Landscape Ecological Applications in Man-Influenced Areas

LANDSCAPE ECOLOGICAL APPLICATIONS IN MAN-INFLUENCED AREAS

LANDSCAPE ECOLOGICAL APPLICATIONS IN MAN-INFLUENCED AREAS Linking Man and Nature Systems Edited by

Sun-Kee Hong Mokpo National University, Jeonnam, Korea

Nobukazu Nakagoshi Hiroshima University, Higashi-Hiroshima, Japan

Bojie Fu Chinese Academy of Sciences, Beijing, P.R. China

and

Yukihiro Morimoto Kyoto University, Kyoto, Japan

A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN-10 ISBN-13 ISBN-10 ISBN-13

1-4020-5487-4 (HB) 978-1-4020-5487-7 (HB) 1-4020-5488-2 (e-book) 978-1-4020-5488-4 (e-book)

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com

Printed on acid-free paper

All Rights Reserved © 2007 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.

To families, children and our colleagues in the world

Contents xi

Editors Contributors

xiii

Preface

xxv xxvii

Foreword Session I. Introduction 1. Landscape ecological applications in man-influenced areas linking man and nature systems –Editorial introduction

1

S.-K. Hong, N. Nakagoshi, B.J. Fu, Y. Morimoto

Session II. Landscape Analysis and Evaluation Method Part 1. Baseline concept 2. Spatial pattern analysis as a focus of landscape ecology to support evaluation of human impact on landscapes and diversity K.J. Koffi, V. Deblauwe, S. Sibomana, D.F.R. Neuba, D. Champluvier, C. De Canniere, N. Barbier, D. Traore, B. Habonimana, E. Robbrecht,

7

J. Lejoly, J. Bogaert

3. Application of landscape ecology in long term ecological research - Case study in China B.J. Fu, D. Niu, G.R. Yu, L.D. Chen, K.M. Ma, Y. Luo, Y.H. Lu, W.W. Zhao 4. Ecological networks, from concept to implementation R.H.G. Jongman

33 57

Part 2. Applications in evaluation 5. Landscape changes in Japan based on national grid maps N. Nakagoshi, J.-E. Kim

vii

71

CONTENTS

viii

6. Challenges faced when creating an evaluation method of biodiversity on an ecosystem level J. Morimoto 7. Identification of the potential habitat for giant panda in the Wolong Nature Reserve by using landscape ecology methodology

81

95

L.D. Chen, X.H. Liu, B.J. Fu, Y.H. Lü, J. Qiu 8. Land use change from traditional to modern eras: Saitama Prefecture, Japan

113

R. Seguchi, R.D. Brown, K. Takeuchi 9. Evaluation and planning of wildlife habitat in urban landscape Y. Natuhara

129

Session III. Landscape Management Part 3. Applications in managing diversity 10. Landscape ecology for biodiversity - Scaling up T.H. Ro, S.-K. Hong 11. A higher-taxon approach with soil invertebrates to assessing habitat diversity in East Asian rural landscapes S.-I. Tanabe, S.K. Kholin, Y.-B. Cho, S.-I. Hiramatsu, A. Ohwaki, S. Koji, A. Higuchi, S.Y. Storozhenko, S. Nishihara, K. Esaki, K. Kimura, K. Nakamura 12. Landscape ecological approach in oil palm land use planning and management for forest conservation in Malaysia S.A. Abdullah, N. Nakagoshi 13. Managing biodiversity of rice paddy culture in urban landscape - Case research in Seoul City I.-J. Song, Y.-R. Gin

149

163

179

193

Part 4. Applications in landscape health 14. Landscape restoration - A case practice of Kushiro Mire, Hokkaido F. Nakamura, Y.S. Ahn 15. Non-indigenous plant species in Central European forest ecosystems S. Zerbe

209 235

CONTENTS

ix

16. Traffic mortality, analysis and mitigation - Effects of road, traffic, vehicle and species characteristics F. van Langevelde, C. van Dooremalen, C.F. Jaarsma 17. Element fluxes and budgets of a plantation embedded in an agroforestry landscape: Implication for landscape management and sustainability

253

273

W. Shen, H. Ren, Y. Lin, M. Li 18. The effects of the regulation system on the structure and dynamics of green space in an urban landscape - The case of Kitakyushu City

291

T. Manabe, K. Ito, D. Isono, T. Umeno 19. Seeding on slopes in Japan for nature restoration

311

H. Yoshida 20. Wetlands and riparian buffer zones in landscape functioning Ü. Mander, K. Kimmel

329

21. Post-fire forest restoration indicated by canopy density in the northern great Hing’an mountains F.-J. Xie, X.-Z. Li, X.-G. Wang, D.-N. Xiao

359

Session IV. Designing for Landscape Creation Part 5. Applications in land planning and strategy 22. Kyoto as a garden city - A landscape ecological perception of Japanese garden design Y. Morimoto 23. Bee-Bo forest: Traditional landscape ecological forest in Korea K.-S. Lee 24. Cultural patterns as a component of environmental planning and design R.D. Brown, R. Lafortezza, R.C. Corry, D.B. Leal, G. Sanesi 25. Comparison of scenarios for the Vistula river, Poland T. Van der Sluis, J. Romanowski, J. Matuszkiewicz, I. Bouwma 26. Trends and future researches in green space design - Toward practical planning K. Nagashima 27. Beijing urban spatial distribution and resulting impacts on heat islands Z. Ouyang, R. Xiao, E.W. Schienke, W.F. Li, X. Wang, H. Miao, H. Zheng 28. Connectivity analyses of avifauna in urban areas H. Hashimoto

375 389 395 417

435 459 479

CONTENTS

x

29. International trends of rural landscape researches for land management and policies J.-E. Kim, S.-K. Hong, N. Nakagoshi

489

Session V. Conclusion 30. Linking man and nature landscape systems - Landscaping blue-green network

505

S.-K. Hong Index

525

Editors Sun-Kee HONG Institute of Island Culture, Mokpo National University, 61 Dorim-ri, Cheonggye-myeon, Muan-gun, Jeonnam 534-729, Korea E-mail: [email protected]

Nobukazu NAKAGOSHI Graduate School for International Development and Cooperation, Hiroshima University, 1-5-1 Kagamiyama, Higashi-Hiroshima 739-8529, Japan E-mail: [email protected]

Bojie FU National Key Lab of Systems Ecology, Research Center For Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China E-mail: [email protected]

Yukihiro MORIMOTO Lab. of Landscape Ecology and Planning, Graduate School of Global Environmental Studies, Lab. of Landscape Architecture, Graduate School of Agriculture Kyoto University, Kitasirakawa-oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan E-mail: [email protected]

Foreword Wolfgang HABER Lehrstuhl für Landschaftsökologie der Techn. Universität München, Weihenstephan, D-85350 Freising, Germany E-mail: [email protected]

xi

Contributors (*corresponding author) Saiful Arif ABDULLAH* (Chapter 12) Institute for Environment and Development (LESTARI), Universiti Kebangsaan Malaysia 43600 Bangi, Selangor Darul Ehsan, Malaysia E-mail: [email protected] Young Sang AHN Laboratory of Forest Ecosystem Management, Graduate School of Agriculture, Hokkaido University, Sapporo 060-8589, Japan E-mail: [email protected] Nicolas BARBIER Université Libre de Bruxelles, Service de Botanique Systématique et de Phytosociologie, 50 Avenue F.D. Roosevelt, CP 169, B-1050 Bruxelles, Belgique E-mail: [email protected] Jan BOGAERT* (Chapter 2) Université libre de Bruxelles, Laboratoire d’Ecologie du Paysage, 50 Avenue F.D. Roosevelt, CP 169, B-1050 Bruxelles, Belgique E-mail: [email protected] Irene BOUWMA ALTERRA Green World Research, P.O. Box 47 6700 AA, Wageningen, The Netherlands E-mail: [email protected] Robert D. BROWN* (Chapter 24) School of Environmental Design and Rural Development, University of Guelph, 50 Stone Road East, Guelph, Ontario, N1G 2W1, Canada E-mail: [email protected] Charles De CANNIÈRE Université libre de Bruxelles, Service de Lutte Biologique et d’Ecologie Spatiale, 50 Avenue F.D. Roosevelt, CP 160/12, Bruxelles, Belgique E-mail: [email protected] Dominique CHAMPLUVIER Jardin Botanique National de Belgique, Domaine de Bouchout, B-1860 Meise, Belgique E-mail: [email protected] xiii

xiv

CONTRIBUTORS

Liding CHEN* (Chapter 7) National Key Lab of Systems Ecology, Research Center For Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China E-mail: [email protected]; [email protected] Young-Bok CHO Natural History Museum, Hannam University, 133 Ojeong-dong, Daedeok-gu, Daejeon 306-791, Korea E-mail: [email protected] Robert C. CORRY School of Environmental Design and Rural Development, University of Guelph, 50 Stone Road East, Guelph, Ontario N1G 2W1 Canada E-mail: [email protected] Vincent DEBLAUWE Université libre de Bruxelles, Laboratoire d’Ecologie du Paysage, 50 Avenue F.D. Roosevelt, CP 169, B-1050 Bruxelles, Belgique E-mail: [email protected] Coby Van DOOREMALEN Resource Ecology Group, Wageningen University, Bornsesteeg 69, 6708 PD Wageningen, Netherlands E-mail: [email protected] Kojiro ESAKI Ishikawa Forest Experiment Station, Hakusan 920-2114, Japan. E-mail: [email protected] Bojie FU* (Chapter 3) National Key Lab of Systems Ecology, Research Center For Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China E-mail: [email protected] Yu-Ri GIN Korea National Park Service (Jirisan Southern Office) 511-1 Masan-myeon Gurye-gun, Jeonnam, Korea E-mail: [email protected] Bernadette HABONIMANA Université du Burundi, Faculté des Sciences Agronomiques, BP 2700 Bujumbura, Burundi E-mail: [email protected]

CONTRIBUTORS Hiroshi HASHIMOTO* (Chapter 28) Lab. Landscape Architecture & Environmental Design Division of Forest and Biomaterials Science Graduate School of Agriculture, Kyoto University E-mail: [email protected] Atsushi HIGUCHI Satoyama Nature School of Kakuma, Kanazawa University, Kanazawa 920-1192, Japan E-mail: [email protected] Shin-Ichi HIRAMATSU Shiramine Elementary School, Hakusan 920-2501, Japan E-mail: [email protected] Sun-Kee HONG* (Chapters 1, 10, 30) Institute of Island Culture, Mokpo National University, 61 Dorim-ri, Cheonggye-myeon, Muan-gun, Jeonnam 534-729, Korea E-mail: [email protected] Dai ISONO Department of Civil Engineering, Faculty of Engineering, Kyushu Institute of Technology, 1-1, Sensuicho, Tobata-ku, Kitakyushu 804-8550, Japan E-mail: [email protected] Keitaro ITO Department of Civil Engineering, Faculty of Engineering, Kyushu Institute of Technology, 1-1, Sensuicho, Tobata-ku, Kitakyushu 804-8550, Japan E-mail: [email protected] Catharinus F. JAARSMA Land Use Planning Group, Wageningen University, Generaal Foulkesweg 13, 6703 BJ Wageningen, The Netherlands E-mail: [email protected] Rob H.G. JONGMAN* (Chapter 4) Wageningen UR, Alterra, P.O. Box 47, 6700 AA Wageningen, The Netherlands E-mail: [email protected] S.K. KHOLIN Institute of Biology and Soil Science, Far Eastern Branch of the Russian Academy of Sciences, Prospect Stoletiya, 159, 690022, Vladivostok, Russia E-mail: [email protected]

xv

xvi

CONTRIBUTORS

Jae-Eun KIM* (Chapter 29) Graduate School for International Development and Cooperation, Hiroshima University, 1-5-1 Kagamiyama, Higashi-Hiroshima 739-8529, Japan E-mail: [email protected] Kai KIMMEL Institute of Geography, University of Tartu, Vanemuise St. 46, 51014 Tartu, Estonia E-mail: [email protected] Kazuya KIMURA Institute of Nature and Environmental Technology, Kanazawa University, Kanazawa 920-1192, Japan E-mail: [email protected] Kouao Jean KOFFI Université libre de Bruxelles, Laboratoire d’Ecologie du Paysage, 50 Avenue F.D. Roosevelt, CP 169, B-1050 Bruxelles, Belgique E-mail: [email protected] Sinsaku KOJI Institute of Nature and Environmental Technology, Kanazawa University, Kanazawa 920-1192, Japan E-mail: [email protected] Raffaele LAFORTEZZA Department of Plant Production Science, University of Bari, Via Amendola 165-A, 70126 Bari, Italy E-mail: [email protected] Frank Van LANGEVELDE* (Chapter 16) Resource Ecology Group, Wageningen University, Bornsesteeg 69, 6708 PD Wageningen, The Netherlands E-mail: [email protected] Diane B. LEAL School of Environmental Design and Rural Development, University of Guelph, Guelph, Ontario, N1G 2W1, Canada E-mail: [email protected] Kyoo-Seock LEE* (Chapter 23) Department of Landscape Architecture, Sungkyunkwan University, 300 Chunchun-dong, Chahngahn-ku, Suwon 440-746, Korea E-mail: [email protected]

CONTRIBUTORS Jean LEJOLY Université libre de Bruxelles, Service de Botanique Systématique et de Phytosociologie, 50 Avenue F.D. Roosevelt, CP 169, B-1050 Bruxelles, Belgique E-mail: [email protected] Minghui LI South China Normal University, Guangzhou 510631, China E-mail: [email protected] Weifeng LI National Key Lab of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China Email: [email protected] Xiu-Zhen LI* (Chapter 21) Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China E-mail: [email protected] Yongbiao LIN South China Botanical Garden, the Chinese Academy of Sciences, Guangzhou 510650, China E-mail: [email protected] Xuehua LIU Department of Environmental Sciences and Engineering, Tsinghua University, Beijing 100084, China Yihe LU State Key Laboratory of Systems Ecology Research Center for Eco-Environmental Sciences Chinese Academy of Sciences, 18 Shuangqing Road, Haidian District, Beijing 100085, China E-mail: [email protected] Yi LUO Institute of Geographical Sciences and Natural Resource Research, Chinese Academy of Sciences, 11 Datun Road, Chaoyang District, Beijing 100101, China E-mail: [email protected]

xvii

xviii

CONTRIBUTORS

Keming MA State Key Laboratory of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, 18 Shuangqing Road, Haidian District, Beijing 100085, China E-mail: [email protected] Tohru MANABE* (Chapter 18) Kitakyushu Museum of Natural History and Human History, 2-4-1, Higashida, Yahatahigashi-ku, itakyushu 805-0071, Japan E-mail: [email protected] Ülo MANDER* (Chapter 20) Institute of Geography, University of Tartu, Vanemuise St. 46, 51014 Tartu, Estonia E-mail: [email protected] Jan MATUSZKIEWICZ Institute of Geography and Spatial Organization, Polish Academy of Sciences, ul. Twarda 51/55, 00-818 Warszawa, Poland E-mail: [email protected] Hong MIAO National Key Lab of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China Junko MORIMOTO* (Chapter 6) Lab. Landscape Science and Planning, Nihon University, 1866 Kameino, Fujisawa 252-0813, Japan E-mail: [email protected] Yukihiro MORIMOTO* (Chapter 22) Lab. of Landscape Ecology and Planning, Graduate School of Global Environmental Studies, Lab. of Landscape Architecture, Graduate School of Agriculture, Kyoto University, Kitasirakawa-oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan E-mail: [email protected] Keiko NAGASHIMA* (Chapter 26) Graduate School of Bioresource and Bioenvironmental Science, Kyushu University, 6-10-1, Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan E-mail: [email protected]

CONTRIBUTORS Nobukazu NAKAGOSHI* (Chapter 5) Graduate School for International Development and Cooperation, Hiroshima University, 1-5-1 Kagamiyama, Higashi-Hiroshima 739-8529, Japan E-mail: [email protected] Futoshi NAKAMURA* (Chapter 14) Laboratory of Forest Ecosystem Management, Graduate School of Agriculture, Hokkaido University, Sapporo 060-8589, Japan E-mail: [email protected] Koji NAKAMURA Institute of Nature and Environmental Technology and Graduate School of Natural Science and Technology, Kanazawa University, Kanazawa 920-1192, Japan E-mail: [email protected] Yosihiro NATUHARA* (Chapter 9) Graduate School of Life and Environmental Sciences, Osaka Prefecture University, 1-1 Gakuen-cho, Naka-ku, Sakai 599-8531, Japan E-mail: [email protected] Danho Fursy NEUBA Université libre de Bruxelles, Service de Botanique Systématique et de Phytosociologie, 50 Avenue F.D. Roosevelt, CP 169, B-1050 Bruxelles, Belgique E-mail: [email protected] Shogo NISHIHARA Graduate School of Agricultural and Life Sciences, Tokyo University, Tokyo 113-8657, Japan E-mail: [email protected] Dong NIU Bureau of Science and Technology for Resources & Environment, Chinese Academy of Sciences, 52 Sanlihe Road, Xicheng District, Beijing 100864, China E-mail: [email protected] Atsushi OHWAKI Graduate School of Natural Science and Technology, Kanazawa University, Kanazawa 920-1192, Japan E-mail: [email protected]

xix

xx

CONTRIBUTORS

Zhiyun OUYANG* (Chapter 27) National Key Lab of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China E-mail: [email protected] Jun QIU Key Lab of Systems Ecology, Research Center For Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China Hai REN South China Botanical Garden, the Chinese Academy of Sciences, Guangzhou 510650, China E-mail: [email protected] Tae Ho RO Korea Environment Institute, 613-2 Bulkwang-dong, Eunpyung-gu, Seoul 122-040, Korea E-mail: [email protected] Elmar ROBBRECHT Jardin Botanique National de Belgique, Domaine de Bouchout, B-1860 Meise, Belgique E-mail: [email protected] Jerzy ROMANOWSKI Centre for Ecological Research, Polish Academy of Sciences, Konopnickiej 1, Dziekanow Lesny, 05-092 Lomianki, Poland E-mail: [email protected] Giovanni SANESI Department of Plant Production Science, University of Bari, Via Amendola 165-A, 70126 Bari, Italy E-mail: [email protected] Erich W. SCHIENKE Department of Science and Technology Studies, Rensselaer Polytechnic Institute, Troy, NY 12180, USA Rui SEGUCHI* (Chapter 8) Laboratory of Landscape Ecology and Planning, The University of Tokyo, Yayoi 1-1-1 Bunkyo-Ku, Tokyo 113, Japan; Isthmus Group Ltd., Landscape Architects, P.O. Box 90 366, Auckland, New Zealand E-mail: [email protected]

CONTRIBUTORS Weijun SHEN* (Chapter 17) South China Botanical Garden, the Chinese Academy of Sciences, Guangzhou 510650, China; Nicholas School of the Environment and Earth Science & Department of Biology, Duke University, Durham, NC 27708-0340, USA E-mail: [email protected] Serge SIBOMANA Université libre de Bruxelles, Laboratoire d’Ecologie du Paysage, 50 Avenue F.D. Roosevelt, CP 169, B-1050 Bruxelles, Belgique; Université du Burundi, Faculté des Sciences Agronomiques, BP 2700 Bujumbura, Burundi E-mail: [email protected] Theo VAN DER SLUIS* (Chapter 25) Netherlands Development Organization SNV. P.O. Box HP 565, Ho, Ghana; Landscape Center, ALTERRA Green World Research, P.O. Box 47, 6700 AA, The Netherlands E-mail: [email protected] In-Ju SONG* (Chapter 13) Department of Urban Environment, Seoul Development Institute (SDI), 391 Seocho-dong, Seocho-gu, Seoul 137-071, Korea E-mail: [email protected] Sergey STOROZHENKO Institute of Biology and Soil Science, Far East Branch of Russian Academy of Sciences, Prospect Stoletiya, 159, 690022, Vladivostok, Russia E-mail: [email protected] Kazuhiko TAKEUCHI Laboratory of Landscape Ecology and Planning, The University of Tokyo, Yayoi 1-1-1 Bunkyo-Ku, Tokyo 113, Japan E-mail: [email protected] Shin-Ichi TANABE* (Chapter 11) Institute of Nature and Environmental Technology, Kanazawa University, Kanazawa, 920-1192 Japan; Echigo-Matsunoyama Museum of Natural Science, 712-2 Matsukuchi, Matsunoyama, Tokamachi, Niigata 942-1411, Japan E-mail: [email protected]

xxi

xxii

CONTRIBUTORS

Dossahoua TRAORE Université de Cocody-Abidjan, Faculté des Sciences et Techniques, Département de Botanique et Biologie Végétale, Laboratoire de Botanique, 22 B.P. 582, Abidjan 22, Côte d'Ivoire E-mail: [email protected] Takashi UMENO Graduate School of Civil Engineering, Kyushu Institute of Technology, 1-1, Sensuicho, Tobata-ku, Kitakyushu 804-8550, Japan E-mail: [email protected] Xiaoke WANG National Key Lab of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China Email: [email protected] Xu-Gao WANG Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China Du-Ning XIAO Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China Rongbo XIAO National Key Lab of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China E-mail: [email protected] Fu-Ju XIE Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China Hiroshi YOSHIDA* (Chapter 19) Graduate School of Global Environmental Studies, Kyoto University, Toko Corporation, 8-9, 5-Chome, Shimbashi, Minato-ku, Tokyo 105-0004, Japan E-mail: [email protected]

CONTRIBUTORS

xxiii

Guirui YU Institute of Geographical Sciences and Natural Resource Research, Chinese Academy of Sciences, 11 Datun Road, Chaoyang District, Beijing 100101, China E-mail: [email protected] Wenwu ZHAO Institute of Land Resources and Management, College of Resources Science and Technology, Beijing Normal University, 19 Xinjiekouwai Street, Beijing 100875, China E-mail: [email protected] Stefan ZERBE* (Chapter 15) Institute of Botany and Landscape Ecology, University Greifswald, Grimmer Strase 88, D-17487 Greifswald, Germany E-mail: [email protected] Hua ZHENG National Key Lab of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Shuang Qing Road 18, Haidian District, P.O. Box 2871, Beijing 100085, China

PREFACE Landscape ecology and its application is disciplinary baseline in the world ecologists. Since last century of ecology, the ecologists had been roles for support many hot issues linking ‘sustainable development’ and ‘nature conservation’ in land use practice. To understand the interactions between landscape change and ecological processes is one of the primary goals of ecology. In populated area, cause and consequences of landscape change are significantly related to human impact in accordance with rapid urbanization and land transformation. Therefore, applying landscape ecology to ecological solution in man-influenced area is emerging issue and new challenge for ecologists. Since two world congresses of INTECOL (2002 Seoul and 2005 Montreal), two Congresses of EAFES (East Asian Federation of Ecological Societies) at 2004 (Mokpo, Korea) and 2006 (Niigata, Japan) and three conferences of IALE-Asia Pacific Chapter (Shenyang, Lanzou and Osaka), editors had been continue to organizing special symposia on issues of landscape ecology and ecological restoration in man-influenced area. Those symposia had been core role as steppingstones and corridors for developing mutual-cooperation of landscape ecologists between Asian and western countries. This book is one of our expressions from the results of exploring landscape systems. Aim of “Landscape Ecological Applications in Man-Influenced Areas” is reviewing landscape ecological applications on multi-scale ecological issues (e.g., resource management, habitat conservation, ecosystem restoration, biodiversity issue, and urban planning including land use policy, etc.) occurring man-influenced area from a global perspective. New dimensions of landscape research were explored in the countries that exposed by many problems on land systems owing to urbanization, population explosion, and global environmental change. This book includes several case studies on landscape analysis and evaluation using spatial analysis and landscape model for sustainable land management. “Linking Man and Nature Systems” as subtitle of this book, it suggests the integrative and ubiquitous solutions (such as socio-ecological integrative planning, large-scale conservation, and ecological restoration) considering harmony of man and nature systems. Such key issues and landscape elements had been reviewed in separate references but there are no textbook to university student and teacher for educational curricula as well as reference book for decision maker and planners. So it is reason why this book should be published. Our thanks are due to all authors for their contributions and expression. Special thanks are due to anonymous reviewers for revising manuscripts. Finally, we express our sincere thanks to each families and Ria Kanter at Springer for patience and cooperation. S.-K. Hong, N. Nakagoshi, B.J. Fu, and Y. Morimoto

xxv

FOREWORD

As a senior European landscape ecologist I take great pleasure to introduce the first book on landscape ecology edited by colleagues from East Asia. To be sure, one third of the 87 authors and co-authors of the book's 30 contributions are from Europe and North America; but only three articles are devoted to European topics, whereas 18 treat East Asian subjects of landscape ecology (the rest deals with more general themes). Thus, the book gives convincing evidence to the most welcome expansion of the originally 'western' landscape idea into the conceptual thinking and investigative activity of East Asian ecologists and environmental scientists. The field of landscape ecology is thus being broadened, the more so as some of the East Asian authors endeavour to include cultural or spiritual assets of their landscapes into its treatment, such as Japan's 'satoyama' or the Chinese 'fengshui' ('poongsoo' in Korea). Furthermore, the great diversity of the topics and approaches of the articles offers interesting reading and a rich information source for landscape ecological themes. As such, it fits well into the present epoch of appraising biodiversity, of which scientific diversity of course is a (not always recognized) part. And this diversity supports one of today's main goals of applied landscape ecology: maintaining or enhancing biodiversity at the landscape scale. Thus, landscape ecology is expanding both geographically and thematically; at the same time, it follows the normal development of every scientific enterprise, that is specialization based on in-depth analysis of the subject on one side, and application of findings to solve human problems on the other. 'Linking man and nature systems', expressed in the book's title as its explicit purpose, is served well by the contributions. These welcome developments, however, also arouse in me some reflections about the very subject of landscape ecology and its future scope. I cannot escape the impression that much of modern landscape-ecological research, or what is designated as such, consists mostly in widening the spatial context of the management of certain ecosystems or the selection and conservation of habitats. In this way, it identifies itself with the recent development of 'macroecology' within general ecology. But I like to stand up for the true landscape approach, keeping in mind the origin of the term and its history. 'Landscape' is distinguished from 'land' by both its cultural aspects and its integrative, 'wholistic' character which, however, tend to be neglected when they found scientific attention and thus became a subject of geographical and lastly ecological investigation. Landscape is, and should remain, a patterned picture in or minds that we want to see realized in our surroundings – which has often happened simply by land use, or xxvii

xxviii

FOREWORD

intentionally by design. Thus it becomes the real configuration of the human environment, both in urban, rural and quasi-natural space contexts, with functional aspects providing basic life support, and structural aspects expressing our culture and giving us pleasure derived from fulfilment of aesthetic and spiritual desires. In this way, landscape ecology inavoidably widens into human ecology. As to its application, the concept of sustainable development has become the challenging task for which landscape or human ecology, respectively, provide the proper focus, clad or disguised into timely approaches like models, scenarios, assessments, networks, or indicators. These reflections, however, do not at all question the great value and utility of this book. On the contrary, the readers will find such ideas, directly or indirectly, expressed in most of its chapters, and so I recommend it to all people concerned with the environment and wish it a large and open-minded readership. Wolfgang Haber Professor Emeritus of Landscape Ecology Technical University of Munich, Germany 26 July 2006

CHAPTER 1

LANDSCAPE ECOLOGICAL APPLICATIONS IN MAN-INFLUENCED AREAS - LINKING MAN AND NATURE SYSTEMS

Editorial introduction

S.-K. HONG1, N. NAKAGOSHI2, B.J. FU3, Y. MORIMOTO4 1

Institute of Island Culture, Mokpo National University, Jeonnam 534-729, Korea; 2 Graduate School for International Development and Cooperation, Hiroshima University, Higashi-Hiroshima 739-8529, Japan; 3Research Center For EcoEnvironmental Sciences, Chinese Academy of Sciences, Beijing 100085, China; 4 Graduate School of Global Environmental Studies, Kyoto University, Kyoto 6068502, Japan

1. INTRODUCTION The characteristics of the structure and function of landscapes and their ecological dynamics are integrated with natural and social factors (Naveh and Lieberman, 1994; Forman, 1995; Zonneveld, 1995). In view of the large-scale perspectives of ecosystem patterns and ecological processes (Turner et al., 2001), special attention should be paid to research on interdisciplinary solutions to the examination of patterns and processes of ecosystems (esp. degraded ecosystems and reserved areas; Liu and Taylor, 2002; Wiens and Moss, 2005). Landscape ecology is therefore a strong emerging concept and implements for those solutions. Landscape ecology does not only support new ecological principles, but also suggests models and designs that facilitate ecosystem creation (Zonneveld and Forman, 1990; Wu and Hobbs, 2002; Jongman and Pungetti, 2004; Pedroli and Pinto-Correia, 2006). A look at the history of landscape ecology reveals that two of its principles are strongly rooted in the academia of the EU and North America (large-scale ecosystem 1 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 1–6. © 2007 Springer.

2

S-K. HONG ET AL.

evaluation and land use planning). However, these two principles and cultural background should be integrated so as to facilitate the linking of various wildlife and systems. The current socio-economic principles are significantly different from those which were in place when the original framework of landscape ecology was formulated in the West. However, these principles have now become a major axis with which to understand landscape ecology and its applications in Eastern countries (Nakagoshi, 1995; Numata, 1996; Takeuchi et al., 2003). Moreover, human activity has now become the main focus of global change ecology, with special attention being paid to developments in these countries’ man-dominated areas (Hong et al., 2004). 2. OBJECTIVE This book focuses on the integration of landscape ecological principles and their application to landscape issues which might emerge in man-influenced areas. Landscape ecology has not only been focused on the conducting of research on spatial heterogeneity and its effect on unit ecosystems, but also on the introduction of strong implements which can be employed for habitat evaluation and the assessment of land use and restoration practices (Turner and Gardner, 1991; Farina, 2000; Gutzwiller, 2002; Wu and Hobbs, 2002; Bissonette and Storch, 2003). The objective of this book is to utilize the many intellectual roots of landscape ecology to integrate the principles of ecology-management-planning (Figure 1).

Landscape Ecology, Management and Planning (LEMP) Landscape Ecology Supporting Principles and Concept for Research and Monitoring

Multi- and Interdisciplinary Cooperation Reaction, and Feedback

Ecosystem Management

Environmental Planning

Application of Landscape Restoration and Conservation Practice

Design and Networking of Social and Culture Systems

Figure 1. Multi-disciplinary objective of landscape ecological application.

LANDSCAPE ECOLOGY FOR LINKING MAN AND NATURE SYSTEMS

3

Current eco-environmental issues such as resource management, habitat conservation, ecological restoration, ecological planning, EIA and the ‘urbanrural gradient’ issue are being resolved through the application of landscape ecology principles (Gutzwiller, 2002; Liu and Taylor, 2002; Bissonette and Storch, 2003). In this regards, the assurance of sustainable harmony between ‘development and conservation’ in man and nature systems should be regarded as an important focal point of any effort to resolve man-related issues. In this book, we attempt to resolve this ‘debate’ by integrating the principles of landscape ecology with practical issues such as management and planning. To achieve such an end, we invited experts in landscape research on such topics such as (1) multifunctional landscape, (2) ecosystem restoration, (3) landscape management, and (4) bio-eco-human networks, from the world over to take part in the compilation of this book. 3. CONTENTS OF THE BOOK Landscape Analysis and Evaluation Method

Section I is divided into two parts, namely general issues and analysis. Part 1 represents an introduction of the major fields of landscape ecology in maninfluenced areas; meanwhile, Part 2 deals with the current state of the development of the techniques and methodologies used for the spatial analysis and evaluation of landscapes. Part 1. Baseline concept Part 1 is a basic introduction to the main contents of the book. It involves a delicate debate of the difference between ‘principles’ and’ practice’ from the standpoint of landscape ecology. While there are many landscape research-related issues in man-influenced areas, we have chosen to focus on three specific ones in this book. In this part, the main principles and concepts associated with landscape analysis and ecological monitoring efforts are discussed at length. The authors also describe the standard landscape methods used to survey landscape patterns and ecological processes. Major keywords such as spatial analysis, long-term ecological database, and its networking in man-influenced areas are also discussed. Part 2. Applications in evaluation As part of the study of landscape evaluation, Part 2 explores instances of the quantification of landscape patterns and ecological processes. “Landscape evaluation” represents the first step towards implementing conservation, planning and restoration practices. Landscape mosaics are strongly influenced by human activities such as agriculture, forestry, and land use (Bunce et al., 1993). Moreover, biological habitats are usually adjacent to fragile landscapes such as urban-to rural areas. Therefore, serious problems occur in very sensitive areas which lie between

4

S-K. HONG ET AL.

natural and densely populated areas (McDonell and Pickett, 1993). Ecosystem modelling, impact assessment, and wildlife habitat models are also addressed in this section. Here, the evaluation of landscape pattern changes can be regarded as one of the main ecological indicators which should be used for sustainable landscape planning. Landscape Management

Many biological disciplines can be linked to landscape ecology from both a theoretical and methodological standpoint (Szaro and Johnston, 1996; Schwartz, 1997). Although the theories associated with restoration, conservation, and wildlife ecology have different historical backgrounds, the need to cooperate in order to ensure the sustainable management of resources, and to develop a shared goal of natural conservation have become increasingly important for landscape researchers. In this part, we discuss the development of new concept which can be used to manifest spatial patterns and ecological processes in multi-scales. Moreover, this section also introduces the important landscape ecological application practices being utilized in various countries. Part 3. Applications in managing diversity Global perspectives on conservation ecology as well as biodiversity issues are discussed in this portion of the book. The dispersal and distribution of biological components are heavily dependent on the landscape configuration and the quality of the landscape matrix. This section includes a discussion of a wide range of biodiversity issues spanning from the species to ecosystem levels. Moreover, the authors discuss the ecological integrity of landscape patterns as viewed through the lens of various cultures. Part 4. Applications in landscape health To date, restoration ecology has been focused on local areas, and more particularly small-scale ecosystems. Restoration ecology, as such, has been applied to those areas that have clear boundaries, such as roads, wetlands, watershed, and forest ecosystems. However, those involved in these restoration efforts, due to their tendency to implement comprehensive plans to change the landscape structure, and their mismanagement of the restoration process, have more often than not wound up weakening the ecological function of other surrounding ecosystems, and in further degrading the ecosystem which they were trying to restore (Barrow, 1991). To resolve these problems and restore a comparatively large-scale region, methods to assess the impact of such restoration efforts on surrounding ecosystems must be developed. These include expanding the scale of these restoration efforts; in other words, moving from the local to the landscape scale. As a conclusion, the practice of ecological restoration is increasingly moving towards the landscape scale in order to deal with these problems.

LANDSCAPE ECOLOGY FOR LINKING MAN AND NATURE SYSTEMS

5

Designing for Landscape Creation

Part 5. Applications in land planning and strategy In Part 5, landscape planning and environmental strategies will be discussed. Biotope creation and nature conservation strategies in man-influenced or mandominated landscapes represent the major issues discussed in this section. In particular, flexible cooperation between governments, citizens, and researchers represents an essential element of any well-laid land development plans. The environmental movement and civic groups are actively pursuing the maintenance of a balance between development and the conservation of nature. Education and the dissemination of information pertaining to ecology, most specifically as it relates to landscape ecology (Farina and Hong, 2004), represent another element in the quest for a sustainable society. Landscape ecology is not an omnipotent principle which can be applied to all ecological fields, but rather an element of the inter-disciplinary cooperation that will be needed to resolve the problems associated with linking man with natural systems. RESUME At the multi-scale level, environmental problems have already moved beyond the social capacity. While the resolution of these problems is a complex one, the development of a spatial understanding land patterns and natural processes had been partly resolved this plethora of emerging problems. Networking and connectivity between man and natural systems as well as the ecological role of landscape bridges such as urban-rural gradients and green-water gradients (such as Feng-shui, windwater theory) are some of the major perspectives which are discussed in this book. Finally, prominent issue discussed herein is that of the search for landscape integrity in man-dominated areas (Chapter 30). AKNOWLEDGEMENTS A big thanks goes out to all the authors who contributed to the compilation of this book. Moreover, we would like to express our heartfelt gratitude to all the anonymous reviewers who took the time to revise the manuscripts. Finally, a special debt of gratitude goes out to the families of these editors and authors, as well as to Ria Kanters of Springer in The Netherlands for her cooperation. REFERENCES Barrow, C.J. (1991). Land degradation. Cambridge University Press, Cambridge. 295p. Bissonette, J.A. and Storch, I. (Eds.) (2003). Landscape Ecology and Resource Management: Linking Theory with Practice. Island Press, Washington. 463p. Bunce, R.G.H., Ryszkowski, L. and Paoletti, M.G. (1993). Landscape Ecology and Agroecosystems. Lewis Publishers. 241p. Farina, A. (2000). Landscape Ecology in Action. Kluwer Academic Publishers, Dordrecht. 317p.

6

S-K. HONG ET AL.

Farina, A. and Hong, S.-K. (2004). A theoretical framework for a science of landscape. In S.-K. Hong, J.A. Lee, B.-S. Ihm, A. Farina, Y. Son, E.-S. Kim and J.C. Choe (Eds.), Ecological Issues in a Changing World (pp. 3-13). Kluwer Academic Publishers, Dordrecht. Forman, R.T.T. (1995). Land Mosaics: The Ecology of Landscapes and Regions. Cambridge University Press, Cambridge. 632p. Gutzwiller, K.J. (Ed.) (2002). Applying Landscape Ecology in Biological Conservation. Springer-Verlag, New York. 518p. Hong, S.-K., Lee, J.A., Ihm, B.-S., Farina, A., Son, Y., Kim, E.-S. and Choe, J.C. (Eds.) (2004). Ecological Issues in a Changing World – Status, Response and Strategy. Kluwer Academic Publisher, Dordrecht, The Netherlands. 425p. Jongman, R. and Pungetti, G. (2004). Ecological Networks and Greenways: Concept, Design, Implementation. Cambridge University Press, Cambridge. 345p. Liu, J. and Taylor, W.W. (2002). Integrating Landscape Ecology into Natural Resource Management. Cambridge University Press, Cambridge. 480p. McDonnell, M.J. and Pickett, S.T.A. (1993). Humans as Components of Ecosystems: The Ecology of Subtle Human Effects and Populated Areas. Springer-Verlag, New York. 364p. Nakagoshi, N. (Ed.) (1995). Grand Designs of Landscape. Kyouritsu Shuppan, Tokyo, 178p. (in Japanese). Naveh, Z. and Lieberman, A. (1994). Landscape Ecology: Theory and Application (2nd Edition). Springer-Verlag, New York. 360p. Numata, M. (Ed.) (1996). Keisoseitaigaku: Introduction of Landscape Ecology. Asakura Shoten, Tokyo, 178p. (in Japanese). Pedroli, B. and Pinto-Correia, T. (2006). Landscape – what’s in it? European landscape research at a turning point. Landscape Ecology, 21, p. 313. Schwartz, M.W. (1997). Conservation in Highly Fragmented Landscapes. Chapman & Hall, New York. 436p. Szaro, R. and Johnston, D.W. (1996). Biodiversity in Managed Landscapes: Theory and Practice. Oxford University Press, New York. 778p. Takeuchi, Y., Brown, R.D., Washitani, I., Tsunekawa, A. and Yokohari, M. (2003). Satoyama – The traditional rural landscape of Japan. Springer-Verlag, Tokyo. 229p. Turner, M.G., Gardner, R.H. and O’Neill, R.V. (2001). Landscape Ecology in Theory and Practice: Pattern and Process. Springer-Verlag, New York. 401p. Turner, M.G. and Gardner, R.H. (Ed.) (1991). Quantitative Methods in Landscape Ecology. SpringerVerlag, New York. 536p. Wiens, J. and Moss, M. (Eds.) (2005). Issue and Perspectives in Landscape Ecology. Cambridge University Press, Cambridge. 390p. Wu, J. and Hobbs, J. (2002) Key issues and research priorities in landscape ecology: an idiosyncratic synthesis. Landscape Ecology, 17, 355-365. Zonneveld, I.S. (1995). Land Ecology: An Introduction to Landscape Ecology as a Base for Land Evaluation, Land Management and Conservation. SPB Academic Publishing, Amsterdam. 199p. Zonneveld, I.S. and Forman, R.T.T. (Eds.) (1990). Changing Landscapes: An Ecological Perspective. Springer-Verlag, New York. 286p.

CHAPTER 2

SPATIAL PATTERN ANALYSIS AS A FOCUS OF LANDSCAPE ECOLOGY TO SUPPORT EVALUATION OF HUMAN IMPACT ON LANDSCAPES AND DIVERSITY

K.J. KOFFI1, V. DEBLAUWE1, S. SIBOMANA1,2, D.F.R. NEUBA3, D. CHAMPLUVIER4, C. DE CANNIERE5, N. BARBIER3, D. TRAORE6, B. HABONIMANA2, E. ROBBRECHT4, J. LEJOLY3, J. BOGAERT1 1

Université libre de Bruxelles, Laboratoire d’Ecologie du Paysage, Bruxelles, Belgique ; 2Université du Burundi, Faculté des Sciences Agronomiques, Bujumbura, Burundi ; 3Université libre de Bruxelles, Service de Botanique Systématique et de Phytosociologie, Bruxelles, Belgique ; 4Jardin Botanique National de Belgique, Meise, Belgique ; 5Université libre de Bruxelles, Service de Lutte Biologique et d’Ecologie Spatiale, Bruxelles, Belgique ; 6Université de Cocody-Abidjan, Laboratoire de Botanique, Abidjan, Côte d'Ivoire

Abstract. The relation between landscape patterns and ecological processes forms a central hypothesis of landscape ecology. Three types of pattern analysis to assess anthropogenic impacts on landscape ecosystems and biodiversity are presented in this chapter. Firstly, the results of an analysis of Acanthaceae data in Central Africa are presented and compared with phytogeographic theories. Phytogeography data reflect the spatial variability of plant diversity, and constitute therefore a major tool in conservation policy development. We investigated if it was possible to proxy the phytogeographic classifications by the spatial distribution of Acanthaceae only. When combined with a classic landscape pattern analysis, this type of study could provide complementary information for the definition of conservation priorities. Secondly, we present an analysis of periodic vegetations in the Sudan. It can be accepted that through an understanding of the underlying mechanisms of the formation of this unique pattern geometry, the knowledge with regard to the functioning and vulnerability of these ecosystems can be deepened. Using high-resolution remote sensing imagery and digital elevation models, the relation between pattern symmetry and slope gradient was explored. In particular, slope gradients that could

7 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 7–32. © 2007 Springer.

8

K.J. KOFFI ET AL.

condition the transition between spotted and tiger bush pattern types were focused. The influence of other sources of anisotropy was also considered. Finally, a complementary approach to the calculation of landscape metrics to analyse landscape pattern is described, using the spatial processes themselves causing landscape transformation. Landscape ecologists agree that there appears to be a limited number of common spatial configurations that can result from land transformation processes. Ten processes of landscape transformation are considered: aggregation, attrition, creation, deformation, dissection, enlargement, fragmentation, perforation, shift, and shrinkage. A decision tree is presented that enables definition of the transformation process involved using patch-based data. This technique can help landscape managers to refine their description of landscape dynamics and will assist them in identifying the drivers of landscape transformation

1. INTRODUCTION ‘Landscape pattern’ refers to features associated with the physical distribution or configuration of patches within a landscape (McGarigal and Marks, 1995). This spatial composition and configuration of landscape mosaics is dealt with in landscape ecology (Wiens, 2002), a branch of science developed to study ecological processes in their spatial context (Antrop, 2001). Changes in the spatial pattern of land use through time are considered to be crucial to the understanding of landscape dynamics and its ecological consequences (Turner and Ruscher, 1988). This central hypothesis of landscape ecology, i.e. that ecological patterns and processes are related (Turner, 1989), is known as the ‘pattern/process paradigm’ (Figure 1). Characteristic patterns of landscapes are supposed to be the result of the operation of ecological processes, that is, processes generate patterns and by analysing these patterns useful inferences about the underlying processes can be made (Coulson et al., 1999). This paradigm should also be applied in reverse order (Bogaert and Hong, 2003). In order to investigate this link between pattern and process quantitatively, and to predict the effects of particular landscape patterns on processes (Levin, 1992), it is useful to characterize these patterns in quantifiable terms. Pattern maps provide unique information because they quantify biologically relevant information that is not necessarily evident from a simple land cover map (Riiters et al., 2000). This focus on pattern has lead to a large number of landscape metrics, of which many have been shown to be correlated (O’Neill et al., 1988) or to exhibit statistical interactions with each other (Li and Reynolds, 1994). Landscape pattern analysis has consequently become a key activity of landscape ecologists. Nowadays, a variety of approaches are encountered in landscape ecology literature to deepen the knowledge on the functioning of landscape ecosystems and its influence on biodiversity. In this contribution, we present three examples of spatial pattern analysis in a landscape ecology perspective to illustrate the use of pattern analysis. Firstly, we present an analysis of the spatial pattern of Acanthaceae species, and link this information to classic theories of phytogeography. This type of analysis is useful when combined afterwards with a landscape pattern analysis (pattern of the abiotic environment or habitat), this to identify biodiversity conservation priorities. A second application of spatial pattern analysis involves the characterisation of unique landscape and vegetation patterns, such as tiger bush patterns in Africa. Through an understanding of the underlying mechanisms of the creation of this

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

9

typical geometry, the functioning and vulnerability of these ecosystems can better be studied. Finally, a complementary approach to the calculation of landscape metrics to analyse landscape pattern is presented, using the spatial processes them causing landscape transformation. This technique can help landscape managers to refine their description of landscape dynamics and will assist them in analysing the drivers of landscape transformation.

Figure 1. Illustration of the pattern/process paradigm. (a) Ecological processes are influenced by landscape pattern, e.g., edge effects as a consequence of patch-matrix interactions in fragmented landscapes; (b) Patterns of landscapes are supposed to be the result of the operation of processes, e.g., habitat fragmentation transforms contiguous vegetations into isolated habitat patches.

2. THE SPATIAL DIMENSION OF SPECIES DIVERSITY: WHERE PHYTOGEOGRAPHY AND CONSERVATION MEET 2.1 Phytogeography and conservation The most efficient method to gain understanding on the geographic distribution of plant species and on the ecological factors controlling this distribution is the analysis of spatial distribution maps of species (Lebrun, 2001). These maps enable the testing of hypotheses regarding the geographic origin of species, their speed of evolution, and their migration pathways. Moreover, a phytogeographic analysis enables to subdivide vast geographic units in smaller phytogeographic entities such as regions, districts, and sectors. When studied for multiple species or plant communities, these maps will reflect the spatial variation of plant (community) diversity, and consequently will be a useful tool in conservation policy development.

10

K.J. KOFFI ET AL.

For Central Africa – defined here as the geographic zone covered by the Democratic Republic of Congo, Burundi and Rwanda – three major phytogeographic theories have been proposed based on plant physiognomic arguments, bioclimatic data (precipitation, dry season length) and using the concept of endemism. This latter notion is central to the study of biogeography (Crisp et al., 2001). A taxon (e.g., a species) is considered endemic to a particular area if it occurs only in that area (Anderson, 1994). Ecologists are interested in areas of endemism because of their importance in conservation: narrowly endemic species are by definition rare, and therefore potentially threatened (Crisp et al., 2001). Robyns (1948) divided Central Africa in 11 districts (Figure 2). White (1979, 1983) subdivided Africa and Madagascar in 20 regional entities, from which the Guineo-Congolian regional centre of endemism, the Zambezian regional centre of endemism, the Afromontane archipelago-like regional centre of endemism, the Guineo-Congolian/Zambezian regional transition zone and the GuineoCongolian/Sudanian regional transition zone are found in Central Africa (Figure 3). Finally, Ndjele (1988) proposed a phytogeographic system subdividing the Democratic Republic of Congo in 13 sectors (Figure 4).

Figure 2. Subdivision of Central Africa in phytogeographic entities according to Robyns (1948). I: Coastal district; II: Mayumbe district; III: Lower Congo district; IV: Kasai district; V: Lower Katanga district; VI: Central Forest district; VII: Ubangi-Uele district VIII: Lake Albert district; IX: Lakes Edward and Kivu district; X: Ruanda-Urundi district; XI: Upper Katanga district.

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

11

Figure 3. Subdivision of Central Africa in phytogeographic entities according to White (1979, 1983). I: Guineo-Congolian regional centre of endemism (IB: Guinean sub-centre; IC: Congolian sub-centre);II: Zambezian regional centre of endemism; VIII: Afromontane archipelago-like regional centre of endemism; X: Guineo-Congolian/Zambezian regional transition zone; XI: Guineo-Congolian/Sudanian regional transition zone.

Figure 4. Subdivision of Central Africa in phytogeographic entities according to Ndjele (1988). I: Central Forest sector; II: Congolo-Sudanian transition sector; III: CongoloZambezian transition sector; IV: Mayumbe Forest sector; V: Lower Guineo/Zambezian transition sector; VI: Bemba sector; VII: Lualaba sector; VIII: Lunda sector; IX: Mountainous sector; X: Lake Mobutu sector; XI: Kivu Uplands sector; XII: Southern Sudania sector; XIII: Southern Atlantic sector of the Guinean coast.

The importance of these models should not be underestimated in the framework of species conservation. Phytogeography data reflect hotspots and spatial variability of plant diversity and constitute therefore a major tool in conservation policy development. Nevertheless, the present models can be

12

K.J. KOFFI ET AL.

considered as rigid due to their complexity: many ecological and botanical parameters are considered to define the districts, centres of endemism, transition zones, mosaics and sectors. Therefore, we investigate in this contribution if it is possible to proxy the aforementioned phytogeographic classifications by the spatial distribution of one single family, i.e. the Acanthaceae. By analysis of the spatial distribution of Acanthaceae species relative to the phytogeographic subdivisions of Robyns (1948), White (1979, 1983) and Ndjele (1988), the indicator value of this family will be analysed. 2.2 The distribution of Acanthaceae in Central Africa: comparison with the phytogeographic theories of Robyns, White and Ndjele A database composed of 9181 herbarium samples of the Acanthaceae family has been used in this study. Each herbarium sample contained, next to the species name, its taxonomic classification and a plant specimen, the geographical coordinates of the observation. These data were used to create spatial distribution maps using ArcView GIS 3.3 software. The herbarium samples represented 48 genus, 310 species, and 6362 different geographical sites. The number of samples per species is quite variable. Nineteen species were represented by more than 100 samples, 35 species by a number of samples between 99 and 50, 141 species by a number of samples between 49 and 10, and 114 species by less than 10 samples. The samples have been collected by 417 scientists between 1888 and 2001 during expeditions financed by scientific institutes such as the Institut National pour l’Etude et la Recherche Agronomique au Congo, the Institut des Parcs Nationaux du Congo Belge, the Centre d’Etudes Médicales de l’Université libre de Bruxelles en Afrique Centrale and the Comité Spécial du Katanga. Remarkable differences in the spatial presence of the species have been found (Figure 5). Certain species have been observed in almost every part of Central Africa, such as Asystasia gangetica subsp. gangetica. Others have been found to be associated with the hydrological network, such as Justicia pynaertii. The distribution of other species e.g. Ruellia tuberosa was related to its use by man. For Justicia diclipteroides subsp. praetervisa, an affiliation with the ecological conditions of the oriental mountainous region has been observed. Firstly, a comparison of the spatial pattern of the Acanthaceae is made with the theory of Robyns (1948) (Figure 6). The Upper Katanga district contains the highest number of Acanthaceae species, and 52 among them are specific or characteristic for this district. In decreasing order follow the district of the Lakes Edward and Kivu and the district of the Central Forest, which contain 14, respectively 10 characteristic species. The Coastal, Lower Congo, and Ubangi-Uele districts contain only one single characteristic species, respectively Barleria elegans, Ruellia togoensis and Lepidagathis peniculifera. The Kasai district does not contain any species that is not found in another district also.

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

13

Figure 5. Spatial distribution of Acanthaceae species: examples. (a) Asystasia gangetica T. Anders. Subsp. micranta (Ness) Ensemu Kelbessa; (b) Justicia pynaertii De Wild.; (c) Ruellia tuberosa L.; (d) Justicia diclipteroides (Lindau) subsp. praetervisa (Lindau) Hedrén.

The presence of the Acanthaceae with regard to the zones defined by White (1979, 1983) is shown in Figure 7. Sixty-two characteristic species are found in the Zambezian regional centre of endemism, 34 are observed for the Afromontane archipelago-like regional centre of endemism, and 15 in the Guineo-Congolian regional centre of endemism. The Guineo-Congolian/Zambezian regional transition zone is characterised by Barleria elegans and Justicia mendoncae, while Acanthus seretii, Lepidagathis peniculfera, Phaulopsis ciliata and P. savannicola characterise the Guineo-Congolian/Sudanian regional transition zone.

14

K.J. KOFFI ET AL. 120 110 100 90 80 70 60 50 40 30 20 10 0 I

II

III

IV

V

VI

VII

VIII

IX

X

XI

Figure 6. Presence of Acanthaceae species in the phytogeographic entities of Robyns (1948). Y-axis indicates the number of species. The white proportion of each bar indicates the number of Acanthaceae species characteristic for the phytogeographic entity considered.

Of the sectors defined by Ndjele (1988), the most characterised by the presence of Acanthaceae species are the Bemba sector and the Central Forest sector, as well as the Mountain sector (Figure 8). The Mayumbe Forest sector, the Lower Guineo/Zambezian transition sector, the Southern Sudanian sector and the Southern Atlantic sector of the Guinean coast are characterized by one single species, respectively Whitfieldia liebrechtsiana, Ruellia togoensis, Lepidagathis peniculfera, and Barleria elegans. The Congolo-Sudanian transition sector and the Lualaba sector are not characterised by a species of the Acanthaceae family not occurring in another phytogeographic entity. 2.3 Discussion Confrontation of the spatial pattern of the Acanthaceae species with the phytogeographic theories of Robyns (1948), White (1979, 1983) and Ndjele (1988) shows that the Zambezian regional centre of endemism, the Guineo-Congolian regional centre of endemism, and the Afromontane archipelago-like regional centre of endemism contain the highest number of herbarium samples. The flora of these zones is well known (Hepper, 1979) since they contain the main cities, research institutes, and universities, and since certain parts are characterized by a temperate climate. It should be emphasized that, outside these well-sampled zones, a large part of Central Africa is still unexplored. This observation has been confirmed for continental tropical Africa as a whole (Lebrun, 1973; Hepper, 1978; Kalanda, 1982; Lebrun and Stork, 1991; Lisowski, 1991). For some species however, a distribution

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

15

throughout the study zone is observed. Their ecological spectrum can consequently be considered as large. Other species are rather bound to particular geographic (ecological) regions, due to their smaller ecological spectrum. When a species is found in more than one single phytogeographic entity, it is considered a transition species. It should be noted that the regions of presence of particular species have been reduced considerably due to climatic changes when species were not able to respond properly to these dynamics (Schnell, 1971). Species restricted to a small geographic region are considered prone to extinction. This is a key issue in conservation biology, where the vulnerability concept can be a more profound interpretation of phytogeographic observations. A wild (sensu not cultivated) plant species is considered vulnerable when it shows an increased extinction risk. The main criteria to evaluate this status are the population size and its phytogeography. According to the UICN (2001), a plant species shows an enhanced extinction risk when the population size is reduced by more than 50% during the last 10 years by reversible causes and by more than 30% by irreversible causes. To study vulnerability on a species base, a method is proposed using six parameters: zonation or altitude range, biotope, morphology, geography, diaspore type, and use by man (Betti, 2001). For each parameter, a score is assigned which increases with the risk of extinction. Finally, the average score is calculated which reflects the overall extinction risk or vulnerability of the species. A species is considered very vulnerable when it is bound to particular altitudinal limits, when it is associated with undisturbed of primary forests, when being a tree, shrub or liana species, when it is an endemic or Afromontane species, when disseminating by sarcochory or desmochory, and when it is used by man for construction or in traditional medicine practices. 130 120 110 100 90 80 70 60 50 40 30 20 10 0 I

II

VIII

X

XI

Figure 7. Presence of Acanthaceae species in the phytogeographic entities of White (1979, 1983). Y-axis indicates the number of species. The white proportion of each bar indicates the number of Acanthaceae species characteristic for the phytogeographic entity considered.

16

K.J. KOFFI ET AL. 120 110 100 90 80 70 60 50 40 30 20 10 0 I

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

XIII

Figure 8. Presence of Acanthaceae species in the phytogeographic entities of Ndjele (1988). Y-axis indicates the number of species. The white proportion of each bar indicates the number of Acanthaceae species characteristic for the phytogeographic entity considered.

The Zambezian region, denoted as the Upper Katanga district by Robyns (1948) and the Bemba sector by Ndjele (1988) represent the highest diversity of Acanthaceae, which confirms the existence of a centre of speciation of the African flora, more specifically in the Bangweolo-Katanga region (Lebrun, 1960, 1976; Ozanda, 1982), a region with a mineralised soil (deposition of copper) and characterized by climate types Aw and Cw of the Köppen classification (Bultot, 1950). Moreover, plateaus (1500-1700 m), open forests with Brachystegia and Pseudoberlinia, and typical herbal steppes are observed (Duvigneaud, 1958), next to dry dense forests (Muhulu), forest galleries, Dembos vegetations and typical plant communities associated with ore-containing sites. At the genus level, the Zambezian region is also the speciation centre of the Thunbergia and Justicia genus, the former being a principal genus of that region as observed by Ndjele (1988). In the framework of conservation, it should be noted that the indicator value of the Acanthaceae is not constant throughout the study area considered. Certain phytogeographic entities are characterized by many species, while the presence or absence of one single species is characteristic to other zones. The use of a restricted number of species to substitute the more complex phytogeographic theories should therefore be executed with caution; nevertheless Barleria elegans, Ruellia togoensis

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

17

and Lepidagathis peniculifera seem to be species with discriminatory properties that can be used in a first approach to characterize regional gradients of plant diversity. For particular phytogeographic entities, no characteristic species have been found, which can indicate that the current approach has to be refined. Comparison with other families and confrontation with other phytogeographic theories and concepts should be effectuated before more profound conclusions can be drawn.

3. CHARACTERISING UNIQUE VEGETATION PATTERNS: ANALYSIS OF TIGER BUSH ORIENTATION USING FOURIER SPACE ANGULAR DISTRIBUTION AND REMOTE SENSING IMAGERY 3.1 Periodic vegetation patterns In water-limited ecosystems, covering about one third of the Earth’s surface (White, 1971; Schlesinger et al., 1990), the vegetation appears discontinuous and usually covers less than 60% of the landscape (Aguiar and Sala, 1999). In those regions between tropical savannah and desert, one can often observe particular vegetation mosaics in which the vegetation cover is not homogeneously or randomly distributed but is contracted into a “periodic” pattern. In the most famous case, the landscape is covered by bands of dense vegetation alternating with bands of bare soil or by strips covered by grass. This type of pattern is known as “tiger bush”, by analogy to the tiger fur pattern. Another common vegetation mosaic consists of bare gaps regularly distributed within a dense matrix of vegetation. This kind of pattern, generally referred to as “spotted bush”, is far less impressive on aerial photographs (Figure 9) and therefore attracted hardly attention of scientists. However, like tiger bush, spotted bush is spatially periodic (Couteron and Lejeune, 2001), i.e. the distance between two successive vegetated bands or a gap is relatively constant throughout the landscape. This distance is referred to as the “wavelength” of the pattern. These periodic vegetation patterns are often not related to pre-existing substratum variability, though environmental factors can potentially distort the symmetry of the pattern. Although wind has often been considered (Ives, 1946; Aguiar and Sala, 1999; Leprun, 1999), a slope gradient is generally assumed to be the leading source of anisotropy in periodic vegetations. It has been observed that, when the slope gradient does not exceed a defined threshold (Valentin et al., 1999), isotropic spotted bush occur. On the other hand, a weak slope has been observed to generate a pattern with bands elongating orthogonally to the gradient (Greenwood, 1957; Boaler and Hodge, 1964; Mabbutt and Fanning, 1987; Montana et al., 1990). This characteristic lead MacFadyen (1950), when he first described such patterns in Somaliland Republic, to propose their use as a slope aspect indicator on aerial photographs. This banded pattern, similar to contour lines on a map, lead to interpret the tiger bush pattern as a water harvesting strategy of vegetation. It was postulated

18

K.J. KOFFI ET AL.

that a weak slope gradient, permitting water sheet flow without channel drainage formation, allows dense vegetation below the bare area to thrive.

Figure 9. Spectral features computed from two typical 51 by 51 pixels sub-windows of (a) spotted and (b) banded vegetation. For visual purposes, both symmetric θ-spectrum values are represented in order to cover all azimuth directions.

Mathematical implementations of vegetation pattern dynamics (Lefever and Lejeune, 1997; von Hardenberg et al., 2001; van de Koppel and Rietkerk, 2004) have lead to a complete review of our understanding of this phenomenon. By

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

19

considering the phytomass growth at any point in space as a function of aridity and of plant-plant or plant-resource interactions, models were able to reproduce the complete range of patterns encountered in the field and agreed on the prediction of a distinct succession of patterns along aridity gradients. Starting from wetter areas, one should invariably encounter continuous cover, spotted bush, bands with no preferred direction (sometimes called “labyrinthine” pattern), gaps consisting of bare soil periodically scattered with dense thickets, and finally desert without any vegetation. Inclusion of an external source of anisotropy in these models can lead to the formation of bands perpendicular to the gradient (Lefever and Lejeune, 1997), i.e. to tiger bush. Slope, and therefore sheet flow, can therefore be considered as a secondary driver in the pattern dynamics. In this contribution, we used high-resolution remote sensing data combined with digital elevation models to explore the relation between pattern symmetry and slope gradient in the Kordofan province of the Sudan. In particular, we tried to evidence critical slope gradients that would condition the transition between spotted and tiger bush pattern types. Additionally, we analysed departure from the expected effect of slope anisotropy on the band orientation, in order to evidence the influence of other sources of anisotropy. 3.2 Quantitative pattern analysis of contracted vegetation in the Sudan using remotely sensed data An area with periodic vegetation patterns was selected in the Western Kordofan state, ~700 km southwest of Khartoum (Sudan). This site was characterised by a sufficiently large wavelength to be detectable on satellite imagery of high resolution and covered a continuous area sufficiently wide to facilitate data handling. Our study area was located ~180 km south-east of the Terminalia brownii arcs and Acacia mellifera whorls described by Wickens and Collier (1971), which are currently strongly damaged (probably due to the high population density). In our site, the vegetation was intact and contracted either in bands – sometimes elongating over several kilometers – or in spotted patterns. The mean annual rainfall ranges from 510 to 590 mm and vegetation belongs to the Sudanian type (White, 1983). Two panchromatic SPOT scenes covering the entire study area with a spatial resolution of 10 m were used. The scenes where taken in the middle of the dry season (December 22, 2001 and January 17, 2002). On panchromatic digital images, brightest pixels usually correspond to bare soil, intermediate gray-scale levels to continuous grass cover sites and darker pixels to woody vegetation. At first approximation, grey-scale levels can be considered a monotonically decreasing function of biomass. We used a SRTM digital elevation model with three arcs second spatial resolution (~90 m) to compute the topography features. Superposition of the digital elevation model with both SPOT scenes was achieved with an average error of less than 30 m in the field. Slope can be considered theoretically as a vector; as such it is determined by intensity (gradient) and by direction (aspect). By convention, we defined the slope aspect as the direction of the steepest decrease of the altitude within the area

20

K.J. KOFFI ET AL.

considered. It ranges from 0° to 360° (0° being the north direction and values increasing clockwise). Preliminary investigation of the images revealed vegetation bands lying on slope gradients of 0.2-1.5% and elongated perpendicularly to the slope (Figure 10). This typical tiger bush formation was characterized by a wavelength of 70-120 m. Another pattern was observed in his vicinity: wide areas of evenly spaced gaps in a continuous vegetation matrix. This spotted bush pattern showed a systematically smaller wavelength (40-60 m). A rectangular area of 2475 km² (10°57’-11°34’N; 28°11’-28°30’E) including the entire tiger bush area together with several wide areas of spotted bush were selected for further analysis. This study site was divided into non-overlapping square-shaped sub-windows of 510×510 m². A zone covered by clouds as well as the border zone between both SPOT scenes was excluded for analysis so that a total of 8029 sub-windows were retained for pattern analysis.

Figure 10. Subset of the land cover map computed from K-means clustering. Contour lines were computed from the 3 arc second SRTM digital elevation model. Equidistance is 5 m. See text for the meaning of each class.

A two-dimensional Fourier transform and the associated computation of the twodimensional periodogram were applied. The use of the periodogram is recommended in the case of patterns showing spatial periodicity, since the amplitude values directly express the proportion of the image variance accounted for by periodic functions of explicit spatial frequencies and orientations. Pixel emissivity

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

21

corrections are not needed since this proportion is invariant to linear grey scale level rescaling. This method has been successfully applied on digitalized aerial views of periodic (Couteron and Lejeune, 2001; Couteron, 2002) as well as non periodic (Couteron et al., 2005; Couteron et al., in press) vegetations. Two selected subwindows and their respective analysis are shown in Figure 9 to exemplify the method. Left side and right side graphs express the partitioning of image variance in typical banded and spotted vegetations respectively. A two-dimensional periodogram is a set of values, in Cartesian co-ordinates, each representing the portion of image variance σ2 in their particular direction and frequency. Pattern information relative to spatial frequency and to spatial orientation was separately captured by summing the periodogram values on either ring-shaped or wedgeshaped concentric frequency regions, in order to compute the r- and θ-spectra respectively (Renshaw and Ford, 1984; Figure 9). These two spectra thus quantify the contributions of successive spatial frequencies (r-spectrum) and spatial orientation (θ-spectrum) to the image variance. Due to sub-window size, the analysis was limited to the first 25 wave numbers in order to avoid aliasing effects. Nonhierarchical, unsupervised clustering using the K-means algorithm and the Euclidean distance (Legendre and Legendre, 1998) was performed on the r-spectra table after standardization to objectively classify the sub-windows into five coarseness-fineness classes. The θ-spectrum consisted of 36 orientations since we partitioned the 0°-180° range of the periodogram into classes of 5°. Due to their intrinsic nature we analyzed those angular data using circular statistics. The basic assumption of circular statistics is that a shift of 360° in data is meaningless. For example the mean direction between 5° and 355° should be 0° and not 180°. It should be emphasized that even when vegetation strips showed a certain orientation, a direction could not be assigned due to the absence of floristic field data indicating plant age or band dynamics. Therefore a shift of 180° in the band orientation was also meaningless. In this case the term “axial data” is used, as opposed to “vectorial data” such as slope orientation. Because each value of a θ-spectrum could be considered as a vector with its direction and intensity, we characterised the pattern orientation as the weighted circular mean of the orientations of each sub-window. This mean orientation was computed as the vectorial sum of the θ-spectra entries. Because these entries have a higher intensity when they express an orientation encompassing a high proportion of the image variance, this sum refers to the axis orthogonal to the maximal elongation of the pattern (i.e. the travelling wave direction). 3.3 Classification of land cover spatial pattern The 8029 standardized r-spectra were submitted K-means clustering following Barbier et al. (in press). Five classes were considered appropriate to obtain a satisfactory separation between the observed landscape structures. The first three classes (C1 to C3) reflected a textural gradient associated with the relative importance of small versus large spatial frequencies in the spectrum (Figure 11). The first class (C1) gathered spectra dominated by very high spatial frequencies

22

K.J. KOFFI ET AL.

(>40 cycles/km) and was dominated by homogeneous savannas. Class C2 grouped sub-windows with spatial frequencies of similar importance all along the spectrum, although a small spike protruding at 17 cycles/km suggested that it contained mixed textures characteristic of classes C4 and C5. The third class (C3) was dominated by small spatial frequencies (<3 cycles/km) thereby corresponding to sub-windows marked by large landscape features, i.e. macroheterogeneity at sub-window scale, such as roads, croplands, hamlets, gallery forests or rocky escarpments. The two remaining classes (C4 and C5) were characterised by a strong pike associated with intermediate frequencies, corresponding to the spatial scales of the spotted and tiger bush respectively, as preliminarily identified.

Figure 11. Mean standardized r-spectrum of each class of land cover pattern (K-means clustering). See text for the meaning of each class.

3.4 Slope direction domain of periodic vegetation and dependency between slope and vegetation pattern orientations The probability to find spotted bush (C4) was the highest on slope directions around 270°. Tiger bush (C5) exhibited a preferential direction between 190° and 230° (Figure 12). Slopes oriented between 30° and 50° seemed to be rarely covered by spatially periodic vegetation.

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

23

Figure 12. Rose diagram of slope direction. The number of sub-window classified as C4 (a) or C5 (b) in each angular bin were divided by the total number of sub-window in that particular direction. Length of pie chart accounts for proportion.

In order to quantify the dependency between each vegetation pattern orientation and the terrain slope direction on which they occur, a circular correlation test was performed. As expected, only the sub-windows classified as C5 (tiger bush) showed a clear correlation (Table 1) and the bands were on average elongated orthogonally to the slope (i.e. their travel direction was parallel to the slope). The V-Test (Fisher, 1993) showed that the shift between slope and pattern mean direction diverged consistently from a uniform distribution towards a distribution centred on zero degrees in the classes C2 and C5. It should be reminded that class C2 contained subwindows composed of a mixture of C4 and C5 classes. The strong dependency between vegetation and slope directions in class C5 was explored in detail. When the shift between the slope direction and C5 bands orientation was investigated, it was observed that its distribution was not uniform around the slope azimuthal circle (Figure 13). The bands were oriented orthogonally to the slope when oriented at 220° (e.g. south-west). From 220° to 340°, the shift increased progressively to reach a maximum at ~300°. In this part of the azimuthal circle, the shift was negative, indicating that bands tended to rotate slightly counter clockwise with respect to the slope direction. Approaching 350°, the shift decreased and between 350° and 70°, the number of sub-windows classified as C5 was too low (<10) and their distribution insufficiently concentrated to compute the circular mean and its confidence interval. Near 100°, the shift reached a new maximum, on the positive direction this time, which suggested that bands tended to rotate slightly clockwise away from the slope direction. Afterwards, it diminished towards 220°. 3.5 Discussion Using the Fourier space signature, we precisely mapped the different kinds of vegetation patterns observable on very high-resolution remotely sensed data for a study site in the Sudan. Each pattern class was characterised by a dominant peak in its frequency domain as characterised by the r-spectra. Therefore, due to their specific wavelengths, both periodic vegetations patterns, known as tiger bush and

24

K.J. KOFFI ET AL.

spotted bush, were clearly separated from other landscape structures. Furthermore, the use of the θ-spectrum allowed us to precisely determine band orientations of all sub-windows classified as tiger bush pattern. Table 1. Dependency between slope and vegetation pattern orientation. Πn : Circular rank correlation coefficient between the mean orientation of vegetation pattern and the direction of slope gradient; N: number of samples. Circular statistics were computed following Fisher (1993). Class C1 C2 C3 C4 C5

Circular rank correlation N Πn P 899 0.011 <0.005 2377 0.149 <0.005 1316 0.021 <0.005 1359 0.019 <0.005 2078 0.535 <0.005

V-Test (expected mean: 0°) V p -0.017 0.77 0.364 <0.005 0.007 0.361 -0.018 0.829 0.803 <0.005

Figure 13. Mean angular discrepancy between slope direction and C5 vegetation pattern orientation. Zero discrepancy indicates that bands of vegetation are orthogonal to the direction of slope. Positive and negative ones indicate a clock and counter-clock shift respectively of the vegetated bands. The mean shift (triangle) of every 20° bin is plotted between his 95% confidences bounds (diamonds).

It is commonly accepted that the tiger bush bands elongate orthogonally to the slope gradient so as to act as a rain gauge system (Valentin et al., 1999). Our results proved that such an assumption is exact, but only as a first approximation. Indeed, the band orientations were well correlated to the slope gradient and their discrepancy

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

25

rarely exceeded 10° in magnitude. Nevertheless, departure from orthogonality showed a systematic and significant structure as the band orientation turned slightly clockwise and counter clockwise on the slopes oriented southeast and northwest respectively. In between, when the slope orientation was southwest, the band travelling waves were orientated exactly in the same direction as the slope. On slopes oriented northeast, too few tiger bush are observed to precisely determine if there was a discrepancy. In some sense, the vegetated bands tended to rotate a little in order to face their upslope front with the northeast direction. Thereby they acted like if a weak anisotropy factor oriented northeast to southwest added his effect to the slope affect it. At our knowledge, only wind could have been responsible for this phenomenon. As observed for dune orientation in central Sudan as well as in the entire Sahelian band, two dominant winds are alternatively blowing in this orientation but in opposite direction: the Harmattan winds and the African monsoon winds. Abiotic processes driven by wind and water include seed dispersion, and redistribution of fine soil particles, associated mineral nutrients and litter that is consequently concentrated underneath vegetated patches. Whether seeds or soil particle movements were involved in this case needs further field investigation. In this regard, the relative decrease of tiger bush presence observed on slopes oriented northeast could be interpreted in several ways. The opposition of slope and wind in the case of a weak slope gradient could lead to an isotropic environment. Therefore, bearing in mind that spotted bush (spots of bare soils periodically distributed in a vegetation matrix) are known to occur when the slope gradient does not exceed a certain threshold, we could expect to find an increasing proportion of this type of pattern when the slopes are oriented northeast. However, this was not the case, since the relative occurrence of spotted bush was the highest on westerly oriented slopes. Probably, another factor, i.e. the slope gradient, influenced pattern formation. In our study area, the mean slope was steeper when its direction was northeast to east (data not shown) reaching 0.65% versus 0.45% elsewhere. This could explain why there was so few-spotted bush on this slope direction. However, the reason why tiger bush did not appear remains unclear because it occurs on slope gradients as much as 1.5% on any direction. In the light of our results, previous work mentioning periodic vegetation bands that are not aligned along the contour (Dunkerley and Brown, 2002) should be reinvestigated for wind and slope interaction. Consequently, the dichotomous distinction between bands and wind-induced anisotropy (d'Herbès et al., 2001) is likely to be tempered in favour of a continuum. Analysis of remotely sensed data using the Fourier power spectrum enabled us to identify the periodic vegetation patterns and, for one of them called “tiger bush”, to quantitatively test the dependence between the vegetation orientation and the anisotropic factors likely controlling them. It can be concluded that vegetation pattern orientation was mainly driven by slope and, to a lesser extent, by another factor that was assumed to be wind.

26

K.J. KOFFI ET AL.

4. A TYPOLOGY FOR SPATIAL LANDSCAPE TRANSFORMATION PROCESSES 4.1 Spatial processes of land transformation The conversion of native landscapes for human activities results in widespread changes in landscape spatial structure (Collinge, 1998). There appears to be a limited number of common spatial configurations that can result from land transformation processes (Franklin and Forman, 1987; Collinge and Forman, 1998). Based upon pattern geometry, eight processes have been described for landscapes composed of two classes, one representing the class of interest and the other one representing the underlying matrix: attrition, bisection, dissection, dissipation, fragmentation, incision, perforation, and shrinkage (Forman, 1995; Collinge, 1998; Collinge and Forman, 1998; Jaeger, 2000). These transformation processes described only land cover loss and were initially developed to assess the impact of pattern change on species diversity. A kind of hierarchical relationship between the processes was considered in Forman (1995) and Jaeger (2000). In Collinge (1998) and Forman and Collinge (1998), all processes were considered equivalent.

4.2 Identification of the spatial process involved in landscape pattern dynamics A decision tree model (Figure 14) was presented to enable determination of the dominant land transformation process using area, boundary length and patch number statistics (Bogaert et al., 2004) of the class of interest. These characteristics were recognized as key elements for pattern description that would encompass most of the phenomena of observed patterns in a landscape (Giles and Trani, 1999). The aim was to provide environmental scientists with a quick protocol to determine the transformation process (es) present in the study area of interest (Bogaert et al., 2004). The models includes processes associated with land cover increase, or with no change of the land cover extent, an evident novelty relative to Forman (1995), Collinge (1998), Collinge and Forman (1998), and Jaeger (2000). Some of the aforementioned processes were not retained because of their overlap with other ones, and every process was given an unequivocal definition. The following ten processes of landscape transformation were considered (Bogaert et al., 2004): aggregation (patch mergence), attrition (reduction of the number of patches), creation (formation of new patches), deformation (change of patch shape), dissection (subdivision of patches using equal-width lines), enlargement (patch size expansion), fragmentation (breaking up of patches into smaller parcels), perforation (gap formation), shift (patch repositioning), and shrinkage (reduction of patch size). For a geometrical illustration of the processes, the reader is referred to the seminal paper.

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

27

determine a0, a1, p0, p1, n0, n1

yes

yes n1>n0

a1>a0

no

no

creation

yes n1=n0

tobs=a1 /a0

no yes attrition

no a1
tobs
no

yes

aggregation

no

fragmentation

no

yes

a1>a0

a1
no

dissection

p1>p0

shrinkage

yes

no yes

deformation

p1=p0

perforation

yes

enlargement

shift

Figure 14. Flow chart based on the decision tree model (Bogaert et al., 2004) to identify spatial processes causing landscape pattern transformation. All decisions (diamond-shaped components) are based on either the area (a), the perimeter (p) or the number of patches (n) before (a0, p0, n0) and after (a1, p1, n1) the transformation of the landscape. Comparison of tobs=a1/a0 with a – by the user – predefined area loss ratio (t) enables distinction between fragmentation and dissection, which generate similar patterns.

Input data for the decision tree consists of the total class area (a), the total class perimeter or boundary length (p), and the number of patches (n) observed for the land cover class of interest before (a0, p0, n0) and after (a1, p1, n1) transformation. Since a binary landscape model was assumed, which was considered suitable for most applications, other key parameters mentioned by Giles and Trani (1999), such

28

K.J. KOFFI ET AL.

as the proportion of the dominant class or the number of classes, were not considered. The decision tree algorithm leads to the identification of one single spatial process by comparing the values of a0, p0, n0 with a1, p1, n1 and is consequently based on equality, increase, or decrease of total land cover area, total perimeter length, and of the number of patches. It should be noted that some processes in the decision tree algorithm are defined by less parameters than others. Aggregation, attrition, creation, dissection, enlargement and fragmentation are defined by two parameters only; while for deformation, perforation, shift and shrinkage three parameters are required. Use of all parameters for all processes could have refined the process identification, but would also have generated a more complex model, which could have countered the initial objective, i.e. designing a practical tool. An example for this case is the process of enlargement. It can be expected that perimeter increase coincide with enlargement. The absence of a parameter can also indicate redundancy of the parameter; this was the case for those transformation processes with n1>n0; the considered processes were all associated with perimeter increase, so that incorporation of this parameter would not have provided complementary information. One weakness was nevertheless observed, which requires some flexibility by the user: patch attrition can dominate the analysis as a consequence of the design itself of the decision tree algorithm. The attrition or creation of one single patch can lead to completely different conclusions. In the case of the attrition of a single or a few patches (and assuming this being a marginal phenomenon in the landscape conversion considered), a methodological fix was suggested, i.e. to redo the analysis with omission of the decrease in number of patches (the user should consider n1=n0 instead of n1
SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

29

5. CONCLUDING REMARKS Since the end of the 1930s, when the term “landscape ecology” was launched (Troll, 1939), many gigantic strides in theory, methodology, and applications have been made in this branch of science (Wu and Hobbs, 2002). The initial developments of landscape ecology took place mainly in Central and Eastern Europe, focusing on issues directly related to planning, management, conservation, and restoration of landscapes. This research emphasis on the interactions between human activities on one hand and resources (i.e., the landscape) on the other hand, initiated the development of holistic, interdisciplinary, and somewhat pragmatic views and approaches (Naveh, 2000; Wu and Hobbs, 2002). In contrast, landscape ecology began to develop in North America in the 1980s with an apparent emphasis on spatial heterogeneity and the concomitant effects on ecological processes where quantitative methods, such as spatial pattern analysis and modelling, prevailed (Wu and Hobbs, 2002). The development of landscape metrics and the ongoing polemic on their use (Bogaert et al., 2002) since the publication of the seminal paper of O’Neill et al. (1988) exemplify this development. The importance of the spatial character of problems and research is widely accepted (Bastian, 2001), but agreement likewise exists that spatial relations remain only one of the relevant foci of landscape ecology. Bridging this gap between landscape ecologists was considered an urgent need both for theoretical and practical reasons, this to enable the discipline to be really effective in terms of addressing world environmental and ecological problems (Farina, 1993). Three types of spatial pattern analysis to assess anthropogenic impacts on landscape ecosystems and biodiversity are presented in this chapter. Firstly, the results of an analysis of the spatial distribution of Acanthaceae data in Central Africa are presented. Their spatial pattern is confronted with phytogeographic theories to test if it was possible to proxy the current phytogeographic classifications by the distribution of Acanthaceae only. Secondly, an analysis of periodic vegetations in the Sudan is presented. Using high-resolution remote sensing imagery and digital elevation models, the relation between pattern symmetry and slope gradient was explored. Finally, a complementary approach to the calculation of spatial metrics to analyse landscape pattern is described, using the spatial processes themselves causing landscape transformation. A decision tree is presented that enables definition of the transformation process involved using patch-based data.

ACKNOWLEDGEMENTS The authors acknowledge the Government of Ivory Coast for the doctoral fellowship of K.J. Koffi. V. Deblauwe is indebted to the FRIA for his doctoral fellowship. The research of S. Sibomana is supported by a CTB fellowship. We are grateful to the European OASIS project for providing us the remotely sensed data SPOT. Copyright – © CNES (2001, 2002), distribution Spot Image S.A. The authors

30

K.J. KOFFI ET AL.

acknowledge the Fonds voor Wetenschappelijk Onderzoek – Vlaanderen (G.0019.04), the Fonds National de la Recherche Scientifique (1.5.028.05), the Université libre de Bruxelles (Crédit extraordinaire de recherche), and the Politique Scientifique Fédérale (SR/00/045) for financial support. REFERENCES Aguiar, M.R. and Sala, O.E. (1999). Patch structure, dynamics and implications for the functioning of arid ecosystems. Trends in Ecology & Evolution, 14, 273-277. Anderson, S. (1994). Area and endemism. Quarterly Review of Biology, 69, 451-471. Antrop, M. (2001). The language of landscape ecologists and planners – A comparative content analysis of concepts used in landscape ecology. Landscape and Urban Planning, 55, 163-173. Barbier, N., Couteron, P., Lejoly, J., Deblauwe, V. and Lejeune, O. Self-organised vegetation patterning as fingerprint of climate and human impact on semiarid ecosystems. Journal of Ecology, in press. Bastian, O. (2001). Landscape ecology: towards a unified discipline? Landscape Ecology, 16, 757-766. Betti, J.L. (2001). Vulnerabilité des plantes utilisées comme antipaludiques dans l’arrondissement de Mintom au Sud de la Réserve du Dja (Cameroun). Systematics and Geography of Plants, 71, 661678. Boaler, S.B. and Hodge, C.A.H. (1964). Observations on vegetation arcs in the Northern region, Somali republic. Journal of Ecology, 52, 511-544. Bogaert, J., Ceulemans, R. and Salvador-Van Eysenrode, D. (2004). A decision tree algorithm for detection of spatial processes in landscape transformation. Environmental Management, 33, 62-73. Bogaert, J. and Hong, S.-K. (2004). Landscape ecology: monitoring landscape dynamics using spatial pattern metrics. In S.-K. Hong, J.A. Lee, B.-S. Ihm, A. Farina, Y. Son, E.-S. Kim and J.C. Choe (Eds.), Ecological Issues in a Changing World (pp. 109-131). Kluwer Academic Publishers, Dordrecht. Bogaert, J., Myneni, R.B. and Knyazikhin, Y. (2002). A mathematical comment on the formulae for the aggregation index and the shape index. Landscape Ecology, 17, 87-90. Bultot, F. (1950). Carte des Régions Climatiques du Congo Belge Etablie d’après les Critères de Köppen (Communication n°9 du Bureau Climatologique). Publication I.N.E.A.C. Collinge, S.K. (1998). Spatial arrangement of habitat patches and corridors: clues from ecological field experiments. Landscape and Urban Planning, 42, 157-168. Collinge, S.K. and Forman, R.T.T. (1998). A conceptual model of land conversion processes: predictions and evidence from a microlandscape experiment with grassland insects. Oikos, 82, 66-84. Coulson, R.N., Saarenmaa, H., Daugherty, W.C., Rykiel, E.J.Jr., Saunders, M.C. and Fitzgerald, J.W. (1999). A knowledge system environment for ecosystem management. In J.M. Klopatek and R.H. Gardner (Eds.), Landscape Ecological Analysis – Issues and Applications (pp. 57-79). Springer, New York. Couteron, P. (2002). Quantifying change in patterned semi-arid vegetation by Fourier analysis of digitized aerial photographs. International Journal of Remote Sensing, 23, 3407-3425. Couteron, P., Barbier, N. and Gautier, D. Textural ordination based on Fourier spectral decomposition: a method to analyze and compare landscape patterns. Landscape Ecology, in press. Couteron, P. and Lejeune, O. (2001). Periodic spotted patterns in semi-arid vegetation explained by a propagation-inhibition model. Journal of Ecology, 89, 616-628. Couteron, P., Pelissier, R., Nicolini, E.A. and Paget, D. (2005). Predicting tropical forest stand structure parameters from Fourier transform of very high-resolution remotely sensed canopy images. Journal of Applied Ecology, 42, 1121-1128. Crisp, M.D., Laffan, S., Linder, H.P. and Monro, A. (2001). Endemism in the Australian flora. Journal of Biogeography, 28, 183-198. Curtis, J.T. (1956). The modification of mid-latitude grasslands and forests by man. In W. L. Thomas (Ed.), Man’s Role in Changing the Face of the Earth (pp. 721-736). University of Chicago Press, Chicago. d'Herbès, J.M., Valentin, C., Tongway, D.J. and Leprun, J.C. (2001). Banded vegetation patterns and related structures. In D.J. Tongway, C. Valentin, J.M. d'Herbès and J. Seghieri (Eds.), Banded

SPATIAL PATTERN: A FOCUS OF LANDSCAPE ECOLOGY

31

Vegetation Patterning in Arid and Semi-Arid Environments. Ecological Processes and Consequences for Management (pp. 1-19). Springer, New York. Dunkerley, D.L. and Brown, K.J. (2002). Oblique vegetation banding in the Australian arid zone: implications for theories of pattern evolution and maintenance. Journal of Arid Environments, 51, 163-181. Duvigneaud, P. (1958). La végétation du Katanga et de ses sols métallifères. Bulletin de la Société Royale Botanique de Belgique, 90, 127-286. Farina, A. (1993). From global to regional landscape ecology (editorial comment). Landscape Ecology, 8, 153-154. Fisher, N.I. (1993). Statistical Analysis of Circular Data. Cambridge University Press, Cambridge. Forman, R.T.T. (1995). Land Mosaics: The Ecology of Landscapes and Regions. Cambridge University Press, Cambridge. Franklin, J.F. and Forman, R.T.T. (1987). Creating landscape patterns by forest cutting: ecological consequences and principles. Landscape Ecology, 1, 5-18. Giles, R.H.Jr. and Trani, M.K. (1999). Key elements of landscape pattern measures. Environmental Management, 23, 477-481. Greenwood, J.E.G.W. (1957). The development of vegetation patterns in Somaliland Protectorate. Geographical Journal, 123, 465-473. Hepper, N.F. (1978). The present stage of botanical exploration of tropical Africa. In I. Hedberg (Ed.) Systematic Botany, Plant Utilisation and Biosphere Conservation (pp. 41-46). I. Almqvist & Wiksell International, Stockholm. Hepper, N.F. (1979). Deuxième édition de la Carte du Degré d’Exploration Floristique de l’Afrique au Sud du Sahara. In G. Kunkel (Ed.), IXe Réunion plénière de l’AETFAT, Las Palmas. Ives, R.L. (1946). Desert ripples. American Journal of Science, 244, 492-501. Jaeger, J. (2000). Landscape division, splitting index, and effective mesh size: new measures of landscape fragmentation. Landscape Ecology, 15, 115-130. Kalanda, K. (1982). Etude Taxonomique et Phytogéographique du Genre Vernonia Schreb. (Asteraceae) au Zaïre. Thèse doctorale, Université de Kisangani, Zaïre. Lebrun, J. (1960). Sur une méthode de délimitation des horizons et étages de végétation des montagnes du Congo Oriental. Bulletin du Jardin Botanique de l’Etat, 30, 75-94. Lebrun, J.P. (1973). Enumeration des Plantes Vasculaires du Sénégal. Etude botanique no. 2, Institut d'Elevage et de Médecine vétérinaire des Pays tropicaux, Maisons-Alfort. Lebrun, J.P. (1976). Richesse spécifique de la flore vasculaire des divers pays ou régions d’Afrique. Candollea, 31, 11-15. Lebrun, J.P. (2001). Introduction à la Flore d’Afrique. CIRAD. Ibis Press. Lebrun, J.P. and Stork, A.L. (1991). Énumération des Plantes à Fleurs D'Afrique Tropicale. Volume. 1. Généralités et Annonaceae à Euphorbiaceae et Pandanceae. Conservatoire et Jardin botaniques. Geneva. Lefever, R. and Lejeune, O. (1997). On the origin of tiger bush. Bulletin of Mathematical Biology, 59, 263-294. Legendre, P. and Legendre, L. (1998). Numerical ecology (2nd English ed.). Elsevier, Amsterdam. Leprun, J.C. (1999). The influences of ecological factors on tiger bush and dotted bush patterns along a gradient from Mali to northern Burkina Faso. Catena, 37, 25-44. Levin, S.A. (1992). The problem of pattern and scale in ecology. Ecology, 73, 1943-1976. Li, H. and Reynolds, J.F. (1994). A simulation experiment to quantify spatial heterogeneity in categorical maps. Ecology, 75, 2446-2455. Lisowski, S. (1991). Les Asteraceae dans la Flore D’Afrique Centrale. Fragmenta Floristica et Geobotanica 36(1 & 2). Polish Academy of Sciences, W. Szafer Institut. Mabbutt, J.A. and Fanning, P.C. (1987). Vegetation banding in arid Western Australia. Journal of Arid Environments, 12, 41-59. MacFadyen, W.A. (1950). Vegetation patterns in the semi-desert plains of British Somaliland. Geographical Journal, 116, 199-210. McGarigal, K. and Marks, B.J. (1995). FRAGSTATS: Spatial Pattern Analysis Program for Quantifying Landscape Structure. Department of Agriculture, Pacific Northwest Research Station, PNW-GTR351, Oregon.

32

K.J. KOFFI ET AL.

Montana, C., Lopez Portillo, J. and Mauchamp, A. (1990). The response of two woody species to the conditions created by a shifting ecotone in an arid ecosystem. Journal of Ecology, 78, 789-798. Naveh, Z. (2000). What is holistic landscape ecology? A conceptual introduction. Landscape and Urban Planning, 50, 7-26. Ndjele, M.B. (1988). Les Eléments Phytogéographiques Endémiques dans la Flore Vasculaire du Zaïre. Thèse doctorale, Faculté des Sciences, Université libre de Bruxelles, Belgique. O’Neill, R.V., Krummel, J.R., Gardner, R.H., Sugihara, G., Jackson, B., DeAngelis, D.L., Milne, B.T., Turner, M.G., Zygmunt, B., Christensen, S.W., Dale, V. and Graham, R.L. (1988). Indices of landscape pattern. Landscape Ecology, 3, 153-162. Ozanda, P. (1982). Les Végétaux dans la Biosphère. Dion éditeurs, Paris. Renshaw, E. and Ford, E.D. (1984). The description of spatial pattern using two-dimensional spectral analysis. Vegetatio, 56, 75-85. Riitters, K., Wickham, J., O’Neill, R.V., Jones, B. and Smith, E. (2000). Global-scale patterns of forest fragmentation. Conservation Ecology, 4, www.consecol.org/vol4/iss2/art3. Robyns W. (1948). Les territoires phytogéographiques du Congo Belge et du Ruanda-Urundi. Atlas général du Congo Belge. Académie Royale des Sciences Coloniales, Bruxelles. Schlesinger, W.H., Reynolds, J.F., Cunningham, G.L., Huenneke, L.F., Jarrell, W.M. and Virginia, R.A. (1990). Biological feedbacks in global desertification. Science, 247, 1043-1048. Schnell, R. (1971). Introduction à la Phytogéographie des Pays Tropicaux. Volume II : Les Milieux. Les Groupements Végétaux. Editions Gauthier-Villars, Paris Troll, C. (1939). Luftbildplan und ökologische Bodemforschung. Zeitschrift der Gesellschaft für Erdkunde zu Berlin, 241-298. Turner, M.G. (1989). Landscape ecology: the effect of pattern on process. Annual Review of Ecology and Systematics, 20, 171-197. Turner, M.G. and Ruscher, C.L. (1988). Changes in landscape patterns in Georgia, USA. Landscape Ecology, 1, 241-251. UICN, (2001). Catégories et critères de l’UICN pour la Liste Rouge. UICN, Gland. Valentin, C., d'Herbes, J.M. and Poesen, J. (1999). Soil and water components of banded vegetation patterns. Catena, 37, 1-24. van de Koppel, J. and Rietkerk, M. (2004). Spatial interactions and resilience in arid ecosystems. American Naturalist, 163, 113-121. von Hardenberg, J., Meron, E., Shachak, M. and Zarmi, Y. (2001). Diversity of vegetation patterns and desertification. Physical Review Letters, 87, art-198101. White, F. (1979). The Guineo-Congolian region an its relationships to other phytochoria. Bulletin du Jardin Botanique National de Belgique, 49, 11-55. White, F. (1983). The Vegetation of Africa. A descriptive Memoire to Accompany the UNESCO/AETFAT Vegetation Map of Africa. UNESCO, Paris. White, L.P. (1971). Vegetation stripes on sheet wash surfaces. Journal of Ecology, 59, 615-622. Wickens, G.E. and Collier, F.W. (1971). Some vegetation patterns in the republic of the Sudan. Geoderma, 6, 43-59. Wiens, J.A. (2002). Riverine landscapes: taking landscape ecology into the water. Freshwater Biology, 47, 501-515. Wu, J. and Hobbs, R. (2002). Key issues and research priorities in landscape ecology: an idiosyncratic synthesis. Landscape Ecology, 17, 355-365..

CHAPTER 3

APPLICATION OF LANDSCAPE ECOLOGY IN LONG TERM ECOLOGICAL RESEARCH Case study in China

B.J. FU1,2, D. NIU1,2 , G.R.YU3, L.D. CHEN1, K.M. MA1, Y. LUO3, Y.H. LU1, W.W. ZHAO4 1

State Key Laboratory of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085; 2 Bureau of Science and Technology for Resources & Environment, Chinese Academy of Sciences, Beijing 100864; 3 Institute of Geographical Sciences and Natural Resource Research, Chinese Academy of Sciences, Beijing 100101; 4Institute of Land Resources and Management, College of Resources Science and Technology, Beijing Normal University, Beijing 100875, China

Abstract. The consequences of the growing world population imply an increasing demand on housing, industry, roads, airports, recreation, land, water resource, etc. Their impacts on natural landscapes and ecosystems are often significant by changing landscape pattern. The current environmental effects due to human activity, such as global change, ecosystem degradation, biodiversity loss, and others, have already arisen the attention from scientists. The long-term and potential environmental effects, however, were difficult to be identified without enough scientific data across both spatial and temporal scales. A primary goal of landscape ecology is to understand the reciprocal relationship between spatial pattern and ecological flows or processes. Achieving this goal may require the extrapolation of results obtained from small-scale experiments to broad scales. Scientific monitoring data makes a solid and convincing base for studying the dynamics of landscape spatial pattern and ecological processes and the overall environmental effects of human activities. In this paper, the Chinese Ecosystem Research Network (CERN) was introduced. It lays a foundation for landscape ecological research by offering long-term monitoring data. At the same time, how to effectively use the CERN to deepen landscape ecological researches was discussed in detail. Finally, perspectives for the development of landscape ecological researches in China were enumerated with special attention to ecological monitoring and the coupling of landscape pattern and ecological processes.

33 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 33–56. © 2007 Springer.

34

B.J. FU ET AL.

1. INTRODUCTION The consequences of the growing world population imply an increasing demand on housing, industry, roads, airports, recreation, land, water resource, etc. Their impacts on natural landscapes and ecosystems are often significant by changing landscape pattern (Norse, 1992; Verburg et al., 1999; Defries, et al., 2002). The impacts cover direct and indirect, secondary, cumulative, short, medium and longterm, permanent and temporary, positive and negative effects (Brismar, 2002). The current environmental effects due to human activity, such as global change, ecosystem degradation, biodiversity loss, and others, have already arose the attention from scientists and governments, and much endeavour were paid (Cooper and Sheate, 2002; Döös, 2002; Meadows and Hoffman, 2003). The long-term and potential environmental effects, however, were difficult to be identified without enough scientific data across both spatial and temporal scales. The impacts on ecosystems and environment would become much stronger with the intensification of human activities. Scientific monitoring data makes a solid and convincing base for studying the dynamics of landscape spatial pattern and ecological processes and the overall environmental effects of human activities. Historically, these studies were difficult to conduct because of the dominance of short term funding programs. The complexity of the environment and the changing nature require additional research efforts that are not only long term, but also address questions of sale dependency, complex assemblages of species and their interactions, and the role of humans in environmental change. Long-term ecological research offers an important means to the traditional ecological research on data acquisition. Landscape ecology emphasizes broad spatial scales and the ecological effects of the spatial pattern of ecosystems (Turner, 1989). Specifically, it considers four pats: (a) the development and dynamics of spatial heterogeneity, (b) interactions and exchanges across heterogeneous landscapes, (c) the influences of spatial heterogeneity on biotic and abiotic processes, and (d) the management of spatial heterogeneity. Landscape pattern is spatially correlated and scale-dependent. Thus, understanding landscape structure and functioning requires multi-scale and longterm information. However, the explicit effects of spatial patterns on ecological processes have not been well studied. The spatial patterns observed in landscapes are resulted from complex interactions between physical, biological, and social forces. Spatial heterogeneity is ubiquitous across all scales and forms the fundamental basis of the structure and functioning of landscapes. To understand how landscapes affect, and are affected by, biophysical and socioeconomic activities, we must be able to quantify spatial heterogeneity and its scale dependence. Therefore, the theory on landscape pattern, heterogeneity and scaling, could provide a guideline on establishment of long-term ecological research monitoring stations. A primary goal of landscape ecology is to understand the reciprocal relationship between spatial pattern and ecological flows or processes (Turner, 1989; Opdam, et al., 2002; Wu and Hobbs, 2002). This goal is difficult to accomplish, however, because the broad spatial-temporal scales involved make experimentation and hypothesis testing more challenging. Thus, achieving this goal

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

35

may require the extrapolation of results obtained from small-scale experiments to broad scales. Understanding the relationship between landscape pattern and ecological processes is critical for any in-depth research in landscape ecology (Wu and Hobbs, 2002; Haase and Halle, 2004). One of the trademarks of landscape ecology was its extensive use of landscape metrics among numerous methods for spatial pattern analysis (O’Neil et al., 1998; Turner and Gardner, 1990; Gustafson, 1998; He, 2000). Among these landscape metrics, most of them were introduced to describe current landscape pattern, the ecological understanding related to landscape pattern change was less studied than expected (Haines-Young and Chopping, 1996; Li and Wu, 2004). There are two reasons accounting for this. One is that it is easy to describe the landscape pattern than to study the dynamic of landscape change, as well as the ecological processes. The second reason is that to study the ecological processes requires much more monitoring data and takes long time. In most programs, short-term environmental effects were better addressed than the long-term change due to budget limitation. Long-term monitoring data are now recognized as crucial to our understanding of environmental change and management, and to the studies on landscape pattern and ecological processes. Thus, the establishment of the Long-term monitoring stations would lay a good foundation for landscape ecological research. 2. FOUNDATION AND FOCUSES OF LANDSCAPE ECOLOGY 2.1 Landscape heterogeneity, change and driving forces Landscape heterogeneity is a key concept in landscape ecology, which is defined as the complexity of patch mosaic in space (Farina, 1998); or the uneven and nonrandom spatial configuration of landscape structures (Forman, 1995). It is even considered that the goal of landscape ecology is to study landscape heterogeneity. Landscape heterogeneity has remarkable impact on ecological functions and processes. For example, it influences the flow and transportation of resources, species, or disturbances in a landscape. Meanwhile, Landscape heterogeneity is also the result of the interactions of some basic ecological and biophysical processes both at spatial and temporal scales. The main sources of landscape heterogeneity are natural disturbances, human activities, and the succession and developing history of vegetation. Landscape heterogeneity mainly focuses on the three aspects, (1) spatial heterogeneity, i.e., the spatial complexity of landscape structures, including gradient distribution and mosaic structures; (2) temporal heterogeneity, i.e., the differences of landscape pattern at different time; and (3) functional heterogeneity, i.e., the functional differences of landscape structures, for example, the spatial configuration of material, energy and species flow. Landscape is changing, and this change will lead to profound influences on the ecology and the environment of a landscape. The result of landscape change not only altered the spatial configuration of the landscape, hence influenced the energy

36

B.J. FU ET AL.

distribution and material cycling, some irrational landscape change also will cause some serious environmental problems, such as land degradation and non-point source pollution, thus has profound influence on the society and economy. Landscape change is the joint result of natural processes and human activities, where natural driving forces often determine the landscape at large spatial scales, causing the large-scope change. These factors include the formation of geomorphology, the influence of climate, the influence of organisms, the development of soil, and natural disturbances. Meanwhile, anthropogenic driving forces are population, technology, politics, economy, and culture etc., which are considered the more and more important factors among the abundant driving forces to landscape change in nowadays. We can see that more positioning data on the ecological function and process is required for the study of landscape heterogeneity; only using the remotely sensed data is obviously not enough. At the same time, it is the same for landscape change study, long-term time series data could provide more reliable evidence. 2.2 Landscape pattern and processes Landscape pattern refers to the configuration of patches at different size and shape in space. Landscape pattern represents landscape heterogeneity in space, and determines the rate and intensity of ecological processes in a landscape. Meanwhile, landscape pattern is also the result of the interactions among these ecological processes at different scales, including disturbances (Turner and Gardner, 1991). It is obvious that remotely sensed data can only provide information on landscape pattern, but could not supply detail information on the ecological processes in landscape. Therefore, we must have the support of on site monitoring data and combine the RS data with the monitoring data to give a clearer explanation on the interactions between landscape pattern and ecological processes. 2.3 Landscape analysis and scale change Scale is the focal issue in landscape ecology from the very beginning. Landscape heterogeneity is always existed, although it will change in intensity along with observing scales. In contrast, homogeneity is relative, which only emerged at some special scales. We need to clarify spatial scale when studying spatial heterogeneity. Landscape pattern also has strong scaling characteristics. It is even considered that there is no pattern if you do not mention scale. Meanwhile, various ecological processes happen at different scales, how they correlate with landscape pattern at different scales is another important issue in landscape ecology. Moreover, the interactions of pattern and process at different scales determined that scale is also needed to be considered in landscape change study. With different observing scales, the processes and results of landscape change will differ. As modern landscape ecology is moving from spatial to temporal, from phenomena to driving forces, and from single scale to multiple scales, long-term monitoring networks becomes more and more important to landscape ecologists.

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

37

Each research station could provide positioning data for landscape change and pattern and process studies at landscape and smaller scales; while the networks of many research stations would support the scaling studies from landscape to regional scales. Such a novel study manner would explore a completely new era of landscape ecology. 3. INTRODUCTION TO CHINESE ECOSYSTEM RESEARCH NETWORK (CERN) The Chinese Ecosystem Research Network (CERN), one of the founding members of the International Long Term Ecosystem Research Network (ILTER), and Global Terrestrial Observation System (GTOS), consists of 36 field research stations, including 14 stations for agriculture, 9 for forest, 2 for grassland, 5 for desert, 1 for marsh, 2 for lake and 3 for marine ecosystems (Figure 1), in addition to five disciplinary centers and a synthesis center. All the CERN stations engage in monitoring work, research, experiment and demonstration, while the disciplinary centers are responsible for the calibration of monitoring instruments and data quality control. The synthesis center is responsible for data exchange and inter-disciplinary research. For years, through its long-term monitoring, research and experiment, demonstration and extension, it has served as an important facility to control desertification, soil erosion, salinization, and eutrophication.

Figure 1. Location of Chinese Ecosystem Research Network (CERN).

38

B.J. FU ET AL.

The Mission of CERN is to promote ecosystem conservation and improvement, environmental quality enhancement and agricultural development, and to advance the studies in ecology and related inter-disciplines. Its mandate includes monitoring, research and demonstration on typical ecosystems in China 4. CERN-BASED RESEARCH ON PATTERN AND PROCESS FROM LARGE SCALE 4.1 Landscape pattern and ecological process study across water gradient in China 4.1.1 Spatial Distribution, Representation, Scientific Themes and Research Missions The rainfall in China is unevenly distributed and declines gradually from the southeast to the northwest, with an annual average rainfall over 2000mm in southeastern China and less than 50 mm in northern Xinjiang (Figure 2). Consequently, the production of cropland ecosystems and its interaction with climate/water varies greatly in different parts of China. Based on rainfall distribution, major crops and soil types, twelve field research stations on cropland ecosystems are established from the northwest to the southeast (Figure 3), which roughly correspond to the water gradient from north to south. The annual average rainfall for these stations is shown in Figure 3.

Figure 2. Distribution map of annual average rainfall in China [data source from China Institute of Water Resources and Hydropower Research (IWHR)].

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

39

Figure 3. Annual average rainfall for CERN-based stations of cropland ecosystems.

Among them, there are four experimental stations located in Southern China along Yangtze River. Yingtan Ecological Experimental Station of Red Soil (N 28° 15'20'', E 116°55'30'') is located at Yujiang County, Jiangxi Province, representing the hilly area of red soil in southeastern China (coverage of 113.3 km2, about 11.8% of China). It lies in the tropical/sub-tropical monsoon climate zones, with an annual rainfall of 1785 mm and annual mean temperature of 17.8 ℃ . The annual accumulated temperature above 10℃ reaches 5528℃ and 262 days are recorded frost-free. As a region rich in water and thermo resources, it plays a significant role in agricultural production and economic development. As the main soil developed from Quaternary red clay, it is an ideal place to conduct research on red soil ecosystems. Taoyuan field research station of Agro-ecosystem Research, representative of red soil in the hilly areas around Dongting Lake, focuses its research on the structure, succession of red soil agro-ecological areas and their interactions with the productivity, the principles and relevant technologies for agricultural resources management, optimum allocation and sustainable development of regional agriculture. Changshu field research station (N31°33', E120°38') is situated in the Yangtze River Delta, representing the cropland ecosystems in the fast-growing economic area in the Yangtze River delta. Yanting field research station, representative of purple soil (N31°16’, E 105°27’), is located in the hilly area of Sichuan Basin, with an elevation of 420m. It has a mid-subtropical monsoon climate, with an annual mean temperature and precipitation of 17.3°C and 826 mm respectively. The purple soil was developed from Cretaceous and Jurassic purplish shale, and the vegetation was dominated by mixed forest of Alnus cremastogyne and Cypressus funebris with few arbors and grass.

40

B.J. FU ET AL.

There are three field research stations in Northern China. As an important field station located in Huang-Huai-Hai plain, Fengqiu field research station (N35o00’, E114o24’) focuses on long-term monitoring, data accumulation and regional ecosystem dynamics research. Another one in Huang-Huai-Hai alluvial plain, Yucheng field research station is situated in Yucheng, Shandong Province in the alluvial plain of the Yellow River, with a semi-humid, warm temperate monsoon climate. Luancheng field research station (N 37°50', E 114°40'), with an elevation of 50.1m, lies in the semi-humid and warm temperate monsoon climate zone in eastern China. It represents the high-yield agro-ecosystem in cinnamon (gray-yellow) soil at the foot of the Taihang Mountain. There are two field research stations in the Loess Plateau region. The first one is located in Changwu County, Shaanxi Province. Changwu field research station (N35°12′, E107°40′) lies in the central southern part of the Loess Plateau. The elevation ranges from 940m to 1220m. It has a semi-humid, warm-temperate and continental monsoon climate, with an annual average precipitation and temperature of 584mm and 9.1°C, respectively, and 171 frost-free days. It is also known as a typical dryland agricultural area, with the ground water depth reaching 50~80m. Geomorphologically, it represents the hilly and gully region in the Loess Plateau, in which 65% is covered with gully slopes and 35% with highlands. Soil is characterized by dark loessial soil (or Heilu soil), a porous soil. The second one, Ansai field research station (N36°51'30″, E 109°19’23″) is situated in Ansai county, Shaanxi province in the central Loess Plateau, with an elevation ranging from 1068m to 1309m. As a representative of loess hilly areas in the Loess Plateau, it lies in the transitional zone from semi-humid and warm temperate zone to semi-arid climate zone and a transitional area between steppe and deciduous broad-leaved forests in warm-temperate area, with an annual mean temperature and rainfall of 8.8°C and 500mm, respectively. However, this area suffers from serious water and soil erosion mainly due to human activities. In northeastern China, there are two field research stations. Shenyang agroecological experimental station lies in Shenyang, Liaoning province, a representative of agro-ecosystem in the lower reaches of Liaohe plain, which is known as one of the major food production bases in China. Hailun agro-ecological experimental station (N47°26’, E126°38’) is located in the western suburb of Hailun, northeastern Songnen plain, with an elevation of 240m. It has a temperate and continental monsoon climate with dry and cold winter but warm and rainy summer. The rainfall and heat occur in the same season. As one of the three largest areas of black soil in the world, the Songnen plain ranks among the most important food production bases in China because of its high soil organic matter. There is only one field research station in northwestern China. It is named as Fukang Desert Ecological Research Station (N43o45’, E87 o45’), which is situated at the southern edge of Junggar Basin, Sangonghe River-basin of Fukang city, Xinjiang Uygur Autonomous Region. It ranges from the highest peak of eastern Tianshan Mountain---Bogda Peak, at an elevation of 5445m to the southern fringe of Gurbantunggut Desert at an elevation of 460m. A natural landscape belt of 80 km is formed in the site, from alpine glacier, high mountain and sub-alpine meadow to mid-montane forest, lower mountain steppe and desert. It is broadly a representative

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

41

of inland desert of temperate zone in the center of Eurasia. As an area dominated by saline alkali arid soils, it has a continental arid climate, with a mean temperature of 6.6°C. Its annual mean precipitation is 164 mm but the annual potential evaporation reaches 2000mm. The primary goals of these field research stations are to fully understand the structure and function of agro-ecosystems based on long-term observation, experimentation and research on various ecological factors. It is also aimed to reveal the matter cycling and energy transfer patterns between environmental resources and human activities, to establish practical models for wise use of agricultural resources and effective management of agro-ecosystems, and to provide a scientific basis for ecological protection in the representative regions. 4.1.2 Experimental Facilities, Measurement Items and Research Approaches The field research stations of agro-ecosystems under CERN are generally equipped with the following facilities including the soil physio-chemical laboratory, physio-ecological laboratory, laboratory for matter cycling simulation, computer center, library, meteorological observation site, observation site on microclimate gradient, experimental site on nutrient balance in cropland, runoff field and lysimeter. Some of these stations are also provided with plastic tunnel, glass greenhouse and other instruments, including the graphite furnace, atomic absorption spectroscopy, IR gas analyzer, UV-visible spectrometer, oxygen-bomb calorimeter, leaf area meter, photosynthesis evapotranspirometer and TDR soil water measurer, etc. The major items to measure include the routine meteorological observation items and the ecological indicators for crops, such as group dynamics, LAI, dry matter accumulation process, soil water and temperature profile, soil nitrogen, ground water changes, irrigation and runoff. A commonly agreed observation standard and indicators system is adopted by these stations for them to conduct comparative research across the network at regional scale. In addition, demonstration sites are available in these field stations to present and disseminate the best practices in agro-ecosystems. In general, the field research stations under CERN focused on data accumulation based on long-term monitoring of major ecosystems and their dynamics, and the research on the structure, function of these ecosystems and their responses to global change; quality assessment and health diagnosis of cropland ecosystems; wise use of agricultural resources and sustainable development in these regions; mechanism and regulation of productivity in ecosystems. In these stations, a variety of tools are used at different scales. For example, at the site level, emphasis is paid on the test and observation data to explore the process of crop/water and crop/nutrient interactions in cropland ecosystems. Mathematical models are developed to improve the ability to predict the dynamics of cropland ecosystems. Other models related to SPAC, crops, nutrient and water dynamics, are also applied in this process. At the regional scale, mathematical models, combined with GIS and remote sensing technologies, are used to study the spatial variability of the components and processes in cropland ecosystems, their

42

B.J. FU ET AL.

long-term changes under the climatic factors and the trends under global change, based on the data collected from the sites. Therefore, the layout, monitoring and research foci of the agricultural stations under CERN have been shifted from static to dynamic, and from micro to macro-scales. An observatory of agro-ecosystem is established to conduct comparative synthesis research, which is made possible by the sufficient and highly representative sites with standardized measurement and testing methods, instruments and facilities. Research programs of common interests have been implemented to conduct long-term observation, to promote data sharing and synthesis on research findings, to provide information related to environmental change and ecosystem succession, and to develop holistic assessment and prediction models. Significant achievements have been made on the researches at the sitelandscape-regional scales based on CERN so far. For instance, the Program of Water Circulation and Regional Variation for Cropland Ecosystems in Northern China, funded by the Natural Science Foundation of China (NSFC), is based on the agricultural sites under CERN in northern China and aimed to address the issues concerning water consumption process, production process and water use efficiency for the cropland ecosystems, and to fully understand the crop production/water interactions in northern China.

Figure 4. NECT and NSTEC under IGBP Terrestrial Transect (adapted from the synthesis of ‘Peng and Ren, 2000’ and ‘Wang and Gao, 2003’).

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

43

4.2 Landscape pattern and ecological process study across thermo-gradient in China Two major transects are established with the field research stations of CERN, one along thermo-gradient extending from north to south and the other along water gradient from east to west. Some backbone sites along the North East China Transect (NECT) and North-South Transect of Eastern China (NSTEC) under the IGBP Terrestrial Transect, representing the major areas and ecosystems in eastern China, serve as a fundamental platform for long-term synthesis research on ecosystem pattern and process across the site-landscaperegion scales (Figure 4). NECT is mainly a rainfall-driven transect extending from east to west, while NSTEC is a thermo-gradient transect from north to south, as well as the significant change of water gradient (Figure 5). They provide unique conditions for scientists to carry out research on ecosystem pattern and process in eastern China, and crossscale studies on ecosystem change at the site-landscape-region scales. Transect Approach is a unique and valuable tool for global change and landscape ecological study, which aims to link the sites and regions (sitetransect-region), to couple and transfer in various spatio-temporal scales, and to conduct integrated analysis on the interactions between ecosystem structure, function, process, and their driving forces. It can be also used to maximize the efficient integration and use of data resources from different sites, and enhance the scientific understanding by analyzing the temporal succession with spatial pattern and experiments. The transect-based research along environmental gradient is currently focused on the ecosystem pattern and process changes driven by environmental and human factors, and is designed to address some new scientific issues including water, carbon and nitrogen cycling process and their coupling; biodiversity/ecosystem functions interaction; and ecosystem response and adaptation. More emphasis is placed on the comprehensive monitoring of matter, energy and stable isotope fluxes of ecosystems based on observation, test and transect survey. It is also a multi-process and cross-scale approach that integrates the comprehensive monitoring with field experiments, ground monitoring and research with remote sensing measurement, and monitoring data with ecological modelling. So far, studies have been conducted along NECT and some results were achieved, including the development of databases covering vegetation, soil, topography, climate, remote sensing, plant physiology, paleoclimate and paleophyte, the implementation of dynamic monitoring and experiments in representative ecosystems along the transect (Zhou, 2002; Tang, 2003), the development of ecological models at various scales such as the stomatal conductance, climate-vegetation, natural vegetation NPP, biogeochemistry and remote-sensed monitoring ( Zhang et al., 1997; Zhou et al., 2003; Gao et al., 1997), preliminary study on the potential response of NECT to global change, as well as the feedback of natural vegetation NPP and typical steppe and broad-leaved Korean pine ecosystems to climate change. Similarly, comprehensive measurements have been made along NSTEC, covering water, soil, atmosphere and biomass among different cropland ecosystems.

44

B.J. FU ET AL.

Figure 5. Temperature and Precipitation Changes of NECT and NSTEC under IGBP Terrestrial Transect.

Other achievements include the development of a distribution pattern map on biomass and productivity of agricultural ecosystems in the transect to simulate their potential responses to global change (Peng, 2002; Sun et al., 2003), the application of transect approach to study the structure, function and process of agricultural ecosystems in eastern China, the revealing of land use/cover change along NSTEC and its natural and socio-economic implications. In addition, the bio-geographical, biogeochemical and function-process coupling models are developed for the major agricultural ecosystems at various scales (patch, landscape and regional scales). Some policy options are proposed to address the potential impact of global change and greenhouse gases on agricultural production. However, due to financial limitation in transect research, the long-term monitoring, comparative experiment and data accumulation along the two transects,

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

45

as well as the transect-based integrated research, are not fully implemented. Therefore, more efforts are required to establish a platform which integrates monitoring, experiment and research, based on site-transect-network, to address the scientific issues such as the response and adaptability of ecosystems to global change; carbon-nitrogen-water cycling process and their coupling relationships, biodiversity/ecosystem function interactions, the effect of human activities and natural process on ecosystem patterns, modelling and forecasting of ecosystem pattern, structure and function at regional scale. An integrated system for data-model integration will also be made available to simulate the pattern, process, structure and function of interactions across community, landscape and regional scales, which can be applied to quantifying the vegetation pattern, carbon, nitrogen and water flux and balance; change of biodiversity and ecosystem productivity, the impact of global change on regional ecological function, and food security. 4.3 Biodiversity monitoring and management in China Over the past few decades, the growing human activities have intensified the global climate change, landscape fragmentation and environmental pollution, and the rate of species extinction has accelerated globally (Brook et al., 2003). As the most populous country in the world, China is challenged with biodiversity loss. Biodiversity is the earth’s life-support system as a result of about 4 billion years of evolution (Chen, 1993), including the plants, animals, microorganisms, the ecosystems and ecological processes. The loss of biodiversity not only results in the damage of food chain and the disconnection of ecological relationships among various species, but also alters the ecosystem functions. Since it takes a rather long time for the succession of biodiversity and ecosystems, transects along various environmental gradients need to be applied as the natural laboratories to explore the driving mechanisms of biodiversity, vegetation and ecosystem distribution, to predict their dynamics, intensity and scope within limited periods, to analyze the temporal succession with spatial pattern, and to define the roles of various driving forces in biodiversity and ecosystem change. NSTEC and NECT, the two major transects in eastern China, are ideal for scientists to conduct research on biodiversity and ecosystem change and bio-resources management, since the vegetation cover types distributed along the NSTEC from the north to south are the coniferous forests in cold temperate zones to tropical rainforests, while along NECT the deciduous broad-leaved and coniferous mixed forests, grasslands and deserts can be found from east to west (See Figure 6). A number of important results were achieved in the following fields along NECT and NSTEC, including the eco-geographical characters and its relationships with vegetation gradient (Zhang et al., 1997; Ni et al., 1999; Teng et al., 2000; Ni and Zhang, 2000), climate variability and vegetation, vegetation and landscape change, land use/cover change, and agricultural regional change under the impact of global change (Peng and Ren, 2000; Peng, 2001, 2002; Zhou, 2002, 2003; Tang, 2003; Ni and Wang, 2004); the spatial distribution characteristics (Chen et al., 2002) and temporal dynamic changes (Chen et al., 2000, 2001, 2002, 2003) for major tree

46

B.J. FU ET AL.

species in eastern forest areas, C4 plant distribution/climatic factors interaction (Tang, 1999); spore and pollen/vegetation interaction, major plant functional types, spatial distribution and their interactions with climate (Ni, 2003; Wang and Ni, 2005); NDVI seasonal changes/precipitation interaction (Tang and Chen, 2003); thermo driving effect and the water-thermo effect of forests (Zhou, 1997); the productivity pattern of agricultural ecosystems (Qian et al., 2001); and NDVI change of typical vegetation in central southern section of the transects (Sun et al., 2003).

Figure 6. Major Natural Vegetation Types along NECT and NSTEC.

As for the predictive trend in vegetation, such issues were addressed, including the spatial distribution of biomes and its response to global change along NECT (Li, 1995; Tang et al., 1998; Ni, 2000); the remote sensing information-driven model of regional vegetation and its response to global change along NECT (Gao and Zhang, 1997; Gao and Yu, 1998); the potential changes of vegetation and agricultural ecosystem patterns along NSTEC (Peng et al., 2002); the vegetation status under land use constraint and the response of its productivity to global change (Zhang, 1997; Gao and Yu, 1998). In spite of this, we are still uncertain about some issues, such as how does the biodiversity of major grassland and forest ecosystems in China changes with the temperature and precipitation gradients; how will the abrupt changes of water-thermo

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

47

factors and the increasing efficiency of nutrients along temperature and water gradients influence the biodiversity and ecosystem function of grassland and forests; how the climate change and human activities will contribute to the change of ecosystems/vegetation patterns and how these patterns will adapt to the climate change. In such cases, the field stations along NECT and NSTEC need to be organized to conduct biodiversity inventory, ecosystem observation and research, with an objective to study the response and adaptation mechanisms of biodiversity and ecosystem function to global change (water, temperature, nitrogen and their composites), and to understand the pattern of biodiversity with the water-thermo gradients; the response and adaptation of grassland biodiversity and ecosystem function to the abrupt changes of water, heat and nutrients, as well as the impact of their abrupt changes on forest biodiversity and ecosystem functions. Comparative synthesis research on the environmental driving mechanism for the changes of ecosystem/vegetation pattern along the two transects is also necessary to predict the adaptability of natural and human-induced changes of ecosystem/vegetation pattern to climate change, and to identify the thresholds of change for ecosystem/vegetation pattern, the intensity and contribution of human activities. 4.4 Landscape pattern analysis and ecological process at multiple scales 4.4.1 Agricultural landscape and ecological processes study in China Agricultural landscapes are mosaics of natural and human-managed patches that vary in size, shape and arrangement. The spatial pattern exerts important influence on many ecological processes. Based on field research stations and the principle of landscape ecology, the research of agricultural landscape and ecological processes can be carried out in China. Agricultural landscape and ecological processes study in the Loess Plateau There are two agricultural stations (Ansai and Changwu) in the Loess Plateau, which are used for long term monitoring and research. These two stations focus more on soil erosion, rational land use, soil water balance, nutrients cycling and deteriorated ecosystem restoration. Based on the two stations, the scale-pattern-process theory in landscape ecology can be applied in the study of agricultural landscape and its effects on soil and water loss, soil nutrient and soil moisture at multiple scales. At slope scale, there are lots of research plots with different slope lengths, gradients and shapes as well as different crops and management practices in Ansai and Changwu stations. Many researches were carried out to detect the effects of land use or topographical characters on soil erosion, runoff, nutrient loss, and soil moistures by using the monitoring data (Fu et al., 2003; Tian et al., 2005). For example, some research found that farmland was the most susceptible land use type on runoff, soil loss and nitrogen loss, and landscape structure at slope scale may alter soil moistures reference catchments, landscape pattern change and its driving forces can be derived by landscape indices and GIS techniques (Chen et al., 2001;

48

B.J. FU ET AL.

Zhang et al., 2004). For example, some research found that farmland was the most susceptible land use type on runoff, soil loss and nitrogen loss, and landscape structure at slope scale may alter soil moistures and nutrient obviously (Fu et al., 1999; Wang et al., 2001). At small watershed scale, by using the long term monitoring catchments of CERN and loss often consider landscape structure and soil erosion model (Yu, 1996; Liu et al., 2001). As for soil nutrient and moistures, it was found that land use, landscape position and management are main driving factors at small watershed scale (Wang et al., 2001; Wen et al., 2005; Qiu et al., 2003). At regional scale, there are some methods on the researches of landscape pattern and ecological process for Loess Plateau. One is to combine the monitoring data and scientific results of the two stations, and draw some conclusions for the loess area. Another way is using up-scaling methods to study the ecological mechanism of large area (Chang et al., 2005). Furthermore, some models at watershed scale may be used at regional scale and integrated study of multiple scales may be realized for some ecological process, such as land use and soil loss. Agricultural landscape and ecological processes study in Northern China Northern China is the main region of wheat production in China, and three field research stations are established in this region (Yucheng, Fengqiu and Luancheng). Committing to the long term on-site experiments and observations, these stations dealt with the research on ecological and environmental factors and ecological processes, dynamics of structure and function of agro-ecosystems, rational utilization of agricultural resources. The study of agricultural landscape and ecological processes in this area focused on the following aspects: movement and utilization of water, nutrient distribution and leaching, and development of agricultural productivity. At plot scale, the data of field experiments of CERN stations can be used to address the issue of the effects of crop growth on water movement and nutrient leaching. For example, movement and utilization of water may involve soil water, crop water evapotranspiration and demand, crop water consumption and water use efficiency (Zhao et al., 2002; Wang et al., 2003). And the possible reasons for nutrient leaching may be the factors such as spatial distribution of soil type, topographical feature, fertilize and irrigation regimes. As for the crop yield at plot scale, not only the function of water and fertilizer, but also plant structure needs to be explored. Some research has verified that plant structure will change crop yield obviously (Wu and Yang, 1998). At regional scale, spatial distribution of nutrient can be studied by considering soil types (Zhu et al., 2004) and other environmental factors. As for water movement, nutrient leaching and agricultural production, some regional models, such as hydrological model and crop growth model, may offer support to landscape pattern analysis and ecological process research at regional scale. Furthermore, comparative studies resulted from different stations can conduct the research of landscape pattern and ecological process, which may offer valuable base for regional agricultural development.

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

49

Agricultural landscape and ecological processes study in Southern China There are four field research stations under the CERN in south China rice cultivating region, including Changshu, Taoyuan, Yingtan, and Yanting. The geomorphological features of these stations are alluvial plain (Changshu) of the Yangze river delta, red soil hilly areas (Taoyuan and Yingtan), and purple soil area (Yanting). Changshu agricultural experimental station is mainly focused on the theories and practices of intensive agriculture and the pertinent environmental impacts in rapid urbanizing area. The other three stations pay more attention to red soil agroecosystem utilization and sustainable management in the hilly areas. Whereas, the common missions of the stations are oriented to conduct research at the field scales and performing demonstrations of 1) optimization of nutrient cycling (Ding et al., 2004; Wang et al., 2004) and management for farmland ecosystems, 2) construction of models of efficient eco-agriculture, 3) the comprehensive control of land or environmental degradation (Li, 1993), and 4) long-term monitoring of agroecological processes and the environmental variables of the surface and underground water, soil, atmosphere, climate, and biological communities (Zong, 1998a, b; Shi et al., 2002). All these researches, although conducted at small scales, are helpful to improve the understanding of the patterns and processes of the agro-ecosystems and to develop techniques for the wise use of resources and effective control of environmental pollution. At small watershed or regional scales, the researches are extended to landscape dynamics and the driving forces (Yue et al., 1997; Zhang et al., 2003), ecologicalenvironmental-economic quality assessment (Wang et al., 2002; Zhang and Li, 2002), and the management and effects of ecological rehabilitation or reclamation (Lou, 1997). These researches integrated both natural environment and human dimensions in identifying and solving practical problems and thus more relevant for multi-level land use decision making from farmer households and other stakeholders to governments. 4.4.2 Forest landscape dynamics and management in China There are nine field research stations in CERN involved in forest observation. Together with the forest experiment stations from China Forest Research Network (CFERN), 23 stations are involved in forest ecosystems monitoring and ecological research. The forest experiment stations in CERN represent the typical forest ecosystems of the humid needle-broadleaved mixed forest ecoregion of the temperate zone, humid and semi-humid deciduous broadleaved mixed forest ecoregion of warm temperate zone, humid evergreen broadleaved forest ecoregion of the sub-tropical zone, Tibet plateau forest and extremely cold meadow ecoregion, and humid and semi-humid evergreen broadleaved forest ecoregion of south Asia monsoon zone (Fu et al., 2004). The zonal and azonal geographical differentiation of the main forest ecosystem types in China is captured by the CERN and CFERN forest experiment stations. The forest ecosystem types under observation vary

50

B.J. FU ET AL.

widely from pure natural to artificial and from climax to the degraded. The monitoring items of these stations include the dynamics of the climatic and atmospheric conditions, the pattern and processes of biological entities from species to communities, hydrological cycle and water quality, and the dynamics of soil conditions. Apart from these long-term monitoring, the experiment stations usually have their special scientific interests. For example, Huitong forest ecological experiment station (Hunan province, south China) is characterized for the ecological researches on forest plantations (Yu, 2000). These experiment stations form a concrete basis for conducting ecological studies at various spatiotemporal scales. At specific geographical locations, many ecological issues such as C and N cycling (Shen et al., 2003; Tian et al., 2004), the dynamics of biotic and abiotic factors in the processes of forest ecosystem succession or restoration (Tong et al., 2004; Xu et al., 2005), forest ecosystem functions and biodiversity (Xu and Liu, 2005), and forest landscape ecological analyses (Ma and Fu, 2000a, b; Liu et al., 2003; Chang et al., 2004) have become hot research topics recently. The researches based on these experiment stations in the past decade were also summarized into ten branches including carbon cycling, biodiversity, biomass, photosynthesis and hydrology, ecosystem functioning, restoration ecology, ecological modelling, urban forestry, global climate change, and other basic comprehensive forest ecological studies (Feng, 2004). At regional or national scales, these experiment stations facilitate comparative studies (Liu et al., 2003), ecosystem assessments, and long-term environmental change analyses. 5. CONCLUDING REMARK 5.1 Roles of CERN in landscape ecological study Landscape ecology deals with the structure, dynamics, functioning, and scaling of landscapes (Wiens, 1999). It should be regarded as a multidisciplinary, better a trans-disciplinary science where different views and approaches from both natural and social sciences should be involved in a holistic manner (Bastian, 2001). Ten important research topics were proposed for future development of landscape ecology in theory and methodologies (Wu and Hobbs, 2002). CERN, as an important research infrastructure on long term monitoring, serves for at least eight of these research topics concerning data acquisition and the core issues of landscape pattern and ecological process relationships and their scaling. These experiment stations cover agro-ecosystems, forest ecosystems, grassland ecosystems, desert ecosystems, everglade ecosystems, wetland ecosystems, and ocean ecosystems. The functioning of these stations have contributed greatly to landscape ecological studies in China from the aspects of land use structure and ecological processes, human disturbance and landscape health, landscape management and sustainable resource utilization, integration of physical and human factors, and landscape modeling in various regional contexts (Fu and Lu, 2006). All

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

51

these contributions are indispensable for the achievements ever made in landscape ecological studies of China. Furthermore, these experiment stations are also important bases for international research collaborations such as Millennium Ecosystem Assessment. 5.2 Further the study of the interactions between landscape pattern and ecological processes in landscape ecological research Much progress has been made on theory and practice since the birth of landscape ecology as a scientific discipline. However, many problems are still waiting to be addressed in further study. Within which the interactions between landscape pattern and ecological processes remain to be poorly addressed core issue in landscape ecological research both in the international context and in China. However, only after the interactions between landscape pattern and ecological processes are well understood, the landscape ecological theories and applications can be successfully bridged to the societal needs such as resource use and conservation, ecological rehabilitation, land use planning and design, and ecosystem management. China is a developing country with large population and diverse landscape types. It is currently in a stage of rapid socioeconomic development. The drive for socioeconomic development is strong, and environmental quality, resource availability and ecological security are also important concerns for regional sustainable development. Many new problems concerning natural resources utilization, ecological conservation and rehabilitation, and the balancing of interregional development are waiting for solutions and call for the active participation from landscape ecologists. Despite all the past encouraging progresses, landscape ecology in China still lags behind the change of actual societal needs. Facing the theoretical and practical challenges, the core issue of landscape pattern and ecological processes need to be lucubrated in China. Any kinds of landscape ecological research, including theoretical development, modelling, planning and design, need to be based on or verified by certain scientific data from long term monitoring, survey and experiments (Fu and Lu, 2006). Therefore, Long-term monitoring, experiments, surveys, and simulation need to be strengthened. Field experiment stations have been set up since the 1950s to conduct research on such landscape ecological issues as water and thermal balance, agroecology, desertification and soil erosion control (Huang et al., 1990). Many scientific data have been produced from these stations. So far, the environmental and ecological monitoring, experimenting, and data sharing mechanisms have been established and put into operation, such as CERN, CFERN, China Meteorological Data Sharing Services Network (http://data.cma.gov.cn), Data sharing Network of China Earth System Science (http://eng.wdc.cn:8080/Metadata/index.jsp). Therefore, the researches on the interactions between landscape patterns and ecological processes have been receiving more and more data support from both the various kinds of field stations for monitoring and research and the web-based data sharing mechanisms. Further efforts need to be directed to the improvement of spatial distribution and data quality of the field monitoring and research stations.

52

B.J. FU ET AL.

Consequently, the core issues of the relationships between landscape pattern and ecological process and scaling will be more informative during landscape ecological research and applications. At larger spatial scales, systematic ecological modelling is necessary in understanding and regulating the relationships between landscape pattern and ecological processes. The hierarchical patch dynamics (HPD) paradigm and the scaling ladder strategy (Wu and David, 2002) are insightful as theoretical framework for this kind of modelling initiatives. Meanwhile, the land unit approach (Zonneveld, 1989) and land evaluation are believed to be helpful in the integration of landscape pattern with ecological processes. Therefore, multidisciplinary and transdisciplinary approaches need to be encouraged in disentangling the overwhelming complexity of real world landscapes through long term and widely distributed monitoring and experimentation based ecological modelling. AKNOWLEDGEMENT This project was supported by the National Natural Science Foundation of China (Grant No. 40321101 and 30570319). REFERENCES Bastian, O. (2001). Landscape ecology – towards a unified discipline? Landscape Ecology, 16, 757-766. Brook, B.W., Sodhi, N.S. and Ng, P.K.L. (2003). Catastrophic extinctions follow deforestation in Singapore. Nature, 424, 420-423. Chang, X., Chen, X., Liu, G.B. and Qiu, H.J. (2005). Use of the regional models to the scale conversion research aimed at the hilly and gully regions of the Loess plateau. Agricultural Research in the Arid Areas, 23(2), 142-147 (in Chinese) Chang, Y., Bu, R., Hu, Y., Xu, C. and Wang, Q. (2004). Dynamics of forest landscape boundary at Changbai Mountain. Chinese Journal of Applied Ecology, 15, 15-20. Chen, X.W., Zhang, X.S., Zhou, G.S. and Chen, J.Z. (2000). Spatial characteristics and change for tree species (genera) along North East China Transect(NECT). Acta Botanica Sinica, 42(10), 1075-1081. Chen, L.D., Wang, J., Fu, B.J. and Qiu, Y. (2001). Land use change in a small catchment of northern Loess Plateau, China. Agric. Ecosyst. Environ. 86, 163– 172. Chen, L.Z. (1993) Biodiversity in China—Current Status and Conservation Strategies. China Science Press, Beijing. Chen, W.X., Zhou, G.S. and Zhang, X.S. (2003). Spatial characteristics and change for tree species along the Northeast China Transect (NECT). Plant Ecology, 164, 65-74. Chen, X.W. (2001). Change of tree diversity on Northeast China Transect (NECT). Biodiversity and Conservation, 10, 1087-1096. Chen, X.W. (2002). Simulation of shift of range for Betula costata Trautv. And Juglans mandshurica Maim. Along Northeast China Transect (NECT). Polish Journal of Ecology, 50, 397-402. Cooper, L.M., and Sheate, W.R. (2002). Cumulative effects assessment: A review of UK environmental impact statements. Environmental Impact Assessment Review, 22, 415–439 Dale, V.H. (2003). Opportunities for using ecological models for resource management. In Dale, V.H. (Ed.). Ecological Modeling for Resource Management (pp. 3-19). Springer-Verlag, New York. Defries, R.S., Bounoua, L., and Collatz, G.J. (2002). Human modification of the landscape and surface climate in the next fifty years. Global Change Biology, 8, 438-458 Ding, G., Liu, J., Peng, Y., Wang, M., Wang, S., Cheng, H., et al. (2004). Preliminary results of NO2 flux measurement in low atmosphere at rice paddy in Changshu using conditional sampling. Journal of Applied Meteorological Science, 15(4), 456-467. Döös, B.R. (2002). Population growth and loss of arable land. Global Environmental Change, 12, 303–311

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

53

Feng, L. (2004). The new development of Chinese located forest ecosystem research. Journal of Inner Mongolia Agricultural University, 25(4), 58-61. Fu, B.J. and Lu, Y.H. (2006). The progress and perspectives of landscape ecology in China. Progress in Physical Geography, 30(2), 232-244. Fu, B.J., Ma, K.M., Zhou, H.F., and Chen, L.D. (1999). The effect of land use structure on the distribution of soil nutrients in the hilly area of the loess Plateau, China. Chinese Science Bulletin, 44(8), 732-736. Fu, B.J., Wang, J., Chen, L.D., and Qiu, Y. (2003). The effects of land use on soil moisture variation in the Danangou catchment of the Loess Plateau, China. Catena, 54(1-2), 197-213. Gao, Q. and Yu, M. (1998). A model of regional vegetation dynamics and its application to the study of Northeast China Transect (NECT) responses to global change. Global Biogeochemical Cycles, 12, 329-344. Gao, Q., Yu, M. and Zhang, X.S. (1997). A simulation study of the responses of North East China Transect (NECT) to global change - a model of regional vegetation dynamics driven by remote sensing information. Journal of Integrative Plant Ecology, 39(9), 800-810. Gao, Q. and Zhang, X.S. (1997). A simulation study of responses of the Northeast China Transect to elevated CO2 and climate change. Ecological Applications, 7, 470-483. Gustafson, E.J. (1998). Quantifying landscape spatial pattern: what is the state of art? Ecosystems, 1, 143-156. Haase, D. and Halle, L. (2004). Development and perspectives of landscape ecology. Landscape Ecology, 19, 567–569. Haines-Young, R.H. and Chopping, M. (1996). Quantifying landscape structure: a review of landscape indices and their application to forested landscapes. Progress in Physical Geography, 20, 418-445. He, H.S. (2000). An aggregation index (AI) to quantify spatial patterns of landscapes. Landscape Ecology 15, 591-601. Huang, B.W., Zuo, D.K. and Chen, F.Z. (1990). Field experimentation of physical geography in China. Acta Geographica Sinica, 45, 225-234. Li, H.B. and Wu, J.G. (2004). Use and misuse of landscape indices. Landscape Ecology, 19, 389–399 Li, J. (1993). Approach on rational development and utilization of water resources in the red earth hilly area of Qianyanzhou. Natural Resources, 15(3), 48-54. Li. S.C. and Cai, Y.L. (2005). Some scaling issues of geography. Geographical Research, 24, 11-18. Li, X. (1995). Modeling the response of vegetation in Northeast China transect to global change. Journal of Biogeography, 22, 515-522. Liu, C.S., Qi, S., and Shi, M.C. (2001). Process of study on relationship between land use change and soil erosion. Journal of Soil and Water Conservation, 15(5), 10-17. (in Chinese) Liu, S.L., Ma, K.M., Fu, B.J., Kang, Y.X., Zhang, J.Y., and Zhang, X.Y. (2003). The relationship between land form, soil characteristics and plant community structure in the Donglingshan mountain region, Beijing. Acta Phytoecologica Sinica, 27, 496-502. Liu, S.R., Sun, P.S. and Wen, Y.G. (2003). Comparative analysis of hydrological functions of major forest ecosystems in China. Acta Phytoecologica Sinica, 27, 16-22. Lou, H. (1997). The administration and management system for comprehensive development of hilly red soil areas-a case study from the experimental region in Qjanyan, Jiangxl province. 5(4), 62-64. Ma, K.M. and Fu, B.J. (2000a). Spatial neighboring and distribution regularities of landscape types in Donglingshan mountain region, Beijing, China. Acta Ecologica Sinica, 20. 748-752. Ma, K.M. and Fu, B.J. (2000b). Landscape pattern and fragmentation in Donglingshan mountain region. Acta Phytoecologica Sinica, 20, 320-326. Meadows, M.E. and Hoffman, T.M. (2003). Land degradation and climate change in South Africa. The Geographical Journal, 169 (2), 168-177 Ni, J. (2000). Modeling vegetation distribution and net primary production along a precipitation gradient, the Northeast China Transect (NECT). Ekologia-Bratislava, 19, 375-386. Ni, J., Li, Y.Y. and Zhang, X.S.(1999). The Scientific implications of North East China Transect (NECT) in global change research from the perspective of eco-geographical characteristics. Acta Oecologica, 19(5), 662-629. Ni, J. and Wang, G. (2004). Northeast China Transect (NECT): ten-years synthesis and future challenge. Acta Botanica Sinica, 46, 379-391.

54

B.J. FU ET AL.

