing historic
old city has no further spaces left without implementing GI to solve problems.
In this master plan, strategic approach such as utilizing road as GI street was
taken specifically for Copenhagen. Cloudburst master plan created frameworks to
accelerate GI implementation and there are three essential points. First, precautions
against cloudburst are developed based on “research.” The City of Copenhagen
worked on investigating multiple sets of data and identified high-risk areas. In
addition, they model and map the large-scale catchment base of stormwater and
246 T. Fukuoka

Disaster
Reduction

Healthy City
Cultural City
Walkable

Green
Infrastructure

Soft Infrastructure
Bio Diversity
Community

Quality of
Urban Life

Fig. 15.3 GI goals with an envision

visualize vulnerable areas as well as their risks. Second, GI function and GI

implemented places bridge over cloudburst by creating “Cloudburst Toolkit.” This
toolkit touches “design and quality” of spaces and aims to provide human-scale
experiences to the place. Hot spots for future GI implementation projects are
identified and rough images of designed places are illustrated according to the
typologies based on the toolkit. Third, socioeconomic cost-benefit analysis was
conducted, and it clarified that GI solutions have 50% of potential in saving over
conventional gray infrastructure or piped solutions. As explained above, these three
points helped utilize GI visions and frameworks for site-scale GI implementation.
15 Toward Holistic Urban Green Infrastructure Implementation 247

Fig. 15.4 Cloudburst caused significant damages to Copenhagen (Ramboll Studio Dreiseitl)

Figure 15.5 illustrates a basic concept for transforming road into GI street. Two
lanes 16 m wide road would be reduced into one lane road, and the rest of the space
was converted into planted areas and pedestrian walkways. Inner 11 m space has a
preventive function as a floodwater storage area to tackle once in 100-year storm
event. Planted area has the function to store stormwater underground and to provide
green and spaces for people. All GI streets had a plan to connect the stormwater
flow and projected frameworks to reduce the flood risk. Similar to Singapore’s
ABC Water Design Guidelines, future site-scale GI projects were identified, and
would be gradually implemented. As a summary, the cloudburst project suggests
one possibility to implement GI into existing city fabric by transforming existing
roads to GI. In 2018, City of Copenhagen selected 50 GI projects planned to
be built, and 5 projects already completed. The speed of GI implementation is
slower than expected due to the complicated coordination with existing underground
infrastructure. The other raised issue is how cloudburst can be evaluated as a climate
change adaptation project (Nakajima and Hoshino 2017). Especially, water quality
control is yet unsolved to meet with EU water quality framework. Lessons learned
from the cloudburst are how to set GI goals as well as evaluation framework to
create multifunctional space in the urban settings.
248 T. Fukuoka

Fig. 15.5 GI implementation proposals in cloudburst master plan (Ramboll Studio Dreiseitl)

15.6 Kashiwanoha Aqua Terrace: “Closed Retention Pond

to GI Open Space”

Kashiwanoha Aqua Terrace is the place that was redeveloped from flood control
pond into multifunctional GI open space. It is located in Kashiwa, northeast of
Tokyo, where 273 ha smart city is being developed. Aqua Terrace is a part of phase
2 developments with the concept of “Innovation Campus.” The retention pond was
formerly fenced around and had no access to the public and purely functioned
as a flood control pond. Site area is about 2.4 ha big and retention capacity is
approximately 73.720 m3 . Within 25 ha of phase 2 development, there was no
central open space but the existing retention pond (Fig. 15.6). Thus, the idea of
renovating closed retention pond to GI open space was strategically proposed.
To realize Kashiwanoha Aqua Terrace, the public-private partnership was formed
by Kashiwa City, Kashiwanoha Urban Design Center (UDCK) and Mitsui Fudosan,
a private developer. Prior to GI open space implementation, the master plan of
vigorous “Innovation Campus” was issued in 2008 to illustrate “District of Mixed
Usage” and “Walkable Street and Public Open Space Network” at the conceptual
level. In 2015, the landscape design for Aqua Terrace started to connect the existing
streets and the pond through applying seamless circulation loop, and multilevel
seating on the retention pond slope. 800-m-long jogging loop around the pond
provides a place for walking and jogging, while permeable paving provides cooling
surface in a hot summer time. Multilevel seating on slope provides places of diverse
15 Toward Holistic Urban Green Infrastructure Implementation 249

