Principles and Practice of Forest Landscape Restoration · Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (2024)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (1)

Edited by A.C. Newton and N. Tejedor

Principles and Practice of Forest Landscape RestorationCase studies from the drylands of Latin America

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (2)

About IUCN

IUCN, International Union for Conservation of Nature, helps the world find pragmatic solutions to our most pressing environment and development challenges.

IUCN works on biodiversity, climate change, energy, human livelihoods and greening the world economy by supporting scientific research, managing field projects all over the world, and bringing governments, NGOs, the UN and companies together to develop policy, laws and best practice.

IUCN is the world’s oldest and largest global environmental organization, with more than 1,000 government and NGO members and almost 11,000 volunteer experts in some 160 countries. IUCN’s work is supported by over 1,000 staff in 60 offices and hundreds of partners in public, NGO and private sectors around the world.

www.iucn.org

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (3)

Principles and Practice of Forest Landscape RestorationCase studies from the drylands of Latin America

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (4)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (5)

Principles and Practice of Forest Landscape Restoration

Edited by A.C. Newton and N. TejedorCase studies from the drylands of Latin America

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (6)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (7)

This book is dedicated to the memory of Margarito Sánchez Carrada, a student who worked on the research project described in these pages.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (8)

The designation of geographical entities in this book, and the presentation of the material, do not imply the expression of any opinion whatsoever on the part of IUCN or the European Commission concerning the legal status of any country, territory, or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries.

The views expressed in this publication do not necessarily reflect those of IUCN or the European Commission.

This publication has been made possible by funding from the European Commission under the ReForLan project: INCO Contract CT–2006–032132.

Published by: IUCN, Gland, Switzerland

Copyright: ©2011 International Union for Conservation of Nature and Natural Resources

Reproduction of this publication for educational or other non-commercial purposes is authorized without prior written permission from the copyright holder provided the source is fully acknowledged.

Reproduction of this publication for resale or other commercial purposes is prohibited without prior written permission of the copyright holder.

Citation: Newton, A.C. and Tejedor, N. (Eds.). (2011). Principles and Practice of Forest Landscape Restoration: Case studies from the drylands of Latin America. Gland, Switzerland: IUCN. xxvi + 383 pp.

ISBN: 978-2-8317-1340-3

Photos: Front cover: ©N. Ramirez-Marcial; back cover: ©C. Echeverria

Layout by: Bookcraft Ltd, Stroud, Gloucestershire, UK

Printed by: Page Bros Ltd, Norfolk, UK

Available from: IUCN Publications Services, Rue Mauverney 28, 1196 Gland, Switzerland Tel: +41 22 999 0000, Fax: +41 22 999 0020 E-mail: [emailprotected] www.iucn.org/knowledge/publications_doc/publications/

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (9)

vii

CONTENTS

List of boxes ixForeword xiiiAcknowledgements xivContributors xvAbbreviations xxiii

1 Introduction 1 A.C. Newton and N. Tejedor

2 Assessing the current extent and recent loss of dryland forest ecosystems 23 J.M. Rey Benayas, L. Cristóbal, T. Kitzberger, R. Manson, F. López-Barrera, J. Schulz,

R. Vaca, L. Cayuela, R. Rivera, L. Malizia, D. Golicher, C. Echeverría, R. del Castillo, J. Salas

3 Assessing fragmentation and degradation of dryland forest ecosystems 65 C. Echeverría, T. Kitzberger, R. Rivera, R. Manson, R. Vaca, L. Cristóbal,

G. Machuca, D. González, R. Fuentes

4 Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile 103

C. Smith-Ramírez, G. Williams-Linera, R.F. del Castillo, N. Ramírez-Marcial, R. Aguilar, N. Taylor, D. Golicher, P. Becerra, J.L. Celis-Diez, J.J. Armesto

5 Experimental analysis of dryland forest restoration techniques 131 G. Williams-Linera, C. Alvarez-Aquino, A. Suárez, C. Blundo, C. Smith-Ramírez,

C. Echeverria, E. Cruz-Cruz, G. Bolados, J.J. Armesto, K. Heinemann, L. Malizia, P. Becerra, R.F. del Castillo, R. Urrutia

6 Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico 183

R.F. del Castillo, R. Aguilar-Santelises, C. Echeverría, E. Ianni, M. Mattenet, G. Montoya Gómez, L. Nahuelhual, L.R. Malizia, N. Ramírez-Marcial, I. Schiappacasse, C. Smith-Ramírez, A. Suárez, G. Williams-Linera

7 Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration 205

A.C. Premoli, C.P. Souto, S. Trujillo A., R.F. del Castillo, P. Quiroga, T. Kitzberger, Z. Gomez Ocampo, M. Arbetman, L. Malizia, A. Grau, R. Rivera García, A.C. Newton

8 Landscape-scale dynamics and restoration of dryland forest ecosystems 229 A.C. Newton, E. Cantarello, N. Tejedor, T. Kitzberger, C. Echeverría, G. Williams-

Linera, D. Golicher, G. Bolados, L. Malizia, R.H. Manson, F. López-Barrera, N. Ramirez-Marcial, M. Martinez-Icó, G. Henriquez, R. Hill

9 Identifying priority areas for dryland forest restoration 273 D. Geneletti, F. Orsi, E. Ianni, A.C. Newton

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (10)

viii

Principles and Practice of Forest Landscape Restoration

10 Development of policy recommendations and management strategies for restoration of dryland forest landscapes 307

M. González-Espinosa, M.R. Parra-Vázquez, M.H. Huerta-Silva, N. Ramírez-Marcial, J.J. Armesto, A.D. Brown, C. Echeverría, B.G. Ferguson, D. Geneletti, D. Golicher, J. Gowda, S.C. Holz, E. Ianni, T. Kitzberger, A. Lara, F. López-Barrera, L. Malizia, R.H. Manson, J.A. Montero-Solano, G. Montoya-Gómez, F. Orsi, A.C. Premoli, J.M. Rey-Benayas, I. Schiappacasse, C. Smith-Ramírez, G. Williams-Linera, A.C. Newton

11 Synthesis: principles and practice of forest landscape restoration 353 A.C. Newton

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (11)

ix

LIST OF BOXES

Box 1.1 Definition of some key concepts relating to forest restoration, from Lamb and Gilmour (2003) 1

Box 1.2 Elements of FLR, according to Mansourian (2005) 4

Box 1.3 Characteristics of FLR approaches, after Maginnis et al. (2007) 6

Box 1.4 Examples of FLR initiatives in different parts of the world 8

Box 1.5 Area de Conservación Guanacaste (ACG) 8

Box 1.6 Principal partners of the ReForLan project 10

Box 2.1 Methodology used to assess the amount and the drivers of forest change 24

Box 2.2 Vegetation cover change in mountain ranges of central Chile (1955–2008) 33

Box 2.3 Land-use change in the Yungas Biosphere Reserve and its area of influence, Argentina (1975–2008) 35

Box 2.4 Historical distribution of the dryland forest in central Chile during the Spanish conquest in the 16th century 41

Box 2.5 Historical reconstruction of land-use patterns from 1920 to 1960 on communal lands of Paso de Ovejas, Veracruz, Mexico 44

Box 2.6 Tuning up coarse-grained potential vegetation maps for estimation of historical forest loss in tropical Mexico 48

Box 2.7 Different sets of drivers across study regions 51

Box 3.1 Landscape features associated with the passive recovery of Mediterranean sclerophyllous woodlands of central Chile 65

Box 3.2 Landscape connectivity in the highly fragmented drylands of the Central Valley of Chiapas 70

Box 3.3 Estimating forest degradation in dryland landscapes in central Chile using MODIS products 92

Box 3.4 Human-caused forest fires in Mediterranean ecosystems of Chile: modelling landscape spatial patterns on forest fire occurrence 96

Box 4.1 Altitudinal variation in vegetation structure and diversity of tree species in the tropical dry forest region of central Veracruz 109

Box 4.2 Diversity of woody vegetation in the Central Depression of Chiapas, Mexico 110

Box 4.3 Tree species diversity and forest structure in subtropical dry forest of northwestern Argentina 114

Box 4.4 Patterns of diversity of fungi in an altitudinal gradient 116

Box 4.5 Species diversity in northwestern Patagonian dryland forests: implications for restoration 116

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (12)

x

Principles and Practice of Forest Landscape Restoration

Box 4.6 Avian-generated seed rain and germination in the patchy shrubland of central Chile 119

Box 4.7 Effect of fragmentation on plant communities of central Chile 121

Box 4.8 Tree species diversity driven by environmental and anthropogenic factors in tropical dry forest fragments of central Veracruz, Mexico 122

Box 5.1 Holistic ranching and landscape restoration in Chiapas, Mexico 135

Box 5.2 The role of cattle in tropical dry forest regeneration in Chiapas, Mexico 136

Box 5.3 Evaluation of commercial plantations and their management in the tropical dry forest region of Paso de Ovejas, Mexico 139

Box 5.4 Terraces to reforest degraded lands in Oaxaca, Mexico 140

Box 5.5 Replacement of a forest stand of exotic species by native plants in the dryland landscape in central Chile 143

Box 5.6 Early secondary succession as passive restoration in initial stages of ecological restoration of tropical dry forest 145

Box 5.7 Tropical dry forest restoration in Chiapas, Mexico, and basic knowledge for native tree species: phenology, seed germination and seedling growth 148

Box 5.8 Soil seed bank, seed removal, and germination in early secondary succession of a tropical dry forest region in central Veracruz, Mexico 149

Box 5.9 Effects of avian ingestion on seed germination in central Chile 150

Box 5.10 Tree-seedling establishment in fragmented Mediterranean forests of central Chile 153

Box 5.11 Post-fire restoration of native tree species: effects of wood shaving application 154

Box 6.1 Firewood consumption for pottery and projections for woody biomass production from Bursera simaruba 191

Box 6.2 Patterns of firewood use in a tropical dry forest landscape in central Veracruz 191

Box 6.3 Taking local knowledge into consideration when selecting tree species for dry forest restoration in central Veracruz, Mexico 193

Box 6.4 Willingness to reforest with native species in rural communities of central Chile 195

Box 6.5 Traditional knowledge in the drylands of central Mexico: an endangered resource? 195

Box 6.6 Assessing the value and commercial potential of non-timber forest products: the CEPFOR project 200

Box 7.1 Genetic hotspots: the quest for preservation of Chile’s evolutionary history 207

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (13)

xi

List of boxes

Box 7.2 Testing forest connectivity using genetic distance: spider monkeys and dry forest restoration, Nicaragua 209

Box 7.3 Genetic variability in populations of Amelanchier denticulata 214

Box 8.1 Anthropogenic impact on dry forests in Chiapas, Mexico 231

Box 8.2 Effects of fire on sclerophyllous forests in the Biosphere Reserve La Campana-Peñuelas in central Chile 252

Box 8.3 Effects of climate change on subtropical forests of South America 261

Box 8.4 Effects of climate change on dryland ecosystems of central Chile 264

Box 9.1 Weight assessment of forest restoration criteria through experts’ interviews in the Upper Mixtec region, Mexico 281

Box 9.2 Use of biotic, abiotic and cultural variables for tropical dry forest conservation and restoration in central Veracruz, Mexico 288

Box 9.3 Selecting forest restoration priorities at the watershed level in central Chile 291

Box 9.4 Priority areas for implementing the CDM to forest restoration projects in conservation corridors of the Andes 292

Box 9.5 A spatial optimization model for combining ecological and socioeconomic issues 300

Box 10.1 Contribution of livelihoods analysis to the establishment of priorities on restoration of tropical dry forest: a case study in the Central Depression of Chiapas, Mexico 315

Box 10.2 Assessing the value of dry forests in two communities of the Central Depression of Chiapas, Mexico 318

Box 10.3 Hydrological services and environmental decision-making in Latin America 324

Box 10.4 Using multicriteria decision-support tools in Rural Sustainable Development Councils in Chiapas, Mexico 326

Box 10.5 Lessons learned about social management in the Andean forests of Bolivia 330

Box 10.6 Sustainable forest management in Yungas forests: a protocol to develop a forest management plan and implementation in an experimental farm 332

Box 10.7 Public policy and land-use change in central Veracruz (Mexico): an important link in efforts to restore a tropical dry forest landscape 333

Box 10.8 What’s next? Design and implementation of policy instruments for forest restoration and management in Latin America 336

Box 10.9 The connection between university research and education/teaching and a rural community in Mexico: the case study of the Barrancas Environmental Restoration Research Station, Morelos, Mexico 339

Box 10.10 Guidelines for restoration of native species in Mapuche communities 343

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (14)

xii

Principles and Practice of Forest Landscape Restoration

Box 11.1 Mapping and valuation of ecosystem services in dryland forest landscapes 357

Box 11.2 Where should biodiversity be restored in the drylands of Latin America? Findings of the ReForLan expert workshop 365

Box 11.3 An integrated approach to identifying restoration priorities in dryland forest landscapes 368

Box 11.4 Implications of REDD+ for forest landscape restoration 374

Box 11.5 Indicators for monitoring the implementation of Forest Landscape Restoration initiatives 379

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (15)

xiii

FOREWORD

Every year, a forest area the size of Greece1 is lost. More than 80% of the world’s forests have been cleared, fragmented or degraded. The world’s biodiversity and climate, and the liveli-hoods of hundreds of millions of people are under serious threat.

On behalf of the Global Partnership on Forest Landscape Restoration (GPFLR), the International Union for Conservation of Nature (IUCN), the World Resources Institute and South Dakota State University have begun to map the extraordinary potential of deforested and degraded landscapes for restoration to address the challenges facing societies around the world today and in the future.

This latest research tells us that there are more than 1 billion hectares of lost or degraded forest lands worldwide where restoration opportunities may be found. With this comes substantial po-tential to not only sequester large volumes of carbon but also to help lift people out of poverty and reduce the vulnerability of rural people and ecosystems through restoration of forest landscapes.

The important role of landscape restoration has been recognized through recent inter-national decisions on climate change and biodiversity. In October 2010, nearly 200 govern-ments attending the Conference of the Parties to the Convention on Biological Diversity in Nagoya, Japan adopted a target calling for restoration of at least 15% of degraded ecosystems by 2020. Just two months later, in December 2010, Parties to the UN Framework Convention on Climate Change, convened in Cancun, Mexico, adopted the goal to slow, halt and reverse forest cover and carbon loss through REDD+ actions.

In February 2011, the UN Forum on Forests called on Member States and others to build on the work of the GPFLR to further develop and implement forest landscape restoration.

If these international commitments are to be translated into action, further evidence of the effectiveness of forest landscape restoration and guidance on how to implement this approach will be needed.

The book Principles and Practice of Forest Landscape Restoration: Case studies from the drylands of Latin America, edited by A.C. Newton and N. Tejedor, is an excellent com-pendium of case studies and analysis, which will be of interest and use to people who wish to move forest landscape restoration forward, no matter what country they operate in.

Practitioners and policy-makers working on forest landscape restoration are learning all the time, through experience, and from each other. It is important to continue to connect partners and collaborators around the world, from Scotland to Sudan and Moldova to Mexi-co, in a growing community of practice, enabling them to spread best practice, build coop-eration and exchange new ideas and solutions.

In this International Year of Forests, this new publication will be an essential contribution to expanding the body of knowledge on forest landscape restoration and strengthening the network of forest landscape restoration experts around the world.

Julia Marton-Lefèvre, IUCN Director General

1 FAO, 2005

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (16)

xiv

ACKNOWLEDGEMENTS

This research was supported by the European Commission of the European Communities through the ReForLan Project, INCO Programme Framework 6, contract CT2006–032132. We are grateful for the useful comments to Chapter 2 provided by Alejandro Brown and Sil-via Pacheco. We are also grateful to Rosario Landgrave and Ignacio González who helped in the imagery processing and classification of the central Veracruz vegetation, also in Chapter 2. Finally, we extend our thanks to all the researchers who contributed to the research activi-ties described in Chapter 8 and to Gillian Myers.

We should also like to thank the many ReForLan research workers, too numerous to name individually, who helped co-author Chapter 10 and contributed to activities that allowed us to envisage the relevance of interactions among public policies, decision-support tools and management plans. Finally, we should like to thank all participants in the ReForLan work-shops who contributed to the discussions that are summarized in Chapter 11.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (17)

xv

CONTRIBUTORS

Aguilar Santelises, Remedios, Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, Unidad Oaxaca (CIIDIR–Oaxaca), Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlán, 71230 Oaxaca, México. E-mail: [emailprotected]

Alfaro Arguello, Rigoberto, Departamento de Agroecología, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Allendes, Juan L., Departamento de Ciencias Ecológicas, Universidad de Chile, Las Palmeras 3425, Chile. E-mail: jrallend@gmail

Altamirano, Adison, Departamento de Ciencias Forestales, Universidad de La Frontera, P.O. Box 54-D, Temuco, Chile. E-mail: [emailprotected]

Alvarez Aquino, Claudia, Instituto de Investigaciones Forestales, Universidad Veracruzana, Apartado Postal No. 551, CP 91000, Xalapa, Veracruz, Mexico. E-mail: [emailprotected]

Aramayo, Ximena, ECOBONA, Bolivia. E-mail: [emailprotected]

Arbetman, Marina, Laboratorio Ecotono, Universidad Nacional del Comahue, Quintral 1250, 8400, Bariloche, Argentina. E-mail: [emailprotected]

Arjona, Fabio, Conservación Internacional Colombia, Medellín, Colombia.E-mail: [emailprotected]

Armesto, Juan, J., Pontificia Universidad Católica de Chile, Santiago, Chile.E-mail: [emailprotected]

Ascension Hernández, Estela, Instituto de Ecologia, A.C., Carretera Antigua a Coatepec No. 351, Xalapa, Veracruz, 91070, Mexico. E-mail: [emailprotected]

Badinier, Capucine, Fundación ProYungas, Av. Aconquija 2423, (4107) Yerba Buena, Tucumán, Argentina. E-mail: [emailprotected]

Barradas Sánchez, Laura P., Instituto de Investigaciones Forestales, Universidad Veracruzana, Apartado Postal No. 551, CP 91000, Xalapa, Veracruz, Mexico. E-mail: [emailprotected]

Becerra, Pablo, Pontificia Universidad Católica de Chile, Santiago, Chile.E-mail: [emailprotected]

Becerra Vázquez, Ángel G., Departamento de Ecología y Sistemática Terrestres, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29290, Mexico. E-mail: [emailprotected]

Bichier, Peter, Departamento de Agroecología, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (18)

xvi

Principles and Practice of Forest Landscape Restoration

Birch, Jennifer, School of Applied Sciences, Bournemouth University, Talbot Campus, Poole, Dorset BH12 5BB, UK. E-mail: [emailprotected]

Black, Thomas, Centro Andino para la Economía en el Medio Ambiente – CAEMA, Medellín, Colombia. E-mail: [emailprotected]

Blundo, Cecilia, Fundación ProYungas, Av. Aconquija 2423, (4107) Yerba Buena, Tucumán, Argentina e Instituto de Ecología Regional, Universidad Nacional de Tucumán, CC34 (4107) Yerba Buena, Tucumán, Argentina. E-mail: [emailprotected]

Bolados Corral, Gustavo, Departamento Manejo de Bosques y Medio Ambiente, Facultad de Ciencias Forestales Universidad de Concepción, Casilla 160-C, Concepción, Chile. E-mail: [emailprotected]

Brown, Alejandro, Fundación ProYungas, Av. Aconquija 2423, (4107) Yerba Buena, Tucumán, Argentina. E-mail: [emailprotected]

Bustamante, Cesar M., Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]

Buzza, Karina, Fundación ProYungas, Av. Aconquija 2423, (4107) Yerba Buena, Tucumán, Argentina. E-mail: [emailprotected]

Callejas, Jonathan, Instituto de Investigaciones Histórico-Sociales. Universidad Veracruzana, Xalapa, Veracruz, México. E-mail: [emailprotected]

Camus, Pablo, Instituto de Historia, Facultad de Historia, Geografía y Ciencia Política, Pontificia Universidad Católica de Chile, Casilla 306–22, Santiago,Chile. E-mail: [emailprotected]

Cantarello, Elena, School of Applied Sciences, Bournemouth University, Talbot Campus, Poole, Dorset BH12 5BB, UK. E-mail: [emailprotected]

Cayuela, Luis, Centro Andaluz de Medio Ambiente, Universidad de Granada, Spain.E-mail: [emailprotected]

Ceccon, Eliane, Centro Regional de Investigaciones Multidisciplinarias. Universidad Nacional Autónoma de México Av. Universidad s/n, Circuito 2 Colonia Chamilpa, Cuernavaca, Morelos, México, 62210. E-mail:[emailprotected]

Celis Diez, Juan L., Instituto de Ecología y Biodiversidad, Pontificia Universidad Católica de Chile, Santiago, Chile. E-mail: [emailprotected]

Chambers, Carol, Northern Arizona University, South San Francisco Street, Flagstaff, Arizona 86011. E-mail: [emailprotected]

Christophers, Carolina, Departamento de Manejo de Recursos Forestales, Universidad de Chile, Casilla 9206, Santiago, Chile. E-mail: [emailprotected]

Church, Richard L., Department of Geography, 1832 Ellison Hall,University of California, Santa Barbara, Santa Barbara, CA 93106–4060, USA. E-mail: [emailprotected]

Cristóbal, Luciana, Fundación ProYungas, Av. Aconquija 2423, (4107) Yerba Buena, Tucumán, Argentina. E-mail: [emailprotected]

Cruz Cruz, Efraín, Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias C.E. Valles Centrales de Oaxaca, Melchor Ocampo No.7 Sto. Domingo Barrio Bajo, Villa de Etla, Oaxaca, México. E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (19)

xvii

Contributors

del Castillo, Rafael F., Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, Unidad Oaxaca (CIIDIR-Oaxaca), Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlán, 71230 Oaxaca, México. E-mail: [emailprotected]

Domínguez Morales, Lesvia, Departamento de Agroecología, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Echeverría, Cristian, Departamento Manejo de Bosques y Medio Ambiente, Facultad de Ciencias Forestales Universidad de Concepción, Casilla 160-C, Concepción, Chile. E-mail: [emailprotected]

Eliano, Pablo M., Asociación Foresto-industrial de Jujuy, Jujuy, Argentina.E-mail: [emailprotected]

Ferguson, Bruce, Departamento de Agroecología, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Fuentes, Castillo Taryn, Facultad de Ciencias Forestales, Universidad de Chile, Santiago, Chile. E-mail: [emailprotected]