Ni，J. and Zhang, X.S. (2000) Climate variability, ecological and the Northeast China Transect (NECT). Journal of Arid Environments, 46, 313-325. Norse, D., (1992). A new strategy for feeding a crowded planet. Environment, 34(5), 6–11, 32–39. O’Neill, R.V., Kreummer, J.R. and Gardner, R.H. (1998). Indices of landscape pattern. Landscape Ecology, 13, 153-162. Opdam, P., Foppen, F. and Vos, C. (2002). Bridging the gap between ecology and spatial planning in landscape ecology. Landscape Ecology, 17, 767–779. Peng, S.L., Zhao, P. and Ren, H. (2002). Potential changes of vegetation and agricultural ecosystems pattern along NECT under the stress of global change. Earth Science Frontiers, 9(1), 217-226. Peng, S.L. (2001). Significant progress on the study of the interaction between terrestrial agricultural ecosystems in eastern China and global change. China Fundamental Sciences, 7, 18-20. Peng, S.L. and Ren, H. (2000). The North-South Transect of Eastern China (NSTEC) for global change studies. GCTE News, 16, 6. Qian, J.P. (2001). The Analysis of productivity pattern for agricultural ecosystems along NECT with remote sensed vegetation index. Chinese Journal of Ecology, 20(6), 46-49,77. Qiu, Y., Fu, B.J, Wang, J. and Chen, L.D. (2003). Spatiotemporal prediction of soil moisture content using multiple-linear regression in a small catchment of the Loess Plateau, China. Catena, 54(1-2), 173-195. Shen, W.J., Peng, S.L., Wu, J.G. and Lin, Y.B. (2003). Simulation studies on carbon and nitrogen accumulation and its allocation pattern in forest ecosystems of Heshan in low subtropical China. Acta Phytoecologica Sinica, 27(5), 690-699. Shi, Y., Dai, C. and Lu, L. (2002). Investigation on microorganism community in different kind of red soil from CAS ecological experiment station. Journal of Nanjing Normal University (Natural Science), 25(2), 32-36. Sun, L., Guo, Q.X., Wang, X.C. and Zhou, X.F. (2003). Change of NDVI for typical vegetation types along the Central Southern Section of North-East China Transect (NECT). Chinese Journal of Applied and Environmental Biology, 9(5), 449-454. Tang, H.P. (1999). C4 plant distribution and its relatedness with environmental factors along NECT. Chinese Science Bulletin, 44(4), 416-421. Tang, H.P. and Chen, Y.F. (2003). The seasonal changes of NDVI along NECT and their interaction with climatic factors. Quaternary Sciences, 23(3), 318-325. Tang, H.P., Chen, X.D. and Zhang, X.S. (1998). Preliminary study on the biomes along NECT and their response to global climate change. Acta Botanica Sinica, 22(5), 428-433. Teng, L., Ren, H. and Peng, S.L. (2000). Natural conditions of terrestrial agricultural ecosystems in eastern China along North-South Transect of East China (NSTEC). Ecological Science, 19(4), 1-10. Tian, D.L., Fang, X. and Xiang, W.H. (2004). Carbon density of the Chinese fir plantation ecosystem at Huitong, Hu’nan province. Acta Ecologica Sinica, 24, 2382-2386. Tian, F.X., Wang, Z.L., Niu, Z.H., and Wang, Y. (2005). Experimental research on soil erosion process in Loess hill slope. Agricultural Research in the Arid Areas, 23 (6), 141-146 (in Chinese) Tong, F., Wang, Q., Liu, X. and Xiao, Y. (2004). Dynamics of soil fauna communities during succession process of secondary forests in Changbai Mountain. Chinese Journal of Applied Ecology, 15, 1531-1535. Turner, M.G. (1989). Landscape ecology: the effect of pattern on process. Annu. Rev. Ecol. Syst. 20, 171-97. Turner, M.G. and Gardner, R.H. (1990). Quantitative methods in Landscape Ecology. Springer-Verlag, New York. Verburg, P.H., Veldkamp, A. and Bouna, J. (1999). Land use change under conditions of high population pressure: the case of Java. Global Environmental Change, 9, 303–312. Wang, J., Fu, B.J., Qiu, Y. and Chen, L.D. (2001) Soil nutrients in relation to land use and landscape positin in the semi-arid small catchment on the loess plateau in China. Journal of Arid environments, 48, 537-550. Wang, J., Li, X.G. and Liu, E.M. (2003). The relationship between relative evapotranspiration and leaf area index and surface soil water content in winter wheat field of North China Plain. Chinese Journal of Eco-agriculture, (11) 2, 32-34. (in Chinese) Wang, G.H. and Ni, J. (2005). Response of plant functional types to an environmental gradient on the Northeast China Transect (NECT). Ecological Research (in press).

LANDSCAPE ECOLOGICAL APPLICATION FOR ECOLOGICAL MONITORING

55

Wang, M.Z., Chen, S.F. and Xiong, S.J. (2002). Development and difference of rural eco-economy in low hill red earth areas. Chinese Journal of Eco-Agriculture, 10(1), 104-106. Wang, R. and Gao, Q. (2003). Climate driven changes in shoot density and shoot biomass in Leymus Chinensis (Poaceae) on the northeast China transect (NECT). Global Ecology and Biogeography, 12, 249-259. Wang, X., Zhu, J., Gao, R. and Hosen, Y. (2004). Dynamics and ecological significance of nitrogen wetdeposition in Taihu lake region-taking Changshu agro-ecological experiment station as an example. Chinese Journal of Ecology, 15(9), 1616-1620. Wen, Z.M., Jiao, F., Liu, B.Y., Bu, Y.J. and Jiao, J.Y. (2005) Natural vegetation restoration and soil nutrient dynamics of abandoned farmlands in forest-steppe zone on Loess Plateau. Chinese Journal of Applied Ecology, 16(11), 2025-2029. (in Chinese) Wiens, J.A. (1999). Landscape ecology: the science and the action. Landscape Ecology, 14, 103. Wu, G. and Yang, X. (1998). Relation between structure of tree-belt and wheat yield in paulownia-wheat integrated system. Acta Ecologica Sinica. 18(2), 167-170. (in Chinese) Wu, J. and David, J.L. (2002). A spatially explicit hierarchical approach to modeling complex ecological systems: theory and applications. Ecological Modelling, 153, 7-26. Wu, J. and Hobbs, J. (2002) Key issues and research priorities in landscape ecology: an idiosyncratic synthesis. Landscape Ecology, 17, 355-365. Xu, G.L., Zhou, G.Y., Mo, J.M., Zhou, X.Y., Zhou, X.Y. and Peng, S.J. (2005). The responses of soil fauna composition to forest restoration in Heshan. Acta Ecologica Sinica, 25, 1670-1677. Xu, H. and Liu, W. (2005). Species diversity and distribution of epiphytes in the montane moist evergreen broad-leaved forest in Ailao Mountain, Yunnan. Biodiversity Science, 13(2), 137-147. Yu, Q.G. (1996). Land use and soil erosion study by use of remote sensing information in Hongtagou watershed on rolling Loess Plateau．Journal of Soil Erosion and Soil Conservation, 2(2), 24-31(in Chinese) Yu, X. (2000). Introduction to Huitong forest experiment station of the Chinese Academy of Sciences. Research Progress of CERN, 11(2), 33-34. Yue, T., Cheng, T. and Zhang, H. (1997). Analysis on landscape dynamics and its driving forces and effect-a case study in Qianyanzhou of Taihe county, Jiangxi province. Natural Resources, 19(6), 1926. Zhang, X.S., Gao, Q. and Yang D.A. (1997). Gradient analysis and prediction along NECT. Acta Botanica Sinica, 39(9), 785-799. Zhang, D.X., Li, X.W., Liu, S.X. and Shi, X.Z. (2003). Analysis on the dynamics of cultivated land quantity in Changshu city based on remote sensing data. Geography and Geo-Information Science, 19(3), 38-41. Zhang, D.X., Pan, X.Z., Li, X.W., Shi, X.Z. and Yu, D.S. (2004). Evaluating loss of rice yield caused by city expansion in recent 17 years-a case study of Changshu city from 1984-2001. Scientia Geographica Sinica, 24(1), 31-36. Zhang, H. and Li, J. (2002). Herbivorous animal husbandry in the middle of Jiangxi province: benefit evaluation and development potentiality analysis. Resources Science, 24(4), 61-67. Zhang, Q.J., Fu, B.J., Chen, L.D., Zhao, W.W., Yang, Q.K., Liu, G.B. and Gulinck, H. (2004). Dynamics and driving factors of agriculture landscape in the semi-arid hilly area of the Loess Plateau, China. Agriculture, Ecosystem and Environment, 103, 535-543. Zhao, Q.J., Luo, Y. and Ouyang, Z.Y. (2002). Water use and irrigation management of winter wheat in the North-western Plain of Shandong Province. Progress in Geography, 21(6), 600-608. (in Chinese) Zhou, G.S. (2002). NECT and Global Change: Desertification, Human Activities and Ecosystems. China Meteorological Press, Beijing , 415p. Zhou, G.S., Wang, Y.H. and Jiang, Y.L. (2002). Global change and NECT. Earth Science Frontiers, 9(1), 198-216. Zhou, G.S., Wang, Y.H. and Xu, Z.Z. (2003). Progress of the study on carbon cycle along NECT. Progress in Natural Science, 13(9), 917-922. Zhou, G.Y. (1997). The Role of water and thermo effect of forests in the transect-based research on global change along NECT. Progress on the research based on natural resources and ecological networks, 8(2), 4-10.

56

B.J. FU ET AL.

Zong, H. (1998a). Survey and prospect of the physical-chemical laboratory in Yingtan station of CERN. Research Progress of CERN, 9(2), 10-14. Zong, H. (1998b). The main concerns of long term monitoring research in upland red soil, CERN Yingtan station. Research Progress of CERN, 9(3), 34-40. Zonneveld, I.S. (1989) The land unit-a fundamental concept in landscape ecology, and its applications. Landscape Ecology, 3(2), 67-86.

CHAPTER 4

ECOLOGICAL NETWORKS, FROM CONCEPT TO IMPLEMENTATION

R.H.G. JONGMAN Wageningen UR, Alterra, Post box 47, 6700 AA, Wageningen, the Netherlands

Abstract. The conceptual and theoretical core of landscape ecology links natural sciences with related human sciences and human activity with landscape pattern, process and change and its impacts. Generating ecological networks means modeling species and landscape patterns. The concept of ecological networks is especially applicable in highly fragmented landscapes where species behave as metapopulations. Analysis of habitat availability is an important precondition for planning ecological networks. However, also the communication with the stakeholders is crucial when ecological networks have to be realized. As ecological network planning means biodiversity management outside protected nature reserves and parks, it also means confrontation between interests and finding ways for cooperation between all users of the wider landscape.

1. INTRODUCTION Landscape ecology integrates ecology and geography and in this way it deals with spatial variation in landscapes at a variety of scales. It includes the biophysical and societal causes and consequences of landscape heterogeneity. Above all, it is broadly interdisciplinary. This is surely the case in planning ecological networks. The planning of ecological networks includes not only the ecological modelling, but also the societal debate on implementation and societal benefits and costs. The conceptual and theoretical core of landscape ecology links therefore natural sciences with related human sciences such as the spatial pattern or structure of landscapes and its relationship with processes in the landscape, the relationship of human activity with landscape pattern, process and change and its impacts. Large parts of the European landscapes, but also parts of the African, Japanese, Indonesian, Chinese and Andean landscapes have been in traditional agricultural use for centuries. The history of these landscapes is different from natural landscapes. In 57 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 57–69. © 2007 Springer.

58

R.H.G. JONGMAN

agricultural landscapes the intensity of disturbances is greater; the decisions made by man are the main influence on land use patterns. In the nineteenth century markets often were of local or regional importance. Farmers did sell their products in the region that could be reached with the transport available (Hubert, 1991). In regions such as northern Portugal special regulations existed and still exist to maintain the balance between farmers. At present we live in a world with a world market. Our meat comes from Argentina, cloths from China and cars from Europe or Japan. That also has changed the way we deal with our landscapes and the space that is left for animals. Land use changes are of all times. Under the influence of changes in food demand, caused by demographic events, the cultivated area of Europe has shown considerable fluctuations (Rabbinge et al., 1996). Periods of expansion and periods of contraction of cultivated area occurred all over the world and during all ages. The idea that we are facing a new period of contraction in Europe is therefore not exceptional and to some extent supported by the characteristics of the present situation of increasing technology and stable or even decreasing population. We see that the world market makes the expansion occur in other parts of the world such as China and Latin America. We see two trends in the landscape occurring, homogenisation and fragmentation (Jongman, 2002) where homogenisation means that land is becoming more homogeneous, field become larger and forests become larger. Fragmentation means that the land is ever more dissected by infrastructure and urban constructions. The important question that follows from this is how to mitigate its negative effects for biodiversity. In the last decades landscape ecological principles have become part of biodiversity conservation. Site based nature conservation can only be successful if the conservation sites are huge as it is in Russia. Even then larger carnivores are threatened. Species have especially difficulties to survive in fragmented landscapes. This made nature conservation change from site protection towards conservation of ecological networks including the wider landscape based on principles from population dynamics (McArthur and Wilson, 1967; Opdam, 1991). Nowadays nature reserves and national parks are considered as units within which the biological diversity of species only can be maintained on the long run if they are connected with other larger units (Jongman, 1995). Development of ecological networks is a process that integrates landscape ecological science and societal processes in the phase of problem statement, modelling and planning. It also integrates disciplines, as it has to deal with ecological, institutional and socio-economic aspects. These interactions are always ongoing and makes the process challenging. 2. ECOLOGICAL NETWORKS: NATURE IN THE WIDER COUNTRYSIDE If nature cannot survive in land especially set apart and we have to accept, that birds, mammals, insects and plants move through the countryside, then we also have to accept that they need space. In a planned and intensively used land as we have in

ECOLOGICAL NETWORK

59

many parts of the world this means that nature has to be one of the backbones of spatial planning: it makes spatial planning the director of environmental conservation. It means concretely that beside official site protection through national parks and nature reserves multifunctional zones should be developed and maintained: ecological corridors, greenways or landscape linkages that have aesthetic functions, contribute to an attractive living environment, have an educational function, a recreational function and last but not least an ecological function (Jongman, 2004) Ecological networks are the result of science based nature conservation, of nature conservation planning. Its basis is founded in biogeography, population dynamics, landscape ecology and land use science. The planning process contains ecological elements, but requires also political, land use planning and awareness components. Without the incorporation of these aspects ecological networks cannot survive as a concept and cannot be realised. This means that they should be based on scientifically based models, on tested scenarios and in participative planning procedures. 3. THE MODELLING OF ECOLOGICAL NETWORKS Natural species can migrate over long distances and they also move through the landscape in search of food, shelter and new breeding sites. They travel at different scale levels constructing their own pathways and their own network. Migrating species are especially vulnerable. They cannot be identified as being present at every moment and they often compete with human land use. In Europe many species are adapted to the cultural landscapes of Europe as accessible and non-hostile land with food and shelter. The role of ecological networks will be to maintain and where needed to restore these functions of the landscape. An ecological network should be geared towards an ecosystem (forest, marshland, moors) or species. A strategic choice of target species benefits many more species than an arbitrary sole species in the network design. There are focal species that have broad-scale effects on the ecosystem level (Simberloff, 1998; Dale et al., 2000): turnstone species (top predators, like wolf, brown bear, otter) ecological engineers (beaver, red deer) and umbrella species (red deer). The concept that can be used for assessments in man-dominated landscape in general and for designing ecological networks is the metapopulation concept (Levins, 1970; Opdam, 1988; Hanski and Gilpin, 1997). A metapopulation is a set of populations in a habitat network connected by inter-patch dispersal. A habitat network is a set of habitat patches close enough to have a reasonable level of interpatch dispersal. Habitat is a species-specific term for the set of conditions a species needs to feed, survive and reproduce. Several approaches are available with their advantages and drawbacks: (1) an empirical approach (census based) (2) a fully mechanistic approach (PVA model based), (3) a statistical approach (landscape index-based) and (4) a spatial standard based approach that is a mixture of the first three (Verboom and Pouwels, 2004).

60

R.H.G. JONGMAN

In highly fragmented landscapes, the occurrence of a species at a certain moment in time does not necessarily mean that the species is part of a sustainable population. The reason is that metapopulation dynamics, such as local extinctions and recolonization processes are taking place constantly and reduce the value of single observations. In conservation planning for metapopulations of more than one and mostly many species, it would for example not be a sound strategy to conserve all the patches where a species is found at a certain moment in time and neglect others patches. Moreover, what we see as distribution patterns of species is the result of historical developments in land use and populations can be in a process of adapting to the present day landscape. Probably, the populations are lagging behind the landscape changes (Tilman et al., 1994). Therefore ecological networks cannot be based entirely upon species distribution data but have to be based on a more general long-term strategy. Another method of assessment is using spatially realistic Population Viability Analysis (PVA) to determine the management perspectives for certain species, usually key species, indicator species, or endangered species of specific interest (Lande, 1988; Lankester et al., 1991; Lindenmayer and Possingham, 1994, 1995). As opposed to the distribution data based approach dynamic population processes are taken into account. However, it is time consuming to unravel the life history of species to point out the relevant parameters and find the right values for them. Moreover, such models can hardly be calibrated and/or validated because of their stochastic nature, their long time horizon, and chance fluctuations in real metapopulations and in real landscapes. Because it is so time consuming, such a PVA can be performed for only one or at the most a small number of species (Verboom and Pouwels, 2004). An approach that that combines the advantages of the above methods without their major drawbacks is based upon Ecologically Scaled Landscape Indices (ESLI, Vos et al., 2001), Landscape Cohesion Assessment (Opdam et al., 2003) and the key patch approach (Verboom et al., 2001). Ecologically scaled landscape indices (ESLI’s) take landscape characteristics into account as encountered by the species in the landscapes and it is estimated in carrying capacities. These ESLI’s have a greater power for predicting sustainability of populations than distribution statistics and landscape statistics alone (Vos et al., 2001). At the Alterra institute this approach is elaborated under the name LARCH. In this approach the area or ecosystem is assessed on presence of habitat for the selected target species. Based on the quality and quantity of habitat it is defined what potential populations are, and if these populations can be considered viable. A population is considered sustainable or persistent if the chances of extinction are less than 5 % in 100 years (Shaffer, 1987). In a the Italian part of the Life-Econet project for Emilia Romagna three ecosystem types were selected, which cover most important natural habitat types in the study area: woodland, wetland, and grassland (Bolck et al., 2004). To assess whether these ecosystem types might function for specific wildlife species, species were selected, which can be considered representative for these ecosystems. The selected species operate at a scale that is appropriate for this landscape, with

ECOLOGICAL NETWORK

61

different dispersal ranges and some also sensitive for barriers. For these species was assessed whether the ecosystem still functions as an ecological network. It has been verified with the help of the LARCH model what at present the connectivity and fragmentation is in the provinces of Bologna and Modena. The result of the analysis shows that the provinces have a serious fragmentation problem. Obviously, the area remaining with natural habitat (only 5%) is too small for many species present. The natural areas can only partly function as a network: many species suffer from fragmentation. In the scenarios for future development shapes of the new corridors have been included as planned in the scenario (strips of land with edges, trees, grass and little pond; a "generic" project in order to test the possible connectivity of the scenario. One of the main conclusions is that habitat requirements for most selected species are high. Table 1. Selected species for analysis with LARCH for Emilia Romagna(It); species sensitive for barriers are shaded. Dispersal capacity Habitat type Woodland

Wetlands/ marshland

Barrier sensitivity

small range (0-10 km)

large range (10-50 km)

Sensitive

-

European polecat (Putorius putorius)

Not sensitive

Red-backed shrike (Lanius collurio)

Turtledove (Streptopelia turtur)

Sensitive

-

Sensitive

Italian crested newt (Triturus carniflex) Banded demoiselle (Calopteryx splendens) -

Not sensitive

Stonechat (Saxicola torquata)

Not sensitive Grassland

Bittern (Botaurus stellaris) Yellow wagtail (Motacilla flava) Quail (Coturnix coturnix)

To be effective in conservation planning this ecological knowledge and modelling results must be translated into technical solutions and policy. Design and management of linkages for conservation can be viewed in a biological way, a socio-political way and as a design problem (Bennet, 1999). Analysis of benefits to flora and fauna is an important first step and an essential basis for evaluating design and management of the landscape and of ecological networks. Within an ecological network ecological corridors are species specific and they can have a variety of functions. Knowledge of the ecological structure and processes in the landscape, combined with the behaviour and ecology of species is

62

R.H.G. JONGMAN

of utmost importance in the design of ecological networks and corridors. In all cases the landscape has to be adapted to its ecological function using forests, hedgerows, streams and small forests for guidance and shelter. The design, the related functions and ecosystem services as translated in benefits and costs are key issues in the further process of implementation of the ecological network. Doing so, means, that calculations should be made of the area of different habitats that is needed for maintaining viable populations of species. These are different for different species in different environments. In the polar areas species need more space than in the tropics because of food availability. Grassland bird species need different areas and population sizes than carnivores and also require different landscape structures. When having stated how much and what type of habitat should be available for different species, then it is still not clear for most of us, what the preferred landscape will be. To understand and communicate these ideas, one needs landscape models. This is a way to summarise and make visible what focal species and habitat requirements mean. In the Netherlands the province of Gelderland summarised the multitude of possible focal species in seven landscape models named a focal species and focusing on linkage of specific habitats (Bolck et al., 2004). The focal species are expected to be a proxy for a group of species and more animal and plant species are expected to make use of this kind of landscape structures. In Gelderland the landscape models consist of a planned zone of 250-500 meter wide with a continuous corridor or steppingstones. For walking species with a high dispersal capacity a corridor for movement and foraging might be sufficient in most cases an on the scale of regional planning. Table 2. Overview of the landscape models for ecological corridors developed by the province of Gelderland and based on habitat requirements and landscape structure (Bolck et al., 2004). Landscape model Badger (Meles meles) Crested newt (Triturus cristatus) Lizard Copper (Lycaena phleas) White Admiral (Limenites camilla) Reed warbler (Acrocephalus schoenobaenus) Ide (Leuciscus idus)

Focal species/ habitat Small mammals amphibians Reptiles, butterflies

Characteristics of the landscape zone Wooded banks, small forests (8%), 500 m wide Corridor and stepping stones, 250 m wide

Reed birds

Corridor and stepping stones (1 and 10 ha) with oligotrophic grassland or heathland, 250 m wide Stepping stones (0,5 and 4 ha) oligotrophic grassland or heathland, 250 m wide Stepping stones, well structured landscape, humid forest, 250 m wide Stepping stones (2,5 and 25 ha), reed marsh

Brooks, streams

Natural banks, spawning places

butterflies butterflies

ECOLOGICAL NETWORK

63

For the crested newt (Triturus cristatus) the landscape can be modelled as an ecological corridor with steppingstones embedded in the landscape. Ponds and other wet landscape elements are essential and also will be favouring other amphibians like the green frog (Rana arvalis), tree frog (Hyla arborea), common spadefoot (Pelobates fuscus) and the grass snake (Natrix natrix). Such a small-scale landscape should consist of a coherent network of linear elements. Additionally at least five small ponds of about 500 m2 are found in every kilometre length of the network. The core of the corridor consists of elements 1 ha with natural vegetation of shrub, humid oligotrophic grassland, deciduous wood, wooded banks, drains, ditches, brooks and banks. Depending on their location and function they will have a minimum width (10-15 m), are not more than 100 m apart and mitigating measures have been taken to cross barriers (tunnels, drainpipes). The landscape should contain a sufficient number of ponds with well-developed water and bank vegetation and open spaces; terrestrial habitat consists of shrub, hedges or wooded banks with sufficient dead wood and holes as hiding places. 4. PLANNING AND IMPLEMENTATION OF ECOLOGICAL NETWORKS The land required to maintain or establish linkages may be in private ownership, public (government) ownership, or it may comprise multiple parcels with a diverse range of owners including private individuals, companies, government agencies or authorities and community or conservation groups. Ideally, a long-term arrangement with the responsible land managers is required to ensure that there is an ongoing commitment to the objective of the linkage (Bennett, 1999). Summarising it can be stated that design principles have to be given to assure a successful realisation and functioning: y Minimal width, the wider the corridor the better it will serve multifunctionality; sizes have been given for different situations from 15 metres to 200 and even 600 or 1000 meters wide, varying for urban and rural situations and terrestrial and riverine corridors; y Well established connectivity for species and man, depending on the longitudinal design of the corridor and the barriers in it; y Differences in use should be taken into account; man is using trails during the day, badgers move during the beginning of the night and dawn, amphibians migrate in the early spring and linkages should be adapted to these habits; y Habitat diversity within the ecological corridor; a greater variety makes it more attractive for different species as well as for man; y Accessibility from the surrounding land makes the ecological corridor multifunctional; Ecological networks can be designed at the national, regional and local level. For implementation in the field the local level is the most appropriate; for regional planning and coordination with other land uses such as agriculture, road planning and urbanisation mostly the regional level is essential as at that level land

64

R.H.G. JONGMAN

use-planning decisions are taken. The national and international levels are more important for the spatial coherency within a country and between countries. At this level also the major policy decisions are taken, although, at least in Europe and in the USA, states in the federation and autonomous regions might have an important role in national decision making for biodiversity planning (Jongman et al., 2004). The different levels of decision making and the involvement of several authorities and NGO’s in implementing ecological networks means, that the institutional and socio-economic aspects of ecological networks are as much important as are the pure ecological aspects. In Italy the Province of Bologna made agreements with the municipalities involved to co-finance ecological corridor projects (Bolck et al., 2004). In the Netherlands NGO’s provinces, municipalities and water boards cooperate in the implementation of ecological corridors. To boost the implementation of ecological corridors the Dutch province of Gelderland started the project Green Connections. Realisation of ecological corridors requires an integrated approach because of the interaction with other land use claims such as agriculture water management, infrastructure housing and recreation. The province has asked water boards – for the water based corridors - and municipalities – for the terrestrial corridors - to play a co-ordinating role in the implementation process. The plans for the further development of ecological networks in Europe are ambitious. The 5th Ministerial Conference “Environment of Europe” concluded that “by 2008, all core areas of the Pan-European Ecological Network will be adequately conserved and the Pan European Ecological Network will give guidance to all major national, regional and international land use and planning policies as well as to the operations of relevant economic and financial sectors”. It is obvious that these targets cannot be met without the active cooperation of relevant land use sectors such as agriculture and forestry, and local and regional planning authorities. The Pan-European Ecological Network will expand beyond the “traditional” domain of nature conservation of protected areas. It will include vast stretches of land over which nature conservation authorities and NGOs have no “jurisdiction”. The targets can only be realized in partnerships between the conservation sector (government and NGO) and the various stakeholders involved (ECNC, 2004). Partnerships are built on mutual interests. The interests of the conservation sector are believed to be clear: conserving biodiversity. Who are the other partners (stakeholders) and what are their interests? It is argued that the integrity of an ecological network as landscape mosaic and perceived as part of an integrated regional or national plan can only be sustained with active support of the “various stakeholders”. Generating active stakeholder support for ecological networks has taken many forms. In the case of the “Life ECOnet” project in Cheshire, United Kingdom the approach to gain support, involved five equally important and co-dependent elements: Technical development of a landscape database in GIS and the application of landscape ecology principles; y Assessing and influencing land use policy and instruments; y Demonstrating integrated land use management;

ECOLOGICAL NETWORK

65

y Engaging stakeholders; y Dissemination. Two important principles were embedded in these elements. The approach, and the resulting ecological network, must allow integration of environmental issues with socio-economic functions of the landscape and the acceptance of the landowners and consumers of the landscape. Secondly, the approach must provide an identifiable product on which the varied skills, knowledge and attitudes of stakeholders can focus (James, undated). It is clear that in this approach the ecological modelling is only the first step. The four following steps involve the integration of ecological knowledge into society and policy decisions. In the case of Estonia the approach to gain support took the form of meetings and public campaigns with emphasis placed on (Sepp and Kaasik, 2002): y Multifunctional nature of ecological networks (e.g. increased environmental health conditions, recreational opportunities); y Conservation of “flagship species” to highlight the importance of biodiversity conservation; and y The accommodation of semi-natural habitats or other “use areas” that allow traditional farming practices in the networks. In many cases in the USA, like the Yellowstone – to – Yukon ecological network, the initiative did not come from government (Yellowstone to Yukon Conservation Initiative, 2006). As most Northern American Greenway plans Y2Y is very much a grassroots initiative enjoying support from a large variety of NGOs and other civil society organisations (360 in total) with the objective to ensure that the eco-region continues to support natural and human communities. In a number of states in the USA (Florida, Georgia, Massachusetts, Rhode Island) the state has embraced these plans into Statewide Greenway Plans based on the integration of biodiversity and civil interest issues (Florida Greenway Commission, 1994). Comparable grassroots-based plans are developed in Portugal around the cities of Lisbon, Porto and Coimbra (Machado et al., 1997). Here the initiative has been a combination of universities and NGO’s. The support from the authorities - and therefore its realisation – is still a difficult process. What these cases have in common is that they focus not only on the conservation of biodiversity but also accommodate the exploitation and consumption of natural resources (Ahern, 2004). Serious efforts are made both to buffer sites of high conservation value from potentially damaging forms of land use and to find ways of reconciling the exploitation of natural resources with biodiversity conservation (Bennett and Wit, 2001). The eastern section of the Netherlands National Ecological Network (EHS) provides a case to illustrate the possible benefits of an ecological network. In this case the ecological network has multifunctional objectives. Conservation and restoration of nature and biodiversity are priorities but they are not the only objectives of the EHS. Such as large claim for space in the densely populated Netherlands can only be justified if it also provides a solution for other problems and needs:

66

R.H.G. JONGMAN

y

Realised environmental objectives through the production of clean water, water management, sustainable use of natural resources (such as timber) and the absorption of CO2; y The protection of important rural, cultural-historical, archaeological and geological values, fulfilling – under conditions – important recreational functions, sustainable agriculture and fisheries, transport over water (under conditions); y An attractive environment for living and business locations by maintaining high-priced qualities such as green space and tranquillity; y Provide space for people to relax and experience nature. Meeting the requirements of nature while at the same time taking into account local stakeholders’ wishes creates public support and willingness of third parties to invest in the areas. The public is prepared to pay taxes and entry fees; farmers (sometimes against payment of conservation subsidies) are willing to consider adapted land management options; owners of country estates and small businesses are interested in investment in nature (Ministry of ANF, 2004). Examples of tangible benefits in the Dutch province of Gelderland are: y Investments in nature-based tourism and recreation generate employment and incomes; y Nature as catalyst for investment prompt estate agents, water utilities and sand extraction companies to expand the acreage of land under conservation management hereby adding to the value of the ecological network while boosting their production and profit; y The enhanced value of nature allows the introduction of innovative and selfsustaining payment mechanisms for farmers to maintain environmentally valuable landscapes (from growing maize to growing nature); y The nature landscape provides clean drinking water and increasingly allows temporary storage of excess river water that may otherwise threaten lowlying population centres. Formulating it in extremes and in line with the recent thinking of “nature has to pay for itself” a multifunctional ecological network may become an opportunity for rural development rather than a (short-term) cost to society. Appropriate planning and control would ensure a rural development that is sustainable and as such contributing to the natural resources the development depends upon, truly a win-win scenario. Having defined an ecological network as a landscape mosaic with both biodiversity conservation as well as sustainable utilisation objectives it will be clear that this cannot be easily planned when the aim is to optimise the balance between the objectives. However, planning of ecological networks cannot be done without including all relevant stakeholders. An adequate institutional context is needed. The landscape or regional scale involves long-term processes, operate across an array of administrative units and embrace a large number and wide range of stakeholders depending heavily on the

ECOLOGICAL NETWORK

67

(harmonised) institutional setting that should be conducive for stakeholder involvement (e.g. respected rule of law, robust zoning arrangements, effective enforcement procedures, financial security for corporate stakeholders, clearly defined public and private ownership patterns (Bennett, 2004; Somma, 2004). It may even be that new multi-stakeholder organizations, platforms or networks are required to ensure the delivery of results that were previously the domain of narrowfocussed (typically environmental) agencies (Miller, 1996). The foundation of a multi-stakeholder process towards the planning and development of an “ecology and development network” requires a shared vision amongst the stakeholders. This vision specifically needs to foster stakeholder participation and convincing of the joint interests. This is especially challenging given the complex social and economic implications of working at a large geographical scale. This also means that far from the ecological models flexibility is needed in setting goals as this is important to make use of positive action in a region or community to join forces in common interest cases. Planners and managers of an “ecology and development network” need a range of social, environmental and economic information at local, regional and national scale (Rientjes, 2000). Information about the importance of ecosystem goods and services is among others required to mobilise public support for the network and to empower local stakeholders to participate meaningfully in the decisions that affect their lives. Integrating socio-economic information with environmental information would provide new perspectives on sustainable use of biodiversity of the network. However, most information remains sector specific and lacks an analytical and holistic perspective. Bringing this information together in a learning environment is vital because the success of the network will depend upon the stakeholders being fully informed on how the project will affect them. A broad-based initiative such as an ecological network brings along process management challenges. Developing a comprehensive proposal that can meet all strategic objectives, collecting and collating data, bringing together all stakeholders and ensuring their commitment, attracting funding, and ensuring effective implementation requires substantial investment in the management process, and the adoption of an integrative and adaptive approach. Lessons drawn show that process management does not have to be the prerogative of government (Bennett, 2004). The ecological network provides environmental goods and services that have a direct use value such as timber and game, recreation and human habitat and indirect use values (watershed protection, climate regulation, erosion control, maintenance of biodiversity). The possible uses have different meanings for different stakeholders and in order to facilitate decision-making about the importance of an ecosystem it is vital to engage in valuation of these goods and services to allow trade-offs (Lette and Rozemeijer, 2005).

68

R.H.G. JONGMAN

REFERENCES Bennett, A.F. (1999). Linkages in the Landscape. The Role of Corridors and Connectivity in Wildlife Conservation. IUCN, Gland, Switzerland and Cambridge, UK. Bennett, G. (2004). Integrating Biodiversity Conservation and Sustainable Use, Lessons Learnt from Ecological Networks. IUCN Gland. Bennet, G. and Wit, P. (2001). The Development and Application of Ecological Networks. A Review of Proposals, Plans and Programmes. AIDenvironment, Amsterdam, 131p. Bolck, M., De Togni, G. Van der Sluis, Th and Jongman, R.H.G. (2004). From models to reality: design and implementation process. In R.H.G. Jongman and G. Pungetti (Eds.), Ecological Networks and Greenways, Concept, Design and Implementation (pp. 128-150). Cambridge University Press. Dale, V.H. (2000). ESA report, Ecological principles and guidelines for managing the use of land. Ecological Applications, 10, 639-670 Florida Greenways Commission (1994). Creating a Statewide Greenways System, for People, for Wildlife, for Florida. Report to the Governor. Tallahassee. Hanski, I.A, Gilpin, M.E. (1997). Metapopulation Biology. Academic Press, London. Hubert, B. (1991). Changing land uses in the Provence (France): Multiple land use as a management tool. In J. Baudry and R.G.H. Bunce (Eds.), Land Abandonment and Its Role in Conservation (pp. 31-52). Options Mediterranéennes Série Séminaires 15. ECNC. (2004). The Pan-European Ecological Network and People – Background paper as well as “Conclusions and Recommendations”of the 2004 seminar on “People and PEEN” in The Hague, the Netherlands (ECNC/LNV). James, P. Undated. Ecological Networks: Creating Landscapes for People and Wildlife. Jongman, R.H.G. (1995). Nature Conservation Planning in Europe: Developing Ecological Networks. Landscape and Urban Planning, 32,169-183. Jongman R.H.G. (2002). Homogenisation and fragmentation of the European landscape: ecological consequences and solutions. Landscape and urban planning, 58, 211-221 Jongman, R.H.G., Külvik, M. and Kristiansen. I. (2004). European ecological networks and greenways. Landscape and Urban Planning, 68, 305-319 Jongman, R.H.G. and Pungetti, G.P. (2004). Ecological Networks and Greenways: Concept, Design and Implementation. Cambridge University Press. Lande R. (1988). Demographic models of the northern spotted owl (Strix occidentalis caurina). Oecologia, 75, 601-607. Lankester, K., van Apeldoorn, R.C. Meelis, E. and Verboom, J. (1991). Management perspectives for populations of the Eurasian badger Meles meles in a fragmented landscape. Journal of Applied Ecology, 28, 561-573. Lette H. and N. Rozemeijer. (2005). Broadening and Diversifying the Financial Basis for Sustainable Forest Management and Nature Conservation. IAC/WUR, the Netherlands. Levins, R., 1970. Extinction. In M. Gerstenhaber (Ed.). Some Mathematical Problems in Biology (pp. 77-107), American Mathematical Society, Providence. Lindenmayer, D.B. and Possingham, H.P. (1994). The Risk of Extinction: Ranking Management Options for Leadbeater’s Possum Using PVA. Centre for Resource and Environmental Studies, ANU, Canberra, Australia. Lindenmayer, D.B. and Possingham, H.P. (1995). Modelling the viability of metapopulations of the endangered Leadbeater’s possum in south-eastern Australia. Biodiversity and Conservation, 4, 9841018. Machado, J.R., Ahern, J., Da Saraiva G, Da Silva M.A., Rocha J., Ferreira, J.C., Sousa, P.M. and Roqueta, R. (1997). Greenways Network for the Metropolitan Area of Lisbon. In J.R. Machado and J. Ahern (Eds.), Environmental Challenges in an expanding Urban World and the role of emerging Information Technologies (pp. 281-289). CNIG Lisbon Portugal. Miller, K. (1996). Balancing the Scales: Guidelines for Increasing Biodiversity’s Chances through Bioregional Management. WRI, Washington, D.C Ministry of ANF (2004). Ecological Networks: Experiences in the Netherlands. “A Joint Responsibility for Connectivity”. Working paper.

ECOLOGICAL NETWORK

69

Opdam, P. (1988). Populations in fragmented landscape. In K.-F. Schreiber (Ed.), Connectivity in Landscape Ecology. Proceedings second International Seminar of the IALE in Münster: Münstersche Geographische Arbeiten, 29, 75-77. Opdam, P. (1991). Metapopulation theory and habitat fragmentation: a review of holarctic breeding bird studies. Landscape Ecology, 5, 93-106 Opdam P., Steingröver, E. and van Rooij, S. (2003). Ecological networks: A spatial concept for multiactor planning of sustainable landscapes. Landscape and Urban Planning, 75, 322-332 Rabbinge R., Van Latesteijn, H.C. and Smeets, P.J.A.M. (1996). Planning consequences of long-term land use scenarios in the European Union. In R.H.G. Jongman (Ed.), Ecological and Landscape Consequences of Land Use Change in Europe (pp. 36-51). Proceedings of the first ECNC seminar on land use change and its ecological consequences. ECNC series on Man and Nature 2. Rientjes, S. (2000). Communicating Nature Conservation. ECNC Sepp, K. and Kaasik, A. (2002). Development of National Ecological Networks in the Baltic Countries in the Framework of the Pan-European Ecological Network. IUCN Central Europe. Shaffer, M.L. (1987). Minimum Viable Populations: coping with uncertainty. In M.E. Soulé (Ed.), Viable Populations for Conservation (pp. 69-86). Cambridge University Press. Somma, D., Aued, M.B. and Bachman, L. (2004). The ecological Networks development in the Yungas, Argentina: planning, economic and social aspects. In R.H.G. Jongman and G. Pungetti (Eds.): Ecological Networks and Greenways, Concept, Design and Implementation (pp. 251- 269), Cambridge University Press. Tilman, D., May, R.M., Lehman, C.L., Nowak, M.A. (1994). Habitat destruction and the extinction debt. Nature, 371, 65-66. Verboom, J., R. Foppen, P. Chardon, P. Opdam, and P. Luttikhuizen. 2001. Introducing the key patch appoach for habitat networks with persistent populations: an example for marshland birds. Biological Conservation, 100, 89-101 Verboom J. and Pouwels, R. (2004). Ecological functioning of ecological networks: a species perspective. In R.H.G. Jongman and G. Pungetti (Eds.), Ecological networks and greenways, concept, design and implementation (pp. 56- 72). Cambridge University Press. Vos, C.C., Verboom, J., Opdam, P., Ter Braak, C.J.F. (2001). Towards ecologically scaled landscape indices. American Naturalist, 183, 24-41. Yellowstone to Yukon Conservation Initiative (2006). 2005 Year End Report Canmore Canada, 10 pp.

CHAPTER 5

LANDSCAPE CHANGES IN JAPAN BASED ON NATIONAL GRID MAPS

N. NAKAGOSHI, J.-E. KIM Graduate School for International Development and Cooperation, Hiroshima University, Higashi-Hiroshima, Japan

Abstract. Landscape changes in Japan based on a national grid map were examined adding current data. For detecting changed regions, we classified all of the prefectures in Japan based on naturalness to know the land use characteristics. The number of cells and percentage of degree of naturalness and limited habitats, such as wetland and tidal land decreased showing nature distraction in Japan. The area of secondary Pinus densiflora forest as representative vegetation in rural landscapes also decreased, which may be dieback due to pine wilt disease and ecological succession. Landscape types (or land use) based on special characteristics of naturalness were classified into six groups; natural forests, secondary forests and grasslands, secondary forests and plantation, agricultural fields, and urban areas. Five prefectures saw and change to plantation type from agricultural field type during 1990s. Japan was shown to have urbanization and decreasing habitat complexity, and pine wilt disease and abandoned rural landscape also led to decreasing Japanese representative cultural landscapes. These land uses and management of landscape led to changes in habitat environments with decreasing habitat complexity, ultimately having an effect on biodiversity, now an important national environmental interest in Japan. Therefore, landscape planning and management have to work to halt the decrease in habitat complexity and quality.

1. INTRODUCTION Landscape changes due to multiple causes are the main issues in the world nowadays. Natural forces such as typhoons (same as cyclones and hurricanes), earthquakes, tsunami, and wildfire are some large and significant factors. However, it is widely known that the main driving force of landscape changes is human impacts (Nakagoshi and Ohta, 1992; Nakagoshi, 1995). Landscape changes consequently led to the destruction of habitat as well as decreasing biodiversity, which in turn affects whole ecosystems. In addition, it also

71 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 71–80. © 2007 Springer.

72

N. NAKAGOSHI AND J.-E. KIM

leads to decreasing historical cultural landscape, cultural amenity, and biological resources, so that landscape changes influence multiple perspectives. Therefore, landscape ecologists have carried out studies to identify land use and land cover changes and their effects on a large scale. In particular, many of them have studied landscape changes and their effects on vegetation (Hong, 1998; Olsson et al., 2000; Jenerette and Wu, 2001; Fujihara et al., 2002; Tasser and Tappeiner, 2002; Jobin et al., 2003; Wilson et al., 2003; Mou et al., 2005 etc.). Many of these studies are useful for sustainable landscape planning and management in decisionmaking. However, national scale maps have been used for fundamental information of land use/cover change and are very useful and widely used for basic information in sustainable planning and management (Haines-Young, 1992; Sanderson and Rushton, 1995; Nakagoshi et al., 1998; Hong et al., 2005). The Japanese Government made a national grid map to gather for fundamental data of natural environmental conservation in the 1980s. It is also useful for environmental impact assessment, and planning and management of Japanese landscapes (Nakagoshi and Naito, 1997; Nakagoshi et al., 1998; Himiyama, 1999). So far surveys for the national grip map have been conducted five times. The first time the survey was carried out in 1973, and the second time was in 1979. The third survey was carried out from 1983 to 1986 (reported in 1988), and fourth one was carried out from 1988 to 1992 (reported in 1994). Finally, the fifth survey was carried out from 1994 to 1998 (reported in 2004). The government in Japanese officially published these reports. The aim of this study is to grasp the trends in landscape changes through surveyed data using the national grip map. In addition, we hope that this paper will provide useful information for landscape planning and management in Japan. 2. DATA SOURCES AND ANALYSIS 2.1 Data source Vegetation maps in Japan are based on the phytosociological method. A standard vegetation map for the whole of Japan, which is an actual vegetation map at a scale of 1:200,000, was produced for the Cultural Agency of the Ministry of Education, Science and Culture from 1969 to 1976. The Japanese Environmental Agency (Ministry of Environment of Japan) succeeded in producing actual and potential vegetation maps on a scale of 1:50,000 in 1973. These maps were based on both aerial photographs and field surveys. Grip maps have been routinely digitized since 1973. The cell size of grid maps is approximately 1 km × 1 km and similar grid maps of actual and potential vegetation maps were made during the third survey. The attribute of a cell is typically the largest vegetation type in the whole cell. However, the Environment Agency used a different method in the fourth survey. Each cell has a central circle that is 250m in diameter (Figure 1). The central circle was taken instead of the whole cell that is due to proper the small vegetation patch.

LANDSCAPE CHANGE ON NATIONAL GRID MAP

73

This method can minimize the risk of missing small vegetation patches and including errors (Nakagoshi et al., 1998). 1km

250m 1km

Figure 1. Cell size and shape construct from Environmental Agency of Japan.

2.2 Statistical analysis Each prefecture has a different historical background, climate, biogeography, geological feature and so on. All 47 prefectures of Japan were analysed to make a group according to vegetation naturalness criteria and show the classification of landscape characteristics of Japan. Since the third survey, cluster analysis using similarity index has been carried out to identify landscape characteristics with land use changes in Japan. The similarity index was used to show the percentage of naturalness area (Nakagoshi and Naito, 1997). This analysis used Morisita’s similarity index Cλ(p) (Morisita 1959). The formula is as follows: n

2

∑p

1i p2i

i =1

Cλ ( p ) =

n

λ1( p ) + λ 2( p )

∑p ∑p 1i

i =1

∑p

∑p

2 1i

i =1 n

λ1( p ) = (

∑p

1i )

i =1

2i

i =1 n

n

where

,

n

i =1 n

, λ 2( p ) = 2

2 2i

(

∑p

2i )

2

i =1

P1i, P2i = area in percentage of naturalness degree i in the naturalness degree of each prefecture.

74

N. NAKAGOSHI AND J.-E. KIM

3. RESULTS 3.1 Examination of vegetation landscapes The number of cells and percentage of vegetation types according to different degrees of naturalness is shows in Table 1. The naturalness is estimated according to absence of human impacts. This is ranked from 10 to 1 and is based on the concept of potential natural vegetation in a structural comparison of phytocenoses (Zerbe, 1998). If certain vegetation is close to potential natural vegetation it will rank highly. 10, 9, 8, 7, and 5 was continuously decreased from the third survey. 6 (plantations), 3 (orchards) increased between third and fourth survey, but they decreased between the fourth and fifth surveys. 4, 2, and 1 increased continuously. Table 1. Number of cells and percentages of vegetation types according to different degree of naturalness for the three survey periods.

Vegetation type by naturalness 10 Natural grassland 9 Natural forest 8 Secondary forest (replaced forest) 7 Secondary forest 6 Plantations 5 Secondary tall grassland 4 Secondary short grassland 3 Orchards 2 Paddy and upland fields 1 Urban areas, exploited land Natural open land Water surfaces Unidentified Total

3rd survey (1983 – 1986)

4th survey (1988 – 1992)

5th survey (1994 – 1998)

Number of cell

%

Number of cell

%

Number of cell

%

4,038 66,797

1.1 18.2

4,011 66,394

1.1 18.0

3,993 65,824

1.1 17.9

20,046 70,484 91,029 5,737 5,939 6,798 76,945 14,841 1,392 4,170 72

5.4 19.1 24.7 1.6 1.6 1.8 20.9 4.0 0.4 1.1 0.0

19,733 69,030 92,072 5,626 6,498 6,817 77,311 15,420 1,416 4,211 71

5.4 18.7 25.0 1.5 1.8 1.8 21.0 4.2 0.4 1.1 0.0

19,598 68,540 91,414 5,568 7,591 6,788 77,695 15,999 1,420 4,227 70

5.3 18.6 24.8 1.5 2.1 1.8 21.1 4.3 0.4 1.1 0.0

368,470

100

368,610

100

368,727

100

The number of cell and percentage of each vegetation type in whole of Japan are shown in Table 2. Alpine and artic primary vegetation did not change. Subalpine/boreal, Fagetea, and Camellietea primary vegetation continuously decreased from the third survey. Riparian, wetland, salt marsh and desert vegetation as a limited habitat continuously decreased. Monoculture conifer plantations and agricultural fields increased between the third and fourth surveys even in the economic slump of forestry and agriculture. However, by the fifth survey these vegetation types had decreased. Others containing residential and commercial areas also increased due to urbanization.

LANDSCAPE CHANGE ON NATIONAL GRID MAP

75

Table 2. Number of cells and percentages according to vegetation types for the three survey periods. 3rd survey (1983 – 1986)

Vegetation type

4th survey (1988 – 1992)

5th survey (1994 – 1998)

Number of cell

%

Number of cell

%

Number of cell

%

Alpine and arctic primary vegetation Sub-alpine/boreal primary vegetation Sub-alpine/boreal secondary vegetation Fagetea primary vegetation Fagetea secondary vegetation Camellietea primary vegetation Camellietea secondary vegetation Riparian, wetland, salt marsh and desert vegetation Plantation, agricultural field Others

1,158 16,147 1,050 45,148 42,232 6,070 58,112 2,513

0.3 4.4 0.3 12.3 11.5 1.6 15.7 0.7

1,158 16,110 1,074 44,704 41,612 5,964 57,357 2,488

0.3 4.4 0.3 12.1 11.3 1.6 15.5 0.7

1,158 15,947 1,266 44,328 41,860 5,925 57,110 2,478

0.3 4.3 0.3 12.0 11.4 1.6 15.5 0.7

169,234 26,806

46.0 7.3

170,598 27,545

46.3 7.4

170,508 28,147

46.2 7.6

Total

368,470

368,610

100

368,727

100

100

(ha) 10000

Primary forests

Secondary forests

Pinus thunbergii forest

Evergreen Quercus forest

Quercus serrata forest

Carpinus forest

Others

Percea forest

Fagus forest

Pinus densiflora forest

-40000

Quercus crispula forest

-30000

Evergreen Quercus forest

-20000

Castanopsis forest

-10000

Fagus forest

0

Total

-50000

-60000

Figure 2. Changes of area according to representative forests in Japan between the fourth and fifth surveys.

76

N. NAKAGOSHI AND J.-E. KIM

Figure 2 shows the changes of area according to representative vegetation. Vegetation decreased except for some primary vegetation and Pinus thunbergii forest in secondary vegetation. The biggest decrease in area was seen in P. densiflora forest in secondary vegetation. 3.2 Cluster analysis Table 3 shows the summarized cluster analysis since the third survey. Six groups of landscape characteristics were identified. These groups can be arranged from 1 to 6 in relation to land use/cover types. Moreover, these groups were generally based on naturalness and so they are also related to habitat environments. Group 1 (natural forest), 2 (secondary forest and grassland), 3 (secondary forest and plantation), and 6 (urban area) maintained the same ecosystem characteristics. However, five prefectures, Aomori, Aichi, Fukuoka, Kagoshima, and Saga, in group 5 (agricultural field) changed to group 4 (plantation) between the fourth and fifth surveys. 4. DISCUSSION 4.1 National trends The naturalness according to degree of human impact may be strongly influenced by land use/cover types (Haines-Young, 1992). Moreover, naturalness and habitat complexity (or habitat type) may be associated with diversity. Since the third survey highly ranked naturalness/vegetation type from 6 to 10 continuously decreased in terms of number of cells. Those from 1 to 4 continuously increased in terms of number of cells, especially due to increasing urban areas, exploited lands, and secondary short grassland (golf course, ski slope etc). In other words, species diversity may be influenced by decreased habitat quality. According to Ministry of Environment of Japan, 1726 species in plants and 245 species in animals are recorded in Red Data Book due to decreasing habitats (http://www.biodic.go.jp/english/rdb/rdb_e.html). Primary vegetation according to vegetation type may have been destroyed and then changed to secondary vegetation. Decreases in unique and limited habitat types such as riparian, wetland, salt marsh, and desert vegetation have been influenced by concrete pavements of dikes, urbanization, reclamation, and so on. These habitat types generally occupied a limited area. Therefore, the meaning of decreasing limited area may have critical meaning on biodiversity than other habitat types. Most representative vegetation areas decreased, with the area of secondary Pinus densiflora forest seeing the biggest decreased. This is mainly due to the wide spread of pine wilt disease and abandoned rural forest. Pine wilt disease was widespread in Japan, except in Aomori and Hokkaido Prefectures until 1971 due to cool temperatures.

LANDSCAPE CHANGE ON NATIONAL GRID MAP

77

Table 3. Summary of cluster analysis based on naturalness. 3rd (1983-1986) Group 1 Natural forest

Group 2 Secondary forest and Grassland

Group 3 Secondary forest and Plantation

Group 4 Plantation

Group 5 Agricultural fields

Group 6 Urban areas

Hokkaido Okinawa Fukushima Ishikawa Fukui Hyogo Shimane Shiga Kagawa Kyoto Okayama Yamaguchi Hiroshima Iwate Nagano Gifu Yamanashi Akita Gunma Miyagi Tochigi Shizuoka Ehime Oita Tottori Tokushima Mie Kumamoto Nara Miyazaki Wakayama Kochi

Ibaraki Saitama Chiba Nagasaki Yamagata Niigata Toyama Aomori Kagoshima Aichi Fukuoka Saga Tokyo Kanagawa Osaka

4th(1988-1992)

5th(1994-1998)

Hokkaido Okinawa Fukushima Ishikawa Fukui Hyogo Shimane Shiga Kagawa Kyoto Okayama Yamaguchi Hiroshima Iwate Nagano Gifu Yamanashi Akita Gunma Miyagi Tochigi Shizuoka Ehime Oita Tottori Tokushima Mie Kumamoto Nara Miyazaki Wakayama Kochi Aomori Kagoshima Aichi Fukuoka Saga Ibaraki Saitama Chiba Nagasaki Yamagata Niigata Toyama

Hokkaido Okinawa Fukushima Ishikawa Fukui Hyogo Shimane Shiga Kagawa Kyoto Okayama Yamaguchi Hiroshima Iwate Nagano Gifu Yamanashi Akita Gunma Miyagi Tochigi Shizuoka Ehime Oita Tottori Tokushima Mie Kumamoto Nara Miyazaki Wakayama Kochi Aomori Kagoshima Aichi Fukuoka Saga Ibaraki Saitama Chiba Nagasaki Yamagata Niigata Toyama

Tokyo Kanagawa Osaka

Tokyo Kanagawa Osaka

78

N. NAKAGOSHI AND J.-E. KIM

The loss of timber productions has increased to more than 1 million m3 since the 1970’s (Mamiya, 1988). In addition, Mamiya (1988) reported that around 650,000 ha of pine forest has been infected by the disease, which is 25 % of the total area of Japan’s pine forests. Secondary P. densiflora forests have occurred after disturbances such as clearcutting, land clearing, and fire. Moreover, P. densiflora is one of the main component vegetation types of rural forests and is in general found in rural forests due to human impacts (Hong, 1998; Nakagoshi and Hong, 2001; Fujihara et al., 2002; Takeuchi et al., 2003). It used to be a very important resource as timber and fuel. However, economic development since the 1960’s has strongly influenced human life in rural areas. Propane gas, oil, and cheap imported timber have come to be used instead of P. densiflora. This is also one of the major factors in vegetation dynamics. Vegetation dynamics of P. densiflora was show to Quercus spp. and evergreen broad-leaved trees (Nakagoshi, 1995; Kamada and Nakagoshi, 1996; Hong, 1998; Fujihara et al., 2002). Moreover, physical conditions bring about drastic changes of forest environments, such as light intensity and soil moisture change during vegetation dynamics. Finally, changing physical conditions will affect understory vegetation composition. However, that reasons not only accelerate the vegetation dynamics but also changes in landscape characteristics in Japan. 4.2 Regional trends National Grip maps were used by the central government of Japan for land use planning and management. These maps were also used for land use planning and management at a regional and prefectural level (Himiyama, 1999). The benefit of making land use strategy policies at a prefectural level is that appropriate planning can be made to cater to the specific environmental characteristics and circumstances of each prefecture. Therefore, regional and prefectural scales influenced land use changes of Japan. The classification of landscape characteristics based on naturalness was toward plantation from agricultural fields in Aomori, Kagoshima, Aichi, Fukuoka, and Saga prefectures (Figure 3). However, the prefectures classified with natural forest, secondary forest and grassland, secondary forest and plantation, and urban areas did not change. There are thought to be two factors affecting those changes; agriculture and forestry systems. Firstly, traditional agricultural activities were changed to mechanization, increasing product efficiency and increasing cultivated area per person. In addition, the market of agricultural products has come concentrated in metropolitan areas, and also other smaller markets have collapsed. Maintenance of agricultural land has proved difficult due to decreasing farm population and increasing old age of farmers. These phenomena may indicate intensive agriculture in some prefectures. Secondly, in the last 50 years in Japan, the Forestry Agency has encouraged plantations (Masaki et al., 2004). Therefore, monoculture conifer plantations have greatly increased because they grow faster and produce better timber than broadleaved trees (Masaki et al., 2004). It has been planted widely and intensively in

LANDSCAPE CHANGE ON NATIONAL GRID MAP

79

response to the demand for wood-based materials. Moreover, plantations have also occurred instead of abandoned rural forest and dead pine trees, due to pine wilt diseases, for economical purposes. From an ecological perspective, monoculture plantations are still to be evaluated for their ecological functions.

Figure 3. Classification of land use characteristics of Japan since the third survey.

Finally, landscape changes in Japan face decreasing landscape diversity with less naturalness, decreasing primary vegetation, and decrease in area of representative forest, especially, P. densiflora. The main causes for this may be urbanization and abandoned rural landscape. In addition, group 2, secondary forest and grassland group, will change due to decreasing P. densiflora and grassland areas.

REFERENCES Fujihara, M., Hada, Y. and Toyohara, G. (2002). Changes in the stand structure of a pine forest after rapid growth of Quercus serrata Thunb. Forest Ecology and Management, 170, 55-65. Haines-Young, R.H. (1992). The use of remotely sensed satellite imagery for landscape classification in Wales (U.K.). Landscape Ecology, 7, 253-274. Himiyama, Y. (1999). Historical information bases for land use planning in Japan. Land Use Policy, 16, 145-151. Hong, S.-K. (1998). Changes landscape patterns and vegetation process in the Far-Eastern cultural landscapes: human activity on pine-dominated secondary vegetation in Korea and Japan. Phytocoenologia, 28: 45-66. Hong, S.-K., Song, I.-J., Byun, B., Yoo, S. and Nakagoshi, N. (2005). Application of biotope mapping for spatial environmental planning and policy: case studies in urban ecosystems in Korea. Landscape and Ecological Engineering, 1, 101-112.

80

N. NAKAGOSHI AND J.-E. KIM

Jenerette, G.D. and Wu, J. (2001). Analysis and simulation of land-use change in the central ArizonaPhoenix region, USA. Landscape Ecology, 16, 611-626. Jobin, B., Beaulieu, J., Grenier, M., Bélanger, L., Maisonneuve, C., Bordage, D. and Filion, B. (2003). Landscape changes and ecological studies in agricultural regions, Québec, Canada. Landscape Ecology, 18, 575-590. Kamada, M, and Nakagoshi, N. (1996). Landscape structure and the disturbance regime at three rural regions in Hiroshima Prefecture, Japan. Landscape Ecology, 11, 15-25. Mamiya, Y. (1988). History of pine wilt disease in Japan. Journal of Nematology, 20, 219-226. Masaki, T., Ota, T., Sugita, H., Olhara, H., Otani, T., Nagaike, T. and Nakamura, S. (2004). Structure and dynamics of tree populations within unsuccessful conifer plantations near the Shirakami Mountains, a snowy region of Japan. Forest Ecology and Management, 194, 389-401. Morisita M. (1959). Measuring of interspecific association and similarity between communities. Mem. Fac. Sci., Kyushu Univ. Ser. E. (Biol.), 3, 65-80 Mou, P., Jones, R.H., Guo, D. and Lister, A. (2005). Regeneration strategies, disturbance and plant interactions as organizers of vegetation spatial patterns in a pine forest. Landscape Ecology, 20, 971987. Nakagoshi, N. (1995). Changing cultural landscapes in Japan. In B.von Droste, H. Plachter and M. Rösser, (Eds.), Cultural Landscapes of Universal Value (pp. 128-138). Gustav Fisher Verlag, Jena. Nakagoshi, N., Hikasa, M., Koarai, M., Goda, T. and Sakai, I. (1998). Grid map analysis and its application for detecting vegetation changes in Japan. Applied Vegetation Science, 1, 219-224. Nakagoshi, N. and Hong, S.-K. (2001). Vegetation and landscape ecology of East Asian ‘Satoyama’. Global Environmental Research, 5, 171-181. Nakagoshi, N. and Naito, K. (1997). Traditional land use and threatened plants in rural landscape in Japan. Journal of International Development and Cooperation, 3, 1-13. Nakagoshi, N. and Ohta, Y. (1992). Factors affecting the dynamics of vegetation in the landscapes of Shimokamagari Island, southwestern Japan. Landscape Ecology, 7, 111-119. Olsson, E.G.A., Austrheim, G. and Grenne, S.N. (2000). Landscape change patterns in mountains, land use and environmental diversity, Mid-Norway 1960-1993. Landscape Ecology, 15, 155-170. Sanderson, R.A. and Rushton, S.P. (1995). A preliminary method of prediction plant species distributions using the British National Vegetation Classification. Journal of Environmental Management, 43, 265-288. Takeuchi, K., Brown, R.D., Washitani, I., Tsunekawa, A. and Yokohari, M. (Eds.). (2003). Satoyama: The Traditional Rural Landscape of Japan. Springer-Verlag, Tokyo, Japan. Tasser, E. and Tappeiner, U. (2002). Impact of land use changes on mountain vegetation. Applied Vegetation Science, 5, 173-184. Wilson, W.L., Abernethy, V.J., Murphy, K.J., Adam, A., McCracken, D.I., Downie, I.S., Foster, G.N., Furness, R.W., Waterhouse, A. and Ribera, I. (2003). Prediction of plant diversity response to landuse change on Scottish agricultural land. Agriculture Ecosystems and Environment, 94, 249-263. Zerbe, S. (1998). Potential natural vegetation: Validity and applicability in landscape planning and nature conservation. Applied Vegetation Science, 1, 165-172.

CHAPTER 6

CHALLENGES FACED WHEN CREATING AN EVALUATION METHOD OF BIODIVERSITY ON AN ECOSYSTEM LEVEL

J. MORIMOTO Nihon University, Kanagawa, Japan

Abstract. In order to create an accurate evaluation method that properly reflects the value of biodiversity on an ecosystem level, the current and past studies of such methods must be analysed. Biodiversity evaluation models published in journals from 1995 to 2005 were studied and was concluded that in order to create an accurate model, the following four elements need to be included: the first, species from different guilds that maximize the phylogenetic diversity, as surrogates of the ecosystem should be included. Secondly, assess the extinction probability of the selected species by Population Viability Analysis. Thirdly, identify the potential habitat area of selected species by Potential Habitat Analysis. Fifthly, estimate the survival probability of the selected species in the potential habitat area.

1. INTRODUCTION There are three types of values in ecosystem goods and services, which are provided by ecosystem functions; i.e., regulation functions, habitat functions, production functions, and information functions (Figure 1). Biodiversity is one of the ecosystem services supported by ecosystem structures and processes that provide habitat for wild plant and animal species (DeGroot et al., 2002). Moreover, biodiversity is the basis for most ecosystem functions, which means, it contributes directly or indirectly to all ecosystem goods and services. As figure 1 shows, the value of biodiversity is a total value of ecological value, socio-cultural value, and economic value. Farber et al. (2002) and Wilson and Howarth (2002) discuss in detail each concept of these three values. Ecological sustainability, i.e., physical, chemical, and biological capacity of the environment, 81 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 81–93. © 2007 Springer.

82

J. MORIMOTO

controls humans’ socio-cultural and economic activity. Therefore, “ecological value” must be evaluated accurately and objectively because it should be the basis for the other values, i.e., “socio-cultural value” and “economic value”. Then, how has the ecological value of biodiversity, the pivot of ecosystem goods and services, in other words, been evaluated? Many evaluation models for biodiversity have been proposed, especially since CBD (The Convention on Biological Diversity) went into effect in 1993. Evaluation of biodiversity is carried out on all kinds of aspects such as assessing developmental impact on wildlife habitat, prioritizing reserve selection, and assessing mitigation effect. Corresponding to the request from these needs, many researchers are trying to evaluate regional biodiversity in unique ways suitable for their own purposes or local ecosystems. Ecosystem Structure & Process

Ecosystem Functions Regulation Atmospheric chemical composition, Weathering of rock, Vegetation root matrix….

Habitat Suitable living space for wild pants and animals

Production Convention of solar energy into edible plants and animals, Genetic material and evolution…

Information

Ecosystem Services Regulation

Ecological Values

Climate, Water, Soil, Pollination

Habitat Refugium function

Production Food, Row materials, Genetic resources

Information Recreation, Cultural & artistic information, Science and Education

Attractive landscape features, Spiritual and historic value, scientific and educational values

Sociocultural Values Economic Values

Figure 1. Framework for integrated assessment and evaluation of ecosystem functions, goods and services. (Adapted from DeGroot et al., 2002).

I examined these latest studies published in major scientific journals and delineated subjects to be included especially in an evaluation model for biodiversity at the ecosystem level. A method to measure biodiversity at the ecosystem level has not been developed enough compared to that at species level. I examined the precise direction for establishing an evaluation method of biodiversity at the ecosystem level by analysing the latest studies. 2. METHODS Twenty three of papers that include words of biodiversity, species or ecosystem, and evaluation in their abstracts were found in 25 major scientific journals including

BIODIVERSITY EVALUATION ON ECOSYSTEM LEVEL

83

Ecological Applications, Biological Conservation, Landscape and Urban Planning, and Ecology, published from 1995 to 2005. Biodiversity is a comprehensive concept that has compositional, structural, and functional concepts in each hierarchy of gene, species, ecosystem, and landscape (Noss, 1990). Biodiversity evaluation models at ecosystem levels were categorized into four groups based on the concept of biodiversity: (1) models that put values on composition, (2) models that put values on structure, (3) models that put values on composition and structure, and (4) models that put values on functions. 3. OVERVIEW OF CURRENT STUDIES ON EVALUATION OF BIODIVERSITY IN ECOSYSTEMS 3.1 Models that put values on composition Most models were categorized in this group, which puts values on the compositional aspect of biodiversity, i.e., number of species, species assemblage, and the number of ecotope types. They used definite values, indexes, or rankings for evaluation. Oliver et al. (1996) investigated the richness and turnover of the species of ants, beetles, and spiders, which were abundant and belonged to the species-rich taxa of the forest floor. They compared the turnover of this assemblage among four forest types and proposed cost-effective methods for evaluating epifaunal invertebrate biodiversity. Schwab et al. (2002) evaluated biodiversity of 18 meadows with different intensiveness of use by an overall index of biodiversity, i.e., the sum of the eight rankings. Rankings reflect three different types of criteria for evaluation of the sites. The first criterion was species number of angiosperms (a), spiders (s), and beetles (b). The second criterion was Simpson’s Diversity Index. The third criterion was conservation index, which used the angiosperm species characteristic to nutrientpoor grassland and special spiders in terms of rarity and specificity. R = Rsp + Rsd + Rci

(1)

Rsp = Rsp(a) + Rsp(s) + Rsp(b)

(2)

Rsd = Rsd(a) + Rsd(s) + Rsd(b)

(3)

Rci = Rci(a) + Rci(s)

(4)

84

J. MORIMOTO

Where R: Indicators of biodiversity, Rsp: Ranking of species, Rsd: Ranking of Simpson’s Index, Rci: Ranking of Conservation Index Lenders et al. (2001) evaluated floodplain ecosystems in the Netherlands using “target species” for Dutch nature conservation policy and “protected/special attention species” in a variety of national and international policy plans, laws, trends, and directives. “Target species” were required to meet the “itr-criteria” (criteria concerning the international importance of the Netherlands for the species involved (i-criterion), population development trend (t-criterion) and species rarity (r-criterion) in the Netherlands (Bal. et al., 1995) and the Red Data List criteria of the World Conservation Union (IUCN, 1993). “Protected species” approach yielded the additional species that were not particularly threatened or endangered in the Netherlands but equally important from a political and legal point of view as Red Data List species.

TBSindex = 100 ⋅ ATB PTB

(5)

ATB = ∑i =1 Si

(6)

PTB = ∑i =1 Si

(7)

n

N

where, TBS index: Taxonomic group biodiversity index, ABT: Actual taxonomic group biodiversity score, PTB: Potential taxonomic group biodiversity score, n: Real species number, N: Selected important species, Si: Score of species i. Evaluation by ranking is not applicable to other regions because it puts relative values on a target area. While, evaluation by absolute value and index have an advantage at the point of applicability to other regions and cases, so that those values can be compared between them. In turn, absolute value and index have the meaning for the first time when they are compared. In addition to this, selection of representative species and weighting on the specific species need scientific grounds because they directly relate to interpretations of biodiversity. 3.2 Models that put values on structure Geneletii (2003) presented an approach to contribute to Biodiversity Impact assessment of road project that focused on the direct loss of ecosystems. He assessed the impact of 20 km of road development on the surrounding forest environment by the rarity of ecosystems. Indicator to express ecosystem’s rarity was Potential Area

BIODIVERSITY EVALUATION ON ECOSYSTEM LEVEL

85

Remaining (PAR), whose value (V) was provided by dividing the target area in a potential vegetation map into the corresponding area in a present vegetation map. Then, ecosystem-loss impact score of alternative i (ELi) was calculated as follows,

ELi = ∑ Nj=1 ( AjVj )

(8)

where, Aj: predicted area loss for ecosystem type j, Vj: assessed value of ecosystem type j; N: number of ecosystem types The method of valuing area as presented above is also used by Morimoto et al. (2003). They valued the watershed biodiversity in Maryland, USA by Biodiversity Probability Index (BPI) using land cover map. BPI for each watershed represented the sum total HU (Habitat Unit) for forest species, grassland species, wetland species, edge species, and multihabitat species per unit area. Habitat Unit is estimated by the habitat value (V), land area within a watershed (A) and contagion index (Q).

BPI = ∑ HUj A

(9)

HUj = ∑ (Qi ⋅ Ai ⋅ Vij )

(10)

Qi = Ni, i Ni

(11)

where i : Land cover type (Water, Developed land, Edge of developed land, Forest, Edge of forest, Grasslands, Edge of grasslands, Wetlands, Edge of wetlands, Barren land, and Edge of barren land), j : Species category ( Forest, Grassland, Wetland, Edge, and Multihabitat species), HUj: Habitat unit of species category j , Qi : Contagion index of i, Ai : Land area of land cover type i, Vij: Habitat value of land cover type i for species category j , Ni,i : Number of cells of i adjacent to i, Ni : Total number of i. Geneletti (2004) proposed an approach to assess the fragmentation of natural ecosystems caused by linear infrastructures, one of the greatest threats to biodiversity world-wide. Three patch indicators, the core area, the average distance from the surrounding settlements and infrastructures (disturbance), and the average edge-to-edge distance from the surrounding ecosystem patches (isolation), were selected to predict the viability of each ecosystem patch as follows,

86

J. MORIMOTO

Vi = ∑ j =1 (Wj ⋅ Aj ) 3

(12)

where Vi : the viability value of patch i, Wj : the weight of the indicator j, Aj : the value score of the indicator j, calculated using the relevant value function. Venema et al. (2005) presented a deforestation simulation model based on the idea that deforestation and forest fragmentation are underlying drivers of global biodiversity losses. They found that MNN (mean nearest neighbour distance) and MPI (mean proximity index) were the best response signal from the landscape ecology metrics to the deforestation process using the principle component analysis.

MNN = 1 n∑ j =1 min Dij , i ≠ j

(13)

MPI (D ) = 1 n∑ j =1 ∑

(14)

n

n

i∈Mj

Aij D 2ij

where n : the total number of patches in the landscape and for each patch j, Aij : the area of the ith patch of the set Mj, Mj : the set of patches within the threshold distance, D of the jth patch (Dij ≤ D). If all patches have no neighbours within the threshold distance, D, then MPI. = 0. Then the genetic algorithm approach for forest structure optimization was illustrated with the several objective function formulations including MNN minimization, MNN constrained by MSI, MPI maximization with varying threshold distance, constrained edge habitat maximization, and PCA-based optimization. Putting values on structural aspect of biodiversity is relatively easy once the database that can be analysed on GIS is available. However, the real meaning of the evaluation on the structure is unknown until the relationship between the structural characteristics and the compositional or functional property of biodiversity is cleared, i.e., “what kind of animals are sensitive to the forest fragmentation?” or, “how the fitness changes in response to the habitat reduction?”. 3.3 Models that put values on composition and structure Hermy and Cornelis (2000) and Duelli (1997) advocated the model that evaluates the compositional and structural aspect of biodiversity. Hermy and Cornelis (2000) developed a method for the general monitoring of biodiversity in urban and suburban parks. They put values on the compositional aspect of biodiversity, i.e., Shannon index of vascular plants and species richness of animals, and structural aspect of biodiversity, i.e., Shannon-Wiener diversity index of planar, linear, and punctual elements in parks. However these evaluation values were not integrated.

BIODIVERSITY EVALUATION ON ECOSYSTEM LEVEL

87

Duelli (1997) evaluated biodiversity in agricultural areas by species diversity in arthropods and by landscape parameters such as habitat diversity and landscape heterogeneity. These compositional and structural aspects of biodiversity are not integrated either, which remains to be a future challenge. 3.4 Models that put values on functions Conroy and Barry (1996) used the source-sink dynamics model to evaluate biodiversity of a hypothetical landscape. They assumed a simple landscape (R) that had “forest (R (F))” and “unforested (R (O))” habitat and four species that had different habitat requirements.

R = R(F ) ∪ R(O )

(15)

A = A(F ) + A(O )

(16)

where, A: area of a landscape, A (F): area of forest, A (O): area of unforested, open spaces. They estimated total species richness based on the source-sink model.

S ( A, f ) = ∑s =1 Sp (S = s ) S

(17)

where, p(S=s): probability of occurrence of species s, f: proportion of forested habitat. Here each species has a different habitat requirement. For example, the forest species have the following annual rates of population change (λ):

λi (F ) = λi (1) φ 1

(18)

λi (O ) = λi (2);0 ≤ λi (O) π 1

(19)

where, λ(F): annual rates of population change in forested habitat,λ (O): annual rates of population change in unforested habitat,λ (1): annual rates of population change in source habitat,λ (2): annual rates of population change in sink habitat. Equally, species k reside in open space has its annual rates of population change (λk).

88

J. MORIMOTO

λi (F ) = λi (2);0 ≤ λi (O) π 1

(20)

λi (O ) = λi (1) φ 1

(21)

Nˆ i (2) = Nˆ i (1) ⋅ {λi (1) − 1} {1 − λi (2 )}

(22)

Based on the equilibrium theory of source-sink model about abundance for species i in its source habitat ( Nˆ i(1) , Nˆ i (2) ), the total number of species appears in a landscape ( Sˆ ) is estimated. None of evaluation models focused on the functional aspect of biodiversity in the real landscape was found this time. It suggests the difficulty of quantifying the interactions between species and ecosystem processes. 4. SUBJECTS TO BE INCLUDED IN AN EVALUATION MODEL 4.1 Identification of representative species group of ecosystems Some of the evaluation models aiming at compositional aspects of biodiversity identified representative species of ecosystems in some way. Attributions of all creatures cannot be investigated due to cost and time restrictions.

Figure 2. Hypothetical species distributions (circles) in a simple environmental space. (Adapted from Faith et al., 2002).

BIODIVERSITY EVALUATION ON ECOSYSTEM LEVEL

89

Therefore, the method used to select the representative species or species group of ecosystems is important. Faith and Walker (2002) insist that effective indicator species groups tend to be distributed over a range of habitats or environments. Figure 2 shows hypothetical species distributions in a simple environmental space. Species group A consists of five species, and species group B consists of four species. Each group has a different habitat requirement. When we select four areas of high biodiversity, repeated samplings from sections a, b, d, which individually are the most species-rich areas for one species group or for both groups together, will not produce a set of areas that are species rich. However, selecting just four areas, one each from sections a, b, c, and d, would sample all the species from the two groups. This suggests that the species group whose members span a wide range of habitats or environments should be an effective indicator of biodiversity. It can be concluded that different guilds, i.e., species or populations at the same level of a food pyramid and using the same resources, are appropriate as indicators of biodiversity. In contrast, Williams and Hero (2001) and Humphries et al. (1995) insist that subsets of organisms that are most evenly spaced or distributed over the topology of the phylogenetic tree, not the geographical environment, should be the effective indicators of biodiversity. This idea comes from the theory that biodiversity can be evaluated by “option value”: a safety net of biological diversity for responding to unpredictable events or needs, and that option value can be measured by the distance between species in cladistic relationships.

Figure 3. A hypothetical cladogram of four taxa. (Adapted from Faith, 1991). Phylogenetic diversity is measured by summing branch lengths* between species i and j. *PDi,j = 0.5 (Dx,i + Dx,j - Di,j), where, Da,b: Minimum branch length between species a and b.

90

J. MORIMOTO

Figure 3 shows a hypothetical cladogram of four taxa. Length between them is regarded as magnitude of character differences between them. Selecting species whose phylogenetic diversity is high, that is, those distances on the phylogenetic tree that are the longest, would maximize the “option value”. Choosing species using phylogenetic trees is characterized by its longer geohistorical span of biodiversity than the selection of species using guilds. It is still difficult to carry out this identification method in practice because species that have been delineated in phylogenetic trees are limited. However, this method is advantageous to the selection of representative species groups because it considers the future of ecosystems. To conserve today’s biodiversity in a shorter time span, and to keep its “option value” maximum, it is necessary to choose species groups from both points of view as evaluation targets of biodiversity. That means, species from different guilds that maximize the phylogenetic diversity, as surrogates of the ecosystem should be included. 4.2 Species’ suitability to environment and response regime to disturbance It may be a serious problem to regard species that use the same resources as completely homogenous. Lindenmayer and Lacy (1995) analysed population vulnerability of two guilds of the mountain brushtail possum, and pointed out that different species, even those within the same guild, may vary in vulnerability to disturbance and environmental perturbation. Alvarez-Buylla et al. (1996) performed population vulnerability analysis (PVA) to predict extinction times of four tropical rain forest tree species. They found that critical phases in a life cycle are different for long- and short-lived species. Moreover, long- and short-lived species had contrasting responses to different temporal and spatial regimes of perturbation. They suggest that it is important to reveal the composition of metapopulations to perform realistic vulnerability analysis. PVA based on the actually measured demographic and environment stochastic enable one to predict the extinction risk of selected species groups. Functional aspects of biodiversity such as habitat suitability and susceptibility to environmental fluctuations are the essence of biodiversity itself. It is desirable to perform PVA for quantitative evaluation of biodiversity. 4.3 Environmental factors related to species distribution Integration of compositional aspects of biodiversity (species richness and species composition) and structural aspects of biodiversity (shape and pattern of habitat and ecotope) enables a potential habitat analysis (PHA).

BIODIVERSITY EVALUATION ON ECOSYSTEM LEVEL

91

Pino et al. (2000) analysed the relationships of spatial distribution of bird species richness with landscape variables such as diversity index of landscape, relative shrubland, cropland, or urban cover, and geo-coordinate. Cropland species were the most dependent on the abundance of relative area of cropland cover and on landscape diversity, whereas forest species exhibited weak correlation with landscape variables. Fairbanks et al. (2001) investigated the relationships between assemblages of bird species in South Africa and 20 environmental variables including climate, topology, and vegetation, landscape type defined by topology and climate, and land-use. He found that bird species assemblages were primarily related to climate variables such as growing season temperature and seasonality of precipitation, and water balance. Environmental factors related to distributions of species and species groups differ among them. Quantifying the relationships of the structural aspect of biodiversity, i.e., pattern of environmental factors, and the compositional aspect of biodiversity, i.e., distribution of organisms will enable estimation of potential habitat. Pattern-based evaluation procedure has a defect in its unrealistic assumption that species suitable to some environmental attributes distribute evenly in the environment with those attributes. However, advantageously, it permits us to predict potential biodiversity even in difficult-to-reach areas. It also enables the assessment of environmental changes effects on species and species groups. 4.4 Integration of functional and compositional aspects of biodiversity into the structural aspect of biodiversity Integration of the functional aspect and structural aspect of biodiversity is behind that of the compositional aspect and structural aspect of biodiversity (for example, Williams and Hero, 2001; Pino et al., 2000; Fairbanks et al., 2001). It will be possible to predict quantitatively the population vulnerability in a potential habitat once the functional aspect of biodiversity, i.e., species’ habitat suitability and susceptibility to the environment changes estimated by PVA, is related to the structural aspect of biodiversity such as pattern of habitat and ecotope. Specifically, it is desirable to investigate the demography of local populations as long as possible in the various habitats and ecotopes with many different attributes, i.e., type, shape, and pattern of environmental factors, in the potential habitat areas predicted by PHA. The investigation period is to be long enough to include yearly fluctuation of species behaviour. Quantifying the relationships between environmental attributes of habitat and ecotope, and parameters that define the population dynamics, will enable one to predict population vulnerability in the potential habitat area. Evaluation of the present and future status of biodiversity would be possible by integration of the functional and compositional aspects into the structural aspect of biodiversity.

92

J. MORIMOTO

5. SUMMARY In order to create an accurate evaluation method that properly reflects the value of biodiversity on an ecosystem level, the current and past studies of such methods must be analyzed. Biodiversity evaluation models published in journals from 1995 to 2005 were studied and was concluded that in order to create an accurate model, the following four elements need to be included: i. Species from different guilds that maximize the phylogenetic diversity, as surrogates of the ecosystem should be included. In order to maintain the current level of biodiversity in the short term, species belonging to different guilds reflect as the best indicator of biodiversity. In addition, species selected widely from the phylogenetic tree can also act as a good indicator of biodiversity maintenance in the long term. ii. Assess the extinction probability of the selected species can by Population Viability Analysis. The Population Viability Analysis or PVA, a type of demographic study, can estimate the vulnerability of a representative species population, quantitatively. iii. Identify the potential habitat area of selected species by Potential Habitat Analysis. Potential Habitat Analysis or PHA can analyze the correlation between the species distribution and environmental pattern quantitatively. iv. Estimate the survival probability of the selected species in the potential habitat area. Combining three aspects of biodiversity, composition, structure, and function of biodiversity will estimate quantitatively the biodiversity on an ecosystem level. In other words, relating parameters derived from PVA (function) of indicator species as described in i (composition) with the structure and pattern of ecotope delineated in PHA will evaluate properly the biodiversity on an ecosystem level. REFERENCES Alvarez-Buylla, E.R., Garcia-Barrions, R., Lara-Moreno, C. and Martinez-Ramos, M. (1996). Demographic and genetic models in conservation biology: Applications and perspectives for tropical rain forest tree species. Annual Review of Ecology and Systematic, 27, 387-421. Bal, D., Beuje, H.M., Hoogeveen, Y.R., Jansen, S.R.J. and Van der Reest, P.J. (1995). Handbook Natuurdoeltypen in Nederland, IKC-Nuurbeheer, Wageningen. Conroy, M.J. and Barry, R.N. (1996). Mapping of species richness for conservation of biological diversity: Conceptual and methodological issues. Ecological Applications, 6(3), 763-773. DeGroot, R., Wilson, M.A. and Boumans, R.M.J. (2002). A typology for the classification, description and valuation of ecosystem functions, goods and services. Ecological Economics, 41, 393-408. Duelli, P. (1997) Biodiversity evaluation in agricultural landscapes: An approach at two different scales, Agriculture, Ecosystem and Environment, 62(2-3), 81-91. Fairbanks, D.H.K., Reyers, B. and van Jaarsveld, A.S. (2001). Species and environment representation: selecting reserves for the retention of avian diversity in KwaZulu-Natal, South Africa. Biological Conservation, 98(3), 365-379. Faith, D.P. and Walker, P.A. (1996). How do indicator groups provide information about the relative biodiversity of different sets of areas? : On hotspots, complementarity, and pattern-based approaches, Biodiversity Letters, 3(1), 18-25.

BIODIVERSITY EVALUATION ON ECOSYSTEM LEVEL

93

Faith, P.D. (1992). Conservation evaluation and phylogenetic diversity. Biological Conservation, 61, 1-10. Farber, S.C., Costanza, R. and Wilson, M.A. (2002). Economic and ecological concepts for valuing ecosystem services. Ecological Economics, 41, 375-392. Geneletti, D. (2003). Biodiversity Impact Assessment of roads: an approach based on ecosystem rarity. Environmental Impact Assessment Review, 23, 343–365 Geneletti, D. (2004). Using spatial indicators and value functions to assess ecosystem fragmentation caused by linear infrastructures. International Journal of Applied Earth Observation and Geoinformation, 5, 1–15. Hermy, M. and Cornelis, J. (2000). Towards a monitoring method and a number of multifaced and hierarchical biodiversity indicators for urban and suburban parks. Landscape and Urban Planning, 49, 149-162. Humphries, C.J., Williams, P.H. and Vane-Wright, R.I. (1995). Measuring biodiversity value for conservation. Annual Review of Ecology and Systematics, 26, 93-111. IUCN (1993). IUCN Red List of Threatened Animals. International Union for Conservation of Nature and Natural Resources, Gland, Cambridge. Lenders, H.J.R., Leuven, R.S.E.W., Nienhuis, P.H., de Nooij, R.J.W. and Van Rooij, S.A.M. (2001). BIO-SAFE: a method for evaluation of biodiversity values on a basis of political and legal criteria. Landscape and Urban Planning, 55, 121-137. Lindenmayer, D.B. and Lacy, R.C. (1995). Metapopulation viability of arboreal marsupials in fragmented old-growth forests: Comparison among species. Ecological Applications, 5(1), 183-199. Morimoto, J., Wilson, M. A., Voinov, H. and Costanza, R. (2003) Estimating Watershed Biodiversity: An Empirical Study of the Chesapeake Bay in Maryland, USA. Journal of Geographic Information and Decision Analysis, 7(2), 150-162. Noss, R.F. (1990). Indicators for monitoring biodiversity: A hierarchical approach. Conservation Biology, 4, 355-364. Oliver, I. and Beattie, A.J. (1996). Designing a cost-effective invertebrate survey: A test of methods for rapid assessment of biodiversity. Ecological Applications, 6(2), 594-607. Pino, J., Roda, F., Ribas, J. and Pons, X. (2000). Landscape structure and bird species richness: implications for conservation in rural areas between natural parks. Landscape and Urban Planning, 49, 35-48. Schwab, A., Dubois, D., Fried, P.M. and Edwards, P.J. (2002). Estimating the biodiversity of hay meadows in north-eastern Switzerland on the basis of vegetation structure. Agriculture, Ecosystem and Environment, 915, -13. Venema, H.D., Calamai, P.H. and Fieguth, P. (2005). Forest structure optimization using evolutionary programming and landscape ecology metric. European Journal of Operational Research, 164, 423– 439. Williams, P.H., Vane-Wright, R.I. and Humphries, C.J. (1991). Measuring biodiversity: taxonomic relations for conservation priorities. Aust. Syst. Bot., 4, 665-669. Williams, S.E. and Hero, J. (2001). Multiple determinants of Australian tropical frog biodiversity. Biological Conservation, 98(1), 1-10. Wilson, M.A. and Howarth, R.B. (2002). Discourse-based valuation of ecosystem services: establishing fair outcomes through group deliberation. Ecological Economics, 41, 431-443.

CHAPTER 7

IDENTIFICATION OF THE POTENTIAL HABITAT FOR GIANT PANDA IN THE WOLONG NATURE RESERVE BY USING LANDSCAPE ECOLOGY METHODOLOGY

L.D. CHEN1, X.H. LIU2, B.J. FU1, Y.H. LÜ1, J. QIU1 1

Key Lab of Systems Ecology, Research Center For Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, P. R. China;2Department of Environmental Sciences and Engineering, Tsinghua University, Beijing 100084, P.R. China

Abstract. Establishing nature reserves is one of the most effective means for protecting endangered species and biodiversity. However, establishing nature reserves may stop gene-exchange of the target population from natural populations given that the previous human activities within the reserves were moved to the surrounding areas of the nature reserves. This may break the connection between nature reserve and natural habitats. In this study, a habitat suitability evaluation of Wolong Nature Reserve for giant panda (Ailuropoda melanoleuca) conservation indicates that among the total area of 202300 ha, only about 2144 ha, or 1.06%, is highly suitable, and as much as 123280 ha (more than 60% of the total area) was unsuitable for giant panda. Moreover, the highly suitable, suitable areas, and moderately suitable areas were spatially fragmented. Based on the evaluation, landscape design for giant panda conservation was performed. It was suggested that both habitat quality and patch size be considered to meet the requirements for sustaining populations when core patches in a nature reserve were designed. Buffer with right width should include all core patches to allow giant panda to move freely. On the subject of corridor design, two cases were to be identified, first, the existing corridor, which was a narrow passage between some patches, had to be protected carefully. Second, those areas, which may become the potential habitats for giant pandas after rational vegetation rehabilitation, should be identified by using GAP approaches. This study indicates that some key areas which may enable the core patches larger to accommodate more giant pandas by vegetation rehabilitation, were much more important than the other places. These areas should receive higher attentions when establishing nature reserve.

95 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 95–112. © 2007 Springer.

96

L.D. CHEN ET AL.

1. INTRODUCTION Alteration and fragmentation of pristine habitats as result of human activity expansion are two of the greatest threats to the maintenance of biodiversity (CBCAS, 1994; Jactel et al., 2002; Fahrig, 2003). The giant panda (Ailuropoda melanoleuca) was not only a rare and ancient animal in China, but also one of the most endangered species in the world. It was considered a precious and mysterious animal from ancient times and its survival was of high scientific value. However, because of economic development and human activity expansion, natural habitats for giant pandas were declining rapidly (Li, 1997; Wang and Ma, 1993; Schaller, 1993; Liu et al., 1999; Ouyang et al., 2001). In the mid-Pleistocene, giant pandas inhabited many regions in China and southeastern Asia as a large and well-connected region (Wang, 1974; Wang and Ma, 1993; Li, 1997). However, with natural habitat loss, giant pandas are currently confined to six separate regions on eastern edge of the Tibetan Plateau (Li, 1997; Taylor and Qin, 1998; Ouyang et al., 2001). Even after nature reserves were established since 1960s, giant panda numbers still decreased and their habitats declined in the past two decades (Li, 1997). The main reasons for the loss of natural habitats and population were (Pan, 1988; Hu, 1996; Qin and Taylor, 1993; Ouyang et al., 2001; Xu et al., 2005): (1) fragmentation and reduction of habitats with extensive deforestation due to road construction and land reclamation; (2) direct disturbance and destruction due to human activity expansion; (3) illegal hunting and (4) food shortage due to flowering, die-off of bamboo during the 1970s and early 1980s. At present, about 27 nature reserves were established to protect giant pandas in China, and some improvements were made, but far below the expectations. For instance, Wanglang Nature Reserve, one of the major reserves for giant panda conservation, was established in 1963 with 196 giant pandas at the beginning (Giant Panda Survey Team of Wanglang Nature Reserve, 1974). However, it decreased to 66 giant pandas in 1975, and further to 27 giant pandas in the mid-1980s. A recent survey shown that approximately 19 (±4) giant pandas were found in Wanglang Nature Reserve in the late 1980s (Feng, 1991). A similar situation appears in Wolong and the other nature reserves (Yang et al., 1994). Approximately 140 giant pandas lived in Wolong Nature Reserve in the 1960s, but only 90 giant pandas remain in the 1990s (Liu, 1996). The rapid loss of giant pandas was due to food shortage with bamboo flowering during the late 1970s and early 1980s. A large nature reserve was essential for protecting the target population. However, habitat suitability and spatial pattern of the nature reserve (or natural habitats) was more important (Li et al., 1999a; Liu et al., 2001; Chave et al., 2002; Steiner and Köhler, 2003). Wanglang Nature Reserve was mainly covered by bare-rock (50%), and forest, grass, shrub covering about 40%, 6% and 4% respectively. The suitable area for giant panda living in Wanglang Nature Reserve covers about 15 %, and in Wolong Nature Reserve was about 25%, and in Jiuzhaigou Nature Reserve was about 15% (Hu et al., 1985). This implied that a large nature reserve with less suitable habitat didn’t help the conservation of the target species. In general, establishing nature

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

97

reserves may reduce human impacts on target populations. However, a potential artificial barrier might be made against gene-exchange in different metapopulations. This was for the reason that former human activities were moved to the surrounding areas of the nature reserve. The grown human activities in the neighboring areas, such as timber cutting, Chinese herb collection and hunting, might gave rise to landscape degradation, or corridor breaking. It would be unfavorable for gene-exchange and gene-diversity conservation. When a new nature reserve was established, besides the size, more attention should be paid to habitat quality, spatial arrangement (patches and corridors) and landscape connectivity between different patches or nature reserves (Chen et al., 2000; Steiner and Köhler, 2003; Fahrig, 2003). The importance of corridors to metapopulations in heterogeneous landscapes is well documented (Bennett, 1990; Henein and Merriam, 1990; Merriam and Lanoue, 1990; Anderson and Danielson, 1997). Empirical data show that patches connected by corridors have lower extinction rates, higher colonization rates, and greater population abundances (Bennett, 1990; La Polla and Barrett, 1993; Bennett et al., 1994). Danielson and Hubbard (2000) indicate that the presence of corridors decreased the probability that P. polionotus (particularly females) would disperse or disappear from a patch. In some circumstances, poor spatial combinations of natural landscape factors and nature reserves resulted in a poor function on target species conservation (Chen et al., 2000; Dauber et al., 2003; Lü, 2003). If a nature reserve were established separately, it might become an isolated island that is enclosed by a heterogeneous environment dominated by humans. The wildlife in the nature reserve would be separated from the other metapopulations. In order to protect endangered species and bio-diversity, much more attention should be paid to nature reserve size and landscape pattern (Steiner and Köhler, 2003). Landscape design and planning in a nature reserve based on scientific habitat suitability evaluation was important for target species conservation (Chen et al., 2000; Chave et al., 2002; Steiner and Köhler, 2003). The objectives of this paper are: the first, to make a habitat suitability evaluation of Wolong nature reserve with reference to the giant panda; the second, to make a landscape design for improving landscape connectivity for giant panda conservation; and thirdly, to find the potential areas that may become suitable for the giant panda after vegetation restoration.

2．METHODOLOGY

2.1 The study area The study area is situated in Wenchuan county in western Sichuan Province, China, ranging from 102o52'-103o24'E to 30o45'-31o20'N (Figure 1). As a national flag nature reserve for giant panda conservation, it was established in 1963, covered 200 km2 at the beginning and was enlarged to more than 2,000 km2 in 1975.

98

L.D. CHEN ET AL.

Figure 1. The location of Wolong Nature Reserve in China.

Apart from the giant panda, the other endangered species, for example, Rhinopethecus roxllanae, Neofelis nebulosa, Panthera pardus, Panthera unci, Cervus albirostris, Budorcas taxicolor, Aguila chrysaetos, Gypaetus barbatus, Macaca mulatta, Macaca thibetana, Selenarcios thibetanus, Ursus arctos pruinosus, Ailurus fulgens, Felis temmincki, Viverricula indina, etc., were also accommodated (WNR and SNU, 1992). This area is located in the transitional zone between the lower Sichuan basin and the Tibetan Plateau. Elevation ranges from 1150m to 6250m above sea level and the topography inclines towards southeast. A steep slope in the northwestern side and a moderately steep slope in the southeastern side of the hills were visible. Within the study area, there are five streams, namely, Pitiao stream, Zheng stream, Gengda stream, Zhong stream and Xigou stream. Annual mean temperature was 8.7oC and mean rainfall was 959 mm. Vegetation changed with elevation, was evergreen broadleaf forest (1150-1600m), evergreen and deciduous broadleaf mixed forest (1600-2600m), conifer and broadleaf mixed forest (2000-2600m), conifer forest (2600-3600m) frigid bush and meadow (3600-4400m), alpine sparse vegetation (4400-5000m) and naval

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

99

(5000-6250m) (Hu et al., 1985; WNR and SNU, 1992; Qin and Taylor, 1993; Ouyang et al., 2001). There are six villages and more than 4300 inhabitants in Wolong Nature Reserve. About 3.84km2 of the land is cultivated for agricultural use, which was mostly located in the valley. Timber cutting, Chinese herb collection and local mining remained, but limited. 2.2 Natural landscape factors affecting giant panda Giant panda prefers to the temperate climate with dense woodland and abundant bamboo (Pan, 1988; WNR and SNU, 1992; Schaller, 1993). In Wolong Nature Reserve, the favorite zone for giant panda ranges from 1600m to 3600m above sea level. Giant pandas normally lived in the higher places in summer and shifted to the lower places in winter (Hu, 1981). Apart from land cover/bamboo, elevation and slope degree were also important landscape factors that affected giant pandas. Human activities, natural enemies, and slope aspect have some impacts on giant pandas (Hu et al., 1985; Liu et al., 1999; Ouyang et al., 2001; L ü et al., 2003). In this study, habitat suitability evaluation, and landscape design were conducted from the natural perspectives by considering the three main natural landscape factors, i.e., land cover (vegetation), elevation and slope degree. 2.2.1 Land cover/bamboo distribution In Wolong Nature Reserve, land cover that was suitable for giant pandas included evergreen broadleaf forest, evergreen and deciduous broadleaf mixed forest, conifer and broadleaf mixed forest and conifer forest. Bamboos available for giant pandas were Gelidocalamus fangiana, Fargesia robusta, Yushania chungii, Fargesia nitida, Phyllostachys nidularia, Fargesia angustissima, etc. However, giant pandas prefer to eat Gelidocalamus fangiana and Fargesia robusta, followed by Yushania chungii; Fargesis nitida, Phyllostachys nidularia and lastly Fargesia angustissima (Ouyang et al., 2001; Li, 1997). Different land cover types combined with the above bamboos were ranked as four grades for giant panda when making habitat suitability evaluation. 2.2.2 Elevation Field observations indicated that the probability of giant panda’s occurrence was dependent on elevation. The range, 2200-2800m above sea level, was the most favorite zone for giant pandas (Hu et al., 1985), the highly suitable zone for the giant panda. This elevation zone was given a high weight when making habitat suitability evaluation. The range of elevation without giant panda’s occurrence suggested unsuitable and thus a low weight was assigned. Four grades may be ranked as shown in Table 1.

100

L.D. CHEN ET AL. Table 1. Weights for the three landscape factors.

Grade I

Food source Gelidocalamus fangiana; Argesia robusta

Elevation 2200-2800m

Slope gradient Less than 20o

Weight (u) 1.0

II

1500-2200m 2800-3500m

20～30o

0.67

III

Yushania chungii; Argesia nitida; Phyllostachys nidularia Fargesia angustissima

3500-4000m 1150-1500m

30～40o

0.33

IV

No bamboo area

>4000m

Above 40o

0.00

2.2.3 Slope degree Field survey indicates that 65% of giant pandas were found in the areas with slopes less than 20o, 25% within 20-30o, and only 12% in the areas with slope above 30o in Wolong Nature Reserve (Hu et al., 1985; Ouyang et al., 2001). Therefore, the area with slope less than 20o may be considered the most suitable for giant pandas, then 20-30o and 30-40o respectively. When the slope reaches 40o or above, it is no longer suitable for giant pandas (Table 1). 2.3 Habitat suitability evaluation The three most important natural landscape factors were important for giant pandas. However, the suitability was also dependent on their spatial pattern. If the three factors were all at their best conditions, the area would be the most suitable. If one of the three factors were at their worst conditions, this area would be unsuitable. Other combinations were in-between the most and lest suitable. This model that describes the habitat suitability index (HSI) can be defined as:

HSI (i.m.n ) = Ui × Um × Un

(1)

Ui = 1 − (l − 1) / 3

(2)

Um = 1 − ( m − 1) / 3

(3)

Un = 1 − ( n − 1) / 3

(4)

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

101

where, HSI was the habitat suitability index for the giant panda, l, m and n were the grade levels for land cover, elevation and slope degree, respectively, Ul, Um and Un were the weights of the three landscape factors in accordance to the grade (Table 1), a, b and c were the total number of grades (l ≤ a, m ≤ b and n ≤ c). From equations (2), (3) and (4), we can found that the weight of suitability (U) would decrease with increasing grade rank. From equation (1), if any weight of suitability of the three landscape factors were 0, the habitat suitability index would be 0. Only were all the value of the weights 1, the HSI would be 1. In other cases, HSI varies between 0 and 1.

Figure 2. Habitat suitability evaluation for giant panda conservation in Wolong Nature Reserve a: bamboo weight map; b: elevation weight map; c: slope degree weight maps; d: habitat suitability evaluation map; Mod: moderately, Marg:marginally.

A Geographical Information System was employed for spatial modeling. Data sources include a Digital Elevation Model (DEM) created from a topographic map of 1:100,000, by which elevation classification and slope degree maps were derived based on Table 1. The updated land cover classification map was derived by

102

L.D. CHEN ET AL.

combining a vegetation map at a scale of 1:500,000 made in 1980s (MBWNR, 1987) and with TM image interpretation. Weighted maps (Figure 2a-c) for each landscape factor were generated by attributing weight to the theme maps. The habitat suitability map was created by spatial modeling of the three component maps. 2.4 Landscape design for wildlife conservation Landscape design was based on habitat suitability evaluation. Four stages were followed: habitat suitability evaluation, core patch design, buffer design and key area design (Figure 3). Analysis of biological

Evaluation of current

nature of target species

Nature Reserve

Determination of landscape

Identification of requirements

parameter to affect target species

of target species on landscape factor

Determination of criteria of

Analysis and mapping of

GIS

Landscape parameter

Landscape factors for evaluation

Habitat suitability evaluation Determination of

Identification & delineation on core patches

GIS

Determination of

Identification and delineation of buffer areas

Identification and delineation of corridors

requirements of core patches

GIS

requirements of core patches

Determination of

GIS

requirements of core patches

Figure 3. Flowchart for landscape design and planning.

2.4.1 Habitat suitability evaluation Landscape pattern design should be in accordance with the demand of the target species. This means both the individual landscape factor and their spatial

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

103

combination should be favor the target population. In first, the biological nature of the target species and their affecting factors had to be analyzed. Then, the requirements of target species on the environment, and which parameters would be used for habitat suitability evaluation, should be decided. After that, the weighted parameter maps of each landscape factor can be derived based on GIS. By combining all the weighted parameter maps with GIS spatial modeling, a habitat suitability evaluation map would be generated. 2.4.2 Core patch design

Although much work was undertaken to select the location of nature reserves, few methods were available on designing the interior structure of a reserve. Nature reserve was normally composed of core patches, buffer zone, and research area (Li et al., 1999b). However, how to determine the above units was still lacking. In this study, the core patch design was based on the habitat suitability evaluation. In the first, all potential polygons, which may be used as core patches, were delimited out and then the area of each polygon was calculated using GIS. After considering the home range and minimum viable population of giant pandas, only those patches, which were large enough to accommodate a certain number of giant pandas, were ascribed to the core patches. 2.4.3 Buffer design

Core patches and their size was significant for target species conservation, however, a large core patch in most cases was not enough to protect the target species. Adverse impacts from outside may produce a strong effect. Buffer zones around reserves were prescribed as a palliative measure (Nepal and Weber, 1994). In this study, the buffer in Wolong nature reserve was designed based on GIS-distance analysis. The width of buffer and the spatial arrangement of core patches were required to concern. As for Wolong nature reserve, we suggest that the buffer enclose all the core patches, and also, the width of the buffer should be no less than a certain size (here 3 km is suggested). 2.4.4 Corridor (or key areas for vegetation rehabilitation) design

GAP analysis was employed in this study to identify the potential corridor (or potential key area for vegetation restoration) for improving landscape connectivity. GAP analysis was a powerful tool for identifying sites that ought to be protected but currently fall outside existing conservation networks (Mckendry and Machlis, 1991; Caicco et al. 1995). GAP analysis appears to offer the most practical guidance for reserve selection. Being able to identify gaps in an existing network is a simple and appealing concept that could easily be adopted by managers. In this study, the current core patches identified at the beginning, were compared with the assumed future potential

104

L.D. CHEN ET AL.

core patches, which were derived on the assumption that all potential vegetation rehabilitation were achieved. Any differences would be promising areas for the corridor (key areas for vegetation rehabilitation) design.

3. RESULTS AND DISCUSSION 3.1 Habitat suitability evaluation

Based on habitat suitability evaluation, we found that Wolong nature reserve was in poor situation for giant panda conservation. In this study, five suitability classes (Table 2 and Figure 2d) were grouped: (1) S1－highly suitable: the suitability index was 1, which means that all the three landscape factors were in their best situation. (2) S2－suitable: the habitat suitability index was more than or equal to 0.45, and smaller than 1. In this case, at least one of the three landscape factors was at their best situation. (3) S3 －moderately suitable: the habitat suitability index was less than 0.45 and bigger than or equal to 0.30. (4) S4－marginally suitable: the habitat suitability index was smaller than 0.30, but greater than 0.00. (5) S5－unsuitable: it means that at least one of the three landscape factors is absolutely unsuitable for giant pandas. Table 2 indicated that about 21.44 km2, c.a. 1.06% of the total area is highly suitable for giant pandas in Wolong nature reserve. Suitable, moderately suitable and marginally suitable areas for giant pandas were 311.34km2 (15.39%), 119.15km2 (5.89%) and 338.04km2 (16.71%), respectively. More than 60% of the nature reserve (1232.8km2) was unsuitable. Table 2. Habitat suitability evaluation in Wolong Nature Reserve.

Suitability Class S1 S2 S3 S4

Value of suitability index 1 <1 and ≥ 0.45 <0.45 and ≥ 0.30 <0.3 and >0.0

S5

0.0

Meaning Highly suitable Suitable Moderately suitable Marginally suitable Unsuitable

Area (km2) 21.44 311.34 119.15

Percentage of total area (%) 1.06 15.39 5.89

338.04

16.71

1232.8

60. 06

3.2 Landscape fragmentation

The habitat fragmentation in Wolong nature reserve was clearly observed in figure 2d. Although the highly suitable, suitable and moderately suitable areas for giant pandas were about 45193ha, they were distributed in 348 patches. Previous researches showed that the home range of a giant panda is normally around

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

105

389~640ha (Yang et al., 1994). However, a recent survey by radio-tracking indicated that the home range of a giant panda was about 480 ha and around 30% of them were shared (Ouyang et al., 2001). If this research result were taken into account, the actual range for a giant panda would be 336 ha. The numbers of giant pandas accommodated within each patch were estimated (Figure 4a). In Wolong nature reserve, only one patch could accommodate more than 50 giant pandas, two patches for 10~20 giant pandas, one patch for 5~10 giant pandas, and five patches for 1~5 giant pandas at present. Most of the patches were too small for giant panda living (Table 3). This situation is unfavorable for giant panda conservation and may induce adverse effects on the exchange of gene-diversity between different populations. These results were based on natural landscape evaluation, however, if the human impacts on giant panda were considered (Liu et al., 1999, 2001; Ouyang et al., 2001), the landscape fragmentation may become much more remarkable. Establishing nature reserves and promoting the exchange of gene-diversity between different groups by constructing corridors between different habitats was a feasible approach to wildlife conservation. To establish nature reserves as many as possible was obviously helpful for giant panda conservation. However, finding the key areas to build nature reserves or corridors would be of practical importance (Yu 1996).

Figure 4. Potential Core Patches for the giant panda in Wolong Nature Reserve. a: indicates the size of the potential core patches and the number of giant pandas accommodated; b: indicates the potential core patches in the WNR.

106

L.D. CHEN ET AL. Table 3. Distribution of potential core patches in Wolong Nature Reserve.

Patch type based on its capacity for giant panda living Less than 1 giant panda 1~5 giant pandas

Number of patches 340 5

Area (ha) 10908

Percentage of total nature reserve (%) 5.39

Percentage of current suitable area (%) 24.14

4043

2.00

8.95

5~10 giant pandas

1

2287

1.13

5.06

10~20 giant pandas

2

7797

3.85

17.25

20～50 giant pandas More than 50 giant pandas Total

0

0

0

0

1

20158

9.96

42.95

348

45193

23.34

100.00

3.3 Core patch design

Due to landscape limitation, giant pandas were confined to certain areas and separated into different groups (Li, 1997; Pan, 1988; Schaller, 1993). Gene-diversity was decreasing with natural habitat isolation. Improving landscape connectivity between different habitats was important for giant pandas, such as constructing corridors, setting buffers and enlarging core patches (Li, 1997). A large patch was good for wildlife conservation. In general, more species were accommodated in a large patch compared with a small one. However, it was inadvisable to establish a large reserve area with poor quality. In Wolong nature reserve, the total area was more than 2000 km2, but highly suitable, suitable and moderately suitable areas for giant pandas were only about 45193ha, and in high fragmentation. When designing core patches, both size and habitat suitability should be considered. In this study, the following two requirements were used for designing core patches: (1) appropriate habitat suitability (highly suitable, suitable and moderately suitable), and (2) sufficient areas to meet population viability (here 5 giant pandas were assumed). 3.3.1 Design procedure

Based on the habitat suitability evaluation map, GIS (ILWIS) was employed to delineate all potential patches that were suitable for giant pandas and the area of each patch was calculated. If 336ha was taken as the minimum activity range for each giant panda (Ouyang et al., 2001), the area of each core patch should be no less than 1680ha. Then a weighted patch map can be obtained (Figure 4a) by replacing the potential patches with the number of giant panda accommodated. The core areas

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

107

in Wolong nature reserve can be designed as Figure 4b, where four large core patches were observed in the east part of Wolong nature reserve. 3.4 Buffer design

A buffer was normally designed around core patches with an appropriate width. Since there were only four core patches in Wolong nature reserve, and they were close to each other, it would be of no significance to have a buffer around each patch. In such case, core patches should be considered as a whole when designing the buffer so that all the pandas can move freely among the patches. The following two criterions were recommended: (1) all core patches should be encircled by the buffer, and (2) the outside edge of the buffer to each core patch should be no less than 3 km. Design procedure: A distance map to each core patch was derived by using GIS (Figure 5a), based on the result, a 2-km range thus met the first requirement. When the second requirement was taken into account, 3-km distance to each patch would be the choice in buffer design (Figure 5b).

Figure 5. Buffer design in Wolong Nature Reserve. A: indicates the distance to the core patches; B: shows the designed buffer area in the Wolong nature reserve.

108

L.D. CHEN ET AL.

3.5 Potential areas for landscape connectivity improvement

Establishing suitable corridors between different habitats could promote geneexchange, which will be beneficial to the whole population conservation (Bennett, 1990). However, where and how to construct corridors are often not clear. Surely, the more corridors were built, the better the situation, and the wider the corridors were built, the better the situation. However, this often gave rise to exorbitant costs and some practical difficulties. Although the above designed buffers had already included all the core patches, it didn’t mean that giant pandas could move freely between core patches because of some unsuitable landscape factors. Therefore, it is necessary to identify the corridors and try to protect them. In this study, two cases were recognized, currently existing corridors that need to be protected, and potential corridor (key areas) that need to be re-vegetated for improving landscape connectivity. Are there any corridors between different core patches? How to define them? In Figure 2b, within the core patches, some parts were connected with the core patches by narrow aisles. If these aisles were destroyed, the core area would be divided into many small patches and become extremely fragmented. These long-and-narrow passages could be thought as the existing corridors and had to be cautiously protected (Figure 6a).

Figure 6. Potential key areas for vegetation rehabilitation to improve landscape connectivity in Wolong Nature Reserve.

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

109

On the other hand, some potential areas also existed although they are currently unsuitable for giant pandas. We could change these areas into suitable habitats and improved the landscape connectivity by vegetation restoration. These potential areas should meet the following requirements: the first, Suitable topographic conditions, for example, slope gradient and elevation should be at lest marginally suitable or better (grade I, grade II and grade III). This was because topographic conditions were difficult to be modified. The second, poor land cover or low suitability (marginally suitable or unsuitable). Such areas could be converted into suitable areas by vegetation rehabilitation. The third, these areas could help expand core patches if vegetation were properly planted. Although all the poor vegetation areas could be turned into high suitability for giant pandas by human modification, some places were more vital than the other places because these areas could be combined with the current core patches to create a larger patch, by which more giant pandas could be accommodated. However, the other areas couldn’t since they were far from the current core patches. In the latter case, even if the vegetation were rehabilitated into a good situation, it was separated from the core patches and was not much use for enlarging giant panda’ habitat. In this study, GIS was used to identify those areas, which were currently unsuitable but may become a suitable habitat for giant pandas by vegetation rehabilitation (Figure 6a). We found that the former four core patches would become two large ones after vegetation rehabilitation. In addition, the size of the largest one would be more than 62590 ha which might accommodate more than 180 giant pandas (Table 4). Table 4. Distribution of potential core patches in WNR after appropriate vegetation rehabilitation.

Patch type based on its capacity for giant panda living Less than 1 giant panda 1~5 giant pandas

Number of patches 56

Area (ha) 440

Percentage of total nature reserve (%) 0.22

Percentage of potential suitable area (%) 0.66

0

0

0

0

5~10 giant pandas

0

0

0

0

10~20 giant pandas

1

3668

1.81

5.50

20～50 giant pandas More than 50 giant pandas Total

0

0

0

0

1

62590

30.94

93.84

58

66698

32.97

100.00

110

L.D. CHEN ET AL.

At the same time, one patch that might accommodate more than 10 giant pandas would be appeared in the northwest part of Wolong nature reserve. Therefore, those areas, which are outside the former core patches and now fall in the enlarged core patches, would be vital for giant panda conservation. They would be the key areas for future vegetation restoration. However, those areas that were separated from the enlarged core patches were less important (Figure 6b).

4. CONCLUSION While establishing nature reserves as many as possible, it was significant to improve the landscape connectivity between different nature reserves or core patches (Baudry et al., 2003; Bayliss et al., 2005). Scientific landscape pattern design and planning based on habitat suitability evaluation was crucial for target species conservation, for example, locating core patches, buffer zones, existing corridors and potential areas for vegetation rehabilitation. Sometimes, it was impossible to conduct vegetation rehabilitation because of natural limitation. In the other cases, the areas suitable for vegetation rehabilitation were divided into two parts. The first one was those areas, which would contribute to expanding the current habitat after they were properly rehabilitated. These areas should be rehabilitated in the first place. The second one was those areas, which are suitable for vegetation rehabilitation, but they were no much use for target species conservation since they couldn’t contribute to expanding the core patches. When nature reserves were established, the spatial arrangement of nature reserves within a region and gene-exchange between different nature reserves had to be considered for improving landscape connectivity. Since the giant panda was a precious endangered species in the world, actions from governments, scientists, international organizations, and individuals were required. GIS was proved a powerful tool in habitat suitability evaluation, landscape design and planning, particularly for identifying the core patches, buffer zones and key areas for vegetation rehabilitation. GAP analysis was a very useful approach for identifying the key areas to enlarge current habitats by right vegetation rehabilitation ACKNOWLEDGEMENTS Financial support for this research came from the National Natural Science Foundation of China (40321101) and the National Basic Research Program of China (contract no.: G2000046807). REFERENCES Anderson, G.S., and Danielson, B.J. (1997). The effects of landscape composition and physiognomy on metapopulation size: the role of corridors. Landscape Ecology, 12, 261-271. Baudry, J., Burel, F., Aviron, S., Martin, M., Ouin, A., Pain, G., et al. (2003). Temporal variability of connectivity in agricultural landscapes: do farming activities help? Landscape Ecology, 18, 303-314.

POTENTIAL HABITAT IDENTIFICATION OF GIANT PANDA

111

Bayliss, J.L., Simonite, V., and Thompson, S. The use of probabilistic habitat suitability models for biodiversity action planning. Agriculture, Ecosystems and Environment (in press). Bennett, A.F. (1990). Habitat corridors and the conservation of small mammals in a fragmented forest environment. Landscape Ecology, 4, 109-122. Bennett, A.F, Henein, K., and Merriam, G. (1994). Corridor use and the elements of corridor quality: chipmunks and fencerows in a farmland mosaic. Biology Conservation, 68, 155-166. Caicco, S.L., Scott, J.M., Butterfield, B., and Csuti, B. (1995). A gap analysis of the management status of the vegetation of Idaho, USA. Conservation Biology, 9, 498-511. Chave, J., Wiegand, K., and Levin, S. (2002). Spatial and biological aspects of reserve design. Environmental Modeling and Assessment, 7, 115-122 Chen L.D., Fu B.J. and Liu X.H. (2000). Landscape pattern design and wildlife conservation in nature reserve - case study in Wolong Nature Reserve. Journal of Natural Resources, 15, 164-169. Chen, L.D., Liu, X.H., and Fu, B.J. (1999). Evaluation on giant panda habitat fragmentation in Wolong Nature Reserve. Acta Ecologica Sinica, 19, 291-297. Committee of Bio-Diversity of Chinese Academy of Sciences (CBCAS). (1994). Approaches and Principles to Study Bio-Diversity. Science and Technology Press, Beijing, China (in Chinese) Danielson, B.J., and Hubbard, M.W. (2000). The influence of corridors on the movement behaviour of individual Peromyscus polionotus in experimental landscapes. Landscape Ecology. 15, 323-331. Dauber, J., Hirsch, M., Simmering, D., Waldhardt, R., Otte, A., and Wolters, V. (2003). Landscape structure as an indicator of biodiversity: matrix effects on species richness. Agriculture, Ecosystems and Environment, 2083, 1-9 Fahrig, L. (2003). Effects of habitat fragmentation on biodiversity. Ann. Rev. Ecol. Evol. Syst. , 34, 487–515 Feng, W.H. (1991). Collected papers on giant panda studies. J. Sichuan University, (special issue), 7-13. Giant Panda Survey Team of Wanglang Nature Reserve. (1974). Preliminary survey of giant panda in Wanglang Nature Reserve, Pingwu county, Sichuan. Acta Ecologica Sinica. 20, 162-169. Henein, K., and Merriam, M. (1990). The elements of connectivity where corridor quality is variable. Landscape ecology, 4, 157-170. Hu, J.C. (1981). Study on biological nature of giant pandas, Gordon Monkey, Taking in Wolong Nature Reserve. Chengdu, China. Sichuan People Press. Hu, J.C. (1996). Progress on giant panda studies in the years 1990-1994. Journal of Sichuan Teachers College, 17, 1-8. Hu, J.C., Schaller, G.B., Pan, W., and Zhu, J. (1985). Giant pandas in Wolong Nature Reserve. Chengdu, China: Sichuan People Press. Jactel, H., Goulard, M., Menassieu, P., and Goujon, G. (2002). Habitat diversity in forest plantations reduces infestations of the pine stem borer Dioryctria sylvestrella. Journal of Applied Ecology, 39, 618-628 La Polla, V.N., and Barrett, G.W. (1993). Effects of corridor width and presence on the population dynamics of the meadow vole (Microtus pennsylvanicus). Landscape Ecology, 8, 25-38. Li, C.B. (1997). A Study of Staple Food Bamboo for The Giant Panda. Guizhou Science and Technology Press, Guiyang, China. Li, W., Wang, Z., Ma, Z., and Tang, H. (1999a). Designing the buffer zone in a biological reserve based on suitable habitats: Yancheng Reserve and the red crowned crane. Biological Conservation, 90, 159-165. Li, W., Wang, Z., Ma Z., and Tang, H. (1999b). Designing the core zone in a biosphere reserve based on suitable habitats: Yancheng Biosphere Reserve and the red crowned crane (Grus japonensis). Biological Conservation, 90, 167-173. Liu, J., Linderman, M., Ouyang, Z., An, L., Yang, J., and Zhang, H. (2001). Ecological Degradation in Protected Areas: The Case of Wolong Nature Reserve for giant pandas. Science, 292, 98-101 Liu, J., Ouyang, Z., Taylor, W.W., Groop, R., Tan, Y., and Zhang, H. (1999). A framework for evaluating the effects of human factors on wildlife habitat: the case of Giant Panda. Conservation Biology, 13, 1360Liu, X.H., Bronsveld, M.C., Toxopeus, A.G., Kreins M.S., Zhang, H., Tan, Y. et al. (1997). GIS application in research of wildlife habitat change - a case study of the giant panda in Wolong Nature Reserve. The Journal of Chinese Geography, 7, 51-60.

112

L.D. CHEN ET AL.

Lü, Y.H., Chen, L.D., Fu, B.J. and Liu, S.L. (2003). A framework for evaluating the effectiveness of protected areas: the case of Wolong Biosphere Reserve. Landscape and Urban Planning, 63, 213-223. Lü, Y.H., Fu, B.J., Liu, S.L., and Chen, L.D. (2003). Comprehensive evaluation of the function of Wolong Nature Reserve. Acta Ecologica Sinica, 23, 571-579. Mckendry, J.E., and Machlis, G.E. (1991). The role of geography in extending biodiversity gap analysis. Applied Geography II, 135-152. Nepal, S.K., and Weber, K.E. (1994). A buffer zone for biodiversity conservation viability of the concept in Nepal Royal Chitwan National Park. Environmental Conservation. 21, 333-341. Ouyang, Z.Y., Liu, J.G., Xiao, H., Tan, Y.C. and Zhang, H.M. (2001). Assessment of giant panda habitat in Wolong Nature Reserve. Acta Ecologica Sinica, 21, 1869-1874. Pan, W.S. (1988). Qinling Mountain’s Natural Habitat for Giant Pandas. Peking University Press, Beijing. Qin, Z.S. and Taylor, A. (1993). Forest Dynamics, Bamboos and Environment in Wolong Nature Reserve. Beijing Forestry Press, Beijing, China. Schaller, G.B. (1993). The Last Panda. The University of Chicago Press, Chicago. Steiner, N.C. and Köhler, W. (2003). Effects of landscape patterns on species richness - a modelling approach. Agriculture, Ecosystems and Environment, 2086, 1-9 Taylor, A., and Qin, Z.S. (1998). Forest landscape dynamics and panda conservation in southwestern China. In Nature's Geography (pp. 56-74), The University of Wisconsin Press. Wang, J.X., and Ma, Z.G. (1993). Ecological Studies on Giant Panda's Main Feed Bamboo. Sichuan Science and Technology Press, Chengdu, China. Wolong Nature Reserve (WNR) and Sichuan Normal University (SNU). (1987). The vegetation and resource plants in Wolong. Sichuan Publishing House of Science and Technology, Chengdu, China. Wolong Nature Reserve (WNR) and Sichuan Normal University (SNU). (1992). The animal and plant resources and protection of Wolong Nature Reserve. Sichuan Publishing House of Science and Technology, Chengdu, China. Xu, J.Y., Chen, L.D., Lü, Y.H., and Fu, B.J. (2005). Harmonization of protected areas management and local development: methods, practices and lessons. Chinese Journal of Ecology, 24, 102-107. Yang, G., Hu, J.C., Wei, F.W., and Wang, W. (1994). The number and activity of the giant panda Population in Dafengding nature reserve, Mabian. Journal of Sichuan Teachers College, 15, 114-118. Yu, K.J. (1996). Security patterns and surface model in landscape planning. Landscape and Urban Planning, 36, 1-17.

CHAPTER 8

LAND USE CHANGE FROM TRADITIONAL TO MODERN ERAS: SAITAMA PREFECTURE, JAPAN

R. SEGUCHI1, R. D. BROWN1, K. TAKEUCHI2 1

School of Environmental Design and Rural Development, University of Guelph, Guelph, Ontario, Canada N1G 2W1; 2Laboratory of Landscape Ecology and Planning, The University of Tokyo, Yayoi 1-1-1 Bunkyo-Ku, Tokyo, 113 Japan

Abstract. This study investigated changes in land use from the traditional era (1880s) to the modern era (1990s) in Saitama Prefecture, Japan. Shortly after Japan opened their country to foreign trade, but before outside influence had affected land use, the first mapped survey of Japan was undertaken. These original 1880s maps, which revealed traditional land use patterns, were compared with modern land use through geographic information system analysis. The amount of paddy field remained essentially unchanged between the 1880s and the 1990s, but forest vegetation decreased from 39% to 20%, and urban development land increased from 5% to 24% of the area. A comparison of land use change on each landform type indicated that land use patterns were determined more by the capability of the land in the 1880s than in the 1990s. An analysis of the transition in vegetation revealed that natural succession was underway in much of the woodland

1. INTRODUCTION For at least two millennia, humans have used rural landscapes in Japan intensively. Ever since the beginning of the Yayoi Era (circa 300 B.C.E.) trees have been extracted as a source of fuel and as building materials. By the end of the Nara Era (circa 800) forests surrounding human settlements had become primarily secondary woodlands, consisting mainly of Pinus densiflora and Quercus serrata. Those secondary woodlands were maintained at an early stage of natural succession through constant disturbance by human processes such as extraction of wood for

113 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 113–128. © 2007 Springer.

114

R. SEGUCHI ET AL.

building materials, fuel and charcoal, collection and composting of fallen leaves, and harvest of understory plants (Takeuchi et al., 2003). Total extraction of the woods was undertaken every decade through coppicing. Human processes also created a variety of other environments in the rural landscape, including grasslands, ponds, paddy fields, and mono-cultural plantation woodlots. This disturbance of natural forest succession and resulting variation of landscape provided habitat for the establishment of unique species diversity (Buckley, 1995) that continued until the middle of the Showa Era (1950s). The 1960s saw the beginning of rapid change in Japan. Inexpensive and readily available fossil fuels, chemical fertilizers, and building materials meant that the traditional roles of the woodlands became redundant (e.g. Kameyama et al., 1996). Rapid economic growth and mass migration from rural to urban areas exacerbated the situation, resulting in abandoned rural areas and unmanaged woodlands (Takeuchi et al., 2003). Over the past four decades many of the unmanaged secondary woodlands have been undergoing a successional transformation (Suminami 1998, Takeuchi et al., 2003), while mono-cultural plantations of Chamaecyparis obtuse and Cryptomeria japonica have expanded in size (Higuchi 1996). The 1991 publication of Jinsokusokuzu (land use maps) from a century earlier has enabled scholars to study traditional landscapes and their patterns. Bessho et al. (2001) used the Jinsokusokuzu to analyse vegetation change in the Tama Hills through the use of a geographic information system (GIS), while Ichikawa et al. (in press) used them to study the transformation of satoyama in the urban fringe of Tokyo. Ogura (1993, 1994) analysed these maps to determine that the natural landscape of the 1880s was mainly young forest vegetation. Sprague et al. (2000) argued that the land use of the 1880s had been sustainable over a long period of time and might provide insight into future sustainable development. The objective of this study was to measure land use change and vegetation transition from the traditional era (represented by the 1880s) to the modern era (represented by the 1990s) with a focus on the effect of landform. The Hiki Region of Saitama Prefecture, a hilly landscape near Tokyo, was used as a case study. Through this study we hope to provide insight into whether land use patterns in the traditional era were more determined by landform than those of the modern era. 2. STUDY SITE The study site is a 10.4 km by 12.6 km rectangular area that is featured in six map sheets of original Jinsokusokuzu in the Hiki region, Saitama Prefecture, Japan (Figure 1). It is a hilly area that spreads between the Chichibu Mountains in the west and the Ara River in the east. Rolling hills of elevation up to 100m have been eroded away by small streams that run in intricate patterns throughout the area, forming a number of small valleys (Ishizuka et al., 1986, Takeuchi et al., 2003). The hills spread across the northern and southern thirds of the site with a valley plain between.

LAND USE CHANGE IN SAITAMA, JAPAN

115

Figure 1. Location of study site in Saitama Prefecture, Japan.

The hills are mainly covered by Brown forest soils. These dry and nutrient-poor soils originate on tertiary sandstone, mudstone, or tuff (Saitama Prefecture, 1974). This poor soil quality precluded agricultural production and allowed only coppice production. The main forest vegetation of the Hiki Region is Pinus. Humans utilized vegetation on the hills as a natural resource supply for centuries (Matsui et al., 1990). The valley plain is covered with Grey lowland soils. These are fine-textured, high viscosity soils with abundant acid accumulation that provides for high land productivity. The upland is mostly covered with Aeolian volcanic ash soils. These provide medium land productivity and are often used for field crops and residential areas (Saitama Prefecture, 1974). The natural vegetation of the area was Quercus

116

R. SEGUCHI ET AL.

glauca and Quercus myrsinaefolia but is now primarily Pinus densiflora and Quercus serrata woodlands (Ishizuka et al., 1986, Miyawaki et al., 1975). 3. DATA SOURCES Jinsokusokuzu survey maps were produced between 1880 and 1886 by the Japanese Army for the purpose of collecting local geographical information in order to keep the capital city under military control (Figure 2). Maps were initially produced in colour, but after approximately 1000 of these maps had been produced the process was changed to black line maps with symbols. The colour maps were stored for almost a century before they were released through publication in 1991 (Geographical Survey Institute 1991). The colours on the maps represent vegetation types: forests are green; paddy fields are yellow; and field crops are pale yellow. In addition, the vegetation types are described in written form on the maps. This information provides a valuable snapshot of land use in the traditional era as the land use of the 1880s was very similar to that of the Edo Period (1603-1867) a time when traditional Japanese lifestyle was at its most established (Moriyama, 1997). Information on modern land use was digitally available through a variety of sources. Survey of the vegetation of Japan has been conducted by Japanese Ministry of the Environment every five years since 1973 (Japanese Ministry of the Environment, 1999). The two most recent inventories (1988-1992 and 1993-1998) were available in digital format but must be overlain on previous maps as they contain only information on vegetation change. Every prefecture conducted the Fundamental Land Classification Survey of Japan. Saitama prefecture has provided three maps used in this study: Landform Classification; Surface Geology; and Soils in 1974. 4. METHOD 4.1 Data Collection The Jinsokusokuzu maps were digitised to allow comparison between eras. Extracted information from the six Jinsokusokuzu that covered the study site were manually traced onto transparent sheets and converted to digital format using Adobe Photoshop then merged into one map. The lines of the scanned image were traced, given geographical references and geological correlation, converted into polygon data, and given a legend, using Microsoft TNTmips. Road crossings that still exist today were used for geographical reference. The resulting maps were exported to ArcView in shape file format to allow GIS analysis. Two maps were produced using the process described above: a polygonal map of traditional land use (Figure 3) and a point data map of traditional era vegetation (Figure 4).

LAND USE CHANGE IN SAITAMA, JAPAN

Figure 2. Black and white photograph of a Jinsokusokuzu produced in the 1880s and published in 1991.

117

118

R. SEGUCHI ET AL.

Figure 3. Polygonal map of traditional land use (1880s) of study site, Saitama Prefecture, Japan.

Areas of woody vegetation on the Jinsokusokuzu contained written descriptions of the dominant plant species of the area. However, they did not provide enough information to gain solid spatial data for each description so a vegetation point data map was created. Analysis was undertaken on the assumption that these specific points at least were covered with the described species.

LAND USE CHANGE IN SAITAMA, JAPAN

119

Figure 4. Point data map of traditional era vegetation (1880s) of study site, Saitama Prefecture, Japan.

Six descriptions appeared on the maps of the study site: matsu (Pinus densiflora); nara (Quercus serrata); kunugi (Quercus acutissima); take (Phyllostachys spp.); sugi (Cryptomeria japonica); and boku of which there is no agreement as to the interpretation. For the purposes of this study we simply used the term boku. The modern era vegetation and land use map (Figure 5) was obtained by combining the Actual Vegetation Maps from the 3rd, 4th, and 5th National Surveys. The vegetation map also includes information on other land uses that were similarly

120

R. SEGUCHI ET AL.

extracted for analysis. The original legend of the 1990s maps were simplified and reclassified so as to correspond with the land use legend of the Jinsokusokuzu.

Figure 5. Polygonal map of modern land use (1990s) of study site, Saitama Prefecture, Japan.

The Landform Map, available in digital format, consisted of 100m grid data digitised from the analogue maps of the National Land Agency. The grid data were converted to polygonal data (Figure 6). The classification of the original legend was used without reclassification. The three major landforms on the site were loam upland (29%), hills (29%), and valley plain (24%).

LAND USE CHANGE IN SAITAMA, JAPAN

121

Figure 6. Polygonal map of landforms of study site, Saitama Prefecture, Japan.

4.2 Data Analysis The three sets of maps (traditional era, modern era, and landforms) were analyzed using ESRI ArcView. The area covered by each land use was compared between the 1880s and the 1990s. The land use maps were then cropped according to landform, using the geoprocessing function of ArcView. For each landform, the area occupied by each land use was calculated and compared over time. This revealed any tendency for certain landforms to be occupied by specific land uses.

122

R. SEGUCHI ET AL.

Next the transformation of forest vegetation was examined. The vegetation points from the Jinsokusokuzu were overlaid on the modern era vegetation map. If the point remained covered by forest vegetation, the points from the 1880s were labelled according to its vegetation type of the 1990s. In this way the changing tendency of vegetation were revealed. The significance of landform in vegetation transformation was investigated by clipping the vegetation maps according to landform. 5. RESULTS AND DISCUSSION Forest vegetation occupied 39% of the total area in the 1880s, but had decreased to 20% in the 1990s (see Figure 7). Field crops had also decreased in total area, from 23% to 9%, while the area occupied by paddy field remained constant. Orchards, mulberry, and tea fields had increased markedly from 1% to 22% and urban development area increased from 5% to 24%.

Figure 7. Land use of the site in 1880s and 1990s, Saitama Prefecture.

In the 1880s forest vegetation occupied more than 50% of the mountains, hills, loam upland, and nearly 50% of cliffs, while paddy field covered close to 50% of natural levees, flood plains, dry riverbeds, and river terraces. By the 1990s mountains were the only landform to be more than 50% occupied by forests. An index of frequency was used to reveal land uses that are concentrated on a specific landform compared to the total site. It was calculated by dividing the ratio of a type of land use occupying a certain landform by the ratio of that land use occupying the entire site. When the value is higher than 1.0 it means that the land use is more concentrated on that landform than on the entire site. The larger the value, the higher the concentration (Table 1 a and b). In the 1880s paddy fields were concentrated on former river courses, deltas, valleys on upland, and valley plains, while field crops were concentrated on flood plains and dry riverbeds, and mulberry

LAND USE CHANGE IN SAITAMA, JAPAN

123

and tea fields were concentrated on natural levees and deltas. The urban development areas were concentrated on natural levees and river terraces while forest vegetation was not strongly concentrated on any landform type. By the 1990s forest vegetation was concentrated on mountains and hills, while paddy fields were concentrated on the same landforms as in the 1880s. Orchards, tea and mulberry fields were concentrated on loam uplands and river terraces, field crops were on flood plains, and urban development areas did not show strong concentration on any specific landform type.

Table 1a. Index of frequency for land uses on each landform in the 1880s.

Table 1b. Index of frequency for land uses on each landform in the 1990s.

The vegetation transformation of the entire site is shown in Figure 8. There were a total of 385 forest vegetation description points identified on the Jinsokusokuzu. In the 1880s, 215 of those points were Pinus densiflora, 76 were Quercus acutissima, and 59 were Quercus serrata. Of the 385 points, 128 remained as forest vegetation in the 1990s. Of those 128 points, 62 were Pinus densiflora, and 53 were Quercus serrata. There was no vegetation description of Quercus myrsinaefolia in the 1880s, but there was one point of this in the 1990s.

124

R. SEGUCHI ET AL.

Figure 8. Vegetation points on overall site. The column on the left is the 1880s vegetation and the one on the right is the 1990s vegetation. The arrows and the corresponding numbers indicate how each vegetation point changed.

Hills cover 29% of the total area, and 143 of the forest vegetation points in the 1880s were on this landform. The nature of the vegetation transformation on hills is shown in Figure 9. By the 1990s, 86 of these points remained as forest vegetation, while 57 changed to other uses, 30 of which were golf courses. Figure 9 reveals that the vegetation change was quite dynamic. For example, although 16 of the points in the 1880s and 53 points in the 1990s were Quercus serrata, only 7 of the points had remained the same. The other 28 points of Quercus serrata in the 1990s had transformed from a different 1880s forest type, majority of which was succession from Pinus densiflora. Pinus densiflora is a plant of primary succession that is hardy on dry soil and can tolerate high levels of disturbance, but does not compete well with other tree species (Harada and Isogai, 2000). According to Ishizaka et al. (1986) and Bessho et al. (2001), it is on moist soil that Quercus serrata forests first take over Pinus densiflora-dominated hill slopes. The potential vegetation of the hills in the study site is Abies firma – Quercus glauca, and that of the loam uplands is Quercus myrsinaefolia (Miyawaki et al., 1975).

LAND USE CHANGE IN SAITAMA, JAPAN

125

Figure 9. Vegetation points on hills. The column on the left is the 1880s vegetation and the one on the right is the 1990s vegetation. The arrows and the corresponding numbers indicate how each vegetation point changed.

Figure 10. Vegetation points on loam upland. The column on the left is the 1880s vegetation and the one on the right is the 1990s vegetation. The arrows and the corresponding numbers indicate how each vegetation point changed.

126

R. SEGUCHI ET AL.

Loam uplands cover 29% of the total area and contained 196-wooded vegetation points. The vegetation transformation on loam uplands is shown in Figure 10. Only 31 of these points remained as forest vegetation in the 1990s. Of those points that did not remain as forest, 58 became mulberry fields and 59 became residential. 6. SUMMARY Overall, from the 1880s to the 1990s, the amount of forest vegetation decreased by almost half, from 39% to 20% of the total area, while field crops decreased from 23% to 9%. In contrast, urban development increased almost five-fold, from 5% to 24% of the total area. In the 1880s landform had a strong influence on land use. Mountains, hills, and cliffs were primarily forested, field crops occupied uplands, and lowlands were used for paddy fields. By the 1990s much of this relationship had been lost. Varieties of land uses tended to share land on every landform unit. Lowlands was the only landform that didn’t change land use, remaining primarily in paddy fields. While in the 1880s urban development was concentrated on river terraces and natural levees, by the 1990s there was no relationship between landform and urban development. Forest vegetation in the 1880s was distributed mostly on hills and uplands. Although forest vegetation decreased in area by the 1990s, 90% of it was still on hills and uplands. Forest vegetation that had been dominated by Pinus densiflora in the 1880s had changed to a mixed forest of Pinus densiflora and Quercus serrata by the 1990s. The forest vegetation in the 1990s had not reached a level of natural succession that might be expected, but indications are that the process is underway. 7. CONCLUSIONS The comparison between land use practices of the 1880s and the 1990s in relation to landform revealed that land use in the 1880s was more determined by landform than in the 1990s. In other words, land use was more determined by the capability of the land in the 1880s than in the 1990s. The land use pattern in the 1880s was very clear. Woodlands spread from mountains to hills to loam uplands. Paddy fields were located on the lowlands. The land between mountains and lowlands were used for field crops, residential areas, and other miscellaneous uses. By the 1990s the tendency for one variety of land use to occupy a certain landform was less pronounced, with the only strong relationship that of paddy fields on lowlands. Urban development spread across the whole site. The amount of forested land decreased between the 1880s and the 1990s, and the quality changed. Woodlands of the 1990s have largely been abandoned, while those of the 1880s were highly disturbed through human use including coppicing and harvesting of leaf litter and understory plants. This change in management regime has allowed natural succession to take place, albeit slowly. The constant disturbance of the woodlands in the centuries leading up to the 1880s had produced a highly specialized ecology, different from the original primeval forest condition, and also

LAND USE CHANGE IN SAITAMA, JAPAN

127

different from the uniform forestry plantations that have often replaced it (Buckley, 1992). However, this specialized ecology requires a continuation of intense human intervention (Takeuchi et al., 2003) a requirement that is unlikely to be met given current trends in lifestyle and globalisation. 8. LIMITATIONS The use of digital data allows effective analysis of large areas. However, it contains inevitable limitations. The original Jinsokusokuzu were one of the earliest attempts at mapping using modern survey methods in Japan (Nagaoka, 1991) and were neither as accurate nor as precise as modern standards. In the process of digitising the Jinsokusokuzu certain information was extracted and traced manually. Some degree of error is unavoidable. The landform map was based on a 100m grid and when converted to vector format to allow overlay with other maps provided only rough polygons with a lower resolution. However, the error is probably minimal since landforms, by their very nature, can generally not be precisely delineated. The vegetation transition analysis was conducted based on point data that was gained from written descriptions on the Jinsokusokuzu. The number of points does not represent the area covered by that specific vegetation type, and the points were not evenly distributed across the site. 9. DISCUSSION The type of analysis done in this study is valuable for illustrating the ways that land use has changed over time. The results reveal that, in traditional times in Japan the characteristics of landforms were important indicators of land use. People tended to use the land for activities that were suited to the specific landforms. In recent times, however, land use has less relationship with landform, with most land uses occurring on almost any landform. There has been considerable recent discussion and concern regarding the abandonment of rural landscapes in Japan (e.g. Takeuchi et al., 2003) and the effect on the natural environment. Abandoned landscapes are steadily becoming less favourable environments for the unique flora and fauna that have long been part of the disturbance regime of rural Japan. However, it is increasingly unlikely that there will be a return to traditional lifestyle and land use patterns in Japan. The increasingly open global market is making agriculture less viable in Japan, resulting in more abandonment. Rural people, particularly the young, are continuing to migrate into cities, resulting in a rural population that is both decreasing and aging. The results of this study suggest that abandonment of the woodlands of rural Japan might actually have a positive effect on the natural environment. As natural succession takes place the landscape will move more towards its potential vegetation. The victims of this transition will be the unique flora and fauna that was part of the heavily managed landscape. The only way these unique environments will survive in modern Japan may be by conservation efforts. If that is the case whether or not they survive will be in our hands.