Fig. 15.6 Retention pond prior to redesign (Nikken Sekkei Ltd.)

activities for people and helps connect people and water. The challenge of Aqua
Terrace design was to control the fluctuating water levels. The water level is usually
kept at 40–80 cm; however, maximum of 4 m water level could be expected at
once in 50-year storm event around this designed area. Trees and site elements
were set carefully, and the lighting can endure against the floodwater. Instead
of installing physical fences or gates against the dangerous water level after the
storm event, lighting pole would blink red light when the water level has risen
more than 1 m (Fig. 15.7). GI discussion tends to set aside the power of design;
however, Kashiwanoha Aqua Terrace demonstrates an excellent design to weave
GI’s function and character into the place designed for people.
Regarding management of Kashiwanoha Aqua Terrace, UDCK plays a central
role to bring all players and volunteers together. Volunteers help with planting
management as well as cleaning up the walkway after the high water overflow.
Events vary from “Outdoor Movie Night” to natural environment observation event
for local children. Toward sustainable GI implementation, it is important to set
organizational structures in order to manage GI properly. Lastly, Kashiwanoha Aqua
Terrace represents a good GI open space model and illustrates the possibility of
public open space in the era of mature society.
250 T. Fukuoka

Fig. 15.7 Kashiwanoha Aqua Terrace as GI open space (Forward Stroke Inc.)

15.7 Minami-Machida Grandberry Park: Creating Livable,

Sustainable City with Open Spaces

15.7.1 GI Visions and Frameworks

The Minami-machida Grandberry Park is a complex project in Machida city,

southwest Tokyo. The train station, shopping mall, urban sports park, and Sakai
River waterfront were integrated under the concept of “park life.” The plan was
to create a walkable community with seamless connections between the station,
commercial facilities, parks, and the Sakai River. The inviting design starting from
the station through the mall up to the sports park promotes “active lifestyle.”
This complex is located in the suburban residential area that was developed about
40 years ago with the station placed in the south and a semi-industrial area in the
north. The urban redevelopment builds open spaces at its core. There are three
major sites. Grandberry Park (shopping mall, privately owned, 8.3 ha), park life site
(privately owned and operated facilities on former roads and public land, 0.5 ha),
and Tsuruma Urban Sports Park (publicly owned, 7.1 ha), and a necklace of 14 open
spaces in three large areas (Fig. 15.8). As illustrated, seven open spaces colored in
blue are privately owned open spaces, and the other seven open spaces colored in
green are publicly owned. In addition, yellow-green colored “park life site” is built
by public-private partnership. Each of the 14 plazas has a variety of designs and
unique atmosphere. For example, in the large Oasis Plaza, children can play on a
15 Toward Holistic Urban Green Infrastructure Implementation 251

Fig. 15.8 Minami-machida Grandberry Park illustrative plan (Machida-Shi, Tokyo)

100-meter-long water carpet in summer, and in winter a skating rink is set up for
multifunctional use.
How does this Minami-machida Grandberry Park function as GI? GI is strate-
gically conceptualized in its vision and frameworks as networked open spaces.
Formerly, the site was divided by the river, roads, urban sports park, shopping
mall, and station. City of Machida (public) and Tokyu Corporation (private) formed
public-private partnership in order to create a livable and sustainable city with
open spaces. The following are the key frameworks related to GI. First, “Walkable
City Network” is proposed from creating seamless connection of spaces through
the district. A big decision was made to close the existing city road and create a
6-m-wide pedestrian street, which enhanced the physical integrity of Grandberry
Park (shopping mall) and Tsuruma City Sports Park. In addition, fluid pedestrian
circulation allows to connect the north and the south district that were formerly
divided by railroad. Second, “Open Space Network” in both public and private land
was created by proposing 14 open spaces. This necklace like open space structure
allowed to promote not only walkability but lively urban lives. Third, “Sustainable
Development Goals” raised by challenging to obtain “LEED-ND Gold” as common
goals for all stakeholders. This vision helped implement sustainable stormwater
management and other site-scale GI later on (Fukuoka 2020).
252 T. Fukuoka