Fuentes, Rodrigo, Departamento Manejo de Bosques y Medio Ambiente, Facultad de Ciencias Forestales Universidad de Concepción, Casilla 160-C, Concepción, Chile. E-mail: [emailprotected]

Garibaldi, Toledo María, Instituto de Ecología, A.C., Carretera Antigua a Coatepec No. 351, Xalapa, Veracruz, 91070, Mexico. E-mail: [emailprotected]

Geneletti, Davide, Dipartimento di Ingegneria Civile e Ambientale, Università degli Studi di Trento, Trento, Italy. E-mail: [emailprotected]

Gobbi, Miriam, Universidad Nacional del Comahue, Quintral 1250, 8400, Bariloche, Argentina. E-mail: [emailprotected]

Golicher, Duncan, School of Applied Sciences, Bournemouth University, Talbot Campus, Poole, Dorset BH12 5BB, UK. E-mail: [emailprotected]

Gómez Alanís, Cristina, Instituto de Ecologia, A.C., Carretera Antigua a Coatepec No. 351, Xalapa, Veracruz, 91070, Mexico. E-mail: [emailprotected]

Gómez Ocampo, Zaneli, Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, Unidad Oaxaca (CIIDIR-Oaxaca), Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlan, 71230 Oaxaca, México. E-mail: [emailprotected]

González, David, Departamento Manejo de Bosques y Medio Ambiente, Facultad de Ciencias Forestales Universidad de Concepción, Casilla 160-C, Concepción, Chile. E-mail: [emailprotected]

González Espinosa, Mario, Departamento de Ecología y Sistemática Terrestres, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Grau, Alfredo, Instituto de Ecología Regional, Universidad Nacional de Tucumán, CC34, (4107) Yerba Buena, Tucuman, Argentina. E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (20)

xviii

Principles and Practice of Forest Landscape Restoration

Gutiérrez, Víctor, Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]

Hagell, Suzanne, Northern Arizona University, South San Francisco Street, Flagstaff, Arizona 86011, USA. Paso Pacífico PO Box 1244 Ventura, CA 93002–1244, USA. E-mail: [emailprotected]

Heinemann, Karin, Universidad Nacional del Comahue, Bariloche, Argentina.E-mail: [emailprotected]

Henríquez, Miguel, Departamento Manejo de Bosques y Medio Ambiente, Facultad de Ciencias Forestales Universidad de Concepción, Casilla 160-C, Concepción, Chile. E-mail: [emailprotected]

Henríquez Tapia, Gabriel, Departamento Manejo de Bosques y Medio Ambiente, Facultad de Ciencias Forestales Universidad de Concepción, Casilla 160-C, Concepción, Chile. E-mail: [emailprotected]

Hernández, Jaime, Laboratorio de Geomática, Antumapu, Universidad de Chile.E-mail: [emailprotected]

Hernández, Rocío C., Departamento de Ecología y Sistemática Terrestres, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29290, México. E-mail: [emailprotected]

Herrera Hernández, Obeimar B., Departamento de Gestión de Territorios, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29290, Mexico. E-mail: [emailprotected]

Hill, Ross, School of Applied Sciences, Bournemouth University, Talbot Campus, Poole, Dorset BH12 5BB, UK. E-mail: [emailprotected]

Holmgren, Milena, Resource Ecology Group, Wageningen University, Wageningen, The Netherlands. E-mail: [emailprotected]

Holz, Silvia C., El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Huerta Silva, Margarita, Departamento de Ecología, Universidad de Alcalá, 28871 Alcalá de Henares, Spain. E-mail: [emailprotected]

Ianni, Elena, Dipartimento di Ingegneria Civile e Ambientale, Università degli Studi di Trento, Trento, Italy. E-mail: [emailprotected]

Jiménez Fernández, Jaime A., Departamento de Ecología y Sistemática Terrestres, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29290, México. E-mail: [emailprotected]

Kitzberger, Thomas, Laboratorio Ecotono, Universidad Nacional del Comahue, Quintral 1250, 8400, Bariloche, Argentina. E-mail: [emailprotected]

Laguado, William G., Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]

Lallement, Mailen, CRUB-U.N.COMAHUE, San Carlos de Bariloche, Argentina.E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (21)

xix

Contributors

Lara, Wilson, Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]

López Barrera, Fabiola, Instituto de Ecología, A.C., Carretera Antigua a Coatepec No. 351, Xalapa, Veracruz, 91070, México. E-mail: [emailprotected]

Lorea, Francisco, Instituto de Ecologia, A.C., Carretera Antigua a Coatepec No. 351, Xalapa, Veracruz, 91070, Mexico. E-mail:[emailprotected]

Machuca, Guillermo, Universidad de Concepción, Casilla 160-C, Concepción, Chile.E-mail: [emailprotected]

Malizia, Lucio R., Fundación ProYungas, Alvear 678, piso 2, oficina 23, (4600) San Salvador de Jujuy, Jujuy, Argentina y Facultad de Ciencias Agrarias, Universidad Nacional de Jujuy, Alberdi 27, (4600) San Salvador de Jujuy, Jujuy, Argentina. E-mail: [emailprotected]

Manson, Robert H., Instituto de Ecologia, A.C., Carretera Antigua a Coatepec No. 351, Xalapa, Veracruz, 91070, Mexico. E-mail: [emailprotected]

Martínez, Icó Miguel, Departamento de Ecología y Sistemática Terrestres. El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Mattenet, Mauricio, Fundación ProYungas, Av. Aconquija 2423, (4107) Yerba Buena, Tucumán, Argentina. E-mail: [emailprotected]

Maturana, Viviana, Pontificia Universidad Católica de Chile, Santiago, Chile.E-mail: [emailprotected]

Miles, Lera, UNEP World Conservation Monitoring Centre, 219 Huntingdon Road,Cambridge, CB3 0DL, UK. E-mail: [emailprotected]

Miranda, Alejandro, Departamento de Ciencias Forestales, Universidad de la Frontera, Temuco, Chile. E-mail: [emailprotected]

Montero, Solano José A., Programa de las Naciones Unidas para el Desarrollo, Oficina Chiapas, México. E-mail: [emailprotected]

Montoya Gómez, Guillermo, Departamento de Gestión de Territorios, El Colegio de la Frontera Sur (ECOSUR), San Cristobal de Las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Nahuelhual, Laura, Departamento de Economía Agraria, Facultad de Ciencias. Universidad Austral de Chile. Casilla 567, Valdivia, Chile. E-mail: [emailprotected]

Newton, Adrian C., School of Applied Sciences, Bournemouth University, Talbot Campus, Poole, Dorset BH12 5BB, UK. E-mail: [emailprotected]

Orsi, Francesco, Dipartimento di Ingegneria Civile e Ambientale, Università degli Studi di Trento, Trento, Italy. E-mail: [emailprotected]

Ortiz Escamilla, Juan, Instituto de Investigaciones Histórico-Sociales. Universidad Veracruzana. Xalapa, Veracruz, México. E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (22)

xx

Principles and Practice of Forest Landscape Restoration

Otterstrom, Sarah, Paso Pacífico PO Box 1244, Ventura, CA 93002–1244, USA. Km 15 Carretera Ticuantepe Centro Comercial MercoCentro, Modulo #5, Ticuantepe, Nicaragua. E-mail: [emailprotected]

Pacheco, Silvia, Fundación ProYungas, Av. Aconquija 2423, (4107) Yerba Buena, Tucumán, Argentina. E-mail: [emailprotected]

Parra Vázquez, Manuel R., Departamento de Gestión de Territorios, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Pascacio, Damián Guadalupe, Departamento de Agroecología, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Pedraza Pérez, Rosa. A., Instituto de Investigaciones Forestales, Universidad Veracruzana, Apartado Postal No. 551, CP 91000, Xalapa, Veracruz, Mexico. E-mail:[emailprotected]

Peredo, Bernardo, School of Geography and the Environment, Oxford University. E-mail: [emailprotected]

Ponce González, Oscar O., Instituto de Investigaciones Forestales, Universidad Veracruzana, Apartado Postal No. 551, CP 91000, Xalapa, Veracruz, México. E-mail: [emailprotected]

Premoli, Andrea, Laboratorio Ecotono, Universidad Nacional del Comahue, Quintral 1250, 8400, Bariloche, Argentina. E-mail: [emailprotected]

Quiroga, Paula, Laboratorio Ecotono, Universidad Nacional del Comahue, INIBIOMA – CONICET, Quintral 1250, Bariloche, Argentina. E-mail: [emailprotected]

Ramírez, Luis Josué, CIIDIR Oaxaca, Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlán, Oaxaca 71230, Mexico. E-mail:[emailprotected]

Ramírez Marcial, Neptalí, Departamento de Ecología y Sistemática Terrestres, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Ramos Vásquez, María Elena, El Colegio de Veracruz, Xalapa, Veracruz, Mexico. E-mail: [emailprotected]

Reid, Sharon, Centre for Advanced Studies in Ecology and Biodiversity (CASEB), Departamento de Ecología, Chile. E-mail: [emailprotected]

Rey Benayas, José M., Departamento de Ecología, Universidad de Alcalá, 28871 Alcalá de Henares, Spain. E-mail: [emailprotected]

Rivera García, Raúl, Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, Unidad Oaxaca (CIIDIR–Oaxaca), Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlán, 71230 Oaxaca, México. E-mail: [emailprotected]

Rivera Hutinel, Antonio, Centro de Estudios en Ecología y Limnología, GEOLIMNOS, Chile. E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (23)

xxi

Contributors

Rueda Pérez, Milka, Departamento de Agroecología, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Salas, Christian, Departamento de Ciencias Forestales, Universidad de La Frontera, P.O. Box 54-D, Temuco, Chile. School of Forestry and Environmental Studies, Yale University, USA. E-mail: [emailprotected]

Salas, Javier, Departamento de Geografía, Universidad de Alcalá, 28871 Alcalá de Henares, Spain. E-mail: [emailprotected]

Santacruz, Alí M., Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]

Schiappacasse, Ignacio, Departamento de Economía de Recursos Naturales y Medio Ambiente, Facultad de Economía y Administración, Universidad de Concepción, Casilla 160-C, Concepción, Chile. E-mail: [emailprotected]

Schulz, Jennifer, Departamento de Ecología, Universidad de Alcalá, 28871 Alcalá de Henares, Spain. E-mail: [emailprotected]

Sierra, Andrés, Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]

Smith Ramírez, Cecilia, Instituto de Ecología y Biodiversidad (IEB), Universidad de Chile, Center for Advanced Studies in Ecology and Biodiversity (CASEB), P. Universidad Católica de Chile and Fundación Senda Darwin (FSD), Alameda 340, casilla 114-D, Santiago, Chile. E-mail: [emailprotected]

Souto, Cintia P., Laboratorio Ecotono, Universidad Nacional del Comahue, Quintral 1250, 8400, Bariloche, Argentina. E-mail: [emailprotected]

Suárez, Alfonso, Instituto de Ciencias Agropecuarias, Universidad Autónoma del Estado de Hidalgo, Avenida Universidad Km 1, Santiago Tulantepec, Hidalgo 43600, México. E-mail: [emailprotected]

Suzart de Albuquerque, Fabio, Centro Andaluz de Medio Ambiente, Universidad de Granada, Spain. E-mail: [emailprotected]

Taylor Aquino, Nathaline, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Tejedor Garavito, Natalia, School of Applied Sciences, Bournemouth University, Talbot Campus, Poole, Dorset BH12 5BB, UK. E-mail: [emailprotected]

Tognetti, Celia, Universidad Nacional del Comahue–CRUB, Quintral 1250, Bariloche, Argentina. E-mail: [emailprotected]

Torres, Rodrigo, Facultad de Educación, Universidad Pedro de Valdivia, Alameda 2222, Santiago, Chile. E-mail: [emailprotected]

Trujillo, Sonia, Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, Unidad Oaxaca (CIIDIR–Oaxaca), Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlan, 71230 Oaxaca, México. E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (24)

xxii

Principles and Practice of Forest Landscape Restoration

Uribe Villavicencio, David, Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, Unidad Oaxaca (CIIDIR–Oaxaca), Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlán, Oaxaca 71230, México. E-mail: [emailprotected]

Urrutia, Rocio, Instituto de Silvicultura, Facultad de Ciencias Forestales, Universidad Austral de Chile, Campus Isla Teja, Valdivia, Chile. E-mail: [emailprotected]

Vaca, Raúl, El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de las Casas, Chiapas 29920, México. E-mail: [emailprotected]

Valenzuela Garza, Ricardo, Ciencias Biológicas, Instituto Politécnico Nacional, Prolongación de Carpio y Plan de Ayala, México DF, Mexico. E-mail: [emailprotected]

Vázquez Mendoza, Sadoth, Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, Unidad Oaxaca (CIIDIR–Oaxaca), Instituto Politécnico Nacional, Hornos 1003, Santa Cruz Xoxocotlán, 71230 Oaxaca, México. E-mail: [emailprotected]

Villablanca, Carola, Laboratorio de Geomática, Antumapu, Universidad de Chile.E-mail: [emailprotected]

Williams-Linera, Guadalupe, Instituto de Ecologia, A.C., Carretera Antigua a Coatepec No. 351, Xalapa, Veracruz, 91070, Mexico. E-mail: [emailprotected]

Yaitul, Valeska, Departamento de Ciencias Forestales, Universidad de La Frontera, P.O. Box 54-D, Temuco, Chile. E-mail: [emailprotected]

Yepes, Adriana, Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]; [emailprotected]

Zapata Arbeláez, Beatriz, Centro de Investigación en Ecosistemas y Cambio Global – Carbono & Bosques, Medellín, Colombia. E-mail: [emailprotected]

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (25)

xxiii

ABBREVIATIONS

ACG Conservation Area of Guanacaste

CBD Convention on Biological Diversity

CODERS Consejos de Desarrollo Rural Sustentable

ECOBONA Programa Regional para la Gestión Social de Ecosistemas Forestales Andinos

EDIEM Estación de Investigaciones Ecológicas Mediterráneas

ENSO El Niño Southern Oscillation

FCPF Forest Carbon Partnership Facility

FLR Forest Landscape Restoration

FRIS Forest Restoration Information Service

GAM Generalised Additive Model

GIS Geographical Information System

IUCN International Union for Conservation of Nature

LGDFS The General Law on Sustainable Forestry Development

LGEEPA The General Law of Ecological Equilibrium and Environmental Protection

LPI Large Patch Index

MBA Man and Biosphere Program

RBYungas Biosphere Reserve of the Yungas

REDD Reducing Emissions from Deforestation and Forest Degradation

SL sustainable livelihoods

SSDF Subtropical Seasonally Dry Forest

TDF Tropical dry forests

TEEB The Economics of Ecosystems and Biodiversity

UNESCO United Nations Educational, Scientific and Cultural Organization

UNFCCC United Nations Framework Convention on Climate Change

WWF Worldwide Fund for Nature (also known as World Wildlife Fund in North America)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (26)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (27)

1

Box 1.1 Definition of some key concepts relating to forest restoration, from Lamb and Gilmour (2003)

Reclamation: recovery of productivity at a degraded site using mostly exotic tree species. Species monocultures are often used. Original biodiversity is recovered but protective function and many of the original ecological services may be re-established.

Rehabilitation: re-establishing the productivity and some, but not necessarily all, of the plant and animal species originally present. For ecological or economic reasons, the new forest may include species not originally present. In time, the original forest’s protective function and ecological services may be re-established.

Ecological restoration: re-establishing the structure, productivity and species diversity of the forest originally present. In time, ecological processes and functions will match those of the original forest. The Society for Ecological Restoration defines it as “the process of assisting the recovery of an ecosystem that has been degraded, damaged or destroyed”.

Human well-being: Ensuring that all people have a role in shaping decisions that affect their ability to meet their needs, safeguard their livelihoods and realize their full potential.

1 INTRODUCTION

A.C. Newton and N. Tejedor

The widespread loss and degradation of native forests is now recognized as a major environmental issue. The problem is so acute that it is justifiably referred to as a ‘deforestation crisis’ (Spilsbury, 2010). Recent reviews indicate that while the rate of deforestation is slowing in some countries, the overall rate of forest loss remains high, estimated at around 130,000 km2/year during the decade 2000–2010 (FAO, 2010; Secretariat of the Convention on Biological Diversity, 2010). Deforestation figures fail to provide a complete picture, however, as many remaining forests are being severely degraded through the use of fire, cutting and herbivory. Accurate data on the extent of forest deg-radation at the global scale are difficult to obtain, but an indication of its impact is provided by a recent estimate of the amount of carbon stored in forest vegetation. Over the period 1990–2005, global forest carbon stocks declined by almost double the decline in forest area (UNEP, 2007).

In response to forest loss and degradation, increasing efforts are being directed towards eco-logical restoration. Forest restoration refers to the process of assisting the recovery of a forest ecosystem that has been degraded, damaged or destroyed (Mansourian, 2005). This may involve the re-establishment of the characteristics of a forest ecosystem, such as composition, structure and function, which were prevalent prior to degradation (Jordan et al., 1987; Hobbs and Norton, 1996; Higgs, 1997). Ecological restoration has been defined in a variety of ways in the past; earlier definitions indicated that the purpose of restoration is the comprehensive re-creation of a speci-fied historical ecosystem, including structural, compositional, and functional aspects. Such defini-tions emphasize the importance of historical fidelity as an endpoint of restoration. In contrast, more recent definitions allow a more flexible set of objectives, noting that cultural values may be important and that a range of ecological variables can be acceptable as endpoints (Higgs, 1997). Some definitions of concepts relating to forest restoration are presented in Box 1.1.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (28)

2

Principles and Practice of Forest Landscape Restoration

Definition of Forest Landscape Restoration

Forest Landscape Restoration (FLR) represents a relatively new approach to forest restora-tion. It was first developed at a workshop led by WWF and IUCN involving a broad range of governmental and non-governmental partners (WWF and IUCN, 2000). This meeting defined FLR as “a planned process that aims to regain ecological integrity and enhance human well-being in deforested or degraded forest landscapes” (Maginnis et al., 2007; Mansourian, 2005).

Since then, the FLR concept has continued to develop. For example:

Dudley and Aldrich (2007) describe FLR as: “A forward looking approach that aims to strengthen the resilience of forest landscapes and to keep a variety of future options open for benefiting from forests and their products, rather than always aiming to restore forests to their original state”;Aldrich et al. (2004) suggest that: “Forest Landscape Restoration focuses on re-establish-ing functions and key ecosystem processes across a whole landscape rather than at just planting or restoring individual sites. As such, it looks at a mosaic of land uses including agricultural lands and forest types ranging from plantations to natural forests; aiming for a balanced mixture of protection, management and restoration providing biodiversity, eco-logical, economic and social benefits and resisting detrimental change”;Bekele-Tesemma and Ababa (2002) state that: “FLR provides a framework for governments, the private sector, NGOs and conservationists to work with one another in making in-formed decisions about sustainable land use”; andSaint-Laurent (2005) summarizes FLR as “getting the right mixture of approaches, at the right scale, to deliver the forest goods and services that people and societies need”.

As noted by Dudley et al. (2005), the FLR approach was developed in response to the widespread failure of more traditional approaches to forest restoration, which have often been site-based, focusing on one or a few forest products, relying heavily on tree planting of a limited number of species, and failing to address the root causes of forest loss and degradation. FLR represents a significant departure from such approaches, as presented in Boxes 1.2 and 1.3. Further information about the FLR approach, including experience gained in its practical application, is provided by Dudley and Aldrich (2007), ITTO and IUCN (2005), IUCN (2008b), Mansourian et al. (2005), Pfund and Stadtmüller (2005) and Rietbergen-McCracken et al. (2007).

The development and application of FLR has become a major activity of the WWF and IUCN forest programmes. The approach was further supported by development of the Glo-bal Partnership on Forest Landscape Restoration (http://www.ideastransformlandscapes.org/), which now involves more than 25 governments and international and non-govern-mental organizations, including WWF and IUCN.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (29)

3

Introduction

Tropical sub-deciduous forest (El Ocote Biosphere Reserve). Photo: R. Vaca

Fires in dry forest areas of Chile. Photo: C. Echeverria

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (30)

4

Principles and Practice of Forest Landscape Restoration

Box 1.2 Elements of FLR, according to Mansourian (2005)It is implemented at a landscape scale rather than a single site - that is to say, planning for forest restoration is done in the context of other elements: social, economic, and biological, in the landscape. This does not necessarily imply planting trees across an entire landscape but rather strategically locating forests and woodlands in areas that are necessary to achieve an agreed set of functions (e.g. habitat for a specific species, soil stabilization, provision of building materials for local communities).It has both a socioeconomic and an ecological dimension. People who have a stake in the state of the landscape are more likely to engage positively in its restoration.It implies addressing the root causes of forest loss and degradation. Restoration can sometimes be achieved simply by removing whatever caused the loss of forest in the first place (such as perverse incentives and grazing animals). This also means that without removing the cause of forest loss and degradation, any restoration effort is likely to be in vain.It opts for a package of solutions. There is no single restoration technique that can be ap-plied to all situations. In each case a number of elements need to be covered, but how to do that depends on the local conditions. The package may include practical techniques, such as agro forestry, enrichment planting, and natural regeneration at a landscape scale, but also embraces policy analysis, training and research.It involves a range of stakeholders in planning and decision-making to achieve a solution that is acceptable and therefore sustainable. The decision of what to aim for in the long term when restoring a landscape should ideally be made through a process that includes repre-sentatives of different interest groups in the landscape in order to reach, if not a consensus, at least a compromise that is acceptable to all.It involves identifying and negotiating tradeoffs. In relation to the above point, when a consensus cannot be reached, different interest groups need to negotiate and agree on what may seem like a less than optimal solution if taken from one perspective, but a solution that when taken from the whole group’s perspective can be acceptable to all.It places the emphasis not only on forest quantity but also on forest quality. Decision-makers often think predominantly about the area of trees to be planted when considering restoration, yet often improving the quality of existing forests can yield bigger benefits for a lower cost.It aims to restore a range of forest goods, services, and processes, rather than forest cover per se. It is not just the trees themselves that are important, but often all of the accompanying elements that go with healthy forests, such as nutrient cycling, soil stabilization, medicinal and food plants, forest dwelling animal species, etc. Including the full range of potential benefits in the planning process makes the choice of restoration technique, locations, and tree species much more focussed. It also allows more flexibility for discussions on tradeoffs with different stakeholders, by providing a diversity of values rather than just one or two.Forest landscape restoration goes beyond establishing forest cover per se. Its aim is to achieve a landscape containing valuable forests, for instance partly to provide timber, partly mixed with subsistence crops to raise yields and protect the soils, as well as partly improving biodiversity habitat and increasing the availability of subsistence goods.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (31)

5

Introduction

Cattle under Acacia pennatula, Chiapas, Mexico. Photo: B. Ferguson

Amatenango del Valle, Central Valley of Chiapas, Mexico. Photo: R. Vaca

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (32)

6

Principles and Practice of Forest Landscape Restoration

Dryland forests

The problem of environmental degradation is most intense in arid and semi-arid areas (Geist and Lambin, 2004), which together cover nearly 30% of the Earth’s surface and comprise half the surface area of the world’s developing countries (UNDP, 2004). Despite their aridity, dryland areas are of global importance for biodiversity, being the centres of origin of many agricultural crops and other economically important species. Rural communities in dryland areas are often highly dependent on forest resources to support their livelihoods, particu-larly fuelwood and fodder. However, in many areas dryland forests have been subjected to un-sustainable land-use practices, including expansion of rangeland for livestock, overharvesting (particularly for fuelwood), conversion to agriculture and rapid growth of urban settlements. These processes have resulted in the widespread deforestation and degradation of dryland forest ecosystems, which has resulted in negative impacts on biodiversity, soil fertility and water availability, and on the livelihoods of local people (UNDP, 2004). Such degradation presents a major challenge to policy initiatives aiming to support sustainable development. Restoration of dryland forest ecosystems is therefore an urgent priority if such policy goals are to be achieved, yet this issue has been neglected by the scientific research community.