128

R. SEGUCHI ET AL.

ACKNOWLEDGEMENTS Thanks to Dr. Shori Yamamoto of the National Institute of Rural Engineering in Japan for his advice, and for providing databases for this research. REFERENCES Bessho, H., Tsunekawa, A. and Takeuchi, K. (2001). Tamakyuryo Tsurumigawa ryuiki ni okeru GIS wo mochiita Satoyama no shokuseihenka (A study on vegetation change of Satoyama in the basin of Tsurumiriver, Tama hills by using GIS). Theory and applications of GIS, 9(2), 83-90. (In Japanese) Buckley, G.P. (1995). Ecology and Management of Coppice Woodlands. Chapman & Hall, London, 336pp. Geographical Survey Institute. (1991). Meiji zenki tegaki saishoku Kanto kansokuzu, Shiryohen (Manually Rendered Survey Map of Kanto of the Early Meiji, Attachment material). Japan Map Center, Tokyo, 253pp. (In Japanese) Harada, H. and Isogai, T. (2000). Matsu to Shii (Pinus and Castanopsis cuspidate). Iwanami Shoten, Tokyo, 106pp. (In Japanese) Higuchi, H. (1996). Hozenseibutsugaku (Conservation Biology). University of Tokyo Press, Tokyo, 253pp. (In Japanese) Ichikawa, K., Okubo, N., Okubo, S. and Takecuhi, K. (in press) Transition of the satoyama landscape in the urban fringe of the Tokyo metropolitan area from 1880 to 2001. Landscape and Urban Planning. Ishizaka, T., Takeuchi, K., Okazaki, M. and Yoshinaga, S. (1986). Kita Hikikyuryo ni okeru chikei, dojou no hairetsu to shokusei bunpu (Vegetation distribution influenced by the landform and soil arrangement in the Hiki-kita Hills, Saitama Prefecture, Central Japan). Applied Phytosociology, 15, 1-16. (In Japanese) Japanese Ministry of the Environment (formerly The Japanese Environment Agency), Nature Conservation Bureau. (1999). The Dataset for GIS on the Natural Environment, Japan ver.2 (CD Rom) Kameyama, A. (1996). Zoukibayashi no shokuseikanri (Vegetation Management of Coppice). Softscience, Tokyo, 303pp. (In Japanese) Matsui, K., Takeuchi, K. and Tamura, T. (1990). Kyuryochi no shizenkankyou: sono tokusei to hozen (Natural Environment of the Hills : It’s Character and Conservation). Kokon Shoin, Tokyo, 202pp. (In Japanese) Miyawaki, A., Okuda, S. and Inoue, K. (1975). Saitamaken nanseibu no shokusei (Vegetation of SouthEastern Saitama). Saitama Prefecture, 86pp. (In Japanese) Nagaoka, M. (1991). Meijizenki no tegaki saishoku Kanto kansokuzu – Daiichigunkan chihou nimanbun no ichi jinsokusokuzu kaidai (Manual survey maps of Kanto area in the early Meiji Era – Reading Jinsokusokuzu 1:20,000). Journal of The Geographical Survey Institute, 74, 22-32. (In Japanese) Ogura, J. (1993). Meiji chuki ni okeru Bousoukyuuryou no shokuseikeikan (Vegetation of the Hills of the Boso Penninsula in the Middle Meiji Era). Journal of the Japanese Institute of Landscape Architecture, 56(5), 25-30. (In Japanese) Ogura, J. (1994). Meiji 10 nendai ni okeru Kantochiho no shinrinkeikan (Forests of the Kanto Region in the 1880s). Journal of the Japanese Institute of Landscape Architecture, 57(5), 79-84. (In Japanese) Saitama Prefecture. (1974). Fundamental Land Classification Survey, Kuma gaya. Saitama Prefecture, Urawa. 51pp. (In Japanese) Sprague, D.S., Goto, T. and Moriyama, H. (2000). Jinsokusokuzu no GIS kaiseki ni yoru Meiji shoki no nouson tochiriyou no bunseki (GIS Analysis Using the Rapid Survey Map of Traditional Agricultural Land Use in the Early Meiji Era). Journal of the Japanese Institute of Landscape Architecture, 63(5), 771-774. (In Japanese) Suminami, Y. (1998). Ryokuchihozenseido to satoyama (Satoyama in Conservation System of Open Space). Journal of the Japanese Institute of Landscape Architecture, 61(4), 290-292. (In Japanese) Takeuchi, K., Brown, R.D., Washitani, I., Yokohari, M. (2003) Satoyama: The Traditional Rural Landscape of Japan, Springer, Tokyo, New York. 229 pp.

CHAPTER 9

EVALUATION AND PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

Y. NATUHARA Osaka Prefecture University, Osaka, Japan

Abstract. We compared the response of various taxonomic groups, birds, butterflies, ants, trees and ferns, in the large city of Osaka, Japan in order to examine relationships between the abundance and arrangement of the habitats, and life history trait of the species. We presented species specific responses to habitat fragmentation. Species richness decreased more rapidly in birds than ants from the urban to rural ends of the urban gradient, and butterflies were intermediate. Birds were influenced by the habitat area and distance to species source. In contrast, ants were less influenced by habitat area, but were susceptible to the history of the isolated habitats. In ants, trees and ferns, some rare species occurred even in small habitats and the small habitats contributed to species diversity in the urban areas. Simultaneously, variation of the life history affected the distribution of species. For example, Parus major could breed in urban area by using scattered trees in an urban matrix; their home range enlarged in the urban area to secure sufficient food. One of the major goals of urban landscape ecology is to use scientific information to restore and preserve biodiversity in urban ecosystems. Some examples of planning and adaptive management for wildlife habitats in urban landscapes were introduced

1. INTRODUCTION Loss of biodiversity is one of most important problem in the global environmental issues. Although biodiversity in urban landscape does not contribute the global biodiversity very much, it has important functions for human life. The functions include improvement of urban ecosystem, remediation of urban climate, fostering culture, providing a place of environment education, etc. Urban ecosystem studies of Japan started in 1971 as a part of a project, “human survival and environment" (Numata, 1976). He defined urban ecosystem as one of man-modified and man-made ecosystems, and he proposed three approaches to urban ecosystems; (1) Examining the flow of energy, matter, people and

129 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 129–147. © 2007 Springer.

130

Y. NATUHARA

information, (2) studying the impact of an urban environment on the terrestrial and aquatic ecosystems, and (3) studying reactions to the abiotic environment (in the Clementian sense). He and his colleagues reported a series of urban ecological studies of the metropolitan Tokyo (see Numata, 1982). Urban ecological study in Japan reaches over many fields, such as pollution of air and water, urban climates, water cycle, etc, and one of important subject is impact of urbanization on flora, fauna and biological community. Early studies focused on biological indicator of urbanization; lichen (Taoda, 1973), alien plants (Hotta, 1977), soil arthropods (Aoki, 1979), and retreat of wild animals (Chiba, 1973) were used as indicators. Through comprehensive study of urban ecology, Numata (1987) noticed importance of landscape ecology in urban ecosystem studies. He emphasized this approach as holistic/integrated approach to anthropocentric ecosystem. Though importance of the fusion between urban ecology and landscape ecology was emphasized, only a few papers have been published in Japan. Some of them are of pattern analysis of urban landscapes (Yokohari and Fukuhara, 1988; Hasebe and Suzuki, 1997; Ochi et al., 2000), and others are of function of urban landscapes as habitat (Hamabata, 1980; Higuchi et al., 1982; Hashimoto et al., 1994; Hattori et al., 1994; Toyama and Nakagoshi, 1994; Imai and Natuhara, 1996; Natuhara and Imai, 1996; Yabe et al., 1998; Goto et al., 1999; Hashimoto et al., 2005a). In this essay we identify effects of habitat islands on wildlife in urban and suburban areas, and test hypotheses that (1) habitat fragmentation in urban gradient generaly increase species diversity, but (2) the effects of urbanization depend on the life histories of the species, by comparative study of different taxonomic groups. We also explore methods for conservation of species diversity in the urban area. 2 Osaka Prefecture occupies an area of 1864 km in the western region of Japan and has a human population of 8,830,000. The annual mean temperature and precipitation are 16.2°C and 1,400 mm, respectively. The monthly mean temperature varies from 5.5°C in January to 28.2°C in August and that of precipitation from 34 mm in December to 206 mm in July. Elevations in Osaka vary from sea level at the Osaka Plains to 1125 m at Mt. Kongo. The dominant potential natural vegetation types are evergreen broad-leaved forest, consisting of Castanopsis cuspidata var, Sieboldii and Quercus glauca from plain to hilly areas, and deciduous forest, of Fagus crenata in the upper mountain. Analyses of pollen in geographic strata reveal that lowland forests were cleared and changed to paddies 2000-3000 years ago. 2. HABITATS ALONG URBAN GRADIENTS Urbanization provides an environmental gradient from a highly developed core to a rural or natural area (Numata, 1976, McDonnell and Pickett, 1990). Analysing changes in a biological community along such a gradient provides a scientific basis for planning ecological cities, but also facilitates testing of hypotheses through management of urban landscapes. In the Osaka Prefecture, as in other cities, building areas decrease from the centre of the urban core to the suburbs, while the area of cultivated field increases and then decreases with increasing forest area

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

131

(Figure 1). In urban areas, fragmented habitats appear as islands in a matrix of build, cultivated and natural areas. Many studies have used island-biogeography theory to analyse fragmented habitats in urban areas, reporting that important variables affecting the species were forest area, degree of isolation, age since isolation, and combinations of these factors (Willis, 1979; Higuchi et al., 1982; Howe, 1984; Opdam et al., 1984; Askins et al., 1987; Soule et al., 1988; Bolger et al., 1991; Haila et al., 1993; Ichinose and Katoh, 1994). The determinants of species diversity are often too complex to be modelled by area alone. The incorporation of other variables such as a measure of habitat heterogeneity or resource availability may be necessary (Boecklen and Gotelli, 1984). Another important finding was that bird species in isolated habitats tended to show high nestedness (Simberloff and Abele, 1976; Patterson and Atmar, 1986; Hashimoto et al., 2005a). Species are not distributed randomly among the isolated habitats, with their response to fragmentation influenced by their population and life-history traits (With et al., 1997; Natuhara and Imai, 1999). At the same time, urbanization creates land mosaics (Forman, 1995; Natuhara and Imai, 1996). A gradient can be seen in the land covers, and in intermediate zone, mosaics of forest and open field are detected (Figure 1).

Figure 1. Urban gradient in the study site, Osaka, Japan.

132

Y. NATUHARA

Formation of the mosaic landscape sometimes increases the biodiversity, though urbanization is generally a major cause of the loss of biodiversity. We examine this mosaic model of habitat and its meaning in the urban area. The affects of landscape mosaics on biodiversity are the result of various mechanisms, and are exerted at various spatial scales (Hansson, 1979; Helle andMuona, 1985; Yahner, 1988). Species richness may increase in a mosaic of habitats by the following mechanisms: the formation of a new habitat at a boundary between neighbouring elements, such as vegetation of forest edges with shrub layer forming the mantle and the lianas forming the veil, and the mosaic effect per se, i.e., a mosaic of forests and open lands can contain habitats for both forest species and open land species. Several butterflies use a set of resources in the habitat, e.g. food plants for larva, nectar source for adults, and their habitat must be a mosaic of different vegetation. However, some species may disappear from the mosaic due to the fragmentation of habitat, and less mobile species cannot move beyond a barrier between patchy habitats. Furthermore, the quality of urban and suburban habitats are changing in the study area; the traditional coppicing in the Satoyama landscape (Natuhara et al., 1999; Fukamachi et al., 2001), which is a mosaic of secondary forests, grasslands, farmlands and irrigation ponds, has been abandoned over the last 40 years. Potentially, lucidophyllous forests are developing in the southwestern Japan; most of them have changed to deciduous Quercus forest by human use over the past several thousand years. Furthermore, the forests had been fragmented by farmlands, and both forests and farmlands were changed to buildings and asphalt by urbanization. The gradient of urban habitats includes the following components: the reduction, isolation, mosaic formation, and the ecological succession of habitats. 3. FAUNAL CHANGE ALONG THE URBAN GRADIENT The intensity of habitat fragmentation by urbanization changes from the suburbs to the urban core, influencing the arrangement of habitat. Forests are fragmented by farmlands in rural areas and both of forests and farmlands change to buildings and asphalt in the urban core. Examination of faunal change along this environmental gradient from continuous large habitats to habitat mosaics is important for conservation planning. In the intermediate level of forest reduction (% forest cover is from 65% to 10%), increasing farmland area forms land mosaics (Figure 1) and the number of fragmented forest increased to between 10 and 35. In this section we focus on the effects of the habitat mosaics in Osaka on birds and butterflies assemblages at the landscape scale. For bird assemblages, we recorded the proportion of nine types of land use (forest, scatter forest, farmland, grassland, bare ground, residence, city centre, pond and river, and sea) and the presence/absence of each of 76 breeding birds in 5-km square quadrats on a map of the Osaka Prefecture (Natuhara and Imai, 1996). For butterflies, 78 butterflies were recorded along 19 transect routes in various land covers (Natuhara, 2000)

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

133

The number of bird species is shown against percentage of forest (Figure 2). The highest richness is detected at around the 50% level. The relationship varies among species group with different habitat use. Species richness changed according to the proportion of forest in the 5-km square and it showed a unimodal curve. However, the response was different in forest-interior species from forest-edge species; the former increased approximately linearly with increasing proportion of forest.

Figure 2. Relationships between percentage of forests in a grid and (C) species richness of birds, and (D) species richness of forest interior birds and forest edge birds.

The changes in bird diversity along the Osaka urban-gradients were also detected by ordination (Natuhara and Imai, 1996) and in butterfly (Natuhara, 2000). The proportion of forest and that of farmland are two major environmental gradients according to Principal Components Analysis of the nine types of land use. Ordination by Canonical Correspondence Analysis showed that breeding bird distribution differentiated along the two major clines, forest and farmland. Avifauna changed successively along those environmental gradients (Natuhara and Imai, 1999). There were no discrete boundaries of the distribution of bird-species groups. We tentatively classified five groups of quadrats on the ordination plane of the sample score and five groups of bird species in CCA-ordination plane. The occurrence of these bird groups correlated with land use; the first group with forest area, the second one with scatter forest, the third with farmland, and the fourth had a relation to seashore. Although diversity of land use seemed to increase species richness in the third group, less diverse and forest-rich group contained as many species as the third group. These effects of mosaics may appear on various spatial scales, and butterfly assemblages can be studied on several spatial scales, such as the region, landscape, and local habitat. On the landscape and smaller scales, the landscape mosaic enhanced the species richness of butterflies, however, the diversity indices and specialist butterflies (univoltine-tree feeder) decreased in the mosaic landscape (Natuhara et al., 1999).

134

Y. NATUHARA

4. URBAN HABITATS AS ISLANDS We compared species-area curves among five taxonomic groups, birds, butterflies, ants, trees and ferns. Generally the relationships are summarized by the power function (1) and log function (2), although the logistic curve was also used. S = c Az

(1)

S = b log A + c

(2)

where S is species richness, A is habitat area, and b, c and z are constants. Coefficient of determinants was high for all groups but ants were high in both models (Table 1). In the power function analysis, the species richness in 1-ha areas was highest in trees, followed by ants, ferns, butterflies, with the lowest in birds, though statistically significant regression could not be estimated for ants. The slope was highest in ferns, then trees, butterflies, birds and lowest in ants. The slope was highest in trees when using the log function. Age of habitats may affect the species richness. When habitats were classified by their age since development (i.e., younger than 10 year-old, between 11 and 50, and older, Figure 3), there was no significant difference in the species richness between other age groups in birds and butterflies with the exception of the Nanko power station, by the analysis of covariance. The Nanko power station had extremely few bird species and many butterfly species. This habitat consisted of three, seven and ten year-old woods; most trees were evergreen, and shorter than 2 m in the youngest stands and grew 10 m with high density (20 / 100 m2) in the older stands. In contrast, the species-area curve for ants varied among the three age groups. The effect of distance to species source or continuous large forests on the species richness of birds, butterflies and ants in Osaka was analyzed using linear regression. The effect of distance was greatest in butterflies and least in ants, and there are no significant correlation in trees and ferns (Table 2). Moving ability and habitat size of the species influences differences in proportion of occurrence in urban area and the distance effect. But these figures are biased, because species-distance relationships are not linear. There are 55 forest birds bred in the Osaka, but the intercept is 13.7. This means 75% of bird species do not breed at fragmented habitats even if the location is close to the continuous habitats. The intercept of butterflies is 43.7 that is about 56% of butterfly species recorded by transect counts in Osaka Prefecture. The intercept of ants is 17.7 that is 20% of 87 ant species recorded in Osaka Prefecture (Natuhara, 1998). But the distribution of ants is random among fragmented habitats, and total number of species recorded in the fragmented habitats was 50 or 57% of total species. Possible causes of no species-distance correlations in trees, ferns and ants in Kyoto are recruitment of seeds from street trees and gardens in the urban matrix, and long dispersal distance of spores of ferns. In ants, dominance of passive dispersal with gardening materials causes long distant dispersal. Furthermore,

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

135

possible metapopulation structure in urban habitats for these taxonomic groups may reduce the slope of species-distance curve (Hanski and Gyllenberg, 1997).

Figure 3. A comparison of species-area curves between birds and ants in Osaka, Japan. Simboles indicates years after plantation of forests; Squares and thin lines: more than 50 years, circles and thick lines: 10-50 years, and triangles and broken lines: less than 10 years.

For most bird species, urban forests are too small to keep enough large size of population for a long-term persistence. Thus, populations in urban forests are sink that are supported by source populations in mountainous area. On the other hand, many species of ants are able to keep large populations in the urban forests, and a metapopulation is formed. In our original analysis of the butterfly data, the distance relationship was not significant (r2 = 0.294, P = 0.0884), but when the Nanko power station site was removed from the data, a significant regression was estimated (Table 1). This site is located at a harbour far from the species source, but many butterflies were recorded at this site because Rutaceae trees, which are important food plants for butterflies, are more abundance than at other urban parks. Furthermore, some butterfly species migrate along the seacoast where the Nanko power station is located. Table 1. A comparison of species-area relationships in taxonomic groups in Osaka, Japan.

Group Birds1) Butteflies2) Ants (Osaka)3) Ants (Kyoto)4) Trees5) Ferns6)

c 1.595 6.580 10.751

S = c Az z 0.235 0.293 0.0495

r2 0.880 0.762 0.0477

c 1.2962 4.0877 11.453

15.109

0.1011

0.1345

16.046

1.398

0.1235

37.8 9.387

0.3007 0.4077

0.671 0.636

43.4 13.5

11.295 5.3468

0.780 0.668

S = b Ln A + c b 3.7167 4.5939 0.696

r2 0.6872 0.6985 0.0832

1) Natuhara and Imai (1999), 2) Imai and Natuhara (1996), 3) Natuhara (1998), 4) Yui et al. (2001), 5) Murakami and Morimoto (2000), 6) Murakami et al. (2003)

136

Y. NATUHARA

The species composition of birds in smaller habitats was a nested subset of the larger habitats (Table 2). The occurrence pattern varied according to the species. Fifteen species possibly breed in urban parks. Streptopelia orientalis, Hypsipetes amaurotis, Zosterops japonica, Passer montanus, and Sturnus cineraceus were recorded at all forest, and increasing forest areas in the urban areas could increase the other ten species. In areas with lower amounts of forest, the numbers of carnivorous and insectivorous species decreased more rapidly than those of granivores and omnivores. Hashimoto et al. (2005a) also reported that the woodland bird species occurrence pattern in woodland of Kyoto City was highly nested. Distribution of species in urban habitats depends on their life history. The most important trait is the variety of microhabitat use by the species. Nakamura (1988) classified the habitats in the study area into open spaces in forests, open space edges, forest edges, and forest interiors, and reported an occurrence pattern of species in those microhabitats. Species using the forest edge, P. major, Aegithalos caudatus, and Emberiza cioides occurred in middle size urban forest in the present study (Table 2). Cettia diphone, which uses not only forest edges but also open space in forests, occurred only in large forest. Forest interior species did not occur in the urban forest. The distribution of carnivorous birds and insectivorous birds were limited to large forests. This is understandable for carnivores because they need a large home range. Canaday (1997) discussed the causes of the negative relationship between the number of insectivorous species and the degree of human impact. His list of important factors includes: microclimatic changes that have altered the insect prey base, greater habitat sensitivity among insectivores resulting from their high degrees of ecological specialization, changes in predation upon these birds, and interference competition from opportunistic, disturbance-adapted omnivores (Canaday, 1997). In California, 6 of 12 species not found in developed area were insectivores, and no insectivore persisted in developed areas (Blair, 1996). In Canada, Lancaster and Rees (1979) reported that 63% of species were insectivores in forest, in contrast to 1% of species in commercial industrial areas. Passer major and Carduelis sinica, in particular, increased in occurrence as urban forest patches increased from 1 to 20 ha. Habitats for these species at forest edges or scattered forest may emerge in this size of forest. Similarly, habitats in the forest interior and open spaces in forests for C. diphone and Dendrocopos kizuki may emerge in forests of more than 20 ha. Habitat use of these species in urban parks has not been studied so far. A microhabitat use of these species in the urban parks and comparison of microhabitat structures among urban parks are needed to confirm the habitat diversity hypothesis. Generalist or forest edge species adapt their behaviour to urban landscapes. Parus major minor, are found in both forests and urban parks in Japan. However the population density is very low in downtowns of Tokyo and Osaka. In our study of 85 parks in Osaka, P. major were present in only 12, with an additional 3 occurrences noted in areas outside of the study sites (Hashimoto et al., 2005b). Although these 3 occurrences were not study sites, we counted them in the number of surrounding habitats. The average areas of parks where P. major were recorded and not recorded were 26.0± 42.1 S.D and 2.1 ± 2.9 ha, respectively. The smallest

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

137

park was 0.56 ha. Breeding behaviour was compared between two areas with different percentage of wood cover (Inoue et al., 2005). Interestingly, home range areas of breeding pairs were larger in the area with lower percentage of wood cover. Mean home range size was 4.70 ha at the site where the percentage of tree cover was 31.5% and 0.76 ha at the other site where the percentage of tree cover was 76.6%. P. major breeding in the former site probably required a larger home range to get food for their fledglings. In comparison with birds and butterflies, ants, trees and ferns did not show nestedness, partly because they usually need smaller areas as habitats. Peintinger et al. (2003) also reported a random pattern in distributions of vascular plants, bryophytes, butterflies and grasshoppers; species richness increased with the area of habitat islands, but overlap among them was so low that even small habitat islands contributed to overall species richness.

Table 2. Nestedness of occurrence in urban habitats for bird species in Osaka, Japan. Area of wood lands (ha) ≤2

20.1-

5.1-20

2.1-5

8

7

6

7

(ha)

38.5

7.99

3.12

0.96

Picus awokera Parus varius) Phasianus colchicus Cettia squameiceps Eophona personata Cettia diphone

13

0

0

0

25

0

0

0

63

0

0

0

Dendrocopos kizuki

88

0

0

0

Emberiza cioides

50

14

0

0

Bambusicola thoracica

38

29

0

0

Aegithalos caudatus

75

29

0

0

Corvus macrorhynchos

88

14

33

0

100

43

17

14

63

43

33

14

100

29

33

14

88

29

33

29

100

57

67

71

88

100

83

71

100

100

100

100

Number of sites Average area Species

Corvus corone Lanius bucephalus Parus major Carduelis sinica Zosterops japonica Sturnus cineraceus Streptopelia orientalis Hypsipetes amaurotis Passer montanus

Percentage of occurrence sites

138

Y. NATUHARA

From the above results, we conclude that species requirements of space and mobility strongly affect their distribution in urban habitats. The hypothesis of random immigration and extinction in the equilibrium theory of island biogeography is not realistic (Patterson and Atmar, 1986). Occurrence of bird species is not random, but each species follows a specific pattern (Lancaster and Rees, 1979; Higuchi et al., 1982; Ambuel and Temple, 1983; Blake, 1991; Hansen and Urban, 1992; Canaday, 1997), that is the nested subset pattern in which most species present in small habitats are also occurring in larger habitats. The distribution of species in urban habitats affects strategies for conservation. Abundance in a species' source habitat, its body size (Soule et al., 1988; Bolger et al., 1991) and necessity of a forest interior (Blake, 1991) are important factors among life history characteristics that influence the nested subset pattern. Consider two hypothetical patterns of species in habitats: nested subset and heterogeneous. In the former case, habitats holding fewer species do not contribute to species diversity at a regional scale. In the latter case, few species occur in multiple habitats because of differences in their habitat requirements, and habitats holding fewer species contribute species diversity at a regional scale. By simulating random arrangement of samples, we detected that several small habitats contain more species than a single large habitat of the same size in ants and the opposite results in birds and butterflies. However, this pattern of butterfly distribution is not universal. No nestedness was detected in butterflies inhabiting mountainous wetlands (Peintinger et al., 2003). 5. HABITATS AT FINE SCALES Fine-scale heterogeneity also enhances the species diversity of birds (MacArthur and MacArthur, 1961). In our study, ant species diversity increased with increasing microhabitat diversity. Yui et al. (2001) reported that species richness of ants was positively correlated with the number of microhabitat types, such as stones, herbaceous patches, etc. They explained the species richness of ants at two different scales by using a Structural Equation Model (Figure 4). The area and shape index of forest affected the habitat quality at larger scales, and management intensity and microhabitat diversity in the forest were important at smaller scale. The effect of habitat quality was stronger at smaller scales than at larger scales. The mosaic at a fine scale is also important for butterflies. The effect of gap clearance in the study area on butterfly assemblages was monitored by Chikamatsu et al. (2002). The average number of species and number of individuals were higher at the gaps (11.3 species, 40 individuals) than the interior of forest (3.2 species, 7 individuals). Among four gaps where the sky factor (proportion of open area to allsky area observed from the forest floor) was measured using hemispherical photographs, population density and species richness with correlated with the sky rate. Pollard and Yates (1993) summarized the butterfly monitoring in Monks Wood that clearance at the edges of the ride is beneficial to butterflies. For the scale of individual movement, butterflies can sample a much more diverse array of habitat types in fine-grained landscapes (Debinski et al., 2001).

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

139

Figure 4. Structural Equation Model for species richness of ants in urban habitats. Latent variables, Env.1: Area and shape index of woodlots and Env.2: Management and micro habitat diversity of woodlots, determine species richness of ants (after Yui et al., 2001).

6. PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE One of goals of urban landscape ecology is to restore ecosystems in urban areas. The latest style of regional ecosystem planning in Japan is that of the Grand Design of the Urban Environmental Infrastructure in the Metropolitan Area (Ministry of Land Infrastructure and Transport, 2004). Objectives of the plan are enhancement of the following five functions of nature: protection of biodiversity, chance to contact nature, beautify landscape, reduction of environmental load, and disaster prevention. The evaluation procedure for protecting areas of biodiversity are as followings: (1) Extract cohesive habitat as protection area. At first ecotopes were classified to 36 types by terrain and vegetation in town block size. Then ecotope types were ranked from one to five by the incidence of wildlife species, including mammals, butterflies and freshwater fish. (2) Evaluation of ecological networks using “guide species” which are selected so as to represent variation of habitats. Connectivity was evaluated from potential habitats of each of the guide species estimated by their home-range sizes and habitat environments. Finally, 25 areas were selected as nature conservation areas based on the total score of five functions. Another system is Guideline for Nature Conservation established in more than 14 prefectures. This guideline aims to grasp and evaluate the present situation of nature to identify regions to conserve, and to show the goal and course of the

140

Y. NATUHARA

integrated and designed policy (Hokkaido Prefecture, 1989). Evaluation items are the degree of human disturbance of vegetation, endangered species, and the number of endangered species (Iwate prefecture, 1995) in each area, which is a 1-km grid (Okinawa prefecture, 1999), or a polygon based on ecotope classification and basin (Fukui prefecture, unpublished). In smaler scale, fauna in urban areas are influenced by area and shape of habitat, percentage of forest in the surrounding areas, and distance to the species source. Among these, as mentioned in the previous section, area is the most important factor for most groups. An isolated forest does not perform well as a habitat for many birds. Askins et al. (1987) reported that there were no differences in population of edge species between larger forests (>187 ha) and smaller ones, but significant difference in populations of forest interior species. In particular, insectivorous interior species are the most sensitive to forest fragmentation. However, edge species can inhabit forest larger than 20 ha. If a larger area of forest cannot be supported in urban areas, increasing the percentage of forest in local areas by planting small groves can help colonization of forest edge species. It is important to realize that urban areas, which represent the fragmentation of potential habitats, can support a diversity of several groups such as butterflies, even disturbance avoider species, if we create and maintain suitable environments (Hogsden and Hutchinson, 2004). In Osaka, planted forest (age >10 years) was not inferior for birds to older forest with the same area (Natuhara and Imai, 1999). Vale and Vale (1976) reported that garden plantings seem most influential in determining the distribution and density of birds. The horticultural plantings are typically more luxuriant and provide more diverse habitats than the pre-suburban environments. This is also true for insects; distribution and abundance are more likely to be limited by the availability of suitable habitat than by their migration ability (Wood and Pullin, 2001). Several methods for increasing urban biodiversity were tested with replication (Gaston et al., 2005), and they found some of the methods, such as bamboo sections as a nesting site for solitary bees and wasps. From an ecological viewpoint, it is better to design habitats for individual species than for species richness or abundance as a whole. Hashimoto et al. (2005b) focused on P. major for reasons mentioned in the previous section? By reason of mentioned previous section, and found the best fitting logistic regression model for describing the distribution of P. major in Osaka was logit P = - 18.144 + 3.799 A250 + 0.688 N1 Where P is the probability of occurrence, A250 is the area of tree cover within a radius of 250 m, and N1 is number of other habitats within 1 km (Figure 5). More tree cover is needed for Great Tits if the number of nearby habitats is small. By applying the model it was found that to achieve a probability of occurrence of 0.5 when the number of habitats within a 1 km buffer was 0, 1, 2 and 3, tree areas of 6.0 ha (31%), 4.0 ha (20%), 2.6 ha (13%) and 1.8 ha (9%) are required.

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

141

Figure 5. Estimation of incidence probability of P. major by area of tree crown within and number of neighboring habitats. When there are three neighboring habitats within 1 km, incidence probability is 0.5 at habitat of 1.2 ha. (after Hashimoto et al., 2005b).

From the model and its examination (Hashimoto et al., 2005b), a minimum of 1.8 ha tree cover in a radius of 250 m or 9% of the area, with at least 3 other habitats within 1 km are factors necessary to provide habitat for P. major in urban areas. Thus, more than 10 % of tree cover is a realistic target figure for an ecologically sustainable environment for urban areas. The tree cover of Osaka City in 1991 was only 4.1 %, but its target figure for 2005 is 15% (Osaka City, 1995), which includes trees in large parks. A target of 10% tree cover for areas outside of large parks is required to maximize avian biodiversity. The target of 10% tree cover is rather high and difficult to achieve in the urban area of Osaka, but there will be chances to create habitat for P. major by using combinations of park networks, roadside trees and rooftop gardens. 7. ADAPTIVE MANAGEMENT OF WILDLIFE HIBITAT IN URBAN LANDSCAPE Ecosystems are unpredictable and it may be useful to modify plans in response to the results of biodiversity monitoring. Adaptive management is a useful method for conservation of biodiversity in urban landscapes. An attempt of the adaptive management for an urban wildlife habitat was reported (Natuhara et al., 2005). Wild Bird Park in Osaka Port was established on the reclaimed land in 1983 by the City of Osaka, and has been managed by the Port and Harbor Bureau and Osaka Port Development and Technology Association. The park has a planted area of 6.5 ha and a sandy area of 12.8 ha, which includes two pools (4.6 ha comprise the north pool and 3.8 ha comprise the south pool) of rainwater and a lagoon (1.4 ha). The lagoon connected with the sea through six

142

Y. NATUHARA

Hume concrete pipes and six steel pipes that pass through the dike. A tidal flat of only 0.2 ha has emerged at the east side of the lagoon. Two pools did not connect with sea and were filled with rainwater. The park functions as a sanctuary for ducks. However, fewer shorebirds came to the site than we expected. We assumed that the area of tidal flat is too small to attract shorebirds and concluded to restore one of pools to a tidal flat for shorebirds. The north pool had been separated from the lagoon by a mound of soil. The mound between the lagoon and the north pool was breached in 1995 to restore tidal flat that provides habitat for shorebirds (Figure 6). The north pool became to be affected the tide through the Hume concrete pipes and to be filled with sea water at high tide and dry at ebb tide. Thus, the area of tidal flat increased to 2.6 ha one of the ponds was restored to a tidal flat in 1995 after consultation among the manager, NPO, and scientist as a result of the monitoring. The area of tidal flat increased from 0.2 to 2.6 ha, and the number of shorebirds increased from 205 (the average of 1991– 1995) to 1042 (1996). The species composition of benthic animals had also changed; the dominant group was Chironomid larvae before the repair and Polychaetes after the repair, and the species richness increased. Natural ecosystems are often unpredictable and are difficult to manage.

Figure 6. Changes in the landscape of the wild bird park after the reform in 1995. Black areas show the tidal flat.

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

143

An artificial tidal flat that we constructed at the wild bird park in Osaka has not been a good habitat for shorebirds for 13 years. However we could change the management plan and restored the tidal flat in 1995, and which made the park attractive for shorebirds. The success of this plan depended on continuous monitoring and on the council system among the manager, the citizen’s group and scientists. The adaptive management is a management-asexperiment in dynamic situations where controls and strict replication are not possible (Holling, 1978). General process of the adaptive management is setting goal(s), building models, planning, carrying out the plan, and adapting the plan according to monitoring and evaluation of the result. Modelling habitat change is important for restoration of salt marshes (Boumans et al., 2002). The Wild Bird Park was designed taking account to the habitat change by subsidence. However the actual change was not as same as the estimate, and the management was modified. Our experiment demonstrates the efficacy of the approach of the adaptive management. Table 3. Changes in spring shorebird numbers in the wild bird park.

Charadrius. dubius C. alexandrinus C. mongolus Pluvialis fulva P. squatarola Arenaria interpres Calidris ruficollis C. acuminata C. alpina C. tenuirostris Limicola falcinellus Tringa erythropus T. nebularia T. glareola Heteroscelus brevipes Actilis hypoleucos Xenus cinereus Limosa limosa L. lapponica N. phaeopus Gallinago gallinago Phalaropus lobatus Other species Total Species richness

1950-56 20 200 15 31 10 250 283 21 107 4 75 1 1 100 2 8 1 50 400 3 49 35 1665 33

74-80 19 360 7 4 5 80 155 58 500 6 1 31 11 33 153 3 10 1 40 6 1 21 12 1516 31

Period 84-89 35 397 6 10 1 9 67 9 133 1 1 1 1 2 8 1 6 13 30 2 633 17 1381 30

90-95 26 30 5 2 1 3 31 4 51 1 1 1 2 7 1 2 3 3 28 10 1 23 226 29

96-2001 56 700 20 15 4 33 1450 10 678 10 1 1 3 3 60 3 5 17 4 25 7 6 44 3148 36

Maximum numbers observed at a census day in each period of years are shown. Species that recorded more than 10 individuals at least once are shown Source: Kobayashi 1959; Takada 1998, 2002.

144

Y. NATUHARA

Another example of adaptive management is that of an urban forest. In 2001, 6artificial gaps (15 x 15 m) were created in an urban forest (98.5 ha) at the Expo'70 Commemorative Park in an urban area of Osaka. Thirty years have passed since the completion of land reclamation and the planting of broad-leaved evergreen trees that were established at the reclaimed land by planting of broad-leaved evergreen trees. Planners thought the broad-leaved evergreen forest was the potential natural vegetation of the site. However, low penetration of solar radiation to the forest floor seemed to restrict the diversity of forest floor vegetation (Nakamura et al., 2005) and the effect of gap clearance on butterfly assemblages was monitored (Chikamatsu et al., 2002; Yamamoto and Natuhara, 2005). Butterflies were recorded in 15 x 15 m quadrats for 10 minutes at six gaps, six plots of interior forest adjacent to the gaps, a vegetable garden, and an area of turf. Average number of species, S and number of individuals, N were 13 and 55, respectively at a vegetable garden, 11.3 and 40 at gaps, 3.2 and 7 at forest interiors, and 2 and 6 at the turf site. Five species, including Papilio bianor were recorded only at the gaps, and the gaps changed the species composition of butterflies and increased the species diversity in the park as a whole. Future monitoring and a clearance program are planned because population density of some species of butterflies and birds change with years after coppicing (Fuller et al., 1989; Warren, 1987). ACKNOWLEDGEMENTS I should acknowledge the following persons for their help: Mark McDonald, Michael McCarthy, Hiroshi Hashimoto, Yukihiro Morimoto, Hisayuki Maenaka, and Akihiro Nakamura. REFERENCES Ambuel, B. and Temple, S.A. (1983). Area-dependent changes in the bird communities and vegetation of southern Wisconsin forests. Ecology, 64, 1057-1068. Aoki, J. (1979). Difference in sensitivities of oribatid families to environmental change by human impacts. Revue D'Ecologie et de Biologie du Sol., 16, 415-422 Askins, R.A., Philbrick, M.J. and Sugeno, D.S. (1987). Relationship between the regional abundance of forest and the composition of forest bird communities. Biological Conservation, 39, 129-52 Blair, R.B. (1996). Land use and avian species diversity along an urban gradient. Ecological Applications, 6, 506-519. Blake, J.G. (1991). Nested subsets and the distribution of birds on isolated forest. Conservation Biology, 5, 58-66 Boecklen, W.J. and Gotelli, N.J. (1984). Island biogeographic theory and conservation practice: speciesarea or specious-area relationships. Biological Conservation, 29, 63-80. Bolger, D.T., Alberts, A.C. and Soule, M.E. (1991). Occurrence patterns of bird species in habitat fragments: Sampling, extinction, and nested species subsets. American Naturalist, 137, 155-66. Boumans, R.M.J., Burdick, D.M. and Dionne, M. (2002). Modeling habitat change in salt marshes after tidal restoration. Restoration Ecology, 10, 543–555. Canaday, C. (1997). Loss of insectivorous birds along a gradient of human impact in Amazonia. Biological Conservation, 77, 63-77. Chiba, S. (1973). Changes in animals habitats and their retreat in Tokyo. Natural Science and Museum, 40, 69-73

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

145

Chikamatsu, M., Natuhara, Y., Mizutani, Y. and Nakamura, A. (2002). Effect of artificial gaps on the butterfly assemblage in urban woods. Journal of the Japanese Society of Revegetation Technology, 28, 97-102 (in Japanese with English abstract) Debinski, D.M., Ray, C. and Saveraid, E.H. (2001). Species diversity and the scale of the landscape mosaic: do scales of movement and patch size affect diversity? Biological Conservation, 98, 179-190. Forman, R.T.T. (1995). Land Mosaics: The Ecology of Region and Landscape. Cambridge University Press, Cambridge. Fukamachi, K., Oku, H. and Nakashizuka, T. (2001). The change of a satoyama landscape and its causality in Kamiseya, Kyoto Prefecture, Japan between 1970 and 1995. Landscape Ecology, 16, 703-717 Fuller, R.J., Stuttard, P. and Ray, M. (1989). The distribution of breeding songbirds within mixed coppiced forest in Kent, England, in relation to vegetation age and structure. Ann. Zool. Fennici, 26, 265-275 Gaston, K.J., Smith, R.M., Thompson, K. and Warren, P.H. (2005). Urban domestic gardens (II): experimental tests of methods for increasing biodiversity. Biodiversity and Conservation 14, 395-413. Goto, S., Morioka, T. and Fujita, S. (1999). Evaluation of ecological networks in urban area by habitat analysis of indicator species. Environmental Information Science, 13, 43-48 Haila, Y., Hanski, I.K. and Raivio, S. (1993). Turnover of breeding birds in small forest fragments: The "sampling" colonization hypothesis corroborated. Ecology, 74, 714-25. Hamabata, E. (1980). Changes of herb-layer species composition with urbanization in secondary oak forests of Musashino plain near Tokyo – Studies on the conservation of suburban forest stands 1. Japanese Journal of Ecology, 30, 347-358 (in Japanese with English Abstract) Hansen, A.J. and Urban, D.L. (1992). Avian response to landscape pattern: The role of species' life histories. Landscape Ecology, 7, 163-180 Hanski, I. and Gyllenberg, M. (1997). Uniting two general patterns in the distribution of species. Science, 275, 397-400. Hansson, L. (1979). On the importance of landscape heterogeneity in northern regions for the breeding population densities of homeotherms: a general hypothesis. Oikos, 33, 182-189. Hasebe, H. and Suzuki, M. (1997). GIS analysis of open space transition in urbanization process of EdoTokyo. Journal of Japan Institute of Landscape Architecture, 60, 633-638 (in Japanese) Hashimoto, H., Murakami, K. and Morimoto, Y. (2005a). Relative species - area relationship and nestedness pattern of forest birds in urban area of Kyoto City. Landscape Ecology and Management 10, 25-36. (in Japanese with English abstract) Hashimoto, H., Natuhara, Y. and Morimoto, Y. (2005b). A habitat model for Great Tits, Parus major minor, using a logistic regression model in the urban area of Osaka, Japan. Landscape and Urban Planning, 70, 245-250 Hashimoto, Y., Kamihogi, A. and Hattori, T. (1994). A study on the conservation of fragmented forests as inhabitant arthropods in the new town, using ant-biodiversity for indicator as arthropods-biodiversity. Journal of Japan Institute of Landscape Architecture, 57, 223-228 (in Japanese) Hattori, T., Kamihogi, A., Kodate, S., Kumadaki, E., Fujii, T. and Takeda, Y. (1994). A study on the actual conditions of the fragmented forests in flower town and their conservation. Journal of Japan Institute of Landscape Architecture, 57, 217-222 (in Japanese) Helle, P. and Muona, J. (1985). Invertebrate numbers in edges between clear-fellings and mature forests in northern Finland. Silva Fennica, 19, 281-294. Higuchi, H., Tsukamoto, Y., Hanawa, S. and Takeda, M. (1982). Relationship between forest areas and the number of bird species. Strix, 1, 70-78 (in Japanese with English abstract) Hogsden, K.L. and Hutchinson, T.C. (2004). Butterfly assemblages along a human disturbance gradient in Ontario, Canada. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 82, 739-748. Holling, C.S. (1978). Adaptive Environmental Management and Assessment. John Wiley & Sons, Chichester. Hotta, M. (1977). Distribution of Taraxacum spp. in Kinki District, Japan. Bulletin of Osaka Museum of Natural History, 1(12), 117-134 (in Japanese) Howe, R.W. (1984). Local dynamics of bird assemblages in small forest habitat islands in Australia and North America. Ecology, 65, 1585-1601.

146

Y. NATUHARA

Ichinose, T. and Katoh, K. (1994). The factors that influenced bird communities on the fragmented woodlots in the Tokorozawa City, Saitama Prefecture. Journal of the Japanese Institute of Landscape Architecture, 57, 235-240. Ide, H. and Kameyama, A. (1993). Landscape Ecology. Asakura-shoten, Tokyo. (in Japanese) Imai, C. and Natuhara, Y. (1996). Comparison of butterfly fauna among urban greeneries in and around Osaka City and an application to island biogeography theory. Japanese Journal of Environmental Entomology and Zoology, 8, 23-34. (in Japanese with English abstract) Inoue, N., Natuhara, Y. and Hashimoto, H. (2005). The effect of percentage of tree cover on breeding performance of Great tit (Parus major). Landscape Ecology and Management, 9(2), 33-39. (in Japanese with English abstract) Lancaster, R.K. and Rees, W.E. (1979). Bird communities and the structure of urban habitats. Canadian Journal of Zoology, 57, 2358-2368. MacArthur, R. and MacArthur, J. (1961). On bird species diversity. Ecology, 42, 594-598. McDonnell, M.J. and Pickett, S.T.A. (1990). Ecosystem structure and function along urban-rural gradients: an unexploited opportunity for ecology. Ecology, 71, 1232-1237. Ministry of Land Infrastructure and Transport of Japan (2004). http://www.mlit.go.jp/kisha/kisha04/02/020315/02.pdf (accessed September 15, 2005) Murakami, K., Matsui, R., Maenaka, H. and Morimoto, Y. (2003). Relationship between species composition of Pteridophytes and micro-landform types in fragmented forest patches in Kyoto City area. Journal of the Japanese Institute of Landscape Architecture, 66, 513-516. (in Japanese with English abstract) Murakami, K. and Morimoto, Y. (2000). Landscape ecological study on the woody plant species richness and its conservation in fragmented forest patches in Kyoto City area. Journal of the Japanese Society of Revegetation Technology, 25, 345-350 (in Japanese with English abstract) Nakamura, A, Morimoto, Y. and Mizutani, Y. (2005). Adaptive management approach to increasing the diversity of a 30-year-old planted forest in an urban area of Japan. Landscape and Urban Planning, 70, 291-300. Nakamura, N. (1988). Mori to Tori to (Forests and birds). Shinano Mainichi Shinbunsha, Nagano. (in Japanese) Natuhara, Y. (1998). Ant faunae in Osaka City and three other sites in Osaka Prefecture. Bulletin of Myrmecological Society of Japan, 22, 1-5. (in Japanese with English abstract) Natuhara, Y. (2000). Changes in butterfly assemblage along the urban-forest gradient. Journal of the Japanese Institute of Landscape Architecture, 63, 515-518. (in Japanese with English abstract) Natuhara, Y. and Imai, C. (1996). Spatial structure of avifauna along urban-rural gradients. Ecological Research, 11, 1-9. Natuhara, Y. and Imai, C. (1999). Prediction of species richness of breeding birds by landscape-level factors of urban woods in Osaka Prefecture, Japan. Biodiversity and Conservation, 8, 239-253. Natuhara, Y. Imai, C. and Takahashi, M. (1999). Pattern of land mosaics affecting butterfly assemblage at Mt Ikoma, Osaka, Japan. Ecological Research, 14, 105-118. Natuhara, Y., Kitano, M., Goto, K., Tsuchinaga, T., Imai, C., Tsuruho, K. and Takada, H. (2005). Creation and adaptive management of a wild bird habitat on reclaimed land in Osaka Port. Landscape and Urban Planning, 70, 283-290. Numata, M. (1976). Methodology of urban ecosystem studies. In Science for Better Environment. Proceeding of the International Congress on the Human Environment (1975, Kyoto, pp. 221-228). HESC, Tokyo. Numata, M. (1982). Changes in ecosystem structure and function in Tokyo. In Bornkamm, R., J.A. Lee and M.R.D. Seaward (Eds.), Urban Ecology (pp. 139-147). Blackwell, Oxford. Ochi, S, Ikegami, Y. and Nakagoshi N. (2000). Analysing landscape change at patch level on urbanizing region. Journal of Japan Institute of Landscape Architecture, 63, 775-778 Opdam, P., van Dorp, D. and ter Braak, C.J.F. (1984). The effect of isolation on the number of forest birds in small woods in the Netherlands. Journal of Biogeography, 11, 473-8. Patterson, B.D. and Atmar, W. (1986). Nested subsets and the structure of insular mammalian faunas and archipelagos. Biological Journal of the Linnean Society, 28, 65-82. Peintinger, M., Bergamini, A. and Schmid, B. (2003). Species-area relationships and nestedness of four taxonomic groups in fragmented wetlands. Basic and Applied Ecology, 4, 385-394.

EVALUATION PLANNING OF WILDLIFE HABITAT IN URBAN LANDSCAPE

147

Pollard, E. and Yates, T.J. (1993). Monitoring Butterflies for Ecology and Conservation. Chapman and Hall, London. Simberloff, D.S. and Abele, L.G. (1976). Island biogeography theory and conservation practice. Science 121, 285-286. Soule, M.E., Bolger, D.T. Alberts, A.C., Wright, J., Sorice, M. and Hill, S. (1988). Reconstructed dynamics of rapid extinctions of chaparral-requiring birds in urban habitat islands. Conservation Biology, 2, 75-92. Taoda, H. (1973). Bryo-meter, an instrument for measuring the phytotoxic air pollution. Hikobia, 6, 224228. Toyama, M. and Nakagoshi, N. (1994). A study on structure of urban greenery spaces and inhabitant ants. Journal of the Japanese Society of Revegetation Technology 20, 13-20. (in Japanese with English abstract) Vale, T.R. and Vale, G.R. (1976). Suburban bird populations in west-central California. Journal of Biogeography, 3, 157-165. Warren, M.S. (1987). The ecology and conservation of the heath fritillary butterfly, Mellicta athalia. III. Population dynamics and the effect of habitat management. Journal of Applied Ecology, 24, 499-513 Willis, E.O. (1979). The composition of avian communities in ramanescent woodlots in southern Brazil. Papeis Avulsos de Zoologia (Sao Paulo), 33, 1-25. With, K.A., Gardner, R.H. and Turner, M.G. (1997). Landscape connectivity and population distributions in heterogeneous environments. Oikos, 78, 151–169. Wood, B.C. and Pullin, A.S. (2002). Persistence of species in a fragmented urban landscape: the importance of dispersal ability and habitat availability for grassland butterflies. Biodiversity and Conservation, 11, 1451-1468. Yabe, K., Yoshida, K. and Kaneko, M. (1998). Effects of urbanization on the flora of open space in Sapporo City. Journal of Japan Institute of Landscape Architecture, 61, 571-576 (in Japanese with English abstract) Yahner, R.H. (1988). Changes in wildlife communities near edges. Conservation Biology, 2, 333-339. Yamamoto, K. and Natuhara, Y. (2005). The change of butterfly assemblages after artificial gap formation in an urban park. Journal of the Japanese Institute of Landscape Architecture 68, 585-588. Yokohari, M. and Fukuhara, M. (1988). Analysis of mixed land use in urban fringe using Landsat TM data. Journal of Japan Institute of Landscape Architecture, 51, 335-340 (in Japanese with English abstract) Yui, A., Natuhara, Y., Murakami, K. and Morimoto, Y. (2001). Factors influence the species richness of ants in urban woods. Journal of the Japanese Society of Revegetation Technology, 27, 78-83. (in Japanese with English abstract)

CHAPTER 10

LANDSCAPE ECOLOGY FOR BIODIVERSITY Scaling up

T.H. RO, S.-K. HONG Korea Environment Institute, Seoul 122-040, Korea; Institute of Island Culture, Mokpo National University, Jeonnam 534-729, Korea

Abstract. Biological diversity has been emerged as a core concept in management and conservation of diverse ecological systems. Scaling up for biodiversity conservation in landscape system is also emerging issue in ecologists. In order to conserve biodiversity from the genetic level to ecosystem and landscape levels, multi-scale strategies and efforts are being adopted and executed in many countries. In this paper, comprehensive and necessary considerations arisen from the view of landscape ecology were discussed for the present situations of wildlife conservation and management in Korea compared with other countries. Especially, the conservation strategy and policy of biodiversity were addressed in broad senses including habitat protections, legal approaches, landscape design and ecological network programs.

1. WORLD’S WILDLIFE: STATUS AND CRISIS Landscape factors—the natural and artificial types of lands including types of utilization of lands by man—are the parts of a general landscape that looks common in sight and provide species with diverse ecological systems such as habitats. The spatial factors in landscape are naturally arranged and built by the environmental inclination. However, the habitats of living things ruined by inconsiderable utilization of land by man has recently become a threat to the survival of living things as the landscape mosaic becomes simpler (Ro et al., 2000). The phenomena in biodiversity that are caused by this rapid change of landscapes are mainly classified into two: the acceleration of the extinction rate of species and the increase of biological invasion by alien species. These two characters of the phenomena occur simultaneously and complement each other (Spellerberg, 1996; Szaro and Johnston, 1996). Therefore, the landscapes plan for biodiversity conservation strategy of natural wildlife should be made under the policy that can supplement these two. The followings are predictions of the species extinction (UNEP, 1995; Washitani, 1999). 149 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 149–161. © 2007 Springer.

150

T.H. RO AND S.-K. HONG y y

y y y y y

Around 15~20% of species on earth will disappear from 1980 to 2000. Of the 250,000 species of vascular plants, 1,000 have already been extinct for the last 100 years and 60,000 more are predicted to be extinct within 50 years. There is a possibility that 1/4 of the world’s total biodiversity might be extinct 20~30 years from now. Thousands of unclassified insects become extinct every year. 17,500 species or 0.1% of the total species become extinct every year. Half of the terrestrial species will be extinct within 50 years. Half of species on the earth will be extinct from 1990 to 2015.

The direct cause of extinction is the destruction of habitats, often by lumbering, fire, reclamation, and dredging. Habitats are becoming more limited because the diverse natural ecosystems are converted into farmland, or into urban ecosystems of cities in worse cases. The limited population density is threatened by the reduction of population caused by capture and over collection. And the possibility of extinction of living things increases due to environmental pollution such as water pollution, air pollution and soil pollution from insecticides and herbicides. Some lakes in north Europe have become ‘lakes of death’ because all the organisms in the lakes died due to acid rain. The artificial changes of natural environment occurring in the entire world can be summarized as follows (Spellerberg, 1996; UNEP, 1995; Szaro and Johnston, 1996). y Extinction by natural causes y Isolation between terrestrial ecosystem and water ecosystem in the area y Fragmentation and simplification of habitats y Change of habitat quality y Destruction of habitats, capture of specific species, and environmental pollution These are the main causes of the extinction of organisms. Most of these are due to the increase of human population. 2. APPLICATION OF THE PRINCIPLE OF LANDSCAPE ECOLOGY The landscape mosaic has different ranges of organisms with different distribution structures of species according to the landscape elements (such as forests, pasture, stream, and farmlands) that compose the landscape structure. Therefore, species migration is also functionally different according to juxtaposition of landscape elements. The heterogeneous landscape, composed of different types of landscapes, increases the number of large-sized mammals or the edge species that inhabit neighbouring landscape elements. Thus, the diverse species of living things grow in number. Likewise, the increase and decrease of species in a landscape both influence the landscape heterogeneity and are influenced by the landscape heterogeneity. Landscape heterogeneity varies according to the degree of disturbance and landscape stability (Turner and Gardner, 1991; Hong, 1999). The characters of landscape influence landscape stability, which is the capability to cope with the disturbance (Cox, 1993; Forman, 1995; Ro et al., 2000).

LANDSCAPE ECOLOGY FOR BIODIVERSITY

151

From the landscape ecological viewpoint, migration and distribution of organisms have close relations with structure, pattern, and arrangement of landscape elements that compose a landscape. Landscape ecology has similar approaches such as those of the existing ecology. In one sense, landscape ecology is based on the principle of the existing ecology. However, it is different from the existing ecology in that landscape ecology observes the interaction between the organisms in the ecosystem complex that is composed of ‘landscape elements,’ which are a spatial measurement in the landscape scale (Forman and Collinge, 1986; Zonneveld, 1989). Various fields related to ecology such as forestry and landscape architecture have used the principle of landscape ecology for the basic forest planning and management, park, suburban development, and design for the river corridors and habitat. Meanwhile, conservation biologists including the managers of national parks and resorts use the principle of landscape ecology for protecting land, restoring it, and conserving biodiversity. Landscape ecology contributes to the development of policies, management, conservation, design, and planning of landscape under the hypothesis that natures and human beings maintain mutual relations. Landscape ecology also suggests beneficial understanding and prediction regarding the conflicts that people may have with natural environments such as high forest productivity, conservation of species, clean water, establishment of dams, and development of residential areas and resorts. Therefore, landscape ecology includes not only interaction between organism and environment in which the ecology is currently interested but also multi-disciplinary fields about human beings (Hong, 2001; Farina and Hong, 2004). The ultimate goal of landscape ecology is to carry a potential role for sustainability. This is possible only through proper prescription focused on landscape ecology conserving a landscape designed by the land development that can effectively satisfy the ecological integrity and the basic desires of human beings from generation to generation. 3. BIODIVERSITY CONSERVATION: ISSUE OF SCALES In order to conserve biodiversity properly and effectively, the conservation of species and population of animals and plants as well as the protection of habitats and environment must be given high importance. There are differences in the ways of conservation. For instance endangered species may be bred in a zoo, a botanic garden, or in protection cages (ex situ conservation) for the conservation purpose. Conservation of the population can also be done in the original habitats (in situ conservation) (Naveh, 1994). Taxonomy and ecology have taken an important role to protect the natural environment. However, the researches from academic fields do not have enough relation with population genetics that provides important knowledge about the protection and conservation of organisms. What makes landscape ecology different from the existing ecosystem and classical application and categorization to protect nature is that landscape ecology positively adopts the accomplishment and technology of new academic fields such as modelling and spatial ecology. Furthermore, relations of the academic system with other academic fields can be

152

T.H. RO AND S.-K. HONG

explored in researches (Cox, 1993; Forman and Collinge, 1996, Farina and Hong, 2004). That is why landscape ecology is considered worldwide as an important academic field. The followings are researches about biodiversity from the viewpoint of landscape ecology (Washitani, 1999). y Understanding of general theology: the mechanism of existence and conservation of biodiversity is an important theme of research in landscape ecology. y Research on the conservation of species, population, and communities of living things. y Research on the habitats, the conservation of landscape, restoration, and management of landscapes. Through these researches, landscape ecology makes an important turning point that changes the existing standard of evaluation on the value of nature from centering on naturalness to centering on biodiversity. 4. ALTERNATIVES FOR THE CONSERVATION STRATEGY 4.1 Hierarchical Approach Biodiversity can be defined as a concept with hierarchical character, structural character, and functional character. These can be categorized based on four levels of 1) gene, 2) species or population, 3) community or ecosystem, and 4) landscape. The strategy for conservation is set up and carried out based on such a systematic level (Table 1). Table 1. Characters and factors in the concept of biodiversity.

Hierarchy

Composition

Structure

Landscape

Landscape type

Landscape pattern

Communityecosystem

Community type, Ecosystem type

Relationship, Habitat structure

Speciespopulation Gene

Species, Population Genetic composition

Population structure Genetic structure

Function Process and distribution of landscape, Land use pattern Interaction among species, Ecosystem process Life history Population process Genetic process

4.1.1 Species population and community levels To determine the worldwide state of extinction of wildlife, it is necessary to research on the actual condition about which species exist under specific circumstances. For this purpose, the countries in the world classify the species that

LANDSCAPE ECOLOGY FOR BIODIVERSITY

153

could possibly be extinct out of the various fauna and flora based on a certain standard of evaluation (Red List or Red Data Book). The standards used as the degree of extinction in Red Data Book are decided by IUCN. y Extinct species (Ex): The species that cannot be found in the outdoor research field for the past 50 years. y Endangered species (En): The species that are considered to be extinct without policies for protection. y Vulnerable species (V): The species that have a possibility to be extinct eventually, even if there is no danger of being extinct in a short period. y Rare species (R): The species that have especially low population, although there is no cause to reduce the population. y Unknown species (U): The species whose state of survival cannot be evaluated because of impossibility to figure out the number of population. The probability of extinction is decided according to the Population Viable Analysis (PVA) in order to make the standard based on fixed quantities while excluding the subjectivity of such categories. Recently, IUCN (2000) adopted the idea and published the Red Data Book based on the new standard. In this new standard, the species about which material collected based on population are categorized into three, namely, extinct species, threatened species, and low risk species (LR). Table 2. The standard of evaluation by IUCN about extinct and endangered species. Category A. Rapid reduction B. Small distribution range (short, continuing reduction, big change) C. Small group (continuing reduction) D1. Very small group D2. Very small distribution range E. Probability of extinction

Extinct species Reduction to less than 20% within 10 years or in third generation

Vulnerable species Reduction to less than 80% within 10 years or in third generation

Distribution range is less than 100 km2 or habitat is less than 10 km2

Endangered species Reduction to less than 50% within 10 years or in third generation Distribution range is less than 5,000 km2 or habitat is less than 500 km2

The population of grown-ups is less than 250 The population of grown-ups is less than 50

The population of grown-ups is less than 2,500 The population of grown-ups is less than 250

The population of grown-ups is less than 10,000 The population of grown-ups is less than 1,000 Distribution range is less than 100 km2 or 5 places Having probability of extinction with over 10% within 100 years or in fifth generation

-

-

Having probability of extinction with over 50% of reduction within 10 years or in third generation

Having probability of extinction with over 20% of reduction within 20 years or in fifth generation

Distribution range is less than 20,000 km2 or habitat is less than 2,000 km2

154

T.H. RO AND S.-K. HONG

The extinct species category is divided into extinct (EX) and extinct in the wild (EW). The category of low risk species (LR) includes conservation dependent species (cd), near threatened species (nt), and minimum concern species (lc). Extinct and endangered species that have a problem in conservation are classified in detail as follows (IUCN, 2000; Washitani, 1999). 4.1.2 Habitats and landscape levels The strategy for biodiversity conservation is classified into five categories, which are the designation of reservation area (Table 3), the ecological network plan, the conservation by law and regulation, the conservation outside the reservation area, and the conservation of the facilities. The most typical and classical way of conservation of species is the restriction of human activities by designating core areas for conservation of wildlife as reservation areas. An ecological network plan is based on the idea to conserve the remaining habitats both in highly valued area and in low valued area for nature, and to connect them ecologically by coordinating the strategies for conservation for inside reservation area and outside reservation area. Furthermore, the connection by protection outside reservation areas, conservation in artificial facilities, and conservation by laws and regulations are also very important. 4.2 Ecological design 4.2.1 Necessity A habitat patch, a spatial element that composes the landscape mosaic, is distinguished from neighboring areas and relatively equalized areas for habitats of wildlife. The patches of habitats have well-known characters such as large, small, round, long, straight, or bent borders. Contrary to habitat patch, the habitat corridors have lined structures. Those characters widely give the ecological meanings to productivity, diversity, soil, and humidity (Forman and Godron, 1986; Forman, 1995; Farina, 2000). The mosaic with moving habitat patches can generally be considered as a part of a land transformation process or a part of a process of landscape change. The corridors and matrix of the landscape are also in the state of moving. The species and the process of ecosystem are also changed. In fact, community succession is one of the courses that decide speed and direction of change inside the habitat patches caused by human activities (artificial). The lands on the slopes of a mountain can be forests, farm lands, swimming pools, or exposed rocky places according to what kind of force caused the main force among biological succession, farming tractors, excavation equipment, or erosion. Therefore, the patches can be changed into the various directions. Furthermore, landscape can be degenerated, grown, or can remain in stability. In the meantime, a landscape can be diverged into many directions and have other shapes.

LANDSCAPE ECOLOGY FOR BIODIVERSITY

155

For instance, in case of the forest that disappeared because of edge effect and soil erosion by cutting off with construction of forest roads, the success of regeneration process depends on the patch size of the forest. Table 3. Kinds and contents of the reservation area according to IUCN.

Area Absolute reservation area

Category Absolute nature and reservation area for wild life

Reservation area for biological resource and anthropology

Contents -Areas that are absolutely reserved for scientific research, education, and environmental monitoring -The areas should be protected from human beings’ influence as much as possible as possible -Areas with beautiful views of nature reserved for protecting ecosystem, scientific research, education, and recreation -Commercial utilization is not allowed. Areas that are under special care because of biological, geographical and literal particularities Areas that are under control to conserve the special biological communities -Areas with cultural or ecological particularities that allow classical utilization but not destructive utilization -The territories can be used for recreation or tours. Areas that are under restriction for the conservation of natural resources as long as the restriction does not conflict with any government policy -Areas that allow livelihood of human beings in a traditional way -Traditional farming is allowed.

Control area for the multi-purpose utilization

Areas that allow the utilization of natural resources such as water, wildlife, pasture, lumbering, and tours

National parks

Natural monument, Landmark of nature Habitats for wildlife, Reservation area Landscape reservation area

Control area

Natural resource reservation area

In the reservation of lakes and aquifer connected to rivers, the quality of water depends on neighbouring patches of natural vegetation with a large size. Therefore, when the reservation areas of nature are designed, it should be decided whether one large sized patch or several small patches would be ideal.

156

T.H. RO AND S.-K. HONG

4.2.2 Concept The most general methods for the designation of reservation area are the policies of government and the purchase of lands by individuals or nature protection institutions that can restrict human activities. Recently, the designation for reservation area by government or environment protection organizations has widely accelerated. Legal reservation area in the worldwide occupies 5.9% of the whole earth. With regard to the limitedly permitted human activities inside a reservation area, IUCN established the international uniform standard and the categories which consist of five absolute reservation areas and three control areas in 1978 as explained in Table 3. The principle of the plan for the designation of reservation areas for biodiversity was proposed by Diamond in 1975 based on island biogeography (MacArthur and Wilson, 1967; Diamond, 1976; Forman and Godron, 1986; Spellerberg, 1996; Jongman and Pungetti, 2004). It is not clear whether a largesized patch of habitat or a small-sized patch of habitat is better in the ecological sense (large or small; LOS). It is also uncertain whether only one large-sized patch or several small-sized patches are better in an ecological sense (single large or several small; SLOSS). It is not a simple job to compare ecologically a large-sized patch of habitat and a small-sized patch of habitat. The types and environment of habitats need to be identical. If not, the size can be confused according to both the types of habitats and the diversity of habitats. There are many different opinions about LOS and SLOSS. Some people consider that one large-sized patch is more beneficial to conserve organisms. On the other hand, some insist that several small-sized patches are more beneficial. The plan of Man and Biosphere (MAB) of UNESCO, which designates 329 Biosphere Reserves (BRs) in the world and coordinately carries out for human activities, research, and conservation of natural environment, is a pioneering example in the approach of reservation areas. BRs areas include three kinds of zones. In the core zone, the communities and ecosystem should be absolutely protected. The core zone should be ecologically connected to another zone through an ecological corridor. The core zone is surrounded by the buffer zone. In the buffer zone, a plan for the multi-utilization of land should be made in order to achieve harmony between human activities and biodiversity. In the buffer zone, traditional human activities, such as the collection of herbs, and non-destructive research and monitoring are allowed. However, such plans for the reservation areas still have problems. Therefore, in order to designate “nature reserve area,” the following should be resolved first. y Prior nature reserve area cannot represent the entire natural communities. Generally, nature reserve areas are established for breeding and habitation of special animals and plants that need to be conserved. In spite of this, nature reserve areas cannot have perfect ecosystems for habitation of specific organisms. y Most nature reserve areas are too small to conserve the population and all the process of ecosystem. Each species has a different ecological characteristics and life pattern. The minimum sized population required for species to be

LANDSCAPE ECOLOGY FOR BIODIVERSITY

y

y

157

conserved from the viewpoint of environmental biology, or the Minimum Viable Population (MVP), is closely related with the habitat size. Large-sized patches are fragmented because of human impact. In such a case, there is a possibility that species might be changed to metapopulation which cannot conserve the population size for survival. Usually, the migration of animals expands outside nature reserve area. In particular, the habitats of birds and its range of territory are flexible. Therefore, the nature reserve area should include breeding and feeding places as well as habits. The nature reserve area does not consider the influence caused by utilization of lands in surrounding areas. It is the best for the nature reserve area to be built by using the resources of natural ecosystem while reserving the natural environment. An ecotope map has already been made and used in the developed countries in order to find the geographically proper place for conservation. An ecotope map is composed of the biological biotope map used as a control unit of environmental plan and conservation of minimum number of species for the habitats of correlatively similar kinds of animals and plants (Whittaker, 1975; Zonneveld, 1989; Haber, 1994; Riitters et al.,1997; Farina, 2000; Hong, 2002a, b). In addition, it also contains the drawings for the general contents about physiotope and anthrotope of ecosystem to support the ecological function of biotope according to its units. The physiotopes such as soils, weather factors, and temperatures, which are very important for habitats and activities of organisms, are drawn and measured by number (Hong et al., 2004; Nakagoshi et al., 2004). Lastly, the plan for “ecological network system” can be made based on an ecotope map that can help understand both ecological structure and function (Choi, 2004; Hong et al., 2005).

4.2.3 Design In order to conserve biodiversity, a proper nature reserve area should be designed and protected. If necessary, the area should be properly controlled. How the fragmentation and isolation of habitats influence biodiversity should also be seriously considered? In other words, the problems that can occur in the fragmentation process, such as the changes in environmental condition for the population and communities that need to be conserved or the biological invasion in fragmented habitat patch, should be fully considered. Even though there are general guidelines regarding the location, number, and type of habitat patch in nature reserve areas, the first concern for the design should be focused on the relationship among the habitat area, the effect of conservation and the decision of patch type in order to minimize an edge effect. A buffer zone is another important consideration to prevent disturbance and invasion of exotic organism caused by an edge effect. Lastly, it should be fully realized that area-perimeter ratio is reduced, as the area of the habitats patch gets smaller when a large-sized patch changes into several small-sized fragmented patches. Until now, the relations with biodiversity conservation are explained by mainly centering on the structure and function of the habitat patch among the spatial factors that compose landscape mosaic. However, because the

158

T.H. RO AND S.-K. HONG

habitat patches and corridors in landscape mosaic are spatially connected with each other in matrix, it is necessary to mention these two spatial factors. In the ecological functions of the corridors of habitats such as rivers and roads, which serve as sources and sinks for species, the functions are major issues that should be considerably studied. 5. CONCLUSION The principle of “network system” is used, from the viewpoint of landscape ecology (Naveh, 1994; Jongman and Pungetti, 2004), as the strategy to obtain various types of patches where the diverse organisms live. It is also used to improve the quality of biodiversity and to reinforce the recent policies for conservation and management of wildlife such as the establishment of nature reserves area. The strategy, so called “ecological network”, “biotope network” or “habitat network”, is for the improvement of ecosystem quality in an entire region. It is based on the assumption that ecological corridors are effective for migration and distribution of animals (Hong et al., 2004). Through the networking system, the habitats existing in the region can be kept eco-functionally in the connected system. Such concepts and methods of ecological networks are being actively preceded in European countries starting from Germany. It has also been especially developed, being classified to land use planning and landscape ecological planning. y Core areas should be regions with high biodiversity and high naturalness that have the typical and representative habitats where the rare and endangered species live. Core areas should also be larger than the minimum-sized area required for the survival of organisms. In Europe, domestically and internationally, the required minimum size of the core area is 500 ha. y Ecological corridors should be areas with a good connection with core areas. They should function as temporary habitats and routines for migration and distribution. In choosing ecological corridors, the size of core areas for connection and the distance from adjacent habitats and existence of obstacles should be considered. y Nature development areas should have some naturalness and should function as buffer zone that prevents the habitats in core area and corridors from artificial influence. They are also the restoration places for nature to reinforce and expand the ecological networks. Ecological network concept is generally accepted as one of broad disciplines that cover diverse type of ecological corridors such as river continuum (blue network) and forest connectivity (green network) (Jongman and Pungetti, 2004, see Chapter 30). The continuity or connectivity among various sized ecosystems in aquatic and terrestrial systems is the matter of sustainability. In Korea, the blue-green network concept has been applied to the early stage of urban planning for the construction of Administrative Multifunctional City (73.14 km2) in middle part of Korea. As a part of environmental impact assessment (EIA) procedures, Preliminary Environmental Review (PER) was conducted, and PER statement described and analysed natural environment characteristics of several candidate locations.

LANDSCAPE ECOLOGY FOR BIODIVERSITY

159

Figure 1. Ecological network analysis and conservation strategy for Administrative Multifunctional City, which is planned to locate in southern part from Seoul, Korea (supported from Korea Environment Institute, Seoul).

An example of blue-green network analysis conducted in this task is shown in Figure 1A-F. Patch analysis has been applied to identify the priority of conservation area in land use planning step. Conservation priority of forest patch has provided major green axes in and around the candidate city’s boundary (Figure 1B). Fauna and flora survey carried out to obtain biological and ecological information. Based on the information, core area and corridors are identified and planned to be protected

160

T.H. RO AND S.-K. HONG

(Figure 1C and D). At same time, the lists of wildlife animals and plants that should be reserved in the area are also screened, and their conservation program is proposed. The conceptual blue-green network plan is produced by overlap the river continuum and green axes in the area (Figure 1E). Finally, detailed land use plan is shaping up depending on the functional schemes of the new city In the operated plans for conservation and restoration of biodiversity in many countries such as Netherlands and Germany (Jongman and Pungetti, 2004), there is a concept that inter-habitat networking are established in the regions and in a country even to the extent of the entire continent. In Europe, the plan, together with social economic methods for the establishment of networks and conservation of important habitats, proceeds at the spatially hierarchy level of European continent, countries, provinces, regions, territories and villages. As the cases of the developed countries are shown in the above, the biodiversity conservation needs the systemic measurement that is used not only for the populations’ conservation but also for the conservation of habitats at the level of region and landscape (Cox, 1993; Forman and Collinge, 1996; Washitani, 1999). It is also used for the common sharing management and conservation through international relations for organisms with a wide range of activities. Interdisciplinary theories and concepts of landscape management and planning for nature-human system are emerging issues (Zonneveld, 1989; Ro et al., 2000; Farina and Hong, 2004). The issues are effectively concerned with nature resource management and sustainable development in environmental policy. Finally, authors suggested that landscape ecology has to provide the baseline framework not only for ecological research and monitoring but also general protocol of environmental policy in a changing world. REFERENCES Choi, Y.-K. 2004. Linking planning system between spatial development plan and environment plan toward sustainable development. In S.-K. Hong, J.A. Lee, B.-S. Ihm, A. Farina, Y. Son, E.-S. Kim and J. C. Choe (Eds.), Ecological Issues in a Changing World: Status, Response and Strategy. Kluwer Academic Publishers, Dordrecht, The Netherlands. Cox, G.W. 1993. Conservation Ecology. Wm. C. Brown Publishers. 352p. Diamond, J.M. 1976. Island biogeography and conservation: Strategy and limitation. Science, 193, 10271029. Farina, A. 2000. Landscape Ecology in Action. Kluwer Academic Publishers. Dordrecht, The Netherlands. Farina, A. and Hong, S.-K. 2004. A theoretical framework for a science of landscape. In S.-K. Hong, J.A. Lee, B.-S. Ihm, A. Farina, Y. Son, E.-S. Kim and J.C. Choe (Eds.), Ecological Issues in a Changing World: Status, Response and Strategy. Kluwer Academic Publishers, Dordrecht, The Netherlands. Forman, R.T.T. 1995. Land Mosaics: The Ecology of Landscapes and Regions. Cambridge University Press. Cambridge. Forman, R.T.T. and Collinge, S.K. 1996. The spatial solution to conserving biodiversity in landscapes and regions. In R.M. DeGraaf and R.I. Miller (Eds.), Conservation of Faunal Diversity in Forested Landscapes (pp. 537-568). Chapman & Hall. Forman, R.T.T. and Godron, M. 1986. Landscape Ecology. John Wiley & Son, New York. Hong, S.-K. 1999. Cause and consequence of landscape fragmentation and changing disturbance by socio-economic development in mountain landscape system of South Korea. J. Environ. Sci., 11, 181-187.

LANDSCAPE ECOLOGY FOR BIODIVERSITY

161

Hong, S.-K. 2001. Factors affecting landscape changes in central Korea: Cultural disturbance on the forested landscape systems. In I.S. Zonneveld and D. van der Zee (Eds.), Landscape Ecology Applied in Land Evaluation, Development and Conservation (pp.131-147), ITC Publication No. 81, IALE publication MM-1, Enschede, The Netherlands. Hong, S.-K. 2002a. Necessity of landscape ecology in environment impact assessment of natural resource and action of IALE network. Nature Conservation, 120, 32-38. (in Korean) Hong, S.-K. 2002b. Man-influenced vegetation in Korea: Landscape ecology, management and planning. In D. Lee, J. Virginia, J.C. Choe, Y. Son, S. Yoo, H.-Y. Lee, S.-K. Hong and B.-S. Ihm (Eds.), Ecology of Korea (pp. 156-178). Bumwoo Publishing Company, Seoul. Hong, S.-K., Kim, S., Cho, K.-H., Kim, J.-E., Kang, S. and Lee, D. 2004. Ecotope mapping for landscape ecological assessment of habitat and ecosystem: A case study at man-influenced rugged landscape. Ecological Research, 19, 131-139. Hong, S.-K., Song, I.-J., Byun, B.-S., Yoo, S. and Nakagoshi, N. 2005. Applications of biotope mapping for spatial environmental planning and policy: Case studies in urban ecosystems in Korea. Landscape and Ecological Engineering, 1(2), 101-112. IUCN. 2000. 2000 IUCN Red List of Threatened Species. Gland, Switzerland and Cambridge, UK. 61pp. Jongman, R. and Pungetti, G. 2004. Ecological networks and greenways: concept, design, implementation. Cambridge University Press. 345p. MacArthur, J.W. and Wilson, E.O. 1967. The Theory of Island Biogeography. Princeton University Press, NJ. Nakagoshi, N., Watanabe, S. and Koga T. 2004. Landscape ecological approach for restoration site of natural forests in the Ota river basin, Japan. In S.-K. Hong, J.A. Lee, B.-S. Ihm, A. Farina, Y. Son, E.-S. Kim and J. C. Choe (Eds.), Ecological Issues in a Changing World: Status, Response and Strategy. Kluwer Academic Publishers, Dordrecht, The Netherlands. Naveh, Z. 1994. From biodiversity to ecodiversity: A landscape-ecology approach to conservation and restoration. Restoration Ecology, 2, 180-189. Riitters, K.H., O’Neill, R.V. and Jones, K.B. 1997. Assessing habitat suitability at multiple scale: A landscape-level approach. Biological Conservation, 81, 191-202. Ro, T.H., Hong, S.-K., Kang, D.-S. and Kwon, O.-S. 2000. Ecology: Nature and Man, Academy Press, Seoul. Spellerberg, I.F. 1996. Conservation Biology. Longman. England. 242pp. Szaro, R.C. and Johnston, D.W. 1996. Biodiversity in Managed Landscapes. Theory and Practice. Oxford University Press. 778pp. Turner, M.G. and Gardner, R.H. 1991. Quantitative Methods in Landscape Ecology: The Analysis and Interpretation of Landscape Heterogeneity. Springer-Verlag, New York. UNEP. 1995. Global Biodiversity Assessment. Cambridge University Press. 1140p. Washitani, I. 1999. Ecology of Biological Conservation. Kyoritsu Publisher, Tokyo, Japan (in Japanese) Whittaker, R.H. 1975. Community and Ecosystems. McMillan Publishing, New York. Zonneveld, I.S. 1989. The land unit - A fundamental concept in landscape ecology, and its applications. Landscape Ecology, 32, 67-86.

CHAPTER 11

A HIGHER-TAXON APPROACH WITH SOIL INVERTEBRATES TO ASSESSING HABITAT DIVERSITY IN EAST ASIAN RURAL LANDSCAPES S.-I. TANABE1, S.K. KHOLIN2, Y.-B. CHO3, S.-I. HIRAMATSU4, A. OHWAKI5, S. KOJI5, A. HIGUCHI6, S.Y. STOROZHENKO2, S. NISHIHARA7, K. ESAKI8, K. KIMURA1, K. NAKAMURA1 1

Institute of Nature and Environmental Technology, Kanazawa University, Kanazawa, 920-1192 Japan, 2Institute of Biology and Soil Science, Far Eastern Branch of the Russian Academy of Sciences, Vladivostok-22, 690022, Russia, 3 Natural History Museum, Hannam University, Daejeon, 306-791, Korea, 4Shiramine Elementary School, Hakusan, 920-2501 Japan, 5Faculty of Science, Kanazawa University, Kanazawa, 920-1192 Japan, 6Satoyama Nature School of Kakuma, Kanazawa University, Kanazawa, 920-1192 Japan, 7Graduate School of Agricultural and Life Sciences, Tokyo University, Tokyo, 113-8657 Japan, 8Ishikawa Forest Experiment Station, Hakusan, 920-2114 Japan

Abstract. Rural biodiversity in East Asia is at risk due to the loss of habitat diversity, and good indicators are needed to evaluate diverse habitats in rural landscapes. We examined whether the higher taxa (classes and orders) of soil invertebrates discriminated among several types of secondary forests such as broad-leaved deciduous forests, conifer forests and bamboo forests, primary forests, grasslands and/or wetlands, better than species assemblages of a well-established indicator, ground beetles (Coleoptera, Carabidae and/or Staphylinidae), in three East Asian regions (Japan, South Korea and the Russian Far East). We collected soil invertebrates with pitfall traps and used community composition and an ordination technique to test their performance as indicators. In Japan, the higher taxa of soil invertebrates discriminated finely among a wide range of habitats, and soil moisture seemed to be an important factor underlying habitat arrangement by these taxa along an ordination axis. While species assemblages of ground beetles detected large faunal differences among grasslands, wetlands and a composite group of three forest-type habitats (oak, conifer and bamboo forests), it failed to discriminate among any of the three forest-type habitats. When the analysis included only these types of forests, ground beetles were found to be able to discriminate finely among them, indicating limited performance in relation to the range of habitats

163 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 163–177. © 2007 Springer.

164

S.-I. TANABE ET AL.

covered. In the other two countries, the higher taxa of soil invertebrates showed a performance similar to that of species assemblages of ground beetles, possibly because of the narrow range of habitats analyzed. We conclude that the higher taxa of soil invertebrates are an effective tool for assessing the diversity of rural habitats across the East Asian region, where taxonomic knowledge at the species level is still insufficient. Our results may be applied broadly to other regions where agricultural intensification and land abandonment have caused quantitative and qualitative changes in rural landscapes.

1. INTRODUCTION Identification of present and future threats to biodiversity is an important first step in realizing effective conservation (Margules and Pressey, 2000). Human-modified landscapes in rural areas have received little attention for conservation planning, but have recently become a matter of great concern due to a widespread decline in biodiversity (Benton et al., 2003; Kato, 2001; Krebs, 1999; Pykälä, 2000; Washitani, 2001). Nevertheless, an overall picture of threats to rural biodiversity remains obscure in many regions of the world because of a very restricted understanding of the losses in rural biodiversity sometimes referred to as the ‘second Silent Spring’ (Krebs et al., 1999). Figuring among such regions is East Asia, where rural landscapes have suffered conspicuous changes due to rapid industrial and economic development (Hong, 1998; Nakagoshi and Hong, 2001). Undoubtedly, rural landscapes in East Asia are of high priority and biodiversity-oriented research is essential to understand and put the current status of rural landscapes and their biodiversity on the front of real planning process. A decline in rural biodiversity results from a loss of habitat diversity across various spatial scales through agricultural intensification with the attendant removal of non-cropped habitats (Benton et al., 2003), through the abandonment of traditional management, which causes qualitative changes in semi-natural habitats (Buckley, 1992; Hong, 1998; Nakagoshi and Hong, 2001; Washitani, 2001), or through the total loss of habitats due to changes in land use. In East Asia, the loss of habitat diversity due to the abandonment of traditional land use has emerged as a threat to biodiversity in only a few countries such as Japan and South Korea (Hong, 1998; Kato, 2001; Washitani, 2001). In Japan, where information on rural biodiversity is much more readily available than in other East Asian countries, a large number of species previously common to rural areas are now on the national red list and this situation characterizes the current crisis of biodiversity in Japan (Kato, 2001; Washitani, 2001). It is highly likely that this threat is also present in other countries where rapid agricultural modernization has strangled the traditional agriculture that has sustained agricultural life over centuries. Assessing habitat diversity is an integral part of any conservation effort (Hughes, 2000; O'Neil, 1995). In general, there are various kinds of human-modified habitats in rural landscapes, including ponds, wetlands, grasslands, fallow lands, plantations and woodlands as well as cultivated fields. Differences in the methods and histories of management of these habitats may enhance further habitat diversity in rural areas. A priori land classification based on types of vegetation or habitats is a useful tool for reserve selection but requires biological survey to examine the relationships between fauna and land classes before applying the classification to reserve selection (Pressey, 1994). As a result, conservation planning of rural landscapes needs good indicators of

INVERTEBRATES IN EAST ASIAN RURAL LANDSCAPES

165

habitat type across rather heterogeneous habitats. The search for such indicators across a wide range of habitats has been limited, even on a spatially restricted scale such as a rural landscape. Invertebrates are ubiquitous, taxon-rich and dominant organisms throughout the world (Wilson, 1987), and there has been a recent increase in awareness of their usefulness as indicators in conservation planning (Kremen et al., 1993; McGeoch, 1998). Soil invertebrates living in and on the ground have proved to be effective in assessing various kinds of human disturbances (Paoletti and Bressan, 1996). Identification at the species level represents a major obstacle to the use of soil invertebrates as indicators (Oliver and Beattie, 1996), however, higher-taxon indicators of soil invertebrates often show a performance similar to that of species-level indicators (Paoletti and Bressan, 1996) and thus can be potential surrogates for soil invertebrates in practical conservation. Furthermore, such a higher-taxon approach can greatly reduce the costs necessary for biodiversity surveys in terms of money, time and labor (Balmford, 1996). We examined the performance of soil invertebrates in higher taxonomic resolutions (classes and orders) as indicators of habitat diversity in rural areas across the northern part of East Asia, including Japan, South Korea and the Russian Far East. To evaluate the effectiveness of such higher-taxon indicators in comparison with species-level indicators, we selected ground beetle assemblages (Coleoptera, Carabidae and Staphylinidae) as a control group. Ground beetles have been widely used as indicators (Luff, 1996; Niemelä, 2001; Rainio and Niemelä, 2003) and provide rich information for the assessment of habitat diversity on various spatial scales (Niemelä et al., 1992; Luff et al., 1989, 1992; Rykken et al., 1997; Blake et al., 2003; Gutiérrez et al., 2003; Scott and Anderson, 2003). We used community composition and multivariate analysis as a measure and a method to test performance (Luff et al., 1989, 1992; O'Neil et al., 1995; Blake et al., 2003; Scott and Anderson, 2003). Recent studies validate the use of community composition in evaluating indicator performance (Howard et al., 1998; Oliver et al., 1998; Su et al., 2004). 2. METHODS 2.1 Study Areas and Habitat Types We defined rural areas as being situated between urban and mountainous areas and focused primarily on rural habitats consisting of secondary woodlands, grasslands and wetlands near human settlement (Takeuchi et al., 2003). We selected several habitat types typical of rural areas in each country under consideration (Table 1). In Japan, we established 50 study sites, which were distributed across 16 localities in Ishikawa Prefecture, Central Japan (Figure 1) and represented five types of rural habitats: oak forests, conifer forests, bamboo forests, grasslands and wetlands (Table 1). Conifer and bamboo forests were man-made plantations, which are usually not regarded as typical rural habitats but which were included in the present study because of their prevalence (conifer) and their rapidly increasing area (bamboo). There are few secondary grasslands in Ishikawa as in many other parts of Japan due to succession after abandonment (Takeuchi et al., 2003), resulting in our selection of mostly

166

S.-I. TANABE ET AL.

secondary grasslands, including two man-made grasslands, in public areas as study sites (Table 1).

Figure 1. Map of study localities in Japan, South Korea and the Russian Far East.

In South Korea, we established 21 sites in the suburbs of two cities 200 km apart from each other, Jeongeup (100 km southwest of Daejeon) and Yeongcheon (35 km east of Daegu) (Figure 1). We selected six habitat types: pine forests, oak forests, pseudoacacia forests, bamboo forests, shrubs and grasslands (Nakagoshi and Hong, 2001) (Table 1). In the Russian Far East, we established seven sites in the suburbs of Vladivostok (Figure 1), including primary forests, oak forests, mixed deciduous forests and grasslands; one mixed deciduous forest was situated in a city park (Table 1). In addition, we established one site in a primary forest near the Ussuriisk Nature Reserve to evaluate the effect of urbanization on primary forests. 2.2 Sampling We used pitfall traps to collect ground-active soil invertebrates. In Japan and South Korea the traps consisted of plastic bottles (500 ml, diameter 9 cm, depth 11 cm), partly filled with a 50% solution of ethylene glycol and covered with rims to prevent flooding due to rain. Each site contained one to four plots, with a distance between plots of 20-30 m, and five traps were placed linearly at 5-m intervals in each plot at each site. In total, there were 480 and 105 traps in Japan and South Korea, respectively. We collected invertebrate samples for two weeks from mid to late September 2003 in Japan and from 22 July to 5 August 2004 in South Korea. In the Russian Far East, we used plastic cups (volume 250 ml, diameter 7 cm, depth 9.5 cm) with water and a few drops of detergent as a collecting medium; there were no lids on the traps. In seven of eight study sites, we placed 15 traps linearly at 5-m intervals and carried out sampling for one day in mid June, early August, early September and mid October 2003. In the primary forest near the Ussuriisk Nature Reserve, we placed a set of five traps arranged in a 3 x 3 cross at 0.5-m intervals between traps as a

Table 1. Summary of habitat types in Japan, South Korea and the Russian Far East.

168

S.-I. TANABE ET AL.

secondary trapping unit and collected specimens for two days from 30 June to 1 July2003, for one week from 13 to 19 July 2003, and for eight days from 22 to 29 August 2003. We identified first centipedes (Chilopoda), millipedes (Diplopoda), land snails (Gastropoda), earthworms (Oligochaeta) and leeches (Hirudinoidea) at the class level and other invertebrates at the order level, as well as carabid beetles at the species level in all three countries. In the Russian samples from the Ussuriisk Nature Reserve, we identified all Coleopteran specimens at the species level (Storozhenko et al., 2003); in the Korean samples, we also identified 85% of rove beetles (Staphylinidae) at the species level and other rove beetles (15%) at the subfamily level (Aleocharinae, Pselaphinae, Scaphidiinae and Tachyporinae). For specimens in Tachyporinae, we were able to distinguish one species, Lordithon aitai, from the others, Tachyporinae spp. We treated these subfamilies and Tachyporinae spp. as a single species and, as in the case of the species data of rove beetles, combined them with the carabid dataset in subsequent analysis. We eliminated springtails (Collembola) and mites (Acarina) from the Japanese samples during sorting because of the considerable amount of time needed to sort a significant number of such very tiny specimens. 2.3 Data Analysis We used an ordination technique to examine the performance of two datasets, one based on the higher taxa of soil invertebrates (hereafter referred to as the invertebrate dataset) and the other on species of ground beetles (carabid dataset), with regard to the classification of study sites according to habitat types (Luff et al., 1989, 1992; Oliver et al., 1998; O'Neil et al., 1995; Scott and Anderson, 2003; Basset et al., 2004). We employed unconstrained ordination methods, such as principal components analysis (PCA) and detrended correspondence analysis (DCA), to analyze datasets from South Korea and the Russian Far East. For the Japanese datasets, we adopted constrained methods, such as redundancy analysis (RDA) and canonical correspondence analysis (CCA), with altitudes, longitudes, latitudes and habitat types as environmental variables, because these datasets consisted of samples collected across a wide range of localities (Figure 1) and altitudes (Table 1). In both the unconstrained and constrained methods, we finally selected either an ordination technique for species response to an underlying environmental gradient to be linear (PCA and RDA), or unimodal (DCA and CCA) based on simple criteria. After carrying out DCA on each dataset and checking the length of the largest gradient among the resultant ordination axes, we selected PCA and RDA if the largest gradient was shorter than 2.0, and DCA and CCA if it was larger than 4.0 (Jongman et al., 1995; Lepš and Šmilauer, 2003). Regardless of the length of the largest gradient, we chose DCA over PCA in those cases in which we visually confirmed an artifactual distortion of the ordination diagram due to the arch effect, in which the second axis was an arched function of the first axis (Jongman et al., 1995; Lepš and Šmilauer, 2003). If necessary, a partial RDA or CCA was carried out using one or several environmental variables as a covariable or covariables in order to help to interpret results obtained by the constrained method. Partial constrained methods enable us to examine effects of environmental variables of interest after partialling out the effect of covariables (Gutiérrez et al., 2003; Lepš

INVERTEBRATES IN EAST ASIAN RURAL LANDSCAPES

169

and Šmilauer, 2003). We carried out these analyses using CANOCO version 4.5 (ter Braak and Šmilauer, 2002). When necessary, we standardized abundance data for each site with the number of traps and days because the sampling effort varied among sites or because some trap samples were lost due to flooding, especially in wetlands, and wildlife disturbance. For the purposes of analysis, we pooled trap samples in each site, excluded study sites whose total number of individuals collected was < 10 and transformed abundance data to log10(x + 1). Where possible, we carried out non-parametric analysis of variance using the Kruskal-Wallis H-test or the Mann-Whitney U-test on site scores, derived from ordination analyses, along the first and second axis to evaluate the classification of study sites among habitat types or localities (Jongman et al., 1995). For constrained ordination in Japan, we analyzed the statistical significance of the ordination by Monte-Carlo randomization F-test with 499 permutations. 3. RESULTS In total, we collected 29865, 14142 and 9270 invertebrates, including 24, 21 and 21 higher taxa (classes and orders), and 1423, 612 (116 carabids and 496 rove beetles) and 733 ground beetles, comprising 43, 38 (21 for carabids and 17 for rove beetles) and 59 species, in Japan, South Korea and the Russian Far East, respectively. In the Japanese fauna, the most abundant higher taxa were Hymenoptera, Coleoptera and Isopoda, which comprised 68% of the invertebrates collected. Two carabid species, Carabus maiyasanus maiyasanus and Synuchus nitidus, dominated beetle samples at 47% of collected specimens. In South Korea, Hymenoptera, Collembola and Aranea comprised 77% of invertebrate specimens, and slightly over half (58%) of ground beetle specimens consisted of a single species of rove beetles, Oxytelus sp. In the Russian Far East, Collembola, Hymenoptera and Coleoptera comprised 67% of invertebrates collected and approximately half (49%) of carabid specimens consisted of four species, Agonum mandli, Carabus venustus, Pterostichus vladivostokensis and Nebria coreica. Ordination analysis showed that while the carabid dataset in Japan detected large faunal differences among grasslands, wetlands and a composite group of three forest-type habitats (oak, conifer and bamboo forests) (Figure 2A; F = 3.1, P = 0.002; H = 21.6, P = 0.0002 for the first axis; H = 9.9, P = 0.04 for the second axis), it failed to discriminate among any of the three forest-type habitats (Figure 2A; H = 4.0, P = 0.13 for the first axis; H = 2.3, P = 0.32 for the second axis). The first and second axes explained 21.3% and 9.7% of the variation in carabid faunal composition and 53.3% and 24.6% of the variance in the relationship between carabid species and the environment, respectively. After excluding grasslands and wetlands from the ordination analysis, however, this dataset effectively discriminated among the three forest habitats (Figure 2B; F = 1.51, P = 0.006; H = 10.5, P = 0.005 for the first axis; H = 13.7, P = 0.001 for the second axis), separating oak forests from conifer and bamboo forests. On the other hand, the invertebrate dataset successfully classified study sites among the forest habitats without exclusion of any habitats (Figure 2C; H = 15.6, P = 0.0004

170

S.-I. TANABE ET AL.

Figure 2. CCA ordination plots of all study sites (A) and sites in forest-type habitats (oak, conifer and bamboo forests: B) based on species assemblages of ground beetles and RDA (C) and partial RDA (D) ordination plots, the latter using altitude, latitude and longitude as covariables, of all study sites based on the higher taxa of soil invertebrates in Japan.

for the first axis; H = 9.7, P = 0.008 for the second axis), separating oak forests from conifer and bamboo forests, as well as among all habitats (Figure 2C; F = 3.14, P = 0.002; H = 31.4, P < 0.0001 for the first axis; H = 20.4, P = 0.0004 for the second axis). The first and second axes explained 23.3% and 6.3% of the variation in invertebrate faunal composition and 67.7% and 18.5% of the variance in the relationship between higher taxon and the environment, respectively. Two types of man-made plantation (conifer and bamboo, Table 1) showed quite similar invertebrate fauna. Soil moisture seemed to be an important factor underlying the arrangement of habitat types in the invertebrate ordination along the first axis, changing leftward from grasslands as a dry extreme to wetlands as a wet one (Figure 2C). Altitude had a large effect on invertebrate fauna, indicated by a long arrow in the ordination diagram, while the effect of geographical location (latitude and longitude) was small (Figure 2C). The partial RDA, using altitude as an environmental variable and the remaining ones as covariables, revealed that the effect of altitude was significant (F = 2.92, P = 0.02), although altitude explained only 5.7% of the variation in invertebrate faunal composition. In addition, the pattern of habitat classification based on the invertebrate dataset were nearly the same even after partialling out the effects of altitude, latitude and longitude in the ordination analysis (Figure 2D), indicating the significant effect of habitat types on invertebrate fauna (F = 3.07, P = 0.002).

INVERTEBRATES IN EAST ASIAN RURAL LANDSCAPES

171

For South Korea, the carabid dataset detected faunal differences between two localities and among habitat types in one locality, Yeongcheon, along the first axis (Figure 3A; U = 14.0, P = 0.018 for the former; H = 7.58, P = 0.023 for the latter). Similarly, the invertebrate dataset discriminated not only between two localities along the first axis (Figure 3B; U = 2.0, P = 0.0004) but also among habitat types in Yeongcheon along the second axis (Figure 3B; H = 8.14, P = 0.017, after excluding one grassland site from analysis). However, the differences among habitat types in Yeongcheon were larger and clearer in the ordination plot of the invertebrate dataset than in that of the carabid dataset. Ordination results in the Russian Far East showed a similar performance between the carabid and invertebrate datasets, both of which classified sites among grasslands, mixed deciduous forests and the others with oak and primary forests along the first axis (Figure 4A and B). We performed no statistical tests on the Russian ordination results because of the small sample sizes in most types of habitats.

Figure 3. DCA ordination plots of study sites based on species assemblages of ground beetles (A) and the higher taxa of soil invertebrates (B) in South Korea. Capital letters indicate habitat types: pine forests (P), pseudoacacia forests (A), oak forests (OF), bamboo forests (B), shrub (S) and grasslands (G).

Figure 4. DCA and PCA ordination plots of study sites based on species assemblages of ground beetles (A) and the higher taxa of soil invertebrates (B), respectively, in the Russian Far East.

172

S.-I. TANABE ET AL.

4. DISCUSSION 4.1 Indicator Performance As a whole, the present study showed better performance in the higher taxa of soil invertebrates than in species assemblages of ground beetles as an indicator of diverse rural habitats. In Japan, while ground beetles identified major differences among grasslands, wetlands and forests, they failed to discriminate among several types of forest habitats and classified all of such types into a single group. On the other hand, the higher taxa of soil invertebrates effectively classified study sites among not only the forest-type habitats but also the other habitat types. After excluding the major differences among grasslands, wetlands and forests, ground beetles were found to be able to successfully discriminate among the forest-type habitats. These results clearly indicate the limited performance of ground beetles in relation to the range of habitat types: ground beetles can discriminate finely among similar habitats or within a limited range of habitats but only roughly among heterogeneous habitats or within a wide range of habitats. There is considerable evidence supporting fine resolution in the classification of similar habitats by carabid fauna, for example, in grasslands (Rushton et al., 1991; Luff et al., 1992; Asteraki et al., 1995; Luff, 1996; Dennis et al., 1997; French and Elliott, 1999), woodlands (Niemelä et al., 1988; Niemelä et al., 1992; Baguette, 1993; Coll et al., 1995; Niemelä et al., 1996; Humphrey et al., 1999; Jukes et al., 2001; Koivula et al., 2002; Similä et al., 2002), heathland (Gardner, 1991), moorland (Holmes et al., 1993; McCracken, 1994; Sanderson et al., 1995) and a limited range of habitats (Thiele, 1977; Bedford and Usher, 1994; Butterfield et al., 1995; Niemelä et al., 1996; Fahy and Gormally, 1998; Ings and Hartley, 1999; Fournier and Loreau, 2001; du Bus de warnaffe and Lebrun, 2004;). In contrast, few studies have examined a wide range of habitats; nevertheless, there is some evidence for the rough distinction of heterogeneous habitats by carabid fauna (Luff et al., 1989; Turin et al., 1991; Blake et al., 2003; Scott and Anderson, 2003). In South Korea, both ground beetles and invertebrate higher taxa differed between the two localities, while no such local differentiation was detected in either fauna of Japan. The number of localities studied was much larger (15 in Japan and two in Korea) and the arrangement of localities was geographically more continuous, with a shorter range between the two most distant sites in Japan (170 km) than in Korea (200 km). The discrepancy in the results between the two countries may be attributable primarily to these differences in study design. In addition, the composition of the studied habitats differed greatly between the two localities in Korea and this may also have contributed to faunal variation between the two localities. For habitat classification in Korea, the higher taxa of soil invertebrates were found to show a performance similar to that of ground beetles in discriminating some types of forest habitats in Yeongcheon region. This is consistent with the present results for classification of forest-type habitats in Japan, which show fine classification among similar habitat types by carabid fauna and similar performance in discriminating among the habitat types between the invertebrate and the carabid datasets. The Korean ground beetles consisted largely of rove beetles, which represented 81% of the total number of individuals collected; in addition to carabid beetles, rove beetles

INVERTEBRATES IN EAST ASIAN RURAL LANDSCAPES

173

can also be a potential indicator of habitat type (Bohac, 1999). We have no explanation for the scarcity of carabids in the Korean samples. However, if this represents a decline in their population in rural areas, it indicates high conservation value for carabids in Korea. Many of soil invertebrates have proven to be closely associated with moist habitats such as damp soil, mud and decomposing organic matter and be sensitive to changes in moisture and relative humidity of their habitats (e.g. Coleman et al., 2004; Lensing et al., 2005). In the present study, the ordination analysis revealed that faunal composition of invertebrate higher taxa collected in Japan gradually changed from grasslands through forests to wetlands along the first axis. Although we measured no abiotic factors in this study, soil moisture may be an important factor that underlies the arrangement of habitat types based on the invertebrate dataset. Ground beetles are also known to respond well to soil moisture in terms of species abundance, diversity and composition (Baguette 1993; Asteraki et al., 1995; Sanderson et al., 1995; Koivula et al., 1999; Jukes et al., 2001). However, no such response of ground beetles to soil moisture was detectable in the ordination plot based on the carabid dataset, implying the relatively minor effect of soil moisture on species distribution of ground beetles across highly heterogeneous habitats in rural landscapes. 4.2 Implications for Conservation Invertebrates are numerous everywhere and perform various ecological functions and essential roles in all ecosystems on earth. Species identification poses a crucial limitation for using invertebrates as indicators in conservation planning. For assessment of habitat diversity, however, the present study clearly shows that the higher taxa of soil invertebrates can finely discriminate among diverse types of rural habitats, even based on samples from a relatively short-term survey. This result highlights the importance of invertebrate higher taxa in assessing the habitat diversity of rural areas across the East Asian region, where taxonomic knowledge of soil invertebrates at the species level is still insufficient and abandonment of traditional management has caused qualitative changes in habitats. Our results may be applied broadly to other regions under similar conditions of land use, for example Europe (Buckley, 1992; Pykälä, 2000). Ground beetles may be less useful than the higher taxa of soil invertebrates in classifying rural habitats, if such a wide range of habitat types as that covered in this study is taken into account. Rather, we suggest using the higher taxa of soil invertebrates as surrogate indicators for assessing the conservation value of various habitats. Habitat classification based on the present carabid datasets was rough but consistent with that based on invertebrate datasets. Such a higher-taxon approach can effectively save money, time and labor (Balmford et al., 1996) and should be one of options in designing biological surveys for conservation planning, especially in regions where available resources are severely limited. We need additional studies on other taxa and types of rural habitats, especially on habitats under management, to refine our results. Temporal variation in structure of invertebrate communities may occur through seasonal changes in both abiotic and

174

S.-I. TANABE ET AL.

biotic environments and thus seasonal replication can also improve the reliability and consistency of the results over time. ACKNOWLEDGMENTS We thank K. Aomori, T. Kishimoto, D. Utsunomiya and S. Iwanishi for helping us during field survey in Japan and S. Takaba, G. Sh. Lafer and J.K. Park for identification of several carabid species in Japan, Russia and South Korea, respectively. We are also grateful to park stuff in Kenko-no-Mori, Shinrin Kouen, Yuhidera Kenmin Sizen-en, Tatsunokuchi Kyuryo Kouen, Kenmin-no-Mori and Kanazawa Castle Park for their permission for sampling, and H. Takahashi, O. Sano, M. Eguchi, K. Ishihara and E. Kinoshita for information for study sites in Japan. This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology as a part of the international project of the Kanazawa University COE program “Long- and Short-Term Dynamics of Pan-Japan Sea Area: Environmental Monitoring and Prediction” and Far Eastern Branch of Russian Academy of Sciences (N04-1-OBH-100). REFERENCES Asteraki, E.J., Hanks, C.B., and Clements, R.O. (1995). The influence of different types of grassland field margin on carabid beetle (Coleoptera, Carabidae) communities. Agriculture, Ecosystems and Environment, 54, 195-202. Baguette, M. (1993). Habitat selection of carabid beetle in deciduous woodlands of southern Belgium. Pedobiologia, 37, 365-378. Balmford, A., Green, M.J.B., and Murray, M.G. (1996). Using higher-taxon richness as a surrogate for species richness: I. regional tests. Proceedings of the Royal Society of London, Series B, 263, 1267-1274. Balmford, A., Jayasuriya, A.H.M., and Green, M.J.B. (1996). Using higher-taxon richness as a surrogate for species richness: II. Local applications. Proceedings of the Royal Society of London, Series B, 263, 1571-1575. Basset, Y., Mavoungou, J.F., Mikissa, J.B., Missa, O., Miller, S.E. and Kitching, R.L. (2004). Discriminatory power of different arthropod data sets for the biological monitoring of anthropogenic disturbance in tropical forests. Biodiversity and Conservation, 13, 709-732. Bedford, S.E. and Usher, M.B. (1994). Distribution of arthropod species across the margins of farm woodlands. Agriculture, Ecosystems and Environment, 48, 295-305. Benton, T.G., Vickery, J.A. and Wilson, J.D. (2003). Farmland biodiversity: is habitat heterogeneity the key? Trends in Ecology and Evolution, 18, 182-188. Blake, S., McCracken, D.I., Eyre, M.D., Garside, A. and Foster, G.N. (2003). The relationship between the classification of Scottish ground beetle assemblages (Coleoptera, Carabidae) and the National Vegetation Classification of British plant communities. Ecography, 26, 602-616. Bohac, J. (1999). Staphylinid beetles as bioindicators. Agriculture, Ecosystems and Environment, 74, 357-372. Buckley, G.P. (1992). Ecology and Management of Coppice Woodlands. Chapman and Hall, London. Butterfield, J., Luff, M.L., Baines, M., and Eyre, M.D. (1995). Carabid beetle communities as indicators of conservation potential in upland forests. Forest Ecology and Management, 79, 63-77. Coleman, D.C., Crossley, Jr. D.A., and Hendrix, P.F. (2004). Fundamental of Soil Ecology. Elsevier Inc., San Diego. Coll, M.T., Heneghan, L., and Bolger, T. (1995). Carabidae fauna in two Irish conifer stands: a comparison with those of some other European forests. Biology and Environment: Proceeding of the Royal Irish Academy, 95B, 171-177.

INVERTEBRATES IN EAST ASIAN RURAL LANDSCAPES

175

Dennis, P., Young, M.R., Howard, C.L., and Gordon, I.J. (1997). The response of epigeal beetles (Col.: Carabidae, Staphylinidae) to varied grazing regimes on upland Nardus stricta grasslands. Journal of Applied Ecology, 34, 433-443. du Bus de warnaffe, G., and Lebrun, P. (2004). Effects of forest management on carabid beetles in Belgium: implications for biodiversity conservation. Biological Conservation, 118, 219-234. Fahy, O. and Gormally, M. (1998). A comparison of plant and carabid beetle communities in an Irish oak woodland with a nearby conifer plantation and clearfelled site. Forest Ecology and Management, 110, 263-273. Fournier, E. and Loreau, M. (2001). Respective roles of recent hedges and forest patch remnants in the maintenance of ground-beetle (Coleoptera: Carabidae) diversity in an agricultural landscape. Landscape Ecology, 16, 17-32. French, B.W. and Elliott, N.C. (1999). Temporal and spatial distribution of ground beetle (Coleoptera: Carabidae) assemblages in grasslands and adjacent wheat fields. Pedobiologia, 43, 73-84. Gardner, S.M. (1991). Ground beetle (Coleoptera: Carabidae) communities on upland heath and their association with heathland flora. Journal of Biogeography, 18, 281-289. Gutiérrez, D., Menéndez, R. and Méndez, M. (2003). Habitat-based conservation priorities for carabid beetles within the Picos de Europa National Park, northern Spain. Biological Conservation, 115, 379-393. Holmes, P.R., Fowles, A.P., Boyce, D.C. and Reed, D.K. (1993). The ground beetle (Coleoptera: Carabidae) fauna of Welsh peatland biotopes-species seemblages in relation to peatland habitats and management. Biological Conservation, 65, 61-67. Hong, S.K. (1998). Changes in landscape patterns and vegetation process in the Far-Eastern cultural landscapes: human activity on pine-dominated secondary vegetation in Korea and Japan. Phytocoenologia, 28, 45-66. Howard, P.C., Viskanic, P., Davenport, T.R.B., Kigenyi, F.W., Baltzer, M., Dickinson, C.J. (1998). Complementarity and the use of indicator groups for reserve selection in Uganda. Nature, 394, 472-475. Hughes, J.B., Daily, G.C. and Ehrlich, P.R. (2000). Conservation of insect diversity: a habitat approach. Conservation Biology, 14, 1788-1797. Humphrey, J.W., Hawes, C., Peace, A.J., Ferris-Kaan, R. and Jukes, M.R. (1999). Relationships between insect diversity and habitat characteristics in plantation forests. Forest Ecology and Management, 113, 11-21. Ings, T.C. and Hartley, S.E. (1999). The effect of habitat structure on carabid communities during the regeneration of a native Scottish forest. Forest Ecology and Management, 119, 123-136. Jongman, R.H.G., ter Braak, C.J.F., and van Tongeren, O.F.R. (1995). Data Analysis in Community and Landscape Ecology. Cambridge University Press, Cambridge. Jukes, M.R., Peace, A.J., and Ferris, R. (2001). Carabid beetle communities associated with coniferous plantations in Britain: the influence of site, ground vegetation and stand structure. Forest Ecology and Management, 148, 271-286. Kato, M. (2001). 'SATOYAMA' and biodiversity conservation: 'SATOYAMA' as important insect habitats. Global Environmental Research, 5, 135-149. Koivula, M., Kukkonen, J. and Niemelä, J. (2002). Boreal carabid-beetle (Coleoptera: Carabidae) assemblages along the clear-cut originated succession gradient. Biodiversity and Conservation, 11, 1269-1288. Koivula, M., Punttila, P., Haila, Y. and Niemelä, J. (1999). Leaf litter and the small-scale distribution of carabid beetles (Coleoptera: Carabidae) in the boreal forest. Ecography, 22, 424-435. Krebs, J.R., Wilson, J.D., Bradbury, R.B., and Siriwardena, G.M. (1999). The second silent spring? Nature, 400, 611-612. Kremen, C., Colwell, R.K., Erwin, T.L., Murphy, D.D., Noss, R.F. and Sanjayan, M.A. (1993). Terrestrial arthropod assemblages: their use in conservation planning. Conservation Biology, 7, 796-808. Lensing, J.R., Todd, S., and Wise, D.H. (2005). The impact of altered precipitation on spatial stratification and activity-densities of springtails (Collembola) and spiders (Araneae). Ecological Entomology, 30, 194-200. Lepš, J. and Šmilauer, P. (2003). Multivariate Analysis of Ecological Data for CANOCO. Cambridge University Press, Cambridge. Luff, M.L. (1996). Use of carabids as environmental indicators in grasslands and cereals. Annales Zoologici Fennici, 33, 185-195.

176

S.-I. TANABE ET AL.

Luff, M.L., Eyre, M.D. and Rushton, S.P. (1989). Classification and ordination of habitats of ground beetles (Coleoptera, Carabidae) in north-east England. Journal of Biogeography, 16, 121-130. Luff, M.L., Eyre, M.D. and Rushton, S.P. (1992). Classification and prediction of grassland habitats using ground beetles (Coleoptera, Carabidae). Journal of Environmental Management, 35, 301-315. Margules, C.R., and Pressey, R.L. (2000). Systematic conservation planning. Nature, 405, 243-253. McCracken, D.I. (1994). A fuzzy classification of moorland ground beetle (Coleoptera: Carabidae) and plant communities. Pedobiologia, 38, 12-27. McGeoch, M. (1998). The selection, testing and application of terrestrial insects as bioindicators. Biological Reviews, 73, 181-201. Nakagoshi, N. and Hong, S.-K. (2001). Vegetation and landscape ecology of East Asian 'SATOYAMA'. Global Environmental Research, 5, 171-181. Niemelä, J. (2001). Carabid beetles (Coleoptera, Carabidae) and habitat fragmentation: a review. European Journal of Entomology, 98, 127-132. Niemelä, J., Haila, Y., Halme, E., Lahti, T., Pajunen, T. and Punttila, P. (1988). The distribution of carabid beetles in fragments of old coniferous taiga and adjacent managed forests. Annales Zoologici Fennici, 25, 107-119. Niemelä, J., Haila, Y., Halme, E., Pajunen, T. and Punttila, P. (1992). Small-scale heterogeneity in the spatial distribution of carabid beetles in the southern Finnish taiga. Journal of Biogeography, 19, 173-181. Niemelä, J., Haila, Y. and Punttila, P. (1996). The importance of small-scale heterogeneity in boreal forests: variation in diversity in forest-floor invertebrates across the succession gradient. Ecography, 19, 352-368. Oliver, I. and Beattie, A.J. (1996). Invertebrate morphospecies as surrogates for species: a case study. Conservation Biology, 10, 99-109. Oliver, I., Beattie, A.J. and York, A. (1998). Spatial fidelity of plant, vertebrate, and invertebrate assemblages in multiple-use forest in eastern Australia. Conservation Biology, 12, 822-835. O'Neil, T.A., Steidl, R.J., Edge, W.D. and Csuti, B. (1995). Using wildlife communities to improve vegetation classification for conserving biodiversity. Conservation Biology, 9, 1482-1491. Paoletti, M.G. and Bressan, M. (1996). Soil invertebrates as bioindicators of human disturbance. Critical Reviews in Plant Sciences, 15, 21-62. Pressey, R.L. (1994). Land classifications are necessary for conservation planning but what do they tell us about the fauna? In D. Lunney, S. Hand, P. Reed and D. Butcher (Eds.), Future of the Fauna of Western New South Wales (Vol. 246, pp. 31-41). Royal Zoological Society of New South Wales, N.S.W., Australia. Pykälä, J. (2000). Mitigating human effects on European biodiversity through traditional animal husbandry. Conservation Biology, 14, 705-712. Rainio, J. and Niemelä J. (2003). Ground beetles (Coleoptera: Carabidae) as bioindicators. Biodiversity and Conservation, 12, 487-506. Rushton, S.P., Luff, M.L. and Eyre, M.D. (1991). Habitat characteristics of grassland Pterostichus species (Coleoptera, Carabidae). Ecological Entomology, 16, 91-104. Rykken, J.J., Capen, D.E. and Mahabir, S.P. (1997). Ground beetles as indicators of land type diversity in the green mountains of Vermont. Conservation Biology, 11, 522-530. Sanderson, R.A., Rushton, S.P., Cherrill, A.J. and Byrne, J.P. (1995). Soil, vegetation and space: an analysis of their effects on the invertebrate communities of a moorland in north-east England. Journal of Applied Ecology, 32, 506-518. Scott, W. A. and Anderson, R. (2003). Temporal and spatial variation in carabid assemblages from the United Kingdom Environmental Change Network. Biological Conservation, 110, 197-210. Similä, M., Kouki, J., M?nkk?nen, M. and Sippola, A.-L. (2002). Beetle species richness along the forest productivity gradient in northern Finland. Ecography, 25, 42-52. Storozhenko, S.Y., Sidorenko, V.S., Lafer, G.S. and Kholin, S.K. (2003). The International Biodiversity Observation Year (IBOY): insects of the forest ecosystems of the Primorye region. A. I. Kurentsov's Annual Memorial Meetings, 13, 31-52. (In Russian). Su, J.C., Debinski, D.M., Jakubauskas, M.E. and Kindscher, K. (2004). Beyond species richness: community similarity as a measure of cross-taxon congruence for coarse-filter conservation. Conservation Biology, 18, 167-173. Takeuchi, K., Brown, R.D., Washitani, I., Tsunekawa, A. and Yokohari, M. (2003). Satoyama: The Traditional Rural Landscape of Japan. Springer-Verlag, Tokyo.

INVERTEBRATES IN EAST ASIAN RURAL LANDSCAPES

177

ter Braak, C.J.F., and Šmilauer, P. (2002). CANOCO Reference Manual and CanoDraw for Windows User's Guide: Software for Canonical Community Ordination (version 4.5). Microcomputer Power, New York. Thiele, H.U. (1977). Carabid Beetles in Their Environments: A Study on Habitat Selection by Adaptations in Physiology and Behavior. Springer-Verlag, Berlin. Turin, H., Alders, P.J., den Boer, P.J., van Essen, S., Hejierman, T., Laane, W. (1991). Ecological characterization of carabid species (Coleoptera, Carabidae) in the Nehterlands from thirty years of pitfall sampling. Tijdschrift voor Entomologie, 134, 279-304. Washitani, I. (2001). Traditional sustainable ecosystem 'SATOYAMA' and biodiversity crisis in Japan: conservation ecological perspective. Global Environmental Research, 5, 119-133. Wilson, E.O. (1987). The little things that run the world (the importance and conservation of invertebrates). Conservation Biology, 1, 344-346.

CHAPTER 12

LANDSCAPE ECOLOGICAL APPROACH IN OIL PALM LAND USE PLANNING AND MANAGEMENT FOR FOREST CONSERVATION IN MALAYSIA

S.A. ABDULLAH, N. NAKAGOSHI Institute for Environment and Development (LESTARI), Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor Darul, Ehsan, Malaysia; Graduate School for International Development and Cooperation, Hiroshima University, 1-5-1 Kagamiyama, Higashi-Hiroshima, 739-8529 Japan

Abstract. Oil palm industry is one of the major revenue for economic development in Malaysia. Oil palm was introduced to Malaysia in 1911 and the first commercial planting began in 1917. Since then and combined with increasing demand of its product both at the local and international market, areas planted with oil palm has been expanded throughout the nation. In 2004, oil palm plantation covered approximately 9.9% of the total Malaysian land area. However, its expansion has been recognised intrude into forested areas. This situation has been debated as one of the major international issues that cause forest degradation and loss of biodiversity. To balance between economic importance and conservation of forested areas a practical approach in planning and managing oil palm land use is necessary. Landscape ecological approach is an emerging pathway to achieve the goal and increasingly recognised as tool in various aspects of planning and managing land use. In this chapter, the land use changes of this premier man-influenced area in Malaysia will be analysed. How landscape ecology concepts can be integrated into planning and managing oil palm land use for forest conservation in Malaysia will be discussed.

1. INTRODUCTION The oil palm (Elaeis guineensis Jacq.) is a palm tree originated from the equatorial belt of West Africa (Hartley, 1988). Today oil palm is planted as a commercial plantation in many countries of tropical region. Historically, the introduction of oil palm as a commercial plantation began in South East Asia and the first establishment was in Sumatra, Indonesia (Hartley, 1988). Since then, land 179 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 179–191. © 2007 Springer.

180

S.A. ABDULLAH AND N. NAKAGOSHI

planted with oil palm increased tremendously not only in Indonesia but also in other tropical countries particularly Malaysia. Until today, the consumption of oil palm products receiving highly demand from many countries around the world. The world consumption of palm oil, for example, was ranked second behind soy oil in 2002 (Oil World, 2002 quoted in Casson, 2003). The major consumer is mainly from the countries of the Asian region such as India (14%) and Indonesia (11%) (Oil World, 2002 quoted in Casson, 2003). Malaysia is one the major producers and exporters of oil palm in the world. From the period between 1995 and 2004, its total palm oil production and export value increased by about 79% and 93%, respectively. Although oil palm production plays an important role in economic development this human land use activity has been recognised to cause degradation of forested areas (e.g, Okuda et al., 2004). Historically, large hectares of forest were converted into oil palm, which predominantly occurred when Malaysian development policy favoured agriculture, that is, in the 1950s to the 1970s (Kumar, 1986). During that period several land schemes were introduced for development of oil palm plantation, which involved a vast clearance of forested areas (Goh, 1982).

Figure 1a. Total palm oil production (tones) of five major countries in the world. Source: Oil World 2020, stated in Malaysian Oil Palm Statistics, Malaysian Palm Oil Board.

The continuing process of this activity until recent years particularly in Sabah and Sarawak has led to diverse environmental problems. River pollution (Abdullah, 1995) and sedimentation (Yusuf and Nordin, 2003) as well as deterioration of aquatic habitat (Azrina et al., in press) are among the issues related to the activity. In addition, many forested areas have been left fragmented (Abdullah, 2003) and wildlife has loss their natural habitat (Zubaid, 1993; Norhayati et al., 2004). Due to this, animals began to encroach to human settlements, resulting in serious conflicts between humanity and wildlife, not only in rural areas but also in developed urban areas (Jasmi, 1997).

OIL PALM LAND USE PLANNING AND MANAGEMENT

181

Figure 1b. Total oil palm export value (tones) of five major countries in the world. Source: Oil World 2020, stated in Malaysian Oil Palm Statistics, Malaysian Palm Oil Board.

Therefore, to ensure the equilibrium between economic development and conservation of forested areas, there is the need for a practical approach in the way the oil palm land use is managed. From the perspective of environmental science, the approach must be towards a sustainable management of the land use. The concept of landscape ecology provides a new insight and opportunity to achieve the goal. Thus, the objectives of this chapter are to analyse the trends and changes of oil palm land use in Malaysia and to discuss how landscape ecology can be integrated into planning and managing oil palm land use for forest conservation in 1 Malaysia . 2. TRENDS AND CHANGES OF OIL PALM LAND USE Oil palm is considered as a major agricultural land use in Malaysia. Due to the high demand of its product both at the local and international market, oil palm plantation areas increased by about 503% from only 641,791 ha in 1975 to 3.9 million ha in 2004 (Figure 2a). The figure of 2004 represents approximately 9.9% of the total Malaysian land area (Malaysian Oil Palm Board, 2004). Of this, 2.2 million ha or 5.6% located in Peninsular Malaysia whereas the remaining were in Sarawak (1.3%) and Sabah (3.0%). In Peninsular Malaysia, the trend of oil palm expansion between 1975 and 2004 was similar as that of the whole Malaysia (Figure 2b). However, a different pattern was shown by Sabah and Sarawak. In these two states, from the 1970s to the 1980s the total area of oil palm increased slowly (Figure 2c and 2d). However, in the 1990s it began to show tremendous expansion but it was more 1

Malaysia has two parts: Peninsular Malaysia and Malaysia Borneo that contains the states of Sabah and Sarawak.

182

S.A. ABDULLAH AND N. NAKAGOSHI

in Sarawak than in Sabah. By 2004, about 17% of the total land area of Peninsular Malaysia covered by oil palm. In the same year, oil palm represents approximately 16% and 4% of the total land area of Sabah and Sarawak, respectively. Generally, in Malaysia the highest rate of oil palm expansion occurred during the decade of the 1990s (128,392.9 ha/year) compared to the 1970s (59,414.4 ha/year) and the 1980s (92,325.3 ha/year) (Figure 3a). However, the highest expansion rate between Peninsular Malaysia and the two states in Borneo occurred in different period. In Peninsular Malaysia, the highest rate of expansion occurred during the 1980s (73,771.0 ha/year) (Figure 3b) whereas in Sabah (66,515.1 ha/year) (Figure 3c) and Sarawak (26,568.1 ha/year) (Figure 3d) it happened during the 1990s. The analysis between 2000 and 2004 revealed that the rate of expansion for the whole Malaysia was generally less than the 1990s but almost similar as in the 1980s (Figure 3a). When compared separately among the three Malaysian regions, the highest expansion rate occurred in Sarawak (35,584.4 ha/year) (Figure 3c), followed by Sabah (32,927.0 ha/year) (Figure 3d) and Peninsular Malaysia (31,221.2 ha/year) (Figure 3b). Except for Sarawak, the rate in Peninsular Malaysia and Sabah was low than the 1990s. From this analysis, it can be suggested that the early phases of oil palm development was predominantly occurred in Peninsular Malaysia but starting from the 1990s, the intensity of this activity has been shifted to Sabah and Sarawak.

Thousands

Total area (ha)

4500 4000 3500 3000 2500 2000 1500 1000 500 0

2500 2000 1500 1000 500

19 96

19 99

20 02

19 99

20 02

19 93

19 96

Year

(d) Sarawak

600 Total area (ha)

500 400 300 200 100

600 500 400 300 200 100

Year

19 93

19 90

19 87

19 84

19 78

19 81

19 75

0

19 75 19 78 19 81 19 84 19 87 19 90 19 93 19 96 19 99 20 02

0

Thousands

(c) Sabah Thousands

19 90

19 87

19 84

19 81

19 78

19 75

20 02

19 99

19 96

19 93

19 90

19 87

19 84

19 81

0 19 78

19 75

Thousands

(b) Peninsular Malaysia

Year

Total area (ha)

Total area (ha)

(a) Malaysia

Year

Figure 2a-d. Trends of oil palm area (ha) from 1975 to 2004 Source: Malaysian Oil Palm Statistics (Malaysian Palm Oil Board).

OIL PALM LAND USE PLANNING AND MANAGEMENT

Figure 3a-d. Expansion rate (hectare/year) of oil palm in 4 decades Source: Malaysian Oil Palm Statistics (Malaysian Palm Oil Board).

183

184

S.A. ABDULLAH AND N. NAKAGOSHI

3. LANDSCAPE ECOLOGY IN OIL PALM LAND USE PLANNING AND MANAGEMENT 3.1 The perspective The future of most landscapes is increasingly being determined by human activities either deliberately or inadvertently, and these modify the existing landscape patterns and processes (Hobbs, 1997). It is becoming increasingly apparent that an understanding of these landscape level patterns and processes is essential for rational land use planning and management for natural forest conservation. Landscape ecology provides a unique opportunity, and indeed an obligation, to provide concepts and techniques to tackle or address environmental issues related to forest degradation due human activities, for example large-scale agriculture. One of the fundamental aspects of landscape ecology is its explicitly consideration of the spatial dimension of ecological processes, thus providing a common language for stronger interactions between ecologist and planners (Leitão and Ahern, 2002). Thus, it is not surprising that it has been attempted to incorporate landscape ecology approach into land use planning and management (e.g Lenz and Stary, 1995; Palmer and Lankhorst, 1998; Gulink and Wagendorp, 2002). 3.2 Landscape spatial pattern of oil palm land use 3.2.1 Spatial relationship with forestland Landscape ecology approach basically involved the measurement of spatial pattern of land use at a landscape scale. One of the important aspects is the spatial relationship between patches of various types of land use that composed the landscape. In Malaysia, oil palm is important component of land uses that modify and determine the landscape pattern of the country. Because this area was originally covered by natural forest, thus understanding of their spatial relationship with forestland is vital. This is to provide an insight about the pattern of oil palm expansion towards forested areas, which is necessary for sustainable planning and managing oil palm land use. We suggest that the spatial relationship between the two land uses can be quantified in three ways. The first assessment is the ratio between oil palm and forest areas. This analysis, however, can be measured at two levels, that is, i) forest institutional status and ii) forest type. Nevertheless, these are applied only for Peninsular Malaysia and not for Sabah and Sarawak due to their different management system. In the first level, forest can be classified into three; permanent forest reserve (PFR), state land forest and wildlife reserve. Permanent forest reserve is gazetted forest under the National Forestry Act 1984 (Amended, 1993). This forest is for timber and non-timber production and environmental

OIL PALM LAND USE PLANNING AND MANAGEMENT

185

protection, managed by the Department of Forestry. State land forest is forest owned by state governments and wildlife reserve is gazetted forest for protection and conservation of wildlife. The wildlife reserve is under the management of the Department of Wildlife and National Parks. In the second level, forest can be classified into three main types: dipterocarp, peat swamp and mangrove forests. However, this forest type classification is only made for permanent forest reserve. For the first classification, the analysis between 1991 and 2000 revealed that the ratio between oil palm and permanent forest reserve declined steadily (Figure 4a). A similar pattern was shown between oil palm and state land forest (Figure 4b). However, over the years the ratio between the agricultural land and wildlife reserve did not show many changes (Figure 4c). These results suggest that the total area of oil palm was gradually exceeded the total area of permanent forest reserve and state land forest. In other words, permanent forest reserve and state land forest was experienced loss of its area. The constant pattern of the ratio between oil palm and wildlife reserve is due to the fact that wildlife reserve is protected area (PA) where human activity is totally prohibited in this area. (a) Oil palm-permanent forest reserve

(b) Oil palm-state land forest

2.8 2.6

Proportion

Proportion

2.7 2.5 2.4 2.3 2.2 2.1

0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Year

(c) Oil palm-wildlife reserve 0.4 0.35 Proportion

0.3 0.25 0.2 0.15 0.1 0.05 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Year

Figure 4a-c. Ratio between oil palm and the forest under the first classification level.

Analysis of the second level classification showed that the ratio between oil palm and dipterocarp forest (Figure 5a) declined over the study period (between 1991 and 2000). The ratio between oil palm and peat swamp forest was almost constant (Figure 5b) and a similar pattern was depicted by ratio between oil palm and mangrove forest (Figure 5c). These results indicate that over the period, the loss of

186

S.A. ABDULLAH AND N. NAKAGOSHI

permanent forest reserve area in Peninsular Malaysia was related to the decline of dipterocarp forest rather than peat swamp and mangrove forests. (b) Oil palm-peat swamp forest 0.06

2.5

0.05

2.4

0.04

Proportion

2.3 2.2

0.03 0.02 0.01

2.1

Year

2000

1999

1998

1997

1996

1995

1994

1993

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

1992

0

2

1991

Proportion

(a) Oil palm-dipterocarp forest 2.6

Year

(c) Oil palm-mangrove forest 0.12

Proportion

0.1 0.08 0.06 0.04 0.02 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Year

Figure 5a-c. Ratio of oil palm and the types of forest under the second level classification.

The spatial relationship can also be evaluated base on percentage of oil palm area bordered with forestland. In this case, Abdullah and Nakagoshi (in preparation) has developed an index called Association Index, which is defined as follows: Association Index, AI (%) = [ TLfr-lui /TLfr ] x 100 where, TLfr-lui is the total length of forest edge bordered with oil palm land use and TLfr is the total length perimeter of forest patches. The value of AI is between 0 and 100. The larger the value of AI means that the association between oil palm and forest areas is increase. Using the state of Selangor, Malaysia as a case study, Abdullah and Nakagoshi (in preparation) calculate the index to measure the association between oil palm and two natural land uses, that is, forest and, wetland forest and marshland. Definition of these land uses can be found in Abdullah and Nakagoshi (2006). Their results showed that within about 30 years (1966 to 1995) oil palm obviously expanded towards both forest and, wetland forest and marshland (Table 1). The third suggestion for the assessment is through the correlation analysis between oil palm and degree of forest fragmentation. Studies by Abdullah and Nakagoshi (in preparation) in the state of Selangor can be used as an example. The negative correlation resulted from regression analysis (Table 2) apparently showed that in 1966, 1981 and 1995 oil palm was one of the major determinants of forest fragmentation in the state.

OIL PALM LAND USE PLANNING AND MANAGEMENT

187

Table 1. Association index value between oil palm and two natural land uses (forest and, wetland forest and marshland) within about 30 years in the state of Selangor (modified from Abdullah and Nakagoshi, in preparation). 1966

1981

1995

Oil palm-wetland forest and marshland

9.7

43.9

61.4

Oil palm-forest

4.9

14.6

22.5

Table 2. Standardized coefficient of forest fragmentation index linear regression model with oil palm as independent variable in the state of Selangor (modified from Abdullah and Nakagoshi, in preparation).

Oil palm

1966

1981

1995

0.432***

0.318**

0.641***

3.2.2 Landscape structure of oil palm The spatial pattern of land use in a landscape can also be quantified using landscape metrics or indices (Riitters et al., 1995; He at al., 2000; Jaeger, 2000). This is the fundamental way to describe the composition and configuration of land use in a particular area or region. Landscape metrics could provide complementary information to conventional land use statistic. This is because the conventional statistical data have not explicitly represents the actual spatial composition and configuration of landscape elements (Herzog and Lausch, 2001). For example, measuring the temporal changes of oil palm using two landscape metrics, that is, mean patch size and number of patches reflects the periodicity and magnitude of this human land use in the state of Selangor (Abdullah and Nakagoshi, in preparation). The two metrics quantify fragmentation and/or expansion of the land use. Fragmentation and/or expansion provide a general picture concerning the connectivity, isolation and percentage of edge of patches of an area. The fragmentation and/or expansion could consequently have an adverse impact on the integrity of various ecological systems of the surrounding areas, such as changes in the ambient condition of wind regimes and evapotranspiration (Hobbs, 1993) and emission of carbon gasses (Okuda et al., 2004). Therefore, these fragmentation and/or expansion metrics could reflect the environmental consequences of human land use activities. Of this rationale, the landscape metrics can be proposed as the basic criteria or indicator for evaluation the existing oil palm land use planning and management.

188

S.A. ABDULLAH AND N. NAKAGOSHI

Other landscape metrics can also be used to measure landscape structure of oil palm patches. Suitable metrics may include compactness, nearest neighbour distance and largest patch index. However, caution and proper use must be assured in the application of landscape metrics (Turner and Gardner, 1991; Forman, 1995; Turner et al., 2001; Farina, 2002). This caution is relates to its suitability to measure the composition and configuration of landscape element. This can be explained in two contexts; being general and specific uses. In the general context, the selection of metrics depends on scale of analysis, that is, either regional analysis (large area, for example a state) or local scale (small area, such as urban or watershed areas). The specific context depends on properties (e.g scale and resolution) of map used to measure landscape structure. The suitability is crucial to produce data/information reliable for representing the conditions and patterns measured accurately. 3.3 The framework for integration The measurement of oil palm and forest spatial relationship, and quantification of their composition and configuration using landscape metrics provide the past and present scenario of the land use. Basically, this information can be used for monitoring and assessing the land use. Furthermore, future scenario can be predicted using various models of landscape ecology (Turner et al., 2001; Verboom and Wamelink, 2005). Nevertheless, to be more effective in planning and managing oil palm land use, the prediction of the future scenario is not merely depends on its spatial structure change. Changes in land use and landscape structure is influenced by many driving factors (Serneels and Lambin, 2001; Li and Yeh, 2004). Therefore, input of the factors in prediction model in oil palm land use planning and management is important (Figure 6). The consideration is to ensure that a more reliable prediction will be produced. In the context of oil palm land use, the develop model may include the policy and institutional factor, which is considered as underlying driving forces of land use change (Lambin, 2001; Geist and Lambin, 2002; Lambin et al., 2003). Land development policies such as the Development Structure Plan, Forest Management Systems and Oil Palm Management Systems are among the necessary input (Figure 6). The Development Structure Plan and Forest Management Systems might involve at both federal and state government levels whereas the Oil Palm Management Systems is mainly links to private or government agencies that manage the oil palm plantations. The factors of geographic conditions such as altitude, slope and soils and rock types must also be taken into consideration to develop the model (Figure 6). Land use capability (LUC) is also need to be included. Socio-economic factors are the other important component. This includes population, population density, land ownership, land prices and the local and international market prices of the commodity (Figure 6). Thus, in this framework it can be said that landscape ecological concept appeared to be as a supportive tool to influence the formulation of policies related to oil palm land use.

OIL PALM LAND USE PLANNING AND MANAGEMENT

189

Driving forces of change Socio-economic Population Population density Land ownership Land prices Oil palm market prices

Geographic Altitude Slope Soils and rock types Aspects

Policy and institutional Land use policy Agricultural policy Forest policy Development structural plan Oil palm management system

M O D E L

Land use and landscape change

Landscape ecology approach Oil palm-forest spatial relationship Landscape structure

F U T U R E S C E N A R I O

support

Past and present scenario

Figure 6. The framework on how landscape ecology can be integrated into model of planning and management of oil palm land use.

4. CONCLUSION Oil palm plantation apparently expanded throughout the nation. However, in recent years it is particularly occurred in Sabah and Sarawak compared to Peninsular Malaysia. Although this human land use has, for a long time, been caused deterioration to natural forest ecosystem, undoubtedly it is still crucial to promote economic development of the country. Thus, in the context of sustainability for this activity, landscape ecology concept is potential to be applied in planning and managing the land use. The rationale is relies on its ability to translate or reflect environmental consequences due land use or landscape change quantitatively. Based on this data and information, by using landscape ecological models a prediction of the future scenario of the land use can be made. However, land use or landscape change is driven by various underlying and proximate factors. Therefore the quantitative data is inadvertently reflects the linkage between the factors and the process of change. Understanding of this relationship is pivotal to influence related policy in order to increase effectiveness in planning and managing oil palm land use. Therefore, model that to be developed is not merely depends on the quantitative data of spatial structure but it also might include the factors that drive the changes. Therefore, in this context the role of

190

S.A. ABDULLAH AND N. NAKAGOSHI

landscape ecology is as supportive tool in providing implication for existing policies related to oil palm land use. Oil palm and forested areas are associated to each other. Thus, the improvement in oil palm land use planning and management to ensure the preservation and protection of natural forest is necessary. This is not only for environmental conservation but also for human well-being. REFERENCES Abdullah, A.R. 1995. Environmental pollution in Malaysia: trends and prospects. Trends in Analytical Chemistry, 14, 191-198. Abdullah, S.A. 2003. Fragmented forest in tropical landscape: the case of the state of Selangor, Peninsular Malaysia. Journal of Environmental Sciences, 15, 267-270. Abdullah, S.A. and Nakagoshi, N. 2006. Changes in landscape spatial pattern in the highly developing state of Selangor, peninsular Malaysia. Landscape and Urban Planning, 77, 263-275. Azrina, M.Z., Yap, C.K., Ismail, A.R., Ismail, A. and Tan, S.G. in press. Anthropogenic impacts on the distribution and biodiversity of benthic macroinvertebrates and water quality of the Langat River, Peninsular Malaysia. Ecotoxicology and Environmental Safety. Casson, A. 2003. Oil Palm, Soybeans and Critical Habitat Loss. A Review Prepared for the WWF Forest Conservation Initiative. WWF Forest Conservation Initiative, Switzerland. Farina, A. 2002. Landscape Ecology in Action. Kluwer Academic Publisher, the Netherlands. Forman, R.T.T. 1995. Some general principles of landscape and regional ecology. Landscape Ecology, 10, 133-142. Geist, H.J. and Lambin, E.F. 2002. Proximate causes and underlying driving forces of tropical deforestation. BioScience, 52, 143-150. Goh, K.C. 1982. Environmental impact of economic development in Peninsular Malaysia: a review. Applied Geography, 2, 3-16. Gulink, H. and Wagendorp, T. 2002. References for fragmentation analysis of the rural matrix in cultural landscapes. Landscape and Urban Planning, 58, 137-146. Hartley, C.W.S. 1988. The Oil Palm. 3rd Edition. Longman, London. He, H.S., DeZonia, B.E. and Mladenoff, D.J. 2000. An aggregation index (AI) to quantify spatial patterns of landscapes. Landscape Ecology, 15, 591-601. Herzog, F. and Lausch, A. 2001. Supplementing land-use statistics with landscape metrics: some methodological consideration. Environmental Monitoring and Assessment, 72, 37-50. Hobbs, R. 1997. Future landscapes and the future of landscape ecology. Landscape and Urban Planning, 37, 1-9. Hobbs, R.J. 1993. Effects of landscape fragmentation on ecosystem processes in the Western Australia wheatbelt. Biological Conservation, 64, 193-201. Jaeger, J.A.G. 2000. Landscape division, splitting index, and effective mesh size: new measures of landscape fragmentation. Landscape Ecology, 15, 115-130. Jasmi, A. 1997. Wildlife Conservation Issues in the Langat Basin. Master Thesis (unpublished). Universiti Kebangsaan Malaysia. Kumar, R. 1986. The Forest Resources of Malaysia: Their Economics and Development. Oxford University Press Pte, Singapore. Lambin, E.F., Turner, B.L., Geist, H.J., Agbola, S.B., Angelsen, A., Bruce, J.W., Coomes, O.T., Dirzo, R., Fischer, G., Folke, C., George, P.S., Homewood, K., Imbernon, J., Leemans, R., Li, X., Moran, E.F., Mortimore, M., Ramakrishnan, P.S., Richards, J.F., Skånes, H., Steffen, W., Stone, G.D., Svedin, U., Veldkamp, T.A., Vogel, C. and Xu, J., 2001. The causes of land-use and land cover change: moving beyond the myths. Global Environmental Change, 11, 261-269. Lambin, E.F., Geist, H.J. and Lepers, E. 2003. Dynamics of land-use and land-cover change in tropical regions. Annual Review of Environment and Resources, 28, 205-241. Leitao, A.B. and Ahern, J. 2002. Applying landscape ecological concepts and metrics in sustainable landscape planning. Landscape and Urban Planning, 59, 65-93. Lenz, R.J.M. and Stary, R. 1995. Landscape diversity and land use planning: a case study in Bavaria. Landscape and Urban Planning, 31, 387-398.

OIL PALM LAND USE PLANNING AND MANAGEMENT

191

Li, X. and Yeh, A.G.-O. 2004. Analyzing spatial restructuring of land use patterns in a fast growing region using remote sensing and GIS. Landscape and Urban Planning, 69, 335-354. Norhayati, A., Abdullah, S.A., Shahrolnizah, A., Md-Zain, B.M., Shukor, M.N., Hazimin, H. and Nordin, M. 2004. Diversity and density of mammals in the peat swamp forests of the Langat Basin, Selangor, Malaysia. Journal of Malaysian Applied Biology, 33, 7-17. Okuda, T., Yoshida, K., Numata, S., Nishimura, S., Suzuki, M., Hashim, M., Miyasaku, N., Sugimoto, T., Tagashira, N. and Chiba, M. 2004. An ecosystem-management approach for CDM afforestation and reforestation activities: the need for an integration ecosystem assessment based on the valuation of ecosystem services for forested land. In Kyoto Mechanism and the Conservation of Tropical Forest Ecosystem. Okuda, T. and Matsumoto, Y. (Eds.), pp. 67-78. Proceedings of the International Symposium/Workshop on the Kyoto Mechanism and the Conservation of Tropical Forest Ecosystems, 29-30 January 2004, Waseda University, Tokyo, Japan. Palmer, J.F., Lankhorst, J.R.K. 1998. Evaluating visible spatial diversity in the landscape. Landscape and Urban Planning 43: 65-78. Riitters, K.H., O’Neill, R.V., Hunsaker, C.T., Wickham, J.D., Yankee, D.H., Timmins, S.P., Jones, K.B. and Jackson, B.L. 1995. A factor analysis of landscape pattern and structure metrics. Landscape Ecology, 10, 23-39. Serneels, S. and Lambin, E.F. 2001. Proximate causes of land-use change in Narok District, Kenya: a spatial statistical model. Agriculture, Ecosystem and Environment, 85, 65-81. Turner, M.G. and Gardner, R.H. 1991. Quantitative Methods in Landscape Ecology. Springer-Verlag, New York, USA. Turner, M.G., Gardner, R.H., O’Neill, R.V. 2001. Landscape Ecology in Theory and Practice. SpringerVerlag, New York, USA. Verboom, J. and Wamelink, W. 2005. Spatial modelling in landscape ecology. In Issues and Perspectives in Landscape Ecology. Weins, J and Moss, M. (Eds.), pp. 79-89. Cambridge University Press, UK. Yusuf, M.A. and Nordin, M. 2003. River water quality assessment and ecosystem health: Langat River Basin, Selangor, Malaysia. In Managing for Healthy Ecosystems. Rapport, D.J., Lasley, W.L., Rolston, D.E., Nielson, N.O., Qualset, C.O., Damania, A.D. (Eds.), pp. 1395-1413. Lewis Publisher/CRC Press, USA. Zubaid, A. 1993. A comparison of the bat fauna between a primary and fragmented secondary forest in Peninsular Malaysia. Mammalia, 57, 201-206.