15.7.2 Design of Places and Public Engagement

By using visions and frameworks as guidelines, series of open spaces are designed
to enhance its concept and character. “Walkable City” visions are developed for
the broader “Active Design” concept. Active Design is a concept to support
development or design for creating healthy cities (Osamura and Fukuoka 2020).
In terms of health, when walkability improved, Tsuruma Park (Urban Sports Park)
would function as the core site where people can participate in sports and become
healthier. The 7.1 ha sports park provides diverse places for sports and the healthy
activities. This park has tennis courts (three omni courts), artificial turf pitches
(equivalent to three futsal pitches), athletic open space with playground (0.65 ha),
clubhouse (studio, café, lockers, and showers), and parking lot for 133 cars. Active
Design for promoting physical activities was implemented all over the park for the
purpose of running, muscle training, jumping, etc. and yoga, dance, karate, etc., and
classes are also open regularly at the studio (Fig. 15.9).
“Open Space Network” was implemented in various forms with unique character.
For example, in the large Oasis Plaza, children can play on a 100-meter-long water
carpet in summer, and in winter a skating rink is set up for multifunctional use.
On the line of flow from the station to the park, plants of mainly native species are
planted in connection with Tsuruma Park and Sakai River. Besides planting, there
are many ideas worked out to allow comfortable usage of outdoor spaces. Various
type of seating, shades, paving, and surrounding shops and façade create a setting
for “park life” for all generations.
Regarding the aspect of sustainability, sustainable rainwater management was
implemented as a part of Open Space Network. All rainwater at the station building
is stored underground for reuse, and a permeable pavement has been installed on
the site. The retention basin under the park is approximately 24,600 m3 . Sustainable
stormwater management was implemented throughout the project by installing rain
gardens and bioswales which promote temporary storage and infiltration.

Fig. 15.9 Tsuruma Urban Sports Park in Minami-machida Grandberry Park (left, before park
renovation; right: after park renovation)
15 Toward Holistic Urban Green Infrastructure Implementation 253

In the process of public engagement and participation, numerous workshops

and activities were held for the citizen during the planning and design phase.
Interestingly, citizen participation workshop was transformed into “Park School”
where citizens themselves plan and implement temporal events working with other
citizens of similar interests. In autumn, “Park School Festival” was successfully
held a few times, and the participation of the citizen transformed it into more
active, motivated activities in the park. In 2021, “Minami-machida-wo-minnna-
no-machie foundation” (Creating Minami-machida Town for All) was founded
by public-private partnership in order to proceed the city to the next stage. As a
summary, the Minami-machida Grandberry Park demonstrates how big their visions
and frameworks are and that the site-scale project was indeed integrated into a whole
GI open space network. In this case, both city and developer envisioned big pictures
such as “creating healthy, active cities” and “sustainable cities,” and those visions
realized the interconnected open spaces. In terms of public engagement, this case
illustrates how participation and engagement process was integrated into the part of
GI implementation in order to achieve a long term, sustainable management.