In 1988, Janzen (1988b) stated that tropical dry forests (TDF) are the most threatened of all major tropical forest types. This statement was largely based on the observation that less than 2% of the tropical dry forest in the Mesoamerican region was sufficiently intact to be considered worthy of conservation, having declined from an original area of 550,000 km2. At that time, only 0.09% of the forests in the Mesoamerican region were accorded some degree of official protec-tion. While the area under protection has subsequently increased, more recent analyses confirm that remaining dry forests are highly threatened. Miles et al. (2006) estimated that 1,048,749 km2 of tropical dry forest remains throughout the tropics, more than half of which (54.2%) is located in South America. In this region, the two most extensive contiguous areas that remain are located in northeastern Brazil; and southeastern Bolivia, Paraguay and northern Argentina. Other notable

Box 1.3 Characteristics of FLR approaches, after Maginnis et al. (2007)

It takes a landscape-level view. This does not mean that every FLR initiative must be large-scale or expensive; rather that site-level restoration decisions need to accommodate landscape-level objectives and take into account likely landscape-level impacts.

It operates on the ‘double filter’ condition. Restoration efforts need to result in both im-proved ecological integrity and enhanced human well-being at the landscape level.

It is a collaborative process. It involves a wide range of stakeholder groups collectively deciding on the most technically appropriate and socioeconomically acceptable options for restoration.

It does not necessarily aim to return forest landscapes to their original state. Rather it is a forward-looking approach that aims to strengthen the resilience of forest landscapes and keep future options open for optimizing the delivery of forest-related goods and services at the landscape level.

It can be applied not only to primary forests but also to secondary forests, forest lands

and even agricultural land.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (33)

7

Introduction

concentrations of tropical dry forest occur within the Yucatan peninsula of Mexico, northern Venezuela and Colombia. Overall, approximately 97% of the remaining area of tropical dry forest is at risk from one or more threats including climate change, habitat fragmentation, fire, increas-ing human population density and conversion to cropland (Miles et al., 2006).

As noted by Miles et al. (2006), the definition of dry forest is somewhat problematic, as dry forests grade into other vegetation types such as wet forests, savannahs and woodlands. Mooney et al. (1995) suggest that in the simplest terms, tropical dry forest may be defined as forests occurring in tropical regions characterized by pronounced seasonality in rainfall distribution, resulting in several months of drought. The forests that develop under such climatic conditions share a broadly similar structure and physiognomy. However, as noted by Mooney et al. (1995), these shared characteristics are difficult to define with precision. Variation in the duration of the rainy season, topography and soil physical characteristics, in particular soil moisture, account for the large differences observed in canopy height, total biomass, productivity and water availability found among these forests (Mooney et al., 1995; Murphy and Lugo, 1995). Dry forests also vary in structure, from relatively open parkland to dense scrub and closed canopy forest (Killeen et al., 1998).

Tropical dry forests have high plant species diversity and endemism (Gentry, 1995; Janzen, 1987; Jansen, 1988b; Kalacska et al., 2004), and although they typically have lower biomass than wet forest, they may be characterized by greater structural and physiological diversity (Mooney et al., 1995). Their growth rate and the regeneration of plants are relatively low and reproduction is highly seasonal. Most plants are out-crossed and dependent on animal pollination (Quesada et al., 2009) and their seed dispersal is principally by wind and animal vectors (Janzen, 1988a).

While the conservation assessment of Miles et al. (2006) focused explicitly on tropi-cal dry forests, other dry forest types occur outside the tropics in Mediterranean and temperate climates. In Latin America, these include the Mediterranean forests and shrub-lands of central Chile, and the temperate dry forests on the eastern side of the Andes in southern Argentina. Both of these areas were also the focus of research described in this book, providing an opportunity to compare and contrast dry forests in different regions of Latin America.

FLR and dry forests

Examples of forest restoration initiatives can be found in the Forest Restoration Information Service (FRIS, 2009), an online database in which more than 200 past and present projects are described. Specific examples of FLR approaches are described by Dudley and Aldrich (2007), Ecott (2002), IUCN (2008a), Mansourian et al. (2005), Rietbergen-McCracken et al. (2007), and the Global Partnership on Forest Landscape Restoration (2009) (Box 1.4). How-ever, relatively few of these examples have been implemented in dryland forest areas; one ex-ample is provided by New Caledonia (Box 1.4). Many FLR initiatives are still at an early stage of development, and therefore evidence of their effectiveness is typically limited (Aldrich and Sengupta, 2005). The most notable example of a long-term, landscape-scale approach to dry forest restoration is the Area de Conservación Guanacaste in Costa Rica, which has re-established forest over some 70,000 ha of former agricultural land since 1985 (Box 1.5). This project provides a powerful demonstration not only of the feasibility of FLR approaches in dry forest, but also the potential benefits that such approaches might provide.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (34)

8

Principles and Practice of Forest Landscape Restoration

Box 1.5 Area de Conservación Guanacaste (ACG)The single most significant ecological restoration initiative in neotropical dry forest is the Area de Conservation Guanacaste (ACG) in Costa Rica, an initiative led by Daniel H. Janzen of the University of Pennsylvania. Dry forest has been re-established on approximately 70,000 ha of former fields and pastures since 1985. Key activities undertaken to support this restoration process were the cessation of anthropogenic fires and all harvest of plants and animals; the purchase of enough farm and ranch land to create one large unit of land; and letting the forest recover through natural processes. The initiative also appointed a highly knowledgeable and committed staff, who were instrumental in the success of the project. The ACG represents one of the most significant forest restoration projects ever undertaken in the tropics, not least for its role in inspiring restoration efforts in other areas. The ACG has also played a major role in drawing attention to the conservation value and importance of tropical dry forests, and in demonstrating the feasibility of their restoration. However, the ACG remains one of very few initiatives that have aimed to restore dry forests at the landscape scale. As such, it was one of the main sources of inspiration for the research described in this book. Further information about the ACG is provided by Allen (1988), FRIS (2007) and Janzen (2002).

Box 1.4 Examples of FLR initiatives in different parts of the world (Aldrich et al., 2004; Ashmole and Ashmole, 2009; Ecott, 2002; Governments of Brazil and the United Kingdom, 2005; Mansourian et al., 2005)

Brazil: FLR approaches are being applied to the restoration of Brazil’s Atlantic forest, which has been highly fragmented and now covers less than 7% of its original area. Restoration efforts are focusing on creation of forest corridors to connect biological reserves.

China: National forest restoration programmes involve 97% of the counties and cities in China. FLR approaches have been implemented in the Minshan conservation area, which covers 33,000 km2 in Sichuan province.

India: Restoration activities in Gujarat have improved water management, protected forested areas, increased planting of local species, and shifted the economic focus away from timber extraction.

Malaysia: Efforts are being made to restore a forest corridor along the Kinabatangan River, which will connect coastal mangroves with upland forests.

Mali: In the Niger Delta, two forests have been restored, leading to increased fishery production, resolution of conflicts, improved social cohesion and the building of local capacity.

New Caledonia: Less than 2% of the original extent of tropical dry forest remains in this Pacific territory. An action programme has been launched to protect and restore these forests, while contributing to social and economic development.

Tanzania: Since 1985, Sukuma agropastoralists in Shinyanga (northern Tanzania) have restored 250,000 ha of degraded land from semi-desert to forest.

United Kingdom: In the Carrifran Wildwood Project, native woodland has been re-established over an entire sub-catchment of the deforested hills of southern Scotland.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (35)

9

Introduction

Research aims and approach

This book presents the results of an international research project, which was designed explicitly to examine application of the FLR approach to dryland forest ecosystems in Latin America. In order for FLR to be transferred into mainstream practice that is adopted and promoted by governments and the private sector, as well as by local communities, information is needed on how the principles of FLR can be implemented in practice, in a cost-effective manner. It is this information need that the project was designed to address.

The ReForLan project (Restoration of Forest Landscapes for Biodiversity Conservation and Rural Development in the Drylands of Latin America; http://reforlan.bournemouth.ac.uk/) was a collaborative research initiative involving ten partners (Box 1.6), under-taken during the years 2007–2009 (Newton, 2008). The overall objective of the project was to identify and promote approaches for the sustainable management of dryland forest ecosystems, by researching ecosystem restoration techniques using native species of economic value. This was achieved by undertaking a programme of multi-disciplinary research analyzing how restoration of degraded lands can be achieved in a way that mitigates the effects of unsustainable land-use practices, contributes to conservation of biodiversity and supports the development of rural livelihoods, according to the FLR approach. The project aimed to achieve a strong link between technology, management and policy research.

The specific project aims were to:

Identify opportunities for enhanced economic productivity and limits to sustainable pro-duction, with a particular focus on identifying incentives for supporting dryland forest restoration by local communities;Analyze the natural resource use systems at local, regional and international levels through an integrated approach, by developing a comparative programme of research in seven study areas distributed in dryland areas of Mexico, Chile and Argentina;Use the information gathered through participatory techniques to inform the planning and implementation of sustainable management strategies for dryland forest resources;Develop appropriate decision-support tools, including information systems, criteria and indicators of sustainability and rehabilitation, together with case studies of practical resto-ration trials, to support dryland ecosystem management and policies; andDisseminate the results through scientific publications, research reports and internet re-sources; strengthen the research capacity of partner organizations both in Europe and Latin America; and provide training and educational resources.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (36)

10

Principles and Practice of Forest Landscape Restoration

The research was implemented as a series of nine interconnected Work Packages (Fig. 1.1), which form the basis of the structure of this book. This book profiles the results of this research, but also presents some additional results provided by other related projects in the region, which were invited contributions to this volume.

Figure 1.1

Overall, the research aimed to identify how dryland forest ecosystems could be restored in ways that both benefit biodiversity and support the livelihood of local communities, and thereby

Box 1.6 Principal partners of the ReForLan projectBournemouth University (BU), School of Conservation Sciences, Poole, Dorset, UK Coordinator: Professor Adrian Newton

Pontificia Universidad Católica de Chile (PUC), Santiago, Chile Contact: Prof. Juan J. Armesto

Universidad Austral de Chile (UACH), Facultad de Ciencias Forestales, Chile Contact: Prof. Antonio Lara

Universidad Nacional del Comahue (UNCO), Laboratorio Ecotono, Bariloche, Argentina Contact: Dra. Andrea Premoli

Fundación Proyungas (FPY), Tucumán, Argentina Contact: Dr. Lucio R. Malizia

El Colegio de la Frontera Sur (ECOSUR), San Cristobal, Chiapas, Mexico Contact: Dr. Mario González-Espinosa

Instituto Politécnico Nacional (IPN), Centro Interdisciplinario de Investigacion para el Desarrollo Integral Regional, Oaxaca, Mexico Contact: Dr. Rafael F. del Castillo

Instituto de Ecologia (IE), Xalapa, Veracruz, Mexico Contact: Dra. Guadalupe Williams-Linera

Universidad de Alcala (UAH), Departamento de Ecologia, Madrid, Spain Contact: Prof. José M. Rey Benayas

Università degli Studi di Trento (UNITN), Dipartimento di Ingegneria Civile e Ambientale, Università degli Studi di Trento, Trento, Italy Contact: Dr. Davide Geneletti

Additional partners included the UNEP World Conservation Monitoring Centre, Cambridge, UK, (Contact: Dr. Lera Miles), Universidad de Concepción, Concepción, Chile (Contact: Dr. Cristian Echeverría), Universidad Veracruzana, Xalapa, Mexico (Contact: Dra. Claudia Alvarez Aquino).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (37)

11

Introduction

contribute to sustainable development objectives. A conceptual framework was developed at the outset of the project to provide a basis for organizing and integrating research activities (Newton, 2008). This was based on the consideration of forest restoration as a potential response to environmental degradation caused by unsustainable land-use practices. Such response options can usefully be viewed according to the DPSIR (Driving forces – Pressures – State – Impacts – Responses) framework, which was developed by the European Environment Agency to help analyze the process of sustainable development (EEA, 1998). The DPSIR framework is based on the fact that different societal activities (drivers) cause a pressure on the environment, which can cause quantitative and qualitative changes in the state of environmental variables. Such changes can produce a variety of different impacts on natural resources and the services that they provide to human communities. Society has to respond to these changes in appropriate ways in order to achieve sustainable development. According to the DPSIR framework, different indicators of sustainability can be developed relating to driver, pressure, state, impact and response variables; the development of such indicators was one of the outputs of the project.

The research approach was based on the application of the DPSIR framework to restoration of dryland forest resources (Fig. 1.2). The underlying drivers responsible for unsustainable land-use patterns can be grouped into demographic, economic, sociopolitical, technological and cultural factors (Geist et al., 2006; Geist and Lambin, 2001). For example, key factors underpinning current patterns of land-use and land-cover change in dryland regions of Latin America include the cur-rent policy context, the structure and function of national and international market chains for ag-ricultural and forest products, and the process of globalization. Such factors influence patterns of land use, such as cultivation of crops and animal husbandry, which can have a major effect on the extent and condition of forest resources. Key variables describing the state of forest resources in-clude forest area, the size distribution and connectivity among forest patches, and the composition and structure of forest stands (Fig. 1.2). The way that human activities influence these patterns will determine their impact on key ecological processes, such as dispersal, growth, survival, competi-tion, succession and gene flow, which affect biodiversity and the provision of the environmental services on which human communities depend (Fig. 1.2). The severity and extent of environmen-tal degradation, and its impact on biodiversity and the provision of environmental services, will determine both the need and scope for forest restoration as a response option.

Figure 1.2

D riv er

Communitydiversity

Species diversity

Genetic diversityForest area

Density and sizedistribution offorest patches

Shape of forestpatches

Harvesting oftree products

Fire

Grazing

Crop cultivation

Connectivity offorest patches

Standcomposition

R esponsePressu re Im pactS tate

Stand structure

Between-patchheterogeneity

Urban growth

Stand density

Productivity offorest products

Carbonsequestration

Hydrology

Soil fertility

Naturalsuccession

Enrichmentplanting

Direct seeding

Managingsecondary

forests

Agroforestry

Mosaics ofmonocultures

Enhancedunderstoreydevelopment

Demographic

Economic

Socio-political

Technological

Cultural

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (38)

12

Principles and Practice of Forest Landscape Restoration

Study areas

The research focused on seven dryland areas where native forests have been subjected to intense human pressure in recent decades, resulting in severe deforestation and degradation. Each of these areas is characterized by high biodiversity of international conservation im-portance, with many endemic, threatened species. These areas are also characterized by the presence of substantial and increasing rural populations, often including indigenous commu-nities, who rely on native forest resources for provision of a number of forest products. The sustainable management of forest resources in these areas is therefore of key importance to the livelihood of local communities. Although the processes of forest degradation in these areas are similar, the socioeconomic and policy contexts vary, providing scope for compara-tive analysis.

Figure 1.3

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (39)

13

Introduction

The areas where research was primarily conducted are (Table 1.1):

Central Chile, including the Central Valley and Coastal Range, extends over 13,175 km2, covering parts of the Valparaíso, Libertador, Bernando, O’Higgins, and Metropolitan admin-istrative regions. This area lies within the Mediterranean bioclimatic zone, characterized by dry summers and wet winters with strong inter-annual variability. Mean annual tem-perature is 13.2°C and mean annual precipitation is 531 mm, with elevations ranging from sea level to 2260 m a.s.l. A spatially heterogeneous mosaic of shrublands and evergreen sclerophyllous forests is distributed mostly on the slopes and in drainage corridors. The region is acknowledged as one of the world’s 25 biodiversity hotspots (Myers et al., 2000). Logging and land-use changes have led to profound landscape transformations since the mid-sixteenth century and agriculture is mostly concentrated in flat valleys, where major activities include vineyard and fruit cultivation as well as corn and wheat cropping. Forest resources are mainly used for the extraction of fuelwood from native tree and shrub spe-cies, and extensive livestock husbandry in shrubland and forest areas. Timber plantations are located mainly in the flat coastal zone. The study area is home to around 5.2 million inhabitants, which represents ca. 34% of the Chilean population.

Southern Argentina, at the forest-steppe ecotone in northwestern Patagonia, Argentina. A rectangular study area was delimited with corners: NW: 38° 44’, 71° 25’; NE: 38° 46’, 70° 48’; SW: 43° 05’, 71° 35’; SE: 43° 05’, 70° 55’. Total area encompassed is 2.8x106 ha; after classification, total non-grassland area was ca. 0.93x106 ha. Overlaying the glacial topogra-phy, numerous layers of volcanic ash along with sediments from volcanic and pyroclastic rocks constitute soil parent material. Dominant soils are acid and alofanic (Andisols). Ow-ing to the rain shadow effect of the Andes on the westerly winds, mean annual precipita-tion declines from 3000 mm at the continental divide to 400 mm over only 100 km, in the eastern foothills. Approximately 60% of the annual precipitation falls in the winter season (from May to August). Only 5% of the Argentinean population lives in Patagonia and it is one of the least densely populated areas in the world with 1.8–5 inhabitants/km2 (INDEC national statistics, Census 2001). In the first decades of the twentieth century, the regional economy was based almost entirely on the sheep industry, which expanded vigorously. The physiographic characteristics of the area have made it an ideal location for ranching. However, the range of economic activities has increased in the past couple of decades, leading to a shift from sheep ranching towards cattle ranching. Forests, lakes, plentiful rivers, and vast grazing areas provide a unique assortment of livelihood opportunities for the rural population.

Northern Argentina, research was carried out in a forest area of ca. 800,000 ha, which cor-responds to seasonally dry premontane forests (the lower and drier end of Yungas forest) and the transition to the drier Chaco forest. This area is located in the San Martín depart-ment of Salta province. The northwestern extreme of this area is an international border with Bolivia, the southern part borders the Bermejo River and the eastern part reaches the Yungas limit according to Brown and Pacheco (2006). Rainfall ranges from around 1000 mm per year on the western side to 700 mm per year on the eastern side, with a marked seasonal variation determined by rainfall concentration during the summer months (No-vember-March). Maximum temperatures occur in this season, which may exceed 40ºC (Brown and Kapelle, 2001). According to the national statistics (INDEC), San Martin de-partment has 139,204 inhabitants, with a population density of 8.6 inhabitants/km2 (Cen-sus 2002). The study area occupies almost 50% of this department, and includes nearly all

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (40)

14

Principles and Practice of Forest Landscape Restoration

of the most important cities of the area. The forested areas are critical for maintenance of regional biodiversity, environmental services (such as irrigation and land stability), and the sustainable development of the forestry sector. These forests are also important for rural and suburban communities since they provide income through wood, hunting, medicinal plants, honey, and other products. However, in recent decades, deforestation for agricul-tural purposes has been the main activity associated with land-use change in this area, con-sistent with the rest of the premontane forest areas and its transition to dry Chaco forest.

Central Veracruz, Mexico, covers a total area of 160,699 ha and is located along the coastal plain from 0 to 800 m. The boundaries (upper left corner: 19°21’20” and 96°50’52.36”; lower right corner: 19°05’51.43” and 96°06’31.37”) were determined using the following criteria: (1) the area should be centred around the municipality of Paso de Ovejas; (2) it should include the major sub-watersheds (five in all; CONABIO, 1998) comprising the northern and southern lim-its of the study area; and (3) the coast of the Gulf of Mexico and 800 m contours correspond to the altitudinal limits of tropical deciduous dry forest in the region (Rzedowski, 2006). Climate is defined as warm sub-humid (mean minimum and maximum average temperatures are 14 to 36°C, respectively), with rainfall (800–1200 mm) occurring primarily in the summer (95%) and followed by an extended (4–5 months) dry season (García, 1990). Topography is very heterogeneous with steep canyons marking the limits of 119 micro-watersheds. The coastal zone is formed mainly by littoral material deposited by wind, except in the centre, which is dominated by sandstone and conglomerate soils. Soil types are very diverse but dominated by the feozem haplic, litosol, and vertisol pelico varieties. All or part of 12 municipalities, and 490 (<1000 inhabitants) and 15 urban localities are located within the study region. While private lands dominate the region, the 151 communal properties (‘ejidos’) are also important, occupy-ing 41% of the study area.