CHAPTER 13

MANAGING BIODIVERSITY OF RICE PADDY CULTURE IN URBAN LANDSCAPE Case research in Seoul City

I.-J. SONG 1, Y.-R. GIN 2 1 Dept. 2

of Urban Environment, Seoul Development Institute (SDI), Seoul, Korea; National Park Jirisan Southern Office, Korea National Park Service, Jeonnam, Korea

Abstract. Cultivated land in the Kangnam area of Seoul is for the most part dispersed. The large area paddies are distributed around the western boundary of Seoul, but the small area paddies are located at the southeast boundary of Seoul. Review of the land use trends of these cultivated lands, large areas have been converted into development areas or converted for use by green houses for high profits. According to this trend, by identifying the landscape ecology characteristics of the rice paddy culture and finding management alternatives it could be possible to improve the biodiversity enhancement of the region, specifically in terms of space and time. According to the results of the shape index analysis, rice paddies in Seoul comprise a 9.4ha area, larger than the mean area of other types of cultivated land with the exception of land used for green houses. Interestingly, the circumference of paddy use area has a similar trend to that of the cultivated land patches of green houses. Also, of the 68% of the first three neighbouring land use that was analyzed the patches were primarily dedicated to river and wetlands (50%) and forests (23%). This characteristic of rice paddy culture, in particular as major neighbouring land use, is representative of the other types of cultivated land; river and wetland usage is due to its particular characteristics. Therefore, on the basis of the analysis’ results, cultivated land management is important for urban environment conservation and biodiversity.

1. INTRODUCTION The land, if viewed from above, is a mosaic that consists of various types of landscape components. The structure of the mosaic landscape has meaning when it is ultimately connected with functions and accordingly the important facets

193 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 193–208. © 2007 Springer.

194

I.-J. SONG AND Y.-R. GIN

are not only content but also context (Gustafson and Gardner, 1996; Lee, 1997; Forys and Humphrey, 1999). Context is defined as the connection of things identified by three factors; location, adjacency, and neighbourhood. Location is where the area is simply placed in regards to space. Adjacency is identified by the degree to which the area contacts other patches or regions of which it is concerned. Neighbourhood is the regional mosaic that is connected with active interaction. The agro-system consists of artificial bio-communities and interacts with the climate and soil. The agro-system, unlike the natural system that is selfmaintained, is fundamentally unstable. Cultivation, harvest, and biological simplicity interrupt the recycling of organic materials and makes crops more susceptible to diseases. Stability in the agro-system is determined by soil fertility, the recycle of organic materials, biological diversity of crops, etc. Traditionally mixed tillage has kept an ecological balance to some extent. A closed agro-system is provided organic materials from the vegetation communities of forests, hedges, crop diversification, crop rotation, and farm animals. Since the cultivation of land, a part of the urban landscape has been managed for crop production, the primary function of which is agricultural land. In regards to economic motivation, unbalanced input and output of energy makes for depreciation in the value of land converted to other types of land use. The issue of decease of cultural lands is shown as a transfer of development rights of cultivated land to urbanites or land use conversion into cultivated land for green houses making it possible to gain better interest from less investment due to the rapid decrease in the number of farmers in comparison to cultivated land area caused by the phenomenon of depreciated farming (Stayle, 2000). Now, in 1999, the area of cultivated land in Seoul reached 5.75% of total land use (Seoul Metropolitan Government, 2000a, b); the proportion of the rice paddy culture, in regards to cultivated land, is 25.26 % (Song and Gin, 2002). This is the second largest area of cultivated land, closely following the area used by the dry paddy culture that has reached 41.49%. However, review of trends in land use, in regards to cultivated land, show the largest part of such land is likely to be converted for developmental use or the cultivated land of green houses, of the type that supply a high economic return. Accordingly, identifying the characteristics of the ecological landscape in the urban rice paddy culture; in regards to space and time, alternative forms of management and observation for the region must be found, there must be an escape from the planning, management and policy that is currently being made on the basis of real boundaries where a patch of mosaics is circumvented.

MANAGING BIODIVERSITY OF PADDY FIELD IN URBAN LANDSCAPE

195

2. METHODS This study approached, by way of three steps, a process to suggest the management for biodiversity on the basis of characteristics of cultivated land through analysis of ecological landscape in cultivated land. First, a literature review was conducted to draw upon the value and meaning of cultivated land conservation, economic efficiency, and the niche of the ecological landscape. Second, to analyse the characteristics of the ecological landscape, in the cultivated land of Seoul, the patches of cultivated lands using the survey data of biotope types in Seoul were selected. On the basis of the data of the biotope survey, the biotope types of cultivated lands were classified into five types of land use; rice paddy culture, dry paddy culture, pasture, orchard, and cultivated land for green houses.

Figure 1. Study process.

196

I.-J. SONG AND Y.-R. GIN

This study analysed number, size, perimeter, and the shape index of patches. Also, to identify the characteristics of patches, neighbouring land usage was analysed after selecting the typical size of patches in cultivated land. Third, the case study was conducted to manage cultivated land for biodiversity on a representative patch that was selected in the prior steps (Figure 1). The map, in 1/3000 scale, used in ‘the Seoul Metropolitan Biotope Map’, which was used as a base; the program of analysis was Arc View 3.2. 2.1 Landscape Ecological niche and economic efficiency of the rice paddy culture in urban areas The regulations governing permitted activities and standards of management in rural and urban areas, according to the land use of cultivated land, are different. Namely, rural areas and agriculture areas in the National Land Use Law are governed by the Farming and Fishery Development Act and the Agriculture Land Act, classified into cultivated land on the Agriculture Land Act. Otherwise, city urban areas are under the National Land Use Act and ruled by a productive green area such as found in the Urban Planning Act. The productive green area has the same context as a natural green area, or agriculture development (enhancement) in as much as it is appointed to regulate and reasonably adjust the cultivated land. However, the cultivated land in urban areas is slightly different from the cultivated land in rural areas, mainly in the way it is managed and classified according to its purpose, in regards to urban growth restricted areas, land use conversion restricted areas, and urban agriculture enhancement areas (Urban Planning Establishment Guidelines, 2000). The rice paddy culture, one of cultivated land, is charged with several functions such as: 1) water reserves, 2) territorial integrity, 3) atmosphere conservation, 4) biodiversity conservation, and 5) leisure and recreation, but its primary responsibility is crop production. The important purpose for water reserves is flood control, underground water retention, and water purification. Territorial integrity is responsible for protection against soil erosion and soil collapse alleviation. Atmosphere conservation is charged with atmosphere purification and climate alleviation (Kim, 1996; Kwon, 1998; Choi, 2001). Lim (2002) reported on the public value of the rice paddy culture in terms of flood control, water retention, and recreation; calculating into alternative facilities and its cost as seen in the following (Table 1). Cultivated land is defined by being a half-natural area that is interrupted regularly by humans and is primarily used as a productive factor in crop supply. Otherwise, on the ecological landscape aspect, the rice paddy culture makes for a dynamic landscape keeping water in urban areas or dry built areas and becomes a potential place and space where biodiversity is protected, offering a habitat where unique flora and fauna may live. This cultivated land also connects with neighbouring river and wetlands, unlike territorial and river ecosystems. Also such areas act as a stepping-stone for migratory birds. Considering the potential the rice paddy culture has, value would be emphasized as a water reserve, rich in biodiversity conservation, and providing for leisure and recreation.

MANAGING BIODIVERSITY OF PADDY FIELD IN URBAN LANDSCAPE

197

The small size of cultivated land and forests in urban areas have been scattered and large forests at the boundary of urban areas have been fragmented and lost. Isolated urban nature would become the biggest factor in causing a decrease of biodiversity and furthermore, extinction and endangerment of flora and fauna species, thus stimulating a change in genetic information according to the increase of artificial affects on the various species. Cultivated land is valuable in terms of biodiversity because fauna and flora native to these areas are rarely found in other types of habitats and those animals that are appropriate to these habitats hardly ever find alternative ecosystems. Table 1. Major functional value of the rice paddy culture (constant market price in 2000).

Benefit public functions

Alternative facilities

Flood control Water retention

Dam construction Dam construction Approach cost to virtual market Carbon cost per ton Carbon cost per ton

Recreation Air filtration O2 generation

Construction cost and value for alternative facilities (billion won) 9,954 43,176

Cost of depreciation cost and maintenance per ha for a year (million won) 0.077 0.336

15,003

1.290

5,414

0.466

27,700

2.380

*Source: A study on the socio-economic effect of public functions in the rice paddy culture

2.2 Structure of landscape ecology and characteristics of the rice paddy culture in Seoul 2.1.1 Environmental Profile and Characteristics of Seoul Seoul is divided into North and South by the border of the Han-river topography. The area is characterized as a temperate and terrestrial climate in which the yearly climatic difference is large in regards to topographical effects. Yearly precipitation reaches 1488mm and the average temperature is 12.5℃ (the average for latest 10 years; 1991-2000) (Seoul Metropolitan Government, 2001). 2.1.1.1 Structural characteristics of the rice paddy culture in Seoul Cultivated land, which currently composes 5.74% of the total Seoul area, is mainly dispersed in the Kangnam area. This assumes that forests in the Kangbuk

198

I.-J. SONG AND Y.-R. GIN

area restricted the use of cultivated land, which is steeper and higher than in Kangnam. Reviewing the dispersal characteristics of the rice paddy culture, it’s distributed in large size near the western boundary of Seoul and the rice paddy cultures in small size are scattered near the southwestern border (Figure 2).

Figure 2. Dispersal map of cultivated land.

Table 2. Present condition of cultivated land use in Seoul.

(Unit:ha(%))

Type of cultivated land

Number of patch

Sum of area

Mean area

Min. area

Max area

Rice paddy

84(6.76)

788.69(25.46)

9.40

0.18

166.37

Dry paddy

940(75.62)

1,307.12(42.49)

1.39

0.01

52.58

Pasture

2(0.16)

2.60(0.08)

1.30

0.46

2.14

Orchard

71(5.71)

146.14(4.75)

2.06

0.03

33.27

Cultivated land of green House

146(11.75)

831.41(27.03)

19.83

0.03

113.17

Total

1,243(100)

3,075.96(100)

-

-

-

MANAGING BIODIVERSITY OF PADDY FIELD IN URBAN LANDSCAPE Table 3. Shape index analysis of rice paddy patches.

199

(Unit:ha(%))

Patch size

Number of patch

Sum of area

Mean area

Min. area

Max area

Mean shape index

Above 50

4(4.76)

414.21(52.52)

103.55

57.70

166.37

1.85

Less 50above 40

1(1.19)

49.67(6.30)

49.67

49.68

49.67

2.20

40-30

2(2.38)

64.62(8.19)

32.31

32.06

32.57

1.70

30-20

1(1.19)

25.52(3.36)

26.52

26.52

26.52

2.63

20-10

6(7.14)

89.39(11.33)

14.90

11.36

18.58

1.86

10-1

44(52.38)

131.63(16.69)

2.99

1.05

9.25

1.64

Less than 1

26(30.95)

12.64(1.60)

0.49

0.18

0.97

1.41

Total

84

788.69(100)

-

-

-

-

The ratio of dry paddy cultures according to the area of cultivated land type is 42.49% (1,307ha), cultivated land used by green houses is 27.65%, orchards 4.75%, and pastures 0.46%. The mean area used by green houses was 19.83ha and by the rice paddy culture 9.40%. Otherwise, the mean area of the dry paddy culture or pastures was respectively 1.39ha and 1.30ha. Particularly, the rice paddy culture and cultivated land used by green houses were relatively larger than the dry paddy culture and the pasturelands. The number of rice paddy culture patches (84) reached about 6.75% in total (Table 2). The rice paddy culture was divided into seven scales, less than 1ha, above 1-less than 10ha, above 10-less than 20ha, above 20-less than 30ha, above 20-less than 40ha, above 40-less than 50ha, and above 50ha. This was used to analyse each scale of mean area, mean perimeter, and mean shape index. The mean shape index of each scale in the 1.41 to 2.63 range had a simple shape of circumference (bound) compared to that of the dry paddy culture that ranged from 3.18 to 6.26 (Song and Gin, 2002). However, there is a similarity to the cultivated land of green houses in that the shape index of each patch in its scale ranged from 1.42 to 2.38 (Song and Gin, 2003). Also, patches of cultivated land for green houses and the rice paddy culture have less relativity between the surveyed areas and the perimeter of the patches. As the patch area increased the shape index generally increased as well, but it decreased when it reached a certain area. This trend was likely to show similar results with the patch analysis of the dry paddy culture. Cultivated land is likely to adopt a typical pattern through artificial management (Table 3).

200

I.-J. SONG AND Y.-R. GIN

Table 4. Landscape ecological attributes analysis of rice paddy patches between 1ha and 10ha in size. (Unit:ha) Patch

Area

Perimeter

Shape

Number

(ha)

(m)

index

1

1.05

514.96

1.42

3

1.09

756.35

5

1.21

774.31

7

1.24

774.31

9

1.37

822.74

Perimeter length and land use type River and

Transportation

wetland

facility

2.04

Forest

Cultivated land

-

1.60

Forest

-

-

1.96

Cultivated land

River and wetland

-

Commercial and

Transportation

business area

facility

Cultivated land

-

1.99

Cultivated land River and

11

1.45

525.38

1.23

13

1.61

673.11

1.50

wetland Forest

15

1.72

724.69

1.56

Cultivated land

17

1.90

658.70

1.35

Cultivated land

19

1.95

576.48

1.16

21

2.10

617.23

1.20

23

2.40

798.12

25

2.46

27

2.66

29

2.88

1154.96

1.96

31

3.51

1239.08

1.87

Cultivated land

33

3.68

889.533

1.31

Cultivated land

35

4.04

1029.69

1.45

Cultivated land

37

4.77

1272.17

1.64

39

5.43

2081.57

41

6.51

1430.23

43

9.90

-

Cultivated land

Residential area

1.45

Cultivated land Residential area Cultivated land

Forest Transportation facility River and wetland Transportation facility Forest

691.34

1.24

Cultivated land

River and wetland

-

1289.56

2.23

Cultivated land Industrial and urban infrafacility

Residential area

-

River and wetland

-

2021.76

1.52

River and wetland Cultivated land

Industrial and urban infrafacility Transportation facility Transportation facility Transportation facility River and wetland

1.58

Cultivated land

River and wetland

1.89

Industrial and urban infrafacility

Idle land

Residential area River and wetland -

River and wetland Cultivated land Transportation facility Commercial and business area

2.1.1.2 Characteristic analysis of land use neighbouring the rice paddy culture in Seoul The developed space in urban landscape ecology is compared to mosaic, whereas forest, green space areas, cultivated land, and river areas are compared to patches or corridors. The continuity and health of these patches or corridors is

MANAGING BIODIVERSITY OF PADDY FIELD IN URBAN LANDSCAPE

201

affected by the interaction, interchange, and movement between ecosystems according to neighbouring land use (Lee, 2001). In other words, the landscape ecosystem depends upon complexity and diversity. Therefore, the enhancement of biodiversity in the rice paddy culture should find alternatives on the basis of analysis of the biological environment and neighbouring land use. This should be taken into consideration with the analysis of unique characteristics that the rice paddy culture itself has. In areas the above 1-less than 10 ha variety were the most numerous and the largest areas of patches (Table 4). These patches were set as a typical patch of the rice paddy culture in Seoul. Of these, 22 patches (50% of 44 patches in total) were selected as samples after arraying the patches in scale. Next, in a review of neighbouring land use in the selected patches it was noted that the neighbouring patches facing the three largest patches were subjected to different usage (Table 5). The most frequent type of neighbouring land use was the other type of cultivated land reached 68%. Neighbouring the patches under other types of land use were rivers and wetlands, transportation facilities, and forests and their ratios were respectively 50%, 45%, and 23%. The results of the analysis in regards to the other types of cultivated land has shown that distribution of the other types of cultivated land neighbouring the rice paddy culture (Song and Gin, 2002; Song and Gin, 2003) are small but as a group are scattered at the boundary of Seoul. Compared to the analysis of the cultivated land of green houses and the rice paddy culture, a unique characteristic is that river and wetlands appear quite near to most of the rice paddy cultures. This result is likely to be related to the usual characteristics of the rice paddy culture. 3. CASE STUDY 3.1 Site selection Of the 22 selected patches chosen as sample analysis sites, the patch that had the mean area, perimeter and shape index of all patches was selected as a representative sample for the survey. The selected site was located in Macheondong, Songpagu, Seoul, Korea was under the administration of Kyunggido (province) on a 2m width of conc.’s road and its area reached 2.40ha. Deciduous forest was facing the north side of the selected site and other types of cultivated land, including cultivated land and dry paddy culture, were on the southeast side; a residential area was on the south side. On the west side that was under the administration of Kyunggi, the area was located near a few manufacture factories that generated a great deal of noise (Figure 3, 4, and 5). Also a 2m-wide area of road passed through the site and a 1m wide river flowed through the south side of site.

202

I.-J. SONG AND Y.-R. GIN

Figure 3. Selected site.

Figure 4. Land use types in surrounding area of the selected site.

MANAGING BIODIVERSITY OF PADDY FIELD IN URBAN LANDSCAPE

203

3.2 Environmental profile of selected site The field survey was conducted on the basis of the environment survey suggested in ‘ecology restoration and biodiversity enhancement according to biotope types in Seoul (1st year)’(Table 5). In the selected site, 49 species in total were found. Fauna were more likely to be based on the neighbouring ecosystem than on the site ecosystem and most of the neighbouring and areas facing the selected site were much deteriorated. The rice paddy culture is a potential place that contains water resources but the accompanying intensive cultivation of single crops for production enhancement drives the use of chemicals and organic materials, such as herbicide, fertilizer on site. This causes a decrease in the number of living things on site as well as negative physical effects such as soil erosion, pollution, groundwater pollution, and underwater pollution. The water quality of the 1m-wide river that flowed along the south side was worsened by input of non-point pollutants such as sewage disposal and dumps.

Figure 5. Landscape characteristics of selected site surrounding Seoul City (winter season).

3.3 Management of the rice paddy culture It is important for management of the rice paddy culture to find alternatives for lessening negative effects coming from the outside of the rice paddy culture and the management of degree and method of intervention, soil resources and living resources. Therefore, to lessen outside influence, some water filtration methods such as vegetation strip, grassland strip, or a water pollution pond could be implemented at the sites. To enhance biodiversity, proper methods such as conservation headline, sterile strip, grassland bank, nesting strip, hedge, grass baulks, and rotational parcels set-aside for birds, could be applied on site to

204

I.-J. SONG AND Y.-R. GIN

function as habitats, shelters, and corridors (Game Conservation Trust, 1995; Song and Gin, 2002). Table 5. General environmental profile of the selected site, its opportunities and constricts to improve suggested issues.

Factor Vegetation Reptiles and amphibians Birds

Descriptions 9species(naturalized species 3)

Insects

38species

Mammals

1 species

General characteristic

There is 2m width of paved road with Conc. getting through in the middle of site At the north side of site, deciduous forest 1m width of stream flows at the south side of site The other types of cultivated land are facing at the boundary of site

Constrict factor of biodiversity

Lack of connectivity with the neighbouring forest habitat for birds Soil erosion, ground water contamination, underground water pollution by organic materials and chemicals

Opportunities of biodiversity enhancement

Available to irrigate water and other types of species to the site through stream that flows at the north side of site

1 species

4. DISCUSSION The area of cultivated land in Seoul has reached 5.75% in total (Seoul Metropolitan Government, 2000c) but this large area of cultivated land is neither friendly nor beautiful to the citizens because the cultivated land does not fully play its role and remains a development reservation area in the urban area. However, cultivated land in urban areas offer crops and economic interest to urbanities, at the same time offering an aesthetic recreation place characterized by open spaces in the urban scale and recycles sewage.

MANAGING BIODIVERSITY OF PADDY FIELD IN URBAN LANDSCAPE

205

Table 6. Alternatives for on-site management according to major issues.

Critical design elements Abiotic environment

Wind

Soil

Water

Biotic environment

Vegetation

Insects Birds Reptiles and amphibians

Landscape

Patch shape

Major issues Air pollution through wind-blow using chemicals on site Soil pollution by organic materials

Non-point pollution source by through ground-flow, underwater flow Required chemical pollutants filtration Management of harmful weed invasion to neighbouring areas Pest management Required shelter to hide Required pond for connectivity between the surrounding forest and river Ellipse shaped patch

Alternatives for in-site management Fences, hedge 2m height for air filtration Use several times using little chemicals when necessary Set fallow damn to minimize the soil erosion by rain flow Grassed waterway on drainage Grassy bank on the boundary of patch 1 meter wide sterile strip

Crop rotation Conservation headland of 6m width Required deliberate review and prior assessment for introduction to the site Boundary that would be set corridor makes complex to let species move between ecosystem but the other side of boundary makes simple to minimize the bad effect and disturbances to the neighbouring ecosystem (Cont.)

206

I.-J. SONG AND Y.-R. GIN Table 6. (cont.)

Network connectivity Edge and margin

Structure

Required design elements to connect between separated landscape ecologies Edge is simplified for intensive crop culture and using machines

Required various structure of vegetation; trees, shrubs and herbs

Create small size of corridor to move animals and birds Create various types of habitat for birds and vegetation It would play a buffer role to minimize negative effect to the surrounding environment and to enhance landscape Its composition would be native species that live near ecosystems such as forest, grassland, river etc. (for example, Quercus acutissima, Pinus densiflora for tree)

In the light of landscape ecology, urban cultivated land is reconsidered an output of human civilization in terms of the socio-cultural aspect and these functions and roles were slightly different from those in rural areas. More precisely, the cultivated land is focused on crop production but more generally; it is connected with the outer environment and affects the urban environment directly or indirectly. Therefore, the relationship and biodiversity between humans and cultivated land is beyond the interaction between living things and non-living things. The rice paddy culture is representative as a half wetland and contributes to enhancement of biodiversity through a connection with the neighbouring landscape. The major purpose of biodiversity maintenance is to support diversity in the local communities and promote independent health and it should pursue the satisfaction of principles in landscape ecology of the cultivated land. For biodiversity in the cultivated land to thrive, the urban landscape ecology that is dependant on the complexity and diversity of neighbouring ecosystems should enhance diversity considering not only the characteristics of the neighbouring ecosystem but also the ecosystem of the cultivated land itself. REFERENCE Agger, P. and Brandt, J. (1988). Dynamics of small biotopes in Danish agricultural landscapes. Landscape Ecology, 1, 227-240 Ahn, I.C. (2000). Trend and prospective of rice paddy culture in Korea. Journal of Agriculture Business and Policy, 27(3), 18-37. (in Korean)

MANAGING BIODIVERSITY OF PADDY FIELD IN URBAN LANDSCAPE

207

Choi, J.D. (2000). Environment management and non-point pollution of agriculture watershed. Journal of Rural and Environment, 10(2), 29-38. (in Korean) Choi, K.W. (2002). Comparison of environmental function between the rice paddy culture and wetland. Agriculture and Environment, 71, 58-64. (in Korean) Council for Agricultural Science and Tech. (2002). Urban and Agricultural Communities: Opportunities for Common Ground-Task Force Report. No. 138 (pp. 8-9), May 2002. Iowa State Univ., Ames. Environmental Analysis Division Office of Environmental Policy Federal Highway Administration (1989) Guidelines for Implementing the Final Rule of the Farmland Protection Policy Act for Highway Projects. Flury, A.M. and Brown, R.D. (1997). A framework for the design of wildlife conservation corridors with specific application to Southwestern Ontario. Landscape and Urban Planning, 37, 163-186 Forman, R.T.T. (1995). Land Mosaics: The Ecology of Landscape and Regions, Cambridge, Cambridge University Press, UK Forys, E. and Humphrey, S.R. (1999). The importance of patch attributes and context to the management and recovery of an endangered lagomorph. Landscape Ecology, 14, 177-185 Gustafson, E.J. and Gardner, R.H. (1996). The effect of landscape heterogeneity on the probability of patch colonization. Ecology, 77, 94-107 Kim, G.S. (1996). Assessment of public functions in the rice paddy culture. Journal of Agriculture and Engineering, 38(4), 27-33(in Korean) Korea National Statistical Office. (2000). Agricultural Census Report –Whole Country 16p. (in Korean) Kuminoff, N.V., Sokolow, A.D. and Sumner, D.A. (2001). Farmland Conversion: Perception and Reality. Agricultural Issues Centre (AIC Issues Brief), University of California. Kwon, S.K. (1998). Let’s expand the crop paddy culture, at the same time conserve for environment conservation. Journal of Korea Agriculture and Engineering, 40(3)(in Korean) Lee, D.W. (1997). Floating Ecology. Bumyangsa Publication, Seoul. (in Korean) Lee, D.W. (2001). Landscape Ecology. Seoul National University Publication. 349p. (in Korean) Lee, I.Y. (2001) Actual condition of agricultural use of chemicals to the rice paddy culture and the dry paddy culture. Korean Journal of Grasses, 21(1), 58-64. (in Korean) Libby, W.L. (2000). Farmland as a Multi-Service Resource: Policy Trends and International Comparisons. International Symposium on Agriculture Policies Under the New Round of WTO Agriculture Negotiations, Taipei, Taiwan, December 5-8, 2000 Libby, W.L. (2001). Efficiency, Equality and Farmland Protection: an Economic Perspective. Educational Symposium of the American Agricultural Law Association, Colorado Springs, Co, October 11-13 Lim, J.H. (2002). Scio-economic effect about the public functions of the rice paddy culture-comparison analysis between Korea and Japan. Agriculture and Environment, 64, 34-40. (in Korean) Lim, K.S. (1999). Energy balance according to the management method of the rice paddy culture area. Journal of Korea Environment and Agriculture, 18(4), 299-303. (in Korean) Ministry of Construction and Transportation (1997). Improvement of System of Development Restrict Area. (in Korean) Seoul Metropolitan Government (2000a). Biodiversity Enhancement and Ecosystem Restoration According to Biotope Types (1st year). 421p.(in Korean) Seoul Metropolitan Government (2000b). Seoul Biotope Survey and Principle of Eco-City Planning (1st, 2nd year) 245p. and 394p.(in Korean) Seoul Metropolitan Government (2000c). Seoul Urban Planning Guidelines (in Korean) Seoul Metropolitan Government (2001). Seoul Statistical Yearbook (in Korean) Song, I.-J., Hong, S.-K. and Kim, H.-O. (2000). Distribution characteristics of naturalized plants influenced by land use pattern in Seoul metropolitan area. In Brandt, J., Tress, B. and Tress, G. (Eds.), Multifunctional Landscape Ecology (pp. 18-20), Roskilde, Denmark. Sept. 2000. Song, I.-J., Hong, S.-K., Kim, H.-O., Byun, B., Gin, Y.-R. (2005). The pattern of landscape patches and invasion of naturalized plants in developed areas of urban Seoul. Landscape and Urban Planning, 70, 205-219

208

I.-J. SONG AND Y.-R. GIN

Song, I.J. and Gin, Y.R. (2002). A model of biodiversity enhancement through analysis of landscape ecology in cultivated land of Seoul. Journal of Korea Environment and Ecology, 16(3), 249-260 (in Korean) Song, I.J. and Gin, Y.R. (2003). A study on Management of the cultivate land of greenhouses through landscape ecological pattern analysis in Seoul urban area. Journal of Korea Environment and Ecology, 17(1), 56-70 Watts, B.D. (1996). Landscape configurations and diversity hotspots in wintering sparrows. Oecologia 108, 512-527

CHAPTER 14

LANDSCAPE RESTORATION A case practice of Kushiro Mire, Hokkaido

F. NAKAMURA, Y.S. AHN Department of Forest Science, Hokkaido University, Sapporo, Japan

Abstract. Kushiro Mire, a marsh located near the mouth of the Kushiro River, is suffering from the cumulative effects of pollution caused by land-use development in the watershed. A high wash load is of particular concern and accounts for approximately 95% of the total suspended sediment load that flows into the marsh. Researches have found that turbid water floods the margins of the marsh; this is due to riverbed aggregation in a channelized stream reach that provides agricultural drainage. An analysis of Cs137 concentrations determined that the rate of fine sediment deposition was approximately three to eight times higher in the channelized reach than in a reach of the natural river. This rapid sediment deposition has lowered groundwater levels and enriched the nutrient content of the marsh soil. Consequently, woody species are rapidly invading the margins of the marsh, causing concern about a vegetation shift from reeddominated marsh to woodland. To address the physical and biological changes that are taking place in Kushiro Mire, various restoration projects have been planned and are being implemented under the Kushiro Mire Conservation Plan. Three examples of projects in the Kushiro Mire Conservation Plan are a restoration of the straightened river channel to meandering course, a forest restoration near Takkobu Lake, and a wetland restoration of a crane habitat. To develop pasture fields the natural meandering rivers in the Kushiro Mire have been channelized from the marginal areas of the marsh. The channelization projects lost pristine river-floodplain landscapes and inhibiting wildlife species. In the Kayanuma area, a river section extending about 2 km of Kushiro River is planned to restore from a straightened channel to a original meandering stream and floodplains. Monitoring and scientific evaluation will be conducted before and after the project and compared with downstream reference reaches. Fine sediments and nutrients have been accumulating in Takkobu Lake because of agricultural development and soil erosion in the uplands. The number of aquatic species in the lake has also been decreasing. An environmental assessment was undertaken in collaboration with “Trust Sarun”, a non-profit organization, and sites were selected for conservation and restoration work. A larch forest was purchased to prevent it from being clear-cut and thus increasing sediment loading in the lake. The forest will be restored to its natural state. In addition, the Ministry of Environment in the Hirosato District acquired a wetland restoration site that was originally designated as an “ordinary area,” i.e., the least regulated area of a national park. The restoration site is an abandoned agricultural field with an old drainage system developed in the 1960s; it is an important breeding habitat for red-crowed cranes (Grus japonesis). Based on a preliminary investigation, and under careful supervision to avoid disturbing the cranes, soil excavation and seeding experiments were undertaken and biogeochemical processes have been monitored.

209 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 209–233. © 2007 Springer.

210

F. NAKAMURA

AND

Y.S. AHN

1. INTRODUCTION Many of ongoing and planned restoration projects in Japan have addressed disturbed rivers, wetlands, and lakes. The degradation of these aquatic ecosystems often involves multiple causes occurring at larger spatial scales than the restoration sites, which may limit the ability or effectiveness of local habitat improvement. Watershed perspective is then assumed for such ecosystem restoration. The extensively managed landscape of Japan is however complex in ecological and institutional structures. With a lack of coordination among regulatory agencies typical in the Japanese government, it is practically impossible to apply a comprehensive approach at a whole watershed scale. Although this institutional limitation constrains specific restoration actions to be localized at sites, important is to plan and design the actions within the watershed’s context. When addressing watershed degradation, a key issue is material cycling based upon the knowledge of watershed hydrology and geomorphology. This is because much of the ecological degradation in aquatic systems is the result of altered hydrology and material cycling. The Kushiro Mire is the largest wetland complex in Japan, spanning an area of 190 km2. The wetland’s watershed encompasses 2,500 km2, 23 times as large as the wetland area.

Figure 1. The location of the Kushiro Mire and land use of the Kushiro River Watershed.

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

211

The wetland has drawn further attention since 1980, when it was registered under “the Convention on Wetlands of International Importance Especially as Waterfowl Habitat”, commonly referred to as the Ramsar Convention. Located in the downstream end of the watershed, the Kushiro Mire has been subject to cumulative influences of upstream landuse (Figure 1). In this paper, we introduce the ecological status of the Kushiro Mire and its related watershed-scale degradation, describing our conceptual approaches to ecological restoration. Before specifically discussing restoration activities in Kushiro, one should have a clear view of what is ‘restoration’. Literally, it is a process of returning. Scientifically, it is defined as the act of returning an ecosystem that has physical, biological, and chemical characteristics in its pre-disturbance conditions (Lake, 2001; Jungwirth et al., 2002). In this definition restoration may imply the reestablishment of disturbance regimes native to the restoration site in which biological and physical processes can promote heterogeneous landscape and species diversity. However, where overwhelming human perturbations have caused irreversible degradation, ‘rehabilitation’ can be applied by improving structural and functional attributes to facilitate an occurrence of self-sustainable ecosystem (Wissmar and Beschta, 1998). In both cases, we emphasize that a highest priority is ‘passive restoration’ (Wissmar and Beschta, 1998). Passive restoration attempts removing human impacts preventing natural recovery of a damaged ecosystem and then let nature develop its own self-sustainable system. This is ‘passive’ because restoration practitioners must wait for a sufficient period of time to allow for the natural recovery processes. Most common misunderstanding is that active manipulations using construction equipments constitute restoration projects. This approach is referred to as ‘active restoration’, which should represent a last restoration alternative with a lowest priority. Passive restoration, on the contrary, eliminates limiting factors, such as bank revetments, drainage ditches, and fertilizer application, initiating natural recovery towards most stable ecosystems of rivers and wetlands. Humans only assist nature in returning to its pre-disturbance conditions. After implementing passive restoration, monitoring and observation of the natural recovery process should be conducted. If recovery trends are not ascertained in a self-sustaining manner, further actions including active restoration may be necessary. Thus, the principle underlying any restoration project should be ‘passive’, working based on the careful observation of ecosystem responses. Restoration should not be the creation of a new ecosystem that previously did not exist, and otherwise the restoration may exacerbate the extent of current degradation. In Europe a river restoration project aiming for “Space for rivers” has been implemented. Dykes and revetments were removed to restore its historical floodplain. This is an example of passive restoration, assuming that the river will regain its pre-disturbance conditions in natural patterns of hydrological and geomorphic processes across the floodplain. In addition, this restoration also addresses the effects of restoration on flood control and water resource management (Hansen, 2003; Geilen, 2003). As with in Europe, river restoration always involves

212

F. NAKAMURA

AND

Y.S. AHN

the issues of flood control and water resource management in Japan. It is necessary to balance between restoration and public interests in these issues. 2. THE HISTORICAL CHANGES OF THE KUSHIRO MIRE In the Kushiro Mire, wetland habitats have been degraded or lost by various human activities. This section summarizes the existing reports to describe watershed-scale landuse impacts on the wetland habitats. 2.1 Alterations of the Kushiro Mire associated with cumulative watershed-scale impacts A Landsat satellite image, taken on May 11th in 1992, is shown in Figure 2. Three black areas are lakes (Shirarutoro, Touro, and Takkobu). There are similar dark spots along the Kuchoro and Setsuri River, tributary to the Kushiro River. These dark color areas indicate a condition of high water tables (e.g., a floodplain), clearly shown during a snowmelt flooding. The area of inundation along the Kuchoro River is particularly large and dark in color. In extreme rainfall events, many of the Kushiro River tributaries become turbid in dark brown color. In the Kuchoro River, suspended sediment delivery in one flood event (from September 27th to October 1st in 1995; roughly 35 mm rainfall/hour) was estimated to be 1,120 tons (Nakamura et al., 2004a). The annual production of suspended sediment was 7,400 tons, and 95% of the yield was fine suspended sediment of which diameter was less than 0.1 mm. This type of fine suspended sediment load is called washload, which is naturally produced at mountain slopes and in floodplains and partly delivered into the ocean. In a meandering river, washload deposits on point-bars to form natural levees along the river course. It may however cause a serious problem in a channelized river that has high gradients. At the point where the straightened channel intersects with a downstream meandering reach, flood power is greatly reduced, causing fine sediment deposition there to cause riverbed aggradation (Emerson, 1971; Brookes, 1988). In the Kuchoro River, riverbed aggradation exceeding by about 2 m its original riverbed occurred at an entry point of the straightened channel into the wetland. This aggradation decreased the cross section of the river by half, when compared to that immediately after channelization. Overbank flooding with fine suspended sediment has likely occurred at the entry point in high flow events (Nakamura et al., 1997). Thus, the dark color areas in Figure 2 have been confirmed as flooding of turbid water with washload. An algorithm to estimate the degree of turbidity in the wetland was developed using water turbidity index (WTI) determined on Landsat satellite images and field verification of WTI (Kameyama et al., 2001). Using this algorithm, wetland alterations associated with washload were evaluated on a spatial and temporal scale; the area of turbid water has been expanding since 1980. The locations of turbid water flooding have also been changing in these years (Nakamura et al., 2004a).

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

213

Figure 2. A Landsat Thematic Mapper image of the Kushiro Mire, taken on May 11th, 1992. A solid line declineates the Kushiro Mire. Areas in dark gray or black color represent lakes and the wetlands of which water tables are near their surface.

To examine a relationship between watershed landuse and sedimentation rates in the wetland, radioactive fallout (Cesium-137) has been used (Mizugaki and Nakamura, 1999). Cesium-137 was released into the atmosphere during the period of frequent nuclear tests. The peak fallout in Japan occurred in 1963, roughly coinciding with the onset of growing agricultural landuse development. Cesium-137 is insoluble in water and tends to be adsorbed particularly onto fine sediment. Thus, by determining the depth of Cesium-137 peak fallout in wetland sediment, sedimentation rate during the agricultural development can

214

F. NAKAMURA

AND

Y.S. AHN

effectively be estimated. Analyses of sediment core samples indicated that fine suspended sediment has deposited about 200 cm near the Kuchoro River channel. The deposition was 160 and 100 cm at the distances of approximately 30 and 50 m from the channel, respectively. Contrastingly, sediment deposition was 40cm near the Chiruwatsunai River channel, of which upstream regions are relatively undeveloped. During the agricultural development, fine sediment deposition in the Kuchoro River has progressed at the rates of three to eight times as great as those in the Chiruwatsunai River and other regions (DeLaune et al., 1978; Johnston et al., 1984). This abnormal sedimentation rate in the Kuchoro River supports the overbank flooding containing washload detected in the satellite image analyses. Thus, upstream agricultural development in the last half-century and stream channelization have likely changed the sedimentation in the wetland. Plant responses to sediment deposition differ in species and at different life stages including germination, seedling establishment and growth (Jurik et al., 1994; Wang et al., 1994; Smith et al., 1995). In the Kushiro Mire hydrological alteration and nutrient-rich turbid water in the wetland due to watershed agricultural development may have resulted in vegetation change. Nakamura et al. (2002) examined forest stands and wetland soils in 15 to 17 quadrats, comparing between two streams with developed watersheds (the Kuchoro River and Setsuri River) and a stream with its undeveloped watershed (the Chiruwatsunai River). The result of a canonical correspondence analysis (CCA) indicated that water table variation, mean particle size, and electric conductivity were greater in the disturbed watersheds than undisturbed one. Water tables also tended to be lower in disturbed watersheds with Salix spp. as a dominant tree species. In contrast, soil moisture and organic material contents and water tables were greater in the undisturbed watershed (Chiruwatsunai). A dominant tree species was the Japanese alder (Alnus japonica Steud). Decreased reach length in a straightened stream channel usually results in magnified flood peaks with short rising rims, thereby increasing water table variations in riparian wetlands (Nakamura et al., 2002). In the Kushiro Mire increased flood power and turbid water flooding further enhanced coarse sediment deposition in the wetland, lowering ground water tables there and recharging ion rich ground water. Terrestrialization and nutrient enrichment favor the growth of Salix stands which rarely occurred in the wetland. Contrastingly, wetland soils with slow decomposition rates are distributed extensively in the Chiruwatsunai River because of its higher water levels. In these soils with excessive moisture contents, alder trees appear to grow in the place of Salix. In downstream portions of the Kushiro Mire, alder forest expansion has also been recognized in native fens dominated by a reed-sedge community. A recent field experiment found that wetland hydrology was an important parameter to explain this alder expansion (Nakamura, 2003). High water tables manipulated by a revetment appear to have affected the survival of alder trees. A detailed investigation

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

215

implicated that alder trees were killed at where water table reached 80 cm from the bottom of peat accumulation. However, this threshold water table might be deeper if the peat mats were floating (Figure 3).

Figure 3. Survival and the total basal area of Japanese alders in relation to increasing water tables. Water tables above the bottom of peat layers were averaged for the period between September and November in 1998. Survival rates were quantified based on the numbers of trees that have no winter buds on their stems (after Nakamura et al., 2003).

In flood events, stream water accompanying washload can be delivered to backwater swamps behind natural levees. Yachidamo (Fraxinus mandshurica var. japonica Maxim.) and Japanese elm (Ulmus davidiana var. japonica Nakai) often grow on the levees, and alders are common in the backwater swamps. However, as with the Kuchoro River and Setsuri River, wetlands in a developed watershed receive a large amount of sediment loading through extensive flooding of turbid water. Fine sediment deposition beyond natural processes can occur. Habitat conditions in such wetlands are likely to be altered, causing vegetation changes there. In the Kushiro Mire, alder forests have invaded into reed-sedge communities even in interior parts of the wetland. Where rapid sedimentation accelerates terrestrialization, the end-point of vegetation shift might be a Salix dominant forest. Figure 4 presents tree distribution in the Kushiro Mire, which was constructed in the development of an environmental information map described later in this paper.

216

F. NAKAMURA

AND

Y.S. AHN

In the construction, distortion in about 200 aerial photographs was first removed. Using a computer data processing with stereovision, trees with canopies greater than 1m diameters were extracted and shown against a Landsat image. Forest expansion is clearly visible along the edges of the wetlands, particularly at downstream ends of the Kuchoro River and Setsuri River. Rare species including the Japanese crane (Grus japonensis) and Siberian salamander (Salamandrella keyserlingii) often nest or lay eggs in reed-sedge dominated marshes. It has been concerned that the rapid forest expansion due to human disturbances reduces their breeding habitats.

Figure 4. The distribution of tree stands in and around the Kushiro Mire. After removing distortion in about 200 photographs, computer data processing and verification using stereovision extracted trees with canopies greater than 1m diameters. The data is shown against a Landsat image.

2.2 The historical changes of lake environment Eastern three lakes (the Takkobu Lake, Sirarutoro Lake, and Touro Lake) are a small lake into the Kushiro River, located in the eastern margin of the Kushiro Mire

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

217

(Figure 2). It was formed about 3,000 years ago, when the sea level descended because of sea regression. The main stream flows into three lakes, whereas the Kushiro River drains into three lakes under high flow condition because the elevation of the three lakes and the Kushiro River is almost equal. Therefore, when the water level of the Kushiro River rises, a part of water discharge of the Kushiro River flows into the three lakes (Hayashibara et al., 2003; Hokkaido Institute of Environmental Sciences, 2005). Thus three lakes are influenced by inflows from both its catchments and the Kushiro River drainage. Recent and current studies have consistently indicated that water quality and plant community structures in the three lakes are rapidly being altered. These three lakes tend to be in eutrophic conditions; lake phosphorous concentrations are elevating in all lakes. In the Takkobu Lake and Touro Lake, high total nitrogen concentrations (around 1.2 mg/L) and algae bloom in a summer have been observed (Takamura et al. 2003). Furthermore, chlorophyll a concentrations are increasing while species numbers of aquatic macrophytes (floating-attached, submersed, and floating-unattached plants) are clearly declining (Kadono et al., 1992). In addition, abundance of lake balls (Aegagropila linnaei Kutzing) and species numbers of aquatic insects has declined (Kimura and Ubukata, unpublished data). To compare the effects of watershed landuse on water quality, we plotted relationships of total nitrogen, total phosphorus and population for the three lakes (Takamura et al., 2003) against data produced by the International Lake Environmental Committee (2001) (Figure 5). The larger populations, the higher nutrient concentrations. However, nutrient concentrations in the three lakes have relatively higher value despite of the small population size.

Figure 5. Relationships between population and total nitrogen, total phosphorus in catchments. The nutrient concentrations of the three lakes published by Takamura et al. (2003) were added to the relationships derived from the International Lake Environmental Committee (2001).

218

F. NAKAMURA

AND

Y.S. AHN

It has already been suspected that wastewater from a campground and households were the sources of water quality degradation. Elevated concentrations of fine suspended sediment and nutrients may have reduced water transparency, causing the declining numbers of aquatic plant species. Not only these point sources, but also watershed-scale degradation of hydrology and water quality may be another factors that cause the alteration of lake ecosystems. In the Takkobu area, various ecosystems including wetlands, lakes, rivers, and upslope, occur together in a small watershed. Human influence started with deforestation in the Takkobu area began in the 1880s and intensified after 1898. Moreover, agricultural development, construction of drainage networks and roads were undertaken after 1940s. The sedimentation associated with flood event and landuse development is reflected in changes in the physical characters of lake sediment. For example, agricultural activity, deforestation, and road construction can lead to an increase in inorganic sediment (Gurtz et al., 1980; Kreutzweiser and Capell, 2001) and coarse sediment inflow (Walling et al., 1998; Owens et al., 1999), and these can be identified from changes in the physical characters in lake core sediment samples (Kim and Rejmánková, 2002). The changes in the physical characters were defined as a “signal”, which is a valuable time marker (Page et al., 1994; Walling et al., 2003). Lake Takkobu core samples contained two tephra layers and the signal of canal construction in 1898 (Ahn et al., in press). From the refractive indices of dehydrated glasses, the lower tephra layer was identified as Ko-c2 (1694) and the upper tephra layer as Ta-a (1739). A clear peak in the Cesium-137 concentration was detected at all the sampling points. The sediment yield averaged over the last 300 years for Takkobu Lake was reconstructed for four periods using the signal, tephra, and Cesium-137 as marker layers (Figure 6).

Figure 6. The average sediment yields reconstructed over the last 300 years for Takkobu Lake, using the signal, tephra, and Cesium-137 as marker layers.

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

219

The sedimentation rates from 1694–1739 and 1739–1898 reflect the natural sedimentation rates, and those from 1898–1963 and 1963–2004 indicate the rates after land use development. The period of 1898–1963 is characterized by deforestation and canal construction as well as initial agricultural development, and during the period after 1963 deforestation, ditch and road construction, and intensified agricultural development have continued. The average sediment yield under natural erosion condition for the first two periods was 226 tons/year in 1694– 1739 and 196 tons/year in 1739–1898. The development of the Takkobu watershed started in 1880s with partial deforestation and channelization, leading to an increased sedimentation yield 1,016 tons/year in 1898–1963. Continuous deforestation, channelization works, road construction, as well as agricultural development caused a further increase up to 1,354 tons/year in 1963–2004. Compared to the average natural sedimentation yield of 206 tons/year until 1898, initial watershed development accelerated lake sedimentation, indicated by the 5fold sediment yield. With increasing agricultural development since 1960s, sedimentation yields were highest for 1963–2004, 7-fold compared to natural conditions (Ahn et al., in press). Progressive development, timber harvesting, including agricultural conversion of wetlands and stream channelization, has directly influenced the wetlands. Concerning this watershed degradation in the Takkobu area, a local environmental conservation group and non-governmental organization, NPO Trust Sarun Kushiro, have been working on land acquisition in upland forests for restoration of native deciduous forests and ultimate protection of the lake ecosystem. 2.3 Declines of wetland habitat due to landuse development (particularly, in southern areas of the Kushiro Mire) In August 1920 a massive flood caused a great deal of damage on the Kushiro City. This experience initiated a sequence of stream channel manipulation for flood control followed by agricultural development. The Akan River, which formerly drained into the Kushiro River, was disconnected, and in 1921 the Kushiro River was partly channelized. Peat lands along the Kushiro River, which were previously covered by reed (Phragmites communis Trin.) and alders, were drained by agricultural conversion. After establishing the Hokkaido Regional Development Bureau (HRDB) in 1951, the government began agricultural development in the Kushiro Mire as “The 1st five-year program for Hokkaido Integrated Development”. HRDB reported that 52 km2 of the Kushiro Mire was converted for various purposes including agriculture and housing until 1996. This landuse conversion concentrated in southern portions of the wetland, where the Kushiro City was located. Many of the farmlands were later abandoned. The portion of the Kushiro Mire in Hirosato (260 ha) consists of previously abandoned farmlands, in which the Ministry of the Environment (MOE) has been conducting wetland restoration since 2002 as described later in this paper. Drainage ditches are still present in the old farmlands developed in late 1960s. As shown in Figure 4, alder forests have been invading into the wetland. Although we know this

220

F. NAKAMURA

AND

Y.S. AHN

forest expansion is not associated with fine sediment loading, it is still unknown whether the expansion is anthropogenic (revetment construction and agricultural land use) or natural succession. There are many unknown factors regarding to the recruitment habitat of the Japanese alder. Propagation by means of seedling establishment has not been observed in field survey; most of the trees continue vegetative growth by sprouting. In addition, because the wetland in Hirosato has no inflows from nearby streams, disconnected by revetments, sediment loading from upstream is unlikely to occur. 3. CONSERVATION MEASURES AND THE DEVELOPMENT OF INFORMATION DISSEMINATION STRATEGY To address these wetland and watershed degradation, HRDB established “The Committee for Conservation of the River Environment in Kushiro Mire (hereinafter, referred to as the Committee)” in 1999, and proposed 12 measures to conserve the river and wetland ecosystems in the Kushiro Mire in March 2001. The details are available in the Proposal for Conservation of the River Environment in Kushiro Mire. Subsequently, MOE has developed a restoration project, the Kushiro Ecosystem Restoration Project (hereinafter, the Project). Before implementing any restoration, it is important to evaluate current conditions of existing ecosystems (Figure 7). This preliminary assessment identifies degraded ecosystems to be restored and remaining intact ones to be preserved, providing a clear criterion in site selection. Intact ecosystems serve as a reference for understanding the processes and relationships between biotic and physical factors of naturally self-sustaining ecosystems. This understanding is essential for determining causes of the degradation and restoration strategies to allow for natural recovery to occur. For spatial analyses, geographic information systems (GIS) are effective to integrate a variety of data based on aerial photographs, topographic maps, satellite images, and land use information. Existing governmental reports and literature also provide valuable sources for the assessment. If any symptoms that indicate the deterioration of objective ecosystems could be found through the above screening process, we should examine more intensively to identify the controlling or regulating variables in the fields (Figure 7). Once we clarify these key variables deteriorating the ecosystems, we may proceed to set the target of the restoration project. The target ecosystem should be set with an agreement of local public and be feasible to be accomplished by restoration project, considering economical and social constraints. Reference site (intact site) or pictures prior to human disturbance will be instructive information to set the target. So-called ‘adaptive management’, that is using the experiment as management strategy, will be important in this stage, because we cannot predict perfectly the response of restored ecosystems (Figure 7). This process is not just a trial-and-error procedure; rather it should be the validation procedure with a clear hypothesis. If the hypothesis is validated by the experimental results, we can extend the techniques over a broad restoration area. A preliminary assessment in the Project identified ecosystems for potential preservation and restoration in the Kushiro Mire: 1) a largest fen in Japan; 2) raised bogs bisected by the Kushiro River revetment; 3) meandering reaches in which a

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

221

largest freshwater fish species, the Sakhalin taimen (Hucho perryi), survives; 4) wetland habitats for the Siberian salamander and Japanese crane; 5) spring habitats at the breaks of slopes for the Japanese crayfish (Cambaroides japonicus); 6) the eastern three lakes supporting lake ball populations; 7) native forests remained in the upstream watersheds.

Figure 7. General flow of restoration projects in Kushiro Mire.

Currently, although various scientists and agencies study the Kushiro Mire, much of their data and results are independently documented and not open to public. The Project has started a GIS database project for information dissemination, integrating this isolated information as an environmental information map. This map is temporarily available through the Internet (Ministry of the Environment, 2005). The goal of this GIS database project is to construct an environmental information map (hereinafter, the Map) such that 1) integrates spatial-scale information including topography, vegetation, and wildlife distribution, which allows to simultaneously identify these spatial attributes for a location of one’s interest; 2) enables temporal-scale analyses by which digitizing existing topographic maps and aerial photographs and identifying land use activities and vegetation; 3) is available to the public after excluding confidential data, such as the locations of rare species habitats or personal information. The Map also promotes information sharing through an interactive function between the web site and public; users can manipulate and analyse GIS data on the web and both download and upload GIS data files. The Map is so to speak an information infrastructure to support the Project based upon public consensus and scientific data.

222

F. NAKAMURA

AND

Y.S. AHN

4. THE PRINCIPLES OF PRESERVATION AND RESTORATION AND SPECIFIC MEASURES 4.1 Ameliorating watershed-scale human impacts Watershed-scale pollution management in Hokkaido is normally conducted by the HRDB Flood Control Division. Reporting its specific strategies and measures, Nakamura (2003) has discussed the principles of construction engineering, sediment hydraulics, and geomorphology, including two-dimensional flood simulation. This section of the present paper provides only a brief summary of these principles, which are related to the issues in the Kushiro Mire. It is noteworthy that the Committee developed numerical objectives for watershed pollution control, and presented specific measures to achieve these objectives. The long-term goal is to recover the conditions of the Kushiro Mire ecosystem that existed at the time of the Ramsar registration in 1980, when its watershed remained relatively undisturbed before extensive land use development (Nakamura et al., 2003). Furthermore, a more specific goal for the next 20 to 30 years is the reduction of watershed pollution loadings to the levels of 20 years ago, in order to protect wetland conditions existed in 2000 from further degradation. To achieve these goals, pollution loads were estimated using existing data and simulation models. For example, annual sediment load was 800 m3 in 1980 and is currently 1,400 m3. A specific objective for controlling sediment load was then determined as the reduction of the current sediment yield by 40 %. Likewise, 20 % reduction is the target value for a total nitrogen yield. Some of the committee members asserted that agricultural development already advanced in 1980, arguing to set the target values at those in older years (e.g., 1970). However, considering the feasibility of restoration goals, all committee members agreed with the above goals and target values. Proposed strategies to achieve the target values for watershed pollution controls include the provision of filtration grounds and ponds along stream channels and at the ends of drainages, and forest restoration in water source regions. The Committee is currently discussing specific locations and methods. 4.2 Restoration of the straightened river channel to meandering course The natural meandering rivers in the Kushiro Mire has been altered into the straighten channel from the marginal areas of the marsh. The main objectives of the channelization project were to develop pasture fields both side of the river channels and to convey floods safely. Although those objectives could be achieved, we have lost pristine river-floodplain landscapes and inhabiting wildlife species. A largest freshwater fish species, the Sakhalin taimen, is now in danger of extinction because of the channelization projects. As a pilot plan, a river section extending about 2 km of the Kushiro River in the Kayanuma area is planned to restore from a straightened channel to a meandering stream and floodplains (Figure 8). The artificial dike built at the right-side bank (left

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

223

side of the figure) will be removed to allow flooding over the floodplains. The abandoned meandering section created by channelization project will be reconnected with the main channel and the straightened channel will be partially buried with the dike sediment. The restoration project will be monitored and evaluated based upon the scientific research results before and after the project and comparison with reference section existing downstream reaches.

Figure 8. The restoration project of meandering river. River flows from up to bottom of the picture. The meandering section (right side) will be connected to the main channel.

4.3 Restoration programs to improve lake ecosystems Of the three eastern lakes, the Touro Lake and Takkobu Lake are of greater concerns (Takamura et al., 2003) because of algae blooms since 2000 and the rapid decline of aquatic plant diversity. The degree of eutrophication still remains low in the Takkobu Lake (Takamura et al., 2003), assuming greater possibility to prevent further degradation. Criticism against the Project included: restoration constitutes traditional public engineering works just for feeding money to construction industries; restoration is contradictory to the development activities destroying healthy ecosystems in elsewhere. To deal with these arguments, the MOE Committee developed a system of rule-based site selection to maintain the independency of site selection from political and institutional constrains. This rule-based process ensures the objectivity of site selection and establishes a clear rationale of restoration activities, which is

224

F. NAKAMURA

AND

Y.S. AHN

subject to public evaluation. To develop this site selection rule for the lake restoration in Takkobu, the scope of the restoration was first determined as the lake’s watershed. Sites for preservation and restoration were then identified on the watershed-scale GIS database. Although we consider applying this approach to the entire Kushiro River Basin including the Kushiro Mire, it is a challenging task. Unfortunately, GIS data inventory spanning 2,500 km2 of the basin is incomplete, and searching and compiling existing data and reports have been time consuming. Restoration activities in the Takkobu Lake and its watershed are at the stage of experiment (Nakamura et al., 2003). Based on the GIS analyses, three types of potential preservation or restoration sites were extracted in the Takkobu Lake Watershed with its area of about 42 km2: Type (1) delineates forest with species compositions that are resembled to historical stands and wetlands. Type (2) represents non-native forests that are close to the wetlands, including non-native forests, sparsely grown young forests, and secondary grassy fields. Type (3) includes unvegetated areas, such as bare grounds, roads, and clear-cut sites. Young silvicultural stands, farmlands, and secondary grassy fields adjacent to the unvegetated areas are also included to this type, if the site has steep slopes and close to the wetlands. The results of this extraction process are shown in Figure 9 (Nakamura et al., 2003). The total area of type (1) was 1,809 ha, accounting 43.0 % for the Takkobu Lake Watershed. These areas are not remaining pristine ecosystems but should be maintained their current conditions. With considering social and economic conditions, these areas will be protected from further alterations as much as possible. Type (2) sites occupied 13.1 % with 550 ha. A potential restoration measure for nonnative forests in type (2) is the restoration of native forests, and those for secondary fields and abandoned farmlands are tree planting and wetland restoration. Type (3) sites consisted 269 ha of the watershed, making up 6.4 %. The provision of forest buffer zones has been considered to control soil erosion. At present there are many issues to be discussed in evaluation and the accuracy of extraction process. This rule-based site selection should be further improved through findings in literature and field survey. 4.3.1 An approach to the restoration of native forests We consider that controlling pollution loadings impacting wetland ecosystems and restoring indigenous forests warrant the first priority in the Takkobu restoration. For selecting a reference site, GIS data and historical reports were used to examine the composition and structure of native forests, because few pristine forests currently exist in the field. This analysis indicated that deciduous forests dominated by the Japanese oak (Quercus crispula Blume) occurred on volcanic ash deposits, consisting mostly of the Takkobu Lake Watershed (Nakamura et al., 2003). On its adjacent mudstone layer in the south, coniferous-deciduous mixed forests were established. Existing forest stands that are closely resembled to these historical forests were then examined in the field. It was further revealed that Japanese oak dominant stands and Japanese elm stands (other co-dominant species include alders and Yachidamo) still exist on hillslopes and along stream channels, respectively. In

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

225

addition, relatively large size trees of these species occur sparsely across the Takkobu Lake watershed.

Figure 9. Areas prioritized for preservation or restoration in the Takkobu and adjacent watersheds, which were extracted based on vegetation, forest structure, and topography. Areas that do not satisfy any criterion were determined as “restoration sites of a lowest priority” (shown as blanks in the map; 1,647 ha, accounting 40 % for the classified area) (after Ministry of the Environment, 2005).

These remnants of native forests adjacent to the restoration sites serve as reference sites to be achieved by the restoration. To evaluate a recovery process after implementing restoration, indicative parameters include forest species composition, tree size, rates of sprouting, the volume of dead trees, seedling and sapling density, and the species composition and density of forest-dependent animals. Emphasizing passive restoration, factors preventing a natural recovery will be removed before implementing tree planting. The limiting factors in the Takkobu restoration include the infestation of Sasa species suppressing the germination and growth of young trees, deer (Cervus Nippon yesoensis) grazing on tree saplings, and

226

F. NAKAMURA

AND

Y.S. AHN

desiccation. To prescribe site-specific restoration actions, we will further identify 1) those forests that can likely recover by protecting them from future landuse activities, 2) those forests that can maintain natural regeneration by fallowing or excavating surface soils to disturb Sasa covers, 3) those that can recover by controlling deer grazing and desiccation, 4) those sites that can establish forests by tree planting, and 5) those sites that can only recover forest stands by active manipulation including weed control and construction of engineering structures. To avoid disturbing genetic resources, the MOE Committee developed a nursery program, which collects seeds nearby areas of the restoration sites and prepares seedlings for tee planting. 4.3.2 Collaboration with citizens To establish collaboration with a local environmental group, the NPO Trust Sarun Kushiro, MOE entrusted Trust Sarun with field survey in the Takkobu restoration. One of the criticisms against restoration by the governments is that the authority takes a primary leadership while “public participation” is superficial. Participation by the public has been effective in the Takkobu restoration at various stages of the restoration planning, extending from attending seminars and committee meetings to conducting field survey. In general, NGOs do not have political or institutional boundaries as the government and authorized agencies do. Therefore, the NGO’s participation in the Takkobu restoration enabled the planning at the watershed scale, irrespective of the artificial boundaries. Furthermore, ecological restoration requires a long-term planning for years to decades, which is hardly managed solely under an authority’s leadership but can be done with the collaboration with NGOs. Thus, social and political infrastructure is also important for ecological restoration. It is not all advantages in restoration under public leadership. Especially with limited resource information and restoration techniques, restoration planning may not be an easy task for private entities. In any restoration projects, it is important for both citizens and the governments to recognize their limited ability and to complement their limitation each other in developing effective restoration strategies to achieve the desired goals. 4.4 Hirosato Wetland Restoration The MOE Committee members all agreed that the goal of a wetland restoration project in Hirosato is the recovery of wetland landscape that existed before the agricultural development in the late 1960s. The conditions to meet for implementing the wetland restoration are to 1) restore a sustainable wetland ecosystem, 2) minimize the adverse impacts of restoration activities on adjacent wetland ecosystems and farmlands, 3) ensure the protection of local residents from flooding, soil erosion, and water quality degradation, and 4) avoid disturbing nesting Japanese cranes.

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

227

In most years a pair of cranes has nested and succeeded breeding in the wetland in Hirosato. It was concerned that the restoration would disrupt their nesting activities. This issue was debated among a conservation group, the Akan International Crane Center, the Committee members, and MOE. Lastly, all worked out an agreement that a great deal of careful attention must be made in any field survey and restoration actions; from the top of a crane truck with its height of 25 m, the conservation group oversees both surveyors and birds and determines their distance. Depending on the distance, surveyors on the ground need to change the locations and times of work. 4.4.1 Summary of a preliminary survey and its findings Six transects were established in a wetland restoration site, which has vegetation typical in Hirosato. At more than 100 plots over transects, vegetation (species and coverage) and physical and chemical parameters (ground water tables, concentrations of dissolved materials in soil) were inventoried. Using CCA, the relationships between vegetation and environmental gradients were examined (Nakamura, 2003). Dominant plant communities changed over the space from a relatively natural portion of the wetland, an abandoned farmland, and a transitional area between them. Representatives were the community of alder - slender sedge (Carex lasiocarpa), bluejoint reedgrass (Calamagrostis langsdorffii) and starwort (Stellaria radians), respectively. This vegetation shift in space was also correlated with the gradients of groundwater tables and nutrient concentrations in wet soils (positive correlations) and variation in groundwater tables (negative correlation). Particularly, the differences in groundwater tables was prominent among the natural area, old field, and transition between these; the plant communities in old field had extremely low in water tables and high in its variation (Figure 10). Hydrological characteristics at a larger scale than the project area were further examined. The cross sectional profile of groundwater table showed a bell curve, with a highest water table in the center of the wetland. Water table decreased rapidly towards the agricultural drainage and an old channel of the Setsuri River that has been disconnected from its main channel (Figure 11). In addition, water tables in the natural portions of the wetland ranged between zero and –0.4 m whereas those in the old field ranged from –0.3 to –2.0 m with its lowest value near the old channel. The wetland seems to be fed only by precipitation and has no inflows from nearby stream channels. Thus, wetland water can be drained by the agricultural drainage and old Setsuri River channel that has considerably low water tables due to disconnection from its main channel. The old farmland seems to be progressively desiccated (Yamada et al., 2004). Because sufficient amount of water usually can maintain wetland conditions (Wheeler and Shaw, 1995), the drier conditions of the Hirosato wetland definitely suppress the growth of wetland plants. The considerable vegetation change from the natural wetland portion to the old agricultural field reflects the change in the groundwater table, that is, desiccation. Therefore, the Hirosato wetland restoration focuses on this old farmland.

228

F. NAKAMURA

AND

Y.S. AHN

Figure 10. Biplot of mean water table (WL mean) and water table variation (WL sd), as well as its variation in dominant plant communities in the Hirosato wetland. Vertical and horizontal bars show the standard deviation of each mean. (after Nakamura et al., 2003)

Figure 11. Spatial distributions of (a) groundwater tables at an instance of measurement in June and (b) variation in groundwater tables during the period of measurement. (after Nakamura et al., 2003)

4.4.2 Approaches to restoration, the experimental design and implementation To ascertain a cause of wetland degradation in Hirosato, an experimental restoration has been conducted. The experiment constitutes analyses on both temporal and spatial scales; data will be compared between before and after treatment, and between reference (intact sites), control (restoration sites without treatment), and impact (restoration sites with treatment). In addition, replications to allow statistical

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

229

tests are established (Nakamura, 2003). Using this design, called BARCI (BeforeAfter-Reference-Control-Impact), restoration actions and their effectiveness to lead the wetland restoration toward its goal can be evaluated in a scientific framework. In searching a reference site for the old farmland, an aerial photographic interpretation determined that slender sedge - reed community was dominant prior to farming. Therefore, existing slender sedge - reed community that has not been affected directly by human activities was chosen in the Hirosato wetland as a reference site. An effective restoration action in the old farmland might be a recovery of wetland hydrology. The aforementioned preliminary analyses indicated that increasing groundwater tables was required for ameliorating the wetland desiccation to recover native wetland communities. Restoration alternatives for the wetland restoration include: water diversion from the Kushiro River to the the Setsuri River old channel; revetment construction at the agricultural drainage and the Setsuri River old channel to prevent draining wetland water; soil excavation in the old farmland to raise the groundwater levels. Considering water rights and impacts on adjacent existing farmlands, a presently possible action is soil excavation, which can also remove the filled materials in the agricultural conversion. Other restoration alternatives would be considered when some conditions, such as land acquisition and consensus among irrigators, are met. An experimental site for soil excavation was established in the old farmland. Within the experimental area, the effects of excavation on water quality, water table variation, and vegetation recovery should be assessed because it can cause strong disturbance in the nearby natural wetland. Concurrently, the MOE Committee has been discussing social and technical feasibility of water diversion from the Kushiro River because this is a most passive approach, requiring a minimal manipulation of the wetland. Resolution in the issues of water rights and flood control are being discussed. When concluding that the stream water diversion is not practical, the MOE Committee will pursuit revetment construction along the Setsuri River old channel and soil excavation. To examine the presence of viable seeds under the soil layer to be excavated, a seed bank test was performed using seedling emergence technique in soils at the three depths down to 1 m. No slender sedges and reeds but small numbers of other species and seedlings emerged (Nakamura et al., 2004b). Only a few seedlings of a wetland species, Harikougaizekishou (Juncus wallichianus), germinated in the soil at the depth between 25 and 35 cm. Thus, vegetation recovery after excavation will likely to be limited without seeding. Another study of the effectiveness of vegetation recovery with and without seeding will be necessary. The soil excavation experiment was begun in 2002. The depth of excavation was determined so that the water table could be equivalent to that in the reference site. This excavation test was designed to evaluate various alternatives in order to find a most appropriate method for vegetation recovery (Figure 12). For example, one experiment plot was excavated with a slope to test seedling emergence at various soil depths, and another was seeded with reeds. To minimize adverse impacts of excavation on the natural wetland, it was carried out in a winter when soil was frozen. For an access to the experimental site, a 1.5 km-long bridge was manually made by from ice in the winter, which melted away in the following spring.

230

F. NAKAMURA

AND

Y.S. AHN

Figure 12. An experimental excavation site in the Hirosato wetland.

4.4.3 Controversial alder forest control As previously mentioned, regeneration mechanisms of alder forests are largely unknown. In the Hirosato restoration site, it is important to determine the direct cause of forest expansion by revealing habitat conditions optimal for alder growth. However, as shown in Figure 10, Japanese alders and slender sedges co-occur as a dominant community in the interior parts of the wetland, sharing similar hydrologic conditions. Currently available data cannot explain habitat conditions supporting alder forest expansion. Evapotranspiration by alder trees is greater than that of reed dominant communities. Expansion of alder forest may increase evapotranspiration in the Hirosato wetland, contributing to further wetland desiccation. Controlling alder forest expansion may also be necessary for protecting one important native wetland species, the sphagnum moss (Sphagnum imbricatum) because the species is sensitive to a change in light and hydrologic conditions. MOE has been studying the effects of alder control, including cutting trees and sprouts, on vegetation under the trees, evapotranspiration and groundwater tables. We are aware of that cutting trees and sprouts can be a short-term solution but not be a causal treatment. It is also inappropriate to extend the area of alder control at present,

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

231

before its regeneration dynamics and optimum habitat conditions have been clearly understood. Even if the cause of alder forest expansion in Hirosato is human related (e.g., the disruption of water inflow by the revetment and drainage impact), the removal of revetment would be impossible because of the needs of flood control in existing agricultural land use. Given water diversion from the Kushiro River is a potential alternative action, the role of wetland hydrology in the alder forest expansion should be clarified at first. 5. IMPLICATIONS FOR FUTURE RESTORATION To be widely accepted by the public for restoration projects, it is necessary to establish consistent national policies in land conservation, agriculture, and natural resource conservation. The ecological meaning of restoration projects should be established in these policies in the past and future. With erroneously perceiving the purpose of ecological restoration, the government should not pursuit a large-scale engineering construction, which in the past often adversely impacted natural resources. However, the history of development at the expense of degradation or loss of pristine ecosystems should not be ignored. This experience could be helpful for evaluating the need of ecological restoration. The restoration actions should also be undertaken under public understanding and consensus. We emphasize the priority of extraction of remaining intact ecosystems and their preservation. Adjacent degraded ecosystems should be restored as much as possible, to protect a healthy sustainable ecosystem over a large area. Thus, in parallel to promoting restoration projects, greater areas of the nation’s pristine ecosystems should definitely be preserved. Information dissemination is essential to eliminate persistent public distrust of the government activities and to gain public understanding of restoration projects. The entire restoration process, including the development of goals, field experiments, and monitoring, should be open to the public, as the restoration progresses. Information sharing is essential in consensus building with local communities. A restoration goal can be developed only by understanding people’s values and needs in the local communities; for example, whether the public wants to recover a selfsustainable ecosystem, or favour a secondary ecosystem that is maintained by active management. A variety of desired endpoints are possible in restoration, which should be actively discussed by the public in the district.

REFERENCES Brookes, A. (1988). Channelized Rivers: Perspectives for Environmental Management. John Wiley & Sons, Chichester. DeLaune, R.D., Patrick Jr, W.H. and Buresh, R.J. (1978). Sedimentation rates determined by 137Cs dating in a rapidly accreting salt marsh. Nature, 275, 532−533. Emerson, J.W. (1971). Channelization: a case study. Science, 173, 325−326.

232

F. NAKAMURA

AND

Y.S. AHN

Geilen, N. (2003). Restoration project of river rhein in the netherlands. In Proceedings of International Symposium on Renaturalization of A River Basin by Eco-Compatible and Adaptive Management, pp. 86−99. Gurtz, M.E., Webster, J.R. and Wallace, J.B. (1980). Seston dynamics in southern Appalachian streams: Effect of clear−cutting. Canadian journal of fisheries and aquatic sciences, 37, 624−631. Hansen, H. (2003). The restoration project of Skjern River in Denmark. In Proceedings of International Symposium on Renaturalization of A River Basin by Eco-Compatible and Adaptive Management, pp. 74−85． Hayashibara, H., Kimura, N. and Ubukata, H. (2003). The density of aquatic insects in several lentic habitats in and around Kushiro marsh, Hokkaido, Japan: Comparisons among habitats and between years at ordinal and familial levels. Journal of Environmental Education, 6(2), 79−89. (in Japanese with English Abstract). Hokkaido Institute of Environmental Sciences. (2005). Lakes and Marshes in Hokkaido (revised ed.). Sapporo, Japan. (in Japanese). International Lake Environment Committee. (2001, June). World Lakes Database. Retrieved October 1, 2005, from the Web site: http://www.ilec.or.jp/database/database.html. Johnston, C.A., Bubenzer, G.D., Madison, F.W. and McHenry, J.R. (1984). Nutrient trapping by sediment deposition in a seasonally flooded lakeside wetland. Journal of Environmental Quality, 13, 283−290. Jungwirth, M., Muhar, S. and Schmutz, S. (2002). Re-establishing and assessing ecological integrity in riverine landscapes. Freshwater Biology, 47, 867−887. Jurik, T.W., Wang, S. and van der Valk, A.G. (1994). Effects of sediment load on seedling emergence from wetland seed banks. Wetlands, 14, 159−165. Kadono, Y., Nakamura, T., Watanabe, K. and Ueda, Y. (1992). Present state of aquatic macrophytes of three lakes in Kushiro Moor, Hokkaido, Japan. Journal of Phytogeography and Taxonomy, 40, 41−46. (in Japanese with English Abstract). Kameyama, S., Yamagata, Y., Nakamura, F. and Kaneko, M. (2001). Development of WTI and turbidity estimation model using SMA - Application to Kushiro Mire, eastern Hokkaido, Japan -. Remote Sensing of Environment, 77, 1−9. Kim, J.G. and Rejmánková, E. (2001). The paleoecological record of human disturbance in wetlands of the Lake Tahoe Basin. Journal of Paleolimnology, 25, 437−454. Kreutzweiser, D. P. and Capell, S. S. (2001). Fine sediment deposition in streams after selective forest harvesting without riparian buffers. Canadian Journal of Forest Research, 31, 2134−2142. Lake, P.S. (2001). On the maturing of restoration: Linking ecological research and restoration. Ecological Management & Restoration, 2, 110−115. Ministry of the Environment. (2005, September). Nature Restoration Project in Kushiro Shitsugen Wetland. from the Web site: http://kushiro.env.gr.jp/saisei/. (in Japanese). Mizugaki, S. and Nakamura, F. (1999). Sediment accumulation at the marginal areas of the Kushiro Mire, Hokkaido, estimated by Cs-137 fallout. Transactions, Japanese Geomorphological Union, 20, 97−112. (in Japanese with English Abstract). Nakamura, F. (2003). Restoration strategies for rivers, floodplains and wetlands in Kushiro Mire and Shibetsu River, nothern Japan. Ecology and Civil Engineering, 5(2), 217−232. (in Japanese with English Abstract). Nakamura, F., Jitsu, M., Kameyama, S. and Mizugaki, S. (2002). Changes in riparian forests in the Kushiro Mire, Japan, associated with stream channelization. River Research and Applications, 18, 65−79. Nakamura, F., Kameyama, S. and Mizugaki, S. (2004a). Rapid shrinkage of Kushiro Mire, the largest mire in Japan, due to increased sedimentation associated with land-use development in the catchment. Catena, 55, 213−229. Nakamura, F., Nakamura, T., Watanabe, O., Yamada, H., Nakagawa, Y., Kaneko, M., Yoshimura, N. and Watanabe, T. (2003). The current status of Kushiro Mire and an overview of restoration projects. Japanese Journal of Conservation Ecology, 8, 129−143. (in Japanese with English Abstract). Nakamura, F., Sudo, T., Kameyama, S. and Jitsu, M. (1997). Influences of channelization on discharge of suspended sediment and wetland vegetation in Kushiro Marsh, northern Japan. Geomorphology, 18, 279−289.

LANDSCAPE RESTORATION: A CASE PRACTICE OF KUSHIRO WETLAND, HOKKAIDO

233

Nakamura, T., Yamada, T., Nakagawa, Y., Kasai, Y., Nakamura, F. and Watanabe, T. (2004b). Ecological aspects of Hirosato restoration area in the Kushiro Mire: Impacts of artificial disturbances on the relationship between distribution of vegetation and hydrochemical environments. Ecology and Civil Engineering, 7(1), 53−64. (in Japanese with English Abstract). Owens, P.N., Walling, D.E. and Leeks, G.J.L. (1999). Use of floodplain sediment cores to investigate recent historical changes in overbank sedimentation rates and sediment sources in the catchment of the River Ouse, Yorkshire, UK. Catena, 36, 21−47. Page, M.J., Trustrum, N.A. and DeRose, R.C. (1994). A high resolution record of storm−induced erosion from lake sediments, New Zealand. Journal of Paleolimnology, 11, 333−348. Smith, M., Brandt, T. and Jeffrey, S. (1995). Effects of soil texture and microtopography on germination and seedling growth in Boltonia decurrens (Asteraceae), a threatened floodplain species. Wetlands, 15, 392−396. Takamura, N., Kadono, Y., Fukushima, M., Nakagawa, M. and Kim, B.H. (2003). Effects of aquatic macrophytes on water quality and phytoplankton communities in shallow lakes. Ecological Research, 18, 381−395. Walling, D.E., Owens, P.N., Foster, I.D.L. and Lees, J.A. (2003). Changes in the fine sediment dynamics of the Ouse and Tweed basins in the UK over the last 100−150 years. Hydrological Processes, 17, 3245−3269. Walling, D.E., Owens, P.N. and Leeks, G.J.L. (1998). The characteristics of overbank deposits associated with a major flood event in the catchment of the River Ouse, Yorkshire, UK. Catena, 31, 309−331. Wang, S., Jurik, T.W. and van der Valk, A.G. (1994). Effects of sediment load on various stages in the life and death of Cattail (Typha x Glauca). Wetlands, 14, 166−173. Wheeler, B.D. and Shaw, S.C. (1995). A focus on fens – controls on the composition of fen vegetation in relation to restoration. In B.D. Wheeler, S.C. Shaw, W.J. Fojt and R.A. Robertson (Eds.), Restoration of Temperate Wetlands (pp. 49−72). Ontario, John Wiley & Sons. Wissmar, R.C. and Beschta, R.L. (1998). Restoration and management of riparian ecosystems: a catchment perspective. Freshwater Biology, 40, 571−585. Yamada, T., Nakamura, T., Nakagawa, Y., Kamiya, Y., Nakamura, F. and Watanabe, T. (2004). Ecological aspects of Hirosato restoration area in the Kushiro Mire: Effect of pasture developments and river improvements on hydrochemical environments of groundwater. Ecology and Civil Engineering, 7(2), 37−51. (in Japanese with English Abstract).

CHAPTER 15

NON-INDIGENOUS PLANT SPECIES IN CENTRAL EUROPEAN FOREST ECOSYSTEMS

S. ZERBE Institute of Botany and Landscape Ecology, University Greifswald, Grimmer Str. 88, D-17487 Greifswald, Germany

Abstract. In the study presented here, the occurrence of non-indigenous vascular plant species in Central European forest ecosystems is outlined with regard to the current state and future perspectives. A focus is laid on Germany. This analysis is based on numerous ecological investigations on the species and ecosystem level. In total, 29 non-indigenous woody and 25 non-indigenous herb species are recorded within forest stands. Generally, there are much less exotic species, which grow on forest sites compared to habitats more or less strongly altered by human impact like, for example, agricultural and urban-industrial ecosystems. Most of the exotic species found in forests belong to the plant families Rosaceae, Pinaceae, and Asteraceae and have their origin in North America. A wide range of different natural and anthropogenic forest communities are invaded by non-indigenous plants, such as floodplain forests, mixed broad-leaved and conifer forests on nutrient-poor to nutrient-rich sites, and dry oak forests. The establishment of nonindigenous species in forests can affect the ecosystem considerably. This is shown, for instance, for the tree species Robinia pseudoacacia (alteration of the soil conditions) and Prunus serotina (influence on forest regeneration) and the herbs of the genus Fallopia (decrease of species richness on a local scale). Few nonindigenous species in forests, like for example Prunus serotina, can cause problems with regard to land use on a supra-regional scale. In conclusion, the management of non-indigenous species in forests on a local scale, in accordance with regional nature conservation objectives and considering socio-economic aspects might be useful. However, an assessment of a positive or negative impact of non-indigenous species on forest ecosystems has to be based on properly defined values.

1. INTRODUCTION The anthropogenic alterations of flora, ecosystems, and landscapes throughout the world are considered a part of the global change. Many research efforts focus on invasions by non-indigenous organisms because the subsequent biodiversity loss is

235 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 235–252. © 2007 Springer.

236

S. ZERBE

recognized as one of the biggest global environmental problems of our time (Vitousek et al., 1997; Sandlund et al., 1999). Additionally, the costs related to biological invasions, for example for the management of established and invasive nonindigenous species, can be considerably high for society (e.g., U.S. Congress, 1993). In Central Europe, invasions by non-indigenous plants are recorded and investi-gated along the whole range from anthropogenically strongly altered towards natural ecosystems (Kowarik, 2003). Thus, for example, settlements (Pyšek, 1993; Zerbe et al., 2003) and agricultural ecosystems like grassland and fields (Pyšek et al., 2002) have been studied with regard to plant invasions, both concentrating on invasive species as well as invaded habitats. Anthropogenic disturbances of sites and vegeta-tion are considered, additionally to others like dispersal abilities and vectors, one of the driving forces of spread and establishment of non-indigenous plant species (Trepl, 1983; Falinski, 1986; Kowarik, 1995; Rejmánek et al., 2005). Compared to non-forest habitats, there are much less comprehensive studies on plant invasions in Central European forest ecosystems (e.g., surveys from Lohmeyer and Sukopp, 1992 and Kowarik, 2003). Against the background that Central Europe is naturally a woodland area and that the percentage of forest cover is relatively high in many present-day landscapes, plant invasions in forest ecosystems have to be considered an important issue for science as well as for practice, such as forestry and nature conservation. Most studies on plant invasions in forests focus on certain species. Thus, for example, the annual herb Impatiens parviflora (Trepl, 1984) and the tree species Prunus serotina (Starfinger et al., 2003) and Pseudotsuga menziesii (Knoerzer, 1999) have been investigated in detail. Although there are comprehensive surveys on Central European forest vegetation (e.g., Oberdorfer, 1992; Ellenberg, 1996), studies with regard to plant invasions in forest ecosystems based on large vegetation data sets rarely exist. Accordingly, Zerbe and Wirth (2006) analyse a large database of vegetation samples taken in Central European pine forests in order to identify plant invasions and the ecological range of non-indigenous plant species in those forests. This paper will focus on the following questions: (1) Which non-indigenous vascular plant species invade Central European forests? (2) Which forest ecosystems are invaded? (3) How do these plant invasions affect the forest ecosystems and what problems can be identified with regard to land use and nature conservation, respectively? The study presented here, is based on a broad range of ecological investigations and findings on non-indigenous species in Central Europe, and in particular in Germany (e.g., Lohmeyer and Sukopp, 1992; Böcker et al., 1995; Hartmann et al., 1995; Pyšek et al., 1995, 2002; Starfinger et al., 1998; Kowarik, 2003). Here, only those species are considered, which have been introduced to Central Europe after 1,500 A.D. (neophytes according to Schroeder, 1969). 2. WHICH SPECIES INVADE CENTRAL EUROPEAN FORESTS? Compared to heavily disturbed ecosystems like those in urban-industrial areas with a high percentage of non-indigenous plant species (according to investigations

NON-INDIGENOUS PLANT SPECIES

237

from Pyšek, 1998a up to about 60 % exotic species in Central European urban floras), relatively few plant invasions have been recorded in forest ecosystems up to now (Table 1). A considerable number of woody species has been introduced for forestry purposes. Thus, the N American tree species Abies grandis, Picea pungens, P. sitchensis, Pinus strobus, Prunus serotina, Pseudotsuga menziesii, Quercus rubra, and the hybrids of the native Populus nigra L. and N American poplars (= P. x euramericana) have been afforested to a more or less large extent on Central European forest sites (Knoerzer and Reif, 2002). Additionally, tree species with E European and Asian origin, respectively, like Abies nordmanniana and Larix kaempferi are found in managed forests. For many of the mentioned tree species, such as for Pinus strobus (e.g., Zerbe, 1999), Prunus serotina (e.g., Starfinger, 1997), Pseudotsuga menziesii (Knoerzer, 1999), and Quercus rubra (e.g., Zerbe, 1999) spontaneous regeneration in forests has been recorded. Furthermore, a considerable number of exotic tree species occur in Central European urban or landscape parks, where they have often been introduced as ornamental plants (Lohmeyer and Sukopp, 2001). Thus, for example, Ahrens and Zerbe (2001) list some specimens of the N American Thuja plicata D. Don among other non-indigenous tree and shrub species in a park forest south of the city of Berlin, a forest which is built up by Acer platanoides L. and Tilia platyphyllos L. Most of the non-indigenous shrub species, which occur in Central European forests, escaped from cultivations in gardens or on green spaces in and around settlements. The N American Mahonia aquifolium and Symphoricarpos albus, for example, are dispersed by birds mostly into forests adjacent to cities (Kowarik, 1992; Adolphi, 1995; Auge, 1997). This holds also true for Amelanchier lamarckii (Schroeder, 1972). North American blueberries (hybrids of Vaccinium corymbosum and V. angustifolium) spread in NW Germany, where cultivars have been grown commercially (Schepker and Kowarik, 1998). Much less commonly, Spiraea alba (Lohmeyer and Sukopp, 1992; Kowarik, 2003) grows on forest sites. Among all non-indigenous species, the Central Asian annual herb Impatiens parviflora is considered the most successful with regard to plant invasions in Central European forest ecosystems. First records of its spontaneous spread from botanical gardens in Central Europe date back to the 1830ies (Trepl, 1984). Nowadays, this species is found in forests throughout Central Europe (cp. distribution map for Germany from Bundesamt für Naturschutz, 2005). Additionally, also the occurrence of the non-indigenous annual and perennial herbs Conyza canadensis, Fallopia div. spec., Helianthus tuberosus, Heracleum mantegazzianum, Impatiens glandulifera, Lysichiton americanus, and Solidago canadensis has been recorded in forests. Lohmeyer and Sukopp (1992, 2001) list several additional species as so-called agriophytes, which are non-indigenous species not only found on anthropogenic sites but are also considered a part of the natural vegetation in Central Europe. Thus, Allium paradoxum, Aster novi-belgii, Claytonia sibirica, Eranthis hyemalis, Iris versicolor, Ornithogalum nutans, Scilla siberica, Scutellaria altissima, S. columnae, and Tulipa sylvestris contribute to the non-indigenous annual and perennial herbs and Alnus rugosa to the exotic trees found in Central European forests.

Table 1. Survey of non-indigenous vascular plant species (woody and herb species) which have been recorded in Central European forests on a local or regional (○) and supra-regional (●) scale with information on the plant family, origin, and forest communities in which they occur; x = most frequent non-indigenous species in Germany as stated by Kowarik (2003), including all habitats; information mostly based on the comprehensive surveys from A,B: Lohmeyer and Sukopp (1992, 2001), C: Knoerzer and Reif (2002), D: Kowarik (2003), and E: Zerbe and Wirth (2006). Non-indigenous species

Plant family

Origin

Abies grandis (Dougl.) Lindl.

Pinaceae

N America

Abies nordmanniana (Stev.) Spach

Pinaceae

SE Europe

Acer negundo L.

Aceraceae

N America

Aesculus hippocastanum L.

Hippocastanaceae

SE Europe

Simaroubaceae

E Asia

Betulaceae

N America

Rosaceae

N America

Occurrence in forests

Frequency in forests3

Selected references

Trees and shrubs

Ailanthus altissima (Mill.) Swingle Alnus rugosa (Du Roi) Sprengl. Amelanchier alnifolia (Nutt.) Nutt.

mixed forests on moist sites under a broad range of soil nutrient conditions mixed forests on dry to moist sites floodplain forests, pine forests floodplain forests, broadleaved slope forests dry oak forests, floodplain forests

○

C

○

C

●

A, D

●x

A

○

A, D

mire forests

○

B

pine forests

○

B, D, E

Table 1 (cont.) Non-indigenous species

Frequency in forests3

Selected references

Plant family

Origin

Occurrence in forests

Rosaceae

N America

oak forests on acid sites, mire forests

●

A, D

Trees and shrubs (cont.) Amelanchier lamarckii Schroeder Amelanchier spicata (Lamk.) C. Koch Cornus stolonifera Michx. Cotoneaster horizontalis Decne. Laburnum anagyroides Med. Larix kaempferi (Lamb.) Carr.

Rosaceae

N America

acid oak forests

○

A

Cornaceae

N America

mire forests

○

A

pine forests

○

A

○

A, D

●

C

Ligustrum vulgare L. Mahonia aquifolium (Pursh) Nutt. Parthenocissus inserta (Kerner) Fritsch Physocarpus opulifolius (L.) Maxim. Picea pungens Engelm.

Rosaceae

E Asia

Fabaceae

S Europe

Pinaceae

E Asia

Oleaceae

S Europe, Asia

pine forests

○

A, D, E

Berberidaceae

N America

pine forests, dry oak and beech forests

●

A, D, E

Vitaceae

N America

floodplain forests

○

A

Rosaceae

N America

floodplain forests

○

A

Pinaceae

N America

open mixed forests under various site conditions

○

C

dry oak forests open mixed forests on moist and oligotrophic sites

Table 1 (cont.) Non-indigenous species

Frequency in forests3

Selected references

Plant family

Origin

Occurrence in forests

Picea sitchensis (Bong.) Carr.

Pinaceae

N America

mixed forests on moist and oligotrophic sites

○

C

Pieris floribunda Benth. et Hook.

Ericaceae

N America

mire forests

○

A

Pinus strobus L.

Pinaceae

N America

mixed coniferous forests on acid sites

●

B, C, D, E

Populus x euramericana (Dode) Guinier

Salicaceae

N America1

floodplain forests

●

A, D

Prunus serotina Ehrh.

Rosaceae

N America

●x

A, D, E

Pseudotsuga menziesii (Mirb.) Franco

Pinaceae

N America

●

B, C, D

Quercus rubra L.

Fagaceae

N America

pine forests

●x

B, C, D, E

Robinia pseudoacacia L.

Fabaceae

N America

pine forests, dry forests, floodplain forests floodplain forests pine forests, floodplain forests

●x

A, D, E

Trees and shrubs (cont.)

Spiraea alba Du Roi Symphoricarpos albus (L.) Blake Hybrids of Vaccinium corymbosum L. and Vaccinium angustifolium Ait.

Rosaceae

N America

Caprifoliaceae

N America

Ericaceae

N America1

pine and oak forests on acid sites mixed broad-leaved and conifer forests on acid sites

oligotrophic pine forests and mire forests

○

A, D

○x

A, D, E

○

B, D

Table 1 (cont.) Occurrence in forests

Selected references

Plant family

Herbs Allium paradoxum (M.B.) G. Don Aster novi-belgii L. Bidens frondosa L.

Liliaceae

W Asia

floodplain forests

○

A

Asteraceae Asteraceae

N America N America

○ ○x

A D

Claytonia sibirica L

Caryophyllaceae

N America

floodplain forests floodplain forests oak forests on acid sites

○

A

Asteraceae

N America

pine forests

○x

A, D, E

Ranunculaceae

SE Europe

Conyza canadensis (L.) Cronquist Eranthis hyemalis (L.) Salisb. Fallopia x bohemica (Chrtek et Chrtková) J.P. Bailey Fallopia japonica (Houtt.) Ronse Decr. Fallopia sachalinensis (F. Schmidt) Ronse Decr. Helianthus tuberosus L. Heracleum mantegazzianum Somm. et Lev. Impatiens capensis Meerb. Impatiens glandulifera Royle

Origin

Frequency in forests3

Non-indigenous species

floodplain forests

○

A

Polygonaceae

2

E Asia

floodplain forests

○

B, D

Polygonaceae

E Asia

floodplain forests

●x

A, D

Polygonaceae

E Asia

floodplain forests

●x

A, D

floodplain forests

○

x

A, D

x

A, D B A, D

Asteraceae

N America

Apiaceae

W Asia

floodplain forests

●

Balsaminaceae Balsaminaceae

N America S Asia

swampy forests floodplain forests

○ ●x

Table 1 (cont.) Non-indigenous species

Plant family

Origin

Impatiens parviflora DC.

Balsaminaceae

Central Asia

Iris versicolor L. Lupinus polyphyllus Lindl. Lysichiton americanus Hultén and St. John Ornithogalum nutans L. Rudbeckia laciniata L. Scilla siberica Andr. Scutellaria altissima L. Scutellaria columnae All.

Iridaceae Papilionaceae

N America N America

Araceae

N America

Liliaceae Asteraceae Liliaceae Lamiaceae Lamiaceae

SE Europe N America E Europe SE Europe SE Europe

Occurrence in forests

Frequency in forests3

Selected references

Herbs (cont.) beech forests, floodplain forests, pine forests, swampy forests mire forests dry oak forests swampy forests

●x

A, D

○ ○

A D

○

B, D

floodplain forests ○ A, D floodplain forests ○ A floodplain forests ○ A dry broad-leaved forests ○ A oak forests on acid sites ○ A floodplain forests, pine x A, D, E Solidago canadensis L. Asteraceae N America ○ forests x A, D Solidago gigantea Ait. Asteraceae N America floodplain forests ○ Tulipa sylvestris L. Liliaceae SE Europe floodplain forests ○ A 1 hybrids with parents from N America; 2 hybrid with parents from E Asia; 3 assessment on the basis of published studies in Central Europe with a focus on Germany

NON-INDIGENOUS PLANT SPECIES

243

Most of the non-indigenous woody species in Central European forests (Table 1) belong to the plant families Rosaceae and Pinaceae, each with 24 %. About 30 % of the non-indigenous herbs are part of the Asteraceae. This is in accordance with results from Pyšek (1997) who found that this plant family is over-represented among aliens compared to other plant families in a global perspective (also see Pyšek, 1998b). Accordingly, Pyšek (1997) states that the plant species of this family are remarkably successful as invaders in terms of dispersal and establishment. About 60 % of all recorded non-indigenous vascular plant species in forests have their origin in N America (Table 1). This does not reflect the general trend in the Central European flora with a higher percentage of non-indigenous species from other parts of Europe (e.g., S Europe) and from Asia (Kowarik, 2003: Figure 1). The survey given in Table 1 for non-indigenous woody (29) and herb species (25), which have been found in Central European forests on a local, regional, and supra-regional scale, is thought as a minimum list. Here, only those species are presented which have been found within forest stands. If all sites or vegetation structures within wooded landscapes were taken into account, such as forest paths, clear-cuts, and small forest mires for example, other species might add to the number of non-indigenous species in forests. Thus, Lohmeyer and Sukopp (1992) point out that clear-cuts in woodland areas can also be habitats for non-indigenous plant species. Examples are given with Conyza canadensis (see Table 1 for forests), Epilobium ciliatum Raf. (origin: N America), and Erechtites hieracifolia (L.) Raf. (origin N and S America). Furthermore, Dostálek (1997) mapped non-indigenous plants, like the North American Rudbeckia laciniata along roads through a woodland area of the Orlické mountains in the Czech Republic. Similar observations of non-indigenous plants along forest roads and paths made Schepker (1998) in NW Germany (e.g., Heracleum mantegazzianum). These occurrences along forest roads and paths might reflect one possible way of (mostly anthropogenic) dispersal and invasions into forest ecosystems. Additionally, many of the non-indigenous herbs are found in floodplain forests (see Lohmeyer and Sukopp, 1992 and Table 1), a phenomenon which also indicates a way of dispersal and introduction into natural vegetation along rivers and streams (Pyšek and Prach, 1994). 3. WHICH FOREST ECOSYSTEMS ARE INVADED? There is comprehensive knowledge on the forest types, which are invaded by Impatiens parviflora. This species is found in various beech forest communities on meso- to eutrophic sites throughout Central Europe (Trepl, 1984; Zerbe, 1999; Oberdorfer, 2001). According to the large vegetation data set compiled by Oberdorfer (1992) for S Germany, Impatiens parviflora also commonly occurs in floodplain forests. However, there are only few records of this species on sites with stagnating wetness on which Alnus glutinosa forests grow (Zerbe and Vater, 2000), thus indicating an ecological limitation of occurrence on wet sites. Derived from the ecological indicator values given by Ellenberg et al. (1991) for Impatiens parviflora, this species preferably grows on sites with intermediate light supply and soil moisture conditions, respectively, and relatively high nitrogen availability.

244

S. ZERBE

As shown in Table 1, the forest types which are invaded by vascular plant species range from broad-leaved to conifer, from dry to wet, from oligotrophic to nutrient-rich, and from natural to anthropogenic (e.g., plantations) forests. The broad range of invaded forest ecosystems has also been revealed by Kowarik (1995) on a regional scale (city of Berlin) and by Pyšek et al. (2002) on a supra-regional scale (Czech Republic). Both studies show the relatively high percentage of non-indigenous plant species in floodplain forests compared to other forest types. With a focus on natural and anthropogenic pine forests, we analysed a data set of about 2,300 vegetation samples from NE Germany with regard to the occurrence of non-indigenous plant species (Zerbe and Wirth, 2006). Out of a total of 362 taxa recorded in these pine forests along a broad range of soil and climate conditions, only 12 non-indigenous species, including trees, shrubs, annual and perennial herbs, and one bryophyte were found. These exotic species in pine forests commonly grow on sites with relatively high nitrogen availability and soil pH (Figure 1). In general, species-rich forests on nutrient-rich sites seem to be invaded more often by nonindigenous plant species than forests on nutrient-poor acid sites. This is in accordance with the findings of Huennecke et al. (1990), Hobbs and Huennecke (1992), McIntyre and Lavorel (1994), Stohlgren et al. (1999), Deutschewitz et al. (2003), and Cassidy et al. (2004), who point out a positive effect of habitat disturbance and nutrient availability on plant invasions. In particular, atmospheric nutrient depositions, a widespread phenomenon in Central Europe (Hüttl, 1998), can affect the upper soils of forests, thus enhancing the establishment of non-indigenous plant species (Zerbe and Wirth, 2006). However, there are some non-indigenous species with a relatively broad ecological range, such as Prunus serotina and Quercus rubra. Both species quite commonly occur in various pine forest communities with the exception of pine forests on very acid, nutrient-poor, and wet sites (Figure 1). It is evident that some forest types are rarely or even not invaded by nonindigenous species. Thus, our analysis (Zerbe and Wirth, 2006) revealed no plant invasions on nutrient-poor, acid forest mires with species like Eriophorum vaginatum L., Ledum palustre L., Sphagnum L. div. spec., and Vaccinium oxycoccus L. (Fig. 1: community # 19). Similar findings were made by Chmura et al. (2005) in S Poland. Reasons for this observation could be that (1) near-natural forests (e.g., forest mires) are less susceptible for plant invasions than anthropogenic ones, (2) there are limitations in the ecological range of the non-indigenous plant species which have been introduced to Central Europe up to now (present-day exotic species pool), which excludes a possible invasion of certain forest types (e.g., mires), and (3) there are limitations in dispersal into these forests. Although anthropogenic disturbances might enhance the establishment of nonindigenous species like it has been shown, for example, for the city of Berlin by Kowarik (1995) by taking all plant communities into account, there is no evidence that near-natural forests are resistant against plant invasions. Impatiens parviflora, for example, is established on a broad range of near-natural broad-leaved forests throughout Europe.

NON-INDIGENOUS PLANT SPECIES

245

Figure 1. Occurrence of non-indigenous vascular plant species in pine forest ecosystems of NE Germany on a broad range of sites. The environmental conditions of the different pine forest communities (clusters # 1 to 23) were assessed by means of ecological indicator values of the species present according to Ellenberg et al. (1991) for vascular plants and Benkert et al. (1995) for bryophytes (for the methodological approach see Ellenberg et al., 1991 and Dupré and Diekmann, 1998). Medians of the indicator values for light, moisture, soil reaction, and nitrogen were computed; the values are expressed on a 1 to 9 scale, i.e. the higher the value, the higher the species’ demand for the particular factor (from Zerbe and Wirth, 2006).

246

S. ZERBE

Among the non-indigenous herb species recorded in Central European forests (Table 1) there are mostly species, which have a relatively high light demand. Thus, those species are not able to grow in forest communities or forest succession stages with a dense canopy cover (e.g., old-growth beech forests) and are limited to relatively open forests. However, few studies focus on ecological limitations of nonindigenous plant species in forests like it was done for pine forests in NE Germany by Zerbe and Wirth (2006). Kowarik (2003) comes to the conclusion that the relatively low number of nonindigenous plant species in Central European forests compared to non-forest ecosystems is mostly due to limited dispersal into woodland. 4. ECOLOGICAL EFFECTS AND PROBLEMS OF PLANT INVASIONS IN FOREST ECOSYSTEMS It has been well investigated for Robinia pseudoacacia that the introduction and establishment of a non-indigenous plant can have strong influence on ecosystems. Due to the enrichment of the soil with nitrogen by its litter, this tree can completely alter the site conditions and the vegetation on formerly nutrient-poor sites towards nutrient-rich conditions with an accumulation of organic matter (Hoffmann, 1961). This was shown, for example, by Kowarik (1992) who compared acid, nutrient-poor grassland without Robinia pseudoacacia with stands dominated by R. pseudoacacia under formerly similar site conditions in the city of Berlin. Within the Robinia stands, nitrophytic species increased in frequency and abundance and species richness decreased. Forest succession was also influenced by enhancing the establishment and growth of broad-leaved trees with a relatively high nutrient demand, such as species of the genus Acer. Consequently, Robinia pseudoacacia can be a problem in nature conservation with regard to the protection of nutrientpoor vegetation and land-use types, respectively (e.g., Paar et al., 1994). Prunus serotina is considered to inhibit forest regeneration, in particular the rejuvenation of native trees (Spaeth et al., 1994; Schepker, 1998). Additionally, it has been revealed that species richness decreases as a consequence of a dense cover of Prunus serotina under an open pine canopy (Schepker, 1998; Starfinger et al., 2003). As these findings have mainly been recorded for anthropogenic pine forests (conifer plantations) in the Central European lowlands, it may be concluded that Prunus serotina is just a stage within the succession towards more natural broadleaved forests like it is known for indigenous short-lived tree species (for Sorbus aucuparia L. and Betula pendula Roth; see Zerbe, 2001 and Kreyer and Zerbe, 2006). Starfinger et al. (2003) document the invasion history and perception or use of this non-indigenous tree species in Central Europe and come to the conclusion that ‘the mere presence of P. serotina in forests in Central European lowlands does not justify an eradication campaign on the basis of its adverse effects on species conservation goals’ and ‘P. serotina as an ‘aggressive invader’ of forest ecosystems is mostly a symptom of preceding silvicultural practice’. Knoerzer (1999) considers the N American Pseudotsuga menziesii a problem with regard to habitat protection in SW Germany. This non-indigenous tree successfully regenerates on dry rocky mountain sites with a unique vegetation

NON-INDIGENOUS PLANT SPECIES

247

structure. On oligotrophic wooded slopes, the height growth exceeds that of the native trees (e.g., Abies alba Mill., Picea abies (L.) Karst., Pinus sylvestris L., Quercus petraea Liebl.), which build up these mixed oak forests. Additionally, due to the alterations of the organic layer by the litter of Douglas fir, a change of the original vegetation can be observed (Zerbe, 1999; Zerbe et al., 2000). The decrease of species richness on a local scale has been revealed in dominant stands of non-indigenous plant species, in particular with large leaves as a consequence of light competition. This is documented, for instance in stands with Heracleum mantegazzianum (Pyšek and Pyšek, 1995) and non-indigenous Fallopia species (Kowarik, 2003). However, these dominant stands rarely occur within forests, but are more commonly found on anthropogenically disturbed sites in urbanindustrial areas or the agricultural landscape. In Białowieża Forest (E Poland), Falinski (1986) recorded an increase in biomass of the herb layer due to the presence of Lupinus polyphyllus compared to plots without this non-indigenous species. It can be concluded that non-indigenous plant species in forests can affect the ecosystem by - Changing the abiotic site conditions such as the nitrogen availability (e.g., Robinia pseudoacacia) or light conditions on the forest floor (e.g., Lysichiton americanus, Prunus serotina), - Increasing the biomass of the herb layer due to nitrogen enrichment (e.g., Lupinus polyphyllus), - Altering the state of biodiversity such as the decrease of species richness, e.g. by the development of dense stands (e.g., Prunus serotina) or the establishment of the legume Robinia pseudoacacia as well as the increase of species richness by contributing positively to the forest species pool (e.g., Impatiens parviflora), - Influencing forest succession by, e.g. decelerating forest regeneration with native species (e.g. Prunus serotina), and - Changing the composition of the native vegetation to a large extent (e.g., Pseudotsuga menziesii, Robinia pseudoacacia). According to an investigation by Kowarik and Schepker (1998) on the attitude and perception of non-indigenous species by public authorities (e.g., nature conservation, forestry, and water management) in NW Germany, vegetation changes as a consequence of plant invasions are perceived as most important conflict. 5. EMERGING FOREST ECOSYSTEMS WITH NON-INDIGENOUS SPECIES? If the abiotic site factors have been changed irreversibly and/or species and populations have been lost (e.g., after peat mining, deposition of man-made substrates, and as a consequence of excavations) or introduced (e.g., by planting non-indigenous trees), new nature can develop which is described by Hobbs et al. (2005) as “emerging ecosystems”. This holds in particular true for strongly degraded landscapes like mining areas, military training areas, quarries, or urban-industrial areas, where neither natural conditions nor any state of the historical cultural landscape can be regenerated.

248

S. ZERBE

So what about emerging forest ecosystems in Central Europe as a consequence of biological invasions in forests? Up to now, there is no evidence that the Central European woodland vegetation will profoundly change to a large extent due to the introduction and establishment of non-indigenous species. Impatiens parviflora, for example, now has its niche in the herb layer of broad-leaved forests, thus enhancing species diversity but not changing the native forest vegetation on the community level. The widespread establishment of this species is hardly considered a problem with regard to socio-economics or nature conservation (Kowarik and Schepker, 1998). Nevertheless, new forests with non-indigenous plants have developed on the local and regional scale on urban-industrial sites (Kowarik and Körner, 2005). In particular, on those sites where buildings were destroyed during World War II (e.g. Kohler and Sukopp, 1964; Kowarik, 1995) or where industrial areas have been abandoned (e.g., Rebele and Dettmar, 1996; Keil, 2005) non-indigenous trees have been established and form new forest communities, e.g. Robinia pseudoacacia forests. According to Kowarik and Körner (2005), a ‘new wilderness’ develops, which opens new perspectives for urban forestry. On a regional or local scale, new forest ecosystems can also evolve outside settlements, in particular if keystone species (according to Mills et al., 1993 and Jordán et al., 1999) such as Robinia pseudoacacia are introduced. This N American tree species has been established within Central European settlements (e.g., Kohler and Sukopp, 1964; Kowarik, 1992) as well as in woodland areas (e.g., Jurko and Kontris, 1982; Kowarik, 1990; Wilmanns and Bogenrieder, 1995). At sites where native, shade tolerant species like Acer spec., Fagus sylvatica L., or Picea abies are not able to grow and probably would out-compete Robinia pseudoacacia (e.g., warm and dry slopes in the Rhine valley; Wilmanns and Bogenrieder, 1995; see also Klauck, 1986), this tree can build up forests with a relatively open canopy. Due to its ability to live in symbiosis with nitrogen-fixing bacteria and thus accumulate organic matter on formerly nutrient-poor sites, it can profoundly change the soil conditions. Consequently, the whole forest vegetation is influenced by this species, forming new forest communities like the Chelidonio-Robinietum (Jurko, 1963) or the Sambucus nigra-Robinia pseudoacacia community, respectively (Klauck, 1986). Based on the broad ecological knowledge, which has been gathered on Robinia pseudoacacia in Central Europe (Böhmer et al., 2001; Kowarik, 2003), Kowarik (2003: p161) concludes that the forest succession of Robinia stands towards other possible communities is still an open question. With the ongoing transformation of anthropogenic forests (in particular conifer monocultures with Pinus sylvestris and Picea abies) towards natural broad-leaved forests with native beech (Fagus sylvatica) and oak (Quercus petraea and Q. robur L.) in Central Europe (Olsthoorn et al., 1999; Klimo et al., 2000; Zerbe, 2002), most of the non-indigenous species most probably might not be able to compete successfully in the natural forest vegetation. This is due to the shady site conditions, e.g. in beech forest, where light demanding species like Conyza canadensis, Robinia pseudoacacia, Solidago canadensis (Table 1) cannot grow. The occurrence of these species is mostly restricted to anthropogenic forests with an open canopy (e.g., pine and oak forests).

NON-INDIGENOUS PLANT SPECIES

249

On a local scale, non-indigenous species can alter forest vegetation and biodiversity in a considerable way as it was shown above for, e.g. non-indigenous Fallopia species and Pseudotsuga menziesii. If this is considered a problem in terms of changing the native vegetation a necessity for human response might be derived. However, as the example of Prunus serotina shows, the management success in order to control the biological invasion is limited. According to an investigation from Schepker (1998) in NW Germany, the management success (control by mechanical and chemical means) is given with only about 30 %. The discussion on positive or negative impact of an invading non-indigenous species is often controversial due to divergent underlying values with regard to nature conservation and environmental protection, socio-economics, or recreation. Additionally, an assessment of this impact depends on values that are often not properly defined (Starfinger et al., 2003). In conclusion, the management of nonindigenous species in forests on a local scale, in accordance with regional nature conservation objectives and considering socio-economic aspects might be useful. Then, however, a continuous monitoring of the control success is necessary. ACKNOWLEDGEMENTS I am indebted to Ingo Kowarik and Herbert Sukopp (both Berlin) for valuable comments on the manuscript. REFERENCES Adolphi, K. (1995). Neophytische Kultur- und Anbaupflanzen als Kulturflüchtlinge des Rheinlandes. Nardus, 2, 1-272. Ahrens, S. and Zerbe, S. (2001). Historische und floristisch-vegetationskundliche Untersuchungen im Landschaftspark Märkisch-Wilmersdorf als Beitrag zur Gartendenkmalpflege. Landschaftsentwickl. u. Umweltforschg., 117, 1-158. Auge, H. (1997). Biologische Invasionen: Das Beispiel Mahonia aquifolium. In R. Feldmann, K. Henle, H. Auge, J. Flachowsky, S. Klotz and R. Krönert (Eds.), Regeneration und nachhaltige Landnutzung: Konzepte für belastete Regionen (pp. 124-129). Springer, Berlin. Benkert, D., Erzberger, P., Klawitter, J., Linder, W., Linke, C., Schaepe, A., Steinland, M. and Wiehle, W. (1995). Liste der Moose von Brandenburg und Berlin mit Gefährdungsgraden. Verh. Bot. Ver. Berlin Brandenbg., 128, 1-70. Böcker, R., Gebhardt, H., Konold, W. and Schmidt-Fischer, S. (1995). Gebietsfremde Pflanzen. Auswirkungen auf einheimische Arten, Lebensgemeinschaften und Biotope, Kontrollmöglichkeiten und Management. ecomed, Landsberg. Böhmer, H.-J., Heger, T. and Trepl, L. (2001). Fallstudien zu gebietsfremden Arten in Deutschland. UBATexte, 13, 1-126. Bundesamt für Naturschutz (Ed.) (2005). Daten und Informationen zu Wildpflanzen und zur Vegetation Deutschlands. Retrieved Dec. 2005 from http://www.floraweb.de. Cassidy, T.M., Fownes, J.H. and Harrington, R.A. (2004). Nitrogen limits an invasive perennial shrub in forest understory. Biol. Invasions, 6, 113–121. Chmura, D., Woźniak, G., Śliwińska-Wyrzychowska, A. (2005). The participation of invasive plants in the degeneration of coniferous forests of the Silesian Upland. In A. Brzeg and M. Wojterska (Eds.), Coniferous forests vegetation – differentiation, dynamics and transformations. Wyd. Nauk. UAM, Ser. Biologia, 69, 339-342. Deutschewitz, K., Lausch, A., Kühn, I. and Klotz, S. (2003). Native and alien plant species richness in relation to spatial heterogeneity on a regional scale in Germany. Global Ecol. and Biogeography, 12, 299–311.

250

S. ZERBE

Dostálek, J. (1997). Changes in the distribution of synanthropic plants along roads in the territory of the Protected Landscape Area of the Orlické hory Mts. Příroda 10, 159-182. Dupré, C. and Diekmann, M. (1998). Prediction of occurrence of vascular plants in deciduous forests of South Sweden by means of Ellenberg indicator values. Appl. Veg. Sci., 1, 139-150. Ellenberg, H. (1996). Vegetation Mitteleuropas mit den Alpen. 5th ed., Ulmer, Stuttgart. Ellenberg, H., Weber, H.E., Düll, R., Wirth, V., Werner, W. and Paulissen, D. (1991). Zeigerwerte von Pflanzen in Mitteleuropa. Scripta Geobot. 18, 1–247. Falinski, J.B. (1986). Vegetation dynamics in temperate lowland primeval forests. Geobotany, 8, 1-537. Hartmann, E., Schuldes, H., Kübler R. and Konold, W. (1995). Neophyten. Biologie, Verbreitung und Kontrolle ausgewählter Arten. ecomed, Landsberg. Hobbs, R.J., Arico, S., Aronson, J., Baron, J.S., Bridgewater, P., Cramer, V.A., Epstein, P.R., Ewel, J.J., Klink, C.A., Lugo, A.E., Norton, D., Ojima, D., Richardson, D., Sanderson, E.W., Valladares, F., Vilá, M., Zamora, R. and Zobel, M. (2005). Emerging Ecosystems: Theoretical and Management Aspects of the New Ecological World Order. Retrieved Jan. 2005 from http://www.unesco.org/mab/EE/ hobbs.pdf. Hobbs, R.J. and Huenneke, L.F. (1992). Disturbance, diversity, and invasions: Implications for conservation. Conservation Biol., 6, 324–337. Hoffmann, G. (1961). Die Stickstoffbindung der Robinie (Robinia pseudoacacia L.). Arch. Forstwesen, 10, 627-631. Huennecke, L.F., Hamburg, S.P., Koide, R., Mooney, H.A. and Vitousek, P.A. (1990). Effects of soil resources on plant invasion and community structure in Californian serpentine grassland. Ecol., 71, 478–491. Hüttl, R. (1998). Neuartige Waldschäden. Ber. u. Abh. Berlin-Brandenburg. Akad. d. Wissensch., 5, 131215. Jordán, F., Takács-Sánta, A. and Molnár, I. (1999). A reliability theoretical quest for keystones. Oikos, 86, 453-462. Jurko, A. (1963). Die Veränderung der ursprünglichen Waldphytozönosen durch die Introduktion der Robinie. Ceskoslov. Ochr. Prír., 1, 56-74. Jurko, A. and Kontris, J. (1982). Phytocoenological and ecological characteristics of acacia-woods in the little Carpathians. Biológia, 37, 67-74. Keil, A. (2005). Use and perception of post-industrial urban landscapes in the Ruhr. In I. Kowarik and S. Körner (Eds.), Wild Urban Woodlands. New Perspectives for Urban Forestry (pp. 117-130). Springer, Berlin, Heidelberg, New York. Klauck, E.J. (1986). Robinien-Gesellschaften im mittleren Saartal. Tuexenia, 6, 325-334. Klimo, E., Hager, H. and Kulhavý, J. (Eds.) (2000). Spruce monocultures in Central Europe – problems and prospects. Eur. For. Inst. Proc., 33, 1-208. Knoerzer, D. (1999). Zur Naturverjüngung der Douglasie im Schwarzwald. Diss. Bot., 306, 1–283. Knoerzer, D. and Reif, A. (2002). Fremdländische Baumarten in deutschen Wäldern. Fluch oder Segen? In I. Kowarik and U. Starfinger (Eds.), Biologische Invasionen: Herausforderung zum Handeln? Neobiota, 1, 27–35. Kohler, A. and Sukopp, H. (1964). Über die soziologische Struktur einiger Robinienbestände im Stadtgebiet von Berlin. Sitzungsber. Ges. Naturforsch. Freunde N.F., 4, 74-88. Kowarik, I. (1990). Zur Einführung und Ausbreitung der Robinie (Robinia pseudoacacia L.) in Brandenburg und zur Sukzession ruderaler Robinienbestände in Berlin. Verh. Berliner Bot. Ver., 8, 33-67. Kowarik, I. (1992). Einführung und Ausbreitung nichteinheimischer Gehölzarten in Berlin und Brandenburg und ihre Folgen für Flora und Vegetation. Ein Modell für die Freisetzung gentechnisch veränderter Organismen. Verh. Bot. Ver. Berlin Brandenbg., Suppl., 3, 1–188. Kowarik, I. (1995). On the role of alien species in urban flora and vegetation. In P. Pyšek, K. Prach, M. Rejmánek and M. Wade (Eds.), Plant invasions. General aspects and special problems (pp. 85-103). SPB Acad. Publ., Amsterdam. Kowarik, I. (2003). Biologische Invasionen: Neophyten und Neozoen in Mitteleuropa. Ulmer, Stuttgart. Kowarik, I. and Körner, S. (Eds.) (2005). Wild urban woodlands. New perspectives for urban forestry. Springer, Berlin, Heidelberg, New York. Kowarik, I. and Schepker, H. (1998). Plant invasions in Northern Germany: Human perception and response. In U. Starfinger, K. Edwards, I. Kowarik and M. Williamson (Eds.), Plant invasions: Ecological mechanism and human responses (pp. 109-120). Backhuys Publ., Leiden.

NON-INDIGENOUS PLANT SPECIES

251

Kreyer, D. and Zerbe, S. (2006). Short-lived tree species and their role as indicators for plant diversity in pine forests. Restor. Ecol., 14, 137-147. Lohmeyer, W. and Sukopp, H. (1992). Agriophyten in der Vegetation Mitteleuropas. Schriftenr. Vegetationskde., 25, 1-185. Lohmeyer, W. and Sukopp, H. (2001). Agriophyten in der Vegetation Mitteleuropas. 1. Nachtrag. In D. Brandes (Ed.), Adventivpflanzen. Beiträge zu Biologie, Vorkommen und Ausbreitungsdynamik von Archaeophyten und Neophyten in Mitteleuropa. Braunschw. Geobot. Arb., 8, 179-220. McIntyre, S. and Lavorel, S. (1994). Predicting richness of native, rare, and exotic plants in response to habitat and disturbance variables across a variegated landscape. Conserv. Biol., 8, 521–531. Mills, L.S., Soulé, M.E. and Doak, D.F. (1993). The keystone-species concept in ecology and conservation. BioScience, 43, 219-224. Oberdorfer, E. (1992). Süddeutsche Pflanzengesellschaften, Part IV: Wälder und Gebüsche. 2nd ed., Fischer, Jena. Oberdorfer, E. (2001). Pflanzensoziologische Exkursionsflora für Deutschland und angrenzende Gebiete. 8th ed., Ulmer, Stuttgart. Olsthoorn, A.F.M., Bartelink, H.H., Gardiner, J.J., Pretzsch, H., Hekhuis, H.J. and Franc, A. (1999). Management of mixed-species forest: silviculture and economics. IBN Scient Contrib., 15, 1-389. Paar, M., Tiefenbach, M. and Winkler, I. (1994). Trockenrasen in Österreich. Bestandsaufnahme und Gefährdung. UBA-Rep., 107, 1-86. Pyšek, P. (1993). Factors affecting the diversity of flora and vegetation in Central European settlements. Vegetatio, 106, 89–100. Pyšek, P. (1997). Compositae as invaders – better than others? Preslia, 69, 9-22. Pyšek, P. (1998a). Alien and native species in Central European urban floras: a quantitative comparison. J. Biogeogr., 25, 155-163. Pyšek, P. (1998b). Is there a taxonomic pattern to plant invasions? Oikos, 82, 282-294. Pyšek, P. and Prach, K. (1994). Plant invasions and the role of riparian habitats – a comparison of four species alien to Central Europe. J. Biogeogr., 20, 413-420. Pyšek, P., Prach, K., Rejmánek, M. and Wade, M. (Eds.) (1995). Plant Invasions. General Aspects and Special Problems. SPB Acad. Publ., Amsterdam. Pyšek, P. and Pyšek, A. (1995). Invasion of Heracleum mantegazzianum in different habitats in the Czech Republic. J. Veg. Sci., 6, 711-718. Pyšek, P., Sádlo, J. and Mandák, B. (2002). Catalogue of alien plants of the Czech Republic. Preslia, 74, 97-186. Rebele, F. and Dettmar, J. (1996). Industriebrachen. Ökologie und Management. Ulmer, Stuttgart. Rejmánek, M., Richardson, D.M., Pyšek, P. (2005). Plant invasions and invasibility of plant communities. In E. van der Maarel (Ed.). Vegetation Ecology (pp. 332-354). Blackwell Science Ltd. Sandlund, O.T., Schei, P.J. and Viken, A. (1999). Invasive Species and Biodiversity Management. Kluwer, Dordrecht. Schepker, H. (1998). Wahrnehmung, Ausbreitung und Bewertung von Neophyten. Eine Analyse der problematischen nichteinheimischen Pflanzenarten in Niedersachsen. ibidem, Stuttgart. Schepker, H. and Kowarik, I. (1998). Invasive North American blueberry hybrids (Vaccinium corymbosum x angustifolium) in Northern Germany. In U. Starfinger, K. Edwards, I. Kowarik and M. Williamson (Eds.), Plant Invasions: Ecological Mechanisms and Human Responses (pp. 253260). Backhuys Publ., Leiden. Schroeder, F.-G. (1969). Zur Klassifizierung der Anthropochoren. Vegetatio, 16, 225-238. Schroeder, F.-G. (1972). Amelanchier-Arten als Neophyten in Europa. Abh. Naturwiss. Ver. Bremen, 37, 287-419. Spaeth, I., Balder, H. and Kilz, E. (1994). Das Problem mit der Spätblühenden Traubenkirsche in den Berliner Forsten. Allg. Forstzeitschr., 49, 234-236. Starfinger, U. (1997). Introduction and naturalization of Prunus serotina in Central Europe. In J.H. Brock, M. Wade, P. Pyšek and D. Green (Eds.), Plant Invasions: Studies from North America and Europe (pp. 161–171). Backhuys Publ., Leiden. Starfinger, U., Edwards, K., Kowarik, I. and Williamson, M. (Eds.) (1998). Plant Invasions: Ecological Mechanism and Human Responses. Backhuys Publ., Leiden. Starfinger, U., Kowarik, I., Rode, M. and Schepker, H. (2003). From desirable ornamental plant to pest to accepted addition to the flora? – the perception of an alien tree species through the centuries. Biol. Invasions, 5, 323-335.

252

S. ZERBE

Stohlgren, T.J., Binkley, D., Chong, G.W., Kalkhan, M.A., Schell, L.D., Bull, K.A., Otsuki, J., Newman, G., Bashkin, M. and Son, Y. (1999). Exotic plant species invade hot spots of native plant diversity. Ecol. Monogr., 69, 25–46. Trepl, L. (1983). Zum Gebrauch von Pflanzenarten als Indikatoren der Umweltdynamik. Sitzungsber. Ges. Naturf. Freunde N.F., 23, 151-171. Trepl, L. (1984). Über Impatiens parviflora DC. Agriophyt in Mitteleuropa. Diss. Bot., 73, 1–400. U.S. Congress O.T.A. (1993). Harmful Non-Iindigenous Species in the United States. U.S. Government Printing Office, Washington D.C. Vitousek, P.M., D’Antonio, C.M., Loope, L.L., Rejmanek, M. and Westbrooks, R. (1997). Introduced species: a significant component of human-caused global change. New Zealand J. Ecol., 21: 1-16. Wilmanns, O. and Bogenrieder, A. (1995). Die Entwicklung von Flaumeichenwäldern im Kaiserstuhl im Laufe des letzten halben Jahrhunderts. Forstarchiv, 4, 167-174. Zerbe, S. (1999). Die Wald- und Forstgesellschaften des Spessarts mit Vorschlägen zu deren zukünftigen Entwicklung. Mitt. Naturwiss. Mus. Aschaffenbg., 19, 1–354. Zerbe, S. (2001). On the ecology of Sorbus aucuparia (Rosaceae) with special regard to germination, establishment, and growth. Pol. Bot. J., 46, 229-239. Zerbe S. (2002). Restoration of natural broad-leaved woodland in Central Europe on sites with coniferous forest plantations. For. Ecol. Manage., 167, 27-42. Zerbe, S., Brande, A. and Gladitz, F. (2000). Kiefer, Eiche und Buche in der Menzer Heide (NBrandenburg). Veränderungen der Waldvegetation unter dem Einfluß des Menschen. Verh. Bot. Ver. Berlin Brandenbg., 133, 45-86. Zerbe, S., Maurer, U., Schmitz, S., Sukopp, H. (2003). Biodiversity in Berlin and its potential for nature conservation. Landsc. Urban Plan., 62, 139–148. Zerbe, S. and Vater, G. (2000). Vegetationskundliche und standortsökologische Untersuchungen in Pappelforsten auf Niedermoorstandorten des Oberspreewaldes (Brandenburg). Tuexenia, 20, 55-76. Zerbe, S. and Wirth, P. (2006). Ecological range of invasive plant species in Central European pine (Pinus sylvestris L.) forests. Ann. For. Sci., 63, 189-203.

CHAPTER 16

TRAFFIC MORTALITY, ANALYSIS AND MITIGATION Effects of road, traffic, vehicle and species characteristics

F. VAN LANGEVELDE1, C. VAN DOOREMALEN1, C.F. JAARSMA2

1

Resource Ecology Group, Wageningen University, Bornsesteeg 69, 6708 PD Wageningen, The Netherlands; 2Land Use Planning Group, Wageningen University, Gen. Foulkesweg 13, 6703 BJ Wageningen, The Netherlands

Abstract. This chapter focuses on the impact of transportation on wildlife. Measures are frequently applied to mitigate these impacts. Most measures involve technical devices that change the road characteristics. However, also other measures may reduce traffic mortality, such as reduction of traffic volume or speed, and periodic closing of roads. For effectively applying these mitigating measures, insight in the effects of road and traffic characteristics on traffic mortality is needed. We argue that the success of measures that mitigate habitat fragmentation by roads drastically increases when minor roads are integrated in transportation planning. We discuss a strategy based on the concept “traffic-calmed rural areas”, where the effects of minor and major roads are not mitigated separately, but in coherence. To enable transportation planning to include the impacts on wildlife in the planning process, we present a traversability model derived from traffic flow theory that can be used to determine the probability of successful road crossings of animals based on the relevant road, traffic, vehicle and species characteristics. We apply this model in a case study in The Netherlands to evaluate different scenarios. Several levels of traffic calming are compared with the autonomous development, which shows that traffic calming can drastically reduce traffic mortality.

1. INTRODUCTION “Transport’s impact on the environment is multifaceted and can be severe” (Button and Nijkamp, 1999; p xiii). A wide range of impacts has been studied, such as traffic safety (Elvik and Vaa, 2004), noise (Lee et al., 1998), emissions (Sharma and Khare, 2001), and vehicular-related air quality (Sharma et al., 2004). In this chapter, we focus on the impact of transportation on wildlife. Infrastructure is one of 253 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 253–272. © 2007 Springer.

254

F. VAN LANGEVELDE ET AL.

the principal causes for the fragmentation of their habitat (Andrews, 1990; Forman and Alexander, 1998; Spellerberg, 1998; Trombulak and Frissell, 2000). There are at least four negative effects of traffic on wildlife (Van Langevelde and Jaarsma, 2004): destruction or alteration of habitat due to construction, disturbance of habitat along the road or railway (noise, vibrations, car visibility, etc.), barriers created by the road or railway (increased resistance for movements), and barriers by traffic (collision risk during crossing). The first two directly affect the habitat of the species. They result in a decline of habitat area or strips along the road with lower quality of habitat. The latter two effects have an impact on individuals. These four effects may have implications for population dynamics and community structure near the road. We mainly focus on the mortality due to traffic on roads. Here, we define the traversability of a road as the probability of successfully crossing that road by an individual. Measures are frequently applied to reduce traffic accidents (Garret and Conway, 1999; Singh and Satheesan, 2000) and protect biodiversity (Van Bohemen, 1998; Trombulak and Frissell, 2000). Mitigation measures include keeping wildlife off the road (e.g., fences: Romin and Bissonnette, 1996; Putman, 1997), providing alternative routes (e.g., fauna passages and ecoducts: Jackson and Griffin, 1998; Keller and Pfister, 1997) or reducing the risk of collisions (e.g., highway lighting or mirrors: Romin and Bissonette, 1996; Putman, 1997). Most measures involve technical devices that change the road characteristics. However, also other measures may reduce traffic mortality, such as reduction of traffic volume or speed, and periodic closing of roads (during the night or a specific season). For effectively applying mitigating measures that reduce traffic mortality at locations where no passageways or fences are constructed, insight in the effects of road and traffic characteristics on traffic mortality is needed (Andrews, 1990; Kirby, 1997; Forman and Alexander, 1998). We argue that the success of measures that mitigate habitat fragmentation by roads drastically increases when the minor roads are integrated in the planning of the measures (Jaarsma and Willems, 2002a; Van Langevelde et al., in prep). In this chapter, we discuss a strategy based on the concept of a traffic-calmed rural area (Jaarsma, 1997), where the effects of the minor and the major roads are not mitigated separately, but in coherence. For a sound planning and design of measures to mitigate environmental impacts of transportation, quantitative models are available that calculate impacts such as noise and pollution. These models enable to predict the impacts of (alternative) plans for infrastructure in quantitative terms such as numbers of hindered people. This is in contrast with impacts of these plans on plants and animals, where at most the acreage of destroyed habitat by the road construction can be quantified. However, the impacts on wildlife movement, essential for both daily and seasonal activities of individuals of a species and generally affecting its population dynamics, remain unknown. To compare alternative solutions for the road network with respect to wildlife movement, a more quantitative approach is desirable. We developed a model for successful wildlife crossings of a road (Van Langevelde and Jaarsma, 2004; Jaarsma et al., in prep). In this chapter, we present this model and review relevant road, traffic, vehicle and species characteristics to estimate the probability

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

255

of successful road crossing (based on Van Langevelde and Jaarsma, 2004 and revised in Jaarsma et al., in prep.). In contrast to other recent studies on traffic mortality (Van Langevelde and Jaarsma, 1997; Jaarsma and Van Langevelde, 1997; Hels and Buchwald, 2001; Clevenger et al., 2003; Jaeger and Fahrig, 2004), we explicitly derived the model from traffic flow theory. The aim of this chapter is to combine knowledge on movements of animals with knowledge on headway distributions on roads in a traversability model and to illustrate its value in a case study. Therefore, we first shortly review literature on environmental impacts of roads and traffic on wildlife. A model to estimate traffic mortality is discussed based on theory of traffic flows. This model can estimate the change of the number of traffic victims among traversing animals before and after mitigating measures, and/or for alternative infrastructure network solutions relative to the present situation. We then apply the model in a case study in The Netherlands to evaluate different scenarios. In this chapter, we summarize our earlier work. 2. ROADS AND TRAFFIC: IMPACTS ON WILDLIFE Seiler (2002) and Forman et al. (2003) review a wide range of direct and indirect effects of infrastructure on nature. Indirect effects follow the construction of new roads or railways, for example, consequent industrial development or changes in human settlement and land use patterns. We focus on effects that directly impact wildlife and their habitat, as these are usually the most relevant to the transport sector. Direct ecological effects are caused by the physical presence of the infrastructure section and its traffic flows. Generally accepted is the next categorization into five major categories: habitat loss, corridor habitats, disturbance and edge effects, barrier effects, and mortality (Van Langevelde and Jaarsma, 2004). Figure 1 presents a schematic representation. Together, these effects result in habitat fragmentation, i.e., the subdivision of natural habitats into small and isolated patches. It leads to conditions whereby species, as well as their populations, are endangered and extinctions might occur. Habitat fragmentation has been recognised as a significant cause for the decline of biodiversity (Seiler, 2002; Forman et al., 2003), and are thus a major concern for society. 2.1 Habitat loss Habitat loss is an inevitable consequence of infrastructure construction. A part of the surface of a new road is paved and therefore it is consequently lost as natural habitat for plants and animals. Motorways may consume more than 10 hectares of land per kilometre of road. Rural highways and minor rural road occupy (much) less area per kilometre, but collectively they comprise at least 95% of the total road stock. Hence, their cumulative effect in the landscape can be considerably greater (Jaarsma and Willems, 2002a; Seiler, 2002). One should realise that associated features, such as verges, slope cuttings, parking places, and service stations etc., also claim space. So, the total area designed to transport is several times larger than the paved surface. It is estimated to be 5-7% of the land surface in rather densely

256

F. VAN LANGEVELDE ET AL.

populated Western-European countries such as The Netherlands, Belgium or Germany (Jedicke, 1994).

Figure 1. Schematic representation of the five direct ecological effects of infrastructure: habitat loss (land uptake, habitat transformation), corridor habitats (corridor, conduit), disturbance and edge effects (avoidance, pollution, predation), barrier effects (by unsuitable habitat/disturbances, repelled by traffic or road characteristics, physical hindrances), and traffic mortality. Together these impacts lead to the fragmentation of habitat (source: Seiler, 2002; p 32).

For Sweden, where transportation infrastructure is sparser, roads and railways are estimated to cover about 1.5% of land surface (Seiler and Erikson, 1997). The USA devotes about 0.45% of its land area to roads, based on the average road density of about 0.75 km/km2 (Forman et al., 2003). These authors estimate that adding the right-of-way of these roads would roughly double the amount of land devoted to roads. They state that even this crude estimate is a significant underestimation because it excludes private roads in sub-urban areas as well as driveways and parking areas. 2.2 Corridor habitats Road verges considerably vary between different landscapes and countries. Despite verges are highly disturbed environments, numerous inventories indicate the great potential of verges to support a diverse range of plant and animal species (Munguira and Thomas, 1992; Seiler, 2002; Forman et al., 2003). As well as providing a habitat for wildlife, verges may also serve as a conduit for species movement for both generalist species that are tolerant of disturbance and ‘unwanted’ or invasive species spreading into the surrounding habitats. “The overall corridor function of infrastructure verges will most likely be influenced by the ecological contrast between the vegetation/structure in the corridor and the surrounding habitat” (Seiler, 2002; p 41).

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

257

2.3 Disturbance and edge effects Disturbance and edge effects mainly result from pollution of the environment due to infrastructure construction and use. Physical disturbance appears during construction activities, when soil, relief and groundwater flows change, and alter the vegetation. In forested areas, the clearance of a road (i.e., the distance from the road to dense vegetation) changes microclimatic conditions up to 30 metres from the edge of this road (Mader, 1984). Hydrological impacts even may include a much longer distance. Chemical pollutants such as road dust, salt, heavy metals, fertilisers and toxins largely contribute to the disturbance and edge effects. Most of these pollutants accumulate in the close proximity of the infrastructure. Seiler (2002) mentions several studies observing direct effects on vegetation and animals at distances over several hundreds metres away. Where tranquillity is perceived as an increasingly valuable resource (Gillen, 2003), traffic noise is one of the major polluting factors. It is questionable whether wildlife is similarly stressed by noise as humans. However, timid species might interpret traffic noise as an indicator of the presence of humans and consequently avoid noisy areas (Seiler, 2002). Seiler also mentions some studies on traffic noise avoidance for elk, caribou and brown bear. Birds appear to be especially sensitive to traffic noise. For The Netherlands, Reijnen et al. (1995) developed a simple model predicting the distance over which breeding bird populations of woodland birds and grassland birds might be affected by traffic noise. Their model is based on the observed relationship between noise burden and bird densities. In a Swedish study (Helldin and Seiler, 2003), however, these findings could not be verified. This study concluded that habitat changes as a consequence of road construction under some conditions could be more important than traffic noise. 2.4 Barrier effects The barrier effect of infrastructure is the reduction of the number of animal movements crossing this infrastructure. It results from a combination of disturbance, avoidance effects (such as traffic noise, vehicle movement, pollution and human activity) and physical hindrances (such as the infrastructure surface, ditches and fences). The clearance of the infrastructure and the open verge character may also act as a barrier to many species, especially small ones (Oxley et al., 1974). Depending on the species, the number of successful crossings is a fraction of the number of attempted movements. Some species may not experience any physical or behavioural barrier at all, whereas others may not even approach the road (Seiler, 2002). Most infrastructure barriers do not completely block animal movements, but reduce the number of crossings significantly (Mader, 1984; Merriam et al., 1989). “The fundamental question is this: how many successful crossings are needed to maintain habitat connectivity” (Seiler, 2002; p 45). To answer this question, knowledge is needed on (1) movements of specific species in a fragmented landscape, and (2) the chance on a successful road crossing for those species that actually cross the road.

258

F. VAN LANGEVELDE ET AL.

2.5 Mortality Despite millions of individual animals are killed on infrastructure each year, this traffic mortality is not considered as a severe threat to population survival for most common species (Seiler, 2002). In contrast to predation, traffic mortality is, however, non-compensatory and will kill a constant proportion of a population. Traffic mortality is, therefore, one of the major death causes for many species in human-dominated landscapes (Groot-Bruinderink and Hazebroek, 1996; Forman and Alexander, 1998; Philcox et al., 1999; Trombulak and Frissell, 2000). For some species, it is most likely responsible for regional extinction (e.g., badger Meles meles, Lankester et al., 1991; Clarke et al., 1998). Moreover, traffic is considered as one of the most important sources of mortality for many endangered or rare species. Although, the number of traffic victims may seriously reduce the population size of some species (Clarke et al., 1998; Huijser and Bergers, 2000), the effect of traffic mortality on populations is often difficult to measure as other factors, such as area, quality and spatial configuration of the habitat along the road, also play a role. There are complex relationships between the barrier effect and the mortality effect, which determine mortality during movement (i.e., the movement death rate), and the number of successful crossings (i.e., the crossover rate) (Verboom, 1994, see Figure 2). To quantify these effects, relationships between traffic and road characteristics must be found. For instance, a wider road encourages both higher traffic volumes and speeds. This, in turn, reduces the chance of a successful road crossing (as formulated by Van Langevelde and Jaarsma, 2004), as the intervals between vehicles become much smaller. Moreover, the wider the road, the more time an animal needs to cross the road and the less chance it has to actual succeed. In addition, an increase of volume may lead to such a flow of vehicles that individuals are restrained to cross the road. Finally, an increase of volume also determines the noise level increasing the barrier effect. In the next section, we focus on the mortality effect as current knowledge does not allow quantifying the barrier effect of roads (see Verboom, 1994). 3. MODELING TRAVERSING WILDLIFE 3.1 Relevant road, traffic, vehicle and species characteristics What are the relevant road, traffic, vehicle and species characteristics that have an effect on the traversability? Regarding the road characteristics, it is clear that as the road is wider, animal need more time to cross and the probability of successful road crossings decrease. Moreover, wider roads carry higher traffic volumes and allow for higher speeds. A small clearance of a road has a negative impact on the traversability of the road (Oxley et al., 1974; Adams and Geis, 1983; Clevenger et al., 2003). A small clearance can often be found in forested landscapes. High traffic volumes cause high noise loads and a high collision probability, as the intervals to cross between the vehicles are small. An increase of traffic volume

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

259

may lead to such a flow of vehicles that individuals do not cross the road anymore. Traffic volume may, however, largely fluctuate over the day and between seasons.

Figure 2. (A) The crossover rate and (B) the movement death rate as function of the barrier effect and the mortality effect (Verboom, 1994).