15.8 Toward Urban Green Infrastructure Implementation:

Open Space as GI

Through examination of “visions and approaches” (15.1–3) and “site-scale GI

project case studies” (15.4–7), the following are to recommend toward urban
GI Implementation. Firstly, GI visions and frameworks need to provide broader
perspectives from the heat mitigation in urban areas, reduction of water disaster to a
healthy and walkable city. Frameworks also need to illustrate rough spatial images
as well as functions of those spaces. City of Philadelphia demonstrated strategically
implemented GI visions and frameworks and activated multiple planning and
guidelines to accelerate GI implementation under very powerful leadership. On the
other hand, Singapore developed nation-scale, top-down GI visions and frameworks
in ABC Waters Design Guidelines. ABC Waters Design Guidelines set clear
goals and provide typological GI approaches as well as demonstrate GI with
actual model projects. Other good finding was circulative system of ABC Waters
Design Guidelines with “ABC Waters Certification” and “ABC Waters Professional
Program.” This framework allows sustainable GI implementation over time.
Secondly, site-scale GI implementation needs to create places for multifunction
and for people. The site-scale GI often tends to focus too much on monolithic
function related to the stormwater management such as rain garden and bioswale.
In Chap. 2, three GI implementation projects were introduced. Copenhagen’s
cloudburst demonstrates how “the strategies of climate change adaptation” were
introduced into the current urban fabrics through transforming roads into multifunc-
tional streets for all. Kashiwanoha Aqua Terrace developed interesting scheme to
renovate closed, monofunctional retention pond into GI open space for diverse users.
254 T. Fukuoka

At Minami-machida Grandberry Park, GI visions and frameworks can be seen in

creating “Open Space Network,” “Walkable City,” and “Sustainable Development.”
At site scale, GI open space is created to achieve both GI function and its character
as a place for people.
Lastly, this article tried to depict how GI visions and frameworks are translated
into the site scale and physical GI implementation through various case studies.
This article indicates how GI needs to be interwoven into vision, planning, and
design, and creates places for people. As stated above, strategic GI visions and
frameworks can catalyze a change to our city through physical manifestation of
GI implementation.

References

ASLA Award 2016 Copenhagen Cloudburst (2016). https://www.asla.org/2016awards/

171784.html
European Commission (2013) Green Infrastructure Enhancing Europe’s National Capitol
Fukuoka T (2020) Catalyze changes to the City from open spaces. Nikkei BP Publishing, Green
Infrastructure, pp 111–120
Fukuoka T, Katagiri Y, Kato S (2020) City of Philadelphia’s framework towards strategic green
infrastructure planning and implementation. J Jpn Inst Landsc Archit 83(5):673–678
Kato S, Fukuoka T (2016) Stormwater Management at a National Scale under the Asia Monsoon
Climate: Case Study of Singapore’s ABC Water Design Guidelines, International Symposium
on Architectural Interchange in Asia (11th ISAIA) P.1905–1908
Nakajima N, Hoshino Y (2017) A study of the infrastructure development process and support
system for climate change adaptation-in case of the cloudburst plan in municipality of
Copenhagen, DE
Osamura Y, Fukuoka T (2020) Process of active design guidelines in New York City: focusing on
ADG implementation. Reports of the City Planning Institute of Japan Nor.19. pp. 277–279
Public Utilities Board (2014) ABC waters design guidelines, 3rd edn. Public Utilities Board (PUB),
Singapore
Public Utilities Board (2016a) Singapore Water Story. https://www.pub.gov.sg/watersupply/
singaporewaterstory
Public Utilities Board (2016b) Active, beautiful, clean waters program. https://www.pub.gov.sg/
abcwaters/about
Public Utilities Board (2016c) Design Guidelines. https://www.pub.gov.sg/abcwaters/
designguidelines
Public Utilities Board (2016d) Certification. https://www.pub.gov.sg/abcwaters/certification
Public Utilities Board (2016e) ABC Waters Professionals. https://www.pub.gov.sg/abcwaters/
abcwatersprofessionals
United States Environmental Protection Agency (n.d.) What is Green Infrastructure? https://
www.epa.gov/green-infrastructure/what-green-infrastructure
West Philadelphia Landscape Project (n.d.). https://wplp.net
15 Toward Holistic Urban Green Infrastructure Implementation 255

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.