Central Valley of Chiapas, Mexico, also known as Central Depression of Chiapas, is locat-ed in the central portion of the state. The boundaries of the Depression dry forest ecotone are: 17°39’28”N and 14°32’00”S; 90°22’28”E; and 94°14’13”W extending (ver 13,974 km2

(Olson et al., 2001). This dry valley is more than 200 km long and up to 70 km broad. The strata consist of mostly marine limestones and slates (Breedlove, 1981; Challenger, 1998). Dominant soils are luvisols and lithosols, rendzinas in upper sites (SEMARNAT, 1998). The most influential factor on the climate of this region is the topography. The northern and central highlands of Chiapas protect the valley from the effects of the ‘alisios’ and ‘nortes’ winds from the Gulf of Mexico. The Sierra Madre of Chiapas produces a rain shadow effect, reducing rainfall as a result of humid winds from the Pacific Ocean. The region therefore has a summer precipitation regime of convective character (Challenger, 1998). According to the WorldClim database, the mean annual temperature of the ecoregion ranges from 22 to 25.2ºC, and annual precipitation ranges from 750 to 1500 mm. Most of the valley was probably originally covered with tropical deciduous forest, however extensive culti-vation and grazing have led to large tracts of thorn woodland and savannah (Breedlove, 1981; Challenger, 1998). The dry forest of the Central Valley is completely surrounded by moist forested mountain areas, resulting in their relative isolation from other areas of xeric vegetation. The flora contains a number of central Mexican elements, but appears to lack many species commonly found among the dry flora of Oaxaca. Another interesting facet of the flora of this region is the occurrence of many species only found in the dry regions of the Yucatan Peninsula (Breedlove, 1981). Tuxtla Gutierrez, the capital city of the state, is located in the study area and is the largest urban centre in the south of Mexico, with

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (41)

15

Introduction

a population of 490,455 inhabitants. Rural population density varies over the area, but is generally below 20 inhabitants/km². This is comparatively low for the state of Chiapas.

Oaxaca, Mexico, research focused on the Mixteca Alta, which belongs to the mountains and valleys of the western Oaxaca physiographic region. This region is located north-west of Oaxaca central valley, and is characterized by a complex topography of canyons, plateaus, valleys and mountains, with creeks and rivers that drain to the Pacific Ocean through the Balsas River. An area of 11,637 km2 was surveyed. We limited our study area by including all the municipalities with at least 90% of their land area located within the Mixteca Alta. Elevation ranges between 700–3200 m. Rainfall is concentrated during the summer months, mostly in short and intense showers facilitating soil erosion, with a dry season extending from November to April. Mean annual precipitation is 692 mm and the mean annual temperature is 22°C. Superficial geology is complex: Precambric gneiss, Paleozoic schist, Jurassic sandstone and shale, and Cretaceous limestone, limonite, and sandstone. Vegetation types are variable: mixed pine-oak, pine forests, tropical dry forest and shrublands (CONABIO, 1998). Handicraft manufacturing and agriculture (corn, bean and wheat) are the major economic activities. Croplands are mostly concentrated in the valleys, and to a lesser extent on the mountains. The total population in the study area is 340,000 inhabitants (9.8 % of the state population).

Grazing in the drylands of central Chile. Photo: J. Birch

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (42)

16

Principles and Practice of Forest Landscape Restoration

Cou

ntry

Stud

y are

asLo

catio

n: la

t.

& lo

ng.

Elev

atio

n (m

)R

ainf

all

(mm

)So

il / g

eolo

gyFo

rest

type

Land

use

Land

ow

ners

hip

Fore

st ch

arac

teri

stic

sPr

imar

y th

reat

s

Mex

ico

C

hia

pas

14

°32’

00”

and

90

°22’

28”;

17

°39’

28”

and

94

°14’

13”

350–

1200

75

0–15

00

Luvi

sols

an

d

lith

oso

ls,

ren

dzin

as in

u

pp

er s

ites

Dry

tro

pic

alA

gric

ult

ure

, cat

tle

ran

chin

g, f

ore

st

extr

acti

on

Ran

ches

, ej

idos

com

mu

nal

la

nd

s

Seco

nd

ary

dis

turb

ed f

ore

st.

Rem

nan

t d

egra

ded

fo

rest

o

n s

lop

es.

Agr

icu

ltu

re,

catt

le, fi

re

C

entr

al

Ver

acru

z.

Aca

zon

ica,

Pa

so d

e O

veja

s, P

uen

te

Nac

ion

a,

Emili

ano

Z

apat

a.

19°2

1’20

” an

d

96°5

0’52

.36”

; 19

°17’

06.3

6”

and

96

°26’

27.1

0”

40–8

00

800–

1200

M

ain

ly

cam

biso

ls a

nd

ve

rtis

ols

wit

h

con

side

rabl

e ex

pos

ed r

ock

Dry

tro

pic

al

Live

sto

ck

hu

sban

dry

, rai

n-

fed

agr

icu

ltu

re,

irri

gate

d

agri

cult

ure

, fru

it-

tree

pla

nta

tio

ns

Pri

vate

lan

d,

ejid

os

Frag

men

ted

wit

hin

a

lan

dsc

ape

com

po

sed

o

f se

con

dar

y fo

rest

, ag

ricu

ltu

re a

nd

gra

ssla

nd

s.

On

e-th

ird

of

the

stu

dy a

rea

is c

lass

ified

as

dis

turb

ed

seco

nd

ary

fore

st (

22%

) an

d

fore

st (

9.26

%).

Cat

tle,

ag

ricu

ltu

re

O

axac

a.

Mix

teca

O

axaq

ueñ

a,

Cen

tral

Val

ley,

Sier

ra M

adre

d

e O

axac

a

18°1

3’28

”N

–98°

20’3

6”W

to

18°

12’3

0”N

, –9

6°25

’16”

W

and

16°4

4’10

”N,

–96°

26’4

3”W

to

16°

45’1

5”N

, –9

8°20

’55”

W.

1400

–31

2050

0–10

00

Lith

oso

l, ve

rtis

ol,

acri

sol a

nd

ex

po

sed

ro

ck

Dry

tro

pic

alA

gric

ult

ure

, fo

rest

ry, c

attl

e,

shee

p a

nd

go

at

ran

chin

g

Mo

stly

co

mm

un

alTr

op

ical

dec

idu

ou

s fo

rest

, o

ak f

ore

st, p

ine

fore

st,

gras

slan

d, a

nd

sh

rub

lan

ds.

Sm

all f

ore

st r

emn

ants

o

n k

no

lls; o

ne

of

the

mo

st e

xtr

eme

case

s o

f en

viro

nm

enta

l deg

rad

atio

n

in M

exic

o. H

igh

so

il er

osi

on

.

Agr

icu

ltu

re,

fore

stry

, liv

esto

ck, s

oil

ero

sio

n

Ch

ileC

entr

al V

alle

y 33

° 58

S, 7

58’W

200–

1200

70

0–10

00

Cre

tace

ou

s an

d T

erti

ary

form

atio

ns,

m

ain

ly

gran

od

iori

te,

ton

elit

e an

d

adam

elit

e

Scle

rop

hyl

lou

s,

Dec

idu

ou

s d

ryA

gric

ult

ure

, fu

elw

oo

d

har

vest

ing,

liv

esto

ck

hu

sban

dry

, m

inin

g

Pri

vate

lan

d

pri

nci

pal

lyEn

dem

ic C

hile

an

scle

rop

hyl

lou

s dr

y fo

rest

s; o

ne

of

the

five

M

edit

erra

nea

n c

limat

e fo

rest

s in

th

e w

orl

d. H

igh

en

dem

ism

, hig

hly

deg

rade

d

soil.

Fuel

wo

od

h

arve

st,

lives

tock

, u

rban

ex

pan

sio

n,

agri

cult

ure

Tabl

e 1.1

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (43)

17

IntroductionC

ount

rySt

udy a

reas

Loca

tion:

lat.

&

long

.El

evat

ion

(m)

Rai

nfal

l (m

m)

Soil

/ geo

logy

Fore

st ty

peLa

nd u

seLa

nd

owne

rshi

pFo

rest

char

acte

rist

ics

Prim

ary

thre

ats

C

oas

tal R

ange

33°

S, 7

1° 3

0’W

0–19

00

200–

450

Alfi

sol a

nd

in

cep

tiso

ls

der

ived

fro

m

gran

itic

ro

ck

and

mar

ine

terr

aces

re

spec

tive

ly

Scle

rop

hyl

lou

s an

d x

ero

ph

ytic

p

lan

t sp

ecie

s su

ch a

s C

act

us

spp

.

Agr

icu

ltu

re,

pas

ture

, min

ing,

co

mm

erci

al

fore

st p

lan

tati

on

s

Pri

vate

an

d

pu

blic

lan

dH

ard

-leav

ed s

pec

ies;

p

seu

do

-sav

ann

ah. H

igh

en

dem

ism

, hig

hly

deg

rad

ed

soil.

Agr

icu

ltu

re,

exo

tic

tree

p

lan

tati

on

s,

lan

d-c

ove

r ch

ange

, m

inin

g, s

oil

ero

sio

n, u

rban

ex

pan

sio

n

Arg

enti

na

No

rth

wes

t Sa

lta

& J

uju

y P

rovi

nce

s

22°−

24°S

, 63

.5°−

65°W

350–

750

500–

900

Sier

ras

sub

and

inas

Sub

tro

pic

al

seas

on

al

dry

An

dea

n

pre

mo

nta

ne

(Yu

nga

s)

Dry

Ch

aco

Agr

icu

ltu

re,

logg

ing,

ex

ten

sive

ca

ttle

ran

chin

g

Pri

vate

la

nd

; has

th

e h

igh

est

con

cen

trat

ion

o

f et

hn

ic

gro

up

s (9

) in

A

rgen

tin

a.

Hig

hly

sea

son

al. T

he

Ch

aco

is

th

e la

rges

t re

mai

nin

g ar

ea

of

neo

tro

pic

al d

ry fo

rest

an

d h

arb

ou

rs t

he

hig

hes

t p

rop

ort

ion

of

deci

duo

us

spec

ies

in S

ou

th A

mer

ica.

Fo

rest

loss

to

agr

icu

ltu

ral

lan

d h

as c

reat

ed a

n

agri

cult

ura

l gap

5–2

5 km

w

ide

bet

wee

n d

ry C

hac

o

and

An

dean

pre

mo

nta

ne

fore

st.

Agr

icu

ltu

re,

lan

d-c

ove

r ch

ange

, se

lect

ive

logg

ing,

fire

So

uth

wes

t, n

ort

her

n

Pata

gon

ia

39°3

0’ a

nd

43

°35’

S35

0–13

00

600–

1000

Poo

rly

dev

elo

ped

A

nd

iso

ls f

rom

vo

lcan

ic a

sh

Fore

st-s

tep

pe

eco

ton

e m

on

osp

ecifi

c co

nif

er f

ore

st

Shee

p, c

attl

e an

d g

oat

ra

nch

ing,

ex

oti

c p

lan

tati

on

s

Pri

vate

lan

dIs

ola

ted

co

nif

er f

ore

st

pat

ches

. Rem

nan

t fo

rest

on

sl

op

es a

nd

in r

avin

es. H

igh

fi

re f

req

uen

cy.

Cat

tle,

sh

eep

, go

ats,

ex

oti

c tr

ee

pla

nta

tio

ns,

fi

re

Tabl

e 1.1

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (44)

18

Principles and Practice of Forest Landscape Restoration

References

Aldrich, M., Belokurov, A., Bowling, J., Dudley, N., Elliott, C., Higgins-Zogib, L., Hurd, J., Lacerda, L., Mansourian, S., McShane, T. 2004. Integrating forest protection, management and restoration at a landscape scale. WWF International, Gland, Switzerland.

Aldrich, M., Sengupta, S. 2005. Forest landscape restoration – seeing the bigger picture. In: IUCN, WWF (eds.), Arborvitæ: The IUCN/WWF Forest Conservation Newsletter. Issue 28. IUCN/WWF, Gland, Switzerland.

Allen, W.H. 1988. Biocultural restoration of a tropical forest. BioScience 38: 156–161.

Ashmole, M., Ashmole, P. 2009. The Carrifran Wildwood Project. Ecological restoration from the grassroots. Borders Forest Trust, Jedburgh.

Bekele-Tesemma, A., Ababa, A. 2002. Forest landscape restoration: initiatives in Ethiopia. Available from: http://assets.panda.org/downloads/ethiopiaflr.pdf (Accessed 17 November 2009). IUCN/WWF, Gland, Switzerland.

Breedlove, D.E. 1981. Flora of Chiapas. Part I. Introduction to the flora of Chiapas. Academy of Science. San Francisco, USA.

Brown, A.D., Kappelle, M. 2001. Introducción a los bosques nublados del neotrópico: una síntesis regional. In: Kappelle, M., Brown, A.D. (eds.), Bosques nublados del neotrópico. Editorial INBIO, San José, Costa Rica: pp. 25–40.

Brown, A.D., Pacheco, S. 2006. Propuesta de actualización del mapa ecorregional de la Argentina. In: Brown, A.D., Martínez Ortiz, U., Acerbi, M., Corchera, J. (eds.), La situación ambiental Argentina 2005. Fundación Vida Silvestre Argentina, Buenos Aires, Argentina: pp. 28–31.

Challenger, A. 1998. Utilización y conservación de los ecosistemas terrestres de México: pasado, presente y futuro. CONABIO, Instituto de Biología de la UNAM y Agrupación Sierra Madre, S.C., México, D.F, México.

CONABIO (Comisión Nacional para el Conocimiento y Uso de la Biodiversidad). 1998. Subcuencas hidrológicas. Secretaría de Recursos Hidráulicos, Jefatura de Irrigación y control de Ríos, Dirección de Hidrología, México D.F., México.

Dudley, N., Aldrich, M. (eds.). 2007. Five years of implementing forest landscape restoration – lessons to date. WWF International, Gland, Switzerland.

Dudley, N., Mansourian, S., Vallauri, D. 2005. Forest Landscape Restoration in context. In: Mansourian, S., Vallauri, D., Dudley, N. (eds.), Forest restoration in landscapes: beyond planting trees. Springer, New York, USA: pp. 3–7.

Ecott, T. 2002. Forest Landscape Restoration. Working examples from five ecoregions. WWF International, Gland, Switzerland.

EEA. 1998. Europe’s Environment – The 2nd Assessment. European Environment Agency. Office for Publications of the European Communities.

FAO. 2010. Global Forest Resources Assessment 2010. Main report. FAO, Rome.

FRIS. 2007. Forest restoration information service, Area de Conservación Guanacaste, Costa Rica. Available from: http://www.unep-wcmc.org/forest/restoration/docs/CostaRica.pdf (Accessed on 26 April 2007).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (45)

19

Introduction

FRIS. 2009. Forest Restoration Information Service (FRIS): Database. Available from: http://www.unep-wcmc.org/forest/restoration/fris/database.aspx. (Accessed on 21 October 2009).

García, E. 1990. Climas, 1: 4000 000. IV.4.10 (A). Atlas Nacional de México. Vol. II. Instituto de Geografía, UNAM, México D.F., México.

Geist, H., Lambin, E., Palm, C., Tomich, T. 2006. Agricultural transitions at dryland and tropical forest margins: actors, scales and trade-offs. In: Brouwer, F., McCar, B.A. (eds.), Agriculture and climate beyond 2015. Springer, Dordrecht, The Netherlands: pp. 53–73.

Geist, H.J., Lambin, E.F. 2001. What drives tropical deforestation. LUCC Report series 4: 116.

Geist, H.J., Lambin, E.F. 2004. Dynamic causal patterns of desertification. BioScience 54: 817–829.

Gentry, A.H. 1995. Diversity and floristic composition of neotropical dry forests. In: Bullock, S.H., Mooney, H.A. Medina, E. (eds.), Seasonally dry tropical forests. Cambridge University Press, Cambridge, UK: pp. 146–194.

Global partnership on forest landscape restoration. 2009. Introduction to the demonstration portfolio. Available from: http://www.unep-wcmc.org/forest/restoration/globalpartnership/portfolio.htm (Accessed on 21 October 2009).

Governments of Brazil and the United Kingdom. 2005. Forest landscape restoration implementation: Report to the 5th session of the UN forum on forests. Global Partnership on Forest Landscape Restoration, Petrópolis, Brazil.

Higgs, E.S. 1997. What is good ecological restoration? Conservation Biology 11(2): 338–348.

Hobbs, R.J., Norton D.A. 1996. Towards a conceptual framework for restoration ecology. Restoration Ecology 4(2): 93–110.

ITTO and IUCN. 2005. Restoring forest landscapes: an introduction to the art and science of forest landscape restoration. ITTO and IUCN, Yokohama, Japan.

IUCN. 2008a. About Forest landscape restoration. Available from: http://www.iucn.org/about/work/programmes/forest/fp_our_work/fp_our_work_thematic/fp_our_work_flr/fp_forest_landscape_about/ (Accessed 15 October 2009). IUCN, Gland, Switzerland.

IUCN. 2008b. Learning from Landscapes. Available from: http://cmsdata.iucn.org/downloads/a_avspecial_learning_from_landscapes.pdf (Accessed 15 October 2009). IUCN, Gland, Switzerland.

Janzen, D.H. 1987. How to grow a tropical national park: basic philosophy for Guanacaste National Park, northwestern Costa Rica. Cellular and Molecular Life Sciences 43: 1037–1038.

Janzen, D.H. 1988a. Management of habitat fragments in a tropical dry forest: growth. Annals of the Missouri Botanical Garden 75: 105–116.

Janzen, D.H. 1988b. Tropical dry forests the most endangered major tropical ecosystem, In: Wilson, E.O. (ed.), Biodiversity. National Academy Press, Washington: pp. 130–137.

Janzen, D.H. 2002. Tropical dry forest: Area de Conservación Guanacaste, northwestern Costa Rica. In: Perrow, M.R., Davy, A.J. (eds.), Handbook of ecological restoration. Vol. 2, Restoration in practice. Cambridge University Press.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (46)

20

Principles and Practice of Forest Landscape Restoration

Jordan, W., Gilpin, M. and Aber, J. (eds.). 1987. Restoration ecology: a synthetic approach to ecological research. Cambridge University Press, Cambridge.

Kalacska, M., Sanchez-Azofeifa, G.A., Calvo-Alvarado, J.C., Quesada, M., Rivard, B., Janzen, D.H. 2004. Species composition, similarity and diversity in three successional stages of a seasonally dry tropical forest. Forest Ecology and Management 200: 227–247.

Killeen, T.J., Jardim, A., Mamani, F., Rojas, N. 1998. Diversity, composition and structure of a tropical semideciduous forest in the Chiquitania region of Santa Cruz, Bolivia. Journal of Tropical Ecology 14: 803–827.

Lamb, D., Gilmour, D. 2003. Rehabilitation and restoration of degraded forests. IUCN and WWF International, Gland, Switzerland and Cambridge, UK.

Maginnis, S., Rietbergen-McCracken, J., Jackson, W. 2007. Introduction. In: Rietbergen-McCracken, J., Maginnis, S., Sarre, A. (eds.), The forest landscape restoration handbook. Earthscan, London, UK: pp. 1–4.

Mansourian, S. 2005. Overview of forest restoration strategies and terms. In: Mansourian, S., Vallauri, D., Dudley, N. (eds.), Forest restoration in landscapes: beyond planting trees. Springer, New York, USA: pp. 8–13.

Mansourian, S., Vallauri, D., Dudley, N. 2005. Forest restoration in landscapes: beyond planting trees. Springer, New York, USA.

Miles, L., Newton, A.C., DeFries, R.S., Ravilious, C., May, I., Blyth, S., Kapos, V., Gordon, J.E. 2006. A global overview of the conservation status of tropical dry forests. Journal of Biogeography 33: 491–505.

Mooney, H.A., Bullock, S.H., Medina, E. 1995. Introduction. In: Bullock, S.H., Mooney, H.A., Medina, E. (eds.), Seasonally dry tropical forests. Cambridge University Press, Cambridge, UK: pp. 1–8.

Murphy, P.G., Lugo, A.E. 1995. Dry forests of Central America and the Caribbean. In: Bullock, S.H., Mooney, H.A., Medina, E. (eds.), Seasonally dry tropical forests. Cambridge University Press, Cambridge: pp. 9–34.

Myers, N., Mittermeier, R., Mittermeier, C., Da Fonseca, G., Kent, J. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853–858.

Newton, A.C. 2008. Restoration of dryland forests in Latin America: the ReForLan Project. Ecological Restoration 26 (1): 10–13.

Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V.N., Underwood, E.C., D’Amico, J.A., Itoua, I., Strand, H.E., Morrison, J.C., Loucks, C.J., Allnutt, T.F., Ricketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W.W., Hedao, P. and Kassem, K.R. 2001. Terrestrial ecoregions of the world: a new map of life on Earth. BioScience 51: 933–938.

Pfund, J.L., Stadtmüller, T. 2005. Forest landscape restoration (FLR), InfoResources Focus, 2/05 InfoResourses, Zollikofen, Switzerland.

Quesada, M., Sanchez-Azofeifa, G.A., Alvarez-Anorve, M., Stoner, K.E., Avila-Cabadilla, L., Calvo-Alvarado, J., Castillo, A., Espirito-Santo, M.M., fa*gundes, M., Fernandes, G.W., Gamon, J., Lopezaraiza-Mikel, M., Lawrence, D., Morellato, L.P.C., Powers, J.S., Neves, F.D., Rosas-Guerrero, V., Sayago, R., Sanchez-Montoya, G. 2009. Succession and management of tropical dry forests in the Americas: review and new perspectives. Forest Ecology and Management 258: 1014–1024.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (47)

21

Introduction

Rietbergen-McCracken, J., Maginnis, S., Sarre, A. 2007. The forest landscape restoration handbook. Earthscan, London, UK.

Rzedowski, J. 2006. Vegetación de México. 1ra. Edición digital. Nacional Comisión para el Conocimiento y Uso de la Biodiversidad, México D.F., México.

Saint-Laurent, C. 2005. Optimizing synergies on forest landscape restoration between the Rio conventions and the UN forum on forests to deliver good value for implementers. Review of European Community & International Environmental Law 14: 39–49.

Secretariat of the Convention on Biological Diversity. 2010. Global Biodiversity Outlook 3. Secretariat of the Convention on Biological Diversity, Montreal. 94pp.

SEMARNAT, Secretaría de Medio Ambiente, Recursos Naturales y Pesca. 1998. Mapa de suelos dominantes de la República Mexicana. Scale 1:4 000 000. México D.F., México.

Spilsbury, R. 2010. Deforestation crisis (can the Earth survive?). The Rosen Publishing Group, New York.

UNDP. 2004. Sharing innovative experiences. Examples of the successful conservation and sustainable use of dryland biodiversity. Available from: http://tcdc.undp.org/sie/experiences/vol9/content9new.asp. (Accessed 19 October 2009). UNDP, New York.

UNEP. 2007. GEO4 Global Environment Outlook: environment for development. UNEP, Nairobi, Kenya.