The traffic volume that largely determines traffic mortality (called the decisive traffic volume) depends on the time split of the daily traffic flow and the activity period of the animals during the day. The daily traffic pattern has characteristic peaks in the rush hours (7% in the morning and 10% in the afternoon) and an intermediate level during the evening (about 5%). During the night hours, only 1 or 2% of daily volumes passes. As animals are active during dusk and night, they deal with considerably lower hourly volumes than during daylight time. Moreover, most animals only cross a road when traffic volume is rather low, which is the case during dusk and night (Clevenger et al., 2003). With respect to vehicle characteristics, their size (length and width) and speed affects the traversability. Vehicle speed seems to be important because of the better opportunities for both animal and driver to avoid a collision when the vehicle speed is lower. Depending on the road, traffic and vehicle characteristics, different animal species experience differences in traffic mortality, such as in insects (Munguira and Thomas, 1992; Vermeulen, 1994), reptiles and amphibians (Hels and Buchwald, 2001), birds (Clevenger et al., 2003) and mammals (Mader, 1984; Lankester et al., 1991; Clarke et al., 1998). Whether species are vulnerable to traffic mortality depends on characteristics such as their home range size, the period of the day or season during which the animals are active, whether they move large distances during foraging, dispersal or migration, their traversing behaviour (velocity, reaction to approaching vehicles), their body length or the size of the group in which the individuals move. Species of closed and half-open landscapes with a large home range that move large distances are relatively sensitive to traffic mortality since they frequently cross roads that have a low clearance (e.g., Oxley et al., 1974; Adams and Geis, 1983; Groot Bruinderink and Hazebroek, 1996; Clarke et al., 1998). Fast moving mammals (often

260

F. VAN LANGEVELDE ET AL.

large animals) are less vulnerable for traffic mortality. However, these animals often have relatively large home ranges or move large distances. For these animals, the effect of traffic mortality on population dynamics can only be assessed when their daily and seasonal road crossings are considered. 3.2 Traffic flow theory In traffic engineering, the calculation of headway distributions, i.e., the frequency of the length of gaps between successive vehicles in a traffic flow at a given cross-section is commonly based on the assumption of a Poisson distributed process (Haight, 1963, 1966; Drew, 1968; Leutzbach, 1988; Daganzo, 1997). The Poisson distribution is a discrete distribution that describes the number of events during a given time period. Here, the event is a vehicle arriving at a given location. The numbers of events in sequential time periods of an equal length are independent stochastic drawings. For a given traffic volume, the probability of a certain number of arrivals within a fixed time period depends only on the length of this period and is thus constant for periods of equal length. When the number of vehicles in a sequence of fixed time periods is Poisson distributed, their headways are (negatively) exponentially distributed, and independent of each other. To be Poisson distributed, it is necessary that the vehicles approach a certain location without any disturbance, due to for example traffic lights. Also, the traffic volumes should be not too high: say, below 400 to 1000 vehicles h-1. According to the Poisson distribution, the probability P(x) that x vehicles arrive at a given location on a one-way road in time period T (in s) can be described as (λT )x e − λT (1) P( x) = x! where λ is the traffic volume in vehicles s-1. For a successful traversing, x should be equal to 0 during at least the time period T when the animal “occupies” the road for traversing. For x = 0, equation (1) changes into P (0) = Pr{Headway > T } = e − λT (2) In other words, P(0) is the probability that the front of the next car does not arrive within a period of T seconds, given a traffic flow with on average λ vehicles s-1. The relevant length of the time period T depends on road, traffic, vehicle and species characteristics as mentioned above. When the road carries traffic in two directions, with flows λ1 and λ2, then both flows can be described as a Poisson process. The well known mathematical theory learns that the two-way flow on that road, λ = λ1 + λ2, is also a Poisson process. So formula (2) remains the same in this situation, with λ now representing the two-way traffic volume. 3.3 Formulation of the traversability model For the application of headway distributions of traffic flows to traversing animal species, several assumptions are made (Van Langevelde and Jaarsma, 2004). The main difference in road crossing by people and animals is that most people can reasonably

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

261

estimate whether a gap between two successive vehicles is sufficiently large to cross safely. As the strategies used by animals to traverse roads are unknown, it is assumed that they act “blind”, without responding to the presence of a car, if any, and maintain a constant speed during their traverse. Especially in situations with a low clearance, the “blind” traversing is a realistic supposition. Beside the animal, we further presume that also the driver is ‘blind’, because the time available to avoid a collision with a traversing animal is around 1 second or less. So the traversability model does not include “corrections” by human and/or animal when their presence coincides. Two further assumptions for modelling are: (1) when an animal during its movement through the landscape finds a road on its way, it will traverse this road promptly, with a constant speed and at an angle (π/2 – α) with the road axis (for α = 0 the crossing is perpendicular), and (2) the traversing animal will be killed in a collision if the appearing gap in the traffic flow at the start of its traversing is too small, and, in reverse, there is a successful traverse if the gap is at least as large as the animal needs for its traverse. We distinguished two chances for a collision that determine the traversability: a collision can appear (1) when the animal is on the part of the road used by the car, and (2) when the animal hits the side of a car (Jaarsma et al., in prep.). Distinguishing these two chances for a collision, the period δ1 (in s), during which the car hits the animal, is Wc + La cos(α) (3) δ1 = Va where Wc is car’s width (in m), and La and Va are the animal’s length (in m) and speed (in m s-1) respectively. We assume here for reasons of simplicity that car and animal can be represented by a rectangle. For α = 0, i.e., perpendicular traversing, formula (3) reduces to W + La δ1 = c (4) Va So, if the animal traverses the road at an arbitrary moment, it can survive if the front of the next car does not arrive within a period of δ1 seconds. The probability of this event, P1, is (see formula 2) − λδ P 1 = Pr{Headway > δ1} = e 1 (5) The period δ2 (in s) during which the animal can hit the car is Lc + Wa cos(α) (6) δ2 = Vc where Wa is animal’s width (in m) and Lc and Vc are the car’s length (in m) and speed (in m s-1), respectively. So, if the animal traverses the road at an arbitrary moment, it will not hit a car and can survive if the front of the last car has passed at least a period of δ2 seconds ago. The probability of this event, P2, is − λδ (7) P 2 = Pr{Headway > δ2} = e 2 Combining both events, the animal can traverse without a collision with probability Pa that equals the product of formulae (5) and (7)

262

F. VAN LANGEVELDE ET AL.

Pa = e − λδ1 e − λδ 2 = e − λ ( δ1 + δ 2 ) (8) Expressed in the characteristics of animal and car this formula transfers into ⎞ ⎛ Wc + La ⎟ ⎜ L +W cos(α ) ⎟ cos(α ) −λ ⎜⎜ + c a ⎟ Va Vc ⎜⎜ ⎟⎟ ⎠ ⎝ e

Pa = (9) For the perpendicular traversing, α = 0 and formula (9) reduces to ⎛ W + L L +W −λ ⎜⎜ c a + c a V Vc e ⎝ a

⎞ ⎟ ⎟ ⎠

Pa = (10) Based on formula (10), the number of traffic victims of a species a, Da, during time period τ can be estimated by Da = (1 − Pa ) K a ,τ (11)

where Ka,τ is the number of attempts to traverse the road by individuals of species a during the time period τ. The parameter Ka,τ is, however, difficult to measure and depends on several species and landscape characteristics such as home-range size, movement behaviour during foraging or dispersal, road density and the location of the road with respect to, for example, the foraging areas. We therefore suggest the model not to apply to calculate the absolute number of traffic kills of species a during, say, a season, but to use it in a relative way. The traffic mortality can be estimated for two situations with the same number of attempts to traverse the road. For example, the present situation is compared with the planned situation with new road and traffic characteristics and the difference between both is considered to be the difference in impact. 4. APPLICATION OF THE TRAVERSABILITY MODEL 4.1 Integral strategy of Traffic-calmed Areas In order to prevent habitat fragmentation due to infrastructure, mitigating interventions can be applied. These interventions can be directed towards enhancement of the traversability of the roads themselves (decreasing traffic intensity and/or speed), creating wildlife overpasses or underpasses, reducing mortality chance (fences), or quality enhancement of the adjacent habitat (noise reducing walls). However, the applicability of these interventions differs between types of roads. We distinguish here three types of roads by their function (Jaarsma, 1997): (1) motorways, with mainly a flow function that offers fast and comfortable service for through traffic on long distances, (2) rural highways, with an access function for regions and for opening up regions, and (3) minor roads, with mainly local collector and access roads with mixed traffic for destination accessibility. Mitigation measures for major roads (motorways and highways) will not be as effective for minor roads as (Van Langevelde et al., in prep.):

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

263

1. Minor roads have a diffuse victim pattern: many locations with low frequency of accidents. This makes the designation of bottlenecks problematic. 2. These bottlenecks are not only more difficult to designate, also prioritising of interventions is much more complex as many accident locations can be a bottleneck. 3. Interventions for rural highways or motorways are primary directed towards reduction of the barrier effect (without increasing traffic mortality). However, interventions for minor roads should primarily reduce the number of accidents, without increasing the barrier effect. 4. Enhancement of the traversibility of the road itself would more feasible for a minor road than for rural highways or motorways, such as speed limiting interventions and/or temporary closure for vehicles. A condition for such interventions would be the presence of acceptable alternatives for through traffic. 5. A large problem with interventions for minor roads is the lack of specified knowledge. There is much more knowledge on the effects of rural highways and motorways on fauna. We assume this coincides with the difficulty to determine the effects of minor roads. Locations with high victim numbers are generally first nominated for mitigating measures. This is indeed frequently applied to rural highways and motorways (Forman et al., 2003; Van Bohemen, 2005). For minor roads, however, even when such locations with relatively high accident frequencies occur, the low number of victims and the low accident risk result in an (too) important role of coincidence. Therefore, such a method is for minor rural roads not feasible to determine bottlenecks or enhancement after mitigation. Interventions to prevent habitat fragmentation by infrastructure can only be really successful when problems concerning minor roads are also accounted for, because: 1. As soon as interventions are implemented on one road section in a road network, unexpected effects can occur elsewhere. This applies to animals (alterations in movement patterns) and human (alterations of traffic flows). For example, measures on one specific road section can have consequences for other road sections in the network, either positive or negative, because of the hierarchy within the road network, consisting of interconnected networks of motorways, highways and minor roads. 2. A shift can be expected as the number of traffic victims on minor roads will increase when mitigating interventions on rural highways and motorways are implemented. The home range of a lot of species covers more than only one road. Within this context, it is stressed that a lot of species not only live in nature reserves but also in other rural parts of the landscape. 3. Implementing road design or road closing interventions for a certain road section is only possible when alternatives are offered to through traffic. For offering alternatives, rural highways and motorways can play an important role. These effects can be prevented, not by planning based on separated road sections (the ‘road section approach’), but by planning based on a coherent road network (the

264

F. VAN LANGEVELDE ET AL.

‘network approach’; Jaarsma, 1997, 1999, 2004; Van Langevelde and Jaarsma, 1997). Therefore, we recommend an integral strategy, in which the problems concerning habitat fragmentation on minor roads, rural highways and motorways in a region are all accounted for in mutual cohesion (Jaarsma and Willems, 2002a; Van Langevelde et al., in prep.). Such an integral strategy requires a regional approach, say, between 50 and 200 km2, and not on the level of only one or a few specific road section(s). The planning concept “traffic-calmed rural areas” (Jaarsma, 1997) is based on such regional network approach. This concept is originally developed to promote traffic safety and tackle rat-run traffic in rural areas. We argue here that this concept can also be applied to mitigate habitat fragmentation. 4.2 Planning concept “traffic-calmed areas” During the 1970s, the concept of urban residential traffic-calmed areas was developed. These residential precincts are areas within urban areas with restricted rights for motorised traffic. This is expressed in a specific design, directed to a low speed level. This concept has already served as an international model (Macpherson, 1993). The concept of “traffic-calmed rural areas” uses the same ideas derived from built-up areas and transfer them to the rural area (Jaarsma, 1997). The underlying idea is a clear separation between space for living that involves inhabitants and recreationists as well as wildlife, and space for traffic flows. Then, starting positions for (re)designing roads in traffic-calmed areas are the preferred functions and not the appearing traffic flows. Usually, residential functions (inhabitants, recreationists, wildlife) will be emphasised, and not the traffic function for through traffic. Traffic-calmed areas will be accessible by means of minor roads with a moderate (technical) design for low speeds and low traffic volume. Through traffic will find faster alternative routes over rural highways or motorways. On these roads, which additionally give access to the traffic-calmed area, bundled traffic flows appear (Jaarsma, 1997). Reduction of traffic speed and volume due to the bundling of traffic flows will have a positive impact on the traffic safety. Bundling also favours noise load. Opposite to small increases along roads with increased traffic flows, large reduction of noise load occurs along traffic-calmed roads. The most important disadvantage of the traffic-calmed areas is the increase in vehicle mileage because the route along minor roads is often shorter in both length and time than the functional route along motorways and other major roads. In time, however, calculated differences mostly are very small (Jaarsma, 1997; Jaarsma and Willems, 2002a). From explorative research (Jaarsma and Van Langevelde, 1997; Jaarsma and Willems, 2002a and 2002b), it seemed that profit for nature is gained by “overall” decrease of zones with high noise loads and enhanced traversibility, especially for larger mammals. Based on model calculations, a decrease of traffic intensity seems to have the largest impact on road crossings by fauna. With that, the trafficcalmed areas create opportunities for local populations with fewer limitations for exchange.

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

265

4.3 Gerdyksterwei as case study We applied the concept traffic-calmed areas and the traversability model to several regions where bottlenecks appear between wildlife and the road network to compare alternative network solutions and their impacts on traversability for wildlife (e.g., Jaarsma and Van Langevelde, 1997; Jaarsma and Willems, 2002a). Here, we present an example of a former section of the Dutch national road network, the Gerdyksterwei between the Frisian villages Gorredijk and Beetsterzwaag, which is bypassed today by the A7 motorway (Figure 3). Since the A7 motorway is in service, carrying daily about 30,000 motor vehicles, the Gerdyksterwei is intended to be a minor road with modest traffic flows. However, many drivers between Gorredijk and the nearby town of Drachten still prefer the former route above the functional route along the A7. Therefore, daily volumes on the Gerdyksterwei (4,100) and in the center of the village of Beetsterzwaag (5,300) are too high from an environmental point of view. By autonomous developments, these volumes are even expected to increase with a further 1,000 vehicles per day in the next ten years. Although the technical capacity of the Gerdyksterwei is large enough to handle these volumes (the road still has its traditional layout with a broad pavement, based on its former function in the national network), the present volume forces two problems. Within the village of Beetsterzwaag livability of the inhabitants is threatened, and in the rural area the Gerdyksterwei intersects an extended wooded area with a lowland brook (Koningsdiep), which is a core area in the Dutch National Ecological Network. Here, for small and larger mammal species such as hedgehog (Erinaceus europaeus), rabbit (Oryctolagus cuniculus), roe deer (Capreolus capreolus), and (when re-introduced) otter (Lutra lutra), the collision chance when traversing the Gerdyksterwei is considerable by its high traffic volumes. Therefore, the local government investigated the impacts of rural traffic calming for livability and wildlife movement. Within this context, traffic calming means priority being given to nature (in the rural area) and to people (in the village), not to through traffic. The latter is offered an alternative route with a high quality via the A7 motorway. Wildlife can traverse this motorway safely through underpasses. As a consequence of traffic calming, speeds and/or volumes on the Gerdyksterwei must decrease, contrary to the autonomous development. We elaborate 4 levels of traffic calming: (1) mainly legal measures, including a rigid enforcement of the present speed limit of 80 km h-1; (2) implementation of a so-called rural residential area with a legal speed limit of 60 km h-1 and with a few speed humps; (3) the previous, with more measures to reduce speed and in combination with a reduction of the pavement width; (4) the previous, with limited access: between 7 p.m. and 7 a.m. access for local residents only. By these measures, the estimated effective speed on the Gerdyksterwei decreases. Consequently, an increasing part of the through traffic will take the A7 because it offers a faster or at least more comfortable route than the traffic-calmed Gerdyksterwei. Based on travel times between Gorredijk and Drachten, it is estimated that the first level of traffic calming only slightly reduces the future flows.

266

F. VAN LANGEVELDE ET AL.

The implementation of a rural residential area will be more effective in avoiding through traffic. The level with limited access reduces the volumes between 7 p.m. and 7 a.m. from 400 to approximately 100, which means on average only 8 vehicles h-1.

Figure 3. The Gerdyksterwei in its regional context. This road traverses both the dry and the wet National Ecological Netwerk (the Wallebosch and Lippenhuisterheide and the low land brook Koningsdiep, respectively). Despite the presence of the A7 motorway, the Gerdyksterwei still carries a lot of cars travelling from Gorredijk to Drachten and further to the north v.v. The village of Beetsterzwaag also burdens a large part of this traffic flow (elaborated from Jaarsma and Van Langevelde, 1997).

This extra reduction during the night is relevant considering the ecology of the mammals mentioned above, because their decisive period for movement is during the night. Nightly volumes for the other situations are estimated by the assumption that one quarter of the daily flow appears between 7 p.m. and 7 a.m., which is equally spread over these twelve hours. Table 1 presents an overview of the road and traffic characteristics applied into our calculations. The decisive traffic volumes in the table are the average hourly volumes during the night. The impacts of traffic calming on wildlife traversability for the Gerdyksterwei are presented in table 2, showing the resulting changes in traffic mortality per 104 traversings for the roe deer, the otter, the rabbit and the hedgehog. From table 2, we conclude that traffic calming can be an effective method to improve traversability

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

267

for wildlife. Differences between the 4 species are small, however, the small and slow moving hedgehog has a somewhat higher victim reduction. Table 1. Road and traffic characteristics for the Gerdyksterwei in the actual situation and estimated for the autonomous development and the 4 levels of traffic calming (explained in the text). Characteristic

Present situation 7

Autonomous development 7

Level 1 7

Levels of rural traffic calming Level 2 Level 3 Level 4 7 5 5

Pavement width (m)* 80 80 80 60 60 60 Legal speed limit (km h-1) 85 80 72 60 50 50 Estimated effective speed (km h-1) 4,200 5,200 4,500 2,500 1,600 1,300 Average annual daily volume (vehicles d-1) Decisive 84 104 90 50 32 8 volume (vehicles h-1) * pavement width is not included in formula (10), but it affects both effective speed and traffic volume

The table also clearly shows that, if traversability is already considered as an ecological problem in the present situation, measures must be taken since in the autonomous development the situation will worsen. Compared to the autonomous situation, the first level of traffic calming shows a slight improvement of the traversability, but this is still worse than in the present situation. A further development of measures allows for a considerable improvement: the second and the third calming levels show a reduction of traffic kills of about one third and more than 50%, respectively. In this situation, with wildlife movements during the night as decisive period, a total closure for through traffic during the night as in level 4 is very effective. It reduces the number of traffic kills to about 10% of the present value. 5. SYNTHESIS In this chapter, we show the important, but not always distinguished, role of major as well as minor roads and their traffic flows on wildlife, as a part of their environmental impacts. For a generation already, the road network pervades a paradoxical role in our society. On the one hand, people seek to harvest the benefits of an expanding road system, including an improving access to ‘green’ areas. On the

268

F. VAN LANGEVELDE ET AL.

other hand, people have growing concerns about threats of roads to the natural environment, including noise, emissions, vehicular-related air quality and loss of species and wildlife habitat. Table 2. Estimated victims per species for the Gerdyksterwei in the present situation, the autonomous development and the 4 levels of traffic calming (explained in the text). Victims relative to the present situation with a decisive volume of 84 vehicles h-1 Animal species

Roe deer Otter Rabbit Hedgehog

Kills per 104 traverses in present situation (= 100%) 204 184 162 572

Autonomous development

125 126 126 124

Relative kills per 104 traverses (%) Level 1 Level 2 Level 3

112 112 113 109

66 66 67 62

45 46 46 41

Level 4

11 11 12 10

Originally, the road network was built in an era when transportation planners focused on providing safe and efficient transport with little regard for wildlife. “That is changing. … the call for new knowledge and skills is stronger than ever” (Forman et al., 2003; p xiii). Also new legislation, such as the EU Habitat Directive, enforces the transportation community to include ecological impacts into their planning system. More specifically, for relevant (threatened) species in the region the impacts of measures proposed in a transportation plan must be described (Haq, 1997; Iuel et al., 2003). Wildlife traversing a road is an important aspect of habitat fragmentation by infrastructure and its traffic flows. So far, a tool is missing to estimate the impacts on wildlife movements of changes in a regional road network and/or the layout of specific road sections and the resulting changes in traffic volumes and speeds. The traversibility model, as presented in this chapter, enables to include impacts of roads and their traffic flows on wildlife movement and traffic kills among animals. The traversability model can contribute to the conscious integration of nature and engineering in a way that is useful for both human and nature (Van Bohemen, 2005). This model can be used to estimate the changes in traffic mortality for animal species as a result of changes in road and traffic characteristics, by comparing changes in road or traffic characteristics or alternatives for road design and traffic volumes. Then, the model can provide insight in the relative effects of these road and traffic characteristics on population dynamics of wildlife. When data on traffic mortality are available (Groot Bruinderink and Hazebroek, 1996; Garrett and Conway, 1999), the model could be used to predict changes after applying mitigating measures. When numbers of victims are not available, however, model predictions based on road and traffic characteristics and the distribution and size of the local populations of the species could also be useful to determine the locations

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

269

where mitigating measures should be applied. It offers thus a relatively simple addition to the existing toolbox of the planner and it only asks for limited data. So far, the traversability model is not tested in experiments. For such an experiment, among others, reliable numbers of road crossing by individual animals as well as numbers of traffic kills per road section should be gathered. As far as we know, there are no studies on the former. Some studies provide numbers of victims, but due to scavengers or identification problems, especially for small animals, the actual numbers are difficult to measure. This conclusion already holds for seventy years (Stoner, 1936; Hels and Buchwald, 2001; Slater, 2002). It is therefore questionable whether an empirical experiment can provide reliable data for the validation of the absolute numbers of traffic kills as calculated by the model. When the model is used to calculate the relative difference between two situations with different road and/or traffic characteristics, systematic errors in the model by animal behavior, if any, will be eliminated by subtraction. The traversability model is based on a limited number of road, traffic, vehicle and species characteristics. Other characteristics also influence the road crossing, such as road lighting as some animals avoid these roads, whereas others are attracted. Some species will flee or stay when a vehicle is approaching, e.g., the traversing speed will be underestimated when individuals flee. Moreover, some animals restrain from roads when traffic volume increases. We assumed that animals cross roads without any waiting time. This may be valid for landscapes where the clearance is low, but otherwise it is plausible that animals are restrained to cross when a vehicle is approaching. They may also be restrained when traffic volume is high due to the constant noise and visibility of vehicles. Moreover, we assumed that when an animal and a vehicle are at the same location at the same moment, a collision occurs. This might not be true since corrections by humans and animals and also mis-hits where the animal survives a collision (e.g., because they are small enough to survive between the tires of a vehicle) also affect traffic mortality. So far, the assumptions in the traversability model exclude the above-mentioned factors. Relaxing these different assumptions does, however, not drastically change the model but have an effect on the predicted traffic mortality. Environmental impacts of infrastructure such as noise and pollution are estimated with quantitative models. Except for habitat loss, the impacts on nature are difficult to quantify. Maybe that is the reason that, beside large-scale mitigation measures by means of wildlife underpasses and overpasses, there is a lack of attention for impacts on wildlife so far. To bridge this gap, and to enable transportation planning to include the impacts on wildlife in the planning process for a regional road network, the presented traversability model can be subservient. This is illustrated in the case study, where several levels of traffic calming are compared with the autonomous development. We show that traffic calming can drastically reduce traffic mortality. Such a traversability model could thus be a tool for transportation planners and conservationists to prevent traffic accidents and protect biodiversity.

270

F. VAN LANGEVELDE ET AL.

REFERENCES Adams, L.W. and Geis, A.D. (1983). Effects of roads on small mammals. Journal of Applied Ecology, 20, 403-415. Andrews, A. (1990). Fragmentation of habitat by roads and utility corridors: a review. Australian Zoologist, 26, 130-141. Button, K. and Nijkamp, P. (1999). Series preface for Environmental analysis and economic policy. In Hayashi, Y., K. Button and P. Nijkamp (Eds.), The Environment and Transport. (pp. xi-xiii), Edward Elgar Publishing, Cheltenham. Clarke, G.P., White, P.C.L. and Harris, S. (1998). Effects of roads on badger Meles meles populations in south-west England. Biological Conservation, 86, 117-124. Clevenger, A.P., Chruszcz, B. and Gunson, K.E. (2003). Spatial patterns and factors influencing small vertebrate fauna roadkill aggregations. Biological Conservation, 109, 15-26. Daganzo, C.F. (1997). Fundamentals of Transportation and Traffic Operations. Elsevier. Drew, D.R. (1968). Traffic Flow Theory and Control. McGraw-Hill, New York. Elvik, R. and Vaa, T. (2004). The Handbook of Road Safety Measures. Elsevier Science, Amsterdam. Forman, R.T.T. and Alexander, L.E. (1998). Roads and their major ecological effects. Annual Review of Ecology and Systematics, 29, 207-231. Forman, R.T.T., Sperling, D., Bissonette, J.A., Clevenger, A.P., Chutshall, C.D., Dale, V.H. (2003). Road Ecology. Science and Solutions. Island Press, Washington. Garret, L.C. and Conway, G.A. (1999). Characteristics of moose-vehicle collisions in Anchorage, Alaska, 1991-1995. Journal of Safety Research, 30, 219-223. Gillen, D. (2003). The economics of noise. In Hensher, D.A. and K.J. Button (Eds.), Handbook of Transport and the Environment. Handbooks in Transport 4 (pp. 81-96.), Elsevier Science, Amsterdam. Groot Bruinderink, G.W.T.A. and Hazebroek, E. (1996). Ungulate traffic collisions in Europe. Conservation Biology, 10, 1059-1067. Haight, F.A. (1963). Mathematical Theories of Traffic Flow. Academic Press, London. Haight, F.A. (1966). Handbook of the Poisson Distribution. John Wiley & Sons, New York. Haq, G. (1997). Towards Sustainable Transport Planning. Asgate, Aldershot. Helldin, J.O. and Seiler, A. (2003). Effects of roads on the abundance of birds in Swedish forest and farmland. Wildlife Biology, 6, 111-120. Hels, T. and Buchwald, E. (2001). The effect of road kills on amphibian populations. Biological Conservation, 99, 331-340. Huijser, M.P. and Bergers, P.J.M. (2000). The effect of roads and traffic on hedgehog (Erinaceus europaneus) populations. Biological Conservation, 95, 111-116. Iuell, B., Bekker, G.J., Cuperus, R., Dufek, J., Fry, G., Hicks, C., Hlaváĉ, V., Keller, V.B., Rosell, C., Sangwine, T., Tørsløv, N., Wandall and B. le Maire (Eds.) (2003). Wildlife and Traffic: a European Handbook for Identifying Conflicts and Designing Solutions. KNNV Publishers, Utrecht, the Netherlands. Jaarsma, C.F. (1997). Approaches for the planning of rural road networks according to sustainable land use planning. Landscape and Urban Planning, 39, 47-54. Jaarsma, C.F. (1999). Planning of rural road networks. In H.N. van Lier, L.S. Pereira and F.R. Steiner (Eds.), CIGR Handbook of Agricultural Engineering. Volume 1, Land and Water Engineering (pp. 73-86.), ASAE (American Society of Agricultural Engineers), St. Joseph MI (US). Jaarsma, C.F. and Willems, G.P.A. (2002a). Reducing habitat fragmentation on minor rural roads through traffic calming. Landscape and Urban Planning, 58, 125-135. Jaarsma, C.F. and Willems, G.P.A. (2002b). GIS as a tool to address environmental issues in rights-ofway planning and management: the example of rural road networks. In Goodrich-Mahoney, J.W., D.F. Mutrie and C.A. Guild (Eds.), Environmental Concerns in Rights-of-way Management (pp. 553562), Elsevier Science, Amsterdam. Jaarsma, C.F. (2004). Ecological "black spots" within the ecological network: an improved design for rural road networks amelioration. In Jongman, R.H.G. and G. Pungetti (Eds.), Ecological networks and Greenways: Concept, Design, Implementation (pp. 171-187), Cambridge University Press.

EFFECTS OF TRAFFIC MORTALITY ON SPECIES

271

Jaarsma, C.F. and van Langevelde, F. (1997). Right-of-Way Management and habitat fragmentation: an integral approach with the spatial concept of the Traffic Calmed Rural Area. In Hydro-Quebec Centre de Documentation Environment (Ed.), Environmental Concerns in Rights-of-way Management (pp. 383-392), Proceedings of the 6th International Symposium, New Orleans (LA), 2426 February, Elsevier Science. Jaarsma, C.F., van Langevelde, F. and Botma, H. (submitted) Flattened fauna and mitigation: traffic victims related to road, traffic, vehicle, and species characteristics. Jackson, S.D. and Griffin, C.R. (1998). Toward a practical strategy for mitigating highway impacts on wildlife. In Evink, G.L., P. Garrett, D. Zeigler and J. Berry (Eds.), Proceedings of the International Conference on Wildlife Ecology and Transportation (pp. 17-22), Florida Department of Transportation, Tallahassee, Florida. Jaeger, J.A.G., and Fahrig, L. (2004). Effects of road fencing on population persistence. Conservation Biology, 18, 1651-1657. Jedicke, E. (1994). Biotopverbund – Grundlagen und Massnahmen einer neuen Naturschutzstrategie. Eugen Ulmer Verlag, Stuttgart, Germany. (in German). Keller, V. and Pfister, H.P. (1997). Wildlife passages as a means of mitigating effects of habitat fragmentation by roads and railway lines. In Kanters, K.J., A. Piepers and D. Hendriks-Heersma (Eds.), Habitat Fragmentation and Iinfrastructure (pp. 70-80), Proceedings of the International Conference on Habitat Fragmentation, infrastructure and the role of ecological engineering. Maastricht/The Hague. Kirby, K.J. (1997). Habitat fragmentation and infrastructure: problems and research. In Kanters, K.J., A. Piepers and D. Hendriks-Heersma (Eds.), Habitat Fragmentation and Iinfrastructure (pp. 32-39), Proceedings of the International Conference on Habitat Fragmentation, infrastructure and the role of ecological engineering. Maastricht/The Hague. Lankester, K., van Apeldoorn, R., Meelis, E. and Verboom, J. (1991). Management perspectives for populations of the Eurasian badger (Meles meles) in a fragmented landscape. Journal of Applied Ecology, 28, 561-73. Lee, C.S.Y., Fleming, G.G. and Burstein, J. (1998). FHWA Traffic Noise Model: Look-Up Tables. Report FHWA-PD-98-047. Washington, D.C.: U.S. Department of Transportation, Federal Highway Administration. Leutzbach, W. (1988). Introduction to the Theory of Traffic Flow. Springer-Verlag, Berlin. Macpherson, G. (1993). Highway and Transportation Engineering and Planning. Longman, Harlow, UK. Mader, H.J. (1984). Animal habitat isolation by roads and agricultural fields. Biological Conservation, 29, 81-96. Merriam, G., Kozakiewicz, M., Tsuchiya, E. and Hawley, K. (1989). Barriers as boundaries for metapopulations and demes of Peromyscus leucopus in farm landscapes. Landscape Ecology, 2, 227235. Munguira, M.L. and Thomas, J.A. (1992). Use of road verges by butterfly and burnet populations, and the effect of roads on adult dispersal and mortality. Journal of Applied Ecology, 29, 316-329. Oxley, D.J., Fenton, M.B. and Carmody, G.R. (1974). The effects of roads on populations of small mammals. Journal of Applied Ecology, 11, 51-59. Philcox, C.K., Grogan, A.L. and MacDonald, D.W. (1999). Patterns of otter Lutra lutra road mortality in Britain. Journal of Applied Ecology, 36, 748-762. Putman, R.J. (1997). Deer and road traffic accidents: options for management. Journal of Environmental Management, 51, 43-57. Reijnen, R., Veenbaas, G. and Foppen, R. (1995). Predicting the Effects of Motorway Traffic on Breeding Bird Populations. Road and Hydraulic Engineering Division and DLO-Institute for Forestry and Nature Research, P-DWW-95-736, Delft, NL. Romin, L.A. and Bissonnette, J.A. (1996). Deer-vehicle collisions: status of state monitoring activities and mitigation efforts. Wildlife Society Bulletin, 24, 276-283. Seiler, A. (2002). Effects of infrastructure on nature. In COST 341. Habitat Fragmentation due to Transportation Infrastructure. The European review. European Commission, Directorate-General for Research. Brussels, B.: 31-50. Seiler, A. and Erikson, I.-M. (1997). New approaches for ecological consideration in Swedish road planning. In Canters, K.J., A. Piepers and D. Hendriks-Heersma (Eds.), Habitat Fragmentation and Infrastructure (pp. 253-264), Proceedings of the International Conference on Habitat Fragmentation, infrastructure and the role of ecological engineering. Maastricht/The Hague (Neth.).

272

F. VAN LANGEVELDE ET AL.

Sharma, N., Chaudhry, K.K. and Chalapati Rao, C.V. (2004). Vehicular pollution prediction modelling: a review of highway dispersion models. Transport Reviews, 24, 409-435. Sharma, P. and Khare, M. (2001). Modelling of vehicular exhausts – review. Transportation Research D, 6, 179-198. Singh, A.P. and Satheesan, S.M. (2000). Designing railroads, highways and canals in Protected Areas to reduce man-elephant conflicts. Paper presented at the 7th International Symposium on Environmental Concerns in Rights-of-way Management. Calgary, Alberta, Canada, September 9-13. Slater, F.M. (2002). An assessment of wildlife road causalities – the potential discrepancy between numbers counted and numbers killed. Web Ecology, 3, 33-42. Spellerberg, I.F. (1998). Ecological effects of roads and traffic: a literature review. Global Ecology and Biography Letters, 7, 317-333. Stoner, D. (1936). Wildlife casualties on the highways. Wilson Bulletin, 48, 276-283. Trombulak, S.C. and Frissell, C.A. (2000). Review of ecological effects of roads on terrestrial and aquatic communities. Conservation Biology, 14, 18-30. Van Bohemen, H.D. (1998). Habitat fragmentation, infrastructure and ecological engineering. Ecological Engineering, 11, 199-207. Van Bohemen, H.D. (2005). Ecological Engineering: Bridging Between Ecology and Civil Engineering. AEneas, Technical Publishers, Boxtel, the Netherlands. Van Langevelde, F. and Jaarsma, C.F. (1997). Habitat fragmentation, the role of minor rural roads and their traversability. In Canters, K.J., A. Piepers and D. Hendriks-Heersma (Eds.), Habitat Fragmentation and Infrastructure (pp. 171-182), Proceedings of the International Conference on Habitat Fragmentation, infrastructure and the role of ecological engineering. Maastricht/The Hague (Neth.). Van Langevelde, F. and Jaarsma, C.F. (2004). Using traffic flow theory to model traffic mortality in mammals. Landscape Ecology, 19, 895-907. Van Langevelde, F., van Dooremalen, C. and Jaarsma, C.F. Traffic mortality and the role of minor roads. (submitted) Verboom, J. (1994). Een Modelstudie naar de Effecten van Infrastructuur op Dispersiebewegingen van dieren (A model study of the effects of infrastructure of fauna dispersal). Rijkswaterstaat/ IBN-DLO, Delft/ Wageningen, 52 pp. (in Dutch). Vermeulen, H.J.W. (1994). Corridor function of a road verge for dispersal of stenotopic heathland ground beetles (Carabidae). Biological Conservation, 3, 339-49.

CHAPTER 17

ELEMENT FLUXES AND BUDGETS OF A PLANTATION EMBEDDED IN AN AGROFORESTRY LANDSCAPE: IMPLICATION FOR LANDSCAPE MANAGEMENT AND SUSTAINABILITY

W. SHEN 1,2, H. REN1, Y. LIN1, M. LI1 1

South China Botanical Garden, the Chinese Academy of Sciences, Guangzhou 510650, China; 2 Nicholas School of the Environment and Earth Science & Department of Biology, Duke University, Durham, NC 27708-0340, USA

Abstract. Nutrient fluxes and cycling are key processes to sustaining the structure and functioning of patches as well as the landscapes in which they are embedded. In this chapter, we synthesized a 10year study on the nutrient cycling of an Acacia mangium plantation located in the upper slope of a watershed landscape consisting on a tree plantation, a fruit garden, a grassland meadow, and a fishpond. Element fluxes and budgets were analyzed to derive guidelines for managing this landscape. We found that the plantation acted as a nutrient sink during its early stage of development (15-20 years since establishment), as it accumulated a large amount of nutrients on its compartment pools (i.e., plant biomass, forest floor, and soil). Over 1/3-1/6 of total litter on the plantation floor could be moved to fertilize other landscape components (e.g., the fruit garden and the grassland). However, the ability of the plantation to retain nutrients started to decrease at a fast rate when it approached near-mature development stage (20-30 years old). During this stage the risk of N saturation, soil acidification, and nutrient depletion (e.g., P, K, and Mg) increased as the plantation ages, especially under conditions of large atmospheric deposition. In order to sustain its nutrient retention ability and other ecosystem services, we suggested that application of composite P, K and Mg fertilizers are needed after the plantation reached about 20-30 years old. This study also confirmed that positioning the A. mangium plantation on the upper slope of the watershed was an appropriate spatial arrangement for the studied landscape, as it decreases the risk of transporting excessive elements into the lowland fishpond via soil erosion and surface runoff processes. Further studies on element fluxes and budgets of the other three landscape components need to be conducted in order to obtain a comprehensive understanding of interactions between landscape pattern and biogeochemical processes, and to ultimately reach a sustainable management of the landscape.

273 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 273–290. © 2007 Springer.

274

W. SHEN ET AL.

1. INTRODUCTION Nutrient fluxes support and interact with primary and higher tropic levels production, soil physical and chemical processes, and community structure (Vitousek and Howarth, 1991; Attiwill and Adams, 1993; Schlesinger, 1997). Thus, they are fundamentally important to sustain ecosystem structure and functioning. A major topic of nutrient flux studies was how nutrients are transported in the plantsoil-atmospheric continuum and how they impact primary productivity and soil properties. Associated with this topic, detailed nutrient transformation processes, such as atmospheric deposition, canopy re-adsorption, litter fall, litter and soil organic matter decomposition, nutrient storage, distribution, and drainage were studied extensively throughout the 20th century (Attiwill and Adams, 1993; Dobrovolksy, 1994; Schlesinger, 1997). The input-output element budget and the processes and factors determining this balance were also measured both within the system and in nearby watersheds for comparisons (Bormann and Likens, 1967; Likens and Bormann, 1995). However, most of these intra-system nutrient cycling (i.e., nutrient transportation and transformation) and watershed biogeochemistry studies ignored the horizontal spatial variability within the system, and treated the whole ecosystem/watershed as a “black box”. A growing body of evidence is showing that biogeochemical cycles and element fluxes of a given ecosystem may markedly influence the functioning of adjacent landscape components (Turner et al., 2001; Wu and Hobbs, 2001; Tenhunen and Kabat, 1999). Thus, ecosystems like riparian vegetation zones, wetlands, and estuaries may play important role in protecting their adjacent aquatic ecosystems from eutrophication (Haycock et al., 1997; Tenhunen et al., 2000; Turner et al., 2001; Ryszkowski, 2001). A classical example is the work by Peterjohn and Correll (1984), who measured nitrogen (N) and phosphorous (P) exports through surface runoff and subsurface groundwater of an upper land cornfield and a lower land riparian forest. They found that the riparian forest had significantly higher nutrient retention efficiency, and thus could reduce the cornfield generated agricultural pollution to the stream water. Such cross-system element transportation occurs in both human-dominated and in natural landscapes, such as the arctic tundra (Shaver et al., 1991) and temperate lake chains (Kratz et al., 1997). Furthermore, humanaccelerated environmental changes, namely, global climate change, land use/cover change, water, air and soil pollution, loss of biodiversity, and changes in atmospheric chemistry, may largely enforce the significance of studies on landscape or regional level biogeochemical cycles (Matzner et al., 2001, 2004). When it comes to management issues, the interactions between element transportation and landscape pattern must be considered. In this chapter, we synthesized results from a 10-year study on the within-stand nutrient cycling of an Acacia mangium plantation. This plantation was originally established as an experimental forest to restore severely degraded subtropical grass slopes in southern China. This pioneer plantation was also a major component located in the upper slope of an agroforestry landscape that occupies a whole watershed with the area of ca. 3 ha; other components of the watershed landscape

ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM

275

included a fruit garden on the middle slope, a grassland stripe on the lower slope, and a fishpond at the bottom of the watershed (Figure 1). Previous studies had mainly focused on within-system nutrient cycling processes, without considering impacts of element fluxes on the functioning of other landscape components and their implications for managing the whole landscape (Fang et al., 1995; Yao et al., 1995; Liu et al., 2000). This type of agroforestry landscapes is commonly found in hilly lands of China, especially in the Western and Southeastern regions of the country (Fu et al., 1998; Yu and Peng, 1996). Sustainable management strategies are urgently needed for land managers. Thus, the major goal of this chapter was to synthesize the nutritional element flux and budget information, and to explore implications of ecosystem-level information to landscape-level management issues. Detailed landscape characteristics and the specific management issues for the landscape are described and proposed in the next section, followed by descriptions on the sampling and analytical methods employed; then inter-annual and seasonal changes in element concentrations are presented in the fourth section. Element fluxes and budgets are assessed based on element concentrations and corresponding water fluxes in the fifth section. In the last section, the implications of nutrient fluxes and budgets for management issues are discussed, and some specific management suggestions are provided. 2. SITE DESCRIPTIONS AND LANDSCAPE CHARACTERIZATION The study site is located at the Heshan Long-Term Ecological Research Station (HS LTER), Heshan city, Guangdong province, southeastern China (latitude 22°41′N, longitude 112°54′E). Mean annual air temperature is 21.7 °C, with the mean monthly air temperature of 28.7 °C in July and 13.1 °C in January. Mean annual precipitation is 1800 mm, 79% of which falls in the wet season from April through September, and the rest of it falls in the dry season from October to March. Historically, this region was covered by evergreen broadleaved forest, but most of forestlands were almost completely transformed into other land use/cover types because of deforestation, urbanization, fuel gathering, and resultant soil erosion and water loss. A small portion of the unprofitable abandoned lands were re-vegetated with plantations dominated by Pinus massoniana, but most of them were turned into degraded subtropical grasslands that have relatively low primary productivity and are prone to soil erosion by water (Shen et al., 1999). Our study landscape was one of the models designed by Chinese ecologists to reuse the degraded grass slopes in the early 1980s. This agroforestry landscape occupied a whole small watershed (Figure 1), with the total area of 3 ha. The upper portion of the slope, with an area of 1.3 ha, is covered by a plantation of fast growing nitrogen fixing species, Acacia mangium. The middle slope is a fruit garden (Dimocarpus longa), occupying the area of 0.87 ha. The lower portion of the slope is a grassland stripe (Pennisetum purpureum). The bottom of the watershed is a fishpond (0.29 ha) for growing commercial fishes. The whole idea was to create a self-sustain agroforestry system that can efficiently control soil erosion and water loss while in the mean time increase the economic gain of the local people.

276

W. SHEN ET AL.

Specifically, plantation litter could be partially moved into the fruit garden as organic fertilizers, therefore reducing the need for inorganic fertilizers. Grass is grown for feeding fish, and the nutrient-enriched pond mud could be moved to grass field or fruit garden as fertilizers.

Figure 1. View of the agroforestry landscape. The studied Acacia mangium plantation (forest in the legend) is located in the upper slope of the watershed (see text for details).

Since the first of these landscapes was established in the HS LTER site, in 1983, it had been widely used by local farmers (Yu and Peng, 1996). However, there is still several management issues need to be further investigated. For example, what would be the most appropriate area composition and spatial arrangement of the three terrain ecosystem stripes in the landscape? Can the landscape self-sustain it’s functioning without additional inputs of inorganic fertilizers? In this chapter we will focus on the A. mangium plantation, mainly because it was expected to be the major nutrient source for other landscape components (e.g., the orchard), and the plantation is showing obvious declination in recent years, both structurally and functionally (Shen et al., 2003). By analyzing nutrient fluxes and budgets of the plantation, we intended to answer these specific questions: 1) why is the plantation declining? Is it because of the harm of acid rain, N saturation, or simply the genetically fast-growing inherence of the species, 2) what management strategies should be applied in order to maintain the sustainability of the plantation? 3) Is the area of the plantation large enough to provide sufficient organic fertilizers (litter) for the fruit garden? And 4)

ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM

277

How much litter can be moved from the plantation into the fruit garden without sacrificing the health of the plantation itself? 3. METHODS 3.1. Sampling scheme and chemical analysis Water samples of bulk precipitation, throughfall, stem flow, surface runoff, and soil solution were collected in 1991, 1994 and 2002. The sampling methods used in these samplings were mostly similar (see Yao and Yu, 1995; Fang et al., 1995; and Ding et al., 1995 for details). Rainfall was measured with a tipping bucket rain gauge, which is situated in the HS LTER weather station and ca. 150 m away from the plantation. Three funnel-type collectors adjacent to the rainfall gauge were used to collect bulk precipitation samples for chemical analysis. Sixteen throughfall collectors were placed randomly on the plantation floor to collect samples for chemical analysis. One large throughfall-collecting trough (with the surface area of 5 m2) was connected to a rain gauge and 4 rainfall gauges were randomly scattered on the plantation floor to record the amount of throughfall. Stem flow was intercepted using spiral collars of split vinyl tubing. Collars were installed on 7 tree boles, four of them representing 4 diameter classes of the trees were connected to rainfall gauges at the base of the tree for measuring the amount of stem flow and 3 of them were connected to plastic receptacles for collecting samples for chemical analysis. Surface runoff from the east-facing slope of the plantation (1/3 of the total area) was exported into a relatively large concrete trough and was continuously recorded through a V-notched weir. Soil solution was sampled at about 80 cm depth using zero-tension lysimeters in 1994 (Fang et al., 1995) and using ceramic cups in 2002. Most of the water samples for chemical analysis were undertaken monthly in 1991, fortnightly in 1994, and bimonthly in 2002. Water samples from the field were transported to the laboratory, filtered (0.45 μm), maintained at 4 ℃, and analyzed as quickly as possible. Measurements of pH were made with a glass electrode pH meter. Nitrate (NO3-N) and ammonium (NH4-N) were determined photometrically. Sulfate (SO4) was determined by using ion chromatography. Phosphate (PO4) was determined colorimetrically with the wet-ashing procedure. Cation elements (K, Ca, Na, and Mg) were measured by atomic adsorption spectroscopy (see details in Yao and Yu, 1995; Fang et al., 1995). Nutrient storages in plant biomass, litter, and soil pools were measured by Li et al. (1995a) and simulated using an ecosystem model (CENTURY) by Shen et al. (2003). Nine trees representing different diameter classes were harvested, weighted, and dried to establish the allometry equations for the species in 1992. These equations were used to estimate net primary productivity based on the inventory data on diameter and height from 1988 through 1995. The biomass of understorey shrubs

278

W. SHEN ET AL.

and grasses were also estimated based on the harvest in two 25 m2 and three 4 m2 quadrats, respectively. Ten 1m2 litter collectors were randomly located on the plantation floor for litter gathering, which was further sampled and analyzed monthly. Soil samples were extracted from two 1 m deep soil profiles. All plant, litter, and soil samples were transported to laboratory, dried at about 60 ℃, passed through a 1 mm mesh, and used for analyzing the concentration of N, P, S, K, Ca, Na, and Mg. The methods for chemical analysis on plant and soil samples were similar to those used for water samples (see details in Li et al., 1995a). Nutrient storages per unit area of different pools were calculated as concentration times the amount of biomass, litter, and soil per unit area. Subsurface runoff and groundwater seepage was very difficult to be quantified in our study system. Zero-tension lysimeters were installed to measure soil water seepage, but rarely received any water sample throughout year 2002. Therefore an ecosystem model, CENTURY (Parton et al., 1988; Parton et al., 1993) was parameterized, validated, and used to simulate the longterm (115 years since the establishment of the plantation in 1983) C, N, and H2O dynamics of the Acacia plantation (Shen et al., 2003). Subsurface runoff was derived from the simulated H2O budget. Simulated carbon storage and NPP were used to estimate nutrient storage and nutrient uptake based on measured fractional nutrient contents in different plant tissues (see Li et al., 1995a; Shen et al., 2003). 3.2. Element budgets Element fluxes associated with hydrologic flows were calculated as concentration times measured or simulated water fluxes, i.e., bulk precipitation, stemflow, throughfall, surface runoff, and soil seepage. According to Ulrich (1994) and Matzner et al. (2004), the total deposition of an element x is calculated as the sum of bulk deposition (BDx) and interception deposition (IDx):

TDx = BDx + IDx

x = N, P, S, K, Na, Ca, Mg

(1)

IDx comprises particle (IDp) and gas (IDg) portions. In order to calculate ID for x = N, P, K, Ca, Mg, the Na factor ( f Na = IDNa / BDNa ) was used by assuming that Na has low exchange between the plantation canopy, i.e., no Na is adsorbed by canopy. This assumption also applies to S. Thus,

ID p , Na = (TFx + SFs ) − BDx

ID p , x = f Na • BDx

x = Na, S x =N, P, K, Ca, Mg

(2) (3)

ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM

279

where TFx and SFx are throughfall and stem flow fluxes, respectively. TF and SF together can also be called as canopy precipitation ( CPx = TFx + SFx ). The gaseous depositions of N and S were ignored in this study. For any compartment of the ecosystem, element budgets are assessed by the inputs and outputs of the compartment. Three pools were considered in this study. Canopy element budget (CBx) or canopy leaching is the difference between total deposition and throughfall plus stem flow:

CBx = TDx − (TFx + SFx )

(4)

The soil element budget (SB) is defined as:

SBx = TDx + LFx − ( RFx + U x + GE x )

(5)

Where LF is element input into soil associated with litterfall, RF is surface runoff plus soil seepage, U is total plant uptake (i.e., the sum of annual plant biomass increment and annual litterfall), GE is gas emission that is ignored in this study because of lack of data and the small portion it accounts relative to U and RF. At the ecosystem level, element budget (EB) is the difference between total deposition and total output (mainly surface runoff and soil seepage):

EBx = TDx − ( RFx + GE x )

(6)

For N, soil and whole ecosystem budgets should also include biological fixation as part of the input, since A. mangium is an N fixing species. 4. TEMPORAL VARIATION IN ELEMENT CONCENTRATIONS Element concentrations in bulk precipitation (BP), throughfall (TF), stemflow (SF), and soil solution (SS) showed clear seasonal variation trends, with lower concentrations occurring in the wet period or growing season (April through September; Figure 2) and larger concentrations in the dry season. This was not the case for element concentrations in surface runoff (SR), which had higher element concentrations in the wet season (Figure 2). The lower wet-season element concentrations in BP, TF, SF, and SS may be attributed to frequent washing-off effects of rainfall in wet season, during which about 80% of the annual rainfall occurs in the research site (Shen et al., 1999). Concentrations of S, N, and Ca showed greater seasonal variability than other element species (Figure 2). The seasonal variation of Ca concentrations was the highest among all element species in 1991, while sulfate concentration showed the largest seasonal variation in 2002 (notice that sulfate was not measured in

280

W. SHEN ET AL.

1991). The absolute magnitude of sulfate concentrations in 2002 were also surprisingly high (e.g., 14.7 mg mg l-1 in October, 2002) compared with measurements from North American and Europe, so does N concentrations (predominantly nitrate). For example, the reported sulfate concentrations in BP were usually less than 5.0 mg l-1 in North American and Europe even back in 1980s and this value was much higher than in nowadays (see Barrie et al., 1984; Lynch et al., 1995; and Matzner et al., 2004). The pH value in BP varied from 3.8 to 4.8, with high values occurring in wet, hot summer and low values in dry, cool winter. pH values in TF, SF, SS, and SR were generally larger than in BP, but exhibited similar seasonal variation trend. Regarding the high sulfate and nitrate concentrations and low pH values recorded, the studied A. mangium plantation had been experiencing severe acid deposition in the past few decades, most likely due to increased industrial emissions resulting from the rapid industrialization in the region since the early 1980s. Element concentrations and seasonal variability generally increased in the following order: BP
ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM 12 Bulk deposition (1991)

2.5

Concentration (mg l-1)

Concentration (mg l-1)

3

2 1.5 1 0.5 0

1 2

3 4

5 6

7 8

6 4 2 0

1 2

3 4

5 6

7

8 9 10 11 12

20 Bulk deposition (2002)

16

N Na Mg

12

Concentration (mg l-1)

Concentration (mg l-1)

N K Na Ca Mg

8

9 10 11 12

20 K Ca S

8 4 0

16

Throughfall (2002)

12 8 4 0

Mar

May

Sep

Oct

Mar

60

May

Sep

Oct

Sep

Oct

Sep

Oct

60 Stemflow (2002)

Concentration (mg l-1)

Concentration (mg l-1)

Throughfall (1991)

10

0

0

50

281

40 30 20 10

50

Soil solution (2002)

40 30 20 10 0

0 May Sep Month

Mar

Oct

Figure 2. Seasonal variation in element concentrations in 1991 and 2002. Note that data of 1991 were obtained from Yao et al. (1995), who did not measure SO4-S.

May

60 Concentration (mg l-1)

Mar

50

Surface runoff (2002)

40 30 20 10 0 Mar

May Month

282

W. SHEN ET AL.

Figure 3. Comparing element concentrations in bulk precipitation (BP), throughfall (TF), stem flow (SF), soil seepage (SS), and surface runoff (SR) of the plantation. Note that data of 1991 and 1994 were obtained from Yao et al. (1995) and Fang et al. (1995), respectively.

ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM

283

5. ELEMENT FLUXES AND BUDGETS Table 1 summarizes the amount of element fluxes following water pathways, nutrient uptake by plants, element fluxes associated with litterfall, element storage in different compartment pools, and element budgets for canopy, soil and the whole ecosystem in two comparable years with very similar amount of water fluxes (1994 and 2002). Element fluxes also showed obvious seasonal variation trend, but with higher fluxes appearing in wet season due to larger water fluxes (data not shown). Table 1. Element fluxes and budgets (in kg ha-1 y-1) of the A. mangium plantation in 1994 and 2002.

Year 1994

H2O

pH

N

P

Bulk deposition

1676

4.40

52.5

1.2

118.3

6.7

17.3

4.4

N/A

Throughfall

1195

4.87

57.0

0.7

209.1

45.5

50.0

12.8

N/A

Stem flow

126

3.86

Annual fluxes in 1994

S

K

Ca

Mg

Na

a

Interception Total deposition

7.4

0.1

56.4

6.2

6.9

2.3

N/A

1.68

0.04

147.3

0.21

0.55

0.14

N/A

54.14

1.28

265.6

6.92

17.8

4.50

N/A

Seepage

267.2

5.15

13.55

0.21

17.98

3.05

7.88

1.87

N/A

Surface runoff

38.0

6.64

1.28

0.02

2.37

0.75

0.27

0.38

N/A

N-fixation

b

20.60

Compartment pools c (kg ha-1) Plant

249.3

31.92

N/A

157.9

343.8

86.19

N/A

Litter

201

12.00

N/A

20.00

73.00

19.00

N/A

5223

573.0

N/A

294.0

119.0

35.00

N/A

Biomass increment c

Soil

28.36

4.20

N/A

17.85

40.29

10.49

N/A

Litterfall c

96.74

5.53

N/A

9.80

34.98

9.08

N/A

Plant uptake

135.3

9.3

N/A

72.4

114.3

30.2

N/A

Return to soil

181.7

6.3

265.6

61.5

91.8

24.2

N/A

Canopy budget

-10.22

0.46

0.00

-44.77 -39.03 -10.63

N/A

Soil budget

31.55

-4.06

N/A

-14.73 -30.62

-8.24

N/A

Ecosystem budget

59.91

1.05

245.22

2.25

N/A

3.12

9.67

Table 1 (cont.)

284

W. SHEN ET AL.

Table 1 (cont.) Year 2002

H2O

pH

N

P

S

K

Ca

Mg

Na

1.01

120.7

16.81

32.38

1.81

40.64

Annual fluxes in 2002 Bulk deposition

1653

4.18

56.30

Throughfall

1179

d

4.26

56.94

0.79

147.6

35.84

28.30

4.37

35.27

Stem flow

124.0 d

3.71

11.05

0.10

44.58

6.26

6.04

0.97

6.63

Interception

1.75

0.03

71.46

0.52

1.00

0.06

1.26

Total deposition

58.05

1.04

192.2

17.33

33.38

1.86

41.90

37.13

0.08

73.28

13.21

26.28

8.23

20.54

10.76

3.13

5.08

0.36

2.94

N/A

311.2

490.9

182.19

N/A

Seepage d d

255.2 d

5.59

d

7.13

1.38

0.00

20.60

10.50

Plant

341

54.30

Litter

189.2

7.20

N/A

56.87

47.63

9.86

69.87

Soil

4986

592.2

N/A

280.7

113.6

33.4

N/A

28.36

22.45

3.07

N/A

20.07

19.36

9.14

Surface runoff

38.00

N-fixation d Compartment pools d (kg ha-1)

Biomass increment d

d

96.74

202

7.69

N/A

60.77

50.90

10.54

135.3

225.2

10.6

N/A

105.6

71.2

23.2

Return to soil

181.7

263.7

N/A

84.0

121.9

102.6

31.7

Canopy budget

-10.22

-0.61

0.15

0.00

-24.77

-0.96

-3.48

Soil budget

31.55

16.91

8.58

N/A

-19.08 -17.34 -15.86

Ecosystem budget

59.91

39.37

0.95

108.1

Litterfall

Plant uptake d

a

b

0.99

2.03

-6.73

c

Note: derived from Fang et al. (1995); obtained from Ding et al. (1995); derived from Li et al. (1995a); d derived from the CENTURY model simulated values (Shen et al., 2003); e provided by Dr. Zhian Li, the measured litterfall was 10.66 t ha-1 y-1 for 2002.

The pH values in BP, TF and SF decreased by 4-12.5% from 1994 to 2002. Whereas pH values in SS and SR increased by 8.5% and 7.3%, respectively (Table 1). These indicate that the atmospheric acid deposition had been becoming more severe; but soil-buffering capacity on acidity had been improved with the development of the plantation and resultant improvements of soil conditions. Fluxes of most element species associated with bulk deposition increased from 1994 to 2002, although N and S fluxes only slightly increased by 7.2% and 2% compared with K (150.9%) and Ca (87.2%) fluxes. The increases in element fluxes further indicated that the atmospheric deposition into the studied plantation had been increased during the study period. In contrast, element fluxes with throughfall and stem flow decreased from 1994 to 2002. This trend was also clearly reflected in decreased canopy budgets, which became less negative for most of the elements. We hypothesized that this was mainly due to increased canopy adsorption efficiency (Shen et al., 2003).

ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM

285

Element fluxes with soil seepage and surface runoff largely increased by about 23 times in 2002 than in 1994. Increase in N fluxes with SS was smaller (174%) compared to S (307%), K (333%), Ca (233%), and Mg (334%) fluxes. Similar variation trend occurred in fluxes with SR, where N fluxes increased ca. 7.4% compared with 353% for S, 318% for K, 181% for Ca, and 5.5% for Mg, respectively. These increases in element output fluxes are mainly ascribed to increased total atmospheric deposition. However, P fluxes with SS and RS seemed slightly decreased. This could be due to increased plant uptake (see Li et al., 1995a) and/or measurement errors, since P concentrations were frequently under detection limit, especially in 2002. Overall, increases in element fluxes with SS were larger than the ones with RS because of larger water fluxes associated with SS. Annual element fluxes were relatively small compared to element storages in compartment pools, of which soil pool was the largest element storage. The storage of elements in the plant pool markedly increased from 1994 to 2002, reflecting the growth of the plantation during the last 11 years. In contrast, litter and soil element pool decreased from 1994 to 2002 because of increased plant uptake (for N and P) or outputs with SS and SR (for K, Ca, and Mg). Interestingly, annual biomass increment slightly decreased from 1994 to 2002 but litterfall increased dramatically, which means that a large amount of nutrients uptaken by plants was returned to the soil as litterfall. We do not know it this was one of the consequences resulting from severe acid deposition or simply the inherent development characteristics of the species. In terms of element budgets of the canopy and soil pools, the plantation acted as a source of N, k, Ca, and Mg, and a small sink for P in both of the two years (1994 and 2002); but the annual canopy leaching of N, K, Ca, and Mg decreased by 94%, 45%, 98%, and 67%, respectively, from 1994 to 2002. This decrease in canopy leaching was the result of decreased living foliar biomass resulting from increased annual litterfall. The soil nutrient pool was a sink for N, but sources for K, Ca, and Mg in both of the two years, but the soil N budget decreased by 46% from 31.6 kg N ha-1 yr-1 in 1994 to 16.9 kg N ha-1 yr-1 in 2002. This decrease was mainly caused by the decrease in N fixation, which decreased from 20.6 kg N ha-1 yr-1in 1994 to 10.5 kg N ha-1 yr-1 in 2002. The soil K and Mg budgets became more negative (from 14.7 to -19.1 kg N ha-1 yr-1 for K and -8.2 to -15.9 kg ha-1 yr-1) from 1994 to 2002, indicating that K and Mg leaching were increased from the plantation soil via soil seepage and surface runoff. In contrasting, soil Ca budget increased from -30.6 kg N ha-1 yr-1 in 1994 to -17.3 kg N ha-1 yr-1 in 2002, mainly due to decreased plant uptake (see Table 1). At the whole ecosystem level, the A. mangium plantation acted as a sink for most of the elements examined in both of the two years, but a general decreasing trend was observed from 1994 to 2002. The nutrient retention ability of the system decreased markedly, with N retention decreased by 34%, P by 10%, S by 56%, K by 68%, Ca by 79, and Mg by 400%. This indicates that as the plantation ages, its ability to retain nutrients decreases, and the risk of nutrient losses from the system increases.

286

W. SHEN ET AL.

6. IMPLICATIONS FOR ECOSYSTEM MANAGEMENT Use of input-output nutrient budgets as a diagnostic tool to identify trends in soil fertility that determine ecosystem productivity could date back to the beginning of modern research on biogeochemistry of watershed ecosystems (Ulrich, 1987; Likens et al., 1995). The input-output budget analysis can be performed at various temporal and spatial scales. As in this study, we assessed element budgets of the whole forest ecosystem, as well as its two major compartments: canopy and soil, on annual bases. Nutrient budgets, be it balanced, positive, or negative, could have different indications. Taking the element budget of soil pool as an example, a balanced budget indicates that the element of interest is not depleted from the soil, but this does not mean no changes occur in the ecosystem; A positive budget means that the element of interest is accumulating in the soil, but this is not always favorable; a negative budget means that the element considered is depleted in the soil, but it has to be related to the available soil nutrient stocks and to the rate of flux into available pool to be a real index of risk for the ecosystem (Ranger and Turpault, 1999). In this section, we will discuss what the budget information of different compartment pools could mean to the management of the plantation and of the agroforestry landscape in which it is embedded. Our budget analyses showed that the A. mangium plantation accumulated a large amount of nutrients in plant, litter, and soil pools (Table 1). This confirms that the fastgrowing tree species, A. mangium can tolerate the relatively poor soil conditions due to its ability to fix atmospheric N and its high resource use efficiency (Li et al. 1995a). By comparing soil organic matter (SOM) and total N content before and after afforestation, Li et al. (1995a) found that SOM increased by ca. 38% (from 1.15 g kg-1 to 1.58g kg-1) and soil total N increased by 61.3% (from 0.58 to 0.93 g kg-1). Due to structural and functional developments and soil condition improvements, the A. mangium plantation showed a certain extent of buffering capacity on soil acidification resulting from severe acid rain (see descriptions in Section 4). Therefore, we think that A. mangium is a very good pioneer species that can be used to restore degraded subtropical grass slopes, but under environmental stresses such as severe acid rain, A. mangium plantation may only function properly during the first 20-30 years after its establishment. This means that human-aided management strategies are need in order to sustain better services for longer time periods. What management strategies should be applied based on the nutrient budget of the A. mangium plantation? First of all, the observed data clearly showed that it has been experiencing severe acid deposition in the last 2-3 decades, as a result of dramatically increased industrial pollution since China initiated its economic reform in the early 1980s. Therefore liming of the plantation soil is suggested. Further studies are needed to determine the critical load of acidification and the amount of lime that should be added to the plantation. Secondly, our budget analysis showed that elements such as K, Mg were depleted from the plantation soil. Considering P is a common limiting nutrient to such subtropical forest soils (Li et al., 1995b), we suggest that applying a certain amount of composite fertilizers enriched with P, K, and Mg may be helpful in sustaining the health of the plantation. Although the budget information in this study has some referring values, the exact amount of

ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM

287

fertilizers that should be applied to the plantation needs more detailed study. Thirdly, this study showed that the A. mangium plantation was not able to sustain its structure and functioning for a relatively long time period (e.g., >50 years). Obvious canopy declinations had been reported in previous studies (Shen et al., 1999). Therefore structural transformation of the plantation is needed at the age of ca. 2030 years. Native tree species such as Schima superba, which has relatively high productivity and better resistance to acid deposition, can be introduced deliberately into the A. mangium plantation, and thus gradually replace the pioneer community (Shen et al., 2003). What can the A. mangium plantation provide for the sustainability of the agroforestry landscape? One of the main intentions of establishing a forest stripe in the upper slope of the watershed was to prevent soil erosion and to improve land productivity. Another intention was to provide some amount of litter that can be used as fertilizer for the other landscape components (i.e. the fruit garden and the grass meadowland). The element budget analyses in this study confirms that this idea is adoptable and practical, for the reason that there was a relatively large amount of litterfall (ca. 200 kg ha-1) stored in the plantation floor, and a large annual litterfall rate (ca. 100-200 kg ha-1 y-1; see Table 2). Based on the information of ecosystem budget, pool size, and plant uptake of nutrients (especially N), we propose here that about 1/3 – 1/6 of the total litterfall stored in the plantation floor could be moved into the fruit garden or the grass meadowland as organic fertilizers. Whether the plantation should be placed in the upper slope of the watershed or in the lower slope (adjacent to the fish pond, see Figure 1) as a riparian forest was an interesting argument during the beginning stage of establishing the agroforestry landscape. In one hand, the upper slope with relatively poor soil condition was considered more labor- and resource-costly to grow fruit trees therefore the A. mangium might be more appropriate for the upper slope. In the other hand, riparian forest could act as a buffer zone, retaining and keeping excessive upper slope originated nutrients from entering into downstream water bodies (Haycock et al., 1997). Based on the budget analyses of this study, we think that the positioning the A. mangium plantation in the upper slope of the watershed was correct, mainly because this plantation was enriched with N due to its N fixation ability plus the large atmospheric deposition (see Table 1). If it is placed adjacent to the water body, excessive nutrients, especially N might directly drain into the fish pond, plus nutrients leaching out from the fruit garden, may increase the chance of water eutrophication of the fish pond. However, for large watersheds that are occupied by similar agroforestry landscape as in this study, riparian forest may be critical; extra cautions need to be put on the practice of constructing riparian vegetation dominated by N-fixing species, especially in areas with large atmospheric N deposition. 7. CONCLUDING REMARKS By comparing nutritional element fluxes and budgets of the A. mangium plantation chronologically, we found that it acted as a sink of elements at the early stage of development (15-20 years since establishment), during which the plantation

288

W. SHEN ET AL.

accumulated a large amount of nutrients in its compartment pools: plant biomass, forest floor, and soil. But its nutrient retention ability decreased fast when the plantation approached the near-mature stage, partly because of the inherent fastgrowing nature of the tree species, partly because of large atmospheric deposition resulted from markedly increased industrial emissions since China initiated the Economic Reform Policy in the region. Therefore, under large atmospheric element deposition, near-mature A. mangium plantation may increase the risk of N saturation and suffer from acid deposition and nutrient depletion (e.g., P, K, and Mg; Aber et al., 1989; Aber et al., 1998; Godbold and Huttermann, 1994). Nevertheless, the pioneer A. mangium plantation is valuable in affforesting degraded hilly land with poor soil conditions in subtropical regions of China. In order to sustain its nutrient retention ability and other ecosystem services, we think that application of composite P, K and Mg fertilizers and man-aided structural transformation of the plantation (i.e., some native tree species that are tolerant to acid deposition and under-canopy low light conditions) are needed for this type of forests at the age of ca. 20-30 years. This study confirmed that positioning the A. mangium plantation in the upper slope of the watershed was an appropriate spatial arrangement for the agroforestry landscape, because this may decrease the risk of transporting excessive elements into the lowland fishpond via soil seepage and surface runoff. The A. mangium plantation could also provide some nutrients to other components of the agroforestry landscape because of the large amount of litter stored on the plantation floor. Based on element budget information of the plantation, we suggested that about 1/3 – 1/6 of litter could be move into the fruit garden or the grassland as organic fertilizers. For other management issues such as determination of the best area composition of the four landscape components (plantation, fruit garden, grassland, and fishpond) and maintenance of sustainability of the landscape, a comprehensive study on element fluxes and budgets of the other three landscape components is necessary. Combining experimental approaches as applied in this study with spatially interactive landscape modeling approach may be valuable in exploring not only these management issues but also in understanding general mechanisms of interactions between landscape pattern and biogeochemical processes. ACKNOWLEDGEMENTS We thank Dr. Fernando Maestre for his valuable comments and assistance with improving the language of the first manuscript. This research was supported by Natural Science Foundation of China (30570274 and 30100021) and Guangdong Sci-Tech Planning Programme (2005B33302012). REFERENCES Aber, J.D., McDowell, W., Nadelhoffer, K.J., Magill, A., Berntson, G., Kamakea, M. et al. (1998). Nitrogen saturation in temperate forest ecosytems: Hypotheses revisited. BioScience, 48, 921-934.

ELEMENT FLUXES AND BUDGETS OF A FOREST ECOSYSTEM

289

Aber, J.D., Nadelhoffer, K.J., Steudler, P. and Melillo, J.M. (1989). Nitrogen saturation in Northern forest ecosystems: Excess nitrogen from fossil fuel combustion may stress the biosphere. BioScience, 39, 378-386. Attiwill, P.M. and Adams, M.A. (1993). Tansley Review No. 50: Nutrient cycling in forests. New Phytologist, 124, 561-582. Barrie, L.A. and Hales, J.M. (1984). The spatial distributions of precipitation acidity and major ion wet deposition in North America during 1980. Tellus, 36B, 333-355. Bormann, F.H. and Likens, G.E. (1967). Nutrient cycling: Small watersheds can provide invaluable information about terrestrial ecosystems. Science, 155, 424-429. Ding, M., Peng, S., Yu, Z., Li, Z. and Fang, W. (1995). Nutrient cycling in a complex ecosystem of forest, orchard, grassland and fishpond in Heshan, Guangdong. Acta Ecologica Sinica, 15, 82-92 Dobrovolsky, V.V. (1994). Biogeochemistry of the World's Land. CRC Press, Boca Raton, Florida. Fang, W., Ding, M., Lu, D., Li, Z., Cai, X., Zhou, G. et al. (1995). Hydrological dynamics and nutrient migration with precipitation of artificial Acacia mangium forest in low subtropical downland. Acta Ecologica Sinica, 15 (Suppl. A), 115-123 Fu, B., Ma, K., Zhou, H. and Chen, L. (1998). Effects of land use structure on soil nutrient distribution in the Loess Plateau. Chinese Science Bulletin, 43, 2444-2448 Godbold, D.L., and Huttermann, A. (Eds.). (1994). Effects of Acid Rain on Forest Processes. John Wiley & Sons, Inc., New York. Haycock, N., Burt, T., Goulding, K. and Pinay, G. (Eds.). (1997). Buffer Zones: Their Processes and Potential in Water Protection. The Proceedings of the International Conference on Buffer Zones, September 1996. Quest Environmental, Harpenden, Hertfordshire, UK. Kratz, T.K., Webster, K.E., Bowser, C.J., Magnuson, J.J. and Benson, B.J. (1997). The influence of landscape position on lakes in northern Wisconsin. Freshwater Biology, 37, 209-217. Li, Z., Ding, M., Fang, W., Weng, H. and Cai, X. (1995a). The nutrient storage and distribution in artificial Acacia mangium forest. Acta Ecologica Sinica, 15 (Suppl. A), 103-114 Li, Z., Fang, W. and Lu, D. (1995b). Physical and chemical properties of soils in Heshan hilly land. Acta Ecologica Sinica, 15 (Suppl. A), 93-102. Likens, G.E. and Bormann, F.H. (1995). Biogeochemistry of A Forested Ecosystem, 2nd ed. SpringerVerlag, New York. Liu, J., Wen, D. and Zhou, G. (2000). Chemical properties of the rainfall in the coniferous and broadleaved forests in acid rain area of Heshan, Guangdong. China Environmental Science, 20, 198-202. Lynch, J.A., Bowersox, V.C. and Grimm, J.W. (2000). Changes in sulfate deposition in eastern USA following implementation of Phase I of Title IV of the Clean Air Act Amendments of 1990. Atmospheric Environment, 34, 1665-1680. Matzner, E., Alewell, C., Bittersohl, J., Lischeid, G., Kammerer, G., Manderscheid, B. et al. (2001). Biogeochemistry of a spruce forest catchment of the Fichtelgebirge in response to changing atmospheric deposition. In J.D. Tenhunen, R. Lenz and R. Hantschel (Eds.), Ecosystem Approaches to Landscape Management in Central Europe (Vol. Ecological Studies, pp. 463-503). Springer, Berlin. Matzner, E., Zuber, T., Alewell, C., Lischeid, G. and Moritz, K. (2004). Trends in deposition and canopy leaching of mineral elements as indicated by bulk deposition and throughfall measurements. In E. Matzner (Ed.), Biogeochemistry of Forested Catchments in a Changing Environment. SpringerVerlag, Berlin. Parton, W.J., Scurlock, J.M.O., Ojima, D.S., Gilmanov, T.G., Scholes, R.J., Schimel, D.S., et al. (1993). Observations and modeling of biomass and soil organic matter dynamics for the grassland biome worldwide. Global Biogeochemical Cycles, 7, 785-809. Parton, W.J., Stewart, W.B., and Cole, C.V. (1988). Dynamics of C, N, P and S in grassland soils: a model. Biogeochemistry, 5, 109-131. Peterjohn, W.T. and Correll, D.L. (1984). Nutrient dynamics in an agricultural watershed: observations on the role of a riparian forest. Ecology, 65, 1466-1475. Ranger, J. and Turpault, M.-P. (1999). Input-output nutrient budgets as a diagnostic tool for sustainable forest management. Forest Ecology and Management, 122, 139-154. Ryszkowski, L. (Ed.). (2001). Landscape Ecology in Agroecosystems Management. CRC Press. New York.

290

W. SHEN ET AL.

Schlesinger, W.H. (1997). Biogeochemistry, An Analysis of Global Change (Second Edition). Academic Press, San Diego. Shaver, G.R., Knadelhoffer, K.J. and Giblin, A.E. (1991). Biogeochemical diversity and element transport in a heterogeneous landscape, the north slope of Alaska. In M. G. Turner and R. H. Gardner (Eds.), Quantitative Methods in Landscape Ecology (pp. 105-125). Springer-Verlag, New York. Shen, W., Peng, S., Wu, J., and Lin, Y. (2003). Simulation studies on carbon and nitrogen accumulation and its allocation pattern in forest ecosystems of Heshan in low subtropical China Acta Phytoecologica Sinica, 27, 690-699 Shen, W., Zhou, G., Peng, S., Yu, Z., Lin, Y. and Zeng, Y. (1999). Surface runoff in five ecosystems of Heshan subtropical hilly land Journal of Tropical and Subtropical Botany, 7, 273-281 Tenhunen, J.D. and Kabat, P. (Eds.). (1999). Integrating Hydrology, Ecosystem Dynamics, and Biogeochemistry in Complex Landscapes. John Wiley & Sons, Chichester. Tenhunen, J.D., Lenz, R. and Hantschel, R. (Eds.). (2000). Ecosystem Approaches to Landscape Management in Central Europe. Springer, Berlin. Turner, M.G., Gardner, R.H. and O'Neill, R.V. (2001). Landscape Ecology in Theory and Practice: Pattern and Process. Springer-Verlag, New York. Ulrich, B. (1987). Stability, elasticity, and resilience of terrestrial ecosystems with respect to matter balance. In E.-D. Schulze and F. Zwolfer (Eds.), Potentials and Limitations of Ecosystem Analysis. Ecological Studies (Vol. 61, pp. 11-49). Springer, Berlin. Ulrich, B. (1994). Nutrient and acid-base budget of Central European forest ecosystems. In D.L. Goodbold and A. Huttermann (Eds.), Effects of Acid Rain on Forest Processes. John Wiley & Sons, New York. Vitousek, P.M. and Howarth, R.W. (1991). Nitrogen limitation on land and in the sea: How can it occur? Biogeochemistry, 13, 87-115. Wu, J., and Hobbs, R.J. (2002). Key issues and research topics in landscape ecology. Landscape Ecology, (in review). Yao, W. and Yu, Z. (1995). The nutrient content of throughfall inside the artificial forests on downland Acta Ecologica Sinica, 15 (Suppl.), 124-131. Yu, Z. and Peng, S. (Eds.). (1996). Ecological Studies on Vegetation Rehabilitation of Tropical and Subtropical Degraded Ecosystems. Guangdong Science & Technology Press, Guangzhou, China (in Chinese).

CHAPTER 18

THE EFFECTS OF THE REGULATION SYSTEM ON THE STRUCTURE AND DYNAMICS OF GREEN SPACE IN AN URBAN LANDSCAPE The case of Kitakyushu City

T. MANABE1, K. ITO2, D. ISONO2, T. UMENO3 1

Kitakyushu Museum of Natural History and Human History, Kitakyushu, 8050071, Japan; 2 Department Civil Engineering, Faculty of Engineering, Kyushu Institute of Technology, Kitakyushu, 804-8550, Japan; 3 Graduate School of Civil Engineering, Kyushu Institute of Technologym, Kitakyushu, 804-8550, Japan

Abstract. The effects of a regulation system on conserving the green space were evaluated in Kitakyushu City, southern Japan. Nearly half of the city is under the Urbanization Control Area that should restrain urbanization, and about 30% of the city is specified as Scenic Zones and Green Conservation Areas where their use is restricted by a regulation system. Area of green spaces within the Urbanization Control Area decreased slightly from 1984 to 2001, although those within Urbanization Promotion Area decreased largely. The specification for the Urbanization Control Area, therefore, plays a role in conserving area of green spaces. Specifying Scenic Zones as well as Green Conservation Areas also have value in retaining green spaces. Some woodland was, however, transferred to residential areas within the Green Conservation Area. This decrease in woodlands was due to constructing a City Planning Road, suggesting that area of green spaces even within a Green Conservation Area depends on decisions made by the municipalities. The habitat function of the forests for dominant canopy and sub-canopy evergreen broad-leaved trees was also evaluated by examining the relationships between stem densities at different growth stage (seedling, sapling and mature). Success of seedling recruitment of Castanopsis cuspidate (Fagaceae), for which the seed-dispersal type is classified as chasing dispersal, was depended largely on the existence of conspecific mature trees. Thus, the forests with a low density of conspecific matures have low habitat function for the species even if safe-sites for seedling recruitment exist. There were no clear relationships between densities at each growth stage for Persea thunbergii, Neolitsea sericea, and Cinnamomum japonicum of the family Lauraceae that the seed-dispersal type is classified as endozoochory. This finding might suggest that the habitat function of the forests for these three was not controlled by the ‘dispersal limitation’ as seen in Castanopsis but by the micro-environmental conditions of the forests.

291 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 291–309. © 2007 Springer.

292

T. MANABE ET AL.

1. INTRODUCTION Most secondary vegetation, such as forests and grasslands, in suburban and farm districts was profoundly related to our daily life. In Japan, secondary vegetation has been maintained by traditional methods and has persisted under human impact and created its own particular ecosystems (Kamada et al., 1991). The traditional landscape in districts containing secondary vegetation is called ‘Satoyama’. However, the ‘fuel revolution’, a rapid transformation of the main energy source from biomass to fossil fuels in the mid-1960s, brought about large lifestyle changes and diminished the socio-economic value of secondary vegetation. This also led to a large decrease and a drastic isolation of the vegetation in the landscape (Iida and Nakashizuka, 1995; Fukamachi et al., 1996). Furthermore, these landscape changes altered the species composition and community structure of the vegetation through ecological succession (Kamada and Nakagoshi, 1990, 1996; Hong et al., 1995; Manabe et al., 2003). These qualitative and quantitative changes in the Satoyama landscapes have driven some organisms living in the Satoyama to local extinction (Senior, 2005). The decrease in green space in urban areas, especially within large cities, was already a big problem before the 1960s. Green space in cities was conserved under regulations such as the City Park Law enacted in 1956, the City Planning Law in 1964, the Urban Green Space Conservation Law in 1973 (revised as the Urban Green Space Law in 2004) and so on. These regulations focused mainly on the socio-economic functions of the green spaces, such as disaster prevention, recreation and scenery. However, the idea that the green spaces in the urban landscapes could have other ecological function, such as habitat, conduit and a source of wildlife (Forman, 1995), was accepted only recently. To realize such an idea, new management systems should be prepared and proposed. For example, vegetation corridors should be used to connect small green spaces to large green spaces with high habitat function in order to conserve high species diversity within the urban area. When making such ecological networks, present and potential habitat function for wildlife must be evaluated by appropriate methods (Kaku et al., 2004). Further, there is a need to confirm that green space is conserved under present regulations. The goal of our study is to construct relevant management systems for urban green space for humans and various wildlife in Kitakyushu City and to develop a system as a model for a large city in warm-temperate Japan. In this chapter we describe: 1) the changing patterns of landscape structure during the 17 years from 1984 to 2001, 2) the effects of a regulation system on conserving the green space in the city. We also describe 3) the changing pattern of green spaces during the 78 years from 1922 to 2000, and then evaluate 4) the habitat function of the green spaces for woody species in the mid-northern part of the city.

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

293

2. KITAKYUSHU CITY In 1963, Kitakyushu City was created by merging five cities, with each having its own peculiar socio-economic and cultural backgrounds; the cities were Moji City, Kokura City, Tobata City, Wakamatsu City and Yahata City. Kitakyushu City has an area of 485.55 km2 and a population of about 1 million in 2005 (Figure 1).

Figure 1. Study site (Kitakyushu City).

The first public iron-manufacturing foundry was constructed in the city (ex-Yahata City) in 1901; since then, the city has developed as ‘the town of iron foundries’. Many forests of various areas, shapes and connectivity are in the city. The main potential natural vegetation is evergreen broad-leaved forest dominated by evergreen canopy species such as Castanopsis cuspidata, Quercus salicina, Distylium racemosum and Persea thunbergii, and evergreen sub-canopy species such as Cinnamomum japonicum, Camellia japonica and Cleyera japonica. But now, the secondary broad-leaved forest is the dominant forest ecosystem, and the natural vegetation can only be seen in some shrine/temple forests. 3. CHANGING PATTERNS OF LANDSCAPE ELEMENTS IN THE CITY What landscape elements exist in the city? How did the distribution of the elements change? What types of factors affect the distribution and dynamics of the elements? In this section, we describe the distribution and changing patterns of landscape elements in the city, and then discuss the effects of the regulation system on the distribution and dynamics of the elements, especially of green spaces.

294

T. MANABE ET AL.

3.1 Landscape elements The digital land cover maps in 1984, 1989, 1995 and 2001 were made by using remote sensing data (Landsat data). Changing patterns of the landscape were analysed by overlapping these maps using GIS (Arc View 8.3). In this analysis, six elements were drawn for land cover types: woodlands (broad-leaved forests, needleleaved forests, needle-leaved plantations, phyllostachys forests and shrubs), grassland, cultivated land, denuded land, open water and residential areas (housing lots and roads) and others (lands that could not be identified because their surface was covered by clouds). Table 1. Area (ha) of each landscape element in each year.

Elements Woodlands

Year 1984

1989

1995

2001

21087.54

19311.93

20816.46

19994.22

Grasslands

4339.62

4429.44

6823.08

5127.12

Cultivated lands

8053.47

9399.42

6996.78

6389.91

1461.6

2238.84

2759.67

1997.28

Denuded land Open water

2349.27

849.33

1070.55

698.85

Residential area

10867.5

11004.84

9692.46

13951.62

0

925.2

0

0

Others

Woodlands, which covered about 40% of the city, were dominant in the area, followed by residential areas and cultivated land in every year studied (Table 1). Cultivated land largely decreased from 1989 to 1995, and residential areas increased from 1995 to 2001. These changes indicate that cultivated lands have been transferred to residential areas since the 1980s, and the area of other landscape elements changed slightly during this period. The changes in landscape elements are ascribed to the structural changes of industry and to socio-economic changes; these changes were caused by a large increase in food imports, a policy of reducing the amount of paddy fields, an increase in the proportion of elderly people engaged in agriculture and so on. 3.2 Lands limited in their use by the regulation system The area that should be developed, improved or conserved comprehensively is specified as the City Planning Area by administrative organs of each prefecture based on the City Planning Law. The area avoids disordered development under the protection of the City Planning Law and other relevant laws. The City Planning Area is classified into two types: the Urbanization Promotion Area, which contains areas that are already urbanized and that should be urbanized during the next decade, and the Urbanization Control (Coordination) Area, which should have restrained

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

295

urbanization. The City Planning Area and Urbanization Control Area occupied more than 99% and 58% of the city in 2001, respectively (Table 2). Table 2. The ratio (%) of each zone classified by each regulation to the total city area in 2001.

Zone

City Planning Area Urbanization Urbanization Promotion Area Control Area Scenic Zone 1.98 27.71 Green Conservation Area 0.14 0.02 Nature Park *) 0.10 14.03 Urban Park 1.46 0.80 Others 37.89 15.70 Total 41.57 58.26 *) National park and quasi-national park

Others

0.17 0.17

The Scenic Zone is specified by the municipalities at sites of scenic value, and has importance for conserving urban environments and for city planning within the City Planning Area. The area is conserved under the protection of the City Planning Law and other relevant laws, and construction of buildings in the area can be restricted by municipal regulations. In 1967, the city specified 15 Scenic Zones (recent Scenic Zones) on the basis of the City Planning Law (revised in 1968). Their total area was 12,870 ha (28% of the city area), and 94% of the zones were located within the Urbanization Control Area (Table 2). In the City, the Scenic Zone (old Scenic Zone) was specified in 1936 based on the City Planning Law enacted in 1918. Recent Scenic Zones are far wider than the old Scenic Zones (Figure 2). Recent Scenic Zones decreased in the centre of the city, although they increased largely in mountain and coastal regions. Increases in the area in mountain regions were largely due to city planning by the municipality to conserve the panoramic view from the centre of the city. It is valuable for conserving the area of green spaces. It should also be noted that new Scenic Zones decreased at the centre of the city, where spaces might have become cores of the network of green spaces in urban regions. Green Conservation Areas, that have good natural environments with scenic value, cultural value, importance for residents’ ordinary life and habitat function for wildlife, can be specified by municipalities within the City Planning Area. The areas are also conserved under the protection of the City Planning Law and other relevant laws. Green Conservation Areas have been specified since 1974, and there were 17, covering 76.8 ha, in 2001; 85.5% of the areas were within the Urbanization Promotion Area.

296

T. MANABE ET AL.

Figure 2. Old Scenic Zones designated in 1936 and new Scenic Zones designated in 1967. Dashed lines and solid lines indicate Scenic Zones specified in 1968 and in 1918, respectively. The narrow line indicates boundaries of each ward.

There were 1559 City Parks specified by the City Park Law in the city in 2001. They covered 1055.7 ha, and about 65% in area were located within the Urbanization Promotion Area. Nature parks, which are contained within the National Park and Quasi-National Park system, are also restricted in land use by a regulation system. One National Park with an area of 6756.2 ha and three QuasiNational Parks of 47.3 ha are specified in the city, and 99.3% of the area of these parks is located within the Urbanization Control Area. In summary, nearly half of the city is under the Urbanization Control Area that should restrain urbanization, and about 30% of the city is specified as Scenic Zones and Green Conservation Areas where their use is restricted by a regulation system. On the other hand, Kaku et al. (2004) showed that green spaces within the Urbanization Control Areas in the mountain regions made up the vast continuous green zones in the city. In contrast, many small green spaces, most of which are City Parks, are distributed patchily within the Urbanization Promotion Area in the middle of the city. Thus, connectivity of green spaces in the region is very low, suggesting some ecological functions are not always high. In order to improve the ecological functions of the city, especially at the centre, we must pay attention to the connectivity as well as to the area of the green spaces.

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

297

3.3 Does the regulation system contributes to conserving green space? Large parts of the land are restricted in their usage by the regulation system in Kitakyushu City. Does this regulation system really contribute to conserving green space? In this section, we summarize distribution patterns of green spaces, and also discuss the effects of the regulation system on conservation of green spaces in the city. Hereafter, we define green space as the land covered by woodland and grassland. Table 3. Area (ha) of each landscape element at each City Plannning Area. Landscape element

Year 1984

1989

1995

2001

Woodlands

1885.3

1287.5

1602.9

1060.2

Grasslands

1739.5

1450.4

3663.5

2151.8

Cultivated lands

4135.0

4977.3

2987.6

2612.1

Urbanization Promotion Area

Denuded land

720.2

1328.3

2101.7

1411.3

Open water

1739.0

545.8

438.0

188.2

Residential area

9800.8

10094.0

9226.0

12596.2

0.0

336.6

0.0

0.0

19152.6

17978.8

19164.3

18888.3

Others Urbanization Control Area Woodlands Grasslands

2595.1

2973.9

3151.4

2964.8

Cultivated lands

3907.8

4407.2

3991.1

3761.5

740.5

908.1

654.7

584.6

Denuded land Open water Residential area Others

595.9

292.3

628.9

504.8

1063.6

906.7

465.2

1351.6

0.0

588.6

0.0

0.0

About 80% of green spaces were located within the Urbanization Control Area, and woodlands occupied most parts of these green spaces in 2001 (Table 3). More than 90% of woodlands were located within the Urbanization Control Area in each year. These patterns were due to the distribution patterns of the Urbanization Control Area of the city; most of the area is mountainous and contains continuous secondary forests and plantations. The specification for the Urbanization Control Area, therefore, plays a role in conserving areas of green space, especially woodlands. In Scenic Zones, green spaces cover about 80% of the area, and woodlands about 90% of this area each year (Table 4). These patterns were due to the distribution patterns of Scenic Zones, most of which are distributed within the Urbanization

298

T. MANABE ET AL.

Control Area containing more than 80% of woodlands in the city. Specifying Scenic Zones has value in retaining green spaces, especially woodlands. On the other hand, some residential areas changed into woodlands at the Scenic Zone within the Urbanization Promotion Area. This is attributed to the city municipal regulations that require developers to turn more than 10% of the area that is converted into residential area and the owner of each house to turn more than 20% of each housing area into green space in the Scenic Zone. The specification of a Scenic Zone, therefore, would very likely conserve an area of green space. The municipal regulation, however, means that existing woodlands are not always conserved when the area is converted into housing lots, as long as more than 20% of the area is planted, even though the woodlands are within the Scenic Zone. Table 4. Area (ha) of each landscape element at Scenic Zone and Green Conservation Area. Landscape element

Year 1984

1989

1995

2001

11115.6

10736.1

11391.6

11194.9

a) Scenic Zone Woodlands Grasslands

1073.3

1030.0

1020.2

967.8

Cultivated lands

1140.0

1319.8

1091.5

925.5

Denuded land

201.4

266.0

162.4

191.3

Open water

410.9

213.6

384.4

328.3

Residential area

355.0

389.8

246.2

688.6

0.0

341.2

0.0

0.0

Woodlands

39.8

52.5

55.5

50.7

Grasslands

6.3

4.8

4.8

5.6

10.4

12.9

6.8

8.1

0.2

0.5

0.8

1.2

14.2

0.5

8.3

5.7

Others b) Green Conservation Area

Cultivated lands Denuded land Open water Residential area

6.0

5.8

0.6

5.7

Others

0.0

0.1

0.0

0.0

Green spaces exceeded 55 ha within the Green Conservation Area in each year except for 1984 (Table 4). This is ascribed to the effects of the Urban Green Space Conservation Law (now revised to the Urban Green Space Law), which obligates the municipalities to buy green space within the Green Conservation Area. Actually, more than 77% of the woodlands were not transferred to other landscape elements, although most grassland was transferred to other elements during three periods (1984–1989, 1989–1995 and 1995–2001; Table 5). Thus, the specification of a Green Conservation Area would strongly conserve the area for green spaces,

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

299

especially woodlands, in the City Planning Area, because most of the Green Conservation Area is located within the City Planning Area. However, to a large extent, it is difficult to specify a Green Conservation Area in the City Planning Area, because the high land prices in the City Planning Area limit purchases by the municipalities. Table 5. The changing ratio (%) of landscape elements at Green Conservation Area in each period.

Landscape element

Period 1984 - 1989

1989 - 1995

1995 - 2001

85.1

78.6

77.5

a) From woodlands to Woodlands Grasslands Cultivated lands

2.7

3.8

5.7

11.1

7.2

7.3

Denuded land

0.2

0.3

1.6

Open water

0.2

9.8

2.8

0.7

0.3

5.2

Residential area b) From grasslands to Woodlands

45.7

64.2

45.3

Grasslands

12.9

9.4

17.0

Cultivated lands

20.0

7.5

17.0

Denuded land

1.4

1.9

0.0

Open water

0.0

17.0

1.9

20.0

0.0

18.9

Residential area

From 1995 to 2001, 5.2% of woodlands were transferred to residential areas within the Green Conservation Area. This decrease in woodlands was due to constructing a City Planning Road in the Green Conservation Area. Not surprisingly, this development was not illegal because the Urban Green Space Conservation Law permits developments with high public need, even if the sites are within the Green Conservation Area. We should note that, on occasions, the destiny of green space within a Green Conservation Area depends on decisions made by the municipalities. In this section, we summarized distribution patterns of green space and discussed the effects of the regulation system on quantitative conservation of green space. In the next section, we summarize changing patterns of green space in the case study region in the city during several decades, and then we discuss habitat function of woodlands in the region from a qualitative viewpoint.

300

T. MANABE ET AL.

4. CHANGING PATTERNS OF LANDSCAPE ELEMENTS IN THE CASE STUDY REGION 4.1 Case study region The region for this case study is at the mid-northern part of the city (Fig. 3) and contains areas that had revoked the Scenic Zone (Fig. 2). The region contains the centre of politics and economy of the city, and some parts of a Quasi-National Park with a wide area of secondary forests.

Figure 3. Study area (mid-northern region of Kitakyushu City).

Some secondary broad-leaved forests of various shapes and areas still exist at the centre of the region (Suzuki et al., 2004). Among them, the following four forests of different areas, all dominated by evergreen broad-leaved trees, were chosen as study sites for understanding vegetation structure and evaluating their habitat function for woody species. Yamada Green Park, one of the four study sites, was used as an ammunition storehouse by the Japanese Army from the beginning of the 1940s until the end of World War II; it was then managed by the US Army and the Self-Defence Forces until 1995, when it was used as a park. The area has been free from severe human disturbance for more than half a century (Ota, 1992). Most of the area, about 345 ha, has therefore been covered by broad-leaved forests, and has the largest and most continuous forest area of the four-forested areas.

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

301

The Central Park area is surrounded by mountains, such as Mt Konpira, and centres on Central Park, which was established out of ex-Tobata City in 1940. It contains some institutions such as a gymnasium, a welfare park and so on. Art-forest Park surrounds the municipal Art Museum established in 1974. Itouzu Park was reestablished in 2002 as a zoo that made use of secondary forests following the replacement of the recreation park with a zoo in 1998. 4.2 Change in green space The digital green space maps in 1922, 1936, 1948 and 1960 were made by using topographic maps published by the Geographical Survey Institute. In those maps, because the boundaries around each element within the maps were unclear, green spaces were classified into two categories: 1) woodlands, which contained the land covered by woodland (broad-leaved forests and needle-leaved forests), phyllostachys forests, shrubs, grasslands and orchards and 2) cultivated lands, which contained paddy fields and other fields. The digital green space maps in 1961, 1974 and 2000 were made by using aerial photos. In those maps, green spaces were classified into three categories: 1) woodlands, which contained broad-leaved forests, needle-leaved forests and phyllostachys forests, 2) grasslands and 3) cultivated fields, which contained paddy fields, other fields and orchards. Changing patterns of green spaces were analysed by overlapping these maps using GIS (Arc View 8.3). See Mitsuda et al. (2003) and Suzuki et al. (2004) for detailed descriptions of the study designs. Woodlands showed only a slight decrease in area between 1922 and 1960 (Table 6) for two reasons. First, woodland had important socio-economic values in this period, and little was transferred to other landscape elements such as residential areas. Next, some parts of the region were specified as Scenic Zones in 1936 (Fig. 2). Table 6. Area (ha) of green space in mid-northern region in each year. Values were estimated for using topographic maps (A) and aerial photos (B). Category A

Year 1922

1936

1948

1960

2217.3

2125.8

2125.8

2071.1

830.6

525.5

525.5

389.1

1961

1974

2000

Woodlands

1591.8

1183.6

1349.6

Grasslands

22.3

82.1

83.0

275.7

155.2

54.0

Woodlands Cultivated lands B

Cultivated lands

Woodlands mainly decreased between 1961 and 1974. This period included the age of the fuel revolution mentioned above, and was when the 1962 Japan’s

302

T. MANABE ET AL.

Comprehensive National Development Plan was pushed forward. During this period, therefore, the value of woodlands greatly decreased, and a wide area of woodland was cut all over Japan. On the other hand, woodland increased slightly between 1974 and 2000 largely because of planting in City Parks and the succession of grasslands to forests in mountain regions following the fuel revolution. The area of cultivated land decreased mainly between 1922 and 1936. It was located on relatively flat land and was easy to transfer to residential area during urbanization. There was no decrease in area between 1936 and 1948, which may have been the result of the need to grow food during World War II. The area of cultivated land decreased successively so that, by 2000, the area was about 20% of that in 1961. The decrease in area of cultivated land, especially after 1961, is largely attributed to the structural changes in industry and the socio-economic changes in Japanese society mentioned earlier. These changes led to a decrease in farming and an increase in cultural land transferred to other elements. Secondary forests with a relatively wide area, however, still existed (Suzuki et al., 2004), although the area of green space decreased in the region where land had been developed since the mid-1920s. This might be attributed to the regulation system, which specified some areas as Scenic Zones. Further, tall broad-leaved trees already dominated the forests in Itouzu Park and some parts of Yamada Green Park, although both were isolated by 1961 (Suzuki et al., 2004). On the other hand, broad-leaved forests covered mainly the mountains of the Central Park and Art-forest areas, although most areas in 1961 consisted of various types of patches such as clear-cut sites, grasslands, shrubs and broad-leaved tall forests. This change was due to the progress of secondary succession. The same tendency was reported in suburban areas (Manabe et al., 2003) and rural areas (Kamada and Nakagoshi, 1990, 1996; Hong et al., 1995), indicating that secondary succession changed the vegetation structure of abandoned forests at suburban and rural regions and urban areas. What type of ecological characteristics do these forests have? In the next section, we summarize the characteristic vegetation structure of these forests, and then we discuss the habitat function of the forests for woody species. 4.3 Characteristics of vegetation structure Eight 50 m × 20 m plots were set up: four in Yamada Green Park (Y2–Y4), two in Central Park (C1 and C2), and one each in Art-forest Park (AP) and Itouzu Park (IP). The plots were situated on slopes facing south to east; the mean slope ranged from 10.0° (AP) to 23.1° (Y3) and the maximum slope from 22° (C2) to 48° (Y3). The stems of all trees and shrubs taller than 2 m (mature stems) were identified to species and then DBH (diameter at breast height) was measured to the nearest 0.1 cm. They were also classified into two groups based on their vertical position: stems in the canopy layer (canopy stems) and stems under the canopy layer (understorey stems). In addition, each 50 m × 20 m plot was divided into 10 contiguous 10 m × 10 m quadrats, and one 5 m × 5 m subquadrat within each quadrat, except for C2. A 2 m × 2 m mesh was set within each subquadrat. Species were identified and height was measured for all saplings (height 0.3–2 m) in each subquadrat and for all seedlings (height < 0.3 m) in each mesh.

Table 7. Characteristics of the study plot.

Study sites and study plots Y2 large 130 E

Yamada Green Park Y3 Y4 large large 120 80 SSE ENE

Y5 large 50 ESE

Centoral Park C1 C2 midium midium 90 110 SE SE

Arts Park AP midium 70 SE

Itouzu Park IP small 30 S

Patch size Above sea level (m) Slope direction Slope degree (°) mean 18.8 23.1 20.0 21.1 18.7 14.3 10.0 16.1 maximun 48 40 39 45 45 22 34 40 Vegetation in 19611) TBF TBF TBF TBF SBF SBF SFG TBF Nunmber of species 22 21 20 20 19 14 20 20 2 -1 Total basal area (m ・ha ) 55.7 59.2 49.9 61.0 46.5 39.6 50.6 53.5 -1 Density (ha ) 3,640 3,100 2,890 3,070 3,210 2,380 2,350 2,140 1) TBF, tall broad-leaved forests; SBF, small broad-leaved forests; SFG, small broad-leaved forests mixed with grasslands 2) L, evergreen Lauradsea species; C, Castanopsis cuspidata *) BA, cumulative basal area of canopy stems (m2・ha-2); DEN, density of understory stems (ha-2); Dmax, maximum DBH (cm)

304

T. MANABE ET AL.

In this study, Castanopsis cuspidata includes C. cuspidata var. sieboldii, because the variety could not be determined because of the presence of hybrid and intermediate forms. The plots contained a total of 43 woody plant species, of which 32 were evergreen broad-leaved species and 11 were deciduous broad-leaved species. Evergreen broad-leaved species occupied 75.5%–99.8% in the basal area (BA) and 82.6%–97.7% in stem number. Among them, two evergreen canopy species, Persea thunbergii and Ilex integra, two evergreen understorey species, Eurya japonica and Ligustrum japonicum, and one deciduous understorey species, Ficus erecta, occurred in all plots. Four evergreen subcanopy species, such as Neolitsea sericea and Ilex rotunda, occurred in seven plots, and two canopy species, Castanopsis cuspidata (evergreen) and Rhus succedanea (deciduous), and one understorey species, Aucuba japonica (evergreen), occurred in six plots. Castanopsis cuspidata had the largest cumulative BA in four plots. Evergreen species of the family Lauraceae had the largest cumulative BA in other plots and, among them; the largest was Persea thunbergii in three plots and Neolitsea sericea in one plot. Some community parameters were similar among plots (Table 7). The number of species and the total BA of a plot ranged from 19 (C1) to 22 (Y2) and from 49.9 m2 ha–1 (Y4) to 61.0 m2 ha–1 (Y5) except for C2 (14 and 39.6 m2 ha–1). The density of stems, however, varied between plots and ranged from 2140 ha–1 (IP) to 3640 ha–1 (Y2). Sorensen’s index, which is based on the presence or absence of species, indicated that Y4 and B1 were the most similar pair of plots and IP and AP were the most dissimilar pair for canopy stems (Table 8). C1 and C2 had relatively high similarity to other plots. In the understorey, C1 and C2, were the most similar, and Y5 and IP the most dissimilar pairs. Y2 and C2 had relatively high similarity to other plots. The similarity among plots was slightly larger for understorey stems than for canopy stems; the indices for canopy stems and understorey stems ranged from 0.31 to 0.85 and 0.41 to 0.81, respectively. In old-growth evergreen broad-leaved forests in southwestern Japan, the dominant species are Castanopsis cuspidata, Quercus salicina and Distylium racemosum for the canopy layer, and Cleyera japonica and Camellia japonica for the sub-canopy layer (Tanouchi and Yamamoto, 1995; Manabe et al., 2000). The urban evergreen broad-leaved forests studied showed some peculiar structural characteristics compared with those of old-growth forests; deciduous broad-leaved species were more abundant and some dominant sub-canopy species in old-growth forests were less dominant. A decrease in species diversity in small, isolated forests was reported for suburban and urban forests in Japan, e.g. the Kanto region (Iida and Nakashizuka, 1995) and Kyoto City (Murakami et al., 2005; Imanishi et al., 2005). The effect of a decrease in forest area on species diversity varied among the species studied (Murakami et al., 2005; Imanishi et al., 2005). The number of species per plot was similar among the plots studied except for one plot (C2), although the area and vegetation types in 1961 at these sites differed greatly, as mentioned above. This does not mean that isolation of forests affects their species diversity in this urban area; rather, it means that the number of species growing in a given area is similar throughout the region. We did not calculate the maximum number of species that

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

305

could occur at a study site by using species-area curves as an estimate (Iida and Nakashizuka, 1995). Table 8. Similarity among plots for species composition at the canopy layer (A) and the understorey layer (B) by using Solensen's similarity index. Values > larger than the mean values of all pairs (a=0.55, b=0.67) are given by bold. Y3 *

Y4 *

C1 *

IP *

Y2

Y5

C2

A. Canopy layer Y4

0.31

C1

0.63

0.71

I2

0.60

0.36

0.43

Y2

0.35

0.44

0.57

Y5

0.67

0.46

0.75

0.40 0.40

0.59

C2

0.67

0.62

0.75

0.60

0.59

0.83

B1

0.40

0.88

0.74

0.31

0.40

0.53

0.53

B. Understory layer Y4 *

0.74

C1 *

0.54

0.59

IP

*

0.58

0.58

0.65

Y2

0.68

0.68

0.70

0.54

Y5

0.56

0.56

0.58

0.41

0.67

C2

0.62

0.67

0.81

0.67

0.67

0.59

AP

0.61

0.61

0.74

0.67

0.56

0.49

0.77

*) Y3, Y4, C1 and I are the plots where Castanopsis cuspidata is the most dominant for cumulative basal area.

Table 9. The number of species at each growth stage at each plot. Plot

Growth stage Mature

Sapling

Seedling

Y2

22

17

20

Y3

21

12

27

Y4

20

14

21

Y5

20

19

24

C2

19

11

22

AP

20

18

23

IP

20

28

9

306

T. MANABE ET AL.

Our results show that some important components of old-growth evergreen broad-leaved forests, such as Camellia japonica and Cleyera japonica, are hardly able to maintain their populations in the study area. Components that are less tolerant of human disturbances might find it difficult to grow in secondary forests. The lack of these species at the study sites might not be due to the isolation of the forests but rather to the past human disturbance regime in the area. Thus, the effect of isolation on the occurrence of tree species larger than 2 m in height was not clear in the forests studied. The number of species at each growth stage in a plot differed greatly between plots as the growth stages became younger (Table 9). This finding suggests that the differences between plots in some environmental conditions became larger more recently, and that younger individuals recruited and established more recently. The reasons for these differences between plots may be the degree of isolation, such as the area of each forest or the distance between forests, and qualitative and/or quantitative changes in human disturbance. Saplings and seedlings of typical pioneer species such as Zanthoxylum ailanthoids and Mallotus japonicus grew vigorously in a particular plot. These species usually regenerate at disturbed sites, such as in canopy gaps in old-growth forests and in clear-cut sites, which suggests that some micro-environmental conditions became similar to those of disturbed sites in the forests. This was especially evident in the plot in the Itouze Park, which had the smallest area and a high seedling density of Z. ailanthoids with the lowest seedling diversity. This suggests that the micro-environmental conditions, such as light, humidity and temperature, of the forest floor were changed largely by an edge effect (Murcia, 1995) in the plot, as seen in the isolated shrine/temple forests of Hyogo Prefecture (Ishii et al., 2004). 4.4 Habitat function for major woody species In this section, we discuss the habitat function of the forests mentioned above for major, canopy and sub-canopy species, i.e. Castanopsis cuspidata, Persea thunbergii, Neolitsea sericea and Cinnamomum japonicum. In this study, the habitat function for woody species is evaluated by examining the relationships between stem densities at each growth stage. It is based on the idea that, if seedling and mature tree densities are high and low, respectively, the site has recent low habitat function, i.e. a low density of seedlings would indicate unsuitable conditions for establishing the species, even if there is a high density of mature individuals that could supply seeds. The relationships (Table 10) were classified into three types: 1) the densities were high for every growth stage (Castanopsis cuspidata at IP, C1 and Y3, Persea thunbergii at Y2 and C1 and Neolitsea sericea at C1), 2) high densities for matures but low densities for saplings and seedlings (P. thunbergii at C2) and 3) low densities (or no plants) for matures but high densities for saplings and seedlings (P. thunbergii at C1, N. sericea at Y2 and Cinnamomum japonicum at IP).

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

307

Table 10. Density (0.1ha-1) of stems of main tree species at each growth stage at each plot. Species

Growth Stage

Plot Y2

Y3

Y4

Y5

Mature

0

Saplings

8

Seedlings

C1

AP

IP

28

78

0

32

196

0

39

27

80

368

24

76

50

1,500

4,550

0

4,950

Mature

52

38

11

80

5

9

8

Saplings

56

24

0

84

8

8

28

Seedlings

775

700

50

750

450

925

375

36

5

17

46

36

6

0

Castanopsis cuspidata

150 38,925

Peresea thunbergii

Neolitsea sericea Mature Saplings

292

92

212

460

120

28

4

1,050

450

625

2,100

5,350

700

50

Mature

29

0

9

2

15

13

5

Saplings

228

4

40

56

840

168

96

Seedlings

325

0

25

125

1,200

225

625

Seedlings Cinnamomum japonicum

There was a tendency for the seedling density of Castanopsis cuspidate to depend largely on the existence of conspecific mature trees, although there were no clear relationships between densities at each growth stage for Persea thunbergii, Neolitsea sericea, and Cinnamomum japonicum of the family Lauraceae. This difference is partly ascribed to the patterns of seed dispersal. It is difficult for C. cuspidata to disperse its seeds long distances because its seed dispersal type is classified as chasing dispersal. Success of seedling recruitment of this species is, therefore, affected by the existence of conspecific trees that can become a seed source. Thus, the species hardly maintains its population in forests that have a low density of recent, conspecific mature trees as Y2, Y5 and AP, which means that these forests have low habitat function for the species even if safe-sites for seedling recruitment and establishment exist. Habitat function for the species, however, might be improved if these forests could be connected with other forests with high densities of conspecific mature trees. The seed dispersal type of three species of Lauraceae is classified as endozoochory and most seeds are dispersed by birds (Kominami et al., 2003), which suggests that the ‘dispersal limitation’ as seen in Castanopsis cuspidata does not always operate success of seedling recruitment strongly. Thus, the habitat function of the forests might be controlled by their micro-environmental conditions.

308

T. MANABE ET AL.

5. SUMMARY Green spaces in Kitakyushu City still exist, although they have largely decreased since the development of the ‘town of iron manufacture’ in 1901. This is attributable to the effects of regulation systems that, to a large extent, restricted land use. Green space conservation might be effective in areas specified as Scenic Zones and Green Conservation Areas within Urbanization Control Areas. Further, Green Conservation Areas have the ability to conserve green space, even if they are within an Urbanization Promotion Area. Further, some forests that were located almost in the centre of the city might have a favourable habitat function for certain tree species. These findings suggest that habitat function as well as species diversity in the city could be improved by connecting isolated small forests having low ecological function with large forests having high ecological function. On the other hand, attention must be paid to the fact that the destiny of green space depends on the decisions made by the municipalities. Hence, scientists, citizens and administrators must understand that green space, even in a large city, can have various useful functions for wildlife and residents. The city must create and practise effective management systems based on the actual regulation systems. ACKNOWLEDGEMENTS This study was partly supported by a Grant in Aid of Science Research (90359472) from the Japan Society for the Promotion of Science (Head of the project: T. Manabe) and projects for contributions to the region from the Ministry of Education, Culture, Sports, Science and Technology and the Kyushu Institute of Technology (head of the project: K. Ito). We thank the members of the Yamada Green Park and the Itouzu Park for permitting this study. We also thank J. Kaku, H. Kashima,T. Suzuki and D. Hashimoto for their assistance. REFERENCES Forman, R.T.T. (1995). Land Mosaics- The Ecology of Region and Landscape. Cambridge: Cambridge Univ. Press. Fukamachi, K., Iida, S. and Nakashizuka, T. (1996). Landscape patterns and plant species diversity of forest reserves in the Kanto region, Japan. Vegetatio, 124, 107–114. Hong, S-K., Nakagoshi, N. and Kamada, M. (1995). Human impacts on pine-dominated vegetation in rural landscapes in Korea and western Japan. Vegetatio, 116, 161–172. Iida, S. and Nakashizuka, T. (1995). Forest fragmentation and its effect on species diversity in sub-urban coppice forests in Japan. Forest Ecology and Management, 73, 197–210. Imanishi, A., Imanishi, J., Murakami, K., Morimoto, Y. and Satomura, A. (2005). Herbaceous plant species richness and species distribution pattern at the precincts of shrines as non-forest greenery in Kyoto City. J. Jpn. Soc. Reveget. Tech., 31, 278–283. (In Japanese with English abstract.) Ishii, H.T., Iwasaki, A. and Sato, S. (2004). Seasonal variation of edge effects on the vegetation, light environment and microclimate of primary, secondary and artificial forest fragments in southeastern Hyogo Prefecture. Proceedings of the IUFRO International Workshop on Landscape Ecology 2004, Tsukuba, Japan. Kaku, J., Ito, K., Isono, D., Mitsuda, Y. and Umeno, T. (2004). Basic study about the evaluation of urban green and the ecological networks—A case study on Murasaki River basin in Kitakyushu. Kyushu J. Forest Research, 57, 163–166. (In Japanese)

STRUCTURE AND DYNAMICS OF URBAN GREEN SPACE

309

Kamada, M. and Nakagoshi, N. (1990). Patterns and processes of secondary vegetation at a farm village in southwestern Japan. Jpn. J. Ecol., 40, 137–150. (In Japanese with English synopsis.) Kamada, M. and Nakagoshi, N. (1996). Landscape structure and the disturbance regime at three rural regions in Hiroshima Prefecture, Japan. Landscape Ecology, 11(1), 15–25. Kamada, M., Nakagoshi, N. and Nehira, K. (1991). Pine forest ecology and landscape management: a comparative study in Japan and Korea. In N. Nakagoshi and F.B. Golley (Eds.), Coniferous Forest Ecology from an International Perspective (pp. 43–62). The Hague: SPB Academic. Kominami, Y., Sato, T., Takeshita, K., Manabe, T., Endo, A. and Noma, N. (2003). Classification of birddispersed plants by fruiting phenology, fruit size, and growth form in a primary lucidophyllous forest: an analysis, with implications for the conservation of fruit–bird interactions. Ornithological Science, 2, 3–23 Manabe, T., Kashima, H. and Ito, K. (2003). Stand structure of a fragmented evergreen broad-leaved forest at a shrine and changes of landscape structure surrounding a suburban forest, in northern Kyushu. J. Jpn. Rev. Tec., 28, 438–447. Manabe, T., Nishimura, N., Miura, M. and Yamamoto, S. (2000). Population structure and spatial patterns for trees in a temperate old-growth evergreen broad-leaved forest in Japan. Plant Ecology, 151, 181–197. Mitsuda, Y., Manabe, T., Ito, K., Kashima, H. and Suzkuki, T. (2003). The methods of making the digital vegetation maps by using digital orthophotographs—In the case of the Yamada Green Park in Kitakyushu City. Bull. Kitakyushu Mus. Nat. Hist. Hum. Hist., Ser. A, 1, 57–65. (In Japanese with English abstract.) Murakami, K., Maenaka, H. and Morimoto, Y. (2005). Factors influencing species diversity of ferns and fern allies in fragmented forest patches in the Kyoto City Area. Landscape and Urban Planning, 70, 221–229. Murcia, C. (1995). Edge effects in fragmented forests: implications for conservation. Trends in Ecology and Evolution, 10(2), 58–62. Ota, M. (Ed.). (1992). Nature of Yamada Park, Kitakyushu City, Japan. Kitakyushu, Kitakyushu Museum of Natural History. Senior, K. (2005). Resource efforts for the Satoyama. Frontiers in Ecology and the Environment, 3(2), 68. Suzuki, T., Manabe, T., Ito, K. and Umeno, G. (2004). Analysis of landscape changes by using digital vegetation map in mid-northern region in Kitakyushu city. Bull. Kitakyushu Mus. Nat. Hist. Hum. Hist., Ser. A, 2, 79–85. (In Japanese with English abstract.) Tanouchi, H. and Yamamoto, S. (1995). Structure and regeneration of canopy species in an old-growth evergreen broad-leaved forest in Aya district, southwestern Japan. Vegetatio, 117, 51–60.

CHAPTER 19

SEEDING ON SLOPES IN JAPAN FOR NATURE RESTORATION

H. YOSHIDA Graduate School of Global Environmental Studies, Kyoto University, Kyoto, Japan; Toko Corporation, Tokyo, Japan

Abstract. In the past few decades, improved seeding technologies have been developed in Japan. This paper discusses the restoration of degraded slopes by the application of seeding, which is known as “restoration of natural vegetation on slopes“ in Japan. First, the technical aspects of revegetation work are described; second, historical changes in approaches to seeding are discussed; third, typical case studies of restoration of engineered slopes (road cuts) using the “thick-growth-media spraying method,” a popular approach in Japan, are presented. The article concludes with a discussion of the basic principles of restoration of natural slopes based on previous Japanese studies and the author’s experience.

1. INTRODUCTION Afforestation work done to prevent flooding and erosion of sandy soil has been applied in Japan. This work has been necessary because much of the country is situated on steep slopes and in a region with abundant rainfall. However, afforestation alone cannot restore degraded vegetation on engineered slopes (road cuts), steep slopes, rocky ground with shallow soil, strongly acidic soils, or sandy soils. Therefore, Japanese researchers have worked toward improved revegetation techniques that focus on restoration, correction, and management of vegetation. In Japan, revegetation technology that combines work on the underlying foundation as well as on the vegetation has been developed to promote successful restoration of the abovementioned difficult sites. The term “revegetation work” was first used by Kurata in 1953, who defined this as “planning, constructing, and management activities related to the restoration, creation, and protection of nature.” This

311 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 311–328. © 2007 Springer.

312

H. YOSHIDA

integrated science has become known as revegetation engineering technology (Japan Society of Revegetation Technology, 1990). Revegetation work aims to restore various ecosystem functions that were lost when the original plant communities were destroyed. This technology has been based on the principle of creating a suitable environment for plant growth. Figure 1 shows an overview of the Japanese system of revegetation work, which is composed of three key stages: the creation of a foundation before introducing the new plants, seeding or planting of the species that will be used for revegetation, and monitoring and tending the introduced species. Recently, seeding has been studied as an alternative method to restore degraded slopes. This article describes revegetation based on slope seeding, a popular technique for slope protection. The contents of this paper were first presented at the East Asian Federation of Ecological Societies (EAFES) International Congress (Yoshida and Morimoto, 2004), and the details of historical changes were published in the Journal of the Japanese Society of Revegetation Technology (Yoshida, 2005).

Figure 1. A technical overview of the Japanese system for revegetation work.

2. HISTORICAL CHANGES IN SEEDING TECHNOLOGY Seeding technology has an 80-year history in Japan. However, the goal has changed little since the first use of this method. Seeding began with afforestation of slopes, particularly for erosion control. Recently, the main purpose of revegetation work has changed. Rather than revegetating solely for erosion control, this technique has been applied as a method for restoration of natural environments. Based on previous studies, I have classified the history of seeding work into five periods.

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

313

2.1 Birth of seeding work The period between 1927 and 1948 began with the first seeding trial, done on the Korean peninsula, and extended to the development of a method called “mixed seeding on slopes” (MSS), and I have named this “period of early development of seeding” (the first period). The idea of “seeding work” at this time derived from afforestation. The initial technique was originally applied on the Korean peninsula in 1927 using Alnus firma, and shortly afterward, using Robinia pseudoacacia and Alnus hirsuta var. sibirica (Kurata, 1979). At the time, seeding work mostly took the form of linear revegetation techniques, with the seeds planted directly in shallow trenches. The currently used MSS revegetation method was developed in 1939. This method emerged as an economical and rapid way to artificially recreate a forest (Sato and Ono, 1942). The goal was to restore bare mountainsides with herbaceous and woody plants. Trial and error revealed that on sites where various types of soil-improving trees such as R. pseudoacacia, Lespedeza bicolor var. japonica, Amorpha fruticosa, A. hirsuta var. sibirica, and Acacia dealbata had not been planted, growth of Quercus acutissima, Quercus variabilis, and Aleurites cordata decreased. These results clearly defined the necessity of using soil-improving trees for revegetation work. However, a simple method of covering bare mountainsides with seeded straw mats that was tested did not spread widely because rainfall and frost heaving caused much erosion during the winter (Kurata, 1979). Kurata (1959, 1979) described the MSS method as a starting point for the development of modern revegetation technology. This method incorporated several improvements compared with previous revegetation strategies at afforestation sites: the formerly linear revegetation technique was changed into a planar approach that covered more of the site, mixed seeding provided both herbaceous (fast-growing groundcover) and woody vegetation (long-term stabilization), the use of several species rather than monocultures for revegetation became more common, and seeding became more widely used. 2.2 Spread of rapid revegetation techniques After importing exotic grasses such as Festuca elatior var. arundinacea (Kentucky 31 fescue) and Eragrostis curvula (weeping love grass) from the United States, Kurata proposed a “rapid revegetation technique” that used the newly imported exotic species, organic matter, and soil-improving trees and grasses during the early stages of revegetation. This work was first done in 1951, and Kurata officially used the term “revegetation work” in 1953 (Nitta, 1995). Seeding using hydroseeders attracted considerable attention as an efficient erosion-prevention method on the bare slopes created by large-scale exploitation forests (Nitta, 1959; Nitta and Kobashi, 1961). Furthermore, the newly imported exotic species adapted well to the damaged landscape in terms of their ability to grow easily and rapidly. Since this method proved to be reliable, it spread widely throughout Japan. A similar, technique developed in the 1970s called the “organic

314

H. YOSHIDA

thick-growth-media spraying method” (TGM) allowed revegetation on rocky slopes that lacked soil (Kikuchi, 1980; Kurata, 1979; Oda, 1976); this approach applies a 3to 10-cm-thick base on which plants can grow. TGM is currently a popular seeding technique in Japan, as it provides high retention of water and nutrients and generally adapts well to steep slopes. This method uses a mix of a base material such as bark compost or peat moss, seeds, chemical fertilizer, and an erosion-control substance such as cement or vinyl acetate resin. This material is sprayed on the slope from a distance of 1 m using a spraying machine capable of handling the thick base. Rapid revegetation using exotic grasses prevented erosion on many slopes and collapsed hillsides. This has contributed greatly toward preventing destruction of roadside environments during development and toward restoring the environment. Unfortunately, the popularity of revegetation using exotic grasses reduced attention on revegetation with woody plants. It is not surprising that many identified the concept of slope seeding as a restoration method for degraded environments that only used exotic species. Because exotic grasses were mainly used for seeding work between 1949 and 1958, I have named this “the early modern period of revegetation work” (the second period). Between 1959 and 1985, seeding mainly used a spraying machine, thus I have named this period “the period of rapid revegetation using exotic grasses” (the third period). 2.3 Spread of rapid reforestation techniques The landscape produced by the two rapid revegetation techniques could not prevent slope erosion and coincided with a period of growing recognition of the importance of landscape esthetics. It became ecologically important to achieve more lasting restoration results that would function similarly to natural forests (Yamadera, 1986, 1989). Thus, Yamadera described the significance of revegetation work as “lending a hand to allow nature to restore itself naturally” and recommended three characteristics of this approach to achieve the abovementioned goals: restoring ecologically suitable vegetation in the degraded landscape, valuing respecting and taking advantage of natural successional pathways, and restoring the environment’s ecological functions as naturally as possible (Yamadera, 1995; Yamadera et al., 1993). He further recommended certain woody species for seeding on slopes: Betula ermanii, Betula platyphylla var. japonica, Clerodendron trichotomum, Mallotus japonicus, Rhus javanica, Quercus mongolica var. grosseserrata, Ligustrum japonicum, Rhaphiolepis umbellata, Pittosporum tobira, Camellia sasanqua, Camellia japonica, Cyclobalanopsis spp., and Sorbus commixta (Yamadera, 1986). The guidelines published by the Ministry of the Environment’s Conservation of Nature Bureau (1982) systematically arranged the woody species into groups for the purpose of revegetation by seeding. After the use of mixed seeding with leguminous shrubs and birch (Betula) species were introduced in the “Standard of road earthworks, Slope protection works” in 1986 (Japan Road Association, 1986), leguminous shrubs began to be widely used.

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

315

To successfully blend the seeds of exotic grasses and leguminous shrub species, researchers had to decrease the amount of exotic grass seeds. Whereas the original method (grasses only) was expected to grow 5000 to 10 000 plants per 1 m2, the new method was expected to grow only 1–10% of this total. Since it was extremely difficult to form a mixed grass and woody plants using the rapid revegetation technique, many studies had to be done to provide some ratios of the numbers of seeds of grasses and other species that have been found to be successful. Following the introduction of revegetation using leguminous shrubs, many guidebooks were published that presented specific design techniques; Road Management Technology Centre (1996), Japan Highway Landscape Association (1986), Ministry of Construction River Bureau Sand Control Department (1996), and Japan Association of Agricultural Engineering Enterprises (1990). As shown in Table 1, the main purpose of these research and development efforts focused on the formation of communities of leguminous shrubs on engineered slopes. Because using mainly soil-improving trees enhanced the development of communities of woody plants, I have named the period after 1986 “the spread of rapid reforestation using mainly leguminous species” (the fourth period). Table 1. Examples of seeding work during the fourth period. Study Egashira (1988) Nakaide et al. (1990) Seino et al. (1995)

Introduced plant species or content Lespedeza bicolor var japonica community on the slopes of an expressway L. bicolor var japonica , A. fruticosa , A. hirusuta var sibirica , and A. firma in an industrial park B. platyphylla var japonica , A. hirusuta var sibirica , A. firma , and L. bicolor var japonica in a cold zone with significant snowfall Oyama et al. (1996) A. fruticosa in the seaside sandy land beside the sea Takeuchi and Nishizawa (1998) Use of shrubs in family Leguminosae in a nature park Tsuchimuro et al. (1998) Use of shrubs in the Leguminosae on the slope of a woodland path

Table 2. Trial examples of seeding work on the nature restoration. Study Yoshida (1990) Yoshida (1991) Yoshida and Hosaka (1992) Yoshida et al. (1995) Yoshida (1998)

Introduced plant species or content Trial introduction of Ligustrum lucidum ; results were sufficiently promising to perform additional trials (see below) Trial introduction of L. lucidum and Rhaphiolepis umbellata on a soft rocky slope Mixed seeding of Betula spp. and L. lucidum Trial introduction of Acer palmatum var matsumurae and Sorbus commixta Proof of successful establishment of a multi-stratum community of woody plants by applying the TGM method to cut slopes; L. lucidum , R. umbellata , Melia azedarach var. japonica , Rhus seccedanea , Rhus javanica , Indigofera pseudo-tinctoria , Alnus hirusuta var. sibirica , Alnus firma , Betula platyphylla var. japonica , Betula ermanii , S. commixta , Quercus mongolica var. grosseserrata , and Fagus crenata

2.4 Development of nature restoration techniques Yoshida and Yamadera (1989) introduced a method that applied TGM to vegetation in the mid- to late stages of ecological succession and in the climax stage. This approach built on older mixed-species approaches by specifically incorporating a mixture of seeds that included pioneer species (to provide rapid early coverage of

316

H. YOSHIDA

the slope), mid-successional species (to provide cover for late-successional species), and climax species (to provide permanent stabilization and restore nature). Table 2 summarizes five studies of this method during the 1990s. This approach supports the production of a multiple-stratum community of woody plants by applying the TGM method on engineered slopes. As seen in Table 3, many studies have been done since 1993 to revegetate engineered slopes using communities of woody plants. Slope seeding using mainly the endemic species listed in Table 3 has been recommended in various guidelines (Ehime Prefecture Engineering Works Part, 1999; Japan Road Association, 1999; Shikoku Construction Bureau of Japan Highway Public Corp, 1998; Shikoku Construction Bureau of Ministry of Land Infrastructure and Transport, 2002). After 1996, the species chosen for revegetation have changed from soilimproving trees to endemic species such as evergreens. Accordingly, I have named this period “the period of development of nature restoration techniques using endemic woody species” (fifth period). Table 3. Examples of seeding work in the fifth period. Study Yamamoto et al. (1993) Hasegawa and Nishizawa (1998) Saito et al. (1999) Shibata (2001) Akita et al. (2000) Ishida et al. (2000) Akita et al. (2001) Ishida et al. (2001) Inoue and Yoshida (2002)

Kida et al. (2002) Sasaki et al. (2003) Furuta et al. (2004)

Furuta and Yoshida (2004) Yamanishi et al. (2004) Yoshida and Furuta (2004)

Yoshida and Morimoto (2005)

Introduced plant species or content Quercus glauca and Quercus phillyraeoides Acer palmatum var matsumurae and Sorbus commixta in a nature park Quercus acutissima and Quercus serrata Q. phillyraeoides, Ligustrum japonicum, Rhaphiolepis umbellata , and Pittosporum tobira L. japonicum, R. umbellata, and Camellia japonica in a sedimentation area characterized by lapilli materials C. japonica , Camellia sasanqua , L. japonicum , and R. umbellata R. umbellata , Q. phillyraeoides , Clerodendron trichotomum , Buddleja venenifera , Melia azedarach var japonica , Eurya emarginata , and Ardisia sieboldii in the Yakushima World A. palmatum var matsumurae , S. commixta , and Rhus seccedanea at a gabion-stabilized slope Formation of a mixed evergreen and deciduous broad-leaved tree community (See section 3.3); L. japonicum , R. umbellata , Rhus seccedanea , Acer palmatum , Elaeagnus umbellata , Zizyphus jujuba var. inermis , Hibiscus syriacus , and Hibiscus mutabilis Myrica rubra , Prunus lannesiana var speciosa , A. palmatum var matsumurae , L. lucidum , and R. umbellata Q. glauca , R. seccedanea , Euptelea polyandra , and Q. phillyraeoides Revegetation examples using a double-layer spraying system; C. japonica , L. japonicum , Mallotus japonicus , Rhus javanica , Elaeagnus umbellata , Lespedeza bicolor var. japonica and Indigofera pseudo-tinctoria Formation of a Q. serrata–Quercus mongolica var. grosseserrata community Formation of a Pinus thunbergii forest in a seaside area Endemic woody species introduced several years after work at 10 sites; Acer japonicum , A. palmatum , A. palmatum var. matsumurae , A. palmatum var. amoenum , Alnus hirusuta var. sibirica , Alnus firma , C. japonica , C. sasanqua , E. umbellata , Kerria japonica , Melia azedarach var. japonica , Prunus jamasakura , Myrica rubra , Q. glauca , Quercus myrsinaefolia , Quercus variabilis , R. javanica , R. seccedanea , Thea sinensis , and Viburnum dilatatum Effects of mixed seeding of Chinese-grown Indigofera spp. and evergreen broadleaved trees; C. japonica , C. sasanqua , R. umbellata , and Ligustrum obtusifolium

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

317

3. FORMATION OF A COMMUNITY OF WOODY PLANTS ON ENGINEERED SLOPES Three examples of nature restoration methods based on seeding using the organic TGM method on engineered slopes are discussed in this section. In these examples, no exotic woody or grass species were used. 3.1 Formation of an evergreen tree community (Case Study A) In this example, researchers seeded evergreen trees on a rocky-engineered slope in Mie Prefecture. They sprayed a 6-cm-thick organic base to serve as a base for plant cultivation (Photo 1). Germination of all plants was confirmed within 1 month. Indigofera pseudotinctoria had become the dominant species within 5 months and had attained a coverage class (C) of 3 based on the Braun-Blanquet method, an average height growth (H) of 0.3 m, and a density (D) of 35 plants/m2. Hereafter, I will use these abbreviations and omit the units. After 2 years and 5 months, I. pseudotinctoria had reached C5, H1.3, and D5.0. Melia azedarach var. japonica (C5, H3.1, D1.5) became the dominant species in the upper layer and Ligustrum lucidum (C5, H1.7, D9.0) was dominant in the lower layer. After 3 years and 5 months, I. pseudotinctoria had declined to C2, H1.8, and D3.0. However, evergreen M. azedarach var. japonica (C3, H3.4, D1.3) became the overstory species, and R. umbellata (C4, H1.3, D11.8) and L. lucidum (C4, H2.3, D4.5) became the dominant understory species after 9 years and 7 months. Currently, after 15 years and 1 month, M. azedarach var. japonica (C2, H4.7, D0.1) dominates the overstory, L. lucidum (C3, H3.1, D5.2) and Rhus seccedanea (C2, H3.1, D0.1) dominate the middle of the canopy, and R. umbellata (C4, H2.4, D12.8) had become the dominant understory species. The initial dominant species, I. pseudotinctoria, declined completely over time (Yoshida and Furuta, 2004). 3.2 Formation of a deciduous broadleaved tree community (Case Study B) This example illustrates the restoration of a deciduous broadleaved tree community in a high-altitude (1220 m) cold area on a engineered slope. This study was done in Iwate Prefecture with a 3-cm-thick organic base sprayed to support plant cultivation (Photo 2). Lespedeza bicolor var. japonica had become the dominant species (C2, H0.1, D79.5) after 10 months. Germination of B. ermanii and S. commixta were confirmed after 1 year and 10 months; their seeds have a long dormancy period. With woody Betulaus species growing well, L. bicolor var. japonica became dominant (C5, H1.5, D3.9). In addition, A. hirsuta var. sibirica (C3, H1.5, D0.5), B. platyphylla var. japonica (C3, H1.0, D2.4), and B. ermanii (C3, H0.6, D1.4) became dominant after 4 years. The growth of these species was generally very slow, perhaps due to the fact that snow remained at the study site until June. However, by 8 years and 11 months, B. platyphylla var. japonica (C4, H2.2, D6.6) dominated the overstory and B. ermanii (C4, H1.8, D6.9) and S. commixta (C2, H0.8, D1.0) began dominating the understory. After 11 years and 11 months, B. platyphylla var. japonica (C4, H3.0,

318

H. YOSHIDA

D5.3) and B. ermanii (C3, H2.5, D5.3) dominated the overstory, followed by A. hirsuta var. sibirica (C2, H4.0, D0.8) and S. commixta (C2, H2.5, D1.0), Q. mongolica var. grosseserrata (C1, H1.5, D0.2), Fagus crenata (C1, H1.5, D0.8), and L. bicolor var. japonica (C1, H1.5, D0.2). After 13 years and 8 months, pressure from downhill creep of the snowpack had distorted the shape of the stem where it joined the roots. However, B. ermanii (C4, H3.3, D2.0) and S. commixta (C4, H3.2, D1.5) had become the overstory species, followed by B. platyphylla var. japonica (C3, H3.4, D2.0), A. hirsuta var. sibirica (C1, H3.1, D0.2), Q. mongolica var. grosseserrata (C1, H0.8, D0.2), F. crenata (C1, H2.2, D0.8), and L. bicolor var. japonica (C1, H0.7, D0.2) (Yoshida and Furuta, 2004). 3.3 Formation of a mixed evergreen and deciduous broadleaved tree community (Case Study C) In this example, researchers revegetated road engineered slopes in industrial park in Nagasaki Prefecture using a 3-cm-thick organic base layer. A variety of revegetation species were used to create a park-like structure. These included trees chosen for their autumn coloration, as well as flowering and fruiting trees (Photo 3). Germination of all plants was confirmed after 7 months. Lespedeza bicolor var. japonica (C4, H1.0, D9.0) and I. pseudotinctoria (C3, H1.5, D1.1) dominated the site, with Zoysia japonica also abundant (C5, H0.1, not counted D) by 1 year and 4 months. Four months later, Z. japonica had declined as a result of competition from woody species. After 3 years and 1 month, Elaeagnus umbellata (C3 to 4, H2.1, D8.0), L. bicolor var. japonica (C3 to 4, H1.6, D1.9), I. pseudotinctoria (C3 to 4, H2.3, D1.1), and Acer palmatum (C3 to 4, H0.7, D2.0) had become the dominant overstory species, and R. umbellata (C3 to 4, H0.5, D18.4) had become the dominant understory species. After 6 years and 2 months, R. seccedanea (C2 to 4, H5.0, D0.6) had become the dominant overstory species, A. palmatum (C2 to 4, H2.0, D1.1) and I. pseudotinctoria (C2 to 3, H2.9, D0.6) were the dominant species below the canopy, and E. umbellata (C3 or less, H3.2, D0.1), Hibiscus mutabilis (C1 or less, H1.8, D0.3), R. umbellata (C4 to 5, H1.5, D11.7), and L. japonicum (C1 to 2, H1.5, D0.6) had become the dominant understory species.

Photo 1. An example of an evergreen tree community established using the TGM method (Case Study A). Left: before, Right: After 15 years.

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

319

Photo 2. An example of a deciduous broadleaved tree community established using the TGM method (Case Study B). Left: before, Right: After 12 years.

Photo 3. An example of formation of an evergreen and deciduous broadleaved tree community established using the TGM method (Case Study C). Left: After 4 months, Right: After 6 years.

As a result of this restoration work, a beautiful natural landscape could be observed from the spring green foliage to the colors of autumn leaves. In addition, several bird nests were confirmed in the plant community. The newly established plant community became a habitat for some wild animals (Inoue and Yoshida, 2002). 4. BASIC PRINCIPLES OF REVEGETATING SLOPES FOR NATURE RESTORATION The main objective of slope revegetation in Japan has changed rapidly in recent years. In the past, revegetation was used primarily to prevent erosion. However, the goal has now become to restore nature. As demonstrated by the case studies in the previous section, seeding allows the creation of endemic communities of woody plants. In particular, TGM makes this possible by allowing rapid restoration of devegetated surfaces that would otherwise be difficult to revegetate. In a natural community, landscape restoration through revegetation serves a variety of functions. It can improve slope stability, restore the ecosystem quickly, conserve biodiversity, and blend the revegetated site with the surrounding landscape.

320

H. YOSHIDA

As Yamadera et al. (1993) pointed out, it is very difficult to restore a community when restoration is impeded by human interference. To create or restore a natural community, it is necessary to develop a plant community in which the species can survive without human assistance. Humans, therefore, should support rather than control the survival of these species. It is important to select a suitable slope revegetation method. For successful revegetation, a suitable goal must be chosen based on the slope’s condition, and a variety of species should be used for revegetation rather than a monoculture; pioneer and mid-successional plants should be combined with climax species, and endemic plants should be used. Seeding should also be chosen so as to stabilize the slope and encourage natural succession of the forest that develops. 4.1 Setting a suitable revegetation goal for the actual slope conditions Generally speaking, a engineered slope is not a suitable environment for plant growth, since this type of landscape is often characterized by rocky conditions and shallow or absent soil. Therefore, it is not surprising that engineered slopes are one of the most difficult landscapes to restore. As a result, it is important to combine revegetation with the establishment of an appropriate foundation to make the growing environment less challenging for plants. If the goal is to reproduce a community of endemic woody plants, defining this goal during the early stages of the revegetation process is very important. Specifically, a suitable plant community such as the one shown in Table 4 should be chosen based on two conditions: first, the main species should eventually dominate the study site, and second, the outward form should resemble a natural vegetation community (Slope Revegetation Research Group, 2004). Constructed solid structures covered with enough soil to permit planting will allow vegetation to recover on the slope. However, this method is very expensive and the vegetation community requires considerable maintenance. As a result, it is important to find a more cost-effective method, such as the various TGM approaches. Table 4. First revegetation target of plant community (Slope Revegetation Research Group, 2004).

Main species Hearbaceous species and grasses Woody pioneer species Mid- to late successional species

Outward form of community Grassy plain Bushes (shrubs) or forest Bushes (shrubs) or forest

4.2 Seeding with mid- to late-successional and climax species in addition to pioneer plants Revegetation must eventually establish a self-sustaining plant community. This means that it is necessary to seed with species that can rapidly cover and begin to improve damaged sites and also fertilize the soil (i.e., pioneer species), especially in

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

321

areas where the site has lost its topsoil. As well, mid-successional species should be included to develop an increasingly complex plant community, and climax species should be included to help form the final community.

Photo 4. A photograph of revegetation after 6 years (1981). Rapid revegetation using exotic grasses was applied on site A and rapid reforestation using mainly leguminous species was applied on site B. 　

(dividing line)

D iagonaldistance (m )

8

6

4

2

0 -12

-10

-8

-6

-4

(Festuca arundinacea com m unity) A m orpha fruticosa (<1m ) A .f.(4m <)

-2 0 2 H orizontaldistance (m )

4

6

8

10

12

(A m orpha fruticosa com m unity)

A .f.(1-2m ) A .f.(2-3m ) A .f.(3-4m ) 施工24年後の木本植物の成立位置 Ligustrum japoni cum Q uercus phyllyraeoides others

Figure 2. Comparison of two different plant communities after 24 years.

In the Amorpha fruticosa community (site A of Photo 4), plant succession occurs faster than in the Festuca arundinacea community (site B of Photo 4). The presence of a community of woody plants hastens natural invasion by evergreen trees by means of seed dispersal by birds and small animals that cache seeds. When comparing plant succession in areas where grasses (Festuca arundinacea) and woody (e.g., A. fruticosa) plant communities have been established using the TGM method on rocky engineered slopes (Photo 4). After 24 years, the woody plant community became more successfully established (Figure 2). Studies have shown that the woody plant community acts as a habitat for birds and small animals, and

322

H. YOSHIDA

that these species disperse seeds, allowing natural succession of the evergreens (Yoshida, 2000, 2003). The speed of the recovery depends on when the seeding work is done. Figure 3 provides general guidelines for the vegetation sere at different time periods. When restoring engineered slopes near an expressway (Hoshiko, 1999), nature restoration techniques using endemic woody species should be used, with a combination of seeds from pioneer, mid- to late-successional, and climax species. This approach improves restoration of the slope. Compared to an approach in which only exotic grass species are selected, this approach produces not only higher slope stability but also restores the landscape more rapidly, with minimal subsequent tending required. Current slope revegetation work introduces herbaceous plants or woody pioneer plants and leaves the resulting community to undergo natural succession. However, successful natural invasion that would blend the resulting community with the surrounding landscape is very difficult to achieve. Therefore, it is necessary to include seeds of species from later successional stages that are found in the surrounding natural environment.

Figure 3. Illustration of the vegetation sere at different time periods after revegetation.

Nature restoration techniques using a mixture of endemic woody species can restore shade-intolerant mixed woodland 32 to 33 years earlier than revegetation using only grassy species and 7 to 19 years earlier than reforestation using mainly leguminous species.

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

323

4.3 Seeding work as an effective method for slope stabilization Slope stability is one of the most important objectives of slope revegetation; this is true both because collapse of slopes can have serious negative environmental and engineering consequences and because unstable slopes will prevent successful revegetation. Past studies such as following photographs have demonstrated that seeded plants develop better than planted vegetation. After 7 years of revegetation work, Yamadera et al. (2002) found that trees that developed from seed formed stronger root systems than planted trees (Photo 5). In addition, Fukunaga (1996) suggest that seeded vegetation becomes established quickly (Photo 6). Furthermore, the main roots of seeded trees can often expand into gaps in the underlying rocks and lateral roots can expand more widely, possibly intertwining with the roots of other trees (Fukunaga et al., 1997). The differences in the two types of root system are shown in Figure 4. When designing a restoration system for slopes, it is necessary to form plant communities that will develop strong root systems (i.e., to favor the use of seeds) rather than focusing on the aesthetics of the planting work.

Photo 5. The root conditions of planted (left) and seeded (right) Quercus phillyraeoides tree after 7 years. (Yamadera et al., 2002).

Photo 6. The root conditions of seeded and planted Rhaphiolepis umbellata trees (Fukunaga, 1996).

324

H. YOSHIDA

The left photograph shows the results 1 year after work and the right photograph displays the root condition after 2 years. The planted trees (left) have less-abundant roots with less intertwining between trees, and often fail to penetrate deeply into the soil. In contrast, the roots of seeded trees (right) are more abundant, form a denser network, may become intertwined, and penetrate more deeply into the soil.

Figure 4. Conditions of the roots of planted and seeded trees.

Figure 5. A comparison of the thick-growth-media (TGM) method and the double-layer spraying system (DLSS).

5. CONCLUSIONS Seeding in Japan has a long history of the development of increasingly effective revegetation techniques. Further improvement of slope restoration technology has been reported, including mixed seeding using the TGM method (Yoshida, 2002) but supplemented by the development of a double-layer spraying system (DLSS) to improve the use of seeds of endemic woody plants and reduction of the amount of seeds required (Figure 5). In the DLSS method, spraying occurs in two layers that are produced during the same pass with the spraying equipment: the first layer contains no seeds and serves primarily as the base for cultivation, and the second (top) layer contains the seeds. This offers the advantage of creating both layers using a single spraying process. In the past, the TGM method required workers to spray each

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

325

layer separately if managers decided that a double layer was required. Because this was more difficult and expensive, double-layer spraying based on previous TGM methods has generally not been applied; that is, workers sprayed only a single layer containing both seeds and the base layer. Thus, the DLSS method makes more effective use of available seed and reduces the cost of the operation (Yoshida et al., 2004). The technique of restoration of slopes is developing continuously. The Japan Society of Revegetation Technology (2002) suggested revegetation as a means to conserve biodiversity and proposed the basic principles of slope revegetation for nature restoration (Slope Revegetation Research Group, 2004). However, to expand this technology, the development of technology for the storage of seeds of woody plants, especially recalcitrant seeds (Akimoto et al., 2001, 2002, 2004), and technology for early detection of germinability of these seeds (Cho et al., 2004) are both necessary. Most revegetation work has used seeds buried in the soil (i.e., in the soil seed bank). However, exotic species may also be present in this soil and can have a potentially devastating impact on Japanese ecosystems. Slopes consisting of naturalized and exotic plant species have limited usefulness in terms of slope restoration because they are less well adapted to the environment than endemic species. Therefore, seeding with a mixture of endemic species can be an effective and powerful tool for restoring slopes if seeding is combined with the development of artificial seed banks containing only the seeds of endemic species. Japan’s Invasive Alien Species Act was implemented in June 2005. Slope revegetation has emerged as a necessary technology to help combat invasive species by creating more vigorous ecosystems that can better exclude the invaders. I hope to contribute toward the technical development of slope restoration as a means to prevent both erosion and invasion by damaging foreign species. ACKNOWLEDGEMENT I would like to thank Professor Y. Morimoto and Assistant Professor Shibata of Kyoto University for their detailed comments and suggestions. REFERENCES Akimoto, T., Cho, S. and Esashi, Y. (2001). Stimulation of the aerobic respiration and deterioration of woody plant seeds by acetaldehyde during the storage period. Journal of the Japanese Society of Revegetation Technology, 27(1), 227-230. (in Japanese) Akimoto, T., Cho, S. and Esashi, Y. (2002). Possible involvement of endogenously evolving acetaldehyde in the low storability of woody recalcitrant seeds. Journal of the Japanese Society of Revegetation Technology, 28(1), 177-180. (in Japanese) Akimoto, T., Cho, S., Yoshida, H., Furuta, H. and Esashi, Y. (2004). Involvement of acetaldehyde in seed deterioration of some recalcitrant woody species through the acceleration of aerobic respiration. Plant and Cell Physiology, 45, 201-210. Akita, M., Hiramatsu, H., Horie, Y. and Ushijima, K. (2000). The application of woody plants in Sakurajma (Pyroclastic materials). Journal of the Japanese Society of Revegetation Technology, 25(4), 507-508. (in Japanese)

326

H. YOSHIDA

Akita, M., Shimoshinbara, H. and Ushijima, K. (2001). Instance of woody planting by natural growth seeds in Yakushima. Journal of the Japanese Society of Revegetation Technology, 27(1), 250-253. (in Japanese) Cho, S., Esashi, Y. and Yoshida, H. (2004). Short-term seed quality testing method for woody plant species. Journal of the Japanese Society of Revegetation Technology, 30(1), 261-264. (in Japanese) Egashira, M. (1988). Introduction of woody plant species by seeding works. Convention press of Japanese Society of Revegetation Technology, 19, 22-25. (in Japanese) Ehime Prefecture Engineering Works Part (1999). Technical Manual of Slope Reforestation Works by Endemic Species. (in Japanese) Fukunaga, K. (1996). Difference of root growing characteristics on seeded and planted plants (III), Rhaphiolepis umbellata introduced after 2 years. Convention Press of Japanese Society of Revegetation Technology, 27, 40-43. (in Japanese) Fukunaga, K., Okiyama, Y., Yokota, M., Kikuchi, T. and Ishida, K. (1997). Root growing characteristics of seeded plants on the hard cut slope (I), -Ligustrum lucidum introduced after 7 years -. Convention Press of Japanese Society of Revegetation Technology, 28, 13-16. (in Japanese) Furuta, T. and Yoshida, H. (2004). Revegetation example for the purpose should be formed Quercus serrata - Quercus mongolica var. grosseserrata community by seeding works in cold district. Journal of the Japanese Society of Revegetation Technology, 30(2), 377-382. (in Japanese) Furuta, T., Okamura, Y., Hayashi, K., Yoshida, H., Inoue, K. and Adachi, N. (2004). Revegetation examples for natural conservation on the cut slopes using Double Layer Spraying System. Journal of the Japanese Society of Revegetation Technology, 30(1), 281-284. (in Japanese) Hasegawa, T. and Nishizawa, M. (1998). Example of slope reforestation by endemic species in the Daisen Oki Nature Park. Convention Press of Japanese Society of Revegetation Technology, 29, 204207. (in Japanese) Hoshiko, T. (1999). A study on the invasions of woody-plants and seed dispersal from on man-made expressway slopes. Journal of the Japanese Society of Revegetation Technology, 25(2), 102-114. (in Japanese with English abstract) Inoue, K. and Yoshida, H. (2002). Revegetation by woody plant seeding technique considering both the natural landscape and the landscape design in the park. Journal of the Japanese Society of Revegetation Technology, 28(1), 154-157. (in Japanese) Ishida, K., Matsumoto, K., Kunihiro, M., Yoshitake, S. and Horie, Y. (2000). An example of woody planting on a huge cutting slope in recreation facility, Changes of evergreen broad-leaved tree’s growth applied by seeding works. Journal of the Japanese Society of Revegetation Technology, 25(4), 521-524. (in Japanese) Ishida, K., Shioda, Y., Katou, K., Nagashii, K. and Sonoda, E. (2001). Example of woody planting on a wire cylinder for the landscape measure. Journal of the Japanese Society of Revegetation Technology, 27(1), 247-249. (in Japanese) Japan Association of Agricultural Engineering Enterprises (1990). Slope Stabilization Works -Guidance of Design and Construction-. Rural Culture Association, Tokyo. (in Japanese) Japan Highway Landscape Association (1986). Technical Guidelines of Restoration to the Ruined Bare Land. Tokyo, JHLA. (in Japanese) Japan Road Association (1986). Standard of Road Earthworks, Slope Protect Works. Maruzen, Tokyo. (in Japanese) Japan Road Association (1999). Standard of Road Earthworks, Slope Protect Works. Maruzen, Tokyo. (in Japanese) Japan Society of Revegetation Technology (1990). Glossary of Revegetation Technology. Sankaido, Tokyo. (in Japanese) Japan Society of Revegetation Technology (2002). Revegetation for conserving biodiversity. Journal of the Japanese Society of Revegetation Technology, 27(3), 481-491. (in Japanese) Kida, Z., Fukuzumi, Y., Niimoto, T., Nakayama, Y. and Nishizawa, M. (2002). Study on the experimental revegetation on the artificial slopes in Kinki Engineering Office. Journal of the Japanese Society of Revegetation Technology, 28(1), 158-161. (in Japanese) Kikuchi, T. (1980). On Spraying Revegetation Method that vegetated on the rocky soil-lacked land. Journal of the Japanese Society of Revegetation Technology, 7(1), 36-37. (in Japanese) Kurata, M. (1959). Outline of Revegetation Technology. Yokendo Press, Tokyo. (in Japanese) Kurata, M. (1979). Revegetation Technology. Morikita Press, Tokyo. (in Japanese)

SEEDING ON SLOPES FOR NATURE RESTORATION IN JAPAN

327

Ministry of Construction River Bureau Sand Control Department (1996). New Planning and Examples of Slope Collapse Prevention Works -Technical Guideline of Steep Hillside Collapse Prevention Works-. Japan Sabo Association, Tokyo. (in Japanese) Ministry of Environment, Conservation of Nature Bureau (1982). Explanation about Standard of Slope Revegetation on the Nature Park. Japan Highway Landscape Association, Tokyo. (in Japanese) Nakaide, K., Umezawa, H. and Miyamaru, M. (1990). A report of revegetation works in the land reclamation of Komatsu eastern industrial zone. Journal of the Japanese Society of Revegetation Technology, 25(3), 63-72. (in Japanese) Nitta, S. (1959). Seeding on the steep slope by cement-gun. Lecture Press of Japanese Forestry Society, 69, 430-431. (in Japanese) Nitta, S. (1995). Fifty years in postwar days of revegetation works. Course of revegetation works Publication of the 30th anniversary in establishment (pp. 16-31). Japan Revegetation Work Association, Tokyo. (in Japanese) Nitta, S. and Kobashi, S. (1961). Fundamental studies on the rapid stabilizing method by seeding for cut and filled slopes (II). Journal of the Japanese Institute of Landscape Architecture, 24(4), 78-83. (in Japanese) Oda, Z. (1976). On spraying revegetation method for hard grounds. Journal of the Japanese Society of Revegetation Technology, 4(1), 19-21. (in Japanese) Oyama, Y., Tan, S. and Nishizawa, M. (1996). Trial of rapid reforestation on the coast sand slope of Hitachi-Naka port city construction project. Convention Press of Japanese Society of Revegetation Technology, 27, 14-17. (in Japanese) Road Management Technology Centre (1996). Enterprise Guidance of Slope Re-Revegetation (tentative). RMTC, Tokyo. (in Japanese) Saito, Y., Nishihara, Y., Wakasugi, K. and Shigeno, S. (1999). Introduce experiment of Quercus acutissima on the bank slope by seeding works (III). Convention Press of Japanese Society of Revegetation Technology, 30, 336-339. (in Japanese) Sasaki, A., Fujii, K., Fujihisa, M., Kohno, S., Inoue, S., Ezaki, T. and Chun, K.W. (2003). Revegetation of the face of slopes using the native woody plants. Journal of the Japanese Society of Revegetation Technology, 29(1), 269-272. (in Japanese) Sato, K. and Ono, Y. (1942). Test of seed mixed seeding on erosion control afforestation (II). Convention Press of Japanese Forestry Society, 704-713. (in Japanese) Seino, T., Anzai, Y., Kanke, S. and Nishizawa, M. (1995). A case of arboreous plants slope seeding at a natural park in a snow cold region. Journal of the Japanese Society of Revegetation Technology, 21(1), 41-49. (in Japanese) Shibata, T. (2001). A study on the afforestation on cut slopes in the early stage by seeding method (V), Characteristics of dispersal on Myrica rubra, Salix subfragilis and Euonymus japonicus-. Journal of the Japanese Society of Revegetation Technology, 27(1), 243-246. (in Japanese) Shikoku Construction Bureau, Japan Highway Public Corp (1998). Slope Reforestation by Seeding Works, -Manual of Plan, Construction and Management (tentative)-. JH, Takamatsu (in Japanese) Shikoku Construction Bureau, Ministry of Land Infrastructure and Transport (2002). Guidance of Slope Seeding Works by Endemic Arboreous Plants. (in Japanese) Slope Revegetation Research Group, Japan Society of Revegetation Technology (2004). Guidance of basic thinking in slope revegetation for nature restoration, Journal of the Japanese Society of Revegetation Technology, 29(4), 509-520. (in Japanese) Takeuchi, S. and Nishizawa, M. (1998). An example of introduce woody plant species on the coastal road cut slope in the national park. Convention Press of Japanese Society of Revegetation Technology, 29, 102-105. (in Japanese) Tsuchimuro, K., Matsubara, S., Yamanaka, T. and Shimoda, N. (1998). Introduction of woody plant species on the slope of woodland path. Convention Press of Japanese Society of Revegetation Technology, 29, 212-213. (in Japanese) Yamadera, Y. (1986). Proposed rapid reforestation method by seeding works. Journal of the Japanese Society of Revegetation Technology, 12(2), 25-35. (in Japanese) Yamadera, Y. (1989). An Experimental Study on Betterment of Revegetation Technology for Steep Slopes. Kyoto Univ. Doctoral Thesis. (in Japanese with English abstract) Yamadera, Y. (1995). Rapid reforestation method by seeding works Forefront of slope revegetation. In Kobashi, S. and Murai, H. (Ed.), The Most Advanced Slope Revegetation Technology (pp. 148-170). Soft Science Co., Tokyo (in Japanese)

328

H. YOSHIDA

Yamadera, Y., Abo, A. and Yoshida, H. (1993). Design of Slope Revegetation for Restoration of Nature Environment, Basis of Slope Revegetation and Model Design. Japan Association of Agricultural Engineering Enterprises, Tokyo. (in Japanese) Yamadera, Y., Yang, X. and Miyazaki, T. (2002). Study on the different of drawing-out resistance between planted tree and seeded tree. Journal of the Japanese Society of Revegetation Technology, 28(1), 143-145. (in Japanese) Yamanishi, T., Kitami, T., Seshimo, K., Naruse, H. and Tan, S. (2004). The trial of reforestation of Japanese black pine (Pinus thunbergii) using SF Afforestation System (the high level crumb composition spraying system) in J-PARC. Journal of the Japanese Society of Revegetation Technology, 30(1), 227-230. (in Japanese) Yoshida, H. (1990). A method of arboreous plant seeding (Ligustrum lucidum). Journal of the Japanese Society of Revegetation Technology, 16(1), 52-55. (in Japanese) Yoshida, H. (1991). Introduction of evergreen trees in the soft rock slopes by seeding works, introduce example of Ligustrum lucidum and Rhaphiolepis umbellate. Journal of the Japanese Society of Revegetation Technology. 16(4), 46-53. (in Japanese) Yoshida, H. (1998). Stratificational revegetation of arboreous plant community by seeding method. Journal of the Japanese Society of Revegetation Technology, 24(2), 90-98. (in Japanese) Yoshida, H. (2000). A study on the plant succession of the revegetative communities by spraying thick cultivative base. Journal of the Japanese Society of Revegetation Technology, 25(4), 305-310. (in Japanese with English abstract) Yoshida, H. (2002). A study on the mixed seeding technology of woody plants by spraying thick cultivative base. Journal of the Japanese Society of Revegetation Technology, 27(4), 594-604. (in Japanese with English abstract) Yoshida, H. (2003). Characteristics of succession in plant communities revegetated by spraying with plant cultivative base, Journal of the Japanese Society of Revegetation Technology, 29(2), 331-342. (in Japanese with English abstract) Yoshida, H. (2005). Slope revegetation by seeding works and its monitoring method. Journal of the Japanese Society of Revegetation Technology, 30(3), 532-540. (in Japanese with English abstract) Yoshida, H. and Furuta, T. (2004). An application example of developing the woody plant’s communities on cutting slopes by forming a spraying thick cultivative base under the Natural Remedy Seeding Method. Journal of the Japanese Society of Revegetation Technology, 29(4), 482-494. (in Japanese) Yoshida, H. and Hosaka, K. (1992). Compound community planting by thinner cultivation base spraying method, Application example of OS-3 revegetation method. Journal of the Japanese Society of Revegetation Technology, 17(3), 175-181. (in Japanese) Yoshida, H. and Morimoto, Y. (2004). Historical changes of seeding works as nature restoration technologies of man-made slopes in Japan. Proceedings of the First EAFES International Congress: (pp. 220-221), Mokpo, Korea Yoshida, H. and Morimoto, Y. (2005). A study on the effects of mixed seeding of Chinese-grown Indigofera spp. and evergreen broad-leaved trees. Journal of the Japanese Society of Revegetation Technology, 31(2), 269-277. (in Japanese with English abstract) Yoshida, H. and Yamadera, Y. (1989). A Study on the Technical Application of ON Spraying Method. Tokyo Univ. of Agri Research Student Thesis. (in Japanese) Yoshida, H., Kikuchi, T. and Ishida, K. (1995). Plant autumnal tints arboreous by seeding method, Acer palmatum, Sorbus commixta. Journal of the Japanese Society of Revegetation Technology, 20(4), 255-264. (in Japanese) Yoshida, H., Furuta, T., Ito, K. and Takayanagi, H. (2004). Double layer spraying system on the purpose of effective utilization of domestic wild woody plants seed and the reduction of seeding works expenses. Journal of the Japanese Society of Revegetation Technology, 29(3), 438-445. (in Japanese)

CHAPTER 20

WETLANDS AND RIPARIAN BUFFER ZONES IN LANDSCAPE FUNCTIONING

Ü. MANDER, K. KIMMEL Institute of Geography, University of Tartu, Estonia

Abstract. According to the main landscape functions - productional (economic), regulatory (ecological), and social (informative) functions - the role of wetlands and riparian buffers regarding landscape functioning is analysed. An analysis of the literature and authors’ earlier research results and experience has been used to highlight these functions. Special focus is devoted to the regulation of nutrient fluxes by wetlands and riparian buffer zones at landscape level. Examples from Estonia are used to illustrate the relevance of wetlands and riparian buffers in landscapes.

1. INTRODUCTION Estimates of global wetland area range from 5.3 to 12.8 million km2. About half of the global wetland area has been lost; although the 1971 Ramsar Convention has helped 144 nations protect the most significant remaining wetlands (Mitsch and Gosselink, 2000). Despite the likelihood that remaining wetlands will occupy less than 9% of the earth's land area, they contribute more to annually renewable ecosystem services than their small area implies (Zedler and Kercher, 2005). Biodiversity support, water quality improvement, flood abatement and carbon sequestration are key functions that are impaired when wetlands are lost or degraded at regional and landscape level, and productional and social services should also be added to the significant roles of wetlands (Mitsch and Gosselink, 2000; Verhoeven et al., 2005). For instance, the temperate peatlands that cover roughly 3.5 million km2 of land throughout the world contain about 455 Gt of carbon, which is almost equivalent to the carbon stored in all of the living things on the surface of the planet, and represents around 25% of all the soil carbon on earth (Moore, 2002). These bogs are sink for atmospheric carbon, and their carbon uptake accounts for about 12% of current human emissions. They vary considerably in their form and structure and are 329 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 329–357. © 2007 Springer.

330

Ü. MANDER AND K. KIMMEL

an important resource for scientific research, including the study of past environments and climate change, and they are also valuable in environmental education. They are relatively low in biodiversity, but their fauna and flora are distinctive, and many groups are confined to this habitat. For all of these reasons, the future conservation of peatlands and other wetland ecosystems is a matter for concern (Moore, 2002). Likewise, riparian buffer zones represent an area of increasing relevance, as the human modification of the landscape continues unabated (Naiman et al., 2005; Goetz, 2006). As in the case of wetland ecosystems, they regulate material and energy fluxes, especially providing water quality improvement (Kuusemets and Mander, 1999; Correll, 2005), supporting biodiversity (Verhoeven et al., 2006), and providing productional and social services at a regional and landscape level (Naiman et al., 2005; Mander et al., 2005a; Verhoeven et al., 2006). Water quality in most of stream catchments and river basins worldwide is severely impacted by nutrient enrichment as a result of agriculture, forestry or other human activities. Water-resource managers are considering the potential role of riparian zones and floodplain wetlands in improving stream-water quality, as there is evidence at the site scale that such wetlands are efficient at removing nutrients from throughflowing water. However, recent studies have highlighted disadvantages of such use of wetlands, including emissions of greenhouse gases and losses of biodiversity that result from prolonged nutrient loading (Verhoeven et al., 2005). Therefore, further investigation of the complex role of both wetlands and riparian buffer ecosystems is an important task that will provide with important information for sustainable landscape management. The main objective of this paper is to give an overview and estimate wetlands’ and riparian buffer strips’ role at landscape level regarding the main landscape functions. A special focus is given on the regulation of nutrient fluxes by wetlands and riparian buffer zones. 2. MATERIAL AND METHODS A total of 147 literature sources were analysed in order to estimate the relationships between the nitrogen and phosphorus load and retention in various ecosystems. There were 45 and 29 literature sources on nutrient budgets in wetlands and riparian ecosystems respectively (see Mander and Mauring, 1994). In addition, rivers (38 papers) and lakes (35 papers) were analysed as reference ecosystems for the loadretention relationship of N and P. The wetland ecosystem studied included both areas where the water table is near the land surface and free-water-surface wetlands. According to watershed perspective, their variety extended from the fringe wetlands of lakes to riparian wetlands and depressional wetlands (Figure 1a). On the other hand, conditions that promote the formation of hydric soil and growth-emergent aquatic vegetation were the main selection criteria. Floodplain wetlands were represented by different swamps, valley bogs, cypress swamps, delta wetlands, and fens, whereas raised bogs, transitional bogs, prairie potholes, and kettle-holes, marshes, fens and shallow ponds represented

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

331

both surface water and groundwater depressions. In addition, mires, raised bogs, pocosins and blanket bogs, being typically located in conditions of surface water and groundwater slope, belonged to the analysed wetland systems (see also the classification by Bullock and Acreman, 2003). In most cases the water received by wetlands comes from natural or agricultural upland areas. Thus the wetlands located between the upland aquatic ecosystems were considered to be sinks, sources and transformers of nutrients (Figure 1b). Among constructed wetlands, both a free water surface system and vertical and/or horizontal subsurface flow soil filters were analyzed (see classification by Vymazal, 2001). In addition, some constructed wetlands and semi-natural wetlands used as media for the secondary/tertiary treatment of municipal or farm wastewater are taken into consideration. In addition to runoff and wastewater, effluent contributions from precipitation are generally taken into consideration in the budget. Groundwater inflow and N fixation were only measured in some wetlands.

a

b

Figure 1. Various geomorphological types of wetlands in the landscape (left), and a summary diagram of the hydrological and biogeochemical functions of wetlands in the watershed (right) (adapted from Mitsch and Gosselink, 2000).

The riparian buffer zones and buffer strips studied were represented by forest buffers (riparian forests of different site types, in some cases wetland or paludified forests, shelterbelts, hedges) and grassland buffers (grassy slopes, e.g. the downslopes of feedlots, grassland strips in strip-cropping, grassland strips as parts of complex buffer zones) (see Mander and Mauring, 1994). In addition to water quality improvement, we also analysed other services of wetlands and riparian buffer zones, regarding their role as important multifunctional landscape elements. For this purpose we used the system of landscape functions devised by Bastian and Schreiber (1994), which is quite similar to the landscape services classification created by de Groot (1992), and the nature services approach established by Costanza et al. (1997). The main landscape functions are;

332

Ü. MANDER AND K. KIMMEL

Productional (economic) functions –Biomass production –Water supply –Suitability of non-renewable resources Regulatory (ecological) functions –Hydrological functions –Regulation of material and energy fluxes –Meteorological functions –Regulation and regeneration of populations and biocoenoses –Habitat (genetic) functionSocial functions –Psychological (aesthetic and ethical) functions –Informative functions –Human ecological functions –Recreational function Table 1. Relevance of wetlands and riparian buffer zones in landscape functioning. Landscape functions adapted from Bastian and Schreiber, 1994 and de Groot, 2006. nr – not relevant. Landscape functions Productional (economic) functions Biomass production (including bio-energy sources) Water supply Suitability of non-renewable resources Regulatory (ecological) functions Hydrological functions Regulation of material and energy fluxes Meteorological functions Regulation and regeneration of populations and biocoenoses Habitat (genetical) function Social (informative) functions Psychological (aesthetic and ethical) functions Informative functions Cultural-historic value Human ecological functions Recreational function

Wetlands

Riparian buffer zones

++ ++ nr

+ nr nr

+++ +++ ++ +++ +++

++ +++ + + +

++ ++ + + ++

++ + nr nr +

3. RESULTS AND DISCUSSION 3.1 Landscape functions 3.1.1 Productional functions 3.1.1.1 Biomass production Since the very earliest times, Homo sapiens has used wetlands for the gathering of foods and materials. Fish, molluscs, frogs, waterfowl eggs and also wetland plants were the earliest foods of ancient hominids and probably one of the reasons for

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

333

explosive brain growth (Cunnane, 2005). Today food from wetlands also plays an important role, especially in Third World countries, although in our paper we concentrate on wetlands’ functionality in providing bio-energy and building material production. Reed (Phragmites australis) from wetlands is already a well-known and valuable building material, especially for roofs. Likewise, cat-tail (Typha latifolia) biomass can be easily used in various construction work (Wild et al., 2002). The leaf mass of cat-tails shows a high porosity and elasticity in the aerenchyma tissue. At the same time, there is a uniform distribution of bast fibres. Thus the leaves have a high stability and show excellent insulation qualities. Furthermore, the leaf tissue has a high content of polyphenols, so the dry raw material shows a high resistance to decay. A mixture of cat-tail chips and clay material is used for the production of cost-efficient building blocks. Cat-tail wool is an excellent insulation material (Wild et al., 2002; Mauring, 2003). The aboveground biomass will be harvested in winter, when the leaf mass has a minimum water content. The harvesting technique has been adopted from a common reed harvest. There are vehicles with wide tyres or caterpillar tracks and designed for transporting loads in amphibious environments (Wild et al., 2002). Wetlands can be seen as an important basis for the sustainable development. In addition to biodiversity and landscape functions, they can be widely used for wastewater treatment and energy/material production. We assume that a significant proportion of oil shale, the main national fossil energy source, but also of imported fuel and gas, can be replaced by energy production from wetlands. Based on the average biomass production of reed and cat-tails (1.5 kg m-2 yr1 ), the estimated energy value of one hectare of energy reed-bed is approximately 200 GJ. On the other hand, about 300,000 ha of the approximately 594,000 ha of drained agricultural areas in Estonia will be abandoned in coming decades. A major part of this area (excluding all protected areas of about 100,000 ha, forested wetland sites, raised bogs and transitional marshes) is not of interest for further agricultural use and can be adapted to wastewater treatment and/or energy wetlands. However, only about 100,000 ha are most favourable as energy/treatment wetlands. The biomass of reed and/or cat-tails from this favourable area can provide about 27,000 TJ of energy a year. This can cover about 30% of Estonia’s annual heat consumption and 20% of its electrical power production. This rate can be increased by combining biomass use with an innovative technology such as micro-turbines or thermoacoustic-Stirling engines (Mander et al., 1999). In respect of the nutrient cycle, one can presume that the land-use change from drained grassland to treatment and energy wetland can reduce the pollution load with trace gases.

334

Ü. MANDER AND K. KIMMEL

Figure 2. Preferred areas for energy and treatment wetlands in Estonia. Adapted from Mander et al., 1999.

The situation regarding land use in Estonia is beneficial for the development of this wetland function. However, its change during the last century has been remarkable. The main trends in land-use dynamics have been a decrease in agricultural land (from 65% in 1918 to 27% in 1999) and an increase in forested areas (from 21 to 51% respectively). Natural and semi-natural grasslands show the most significant decrease (Mander and Palang, 1999). This trend is especially remarkable after the regaining of independence in 1991. Land amelioration has shifted agricultural activities from former arable lands to marginal areas (natural grasslands, wetlands). This has caused an essential disturbance in the stabilized nutrient cycling in landscapes. Since the early 1990s, many drainage systems have fallen into disrepair. Substantial areas of drainage infrastructure are now in need of rehabilitation. About 70% of polder systems (10,000 ha in total size) are already in a stage of natural succession. With the shift in agricultural policy and accession to the European Union, there is no realistic possibility to return these drained wetland areas to agricultural use. In our estimation there are approximately 300,000 ha of currently or potentially abandoned former wetland areas in Estonia (Mander et al., 1999). The percentage of wet and moist land in Estonia is relatively large. Peat soils cover 21% of Estonian territory, and gleysols cover even more – 33%. Not all of this territory is acceptable for the creation of new wetlands. After the exclusion of protected areas (about 100,000 ha are protected as special mire protection sites or

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

335

located in protected areas of different regimes; Mander et al., 1999), forested sites, raised bogs and transitional marshes, there remains about 594,000 ha of potential territory for energy and treatment wetlands. Some of these sites are in riparian or coastal buffer zones, some are well drained and intensively used in agriculture, some are too small, and some are located in distant places. Presumably reed and cattail plantations are not the only possible land use for these potential areas. About 100,000 ha of the most favorable sites for Phragmites and Typha plantations were selected using additional criteria. Sites located at least 200 m from the coastline, lake shores, riverbanks and also undrained bogs and larger sites closer to the settlement are preferred (Figure 2). In addition, all peat-mining areas are to be considered as potential reclamation sites (Mander et al., 1999). Drained wetlands are the second largest source of unbalanced carbon flow in Estonia. According to the estimations, the average rate of decomposition of the sphagnum in drained wetlands is about 4.3 t CO2-C ha-1 yr-1, whereas the fixation rate of carbon due to the formation of the sphagnum is only 0.9 t CO2-C ha-1 yr-1. The amount of decomposed sphagnum exceeded the amount of sphagnum formed by about 2660 thousand tons of CO2-C annually. According to the data from the beginning of the 1990s, the carbon flow originating from drained wetlands alone exceeded the environmental space of Estonia about 4 times and was 6.7 times higher than the average global per capita emission of carbon from land use changes (Mander et al., 1999). The primary production of Estonian forests does not compensate the carbon flow from the wetlands. Therefore, one of the means to improve the CO2 balance could be the restoration of wetlands and their transformation into productive sites. Measurements of trace gas fluxes (nitrous oxide, methane) in constructed wetlands have shown significantly higher cumulated nitrous oxide fluxes than in drained grasslands (Wild et al., 2002). As regards methane, the situation may be opposite, however, and the global warming potential (GWP; summarises the influence of all greenhouse gas fluxes) of newly created wetlands is more favourable than that of the drained grasslands. The development of vegetation in wetlands can compensate the lack of oxygen in sediments. This demonstrates how root oxygen release from plants such as cat-tail can significantly alter rates of biogeochemical processes such as methanogenesis (Jespersen et al., 1998). In respect of the nutrient cycle, one can suppose that the land-use change from drained grassland to treatment and energy wetland can reduce the pollution load from trace gases. The productional functions of riparian buffer zones have not been comprehensively considered. Therefore there is not much information available about this unction of riparian buffers. However, due to the need for the management of riparian grassland ecosystems and riparian forests, these areas can provide some biomass (see Figure 3). For instance, it is recommended that riparian forest stands (especially alder forests) be managed through regeneration cutting, in order to keep their nutrient removal rate high (Mander et al., 2005b).

336

Ü. MANDER AND K. KIMMEL

3.1.1.2 Water supply This function of wetlands will be considered under the chapter on general hydrological functions in the complex of regulatory functions of energy and material fluxes.

Figure 3. Possible design of riparian buffer zones. Grassed waterway with managed grassland buffer strip (above) and a complex riparian buffer zone with 3 functional parts: grassland buffer strip, managed forest strip and old forest strip (below). Adapted from Lovell and Sullivan, 2006. See also Lowrance et al., 1997.

3.1.2 Regulatory functions 3.1.2.1 Hydrological and meteorological functions This is one of the most important functions of wetlands. Wetlands are significant in altering the water cycle from the global to the local level. Bullock and Acreman (2003) analysed 169 scientific studies published during the period 1930-2002, and

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

337

found 439 statements on the hydrological significance of wetlands. Only 19% of them concluded the wetlands’ influence on water cycle to be neutral or insignificant. However, this study also demonstrated that a wetland’s influence can diff greatly depending on a region’s geomorphology, climate and soil and vegetation cover (see Fig. 4).

Figure 4. Examples of wetland hydrology and surface water-ground water interaction: a) wetland perched above the water table; b) groundwater discharge wetland; c) groundwater recharge wetland; d) riparian wetland with groundwater discharge (adapted from Mitsch and Gosselink, 2000).

Almost all studies showed that floodplain wetlands reduce and delay floods, with examples from all parts of the world (Bullock and Acreman, 2003). However, there are examples of headwater wetlands increasing flood peaks. Likewise, the majority of wetlands studied increase average annual evaporation or reduce average annual river flow. However, about 10% of studies conclude the opposite. Thus in evaluating wetlands’ role, one should carefully study the local conditions. Hydrological regulation in riparian buffer zones is less important than in wetlands (see Table 1), and is related to nutrient transformations. Typically, a higher hydrological load to buffer ecosystems decreases retention time and the intensity of several retention processes (see Mander and Mauring, 1994).

338

Ü. MANDER AND K. KIMMEL

3.1.2.2 Regulation of material and energy fluxes Wetlands Our study of 169 water-related ecosystems demonstrate great differences between the studied systems in terms of nitrogen and phosphorus retention. For instance, the retention capacity of wetlands and riparian buffers is about 300 times higher than that of oligotrophic lakes (Figure 5). On the other hand, smaller systems with longer retention times can be more effective than large systems.

1 - oligotrophic lakes 2 - meso- & eutrophic lakes 3 - middle-size to large rivers 4 - small rivers 5 - riparian forest buffers 6 - riparian grassland buffers 7 - natural wetlands 8 - constructed wetlands

1

-2

-1

Retention (g m d )

1.2

0.8 0.6 0.4

0.27

0.35

0.3

1.05

0.58 0.36

0.003 0.06

0.2

0.0009 0

0.004 1

2

3

0.16 4

5

0.28 6

0.033

0.32 Nitrogen Phosphorus

7

8

Figure 5. Average annual nitrogen and phosphorus retention in selected ecosystems (adapted from Mander and Mauring, 1994).

Most of the wetlands studied had a relatively low nutrient and hydraulic load. In 23 of the total of 45 wetlands, the nitrogen and hydraulic load was below 1 g N m-2 d-1 and 100 mm d-1 respectively (Mander and Mauring, 1994). In 17 wetlands out of 35, the phosphorus load was lower than 0.03 g P m-2 d-1. Three biological processes can remove nitrogen: (1) uptake and storage in vegetation; (2) microbial immobilization and storage in the soil as organic nitrogen; and (3) microbial conversion to gaseous forms of nitrogen (denitrification; Robertson and Tiedje, 1984). All of these processes, especially denitrification, can vary greatly on the spatio-temporal scale, whereas different processes can play a leading role in nitrogen removal (Kadlec and Knight, 1996; Mitsch and Gosselink, 2000). Our study shows that wetlands very efficiently remove nitrogen from inflowing polluted water. The high retention efficiency of buffer strips depends mainly on the

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

339

heterogeneity of the loading events, i.e. the higher the initial load, the higher the retention (mass removal) value. This phenomenon was first documented by Fleischer et al. (1991), and has later been repeated by many other authors. The regression between the logarithmic values of nitrogen load and removal in buffer strips is linear (Figure 6a above): y = -0.30 + 1.07 x (R 2 = 0.83, n = 45, p < 0.001),

(1)

where y is retention and x is load, both in g m-2 d-1. In our study we were unable to find examples of increasing retention capacity. a

b 1 2

y=-0.30+1.07x, R =0.83, n=45 p < 0.001

0

0.8

-1

retention/load

log retention (g N m-2 d-1)

1

-2

summer whole year wastewater inflow

-3 -4

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

0.4 2

R = 0.0013 p = 0.1

0.2

-5 -3

0.6

0

1.5

-3

log load (g N m-2 d-1)

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

log load (g N m-2 d-1)

1

1

2

y=-0.18+1.10x, R =0.94, n=35 p < 0.001

0

0.8

retention/load

log retention (g P m-2 d-1)

2

-1 -2 -3

summer whole year wastewater inflow

-4 -5

0.6 2

R = 0.0561 p = 0.15

0.4 0.2

-6 -5

-4

-3

-2

-1

log load (g P m-2 d-1)

0

1

2

0 -5

-4

-3

-2

-1

0

1

2

log load (g P m-2 d-1)

Figure 6. The load-retention relationship of nitrogen (top) and phosphorus (bottom) in wetland ecosystems. a – the correlation between load and retention, b – relative removal efficiency: retention/load (i.e., y/x of the a-part of the figure) vs. logarithmic load values. Adapted from Mander and Mauring, 1994 and Mander et al., 1997.

The storage of phosphorus in riparian buffer zones depends on the following processes: (1) soil adsorption; (2) removal of dissolved inorganic phosphorus through plant uptake and (3) microbial uptake and, in the case of peatlands (4), the incorporation of organic phosphorus into peat (Richardson, 1985). In absolute terms, soil adsorption and vegetation uptake are on a comparable level (Kadlec and Knight, 1996; Mitsch and Gosselink, 2000). In contrast to nitrogen, phosphorus can be released from wetland soils when these become saturated (Richardson and Marshall, 1986; Vanek, 1992).

340

Ü. MANDER AND K. KIMMEL

A relationship similar to nitrogen was found for phosphorus in wetlands with a close linear correlation between the log-values of load (x) and retention (y) (Figure 6a below): y = -0.18 + 1.10 x (R 2 = 0.94, n = 35, p < 0.001)

(2)

In natural conditions, retention capacity is relatively low and reaches a maximum of only 0.14 g P m-2 d-1. At higher P input (e.g. if the wetland were used for wastewater treatment), the average annual P retention was 0.1-0.35 g P m-2 d-1. At extremely high input concentrations (15.9 g P m-2 d-1), the retention can be be up to 10 g P m-2 d-1, although these systems will be saturated in the course of time (see Devito et al., 1989). The load-retention relationship in various ecosystems has been discussed in several studies. Fleischer et al. (1991) proposed that water pollution problems caused by nitrogen loading from non-point pollution sources can be solved by wetlands, because no limit had been found on the retention capacity of nitrogen in wetland ecosystems. However, the relative removal efficiency y/x for both N and P does not increase significantly when x increases (Figure 6b). The relative removal is similar to the percentage of removal ((input – output)/output*100%), which also did not demonstrate a significant correlation with initial load. However, removal percentage was higher when raw wastewater passed through the wetland. This is probably due to intensive ammonia volatilization and the higher carbon content of untreated wastewater. The presentation in the same plot (load vs. retention) of various ecosystems with very different nutrient concentration dynamics only shows the main trends and does not offer information about retention dynamics in individual wetlands. Arheimer and Wittgren (1994) demonstrated that there is a significant difference between two individual wetlands with different seasonal hydrological and hydrochemical dynamics. Therefore they suggested that the comparison of nutrient retention from different studies and the extrapolation of results from one region to another are only possible if detailed background data is available. In Estonia, many natural/semi-natural wetlands have been used for municipal or farm wastewater (Mander and Mauring, 1997). More than 40 constructed wetlands for wastewater purification have been established in the last decade. In this paper, budgets of the organic matter (BOD), total-N and total-P of eight systems are analysed (Table 2). Except for nitrogen, the efficiency of the sand/plant filter was found to be satisfactory: 82%, 36% and 74% for BOD5, total-N, and total-P respectively. The poor performance with respect to nitrogen may be caused by weak vegetation. In the Phalaris-system, 65% of organic matter, 67% of N and 80% of P was removed. The average output concentrations of this system were always lower than the recommended limits (BOD5 <10 mg O2 l-1, total-N <10 mg l-1, and total-P <2 mg l-1).

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

341

Table 2. Design and purification parameters of selected constructed wetlands for wastewater treatment in Estonia (after Öövel 2006). Name

Type

Wastewater type

Koopsi

Semi-natural wet meadow slope with Phalaris arundinacea FSW channel with helophytes (bioditch) VSSF sand/plant filter Hybrid system (VSSF+HSSF) FSW (cascade of macrophyte ponds) Hybrid system (VSSF+HSSF+FSW) Hybrid system (VSSF+HSSF) Semi-natural floodplain

Effluent from sedimentation pond

Rakke Põlva Kodijärve Põltsamaa Kõo Paistu Tapa 1

Construct -ed year 1989

Area (m2) 2400

Loading (PE1) 500

Effluent from sedimentation pond Effluent from septic tank Effluent from septic tank

1989

140

190

1994 1996

180 331

40 60

Effluent from activated sludge plant Effluent from septic tank

1997

12000

6670

2001

1200

250

Effluent from septic tank

2002

432

64

Effluent from activated sludge plant

2002

651

100000

– population equivalent, 1 PE = 60 g BOD7 d-1. FSW – free surface water VSSF – vertical subsurface flow HSSF – horizontal subsurface flow

Due to high input load, the BOD5, total-N, and total-P values in the outlet of the bio-ditch were high and extremely variable: 5-100, 6-16, and 1-4 mg l-1 respectively (see Table 3). However, outflow values from Põltsamaa and Kodijärve do not fulfil the requirements for treated wastewater quality in Estonia and most EU countries. The best performance was demonstarted by the hybrid constructed wetlands that consisted of various parts, e.g., a combination of VSSF + HSSF + (FSW) constructed wetlands. For instance, the Kõo hybrid system, consisting of 3 steps (VSSF + HSSF + FSW), which was constructed for cattail production for building material, and the lightweight aggregate (LWA)-filled Paistu system (VSSF + HSSF) showed outflow N and P concentrations of <20 mg l-1 and < 2 mg l-1 respectively (Table 3). Thus both constructed wetlands and riparian buffer zones are effective ecotechnological measures to control fluxes of nutrients at watershed level (Kuusemets and Mander, 1999). Riparian buffer zones and buffer strips In the agricultural areas of Estonia, the preferable land-use alternative was a perennial grassland (buffer zone) in combination with a forest or bush buffer strip directly on riverbanks or lake shores (Mander, 1995). New laws on nature conservation and environmental protection in Estonia stress the importance of riparian buffer zones and strips, and provide jurisdiction for the buffer zones of protected areas and all natural/semi-natural ecosystems outside protected areas. For instance, the Act on the Protection of Coastal Areas (1994) and the Act on the

342

Ü. MANDER AND K. KIMMEL

Protection of Surface Waters (1994) state that all building activities are prohibited within 200 m from the coastline and lakeshores (see Mander, 1995). Table 3. Inflow and outflow concentrations, purification efficiency and mass removal in selected constructed wetlands for wastewater treatment in Estonia (average ± standard deviation). For explanation, see Table 2. After Öövel 2006. Constructed wetland Koopsi

Rakke

Põlva

Kodijärve

Põltsamaa

Kõo Paistu

Parameter -1

Inflow (mg l ) Outflow (mg l-1) Efficiency (%) Mass removal (g m-2 d-1) Inflow (mg l-1) Outflow (mg l-1) Efficiency (%) Mass removal (g m-2 d-1) Inflow (mg l-1) Outflow (mg l-1) Efficiency (%) Mass removal (g m-2 d-1) Inflow (mg l-1) Outflow (mg l-1) Efficiency (%) Mass removal (g m-2 d-1) Inflow (mg l-1) Outflow (mg l-1) Efficiency (%) Mass removal (g m-2 d-1) Inflow (mg l-1) Outflow (mg l-1) Efficiency (%) Inflow (mg l-1) Outflow (mg l-1) Efficiency (%) Mass removal (g m-2 d-1)

BOD7 17.0±17.6 4.0±4.0 65±21 1.7±0.6 136±112 22±18 81±9 3.5±2.7 173±114 28±32 82±12 2.1±1.7 124.9±57.6 13.4±15.2 89.0±12.8 1.6±1.5 74.5±84.2 39.2±72.9 51.9±85.7 2.5±4.2 141±111.6 17.4±29.5 87.9±10.9 91.8±46.9 5.5±5.9 90.8±13.1 1.53±1.28

Total N 16.0±5.8 5.0±2.5 67±17 0.7±0.23 17.8±10.8 5.4±4.0 66±12 2.7±2.0 40.5±10.6 24.8±5.9 36±14 1.0±1.2 96.5±29.6 46.2±15.8 52.1±19.0 1.2±1.0 22.4±14.6 15.4±9.7 24.3±43.4 0.5±1.1 50.9±31.8 17.9±20.9 65.5±24.4 64.3±30.1 19.2±6.7 62.8±21.6 0.48±0.42

Total P 4.1±2.7 0.7±0.4 80±12 0.5±0.11 5.2±3.6 1.4±1.3 69±19 1.6±1.2 10.9±4.2 2.6±2.0 74±15 0.4±0.11 13.9±4.3 3.4±1.7 75.2±18.7 0.2±0.2 4.9±2.4 4.4±2.2 -1.4±58.3 0.1±0.2 7.04±4.39 2.03±2.89 72.3±24.6 4.4±2.2 0.4±0.3 88.6±11.3 0.06±0.04

In some countries the complex structure of buffer zones is officially recommended or legislatively stated. In the USA the recommended complex buffer zone consists of three parts that are perpendicular to the stream bank or lake shore (sequentially from agricultural field to water body): a grass strip, a young (managed) forest strip and an old (unmanaged) forest strip (Lowrance et al., 1997; see also Figure 3). Riparian buffer zones are typical elements of territorial ecological networks, playing an important compensating role in all of the hierarchy levels listed above. Also, their functions coincide with the idea of ecological networks. Riparian biotopes generally have the following essential functions: (1) filtering polluted overland and subsurface flow from intensively managed adjacent agricultural fields; (2) protecting banks of water bodies against erosion; (3) filtering polluted air, especially from local sources (e.g., large farm complexes, agrochemically treated fields); (4) avoiding intensive growth of aquatic macrophytes by canopy shading; (5) improving the microclimate in adjacent fields; (6) creating new habitats in

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

343

land/inland water ecotones; and (7) creating more connectivity in landscapes due to migration corridors and stepping stones (Mander et al., 1997). Some of the leading regulatory functions of riparian buffers are highlighted below. - Filtering of polluted overland and subsurface flow from intensively managed adjacent agricultural fields This is the key function of buffer zones and strips. As land-water ecotones, the riparian areas can effectively control nutrient fluxes from adjacent agriculturally used areas (Lowrance et al., 1984; Vought et al., 1994). Their importance has been shown by sources with high pollution loads such as feedlots or manure-accumulating sites (Doyle et al., 1977; Young et al., 1980; Dillaha et al., 1989). Also, riparian buffers can effectively retain sediments from overland flow (Cooper et al., 1987; Magette et al., 1989). It is more complicated to assess the role of buffers in the removal of soluble nutrients. However, several case studies worldwide suggest that different riparian ecosystems can significantly decrease the nitrogen and phosphorus concentration both in overland flow and in groundwater (Peterjohn and Correll, 1984; Knauer and Mander, 1989; Jordan et al., 1992; Osborne et al., 1993; Vought et al., 1994). In riparian wetland ecosystems (both in forests and meadows), the denitrification process has been reported to be the most significant factor in nitrogen removal (Groffman et al., 1991; Haycock and Pinay, 1993; Pinay et al., 1993). Phosphorus, in contrast, can be released from the wetland soils of the riparian zone (Vanek, 1992). However, forest buffer strips on stream banks between sedge fens (or wet meadows) and streams can remove the released phosphorus (Kuusemets et al., 2001). Therefore a combination of grasslands (resp. wet meadows) as buffer zones and bush communities as buffer strips immediately on stream banks is the most ideal structure of riparian buffer communities (see also Lowrance et al., 1997). Nonetheless, it is the same structure as that which naturally forms on floodplains and stream banks. Similarly to wetland ecosystems, uptake and storage in vegetation, microbial immobilization and storage in the soil as organic nitrogen; and denitrification can remove nitrogen. The denitrification value varies greatly (<1-1600 kg ha-1 yr-1) in riparian buffers, but vegetation uptake, especially in riparian meadows, also displays considerable variation (<10-350 kg ha-1 yr-1) (Mander et al., 1997). The storage of phosphorus in riparian buffer zones depends on soil adsorption, the removal of dissolved inorganic phosphorus by plant uptake, and microbial uptake. In absolute terms, soil adsorption and vegetation uptake are on a comparable level, varying from 0.1 to 236 and from 0.2 to 50 kg P ha-1 yr-1 respectively (Mander et al., 1997). As in wetlands, phosphorus can be released from the soils of riparian zones (Vanek, 1992). Table 2 presents the most important processes controlling phosphorus retention in riparian buffers. In general, vegetated buffers along river banks and lake shores will effectively turn the lateral nutrient inflow to internal cycling and release minimal amounts of inputted nutrients. The more complex (consisting of various communities) the buffer is, the more effective the internal cycling. The main release of nitrogen will be in gaseous form due to denitrification (Groffman et al., 1991) (Figure 7). Younger

344

Ü. MANDER AND K. KIMMEL

successional stages (bushes, pioneer communities) and managed ecosystems show a greater capacity to retain nutrients. Several case studies worldwide suggest that different riparian ecosystems can significantly decrease the nitrogen and phosphorus concentration in both overland flow and groundwater (Peterjohn and Correll, 1984; Knauer and Mander, 1989; Jordan et al., 1992; Osborne et al., 1993; Vought et al., 1994). There are three important phenomena regarding retention processes in riparian buffers: (1) the non-linear character of the retention process on a spatial scale (i.e., more materials will be filtered and/or transformed in the parts of the buffer located closer to the source); (2) load-retention relationship (see Wetlands in this sub-chapter); (3) saturation of buffers (i.e., the release of material when initial load decreases).

Figure 7. Schematic view of the buffering of material fluxes in riparian ecosystems. A without buffer, B - meadow buffer, C - forest buffer, D - complex buffer (meadow+forest). Adapted from Mander and Mauring, 1994.

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

345

The non- linear character of purification processes in riparian buffers has been reported by many authors (Figure 8). This is comparable to the approach for the modelling of purification processes in constructed wetlands (Kadlec and Knight, 1996). The load-retention relationship has been documented in several studies on natural buffer strips (Knauer and Mander, 1989), and has also been demonstrated in experimental plots (Dillaha et al., 1988; Magette et al., 1989). In some research and experiments on transects through different buffer strips, both overland and subsurface flow was analyzed, but most research has dealt only with overland flow. The regression between the logarithmic values of nitrogen load and removal in buffer strips is linear (Figure 9a above): y = -0.19 + 0.93 x (R2 = 0.99, n = 25, p < 0.001) a

b 100

1000

80 100

NO3 -N (%)

Total N (mg l -1)

(3)

10

60 40 no groundwater dilution

possible groundwater dilution

20 1 0

5

10

15

20

25

0

30

0

Distance from the feedlot - riparian forest edge (m)

10 15 20 25 30 35 40 45 50 55 60 65 70 Riparian zone width (m)

c

d

100

600

80

500 -1

BOD5 (mg O 2 l )

PO 4-P (mg l-1)

5

60 40 20

400 300 200 100

0

0

0

5

10

15

20

0

5

10

Distance from the field - forest edge (m)

15

20

Distance from the field - forest edge (m)

e 100 Intensively fertilised field

Black alder forest on lakeshore Lake

μg l -1

10

1

0.1 0

40 Pb1

Pb2

80 Cu1

120 Distance (m)160 Cu2

Cd1

Cd2

Figure 8. The non-linear character of purification efficiency of total N (a), NO3-N (b), PO4-P (c), BOD (d), and heavy metals (e) in riparian buffer zones. Adapted from: a – Doyle et al., 1977; b – Vought et al., 1994; c – Mander, 1991; d – Mander, 1991, e – Knauer and Mander, 1990.

346

Ü. MANDER AND K. KIMMEL

There are relevant differences between the removals of different nitrogen forms in buffer strips. The proportion of organic nitrogen typically increases with low nitrogen inputs in overland flow and soil water passing through the buffer strip (Peterjohn and Correll, 1984). At the same time, the concentration of total nitrogen decreases. In the case of very high inputs of organic nitrogen and ammonia, nitrification processes cause an increase in nitrate nitrogen output (Dillaha et al., 1988). The proportion of nitrate nitrogen also increases when the overland flow passes through a buffer strip with leguminous plants, which can fix the atmospheric nitrogen (Young et al., 1980). A relationship similar to that of nitrogen was registered for phosphorus in buffer strips with a close linear correlation between the log-values of load (x) and retention (y): y = -0.19 + 0.93 x (R2 = 0.98, n = 11, p < 0.005) b

a

1.2 2

y=-0.19+0.93x, R =0.96, n=25 p < 0.001

1

0 -0.5 -1

summer whole year wastewater inflow

-1.5 -2 -2.5 -3 -2.5

retention/load

-1

0.5

log retention (g N m d )

1

-2

1.5

0.8 0.6 0.4 0.2

2

R = 0.1106 p = 0.07

0 -1.5

-0.5

0.5 -2

1.5

-2.5

2.5

-1.5

0

summer whole year wastewater inflow

-1 -1.5

-0.5

1.5

0.8

-0.5

-1

0.5

1

2

y=-0.19+0.93x, R =0.98, n=11 p < 0.005

0

0.5

log load (g P m-2 d-1)

1

1.5

retention/load

log retention (g P m -2 d-1)

1

-2 -1.5

-0.5

log load (g N m-2 d-1)

-1

log load (g N m d )

0.5

(4)

0.6 0.4 2

R = 0.1711 p = 0.07

0.2 0 -1.5

-1

-0.5

0

0.5

1

1.5

log load (g P m-2 d-1 )

Figure 9. Load - retention relationship of nitrogen (top) and phosphorus (bottom) in riparian buffer ecosystems. a – the correlation between load and retention, b – relative removal efficiency: retention/load (i.e., y/x of the a-part of the figure) vs. logarithmic load values. Adapted from Mander & Mauring, 1994 and Mander et al., 1997.

Equation 4 was calculated mainly on the basis of results from the vegetation period (Figure 9a below). However, the data from short-term events during the summer with an extremely high input load demonstrate the continuing high capacity for the retention of phosphorus (Magette et al., 1989). Some of the research mentioned in the literature describes the share of the groundwater transport of phosphate phosphorus in output from buffer strips (Peterjohn and Correll, 1984; Knauer and Mander, 1989). It demonstrates that phosphorus can be absorbed more intensively from overland flow than from subsurface water.

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

347

Significant hydraulic load seems to influence retention capacity to a very small degree. High retention values were registered by both high and low hydraulic loads. In buffer strips the influence of hydraulic load was more significant than in wetlands (Mander and Mauring, 1994). On the other hand, nitrogen retention in buffer strips was more significantly influenced by hydraulic load than phosphorus retention. Nevertheless, increasing hydraulic load caused a slight decrease in nutrient retention both in wetlands and buffer strips. The relative removal efficiency y/x for both N and P decreases as x increases. Figure 9b presents the relative removal efficiency curves for N and P. Although the correlation between y/x and load (log x) is not high, the plots clearly demonstrate a decreasing trend in y/x values (see discussion in Wetlands, this sub-chapter). The load-retention relationship in various ecosystems has been discussed in earlier studies (Fleisher et al., 1991; Haycock and Pinay, 1993; Mander et al., 1997). The results show a strong positive correlation between nutrient load and removal. However, buffer zones have upper limits of purification, and these regression formulae cannot be used in planning in the case of high input values. This analysis provides limits for buffer strips as water purification systems. The input concentration range (1.0-5.0 and 0.01-0.15 mg l-1 of N and P, respectively) can be considered to represent natural conditions of the studied buffer strips where water output quality also depends on the natural processes taking place in the buffer and on the increased nutrient content in soils. This can explain the certain increase of total-N and total-P content in groundwater inside both studied buffers. However, the highest increase of N and P content in soil took place in the grey alder (Alnus incana) community, while the nutrient content in groundwater is decreasing. This indicates that the increased N and P content in soil do not directly affect water quality. The increased nutrient content in the soil in the A. incana community can be explained by the very high nutrient content in alder litter. In order to achieve good purification ability, it is important to design a complex buffer zone consisting of different ecosystem strips. The average outflow values from both buffer zones were lower than 3.1 and 0.07 mg l-1 of total-N and total-P in an old (>40 yr) grey-alder forest and 1.5 and 0.06 mg l-1 in an 15-yr-old grey alder stand respectively (Mander et al., 1997). In order to analyse the relation between purification efficiency and input concentration, we calculated the purification efficiency separately for each sampling day and for every buffer strip between two sampling points. The probability of negative removal was calculated by dividing the number of negative removal cases with the total measurement number (%). The comparison of removal efficiency and input concentration shows that the removal of total-N was negative (p<0.01) when the input was less than 1.0 mg l-1 (Figure 10a). For input concentrations between 1.0 and 5.0 mg l-1, removal of total-N showed no significant positive or negative tendency, although positive removal was more common – the probability of negative removal is 30.8 to 40.0% (p>0.05). For input concentrations greater than 5.0 mg l-1, purification efficiency is significantly positive (p<0.01), and for input concentrations greater than 42 mg l-1, purification efficiency is always positive (p<0.001). The relation between

348

Ü. MANDER AND K. KIMMEL

input concentration and the appearance of the negative removal of nitrogen (Nneg) is described by the following logarithmic regression: Nneg = 63.0 –16.84ln(INmax) R2 = 0.78, p < 0.001

(5)

where INmax is the maximum value of the N input class. - Shading effect Large quantities of aquatic plants in watercourses reduce water flow and increase the risk of water flooding adjacent land. Intensive macrophyte growth causes silting of the stream and this, in turn, aggravates the problem. Therefore, according to a law in most countries of the temperate zones, streams and rivers are primarily managed for land drainage and the reduction of flood risk. Ecologically based reasons are typically not considered. The mechanical removal of aquatic macrophytes and some of the marginal vegetation once or twice a year, combined with the occasional dredging of the channel, is the most common practice of watercourse management. This method greatly disturbs the trophic structure of the stream ecosystem and, as reported by different authors, leads to more intensive macrophyte growth than before the treatment (Böttger, 1978). a

b 100

100 80

y = 7.9-10.20ln(x) 2

R = 0.72

60 P n eg (%)

60 P n eg (%)

80

y = 7.9-10.20ln(x) 2 R = 0.72

40

40 20

20

0

0 0

1

2

3

4

-20

5

6

0

1

2

3

4

5

6

-20 -1

-1

I (mg P l )

I (mg P l )

Figure 10. The relation between input concentration (I, mg l-1) and the probability of the negative removal of (a) nitrogen and (b) phosphorus (Nneg and Pneg, %, respectively). Adapted from Kuusemets et al., 2001 and Mander et al., 2005.

Several authors have suggested that trees and bushes on stream banks can be used as a very effective control of macrophyte growth in channels (Niemann, 1963; Lohmeyer and Krause, 1975; Krause, 1977; Böttger, 1978; Binder, 1979; Dawson and Kern-Hansen, 1979; Bobrowski and Böttger, 1983). This practice is more effective in the long-term perspective and is even half as expensive than the mechanical treatment of watercourses (Krause, 1978). Likewise, Dawson and KernHansen (1979) refer to American studies that demonstrated that stream bank stability increased significantly to 85 % and 95 % after trees and bushes were planted.

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

349

The data and experience derived in Estonia completely coincide with the results of other studies. Figure 11 shows a highly significant correlation (p <0.001) between the shading rate of the stream surface (Sn; relative light intensity measured at the water surface compared to that in the open) and aquatic macrophyte biomass (B) in lowland watercourses in Southern Estonia (Mander, 1995). The ecologically recommendable level of aquatic plant biomass -2 (about 100 g DW m , shading rate about 0.2; Figure 11) is suggested because too much shade is detrimental to the aquatic fauna and can lead to an accentuated local accumulation of leaves (Dawson and Kern-Hansen, 1979). Leaves shed in the autumn can exert an oxygen demand locally (Slack and Feltz, 1968) but leaves also are an important source of food for aquatic organisms. On the other hand, the technologically acceptable level of aquatic plant biomass -2 (about 300 g DW m , shading rate about 0.5; Figure 11) means that macrophyte growth still enables water flow from drained areas and does not cause a substantial decrease in oxygen supply in critical periods (e.g. in the autumn and in the morning hours).

-2

Aquatic macrophyte biomass (B; g DW m )

700 y = 637.1x - 31.9

600

2

R = 0.885; p <0.05

500 400 Technologically acceptable level

300 200

Ecologically recommendable level

100 0 0

0.2

0.4

0.6

0.8

1

Shading rate (Sn)

Figure 11. The influence of shading (Sn) on the biomass of aquatic macrophytes (g DW m-2) in lowland ditches of agricultural landscape in Estonia (adapted from Mander et al., 2005). Biomass below an ecologically recommendable level causes disturbances in stream benthos ecosystem, particularly decreasing biodiversity. The biomass of macrophytes exceeding the technologically allowable limit creates significant obstacles to water runoff.

350

Ü. MANDER AND K. KIMMEL

This level coincides with the results obtained by other researchers (250 gDWm-2; Jorga and Weise, 1977). Simplified schemes in Figure 11 show the typical forest and bush locations on stream banks leading to respective shading rate and macrophyte growth. However, different authors recommend different tree and bush patterns on stream banks. For instance, Dawson and Kern-Hansen (1979) recommend larger trees on the northern bank of small streams (width 0-3 m, depth <1 m). Our investigations, in contrast, show that this will lead to similar intensive macrophyte growth to that in open lowland streams (Figure 11). Therefore, if the water is flowing to east or west, higher vegetation is to be recommended for the southern banks (see also Jorga et al., 1982). In any case, all investigations suggest that light should be reduced to about half of that presently available in the open (Sn = 50 %). - Filtering effect of canopies In the event of local atmospheric pollution sources, e.g., the application of fertilizers and pesticides from planes or helicopters, the filtering of pollutants by canopies plays an important role in stream protection. Figure 12 shows the results of experiments carried out by the author in Estonia in June 1982.

-2

Loading rate (g NH4CONH2 m )

6 5 4 3 2

riparian grey alder forest

Ser 1 Ser 2 Ser 3 Ser 4

1 barley field

0

-20

ditch

0

20

40

60

80

100

120

Distance (m)

Figure 12. The filtering effect of canopies of a riparian grey alder (Alnus incana) stand. The amount of urea reaching the ground by fertilising a barley field by airplane (g NH4CONH2 m- 2; the treatment rate was the same both in the field and above the riparian forest; adapted from Mander, 1995). Series according to wind direction: 1 –towards the riparian stand, 2 and 3 – parallel to the riparian stand, 4 – towards the field.

During these experiments, the carbamide (urea; NH 4 CONH 2 ) solution was applied as fertilizer on summer-barley fields and adjacent alder forest buffer strips, using planes (Mander, 1995). The mean application value (50 kg ha -1 ) was the same in both fields and adjacent buffer strips. Water samples taken from 0.9 m 2 gauges installed at different distances from the forest edge in the field and under the tree

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

351

canopies showed a significantly lower loading under the buffer strips. Although the wind direction was different, only 5-10% of the carbamide was found to reach the soil surface under the canopies in comparison with the values in the open field. Even the outwash from canopies caused by intensive rains in the following 1-2 days (2030 mm) did not exceed 20-40% of the load in the open area (Figure 12). This experiment also demonstrates a significant edge effect: higher loading values were found directly in the vicinity of the forest edge. Although leaves presumably absorbed some amount of carbamide, the experiments suggest that the filtering of atmospheric fluxes of pollutants is an essential function of forest buffer zones. Basically, the effectiveness of filtering can be estimated using similar forest stand parameters as those used in the case of the estimation of shading effectiveness. 3.1.2.3 Regulation and regeneration of populations and biocoenoses and habitat (genetic) function Wetland ecosystems support up to 15% of global biodiversity (Mitsch and Gosselink, 2000). The most important role is played by tropical and sub-tropical wetlands and floodplains, whereas isolated wetlands in surface water and/or groundwater depressions of the temperate zone are less biodiversity. However, they provide habitats with very specialized species (Moore, 2002). Wetland losses not only create many hydro-engineering and eutrophication problems in floodplains and coastal zones (see Bildstein et al., 1991; Turner, 1997), but their losses (Brinson and Malvarez, 2002) and eutrophication by both nitrogen (Vitousek and Howarth, 1991) and phosphorus (Wassen et al., 2005) cause a direct decrease in habitats and species diversity. In the following, Estonian floodplain communities are briefly characterized as an example of floodplain biodiversity. In Estonia, forests of the floodplain site type group have survived only fragmentarily, in the valleys of some larger rivers (Paal, 1998). The maintenance of natural water movements and groundwater level is required in order to protect these forests. Moreover, the protected areas must be surrounded by a buffer zone, the extent of which will depend on the general water conditions of the surrounding landscape. Floodplain forests and also minerotrophic swamp forests are the most speciesrich of the forest communities in the boreal zone. In the case of forest fires, they often serve as sanctuaries where the gene pool of many species is able to survive (Sjöberg and Ericson, 1992). On the other hand, floodplain meadows play a similarly important role in supporting biodiversity. However, the species diversity of floodplain meadows is continuously decreasing due to overgrowing by scrub and forests. All Estonian flood plain forests and meadows are considered to be maintained as nature reserves (Paal, 1998). The endangered vascular plant species of the floodplains that belong to the Red Data Book of Estonia and have almost disappeared in most parts of Europe include Ligularia sibirica, Angelica palustris, Crepis mollis, Equisetum scirpoides, Gladiolus imbricatus, Trisetum sibiricum, Gentiana pneumonanthe, Sparganium glomeratum, Carex aquatilis, Conioselinum tataricum. The first two are included in Annex II of EU Nature Directive.

352

Ü. MANDER AND K. KIMMEL

Most of the wetland bird species are not strictly confined to one specific habitat. Nevertheless, of different wetland habitats, the highest density of breeding birds (up to 240 pairs/10 ha) has been recorded in floodplain willow shrublands (Masing et al., 2000). Floodplains are important for migrating birds, including several highly endangered species of waterfowl, such as smews (Mergus , II protection category) and Bewick´s swans (Cygnus columbianus II). The characteristic, though less numerous, species of breeding birds of the floodplains also include several endangered and protected species, such as Philomachus pugnax (I), Limosa limosa (II), Gallinago media (II), Circus pygargus (III), Porzana porzana (III), Crex crex (III), Tringa totanus (III) (Kuresoo, 1996). Floodplain meadows should be considered one component among the other alluvial biotopes (alluvial forests, water bodies, reedbeds). Several protected bird species, such as Ciconia nigra (I), Aquila clanga (I), Aquila pomarina (I), Ciconia ciconia (III) depend on general condition and the supply of suitable food in the floodplain meadows (Kuresoo, 1996). Typical nesting birds of Estonian floodplain forests are Aquila clanga (I), Dendrocopus leucotus (II), and Picus canus (III) (Paal, 2004).

Figure 13. Habitat type structure and density (individuals per hectare) of Clouded Apollo (Parnassius mnemosyne) in the Ahja River valley in Southern Estonia. Adapted from Meier et al., 2005.

Several endangered butterfly species such as Parnassius mnemosyne (II, Annex IV of EU Nature Directive), Lopinga achine (II, Annex IV), Euphydryas aurinia (III, Annex II), and dragonflies such as Aechna viridis (III, Annex IV) Leucorrhinia

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

353

pectoralis (III, Annex IV) are found in the Estonian floodplain meadows (Pedmanson, 1996). Interestingly, the population of Parnassius mnemosyne shows a significant increasing trend (Meier et al., 2005; Figure 13), which may be a result of the improved connectedness of floodplains as elements of the ecological network, and less pollution loads in river valleys due to a significant decrease of fertilization since the beginning of the 1990s (Mander and Palang, 1999). 3.1.3 Social functions Although very little information is available in academic publications concerning the social (psychological, aesthetic and ethical, informative, cultural-historical, human ecological, recreational) functions of wetlands and riparian buffers, local and regional interest in cultural heritage and other aspects is significantly increasing (see Smardon, 2006). Due to Ramsar activities, the worldwide development of (eco)tourism and overall globalization, the wetlands are acquiring greater importance at a global level (see Kerstetter et al., 2004). 4. CONCLUSIONS Both wetlands and riparian buffer zones are important multifunctional elements in landscapes, providing a variety of services for society. Among these functions, the regulation of energy and material fluxes, as well as habitat function and biodiversity support are the most relevant, and a large variety of literature and knowledge is available about these. However, their biomass production and various social (psychological, aesthetical, cultural) functions are of rising importance. In particular, the complex process of the restoration of wetlands and riverine ecosystems requires that more attention be paid to the socio-economic aspects of both ecosystem types. ACKNOWLEDGEMENTS This study was supported by the EU 5 FP RTD project PRIMROSE (EVK1-200000728) “Process Based Integrated Management of Constructed and Riverine Wetlands for Optimal Control of Wastewater at Catchment Scale”, Estonian Science Foundation grants Nos. 5247 and 6083, and the Target Funding Projects Nos. 0180549s98 and 0182534s03 of the Ministry of Education and Science of Estonia.

REFERENCES Arheimer, B. and Wittgren, H.-B. (1994). Modeling the effects of wetlands on regional nitrogen transport. Ambio, 23, 378-386. Bastian, O. and Schreiber, K.-F. (Hrsgb.) (1994). Analyse und Ökologische Bewertung der Landschaft. Gustav Fischer Verlag, Jena, Stuttgart, 425 S. Bildstein, K.L, Bancroft, G.T., Dugan, P.J., Gordon, D.H., Erwin, R.M., Nol, E., Payne, L.X. and Senner, S.E. (1991). Approaches to the conservation and coastal wetlands in the Western Hemisphere. Wilson Bull., 103 (2), 218-254.

354

Ü. MANDER AND K. KIMMEL

Bobrowski, U. and Böttger, K. (1983). Changes in vegetation of the Schierensee brook (Nature Park Westensee, Schleswig-Holstein) caused by planting trees. Landschaft + Stadt, 15, 60-71. (In German with English summary). Böttger, K. (1978). Ökologischer Gewässerschutz eines Norddeutschen Tieflandbaches. Erste Gestaltungsmaßnahmen am Wiesenabschnitt des Unteren Schierenseebaches im Naturpark Westensee, Schleswig-Holstein. Schr. Naturw. Ver. Schlesw.-Holst., 48: 1-12. Brinson, M.M. and Malvarez, A.I. (2002). Temperate freshwater wetlands: types, status, and threats. Environ. Conserv. 29 (2), 115-133. Bullock, A. and Acreman, M. (2003). The role of wetlands in the hydrological cycle. Hydrology and Earth System Sciences, 7 (3), 358-359. Cooper, J.R., Gilliam, J.W., Daniels, R.B. and Robarge, W.P. (1987). Riparian areas as filters for agricultural sediment. Soil Sci. Soc. Am. J., 51, 416-420. Correll, D.L. 2005. Principles of planning and establishment of buffer zones. Ecological Engineering, 24 (5), 433-439. Costanza, R., d'Arge, R., de Groot, R.S., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O’Neill, R., Paruelo, J., Raskins, R., Sutton, B. and van den Belt, M. (1997). The value of the world's ecosystem services and natural capital. Nature, 387 (6630), 253-260. Cunnane, S.C. 2005. Survival of the Fattest: The Key to Human Brain Evolution. World Scientific Publishing Company, 368 pp. Dawson, F.H. and Kern-Hansen, U. (1979). The effect of natural and artificial shade on the macrophytes of lowland streams and the use of shade as a management technique. Int. Revue Ges. Hydrobiol. 64, 437-455. De Groot, R.S. (1992). Functions of Nature: Evaluation of Nature in Environmental Planning, Management and Decision-Making. Wolters Noordhoff BV, Groningen, the Netherlands, 345 pp. Devito, K.J., Dillon, P.J. and Lazerte, B.D. (1989). Phosphorus and nitrogen retention in five Precambrian shield wetlands. Biogeochemistry, 8, 185-204. Dillaha, T.A., Sherrard, J.H., Lee, D, Mostaghimi, S. and Shanholtz, V.O. (1988). Evaluation of vegetated filter strip as a best management practice for feed lots. J. Water Pollut. Control Fed., 60, 1231-1238. Doyle, R.C., Stanton, G.C. and Wolf, D.C. (1977). Effectiveness of Forest and Grass Buffer Strip in Improving the Water Quality of Manure-Polluted Runoff. ASAE Paper No 77-2501, St Joseph, MI, 32 pp. Fleischer, S., Stibe, L. and Leonardson, L. (1991). Restoration of wetlands as a means of reducing nitrogen transport to coastal waters. Ambio, 20, 271-272. Goetz, S.J. (2006). Remote sensing of riparian buffers: Past progress and future prospects. Journal of the American Water Resources Association, 42 (1), 133-143. Groffmann. P.M., Axelrod, E.A., Lemunyon, J.L. and Sullivan, W.M. (1991). Denitrification in grass and forest vegetated filter strips. Journal of Environmental Quality, 20, 671-674. Haycock, N. and Pinay, G. (1993). Nitrate retention in grass and poplar vegetated riparian buffer strips during the winter. Journal of Environmental Quality, 22, 273-278. Jespersen, D.N., Sorrell, B.K. and Brix, H. (1998). Growth and root oxygen release by Typha latifolia and its effects on sediment methanogenesis. Aquatic Botany, 61 (3), 165-180. Jordan, T.E., Correll, D.L. and Weller, D.E. (1992). Nutrient interception by riparian forest receiving inputs from adjacent cropland. Journal of Environmental Quality, 22, 467-473. Jorga, W. and Weise, G. (1977). Growth of submerged macrophytes in slowly running waters related to oxygen balance. Int. Revue Ges. Hydrobiol., 62, 209-234. Jorga, W., Heym, W. and Weise, G. (1982). Shading as a measure to prevent mass development of submersed macrophytes. Int. Revue Ges. Hydrobiol., 67, 271-281. Slack, K.V. and Feltz, H.R. (1968). Tree leaf control on low flow water quality in a small Virginia stream. Environ. Sci. Technol., 2, 126-131. Kadlec, R.H. and Knight, R.L. (1996). Treatment Wetlands. CRC Press/Lewis Publishers, New York, 839 p. Kerstetter, D.L., Hou, J.S. and Lin, C.H. (2004). Profiling Taiwanese ecotourists using a behavioral approach. Tourism Management, 25 (4), 491-498.

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

355

Knauer, N. and Mander, Ü. (1989). Untersuchungen über die Filterwirkung verschiedener Saumbiotope an Gewässern in Schleswig-Holstein. 1. Mitteilung: Filterung von Stickstoff und Phosphor. Z. f. Kulturtechnik und Landentwicklung, 30, 365-376. Knauer, N. und Mander, Ü. (1990). Untersuchungen über die Filterwirkung verschiedener Saumbiotope an Gewässern in Schleswig-Holstein. 2. Mitteilung: Filterung von Schwermetallen. Z.f. Kulturtechnik und Landentwicklung, 31, 52-57. Krause, A. (1977). On the effect of marginal tree rows with respect of the management of small lowland streams. Aquat. Bot., 3, 185-192. Kuresoo, A. (1996). Vertebrates of floodplain meadows. In Leibak, E., Lutsar, L. (Eds.) Estonian Coastal and Floodplain Meadows (pp. 57-60). Eestimaa Looduse Fond. Kuusemets, V. and Mander, Ü. (1999). Ecotechnological measures to control nutrient losses from catchments. Water Science and Technology, 40 (10), 195-202. Kuusemets, V., Mander, Ü., Lõhmus, K. and Ivask, M. (2001). Nitrogen and phosphorus variation in shallow groundwater and assimilation in plants in complex riparian buffer zones. Water Science and Technology, 44 (11-12), 615–622. Lohmeyer, W. and Krause, A. (1975). Über die Auswirkungen des Gehölzbewuchses an kleinen wasserläufen des Münsterlandes auf die Vegetation im Wasser und an den Böschungen im Hinblick auf die Unterhaltung der Gewässer. Schriftenreihe für Vegetationskunde 9: 1-105. Lovell, S.T. and Sullivan, W.C. (2006). Environmental benefits of conservation buffers in the United States: Evidence, promise, and open questions. Agriculture, Ecosystems and Environment, 112 (4), 249-260. Lowrance, R.R., Todd, R., Fail, J., Hendrickson, O., Leonard, R. and Asmussen, L. (l984). Riparian forest as nutrient filters in agricultural watersheds. BioScience ,34, 374-377. Lowrance, R.R., Altier, L.S., Newbold, J.D., Schnabel, L.L., Groffman, P.M., Denver, J.M., Correll, D.L., Gilliam, J.W., Robinson, J.W., Brinsfield, R.B., Staver, K.W., Lucas, W. and Todd, A.H. (1997). Water quality functions of riparian forest buffers in Chesapeake Bay watershed. Environmental Management, 21 (5), 687-712. Magette, W.L., Brinsfield, R.B., Palmer, R.E. and Wood, J.D. (1989). Nutrient and sediment removal by vegetated filter strips. Trans. ASAE, St. Joseph, Mich., 32, 663-627. Mander, Ü. (1991). Eco-engineering methods to control nutrient losses from agricultural watersheds. In Brandt, J. (Ed.) Proceedings of European Seminar of IALE on Practical Landscape Ecology (pp. 5364), Roskilde, Denmark, May, 2-4, 1991. v. IV. Mander, Ü. (1995). Riparian buffer zones and buffer strips on stream banks: Dimensioning and efficiency assessment from catchments in Estonia. In Eiseltová, M. and Biggs, J. (Eds.) Restoration of Stream Ecosystems (pp. 45-64). IWRB Publication No 37, Slimbridge, Gloucester, UK. Mander, Ü., Hayakawa, Y. and Kuusemets, V. (2005a). Purification processes, ecological functions, planning and design of riparian buffer zones in agricultural watersheds. Ecological Engineering, 24 (5), 421-432. Mander, Ü., Kuusemets, V., Lõhmus, K. and Mauring, T. (1997). Efficiency and dimensioning of riparian buffer zones in agricultural catchments. Ecological Engineering, 8: 299-324. Mander Ü., Lõhmus, K., Kuusemets, V., Ivask, M., Teiter, S. and Augustin, J. (2005b). Budgets of nitrogen and phosphorus fluxes in riparian grey alder forest. In Vymazal, J. (Ed.) Natural and Constructed Wetlands: Nutrients, Metals and Management (pp. 1-19). Backhuys Publishers, Leiden. Mander, Ü. and Mauring, T. (1994). Nitrogen and phosphorus retention in natural ecosystems. In Ryszkowski, L. and Balazy, S. (Eds.) Functional Appraisal of Agricultural Landscape in Europe (EUROMAB and INTECOL Seminar 1992, pp. 77-94). Research Centre for Agricultural and Forest Environment, Pol. Acad. Sci. Poznan. Mander, Ü. and Mauring, T. (1997). Constructed wetlands for wastewater treatment in Estonia. Water Science and Technology, 35, 323-330. Mander, Ü. and Palang, H. (1999). Landscape changes in Estonia: reasons, processes, and consequences. In R. Krönert, J. Baudry, I.R. Bowler and A. Reenberg (Eds.), Land-Use Changes and Their Environmental Impact in Rural Areas in Europe (pp. 165-187). MAB Series, Vol. 24. The Parthenon Publishing Group, Paris. Mander, Ü., Strandberg, M., Mauring, T. and Remm, K. (2001). Wetlands as essential basis for sustainable development: Estonian case. In Villacampa, Y., Brebbia, C.A. and Uso, J.-L. (Eds.) Ecosystems and Sustainable Development III (pp. 459-467), WIT Press, Southampton and Boston.

356

Ü. MANDER AND K. KIMMEL

Masing, V., Paal, J. and Kuresoo, A. (2000). Biodiversity of Estonian wetlands. In Gopal, B., Junk W.J. and Davis, J.A. (Eds.), Biodiversity in Wetlands: Assessment, Function and Conservation. Vol 1 (pp. 259-279). Backhuys Publishers, Leiden, The Netherlands. Mauring, T. (2003). The use of reed and cattail produced in constructed wetlands as building material. In: Mander, Ü., Vohla, C. and Poom, A. (Eds.) Constructed and Riverine Wetlands for Optimal Control of Wastewater at Catchment Scale. Publicationes Instituti Geographici Universitatis Tartuensis, 94, pp. 286-288. Meier, K., Kuusemets, V., Luig, J., and Mander, Ü. (2005). Riparian buffer zones as elements of ecological networks: Case study on Parnassius mnemosyne distribution in Estonia. Ecological Engineering, 24 (5), 531-537. Mitsch, W.J. and Gosselink, J.G. (2000). Wetlands, 3rd Edition. John Wiley and Sons, New York, 936 p. Moore, P.D. (2002). The future of cool temperature bogs. Environmental Conservation, 29 (1), 3-20. Naiman, R.J., Decamps, H. and McClain, M.E. (2005). Riparia: Ecology, Conservation, and Management of Streamside Communities. Academic Press, London, 448 pp. Niemann, E. (1963). Die natürliche Ufervegetation in ihrer Bedeutung für Uferbepflanzung und ingenieurbiologische Maßnahmen. Z. Landeskultur, 4, 187-206. Öövel, M. (2006). Performance of Wastewater Treatment Wetlands in Estonia. PhD Thesis. University of Tartu Press, 148 pp. Osborne, L.L. and Kovacic, D.A. (1993). Riparian vegetated buffer strips in water-quality restoration and stream management. Freshwater Biology, 29: 243-258. Paal, J. (1998). Rare and threatened plant communities of Estonia. Biodiversity and Conservation, 7, 1027-1049. Paal, J. (2004). Estonia’s Habitats Valuable for Europe Euroopas Väärtustatud Elupaigad Eestis. Ministry of Environment. (In Estonian). Pedmanson, R. (1996). Invertebrates of floodplain meadows. In Leibak, E. and Lutsar, L. (Eds.) Estonian Coastal and Floodplain Meadows (pp. 61-64). Eestimaa Looduse Fond. Peterjohn, W.T. and Correll, D.L. (1984). Nutrient dynamics in an agricultural watershed: observations on the role of a riparian forest. Ecology, 65, 1466-1475. Pinay, G., Roques, L. and Fabre, A. (1993). Spatial and temporal patterns of denitrification in a riparian forest. Journal of Applied Ecology, 30, 581-591. Richardson, C.J. (1985). Mechanisms controlling phosphorus retention capacity in freshwater wetlands. Science, 228 (4706): 1424-1427. Richardson, C.J. and Marshall, P.E. (1986). Processes controlling movement, storage, and export of phosphorus in a fen peatland. Ecological Monographs, 56 (4), 279-302. Robertson, G.P. and Tiedje, J.M. (1984). Denitrification and nitrous oxide production in successional and old-growth Michigan forests. Soil Sci. Soc. Am. J., 48 (2). 383-389. Sjöberg, K. and Ericson, L. (1992). Forested and open wetland complexes. In Hansson, L. (Ed.) Ecological Principles of Nature Conservation. Applications in Temperate and Boreal Environments (pp. 326-351). Elsevier, London, New York. Smardon, R.C. (2006). Heritage values and functions of wetlands in Southern Mexico. Landscape and Urban Planning, 74 (3-4), 296-312. Turner, R.E. (1997). Wetland loss in the northern Gulf of Mexico: Multiple working hypotheses. Estuaries, 20 (1), 1-13. Vanek, V. (1991). Riparian zone as source of phosphorus for a groundwater-dominated lake. Water Research, 25, 409- 418. Verhoeven, J.T.A., Arheimer, B, Yin, C.Q. and Hefting, M.M. (2006). Regional and global concerns over wetlands and water quality. Trends in Ecology and Evolution, 21 (2), 96-103. Vitousek, P.M. and Howarth, R.W. (1991). Nitrogen limitation on land and sea: how can it occur? Biogeochemistry, 13, 87-115. Vought, L.B.-M., Dahl, J., Pedersen, C.L. and Lacoursière, J.O. (1994). Nutrient retention in riparian ecotones. Ambio, 23, 342-348. Vymazal, J. (2001). Types of Constructed Wetlands for Wastewater Treatment: Their Potential for Nutrient Removal. In Vymazal, J. (Ed.), Transformations of Nutrients in Natural and Constructed Wetlands (pp. 1-93). Backhuys Publishers, Leiden, The Netherlands. Wassen, M.J., Venterink, H.O., Lapshina, E.D. and Tanneberger, F. (2005). Endangered plants persist under phosphorus limitation. Nature, 437 (7058), 550.

WETLAND AND RIPARIAN BUFFER IN LANDSCAPE FUNCTIONING

357

Wild, U. Lenz, A. Kamp, T. Heinz, S. and Pfadenhauer, J. (2002). Vegetation development, nutrient removal and trace gas fluxes in Typha wetlands. Mander, Ü. and Jenssen, P.D. (Eds.), Natural Wetlands for Wastewater Treatment in Cold Climate. Advances in Ecological Sciences, Vol. 12 (pp. 101-125). WIT Press, Southampton, Boston. Young, R.A., Huntrods, T. and Anderson, W. (1980). Effectiveness of vegetated buffer strips in controlling pollution from feedlot runoff. Journal of Environmental Quality, 9, 483-487. Zedler, J.B. and Kercher, S. (2005). Wetland resources: Status, trends, ecosystem services, and restorability. Annual Review of Environment and Resources, 30, 39-74.

CHAPTER 21

POST-FIRE FOREST RESTORATION INDICATED BY CANOPY DENSITY IN THE NORTHERN GREAT HING’AN MOUNTAINS

F.-J. XIE, X.-Z. LI, X.-G. WANG, D.-N. XIAO Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China

Abstract. The restoration of forest landscape has drawn much attention since the catastrophic fire took place on the northern slope of Great Hing′an Mountains in 1987. Forest canopy density, which has close relation to forest productivity, was selected as a key factor to find how much the forest quality was changed 13 years after fire, and how fire severity, regeneration way and terrain factors influenced the restoration of forest canopy density, based on forest inventory data in China, using Kendall Bivariate Correlation Analysis, and Distances Correlation Analysis. The results showed that fire severity that was inversely correlated with forest canopy density grade was an initial factor among all that selected. Regeneration way which did not remarkably affect forest canopy density restoration in short period may shorten the cycle of forest succession and promote the forest productivity of conophorium in the future. Among the three terrain factors, the effect of slope was the strongest, the position on slope was the second and the aspect was the last

1. INTRODUCTION Forest fire as one of the important ecological factors in forest ecosystem, had drawn much attention long before (Garren, 1943; Weaver, 1951; Gill, 1975). Yet the study on forest fire had not developed broadly in China before 1987. The catastrophic fire in the Great Xing’an Mountains in 1987 that brought an ecocalamity provided an opportunity to study the effects of large forest fire on forest ecosystem. Although a number of research achievements have been obtained since 1988 (Xiao et al., 1988; Zhao, 1988; Li et al., 1988; Guan and Zhang, 1989; Zhou et al., 1994; Zhao et al., 1994; Yang et al., 1998), there exist many deficiencies in this field, compared with international progress (Bergeron and Brisson, 1990; Pianka, 1992; Turner, 1997; Grogan et al., 2000). First, most studies still were focused on the fire effect on one of the ecological factors such as vegetation, soil, hydrology, etc., and at fine scales. These studies lacked of 359 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 359–374. © 2007 Springer.

360

F.-J. XIE ET AL.

comprehensive research that takes ecosystem as an integer at landscape or regional scale. Second, with the restoration of forest landscape in the burned area, a lot of research work has stopped since 1995, which results in discontinuity in this field. The destructive fire on May 6, 1987 led to severe economic loss and enormous negative ecological effect. The anthropogenic disturbances after fire aggravated the complexity and uncertainty of the forest ecosystem succession (Nakagoshi, 2001; Luo, 2002; Wang and Li, 2003; Kong et al., 2003). Therefore the research on restoration of forest in the burned area especially on the function restoration at coarse scale should be continued. Forest productivity is an elementary index of the forest ecosystem function that can evaluate how the forest grows (Liu and Fu, 2001). The forest canopy density that is one of the important indexes reflecting forest productivity as well as a key statistic data in forest inventory in China was selected as the research object in this paper. The main aim of this study is to reveal the relationship between the restoration of forest function and influential factors using forest spatial data in 1987 and 2000. Considering the interaction between anthropogenic and natural factors, the regeneration ways (which were sorted as natural regeneration, artificial stimulating regeneration, direct seeding regeneration, planting regeneration) and terrain factors including slope, aspect and position were selected as influential factors in this paper besides fire severity, which was classified as lightly, moderately and severely burned. 2. STUDY AREA AND METHODS 2.1 Study area description Yuying and Fendou forest farms in the middle of Tuqiang Forestry Bureau on the northern slope of Great Hing′an Mountains were chosen as study area, which occupies 1200 km2 land area in total. Several reasons were taken into account for choosing the two forest farms as study area. First, 87.5% of the area was burned, which is convenient for the study of forest restoration at large scale. Second, topographical relief in this area is relatively steep, which is advantageous for the study of effects of terrain factors on forest restoration. Third, the natural regeneration assisted by human promotion and planting measures was taken after fire, which provides a chance to study the effect of human interference on post-fire forest restoration. Tuqiang Forest Bureau lies in the northwest of Great Hing′an Mountains, at upper reaches of Heilong River, and belongs to Mohe County, Heilongjiang Province. Geographic coordinates are between longitude 122°18′05″–123°29′00″ E and latitude 52°15′55″–53°33′40″ N. The average altitude is 500m with gentle undulating hills and open river valleys. The tendency of topography shows that the south is higher and the north is lower. Moreover, east slopes are steeper than west ones. This area has long and cold winter and short and hot summer. Mean annual temperature is -4.94 °C, with lowest temperature recorded at -53 °C. Mean annual

POST-FIRE FOREST RESTORATION

361

precipitation is 432 mm, with relatively dry spring and winter, and moist summer and autumn. The main forest species are Hing’an larch (Larix gmelimii), birch (Betula platyphylla), pine (Pinus sylvestris var. mongolica) and aspen (Populus davidiana). Brown coniferous forest soil is the dominant soil genus with thickness of 10–30 cm. 2.2 Methods Main data sources include forest inventory data (forest stand maps) of Tuqiang Forestry Bureau in 1987 and 2000, from which forest canopy density data was derived. The field investigation for inventory data of 1987 was made before fire, and can be considered as pre-fire data source. Map of fire severity in 1987, management plan of Tuqiang Forestry Bureau, and field survey data were also used. Fire severity map in 1987 and forest stand maps in 2000 were scanned and digitised in ArcView3.3. Forest canopy density and various influential factors were extracted to generate distribution maps of forest canopy density grade, fire severity, regeneration way, slope grade, aspect type and slope position respectively, according to criterion listed in Table 1, and the maps can be found in Figure 1a-f. By overlaying the digital maps of fire severity, regeneration way, slope grade, aspect type and slope position type with the map of forest canopy density grade, the effect of factors above on forest restoration were analysed. Nonparametric Rank Correlation Analyses (Kendall Rank Correlation) was used to test the relationship between fire severity and area distribution of canopy density grades. Distance Correlation Analyses (Similarity Matrix) were adopted to evaluate the effects of regeneration ways and terrain factors on the restoration of forest canopy density. In order to normalize the area difference between various fire severities, regeneration ways, gradients of terrain factors, and distribution of forest canopy density grades. We introduced relative area percentage expressed as: P = Aij / Ai where, Aij is the area of the grade j in type or gradient i; Ai is the total area of type or gradient i. Data analyses were performed using SPSS10.0 and EXCELL. 3. RESULTS AND DISCUSSION 3.1 Changes of forest canopy density grades Two dominant forest canopy density grades appeared in both 1987 (pre-fire) and 2000 (Figure 2). Grade 1 and grade 4 were the main grades pre-fire, while grade 1 and grade 3 were the main grades in 2000. Although grade 1 was the

362

F.-J. XIE ET AL.

dominant one all the way, its area percentage reduced from 32.1% in 1987 (prefire) to 23.7% in 2000. This situation may be related to the strategies of afforestation instead of harvest in suitable areas post-fire, which reduced the area of non-stocked. Another dominant grade dropped from grade 4 to grade 3, and area percentage increased from 7.6% to 39.6%, which showed that the forest was at the junior succession stage. Compared to those pre-fire, area percentage of higher canopy density grades decreased distinctively, for example, area percentages of grade 5 and grade 6 decreased from 18.8% and 15.0% pre-fire to 8.1% and 0.1% in 2000, respectively. All the results showed that the level of forest canopy density decreased compared to that of pre-fire, but the forest matrix had already formed, which can conduce to the further restoration of forest quality. Table 1. Criterion of various factors.

Factors

Grade（type） 1

Canopy density

2 3 4 5 6 Lightly burned

Fire severity Moderately burned Severely burned Natural regeneration Regeneration way

Artificial stimulating regeneration Direct seeding regeneration Planting regeneration

Attribute describe Non-stocked land or young growth and woodland which have not reached the statistic criterion Canopy density 0–0.2 (including 0.2) Canopy density 0.2–0.4 (including 0.4) Canopy density 0.4–0.6 (including 0.6) Canopy density 0.6–0.8 Canopy density ≥ 0.8 Percentage of trees consumed by fire ≤ 30% Percentage of trees consumed by fire 30%–70% Percentage of trees consumed by fire ≥ 70% The generation completely depends on seed trees without any human measures Seeds come from seed trees, but rooting is assisted by wiping the litter away Artificially or aerially seeding Directly plant young coniferous seedlings (Larix gmelimii or Pinus sylvestris var. mongolica ) (Cont.)

POST-FIRE FOREST RESTORATION

363

Table 1. (cont.)

Slope

Flats Gentle slope Moderate slope Steep slope Sharp slope

≤ 5° 5°–15° 15°–25° 25°–35° ≥ 35°

Aspect

Shaded aspect Semi-shaded aspect Semi-sunny aspect Sunny aspect No aspect

North and northeast slope Northwest and east slope West and southeast slope Southwest and south slope Dale and flat

Valley

Area with gradients less than 5° and at the bottom of the catena Various aspects from piedmont to 1/3 height of mountains whose altitudes are less than 900m Various aspects from 1/3 to 2/3 height of mountains whose altitudes are less than 900m Various aspects from 2/3 height to peak of mountains whose altitudes are less than 900m Area on top of mountains and higher than 900m

Position Low slope position Middle slope position Upper slope position Hill top

3.2 Fire severity and forest canopy density The area percentages of forest canopy density grades in different fire severity areas in 2000 have changed distinctly compared with those in unburned area (Figure 3). Among the three burn severities (lightly burned, moderately burned, severely burned), the curve of canopy density grades in lightly burned area resembles the most to that in unburned area, while those in moderately and severely burned areas were more similar to each other, with notable difference from that in the unburned area. In lightly burned area, the dominant grades were grade 3 and 4, occupying 68.9% areas; yet in unburned area grade 4 and 5 were the dominant grades, occupying 78.5% of the area, while the area percentage of grade 1 increased by 11.4% compared with that in the unburned area, which indicated that the forest productivity in lightly burned area generally had already declined, and the area of non-stocked land or young growth below statistic criterion increased. The difference between maximum and minimum area percentage of forest canopy density grades was 53.9% in moderately burned area and 52.5% in severely burned area, which was higher than that (41.6%) in unburned area. The area percentage of grade 1 in moderately and severely burned area was also higher than that in the unburned area. Although the area percentage of dominant grade (grade 3) in

364

F.-J. XIE ET AL.

moderately and severely burned area was above 50%, which is much higher than that in unburned area (9.9%), the area percentages of grade 4 and 5 were much lower than that in the unburned area. Such canopy density grade composition showed that the productivity of forest post-fire in moderately and severely burned area had greatly declined, and the difference of forest productivity had augmented. This can be explained as follows: Firstly, because higher canopy density grades often came from the survived trees after fire, and fire severity was closely related with tree survival rate, the loss was more in moderately and severely burned area than that in lightly burned area. Secondly, the forest regenerated after fire at the same time almost had reached the third grade level. Along with the area percentage of this grade increasing greatly, the difference of area percentage of forest canopy density among grades had expanded remarkably.

area percent (%)

Figure 1. Distribution of forest canopy density types and some influential factors on forest restoration. a. Forest canopy density grades; b. Fire severity grades in 1987; c. Regeneration ways; d. Slope grades; e. aspect types; f. Position types. 50

pre-fire(1987)

40

in 2000

30 20 10 0 1

2

3

4

5

6

Forest canopy density grade

Figure 2. Area percentage of forest canopy density grades in 1987 and 2000.

POST-FIRE FOREST RESTORATION

365

Area percent (%)

60 50 40 30 20 10 0 1

2

3

4

5

6

Forest canopy density grade Lightly burned area

moderatelyburned area

Severely burned area

Unburned area

Figure 3. Area percentage of forest canopy density grades in different fire severities area in 2000.

Finally, in severely burned area, seed dispersal and recolonization was difficult, resulting in the increase of non-stocked land. It is obvious that there exists a correlation between the area percentage of canopy density grade and fire severity. We adopted Kendall Rank Correlation method to test further correlation. The result revealed a moderate significant correlation between fire severity and forest canopy density grade in 2000 (tau-b=-0.422, p<0.01), which means fire severity had generated a negative influence on forest productivity. The results above can be explained from the following aspects: First, the more severe the fire was, the more forest productivity was declined, and the more recovery time was required. As a result, there exists a negative correlation between fire severity and canopy density grade. Second, considering the severe breakage of this catastrophic fire, restoration strategies including natural and artificial regeneration ways were taken, especially in severely burned areas, where artificial regeneration way was completely adopted, and thus reduced the area percentages of coverage grade 1 and 2, which weakened the negative correlation between fire severity and forest productivity. 3.3 Regeneration way and forest canopy density Figure 4 indicated that the relationship between area percentage of forest canopy density grades with regeneration way and that with fire severity, which can be attributed to post-fire management program. The natural regeneration was adopted in lightly burned area, areas of broadleaved under various fire severities and severely burned area with poor site condition, while human stimulating regeneration was introduced in moderately burned area, and severely burned area with good enough site condition. Distance Correlation Analysis revealed that the similarity between forest canopy density grades of post-fire regeneration ways and that of unburned area was very low. However, the similarity among various regeneration ways was significantly high. The

366

F.-J. XIE ET AL.

highest similarity to unburned area was natural regeneration, but the value was only 0.389 (Table 2). It seemed that artificial measures did not have effect on forest restoration. But this was not true. 1) Regeneration measures were prescribed according to fire severity and site condition. The area of natural regeneration mainly involved lightly burned area and broadleaved area, where a great number of trees survived, and broadleaved species have better germination ability. On the contrary, artificial measures were mostly used in severely burned area. Furthermore, the planted forest species was slow growth conifer. Therefore the restoration of forest productivity in natural regeneration area was better than that in artificially restored area in short term. This can also interpret that fire severity is the key factor which influences forest restoration. 2) Artificial measures can exert important effect on restoration of forest landscape pattern. The area proportion of conifer in regions of artificial regeneration (direct seeding and planting) was higher than that of natural regeneration (natural and human promoted generation) in 2000 (Figure 5), which greatly shortened the cycle of succession from broadleaved forest to conifer, especially in severely burned area (Wang, 2003), and can significantly affect the productivity of conophorium at large time scale. 3) The high similarity of productivity restoration between natural regeneration and other regeneration measures in burned areas (Table 2) indicated that the selection of regeneration way was reasonable.

Area percent (%)

60.0 50.0 40.0 30.0 20.0 10.0 0.0 1

2

3

4

5

6

Forest canopy density grade The area of natural regeneration The area of human stimulating regeneration The area of seeding regeneration The area of planting regeneration Unburned area

Figure 4. Area percentage of forest canopy density grades under different regeneration ways in 2000.

POST-FIRE FOREST RESTORATION

367

3.4 Terrain factors and forest canopy density 3.4.1 Influence of slope on forest canopy density restoration Distance analysis revealed that there was a significant correlation between area percentage of canopy density grades and slope gradients in 2000 (Table 3), indicating that slope had significant influence on canopy density restoration. The area proportion of canopy density grade 1 on flats was 43.25%, which was higher than that on slope lands, but the trend for grades 4 and 5 was the other way around (Figure 6 A). Table 2. Similarity matrix of canopy density grade in areas under regeneration ways and unburned area.

Human Direct Natural Planting stimulated seeding Unburned regeneration regeneration restoration regeneration Natural regeneration Human stimulated restoration Direct seeding regeneration Planting regeneration Unburned

1.000

0.941

0.805

0.830

0.398

1.000

0.965

0.981

0.165

1.000

1.000

0.000

1.000

0.108 1.000

Figure 5. Area percentage of forest types under different regeneration ways in 2000.

368

F.-J. XIE ET AL.

The main reason was that the proportion of swamp and bushes on flats was higher than those on slope lands according to field survey, even in the forestland. Trees grow slowly due to the excessive soil moisture. Another reason was that the proportion of burned area was higher, especially that the area proportion of severely burned on flats was the highest among all the slope grades (Figure 7 A), and limited the distribution of higher grades. Table 3. Similarity matrix of canopy density grades on slope gradients.

Flats Flats Gentle slope Moderate slope Steep slope Sharp slope

1.000

Gentle slope 0.420 1.000

Moderate slope 0.271 1.000 1.000

Steep slope 0.221 0.797 0.948

Sharp slope 0.000 0.295 0.517

1.000

0.854 1.000

Area proportion of canopy density grades 1 and 2 increased, yet that of grade 3 declined, and those of grades 4 and 5 increased too with the gradient increasing from gentle slope to steep slope (Figure 6A). Although the proportion of burned area and the fire severity decreased with the gradients increase, the proportion of canopy density grade 1 and 2 did not decrease but showed increasing tendency. On the contrary, area proportion of grade 3 decreased with gradient increase. This pattern demonstrated that the lower the gradient was, the better the forest recovered on these slopes. Slope not only influences the angle of sun incidence and results in change of air and ground temperature, but also is the driven factor for the cycle of soil water and nutrient, which influences the thickness and physicochemical property of soil (Shen et al., 2000). The more precipitous the gradient is, the more severe the soil erosion is, which leads to a poor forest restoration. The reason why area proportions of canopy density grades 4 and 5 increased with slope gradient increasing was that survival trees increased with the proportion of burned area decrease and fire severity declination (Figure 7A). The area proportion of sharp slope was terribly small with only 0.035%, and located in lightly burned area (Figure 7A), where the proportion of grade 4 was comparatively high. 3.4.2 The Influence of aspect on forest canopy density restoration Distance analysis indicated that the similarities of area percentage of canopy density grades among various aspect gradients were very high, with similarity indices all above 0.89, except no aspect land, which revealed that aspect had slight effect on restoration of forest productivity (Table 4). Due to the high proportion of swamp and shrub in no aspect land pre-fire, and global warming and anthropogenic disturbances, permafrost thawing and the expansion of forest wetland changed remarkably (Zhou

POST-FIRE FOREST RESTORATION

369

et al., 2003), the area proportion of non-stocked land was high (Figure 6B). Shaded and semi-shaded, sunny and semi-sunny aspects were more similar respectively. Area proportions of forest canopy density grade 1 to 3 in shaded and semi-shaded aspects were larger than those in sunny and semi-sunny aspects, while grade 4 to 6 presented opposite character. The reasons lay in two aspects: 1) Since proportion of burned area in sunny and semi-sunny aspects was larger than that in shaded and semi-shaded aspects (Figure 7B) and fire severity was more severe in sunny and semi-sunny aspects than that in shaded and semi-shaded aspects, area proportion of high canopy density grades was decreased, while non-stocked land was higher in sunny and semi-sunny aspects. 2) Aspect exerts effect on site condition such as water and soil property through influencing sunlight condition. Sunny and semi-sunny aspects were drier in favour of birch, aspen which adapted arid condition (Xu 1988). As a result, the area proportions of canopy density grade 3 in sunny and semi-sunny aspects land were obviously larger than those in shaded and semishaded aspects. Table 4. Similarity matrix of canopy density grade on aspect gradients.

Shaded aspect Semi-shaded aspect Semi-sunny aspect Sunny aspect No aspect

Shaded aspect 1.000

Semi-shaded Semi-sunny aspect aspect 0.993 0.944 1.000

Sunny aspect 0.940

No aspect 0.006

0.906

0.894

0.000

1.000

1.000

0.108

1.000

0.175 1.000

Table 5. Similarity matrix of canopy density grade on position gradients.

Valley Low slope position Middle slope position Upper slope position Hill top

0.491

Middle slope position 0.003

Upper slope position 0.000

1.000

0.867

0.866

0.848

1.000

1.000

0.982

1.000

0.962

Valley

Low slope position

1.000

Hill top 0.000

1.000

370

F.-J. XIE ET AL.

3.4.3 Influence of slope position on forest canopy density restoration The similarity of area percentage of forest canopy density grades between valley and other slope positions in 2000 was the lowest, while the highest similarity was found between middle slope position and upper slope position (Table 5). The largest area proportion of non-stocked land and no distribution of canopy density grade 5 and 6 were discovered in valley (Figure 6C). High soil humidity in valley is the repellent force to broadleaved species such as birch and aspen, while the growth of conifer is retarded. Besides, severe fire disturbance was another reason leading to large proportion of non-stocked land and low proportion of high grades (Figure 7C). Area proportion of canopy density grades in middle, upper slope and hill tops existed obvious uniform constitute: grades 3, 4 and 5 were the preponderant grades occupying 84.5%, 90.8% and 94.5% respectively (Figure 6C). Good drainage condition and relatively dry soil in those positions were suitable for fast growing tree species such as birch and aspen, which promoted forest restoration. Forest canopy density which has close relationship with forest productivity is an important index in forest inventory in China. Since the method adopted was qualitative estimation, this important index was often misunderstood for the lack of precision. In fact, most foresters have rich experience, and the data came from field survey. Therefore the results should be credible and valuable. 4. CONCLUSION The forest canopy density that is one of the important elements of forest function has decreased notably. The forest canopy density matrix was low canopy density grades (grades 1, 2 and 3), while area percentage of high grades decreased a lot. Restoration of forest canopy density thoroughly will need much time. Fire severity was the initial influence factor on forest restoration, which determined the number of survival trees and seeding source. Even if seeding source can be complemented through artificial regeneration, survival tree rate still was the main factor that determined the area proportion of trees in good growth condition in a short time after fire. It was difficult to recover forest under poor site condition after severe fire. This kind of area can be changed into wetland or shrubs, resulting in the increase of proportion of non-stocked land in severely burned area. Though lightly burned area was completely naturally restored, the restoration status of forest was better than that in the moderately and severely burned area, where artificial regeneration measures were employed. Thus, fire severity was the key factor for the restoration of forest productivity. Regeneration ways were human intervening factors in order to attain better restoration. The selection of regeneration ways was based on fire severity, site condition, and pre-fire species. Artificial regeneration supplied the seeding source, shortened the period from broadleaved to conifer, and restrained the degradation of

POST-FIRE FOREST RESTORATION

371

100

A

Area percent (%)

80

Flats

60

Gentle slope

40

moderate slope Steep slope

20 0 1

2

3

4

5

6

Sharp slope

Forest canopy density grade

Area percent (%)

100

B Shaded aspect Semi-shaded aspect Semi-sunny aspect sunny aspect No aspect

80 60 40 20 0 1

2

3

4

5

6

Area percent (%)

Forest canopy density grade

90 80 70 60 50 40 30 20 10 0

Valley

C

Low slope position Middle slope position Upper slope position Hill top 1

2 3 4 5 Forest canopy density grade

6

Figure 6. Area distribution of forest canopy density grades on terrain（A, Slope; B, Aspect; C, Position）gradient in 2000.

372

F.-J. XIE ET AL.

Figure 7. Area distribution of fire severity on various terrain gradients. A: F-Flats; GSGentle slope; MS-Moderate slope; StS-Steep slope; ShS-Sharp slope. B: SA-Shaded aspect; SSA-Semi-shaded aspect; SSUA-Semi-sunny aspect; SUA-Sunny aspect;NA-No aspect. C: VValley; LSP-Low slope position; MSP- Middle slope position; USP-Upper slope position; HT-Hill top.

POST-FIRE FOREST RESTORATION

373

forestland. Though artificial regeneration ways did not have obvious effect on forest productivity now, it was expected that restoration of conophorium productivity could be improved in the future 50-80 years. Slope, aspect and slope position had all influenced forest canopy density restoration, with slope as the most significant, and aspect as the least. Terrain factors exerted their influences on post-fire forest restoration mainly through influencing the intensity of fire and offering different site conditions. The area proportions of canopy density grade in flats, valley and no aspect land had obvious difference from those on other terrain gradients, which was relative to larger proportion of wetland in this kind of area. Therefore probing into the evolvement of forest wetland will be of great importance on forest landscape restoration further. ACKNOWLEDGEMENT This paper was sponsored by the National Natural Science Foundation of China (No. 30270225，40331008), and the Chinese Academy of Sciences (KSCX2-SW-133). REFERENCES Bergeron, Y. and Brisson, J. (1990). Fire regime in red pine stands at the northern limit of the species range. Canadian Journal of Forest Research, 17, 129-137 Garren, K. H. (1943). Effects of fire on vegetation of the Southeastern United States. Botanical Review, 9 (3), 733-736. Gill, A.M. (1975). Fire and Australian flora: a review. Australian Forestry, 38 , 1-25. Grogan, P., Bruns, T.D. and Chapin, F.S. (2000). Fire effects on ecosystem nitrogen cycling in a Californian bishop pine forest. Oecologia, 122 (4), 537-544. Guan, K.Z. and Zhang, D.J. (1989). Influences analysis of forest fire on Daxinganling Mountains on vegetation. Environmental Science, 11 (5), 82-88. Kong, F.H., Li, X.Z. and Wang, X.G. (2003). Advance on Study of forest restoration in the burned blank. Chinese Journal of Ecology, 22 (2), 60-64. Li, F.Z., Zheng, H.Y. and Lu, Y.B. (1988). Effects of catastrophic forest fire in the Great Xingan Mountains on soil microorganisms. Chinese Journal of Ecology, 7 (suppl.), 60-62. Liu, G.H. and Fu, B.J. (2001). Effects of global climate change on forest ecosystem, Journal of Natural Resources, 16 (1), 71-78. Luo, J.C. (2002). Influence of forest fire disaster on forest ecosystem in Great Xing’anling. Journal of Beijing Forestry University, 24 (5/6), 101-107. Nakagoshi, N. (2001). Forest fire and management in pine forest ecosystem in Japan. Hikobia, 13, 301311. Pianka, E. (1992). Disturbance, spatial heterogeneity, and biotic diversity: fire succession in arid Australia. Res. Explor, 8, 352-371. Shen, Z.H., Zhang, X.S. and Jin, X.Y. (2000). Gradient analysis of the influence of mountain topography on vegetation pattern. Acta phytoecologica Sinica, 24 (4), 430-435. Turner, M.G., Romme, W.H., Gardner, R.H. and Hargrove, W.W. (1997). Effects of fire size and pattern on early succession in Yellowstone National Park. Ecological Monographs, 4 (67), 411-433. Wang, X.G. and Li, X.Z. (2003). Model of vegetation restoration under natural regeneration and human interference in the burned area of northern Daxinganling. Chinese Journal of Ecology, 22 (5), 30-34. Weaver, H. (1951). Fire as an ecological factor in the southwestern ponderosa pine forest. Journal of Forestry, 49, 93-98. Xiao, D.N., Tao, D.L., and Xu, Z.B. (1988). Impacts of an extraordinarily disastrous fire on forest resources and environment. Chinese Journal of Ecology, 7 (suppl), 5-9. Xu, H.C. (1988). Da Hinggan Ling Mountains Forests in China (pp. 7-21). Beijing, Science Press. Yang, S.C., Liu, X.T. and Cao, H.B. (1998). 1994. Vegetation change on burn blank in Da Xing’anling forest area. Journal of Northeast Forestry University, 26 (1), 19 - 22.

374

F.-J. XIE ET AL.

Zhao, D.C. (1988). Vegetations and their restoration after the disastrous fire in the Great Xingan Mountains. Chinese Journal of Ecology, 7 (sup), 5-9. Zhao, K.L., Zhang ,W.F. and Yang, Y.X. (1994). The impact on swamp and countermeasure of the fire in the Da Hinggan Mountains analysed by vegetation. In Zhao, K.Y., Zhang, W.F., Zhou, Y.W. and Yang, Y.X. (Eds.). Environmental Influences and Strategies of Forest Fire in Daxinganling Mountains. (pp. 54-63). Beijing, Science Press. Zhou, M., Yu, X.X. and Feng, L. (2003). Effects of permafrost and wetland in forests in Great Xing’an Mountain on ecology and environment. Journal of Beijing Forestry University, 25 (6), 91-93. Zhou, Y.W., Liang, L.H. and Guo, Z.W. (1994). Changes of the Hydro-Thermal regime of frozen ground after the forest fire in the Northern part of the Da Hinggan Mountains. In Zhao, K.Y., Zhang, W.F., Zhou, Y.W. and Yang, Y.X. (Eds.). Environmental Influences and Strategies of Forest Fire in Daxinganling Mountains (pp. 25-35). Beijing, Science Press.

CHAPTER 22 KYOTO AS A GARDEN CITY A landscape ecological perception of Japanese garden design

Y. MORIMOTO Kyoto University, Graduate School of Global Environmental Studies, Kyoto, Japan

Abstract. This paper will explain the key phrase of “Kyoto as a Garden City,” as a secondary nature based on comprehending three ideas: one, the relationship between the landscape of Kyoto and the site of the Japanese garden, namely, the understanding of the garden culture as a part of the landscape. Two, garden, as secondary nature is the result of the continuous interaction of nature and the garden, rather than nature or the design intention itself. Lastly, the creation of Japanese gardens was based on the fractal property of unconsciously recreating nature. These discussions are inspired by ecological and landscape ecological concepts, such as edge effects, eco-tones, disturbance dependent ecosystems and hierarchical perception of ecosystems. The author concludes that the amenity of traditional Japanese garden is strongly related to the sustainability, which is clarified by landscape ecology.

1. INTRODUCTION Too many papers have been published about Japanese gardens in Kyoto. With forty percent of the traditional Japanese gardens, valued as Japan’s Cultural Asset, exist in Kyoto; it is not surprising that many strongly associate the prefecture of Kyoto to the historical Japanese gardens. These have reflected the world and aesthetic values of the owners, garden designers, and even those of the lower class workers also known as “Kawara-mono” or “homeless” at that time. As a result, those in different fields of study have exchanged highly diverse viewpoints and structures regarding the Japanese gardens. These include garden designers who create gardens, experts of garden history who study excavations and decode Japanese ancient writings, and also literary writers who enjoy and appreciate gardens. I do not mean to make endless comments on well-known articles. But rather, it would be interesting to find out why so many people have been captivated by the Japanese gardens. 375 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 375–388. © 2007 Springer.

376

Y. MORIMOTO

The nature of Kyoto may be the secret of attracting so many people, more than the history, or the actual design of the Japanese gardens. Rather than naming his book “Japanese gardens of Kyoto,” Katsuhiko Mizuno’s photography book is titled as “Kyoto as a Garden” (2001). This book exactly reflects how I perceive the Japanese gardens in Kyoto. The photographs in this book do not simply present the gardens; one, it allows a glimpse into the life found in the gardens, two, it displays the perception of the ancient people towards nature, and three, it exhibits the uniqueness of the climate of Kyoto to be able to incorporate all of those abovementioned characteristics. As a result, without going into details of the arguments or form criticisms by experts, the secret can be found in the phrase: “Kyoto as a Garden City.” This paper will explain this key phrase of “Kyoto as a Garden City,” as a secondary nature based on comprehending three ideas: one, the relationship between the landscape of Kyoto and the site of the Japanese garden, namely, the understanding of the garden culture as a part of the landscape. Two, garden, as secondary nature is the result of the continuous interaction of nature and the garden, rather than nature or the design intention itself. Lastly, the creation of Japanese gardens was based on the fractal property of unconsciously recreating nature. 2. LANDSCAPE 2.1 Landscape and garden The landscape of Kyoto produced the Kyoto garden culture. This landscape is not simply a pretension; it is the result of the long interaction between human beings and nature. As a result, this interaction became a part of the Japanese culture. It may sound too abstract. However, the following section will further develop on this concept. 2.2 The Landscape Structure between Hiei-zan mountain and Dai-mon-ji Landscape consisting of Higashi-yama Mountain, Dai-mon-ji, Hiei-zan Mountain, and Kiyomizu has a historical value. The Higashi-yama Mountain is a part of a gentle mountain range of thirty-six peaks and Hiei-zan Mountain stands as the highest peak in the northeast. South of Hiei-zan Mountain is Dai-mon-ji-yama Mountain where between these two, the region is lower. From the city of Kyoto and the eastern mountains, an impressive, hollow skyline view can be observed. The garden culture is closely tied with the unique geology, topography, and vegetation of this region. From a geological perspective, sedimentary rocks like chert and shale make up the mountains, which surround Kyoto basin. Whereas, granite, from the Mesozoic Era, consists in the soil between Hiei-zan Mountain and Dai-mon-ji. These geological differences had much influence on the topography, vegetation, life style, and culture of the people. When granite, a type of igneous rock, came in contact with the sedimentary rock, a chemical reaction of thermal metamorphism occurred

LANDSCAPE ECOLOGY OF GARDEN CITY IN JAPAN

377

and created a hard rock called “Hornfels,” which means “cow’s horn.” This blackish rock is called “Yase-no-maguro” if it comes from Yase of Takanogawa River and “Shishigatani Stone” if originates in Dai-mon-ji-yama Mountain. The “Shishigatani Stone” has a purplish hue. The smooth, blackish color of hornfel fits in well in the moss garden. Sometimes the re-crystallized pattern of the cordierite can be found in the hornfel, which is very much valued as a garden rock. On the other hand, granite rock weathers quickly and it is very common to deeply weather to sand in the geological stratum. Therefore, the granite mountain erodes easily and become smaller whereas the hornfel peak survives. These are the peaks of Hiei-zan Mountain and Dai-mon-ji. The Shirakawa-zuna Sand is made from eroded granite where its grains of quartz and feldspar are white. This has become an essential element to the Japanese gardens of Kyoto. Granite, which has not weathered as much as “Shira-kawa-zuna Sand,” is called “Shira-kawa-ishi Rock” and is often used to create garden rocks and stone lanterns. This rock is the source of the stone garden culture. The garden of Ginkaku-ji Temple masterfully used the terrain of Shira River flowing into the basin of Kyoto. The open sea or “Gin-sha-dan” expressed by the Shira-kawa-zuna Sand is beautiful. The Shira-kawa-zuna Sand found at the temple not only reinforces the beauty of the landscape but also produces the landscape of traditional pavement. Furthermore, as also seen in the gardens of Ryo-an-ji Temple and Daitoku-ji Temple, the whiteness of Shira-kawa-zuna Sand reflects the light, allowing natural light into the buildings. Currently, the extraction of river sand is prohibited, including the Shira-kawazuna Sand. Therefore, it is a major problem when renovating the rock gardens of Ryo-an-ji Temple. Over the years, weather-beaten sand has become finer and renewal is necessary to make ripple marks. The current replacement of Shirakawa-zuna Sand comes from mountain rocks that have been crushed and sieved. Since the grains of the crushed rocks are rounded off without corners and are all the same size, it lacks the characteristics of the true river sand. The natural river sand has corners and sizes of the grains are not uniformed. Similar problems arose for the garden restoration of Byo-do-in Temple in Uji City, Kyoto. The unique topography of cobblestones piling up in the region where the Uji River flows into the plainfield, allowed the distinguishing landscape of Byo-do-in Temple. Therefore, when the stones were prohibited from removal of the riverbed, the restoration became difficult. 2.3 Ecotone as the source of Japanese gardens Japanese gardens are not randomly located. Gomizuno-O Emperor built the Detached Imperial Palace at Shu-ga-ku-in specifically for the reason of the surrounding landscape. Ridge of Hiei-zan Mountain in the background, this site is surrounded by mountains and terraced paddy fields. The garden of Shu-ga-ku-in Detached Imperial Palace consists of three teahouses: Teahouse of Kami, Naka, and Shimo, which mean upper, middle and lower part of the teahouse. From the top teahouse, one can observe a pond, the great trimmed hedgerow, trees and other common elements found at the Japanese gardens. Moreover, the famous vista of the

378

Y. MORIMOTO

borrowed scenery, of rows of mountains, exists in the background. The beautiful terraced rice paddy found at the foot of the granite mountain makes this landscape of Kyoto very unique. This is the result of fourteen years of search for the perfect site. The highland of the granite mountain is slightly complex and this was the best place for this garden. Unlike granite, the Tanba strata, where the Higashi-yama highland connects to Kita-yama Mountain and Dai-mon-ji from the south, a detailed topography cannot be produced, and thus, difficult to enjoy the terraced paddy field scenery. Nonetheless, it is common to find garden culture at the foot of the mountains or “san-raku.” For example, many historical gardens exist at the foot of Higashi-yama Mountain including: Gin-kaku-ji Temple, Ko-chi-in of Nan-zen-ji Temple, and Shisen-do. This area is considered ecologically unique for the transitional zone between the mountain and the flatland, ecologically known as “Ecotone.” This is where the unique landscape of Japanese gardens in Kyoto originated. There is a theory called “Greenway Hypothesis” (Ahern, 2002). This hypothesis explains that everything significant including natural resource and cultural assets are spread in a zonal distribution rather than in a random scatter. To begin with, diverse species heavily rely on areas where various resources are found, including the ecosystem along the river and coast. This ecotone is necessary for two reasons: one, as a corridor for species and two, as a place to obtain various resources. For example, river species depend on inorganic and organic nutrients from the river and the forest. Forest species, as well, rely on the river and forest ecosystem. This significance of the ecotone has been recently studied and revealed in the field of landscape ecology. From a cultural standpoint, importance of the ecotone could be the same. An ancient road, averting the bog near the foot of Higashi-yama Mountain, delicately runs north between the foot of the mountains and connect to the garden of Shu-gaku-in Detacthed Imperial Palace, and to the garden of Manshu-in. Towards the north, this ancient road meets another ancient road, Saba road, which leads to a small beach. The Saba Road runs straight along the Hanaore Fault, which was used to carry fruits of sea to Kyoto. Though the fault was found in a mountain terrain, this ancient road ran naturally straight. As a result, most of the ancient roads exist along the fault. Moreover, the geography and geology largely contributed towards the zonal distribution of an ecotone. During the Meiji Era, once again, an extraordinary garden culture was born. It is “Ueji,” Jihei Ogawa’s art. Many of his work survive in places other than Kyoto. However, the group of gardens found at the foot of Higashi-yama Mountain expresses his work the best. Not only the distinctive landscape shown at the foot of mountains, but also the body of water from Biwa Lake makes Ueji’s artwork of Aritomo Yamagata’s “Murin-an” and villa gardens of Nomura and Hosokawa remarkable. Biwa Lake contributed towards the modernization of Kyoto as well as the birth of the Second Biwako Canal Project. This added another wealth to the ecotone of “san-roku.” When rivers usually flow from north to south, “Sakasa-gawa River” or “Upside down river” weaves between the san-roku from Sanjou Keihan to the north.

LANDSCAPE ECOLOGY OF GARDEN CITY IN JAPAN

379

This Sakasa-gawa River became a significant element to Ueji’s art, allowing the water flow into his artwork.

2.4 The quality of the water system design of Japanese gardens known by fish The element of water flow makes the garden of Tai-ryu-san-sei-so as one of the masterpiece gardens. Unfortunately, the distinguished characteristics of this garden cannot be easily captured in a photograph. From the parlor, it appears as though the water fall drops into the plain field from the mountains at the back part of the garden. Though not large, it is impressive. The reason is the sound. When sitting in the parlor, the refreshing sound of the waterfall invites the eye into the back. However, in reality, the sound comes beneath the parlor where the actual waterfall cannot be seen. This water flow is simply not a waterway. The size is small but along with the waterfall, it has a stream and virago. Fish know this the best. In a limited water system, nature responds, allowing a rich group of fish species to habit the garden. The secret lies in Ueji’s design of creating a model based on the application of the water flow to diverse groups of nature and the water flowing down from Biwa Lake. Various fries swim down with the body of water from Biwa Lake. However, in a monotonous environment, fries disappear quickly as a large predatory fish eats them. Nonetheless, if refuge could be found in gaps or in shallow water, where large fish cannot exist, fries could escape from the predator. The natural structure of mountain stream allows such ecosystem and creates diverse habitats, not the featureless waterway. Incidentally, the Japanese garden of Heian Shrine displays how the water system of Ueji’s garden is distinguished for the habitat of the wild flora and fauna. The capacity of the garden at Heian Shrine is larger than the garden of Tariyu Mountain Villa or “Tairyu-sanso-teien” and the fish diversity is richer. Furthermore, this environment made by the appropriate current, stagnation and various shorefronts, allows endangered species to exist where even in Biwa Lake, it is difficult to find. The endangered species exhibited at the Museum of Biwa Lake were captured at the Japanese garden of Heian Shrine. The high fish biodiversity implies the richness of the water system as well as the biota. Where the Japanese bitterling species reproduce, a rich water landscape exists: of large bivalve clams and a proper habitat for spawning. On the other hand, in the confined garden pond where the large Nishiki carps live, this type of ecosystem reflects a poor landscape with low fish diversity and poor benthic biota; this definitely is not a rich landscape from the viewpoint of biodiversity. 3. THE ECOLOGY OF JAPANESE GARDEN LANDSCAPE The Japanese garden, with time, displays its beauty. Especially the gardens, which have been taken moderate care by a gardener who knows the life history of

380

Y. MORIMOTO

trees and plants, are the most beautiful. Immediately after the completion of a well-designed garden at fancy Japanese style restaurant, even well made, it is not too difficult to point out some cheesy characteristics. Yet, with five to ten years of appropriate care, even artificial and too formalistic gardens can acquire depth and richness of the Japanese gardens. Why is this so? In the past, there have been simple discussions of trying to explain this reason by an ecological standpoint or by the mystical relation between the gardener and nature. However, one cannot simply discuss about the Japanese gardens without understanding the relationship between nature and art. If this idea is simply divided into two, the essence of what we consider as the characteristics of “Japanese garden,” of the balance of nature and art, could be understood: one, the ecological concept: the phenomenon produced by the interaction between the living species and the environment, and two, concept relating to the natural feature: the human perspective towards nature when man takes care of the garden. 3.1 The interaction of trees and environment First, the ecology of gardens will be explained. The gardener plants few trees. Of course, one plants them in the proper soil with plenty of water. In the beginning, the shape of the tree reflects the environment where the tree spawned. This includes not only the shape of the trees but also the shape and deepness of the green color of leaves. The morphology, little by little, begins to reflect the new environment and the relationship of the gardener and the garden. The first response to the new environment is the direction of leaves. Appearing as though it doesn’t move, leaves and branches are driven towards the direction of growth. Immediately after planting trees, leaves face unpredictable direction. However, with few days, leaves can detect the direction of light. If successfully rooted, it can start growing again, depending on the strength of plants and the condition of the natural temperament: sunlight, soil, rain, and wind. However, the regulation process of plant species may force the leaves and branches to wither after completing its role. However, the leaves and branches do not just wither. The fall of autumn leaves is the proof of tree preparing for budding the following spring. Even cut branches, for propagation, leaves wither and fall as a proof of living. If the trees were able to root into the ground, leaves of trees with cut branches fall and buds grow, if not, leaves remain. The living species not only adapt to its environment but also produce its own environment. At the new environment, other species propagate from elsewhere and thus, a new environment is established. With ten to twenty years, this new garden harmonizes into the landscape and allows comfort into our heart. Even if there is no structure to the garden, the hundred years old tree, by itself, has the ability to create a space. Even after its death, trees also provide a source for inspiration. At the Tadasu-no-mori forest of Shimogamo, such dead trees could be found: One tree is decorated with a thick, sacred, straw rope called “Shimenawa” and the other is the source to the natural habitat garden or “Shizen-yuen.”

LANDSCAPE ECOLOGY OF GARDEN CITY IN JAPAN

381

3.2 Landscape of gardens as secondary nature How is it in the case of a tree where a skilled gardener has taken care for a long time? The colonnades of Japanese black pine are found on both sides of the carriage drive or “Obashiya-michi” of Detached Imperial Palace garden of Shugaku-in. For over hundred years, gardeners have continued to pick new buds of the Japanese black pines, also known as “Midori-tsumi.” Visitors who walk this road are moved not because of the nature or the work done by man; it is the long, uninterrupted interaction between the gardener and the pine. Furthermore, the terraced paddy field view found between the colonnades is the result of the long interaction between man, land, water and rice. This brings about ecological stability and appears in front of us neither artificially, nor wild. Until recently, in the field of ecology, compared to the wilderness area, this type of secondary nature was considered inferior and was not targeted for conservation work. Yet, recent studies have shown that a distinctive biota resulted from the long interactive relationship between man and secondary nature. These studies have also shown the remnant of the Ice Age. The strongest support of this concept lies in Satoyama, ridges and paddy fields, as a refuge for some living species. During the Ice Age, the biota was distributed evenly in the frigid and dry environment on the adjoining continents. With the warm temperature trend, this biota was driven into the laurel forest where it is dark, even during the daytime. The periodic cutting of firewood in traditional rural landscape or “Satoyama,” and mowing and plowing in paddy fields and ridges, allowed sunlight into the landscape. To support this theory, same species have been found in the grassland of Korean Peninsula, northeastern part of the Chinese continent, as well as the paddy fields in Japan. Dr. Koizumi, the first professor of Department of Botany, School of Science, at Kyoto University, looked at the burnet found on the paddy ridge and came up with the following theory. According to the endangered species study done recently by the Japan’s Ministry of Environment, plant species living in the secondary nature faces endangerment more than those habiting the wilderness area. The landscape of the terraced rice paddy, which reflected the lifestyle of the people then, was incorporated into the landscape of the carriage drive of Shugaku-in Detached Imperial Palace. This is the cultural landscape and secondary nature. The botanical landscape of traditional rice cultivating vista and its component is now on the verge of extinction due to the laurel forest taken over after the withering of pine trees and by agricultural modernization. The yearning one feels when looking at the landscape of Obashiyamichi, could be due to the uneasiness one feels towards the possibility of this landscape, which has existed since Yayoi Era of over two thousand years, to disappear. 3.3 Pruning that connects the gardener and the tree When the garden shear is inserted into the tree, depending on the method, each plant species respond differently. Beneath the cut surface, many tree species generate adventitious buds or “Futei-mei.” Maples also produce many adventitious

382

Y. MORIMOTO

buds beneath the cut surface, and, when all of these grow, they appear disheveled. If branches are snapped by hand, the snapped surface has a splinter and decay easily. As a result, it is difficult for the buds to sprout and the maple tree grows small, yet still mimics the large maple tree. The gardeners of Kyoto should be acquainted with such techniques. The luster and shape of leaves only need a month or two to reflect the newly established environment. In shaded areas, leaves are large and colors dark. Where it receives much sunlight, or receives little nutrition, the colors are light and leaves small. With the strong wind, branches and leaves scrape together and trees release ethylene, a plant hormone, allowing the canopy to grow compactly rather than the segments to extend endlessly. This method is applied to the care of Bonsai: branches are stroked and held with a clothespin for a compact growth. The soil condition also has some influences on the leaves. At the Katsura Detached Imperial Palace Garden, a mountain full of fall leaves, “Momiji-yama,” can be found. This steep, small, mountain is made up of fine grain soil, which acts as a spatial screen, essential, for a dramatic recurring landscape. The soil of the Momiji-yama is thick, making the water penetration difficult and water accumulates. As the water accumulates in the thick soil, roots cannot grow, and as a result, the maple tree cannot shoot and develop into a tall tree, producing a well-balanced maple tree form, proportionate to the size of the hill. Thus, even if the gardener were to cut into the tree, the growth does not change. After couple of years, the overall leave density responds to the soil condition and grows into what it is capable of. Yet, generally, the gardener has controlled these characteristics. If neglected, the upper layer of leaves grows thickly and the lower layer of shrubs, underbrush, and flowers grow thinly. In order to prevent such ecological succession, the gardener uses his scissor to take care of the tree. At the gardens belonging to the Japanese Imperial Household Agency of Katsura, Shyugaku-in, and Sentou Imperial Palace, one can find great examples of such methods. One never uses a barbaric pruning tool like the big two-handed scissors to take care of trees, an essential element to the Japanese garden landscape. If such tool were applied, it would lose the uniqueness of the tree form as a part of the Japanese garden. For example, the tree form of Ternstroemia gymnanthera, which makes up the main landscape of the Japanese garden, has been maintained by controlling the number of the buds for many years. The single-handed scissor or “wabasami” has sustained the tree form of azalea in front of Shyo-i-ken, a Japanese teahouse, rather than the big two-handed scissor. This method is time consuming; however, this is what the Japanese call, “Sukashi,” or branch thinning, which contributes towards the beautiful Japanese garden. 3.4 The changing garden of Katsura detached imperial palace Despite putting much care into the trees and gardens for many years, it changes easily. Even the rocks, which do not alter easily, weathers. The stone lanterns used at the garden of Katsura Detached Imperial Palace is appraised for the distinctive figurative artwork. The softness of the base material of the welded tuft makes this easy to sculpture. Yet, this welded tuft weathers easily and loses its shape. Thus, it

LANDSCAPE ECOLOGY OF GARDEN CITY IN JAPAN

383

should not be left in the outdoor landscape. Therefore, one should not complain that the current stone lantern at the Japanese gardens is an imitation lantern. When rocks weather easily, it is natural that the renewal of the wood-constructed teahouse is necessary. In a way, the teahouses of Katsura seem modest; the main materials of the teahouse appear as though it could be from any mountain. The delicate natural tree forms of Quercus variabiris and Lyonia ovalifolia are similar yet different with the original ones. As a result, when a radical innovation was done during the Showa Period, it was extremely difficult to find the perfect material of the teahouses. In order to preserve the Japanese garden, plant succession of each garden must be considered. In 1934, the “Muroto” typhoon hit the Kinki region, leveling the trees of Katsura Detached Imperial Palace. After the catastrophe, the annuals rings of the broken trees were analyzed, and were found out that these were not the original trees from when the imperial garden was originally built. Since the original garden of Katsura Detached Imperial Palace garden was built four hundred years ago, (for over two generations of Hachi-jo-no-miya), the trees have metabolized. Ecologically, this is one frame of plant’s ecological succession. If a site is stable, with the climate of Kyoto basin, it could grow a forest consisting of Castanopsis and Aphananthe aspera. Therefore, the garden care is extremely important as a part of maintaining the ecological succession of the Japanese garden. However, at the site of Katsura Detached Imperial Palace, the Katsura-gawa River can be found; this river often floods. Deciduous broad-leaved trees, which can be found at the Tadasuno-mori Forest such as hackberry, Aphananthe aspera and zelkova species prefer such flooding environment. Tree numbers have widely expanded since Bruno Taut (1880-1938) first introduced Katsura Detached Imperial Palace to western countries. Currently, along the pathway of Katsura Detached Imperial Palace, artworks done by the skilled gardener over generations have been preserved. Unfortunately, the limit of the budget makes these maintenance and care of the garden difficult. Once off the pathway, only bushland stretches across the landscape. From the Shyo-ka-tei, the skyline cannot be seen, when the main purpose of this site is the vista of the mountain. With the wind and birds bringing seeds into the gardens, species, which have not been planted, grow quickly, harmonizing its morphology into the landscape. This makes it difficult for the gardener to cut down these species, which do not belong here. Professor Ayaakira Okazaki, a former professor at Kyoto University’s Department of Landscape architecture, was the first person to lament, yet claim the necessity for a large scale thinning to recover the well-maintained Katsura. Though he passed away in 1995, fortunately, regular thinning continues at the garden. When accompanying Professor Okazaki at the garden of Katsura, he taught me a saying: “Let others do the thinning” even at afforested areas for silviculture. With this teaching and verifying by computer graphics, thinning was done at Katsura Detached Imperial Palace Garden. Though insisting the density be halved for the species that have lost its habitat to revive, nonetheless, I could not get too enthusiastic about the large-scale thinning done when these trees have been cared for

384

Y. MORIMOTO

a long time. As a result, in the end, minimal thinning was done and it was decided to observe the course of result overtime. This type of change happens very quickly. Therefore, gardens are most vulnerable to change and in some way, some figurative arts are left in the hands of the succeeding generations. Not only man, but also, some garden landscapes depend heavily on climate, as it will be explained in the next section. 3.5 Moss garden as a result Normally, in Kyoto, it fogs densely on a clear morning day. However recently, with the city climate, this trend has not been observed as often. Yet still, at Kameoka, this thick morning mist as if being surrounded by the Sea of Milk or dense fog along the river can be experienced. Because of this humidity, the hair moss exists on the forest floor of the chestnut farm at Kameoka and supports as well, the particular garden design by the expert gardeners. The Polytrichum commune, one type of moss found in a Japanese garden and sold at markets, usually are made in the Kanto region where it has been raised in the black volcanic ash. Unfortunately, this P. commune does not blend in well with the landscape of Kyoto. Yet, P. commune of Kameoka blends extremely well with its light color. Furthermore, the morning fog and mist has naturally cultivated the moss gardens of Kyoto. Including various moss species of more than ten types found at the Saihou-ji Temple, most of the mosses found in the gardens of Kyoto were not originally designed specifically for the gardens of Kyoto. At the Kyoto basin, normally, if appropriate amount of light is given and if fallen leaves are swept continuously, within a month during the rainy season, various types of moss species such as hair moss and Rhacomitrium conescens grow. However, unfortunately, it has been difficult for P. commune to grow recently. With the city climate raising the daily minimum temperature, despite the moisture level being the same, the percent humidity has decreased. As a result, the dew condensation in the morning disappeared. This has had a devastating impact on the moss species, especially those whose root systems do not germinate. Nonetheless, gardens found at a higher elevation, dew condenses easily and gardens of Shugaku-in Detached Imperial Palace and Sai-hou-ji temple can still manage. Yet, this changing climate has had an overwhelming impact on the gardens of level ground such as Katsura Detached Imperial Palace. Surprisingly, the P. commune species can only grow in bright sunlight and is extremely vulnerable to the climate. Therefore, as mentioned before, where it hasn’t been thinned enough, P. commune species disappear. Yet, where the sunlight from the west shines, the moss withers and dies due to the lack of moisture. In the past, the appropriate amount of pruning and sweeping of the garden, by itself, created the landscape of the moss garden. It is unfortunate that a substantial amount of annual supplementary planting has become necessary to sustain the landscape. The city warming has forced the traditional secondary nature to transform into something different from the past.

LANDSCAPE ECOLOGY OF GARDEN CITY IN JAPAN

385

4. FRACTAL PROPERTY With the exception of being found in Kyoto, the Katsura Detached Imperial Palace Garden, fall leaves on garden floors, Kita-yama-sugi Cedar, and Tadasuno-mori Forest, each have different purposes to the landscape of Kyoto but these have something in common: it is part of the Japanese garden. Photography books, artwork identified as “Japanese garden,” courtyards of small traditional urbane houses, large gardens of Japanese feudal lords, ancient Pure Land Buddhism Garden and the Murin-An annex. No one considers these various gardens to be out of place when identified as “Japanese Garden.” Neither the western style plane geometric garden like le jardin de Versailles or la parc of a palace, nor the natural landscape that have been influenced by the rural or agricultural landscape in England, nor the Islam patio have anything in common. Yet these are called “Western Garden,” and only because of the gardens exist outside Japan. The secret of a certain feel for commonality exists only in Japanese Gardens because of the “fractal property” (Morimoto and Seo, 2001). B. Mandelbrot, a professor of mathematics at Yale University, established the concept of “fractal property” in the early 1980s. According to Mandelbrot, the partial-ness resembles the whole-ness; it appears as though the shape or object looks the same when changing the magnification of the telescope or the microscope. This is called “selfsimilarity.” For example, if the deeply-indented coastline, ria coast, is drawn on a map and measured using a divider. The shorter the unit length of the divider, the more detailed the shoreline, thus longer the coastline. As a result, a set length does not exist. Like the shape of a thundercloud and the deeply indented coastline of the ria coast, formation of living things appears very complex. In the past, the vague adjective, “complex,” has been used often. It is a revolutionary discovery that this term “complex” allows the feeling of commonality to identify natural forms. Unconsciously, the fractal property should appear in the structure of the gardens naturally if the fractal property were applied into the beauty of the gardens consciously. This theory has been studied and applied to the shape to the pond of Katsura and was accepted within a certain parameter of unit length. More studies were done and found out that all of the ponds have fractal property in addition to the fractal dimensional value reflecting the complexity of the coastline. It has been concluded that the parameter capacity that make up the fractal-ness can be a strong tool for research when analyzing shape of the garden pond (Seo and Morimoto, 1996, 1997). Then, what about the distribution of garden rock and trees? To study the fractal distribution of these elements, mesh method was used. Figure 1 shows the measurement of the fractal dimensional value using the divider method with the pond of Ten-ryu-ji Temple as an example. Figure 2 displays the measurement based on the distribution of garden rock of Katsura Detached Imperial Palace Garden. With such methods, it was concluded that most of the Japanese gardens display the fractal property and fractal dimensional value differing between each time period. For example, the ancient Pure Land Buddhism Garden has a low fractal dimensional value whereas a middle value is placed on gardens made during the Muromachi (1392-1573), Momoyama (1573-1600), and early Edo Period (16001867). Moreover, by scrutinizing the placement of the garden rock, the high dimensional value of pathway method is one distinguishing feature. Yet, the

386

Y. MORIMOTO

frequency of tree species found here differed. This said, the fractal property was found in small sized gardens, but not in large sized gardens. Rather, regular set of tree species were chosen for the larger sized gardens.

Figure 1. Measurement of the fractal dimensional value of the shoreline of the garden pond of Ten-ryu-ji Temple.

This perhaps is due to as the size of garden increases; the nature-oriented design exceeds the capacity of the garden designer. Therefore, the computer allows not only the spontaneous and stylized design but also the larger capacity garden design (Seo et al., 1998). Interesting pattern was found in the size of trees, the tree height, width of tree, and diameter breast height (DBH) of Tadasu-no-mori Forest and garden of Katsura Detached Imperial Palace. At the Katsura Detached Imperial Palace garden, fractal property was found in tree height ranging from 5.6 to 20m, width of tree, 3.0 to 21m and not with DBH. Tadasu-no-mori Forest, on the other hand, fractal property was

LANDSCAPE ECOLOGY OF GARDEN CITY IN JAPAN

387

observed in tree height of 5.0 to 20m, DBH of 40.7 to 120.5cm. As a result, even well maintained Katsura Detached Imperial Palace and semi-natural Tadasu-no-mori forest, it is natural to find the fractal property in a spatial dimension of one tree, ignoring the minimum and maximum number. At the Katsura Detached Imperial Palace, of the growth of new trees, the thinner trees tend to be more thinned often than the thicker trees and thick trees on the verge of collapse were maintained to prevent such. This maintenance method may have reflected the above-mentioned results. In conclusion, the structures of the Japanese gardens are not natural. Rather, it is the result of the strong interaction between man and nature. Moreover, the fractal property analysis supports that the Japanese garden’s distinguished characteristics is based on the secondary nature. It is certain that this is one secret of Kyoto as a Garden City.

Figure 2. Measurement of the fractal dimensional value using mesh method of the distribution of garden rock of Katsura Detached Imperial Palace Garden.

388

Y. MORIMOTO

REFERENCES Ito, S. and Morimoto, Y. (2003). Garden ponds as wildlife habitats for fish from lake Biwa into Kyoto city. Journal of the Japanese Institute of Landscape Architecture, 66, 621-625 (in Japanese with English summary) Mizuno, K. (2001). Kyoto as a Garden, Tokyo: Japan Broadcast Publishing Co. Ltd. pp. 621-625 (in Japanese) Morimoto, Y. (1993). Computer visualization of the vegetation of Katsura imperial palace garden by a plant modeling and visualization system. Journal of the Japanese Institute of Landscape Architecture, 57(2), 113-120. (In Japanese with English summary) Morimoto, Y. and Natubara Y. (Eds). (2005). Living Forest: Theory and Application of Natural Habitat Oriented City. Kyoto, Kyoto University Press. (In Japanese) Morimura, A., Morimoto, Y. and Seo, Y. (2003). Case studies of fractal properties in semi-natural landscapes. GEI, 1, 90-98 (Univ. Human Environment, Okazaki) Seo, Y. and Morimoto, Y. (1996). Fractal on design elements of Katsura imperial garden. Journal of the Japanese Institute of Landscape Architecture, 60, 56-60 (in Japanese with English summary) Seo, Y. and Morimoto, Y. (1997a). Fractal distribution on land use pattern and patch form in an area characterized by slush and burn type of agriculture. Papers on Environmental Information Science, 11, 65-68 (in Japanese with English summary) Seo, Y., and Morimoto, Y. (1997b). Fractal on design elements of Japanese garden. Journal of the Japanese Institute of Landscape Architecture, 60, 615-618 (in Japanese with English summary) Seo, Y., Morimoto, Y. and Morimura, A. (1998). Expert CAD system with fractal method on Japanese garden design. Papers on Environmental Information Science, 12, 137-142 (in Japanese with English summary)

CHAPTER 23 BEE-BO FOREST: TRADITIONAL LANDSCAPE ECOLOGICAL FOREST IN KOREA

K.-S. LEE Department of Landscape Architecture, Sungkyunkwan University, Korea

Abstract. People want to live in good environment. In Korea and China, ancient people tried to locate their housing and villages within good surrounding environment. This was named Poongsoo in Korea and Fengshui in China. The theory describes the harmonious spatial relationship between human settlements and natural environment. Because every place cannot have good conditions for residential location in terms of Poongsoo, ancient Koreans tried to improve their living environment by supplementing forests. It is based on Bee-Bo theory. In Korea, the landscape would be a part of total system that includes the man and nature. They thought the landscape could be damaged easily by improper land use and also be supplemented by careful landscape planning. To supplement insufficient landscape elements, Bee-Bo forest was created. It works as disaster prevention zone, microclimate control zone, biodiversity conservation patch and cultural landscape area. It is based on the philosophy that man and nature are parts of the universe and interaction between them should be managed based on energy equilibrium.