WWF, IUCN. 2000. Forests reborn: A workshop on forest restoration. In: WWF/IUCN International Workshop on Forest Restoration: July 3–5, Segovia, Spain. IUCN, Segovia, Spain.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (48)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (49)

23

2 ASSESSING THE CURRENT EXTENT AND

RECENT LOSS OF DRYLAND FOREST

ECOSYSTEMS

J.M. Rey Benayas, L. Cristóbal, T. Kitzberger, R. Manson, F. López-Barrera, J. Schulz, R. Vaca, L. Cayuela, R. Rivera, L. Malizia, D. Golicher, C. Echeverría, R. del Castillo, J. Salas

Introduction

Land-cover change is regarded as the most important global change affecting ecological systems (Vitousek, 1994). Natural landscapes – i.e. those largely unaffected or hardly affected by human activities – are being rapidly transformed into urban and farmland landscapes throughout the world (Foley et al., 2005; Feranec et al., 2010; López and Sierra, 2010). As the characteristics of land-cover have important impacts on climate, biogeochemistry, hydrology, species diversity, and the well-being of human societies, land-cover change has been identi-fied as a high priority for research and to inform the development of strategies for sustain-able management (Turner et al., 1993; Ojima et al., 1994; Millennium Ecosystem Assessment, 2005a). In recent years, special attention has been given to land-use changes and degradation in drylands. Dryland forests are highly prone to degradation and desertification on account of their limited primary productivity and slow recovery following human disturbance (Mil-lennium Ecosystem Assessment, 2005b), yet these ecosystems play a crucial role in providing services such as climate and water regulation (Maass et al., 2005; Lemons, 2006).

To develop approaches for the conservation and restoration of dryland forests at the re-gional level it is crucial to know their current extent and to understand the main recent and historical changes that have affected them (Schulz et al., 2010). To accomplish this goal, it is necessary to assess what processes may be driving such changes, to reveal the threats to forest ecosystems, and to develop alternative strategies to diminish these threats (Angelsen and Kaimowitz, 1999; Geist and Lambin, 2002; Lambin et al., 2003; Antrop, 2005; Olander et al., 2008). Changes in patterns of forest distribution at a wide variety of spatial scales, from global to local scales, are among the land-cover changes most frequently investigated. At the global scale, forest extent is declining mostly as a result of expansion of farmland (Foley, 2005; FAO, 2010). Additional activities associated with widespread deforestation and forest degradation include industrial tree plantations (often composed of non-native species), log-ging for firewood, and cattle grazing (Lara and Veblen, 1993; Kahn and McDonald, 1997; Geist and Lambin, 2002). However, at local and regional scales, land abandonment resulting mainly from rural-urban migration can contribute towards passively restoring considerable amounts of the original forest extent (Aide and Grau, 2004; Pascarella et al., 2000; Rudel et al., 2005; Grau and Aide, 2008; Parés-Ramos et al., 2008). These landscape processes have rarely been mapped and quantified, and change trajectories among land-cover types have not been sys-tematically examined for particular types of forest.

We addressed these issues in selected areas in Chile, Argentina and Mexico, using standardized research protocols (Box 2.1). The primary advantage of assessing these

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (50)

24

Principles and Practice of Forest Landscape Restoration

different areas is that they include a range of ecological, socioeconomic, and cultural characteristics. We used remote sensing data to measure and monitor land-cover change because of their ability to capture an instantaneous synoptic view of a large part of the Earth’s surface and to provide repeated measurements of the same area on a regular ba-sis (Donoghue, 2002). Land-cover change detection and monitoring is especially useful in those regions where there is a lack of available cartographic information with suffi-cient spatial resolution to examine land-cover change. To investigate the possible causes of change in forest cover, a Geographic Information System (GIS) database incorporating satellite imagery and biophysical and socioeconomic variables was developed for each study area. This information was statistically analyzed to infer likely drivers of forest-cov-er change and to test a series of specific hypotheses relating to the factors responsible for deforestation. For example, we expected that the rate of forest loss would be (i) posi-tively associated with population density and accessibility (i.e. proximity to roads and rivers), and (ii) highest on sites most suitable for agriculture such as those with gentle slopes (Sader and Joyce, 1988; Pfaff, 1999; Lambin et al., 2003).

As part of the contribution to the international ReForLan project (Newton, 2008), we re-port here the first multi-regional assessment of land-use/land-cover changes, with special at-tention to forest loss, spanning a ca. 30-year period (1970s–2000s) in dryland Latin American regions. The results presented should be useful for planning the restoration and conservation of dryland forest in the study areas and elsewhere in the region. Furthermore, they provide an example of the research that needs to be conducted in other regions of the world.

Box 2.1 Methodology used to assess the amount and the drivers of forest change

A total of six areas were considered in this study in central Chile (the Central Valley and the Coastal Range extending to the Pacific Ocean), southern Argentina (northwestern Patagonia), northern Argentina (Salta province), as well as central Veracruz (central Mexico by the Gulf Coast), Central Depression of Chiapas, and Oaxaca (southern Mexico by the Pacific Coast).

Analysis of land-cover/land-use changeRemote sensing data

We acquired a time-series of satellite imagery to analyze land-cover/land-use change in each study area (Table 1). All images were pre-processed, including geometric, atmospheric and topographic corrections. The images were geometrically corrected using standard procedures based on ground and roadway map control points. For the removal of atmospheric effects and variations in solar irradiance, an atmospheric correction was carried out to transform the original radiance images to reflectance images using an algorithm based on the Chavez reflectivity model (Chávez, 1996) in most study areas. Topographic corrections were also performed to reduce shadows on hilly areas when necessary. We employed a variety of methods for this task, such as the C-correction proposed by Teillet et al., (1982) using a digital elevation model (DEM) interpolated from contour lines of 25 m for TM and ETM+ images in central Chile and Oaxaca, and the NASA SRTM model with a resolution of three arc seconds per pixel (ca. 80 m) in central Veracruz. To compare images of different pixel size, the original MSS raster grids were re-sampled to the resolution of the TM raster grids (30 m) in most study areas.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (51)

25

Assessing the current extent and recent loss of dryland forest ecosystems

Box 2.1 (cont.)

Table 1 Time series of satellite imagery (Landsat and SPOT) used to analyze land-cover/land-use change in the selected study areas in Latin America.

Study area Year/Sensor Year/Sensor Year/Sensor Year/Sensor

Central Chile 1975 MSS 1985 TM 1999 TM 2008 ETM+

Southern Argentina 1973 MSS 1985 TM 1997 TM 2003 ETM+

Northern Argentina 1976/77 MSS 1987 TM 1993 TM 2006 TM

Central Veracruz (Mx) 1973 MSS 1990 TM 2000 ETM+ 2007/08 SPOT

Chiapas (Mx) —* 1990 TM 2000 ETM+ 2005 ETM+

Oaxaca (Mx) 1979 MSS 1989 TM 2000 ETM+ 2005 SPOT

*Classification results for 1975 were excluded due to inconsistencies related to Landsat MSS resolution.

Land-cover/land-use classification

All study areas attempted to follow a common protocol for classification of land-cover/land-use types that distinguished eight major classes: (1) forest, (2) shrubland, (3) pasture, (4) bare ground, (5) agriculture, (6) timber plantations, (7) urban areas, and (8) water. However, these pre-defined classes were modified according to local conditions in each study area. The land-cover maps were derived using a supervised classification procedure in all study areas except in northern Argentina, where an ISODATA unsupervised classification technique was used.

To classify the images, field points were taken with a GPS in order to train the spectral signature of the selected land-cover classes (198 in central Chile, 311 in southern Argentina, 1071 in central Veracruz, and 50 in Oaxaca). This information was complemented with high-resolution imagery obtained from Google Earth and aerial ortho-photos to account for areas with restricted accessibility, control points from vegetation and land-use maps. In addition, informal interviews were conducted with land owners and land managers during field surveys to obtain information on previous and current land-cover and land use.

For central Chile and Oaxaca, a region-growing approach was used with the ‘seed’ function, and signature separability of the initial classes for all images was evaluated using the Bhattacharyya distance. Based on this distance, classes were iteratively merged until reasonably high signature separability (Bhattacharyya distance >1.9) was achieved. For northern Argentina, the classifications performed used nine iterations and a convergence of 0.95. For central Veracruz, polygons were created by means of a regional growth algorithm that combined pixels that were judged to be similar based on the values of the spectral bands available for each image, as well as data from scale, texture, and shape indices. For central Chiapas, classifications were performed using an iterative procedure developed for classifying multi-temporal Landsat imagery in complex tropical landscapes (Harper et al., 2007). A unified classification for the three analyzed years was produced. Maximum likelihood classification was the principal method used as it has proven to be a robust and consistent classifier for multi-date classifications (Yuan et al., 2005); however, the Sequential Maximum A-Posteriori classifier (SMAP in GRASS 6.4) produced the best results in Chiapas. Post-classification processing was applied to combine initial classes and to better discriminate between confounding classes in each study area. More details on the classification procedures and used software can be found in Rey Benayas et al., (2010a), Schulz et al., (2010), Cristóbal et al., (in prep.), Gowda et al., (in prep.), Manson et al., (in prep.), Rivera et al., (in prep.), and Vaca et al., (in prep.).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (52)

26

Principles and Practice of Forest Landscape Restoration

Accuracy assessment

Accuracy assessment of the classification results was carried out using independent ground control points sampled in the field (280 in central Chile, 432–520 control points in southern Argentina, 157 in northern Argentina, and 300 in Oaxaca) as well as control points derived from Google Earth and ortho-photos (>400 in central Veracruz, and >2500 in Chiapas). Based on these points, confusion matrices and associated Kappa indexes of agreement for each class were generated (Rosenfield and Fitzpatrick-Lins, 1986). As independent data sources such as previous classifications and ortho-photos were largely absent for central Veracruz, an Index of Ecological Congruence (IEC) was developed following the same logic as the Kappa index, where congruent and incongruent events were considered as feasible or non-feasible land-cover transitions between two time periods given current ecological understanding of land-use patterns and rates of vegetation succession in the region. Overall classification accuracies for the different classified study areas are reported in Table 2.

Table 2 Table 1

Study area Series 1 Series 2 Series 3 Series 4

Central Chile 68.5 77.3 78.9 89.8

Southern Argentina 69.3 80.7 85.0 86.0

Northern Argentina 68.2 82.5 85.3 83.4

Central Veracruz (Mx)* 93.4 94.0 74.8 82.0

Chiapas (Mx) — 74.1**

Oaxaca (Mx) 92 90 93.3 93.5

*Based on an Index of Ecological Congruency (IEC) and not Kappa for Series 1–3 (1973–2000). **Since a unified clas-sification for the three analyzed years was produced, we performed a unique validation analysis.

Change identification

The spatial distribution of land-cover/land-use changes was investigated using the previously classified remotely sensed images in order to obtain a matrix of change directions among land-cover classes (Lu et al., 2004) in each study area. Changes were analyzed by cross-tabulation as proposed by Pontius et al., (2004) to quantify net changes, gains, losses and persistence as well as inter-categorical change trajectories. In Chiapas, the classification was combined with results from the map series IV (2007–9) of Use of the land and Vegetation elaborated by INEGI (scale 1:250,000).

Drivers of forest cover changeGIS analysis and explanatory variables

To analyze the drivers of forest cover change we created binary maps to represent forest versus other cover types. The points where change (deforestation) occurred or not were determined by overlapping the binary land-cover maps from the two dates that spanned the time period studied in each study area. In southern Argentina, since afforestation and not deforestation was detected, we analyzed drivers of afforestation with exotic conifers rather than drivers of deforestation.

Box 2.1 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (53)

27

Assessing the current extent and recent loss of dryland forest ecosystems

We then followed a common protocol (Echeverría et al., 2006) that included randomly selecting a grid of sampling points separated by a minimum distance of 1000 m for extracting values of both the response (change vs. no change) and explanatory variables (biophysical and socioeconomic) in order to reduce problems with spatial auto-correlation. A number of biophysical and socioeconomic variables that may influence forest cover change were selected for analysis including: (1) elevation (m), (2) slope (°), (3) insolation or radiation input, (4) mean annual precipitation (mm), (5) soil quality, (6) distance from rivers and lakes (m), (7) distance from forest edges (only for forest loss), (8) distance from forest (only for forest gain in southern Argentina), (9) distance from human settlements (cities, towns and villages) with different numbers of inhabitants (m), (10) human population density (#/km2), (11) distance from different types of roads (e.g. primary or secondary, paved or unpaved (m), (12) distance from the agricultural frontier (m), (13) distance from irrigation infrastructure (m), and (14) distance from cattle pastures (m). For all explanatory variables, values were extracted using the random sampling points previously selected with forest to non-forest changes for each of the four analysis periods. However, not all of these variables were used in each study area as availability of information depended on local conditions.

Model building

Logistic regressions were performed to explore the effect of the described explanatory variables on deforestation in all study areas except in Chiapas, where a different method was used (see below). To fit the models, we started by using the full set of explanatory variables described above. Before starting the model selection process, a Pearson’s or Spearman’s correlation test was performed to identify the correlated explanatory variables, and the single representative of variables that were highly correlated (typically r >0.7) was selected for further analyses to avoid multicollinearity. A spatial correlogram based on Moran’s index of autocorrelation was used to explore the autocorrelation of data at different geographic distances, and was found to be low in all cases. In southern Argentina, the effect of each predictor variable on the occurrence of afforestation was evaluated by univariate logistic models. In Chiapas, a series of generalized additive models (GAMs) of the binomial family were fitted using the R package ‘MGCV’ (Wood, 2004). These models allow non-linear responses to be modelled.

Model selection

Multivariate, spatially explicit models were developed for each of the three periods of time and the whole study period in most study areas. In central Chile and northern Argentina, we performed a backward stepwise model selection based on the Akaike (1974) Information Criterion (AIC) to determine the set of explanatory variables constituting the best fitting model for each period. The measure used by generalized linear models, including logistic regression, to assess the aptness of fit is called the deviance. Deviance reduction (D2) is estimated as D2 = (Null deviance – Residual deviance)/Null deviance. For southern Argentina, selection of the final multivariate logistic model to explain afforestation was performed using forward, backward and best subset procedures with GIS Landchange Modeler (Idrisi, 2006); then a spatial model of potential afforestation was generated and evaluated by means of the relative operating characteristic (ROC) curve. For central Veracruz, those variables with p≤0.10 in univariate logistic regressions were used to construct a final multiple logistic regression model that best explained the loss of undisturbed forest in the study region. In Chiapas, to complement the GAM analysis and provide a direct interpretation of the strength of the drivers, recursive partitioning was also used as implemented in the R package ‘rpart’ (Therneau and Atkinson, 2009). Recursive partitioning models allow interactive effects between variables to be investigated.

Box 2.1 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (54)

28

Principles and Practice of Forest Landscape Restoration

Major changes in land-cover as a result of land-use intensification

Important changes in land-cover/land-use types occurred during the time period analyzed within each study area, as shown by the mapping and quantification of remote sensing im-agery (Fig. 2.1) and the analysis of change trajectories among land-cover types (Fig. 2.2). The observed changes between land-cover classes differed significantly among the various study areas and between the particular periods of time under study. We detected the following major trends: (1) forest degradation to shrubland (in central Chile and southern Argentina) or secondary forest (in northern Argentina and Veracruz), (2) conversion of shrubland and secondary forest to agricultural land, grassland or bare ground (in central Chile, Veracruz and Chiapas), (3) direct conversion of forest to agricultural land or grassland (in northern Argen-tina and Chiapas), (4) conversion during different periods between forest and shrubland and between shrubland and grassland in southern Argentina, and (5) conversion during different periods between forest and secondary forest and conversion of grassland to agriculture in Veracruz. These changes indicate overall land-use intensification across all study areas during the interval studied. Loss of natural vegetation cover, namely forest and shrubland, is the most consistently observed change. Oaxaca presented by far the most complex patterns of change trajectories, with alternate exchanges between natural vegetation cover and bare ground or agriculture over the consecutive periods of time. Transformation to urban areas was less important as compared with other changes in all areas excepting in Chiapas. Similarly, expan-sion of tree plantations was relatively high in central Chile and southern Argentina, but was of lesser importance in other study areas than the changes indicated above.

Figure 2.1

et al.

et al.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (55)

29

Assessing the current extent and recent loss of dryland forest ecosystems

Land use mosaic in La Sepultura Biosphere Reserve, Chiapas, Mexico. Photo: N. Tejedor

Vineyards in Casablanca valley, Chile. Photo: C. Echeverria

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (56)

30

Principles and Practice of Forest Landscape Restoration

Forest

Eroded forest

Agriculture

Bare ground

1976-77 1987 1993 2006

0.4%

4.2%0.4%

0.4%

0.5%

5.2%

0.4%

9.5%

Forest

Eroded forest

Agriculture

Bare ground

Forest

Eroded forest

Agriculture

Bare ground

Forest

Eroded forest

Agriculture

Bare ground

0.8%

0.4%

Figure 2.2c

Figure 2.2a

Figure 2.2b

Forest

Shrubland

Agriculture

Pasture

Bare ground

Plantation

1973 1985 1999 2008

4%

1.4%

0.9%

1.1%

1.7%

1%

0.5%

3.2%

Urban

Forest

Shrubland

Agriculture

Pasture

Bare ground

Plantation

Urban

Forest

Shrubland

Agriculture

Pasture

Bare ground

Plantation

Urban

Forest

Shrubland

Agriculture

Pasture

Bare ground

Plantation

Urban

1.2%

1.4%

1.5%

0.8%

1.2%

1.5%

1.3%

3.3%

1.1%

Forest

Shrubland

Grassland

Plantation

Urban

Burn

Forest

Shrubland

Grassland

Plantation

Urban

Burn

Forest

Shrubland

Grassland

Plantation

Urban

Burn

Forest

Shrubland

Grassland

Plantation

Urban

Burn

1973 1985 1997 2003

9.8%

3.5%

1.6%5.2%0.7%

1.7%

1.8%

3.7%

1.2%

1.8%

0.4%

1.4%

2.6%

1.5%

1.5%

0.6%

1.2%

Figure 2.2

et al

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (57)

31

Assessing the current extent and recent loss of dryland forest ecosystems

Forest

Agriculture

Fires

Water

Urban

Others

1990 2000 2005

0.36%

0.5%

Pasture

Forest

Agriculture

Fires

Water

Urban

Others

Pasture

Forest

Agriculture

Fires

Water

Urban

Others

Pasture Figure 2.2d

Figure 2.2e

Bare ground

Water

Forest

GrasslandShrubland

Agriculture

Urban

1979 1989 1999 2005

Bare ground

Water

Forest

GrasslandShrubland

Agriculture

Urban

Bare ground

Water

Forest

GrasslandShrubland

Agriculture

Urban

Bare ground

Water

Forest

GrasslandShrubland

Agriculture

Urban

Forest

Secondaryforest

Grasslands

Rain-fedagriculture

1973 1990 2000

3.8% Forest

Secondaryforest

Grasslands

Rain-fedagriculture

Forest

Secondaryforest

Grasslands

Rain-fedagriculture

5.5%

2.4%

7.6%6%

1.6%

10.2%

4.5% 3.2%

6.7%

5.8%

8.8%

5.6%

Figure 2.2f

Figure 2.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (58)

32

Principles and Practice of Forest Landscape Restoration

Changes in land-cover/land-use types as inferred from mapping and quantification of re-mote sensing imagery and analysis of change trajectories among land-cover types indicate ongoing land-use intensification in all study areas. Major changes observed include a reduc-tion in natural vegetation cover, namely forest and shrubland, and a strong increase in hu-man-induced land-cover types such as cropland, pasture, bare ground, urban areas, and tree plantations with exotic species. Beyond this consistent pattern, the observed changes were of different intensity – when they existed at all – in the various study areas. The transforma-tion of forest to the most highly influenced land-cover types, such as farmland, occurred with or without other intermediate land-cover types.

Land-cover change in the Mediterranean climate area of central Chile revealed a general trend towards a continuous reduction in forest and shrubland that, in turn, has led to an in-crease in provisioning ecosystem services such as food and timber production likely at the expense of biodiversity and hydrological services (Schulz et al., 2010; Box 2.2). This process has involved a progressive modification from forest to shrubland vegetation, the predomi-nant vegetation cover in this semi-arid landscape, and a relatively high loss of shrubland as a consequence of conversion to agriculture and timber plantations and, to a lesser extent, urbanization. This can be explained by an increase in local demand owing to population growth and an open market policy initiated after Chile’s economic crisis at the beginning of the 1970s (Silva, 2004). The strong increase in agriculture has been stimulated by a combina-tion of market liberalization, incentives for new export-oriented crops, introduction of new irrigation technologies, and improvements in road infrastructure (Valdés and Foster, 2005).

In contrast to central Chile, we observed an abrupt conversion of forest to agricul-tural land in northern Argentina, which has one of the highest agricultural land conver-sion rates in the country. In the 1970s, almost 95% of the area was covered by forests, and 90% by some type of tree cover. The deforested areas consisted mainly of pastures and rotational cropland plots in rural communities and small areas of premontane for-ests in areas where irrigation was possible. Since the 1980s, when soybean cultivation became highly profitable and began (Adelman, 1994; Brown and Malizia, 2004; Gasparri and Grau, 2009), huge areas of forest were converted at a rate of >20,000 ha/year (Box 2.3). Similarly in central Veracruz, Mexico, following a period of slight increase in forest area in the previous decade, forest area declined markedly in the 1990s with the estab-lishment of powerful federal incentives, notably the Procampo programme, to promote the conversion of forest cover to cattle pasture and croplands (Klepeis and Vance, 2003; Montero-Solano, 2009). The programme originally applied to areas planted with beans, cotton, maize, rice, sorghum, soybeans, or wheat; however farmers supported by the programme could use their land for other crops, raising livestock or silviculture. Pro-campo is currently being phased out under the provisions of NAFTA (i.e. after 15 years) and forest cover appears to have begun to increase once again.

In Chile, there has been a pronounced expansion of timber plantations, mostly as a result of government subsidies for tree-planting that were introduced in 1974 and which stimulated the planting of Pinus radiata and Eucalyptus globulus (Aronson et al., 1998). The expansion of timber plantations did not result in major conversions of native forest, as it did in southern Chile (Echeverria et al., 2006) and the region in southern Argentina studied here. In this latter region, deforestation can primarily be explained by the occurrence of natural and anthropogenic fires which, in many cases, did not regenerate back into forests and remained as stable grasslands or shrublands (Mermoz et al., 2005). However, dryland forest areas in this region are undergoing

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (59)

33

Assessing the current extent and recent loss of dryland forest ecosystems

Box 2.2 Vegetation cover change in mountain ranges of central Chile (1955–2008)

C. Villablanca, J. Hernandez, C. Smith-Ramírez, J. Schulz

The sclerophyllous forest is the most characteristic plant formation in the Mediterranean climate area of central Chile. Sclerophyllous trees and shrubs have hard leaves adapted to long dry summers and occasional morning frost. This plant formation is distributed mainly on south-facing slopes and creeks, where soil humidity is concentrated. The most common tree species at the dry end of the environmental gradient of Chilean sclerophyllous vegetation are quillay (Quillaja saponaria), maitén (Maytenus boaria) and litre (Lithrea caustica). In more humid places such as creeks and sheltered slopes, trees such as peumo (Cryptocarya alba), patagua (Crinodendron patagua), belloto del norte (Beilschmiedia miersii), pitra (Myrceugenia exsucca) and canelo (Drimys winteri) are found (Donoso, 1995).

Chilean sclerophyllous forests have high species richness and endemism (Villagrán, 1995). The high endemism and increasing degree of threat owing to extensive land-cover change have resulted in the inclusion of these forests as one of the 25 global biodiversity hotspots (Myers et al., 2000). However, economic activities and increasing human population in central Chile have resulted in high impacts on natural resources, leading to important losses of biodiversity. Sclerophyllous forests have been eradicated and degraded over large areas of central Chile, especially in the central valley (see Box 2.4), but also in the Andean and coastal ranges. Deforestation and land use change are the result of expanding farmland and silvicultural plantations, as well as urban expansion (Armesto et al., 2010; Schulz et al., 2010). At the same time remnant stands are used for firewood and soil extraction, cattle grazing and trampling, and most importantly, subject to a high frequency of anthropogenic fires. Hervibory of shrub and tree seedlings by rabbits, horses and goats, is a major factor preventing sclerophyllous forest recovery on abandoned lands (Fuentes et al., 1983).

The objective of this study was to assess the changes in land-cover that have occurred in the second half of the twentieth century in the Mediterranean-climate region of central Chile. We chose the period 1955–1975, as an antecedent to the patterns described by Schulz et al. (2010) for the period 1975–2008 in the same region. The study area was the landscape surrounding the Casablanca hills and valleys (2740.2 km2, 32º50’00”–33º27’00” S and 71º36’00”–70º58’00” W (from sea level to 2190 m a.s.l.) and Cantillana hills (4304.18 km2, 33°38’00”–34º15’00” S and 71°27’00”–70º38’00” W, from 145 to 2280 m a.s.l.), (Fig. 1). Plant species richness and composition were documented in sclerophyllous vegetation patches larger than 60 ha. We excluded extensive areas that had been deforested prior to 1955. Such areas are found south and east in the valley between Casablanca and the Cantillana hills. We estimated vegetation cover in 1955 using aerial photographs, and distinguished eight different cover classes: forests, shrubland, farmland, urban areas, bare soil, anthropogenic prairies, and exotic silvicultural plantations. Subsequently, we compared our results for 1955 with the vegetation cover map for 1975, obtained from Landsat images. We identified and quantified the major changes by cover type during this 20-year period.

To estimate land-cover changes between 1955 and 1975 we applied the Land Change Modeller in the IDRISI software programme; we used Landsat images to estimate land-cover changes between 1975 and 2008. The changes reported here are notable because shrublands in central Chile have been considered extremely persistent. The dynamics of land-cover change over these two decades of the twentieth century is shown in Table 1, Figs. 2 and 3. Between 1955 and 2008, 71,290.75 ha of forest were lost. From the total forest cover that existed in 1955, 64.2% (14,3160.96 ha) was below 700 m. Plant species richness is higher in Cantillana, and in Casablanca below 600 m (Universidad de Chile, 2009). However, the areas of greatest species richness were replaced by other land-cover types by the first half of the twentieth century or earlier.

The percentage of change in sclerophyllous forest cover from 1955 to 1975 was comparatively low compared to the period 1975 to 2008, where natural vegetation cover was lost much more rapidly (Schulz et al., 2010). In the study areas, between 1955 and 1975, forest cover decreased by 8.5%, however between 1975 and 2008, forest cover decreased by 45%. Between 1955

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (60)

34

Principles and Practice of Forest Landscape Restoration

and 1975 Matorral (shrubland cover) increased by 5.5%, however, it decreased by 22.7% in the period between 1975 and 2008. This was probably the result of the expansion in agricultural land and the development of an agricultural export industry in Chile coinciding with the opening up of international markets. The development of new methods of irrigation during the 1980s was responsible for the expansion of cultivated land up to altitudes of 500 m on coastal hills; this affected the plant diversity of many remnant patches of sclerophyllous forest and shrubland.

Table 1

Ha Forest ShrublandAgricul-

tural landUrban soil Bare soil Water Prairie Plantations

CB 1955 59,905.51 148,625.93 9,634.07 5574.85 3120.80 1690.97 39,440.08 5031.51

CB 2008 25,549.62 104,176.17 28,550.01 17,471.1 34,329.55 801.46 39,043.49 20,238.92

C 1955 83,255.45 147,079.46 119,254.14 3062.85 14,056.02 1425.32 44,688.32 401.11

C 2008 46,320.59 136,973.54 133,873.81 6971.57 42,607.54 4115.66 32,685.24 578.71

Box 2.2 (cont.)

Figure 1

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (61)

35

Assessing the current extent and recent loss of dryland forest ecosystems

Figure 2

Figure 3

Box 2.2 (cont.)

Box 2.3 Land-use change in the Yungas Biosphere Reserve and its area of influence, Argentina (1975–2008)

S. Pacheco, L. Cristóbal, K. Buzza

The Biosphere Reserve of the Yungas (RBYungas) was created in northwestern Argentina in 2002 under the UNESCO Man and the Biosphere (MAB) programme. It is one of the largest reserves in the country, extending over approximately 1,350,000 ha, and includes two provincial territories (Jujuy and Salta). The reserve was created as part of efforts to implement actions conducive to conserving and managing the Yungas region (subtropical mountain forests) in a sustainable way.

The RBYungas includes mainly subtropical mountain forests, particularly in the northern latitudinal sector, which are functionally connected to the central sector of the Yungas and to the Chaco forests in the surrounding areas. As a result of agricultural activity, these environments are being fragmented, with remaining fragments sometimes connected by areas that act as corridors. The objective of this study was to describe the land-use change process in the northern and central sectors of the Yungas and its transition to Chaco, with special emphasis on the RBYungas for the period 1975–2008.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (62)

36

Principles and Practice of Forest Landscape Restoration

Box 2.3 (cont.)

The productive activities carried out in the 1970s include sugar cane, tobacco and agricultural cultivation. During 1985 these three activities remained important but the production of soybean combined with beans represented almost 30% of production. During 2005, these two new categories accounted for almost 50% of production. During the four years analyzed, 90% of the transformation occurred on slopes below a 5% gradient. The transformation of the forest in steeper areas was mainly for forest plantations, pastures and agricultural plots of small size.

Figure 1

The identification of transformed areas was carried out through visual interpretation of Landsat satellite images. We developed a time series of an area of more than 5 million ha located in Jujuy and Salta. The years included were 1975, 1985, 2005 and 2008. For the four years analyzed, we calculated the total deforested area and the annual transformation rate. For each year of the time series, we identified types of crops, determined their slope and established the ecoregion that was transformed in each case. For the RBYungas we determined the deforested area for each year analyzed and the remaining surface which may be subject to transformation.

From 1975 to 2008, the transformed surface in the study area increased almost 13% (Table 1). During the 1970s, the transformed areas were mainly concentrated in the flat areas of the premontane forest, on the west side of the study area. During the 1980s, the expansion of the agricultural frontier began in the east of the region, mainly in the province of Salta, occupying the Chaco environments (Fig. 1).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (63)

37

Assessing the current extent and recent loss of dryland forest ecosystems

Table 1

Year Transformed (ha) % of RBYungas Annual transformation rate (ha)

1975 359,143 6,7

1985 611,295 11,5 25,215 (1975–1986)

2005 908,845 17 14,878 (1998–2004)

2008 1,034,486 19,4 41,880 (2004–2008)

Transformation within RBYungas

In 1975, 4% of the RBYungas was transformed into agricultural land, mainly concentrated in flat areas; this reached 6.5% in 2008. This represents an annual transformation rate which increased from 930 ha for the period 1975–1985 up to 3274 ha between 2005 and 2008 (Table 2). Almost 90% of the RBYungas surface is represented by steep areas with a slope of a gradient of more than 5%. Only 150,000 ha correspond to flat areas suitable for agricultural use. If we analyze the transformed area taking into account only available flat areas within the RBYungas, we observe that during 2008 more than 50% of this surface was transformed (Table 2).

Table 2

Year Transformed (ha) % of RBYungas % flat areas in RBYungas

Annual transforamtion rate (ha)

1975 53,597 4.0 34.2

1985 62,725 4.7 39.4 930 (1975–1986)

2005 77,164 5.7 49.2 722 (1998–2004)

2008 86,987 6.5 55.3 3.274 (2004–2008)

These results indicate that the transformation of premontane forest has been increasing in recent years. This process began in areas with slopes of less than 5% gradient with the planting of sugar cane which requires irrigation; it resulted in the near disappearance of forests in flat areas at a regional level. Since the 1980s, there has been an expansion of the agricultural frontier towards drier areas of the Chaco owing to changes in technology, an increase in precipitation in the last century and the incorporation of the soybean crop which does not require irrigation. More than 50% of the flat surface of the RBYungas has already been transformed into agricultural land. RBYungas remains connected to its surrounding natural areas through those slopes that remain forested. Currently, local governments have implemented land-use plans, which delimit production areas and areas that must be protected in order to maintain the functional connection between the different environments of the region.

Box 2.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (64)

38

Principles and Practice of Forest Landscape Restoration

Firewood next to a potter’s kiln, Chiapas, Mexico. Photo: B. Ferguson

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (65)

39

Assessing the current extent and recent loss of dryland forest ecosystems

a small but steady conversion to plantations with exotic conifers. In southern Argentina, rapid expansion of urban areas coincided with the abolition of the urban limits by the Ministry of Housing and Urbanism in 1979 and the liberalization of the urban land market (Kusnetzoff, 1987).

Changes in forest extent

Forest loss was consistently detected in all study areas, ranging from an annual rate of –1.7% in central Chile to a negligible –0.12% in the Central Valley of Chiapas, with an aver-age rate of –0.78% across all study areas (Table 2.1). Some study areas have experienced a relatively high proportion of forest loss (15% to 9% of the study area in central Chile, 90% to 70% in northern Argentina, and 11.3% to 6.56% in central Veracruz), whereas others have experienced relatively little deforestation (13% to 11% in southern Argentina, 59% to 57% in Oaxaca, Mexico) or hardly any changes in their forest extent (32% of the study area in Chiapas at both reference dates, as 68% of forest land had already been lost in this region by the beginning of the study period).

Table 2.1

Study area % Forest cover in early or mid-1970s

% Forest cover in 2000s

% Annual rate of change

Central Chile 43.3 33.9 –1.7

Southern Argentina 17.3 16.4 –0.17

Northern Argentina 94.0 73.0 –1.3

Veracruz (Mx) 11.3 6.56 –1.22**

Oaxaca (Mx) 59.3 56.6 –0.18

Chiapas (Mx) 32.1* 31.5 –0.12

However, forest loss varied considerably between the analyzed time periods in most study areas. Thus, northern Argentina showed high forest loss (>5%) between all analyzed periods, central Chile lost forest extent in all periods except from 1985 to 1999 when there was hardly any change, and central Veracruz gained total forest cover at a yearly rate of 0.6% be-tween 1973 and 1990 but then underwent considerable forest loss at a yearly rate of 4.33% between 1990 and 2000.

The analysis of change trajectories revealed that the patterns of conversion of forest cover also differed between study areas. The conversion of forest to agriculture was mediated by an intermediate shrubland state in central Chile (Fig. 2.2) and southern Argentina, whereas in Veracruz, a similar trend was observed for the conversion from primary forest to agriculture mediated by an intermediate secondary forest state. How-ever, in northern Argentina and Oaxaca, and to a lesser extent in Chiapas, most forest loss occurred as a result of direct transformation into agricultural land. In all study areas, loss of forest extent was partially mitigated by forest re-growth as a result of farmland abandonment (Fig. 2.3).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (66)

40

Principles and Practice of Forest Landscape Restoration

Figure 2.3

et al

While forest loss was a consistently observed outcome of land-use intensification, annual rates of deforestation were highly variable between study areas and among different time pe-riods within the same study area. The highest deforestation rate was observed in Chile; nev-ertheless, this rate is relatively low compared to rates in temperate forests in south-central Chile (Echeverria et al., 2006). This is probably attributable to the fact that central Chile has been densely populated since early European colonization and major conversions of forest cover had taken place long before the 1970s (Conacher and Sala, 1998; Box 2.4). Historical data from central Veracruz suggest that dryland forest cover has been suppressed (<33% cover) at least since the start of the previous century when land reforms were implemented following the Mexican revolution (Box 2.5). Similarly, most deforestation in Chiapas occurred long ago (Challenger, 1998) and has been relatively minor in recent times (Box 2.6). Our re-sults suggest that this region has undergone a comparable level of historical deforestation to other dry forest ecoregions in Mexico. Challenger and Dirzo (2009) reported that dry forest loss at the country level between 1976 and 1993 amounted to 177,000 ha per annum (an-nual deforestation rate of 1.6%) and reduced to 44,416 ha per annum (annual deforestation rate of 0.5%) over the next decade (1993–2002 ). Deforestation rates found in our study area in Oaxaca were slightly lower as compared with other areas with similar vegetation in Oaxaca and in Mexico (Aguilar et al., 2000; Velázquez et al., 2003; Díaz-Gallegos et al., 2008). Cayuela et al. (2006), however, reported an annual deforestation rate of 4.8% for the high-lands of Chiapas in the period 1990–2000.

Primary forests are lost because they are directly converted to cropland or grazing land or because they are degraded by permanent grazing pressure, firewood collection and charcoal

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (67)

41

Assessing the current extent and recent loss of dryland forest ecosystems

Box 2.4 Historical distribution of the dryland forest in central Chile during the Spanish conquest in the 16th century

C. Echeverria, R. Fuentes, R. Torres, P. Camus

Historic landscape reconstruction through documentary sources is useful to (i) assess the dynamics of historical land-use changes, (ii) define the potential of dryland forest ecosystems, and (iii) place resource management practices of indigenous people and other local communities in a historical context (Prieto et al., 2003). The coastal zone in central Chile that extends from the cities of Santiago to Valparaiso presents evident signs of environmental degradation that has occurred since the 16th century. Traditionally, Chilean historians have presented an idyllic image of a territory covered by beautiful forests (Barros Arana, 1884), while others, more recently, have provided information indicating an absence of forests over large areas in historic times (Camus, 2002). Primary documentary sources offer a valuable description of the past as they narrate the first impressions of European colonists when they laid eyes on the landscape for the first time.

The reconstruction and mapping of dryland forest in the 16th century, at the onset of European colonization, is key to a clear understanding of the current and historical patterns of land-use change and the development of restoration strategies in central Chile. The objective of this exercise was to reconstruct a picture of the vegetation between Santiago and Valparaíso through a spatially-explicit approach that integrates information from documentary sources and environmental factors into a GIS.

Visual descriptions of vegetation were obtained from field notes taken from travellers through the region. These descriptions were collected from reviews of primary and secondary historical sources written in the 16th century. Most of the visual descriptions used in the present study were gathered in the Casablanca valley, Colina zone and the Aconcagua River route. The main primary documentary sources used in this study were drawn from registries maintained by Santiago’s Cabildo (town council), land measurements and registries, chronicles, letters and travel diaries. However, further descriptions from secondary documentary sources were also used when they clearly referred to descriptions of vegetation that existed before the arrival of the Spanish. Visual descriptions of naturalists such as Charles Darwin and Edward Poeppig were used to model species’ distributions on the Aconcagua-Valparaiso route and its surroundings. We discarded those descriptions that did not have a precise enough spatial reference. As a result, we did not use Claudio Gay’s botanical descriptions.

The descriptions and roads used by travellers were spatially plotted on a 30 m-resolution elevation map. Then, environmental requirements such as aspect and elevation of the description points in relation to vegetation composition were obtained through a review of the literature (Donoso, 1982; Donoso, 1995). This information was used to generate digital maps of habitat suitability for each category of vegetation (group of species or individual species in some cases) through a combination of ranges of elevation with categories of aspect (Table 1). North and south aspects are the major environmental factors that determine patterns of species distribution in dryland landscapes in this part of Chile (Donoso, 1995). South-facing hills are characterized by lower levels of solar radiation/insolation, and therefore higher humidity in the soil and air than north-facing sites.

In some cases the documentary sources provided specific locations for some currently threatened species such as Jubaea chilensis (Grau, 2004) and Porlieria chilensis. This enabled the historical distribution at the species level to be mapped. Similarly, several historical descriptions mentioned the abundant presence of espinales, a disturbed pseudo-savannah dominated by Acacia caven, across the study area. This enabled different sub-categories of espinales to be mapped. Additionally, documentary sources containing detailed descriptions of some of the current main cities in the study area such as Santiago, Valparaiso and Quillota were mapped.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (68)

42

Principles and Practice of Forest Landscape Restoration

Table 1

Vegetation type Altitude (m a.s.l) and aspect

Sample of original texts described by travelers during the 16th century

Historical references

Grassland > 200 North and East

“Las únicas plantas que cabría mencionar eran arbustos pequeños e inaperentes, que ostentaban sólo de vez en cuando todavía en temporada desfavorable una miserable flor y que se presentaban semisecos y polvorientos”

Poeppig, E. 1835

Open espinal 200 – 350 North and East

> 300 South and West

“ … Todos los campos estaban desiertos, solo se veían cubiertos de ciertos arboles espinosos que hacen mui incómodo el camino”.

Frezier, M. 1716

350 – 450 North and East

300 – 350 South and West

“ … me encuentro otra vez en tierras salvajes cubiertas en forma muy rala por acacias y algarrobos cuya compañía, empezaba yo a sospechar, no iba a perder en esta tierra”

Schmidtmeyer, 1824

Dense espinal 350 – 450 North and East

300 – 350 South and West

“ … Numerosos troncos mejor talados a uno o dos pies del suelo, donde ensanchaban su espacio estéril no parecían mejorar su perspectiva, pero indicaban que este valle había estado cubierto tupidamente por ellos antiguamente … “

Schmidtmeyer, 1824

350 – 450 North and East

300 – 350 South and West

“ … noto sobre la vertiente septentrional no crecen sino zarzas … “

Darwin, C. 1839

Sclerophyllous forest

350 – 450 North and East

350 – 1000 South and West

“ … He visto algunos lugares bonitos, que consisten en pequeñas colinas y cañadas de formas suaves, cubiertas de varias clases, mas vegetación que la vista hasta ahora y de verdor ás agradable”

Schmidtmeyer, 1824

350 – 450 North and East

350 – 1000 South and West

Santiago “ … es un hermoso y grande llano como tengo dicho. Tiene a cinco y seis leguas montes de muy buena madera que son unos árboles muy grandes que sacan muy buenas vigas. hay otros árboles que se llaman canela”

Gerónimo de Bibar, 1558

350 – 1000 South and West

Estero de Pocochay como “Maquilemu”, o “bosque de maquis”

Ginés de Lillo, 1605

350 – 1000 South and West

“ … la vertiente meridional está cubierta de un bambú que llega a alcanzar hasta15 pies de altura”

Darwin, C. 1839

Box 2.4 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (69)

43

Assessing the current extent and recent loss of dryland forest ecosystems

Vegetation type Altitude (m a.s.l) and aspect

Sample of original texts described by travelers during the 16th century

Historical references

Jubaea chilensis

500 – 2000 North “Hay palmas y solamente las hay en esta gobernación en dos partes, que es en el río de Maule hay un pedazo que hay de estas palmas, y en Quillota las hay en torno de siete y ocho leguas”

Gerónimo de Bibar, 1558

500 – 2000 North “En algunos lugares se encuentran palmeras y quedo muy asombrado al hallar una de ellas a 4.500 pies de altitud … ”

Darwin, C. 1839

Indigenous settlements

>200 “La población vivía sobre todo en ese Sector, llamado también ‘Camino de Coquimbo’, pues este seguía por Calera a la cuesta del Melón y se juntaba en La Ligua con el de la costa, para continuar al Norte”

Keller, Carlos, 1960

The reconstruction (mapping) of the vegetation revealed that there was once a greater number of species spread over a greater area (Fig. 1). In the 16th century, sclerophyllous forest occupied an area of approximately 115,000 ha, mainly south facing. Espinales covered approximately 670,000 ha on north-facing and flat sites. At the municipal level, 21% of the total area in Casablanca municipality was occupied by sclerohyllous forest and 66% by espinales. In Quilpué, these values were 18% and 68%, respectively; and in Melipilla, 34% and 67%, respectively.

At the species level, the reconstruction of J. chilensis distribution revealed that this once covered approximately 17,000 ha, a much larger area than at present (Fig. 1). The species was distributed in two main populations, one on the coast around the current city of Viña del Mar, and the other in the La Campana mountains (Fig. 1). Nothofa*gus macrocarpa forest covered most of the summit on the Roble hills and the Cantillana hills (Fig. 1), with a total area of 33,800 ha. Towards the east, Porlieria chilensis occupied an area of approximately 5,100 ha of the hills around the Santiago area (Fig. 1).

Table 1

Box 2.4 (cont.)

Figure 1

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (70)

44

Principles and Practice of Forest Landscape Restoration

Although the area of forest coverage in 16th century was larger as compared to the present day, in the eyes of the newly arrived Spanish population of central Chile, tree species were already scarce at that time. The Cabildo therefore issued several directives regulating the cutting of trees. One of the main indigenous settlements in the 16th century was established in the current Quillota valley and occupied approximately 1,800 ha (Fig. 1). Santiago was the main Spanish settlement during the Conquest, extending over 320 ha (Fig. 1). Some documentary sources also provide evidence of the main disturbances affecting vegetation in central Chile in the 18th century after the founding of Santiago. The establishment of Santiago resulted in the expansion of urban areas and rangelands for livestock into the Central Valley during the 18th century. This expansion resulted in the high consumption of tree and shrub species (Cunill, 1995).