1. INTRODUCTION People want to live in good environment. In Korea and China, ancient people tried to locate their housing and villages in good place. The focus of this idea was that the site is to be located within good surrounding environment. It was named ‘Poongsoo’ in Korea and ‘Fengshui’ in China. The traditional village in Korea has been located based on Poongsoo theory. The theory describes the harmonious spatial relationship between human settlements and natural environment. While the topography of ancient China is very flat and they emphasized the water movement in Fengshui theory, Korea has hilly terrain and the wind control considering surrounding topography has been more emphasized in ‘Poongsoo’ theory. In hilly area, energy distribution and wind flow are not the same. Thus the wind plays an important role in energy flow. Because every place cannot have good conditions for residential location in terms of Poongsoo, ancient Koreans tried to improve their living environment by supplementing forests. It is based on Bee-Bo theory. For locating 389 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 389–394. © 2007 Springer.

390

K.-S. LEE

their villages Koreans construct village forest named as Bee-Bo forest. It is worth to look into it with landscape ecological view. Thus, this paper reviews the traditional site planning theory, Poongsoo in Korea and China, and Bee-Bo, the unique characteristic of Poongsoo in Korea which is distinguished from Chinese Fengshui and finally how it was applied to village settlement through Bee-Bo Forest. 2. POONGSOO IN KOREA The essence of Poongsoo theory is to find the site named as ‘Hyeol’ that are good places to collect “Kee (living energy)”. The Kee is the essential idea of the ancient Orient. If the person has a good Kee, he or she is very healthy and if the site keeps lots of Kee, it is a good place for the life. So, the concern of people is focused on how to control and improve Kee in human being and how to find the good place that collects good Kee. The traditional medical science in Korea and China is very concerned about keeping and fostering Kee. It is also true of landscape planning in housing and village development. In ancient Orient they developed unique philosophy ‘Yin and Yang’ theory. Everything has dual aspects. For example, the male and female works as Yang and Yin for life. The sky and the earth do the same in the universe. The light and darkness, the water and land, they all work as Yang and Yin. It is the counterpart of one thing, however, both can supplement each other and the one thing cannot exist without the other thing. In advance, they cooperate and produce the synergy effect in harmony. Ancient Oriental people used to think the water as ‘Yang’ while the land does as ‘Yin.’ The land was classified into ‘Four Protectors.’ It is based on the direction of the site concerned. The surrounding four mountains or hills are named as follows. North; Black Tortoise South; Red Phoenix East; Blue Dragon West; White Tiger North mountain is called North Black Tortoise, south, Red Phoenix, east Blue Dragon, and west White Tiger. So, how these four protectors are arranged is very important criteria in assessing land suitability. In addition, the water was considered as ‘Yang,’ and the flow speed, direction, the streambed type, the amount of water, location was important in assessing the land capability together with value assessment in land (Yin). Figure 1 illustrates how Poongsoo was applied to capital location of Lee Dynasty (1392-1910). As we can see in Figure 1, Han River flows through Seoul City. It flows between the line connecting Mt. Nahk and Mt. Inwang and inner Red Phoenix (Mt. Nahm). In north there is Mt. Bookahk, inner dark tortoise, south Mt. Nahm, inner Red Phoenix, east Mt. Nahk, inner Blue Dragon, west Mt. Inwang, inner White Tiger. In Poongsoo, it is an excellent place for the city to be located. In outer side, there are also outer ‘Four Protectors.’ In north there is Mt. Bookhan, outer Dark Tortoise, south Mt. Gwanahk, outer Red Phoenix, east

TRADITIONAL RURAL LANDSCAPE

391

Mt. Acha, outer Blue Dragon, and west Mt. Dukyang outer White Tiger. Depending on season there is a major wind direction in Korea. Especially this kind of place protects inhabitants from being exposed to cold high pressure flowing from Lake Vikal in winter. Also it receives more solar radiation because it faces southwards.

Figure 1. The diagram of Poongsoo in Seoul.

In 1394, Lee dynasty settled their capital in Seoul, now the capital of Korea. Seoul was the first capital of Paikjhe Dynasty. They settled in Han River Basin in before Christ. As we can see in Figure 1, Han River flows through Seoul City. Koreans believe that the government occupying Han River basin unifies Korean peninsula. It has been verified throughout Korean history. Seoul is located in Han River Basin. In early Lee Dynasty, they afforested at the site where tributaries of Han River join. The site where tributaries join is called ‘Soo-Koo’ which is very important place in Bee-Bo Poongsoo. 3. BEE-BO FOREST USING POONGSOO THEORY The concept of ‘Bee-Bo’ in Poongsoo originated from ‘Tohseon’ Buddhist Monk, Shilla Dynasty (B.C. 57- A.D. 935) in late ninth century. Throughout Korea Dynasty (A.D. 918 - 1392) it was a main land philosophy in housing, village, city location. The insufficient portion of Four Protectors was supplemented by artificial construction using natural materials called Bee-Bo materials. Bee-Bo materials are not confined only to forests. Waters, Buddhist temples, Buddhist towers are also used for Bee-Bo materials. For example, the village is located at place which has

392

K.-S. LEE

good three hills (north, south, and west), however, the eastern hills are short or absent. Then the village people used to make forest to extend the eastern hill (Blue Dragon) or create new Blue Dragon. Then, they work as east Blue Dragon partly. In the folk village forest was formed for Bee-Bo forest like this way. There are many functions of Bee-Bo forest in terms of landscape ecology. Some of them are as follows. 3.1 Disaster prevention zone The stream is a meandering stream flowing near or through village in Korea. Traditionally the riverbed should be meandering type rather than straight linear form because it reduces the flow speed of the stream or river, and protects the village and crop field from stream erosion especially during the flood season. Figure 2 shows Hahmyang Upper Forest that is designated as Natural Monument Number 154. Originally the stream flowed in central part of downtown Hahmyang. In late ninth century, the county governor at Hahmyang, Chee-won Choi (the famous scholar in Shilla Dynasty) moved the stream channel to the suburbs of town and planted lots of trees to make forests beside the new stream for the flood control. During the dry season the forest can also supply the water for drinking and farming. Now it is the oldest forest made by human beings for flood control in South Korea. 3.2 Microclimate control zone One of the most important functions in Bee-Bo forest is windbreak. Especially, in island and coastal area the Bee-Bo forest obstructs the strong wind from flowing directly into the village. It reduces the wind speed and disperses wind direction. As a result, lighter wind can flow into village inside Bee-Bo forest. In Poongsoo, inhabitant should avoid the strong wind flowed into the body directly. The Bee-Bo forest can help people with reducing the chance of facing the strong wind. In storm blowing season, it also decreases the damage from the storm. In summer warm and humid wind is blocked at Bee-Bo forest. The trees and plants absorb the moisture from the wind and store the vapour in the forest. As a result, the air in the village could be drier than outside forest and people feels less uncomfortable. 3.3. Biodiversity conservation patch Many of Bee-Bo forests are forest patches connected to riparian corridor. So there are both riparian ecosystem and land ecosystem in the Bee-Bo forest. So, the edge of Bee-Bo forest is an ecotone area where the biodiversity is high. It works as habitats for small mammals like raccoons, Oriental badger, weasel and stepping-stones for species dispersal. It is also a water supply area for mediumsize mammals like Chinese water deer. It is an indispensable element in econetwork in modern landscape ecological viewpoint. Adjacent to Bee-Bo forest, there are crop field from which nitrogen and phosphorous are discharged during farming. They can be stored and absorbed in Bee-Bo forest. After filtering the

TRADITIONAL RURAL LANDSCAPE

393

rest are discharged into the stream. It reduces eutrophication in the stream as nonpoint source pollution. The forest can reduce the water pollution as a vegetation buffer zone. For the insect, a larva skin is soft and should be protected from being exposed to direct sun light. During their life in the water there is no problem. However, during the pupa stage they live in the land. They need shade to become mature insect. So Bee-Bo forest can be a good habitat for these insects. Ultimately it is helpful to keep and conserve biodiversity. 3.4. Cultural landscape area Many Bee-Bo forests have old big trees. They are hundreds of year old ones. Around these trees many cultural activities have been done. People used to take a rest for cooling during hot summer because it provides shades and there is cool breeze under the tree. Children also play around these trees. In many forests there are village altars to pray for the safety and health of people and village. It has been directly related to life of inhabitants. So, Bee-Bo forests have been maintained reflecting the history, belief and culture of the village. Thus, Bee-Bo forests were also important cultural landscapes.

Figure 2. Hahmyang Upper Forest.

394

K.-S. LEE

4. CONCLUSIONS People want to live in harmony with nature. So, they developed their own land philosophy. In Korea, the landscape would be a part of total system that includes the man and nature. They thought the landscape could be damaged easily by improper land use and also be supplemented by careful landscape planning. To supplement insufficient landscape elements, Bee-Bo forest was created. It works as disaster prevention zone, microclimate control zone, biodiversity conservation patch and cultural landscape area. It is based on the philosophy that man and nature are parts of the universe and interaction between them should be managed based on energy equilibrium. Thus, the rural landscape in Korea had been sustainable until late nineteenth century. From twentieth century many Bee-Bo forests are gone due to the industrialization and commercial development. However, it is now the time to appreciate their value in terms of landscape ecology and to preserve them for descendants.

REFERENCES Brown, R.D. and Gillespie, T.J. (1995). Microclimatic Landscape Design: Creating Thermal Comfort and Energy Efficiency. John Wiley & Sons, New York. Carpenter, P.L., et al. (1975) Plants in the Landscape. W.H. Freeman, San Francisco. Choi, J.-U., and Kim, D.Y. (2005) Perspectives on the landscape ecological function, of Dangsan forest and rural community forests as stream landscape. Environmental Police Study, 4, 31-55. Choi, W.-S. (2000). Cultural Geography of the Bi-Bo in Youngnam Region. Ph.D. Dissertation. Korea University Seoul (in Korean) Forman, R.T.T. (1995) Land Mosaics: The Ecology of Landscapes and Regions. Cambridge University Press, Cambridge. Kim, H.-B, and Jang, D.-S. (1994). The Korean Village Grove. Youl Hwa Dang Publisher, Seoul. (in Korean) Marsh, W.M. (1991). Landscape Planning; Environmental Applications. John Wiley & Sons, New York. McHarg, I. (1969). Design with Nature. Doubleday Press, New York. Ministry of Environment. (2002). River Restoration Guideline, Seoul, Korea (in Korean) Wong, E. (2001). A Master Course in Feng-shui. Shambhala, Boston. Yoon, H.-K. (1976) Geomantic Relationships between Culture and Nature in Korea (Taipei: The Orient Culture Service) Yoon, H.-K. (2004) Environmental Perspectives of Poongsoo. In Lee D. (Ed.), Traditional Landscape Ecology (pp. 48-75), Seoul National University, Korea http://nature.cha.go.kr http://user.eandong.net/chodangryu/ http://www.hwasun.jeonnam.kr http://www.seongju.go.kr/tour/

.

CHAPTER 24

CULTURAL PATTERNS AS A COMPONENT OF ENVIRONMENTAL PLANNING AND DESIGN

R.D. BROWN1, R. LAFORTEZZA2, R.C. CORRY1, D.B. LEAL1, G. SANESI2 1

School of Environmental Design and Rural Development University of Guelph Guelph, Ontario, Canada N1G 2W1; 2Department of Plant Production Science, University of Bari, Bari, Italy

Abstract. Rural landscapes are multi-functional systems. Environmental functions are influenced by both natural and cultural landscape patterns. Beyond the traditional productive functions, rural landscapes are increasingly being recognized as complementary sources of biodiversity and places for cultural identification. Rural landscapes can often be seen as a complex assemblage of structural elements (patches, corridors, and matrix) whose arrangement reflects the magnitude, intensity, and type of human intervention and influence. This chapter describes some of the cultural patterns inherent in selected rural landscapes. It outlines how cultural artifacts and remnant habitat patches can affect ecological functions in two contrasting landscapes: the relatively young agricultural landscapes of southern Ontario, Canada; and longer-established agricultural landscapes of the Apulia region in southern Italy. For these landscapes, we illustrate the effects of cultural settlement patterns on habitat patterns and discuss implications for enhancing ecological attributes through landscape planning and design

1. INTRODUCTION The concept of landscape is often studied in a cultural context because it is a spatial entity perceived and influenced by human activity over time. However, inherent in this holistic and dynamic concept is the recognition that landscapes are seamlessly related to ecological function. The local cultural values and ideals that respond to and evolve landscapes also directly determine the structure, function, and change of the ecosystems therein and those linked at a coarser or finer scale. As with the idea of landscape, ecological function can vary in relation to time and space. Due to their

395 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 395–415. © 2007 Springer.

396

R.D. BROWN ET AL.

relative size and nature1, rural landscapes have been the focus of several studies examining how cultural activities relate to ecological function within a landscape (Tress and Tress, 2002; Correy and Nassauer, 2002). Rural landscapes can often be regarded as a complex assemblage of remnant habitat patches and green corridors set within an agricultural matrix. The structure of these landscape elements can affect many ecological processes, including movement and persistence of species, spread of disturbances, and flow of matter and nutrients across the landscape (Forman, 1995; Forman and Godron, 1981; Turner, 1989). This chapter focuses on cultural artifacts and patterns inherent in selected rural landscapes of Europe and North America, and illustrates how they affect ecological function. In this context, a cultural artifact is considered to be any tangible object or element present in a landscape that has either been made or modified by human activity (Korr 1997), and while it is not a process, it may be the result of a process (Leal 2005). The repetition and arrangement of cultural artifacts creates patterns – measurable physical expressions of the relationship between a culture and the environment it inhabits. Whether historical or recent, cultural artifacts are often multi-functional. Understanding the role of these patterns at the landscape scale allows landscape architects and planners to guide human activity so as to achieve the desired production while maintaining or enhancing ecological functions such as water movement, species diversity, and microclimate. The data presented here were collected from two contrasting cultural landscapes: the long-established agricultural landscapes of Apulia, in southern Italy; and the relatively recent agricultural landscapes of southern Ontario, Canada. Exploring the physical patterns of such different cultural landscapes not only provides insight into the place-specific relationship between people and their landscape but also demonstrates the potential for place-specific landscape patterns to improve ecological function. Corry and Nassauer (2002) have identified three sets of cultural values and traditions that affect the structure and, consequently, the ecological function of rural landscapes: (a) land division, settlement patterns, and ownership traditions; (b) applied science and technology; and (c) stewardship values and landscape aesthetic values. These three categories provide the framework for our discussion of cultural patterns and their effect on ecological function. 2. CULTURAL PATTERNS IN RURAL LANDSCAPES 2.1 Land division and settlement patterns 2.1.1 Italy The patterns of land division and settlement have commonly fragmented primeval ecosystems along lines that coincide with road networks, farm boundaries, and field 1

Nature here refers to the defining characteristics such as shape, texture, connectivity, and land use.

CULTURAL PATTERNS IN PLANNING AND DESIGN

397

patterns. As in many Mediterranean regions, the landscape in the Apulia region of southern Italy has been influenced by many different civilizations and their land use practices over millennia, resulting in a deep and textured landscape history that continues to evolve (Makhzoumi and Pungetti, 1999). Traditionally, rural activities have played a major role in shaping these landscape mosaics. Cereal cultivations, olive groves, vineyards, citrus groves, inter-cropped orchards, and other multi-use farming systems are examples of the human exploitation of this region. The traditional farming system was often labour-intensive, technological resources were few, and there was a low level of productivity. This gradually led to complex and heterogeneous cultural-rural patterns typified by a fine-grained texture, made by relatively small patches and corridors, and high species diversity as a consequence of cyclical perturbations introduced by rotational grazing, cutting and coppice regimes, and fire management together with cultivation and other human land uses (Naveh, 1995). The conventional subdivision of properties into small units, due to ownership succession and transfers of property, has further augmented this structural heterogeneity, thus influencing the contemporary cultural patterns of rural landscapes (see Figure 1).

Figure 1. Fine-grained, fragmented landscape of Apulia, Italy (© Ufficio Informatico and Servizio Cartografico, Regione Puglia, 1997: with permission).

The physical expression of land division and property ownership is further reinforced visually by the presence of limestone walls and other structures and facilities such as terraces, hedgerows, and canals along property lines. These landscape artifacts have the dual function of protecting soils from being eroded by wind and water and delineating farm property boundaries. As a physical consequence of the nature and physical properties of these artifacts, they can potentially play an important role in the ecological function of the landscape at a

398

R.D. BROWN ET AL.

broader scale. For example, indigenous, forest-interior tree species, such as Quercus pubescens and Quercus trojana, find microclimates conducive to good growing conditions adjacent to stone artifacts. The presence of these species not only increases biodiversity and structural complexity in the landscape, they may also provide valuable habitat for species that would otherwise be displaced in this highly deforested landscape (Figure 2). Furthermore, the linear arrangement and frequency of these stone artifacts may provide an opportunity to connect the habitat patches thereby creating a network that enhances the ecological function of the landscape.

Figure 2. Mediterranean forest vegetation (Quercus pubescens and Quercus trojana) along stone artifacts in the Apulia region of Italy (photograph by Raffaele Lafortezza).

2.1.2 Canada Before European settlement, the land that is now southern Ontario was occupied by indigenous people who cleared small areas in the vast forests to grow agricultural crops. These disturbances were small and widely-dispersed. When the population in southern Ontario increased rapidly in the late 1700s due to immigration there was an immediate need for the land to be surveyed. Most of the surveying was conducted by relatively unskilled labourers (Hart, 1998) and as a consequence southern Ontario ended up with 5 major land division systems and 166 variations, resulting in a

CULTURAL PATTERNS IN PLANNING AND DESIGN

399

unique, oddly-patterned landscape, especially at the boundaries where different survey systems meet (Hart, 1998). In contrast to much of the mid- and western parts of North America where the land was divided in 1-square-mile2 grids (640 acres or 259ha) aligned with cardinal compass directions (Johnson, 1976), southern Ontario was divided into grid blocks and farms that varied in size, dimension, and compass direction from township to township (Hart, 1998). Whereas mid-western farms and fields were principally square (the norm for early settlement was 1/2 mile by 1/2 mile in dimension, with fields that were commonly square divisions within the farm), farms in southern Ontario were long, narrow rectangles (for example, in the 1 000 acre [404ha] blocks farms could be 80 by 200 rods3, with farm buildings at the frontcentre and fields that were commonly half the width of the farm). Hence the distance among field boundaries and the shape of fencerow vegetation networks differs among land division systems. As villages, townships, roads, farms, and fields in Ontario were laid out along the poorly-surveyed lines of the various systems, a number of odd patterns resulted: “townships had to respect the boundaries of those that already existed, which necessitated some triangles, parallelograms, and other weird angles, and the complexity was compounded when concessions [rows] of narrow farms were laid out at right angles to some of the major colonization roads that sliced diagonally across the countryside” (Hart, 1998, p.162-3). Southern Ontario land division patterns created acute corners in fields, farms, roadsides, and townships (Figure 3). Acute corners are often not easily managed, and small patches of biodiversity are often associated with the intersecting boundaries of land division or management. Early settlers in Ontario built roads in front of their farms (Hart, 1998) using timber cleared from the fronts of farm lots. Farmsteads were often established close to the roads, clustering houses, barns, and human activity along front property lines. The clearing of forests to make arable land began at the road’s edge and continued toward the back property line of the farm. Since woodlots are a source of fuel, construction timber, hunting, trapping, and maple syrup, almost every farm retained a woodlot along back property lines. Cumulatively a row of woodlot corridors forms across neighboring farms, blocks, and townships in a pattern that begins to identify settlement systems (Figure 4). These corridors and patches offer an opportunity for enhancing landscape ecological integrity. It is important that as the landscape changes and develops over time that these critical components are recognized and conserved.

2

Imperial measurement units are cited as the traditional land division unit basis for these farms. Metric quantities are not part of the cultural tradition of these landscapes.

3

A rod is a system of measure equalling sixteen and one-half feet, or approximately 5 metres. It is the cultural measurement standard for southern Ontario farms.

400

R.D. BROWN ET AL.

Figure 3. Aerial photograph of a portion of rural southern Ontario showing the confluence of four different land survey systems. Note acute angles where survey boundaries meet (© County of Oxford, Ontario, 2003: with permission).

Figure 4. Oblique aerial photograph of a portion of rural southern Ontario showing pattern of cropping and remnant woodland patches (photograph by Robert C. Corry).

CULTURAL PATTERNS IN PLANNING AND DESIGN

401

2.2 Applied science and technology Applied science and technology are influential agents for driving the modification of rural landscapes. From clearing timber with axes and saws, to pulling stumps, to draining, leveling, fencing, cultivating, and building, science and technology have facilitated the removal and re-shaping of ecosystems. Machines have substantially eased the physical labor of land conversion, and allowed fewer people to convert larger areas in shorter times. Chemically supported agriculture has homogenized crop production and protection techniques, and in conjunction with land drainage and cultivation, have caused extensive ecosystem change.

2.2.1 Italy In the past few decades, the mosaic landscape pattern of Apulia has become increasingly homogenous due to the intensification of agricultural practices including the use of science and technology, and the development of global trade and economic pressures. Larger, wider machines, more chemical fertilizers and pesticides, and crop variety improvements together with advances in drainage and irrigation greatly impacted on the spatial organization of cultural patterns, in terms of the average size and shape of fields, and land use. In a recent study, Lafortezza and Sanesi (2003) provided evidence that intensification of olive production has been changing the pattern of the agricultural landscape. Traditionally, olive trees were planted far enough apart

Figure 5. Ancient olive grove with trees spaced far apart (photograph by Diane B. Leal).

402

R.D. BROWN ET AL.

to allow for harvesting by hand (Figure 5). While the olives used in making the finest grade olive oil are still picked by hand, this is a rare practice and most of Apulia’s olives are now harvested by machine. This requires a regular arrangement and increased production pressure encourages farmers to plant trees closer together. These recent practices are substantially changing the pattern of olive production lands (Figure 6).

Figure 6. Contempory olive grove. Note the close and regular spacing between trees. This regular arrangement intensifies production while accomodating mechanized harvesting (photograph by Diane B. Leal).

Traditional extra-moenia (i.e. the landscape located outside the city walls) landscapes are also being altered because of land use change. These traditional small garden plots typically have relatively regular edges and are located behind ditches, stonewalls or fences (Figure 7). Due to a growing tourist industry and other development pressures, these small gardens are now being converted to roads and parking lots. The change in land use is having severe impacts on the water drainage patterns in this area. During heavy rainfall, water that was previously infiltrated or slowed down by the presence of the extra-moenia perimeter gardens now runs off at greater speed and in greater quantity over the new impervious infrastructure causing erosion, instability and washouts (Figure 8).

CULTURAL PATTERNS IN PLANNING AND DESIGN

403

Figure 7. Some small garden plots, such as those in this photograph, still remain in the lower part of Ostuni, Italy, but many have been removed to make way for roads and parking lots (photograph by Raffaele Lafortezza).

Figure 8. The boundary wall of the city of Ostuni, Italy has been damaged by erosion (repair is seen as white). As historical terraced gardens are converted into paved parking lots, the speed and volume of water flow after heavy rains has increased (photograph by Diane B. Leal).

404

R.D. BROWN ET AL.

2.2.2 Canada In southern Ontario, local, provincial, and federal governments have supported drainage projects to convert wetlands into usable agricultural land and promote settlement. Subsurface farm drainage (using clay tile or plastic tubing) has effectively destroyed and fragmented wetland ecosystems and efficiently concentrated the flow of nutrients and pollutants to streams and rivers (Spaling, 1995). Farm machinery has become increasingly large, powerful, and capable of managing larger fields and farms with fewer operators. For example, corn planting equipment has increased from relatively narrow 4-row planters to widths up to 24 rows wide (Corry and Nassauer, 2002). While large machines increase the area of corn that can be planted in a day, a wider planter cannot readily access small fields or field corners. As with farms in the mid-western USA, larger equipment often has a longer turning radius and cannot easily access rectangular or acute corners or the odd patterns of southern Ontario settlement. Odd patterns can be difficult to manage with mechanized farm equipment. In a landscape that is intensively managed for livestock or crop production, odd patterns lead to areas that are not easily sprayed, fertilized, cultivated, mown, or fenced. It is in these areas that some of the most diverse ecosystems may be found, and disturbance may be reduced by inaccessibility to modern technologies. Odd patterns and acute angles can be difficult or costly to construct fences upon (especially if fences are made of posts and stretched wire). Early settlers created fences from locally-available material such as stumps, logs, stones, and vegetation (both spontaneous and planted). As steel wire – especially barbed wire – became available and machines could assist fence construction, corridors of woody fencerow vegetation were replaced with wire fences that were usually accompanied by strip of herbaceous vegetation. With the advent of electrified fencing for livestock control, multi-wire fences were again replaced by single or multiple electrified wires, and herbaceous vegetation strips could be mechanically mown or chemically-sprayed. As livestock became more concentrated in housed and fenced yards, the electric fences were removed and the former fencerow converted to cultivation. Farm tractors and equipment have increased in mass, with related threats of soil compaction (especially on wet or undrained soils). Large machines may avoid areas where drainage remains poor despite substantial effort to remediate soil wetness. Wet patches may remain as perennial vegetation, and are likely to concentrate relatively high diversity in small areas. These areas may connect to similar ecosystems along land division and settlement lines: roadsides, field boundaries, railways. 2.2.3 Agriculture and Microclimate In addition to the relationship between cultural and ecological patterns, there is an inextricable and reciprocal relationship between many of the cultural patterns in the landscape and microclimate. A recent study (Leal, 2005) identified a wide range of landscape elements that modify the microclimate to provide more

CULTURAL PATTERNS IN PLANNING AND DESIGN

405

productive, energy- and water-efficient, and thermally-comfortable environments. Cultural patterns in the rural landscape were found to have, at their root, a functional role in modifying the microclimate. For example, individual Quercus ilex or Quercus trojana trees are often found in the middles of pastures and fields (Figure 9). These trees provide shelter from the hot, intense midday and afternoon sun for both livestock and fieldworkers. With appropriate planning and management these individual trees could become stepping stones in the landscape.

Figure 9. Shade trees are often found in agricultural fields in Apulia. This Quercus provides shade for workers in a field near Martina Franca, Italy (photograph by Diane B. Leal).

It is also a common practice in the olive growing regions of Apulia to pile white rocks around the base of individual olive trees (Figure 10). The stones shade the soil from solar radiation and their light colour reflects a large amount of incident solar radiation. The energy input into the soil is consequently reduced, minimizing water loss through evaporation from soil near the root zone (Oke, 1979; Miller 1980). The stones also serve as insulating devices, moderating the soil temperatures so that they remain close to the year round average (Oke, 1979; Miller, 1980).

406

R.D. BROWN ET AL.

Figure 10. White rocks piled around the base of individual olive trees provide an effective mulch that increase water efficiency (drawing by Diane B. Leal).

Another example of the relationship between cultural patterns and microclimate is illustrated in the orientation and pattern of vineyards (Figure 11). For example, while the tidy rows of grape vines ripple throughout the landscapes of Apulia, their orientation is neither random nor irrelevant. It is important that the rows not shade one another and that leaves receive the maximum amount of sunlight necessary for maximum and high quality yields. In fact, temperature and solar exposure are the two primary concerns in the layout of a vineyard (Winkler et al., 1974). This is because light quality is a very important determinant of vine growth and grape quality (Gladstones, 1992). Well exposed leaves generally produce grapes with better colour and flavour so long as they are protected from sun burn4. In contrast, shaded leaves produce lower yields (Winkler et al., 1974). The optimal orientation for the rows is NNE to SSW. This means that both faces of each row will receive sunlight. The greatest potential for sunburn damage to grapes occurs in the afternoon when the air temperature is at its maximum. This orientation minimizes burning the 4

When sunlight falls incident upon the upper most leaves, they absorb most of the visible light necessary for plant growth and seed production. The remaining light that is not absorbed is mostly solar infra-red. This is poorly absorbed by the leaves (which is why the upper unshaded leaves did not absorb them in the first place) (Gladstones, 1992).

CULTURAL PATTERNS IN PLANNING AND DESIGN

407

grapes because in the mid-afternoon, the sun strikes the tops of vine leaves from the south west but the grapes remain shaded by the exposed overhead leaves and shadowed by the preceding row (Winkler et al., 1974).

Figure 11. Vineyards in Apulia are oriented so as to capture the maximum solar radiation which yields the highest quantity and quality of fruit (photograph by Diane B. Leal).

Microclimate has a pervasive effect on ecological functions, being an essential component of everything from movement and persistence of species, to the flow of matter and nutrients. Cultural patterns strongly influence microclimate and vice versa, while microclimate strongly influences ecological functioning within a landscape. 2.3 Stewardship and landscape aesthetic values Stewardship and landscape aesthetic values are unique in that they are not rationalized as necessarily economic choices, but rather include individual and community values (Corry and Nassauer, 2002). The management of the farm landscape is sometimes also a way of demonstrably communicating a farmer’s values and participation in the rural community (Leopold, 1949). Because the farmer’s values are not purely economic, they are an opportunity to protect or enhance ecological patterns and functions which may not contribute to profitability but which convey the farmer’s interests in stewardship and aesthetics. Farmers tend

408

R.D. BROWN ET AL.

to value productive cropping practices that also demonstrate that soil and water are being cared for. Soil erosion and weed growth are seen as signs of poor care, whereas crop residues and buffer strips appear as good care. Some of these management practices directly imply diverse ecosystems or caring for ecological quality. In contrast to stewardship values are the common values for good landscape care and aesthetics. Mown roadsides and fencerows are likely to give the impression of a tidy farm. Crop fields may be managed for weed elimination, far beyond economic necessity. Left-over and un-cropped patches are expected, by societal pressures, to be either removed or put into order by straightening edges and eliminating weeds. While stewardship values tend to increase biodiversity and care for ecological systems, aesthetic values, especially those that appreciate tidiness, decrease habitat diversity (Corry and Nassauer, 2002). 2.3.1 Italy Across Apulia, traditional structures such as terraces and stone walls (Figure 12) were constructed primarily to protect the sloping soils against erosion thus facilitating agricultural production on hillsides. These cultural artifacts communicate the fundamental role of farmers in controlling and preserving landscape resources, thereby establishing the cultural pattern of these landscapes. Stone walls bordering fields not only create a sense of separation and ownership from the neighbouring landowners (as mentioned previously), but also visually convey the image of good stewardship. A close look at patches of olive grove, vineyard, and cherry orchard gives evidence of the landowners caring for the land (Figure 13). Weeds are mowed or the ground is cultivated at regular intervals to avoid competition for the limited resources, especially soil moisture. Trees are constantly pruned, shaped and managed following traditional cultural techniques to facilitate production and fruit quality over the long-term. Cultural artifacts, cropping patterns, and remnants of vegetation contribute to the aesthetic beauty of landscapes in this region. Landscape aesthetic values emphasize cultural patterns of fields and stone walls which preserve vernacular identity and ecological diversity of Mediterranean rural-cultural landscapes. However, the desire for neatness in the landscape limits the potential for biodiversity and ecological richness. There is an opportunity for cultural values to be nudged in a direction that allows the beauty to remain, but includes habitat for native plants and animals.

CULTURAL PATTERNS IN PLANNING AND DESIGN

409

Figure 12. Terraces and stone walls in the landscape allow agriculture production on steep slopes without causing erosion (photograph by Robert D. Brown).

Figure 13. A cherry orchard in Apulia that is cared for in a manner that demonstrates neatness (photograph by Robert C. Corry)

410

R.D. BROWN ET AL.

2.3.2 Canada In southern Ontario buffer strips and fencing along stream banks are seen as contributing to good management of soil, water, and livestock. Grassed waterways and reduced tillage practices in crop fields are valued for conserving soil, moisture, and energy. Many of these stewardship practices are not extensive in Ontario, where a relatively cold climate and a short agricultural history have not invited full exploration of the application of appropriate stewardship practices. For example, contour cropping is a striking visual image that communicates good stewardship and is common in the mid-western and Pacific north-western USA where steep slopes have been cultivated (Steiner, 1990). Ontario’s steeper slopes have not been extensively cultivated and contour cropping is not part of the stewardship norm. Landscape aesthetic values in Ontario emphasize tidiness (Figure 14). Roadsides in southern Ontario are mown both by private landowners and by municipalities, not only for aesthetic reasons but also the practical reason of providing a location to pile snow removed from roads during winter. Orchards and vineyards are mown and ordered to facilitate production, but also to appear well cared-for.

Figure 14. Roadsides in southern Ontario are often carefully mown by adjacent landowners or by the local government (photograph by Robert D. Brown)

Farmers in some parts of southern Ontario continue to remove the remaining farm woodlots to demonstrate their agricultural knowledge. Wet or unproductive areas of farms are improved through drainage. Where limitations remain, field edges

CULTURAL PATTERNS IN PLANNING AND DESIGN

411

are re-shaped to eliminate irregularity, and herbaceous patches are mown to show care (see Figure 15).

Figure 15. Photograph of grass buffer strip along a stream in cultivated agriculture. Note that grass has been mown. Source: Natural Resources Conservation Service (Photograph by Lynn Betts).

The closely mown roadsides and buffer strips offer little in the way of microclimatic diversity. In contrast, the remnant habitat patches, particularly when they are of a fairly large size, can offer interior habitat with microclimates that support a diversity of species (see Figure 16). These habitat patches, although modest in size by Canadian standards, are enormous in comparison to those in Apulia and provide a significant opportunity to improve ecological function in the landscape. The structure and function of the southern Ontario landscape is remarkably different from that of Apulia, and offers a different range of opportunities. Large remnant habitat patches are available for conservation or preservation. Remnant habitat corridors along waterways and topographic features offer opportunities to enhance connectivity. Furthermore, there are many examples of highly functioning ecological areas integrated within the cultural patterns in the southern Ontario landscape. This combination provides an opportunity for increased landscape ecological integrity through careful planning and design. However, the need for neatness is limiting the potential for landscape ecological integrity. As in Apulia, the biodiversity of the landscape could be enhanced through an aesthetic that preserved the beauty but included habitat for a wider range of native plants and animals.

412

R.D. BROWN ET AL.

Figure 16. Remnant habitat patches persist throughout the southern Ontario landscape. When adjacent landowners keep woodlots at the backs of their properties the cumulative effect can be to provide a fairly large, contiguous patch that are connected through a series of corridors (Photograph by Robert C. Corry).

3. ENHANCING ECOLOGICAL ATTRIBUTES THROUGH PLANNING AND DESIGN Cultural patterns are critical to the form and function of ecological patches in rural landscapes. Depending on the geographical region and the farming tradition, these patterns are the consequence of humans’ adaptation of primeval ecosystems that have been shaped and spatially arranged to maximize productivity. In rural landscapes, cultural patterns can be described through consideration of the prevalent land-use of the agricultural region and the organization of settlements and infrastructures. These patterns are also inclusive of a wide range of artifacts in the form of small patches of remnant or diverse vegetation which contribute to the biodiversity and ecological functionality of these landscapes (e.g. habitat for wildlife species and opportunities for species dispersal in the rural landscape). Cultural values and traditions underlie the human activities that affect the structural and spatial attributes of landscape elements (Corry and Nassauer, 2002). Among these attributes, patch size, shape, core area, and spatial arrangement may

CULTURAL PATTERNS IN PLANNING AND DESIGN

413

convey evidence of the cultural ‘manipulation’ of the landscape. Such attributes are common descriptors of landscape configuration (McGarigal and Marks, 1995) as they influence the interaction of the patch with the surrounding matrix and functional elements (Lafortezza and Brown, 2004). Fragments of habitat patches within rural landscapes may play different functional roles depending on their extent, boundary complexity, relative amount of interior-to-edge habitat, and spatial location and connectivity. As an example, remnants of vegetation of relatively large size and regular shape will support more diverse species with interior habitat requirements than would persist in smaller and narrower patches. In addition, the higher the level of spatial distribution and connectivity among remnants the higher the probability that species can successfully disperse among the patches (Lafortezza et al. 2004). The influence of structural and spatial attributes on ecological functions becomes relevant in the case of highly fragmented landscapes where even a single forest tree can provide nesting sites or favourable microclimate conditions.

CULTURAL PATTERNS IN RURAL LANDSCAPES Cultural values & traditions

Structural & spatial attribues

Ecological functions

Land division & settlement

Size

Microclimate modification

Shape Species dispersal and diversity

Applied Science & technology Core area Stewardship & landscape aesthetic

Spatial arrangement

Water and nutrients flows

Integrative planning and design (e.g., FDNP)

Figure17. Conceptual model for analyzing cultural patterns in rural landscapes.

The close interdependency among cultural values and traditions, structuralspatial attributes, and ecological functions is illustrated in Figure 17. Cultural patterns in rural landscapes are the outcome of these complex and mutual interactions. A key question for landscape architects and planners is how to enhance the ecological functionality of rural landscapes while keeping crop production sustainable and economically viable. In their recent paper Lafortezza and Brown (2004), provided a tool (namely FDNP - Framework for the Design of New Patches)

414

R.D. BROWN ET AL.

for incorporating structural and spatial attributes in the landscape planning and design process, with a specific focus on the design of new patches in the rural landscape. The FDNP considers ecological context in recommending size, shape, location, and composition of new habitat patches that have the potential to support ecological function in rural landscapes (Lafortezza and Brown, 2004). The novelty of this approach lies in its ability to analyze cultural patterns in both structural and functional terms: new patches are planned and designed considering the pattern of the neighbouring fragments of vegetation (expressed in terms of number of patches, average size, shape, and core area) and their spatial arrangement in relation to the behavioural aspects of a guild of species (i.e. functional connectivity). Landscape architects and planners can use this approach when designing new elements in relation to the cultural patterns expressed by rural landscapes. 4. SUMMARY AND DISCUSSION In this chapter we have demonstrated how some key cultural factors have changed natural patterns of rural landscapes into pervasive and persistent cultural landscape patterns. These factors include human tendencies for owning, occupying, and controlling land, producing an economic return and improving land, and using the landscape to communicate aesthetic and stewardship values that connect people to their communities. The examples of Italian and Canadian rural landscapes share common cultural factors, yet the results are place-specific. Where Italian landscapes have smaller management units and long associations with hand-made artifacts like stone walls, Canadian landscapes have larger management units and artifacts that are machine-derived (like grass buffer strips). Though the resulting patterns differ, the driving cultural factors are categorically the same. The primeval rural landscapes in Italy and Canada had diverse habitats which accommodated large populations of now-threatened or rare species of plants and animals. The ways in which rural landscapes have been divided, settled, and managed for agricultural production have fragmented ecosystems and increased habitat loss. The remaining natural patterns, associated with small patches and linear features, have dramatically limited inter-patch movements of once-common species. However, understanding the driving factors and their resilience may lead to opportunities for establishing more heterogeneous landscapes made by closely intertwined patterns of both cultural and natural elements. Any intervention intended to have a positive ecological effect should include consideration of land division and settlement, applied science and technology, and stewardship and landscape aesthetic values. Landscape ecology may provide the theory for integrating factors affecting landscape patterns in a unique and holistic framework. We conclude that cultural factors yield long-lasting landscape patterns that must be considered in subsequent planning and design decisions.

CULTURAL PATTERNS IN PLANNING AND DESIGN

415

REFERENCES Corry, R.C. and Nassauer, J.I. (2002). Managing for Small Patch Patterns in Human-dominated Landscapes: Cultural Factors and Corn Belt Agriculture. In J. Liu and W. Taylor (Eds.), Integrating Landscape Ecology into Natural Resource Management (pp. 92-113). Cambridge University Press, Cambridge, Massachusetts. Forman, R.T.T. (1995). Land Mosaics –Ecology of Landscapes and Regions. Cambridge University Press, Cambridge. Forman, R.T.T. and Godron, M. (1981). Patches and structural components for a landscape ecology. BioScience, 31, 733–740. Hart, J. F. (1998). The Rural Landscape. Johns Hopkins University Press, Baltimore, Maryland. Johnson, H.B. (1976). Order upon the Land: The U.S. Rectangular Land Survey and the Upper Mississippi Country. Oxford University Press, New York. Lafortezza, R. and Sanesi, G. (2003). The spatial pattern of the Mediterranean landscape: towards a new definition. Proceedings of the 18th Annual Symposium of US-IALE, Banff, Canada. Lafortezza, R., Sanesi, G., Pace, B., Corry, R.C and Brown R.D. (2004). Planning for the rehabilitation of brownfield sites: a landscape ecological perspective. In A. Donati A., C. Rossi and C.A. Brebbia (Eds.), Brownfield Sites II. Assessment, Rehabilitation and Development (pp. 21-30). WIT Press, Southampton, UK. Lafortezza, R. and Brown, R.D. (2004). A framework for landscape ecological design of new patches in the rural landscape. Environmental Management, 34(4), 461-473. Leal, D.B. (2004). Exploring the Relationship Between Climate and Culture in Apulia, Italy. Master of Landscape Architecture Thesis. University of Guelph. Leopold, A. (1949). A Sand County Almanac, and Sketches Here and There. Oxford Univ. Press, New York. Makhzoumi, J. and Pungetti, G. (1999). Ecological Landscape Design and Planning: The Mediterranean Context. E and FN Spon, London. Naveh, Z. (1995). Interactions of landscapes and cultures. Landscape and Urban Planning, 32, 43-54. Spaling, H. (1995). Analyzing cumulative environmental effects of agricultural land drainage in southern Ontario, Canada. Agriculture Ecosystems and Environment, 53(3), 279-292. Steiner, F. R. (1990). Soil Conservation in the United States: Policy and Planning. Johns Hopkins University Press, Baltimore, Maryland. Tress, B. and Tress, G. (2002). Scenario visualisation for participatory landscape planning - a study from Denmark. Landscape and Urban Planning, 982, 1-18. Turner, M.G. (1989). Landscape ecology: the effect of pattern on process. Ann Rev Ecol Syst, 20, 17–197.

CHAPTER 25

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND T. VAN DER SLUIS1, J. ROMANOWSKI2, J. MATUSZKIEWICZ3, I. BOUWMA1

1

2

ALTERRA Green World Research, The Netherlands; Centre for Ecological Research PAN, Poland; 3Institute of Geography and Spatial Organisation PAN, Poland

Abstract. The Vistula River is considered one of the most valuable rivers in Western and Central Europe. The Vistula floodplain area forms an extensively managed landscape, with high nature values and biodiversity. This riverine landscape is rapidly undergoing changes in land use and spatial developments. The floodplain area from Warsaw to Płock was selected for an assessment of ecological impacts through scenario analysis. Five scenarios were developed which contain elements of expected or possible developments. Effects of each scenario on indicative fauna species were analyzed with the computer model LARCH, an expert system for scenario analysis and policy evaluation. The analysis shows that fragmentation presently does not threaten the favourable conservation status of the species assessed. Most of the species have either nearly sustainable, sustainable or highly sustainable networks. The analyses conducted for the 16 species characteristic for the Vistula valley show potential threats of infrastructure development reflected in scenario 1, “Maximum river regulation and infrastructure development”. Scenario 1 showed pronounced effects on species dependent on steep banks and sandbanks. Scenario 3, “Renaturalisation” showed positive effects for most species, notably species of steep banks and sandbanks. Scenario 5, “Reforestation” showed positive effects on species depending on forest habitat, while species typical for meadows decreased. The designation of the 2 Natura 2000 areas has strongly reduced the options for building dams in the area, as interventions are only allowed if they have no significant effects on the area. An evaluation of developments that affect the biodiversity and spatial cohesion (fragmentation) of habitats is essential to come to balanced developments, taking into account both environmental and societal needs. Stakeholder involvement and scenario modelling should be widely used in the process of decision making for spatial development.

417 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 417-433. © 2007 Springer.

418

T. VAN DER SLUIS ET AL.

1. INTRODUCTION Riverine landscapes are highly dynamic ecosystems, due to their hydrological regime and natural variation in morphology. Rivers have always attracted man, for they offer opportunities for work, transport, and supply of drinking water. Man has had a large impact on the hydrology and the landscape by engineering works for flood protection, boat traffic or hydropower, and urban settlement and agricultural land use, owing to transport opportunities and fertile soils. The Vistula is one of the largest remaining, more or less natural rivers in Central and Western Europe. This riverine landscape is rapidly undergoing changes in land use and spatial developments. For instance, hydro-technical measures aimed at flood control and water transport are presently being considered in the study area. Furthermore, current agricultural land use will change considerably because of the accession of Poland to the EU and the influence of the EU-Common Agricultural Policy. Finally, the study area is under influence of urbanization, due to its close proximity to Warsaw. All these developments are likely to have a large impact on biodiversity. In the Vistula Econet Development and Implementation project (VEDI) a multi-stakeholder planning approach was followed. Stakeholders’ participation in planning is one of the requirements of the Water Framework Directive and requires discussion of development plans at different levels (Jonsson 2005). The Vistula floodplain area from Warsaw to Płock was selected for an assessment of ecological impacts through scenario analysis. With model simulations it is assessed how different scenarios may influence natural values of the Vistula River. The scenario outcomes are used for discussion of riverine development scenarios with the different stakeholders. 2. STUDY AREA 2.1 Description of the vistula valley The study area is located in the centre of Poland (Figure 1). The area comprises a part of the Vistula valley of about 1545sq km, from the border of Warsaw to the dam at Włocławek, a length of 135 km and a width ranging from 20 to 6 km. The Vistula is the biggest river in the Baltic Sea catchment area, which has best retained its natural characteristics in its middle stretch through the Mid-Polish lowlands. More importantly, the Vistula is also the last big European river, which is allowed to meander here more or less undisturbed between its winter dikes. Therefore, The Vistula River has still retained a semi-natural character for most of its length and is considered one of the most valuable rivers in Western and Central Europe. The international recognition of this unique river valley is reflected by the recommendation of Contracted Parties of the Ramsar Convention on Wetlands to conserve the middle reach of Vistula River. The importance for both the national and international ecological networks is reflected in the EECONET- Poland programme and the Natura 2000 network in Poland. Two Natura 2000 areas in this part of the valley have been proposed to the European Commission.

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

419

The Vistula is fed by several tributaries: Narew, Bzura, Skrwa Prawa and Skrwa Lewa. This stretch of the Vistula is characterised by variable flow and water levels. The annual range of extreme floods is up to 8 m. Annual floods are 4-6 m higher than mean flow and usually occur at the end of winter and early spring when snow melts. Lowest flows occur at the end of summer and in early autumn, especially in September. Extreme water levels often occur in June or July, caused by rainstorms. Ice phenomena also play an important role in the water regime of the Vistula. Arable lands, meadows and pastures, gardens and orchards cover over 46% of the area. Farms usually measure 5-20 ha. Main crops are cereals 69 %; potatoes 10 %; industrial crops 5.5 % and the remainder being vegetables. Forests cover 34% of the study area, 69% of which are coniferous forests and 31% broad-leaved forests. Grasslands (mostly extensive meadows) cover 17%. Approximately one third of the study area is protected by existing nature reserves and Kampinos National Park under the following categories: 1) Nature reserves; 2) National Parks: 3) Landscape Parks; 4) Natura 2000 Network; and 5) other categories. The area is part of the Ecological Network of Poland (EECONET- Poland; Figure 2). The area includes core areas and ecological corridors of the national network. The Vistula river basin forms an important ecological corridor connecting southern and northern parts of the Central European Lowland. The Vistula floodplains with its meanders, islands and ox-bows are also a specific habitat for many aquatic and terrestrial species.

Figure 1. The study area and the Vistula River in Poland.

420

T. VAN DER SLUIS ET AL.

Figure 2. Eeconet, the Polish Ecological network (Liro et al. 1995).

2.2 Spatial planning and developments Water management in the Vistula River valley is regulated by the Water Act of 2001. Main institutions involved in water management are: 1) The Regional Board for Water Management (under the Ministry of Environment) and 2) The Provincial Board for Water Purification and Water Installations. Main tasks of the Regional Water Boards are supervision of water resources, and coordination of measures for flood protection and against drought, and measures to reconstruct degraded aquatic ecosystems. The Provincial Boards manage water purification installations and maintain river dikes. Several integrated management plans existed for the Vistula Valley, but none has ever been firmly and systematically implemented by successive governments, scientific and technical advisory units. At this moment, flooding issues are addressed including potential development of two new dams at Płock and Wyszogród (both within this project study area). Additionally proposals to manage the problems that have arisen with the operation of the aging Włocławek Dam are considered at the governmental level. Potential options for mitigating the threats from the Włocławek Dam include: o The decommissioning of the Włocławek Dam and restoration of the Vistula into a free-flowing river. o The modernisation of the existing Włocławek Dam o Construction of a new dam at Nieszawa.

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

421

Several elements of these river management proposals and other potential developments in agriculture, forestry and urbanisation have been incorporated into development scenarios analysed with the model LARCH. 3. ECOLOGICAL NETWORKS To define the ecological network and its functioning an approach has been developed for spatial analysis, based on the theory of metapopulations. The metapopulation theory states that in fragmented landscapes populations of animal species do not live in a continuous habitat but in a network of habitat patches, which are mutually connected by dispersal movements (Levins 1970, Andrén, 1994, Hanski and Gilpin, 1997, Opdam, 2002). Whether an ecological network can sustain a persistent population or not, depends on: o characteristics of a species: habitat preference, home range, dispersal capacity o the amount, shape and area of habitat patches in a landscape o connectivity of the landscape, which defines how easily species can move to other habitat patches (spatial configuration of habitat patches). The network function of a landscape can be tested for different ecosystem types with e.g. landscape ecological models. 4. METHODS 4.1 The model LARCH The assessment of the wildlife populations and population viability in the study area is done with the Landscape Ecological model LARCH (acronym for: Landscape ecological Analysis and Rules for the Configuration of Habitat). LARCH is a tool to visualise the viability of metapopulations in a fragmented environment, whether available (fragments of) habitat are large enough for species to survive. With LARCH, effects of each scenario on indicative fauna species were analyzed and compared with the current situation. LARCH is a spatial model, designed as an expert system, used for scenario analysis and policy evaluation. It assesses the status of wildlife populations, in particular their viability in the larger landscape context. The model has been fully described elsewhere (Pouwels et al., 2002, Groot Bruinderink et al., 2003, Chardon et al., 2000, Verboom et al., 2001, Van der Sluis et al., 2001, 2003, 2004, Van Rooij et al. 2003). LARCH provides information on the metapopulation structure and population viability in relation to habitat distribution and carrying capacity (Figure 3). It assesses spatial cohesion of potential habitat, and provides information on the best ecological corridors in the landscape.

422

T. VAN DER SLUIS ET AL.

Figure 3. Set-up of the model LARCH.

The principles of LARCH are simple: a species is selected, relevant for nature conservation or an indicator species representing a suite of species, to assess the natural areas. The size of a natural area (habitat patch) determines the potential number of individuals of a specific species it can contain. The distance to neighbouring areas determines whether it belongs to a network for the species. The carrying capacity of the network determines whether it can contain a key population (a relatively large, local population in a network, which is persistent under the condition of one immigrant per generation) or a Minimum Viable Population (a population with a probability of exactly 95% to survive 100 years under the assumption of zero immigration). If that is the case, the network population is viable or sustainable for the species (a viable population is defined here as a population with a probability of at least 95% to survive 100 years). Furthermore, the connectivity of the landscape is assessed (the connectivity defines how easily species can move to other habitat patches, and is defined by the spatial configuration of habitat patches) LARCH requires input in the form of habitat data (e.g. a vegetation or land use map) and ecological parameters for indicator species (e.g. home range, dispersal distance, carrying capacity for all habitat types). LARCH parameters are based on literature and empirical studies. Simulations with the dynamic population model METAPHOR have been carried out to validate parameters and standards for the model (Foppen et al., 1999, Verboom et al., 1993, 2001, Vos et al., 2001, 2002, Chardon, 2001). Actual species distribution or abundance data are not required per se since the assessment is based on the potential for the ecological network of a

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

423

species. LARCH models the habitat, and evaluates the network population, the viability of the network population and spatial cohesion. The results are either provided in the form of maps, or as tabular data. 4.2. Development scenarios Five scenarios were developed in consultation with the stakeholders for the Middle Vistula River. The scenarios contain elements of expected or possible developments. Some elements may seem rather extreme, these are however very illustrative of specific consequences. Scenarios are thus tools to be used in development planning and to guide in evaluation of planning choices (Ahern 2002, Bolck et al. 2004). The outcomes of scenario analysis can guide planners and nonprofessional alike, and make the outcome of long term planning processes more transparent. Scenario analysis identifies advantages and disadvantages, while choices must be made by policy-makers, preferably in consultation with stakeholders, and at least informing stakeholders and the general public (Jonsson 2005). The following scenarios were defined: o Scenario 1: Maximum river regulation and infrastructure development. Construction of two new dams; removal of all trees in the river area inside of the dikes; and development of other infrastructure like roads, dikes, motorway etc. o Scenario 2: Medium intensity regulation. Minor river regulation; partial removal of trees; and redirection of the main channel. o Scenario 3: Renaturalisation. Removal of some of the dikes (where possible); removal of the present dam at Włocławek; and removal of some of the settlements in the river valley; succession of managed forests towards more natural ones o Scenario 4: Restoration and protection of meadows and pastures. Supporting small-scale farming through agro-environmental programmes, resulting in a small-scaled landscape. o Scenario 5: Reforestation. Conversion of low-productivity agricultural fields into forest plantations and natural succession of abandoned areas. Detailed maps were prepared for each of the developed scenarios, which outlined the changes on the vegetation in the study area based on experts’ assessments. A review of the LARCH model outputs has shown very limited effects of two scenarios (Scenario 2: Medium intensity regulation and Scenario 4: Restoration and protection of meadows and pastures) and therefore only results for the remaining three scenarios, 1, 3 and 5, are presented. 4.3 Ecosystems and species selection The habitat types selected are considered most indicative for natural river valleys, and thus relevant for evaluation of river management scenarios. The habitat types are aquatic (river, oxbows, and ponds) and terrestrial (riparian forest,

424

T. VAN DER SLUIS ET AL.

meadows, sandy islands and steep riverbanks). The ecosystems that are evaluated, in fact, form the landscape. The situation is assessed for relevant species for each ecosystem. The selected species include some short-range species, e.g. reptiles, amphibians, mammals, which are all vulnerable for fragmentation. In addition, bird species are included in the analysis. In total 16 species were selected, representing five different habitats (Table 1). Table 1. Selected species for LARCH analysis. Priority habitat type

English name

Scientific name

Steep banks

Kingfisher Sand martin Little ringed plover Little tern Great crested newt Banded demoiselle Beaver Bank vole Pine marten Middle spotted woodpecker Elk Black stork Sand lizard Common root vole Large copper Corncrake

Alcedo atthis Riparia riparia Charadrius dubius Sterna albifrons Triturus cristatus Calopteryx splendens Castor fiber Clethrionomys glareolus Martes martes Dendrocopus medius Alces alces Ciconia nigra Lacerta agilis Microtus oeconomus Lycaena dispar Crex crex

Sand banks Semi-aquatic

Forests

Meadows

4.4 Species modelling The selected species were modelled with a set of parameters, based on expert knowledge and, where possible, field data. In practice, it is difficult to find field data of the required area and species, in particular for parameters such as dispersal distance or density; in some cases, data could be used from other regions in Poland, sometimes data from other Western European countries were adjusted for the initial model. If none of these was available, expert judgement was used to estimate the best parameters. The model was prepared using parameters for each species, resulting in a habitat model (distribution map) and population figures. The outcome of the model was crosschecked with the species expert, and, where possible, with existing distribution maps. If the results were not in accordance with the current distribution, the parameters were adjusted in discussion with the species expert. In an iterative process, the model was adjusted, until the model result was satisfactory. Once the model was considered reliable, the model was run with the scenario maps.

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

425

5. RESULTS At present, the area is not heavily fragmented, and therefore most of the species assessed have a favourable conservation status. Most of the species populations are nearly sustainable, sustainable or highly sustainable, and the networks hold key populations or even Minimum Viable Populations, as illustrated for the Pine marten (Figure 4). Scenario 1, “Maximum river regulation and infrastructure development”, has a pronounced effect in comparison to the present situation, in particular on the species dependent on steep banks and sandbanks (respectively SB and SA in Figure 5). These habitat types are small in area, and form a fragmented pattern along the river. The loss of these habitats not only results in a strong decrease in numbers, it also changes the viability of these small populations. Except for the Sand martin, the viability of the populations pass threshold levels, i.e. the longterm viability is not guaranteed anymore for the Little tern, Little ringed plover and the Kingfisher. Species of semi-aquatic and forest habitats are to a limited extent affected. There is a decrease in population number but not in viability of the overall population. The most pronounced decrease occurs for the Beaver due to destruction of its habitats: from one large population it changes into small and fragmented local populations (Figure 6). However, the population is not endangered because there remains sufficient habitat, also away from the river.

Figure 4. Potential local populations of the Pine marten in the Vistula Valley based on LARCH modelling.

426

T. VAN DER SLUIS ET AL.

Although the changes in numbers may not seem more pronounced than for other species (Figure 5), at population level this decrease can have a large impact on viability due to specific threshold levels related to local extinction patterns. Effects on other species and habitats seem negligible at the regional scale, although locally changes will occur where e.g., riparian forests disappear or new roads are constructed. Scenario 3, “Renaturalisation”, affects most of the selected species positively, especially the species of steep banks and sandbanks (Figure 5). The connectivity increases much for a species like the Little ringed plover and due to removal of the dam and consequently formation of new islands, this part of the Vistula river would form a network. The Little tern population becomes highly sustainable due to these changes. In addition, semi-aquatic and forest species benefit. Several of them like the Great crested newt, Beaver (Figure 6), Pine marten and Middle spotted woodpecker become highly sustainable in the whole area, which form mostly Minimum Viable Populations. However, the species of meadows show a more ambivalent picture: some benefit from the “Renaturalisation” but species, like the Corncrake (Figure 7), depending on extensively used grasslands, and Sand lizard, a species typical of open, well-structured vegetation, decline. Scenario 5 “Reforestation” logically positively influences the species dependent on forest habitat (Figure 5). A positive effect is in particular noted for the Black stork, which shows a potential increase in population size of approximately 20 % (resulting in a Minimum Viable Population). Some of the species typical for meadows like Sand lizard, Corncrake (Figure 7) and Large copper show a decrease because of the conversion of arable fields and grasslands into forest. Except for the change in population number and viability, also the connectivity of the landscape was assessed. Scenario 1 „Maximum river regulation” has the strongest effect on the Little ringed plover from all analysed scenarios. The population is reduced by 50% as compared to the present situation, due to the disappearance of almost all islands, which constitute the major breeding sites of the species. The consequences of the development of two reservoirs are a further fragmentation of the species range and reduction of connectivity (Figure 8). The remaining three scenarios have slightly positive effects (Table 2). Scenario 3 „Renaturalisation”, leads to an increase in range and number of the Little ringed plover populations due to new nesting sites on newly formed (restored) islands on the section of river where the Włocławek reservoir is liquidated. The improved connectivity covers all of the Vistula River within the study area (Figure 8). Formerly fragmented populations are well connected under the conditions of Scenario 3. 6. DISCUSSION The study area is part of the Ecological Network EECONET – Poland (Liro et al., 1995); it ensures connection between core areas (Kampinos National Park) via ecological corridors (Vistula floodplain) to the remaining National and European Network.

le

Species

ne

(F o)

sto rk (F o) co pp er (M e) Ro ot vo le (M e) Co rn cr ak e ( M Sa e) nd liz ar d (M e)

Bl ac k

)

Fo )

k( Fo

(F o)

Aq )

n(

El

ar te

ck er

m

wo od pe

Pi

La rg e

sp ot te d

vo le

q)

Aq )

(S a)

(A

Be av er (

lle

ne wt (

ois e

d

Ba nk

de m

re ste

)

B)

(S a)

te rn

er

(S

M idd

Species

idd

Ba nd ed

le

plo v

Lit t

r in ge d

r( SB

ar t in

Relative change (%)

le

Pi ne m

El ar te

k( Fo )

n( wo Fo od ) pe ck er (F o) Bl ac ks to rk La (F rg o) e co pp er (M e) Ro ot vo le (M e) Co rn cr ak e ( M Sa e) nd liz ar d (M e)

sp ot te d

ng f is he r( SB Sa nd ) m Lit ar tle t in r in (S ge B) d plo ve r( Sa ) Lit tle Gr ea te rn tc (S re ste a) d Ba ne nd wt ed ( A de q) m ois el le (A q) Be av er (A q) Ba nk vo le (F o)

Ki

Relative change (%)

M i dd

Species

M

le

m

ng f is he

Sa nd

Gr ea tc

Lit t

Ki

le sp ot te d

ne m

El ar te

k( Fo )

n( Fo od ) pe ck er (F o) Bl ac ks to rk La (F rg o) e co pp er (M e) Ro ot vo le (M e) Co rn cr ak e ( M Sa e) nd liz ar d (M e) wo

Pi

ng fis he r( SB Sa nd ) m Lit ar tle t in r in (S ge B) d plo ve r( Sa ) Lit tle Gr ea te rn tc re (S ste a) d Ba ne nd wt ed ( A de q) m ois el le (A q) Be av er (A q) Ba nk vo le (F o)

Ki

Relative change (%)

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

Present

Present

Present

427

The results of the study indicate that at present most of the analysed species form thriving local populations arranged in sustainable networks (for details see Romanowski et al. 2005).

120

100

80

60

40

20

0

200 Scenario 1

180

160

140

120

100

80

60

40

20

0

Scenario 3

140

120

100

80

60

40

20

0

Scenario 5

Figure 5. Relative change of population size (number of reproductive units) for scenario 1 (maximum regulation), 3 (renaturalisation) and 5 (reforestation).

428

T. VAN DER SLUIS ET AL.

The LARCH spatial cohesion analysis shows a good connectivity of habitats for forest and semi aquatic species like the Middle spotted woodpecker, Bank vole and Beaver. Although species are analysed, these should be regarded as exemplary for a group or guild of species. It may seem less important whether a species like a Little ringed plover shows a decline as a result of certain scenario’s, more importantly it shows that the group of ‘middle-sized birds’, dependent on islands and sandbars, is likely to decline. Furthermore, it should be stressed that the model is mostly relevant to compare the effects of different scenarios – it indicates the relative effects of the scenarios. Although the model does provide population numbers, they should not be seen as an exact prediction of populations’ numbers if one of the scenarios is realised. One of the strengths of the model and scenario approach used is that the meaning of the results exceeds the species level; it shows how ecosystems can be affected. The results give insight in the changes at population- and metapopulation level. Not only is shown the loss in habitat or perhaps a decline in numbers, but this is translated to an overall impact of the population of species: comes a species in a ‘danger zone’ and does this affect the population viability, and therefore the longterm biodiversity. The scenarios developed are hypothetical scenarios and meant as an illustration. Still several elements of scenario 1, ‘maximum regulation’, are considered by the Polish Regional Board for Water Management and the Ministry of Infrastructure. The results point out that this scenario is not likely to be implemented due to the impacts it will have on the Natura 2000 sites, which is not in line with (European) regulations. Scenario 3, ‘renaturalisation’, may be preferred by conservation NGOs. Local acceptance may be low, and in addition, many funds would be required to compensate land and property owners in the Vistula River Valley, funds that are currently not available. Scenario 5, ‘reforestation’, is likely to occur to some extent. Even if the current afforestation programme is more limited in extent, with expected changes in agriculture more marginal land will be abandoned over time and natural encroachment will take place. At low costs, there may still be considerable gain for several wildlife species. A factor, which is hard to assess, is how much control there will be on spatial developments and in particular the urbanisation. These scenario’s show the need of proper, integrated planning, to avoid the spread of built-up areas all over the Vistula Valley, which will be detrimental for the natural values of the area, but also costly in regard of required infrastructure development. The Advisory Group accepted the generated model results. Acceptance is usually dependent on the factors: confidence in practical applications, confidence in the people involved in or providing materials, social characteristics of the participants and the communication of the data (Olsson and Berg 2005). In particular, the social characteristics of the stakeholders involved have been of importance in this study.

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

429

Figure 6. Potential local populations of the Beaver in the Vistula Valley near Płock; present situation, scenario 1 (maximum regulation) and 3 (renaturalisation).

430

T. VAN DER SLUIS ET AL.

Figure 7. Potential local populations of the Corncrake in the Vistula Valley, near Płock; present situation and scenario 3 (renaturalisation).

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

431

Figure 8. Spatial cohesion of the Little ringed plover habitats for scenario 1 (maximum regulation) and scenario 3 (renaturalisation).

7. CONCLUSIONS The analyses conducted for the 16 species characteristic for the Vistula valley show potential threats of infrastructure development reflected in scenario 1: “Maximum river regulation and infrastructure development” (Romanowski et al. 2005). The results provide additional indications of the effects of possible measures on two Natura 2000 areas designated in the study area. The changes in the river regulation (scenario 1) will have a profound effect on the bird species of the Dolina Środkowej Wisły - the Middle Vistula valley. The effects of scenario 1 on both Natura 2000 habitats and species of the Kampinos National Park are limited due to existing protection of this large area. It is likely that plans for large-scale water regulation will conflict with the protection of the Middle Vistula valley area. The designation

432

T. VAN DER SLUIS ET AL.

of the Natura 2000 site has strongly reduced the options for building dams in the area, as interventions are only allowed if they have no significant effects on the area. An evaluation of developments that affect the biodiversity and spatial cohesion (fragmentation) of habitats is essential to come to balanced developments, taking into account both environmental and societal needs. Stakeholder involvement and scenario modelling should be widely used in the process of decision making for spatial development (participatory spatial planning), in particular in Poland where many changes are likely to occur in the near future. Ecological modelling is a powerful tool in the decision making process. The discussions with stakeholders in the VEDI-project showed that the LARCH model is a practical tool, especially on a regional and local scale for the evaluation of ecological consequences of developments in the Vistula valley. It can illustrate clearly spatial changes in wildlife populations because of habitat changes. Furthermore, the approach proved to be very useful to give insight in the consequences, also to e.g. water managers, which was also demonstrated in other studies (Jonsson 2005). Choices are made more explicit, and the modelling results facilitate the discussion of different scenarios, from different disciplines views are exchanged on spatial developments. It is recommended that modelling should be applied in the decision making process regarding all spatial developments that concern Natura 2000 areas, also as a proper evaluation tool to meet EU-requirements in regard of Environmental Impact Assessments. REFERENCES Ahern, J. (2002). Greenways as Strategic Landscape Planning: Theory and Application. PhD-Thesis, Wageningen, Netherlands. Andrén, H. (1996). Population responses to habitat fragmentation: statistical power and the random sample hypothesis. Oikos, 76, 235-242. Bolck, M. de Togni, G., van de Sluis, T. and Jongman, R. (2004). From models to reality: design and implementation process. In R. Jongman and G. Pungetti (Eds.), Ecological Networks, and Greenways – Concept, Design, Implementation (pp. 128-150). Cambridge studies in Landscape Ecology. Chardon J.P., Foppen R.P.B. and Geilen N. (2000). LARCH-RIVER, a method to assess the functioning of rivers as ecological networks. European Water Management, 3 (6), 35-43. Foppen, R, Geilen, N. and Van der Sluis, T. (1999). Towards a Coherent Habitat Network for the Rhine. IBN-research report 99/1, ISSN: 0928-6896 Groot Bruinderink, G.W.T.A., van der Sluis T., Lammertsma D.R. and Opdam P. (2003). The design of a tentative, coherent ecological network for large mammals in Northwest Europe. Conservation Biology, 17 (2), 549-557. Hanski, I. and Gilpin M.E. (Eds.). (1997). Metapopulation Biology. Academic Press, London. Jonsson, A. (2005). Public participation in water resources management: stakeholder voices on degree, scale, potential and methods in future water management. Ambio, 34 (7), 495-500. Levins, R. (1970). Extinction. In M. Gerstenhaber (Ed.). Some Mathematical Problems in Biology (pp. 77-107). American Mathematical Society, Providence. Liro, A., Głowacka, I., Jakubowski, W., Kaftan, J., Matuszkiewicz, A.J. and Szacki, J. (1995). National Ecological Network EECONET-Poland (pp. 1-66). IUCN Poland, Warsaw. Matuszkiewicz, J.M. and Solon, J. (1998). Charakterystyka zróżnicowania typologiczno-przestrzennego roślinności rzeczywistej oraz rozpoznanie specyficznych siedlisk i ekosystemów. In Matuszkiewicz J.M. (Ed.) Przyrodnicze podstawy opracowania optymalnej koncepcji zagospodarowania obszaru doliny dolnej Wisły na odcinku od ujścuia Narwi do dolnego stanowiska poniżej zapory we

COMPARISON OF SCENARIOS FOR THE VISTULA RIVER, POLAND

433

Włocławku. niepublikowane opracowanie dla Okręgowej Dyrekcji Gospodarki Wodnej w Warszawie, Warszawa. Matuszkiewicz, J., Romanowski, J., Kowalska, A. and Kowalczyk, K. (2005). Description of the Vistula Valley. In van der Sluis, T. (Ed.), (2005). Evaluation of Ecological Consequences of Development Scenarios for the Vistula River (pp. 1-127). Wageningen, Utrech, Warsaw. Olsson, J.A. and Berg, K. (2005). Local stakeholders’ acceptance of model-generated data used as a communication tool in water management: the Rönneå study. Ambio, 34 (7), 507-512. Opdam, P. (2002). Assessing the conservation potential of habitat networks. In K.J. Gutzwiller (Ed.), Concepts and Application of Landscape Ecology in Biological Conservation: Integrating the Metapopulation Concept into Biological Conservation (pp. 381-404), Springer Verlag, New York. Pouwels, R., Jochem, R., Reijnen, M.J.S.M., Hensen, S.R. and Greft, J.v.d., (2002). LARCH for Spatial Ecological Assessments of Landscapes (in Dutch). ALTERRA Report 492. Wageningen, the Netherlands. Romanowski, J., Matuszkiewiecz, J., Kowalczyk, K., Kowalska, A., Kozłowska, A., Solon, J., Bouwma, I.M., Middendorp, H., Reijnen, R., Rozemeijer, R. and T. van der Sluis (Ed). (2005). Evaluation of Ecological Consequences of Development Scenarios for the Vistula River (pp. 1-127). Warsaw, Wageningen, Utrecht. Van der Sluis, T., van Rooij, S.A.M. and Geilen, N. (2001). Meuse-Econet. Ecological Networks in Flood-Protection Scenarios: A Case Study for the River Meuse. Intermeuse-report no. 3, IRMASponge nr. 9. ALTERRA, Wageningen, the Netherlands. 59 p. Van der Sluis, T., Baveco, H., Corridore, G., Kuipers, H., Knauer, F., Pedroli, B. and Dirksen, J. (2003). Networks for Life - An Ecological Network Analysis for the Brown Bear (Ursus arctos) - and Indicator Species in Regione Abruzzo. ALTERRA Report nr. 697. ALTERRA, Green World Research. Wageningen. Van der Sluis, T. and Pedroli, B. (2004). Ecological Network Analysis for Umbria (Italy). RERU, Rete Ecological Delle Regione Umbria. ALTERRA report 1013, Wageningen, the Netherlands. Van Rooij, S.A.M., Steingröver, E.G. and Opdam, P.F.M. (2003). Corridors for Life. Scenario Development of an Ecological Network in Cheshire County. ALTERRA Report nr. 699. ALTERRA Green World Research. Wageningen. Verboom, J., Metz, J.A.J. and Meelis, E. (1993). Metapopulation models for impact assessment of fragmentation. In Vos C.C. and Opdam P.F.M. (Eds.). Landscape Ecology of a Stressed Environment (pp. 172-191). IALE studies in Landscape Ecology 1. Chapman and Hall, London. Verboom J., Foppen R., Chardon P., Opdam P. and Luttikhuizen, P. (2001). Introducing the key patch approach for ecological networks with persistent populations: an example for marshland birds. Biological conservation, 100 (1), 89-101. Vos, C.C., J. Verboom, Opdam, P.F.M. and Ter Braak, C.J.F. (2001). Toward Ecologically Scaled Landscape Indices. The American Naturalist, 183, 24-41 Vos, C.C., Baveco, H. and Grashof-Bokdam, C.J. (2002). Corridors and species dispersal. In Gutzwiller, K.J. (Ed.), Concepts and Application of Landscape Ecology in Biological Conservation, Springer Verlag, New York.

CHAPTER 26

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN Toward practical planning

K. NAGASHIMA Kyushu University, Fukuoka, Japan

Abstract. This paper reviews the essential planning process of green space design based on the concept of ecosystem management together with the recent technical trends and required data set applicable at each process. The techniques reviewed in this paper contribute toward the accomplishment of the goals of protecting the entire spectrum of biodiversity, ensure their persistence, and accommodate human use within these constraints; these are the three main goals of ecosystem management. The future researches to be carried out for supporting practical green space planning have also been discussed.

1. INTRODUCTION In response to the current increase in the replacement of natural ecosystems and the deepening biodiversity crisis, “ecosystem management approach” has been proposed as a new “paradigm” of management and improved framework for the long-term protection of resources (Christensen et al., 1996; Cortner and Moote, 1999; Brody, 2003). In this approach, entire ecological systems and their ecological processes form the framework for the management efforts. Within the overall goal of sustaining ecological integrity, Grumbine (1994) defined five specific goals by reviewing papers published on ecosystem management: (1) represent all native ecosystem types across their natural range of variations, (2) maintain viable populations of all native species in situ, (3) maintain the evolutionary and ecological processes, (4) manage over periods of time long enough to maintain the evolutionary potential of species and ecosystems; and (5) accommodate human use and occupancy within these constraints. Although the fifth goal acknowledges that the vital role will have to be played by humans, the first four goals are derived from current scientific knowledge that aims to reduce the biodiversity crisis and can be 435 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 435–457. © 2007 Springer.

436

K. NAGASHIMA

integrated into two conservation objectives, namely, representativeness and persistence. In order to meet these two goals and eventually provide scientific bases for implementing ecosystem management, the emphasis of recent scientific researches has been on developing a theory and techniques to identify areas of high conservation interest and to design efficient networks of natural reserves. The techniques of gap analysis and site-selection algorithms are the representative achievements employed by these researches. Although the abovementioned scientific approaches and techniques have been well established and have a great potential to transform the manner of allocation and protecting land for conservation, only a limited amount of this knowledge has been applied to practical spatial planning (Prendergast et al., 1999; Cabeza and Moilanen, 2001; Opdam et al., 2002). In other words, most of the proposed approaches on reserve design remain theoretical (Prendergast et al., 1999). Several technical issues such as the difficulty in handling the complexity of land ownership (Prendergast et al., 1999) were cited as reasons for not applying these approaches. However, the communication gap between scientists and land-use managers was cited as the main problem (Prendergast et al., 1999; Cabeza and Moilanen, 2001). The research results are scattered and fragmented throughout the literature, and most land-use managers have little opportunity to access this literature (Prendergast et al., 1999). Therefore, land-use managers do not know how science can contribute to conservation planning. Scientists are unaware of several management goals and constraints; this leads to a wide gap between the theoretically selected set of reserves and the practically optimal set of reserves (Cabeza and Moilanen, 2001). This study aims to contribute in bridging this gap between land-use managers and scientists by collating the fragmented scientific information and by reviewing the recent techniques and methods with regard to green space design together with other required data. The entire information is presented based on the planning processes essential for implementing the ecosystem management. A recent paper published by Margules and Pressey (2000) also reviewed the conservation planning process and scientific techniques, and they elaborated on the technical trends of conservation planning. However, they did not discuss this process in relation to ecosystem management, and the process lacked the steps necessary for consensus building that is crucial for ecosystem management. This review builds on and extends the review of Margules and Pressey (2000) by revising the process based on the concept of ecosystem management, and it includes more information and recent discussions on data collection and analytical techniques. Here, the term “green space(s)” is used instead of “natural reserves” because habitat restoration and off-reserve systems may be important strategies for conservation. The drawbacks of the current techniques presented in this paper might aid in identifying the future direction of research in ecology or landscape ecology that supports practical planning based on the concept of ecosystem management.

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

437

2. PLANNING PROCESSES FOR IMPLEMENTING ECOSYSTEM MANAGEMENT Ecosystem management is an approach that integrates the scientific knowledge of ecological relationships within a complex socio-political framework in order to restore and sustain the health, productivity, and biodiversity of ecosystems along with the overall quality of life. As mentioned earlier, its specific goals are protecting the entire spectrum of biodiversity (representativeness), promoting the long-term survival of species and ecosystems by maintaining the evolutionary and ecological processes (persistence), and including human use within these constraints. Although scientific knowledge related to the first two goals have been accumulated, it cannot be considered accurate because of the complexity of the ecosystem and biodiversity and because the impracticability of investigating the species distribution across the entire planet. Hence, scientific knowledge can be considered provisional. Ecosystem management, therefore, applies the concept of adaptive management that considers “management” as a learning process that enables the revision of the management plan by incorporating the new information in the previous management results. This emphasizes the importance of the monitoring and assessment of the management results. The simultaneous quantitative evaluation of the management plan implies the identification of specific and quantitative targets; consequently, the collection of quantitative data becomes the key process for implementing an ecosystem management. Ecosystem management is a human boundary-spanning problem (Grumbine, 1994; Szaro et al., 1998; Brody, 2003). Ecological boundaries do not coincide with ownership boundaries, and it often extends across them. Thus, in order to incorporate human requirements and values into the management plan, the involvement of various government agencies, private owners, and diverse stakeholders in both public and private lands is necessary. Consensus building and cooperative determinations by administrative interactions were absent in the traditional resource management system (Szaro et al., 1998); this could be designated as a distinguished component of the ecosystem management. Ecosystem management does not identify one method as the correct approach to manage resources. Instead, it develops better options and sustainable solutions by combining human requirements and values with our best understanding of the environment (Szaro et al., 1998). However, the abovementioned concepts of ecosystem management suggest that data collection, identification of specific targets, consensus building, and monitoring and assessment are the important processes for implementing the ecosystem management (for more information on the concept of ecosystem management see Grumbine, 1994; Szaro et al., 1998; Kakizawa, 2000; Brody, 2003). Although the term ecosystem management has not been used, several studies on spatial planning or reserve design have discussed the planning processes from a scientific perspective in order to accomplish the same scientific goals, i.e., representativeness and persistence (e.g., Margules and Pressey, 2000; Opdam et al., 2002). The first step in the planning processes indicated by these studies is defining

438

K. NAGASHIMA

the problem, and this requires clearly defined goals and quantitative measures for assessing the current design of green spaces. After the existing green spaces have been reviewed and the difficulties in achieving the targets have been identified, possible solutions must be discussed. Additional reserve selection and the identification of restoration areas might be included in this step. Several options might be proposed at the decision-making stage. When the most appropriate option has been decided, the plan moves to the implementation stage. The management result will be monitored and evaluated regularly; this creates an ongoing feedback loop of useful information. Collating the processes based on the concept of ecosystem management and the studies on spatial planning, the essential processes for green space design may be elaborated as follows: (1) data collection and identification of management targets, (2) assessment of existing green spaces and defining the problems, (3) proposing possible solutions, (4) decision making by involving all stakeholders, (5) implementing the conservation actions, and (6) monitoring and assessment. This process might not be unidirectional. There will be many feedbacks and reasons for revising the green space management plan. For example, if it was difficult to build a consensus on one particular idea among the various options, another set of options that consider the requirements of stakeholders must be re-examined, thus indicating a feedback from step 4 to step 3. If unpredictable problems were encountered, the process must be restarted and data collection and target identification must be conducted again. Further, new scientific knowledge has the potential to revert the process to the previous steps. Recent studies on identifying areas or designing efficient networks of high conservation priorities discuss the techniques and tasks related to most of the steps in the process; each of which is reviewed below. The review might be useful in aiding not only land-use managers to understand how science can contribute to practical planning but also to students studying ecology or landscape ecology to understand the research trends in green space design. 3. DATA COLLECTION AND TARGET IDENTIFICATION Since every conceivable item within an ecosystem cannot be measured, surrogates have to be used for measuring the biodiversity as well as the ecological processes that ensure persistence. The conditions for appropriate surrogates summarized by Carignan and Villard (2002) are as follows: (1) provide an early warning of natural responses to environmental impacts, (2) directly indicate the cause of change rather than only the existence of the change, (3) provide a continuous assessment over a wide range and intensity of stresses, which allows the detection of numerous impacts on the ecosystem, and (4) ensure that the surrogates are cost-effective and can be accurately estimated by all personnel involved in the monitoring. Valuable indicators may possess some or all of these characteristics. For practical use, targets must be as specific and quantitative as possible; this requires an understanding of the threshold of each surrogate, thereby ensuring the persistence and representativeness of biodiversity. The quantitative targets,

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

439

therefore, might be reflected in the available data, or conversely, the determined target might require additional data collection. 3.1 Measurement and target setting for biodiversity conservation A common surrogate to measure biodiversity is the indicator taxon; this surrogate assumes that the areas occupied by many species of a well-studied indicator taxon coincide with a species-rich area for other taxa (Schall and Pianka, 1978; Landres et al., 1988; Pearson and Cassola, 1992; Prendergast et al., 1999; Carignan and Villard, 2002). In practice, the available distribution data is often limited to this type of a well-studied taxon, which leads to its frequent use as a surrogate. Although the tendency that supports this assumption is observable at the global or continental scale (Buzas, 1972; Reid, 1998; Carignan and Villard, 2002), the results of few recent studies support this assumption at finer scales (e.g., Kremen, 1992; Prendergast et al., 1993; Pimm and Lawton, 1998; Allen et al., 2001). Prendergast et al. (1993) examined the extent to which hotspots overlap in the case of five groups in Britain, namely, butterflies, dragonflies, liverworts, aquatic plants, and birds. The maximum overlap observed between butterflies and dragonflies was only 34%, and the association was even weaker for the other pairwise comparisons. Further, Kremen (1992) revealed that butterflies couldn’t be indicators for plant species richness. The results of the studies investigating whether or not the sets of complementary sites for different taxa overlap were also not encouraging. By using the data obtained from more than 9000 species of wellstudied birds, butterflies, mammals, and vascular plants and from lesser-known termites, antlions, and two types of beetles mapped onto grid squares that were 25 km by 25 km in South Africa, the pairwise comparisons of the complementary sets reveal a mean overlap of only 10% of cells with a maximum overlap of 21% (van Jaarsveld et al., 1998). In multiple comparisons, no grid square was common to all taxa. However, an identical study conducted in Uganda gave more promising results (Howard et al., 1998). Similar to earlier studies, a low spatial congruence in the species richness of woody plants, large moths, butterflies, birds, and small mammals was found. Nonetheless, the complimentary sets of priority forests selected using the data on a single taxon often captured the species richness in other groups with the same efficiency as that captured by using the data on all taxa at once. In Uganda, different taxa exhibit similar biogeography; therefore, priority forests for a single taxon collectively represent the important forest types for the other taxa. If hotspots or sets of complementary sites for different taxa show little concordance, the same results are expected in case the umbrella species and flag species are used as surrogates; in fact, this is indeed true in the case of flag species. A flag species is normally a charismatic large vertebrate. An entire conservation campaign can be based on it because it arouses public interest and sympathy. However, it need not be a good indicator (Simberloff, 1998). The areas specifically proposed for conserving the Florida panther, a species that has become the symbol of an entire conservation campaign in Florida, included at least 24 of the 51 threatened taxa, but the conservation areas for other species such as the Florida

440

K. NAGASHIMA

black bear were better indicators (Simberloff, 1998). Williams et al. (2000) also found that areas selected using flagships was not a better indicator of biodiversity than the same number of areas selected randomly. The umbrella species is a species with such demanding habitat and large area requirements; therefore, ensuring its conservation is considered to be effective in protecting many other species (Simberloff, 1998). Although some authors thought that this concept might be useful for conservation planning (Wilcove, 1993; Fleishman et al., 2000; Suter et al., 2002), several studies showed that the scope of this concept was limited or ineffective (Berger, 1997; Poiani et al, 2001; Rubinoff, 2001; Bifolchi and Lode, 2005). Suter et al. (2002) tested the usefulness of Capercaillie as an umbrella species in the Swiss Prealps by analyzing the relationships between Capercaillie occurrence and avian biodiversity. Study plots with Capercaillie did not possess significantly higher bird diversity than plots without it. However, the species richness and abundance of birds that are more or less restricted to the subalpine forests and that at the same time are on the red list was considerably higher in plots with Capercaillie than in those without it. The author maintains that the concept of using umbrella species is effective, but it is restricted to species with identical habitat requirements, which conflicts with the assumption drawn based on the concept of umbrella species, i.e., the requirements of a single species can encompass those of an entire ecosystem. Martikainen et al. (1998) also indicated that the assumed umbrella species and the target species for conservation must share similar ecological requirements, despite the fact that a large area theoretically provides habitats for a larger number of species. Based on these studies, Bifolchi and Lode (2005) argued to review the umbrella species concept. Results of several other studies of umbrella species recommended the use of multiple species or functional groups as more comprehensive alternatives to the single-species approach in order to overcome this drawback (Wallis de Vries, 1995; Lambeck, 1997; Poiani, 2001). Another frequently cited criterion for site selection by conservationists is the existence of rare species; this criterion is used assuming that areas with rare species and high biodiversity can be protected simultaneously (Noss, 1990; Pearson and Cassola, 1992; Prendergast et al., 1993; Williams et al., 1996; Prendergast et al., 1999). However, these assumptions appear insupportable because a sufficient number of rare species do not exist in most of the species-rich areas (Prendergast et al., 1993; Reid, 1998; van Jaarsveld et al., 1998; Carignan and Villard, 2002). Hence, most studies caution against the strategy that is solely based on rarity. On the other hand, some studies recommend selecting hotspots of rarity to provide high priority for site selection because sites comprising species with narrow distributions must be included in a protected area network in order to cover all the species (Csuti et al., 1997; Reid, 1998). As the small overlap between the species-based indicators and high biodiversity areas revealed, several studies proposed the utilization of higher-taxon richness as indicators of biodiversity. Williams and Gaston (1994) reported that family richness was a good predictor of species richness for a variety of groups and regions, including British ferns and British butterflies in grid squares measuring 100 km by 100 km, Australian passerine birds in 5° by 5° grid squares and 10° by 10° grid

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

441

squares, and North and Central American bats in grid squares of ca. 611,000 km2. Reid (1998) also indicated the possibility of higher-taxon richness to serve as surrogates of species richness by introducing studies that proved the reliability of genera and family richness both at a continental scale and a finer scale (18 30,000 ha) (Balmford et al., 1996a and b in Reid, 1998). In contrast, van Jaarsveld et al. (1998) stated that genera and families show little overlap with species-based complementary sets across the taxa. Pimm and Lawton (1998) also declared that genus and family data couldn’t result in efficient species-level conservation. The discussion of the utility of high-taxon surrogates continues to be a controversial issue and conclusions cannot be drawn based on it. Further, ecologists challenged the use of habitat or environmental diversity as surrogates for measuring biodiversity. The use of characteristics of the physical environment as predictors of the overall biodiversity appears attractive because data for physical variables may already be available or may be relatively inexpensive to collect (Williams and Gaston, 1994). It is also considered that habitat diversity might be able to integrate a greater number of ecological processes that contribute to the maintenance of ecosystem function although this issue is being actively debated (Margules and Pressey, 2000). According to the rationale that the area representing all environmental surrogate classes will cover the entire range of regional environmental variation, and therefore, might contain all the species found in the region, surrogates such as climatic and physiographic variables (Belbin, 1993), landform-vegetation classes (Awimbo et al., 1996), land systems (Pressey and Nicholls, 1989b) and land classes (Pressey and Logan, 1995), physio-chemical variables (Faith and Norris, 1989), and environmental domains (Bedward et al., 1992) have been utilized. However, examining its usefulness as a surrogate for biodiversity has just begun. Wessels et al. (1999) investigated the informativeness of land facets as biodiversity surrogates at a local scale in South Africa by testing whether the land facets represent different bird and dung beetle communities. The results adequately represent distinct bird and dung beetle assemblages, which support the assumption of environmental surrogates. On the other hand, a study in Europe investigating whether or not an area selected by environmental diversity would represent biodiversity revealed that the approach was effective only for plants and not for birds, mammals, amphibians, and reptiles (Araujo and Humphries, 2001). In addition, some authors state that if the aim of conservation is maximizing the number of species to protect, sites on ecotones between biogeographical units may host a greater number of species (Brown, 1991 and Prendergast, 1994 in Prendergast et al., 1999). The conflicting results from a large body of literature assessing the utility of each of the abovementioned surrogates are attributable to the differences in the analytical methods, geographical scales, and biogeographical histories. Understanding how these factors affect the concordance of areas selected by using indicators and areas with high biodiversity for reliable generalization is still under development (Margules and Pressey, 2000). Inventory efforts to improve data quality that avoid biases toward particular taxa or sites are other important issues to be considered while examining appropriate surrogates for biodiversity measurement. However, the

442

K. NAGASHIMA

studies introduced here contributed to the identification of constraints of the currently available surrogates. In most cases, the decision regarding the use of a particular surrogate depends on the available data. Nonetheless, a combination of surrogates, for instance, simultaneously using climate, geographical variables, vegetation types, and rare species, seems to be the most practical for biodiversity measurement (Margules and Pressey, 2000) since the shortcoming of each surrogate would be compensated. The goal of representativeness is to protect the entire spectrum of biodiversity; this suggests the sampling of a complete multivariate pattern that is generated by a classification of all the areas. The established targets for representativeness diversified as the methods for biodiversity measurement were being developed. Most exercises set targets based on the occurrence of species (e.g., Sarakinos, 2001). Targets based on the spatial extent of communities, environmental classes, and habitat types have also been set (Margules and Pressey, 2000). The recent improvement in computerized approaches allows the selection of sets of green spaces that meet several targets; therefore, increasing attention is paid to the use of multicriteria targets (Cabeza and Moilanen, 2001). 3.2 Data collection and target setting for protecting ecological processes those ensure persistence Measurement of the ecological processes and their target setting are more problematic. Green space design is a spatial practice; therefore, the measurement and protection of ecological processes should be based on spatial surrogates rather than the processes themselves (Margules and Pressey, 2000). In other words, knowledge about the relationship between the landscape patterns and ecological processes is required, which is the underlying concept of landscape ecology (Opdam et al., 2002). In landscape ecology, maps are assumed to be the simplified reflections of landscape functioning, and many studies have focused on the analysis or description of spatial geometry by using several indices. Nevertheless, this assumption was rarely tested and several recent studies revealed that the indices describing the landscape pattern indicated a weak relationship with ecological processes (Schumaker, 1996; Vos et al., 2001; Opdam et al., 2002). Further, it is considered that most of the contemporary work on landscape patterns has failed to provide an understanding of its significance or elucidate the meaning of the patterns (Hains-Young, 1999; Opdam et al., 2002). On the other hand, Margules and Pressey (2000) cited seven important theories on ecological and evolutionary processes: biogeographical theory, metapopulation dynamics, source-pool effects and successional pathways, spatial autecological requirements, source-sink population structures, effects of habitat modification, and species as evolutionary units. These theories may provide guidelines for target setting and measurement for ecological processes. Each of these seven theories is presented below by summarizing Margules and Pressey (2000). Necessary data and related information are also added.

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

443

Biogeographical theory: The equilibrium theory with regard to island biogeography and the associated biogeographical theory, which indicated that bigger, closer, and circular patches of green spaces are better, (Margules and Pressey, 2000), indicated the importance of measuring the size, shape, and connectivity of the patches. Metapopulation theory: This theory claims to consider the natural ranges of species so that some populations might escape the impact of unpredictable events and hence, reduces the risk of extinction. It also indicates the importance of retention of landscape linkages in order to promote the dispersal and exchange of individuals between geographically separate subpopulations and the retention of patches of suitable but currently unoccupied habitats (Margules and Pressey, 2000). In order to act on those recommendations, the rate of extinction, migration, and recolonization of the target or indicator species must be collected. In addition, the movement of the target or indicator species must be investigated in relation with the landscape element. Vos et al. (2001) investigated the movement speed, turning angle, and the probability of crossing boundaries in radio-tagged tree frogs in an agricultural landscape. Such data indicated that the landscape elements influenced the direction and speed of movement of these frogs. Opdam et al. (2002) emphasizes that this kind of data need to be generalized to other landscape patterns and aggregated to the multi-species level for the quantitative application in landscape planning. Source-pool effects and successional pathways: The species composition of an area varies as the successional stage progresses. In most cases, landscapes comprise areas at various successional stages as a result of patchy and periodic disturbances and many species exploit the temporal and spatial variation of the natural disturbance regimes (Margules and Pressey, 2000). Therefore, it is desirable if the successional stages are recorded and the designed green spaces represent all the successional stages. Spatial autecological requirements: One patch of green space might contain species that would not persist if they were isolated, which implies the requirements of sustainable populations across entire landscapes. Viable population sizes and structures such as age classes and sex ratio for this kind of taxa should be included in the green space design. Many species exploit the temporal variation by moving between different habitats (Margules and Pressey, 2000). The key habitat combinations differ among species, which further complicates the green space design. The focal-species approach might be one of the methods to incorporate the key habitat combinations into the green space design. Focal species is the most sensitive to each of the threatening processes, such as simplification of the patch structure and patch size reduction in a given landscape. The critical habitat

444

K. NAGASHIMA

requirements of each focal species are used to define the amount and configuration of habitats that must be maintained in the landscape. This approach assumes that green spaces designed to meet the requirements of the most demanding species will also encompass the requirements of other species (Lambeck, 1997; Watson et al., 2000; Brooker, 2002; Freudenberger and Brooker, 2004). Source-sink population structures: Population sources where the reproduction rate of species exceeds its mortality rate is the key for ensuring the persistence of rare species because the net dispersal away from the sources may sustain population in sinks where its reproduction rate is lower than mortality rate (Margules and Pressey, 2000). Therefore, identifying if a certain patch of green space is a source or sink by calculating the reproduction and mortality rates of the target species is recommended for green space design. Effects of habitat modification: The persistence of species within isolated habitats surrounded by artificial land-uses might be at risk. Establishing buffers and habitat restoration methods might be helpful in reducing this risk (Margules and Pressey, 2000). However, it is difficult to determine the extent to which the patch could be regarded as isolated. The utilization of the focal-species approach, which is sensitive to patch size and patch isolation, might help to identify the threshold for these parameters (e.g., Lambeck, 1999 in Freudenberger and Brooker, 2004; Watson et al., 2001; Brooker, 2002). Species as evolutionary units: The treatment of species as dynamic evolutionary units is been actively debated. Margules and Pressey (2000) suggest that there are at least two implications for planning. First, the areas occupied by taxa that appear from phylogenies to be actively radiating, or are most phylogenetically distinct, might be targeted for protection. Second, with an understanding of the physical and biological processes that lead to the active diversification of taxa, it is possible to identify and set targets for evolutionary templates. Each of the seven aspects had been studied in independent research field. For planning application, integration of those areas is needed. However, the best way for integration is not clear yet (Margules and Pressey, 2000). If attempts to explore the connection of their ecological theories and landscape configuration in each of the study area, as like the focal species approach in Spatial autecological requirements and Effects of habitat modification or like Opdam et al. (2002) argued in the metapopulation theory, and if there are increased challenges to generalize to the landscape level and multi-species level, some promising directions, hopefully, will be identified.

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

445

4. ASSESSMENT OF EXISTING GREEN SPACES AND DEFINING PROBLEMS The extent to which the targets for representativeness and persistence have already been achieved under the existing green space plan or policy have to be examined because it is considered as the basic information required to define the drawbacks of the current policy and to recognize the scope for the possible solutions. The U.S. gap analysis program (GAP), which is now at an operational status in each of the 48 contiguous states (Jennings, 2000), is well suited for the systematic reviews of the existing green space plan (Margules and Pressey, 2000). This program was originally designed to provide a quick overview of the distribution and conservation status of several biodiversity components (Scott et al., 1993). However, the demand and application of the GAP information expanded its intent to identify candidate conservation areas (Margules and Pressey, 2000) or to understand the vulnerability of plant communities and vertebrate species (Stoms, 2000). Gap analysis is a course-filter approach (Noss, 1987), which focuses on the community-based units of habitats as well as on each individual species (Jennings, 2000). It examines the element of biodiversity (e.g., vegetation type, habitat, and species) that is either sufficiently represented or not in the existing conservation network. The process of gap analysis compares the maps that show distributions of species and vegetation types of interest with the conservation area map or land ownership map. In GAP, land is classified into “status levels” that indicates the degree to which an area is managed to maintain its biodiversity (Scott et al., 1993). Status 1 indicates an area maintained in its natural state with natural disturbance regimes. Status 2 is similar to Status 1; however, at some instances, it may be used such that the quality of its habitat is slightly degraded. Status 3 includes all non-designated public lands such as national forests, which are managed for multiple uses. Status 4 indicates a land that is managed for human uses, including all privately owned land. The extent of representation of each biodiversity element will be represented in percentages. For example, the study assessing the extent and degree of protection of the vegetation in Idaho revealed that the percentages of montane forest, subalpine parklands, and subalpine forests under Status 1 were 17%, 24%, and 33%, respectively, and the amounts of the remaining vegetation types in the foothills and plain woodlands, forest-steppe transitions, and shrub-steppe and grassland complexes were less than 5% (Caicco et al., 1995). A study conducted in Wyoming (Merrill et al., 1996) showed that a smaller percentage of habitats for amphibians (8.8%) and reptiles (2.6%) were included in areas of Status 1 and Status 2 than those for birds (14.4%) and mammals (14.5%). Further, 6 amphibians (50% of all amphibians), 8 reptiles (31%), 25 mammals (22%), and 41 birds (14%) were designated as gap species. At the same time, these results imply that the highest priorities for future protection should be given to the vegetation types or species that were not well represented in the current conservation areas. Although gap analysis is a useful method for assessing the existing green space plan or policy, it has several limitations. First, no independent assessments are being performed on the findings of the gap analysis (Jennings, 2000). By quoting the

446

K. NAGASHIMA

results of Smith et al. (1998), who found that approximately 4% of the ovenbird habitat was represented in the conservation lands of Arkansas, Jennings (2000) stated that the accuracy of the percentage had not been evaluated. The reliability of these results has not been sufficiently researched. The second limitation is that the results do not reflect the previous habitat loss. Jennings (2000) has indicated again that if an element has already been extirpated from 70% of its previous distribution and, at present, the gap analysis shows a 10% representation in conservation areas then only 3% of the element’s previous distribution is represented. This might affect the potential for adapting to changing environments or the likelihood of metapopulation persistence. Strittholt and Boerner (1995) also suggest the importance of using maps that show the pre-settlement conditions in addition to the current vegetation in the U.S. in order to achieve the long-term conservation goals of representation and integrity. The third limitation is that the gap analysis only provides information on representativeness and does not measure the processes and persistence (Jennings, 2000; Margules and Pressey, 2000). This is due to the difficulty of the process and persistence measurements. Gap identification of the process and persistence of the biodiversity elements that examines areas may not maintain their process and persistence sufficiently, is expected to develop along with the accumulation of knowledge with regard to process and persistence measurement. This reviewing process might expose problems that can be included as new targets that should be tackled. In this case, the planning process should revert back to the measurement or target setting stage in order to deal with the problem as quickly as possible. 5. PROPOSING POSSIBLE SOLUTIONS Once the unrepresented biological features were clarified, the next step was to redesign the existing green space plan. The most practical tool to do this might be the site-selection algorithms that can identify minimum or near-minimum solutions, in terms of the number or area of sites, to the problem of representing all the targeted natural features in a region (Pressey et al., 1997). Site-selection algorithms have primarily been used in Australia (e.g., Margules et al., 1988; Nicholls and Margules, 1993; Margules et al., 1994; Pressey and Possingham, 1997) and South Africa (e.g., Rebelo and Siegfried, 1990; Turpie, 1995; Lombard et al., 1997). The same approach has been increasingly applied in other regions such as the U.S. (Dobson et al., 1997), Canada (Sarakinos et al., 2001), and Paraguay (Andelman and Willing, 2002). The algorithms utilized in these studies are called iterative selection procedures because they iterate through a list of candidate sites, choosing the best site at each step that contain features most complementary to those in the sites that have already been selected (Nicholls and Margules, 1993). The most common target features applied to the algorithms are species (e.g., Turpie, 1995; Dobson et al., 1997; Lombard et al., 1997; Sarakinos et al., 2001), which utilize the species distribution data. Few studies also utilized landscapes (e.g., Pressey and Nicholls, 1989) and environmental domains (e.g., Kirkpatrick and Brown, 1994). Utilizing both

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

447

distribution data and environmental domains has also been suggested (Kirkpatrick and Brown, 1994). Two main approaches are utilized in these studies. One is the “rarity algorithms” that begins with sites that contain unique features and progressively adds sites that contain the rarest unrepresented feature. The other is the “richness algorithms” that first selects the site having the greatest number of unreserved features and then adds sites that contain the greatest number of remaining unreserved features (Pressey et al., 1997). Both these algorithms require a resolution if a choice needs to be made between equally qualified sites. Although random selection such as selecting the first sites on the list could be done, recent studies describe a procedure or series of procedures that distinguish the sites on the basis of one or more criteria, which is expected to derive a more efficient solution to achieve the goal (Pressey et al., 1997). Taking in to account the proximity of sites (e.g., Nicholls and Margules, 1993; Lombard et al., 1997; Sarakinos et al., 2001), the abundance of each species (Turpie, 1995), the target areas of vegetation type, and the alien species distribution (Lombard et al., 1997) can be considered as examples of these criteria. The site-selection algorithms are considered to be a useful tool for green space design or conservation planning because it educates the planners about the approximate minimum cost to achieve the complete set of targets and thus ensures their feasibility. It also can be used to investigate and compare various political scenarios, for instance, to estimate the minimum green space requirements if certain sites are made mandatory or unavailable for conservation (Pressey et al., 1997). Good examples of this are studies comparing the case to begin the selection with existing green spaces that may or may not be under conservation (Margules et al., 1994; Andelman and Willing, 2002). These studies revealed that the minimum set of conservation area was small when the existing reserves were not considered than when the selection was done by considering the existing reserves; this can be attributed to the fact that current conservation areas were not established with the specific aim of protecting biodiversity. Nonetheless, including the existing green spaces under the conservation status from the beginning might be more practical than not considering them because they can become the focal points or spatial constraints around which enlarged reserves or new separate ones can be located (Margules and Pressey, 2000). Further, two more advantages of the site-selection algorithms are addressed. First is the availability of “gap species” data, which is the result of the gap analysis, in order to identify the additional protected green spaces (e.g., Kiester et al., 1996). A short running time even for a large data set is another advantage of practical planning because it enables the exhibition of their ability in case the land-use managers and politicians want to see the results of some alternative scenario suggested by them (Pressey et al., 1996). All the three-abovementioned advantages of the algorithms provide the basis for the negotiation and refinement of the conservation plan; therefore, they are useful tools for decision- making (Pressey et al., 1997). Although algorithms may ensure a comprehensive sample of biodiversity, it is doubtful whether they can ensure persistence. Most studies apply algorithms that select the minimum set of sites in which each natural feature is included at least

448

K. NAGASHIMA

once. In this case, selecting a species as the natural feature involves a risk in that a species may be designated as “conserved” at a site where it is not abundant enough to ensure its long-term survival (Turpie, 1995). Although this rule could be modified to select sites in which each species is included more than once (e.g., three or four times), the abundance of each species is still uncertain. Turpie (1995) challenged this problem by considering abundance while continuing to use the presenceabsence data, which was achieved by counting the species only at three sites where they were the most abundant. The utilization of data collected only at one point time is also cited as a problem (Margules et al., 1994; Pressey et al., 1999; Cabeza and Moilanen, 2001). Based on the high turnover of many communities, Margules et al. (1994) warned that site-selection methods based on static criteria might not protect the persistence of species as long as expected. The awareness of this problem is growing and it is suggested that the spatial population modeling and site-selection algorithm have been integrated (Cabeza and Moilanen, 2001; Lawler and Schumaker, 2004). In order to increase the utility of real planning scenarios, features such as commercial attributes, land ownership, and demands for multiple land use have to be incorporated into the algorithms (Prendergast et al., 1999). There are several studies that incorporate commercial attributes such as land value (Bedward et al., 1992; Ando et al., 1998) and land availability (Dobsont et al., 1997). However, no study has been able to handle the complexity of land ownership and the multiple demands of land-uses thus far; this accounts for the limited application of the site-selection algorithm to practical planning (Prendergast et al., 1999). Other obstacles such as its limited ability to explore alternative configurations for reserve networks (Bedward et al., 1992) and the fact that the algorithm does not provide any information on the potential value of unselected sites (Pressey et al., 1997) are also cited. Therefore, an iterative algorithm is a potentially valuable tool; however, it cannot support all the decisions that have to be taken in conservation planning, unless it is complemented by other analyses (Pressey et al., 1997). The requirement of species distribution data is another limitation for the application of the site-selection algorithm (Prendergast et al., 1999; Margules and Pressey, 2000; Cabeza and Moilanen, 2001). Species is the most common target feature utilized in the site-selection algorithms; however, the algorithms can be applied entirely only when species distributions are known (Prendergast et al., 1999). However, the quality of the species distribution data remains poor. The data on species abundance and distributions are often biased toward the preferred sites or toward a few charismatic species (Williams and Gaston, 1994; Cabeza and Moilanen, 2001). Efforts to incorporate environmental surrogates, which might be easier and cheaper to acquire than the species distribution data, to the site-selection algorithm are few (e.g., Kirkpatrick and Brown, 1994), and these are in the early stages of development (Prendergast et al., 1999). While site-selection algorithms improved continuously, studies utilizing these algorithms only attempted to set priorities on the existing green spaces. In urban areas where green spaces are scarce and fragmented, land-use managers might also pay attention to habitat recreation and ecological restoration. Site-selection

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

449

algorithms play little or no part in these aspects (Prendergast et al., 1999), which might be one of the directions for future studies. 6. DECISION MAKING BY INVOLVING ALL STAKEHOLDERS The design of green spaces proposed by the previous step is preliminary because practical impediments that require a degree of revision might be revealed during the extensive discussions held among all stakeholders while building a consensus. Consensus building is a critical step that reflects the quality of green space design and its implementation result. The feasibility would increase as more people give inputs to the plan, which might lead to ideal implementation. The extensive discussion while building a consensus might also contribute to increase the awareness of stakeholders with regard to the importance of off-reserved sites. The politically conserved areas must be recognized as the core component of regional conservation regimes that are not complete in them. In order to sample all the species, large areas of land are required, but it is not possible to categorize all of them under the conservation estate. An enhanced sympathetic management of species both inside and outside the reserves is essential (Margules et al., 1994), which cannot be achieved without the awareness and understanding of stakeholders. Here, “stakeholders” include governmental agencies, private owners, and other diverse stakeholders of both public and private lands. Not only the local government but also the neighboring jurisdictions or other political organizations might be included in the government agencies to manage the transboundary problems (Brody, 2003). The involvement of specialists in the fields of social economy, engineering, and land-use planning in addition to ecology are also desirable so that reasonable advices can be obtained when revision becomes necessary. For revising the initial selection, again, the site-selection algorithm is a useful tool because it serves well to investigate and compare various scenarios that suggest reverting to the possible solution phase. Displaying the irreplaceability of each green space as a map (e.g., Ferrier et al., 2000; Cowling et al., 2003) might be another useful tool to explore the alternative areas for a certain patch to meet the conservation goal (Margules and Pressey, 2000) under the same scenario. Irreplaceability is defined as the likelihood that the area will be required as part of a conservation system that achieves the set of targets and the extent to which the options for achieving the set of targets are reduced if the area is unavailable for conservation. Irreplaceability is represented in terms of percentage. Areas with 100% irreplaceability indicate that no spatial options are available for achieving the target and therefore, required to be under the conservation estate. Lower irreplaceabilities imply more replacements in the planning area that provide a better scope for alternative selections (Ferrier et al., 2000; Margules and Pressey, 2000). Two more advantages to use irreplaceability are cited by Pressey et al. (1995) as follows. First, the irreplaceability map can show the initial irreplaceability patterns of selected as well as unselected areas; therefore, it can be utilized as a guide to examine the feasibility of modifications to the conservation plan. Second,

450

K. NAGASHIMA

irreplaceability can be incorporated into an interactive algorithm that recalculates and redisplays the pattern as decisions about new green space designs are made. 7. IMPLEMENTING THE CONSERVATION ACTIONS Although the most practical green space plan was decided by a majority vote of all the stakeholders, the complete implementation of the plan on the ground in a short time is usually difficult because of the lack of budget or various vested interests in the area (Margules and Pressey, 2000; Pressey and Taffs, 2001). Scheduling of conservation actions becomes necessary in these cases, which requires information on the most important features that need to be protected first. The strategy most widely utilized for answering this question is focusing on the areas that exhibit a high conservation value and high vulnerability to processes that threaten biodiversity (Cowling et al., 2003). The conservation value is measured by irreplaceability (Ferrier et al., 2000; Margules and Pressey, 2000; Pressey and Taffs, 2001; Noss et al., 2002), endemism (Mittermeier et al., 1998; Balmford et al., 2001), and the diversity, abundance, and persistence of the conservation features (Groves et al., 2002). Vulnerability is an estimate of the likelihood or imminence of habitat loss or degradation. The measures of vulnerability include features such as previous or projected rates of habitat loss (White et al., 1997; Mittermeier et al., 1998), human population density (Thompson and Jones, 1999; Balmford et al., 2001), and land suitability for certain land-use (Pressey and Taffs, 2001). One example of utilizing the information for scheduling is to plot the selected areas on a graph of irreplaceability (y axis) versus vulnerability (x axis) (e.g., Margules and Pressey, 2000; Pressey and Taffs, 2001; Noss et al., 2002). The highest priority for implementation is given to the areas with high vulnerability to losses and high irreplaceability that belongs to the upper right quadrant. However, this approach may prioritize highly fragmented areas with high opportunity costs of implementation, which was the case in South Africa (Cowling et al., 2003). Moreover, prioritizing areas that are highly fragmented implies a risk of ensuring the persistence of conservation features in the region. This can be compensated if a strategy is applied that focuses on large areas of intact habitat, where the vulnerability is low to moderate and it is feasible to accommodate a wide range of pattern and processes features (Cowling et al., 2003). The opportunity cost and management costs are considered relatively low in this strategy (Noss et al., 2002; Frazee et al., 2003); therefore, greater biodiversity returns per unit of investment are considered to be the gain. Further, it might ensure better persistence than the former strategy (Cowling et al., 2003). The abovementioned scheduling strategies are useful tools that provide scientific information for prioritization. However, in practice, various socio-economic factors such as opportunity cost, management cost, funding, institutional capacity, and so on, must be considered to determine the most important and feasible area where the conservation action has to be implemented first (Theobald et al., 2000; Faith et al., 2001; Cowling et al., 2003). The available form of management is also a critical issue for scheduling (Margules and Pressey, 2000). There might be some cases in

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

451

which the most preferred form of management is infeasible, but the other form can be implemented in a short time. Various legislations that provide financial incentives to the landowners, such as subsidy payment and preferential tax treatments (Brody, 2003), might increase chances to obtain their support, especially when acquisition and designation as local authority reserves are infeasible. In this case, the area can be put into the off-reserve system for a while until the most preferred form is applicable. This situation also implies the importance of political factors like the incentive-based legislations. In reality, all these socio-economic and political factors should be integrated in the prioritization strategies for scheduling (Crowling et al., 2003), and this remains as one of the aspects that need to be tackled in the future. 8. MONITORING AND ASSESSMENT As mentioned above, scientific approaches have been well established and have a great potential to provide the basic information for biological conservation and green space design. However, scientific knowledge, the basis for these approaches, is provisional. In fact, several tasks that should be addressed in the future have been indicated for each of the previous steps. Hence, periodic monitoring is essential not only for understanding the status of the selected biological features but also for assessing the adequacy of green space management. The process for monitoring and assessment assumes the same process as the planning processes mentioned above; further, this process is unidirectional (Margules and Pressey, 2000). It requires information on biodiversity and ecological processes that ensure persistence of each of the green spaces (Step 1). The effectiveness of the green space plan might be assessed by understanding the contribution of each green space to the targets (Step 2). If a problem is encountered, the plan should be redesigned and it should undergo the decision-making process for building consensus on the revised plan (Steps 3–4). Further, the newly accumulated scientific knowledge should be utilized for updating the plan. Both the problem encountered and the new scientific knowledge can possibly modify the items or indicators for data collection, which implies reverting to the very first step of the process. 9. CONCLUSION This paper reviews the essential planning processes of green space design under the concept of ecosystem management together with the recent technical trends and the required data set applicable at each process in order to accomplish the goals to protect the entire spectrum of biodiversity (representativeness), ensure their persistence, and accommodate human use within these constraints. The essential processes for green space design were as follows: (1) data collection and identification of management targets, (2) assessment of existing green spaces and defining the problems, (3) proposing possible solutions, (4) decision making by involving all stakeholders, (5) implementing the conservation actions, and (6) monitoring and assessment.

452

K. NAGASHIMA

Surrogates are utilized for measuring biodiversity and ecological processes because it is unrealistic to investigate all the aspects of the biological features in a short span of time. Species-based indicators utilized for measuring biodiversity, such as an indicator taxon, umbrella species, flag species, and rare species, exhibit little concordance with biodiversity hotspots. Other surrogates such as higher taxon richness and habitat or environmental diversity are also proposed; however, their utility as surrogates has just been examined. The most practical measurement of biodiversity is to simultaneously utilize several surrogates such that the shortcoming of each surrogate might be compensated. The measurement of the ecological process is more difficult because spatial surrogates are required. Many studies have been conducted in order to understand the landscape pattern; however, only few studies provided the understanding of the ecological meaning of the patterns, which is one of the future directions of landscape ecology. Biogeographical theory, metapopulation theory, source-pool effects and successional pathways, spatial autecological requirements, source-sink population structures, effects of habitat modification, and species as evolutionary units may aid in setting targets and measuring the ecological processes in order to maintain persistence. However, the integration of these theories is essential for practical use. Gap analysis is a useful tool for assessing the existing green spaces plan; it defines the problems and recognizes the scope for the possible solutions. Further, it examines the element of biodiversity that may or may not be sufficiently represented in the existing conservation network. It provides information on representativeness, but not on persistence. Gap identification of the processes and persistence is another challenge that should be overcome in the future. Site-selection algorithm is considered to be the most practical tool for redesigning the existing green space plan because it can identify minimum solutions to the problem of representing all the targeted natural features in the region. It can also support decision making by providing the basis for negotiation, such as the result of comparing several scenarios. Nonetheless, its utility to ensure persistence is still doubtful. The integration of spatial population modeling and site-selection algorithm is recognized as one of the solutions; however, studies on this algorithm have been limited. Incorporation of commercial attributes, land ownership, and multiple demands for land-use in the algorithm is also required to enhance the algorithm for practical use. Its application to habitat recreation and ecological restoration is another feature to be studied. Involving all the stakeholders in the decision-making phase can be regarded as the distinguished process of ecosystem management. This is a critical step that reflects the feasibility and implementation result of the plan. Extensive discussions among all the stakeholders might result in a high degree of revision in the plan, which implies reverting to the previous step. When consensus is built with regard to the scenario but not on the configuration of the green space, irreplaceability might become a useful criterion to explore an alternative area for a certain patch to attain the goal. Complete implementation of the plan in a short time is often difficult; therefore, prioritization of the targeted areas becomes essential for scheduling the conservation

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

453

actions. Two scientific strategies have been introduced. One is to focus on areas with high conservation value and high vulnerability. The other is to focus on large areas of intact habitat where vulnerability ranges from low to moderate. The former can conserve fragmented areas; however, it poses the risk of maintaining persistence. The latter can accommodate a wide range of patterns and processes, but not fragmented areas with high vulnerability. Both strategies should be utilized in order to compensate for their respective shortcomings. However, techniques to incorporate other factors such as costs and management form are also required for practical use. Scientific knowledge that provides the basis for green space design is provisional. Therefore, periodic monitoring should be conducted both for understanding the status of the selected biological features and for assessing the adequacy of the green space plan. Monitoring requires the same process as the planning process with regular feedbacks and reasons for revising the plan. In summary, this review revealed the following issues as future challenges. First is the development of a generalized methodology for biodiversity as well as ecological process measurement. Several surrogates have been utilized to measure biodiversity, but each of them had a little overlap with the highly diversified area. Methodologies for measuring the ecological processes are still in its early stages of development. Researches should begin with case studies that explore the relationship between the seven related ecological theories and the landscape configuration, and then be generalized to suit the landscape level and multi-species level. Second is the investigation of adequate methods for assessing the existing green space plan and for proposing solutions from the viewpoint of protecting the ecological processes and maintaining persistence. Both gap analysis and siteselection algorithms are useful tools to represent biodiversity, but not persistence. Third is the effort to improve the quality of data. Inventory efforts should be distributed as widely as possible among the sites and taxa. Qualified data is essential for exploring the optimum biodiversity surrogates and analytical methods such as gap analysis and site-selection algorithms. The final issue is to increase the number of studies that focus on habitat recreation or ecological restoration. Despite the efforts for conserving the remaining green spaces, a continuous loss of species and ecosystems is reported and highly fragmented areas are increasing with increasing urbanization. Land use planners in these areas are increasingly interested in the techniques of habitat restoration; therefore, green space design is recommended to provide information on sites and habitat types that have to be recreated. REFERENCES Allen, C.R., Pearlstine, L.G., Wojcik, D.P., and Kitchens, W.M. (2001). The spatial distribution of diversity between disparate taxa: Spatial correspondence between mammals and ants across South Florida, U. S. A. Landscape Ecology, 16, 453-464. Andelman, S.J., and Willing, M.R. (2002). Alternative configurations of conservation reserves for Paraguayan bats: Considerations of spatial scale. Conservation Biology, 16, 1352-1363 Ando, A.A., Camm, J., Polasky, S. and Solow, A. (1998). Species distributions, land values, and efficient conservation Science, 279, 2126-2128. Araujo, M.B. and Humphries, C.J. (2001). Would environmental diversity be a good surrogate for species diversity? Ecography, 24, 103-110

454

K. NAGASHIMA

Awimbo, J.A., Norton, D.A., and Overmars, F.B. (1996). An evaluation of representativeness for nature conservation, Hokitika ecological district, New Zealand. Biological Conservation, 75, 177-186. Balmford, A., Moore, J.L., Brooks, T., Burgess, N., Hansen, L.A., Williams, P. and Rahbek, C. (2001) Conservation conflicts across Africa. Science, 291, 2616-2619. Bedward, M., Pressey, R.L., and Keith, D.A. (1992). A new approach for selecting fully representative reserve networks: addressing efficiency, reserve design and land suitability with an iterative analysis Biological Conservation, 62, 115-125. Belbin, L. (1993). Environmental representativeness: Regional partitioning and reserve selection. Biological Conservation, 66, 223-230. Berger, J. (1997). Population constraints associated with the use of black rhinos as an umbrella species for desert herbivores. Conservation Biology, 11, 69-78. Bifolchi, A. and Lode, T. (2005). Efficiency of conservation shortcuts: An investigation with otters as umbrella species. Biological Conservation, 126, 523-527. Brody, S.D. (2003). Implementing the principles of ecosystem management through local land use planning. Population and Environment, 24 (6), 511-540. Brooker, L. (2002). The application of focal species knowledge to landscape design in agricultural lands using the ecological neighbourhood as a template. Landscape and Urban planning, 60, 185-210. Buzas, M. A. (1972). Patterns of species diversity and their explanation. Taxon, 21, 275-286. Cabeza, M. and Moilanen, A. (2001). Design of reserve networks and the persistence of biodiversity. Trends in Ecology and Evolution, 16 (5), 242-248. Caicco, S. L., Scott, J. M., Butterfield, B., and Csuti, B. (1995). A Gap analysis of the management status of the vegetation of Idaho (USA). Conservation Biology, 9, 498-511. Carignan, V. and Villard, M. (2002) Selecting indicator species to monitor ecological integrity: A review. Environmental Monitoring and Assessment, 78, 45-61 Christensen, N., Bartuska, A., Brown, J., Carpenter, S., D’Antonio, C., Francis, R., Franklin, J., MacMahon, J., Noss, R., Parsons, D., Petterson, C., Turner, M. and Woodmansee, R. (1996). The report of the ecological society of America committee on the scientific basis for ecosystem management. Ecological Applications, 6, 665-91. Cortner, H. and Moote, M. (1999). The Politics of Ecosystem Management. Island Press, Washington DC. Cowling, R.M., Pressey, R.L., Rouget, M. and Lombard, A.T. (2003). A conservation plan for a global biodiversity hotspot: the Cape Floristic Region, South Africa. Biological Conservation, 112, 191216. Csuti, B., Polasky, S., Williams, P.H., Pressey, R.L., Camm, J.D., Kershaw, M., Kiester, A.R., Downs, B., Hamilton, R., Huso, M. and Sahr, K. (1997) A comparison of reserve selection algorithms using data on terrestrial vertebrates in Oregon. Biological Conservation, 80, 83-97. Dobson, A.P., Rodriguez, J.P., Roberts, W.M. and Wilcove, D.S. (1997). Geographic distribution of endangered species in the United States. Science, 275, 550-553 Faith, D.P. and Norris, R.H., (1989). Correlation of environmental variables with patterns of distribution and abundance of common and rare freshwater macroinvertebrates. Biological Conservation, 50, 7799. Faith, D.P., Margules, C.R. and Walker, P.A. (2001). A biodiversity conservation plan for Papua New Guinea based on biodiversity tradeoffs analysis. Pacific Conservation Biology, 6, 304-324. Ferrier, S., Pressey, R.L. and Barrett, T.W. (2000). A new predictor of the irreplaceability of areas for achieving a conservation goal, its application to real-world planning, and a research agenda for further refinement. Biological Conservation, 93, 303-325. Fleishman, E., Murphy, D.D. and Brussard, P.F. (2000). A new method for selection of umbrella species for conservation planning. Ecological Applications, 10 (2): 569-579. Frazee, S.R., Cowling, R.M., Pressey, R.L., Turpie, J.K. and Lindenberg, N. (2003). Estimating the costs of conserving a biodiversity hotspot: a case study of the Cape Floristic Region, South Africa. Biological Conservation, 112, 275-290. Freudenberger, D. and Brooker, L. (2004). Development of the focal species approach for biodiversity conservation in the temperate agricultural zones of Australia Biodiversity and Conservation, 13: 253274 Groves, C.R., Jensen, D.B., Valutis, L.L., Redford, K.H., Shaffer, M.L., Scott, J.M., Baumgartner, J.V., Higgins, J.V., Beck, M.W. and Anderson, M.G. (2002). Planning for biodiversity conservation: putting conservation science into practice. BioScience, 52, 499-512. Grunbine, R.E. (1994). What is ecosystem management? Conservation Biology, 8 (1), 27-38.

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

455

Haines-Young, R. (1999). Landscape pattern: Context and process. In J. A., Wiens and M. R. Moss (Eds.) Issues in Landscape Ecology. (pp. 33-37). International Association for Landscape Ecology, Ontario. Howard, P.C., Viskanic, P., Davenport, T.R.B., Kigenyi, F.W., Baltzer, M., Dickinson, C.J., Lwanga, J.S., Matthews, R.A. and Balmford, A. (1998). Complementarity and the use of indicator groups for reserve selection in Uganda. Nature, 394, 472-475. Jennings, M.D. (2000). Gap analysis: concepts, methods, and recent results. Landscape Ecology, 15, 5- 20. Kakizawa, H. (2000). Ecosystem Management. Tsukiji-shokan Co., Tokyo. Kiester, A.R., Scott, J.M., Csuti, B., Noss, R.F., Butterfield, B., Sahr, K. and White, D. (1996). Conservation prioritization using gap data. Conservation Biology, 10 (5), 1332-1342. Kirkpatrick, J.B. and Brown, M.J. (1994). A comparison of direct and environmental domain approaches to planning reservation of forest higher plant communities and species in Tasmania. Conservation Biology, 8 (1), 217-224. Kremen, C. (1992). Assessing the indicator properties of species assemblages for natural areas monitoring Ecological Applications, 2 (2), 203-217. Lambeck, R.J. (1997). Focal species: a multi-species umbrella for nature conservation. Conservation Biology, 11, 849-856. Landres, P.B., Verner, J. and Thomas, W. (1988). Ecological uses of vertebrate indicator species: a critique. Conservation Biology, 2, 316-328. Lawler, J.J. and Schumaker, N.H. (2004). Evaluating habitat as a surrogate for population viability a spatially explicit population model. Environmental Monitoring and Assessment, 94, 85-100. Lombard, A.T., Cowling, R.M., Pressey, R.L. and Mustart, P.J. (1997). Reserve selection in a speciesrich and fragmented landscape on the Agulhas Plain, South Africa. Conservation Biology, 11, 11011116. Margules, C.R. and Pressey, R.L. (2000). Systematic conservation planning. Nature, 405, 243-253. Margules, C.R., Nicholls, A.O. and Pressey, R.L. (1988). Selecting networks of reserves to maximize biological diversity. Biological Conservation, 43, 663-676. Margules, C.R., Nicholls, A.O. and Usher, M.B. (1994). Apparent species turnover, probability of extinction and the selection of nature reserves: a case study of the Ingleborough limes stone pavements. Conservation Biology, 8, 398-409. Martikainen, P., Kaila, L. and Haila, Y. (1998). Threatened beetles in White-Backed Woodpecker habitats. Conservation Biology, 12, 293-301. . Merrill, E.H., Kohley, T.W., Herdendorf, M.E., Reiners, W.A., Driese, I.K.L., Marrs, R.W. and Anderson, S.H. (1996). The Wyoming Gap Analysis Project final report. Retrieved December 5, 2005, from U. S. Geological Survey GAP homepage: http://www.gap.uidaho.edu/ Mittermeier, R.A., Myers, N., Thomsen, J.B., da Fonseca, G.A.B. and Olivieri, S. (1998). Biodiversity hotspots and major tropical wilderness areas: approaches to setting conservation priorities. Conservation Biology, 12, 516-520. Nicholls, A.O. and Margules, C.R. (1993). An upgraded reserve selection algorithm. Biological Conservation, 64, 165-169. Noss, R.F. (1990). Indicators for monitoring biodiversity: a hierarchical approach. Conservation Biology, 4, 355-364. Noss, R.F., Carroll, C., Vance-Borland, K. and Wuerthner, G. (2002) A multicriteria assessment of the irreplaceability and vulnerability of sites in the greater Yellowstone ecosystem. Conservation Biology, 16, 895-908. Noss, R.F. (1987). From plant communities to landscapes in conservation inventories: a look at the Nature Conservancy (USA). Biological Conservation, 41, 11-37. Opdam, P., Foppen, R. and Vos, C. (2002) Bridging the gap between ecology and spatial planning in landscape ecology. Landscape Ecology, 16, 767-779. Pearson, D.L. and Cassola, F. (1992). World-wide species richness patterns of tiger beetles: indicator taxon for biodiversity and conservation studies. Conservation Biology, 6, 376-391. Pimm, S. L. and Lawton, J.H. (1998). Planning for biodiversity. Science, 279, 2068-2069. Poiani, K.A., Merrill, M.D. and Chapman, K.A. (2001). Identifying conservation priority areas in a fragmented Minnesota landscape based on the umbrella species concept and selection of large patches of natural vegetation. Conservation Biology, 15 (2), 513-522. Prendergast, J.R., Quinn, R.M. and Lawton, J.H. (1999). The gaps between theory and practice in selecting nature reserves. Conservation Biology, 13 (3), 484-492.

456

K. NAGASHIMA

Prendergast, J.R., Quinn, R.M., Lawton, J.H., Eversham, B.C. and Gibbons, D.W. (1993). Rare species, the coincidence of diversity hotspots and conservation strategies. Nature, 365, 335-337. Pressey, R.L. and Logan, V.S. (1995). Reserve coverage and requirements in relation to partitioning and generalization of land classes: Analyses for western New South Wales. Conservation Biology, 9, 1506-1517. Pressey, R.L. and Nicholls, A.O. (1989a). Efficiency in conservation evaluation: Scoring versus iterative approaches. Biological Conservation, 50, 199-218. Pressey, R.L. and Nicholls, A.O. (1989b). Application of a numerical algorithm to the selection of reserves in semi-arid New South Wales. Biological Conservation, 50, 263-278. Pressey, R.L. and Taffs, K.H. (2001). Scheduling conservation action in production landscapes: priority areas in western New South Wales defined by irreplaceability and vulnerability to vegetation loss. Biological Conservation, 100, 355-376. Pressey, R.L., Ferrier, S., Hutchinson, C.D., Sivertsen, D.P. and Manion, G. (1995). Planning for negotiation: using an interactive geographic information system to explore alternative protected area networks. In D. A. Saunders, Craig, J.L. and Mattiske, E.M. (Eds.), Nature conservation: The Role of Networks (pp. 23-33.). Surrey Beatty and Sons, Sydney. Pressey, R.L., Possingham, H.P. and Day, J.R. (1997). Effectiveness of alternative heuristic algorithms for identifying indicative minimum requirements for conservation reserves. Biological Conservation, 80, 207-219. Pressey, R.L., Possingham, H.P. and Margules, C.R. (1996). Optimality in reserve selection algorithms: When does it matter and how much? Biological Conservation, 76, 259-267. Rebelo, A.G. and Siegfried, W.R. (1990). Protection of Fynbos vegetation: ideal and real world options. Biological Conservation, 54, 15-31. Reid, W.V. (1998). Biodiversity hotspots. TREE 13 (7): 275-280. Rubinoff, D. (2001). Evaluating the California Gnatcatcher as an umbrella species for conservation of Southern California coastal sage scrub. Conservation Biology, 15 (5), 1374-1383. Sarakinos, H., Nicholls, A.O., Tubert, A., Aggarwal, A., Margules, C.R. and Sarkar, S. (2001). Area prioritization for biodiversity conservation in Quebec on the basis of species distributions: a preliminary analysis. Biodiversity and Conservation, 10, 1419-1472. Schall, J.J. and Pianka, E.K., (1978). Geographical trends in number of species. Science, 20, 679-686. Schumaker, N. (1996). Using landscape indices to predict habitat connectivity. Ecology, 77, 1210-1225. Scott, J.M., Davis, F., Csuti, B., Noss, R.F., Butterfield, B., Groves, C., Anderson, H., Caicco, S., D’Erchia, F., Edwards, T.C., Jr., Ulliman, J. and Wright, G. (1993). Gap analysis: a geographic approach to protection of biological diversity. Wildlife Monographs, 123. Simberloff, D. (1998). Flagships, umbrellas, and keystones: is single-species management passé in the landscape era? Biological Conservation, 83, 247-257. Stoms, D.M. (2000). GAP management status and regional indicators of threats to biodiversity. Landscape Ecology, 15, 21-33. Strittholt, J.R. and Boerner, R.E.J. (1995). Applying biodiversity gap analysis in a regional nature reserve design for the edge of Appalachia, Ohio (USA). Conservation Biology, 9, 1492-1505. Suter, W., Graf, R.G. and Hess, R. (2002). Capercaillie and Avian biodiversity: Testing the umbrella species concept. Conservation Biology, 16 (3), 778-788. Szaro, R., Sexton, W. and Malone, C. (1998). The emergence of ecosystem management as a tool for meeting people’s needs and sustaining ecosystems. Landscape and Urban Planning, 40, 1-7. Theobald, D.M., Hobbs, N.T., Bearly, T., Zack, J.A., Shenk, T. and Riebsame, W.E. (2000). Incorporating biological information in local land-use decision-making: designing a system for conservation planning. Landscape Ecology, 15, 35-45. Thompson, K. and Jones, A. (1999). Human population density and prediction of local plant extinction in Britain. Conservation Biology, 13, 185-189. Turpie, J.K. (1995). Prioritizing South African estuaries for conservation: a practical example using waterbirds. Biological Conservation, 74, 175-185. van Jaarsveld, A.S., Freitag, S., Chown, S.L., Muller, C., Koch, S., Hull, H., Bellamy, C., Kruger, M., Endrody-Younga, S., Mansell, M.W. and Scholtz, C.H. (1998). Biodiversity assessment and conservation strategies. Science, 279, 2106-2108. Vos, C.C., Verboom, J., Opdam, P.F. and Ter Braak, C.J.F. (2001). Towards ecologically scaled landscape indices. The American Naturalist, 157, 24-51.

TRENDS AND FUTURE RESEARCHES IN GREEN SPACE DESIGN

457

Wallis de Vries, M.F. (1995). Large herbivores and the design of nature reserves in Western Europe. Conservation Biology, 9, 25-33. Watson, J., Freudenberger, D. and Paull, D. (2001). An assessment of the focal-species approach for conserving birds in variegated landscapes in Southeastern Australia. Conservation Biology, 15, 13641373 Wessels, K.J., Freitag, S. and van Jaarsveld, A.S. (1999) The use of land facets as biodiversity surrogates during reserve selection at a local scale. Biological Conservation, 89, 21-28. White, D., Minotti, P.G., Barczak, M.K., Sifneos, J.C., Freemark, K.E., Santelmann, M.V., Steinitz, C.F., Kiester, A.R. and Preston, E.M. (1997). Assessing risks to biodiversity from future landscape change. Conservation Biology, 11, 349-360. Wilcove, D. (1993). Getting ahead of the extinction curve. Ecological Applications, 3, 218-220. Williams, P.H., Burgess, N.D. and Rahbek, C. (2000). Flagship species, ecological complementarity and conserving the diversity of mammals and birds in sub-Saharan Africa. Animal Conservation, 3, 249260. Williams, P., Gibbons, D., Margules, C., Rebelo, A., Humphries, C. and Pressey, R. (1996). A comparison of richness hotspots, rarity hotspots, and complementary areas for conserving diversity of British Birds. Conservation Biology, 10, 155-174. Williams, R.H. and Gaston, K.J. (1994). Measuring more of biodiversity: Can higher-taxon richness predict wholesale species richness? Biological Conservation, 67, 211-217.

CHAPTER 27

BEIJING URBAN SPATIAL DISTRIBUTION AND RESULTING IMPACTS ON HEAT ISLANDS

Z. OUYANG1, R.B. XIAO1, E.W. SCHIENKE2, W.F. LI1, X. WANG1, H. MIAO1, H. ZHENG1 1

National Key Lab of Systems Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. Shuang Qing Road 18, Haidian District, P.O. Box 2871. Beijing, 100085. China; 2Dartment of Science and Technology Studies, Rensselaer Polytechnic Institute, Troy, NY 12180 USA

Abstract. The physical characteristics of the ground surface are regarded as the main factors in the urban heat island phenomena. Over two seasons, this study spatially and quantitatively examines the influence of urban surface features on land surface temperature in Beijing, China through the use of remote sensing (RS) combined with geographic information systems (GIS). Primary data sources include: Landsat Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+), SPOT, QuickBird and Beijing Road vector map. Variables extracted and considered in the study are: (1) percent (surface) imperviousness, (2) Normalized Difference Vegetation Index (NDVI), (3) ratio of water bodies, (4) ratio of tall-building areas, and (5) road density. Results indicate that Beijing’s urban spatial pattern presents a typical concentric distribution: NDVI values increase, but impervious surface and tall-building area decrease from inner city to outskirts. The land surface temperature (LST) pattern is non-symmetrical and nonconcentric, with relatively higher temperature zones clustered towards the south of the central axis and within the fourth ring road. Principal component regressions indicate that a strong linear relationship exists between LST and the studied urban parameters, such as percent imperviousness, NDVI, ratio of water cover, tall building and road density, though they do exhibit seasonal variations. In the August image, the percentage of impervious surfaces exhibits the largest positive correlation with LST, which is able to explain 81.7% of LST variance. NDVI follows in impact with a strong negative correlation. For analysis in May, with an R2 of 0.720, NDVI and water are the two features, which most negatively correlate with LST. As a practical result, these findings can be used to inform future design measures for mitigating urban heat island effects.

459 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 459–478. © 2007 Springer.

460

Z. OUYANG ET AL.

1. INTRODUCTION One of the best-known forms of anthropogenic climate modification is the phenomenon of urban heat islands (UHIs), which was first documented in London by Lake Howard in 1833. There are varieties of reasons for the urban/rural temperature variances. Among these, the main contributing factors are changes in the physical characteristics of the surface (albedo, thermal capacity, heat conductivity), owing to the replacement of vegetation by asphalt and concrete; the decrease of surface moisture available for evapotranspiration; changes in the radiative fluxes in and near the surface flow, owing to the complex geometry of streets and tall buildings, and anthropogenic heat (Dousset and Gourmelon, 2003; Streutker, 2003). Consequently, there are many studies investigating the relationship between surface temperature and land use/land cover. Unger (2001) developed the regression to examine the influence of urban and meteorological factors on the surface air temperature for Szeged, Hungary. While the Dousset and Gourmelon (2003) study demonstrated the physical effects of surface properties, especially in downtown business and industrial districts that display heat-islands larger than 7 °C in variation, as well as the temperate influence of water. Weng (2001) examined LST patterns and their relationship with land cover in the Zhujiang Delta, China. Nichol (2005) examined how different land use presented different diurnal thermal behaviours between day and night. Voogt and Oke (1998) found there were strong directional variations in apparent surface temperature over each of three urban landuse areas (light-industrial, residential, and downtown). These studies have contributed to our understanding of the thermal patterns created by individual land use/cover within a city such as parks, water, industrial facilities, variations of tree cover, etc. Despite these efforts, further research is needed in other urban areas in order to reinforce the absolute and comparative relationship between surface features and urban heat islands. Urban heat island studies are generally conducted in one of two ways: measuring the UHI in air temperature through the use of automobile transects and weather station networks, and measuring the UHI in surface (or skin) temperature through the use of airborne or satellite remote sensing. In situ data have the advantage of a high temporal resolution and a long data record, but have poor spatial resolution. Conversely, remotely sensed data has higher spatial distribution but low temporal resolution and a shorter data record (Streutker, 2003). So, studies on the UHI phenomenon using satellite derived land surface temperature measurements have been widely conducted (Dousset and Gourmelon, 2003; Streutker, 2003; Weng et al., 2004). LST is the basic parameter for the deviation of the thermal behaviour of the environment. It has a number of meteorological and eco-environmental consequences, including changing the incidence of downwind precipitation, increasing urban soil temperature which affects soil metabolism, and influences various behaviours of urban biology, as well as the chemical cycle, and urban energy utilization for indoor climate management (Xiao et al., 2005). In this paper, we integrate RS and GIS to study Beijing’s urban spatial distribution and its impacts on land surface temperature in two different seasons.

URBAN DISTRIBUTION AND HEAT ISLAND

461

Based on previous studies, five affected parameters are considered, including: NDVI, water body ratio, high building ratio, imperviousness ratio and road density. The objectives of this paper are (1) to detect and analyse the spatial patterns of Beijing urban structure; (2) to analyse the spatial variations of land surface temperatures; and (3) to examine the relationship between urban structure and surface temperature. This work can contribute to provide the theoretical basis to alleviate Beijing’s UHI phenomenon and give some informative parameters for ongoing urban planning. . 2. STUDY AREAS AND DATA DESCRIPTION Beijing, the capital of China and one of the largest and oldest cities in the world, covers approximately 16,800 km2 with a population of 13 million. Rapid expansion of urban areas, fast development and lack of appropriate urban planning have had serious impacts on urban thermal environment. It is found that heat island effects in Beijing were evident since at least 1961, when the average daily temperature in the city was 4.6°C higher than that in the suburb. (Song and Zhang, 2003) Beijing’s expansion presents a typical concentric distribution, which forms an obvious ringshaped expansion pattern from the inner centre to the outskirts. Our study area focuses within the fifth-ring road, where the majority of built-up area is located. This area occupies 666.7 km2 comprised by 9 counties and 120 townships, residing on the plain, with an elevation approximately 100m above mean sea level. The local region belongs to the temperate climatic zone, and has a mean annual temperature of 12ºC, and an average annual precipitation of 640mm. The study area was divided into 1000m×1000m grids (Figure 1) in order to compute urban parameters for statistical analysis.

Figure 1. Location of study area.

462

Z. OUYANG ET AL.

Four primary data sources were employed in the study: (1) Landsat TM (August 31, 2001) and ETM+ (May 22, 2002) images were acquired to retrieve urban surface imperviousness and surface temperature in two seasons; (2) QuickBird (July 5, 2002) was utilized for derivation of training/test data to estimate surface imperviousness; (3) Panchromatic SPOT (November 13, 2000) was used to extract building height distribution; and (4) Beijing 1:10,000 topography and road map from Beijing Institute of Surveying and Mapping. All data were geometrically intermatched based on the 1:10,000 topographic maps. Table 1. Summary of the images used in this study.

Date Aug. 31, 2001 May 22, 2002 Nov 13, 2001 Jul 5, 2002

Type of image TM

Spatial resolution (m) 30, 90a

No. of bands 7

Sun elevation (degree) 52

Sun azimuth (degree) 139

ETM+

15b, 30, 60c

8

62.7

129.8

SPOT

10

1

32.4

173.6

QuickBird

0.64d, 2.88e

4

67.8

134.2

a

The thermal band has a resolution of 90m. The panchromatic band has a resolution of 15m. c The thermal band has resolution of 60m. d . The panchromatic has resolution of 0.64m. e . The multi-spectral band has resolution of 2.88 b

3. METHODS The proposed methodology can be divided into the following major component: (1) remote sensing data acquisition and pre-processing, (2) urban biophysical variables extraction, (3) land surface temperature computation, and (4) statistical analysis (Figure 2). LST TM/ETM+ Image

NDVI Water

Quickbird image

Imperviousness

SPOT Image

High-building ratio

Road Map

Road Density

Statistics Analysis

Figure 2. Processing flow for the study.

Relationship Between LST and Urban Spatial Distribution

URBAN DISTRIBUTION AND HEAT ISLAND

463

3.1 Calculation of NDVI NDVI is calculated from the pixel values of the TM image:

NDVI =

ρ (band 4) − ρ (band 3) ρ (band 4) + ρ (band 3)

(1)

ρ

is band reflectivity. where The reflectivity is calculated with the following equation (Chander and Markham 2003):

π ⋅ Lλ ⋅ d 2 ρP = E0 cos θ z where

ρP

(2)

is unit less planetary reflectance,

Lλ is the spectral radiance at E0

sensor’s aperture, d is the earth-sun distance in astronomical units, exoatmospheric irradiances,

is mean solar

θ z is the zenithal solar angle. All above parameters can

be obtained from the literature (Chander and Markham, 2003), except for can be retrieved from the head file of TM or ETM+.

θ z , which

3.2 Extraction of water and road An unsupervised classification with the ISODATA (Iterative Self-Organizing Data Analysis) algorithm was conducted to classify the Landsat TM, and finally water and non-water were identified. Through a standard procedure described by Congalton (1991), classification accuracy was determined to be 84.62% with a Kappa index of 0.8532. Road density was calculated by intersecting the road and grid map. 3.3 Estimation of imperviousness Surface imperviousness was estimated through sub-pixel imperviousness mapping developed by Smith (2000) and Yang et al. (2003) with TM and ETM+. The process involved the following steps: (1) training/validation data development using QuickBird, (2) selection of predictive and regression tree modelling and assessment, (3) final spatial modelling and mapping with Landsat TM/ETM+.

464

Z. OUYANG ET AL.

A pan-sharpened QuickBird image with 0.64-m resolution compiled by a resolution merge of the panchromatic image with a multi-spectral image was used as the base calibration data. An unsupervised classification was conducted to classify the QuickBird, and finally five land cover types were identified, including imperviousness, water, vegetated areas, bare soil, and shadow. A binary map was further generated by recoding the impervious surfaces on the land cover map as 1 and the rest as 0, which achieved a classification accuracy greater than 85% with a standard procedure described by Congalton (1991). A gridnetwork file with the identical grid system as the Landsat TM/ETM+ imagery was created. Each grid cell covers a ground area of 900 m2 (30 m × 30 m). Within a 30 m × 30 m grid, all 0.64-m pixels classified as impervious surface in the binary map were enumerated to determine the percent imperviousness for each grid. In this way, a calibration map with continuous imperviousness estimation (in %) was generated. It has a 30-m grid size. A total of 25,000 sample grids were randomly selected and were broken down into 50% training and 50% test data. A number of predictive variables were reported to be remarkable for establishing meaningful statistical model to estimate imperviousness, such as NDVI, brightness, greenness, et al (Smith, 2000; Xiao et al., 2005; Yang et al., 2003; Yang and Liu, 2005). In this study, in addition to the reflective bands from TM and ETM+, NDVI, and the first three components of the Tasseled Cap transformation, an additional band ratio calculated from the radiance values of TM/ETM+ band 5 and band 1 was added as a possible soil moisture indicator helpful in discriminating between concrete and exposed soil (Smith, 2000). Then, these predictive variables were input into Cubist∗, the regression tree program used in this study to produce impervious surface estimates. 3.4 Computation of modified tall-building area ratio Tall buildings can affect urban vertical thermal characteristics. Considering the effects of building height on surface temperature, we proposed a new method to compute the modified tall-building areas.

MHBA = HBA × h

(3)

Where MHBA is modified tall-building area, HBA is tall-building area, h is building height weight. The building heights were delimited from the shadows that were identified on a panchromatic SPOT-image (Cheng and Thiel, 1995). In this study, all buildings were categorized into three levels, i.e., extremely high buildings (>38.4m), medium-high buildings (25.6-38.4m), and low buildings (12.8-25.6m), then h equalled to 3, 2, and 1 respectively. Buildings lower than 12.8 m were not able to be distinguished from the SPOT image.

∗ Use of any trade, product, or company names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Detailed information on Cubist software is available at http://www.rulequest.com/cubist-info.html.

URBAN DISTRIBUTION AND HEAT ISLAND

465

3.5 Derivation of land surface temperature The LST were derived from the corrected ETM+ TIR band (10.44–12.42 μm) and TM TIR band (11.45-12.50μm) that have spatial resolutions of 60m and 120m, respectively. At first, the digital numbers were transformed into absolute radiance (Chander and Markham, 2003; http://ltpwww.gsfc.nasa.gov/IAS/handbook) using

Lλ = ( Lmax − Lmin ) / 255 * DN + Lmin

(4)

L

where λ is the spectral radiance, Lmin and Lmax (mW cm-2 sr-1 µm-1) are spectral radiances for each band at digital numbers 0 and 255, respectively. For TM 5, Lmin and Lmax are the values 0.124 and 1.530 in (mW cm-2 sr-1 µm-1), respectively (Chander and Markham, 2003). For ETM+ the following reference values are given: Low Gain: Lmin = 0.0, Lmax = 17.04 W/(m2 sr µm) High Gain: Lmin = 3.2, Lmax = 12.65 W/(m2 sr µm) The next step was to convert the spectral radiance to a satellite brightness temperature (i.e., blackbody temperature, TB) under the assumption of uniform emissivity. The conversion formula is:

TB =

K2 K1 ln( + 1) Lλ

(5)

where: TB is effective at-satellite temperature in kelvin, K2, K1are calibration

L

constants in Kelvin, λ is spectral radiance at the sensor’s aperture. For TM, K2 = 1260.56, K1 = 607.76 mW cm-2 sr-1 µm-1. For Landsat-7 ETM+, K2 = 1282.71 K, and K1 = 666.09 mW cm-2 sr-1µm-1. The temperature values obtained above are referenced to a black body. The emissivity corrected land surface temperatures (St) were computed as follows (Artis and Carnahan, 1982; Weng et al., 2004):

St =

TB 1 + (λ × TB ρ ) ln ε

(6)

where: k = wavelength of emitted radiance (for which the peak response and the average of the limiting wavelengths (λ = 11.5 µm) will be used (Markham and

Barker, 1985). ρ =1.438 ×10-2 m K. ε is land surface emissivity, which is obtained by using NDVI Thresholds Method (Sobrino et al., 2004) as following:

466

Z. OUYANG ET AL.

ε = ε soil , when NDVI<0.2;

(7)

, when NDVI>0.5;

(8)

ε = ε veg

ε = ε veg Pv + ε soil (1 − Pv )

ε

, when 0.2 ≥NDVI ≤ 0.5

ε

where soil is the soil emissivity, veg is the vegetation of the emissivity, the vegetation propottion obtained according to (Carlson and Ripley, 1997):

⎡ NDVI − NDVI min ⎤ Pv = ⎢ ⎥ ⎣ NDVI max − NDVI min ⎦

(9)

Pv

is

2

(10)

where NDVImax = 0.5, NDVImin= 0.2. Here, soil and vegetation emissivities was estimated as 0.97 and 0.99, respectively (Xiao et al., 2005). 3.6. Statistical Analysis on the relation of LST and impact factors In the course of determination of model equations we used the mean surface temperature (LST) in both seasons and the earlier mentioned basic parameters, such as: mean NDVI, mean imperviousness (I), road density (R), modified highbuilding area (B), and water body (W) as a percentage by cells. At first, multiple stepwise regressions (MSR) were applied to get the s independent variables with statistical significances (P<0.001) and reveal whether the s independent variables had multicollinearity or not. If there was a multicollineatity, then, principal component regressions (PCR) were adopted alternatively, which is the method of combining linear regression with principal component analysis (PCA). PCA can gather highly correlated independent variables into a principal component. All principal components are independent of each other, so that all it does is to transform a set of correlated variables to a set of uncorrelated principal components (Montgomery and Peck, 1992). MSR and PCR were carried out using SPSS 11.0 (Liu et al., 2003). 4. RESULTS AND DISCUSSION 4.1 Characteristics of urban spatial distribution and LST NDVI can be used to infer general vegetation conditions that are important factors to affect LST spatial variations. It is evident from figure 3-4 that NDVI values increase from inner city to outskirts in both seasons. In the outskirt farmland or in the large parks, NDVI ranges from 0.60 up to 0.75, with higher values associated with greater density and greenness of the plant canopy. In the centre of city, surrounding buildings, road values are close to zero while the difference for water bodies has the opposite trend to vegetation, and the index is negative. Overall distribution is

URBAN DISTRIBUTION AND HEAT ISLAND

467

concentric, with areas of high NDVI forming a band between the 4th and 5th ring roads, and is scattered within the 3rd ring road matching the distribution of parks and recreation sites.

Figure 3. Spatial distribution of NDVI on August 31, 2001.

Figure 5 displays the spatial distribution of water and road. Many of water bodies are located in the large parks and others are mainly paddy field situating in out of fifth ringroad. Roads are densely distributing like a net in the whole study area. Estimation of imperviousness is very good because the regression tree modelling results demonstrates that he correlation coefficient is 0.94 with an average error of 8.59%. Visual inspection revealed that the spatial pattern of modelled imperviousness is quite reasonable. Figure 6 displays red areas (high values) at the centre of the study area corresponding to the concrete cover (building, roads, etc.) and in commercial/industrial areas, old residences, and large transportation facilities

468

Z. OUYANG ET AL.

(such as interstate highways, airports, parking lots, etc.). Green areas of low percent imperviousness values are found in the surrounding areas corresponding to more vegetation or large water bodies.

Figure 4. Spatial distribution of NDVI on May 22, 2002.

The distribution of building heights within the city can be delimited from the shadows on a panchromatic SPOT-image, which meet the requirements of the research. Building density is relatively low, except for the central business district within second ringroad, where there are many parks, remnant dwellings, and large water bodies. The highest building areas are found in zone-2 and zone-3, especially in the north part, because many new tall buildings were constructed there in recent years. There are also many tall buildings along main roads and in some village cities (Figure 7).

URBAN DISTRIBUTION AND HEAT ISLAND

469

Figure 5. Spatial distribution of main road and water body.

Table 2. Total area, land surface temperature and NDVI in different regions.

Area (Km2)

Percent Imperviousness (Std. Deviation)

Aug. 31, 2001

May 22, 2002

Mean LST(Std. Deviation) 28.80 (2.11)

Mean NDVI(Std. Deviation) 0.21 (0.12)

Mean LST(Std. Deviation) 40.02 (3.61)

Zone 1

62.52

67.30 (0.25)

Mean NDVI(Std. Deviation) 0.13 (0.11)

Zone 2

96.15

65.90 (0.26)

0.15 (0.12)

28.43 (2.03)

0.21 (0.12)

39.94 (3.39)

Zone 3

143.07

66.00 (0.28)

0.16 (0.13)

28.64 (2.20)

0.19 (0.12)

40.99 (3.08)

Zone 4

364.97

46.40 (0.32)

0.24 (0.17)

27.06 (3.18)

0.22 (0.14)

40.66 (3.92)

Total

446.70

55.42 (0.31)

0.20 (0.16)

27.76 (2.86)

0.21 (0.13)

40.56 (3.67)

Region

470

Z. OUYANG ET AL.

Figure 6. Spatial distribution of percent imperviousness.

In both seasons, the shapes of LST patterns are non-symmetrical and nonconcentric with a heavy band weighted towards the south of the central axis and within the fourth ring road (Figure 8, 9). On Aug. 31, 2001, mean LST is 27.76°C with a standard deviation 2.86. Across the entire study area, LST values increase from zone-4 to zone-1, but mean value in zone-2 is lower than in zone-3. The main reason, perhaps, is that there are more density built-ups in zone-3. The mean LST in May 2002 is higher than in August, which is 40.56°C, while the spatial pattern is similar which appears increasing from outskirts towards the inner urban areas (Table 2). Main differences between the two are found in the outskirts, where LST is higher in May as compared to that in August, which is likely due to vegetation covering conditions.

URBAN DISTRIBUTION AND HEAT ISLAND

471

Figure 7. Spatial distribution of modified high-building area (MHBA) ratio.

Temperature differences between two date are perhaps caused by following reasons: (1) Vegetation can not develop enough high-density cover in May, while they possess more density greenery in August; (2) Spectral differences between ETM+ and TM that are caused because of factors such as atmospheric absorption and scattering, sensor-target-illustration geometry, sensor calibration, and image data processing procedures (Yang and Liu, 2005). 4.2 Relationship between LST and urban biophysical variables Comparison of the effect of the variables on surface temperature can be accomplished via the correlation coefficients with surface temperature, since the larger the correlation coefficients, the stronger the connections between the variables concerned. Table 3 has listed the correlation matrix of the variables with test of significance of the correlation coefficients.

472

Z. OUYANG ET AL.

Figure 8. Spatial distribution of land surface temperature on August 31, 2001. Table 3. Correlations matrix of variables in the study area.

LST 0108 LST 0205 NDVI 0108 NDVI 0205 I W R B

LST 0108 1.000

LST 0205

NDVI 0108

NDVI 0205

0.642**

1.000

-0.819**

-0.269**

1.000

-0.652**

-0.606**

0.730**

1.000

0.904** -0.364** 0.302** 0.302**

0.488** -0.575** 0.091** -0.170**

-0.915** -0.046 -0.312** -0.501**

-0.710** -0.031 -0.172** -0.033

I

1.000 -0.258** 0.357** 0.478**

W

1.000 -0.171** -0.076**

R

1.000 0.246**

B

1.000

Note: Correlation is significant at the 0.01 level (2-tailed). LST0108, NDVI0108, LST0205, NDVI0205 refer to land surface temperature and NDVI on August 31, 2001, May 22, 2002 respectively. I, R, W, B refer to mean percent imperviousness, road density, ratio of water body and modified high-building area.

URBAN DISTRIBUTION AND HEAT ISLAND

473

Figure 9. Spatial distribution of land surface temperature on May 22, 2002.

It can be seen that percent imperviousness has the largest correlation coefficient (r = 0.904) with LST in August 2001, indicating its largest effect on urban surface temperature. NDVI comes next (r = -0.817), indicating strong negative effect on surface temperature. While in May 2002, NDVI has the largest coefficient (r=0.606), followed by water (r=-0.575), and percent imperviousness (r=0.488). It should be noted that ratio of high-building area is negative with LST which is in contrast to common sense.

474

Z. OUYANG ET AL.

4.3 Statistical model equations To examine more closely the effect, multiple stepwise regressions were conducted (Table 4). As the results indicate, the applied parameters and their order entered in the model are different in the two seasons. Just being same with the results of correlation analysis, percent imperviousness is the most important predictor in August with R2 of 0.817, which means it’s able to explain 81.7% of the abovementioned relationship, while NDVI has more of an impact in May, with R2 of 0.368. Table 4. Values of the stepwise correlation by grid cells and their significance levels. R

R2

Significance level

I

0.904

0.817

0.10%

I, B

0.916

0.839

0.10%

I, B, W

0.925

0.856

0.10%

I, B, W, NDVI

0.936

0.876

0.10%

Date Aug. 31, 2001

May 22, 2002

Parameters entered

NDVI

0.606

0.368

0.10%

NDVI, W

0.849

0.720

0.10%

NDVI, W, B

0.881

0.776

0.10%

NDVI, W, B, R

0.884

0.781

0.10%

Note: I, B, W, R refers to mean percent imperviousness, ratios of modified high-building area, water, and road density.

In general, the process of PCR is similar between August and May, therefore we present here only the process in August. Table 6 shows that the 5th eigenvalue is close to 0 (0.002), condition index is more than 15 (48.211) and the variance proportions of the independent variables I and NDVI are large (0.972 and 0.971). These results clearly indicate that there is collinearity between I and NDVI. Principal component analysis suggests the first two principal components (PCs) can explain 83.401% of the total variation in LST (Table 7). From factor loading matrix, the expression between PCs and primal eight variables is created as follows,

PC = Ax

(11)

where PC is the first four principal components matrix, A is first two principal components loading matrix, x is the matrix constituting by I, B, W, NDVI (Table 7). The standardized principal component regression equation is performed to evaluate the relationship between land surface temperature and the first two PCs:

LST = 0.357PC1 - 0.207PC2

(12)

URBAN DISTRIBUTION AND HEAT ISLAND

475

(R2=0.726, sign. =0.001) Applying equation 11 to equation 12, the final model can be presented as follows:

LST = 27.148 + 3.413I + 3.620B - 8.291W - 6.105NDVI

(13)

where I, B, W is mean percent imperviousness, ratios of modified high-building area, water. Using the same way, LST in May can be modelling by following:

LST = 40.314 - 7.763NDVI - 1.204W + 3.695B + 10.805R (14) where B, W, R is ratios of modified high-building area, water, and road density. It is concluded that LST is positively correlated to percent imperviousness; ratio of tall-building area, but negatively correlated to NDVI and the ratio of water bodies in August. At same time, LST is negatively correlated to NDVI and the ratio of water bodies, and positively correlated with ratio of high-building areas and road density. Table 5. Stepwise multiple regression model between LST and urban biophysical variables. Date Aug. 31, 2001

May 22, 2002

Parameters

B

Constant

28.270

I

4.165

Beta 0.435

57.500

Significance level 0.10%

8.530

0.10%

t

B

-2.951

-0.195

-12.219

0.10%

W

-8.854

-0.292

-14.088

0.10%

NDVI

-11.457

-0.532

-10.569

0.10%

Constant

47.046

233.672

0.10%

NDVI

-24.832

-0.645

-34.812

0.10%

W

-22.668

-0.624

-33.640

0.10%

B

-3.999

-0.220

-11.713

0.10%

R

-2.648

-0.073

-3.760

0.10%

Note: I, B, W, R refers to mean percent imperviousness, ratios of modified high-building area, water, and road density.

5. CONCLUSIONS The spatial distribution of Beijing urban surface features and resulting impacts on the land surface temperature were investigated in the centre of Beijing. The results indicate that: (1) Beijing’s urban spatial pattern presents a typical concentric distribution: NDVI values increase, but impervious surface and tall-building area decrease from inner city to outskirts.

476

Z. OUYANG ET AL. Table 6. Co-linearity diagnosis of the linear regression equation.

Date

Dimension

Eigenvalue

Condition Index

Aug. 31, 2001

1

3.505

1.000

0.000

0.001

0.018

0.003

0.001

2

0.941

1.930

0.000

0.000

0.007

0.414

0.000

3

0.424

2.874

0.000

0.000

0.355

0.012

0.010

4

0.128

5.225

0.001

0.027

0.595

0.001

0.018

5

0.002

48.211

0.999

0.972

0.025

0.570

0.971

Constant

NDVI

W

B

R

1

3.620

1.000

0.003

0.004

0.007

0.021

0.008

2

0.936

1.966

0.000

0.000

0.925

0.008

0.002

3

0.309

3.423

0.009

0.024

0.019

0.924

0.012

4

0.111

5.701

0.007

0.170

0.022

0.045

0.675

5

0.023

12.602

0.982

0.801

0.028

0.002

0.304

May 22, 2002

Variance Proportions Constant

I

B

W

NDVI

Note: I, B, W, R refers to mean percent imperviousness, ratios of modified high-building area, water, and road density. Table 7. The eigenvalue, % of variance and principal component loading matrix. Date Aug. 31, 2001

May 22, 2002

Component

Eigenvalue

Variance (%)

Cumulative (%)

Variables I

B

Factor 1

2.309

57.715

57.715

0.948

0.713

Factor 2

1.027

25.686

83.401

-0.085

Factor 3

0.628

15.708

99.109

-0.276

Factor 4

0.036

0.891

100.000

0.131 NDVI

W

NDVI

-0.196

-0.929

0.077

0.980

-0.234

0.697

-0.017

0.256

0.005

0.042

0.129

W

B

R

Factor 1

1.378

34.462

34.462

-0.378

-0.453

0.637

0.790

Factor 2

1.035

25.867

60.328

0.794

-0.621

0.116

-0.070

Factor 3

0.904

22.594

82.922

0.354

0.580

0.664

-0.033

Factor 4

0.683

17.078

100.000

0.318

0.269

-0.374

0.608

Note: I, B, W, and R refers to mean percent imperviousness, ratios of modified high-building area, water, and road density

(2) The shapes of LST patterns are non-symmetrical and non-concentric with a heavy band weighted towards the south of the central axis and within the fourth ring road. (3) On the basis of our statistical analysis we have proved a strong linear relationship between LST and the studied urban parameters such as percent imperviousness, NDVI, ratio of water, high building and road density, though they demonstrate different characteristics in different seasons. In August, the percentage of impervious surfaces has the largest positive correlation with LST which can

URBAN DISTRIBUTION AND HEAT ISLAND

477

explain 81.7% of LST variance. NDVI comes next with strong negative correlation. In May, NDVI and water are two first negatively correlated with LST with R2 of 0.720. These findings can be used to suggest measures for mitigating urban heat island effects in urban planning. ACKNOWLEDGEMENT This study is supported by the Project of Knowledge Innovation CAS (No. KZCX32SW 2424). REFERENCES Artis, D.A. and Carnahan, W.H. (1982). Survey of emissivity variability in thermography of urban areas. Remote Sensing of Environment, 12, 313-329. Carlson, T.N. and Ripley, D.A. (1997). On the relation between NDVI, fractional vegetation cover, and leaf area index. Remote Sensing of Environment, 62, 241-252. Chander, G. and Markham, B. (2003). Revised Landsat-5 TM radiometric calibration procedures and postcalibration dynamic ranges. Ieee Transactions on Geoscience and Remote Sensing, 41, 26742677. Cheng, F. and Thiel, K.H. (1995). Delimiting the building heights in a city from the shadow in a panchromatic SPOT-Image - Part 1. Test of forty-two buildings. International Journal of Remote Sensing, 16, 409-415. Congalton, R.G. (1991). A review of assessing the accuracy of classifications of remotely sensed data. Remote Sensing of Environment, 37, 35-46. Dousset, B. and Gourmelon, F. (2003). Satellite multi-sensor data analysis of urban surface temperatures and landcover. ISPRS Journal of Photogrammetry and Remote Sensing, 58, 43-54. Liu, R.X., Kuang, J., Gong, Q. and Hou, X.L. (2003). Principal component regression analysis with SPSS. Computer Methods and Programs in Biomedicine, 71, 141-147. Markham, B.L. and Barker, J.K. (1985). Spectral characteristics of the Landsat Thematic Mapper sensors International Journal of Remote Sensing, 6, 697-716. Montgomery, D.C. and Peck, E.A. (1992). Introduction to Linear Regression Analysis. John Wiley & Sons, New York. Nichol, J. (2005). Remote sensing of urban heat islands by day and night. Photogrammetric Engineering and Remote Sensing, 71, 613-621. Smith, A.J. (2000). Subpixel estimates of impervious surface cover using Landsat TM Imagery. In Geography Department, vol. M.A. Scholarly Paper: University of Maryland, College Park. Sobrino, J.A., Jimenez-Munoz, J.C. and Paolini, L. (2004). Land surface temperature retrieval from LANDSAT TM 5. Remote Sensing of Environment, 90, 434-440. Song, Y.L. and Zhang, S.Y. (2003). The study on heat island effect in Beijing during last 40 years. Chinese Journal of Eco-Agriculture, 11, 126-129. Streutker, D.R. (2003). Satellite-measured growth of the urban heat island of Houston, Texas. Remote Sensing of Environment, 85, 282-289. Unger, J., Sumeghy, Z., Gulyas, A., Bottyan, Z. and Mucsi, L. (2001). Land-use and meteorological aspects of the urban heat island. Meteorological Applications, 8, 189-194. Voogt, J. A. and Oke, T. R. (1998). Effects of urban surface geometry on remotely-sensed surface temperature. International Journal of Remote Sensing, 19, 895-920. Weng, Q. (2001). A remote sensing-GIS evaluation of urban expansion and its impact on surface temperature in the Zhujiang Delta, China. International Journal of Remote Sensing, 22, 1999-2014. Weng, Q., Lu, D. and Schubring, J. (2004). Estimation of land surface temperature-vegetation abundance relationship for urban heat island studies. Remote Sensing of Environment, 89, 467-483. Xiao, R., Ouyang, Z., Li, W., Zhang, Z. and Gregory, T.-J. (2005). A review of the eco-environmental consequences of urban heat islands. Acta Ecologica Sinica, 25, 2055-2060.

478

Z. OUYANG ET AL.

Yang, L. M., Xian, G., Klaver, J. M. and Deal, B. (2003). Urban land-cover change detection through sub-pixel imperviousness mapping using remotely sensed data. Photogrammetric Engineering and Remote Sensing, 69, 1003-1010. Yang, X. and Liu, Z. (2005). Use of satellite-derived landscape imperviousness index to characterize urban spatial growth. Computers, Environment and Urban Systems, 29, 524-540.

CHAPTER 28

CONNECTIVITY ANALYSES OF AVIFAUNA IN URBAN AREAS

H. HASHIMOTO Graduate School of Agriculture, Kyoto University, Kyoto, Japan

Abstract. Some resident birds widen their habitat ranges during the non-breeding season. Their wintering habitats may be determined not only by quality of habitats but also by their connectivity to source habitats. In this study, we extract some possible stepping-stone corridors and evaluate which corridors explain the presence/absence of four resident birds–the Bush Warbler, Long-tailed Tit, Varied Tit, and Masked Grosbeak–in possible wintering habitats in Kyoto City. Bird surveys were conducted in 18 woods (> 0.6 ha): almost all-potential breeding patches for these four species in the urban area of Kyoto. I assume the mountain forests around the city and patches in which each bird species was recorded during the breeding season to be their source habitats, and assess whether corridors connect their source and possible wintering habitats. A 15 × 15 m resolution vegetation cover ratio (VCR) map was derived from Terra/ASTER and Quick Bird images. We assumed the minimum VCR for stepping-stones to be 0.25 and created buffer zones from each stepping-stone at four distances: 50 m, 75 m, 100 m, and 125 m. Four types of corridors were extracted from the contiguous buffer zones. The maximum interval between stepping-stones was twice the distance of the created buffer. The maximum interval for suitable corridors evaluated by maximum Cohen’s kappa, maximum overall prediction success, and minimum interval for which more than 80% of the actual wintering habitats connect to the source habitats were, respectively, 200m (kappa 0.25), 200250 m (75% success), and 200m (89% connection) for the Bush Warbler; 100 m (0.12, 58%) and 150 m (80%) for the Long-tailed Tit; 100 m (1.0, 100%, 100%) for the Varied Tit; 100 m (0.29, 64%) and 250 m (100%) for the Masked Grosbeak. Though further analysis may be necessary, these results give us some indication regarding the size of the interval between stepping-stones necessary for planning ecological corridors.

1. INTRODUCTION The concept of an ecological network has become popular among urban planners in Japan, but suitable landscape structures for the ecological corridors of woodland birds have not been verified by actual data. This section describes a preliminary study about stepping-stone corridors for birds in urban areas (Hashimoto et al., 2004). Birds have ability to fly over unsuitable habitats, and 479 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 479–488. © 2007 Springer.

480

H. HASHIMOTO

therefore they do not need contiguous greenways for their movements. They move between habitats using street trees, small parks and so on. Scattered woods distributed within the dispersal range of species can also function as movement corridors for birds. We call this type of ecological corridor a “stepping-stone” corridor. When considering ecological corridors for woodland birds in urban areas, we focused on corridors that facilitate either natal dispersal or seasonal movement of resident birds. This is because, compared with resident birds, long-distance migratory birds should have a greater capacity for long-distance flight. Some resident birds widen their habitat ranges during the wintering season. Their wintering habitats may be determined by the quality of the habitat and also by the connection to source (breeding) habitats. Therefore, the seasonal distribution of these birds should function as a good indicator in evaluating the connectivity of urban woods and ecological networks in urban areas. In this study, we extracted information on potential stepping-stone corridors using remote sensing data and GIS, and we evaluated which corridors best explained the presence/absence of four resident woodland birds–the Bush Warbler (Cettia diphone), Long-tailed Tit (Aegithalos caudatus), Varied Tit (Parus varius), and Masked Grosbeak (Eophona personata) –in possible wintering habitats in Kyoto City. 2. STUDY AREAS AND METHODS 2.1 Study area and materials The study area was Kyoto City, which is located in central Japan, and is surrounded by low montane forests (Figure 1). Bird surveys were conducted in 18 fragmented woods (> 0.6 ha), which represent practically all the possible breeding patches for the four species of birds in the urban area of Kyoto City. The vegetation types of these woods vary between broad-leaved deciduous or evergreen forest, and evergreen coniferous forest; however we did not take into account the differences in vegetation types. We chose four resident woodland birds–the Bush Warbler, Long-tailed Tit, Varied Tit, and Masked Grosbeak–to evaluate stepping-stone corridors in the urban area of Kyoto City. Bush Warblers grow to 14–15.5 cm in length and inhabit dense bamboo thickets, forest undergrowth, edge of cultivation, and shrubby grasslands (CCJB, 2000). Long-tailed Tits are 13.5 cm in length and inhabit deciduous and mixed woods (CCBJ, 2000). Varied Tits are 14 cm in length and inhabit deciduous, mixed, and evergreen broad-leaved forests (CCBJ, 2000). Masked Grosbeaks (Japanese Grosbeaks) are 25 cm in length and inhabit deciduous and mixed woods, open woodlands, and gardens (CCBJ, 2000).

CONNECTIVITY ANALYSES OF URBAN AVIFAUNA HABITAT

481

Figure 1. A vegetation cover ratio map of Kyoto City. Black areas are mountain areas, dark grey areas are those grids having a vegetation cover ratio greater than 0.50 and light grey areas are those grids having a vegetation cover ratio greater than 0.25.

2.2 Procedures 2.2.1 Bird data Bird survey was conducted in 18 woods during the breeding season (twice from May to July) and the wintering season (twice from December to February) of 2002 and 2003. The original data from these surveys were published in Hashimoto et al. (2005). 2.2.2 Creating a vegetation cover ratio map An approximate 15 × 15 m resolution vegetation cover ratio map was derived from Terra/ASTER (ERSDAC) and Quick Bird (Digital Globe) remotely sensed images.

482

H. HASHIMOTO

The supervised data from the tree cover map used to estimate the vegetation ratio was defined as more than 156 of the rescaled NDVI (0–255) derived from a 5 × 5 km Quick Bird image, which is a 2.4 × 2.4 m resolution multi-spectral image sensed in November 2003. A logistic curve model was fitted by regression analysis for the relationship between the rescaled NDVI (0–255) derived from a Terra/ASTER image sensed in June 2003 and the supervised data of vegetation cover ratio derived from the Quick Bird image. The vegetation cover ratio of each grid cell was then estimated by the equation obtained from the rescaled NDVI image derived from the Terra/ASTER image. 2.3 Data analysis 2.3.1 Extraction of stepping-stone corridors We assumed the minimum vegetation-cover ratio for a 15 × 15 m grid for a stepping-stone to be 0.25 (“15 m–0.25” stepping-stones) or 0.50 (“15 m–0.50”), or 0.50 in a 30 × 30 m grid (“30 m–0.50”). Further, we created buffer zones at seven distances from each stepping-stone–50 m, 75 m, 100 m, 125 m, 150 m, 200 m and 250 m. Then 21 types of corridor were extracted from the contiguous buffer zones. Note that the maximum interval between stepping-stones is twice the distance of the created buffers, and the maximum intervals were 100 m, 150 m, 200 m, 250 m, 300 m, 400 m, and 500 m (Figure 2). Distance from stepping-stone (m)

Under 250 m

Under 100 m

Figure 2. Buffer zones for each stepping-stone at seven distances. Grids printed in green are stepping-stones. The maximum interval between stepping-stones is twice the distance of the created buffers.

CONNECTIVITY ANALYSES OF URBAN AVIFAUNA HABITAT

483

2.3.2 Evaluation of extracted corridors We assumed that the mountain forests around the city and patches in which each bird species was recorded during the breeding season were their source habitats, and assessed whether or not the source and possible wintering habitats were connected by stepping-stone corridors. If a patch was connected to source habitats by corridors, we predicted it as a positive (presence) patch during the wintering season. Several indices for evaluating presence-absence models in ecology have been proposed (Manel et al., 2001). In this study, the maximum interval for a suitable corridor was evaluated by Cohen’s kappa, sensitivity (percentage of true positives correctly predicted), and negative predictive power (percentage of predicted absences that were real). The equations for these indices are shown in Table 1.

Figure 3. Distribution of source (“1” and mountain areas) and wintering patches (“2”) for the four bird species—(a) Bush Warbler (Cettia diphone), (b) Long-tailed Tit (Aegithalos caudatus), (c) Varied Tit (Parus varius), and (d) Masked Grosbeak (Eophona personata)—in Kyoto City. “0” indicates unrecorded woods.

484

H. HASHIMOTO

Table 1. Equations of the indices used to evaluate the presence-absence models, where a is the number of positive predictive successes, b is the number of commission errors, c is the number of omission errors, d is the number of negative predictive successes, and n is the total number of a, b, c, and d. Cohen’s kappa [(a + b) – (((a + b)(a + c) + (b + d)(c + d))/n)]/[n – (((a + c)(a + b)+ (b + d)(c + d))/n)] Sensitivity A/(a + b) * 100 Negative predictive power d/(c + d) * 100

Figure 4. Extracted stepping-stone corridors in Kyoto City. Grids printed in green are stepping-stones or forests.

3. RESULTS AND DISCUSSION 3.1 Distribution of the for bird species in Kyoto Seasonal distribution maps for the four bird species are shown in Figure 3. Each species was distributed in large woods during the breeding season and then widened its distribution during the wintering season.

CONNECTIVITY ANALYSES OF URBAN AVIFAUNA HABITAT

485

3.2 Study area and materials The logistic curve model shown below was the best-fit model for the relationship between the vegetation cover ratio and the rescaled NDVI (0 - 255) derived from the Terra/ASTER image (R2 = 0.613). The estimated vegetation cover ratio map is shown as Figure 1. ln {[Vegetation cover ratio] / (1 – [Vegetation cover ratio])}= –15.715 + 0.0958 [rescaled NDVI]

3.3 Distribution of the four bird species in Kyoto On the basis of the buffering of stepping-stone grids derived from GIS, 21 types of stepping-stone corridor were extracted (Figure 4). All 18 woods were connected to the source habitats of each species by 125 m buffer corridors (250 m interval) for 15 m–0.25 stepping-stones or 250 m buffer corridors (500 m interval) for 15 m–0.50 stepping-stones and 30 m–0.50 stepping-stones. Indices for evaluating the model’s performance for each species and each maximum interval are shown in Table 2 (the value of the indices for the 300 and 400 m intervals are identical). Although Manel et al. (2001) suggested that Cohen’s kappa is a simple but effective statistic for evaluating or comparing presence-absence models, sensitivity and negative predictive power may also be effective for evaluation in this study because our analysis considered only connectivity. The quality of habitats and distance between source and wintering habitats also affect the distribution of each species, thus, the predicted presence may have often been incorrect and Cohen’s kappa was very low. In this study, when the sensitivity of an interval was 100%, all wintering habitats were connected to the source habitats by stepping-stone corridors. On the other hand, when the negative predicting power of an interval was 100% or zero (indicated as “–” in Table 2), not all wintering habitats were predicted as absent. Thus, 100% sensitivity and negative predicting power occurred simultaneously. We used sensitivity greater than 80%and Cohen’s kappa to evaluate the maximum tolerable intervals for the following three reasons: (1) ecological data often includes some errors; (2) thresholds should be determined at safety levels for the conservation of wildlife; (3) negative predicting power often becomes small when the number of patches predicted as absent is small. As a result, the maximum tolerable intervals in stepping-stone corridors for each species are 200 m (15 m–0.25) or 250 m (30 m–0.50) for the Bush Warbler, 300 m (15 m–0.50 or 30 m–0.50) for the Long-tailed Tit and 100 m (15 m–0.25) for the Varied Tit. Based on the result of this study it was not possible to determine stepping-stone intervals for the Masked Grosbeak. Masked Grosbeaks have a larger body size than the other three species studied, and they disperse over a wide range in flocks during the wintering season, searching woodland for nuts or seeds (Nakamura and Nakamura, 1995). Thus, a

486

H. HASHIMOTO

250–300 m interval between stepping-stones may be a tolerable distance for the daily movements of this species. The maximum interval for the Varied Tit (100 m) was the least of the four species. However, some Varied Tits winter in large parks in large cities such as Osaka in western Japan (Hashimoto, unpublished); thus, they can potentially fly over intervals of several hundreds meters. Both the quality and the carrying capacity of wintering habitats in a region may also affect the distribution of birds in the wintering season. Shrines and gardens in the foot of mountainous areas in Kyoto have great capacity as wintering habitats for Varied Tits; thus, they may not come to unattractive small woods in the centre of the urban area. Most of the intervals between stepping-stones for the Bush Warbler were less than 200 m or 250 m; however, one individual was recorded in a patch linked with other source habitats by a stepping-stone corridor with an interval greater than 300 m (Figure 3a). It is generally believed that these warblers move in bushes or thickets during their migration and have limited flight abilities. However, populations living in the northern area of East Asia migrate to Japan, Ryukyu Archipelago in the south (Kajita et al., 2002) and to Southeast Asia for wintering (CCJB, 2000). Therefore, some individuals of this species have the ability to fly long distances. Consequently, we have to take into account the fact that some of the Bush Warblers wintering in urban woods could be migrants from the northern part of Japan or Russia. Long-tailed Tits group in flocks consisting of several families that remain together for almost the entire year (Ezaki et al., 1991), and they move around urban districts using street trees and electric cables (personal observations). Although the maximum interval between stepping-stone corridors for this bird was 300 m, a fairy long distance, I consider flocks of Long-tailed Tits in urban areas as evidence of the significant role of street trees as corridors for these birds. 4. CONCLUSION Although this study is a rudimentary analysis of ecological corridors, and other variables such as the productivity of source habitats and the distance between source and wintering habitats, should be analysed, the results indicate the maximum tolerable intervals (gaps) in stepping-stone corridors for each of studied species. Although there are many woods in shrines, temples, and private gardens in Kyoto that play important roles as stepping-stones, the density of urban parks by standard arrangement in Japan (five parks per km2 area) is not sufficient for woodland birds in terms of stepping-stone corridors because the distance between parks is often approximately 500 m. Dense street trees or additional small parks in which small birds can shelter from predators should be considered when planning ecological networks in urban areas.

Table 2. Indices for evaluating the model performance of each species and each maximum interval. Cohen's kappa

Index Maximum 100 m intervals (15 m–0.25 VCR) Bush 0.20 Warbler Long-tailed 0.12 Tit 1.00 Varied Tit Masked 0.29 Grosbeak (15 m–0.50 VCR) Bush 0.20 Warbler Long-tailed -0.32 Tit 0.00 Varied Tit Masked 0.04 Grosbeak (30 m–0.50 VCR) Bush 0.13 Warbler Long-tailed –0.17 Tit 0.00 Varied Tit –0.11 Masked Grosbeak

150 m

200 m

250 m

0.11

0.25

0.08

–0.05

Sensitivity (%) 300 m

500 m

100 m

150 m

200 m

250 m

0.0

60.0

77.8

88.9

0.0

40.0

80.0

Negative predictive power (%) 300 m

500 m

100 m

150 m

200 m

250 m

300 m

500 m

100

33.3

33.3

50.0

–

–

–

80.0

100

63.5

66.7

50.0

–

–

–

0.53

0.33

0.0

100

100

100

100

100

100

100

–

–

–

–0.15

–0.27

0.0

71.4

71.4

71.4

100

66.7

33.3

0.0

–

–

–

0.08

–0.17

0.11

–0.14

0.0

33.3

44.4

44.4

77.8

88.9

100

33.3

28.6

16.7

33.3

0.0

–

–0.32

–0.31

0.03

0.11

0.0

0.0

0.0

25.0

75.0

100

100

60.0

55.6

50.0

66.7

100

–

0.38

0.75

0.53

0.16

0.0

0.0

33.3

100

100

100

100

62.5

71.4

100

100

100

–

0.02

0.21

–0.15

–0.13

0.0

50.0

57.1

55.6

71.4

85.7

100

75.0

62.5

50.0

33.3

0.0

–

0.29

0.08

0 .25

–0.14

0.0

22.2

44.4

44.4

88.9

88.9

100

30.0

37.5

28.6

50.0

0.0

–

–0.17

–0.32

–0.08

0.11

0.0

0.0

0.0

0.0

75.0

100

100

60.0

60.0

55.6

50.0

100

–

0.00 0.32

0.38 0.34

0.33 –0.27

.0 16 –0.13

0.0 0.0

0.0 14.3

0.0 57.1

33.3 71.4

100 71.4

100 85.7

100 100

62.5 50.0

62.5 66.7

71.4 71.4

100 0.0

100 0.0

– –

488

H. HASHIMOTO

ACKNOWLEDGMENTS This work was conducted in collaboration with Mr. J. Dong (Graduate School of Agriculture, Kyoto University), Dr. J. Imanishi, and Prof. Y. Morimoto (Graduate School for Global Environmental Studies, Kyoto University). Dr. K. Murakami (Kishiwada Nature Museum) provided us with essential information about the study sites. The staff of shrine kindly allowed us to conduct our bird surveys within their premises. I am grateful to all these people. This research was partly supported by the Japan Forest Technology Association and by a Grant-in-Aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science and Technology (JSPS Research Fellow 16-945). REFERENCES CCJB (Committee for Check-list of Japanese Birds) (2000). Check-List of Japanese Birds – Sixth Revised Edition. The Ornithological Society of Japan, Hokkaido, 345. (in Japanese) Ezaki, Y., Miyazawa, N. and Sakikawa, A. (1991). Disintegration and reorganization of the flock of Long-tailed Tits Aegithalos caudatus in an urban district in Kyoto, Japan. Japanese Journal of Ornithology, 40, 1–13. Hashimoto, H., Dong, J., Imanishi, J. and Morimoto, Y. (2004). Extraction of stepping-stone corridors for birds in urban areas using remote sensing and GIS. Conservation and Management of Fragmented Forest Landscape (pp. 68-71), Proceedings of IUFRO International Workshop on Landscape Ecology 2004. Hashimoto, H., Murakami, K. and Morimoto, Y. (2005). Relative species – area relationship and nestedness pattern of woodland birds in urban area of Kyoto City. Landscape Ecology and Management, 10(1), 25–35. (in Japanese with English abstract) Kajita, M., Mano, T. and Satoh, F. (2002). Two forms of Bush Warbler Cettia diphone occur on Okinawajima Island: Re-evaluation of C. d. riukieuensis and C. d. restricta by mu1tivariate analyses. Journal of the Yamashina Institute for Ornithology, 33, 148–167. (in Japanese with English abstract) Manel, S., Williams, H.C. and Ormerod, A.S.J. (2001). Evaluating presence-absence models in ecology: the need for prevalence. Journal of Applied Ecology, 38, 921–931. Nakamura, T. and Nakamura, M. (1995). Birds’ Life in Japan with Color Pictures: Birds of Mountain, Woodland and Field. Hoikusha, Osaka, 301. (in Japanese)

CHAPTER 29

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES FOR LAND MANAGEMENT AND POLICIES

J.-E. KIM1, S.-K. HONG2, N. NAKAGOSHI1 1

Graduate School for International Development and Cooperation, Hiroshima University, 1-5-1 Kagamiyam, Higashi-Hiroshima, 739-8529, Japan; 2Institute of Island Culture, Mokpo National University, 61 Dorim-ri, Cheonggye-myeon, Muangun, Jeonnam 534-729, Korea

Abstract. Rural landscapes have a long history of human impacts. The course of history was established by an interrelationship between nature and humans and rural landscape changes occur in the changing inter-relationship between them. In particular, changes in human impacts due to socio-economic changes have been the main driving force in the world since the industrial revolution. These have affected the rural ecosystem entity, such as by decreasing biodiversity, decreasing cultural diversity, destroying amenities, and so on. Theses changes have been occurring since about the 1950’s in European countries, and they led to the start of landscape ecological studies on the effects of rural land use changes. In current, the same phenomenon has also occurred in East Asian countries. Therefore, landscape ecological studies of rural land use changes in European countries give a basic model of Asian rural landscape studies. Korea and Japan is examined as examples of rural landscape studies in East Asian countries. Finally, we believed that this paper is helpful in the understanding of land history and policy of East Asian rural landscapes.

1. INTRODUCTION The word “rural” is derived from the Latin word for “countryside” which is “rusruris”. It is usually applied in relation to farming as an economic activity and used in contrast with the term “urban”. These dictionary definitions, however, are only a partial help in developing a practical working definition of the concept (Golley and Bellot, 1999). Rural landscapes are often used to study the whole system of human activities in agriculture, as well as nature (Kamada and Nakagoshi, 1996; Nassauer, 1995; Hong, 489 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 489–504. © 2007 Springer.

490

J.-E. KIM

ET AL.

1998; Moser et al., 2002; Lee, 2004). Rural landscape has been established by the reciprocal relationship between nature and humans over a long period of time. Actually, rural forests serve as the sink of resources used by humans for daily life and agriculture (Hong, 1998; Bastian, 1999; Nakagoshi and Hong, 2001; Lee, 2004). Humans have survived by creating agricultural fields by burning woodlands, expanding agricultural fields by managing water, gathering wood and grass for fertilizer, fuel, or construction materials, and fishing in rivers and lakes. As a result, some species have died out, while others have had their habitat reduced. On the other hand, other species have taken advantage of the new conditions and have expanded their habitat. Many of the successful species were those that adapted to natural disturbances, such as wild fire caused by volcanic activity, landslides caused by earthquakes, and strong winds and flooding caused by typhoons and hurricanes etc. The ability of these species to deal with natural disturbances affected their ability able to deal with later human disturbances. Rural landscape is sometimes called “cultural landscape”, because human activities in rural ecosystems create special features in rural landscape, such as hedgerows, pollards, terraced paddy fields and so on (Ausad, 1988; Nassauer, 1995; Oreszczyn, 2000; Nakagoshi and Hong, 2001). These special scenes made by the reciprocal relationship between nature and human activities have a unique and special ecological function. Moreover, these unique and special scenes in rural landscapes lead people to come to rural areas for tourism (Hunziker and Kienast, 1999; Gustafson et al., 2005). However, landscape changes are the main issue nowadays in the world, because changes in human impacts have been inducing landscape changes all over the world. The main driving force in landscape change is the human impact on frequency and intensity of rural landscapes (Brandt et al., 1999; Turner et al., 2001; Smailes et al., 2002). Rural landscape has clearly been affected by changing human impacts in this regard. For example, litter removal through understory management allows the germination of plants with small seeds, species that are important components of well-managed rural forests. On the other hand, bigger seeds that are produced from successionally mature species can germinate in litter (Takeuchi et al., 2003; Mou et al., 2005). This is an indication of one way that understory management can affect rural vegetation composition. Finally, human impacts affect habitat environments and then influence biodiversity at the gene level in rural landscapes (Takeuchi et al., 2003). These changes in agricultural systems and socio-economic environments in rural landscapes have led to changes in rural ecosystems (Antrop, 1997; Poudevigne et al., 1997). These changes have induced the usage of natural resources in rural landscapes. People nowadays do not need as much fuel, housing materials, and charcoal and organic fertilizer and so on from rural forests as before. Changes in rural landscape have occurred frequently during the last millennium all over the world. The rise in agricultural productivity per unit of area (ha) and per man hour has increased dramatically during the last century and is expected to continue to do so during the first period of the this century, at least in the majority of

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES AND POLICIES

491

the agricultural areas of Europe. In other parts of the world, such as Asian countries, these increases have occurred. In this paper, we examined the trend of rural landscape studies in Europe based on previous studies. In addition, we examine East Asian rural landscapes, such as Korea and Japan, to observe strategies and trends of studies and policies in rural landscape changes and management. We also suggest ideas for rural landscape management for the future. 2. TRENDS IN EUROPEAN RURAL LANDSCAPE RESEARCHES The history of agriculture is millennia old (Rabbinge and van Diepen, 2000), and this is one of the main landscape types in European countries. Rural landscape is a man-nature ecosystem. We can distinguish several periods of contraction and expansion of cultivated areas and these expansions and contractions are have been repeated due to changing human impacts (Figure 1). These changes have inevitably influenced the rural landscapes.

Expansion

1000

1100

Expansion

Expansion

1200 1300 1400 Contraction

1500

1600

1700 1800 1900

Contraction

2000

Contraction

Figure 1. Expansion and contraction of cultivated area in Europe (Rabbinge and van Diepen 2000).

European rural landscapes most recently changed drastically around the 1940s. Agricultural areas in Europe have decreased over the last forty years by about 13 % (Rounsevell et al., 2003), the main causes of which are industrialization and urbanization (Antrop, 2000; Waldhardt et al., 2003; Antrop, 2004: Palang et al., 2005). However, agricultural productivity has increased food production but there has been a growing demand for food and a decrease farm labour at the same time. Agricultural efficiency improved due to mechanization, chemical fertilizers, development transportation systems and so on. Therefore, traditional agricultural systems have been modernized to those that do not require much human physical strength. Traditional human activities have made created unique cultural landscapes and suitable habitat environments. Small-scale mosaics of grassland and arable fields, highly diverse of habitats, have undergone a wide range of changes since agricultural production became widely displaced by extensively managed grasslands or forests from around the 1940s (Buhler-Natour and Herzog, 1999; Palang et al., 2005). These developed agricultural systems affect the component

492

J.-E. KIM

ET AL.

biodiversity. In particular, floral diversity has decreased hugely over the last few decades in areas (Waldhardt et al., 2003) due to the abandonment of rural landscapes which cause habitat environment changes. In response to this, scientists of many European countries have carried out various studies to identify means of conducting sustainable rural landscape management which can help to improve the environments due to changing rural landscapes then to be applied in Asian countries. The Common Agricultural Policy (CAP) in the EU was established in 1962. This mainly aimed to increase agricultural productivity by promoting technical progress and stabilizing markets. In 1992, CAP recognized the need of contributing to environmentally sustainable rural management, especially in terms of food quality and the role of farmers. The last reform of the CAP took place in 1999. This led to a more sustainable management policy from the perspective of both natural and cultural heritage for the development of rural areas and improvement of the environments (Ernoult et al., 2003; Pinto-Correia et al., 2006). However, it neglected to allow for a variety of place-related approaches. According to Pinto-Correia et al. (2006), many European countries, in national strategies for rural landscape management and conservation have been particularly pressured to over-unify. Moreover, rural landscape changes due to urbanization in each country or region have shown varieties of time lag and spatial scale (Antrop, 2000; Waldhardt et al., 2003; Antrop, 2004; Palang et al., 2005; Pinto-Correia et al., 2006). These changes of time lag were shown from south to north and west to east in Europe (Antrop, 2004). Therefore, policies of sustainable rural landscape must be concerned not only with ecological perspectives but also other perspectives, such as social, economic, historical, and cultural. 3. TRENDS IN EAST ASIAN RURAL LANDSCAPE RESEARCH 3.1 Korean rural landscapes The traditional Korean rural landscape is a combination of high mountains and gentle slopes of low mountains, and houses, and plains with agricultural fields and streams (Hong, 2001; Hong et al., 2001; Lee, 2004, see Figure 2). Spatial and land use patterns of traditional rural landscapes in Korea have been strongly related to ‘Feng-shui’ theory (Hong, 2001; Hong et al., 2001; Lee, 2004). The concept of Feng (風: wind; ‘Poong’ in Korean; ‘Fu’ in Japanese) -shui (水: water; ‘Soo” in Korean; ‘Sui’ in Japanese) was based on long-term experience, optimum location, that is combined with bio-geoecology and socio-economic environments (Forman, 1995; Zonneveld, 1995; Hong et al., 2001; Lee, 2004; Hong et al., in press). People have traditionally believed that fresh air, clean water, forest products, and fertile land originated from spatial arrangements and the pathways of wind and water for comfortable human life.

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES AND POLICIES

493

Figure 2. Traditional feature of spatial pattern in Korean rural landscape due to Feng-shui theory. A: Yangwha-ri, Chungnam (Photo by J.-E. Kim in 2003), B: A schematic crosssectional view of traditional Korean rural landscape according to theory of Feng-shui.

494

J.-E. KIM

ET AL.

Figure 3. Graveyards as rural landscape in Korea. A: Well managed graveyard enclosed by Pinus densiflora trees. B: Managed family graveyard abandoned for about 6 months (Photo by J.-E. Kim in 2003).

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES AND POLICIES

495

Traditional rural forests used to consist mainly of Pinus densiflora and Quercus spp. According to Choung and Hong (2006), P. densiflora in 1987 occupied 10,723 km2, 16.7 % of the total forest area (64,360 km2) in South Korea. It occurred widely (72.4 %) below 300m above sea level. The lowland areas, near houses and agricultural fields, were usually occupied by P. densiflora, and these were heavily influenced by human activities (Kamada and Nakagoshi, 1993; Hong et al., 1995; Hong, 1998; Nakagoshi and Hong, 2001; Kim et al., 2002). Rural forest was used for timber, fuel, and fertilizer. Human activities, such as thinning and litter gathering, control the intensity of sunlight, soil moisture, temperature, and so on. In addition, graveyards in Korean rural landscapes play an important role in cultural landscapes and ecological functions (Hong, 1998; Rim and Hong, 1999; Nakagoshi and Hong, 2001; Kim et al., 2002). Graveyards usually occupy between forest and village (Figure 2). In order to construct graveyards, wood is cut and soil is piled in mounds and covered by grass (Figure 3), especially Zoysia japonica Steud. The graveyards are cleared once or twice a year to remove regenerating plants. Moreover, shrubs and herbaceous plants invading from the adjacent forest are removed to make the graveyards look attractive. However, if graveyards are abandoned, this greatly influences vegetation dynamics. Flora and fauna affect rural forest ecosystems, not only in habitat environments but also at the gene level (Burel and Baudry, 1995; Jeanneret et al., 2003; Washitani, 2001). In other words, changes in rural landscapes have brought changes of habitat environments, which eventually, affect the biodiversity of rural landscape ecosystems (Burel and Baudry, 1995; Forman, 1995; Jeanneret et al., 2003). Much of the current rural landscape in Korea risks becoming an abandoned rural landscape due to changes in socio-economic environments, such as decreasing human depopulation, increasing agricultural efficiency, and an ageing population, which influences secondary rural vegetation (Hong et al., 1995; Hong, 1998; Nakagoshi and Hong, 2001; Turner et al., 2001). The abandonment of rural forests accelerates vegetation dynamics. The Korean government established in 2000 a national land development program to boost the economy in rural areas, which includes introducing ecotourism and industrial infrastructure in rural areas. However, traditional management practices have been neglected in this program. Therefore, many of the various ecological functions and integrity in rural landscapes are rapidly disappearing. From the 1990s, rural landscape studies began in earnest. Rural landscape studies of Korea have focused on several aspects. But these studies have mainly been concerned with rural landscape amenities for tourists. Korean rural landscape management needs to cover multifunctional aspects, not only economic perspectives, but also ecological perspectives, for landscape planning and management in the future. Researchers should provide the government and farmers with useful information about ecological issues, as well as promoting appropriate forest management activities. In addition, long-term ecological studies are necessary for supporting ecosystem management in Korea.

496

J.-E. KIM

ET AL.

3.2 Japanese rural landscapes Japanese people refer to rural landscapes as “Satoyama ( 里 山 )”. Directly translated, this refers to ‘villages’ and ‘mountains’. The concept of Satoyama landscapes is easily understood by Japanese nowadays, as land consisting of agricultural fields and households with mountains (Figure 4), and they are also recognized as a cultural landscape (Fukamachi et al., 2001; Nakagoshi and Hong, 2001; Takeuchi, 2001; Washitani, 2001; Takeuchi et al., 2003).

Figure 4. Rural landscape elements in Minamikata, Hiroshima Prefecture (Photo by J.-E. Kim in 2003).

Rural Japanese vegetation is composed mainly of secondary vegetation, such as pines or mixed deciduous species (Hong et al., 1995; Kamada and Nakagoshi, 1996; Hong, 1998; Washitani, 2001; Takeuchi et al., 2003). This vegetation is useful for human life, for example by providing fuel (Figure 5), fertilizers, timber and so on. Human activities in rural forests have been able to manage physical habitat environments, soil moisture, light intensity, and so on, well (Kato et al., 1997; Takeuchi, 2001; Fujihara et al., 2002; Takeuchi et al., 2003). Traditional rural landscapes show a spatial heterogeneity according to the different successional stages, provides various habitats, and also maintains many plant species. Japanese rural vegetation is known to include rich and unique flora, due to the variety of habitats and diverse physical conditions, especially in understorey vegetation.

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES AND POLICIES

497

Figure 5. Fuel wood collecting in Minamikata, Hiroshima Prefecture (Photo by J.-E. Kim in 2003).

Figure 6. Pinus densiflora dominated rural forest, Miwa-cho, Hiroshima prefecture (Photo by J.-E. Kim in 2004).

498

J.-E. KIM

ET AL.

Changes in socio-economical environments in the 1960s in Japan were brought about by urbanization and industrialization. Therefore, depopulation and the decreasing size of cultivated area of rural areas have also occurred there. In addition, usage of rural forests as resources has decreased and human impacts in rural forests have also decreased, especially in Pinus densiflora forests (Figure 6). P. densiflora forests have been strongly related to human impacts (Hong et al., 1995; Kamada and Nakagoshi, 1996; Hong, 1998; Nakagoshi and Hong, 2001; Fujihara et al., 2002; Takeuchi et al., 2003; Kim et al., 2005). They play an important role as a suitable habitat for light intolerant species. Therefore, P. densiflora is important for the conservation of habitats as well as species diversity. Decreasing human impacts affect the physical environment in rural landscape, controlling the regeneration of P. densiflora. Moreover, pine wilt disease has led to a clear to decrease in P. densiflora in rural forests. Since the 1930s, pine wilt disease has become widespread in Japan (Figure 7). Mamiya (1988) started that around 650,000 ha of pine forest had already been infected in 1988, which is 25 % of Japan’s total area of pine forests. Attempts to control the disease were disregarded and dieback trees were left in forests and these became a source of contamination to healthy pine trees. Abandoned rural forest and pine wilt disease are the main causes of change in rural forests. In addition, plantation also carried out instead of dieback pine trees in rural areas. This was encouraged from government to economical benefit (Figure 8). All trees in plantation were removed and then replaced with, for example, Chamaecyparis obtuse (Japanese cypress) (Figure 9).

Figure 7. Dieback trees of Pinus densiflora in Minamikata , Hiroshima Prefecture (Photo by J.-E. Kim in 2003).

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES AND POLICIES

499

Figure 8. Clear cutting area for plantation in Minamikata, Hiroshima Prefecture (Photo by J.-E. Kim in 2003).

Figure 9. Chamaecyparis obtusa (Japanese cypress) plantation with Sasa species in Minamikata, Hiroshima Prefecture (Photo by J.-E. Kim in 2003).

500

J.-E. KIM

ET AL.

The flora has been threatened in recent years due to disappearing habitats, and at present 17% of vascular plant flora is endangered, according to the “red data book” (The Investigation Committee for Important Plant Species and Communities for Conservation in Japan 1989). Many threatened species are in areas of rural vegetation which has been managed for rural forest, although some species are now in diverse stages of secession (Kamada and Nakagoshi, 1996; Fukamachi et al., 2001). In the 1980s, ecologists considered the effect of rural landscape changes, but in the 1990s were more concerned with biodiversity due to loss of habitats. Many studies have considered the succession of rural landscapes (Hong et al., 1995; Kamada and Nakagoshi, 1996; Hong, 1998; Nakagoshi and Hong, 2001; Fujihara et al., 2002; Washitani, 2001; Takeuchi et al., 2003; Kim et al., 2005). The Japanese Basic Environment Plan of 1994 recognized the importance of rural landscapes. In this plan, the interrelationships between nature and human were considered to be one of the most important long-term goals. Recently, in 2002, the new Biodiversity Strategy was made to improve national biodiversity strategies. It was adopted at the meeting of the Council of Ministers for Global Environment Conservation held on March 27, 2002. In this strategy, the Satoyama ecosystem was chosen as one of three systems in crisis and was also examined from a cultural perspective. It will be developed by the relevant governmental sectors and in partnership with various NGO groups, citizens, and experts and integrated scholars. 4. CONCLUSION Rural landscapes can form transition zones between the controlled human environment of cities and the wilderness areas of flora and fauna. According to the latest report on endangered species, endangered species are found mostly in traditional rural landscapes and forests that are in their natural state. It is well-known fact that rural areas are important for biodiversity. Therefore, changes in rural landscapes are very important now. Traditional human activities, which contributed to the formation of cultural landscapes, especially in rural landscapes, have a role to play in the sustainable management of spatial patterns, biodiversity, and cultural landscape. Therefore, traditional human activities are needed for conservation from the gene level to the landscape level on a global scale. Conservation and sustainable management and planning in rural landscape studies have introduced diverse approaches, such as holistic (Naveh, 2000, 2001; Palang et al., 2000), systematic (Grossmann and Bellot, 1999; Bürgi et al., 2004), integrated (Moss, 2000; Tress et al., 2005), and multifunctional ones (Brandt et al., 2000; Jongman, 2005; Pinto-Correia et al., 2006). The international dimension of rural landscape has been readily recognized for many years through biodiversity, although cultural landscapes have tended to be thought of almost extensively as a national concern (Figure 10). The protection of rural landscapes, and the management of change within them, is primarily a matter for national and local action. Furthermore, there is clearly concern about the cultural perspectives because each country or region displays different geographical features and types of interaction between nature and humans.

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES AND POLICIES

501

Changes of rural landscape due to change of human impact intensity Similar trends of socio-economic changes zDepopulation zDecrease in economic growth zDecrease in agricultural fields zChanges of usage of forest production zUrbanization

20 years

+

Different trends of unique cultural landscape zHedgerows zTerraced paddy fields zDecrease in small scale agricultural fields, Natural monuments and so on

60 years

40 years

Korea • Diversity of forest production • Wild vegetables and charcoal production increased

Japan • Simplicity of forest production • Continuous decrease of forest production

EU • Collapse of cultural landscape • Continuous decrease of forest production

Rural landscape changes Spatial heterogeneity Biodiversity Culture Degree of abondonment of rural landscapes Time scale of abandoned rural landscape in Korea < Japan < EU → Different trend of rural landscape changes

Figure 10. An international schematic model of rural landscape changes due to different human impact intensities.

The particular richness and diversity of rural landscapes, the attraction of visitors from within and without of rural areas, combined with their many cultural associations, makes landscapes a matter of interest and concern to all. In addition, in the case of Asian countries, and even in the EU also, the intensity of human impacts is showing varies depending on the level of economic development. Therefore, when we make strategies and plan for sustainable rural management, we must consider the dissimilarity of cultural and bio-geological between regions and eras.

502

J.-E. KIM

ET AL.

ACKNOWLEDGEMENTS Our special thanks are due to Mr. Martin Stephen Ward at Hiroshima University proofreading this paper.

REFERENCES Antrop, M. 1997. The concept of traditional landscapes as a base for landscape evaluation and planning: the example of Flanders Region. Landscape and Urban Planning, 38, 105-117 Antrop, M. 2000. Changing patterns in the urbanized countryside of Western Europe. Landscape Ecology, 15, 257-270. Antrop, M. 2004. Landscape change and the urbanization process in Europe. Landscape and Urban Planning, 67, 9-26. Austad, I. 1988. Tree pollarding in Western Norway. In Birks, H.H., Birks, H.J.B., Kaland, P.E. and Moe, D. (Eds.) The Cultural Landscape: Past, Present and Future. Cambridge University Press. Cambridge. Bastian, O. 1999. Description and analysis of the natural resource base. In Krönert, R., Baudry, J., Bowler, I.R. and Reenberg, A. (Eds.), Land-use Changes and Their Environmental Impact in Rural Areas in Europe (pp. 43-64). Parthenon Publishing, Paris. Brandt, J., Primdahl, J. and Reenberg, A. 1999. Rural land-use and landscape dynamics – analysis of ‘driving forces’ in space and time. In Krönert, R., Baudry, J., Bowler, I.R. and Reenberg, A. (Eds.), Land-use Changes and Their Environmental Impact in Rural Areas in Europe (pp. 81-102). Parthenon Publishing, Paris. Brandt, J., Tress, B. and Tress, G. (Eds.) 2000. Multifunctional Landscapes. Interdisciplinary Approaches to Landscape Research and Management. Center for Landscape Research, Roskilde. Buhler-Natour, C. and Herzog, C. 1999. Criteria for sustainability and their application at a regional level: the case of clearing islands in the Cubener Heide nature park (Eastern Germany). Landscape and Urban Planning, 46, 51-62. Burel, F. and Baudry, J. 1995. Species biodiversity in changing agricultural landscapes: A case study in the Pays d’Auge, France. Agriculture Ecosystem and Environment, 55, 193-200. Bürgi, M., Herspergerk, A.M. and Schneeberger, N. 2004. Driving forces of landscape change-current and new directions. Landscape Ecology, 19, 857-868. Choung, H.-L. and Hong, S.-K. 2006. Distribution patterns, floristic differentiation and succession of Pinus densiflora forest in South Korea: A perspective at nation-wide scale. Phytocoenologia, 36, 213-229. Ernoult, A. Bureau, F. and Poudevigne I. 2003. Patterns of organization in changing landscapes: implications for the management diversity. Landscape Ecology, 18, 239-251. Forman, R.T.T. 1995. Land Mosaics: The Ecology of Landscapes and Regions. Cambridge University Press, Cambridge. Fujihara, M., Hada, Y. and Toyohara, G. 2002. Changes in the stand structure of a pine forest after rapid growth of Quercus serrata Thunb. Forest Ecology and Management, 170, 55-65. Fukamachi, K., Oku, H. and Nakashizuka, T. 2001. The change of a Satoyama landscape and its causality in Kamiseya, Kyoto Prefecture, Japan between 1907 and 1995. Landscape Ecology, 16, 703-717. Golley, F.B. and Bellot, J. 1999. Planning as a way of achieving sustainable development. In Golley, F.B. and Bellot, J. (Eds.), Rural Planning from an Environmental Systems Perspective (pp. 3-17). Springer, New York. Grossmann, W.D. and Bellot, J. 1999. Systems analysis as a tool for rural planning. In Golley, F.B. and Bellot, J. (Eds.), Rural Planning from An Environmental Systems Perspective (pp. 315-343). Springer, New York. Gustafson, E.J., Hammer, R.B., Radeloff, V.C. and Potts, R.S. 2005. The relationship between environmental amenities and changing human settlement patterns between 1980 and 2000 in the Midwestern USA. Landscape Ecology, 20, 773-789.

INTERNATIONAL TRENDS OF RURAL LANDSCAPE RESEARCHES AND POLICIES

503

Hong, S.-K. 1998. Changes landscape patterns and vegetation process in the Far-Eastern cultural landscapes: human activity on pine-dominated secondary vegetation in Korea and Japan. Phytocoenologia, 28, 45-66. Hong, S.-K. 2001. Factors affecting landscape changes in central Korea: Cultural disturbance of the forested landscape system. In D. van der Zee and I.S. Zonneveld (Eds.), Landscape Ecology Applied in Land Evaluation, Development and Conservation: Some Worldwide Selected Examples (pp. 131147). ITC, Netherlands. Hong, S.-K., Nakagoshi, N. and Kamada, M. 1995. Human impacts on pine-dominated vegetation in rural landscapes in Korea and western Japan. Vegetatio, 116, 161-172. Hong, S.-K., Song, I.-J. and Choi, W.-S. 2001. Theoretical comparison of modern and traditional urban plan: From an Asia landscape ecological planning perspective. Publicationes Instituti Geographici niversitatis Tartuensis, 92, 209-213. Hong, S.-K. Song, I.-J. and Wu J. Feng-shui theory and urban landscape planning. Urban Ecosystems (DOI 10.1007/s11252-006-3263-2) Hunziker, M. and Kienast, F. 1999. Potential impacts of changing agricultural activities on scenic beauty – a prototypical technique for automated rapid assessment. Landscape Ecology, 14, 161-176. Jeanneret, Ph., Schüpbach, B. and Luka, H. 2003. Quantifying the impact of landscape and habitat features on biodiversity in cultivated landscapes. Agriculture, Ecosystem and Environment, 98, 311320. Jongman, R.H.G. 2005. Landscape ecology in land-use planning. In Wiens, J. and Moss, M. (Eds.), Issues and Perspectives in Landscape Ecology (pp. 316-328). Cambridge University Press, Cambridge. Kamada, M. and Nakagoshi, N. 1993. Pine forest structure in a human-dominated landscape system in Korea. Ecological Research, 8, 35-46. Kamada M. and Nakagoshi, N. 1996. Landscape structure and the disturbance regime at three rural regions in Hiroshima Prefecture, Japan. Landscape Ecology, 11, 15-25. Kato, Y., Yokohari, M. and Brown, R.D. 1997. Integration and visualization of the ecological value of rural landscapes in maintaining the physical environment of Japan. Landscape and Urban Planning, 39, 69-82. Kim, J.-E., Hong, S.-K. and Nakagoshi, N. 2002. Landscape ecology on vegetation types and land use systems of agro-forested regions in Korea. Hikobia, 13, 693-703. Kim, J.-E., Tao, T. and Nakagoshi, N. 2005. Vegetation dynamics in Minamikata, Chiyoda-cho, Hiroshima prefecture. Natural History of Nishi-Chugoku Mountains 10·11, 39-55 (Japanese with English abstract). Lee, D. 2004. Ecological Knowledge Embedded in Traditional Korean Landscapes. Seoul National University Press, Seoul, Korea (in Korean with English abstract). Mamiya, Y. 1988. History of pine wilt disease in Japan. Journal of Nematology, 20, 219-226. Moser, D., Zechmeister, H.G., Plutzar, C., Sauberer, N., Wrbka, T. and Grabherr, G. 2002. Landscape patch shape complexity as an effective measure for plant species richness in rural landscapes. Landscape Ecology, 17, 657-669. Moss, M.R. 2000. Interdisciplinarity, landscape ecology and the ‘Transformation of Agricultural Landscapes’. Landscape Ecology, 15, 303-311. Mou, P., Jones, R.H., Guo, D. and Lister, A. 2005. Regeneration strategies, disturbance and plant interactions as organizers of vegetation spatial patterns in a pine forest. Landscape Ecology, 20, 971987. Nakagoshi, N. and Hong, S.-K. 2001. Vegetation and landscape ecology of East Asian ‘Satoyama’. Global Environmental Research, 5, 171-181. Nassauer J.I. 1995. Culture and changing landscape structure. Landscape Ecology, 10, 229-237. Naveh, Z. 2000. What is holistic landscape ecology? A conceptual introduction. Landscape and Urban Planning, 50, 7-26. Naveh, Z. 2001. The major premises for a holistic conception of multifunctional landscapes. Landscape and Urban Planning, 57, 269-284. Oreszczyn, S. 2000. A systems approach to the research of people’s relationships with English hedgerows. Landscape and Urban Planning, 50, 107-117. Palang, H. Mander, Ü. and Naveh, Z. 2000. Holistic landscape ecology in action. Landscape and Urban Planning, 50, 1-6. Palang, H. Printsmann, A. Gyuró, É.K., Urbanc, M., Skowronek E. and Woloszyn W. 2005. The forgotten rural landscapes of Central and Eastern Europe. Landscape Ecology, 20, 645-655.

504

J.-E. KIM

ET AL.

Pinto-Correia, T. Gustavsson R. and Pirnat J. 2006. Bridging the gap between centrally defined policies and local decisions-Towards more sensitive and creative rural landscape management. Landscape Ecology, 21, 333-346. Poudevigne, I., van Rooij, S., Morin, P. and Alard, D. 1997. Dynamics of rural landscapes and their main driving factors: A case study in the Seine Valley, Normandy, France. Landscape and Urban Planning, 38, 93-103. Rim, Y.-D. and Hong, S.-K. 1999. Landscape ecological studies on structure and dynamics of plant populations on vegetation-landscape patterns in rural regions: I. The effect of patch shape on the initial population structure of pine and oaks. The Korean Journal of Ecology, 22(2), 69-77. Rabbinge, R. and van Diepen, C.A. 2000. Changes in agriculture and land use in Europe. European Journal of Agronomy, 13, 85-100. Rounsevell, M.D.A., Annetts, J.E., Audsley, E., Mayr, T. and Reginster, I. 2003. Modelling the spatial distribution of agricultural land use at the regional scale. Agriculture, Ecosystems and Environment, 95, 465-479. Smailes, P.J., Argent, N. and Griffin, T.L.C. 2002. Rural population density: its impact on social and demographic aspects of rural communities. Journal of Rural Studies, 18, 385-404. Takeuchi, K. 2001. Nature conservation strategies for the ‘SATOYAMA’ and ‘SATOCHI’, habitats for secondary nature in Japan. Global Environmental Research, 5, 193-198. Takeuchi, K., Brown, R.D., Washitani, I., Tsunekawa, A. and Yokohari, M. (Eds.). 2003. Satoyama: The Traditional Rural Landscape of Japan. Springer-Verlag, Tokyo. Tress, B., Tress, G. and Fry, G. 2005. Integrative studies on rural landscapes: policy expectations and research practice. Landscape and Urban Planning, 70, 177-191. Turner, M.G., Gardner, R.H., and O’Neill, R.V. 2001. Landscape Ecology: Theory and Practice. Springer-Verlag, New York. Washitani, I. 2001. Traditional sustainable ecosystem ‘Satoyama’ and biodiversity crisis in Japan: conservation ecological perspective. Global Environmental Research, 5, 119-133. Waldhardt, R. Simmering D. and Albrecht, H. 2003. Floristic diversity at the habitat scale in agricultural landscapes of Central Europe - Summary, conclusions and perspectives. Agriculture, Ecosystems and Environment, 98, 79-85. Zonneveld, I.S. 1995. Land Ecology. SPB Academic Publisher, The Hague.

CHAPTER 30

LINKING MAN AND NATURE LANDSCAPE SYSTEMS Landscaping blue-green network

S.-K. HONG Institute of Island Culture, Mokpo National University, Jeonnam 534-729, Korea

Abstract. Of the various concepts of blue-green networking in man-nature systems suggested with regards to the sustainable use of Korea’s natural ecosystems, this paper, in an effort to complement existing theories, analyses the concept of blue-green networking from the standpoint of landscape ecological planning (LEP). In addition, information related to the sustainable conservation and management of natural ecosystems in Gyeonggi province will be presented through an analysis of similar cases in other countries. In particular, in conjunction with the task of formulating measures that can be used to complement and improve the ecological network in the Gyeonggi area, ecological network systems that have been implemented in Europe and Japan were introduced.

1. BACKGROUND The characteristics of land(scape) use in Asia have been historically consistent with the concept of, Poongsoo (風水, ‘Fengshui’ in Chinese). Poongsoo is a combination of the terms ‘poong’ (風, wind), which here refers to green zones (mountains), and ‘soo’ ( 水, water), which connotes lakes and rivers (see Chapter 23 and 29). As wind is formed in the mountains, and water in green zones, the quality and direction of wind as well as the structure of lakes and rivers can be said to be dependent on the geological structure of mountains (Hong et al., 2001; Hong et al., in press). In turn, water quality and the characteristics of aquatic ecosystems and biota, are formed in accordance with the attributes of forests and valleys (Forman, 1995). As such, the concept of ‘Poongsoo’ is not one that whose scope is limited to the simple traditional land use, but rather one that is consistent with a notion of landscape ecology that can be applied to modern land use management policies (Hong 2001; Hong et al., 2001, Table 1).

505 S.-K. Hong, N. Nakagoshi, B.J. Fu and Y. Morimoto (eds.), Landscape Ecological Applications in Man-Influenced Areas: Linking Man and Nature Systems, 505–523. © 2007 Springer.

Table 1. Comparison in concept, component and application of Poongsoo (Fengshui) and modern landscape ecology (from Hong et al., in press). Landscape elements & geophysical attributes

S H A P E

S H A P E L E S S

Functioning effects related to ecological principles

Human impact on land

Land Planning by Feng-shui(風水) Positive (陽氣)

Negative (陰氣)

Mountain (large nature vegetation)

biodiversity microclimate control temperature large home-ranged animal corridor/habitat patches

perforation dissection fragmentation

sliding • gentle slope

rugged • steep slope

Stream

transport / corridor local climate / water supply

straitness channalization cement

clean • gentle • curved

turbid • speedy • strait

transport corridor

straitness / habitat crossing / road kill increasing invasive plants

Horizontally • curved

vertically • direct

Forest (rural small forest)

productivity habitat patches for small mammals

perforation / dissection fragmentation attrition / loss

sparse

dense

Topography

wind /clim ate / temperature

surrounded • flat

distorted • slope

condensed • soft • nutrient

wet • dry • eroded

Road

acid / deposition desertification soil pollution / erosion

Eco-technology

Pseudo-mountain - plantation

Pseudo-stream - increasing curveness - habitat creation

- unpaved - eco-bridge - eco-road

Pseudo-mountain - plantation / hedgerow - windbreak plantation

Soil

productivity flora and fauna

Air

temperature humidity

air pollution

clear • dry

impure • wet

Suitable land assessment by fengshui method

productivity distribution and richness of species

unplanned development

Sunny place

Shadow place

- planned development after E.I.A

biodiversity

high urban heat island greenhouse effect thermal inversion

warm • temperature

cold • rough

Direction

Temperature

LANDSCAPING BLUE-GREEN NETWORK

507

Korea has long been known as a nation in which the basic notion of Chisan chisu (治山治水), which essentially means that the people’s happiness can be secured through the virtuous management of mountains and water sources, has prevailed. This philosophy has historically been reflected in Korea’s national land use policies. However, the excessive development of industries and land wrought by modernization has resulted in bringing about rapid changes in traditional land use practices. The concept of Poongsoo has also undergone a gradual deterioration as well. Korea has traditionally been referred to as Keumsu kangsan (錦繡江山), which loosely translates to the land surrounded by breathtaking lakes and mountains that almost look as if they had been embroidered on silk cloth. However, it is difficult to find such beautiful ecological landscape factors anywhere these days. The ecological efficiency and beauty of the natural environment can only be maximized when the mountains and water sources can naturally connect to one another. Fortunately, following the lead of Seoul Metropolitan City, the practice of urban biotope mapping (see Chapter 13) is now being implemented across Korea’s main cities (Seoul Development Institute, 2001a, b; Seongnam, 2001; Song, 2001; Hong et al., 2005). Of particular interest is the Blue-Green Network Plan implemented by Gyeonggi Province as part of its efforts to connect green zones to lakes and rivers (Gyeonggi Research Institute, 2005). By focusing on this BlueGreen Network Plan, which is regarded as being adaptable to the Korean ecosystem, from the standpoint of landscape ecological planning (LEP), this paper hopes to help foster the advent of a viable ecological network system. 2. NECESSITY OF BLUE AND GREEN NETWORK The basic nature of a landscape system is one in which green spaces and lakes and rivers cannot be separated from one another. Their existence is premised on the maintenance of networks within the wider landscape mosaic (Forman, 1995). The ability of biota in aquatic ecosystems to breed and survive is closely related to the vegetation structure of the green zone within the waterfront. In turn, the composition of vegetation within the waterfront is closely related to the physical characteristics of the general riparian landscape. Meanwhile, the physical characteristics of the riverbed are dependent on the flow speed, flow rate, and physical characteristics of the inflowing water from the upper reaches of the river (Malanson, 1993). Therefore, green zones and the species that exist within them, play an important role in the improvement of the quality of rivers, as well as in increasing biodiversity. As such, the development of a healthy green network naturally contributes to the ecological health of the blue networks they are connected to. In addition, green spaces, as the main patches of the landscape system, and rivers, in their role as ecological corridors, serve as both habitats and sources of biota (Jongman, 1995 see Chapter 4). As the number of national projects such as the construction of new cities, extension of tourism facilities, and the expansion and construction of roads has increased, landscape changes such as reclamation and the fragmentation of forests, habitats, and aquatic ecosystems have become more common. The main natural

508

S.-K. HONG

forests have been altered to plantation forests that feature a low level of biodiversity, and even those forests with a higher density have become ecologically and environmentally isolated. Specifically, the volume of water flowing in the valleys located in green zones has decreased as the number of large-scale development projects, such as the building of golf courses and ski resorts, which involve the clear-cutting of forests has grown; this dry stream phenomenon has even led to an alteration of the physical characteristics of landscapes. What’s more, the climatic changes wrought by global warming are also expected to lead to an increased amount of incidents involving natural disturbances such as high temperatures, typhoons, storms, and heavy snowfalls in Korea. The serious social and economic damage created by this situation is expected to be further compounded by the fact that the ecological restoration process conducted in land usage areas has not kept pace with the speed at which natural disturbances have emerged. Another unfortunate development has been the fact that international and domestic greenbelts, which had heretofore functioned as buffer zones within regional ecological networks, are increasingly being dissolved as part of developmental theories (Nakagoshi and Rim, 1988). Fortunately, some of these concerns have gradually decreased as ecological network planning has been introduced at the national level, and environmental restoration projects have been carried out alongside development schemes. However, the introduction of such concepts at the regional and city levels remains in its initial stages. Nature Conservation

Urban Planning Ecological Network (EU, 1990) Focusing on bio-organism (special migration, habitat connectivity in urban planning strategy)

Landscape Ecology, Green Network and Greenway (USA, 1980) Focusing on traffic- human settlement (large scale urban park, ecological urban-rural gradient)

Biotop and Biotop Network (Germany-Japan, 1990) Focusing on biological habitat in human settlement (increasing connectivity of small biotope, networking civil movement from urban to rural area, environmental education)

Civil Movement

Figure 1. Historical background and cooperation on principles of ecological network and landscape ecology in East and West of the World.

LANDSCAPING BLUE-GREEN NETWORK

509

Under these circumstances, many areas, including Gyeonggi Province, have begun to devise projects related to ecological networks (Gyeonggi Research Institute, 2003, 2005). In particular, in conjunction with the Ministry of the Environment’s implementation of the National Ecological Network (NEN) (Ministry of Environment, 2002), a steady stream of green network projects, which have their origins in the concept of ecological networks, have now begun to be implemented. 3. ECOLOGICAL NETWORK AND THE CURRENT STATUS QUO The concept of the Korean ecological network, which has been developed in accordance with the theoretical research and actual field studies conducted on Korea’s green network, is closely related to the urban landscape planning methods adopted in European countries (esp., Arbeitsgruppe Methodik der Biotopkartierung im besiedelten Bereich, 1993; Sukopp and Wittig, 1993; Bernd et al., 2001). In Japan, urban landscape planning concepts such as biotope and urban ecological networks began to be applied whole or in part by environmental specialists during the 1990s (Figure 1). One finds no fundamental differences between the notions of ecological networks developed in Europe and Japan. The fact that Japan introduced European notions of ecological networks, a continent that features a different ecological environment from a social, cultural, and historical standpoint, to Japanese landscapes, is indeed significant (Nakagoshi et al., 2004; see Chapter 5). However, as Europe features a national planning and management system in which urban planning is closely related to wider regional plans, the conservation and management of urban ecosystems is inherently related to that of the wider regional ecosystems and landscape systems (Song, 2001). The concept of Blue-Green Network developed in Korea is based on an extended ecological network concept that takes into consideration the aquatic environment and landscape factors. Nevertheless, it is not very different from the original European concept of ecological networks (Gyeonggi Research Institute, 2003, 2005). The simplicity of the biodiversity and landscape structure of the European urban ecosystem, where the concepts of landscape ecological planning (LEP) and ecological networks (EN) are well developed, greatly simplifies the process of investigation and assessment associated with the establishment of conservation plans. However, in the Korean case, even the establishment of basic plans for urban areas has proven to be an arduous task. To date, no basic investigation and assessment of the main terrestrial and aquatic ecosystems, which function respectively as the ecological network’s core habitats and ecological corridors, has been completed. In addition, no in-depth study of the biotic population structure of the representative keystone species within the ecological network, or of the physical and environmental characteristics of their habitats, has been undertaken. Amidst the current circumstances characterized by the lack of any basic investigations, the establishment of ecological or green networks based solely on the spatial arrangement of green areas and green area density might be perceived as an acceptable option. However, such an approach might greatly complicate the task of identifying the main landscape elements, such as patches, corridors, and matrixes, as well as their ecological functions. While efforts have been made to overcome this

510

S.-K. HONG

lack of any basic investigations, this author believes that, given the structure of the Korean landscape system, in which the urban, suburban, agricultural, and mountainous landscapes are socially, culturally, economically, and ecologically related to one another, environmental restoration programs should be based on the eco-functional integrity of the following three networks within regional landscapes: green networks such as mountains and forests, water-front networks such as rivers, and semi-nature networks such as agricultural land and secondary forests. In addition, through the conducting of LEP investigations and research on the surrounding natural and social environment, which serve as the network and matrix, efforts should be made to avoid managing these networks in isolation so as to circumvent any negative effect on overall connectivity or circuitry. 4. APPLICATIONS An ‘ecological network’ is a combination of the term ecology, which refers to the study of biological species and their physical environment, with the network concept, which is based on the vertical and horizontal connection of organizations and information within the ecological system. Moreover, it serves as the source of the fundamental concepts of landscape ecology (Jongman, 1995; Hong, 2004). The Ministry of Environment (2002) has defined an ecological network as an ecologically healthy green space which acts as a link between regions featuring superior ecological landscapes, and which helps to secure faunal and floral corridors. An ecological network is composed of those ecological factors which are needed to secure a healthy physical environment for the biological species that exist within the landscape in which human activities are carried out. Such a network can be divided into core zones, corridors, and buffer zones (Forman, 1995; Mun et al., 2005). The establishment of an ecological network is designed to facilitate the preservation of diverse biological species and landscapes, and the coordination of environmental policies pertaining to the conservation of natural ecosystems (Ahern, 1991). Generally, the term ‘green network’ is understood to refer to ecological networks focusing on green areas. The British notion of a green network is in many ways similar to that of the ‘greenbelt’, in that, it is based on the conservation of the urban environment. Meanwhile, in the United States, the term is closer in meaning to the ‘greenway’ concept, which in turn, is based on the forging of linear green zones designed to facilitate the smooth linking of the social and ecological functions of urban, suburban, and agricultural landscapes (Ahern, 1995). More succinctly, while the basic concepts of a green network are based on ecological network theory, its theoretical principles are founded in the theory of island biogeography (MacArthur and Wilson, 1976), the metapopulation concept (Hanski et al., 1995), and the environmental protection movement. In turn, landscape ecology, one of the basic principles of the ecological network, is based on the concept of the metapopulation, which focuses on the location of ecological corridors and the effective arrangement of stepping stones. The establishment of the patch-corridor-matrix model by Forman (1995) has resulted in expanding the traditional focus of environmental conservation studies on biological species and their habitats to include their relationship with adjacent ecological spaces (such as human system). Specifically, the role of the main landscape factors and ecological corridors in facilitating exchanges between biological species and populations is

LANDSCAPING BLUE-GREEN NETWORK

511

increasingly drawing interest. To this end, as a result of cooperation amongst the European nations, an ecological network by the name of Natura 2000 was set up to facilitate the smooth exchange and passage of important biological species (Jongman, 1995). In accordance with the EU’s stated policy of securing the comprehensive management of European ecosystems, the EECONET (The Pan European Ecological Network) was set up to facilitate the integration of biotarelated landscape factors such as biotope size, biotope distance, corridors, and land use (Mander et al., 2003).

Scaling

Political consideration in practice

Ecological consideration of ecosystem health and ecological integrity in practice

Decision of characteristics of EN

Structural EN Mapping of land use pattern, green space and spatial elements

Functional EN Evaluation of ecosystem function Application of Landscape Ecological Analysis

Classification

Selection of network element (Green and Blue)

Survey and Analysis

Disturbance, pollutant, stream quality, species diversity, behaviors of key species, seed dispersal pattern of dominant species, etc.

Supporting data of ecosystem monitoring

Figure 2. Approaching of ecological network (EN) planning and flow of methodology.

The basic notions and outlines of ecological networks are now well known in Korea (Lee and Hong, 2002; Hong et al., 2004; Mun et al., 2005). There are two overarching approaches to the study of ecological networks (Figure 2). The first, the structural network approach, focuses on the analysis of the structure of landscape or ecosystem (see Chapter 5), while the second, the functional network approach, is concerned with the study of the multifunctional attributes of such landscapes and ecosystems by ecological monitoring system (see Chapter 3). Moreover, while the structural network approach is geared towards arranging network factors in accordance with such elements as land use types, the distribution of green areas, and other ecological data; the functional network approach analyses ecological networks using a landscape ecological method (such as spatial analysis, Bogaert and Hong, 2004; see Chapter 2) which takes into consideration those functional attributes of the ecosystem (such as natural disturbances, spread of pollutants, the water quality of rivers, species diversity, faunal behaviour, and the dispersion of plant seeds), which cannot be found on an ecological map. The term green network as it is commonly understood in

512

S.-K. HONG

Korea is closely related to the structural approach to networks. Therefore, in Korea, the term green network is associated solely with a single layer of the ecological network, thus, meaning that only a partial analysis of the overall ecological network is possible. However, a look at international trends reveals that the concepts of ecological and green networks (or greenways) have gradually been accepted as having the survival and movement of biological populations at their root.

Figure 3. Healing the fragmented landscape is the key way of ecological restoration (photo by author at Mt. Dorak, Yangju, Korea).

5. CHARACTERISTICS OF ECOLOGICAL NETWORKS 5.1 Structure If natural ecosystems can be said to consist of dynamic and complicated factors which weave a web between the natural environment and land use practices, then landscape ecology can be defined as the study of the state of these ecosystems at the landscape level. Land use has impacted the overall functions of landscape systems, including the self-purification and assimilative capacities of the environment. In addition, fragmented landscapes affect the quality of habitats, as well as the potential dispersion and movement essential to the survival of biological populations (Figure 3). From an ecological standpoint, the isolation of biological species is one of the main characteristics of urban and

LANDSCAPING BLUE-GREEN NETWORK

513

agricultural landscapes in Korea. Human activities in forest plantation areas have led to the isolation of natural forests. The current natural habitats or seminatural habitats, of which only patches remain, used to be integral parts of huge natural forests (Song et al., 2004). In turn, the biological species which reside in these natural forests are increasingly finding themselves isolated. The isolation and loss of habitats prevents any increase in biological populations and leads to their eventual extinction. Exchanges between biological populations have been curtailed as movement within forests has abated and stream corridors have disappeared. In this regard, a method designed to increase ecological linkages through the establishment of ecological networks has been introduced as a means of invigorating exchanges between biological species. An ecological network has as its main purpose the securing of a level of biodiversity that is higher than what exists within fragmented natural systems by increasing the connectivity of natural preservation areas. These ecological networks consist of core areas, buffer zones, and of the ecological corridors which connect the core areas to the buffer zones (Jongman, 1995). 5.2 Core areas A core area usually consists of those areas which, in accordance with a nation’s conservation policy, have been identified as protected areas. In Europe, where many countries are connected to one another, the majority of countries have accepted the main categories of protected areas set out by the IUCN, and this despite the fact that most have established their own definitions of what constitutes a protected area. Although Denmark does not have any national parks, much of its energy has gone into strengthening the legal protection afforded to the small-scale protected areas known as biotopes (Bernd et al., 2001). Amongst the EU members, one finds various definitions of what constitutes a core area. In certain instances, agricultural areas, afforested areas, and even recreational forests are included amongst the protected areas. While the European strategies for natural conservation are established in accordance with the EU’s national environmental policy (EU Habitat and Species Directive, 92/43/EEC), the design and composition of ecological networks are implemented at the national level, and as such, reflect each nation’s own characteristics. As a result, some concepts are defined based on the traditions and policies of individual countries (Hong, 2004; Hong et al., 2004; see Chapter 10). 5.3 Buffer zones The role of buffer zones, namely, to facilitate overall ecological management, is carried out by controlling human activities in the lands adjacent to the core areas in order to mitigate the potential impact of such activities and the isolation of ecological species. While local residents are usually allowed to reside in buffer zones, such zones are basically considered to fall within the protected areas in the European case. In cases where land development projects may be attempted, legal action can be taken to have the buffer zone expanded.

514

S.-K. HONG

Figure 4. Buffer zone surrounding core area is shrinking by urban development (photo by author at Mt. Bulgok, Yangju, Korea).

In the case of Korea, all areas, with the exception of the national parks which make up the core area, can be regarded as falling within the buffer zone. Unlike Europe, Korea’s green zones, agricultural lands, and the partially managed areas which surround the national parks, can be included within the buffer zone as well. Of particular concern is the fact that the natural barrier which protect the core areas in Korea have been rapidly disappearing as a result of the easing of the regulations governing regional greenbelts. For instance, landscapes in Gyeonggi Province adjacent to Seoul Metropolitan area have undergone rapid changes as a result of varied use of the land, including for the construction of new cities, the large-scale development of residential areas, and the development of tourism areas (Figures 3 and 4). In addition, the green and aquatic spaces that should be protected as buffer zones have also been damaged, as exemplified by the destruction of the waterfront landscape in Lake Paldang where the Bukhan and Namhan Rivers meet.

5.4 Ecological corridors Ecological corridors are exemplified by their ability to heighten connectivity and to, through their ecological coupling effect, create physical structures. These ecological corridors help the dispersion and movement of biological species, and prevent the extinction of local species. Some ecological corridors (such as hedges,

LANDSCAPING BLUE-GREEN NETWORK

515

stone fences, small-sized forests, waterways, and streams) are created through human activities (Figure 5 and 6).

Figure 5. Hedgerow with stone-fence in traditional village is working as ecological corridor for local fauna (photo by author at Woeamri Fork Village, Asan).

Ecological corridors are characterized by their multi-functionality. Thus, not only do they help to maintain the number of species and increase the size of meta-populations by escalating the movement of isolated species, but they also lessen the chance of genetic extinction by decreasing inbreeding. Furthermore, these corridors provide opportunities for migratory birds to lay eggs, and protect biological species from predators and natural disturbances. In the Korean case, ecological corridors can, in accordance with the characteristics of the regional landscape structure, be divided into stream and green eco-corridors. Unlike what is the case with the corridors which cover green areas, the integrated management of stream eco-corridors and core terrestrial protected areas represents a very important factor when it comes to increasing regional biodiversity through the expansion of landscape diversity. In particular, the invigoration of ecological functions and ecological soundness should be considered when selecting the core areas and ecological corridors of the waterfront ecosystemsa decision which has a tremendous influence on the improvement of overall water quality- in the Gyeonggi area, a region which includes the main streams of the Namhan and Bukhan Rivers that serve as the main water sources for the citizens

516

S.-K. HONG

Figure 6. Line corridor of old trees in village. They are working as wind-breaking forest in the rural landscape (photo by author at Anjwa Island, Sinnan, Korea).

around the capital area. Therefore, rather than focusing on conservation and management efforts which are based on the designing of ecological corridors focusing on green areas, efforts should be made to draw up plans which include the aquatic spaces and waterfront areas, which are physically connected to the core area, within the buffer zone and ecological corridor. 6. THE ROLE OF AQUATIC SPACES WITHIN BLUE NETWORK Streams are dynamic forces which continuously change the ecological structure and its functions. In the upper reaches of streams which feature strong currents, the vegetation includes leaves, branches, and decomposed scraps of wood which provide good sources of energy for the main biological species (Mander 1995; Mander et al., 2005). Therefore, the major organic decomposers become the dominant species in the upper reaches of a stream (see Chapter 20). Energy in the form of organic scraps and plants continuously flows toward the lower reaches of the river. Certain segments of streams and rivers demonstrate a reverse flow due to the piling up of such organic materials. As such, the continuity of stream ecosystems is impacted by many materials (Malanson, 1993). Streams are perceived as linear corridors, and as the integrators of complicated ecosystems. Streams are linked to the surrounding environment in many different ways. Therefore, the continuity of streams should be considered from a temporal and spatial standpoint. The dimensional structure of a stream ecosystem is closely related to the speed at which

LANDSCAPING BLUE-GREEN NETWORK

517

the current flows. Meanwhile, the structure of a riverbed, as well as the composition of the biota, differ based on the overflow of streams. Flood plains represent an important element when it comes to fostering the processes of stream ecosystems, and this is especially true with regards to the complexity and quality of stream ecosystems (Nakamura et al., 2004; see Chapter 14). Fishes’ dispersion and movement methods are established based on the prevailing environmental situation, and the characteristics of the species found within a given environment. Fish, mammals, and plants (seeds and nuts) move through these eco-corridors at various speeds and increments. The exchange, distribution, direction and concentration of energy, materials, and nutrients, all of which can influence the biological species found within the natural environment, are established by the interactions between tributary streams and waterfront ecosystems. The conveyance of materials and nutrients represents an important mechanism in the formation of a new landscape as well as in linking biological species to breeding places and populations. As such, the movement of materials, energy, and species within streams can be explained using a spiralling concept based on the recycled use of materials that are produced by streams within the ecosystem. Although this model was designed to deal with the recycling of nutrients, it nevertheless represents an important concept when it comes to the movement of species. 7. INTEGRATION 7.1 Connectivity of blue and green elements Streams are very dynamic systems that play the important role of corridors within landscapes (Jongman and Pungetti, 2004). When diversity is increased in the lower reaches, dynamic is decreased (Malanson, 1993). This means that stream systems become more complicated when many different ecosystems are present in the lower reaches. People use the upper and lower reaches of a stream for different purposes. Streams cannot be regarded as constituting one overarching system. As streams move through various countries, cities, and administrative areas, various interests may be involved, a situation which creates many difficulties in terms of their management and the use of the adjacent land. Nevertheless, streams’ function as the source of nutrients and organic materials, as well as their role in facilitating the movement of various bio-species, should be regarded as the main spatial planning factors that go into the development of ecological networks. The River Continuum Concept as well as the spiralling concept also plays important roles in facilitating a proper understanding of the temporal and spatial ecological processes that take place in streams and rivers. As such, these should also be taken into consideration when planning an ecological network. The roles of aquatic spaces within the ecological network, especially, in terms of the characteristics of streams, are demonstrated in the connectivity and circuitry of waterways (Forman, 1995). These two characteristics help to define the structural and functional characteristics of streams, and are used as indicators with which to understand the ecological functions of streams as habitats and ecological corridors.

518

S .-K. HONG

Figure 7. Mt. Seorak in Baekdudaegan Mountain Range, (photo by author at Naeseorak, Kangwon Province).

7.2 Arrangement of spaces In Korea, ecological environment planning begins with environmental and cultural investigations of the landscape to be developed in accordance with national land development projects and urban planning policies for terrestrial green zones. As such, the ecological uniqueness and diversity of a region is in many at the mercy of the policies which are established. Gyeonggi Province includes several broad aquatic and bio-species zones which are located in branch streams connected to major streams such as the Bukhan and Namhan Rivers (Gyeonggi Research Institute, 2003). In addition, the green networks connected to the Baekdudaegan Mountain Range exhibit superior qualities in terms of their uniqueness, diversity, and scarcity (Figure 7). Furthermore, the outstanding nature of the ecological system found within the DMZ dissecting the Korean peninsula is known the world over (The Ministry of Environment, 2000). From the standpoint of landscape ecology, streams, which act as the corridors; green zones, which serve as habitat patches; the agricultural areas surrounding them; as well as sustainable land uses; should represent the matrix of any land mosaic. In accordance with this theoretical background, when planning the ecological network in Gyeonggi Province, it is required to establish the balanced ecological system that is based on landscape ecological plans which reflect the environmental and ecological factors and characteristics of the Gyeonggi area.

LANDSCAPING BLUE-GREEN NETWORK

519

Figure 8. Han River joining Bukhan and Namhan Rivers is the main ecological corridor in urban blue-green network, Seoul (photo by author at Namsan Tower, Seoul).

The ecological axis, which can be regarded as being composed of the main protected areas in Gyeonggi Province, can be perceived as consisting of the streams and forests along the Bukhan and Namhan Rivers (Figure 8). The cities and villages formed along the streams and forests where the Bukhan and Namhan Rivers converge can be regarded as contact points, which based on their type and size, act as either eco-corridors or core areas. As mountainous and agricultural areas are considered to be the buffer zone which protects the core areas and eco-corridors, as well as to function as the ecotone, there is a need, based on the spatial characteristics of ecological areas, to increase the biodiversity within the landscape by expanding the Blue-Green Network; which, as mentioned above, can serve as the contact point between streams and green zones. 7.3 Development of an ecological network focusing on the main species and habitats Biological species have no administrative boundaries. Their activity areas delineated by ecological boundaries are defined as either green zones or streams. Therefore, green network plans, which are established from the standpoint of humans, can easily lead to damage being caused to the corridors for wildlife species. In this regard, a long-term debate about ecological networks that focuses on the habitats and behaviors of various biological species is essential. In terms of ecological network plans, there is a need to create an environment, in which the biota that is selected can be harmonized with large and small cities and biotopes, and to form and connect the habitats of the main species that reside within the network

S.-K. HONG

520

(Figure 9). To create such an environment, the ecological behaviours and habitats of the main biota, including a survey of their populations, should be analysed. There have been instances abroad in which the GAP and HEP models have been used to conduct spatial analysis of these habitats (Natuhara, 2005, see Chapter 2 and 9).

C la s s if ic a tio n o f E N N a tio n a l s c a le • N a tio n a l o r in te rn a tio n a l e c o lo g ic a l n e tw o rk

M a j o r S u r v e y I te m s in c a s e B a c k b o n e : L a rg e m o u n ta in r a n g e s ( e .g . B a e k d u d a e g a n ), la rg e riv e r, h a b ita t a n d te rr ito r y o f la r g e a n im a l, m a jo r g re e n z o n e

R e g io n a l s c a le • R e g io n a l e c o lo g ic a l n e tw o rk

A x is : N a tu r e c o n s e r v a tio n a r e a , n a tio n a l p a r k , s tre a m a n d r iv e r , h a b ita t a n d te r rito ry o f m id s m a ll s c a le a n im a l, m a jo r p la n t c o m m u n ity

U r b a n s c a le •G r e e n n e tw o r k o r la n d s c a p e n e tw o rk

L a n d s c a p e e le m e n t: u r b a n p a r k , g re e n b e lt, la rg e b o u le v a r d , p u b lic g r e e n s p a c e , c ro p la n d , b io lo g ic a l c o m m u n ity , f lo ra a n d fa u n a

F o c a l s c a le • S m a ll e c o lo g ic a l n e tw o r k

B io t o p e n e tw o r k : u r b a n fo re s t, p o n d , s m a ll g re e n s p a c e a n d w e tla n d , s p e c ie s a n d its h a b ita t

• B io to p e n e tw o r k

Figure 9. Methodology and scaling in several cases of ecological network (EN) plan.

Figure 10. Dangsanje, the traditional praying performance for thanks giving to old trees in village forest (photo by author at Bupsungpo, Jeonnam, Korea).

LANDSCAPING BLUE-GREEN NETWORK

521

Although the characteristics of biological species are usually assessed based on the characteristics of the relevant green zones, the composition of the main wildlife species can also indirectly be ascertained by a look at the composition of the herbaceous layer which these animals consume (Hong et al., 2004). Therefore, an analysis of the relationship between green zones and biota provides an important scientific basis from which to select conservation areas and setup core areas and corridors. 8. CONCLUSION: HUMAN NETWORKING Regional-level ecological networks, which include greenbelts, can be regarded as an indicator of the soundness of environmental policies pertaining to ecological management. Viewed from the standpoint of main populations and habitats, ecological network planning should be based on the establishment of objective and consistent investigation models which can be used whenever cooperation with other regions is needed in terms of ecosystem management. Although many ecological maps, including urban ecological maps, have been established at the regional and city levels, these have employed different data, mapping methods, and backgrounds. To prevent the emergence of such problems, a unified protocol should be developed by the organizations responsible for ecological mapping in order to ensure consistency at the national level. The activation of cooperative policies in which several adjacent areas join hands to establish ecological maps will also contribute to the development of autonomous ecosystem management by encouraging the autonomous participation of local residents which represents one of the overarching goals of those seeking to establish an ecological network (Figure 10). ACKNOWLEDGEMENTS I would like to express my sincere thanks to Dr. Mi-Young Song at Gyeonggi Research Institute, Korea and Dr. Jae-Eun Kim at Hiroshima University for providing valuable references and comments. This research was partially supported from Korea Research Foundation as titled “A Study of the Korea Maritime Cultures by Regional Groups: Tangible Cultural Properties (KRF-2005-005-J13701)”. REFERENCES Ahern, J. 1991. Planning for an extensive open space system: linking landscape structure and function. Landscape and Urban Planning, 21, 131-145. Ahern, J. 1995. Greenway as a planning strategy. Landscape and Urban Planning, 33, 131-156. Arbeitsgruppe Methodik der Biotopkartierung im besiedelten Bereich. 1993. Flaechendeckende Biotopkartierung im besiedelten Bereich als Grundlage einer Naturschutz orientierten Planung. Natur und Landschaft, 68, 491-526. Bernd M, Nygaard B, Ejrnæs R, and Bruun, H.G. 2001. A biotope landscape model for prediction of semi-natural vegetation in Denmark. Ecological Modelling, 139, 221-233. Bogaert, J. and Hong, S-K. 2004. Landscape Ecology: Monitoring landscape dynamics using spatial pattern metrics, In Hong, S.-K. et al. (Eds.). Ecological Issues in a Changing World–Status, Response and Strategy (pp. 103-131). Kluwer Academic Publisher, Netherlands. Forman, R.T.T. 1995. Land Mosaics: Ecology of Regions and Landscapes. Cambridge University Press. 683p.

522

S .-K. HONG

Gyeonggi Research Institute. 2003. Blue Gyeonggi: Green Program 21. Gyeonggi Province. 450p. (in Korean) Gyeonggi Research Institute. 2005. Urban Green Revegetation Plan in Suwon City. Gyeonggi Province. 100p. (in Korean) Hanski, I., Poyry, J., Pakkala, T., and Kuussaarri, M. 1995. Multiple equilibria in metapopulation dynamics. Nature, 377, 18-21. Hong, S.-K. 2001. Factors affecting landscape changes in central Korea: Cultural disturbance of the forested landscape system. In D. van der Zee and I.S. Zonneveld (Eds.), Landscape Ecology Applied in Land Evaluation, Development and Conservation: Some Worldwide Selected Examples (pp. 131147). ITC, Netherlands. Hong, S.-K. 2004. Ecological network for landscape conservation and restoration: from an international perspective”. J. Korea Env. Res. & Reveg. Tech., 7(5), 11-25. Hong, S.-K., Kim, S., Cho, K.-H., Kim, J.-E., Kang, S. and Lee, D. 2004. Ecotope mapping for landscape ecological assessment of habitat and ecosystem”. Ecological Research, 19, 131-139. Hong, S.-K., Song, I.-J., Byun, B.-S., Yoo, S. and Nakagoshi, N. 2005. Applications of biotope mapping for spatial environmental planning and policy: Case studies in urban ecosystems in Korea. Landscape and Ecological Engineering, 1(2), 101-112. Hong, S.-K., Song, I.-J. and Choi, W.-S. 2001. Theoretical comparison of modern and traditional urban plan: From an Asia landscape ecological planning perspective. Publicationes Instituti Geographici Universitatis Tartuensis, 92, 209-213. Hong, S.-K. Song, I.-J. and Wu J. Feng-shui theory and urban landscape planning. Urban Ecosystems, (DOI 10.1007/s11252-006-3263-2) (in press) Jongman, R. 1995. Nature conservation planning in Europe: developing ecological networks. Landscape and Urban Planning, 32, 169-183. Jongman, R. and Pungetti, G. 2004. Ecological Networks and Greenways: Concept, Design, Implementation. Cambridge University Press. 345p. Lee, S.-E. and Hong, S.-K. 2002. Urban Ecological Network Plan. Sigma Press, Seoul. (in Korean) MacArthur, J.W. and Wilson, E.O. 1976. The Theory of Island Biogeography. Princeton University Press, Princeton, NJ. Malanson, G.P. 1993. Riparian Landscapes. Cambridge University Press. 296p. Mander, Ü., Külvik, M. and Jongman, R. 2003. Scaling in territorial ecological networks. Landschap, 20, 113-127. Mander, Ü. 1995. Riparian buffer zones and buffer strips on stream banks: Dimensioning and efficiency assessment from catchments in Estonia. In Eiseltová, M. and Biggs, J. (Eds.) Restoration of Stream Ecosystems (pp. 45-64). IWRB Publication No 37, Slimbridge, Gloucester, UK. Mander, Ü., Hayakawa, Y. and Kuusemets, V. 2005. Purification processes, ecological functions, planning and design of riparian buffer zones in agricultural watersheds. Ecological Engineering, 24 (5), 421-432. Ministry of Environment. 2000. Sustainable Land Use Planning of Nation. Korea Environment Institute, Seoul, 338p. (in Korean) Ministry of Environment. 2002. Report for Promoting Strategy of National Ecological Network. Unpublished paper (in Korean) Mun, S.-K., Seong, H.-C., Ku, B.-H., Byun, B.-S., Yu, H.-S., Lee, D.-K., Lee, S.-M., Lee, E.-Y., Lee, E.H., Lee, J.-J., Jeon, S.-W., and Jeon, Y.-O. 2005. Environmental Planning. Bomoondang, Seoul, 471p. (in Korean) Nakagoshi, N., and Rim, Y.-D. 1988. Landscape ecology in the greenbelt area in Korea. Münstersche Geographische Arbeiten, 29, 247-250. Nakagoshi, N., Watanabe, S. and Koga, T. 2004. Landscape ecological approach for restoration site of natural forest in the Ota River Basin, Japan. In Hong S.-K. et al. (Eds.), Ecological Issues in a Changing World – Status, Response and Strategy (pp. 301-310). Kluwer Academic Publisher, Dordrecht, Netherlands. Nakamura, F., Kameyama, S. and Mizugaki, S. 2004. Rapid shrinkage of Kushiro Mire, the largest mire in Japan, due to increased sedimentation associated with land-use development in the catchment. Catena, 55, 213−229. Natuhara Y. 2005. Landscape evaluation for ecosystem planning. In Proceedings of the 1st International Symposium on Landscape and Ecological Engineering. 6-8 Oct. 2005, Everland, Korea. Seongnam. 2001. City Biotope Map and GIS Data Base. Seongnam, 300p. (in Korean) Seoul Development Institute. 2000a. Biotope Mapping and Guideline for Establishment of Ecopolis in Seoul (I). 245p. (in Korean)

LANDSCAPING BLUE-GREEN NETWORK

523

Seoul Development Institute. 2000b. Biotope Mapping and Guideline for Establishment of Ecopolis in Seoul (II). 394p. Song, I.-J. 2001. Urban ecology and ecological urban planning. In Understanding of Eco-City. Darakbang Publishing Company, Seoul (in Korean) Song, I.-J. 2001. Guidelines for ecological urban management. Urban Affairs, 36, 97-109. (in Korean) Song, I.-J, Hong, S.-K., Kim, H.-O., Byun, B.-S. and Gin, Y. 2005. Pattern of landscape patches and invasion of naturalized plants in urban developed area of Seoul. Landscape and Urban Planning, 70, 205-219. Sukopp, H. and Wittig, R. 1993. Oekologische Stadtplanung. In Sukopp, H. and Wittig, R. (Eds.) Stadtoekologie (pp. 348-373), Gustav Fischer Verlag, Stuttgart.

Index A

B

abiotic process, 25, 34 Acacia mangium, 274, 275 Acanthaceae, 8, 12, 14, 16, 19 adaptive management, 141–4, 437 adjacency, 194 Aegithalos caudatus, 136, 480 aerial photograph, 17, 72, 216, 220, 221, 229 aesthetic values, 375, 396, 407, 408, 410, 414 Africa, 8, 10, 12, 14, 25, 29, 91, 179, 439, 441, 450 Afromontane, 10, 13, 14, 15 agricultural intensification, 164, 401 agricultural landscape, 47–9, 58, 247, 385, 396, 401, 443, 510, 513 agricultural matrix, 396 agriophytes, 237 agroforestry, 273 agro-system, 194 Ailuropoda melanoleuca, 96 Albedo, 460 Algorithm, 21, 28, 86, 212, 436, 446–50, 463 alien species, 149, 325, 447 Alnus firma, 313 Alnus japonica, 214 Amphibians, 63, 259, 424, 441, 445 animal movements, 255, 257, 261 anthropogenic disturbances, 236, 244, 360, 368 area-perimeter ratio, 157 Asian countries, 164, 491, 492, 501 Asian rural landscape, 163, 492–500 Asteraceae, 243 atmospheric deposition, 274, 280, 284, 285, 287, 288 avifauna, 133, 479

Baltic Sea, 418 Bamboo, 96, 99, 140, 166, 169, 170, 480 BARCI, 229 barrier effects, 255, 257–8, 263 barriers, 63, 97, 132, 254, 257, 514 basal area, 304 Bee-Bo forest, 389, 391–3 Beetles, 83, 165, 168, 172, 173, 439 Beijing, 459 Belgium, 256 Bibo (Bee-Bo), 389–94 Biodiversity, 4, 45–7, 58, 65, 81–91, 132, 139, 149–60 biodiversity conservation, 8, 58, 65, 151–2, 154, 196, 392–3, 439–42 biodiversity impact assessment, 84 biodiversity measurement, 441–2 biodiversity monitoring, 45–7 Biodiversity Probability Index, 85 bio-eco-human network, 3 biogeochemistry, 43, 274, 286 biogeographical theory, 442–3, 452 biogeography, 10, 59, 73, 138, 439, 443, 510 biological conservation, 83, 451 biological indicator, 130 biological invasion, 149, 157, 236, 248, 249 biomass, 19, 43, 44, 50, 247, 277, 285, 332–5, 349, 353 Biosphere Reserves, 156 Biotope, 157, 158, 195, 203, 342, 352, 507, 509, 513 biotope map, 157, 196, 507 biotope network, 158 bird species, 62, 91, 131, 133–8, 484–6 bitterling species, 379 525

526

Biwa Lake, 378–9 Biwako, 378 blue network, 158, 507, 516–17 blue-green network, 158–60, 505–7, 519 boundary, 132, 201, 413, 437, 449 boundary complexity, 413 Braun-Blanquet, 317 Breeding, 59, 132–4, 137, 156, 216, 227, 257, 352, 426, 480–1, 507, 517 Britain, 439 broad-leaved evergreen tree, 144 buffer, 103, 107–8, 157, 158, 284, 287, 329, 411, 444, 485, 513–14, 519 buffer strips, 330–1, 338–9, 341–2, 347, 351, 408, 410–11 building height, 462, 464, 468 Burundi, 10 Bush Warbler, 480, 485, 486 Butterfly, 133, 134, 135, 138, 352 C Canada, 136, 398–9, 404, 410–11, 414, 446 Canonical Correspondence Analysis, 133, 168, 214 Canopy, 359, 363, 365, 367, 368 canopy density, 359–73 canopy species, 293, 304, 306 capacity, 5, 62, 81, 265, 286, 340, 346–7, 379, 386, 421–2, 480, 486 Capercaillie, 440 Carabid, 168–9, 171–3 Carabidae, 165 carrying capacity, 421–2, 486 Castanopsis, 130, 293, 306–7, 383 cat-tails, 333, 335 Central Europe, 235–7, 243, 244, 248, 418–19 Central European forest, 235, 236 CENTURY, 277–8 CERN, 37–9, 41–5, 48, 49, 50–2 Cettia diphone, 136, 480 Chamaecyparis obtuse, 114, 498

INDEX

Channelization, 212, 214, 219, 222–3 channelized stream, 214, 219, 222, 224, 227, 392 chemical analysis, 277–8 chemical fertilizer, 114, 314, 401, 491 China, 33, 38–41, 42–4, 45–50, 51 Circuitry, 510, 517 city planning, 292, 294–5, 299 classification, 12, 16, 21–2, 73, 78, 101, 116, 120, 140, 164, 168–70, 173 climate, 14, 36, 38–41, 43, 45, 49, 50, 67, 73, 82, 91, 99, 129, 130, 194, 196, 197, 244, 274, 330, 376, 384, 392, 410, 442, 460 climax species, 316, 320–2 cluster analysis, 73, 76 Cohen’s kappa, 483, 485 Coleoptera, 165, 168, 169 Colonization, 60, 97, 140, 365, 399, 443 Congo, 10, 12, 13, 14 conifer plantation, 74, 78, 246 connectivity, 5, 61, 63, 97, 103, 108–10, 139, 158, 187, 293, 296, 411, 413, 426, 479–86, 517 connectivity analysis, 426, 428, 479–86, 510, 513, 514, 517 conophorium, 366, 373 conservation ecology, 4, 154 conspecific trees, 307 context, 8, 51, 66, 188, 189, 194, 263, 265, 395, 396, 414, 421 core patch, 102, 103–4, 106–7, 108–10 corridor habitats, 255, 256 countryside, 58–9, 399, 489 cropland, 22, 38, 39, 41–3, 91 crossover rate, 258 Cubist, 464 cultural functions, 353 cultural landscapes, 47, 58, 59, 72, 381, 385, 393, 396, 490–1, 495, 496, 500, 510, 513 cultural landscape, 47, 48, 49, 393, 490 cultural patterns, 395–7, 401, 404, 406, 407, 408, 412, 413, 414

INDEX

cultural values, 82, 295, 395, 396 culture, 36, 72, 129, 193, 194, 195, 196–7, 198, 199, 203–4, 226, 376, 378, 393, 396 D data collection, 116–20, 436, 437, 438–44, 451 decision making, 49, 64, 67, 432, 438, 447, 449–50, 451, 452 deforestation, 86, 96, 218, 219, 275 DEM, 101 Denitrification, 338, 343 Denmark, 513 diameter at breast height, 302, 386 digital elevation model, 19, 29, 101 digital image, 19, 481 dipterocarp, 185, 186 disturbance, 90, 257 Diversity Index, 83, 84, 86, 91 DLSS method, 324, 325 drained wetlands, 334, 335 driving force, 35, 36, 43, 45, 47, 49, 71, 188, 236, 490 Dutch, 64, 66, 84, 265 E EAFES, 312 East Asia, 163–5, 173, 312, 486, 491, 492–500 eco-functional integrity, 495, 510 ecological axis, 519 ecological corridor, 59, 61, 63, 64, 152, 158, 419, 421, 479–80, 486, 507, 509, 510, 514–16 ecological function, 4, 35, 36, 45, 59, 62, 79, 157, 158, 173, 292, 306, 308, 314, 395–6, 397, 398, 407, 411, 413, 414, 490, 495, 509, 510, 515, 517 ecological impacts, 268, 418 ecological integrity, 4, 51, 399, 411, 435 ecological modeling, 43, 50, 52, 57, 65, 432 ecological network, 57–67, 154, 158, 265, 292, 342, 353, 419, 421,

527

422, 426, 479, 480, 507, 508, 509–10, 511, 512–16, 519–21 ecological perception, 375 ecological planning, 3, 158, 507, 509 ecological processes, 1, 3, 4, 8, 29, 34, 35, 36, 45, 47–9, 51, 52, 184, 396, 437, 438, 442–4, 451, 452, 453, 517 ecological restoration, 3, 4, 211, 226, 231, 448, 452, 453, 508 ecological solution, 51, 254, 437 ecological succession, 132, 292, 315, 382, 383 ecological value, 81, 82 Ecologically Scaled Landscape Indices, 60 ecology-management-planning, 2 ECOnet, 64 economic development, 39, 51, 78, 96, 164, 180, 181, 189 economic value, 82 ecoregion, 49 ecosystem assessment, 50, 51 ecosystem creation, 1, 67, 211, 452 ecosystem degradation, 34, 210, 450 ecosystem evaluation, 152, 432 ecosystem level, 4, 59 ecosystem management approach, 435 ecosystem model, 277, 278 ecosystem modeling, 4 ecosystem process, 88, 152, 154, 156, 452 ecosystem restoration, 3, 47, 210, 220 ecosystem-level, 4, 59, 64, 81–3, 85, 92, 275, 279, 285 ecotone, 343, 377–9, 392, 441, 519 ecotope, 83, 90, 91, 139, 140, 157 ecotope map, 157 edge, 40, 85, 86, 96, 107, 133, 136, 140, 150, 155, 157, 186, 255, 257, 306, 351, 392, 399, 413 edge effect, 155, 157, 255, 257, 306, 351 edge species, 85, 133, 136, 140, 150 EECONET, 418, 419, 426, 511 Elaeis guineensis, 179 emerging ecosystems, 247

528

endangered species, 60, 96, 97, 98, 110, 140, 151, 153, 154, 158, 352, 379, 381, 500 endemic woody species, 316, 322 energy flow, 129, 389 engineering technology, 312 environmental biology, 157 environmental change, 34, 35, 42, 50, 91, 274 environmental diversity, 441, 452 environmental impact assessment, 72, 158, 432 environmental policy, 160, 513 Eophona personata, 480, 483 Erosion, 37, 40, 47, 51, 67, 154, 155, 203, 219, 223, 275, 287, 312, 314, 325, 408 ESLI, 60 Estonia, 65, 333, 334, 335, 340, 341, 349, 359 EU, 1, 268, 351, 418, 432, 492, 501, 503 Europe, 29, 58, 59, 64, 160, 173, 211, 236, 237, 280, 396, 441, 491, 509, 513 European landscapes, 57 Eutrophication, 37, 223, 274, 287, 351, 393 Evaluation, 3–4, 7, 52, 81, 82, 83–8, 91, 97, 100–2, 104, 129, 139, 140, 432, 483–4 Evapotranspiration, 48, 187, 230, 460 Evergreen, 49, 78, 99, 130, 134, 144, 304, 317–19, 480 extra-moenia, 402 F Fagus, 75, 130, 248 Fallopia, 247, 249 fast-growing, 39, 275, 276, 313, 370 Fengshui, 389, 390 Fertilizer, 48, 114, 286, 287, 288, 350, 490, 495, 496 fine-grained texture, 397 fine-scale heterogeneity, 138 flagship species, 65, 440

INDEX

flood, 196, 215, 218, 219, 329, 351, 420, 517 flood control, 196, 211, 212, 219, 229, 231, 392, 418 flood protection, 418, 420 floodplain, 84, 211, 330, 337, 351, 352, 353, 419 forest connectivity, 158 forest ecosystem, 4, 46, 49, 50, 235–7, 243–5, 246–7, 248, 378, 498 forest fire, 359, 373 forest management, 188 forest regeneration, 246, 247 forest-type habitats, 169, 172 Fourier, 17, 23, 25 Fourier transform, 20 Fractal, 376, 385–7 fractal property, 386, 387 fragmentation, 28, 45, 58, 61, 85, 104–6, 132, 140, 150, 187, 262, 507 fuel, 114, 275, 292, 301, 302, 333, 399, 490, 495, 496 functional connectivity, 414 functional network, 511 G GAP analysis, 103–4, 110, 436, 445–6, 447, 452, 453 Garden City, 375–6, 387 genetic level, 149 geographic information system, 114, 220 geomorphology, 36, 210, 222 geoprocessing, 121 Gerdyksterwei, 265–7 Germany, 160, 236, 237, 243, 244, 247, 249 Germination, 214, 225, 366, 490 giant panda, 95, 96, 99–101, 106, 109, 110 GIS, 41, 47, 64, 86, 103, 106, 110, 114, 220, 221, 224, 460, 480, 485 global change, 2, 34, 41, 45, 46, 47, 235

INDEX

global climate change, 45, 50, 274 global warming, 335, 368, 508 global warming potential, 335 globalization, 353 government, 5, 34, 72, 156, 219, 220, 265, 404, 495, 498, 500 Great Hing′an Mountains, 360 Green Conservation Area, 295–6, 298, 299, 308 green network, 507–9, 510, 511, 512, 518, 519 green space, 66, 200, 291, 292, 297–9, 301–2, 435, 438, 442, 443, 445–6, 449 green space design, 435, 442, 447, 449, 450, 451, 453 green zones, 296, 507, 508, 518, 519, 521 greenbelt, 508, 510, 514, 521 green-water gradients, 5 greenway, 59, 378, 480, 510 Greenway Hypothesis, 378 grid map, 71, 463 groundwater tables, 227, 229, 230 Grus japonesis, 216 GTOS, 37 Guineo-Congolian region, 10, 13, 14 Gyeonggi Province, 507, 509, 514 H Habitat, 59, 60, 63, 89, 90, 95, 100–2, 104, 130–2, 134–8, 141–3, 154, 163, 219–20, 306–7 habitat diversity, 63, 87, 136, 138, 163, 164, 173, 408 habitat fragmentation, 104, 132, 262, 263 habitat function, 293, 295, 306–8, 353 habitat loss, 96, 255–6, 269, 414, 450 habitat network, 59, 158, 160 habitat ranges, 480, 491 habitat restoration, 436, 444, 453 habitat sensitivity, 136 habitat suitability, 100–3, 104, 106 habitat suitability index, 100, 101, 104

529

heat island, 459 heterogeneity, 35–6, 138, 339 heterogeneous landscape, 34, 97, 150, 211, 414 hierarchical approach, 152–4 highways, 263, 264, 315, 316 hirosato wetland, 226–31 Hokkaido, 76, 219, 222 home range, 103, 104, 105, 137, 259, 263, 421 human activity, 2, 34, 96, 185, 257, 396, 399 human impact, 222, 491, 498, 501 human networking, 521 human system, 160, 510 human-dominated landscapes, 258, 274 human-influenced area, 218 human-modified habitats, 164 Hungary, 460 hydrologic flows, 278 I ILTER, 37 ILWIS, 106 Indonesia, 57, 179, 180 Industrialization, 280, 491, 498 Infrastructure, 85, 139, 255, 256, 257, 262, 269, 402, 423, 425 Integration, 2, 58, 188, 517–21 Interdisciplinary, 1, 29, 57, 160 interdisciplinary solutions, 1 interior species, 133, 136, 140 interior-to-edge, 413 international trend, 489–91, 512 invertebrates, 163, 165, 172, 173 island-biogeography, 131, 138, 156, 443, 510 ISODATA, 463 Isolation, 150, 157, 187, 306, 513 Italy, 64, 396–8, 401–3, 408 IUCN, 84, 153, 156, 513 J Japan, 71, 113, 209, 311 Japanese alder, 214, 220, 230

530

Japanese cypress, 498 Japanese garden, 375–6, 377–6, 379–84 Jinsokusokuzu, 114, 116, 122, 127 K Kangnam, 197, 198 Katsura, 382–5, 386, 387 Kendall Rank Correlation, 361, 365 Kitakyushu, 291, 293 Korea, 43, 158, 164, 201, 313, 389, 390–2, 492, 507, 509, 511–12 Kushiro Mire, 212–16, 219–22 Kyoto, 375–8, 384, 484–6 L lake ecosystem, 218–19, 223–6 land amelioration, 334 land cover, 21–2, 72, 99, 101, 131, 301, 464 land cover map, 8, 85, 294, 464 land mosaics, 131–2, 518 land ownership, 188, 436, 436, 445, 448, 452 land planning, 5, 506 land surface temperature, 460, 462, 465, 474, 475 land transformation process, 8, 26, 154 land use capability, 188 land use change, 113–14, 333, 335, 402 land use planning, 179, 184–9 land use policy, 64, 189 land value, 448 landform, 120–2, 124, 126–7 Landsat, 212, 462, 463, 464, 465 landscape analysis, 3, 36–7 landscape change, 36–7, 71–3, 490–2, 507 landscape configuration, 4, 413, 444, 453 landscape connectivity, 97, 103, 106, 108–9, 110 landscape dynamics, 8, 9, 28, 49–50 landscape ecological application, 1–5

INDEX

landscape ecological planning, 158, 507, 509 landscape ecologists, 8, 29, 36, 51, 72 landscape ecology, 4, 7–9, 29, 33–4, 35–7, 47, 47, 50–1, 57, 95, 149–53, 184–9 landscape elements, 293–9, 300–7, 404 landscape factor, 100, 104, 149, 507, 509, 510, 511 landscape function, 330, 331–2 landscape functioning, 329–33, 442 landscape heterogeneity, 35–6, 87, 150 landscape linkage, 59, 443 landscape management, 273–88, 330, 491, 492 landscape matrix, 4 landscape metrics, 9, 29, 35, 187, 188 landscape modeling, 51 landscape mosaics, 3, 66, 132–3, 150, 154, 158, 397, 507 landscape pattern, 8–9, 36, 38–42 landscape pattern dynamics, 8, 9, 35, 49–50 landscape planning, 5, 390, 495, 509 landscape research, 489–90, 491–2, 493–500 landscape stability, 150 landscape structure, 4, 21, 35, 187–8, 376–7, 479, 515 landscape transformation, 9, 26–8 landscape-level, 154, 184, 275, 329, 330, 444, 453, 500, 512 land-use managers, 436, 438, 447, 448 LARCH, 60, 61, 421–3, 428 Lauraceae, 304, 307 leguminous shrubs, 314, 315 LEMP, 2 LEP, 507, 509, 510 linear corridors, 516 litterfall, 279, 280, 283, 285, 287 living energy, 390 local populations, 91, 264, 268, 425, 427 London, 460 Long-tailed Tit, 480, 485, 486

INDEX

long-term ecological research, 34, 275 long-term monitoring, 35, 36, 37, 40, 41, 44, 49, 50, 51 long-term stabilization, 313 LOS, 156 LST, 460, 465, 466–7, 471, 476 LTER, 275, 276, 277 Lyonia ovalifolia, 383 M Macrophyte, 217, 342, 348, 349, 350 Malaysia, 179 Man and Biosphere, 156 man-influenced area, 1 marsh, 37, 74, 76 Masked Grosbeak, 480, 485 Matrix, 4, 9, 17, 20, 25, 26, 129, 131, 134, 154, 158, 361, 362, 367, 368, 369, 370, 395–6, 413, 471–2, 474, 476, 509–10, 518 Mediterranean, 397, 398, 408 Metapopulation, 421, 443, 444, 452, 510 Microclimatic, 136, 257, 411 Microhabitat, 136, 138 Mitigation, 253, 263 mixed seeding on slopes, 313 model simulations, 418 mortality, 253, 258 mosaic landscape, 132, 133, 193, 401 motorways, 255, 263, 265 movement death rate, 258, 259 multi-functional landscape, 396 multifunctionality, 63, 65, 331, 353, 495 multihabitat species, 85 multi-scale, 4, 5, 34 multi-scale level, 5 N National Ecological Network, 65, 265, 418, 509 Natura 2000, 418, 419, 428, 431, 432, 511

531

natural succession, 113, 126, 127, 220, 314, 320, 322, 334, 423 naturalness, 74, 76, 78, 158 nature conservation, 59, 139, 236, 246, 247, 248, 249 nature reserves, 96, 97, 103, 105, 110, 158, 263, 351, 419 nature system, 1 NDVI, 46, 461, 463, 464, 465, 466, 467, 473, 474, 475, 477, 478 Neighbourhood, 194 Nestedness, 131, 137, 138 Netherlands, 62, 64, 65, 84, 160, 255, 256, 257 network system, 157, 158, 507 networking, 3, 5, 158, 160, 505, 521 N-fixing, 287 NGO, 64, 65, 226, 428, 500 Nitrogen, 41, 43, 45, 47, 48, 217, 222, 243, 246, 247, 280, 338, 340, 343, 344, 345, 347, 348, 351, 392 nitrogen-fixing, 248 non-indigenous, 235, 247 non-point pollution, 205, 340 nutrient budgets, 286, 330 nutrient fluxes, 274, 275, 276, 330, 343 nutrient-poor, 115, 244, 246, 248 O oil palm, 179, 181, 184 opportunistic, 136 orientation, 17 Osaka Prefecture, 130, 132, 134 P paddy fields, 114, 116, 122, 123, 126, 301, 377, 378, 381, 467, 490 Pan-European Ecological Network, 64 Paradigm, 8, 52, 435 Parus major, 136, 137 Parus varius, 137, 480, 483 passive restoration, 211, 225 patch, 8, 36, 154, 392–3

532

patch analysis, 159, 199 patch isolation, 444 patch structure, 443 patch-corridor-matrix, 510 pattern analysis, 7 pattern and process, 36, 38–50 peat, 185, 186, 215, 219, 247, 314, 334, 335, 339 perimeter, 27, 28, 157, 186, 196, 199, 201, 402 periodic vegetation, 17, 22–3, 25, 29 periodogram, 20, 21 persistence, 135, 396, 407, 436, 437, 438, 442, 444, 445, 447, 448, 450, 451 PHA, 90, 91, 92 Phosphorous, 217, 274, 392 Phragmites, 335 Phragmites communis, 219 physical structures, 514 physiotope, 157 phytocenoses, 74 phytogeography, 9–17 phytosociological method, 72 Pinaceae, 238, 239, 240, 243 pine forest, 166, 239, 244, 246, 440, 445, 498 pine wilt disease, 76, 79, 498 Pinus densiflora, 75, 76, 116, 123, 124, 126, 495, 497 Pinus thunbergii, 75, 76, 316 pioneer species, 286, 306, 315, 320 plant biomass, 277, 279, 288, 349 plant-soil-atmospheric continuum, 274 Płock, 418, 420, 429, 430 Poisson distribution, 260 Poland, 417 Pollution, 130, 150, 180, 203, 222, 224, 257, 269, 274, 286, 333, 335, 340, 343, 350, 393 Polytrichum commune, 384 Poongsoo, 390, 391 population density, 136, 138, 144, 150, 188, 450 population sizes, 15, 62, 157, 217, 258, 426, 443 Population Viability Analysis, 60

INDEX

post-fire, 359 potential habitat, 95 Potential Habitat Analysis, 90 potential natural vegetation, 74, 130, 144, 293 precipitation, 39, 40, 41, 46, 91, 130, 197, 227, 275, 279, 331, 361, 460, 461 Preliminary Environmental Review, 158 Principal Components Analysis, 133, 168 Prunus serotina, 236, 237, 246, 247, 249 PVA, 59, 60, 90, 91, 153 Q quantitative pattern analysis, 19–21 Quercus, 75, 78, 132 Quercus pubescens, 398 Quercus trojana, 398, 405 Quercus variabiris, 383 QuickBird, 462, 463, 464 R Ramsar, 211, 222, 329, 353, 418 range of habitats, 89, 165, 172, 173 rarity algorithms, 447 Red Data, 76, 84, 153, 351 red-crowed cranes, 209 reed-sedge community, 214 regeneration, 351, 365–6, 370 regulatory functions, 336, 343 rehabilitation, 103, 211 remnant habitat patches, 396, 411, 412 remote sensing, 19, 29, 41, 43, 46, 294, 460, 462, 480 remote sensing imagery, 17, 29 renaturalisation, 423, 426, 427, 428, 429 representativeness, 436, 437, 438, 442, 445, 446, 451, 452 residential area, 115, 126, 151, 201, 265, 266, 294, 298, 299, 301, 302, 514

INDEX

restoration, 209, 222, 311, 319, 359, 370 restoration ecology, 4, 50 revegetation engineering, 312 revegetation technology, 311, 312, 313, 325 revegetation work, 311, 312, 313, 314, 322, 323, 325 rice paddy culture, 193, 196, 203 richness algorithms, 447 right-of-way, 256 riparian buffer, 329 riparian buffer zones, 329, 341 riparian ecosystem, 330, 343, 344, 392 riparian forest, 274, 287, 335, 423 river, 222, 417 river continuum, 158, 160, 517 river-floodplain, 222 riverine landscape, 418 road, 253, 255 road crossing, 116, 255, 257, 258, 260, 264, 269 road density, 256, 262, 463, 475 road network, 254, 263, 265, 267, 268, 269, 396 Robinia pseudoacacia, 246, 247, 248, 313 Rosaceae, 243 rural biodiversity, 164 rural landscape, 163, 396, 489, 491, 492 rural landscape research, 489, 492 Russia, 58, 165, 166, 168, 171, 174, 486 Russian Far East, 165, 166, 167, 168, 169, 171 Rwanda, 10 S Saitama Prefecture, 113 Salamandrella keyserlingii, 216 Sasa, 225, 226, 314 satellite image, 19, 212, 214, 220 Satoyama, 114, 132, 292, 381, 496, 500 Scale, 38, 47, 138, 151

533

scaling up, 149 scenario analysis, 418, 421, 423 Scenic Zone, 295, 296, 297, 298, 300, 302, 308 secondary ecosystem, 231 secondary forest, 76, 78, 79, 132, 297, 300, 302, 306, 510 secondary vegetation, 76, 292, 496 secondary woodland, 113, 114, 165 seed-dispersal type, 307, 321, 365 seeding, 311, 312, 313 seedling recruitment, 307 self-similarity, 385 self-sustaining ecosystems, 220 semi-natural, 65, 164, 331, 334, 340, 341, 387, 418 semi-natural ecosystem, 341 Seoul, 193, 197, 200 Seoul Metropolitan, 194, 196, 197, 204 shade-intolerant, 322 Shannon index, 86 Shannon-Wiener, 86 shape index, 138, 196, 199, 201 Siberian salamander, 216, 221 Sichuan Province, 97 Similarity Index, 73, 305 Similarity Matrix, 361, 367, 368, 369 site-selection, 436, 446, 447, 448, 449, 452, 453 site-selection algorithms, 436, 446, 447, 448, 449, 452, 453 slope seeding, 312, 314, 316 slope stability, 319, 322, 323 SLOSS, 156 small-scale landscape, 63 socio-cultural value, 81, 82 socio-economic, 2, 44, 58, 64, 65, 67, 188, 248, 249, 292, 293, 301, 302, 451, 490, 495 socio-economic aspects, 54, 64, 249 socio-economic principles, 2 soil erosion, 37, 40, 47, 48, 51, 196, 203, 224, 226, 275, 287, 408 soil invertebrates, 163 soil moisture, 47, 48, 78, 170, 173, 214, 243, 368, 408, 410, 464, 495–6

534

soil organic matter, 40, 274, 286 soil seepage, 278, 279, 282, 285, 288 soil-improving, 313, 315 source-pool effects, 442, 443, 452 source-sink, 87, 88, 442, 448, 452 South Korea, 164–9, 171, 172, 174, 392, 495 spatial arrangement, 97, 103, 110, 276, 288, 412, 414, 492, 509 spatial attributes, 221, 412, 413, 414 spatial cohesion, 421, 423, 428, 431, 432 spatial composition, 8, 187 spatial geometry, 442 spatial heterogeneity, 2, 29, 34, 35, 36, 496 spatial pattern, 7, 21, 184 spatial structure, 26, 188, 189 spatio-temporal scales, 43, 338 species composition, 90, 136, 142, 144, 224, 225, 292, 443 species diversity, 9 species richness, 86, 87, 90, 91, 132, 133, 134, 137, 138, 140, 142, 246, 247, 439, 440, 441 species-area curves, 134, 135, 305 Sphagnum imbricatum, 230 SPOT, 19, 20, 462, 464, 468 spotted bush, 17, 19, 20, 22, 24, 25 SRTM, 19, 20 Staphylinidae, 165, 168 steep slopes, 224, 311, 314, 410 stemflow, 278, 279 stepping-stone, 482, 485 stewardship, 407 stream ecosystem, 348, 516, 517 Structural Equation Model, 138, 139 structural network, 511 successional stage, 322, 344, 443, 496 Sudan, 19, 23, 25, 29 suitable habitat, 96, 109, 140, 491, 498 Sumatra, 179 surface runoff, 274, 277, 278, 279, 285 sustainability, 273

INDEX

sustainable development, 39, 41, 51, 114, 160, 333 sustainable management, 4, 49, 181, 275, 492, 500 T target populations, 96, 97, 103 taxon, 10, 163 temple, 293, 306, 377, 384, 391, 486 Terra/ASTER, 481, 482, 485 Terrestrialization, 214, 215 TGM method, 316, 317, 321, 324, 325 Thematic Mapper, 213 Throughfall, 277–84 Tibetan Plateau, 96, 98 tiger bush, 17, 22 Tokyo, 114, 130, 136 Topology, 89, 91 Tourism, 66, 353, 490, 495, 507 traditional land use patterns, 116, 164, 505, 507 traditional management, 164, 175, 495 traffic, 255, 260, 262, 264 traffic mortality, 253 traffic volume, 254, 258–60, 262, 264–5, 266–9 traffic-calmed roads, 264 traffic-calmed rural areas, 264 transportation, 253, 268, 418, 467, 491, 506 traversability, 262–7 Typha, 333, 33 Typology, 26 U ubiquitous solutions, 34, 165 umbrella species, 59, 439, 440, 452 UNESCO, 156 United Kingdom, 64 urban area, 196, 479 urban development, 122, 123, 126, 151 urban ecology, 130 urban gradient, 130, 132

INDEX

Urban Green Space, 292, 298, 299 urban habitat, 134–8 urban heat, 460, 506 urban landscape, 129, 139, 141, 193, 291 urban park, 86, 135, 136, 486 urban planning, 83, 158, 196, 461, 477, 508, 509, 518 urban surface features, 475 urbanization, 130, 131, 132, 296 urban-rural gradients, 5, 508 USA, 28, 64, 65, 85, 256, 342, 404, 410 Ussuriisk Nature Reserve, 166, 168 V Varied Tit, 480, 485, 486, 487 vascular plant, 86, 137, 150, 236, 243, 351, 439, 500 vectorial data, 21 VEDI, 418, 432 Vegetation, 17, 74, 313, 319, 320, 481 vegetation community, 194, 320 vegetation cover, 481, 482 vegetation cover ratio, 481–2, 485 vegetation dynamics, 78, 495 vegetation map, 72, 85, 102, 119, 122 vegetation matrix, 20, 25 vegetation patch, 72, 73 vegetation pattern, 17–25 vegetation rehabilitation vegetation transformation, 103–4, 108, 109, 110 vegetation type, 74, 76, 116, 480 vertebrate, 163, 173

535

vineyards, 397, 406, 408, 410 Vistula floodplain, 418, 419, 426 Vistula River, 417, 423 Vulnerability, 9, 15, 90, 91, 445, 450 W Warsaw, 418 Washload, 212, 214, 215 water resource, 34, 203, 211, 212, 330, 420 water table, 213, 214, 215, 227, 228, 229, 230, 330 water turbidity index, 212 watershed hydrology, 210 web-based data, 51 wetland, 329 wildlife movement, 254, 265, 267, 268 Wolong Nature Reserve, 95 woodland birds, 136, 257, 439, 480, 486 Y Yamada Green Park, 300, 302, 303, 308 Yellowstone, 65 Yukon ecological network, 65 Z Zambezian region, 10, 13, 14, 16 zero-tension lysimeters, 277, 278 zone, 329, 341, 392, 513 Zoysia japonica, 318, 495