Our results demonstrate that by the 16th century, the landscape in central Chile was dominated by different vegetation types. Sclerophyllous forest and some species were more abundant on south-facing slopes, while espinales covered large areas on the north-facing and flat sites across the study area. The presence of this human-induced vegetation type reveals that the original vegetation had already been disturbed by indigenous people. All this indicates that dryland vegetation has been profoundly and irreversibly transformed since the Conquest until the present time, highlighting the need for ecological restoration. The historical analyses presented here can be used to inform restoration plans.

Box 2.4 (cont.)

Box 2.5 Historical reconstruction of land-use patterns from 1920 to 1960 on communal lands of Paso de Ovejas, Veracruz, Mexico

J. Ortiz, F. López-Barrera, J. Callejas, R.H. Manson

The primary forests in the municipality of Paso de Ovejas suffered few alterations following the Spanish conquest. However, in the 19th century these lands came under the management of the military, merchants and foreign investors. Most forest cover in the lower sections of the municipality, where irrigation was increasingly prevalent, were subsequently replaced by commercial plantations such as sugarcane. Conversely, the highlands, hills, slopes and canyons, which were largely isolated from human infrastructure and roads, suffered relatively few alterations to their natural vegetation cover until the first decades of the 20th century, when they were used as pastures.

Increases in human populations were also a contributing factor to these changes. In 1799, census data registered a population of barely 100 people. A century later this number had increased to 3572. The heads of family included men who worked on the land of the large private properties (haciendas) as labourers, sharecroppers, and day labourers. Each family had permission to use a small part of the cultivated land to grow staples such as corn, beans and chillies, and to raise domestic animals such as pigs or chickens.

After the Mexican revolution and the establishment of the Constitution of 1917, these haciendas were largely dismantled as part of agrarian reforms. Initially, the most productive lands were transformed into communal lands known as ejidos and divided among the peasants that had worked in the haciendas for many years. Under the ejidos systems land was kept in collective trusts for the peasant communities who were allowed to use it for farming and natural resource extraction. In the first thirty years of the 20th century, the population of Paso de Ovejas doubled to more than 7350 people. This population growth continued during the next two decades. By the 1970s, owing to both intrinsic population growth on communal lands and the arrival of new immigrants, the population had doubled again reaching a total of 15,271 people. The greatest increases were observed among the male population, specifically the 25 to 29-year age group.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (71)

45

Assessing the current extent and recent loss of dryland forest ecosystems

The combination of rapid population growth on communal lands, including those used for cultivation and cattle ranching, and those considered too infertile for agriculture, put considerable strain on this new form of land tenure. The ejido system is one of the main legacies of the Mexican revolution; it was incorporated in the 1917 Constitution. Ejido members lived in communities. They were designated land for housing with separate parcels of land designated for cultivation. As family sizes increased, the parcels of land they were assigned came under increasing pressure resulting in intensification and transformation of land uses. Forests came under particular pressure as they were felled for fuel or lumber, and in order to make room for crops, and cattle pastures.

In addition to these changes in land ownership and population demographics, the conversion of forest cover to other land uses was actively promoted by public policies, and related financial and technical assistance provided to ejidos by the State and Federal governments. In 1920 the Law for Idle Lands was established nationally and triggered large-scale deforestation across Mexico. The objective behind this law was to increase the volume of crops to feed the rapidly growing Mexican population. This was followed by the creation of the National Ejidal Credit Bank (1935), the organization of farmers in the National Peasant Confederation (1938), the creation of the Agriculture Bureau and Mexican Fertilizers Bureau (1943), the establishment of the National Agriculture Plan (1953), and the formation of the National Seed Producers Organization (1960), all of which facilitated the conversion of forest cover to other land uses.

As a result, remaining forest cover was increasingly limited to inaccessible lands (steep slopes or rocky soils far from roads and towns). Land-use patterns for 23 ejidos in the municipality of Paso de Ovejas from 1927 to 1968 are described in Table 1. These data were obtained through the revision of historical documents and maps from the National Agrarian Registry in Mexico (Registro Agrario Nacional). Forest cover was found to be absent (13 ejidos) or limited (from 19% to 35%) in most ejidos with only two showing forest coverage of more than 40% of their land area (mainly secondary forests; Table 1).

This study showed that in some tropical areas in Mexico, most deforestation, degradation, and fragmentation of forest cover probably occurred prior to 1920. These patterns of land-use change are largely undetectable using current satellite-based methods and imagery data available since the early 1970s. For example, in Paso de Ovejas, the extension of primary forest was only 2.06% in 1973 and increased to 6.28% in 1990 (Montero et al., in prep). These results highlight the importance of a long-term historical perspective for understanding and interpreting current patterns of land use and the overall impact of public policies on patterns of land-use change in the region.

Table 1 Historical records of type of land use present on 23 communal lands (ejidos) in central Veracruz, Mexico, including the year when they were created, area, and percentage cover of different types of land use.

Ejido name

Year Extension (ha)

Rain-fed agriculture

Irrigated agriculture

Grasslands Primary and secondary forest

Urban

Acazónica 1927 1610 14.7 72.9 12.4

Tierra Colorada

1928 150 100

Plan de manantial

1928 312 79.2 20.8

Palmaritos 1928 482 24.1 52.5 23.4

Box 2.5 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (72)

46

Principles and Practice of Forest Landscape Restoration

Ejido name

Year Extension (ha)

Rain-fed agriculture

Irrigated agriculture

Grasslands Primary and secondary forest

Urban

Paso de Ovejas 1930 1617 100

Bandera de Juárez

1930 472 11.0 10.6 78.4

Loma del Nanche

1930 262 64.9 35.1

Puente Jula 1930 168 100

Cerro de Guzmán

1931 359 54.3 21.4 24.2

Mata Grande 1931 100 76.0 24.0

Paso Panal 1932 370 100

Patancán 1932 180 100

La Víbora 1935 279 100

Cantarranas 1935 1124 66.2 33.8

El Angostillo 1936 900 80.0 20.0

Cocuyo 1936 428 80.4 19.6

El Mango 1937 211 100

Mata Mateo 1937 714 52.9 47.1

Mata Grande 1941 260 67.7 32.3

Rancho Nuevo 1958 630 82.5 15.9 1.6

El Angostillo 1964 900 75.2 22.6 2.2

Acazónica 1964 2182 48.1 40.6 9.2 2.1

Loma del Nanche

1968 156 100

Average 602.87 68.57 2.28 15.74 13.15 0.26

Standard deviation

554.98 31.17 10.94 25.85 20.09 0.69

Table 1

Box 2.5 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (73)

47

Assessing the current extent and recent loss of dryland forest ecosystems

Austrocedrus chilensis stand, Nahuel Huapi, Argentina. Photo: A.C. Newton

Shrubland and steppe vegetation, Nahuel Huapi, Argentina. Photo: J. Birch

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (74)

48

Principles and Practice of Forest Landscape Restoration

Box 2.6 Tuning up coarse-grained potential vegetation maps for estimation of historical forest loss in tropical Mexico

R. Vaca, L. Cayuela, J.D. Golicher

In areas with a long history of disturbance, historical forest loss is a major issue. Most deforestation in these areas has occurred prior to the development of remotely sensed techniques. In such circ*mstances potential vegetation maps can be used as a baseline for the estimation of historical forest loss (e.g. Trejo and Dirzo, 2000), as they represent the area hypothetically covered by forest in the absence of human disturbance (Bredenkamp et al., 1998; Moravec, 1998). However there is a general concern that the resolution of most of the available maps of potential vegetation is too coarse for real-world applications (Bredenkamp, et al., 1998; Hartley et al., 2004).

A typical method for constructing a potential vegetation map involves identifying remnants of vegetation with natural or near-natural character (Zerbe, 1998). The vegetation found in these remnants may be assumed to potentially extend to a wider geographical area with similar environmental conditions (Moravec, 1998; Zerbe, 1998). Map inaccuracies often result from the coarse resolution in the available maps of predictor variables (van Etten, 1998). Coarse scale maps may overlook variability in mountainous and other areas in which fine scaled climatic gradients determine the observed vegetation type (Franklin, 1995). In order to use vegetation maps effectively their resolution must be adjusted to the needs of pure and applied biological and ecological research (Araújo et al., 2005; McPherson et al., 2006). For many applications, data at a fine (≤1 km2) spatial resolution are necessary to capture environmental variability that can be partly lost at coarser-grained resolutions (Hijmans et al., 2005). A particular challenge is thus the generation of a fine-grained map of the potential vegetation over a large area. Climate is widely known to condition the formation of different vegetation types (Woodward, 1987). Thus, when the grain of potential vegetation maps is coarser than that of climatic layers, one possible solution is to use climate to downscale potential vegetation maps through statistical modelling.

Here, we illustrate the usage of climatic information in the downscaling of coarse-grained potential vegetation maps with reference to the state of Chiapas, a region historically affected by human activity located in Southern Mexico. The aim of this study was to define the original distribution of the area occupied by different tropical vegetation types in the region. At present, one widely recognized source of information on the potential distribution of vegetation types of Southern Mexico is Rzedowski’s potential vegetation map (1990) (Olson et al., 2001). This is represented at a scale 1:4,000,000, which is clearly limited as a baseline for the estimation of historical forest loss (Trejo and Dirzo, 2000). In response to this problem, we downscaled Rzedowski’s potential vegetation map for Chiapas, from a 1:4,000,000 to 1 km2 grid resolution, using climatically-based random forests models.

To obtain categorical values for the dependent variable (classes of potential vegetation), we systematically extracted points at one km distance from Rzedowski’s potential vegetation map. We did not obtain samples from aquatic vegetation because this vegetation is ‘azonal’, which means that it is not climatically driven. Once the final potential vegetation map was generated, the distribution of aquatic vegetation was explicitly defined based on a soil map developed for Mexico at a scale 1:1,000,000 (INIFAP and CONABIO, 1995). At each point, we extracted values from 55 climatic variables obtained from the WorldClim site (Hijmans et al., 2005). The dataset consists of 36 grids of monthly mean minimum temperature, maximum temperature and precipitation and a set of 19 bioclimatic variables.

The analysis was performed with 1000 trees. As part of its construction, random forests constructs successive independent trees using a bootstrap sample of the data set, each of which produces a vote (Breiman, 1996). In the end, the set of votes is used to generate a simple majority vote for prediction, or scores that provide basic probability estimates, which may then be used in weighted voting (Fawcett, 2006). We used majority vote prediction rules to generate the downscaled potential vegetation map. We validated the model using 256 inventory plots sampled near the transitions between different vegetation types (as these are the areas that are more inaccurate at the original scale). We used

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (75)

49

Assessing the current extent and recent loss of dryland forest ecosystems

validation results to get insights into the performance acquired for the different vegetation classes and for overall model. Then we used receiver operating characteristics (ROC) curves to select the decision thresholds (cut-off thresholds based on probability estimates) that encompass the distribution limits for classes with low agreement and maximize overall prediction accuracy (Fawcett, 2006). At the end, we generated a potential vegetation map using the decision thresholds for these classes. Both the original and the climatically downscaled potential vegetation maps are shown in Fig. 1. The Kappa Index of Agreement showed an increase in accuracy from 0.40 for Rzedowski’s map (95% confidence intervals between 32.5–48.7) to 0.80 for climatically-derived map (95% confidence intervals between 73.9–86.0). Overall accuracy increased from 55.5% to 85.9%. Estimated Kappa for each vegetation class was increased in all cases. Climatically-based random forests models can prove useful to increase the spatial resolution and accuracy of coarse-grained potential vegetation maps in mountainous areas with strong environmental gradients where important climatic variability is obscured at coarse-grained resolution. In conclusion, the proposed method is suitable to generate maps that can be appropriately used as a baseline for the estimation of the historical forest loss.

Figure 1

Box 2.6 (cont.)

production (Rundel, 1999; Balduzzi et al., 1982; Fuentes et al., 1986; Armesto et al., 2007). In addition, successional recovery of forest is usually constrained by continued pressure, water availability, soil erosion, lack of seed banks, disturbance by human-induced fires and limited regeneration capacities of forest species (Balduzzi et al., 1982; Fuentes et al., 1986; Conacher and Sala, 1998; Rundel, 1999; Armesto et al., 2007).

We found that loss of forest extent is partially mitigated by forest re-growth following the abandonment of farmland. In central Chile, we detected forest recovery on about 2.7% of the study area, a rate similar to that documented in other Mediterranean areas (Serra et al., 2008). Forest recovery in central Veracruz may be partially explained as a by-product of comparing data from different types of satellites, but also by declines in agricultural area that may be linked to a reduction in subsidies and increased competition from the US following the NAFTA free-trade accords over the last decade (Pascual and Barbier, 2007). In Oaxaca,

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (76)

50

Principles and Practice of Forest Landscape Restoration

human migration may explain the decrease observed in cultivated land from 1989–1999 (INEGI, 2000). Deforestation in this region was concentrated in a few patches and land-use dynamics are apparently rapid, as traditional crop management involves a 5–10 year continu-ous cultivation period followed by a 10–15 year fallow period.

Drivers of forest change

We identified a number of biophysical and socioeconomic variables that were associated with changes in forest extent across our study areas (Table 2.2). Interestingly, the change in forest extent was explained by a unique combination of variables in each study area and the same variable may have either a positive or a negative effect in the different study areas, i.e. in different ecological, socioeconomical and cultural contexts (Boxes 2.2 – 2.7). For the entire time period addressed in this research, the biophysical variables with the strongest effects on change in forest extent were slope, insolation, and distance from remnant forest. Some of these variables drove forest change in opposite directions (loss or gain) in the different study areas. Thus, whereas the probability of an area experiencing forest loss was higher on gentle slopes, in accordance with our hypothesis, insolation showed different impacts in Chiapas (positive correlation) from central Chile and Oaxaca (negative correlation). Proxim-ity to human settlements and farmland decreased the overall probability of deforestation in most study areas – an example of the ‘curtain effect’ – contrary to our hypothesis. Similarly human density was not found to have a major impact on deforestation. Distance from forests or roads had different effects in different study areas (Table 2.3), thereby partially confirming our hypotheses.

Table 2.3

Variable C. Chile N. Argentina C. Veracruz Chiapas Oaxaca

Elevation - + + +

Slope - - – - - – - - – -

Insolation - – - + + + - – -

Precipitation - – - -

Distance to river + + +

Distance to forest1 + + + - – -

Distance to settlements +2 + + 3 + + +4 + + +

Human density

Distance to roads + + + - – – +

Distance to agriculture + + + + + + + +

Distance to pasture - -

1 Distance within forest edge for the study area in central Chile 2 Cities >20,000 inhabitants 3 Towns <20,000 inhabitants 4 Access to towns >5,000 inhabitants

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (77)

51

Assessing the current extent and recent loss of dryland forest ecosystems

Box 2.7 Different sets of drivers across study regionsIn central Chile, the multivariate logistic regression model for forest-no forest revealed that the probability of an area experiencing forest loss was highly significant (p <0.001) and positively related to the distance from the nearest forest edge for the four study periods, i.e. deforestation progresses from within forest fragments towards the edge, and produces treeless gaps within the forest. This variable had the strongest partial deviance of the four resulting models, accounting for at least half of the deviance explained by the final models achieved in the stepwise selection procedure. For the whole study period, the main explanatory variables of deforestation after distance from the nearest edge were distance from roads and distance from agriculture, all of which were positively correlated with the probability of deforestation, while insolation and slope were both negatively correlated and less relevant in terms of explained variance.

In northern Argentina, the probability of forest loss was highly related (p <0.001) to slope (negative correlation) and distance from rivers (positive correlation) in all study periods. Other significant explanatory variables were distance from roads and mean annual precipitation, showing that deforestation started close to roads and with relatively high precipitation (for dry forests), and moved away from those ideal conditions to current areas further away from roads and with lower precipitation. Distance from the agriculture frontier and from villages showed a significant positive effect, with higher deforestation rates in areas of higher accessibility and human presence.

In central Veracruz, univariate logistic regression models indicated that only four variables were negatively correlated with the probability of forest transformation. They included, in decreasing order of importance, slope, distance from pastures, distance from irrigation infrastructure, and aspect. When these same variables were incorporated into a multivariate logistic regression, the resulting model was highly significant according to the analysis of deviance, although the percentage of deviance explained was relatively low (19.4%). Slope and distance from pasture were found to be significant at the p <0.001 level, whereas the importance of distance from irrigation infrastructure (p = 0.063) and aspect (p = 0.09) declined.

In Chiapas, a GAM model which included elevation and annual rainfall explained the largest proportion of the deviance (12.2%). A model without rainfall explained 11.6% of the deviance, and slope became the variable most strongly associated with the probability that a pixel remained forested. After slope, radiation was the second most important variable both when taken alone and within a multivariate model. Access to large towns (>5000 inhabitants) had greater explanatory power as measured by deviance and partial deviance than access to small villages (>100 inhabitants). The probability that forest has been lost was found to be associated with high values of insolation during winter months, gentle slopes, accessibility to principal markets, low elevations, and low annual precipitation.

In Oaxaca, the probability of forest loss was highly significant and positively related to distance from crop fields and distance from villages in the four study periods. For the entire period (1979–2005), the main explanatory variables for deforestation with significant positive correlations were elevation and distance from agriculture, from villages and from forest; those with significant negative correlation were insolation and distance from forest.

Dryland forest areas in southern Argentina are undergoing a small but significant change to plantations with exotic conifers. Afforestation with exotic pines during the 1973–2003 pe-riod tended to occur at significantly (p <0.05) shorter distances from roads, urban areas and towns >1000 inhabitants. Distance from rivers or lakes seems to have been a poor predictor of afforestation. Contrary to expectation, it tended to occur at longer distances from small towns. Afforestation tended to occur at low elevations and on gentle slopes and at areas with mean annual precipitation higher than the average of the study area. This model was useful to detect possible land-use conflicts with the afforestation process. For example, it has been

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (78)

52

Principles and Practice of Forest Landscape Restoration

suggested that afforestation with exotic conifers has taken place at the expense of land potentially suitable for passive or active restoration of dryland forests of the native conifer Austrocedrus chilensis. If we compare potential for transition towards afforestation with habitat suitability for this species, we obtain the land-use conflict map presented in Fig. 2.4.

Figure 2.4 Left

CentreAustrocedrus chilensis

Right

The differentiated analysis of drivers of deforestation indicated that the trend towards a reduction in natural vegetation cover was determined by a variety of biophysical and socio-economic contexts that resulted in different patterns of land-use types. For the entire time period addressed in this research, the biophysical variables with strongest effects on change in forest extent were slope (higher deforestation in flat areas), insolation, and distance from forest remnants. Slope and associated topographical barriers are fairly typical drivers identi-fied in studies of tropical forest transformation (Geist and Lambin, 2002). In central Chiapas, fire is used to clear forests and to prevent woody re-growth in cattle pastures; therefore, slopes that receive more insolation in the dry season are more likely to be permanently cleared of woody vegetation than shaded slopes. Moreover, Echeverría et al. (Chapter 3) have

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (79)

53

Assessing the current extent and recent loss of dryland forest ecosystems

shown that fire frequency and extent are increasing in these forest landscapes, making them more vulnerable to desertification.

Climate change, particularly rainfall patterns, may be also linked to deforestation (Grau et al., 2005). In recent years, there has been an increase in rainfall in northwestern Argentina (Villalba, 1995), leading to expansion of the agricultural frontier and contributing signifi-cantly to the rapid increase in deforestation in the region. This has been accompanied by technological improvements (e.g. genetically modified soya) and high international demand that have raised product prices.

In central Chile, forest loss occurred with higher probability inside forest stands than at the border. As a consequence, our analyses also detected a higher probability of deforestation at larg-er distances from roads and agricultural fields. The same pattern has been observed in other stud-ies (Ochoa-Gaona and González-Espinosa, 2000) and reveals hidden pressures from cattle grazing and illegal logging activities such as firewood collection and charcoal production (Armesto et al., 2007). Such hidden pressures are not rare in Latin American countries (Callieri, 1996; Aubad et al., 2008) where the rural population often depends on firewood for household consumption as well as the illegal production of charcoal for income generation. In central Veracruz, clearly, the rapid expansion of irrigated agriculture and cattle ranching increased pressure on native forests. Human activity-related factors affect forest fragment accessibility, as has been reported in other studies (Fujisaka et al., 1996; Wassenaar et al., 2007). Settlers converted land to pasture not only to raise cattle, but also to establish unpaved roads and collect firewood. After the forest fragments near the pastures are degraded, land conversion to agriculture or other pasture is more prob-able. In southern Argentina, deforestation is presumably produced by the occurrence of natural and anthropogenic fires, which in many cases do not regenerate back into forests and remain as stable grasslands or shrublands (Mermoz et al., 2005). However, dryland forest areas in this region are undergoing a small but steady change to plantations with exotic conifers. This trend, in concert with other threats such as anthropogenic fires, livestock grazing and the introduction of exotic herbivores such as hares and rabbits are factors that hinder the restoration of dryland forest, which has been fragmented over centuries by native populations and European colonists (Veblen and Lorenz, 1988). Models of land-use/land-cover change can aid the identification of target areas of low conflict for a more rational planning of restoration efforts.

Implications for landscape planning and management

Human interactions with ecosystems are inherently dynamic and complex, and any categoriza-tion of these is an oversimplification. However, there is little hope of understanding these inter-actions without such simplifications (Ellis and Ramankutty, 2008). Working in multiple regions within Latin America enabled us to identify general trends at the regional scale that might be useful for landscape planning and serve as a basis for analyzing proximate drivers of land-cover change. However, the disadvantage of this approach is that it is more difficult to identify and as-sess patterns and process of land-use change at local scales in the real world (e.g. an individual field). Nevertheless, informal interviews that were conducted alongside the field surveys to estab-lish classifications or to ascertain accurate assessments in most study areas provided an important complementary source of information to interpret the detected changes at the regional scale.

Natural vegetation loss and degradation reduce precipitation infiltration and runoff regulation, which promotes soil erosion, landslides and avalanches, and has a negative

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (80)

54

Principles and Practice of Forest Landscape Restoration

impact on ground water recharge (Conacher and Sala, 1998; Millennium Ecosystem Assessment, 2005b). In addition, vegetation cover is tightly associated with water balances within watersheds, biodiversity conservation, and regional climate regulation (Maass et al., 2005; Feddema et al., 2005; Foley et al., 2005; Pielke, 2005). Land-use decisions therefore have consequences for the structure and function of ecosystems and affect provision of environmental goods and services; these decisions also affect humans in ways that go beyond the immediate land-use situation (Turner et al., 2007). The continuous degradation of vegetation cover could have a strong impact on human livelihoods and well-being in the studied dryland landscapes, as there are increasing water demands for agriculture (Cai et al., 2008) and human consumption owing to large population increases.

Environmental problems such as degradation, loss of biodiversity and decreases in pro-ductivity accumulate over the long term and have non-linear effects at regional to global scales (DeFries et al., 2004; Foley et al., 2005). Consequently, strategies for adapted land use, including the optimization of the spatial configuration of uses and restoration of the natu-ral vegetation cover in critical areas should be developed quickly. Strategies should go be-yond preservation within protected areas and logging restrictions along rivers and streams (Turner et al., 2007). For instance, Rey Benayas et al. (2008) proposed the ‘woodland-islet in agricultural seas’ model to conciliate agricultural production and conservation or restoration of native woodlands. Closer monitoring is needed of livestock to establish guidelines for an adapted carrying capacity, as cattle also graze in forests. The repercussions of unsustainable firewood extraction and charcoal production have hardly been quantified in many regions, but we know that they impact strongly on forest conservation (see Chapter 6).

Land-use planning at the regional scale provides a unique opportunity for the establishment of general strategies that may, on one hand, accept or even promote deforestation at particular selected areas, and on the other hand maintain large forested areas suitable for sustainable timber and non-timber forest uses, and probably to a lesser extent, areas for conservation pur-poses. In northern Argentina, a land-use planning policy has been implemented over 10 million hectares of dry forests, zoning different land uses, from deforestation to conservation. Most of the forest corresponds to an intermediate category, theoretically oriented towards forest uses compatible with its own maintenance in the long run. In practice, most forest is heavily de-graded and major efforts should be made to find economic incentives for the local inhabitants to reverse the degradation process and provide value to the remaining forests (see Chapter 10).

Apart from the need for land-use planning, restoration and rehabilitation are important issues in drylands (Le Houerou, 2000; Vallejo et al., 2006). Long-term land-use intensification may represent unique cultural challenges for restoration efforts owing to the long history of human activity, the period of time during which dryland forest has been reduced and de-graded, and generations of inhabitants have grown accustomed to its absence in the studied regions (Piegay et al., 2005; Hobbs, 2009). In Chile, Holmgren and Scheffer (2001) postulated that there might be a window of opportunity for passive restoration through the exclusion of herbivores in El Niño Southern Oscillation (ENSO) years owing to higher water availabil-ity; this strategy could also be applied in southern Argentina given similar ENSO-climate con-nections. It could be especially interesting to use this strategy to establish buffer zones and corridors between remaining old growth forest, which were detected in this study as stable forest areas. Also, forms of adaptive and multifunctional land use such as mixed agroforestry systems should be encouraged as an alternative to monoculture cropping and crop pasture rotations (Ovalle et al., 1996; Aronson et al., 1998).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (81)

55

Assessing the current extent and recent loss of dryland forest ecosystems

Conclusion

The research described here has provided quantitative estimates of forest extent and char-acterized the changes in land-cover in a wide variety of dryland landscapes under contrast-ing ecological, socioeconomic, and cultural scenarios. In addition, research examined the dynamics and drivers of forest loss that has taken place over the last ca. 30 years. We con-cluded that land-use intensification and limited natural regeneration continue to threaten dryland forest cover in many regions of Latin America, but that deforestation rates have diminished in the recent past compared to trends in the early part of the twentieth cen-tury, in accordance with global trends. The probability of an area experiencing forest loss was found to be higher on gentle slopes, and surprisingly, proximity to human settlements and farmland decreased the probability of deforestation in most study areas. Such analyses can help identify those areas that supported native forest in the past, and might therefore be considered as candidates for restoration. In addition, analysis of the factors responsible for forest loss and degradation can inform the development of restoration strategies and plans, by identifying those threatening processes that need to be addressed if restoration actions are to be successful.

References

Adelman, J. 1994. Frontier development: land, labour and capital on the wheatlands of Argentina and Canada, 1890–1914. Oxford Historical Monographs, Clarendon Press, Oxford, UK.

Aguilar, C., Martínez, E., Arriaga L. 2000. Deforestación y fragmentación de ecosistemas: ¿Qué tan grave es el problema en México? Biodiversitas 30: 7–11.

Aide, T.M., Grau, H.R. 2004. Globalization, migration and Latin American ecosystems. Science 305: 1915–1916.

Akaike, H. 1974. A new look at the statistical model identification. IEEE Transactions on Auto-matic Control 19: 716–723.

Angelsen, A., Kaimowitz, D. 1999. Rethinking the causes of deforestation: Lessons from eco-nomic models. The World Bank Research Observer 14: 73–98.

Antrop, M. 2005. Why landscapes of the past are important for the future. Landscape and Urban Planning 70: 21–34.

Araújo, M.B., Thuiller, W., Williams, P.H., Reginster, I. 2005. Downscaling European species atlas distributions to a finer resolution: implications for conservation planning. Global Ecology and Biogeography 14(1): 17–30.

Armesto, J.J., Arroyo, K., Mary, T., Hinojosa, L.F. 2007. The Mediterranean environment of Cen-tral Chile. In: Velben, T.T., Young, K.R., Orme, A.R. (eds.), The Physical Geography of South America. Oxford University Press, New York, USA: pp. 184–199.

Armesto J.J., Manuschevich, D., Mora, A., Smith-Ramírez, C., Rozzi, R., Marquet, P.A.. 2010. A historical framework for land-cover transitions in south-central Chile during the Anthro-pocene. Land Use Policy 27: 148-160.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (82)

56

Principles and Practice of Forest Landscape Restoration

Aronson, J., del Pozo, A., Ovalle, C. Avendaño, J., Lavin, A. 1998. Land use changes in Central Chile. In: Rundel, P.W., Montenegro, G., Jaksic, F. (eds.), Landscape Disturbance and Bio-diversity in Mediterranean-type Ecosystems. Springer-Verlag Berlin Heidelberg, Germany: pp.155–168.

Aubad, J., Aragón, P., Oalla-Tárraga, M.A., Rodríguez, M.A. 2008. Illegal logging, landscape structure and the variation of tree species richness across North Andean forest remnants. Forest Ecology and Management 255: 1892–1899.

Balduzzi, A., Tomaselli, R., Serey, I., Villaseñor, R. 1982. Degradation of the Mediterranean type of vegetation in central Chile. Ecologie Méditerranée 7: 223–240.

Barros, A.D. 1884. Historia General de Chile. Tomo I. Santiago, Chile. Editorial Universitaria. Centro de Investigaciones Diego Barros Arana, Dibam.

Bredenkamp, G., Chytry, M., Fischer, H.S., Neuhäuslová, Z., van der Maarel, E. 1998. Vegeta-tion mapping: theory, methods and case studies. Applied Vegetation Science 1: 161–266.

Breiman, L. 1996. Bagging predictors. Machine Learning 24: 123–140.

Brown, A.D., Malizia, L.R. 2004. Las selvas pedemontanas de las Yungas: en el umbral de la extinción. Ciencia Hoy 14: 52–63.

Cai, X., Ringler, C., You, J.Y. 2008. Substitution between water and other agricultural inputs: Implications for water conservation in a River Basin context. Ecological Economics 66: 38–50.

Callierie, C. 1996. Degradación y deforestación del bosque nativo por extracción de leña. Ambiente y Desarrollo 12: 41–48.

Camus, P. 2002. Bosques y tierras despejadas en el período de la conquista de Chile. En Reta-mal Ávila, Julio (Coordinador): Estudios Coloniales II. Santiago, Chile. Editorial Biblioteca Americana. Universidad Andrés Bello.

Cayuela, L., Rey Benayas J.M., Echevarria, C. 2006. Clearance and fragmentation of tropical montane forests in the highlands of Chiapas, Mexico (1975–2000). Forest Ecology and Management 226: 208–218.

Challenger, A. 1998. Utilización y conservación de los ecosistemas terrestres de México: pasado, presente y futuro. CONABIO, Instituto de Biología de la UNAM y Agrupación Sierra Madre, S.C., México, D.F, México.

Challenger, A., Dirzo, R. 2009. Factores de cambio y estado de la biodiversidad. In: CONABIO (ed.), Capital natural de México, Vol. II: Estado de conservación y tendencias de cambio. CONABIO, México, D.F, México: pp. 37–73.

Chavez, P.S. 1996. Image-based atmospheric corrections. Revisited and improved. Photogram-metric Engineering Remote Sensing 62: 1025–1036.

Conacher, A.J., Sala, M. 1998. Land degradation in Mediterranean environments of the world: nature and extent, causes and solutions. John Wiley and Sons Ltd, Chichester, UK.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (83)

57

Assessing the current extent and recent loss of dryland forest ecosystems

Cristóbal, L., Pacheco, S., Malizia, L., Echeverria, C., in preparation. Deforestation and fragmen-tation of Yungas Premontane Forest in NW Argentina (1976–2006). Forest Ecology and Management.

Cunill, P. 1995. Transformaciones del espacio geohistórico latinoamericano, 1930–1990, Méx-ico, Fondo de Cultura Económica.

Darwin, C. 1945. Viaje de un naturalista alrededor del mundo. Buenos Aires, Argentina. Li-brería El Ateneo.

De Bibar, G. 1966. Crónica y relación copiosa y verdadera de los reinos de Chile (1558). Tran-scripción paleográfica de Irving A. Leonard. Santiago, Chile. Edición facsimilar y a plana del Fondo Histórico y Bibliográfico José Toribio Medina. 37pp.

DeFries, R.S., Foley, J.A., Asner, G. P. 2004. Land-use choices: balancing human needs and eco-system function. Frontiers in Ecology and the Environment 2: 249–257.

Díaz-Gallegos, J.R., Mas, J.F., Velásquez, A. 2008. Monitoreo de patrones de deforestación en el corredor biológico mesoamericano, México. Interciencia 33: 882–890.

Donoghue, D.N.M. 2002. Remote sensing: environmental change. Progress Physical Geogra-phy 26: 144–151.

Donoso, C. 1982. Reseña Ecológica de los bosques mediterráneos de Chile. Revista Bosque, Valdivia, Chile, Universidad Austral 4(2): 117.

Donoso, C. 1995. Bosques Templados de Chile y Argentina. Variación estructura y dinámica. Editorial Universitaria. Santiago, Chile. 483pp.

Echeverria, C., Coomes, D., Salas, J., Rey-Benayas, J.M., Lara, A., Newton, A. 2006. Rapid defor-estation and fragmentation of Chilean Temperate Forests. Biological Conservation 130: 481–494.

Echeverría, C.T., Kitzberger, T., Rivera, R., Manson, R., Vaca, R., Cristóbal, L., Machuca, G., González, D., Fuentes, R. 2011. Assessing fragmentation and degradation of dryland forest ecosystems. In: Newton, A.C., Tejedor, N. (eds.), Principles and practice of forest landscape restoration: case studies from the drylands of Latin America. IUCN, Gland, Switzerland.

Ellis, E.C., Ramankutty, N. 2008. Putting people on the map: anthropogenic biomes of the world. Frontiers in Ecology and the Environment 6: 439–447.

FAO. 2010. Global forests resources assessment 2010. Food and Agriculture of the United Na-tions Organization, Rome, Italy.

Fawcett, T. 2006. An introduction to ROC analysis. Pattern Recognition Letters 27: 861–874.

Feddema, J.J., Oleson, K.W., Bonan, G.B., Mearns, L.O., Buja, L.E., Meehl, G.A., Washington, W.M. 2005. The importance of land-cover change in simulating future climates. Science 310: 1674–1678.

Feranec, J., Jaffrain, G., Soukup, T., Hazeu, G. 2010. Determining changes and flows in Euro-pean landscapes 1990–2000 using CORINE land-cover data. Applied Geography 30: 19–35.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (84)

58

Principles and Practice of Forest Landscape Restoration

Foley, J.A., DeFries, R., Asner, G.P., Barford, C., Bonan, G., Carpenter, S.R., Chapin, F.S., Coe, M.T., Daily, G.C., Gibbs, H.K., Helkowski, J.H., Holloway, T., Howard, E.A., Kucharik, C.J., Monfreda, C., Patz, J.A., Prentice, C., Ramankutty, N., Snyder, P.K. 2005. Global consequences of land use. Science 309: 570–574.

Franklin, J. 1995. Predictive vegetation mapping: geographic modelling of biospatial pat-ters in relation to environmental gradients. Progress in Physical Geography 19(4): 474–499.

Frezier, M. 1902. Relación del viaje por el mar del sur a las costas de Chile y el Perú durante los años de 1712, 1713 i 1714. Santiago, Chile. Imprenta Mejía.

Fuentes, E.R., Hoffmann, A.J., Poiani, A., Alliende, M.C. 1986. Vegetation change in large clear-ings: patterns in the Chilean matorral. Oecologia 68: 358–366.

Fuentes, E.R., Jaksic, F.M., Simonetti, J. 1983. European rabbits vs. native rodents in central Chile: Effects on shrub seedlings. Oecologia 58: 411–414.

Fujisaka, S., Bell, W., Thomas, N., Hurtado, L., Crawford, E. 1996. Slash-and-burn agriculture, conversion to pasture, and deforestation in two Brazilian Amazon colonies. Agriculture, Ecosystems and Environment 59: 115–130.

Gasparri, N.I., Grau, H.R. 2009. Deforestation and fragmentation of Chaco dry forest in NW Argentina (1972–2007). Forest Ecology and Management 258: 913–921.

Geist, H.J., Lambin, E.F. 2002. Proximate causes and underlying driving forces of tropical de-forestation. Bioscience 52: 143–150.

Ginés de Lillo, 1942. Mensuras de Ginés de Lillo (con introducción de Aniceto Almeyda). En Colección de historiadores de Chile y de documentos relativos a la historia nacional, tomo XLIX. Santiago, Chile. Imprenta Universitaria.

Gowda, J.H., Kitzberger, T., Premoli, A.C., in preparation. A century of land use change and landscape response in the northern Patagonian forest-steppe transition: trends, drivers and legacies. Ecology and Society.

Grau, H.R., Gasparri, N.I., Aide, T.M. 2005. Agriculture expansion and deforestation in season-ally dry forests of north-west Argentina. Environmental Conservation 32: 140–148.

Grau, H.R., Aide, M. 2008. Globalization and land-use transitions in Latin America. Ecology and Society 13(2): 16. < http://www.ecologyandsociety.org/vol13/iss2/art16/>.

Grau, J. 2004. Palmeras de Chile: revisión exhaustiva de las dos palmeras endémicas y reseña de las especies introducidas. Ediciones OIKOS. 203pp.

Harper, G.J., Steininger, M.K., Tucker, C.J., Juhn, D., Hawkins, F. 2007. Fifty years of de-forestation and forest fragmentation in Madagascar. Environmental Conservation 34: 325–333.

Hartley, S., Kunin, W.E., Lennon, J.J. and Poco*ck, M.J. 2004. Coherence and discontinuity in the scaling of species distribution patterns. Proceedings of the Royal Society of London, Series B 271: 81–88.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (85)

59

Assessing the current extent and recent loss of dryland forest ecosystems

Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G. and Jarvis, A. 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25(15): 1965–1978.

Hobbs, R. 2009. Woodland restoration in Scotland: Ecology, history, culture, economics, poli-tics and change. Journal of Environmental Management 90: 2857–2865.

Holmgren, M., Scheffer, M. 2001. El Niño as a window of opportunity for the restoration of degraded arid ecosystems. Ecosystems 4: 151–159.

Idrisi. 2006. Idrisi Andes. Guide to GIS and Image Processing. Clark Labs, Clark University, Worcester, USA.

Instituto Nacional de Estadística, Geografía e Informática (INEGI). 2000. XII Censo General de Población y Vivienda, 2000, México, D.F., México.

Instituto Nacional de Investigaciones Forestales y Agropecuarias (INIFAP) and Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO). 1995. Edafología. Scales 1:250 000 and 1:1 000 000. México D.F.

Kahn, J.R., McDonald, J.A. 1997. The role of economic factors in tropical deforestation. In: Laurence, W.F., Bierregaard, R.O.J. (eds.), Tropical Forest Remnants Ecology, Management and Conservation of Fragmented Communities. The University of Chicago Press, Chicago, USA: pp. 13–18.

Keller, C. 1960. Los orígenes de Quillota. Apartado del Boletín de la Academia Chilena de Historia, N°61. Santiago: p. 19.

Klepeis, P., Vance, C. 2003. Neoliberal policy and deforestation in southeastern Mexico: an as-sessment of the PROCAMPO program. Economic Geography 79: 221–240.

Kusnetzoff, F. 1987. Urban and housing policies under Chile’s military dictatorship 1973–1985. Latin American Perspectives 14: 157–186.

Lambin, E.F., Geist, H.J., Lepers, E. 2003. Dynamics of land-use and land-cover change in tropi-cal regions. Annual Review of Environment and Resources 28: 205–241.

Lara, A., Veblen, T.T. 1993. Forest plantations in Chile: a successful model? In: Marther, A. (ed.), Afforestation Policies, Planning and Progress. Belhaven Press, London, UK: pp. 118–139.

Le Houerou, H.N. 2000. Restoration and rehabilitation of arid and semi-arid Mediterranean ecosystems in North Africa and West Asia: A review. Arid Soil Research and Rehabilitation 14: 3–14.

Lemons, J. 2006. Conserving dryland biodiversity: Science and policy. Science and Devel-opment Network, Policy Briefs. <http://www.scidev.net/en/policy-briefs/conserving-dry-land-biodiversity-science-and-policy.html>.

López, S., Sierra, R. 2010. Agricultural change in the Pastaza River Basin: A spatially explic-it model of native Amazonian cultivation. Applied Geography 29, doi:10.1016/j.apge-og.2009.10.004.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (86)

60

Principles and Practice of Forest Landscape Restoration

Lu, D., Mausel, P., Brondizio, E., Moran, E. 2004. Change detection techniques. International Journal of Remote Sensing 25: 2365–2407.

Maass, J., Balvanera, P., Castillo, A., Daily, G.C., Mooney, H.A., Ehrlich, P., Quesada, M., Miranda, A., Jaramillo, V.J., García-Oliva, F., Martínez-Yrizar, A., Cotler, H., López-Blanco, J., Pérez-Jimén-ez, A., Búrquez, A., Tinoco, C., Ceballos, G., Barraza, L., Ayala, R., Sarukhán, J. 2005. Ecosystem services of tropical dry forests: insights from long-term ecological and social research on the Pacific Coast of Mexico. Ecology and Society 10(1): 17. <http://www.ecologyandsoci-ety.org/vol10/iss1/art17/>.

Manson, R.H., López-Barrera F., Landgrave, R., in preparation. Patterns and drivers of tropical deciduous dry forest transformation in central Veracruz, Mexico.

McPherson, J.M., Jetz, W., Rogers, D.J. 2006. Using coarse-grained occurrence data to predict species distributions at finer spatial resolutions—possibilities and limitations. Ecological Modelling 192(3–4): 499–522.

Mermoz, M., Kitzberger, T., Veblen, T.T. 2005. Landscape influences on occurrence and spread of wildfires in Patagonian forests and shrublands. Ecology 86: 2705–2715.

Millennium Ecosystem Assessment. 2005a. Ecosystems and human well-being. Current state and trends. Island Press, Washington, D.C., USA.

Millennium Ecosystem Assessment. 2005b. Ecosystems and human well-being: Desertifica-tion synthesis. World Resources Institute, Washington, D.C., USA.

Montero Solano, J.A. 2009. El papel de las políticas públicas en el cambio de uso de suelo en el centro de Veracruz: hacia la restauración del paisaje forestal, la conservación de la biodiversi-dad y el desarrollo sustentable. MSc. Thesis, Universidad Anahuac, Xalapa, Mexico.

Montero Solano, J.A., Manson, R.H., López Barrera, F., Ortiz, J., Callejas, J., in preparation. Public policy and land use change in central Veracruz: an important factor in efforts to restore a tropical dry forest landscape.

Moravec, J. 1998. Reconstructed natural versus potential natural vegetation in vegetation mapping: A discussion of concepts. Applied Vegetation Science 1(2): 173–176.

Myers, N., Mittermeier, R., Mittermeier, C., Da Fonseca, G., Kent, J. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853–858.

Newton, A.C. 2008. Restoration of dryland forests in Latin America: The ReForLan project. Ecological Restoration 26: 10–13.

Ochoa-Gaona, S., Gonzalez-Espinosa, M. 2000. Land use and deforestation in the highlands of Chiapas, Mexico. Applied Geography 20: 17–42.

Ojima, D.S., Galvin, K.A., Turner, B.L. 1994. The Global Impact of Land-Use Change. BioScience 44: 300–304.

Olander, L.P., Gibbs, H.K., Steininger, M., Swenson, J.J., Murray, B.C. 2008. Reference scenarios for deforestation and forest degradation in support of REDD: a review of data and meth-ods. Environmental Research Letters 3: 025011.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (87)

61

Assessing the current extent and recent loss of dryland forest ecosystems

Ovalle, C., Avendaño, J., Aronson, J., Del Pozo, A. 1996. Land occupation patterns and vegeta-tion structure in the anthropogenic savannas (espinales) of central Chile. Forest Ecology and Management 86: 129–139.

Parés-Ramos, I.K., Gould, W.A., Aide, T.M. 2008. Agricultural abandonment, suburban growth, and forest expansion in Puerto Rico between 1991 and 2000. Ecology and Society 13(2): 1. < http://www.ecologyandsociety.org/vol13/iss2/art1/>.

Pascarella, J.B., Aide, T.M., Serrano, M.I., Zimmerman, J.K. 2000. Land-use history and forest regeneration in the Cayey Mountains, Puerto Rico. Ecosystems 3: 217–228.

Pascual, U., Barbier, E.B. 2007. On price liberalization, poverty, and shifting cultivation: An example from Mexico. Land Economics 83: 192–216.

Pfaff, A.S.P. 1999. What drives deforestation in the Brazilian Amazon? Evidence from satellite and socioeconomic data. Journal of Environmental Economics and Management 37: 26–43.

Piégay, H., Gregory, K.J., Bondarev, V., Chin, A., Dahlstrom, N., Elosegi, A., Gregory, S.V., Joshi, V., Mutz, M., Rinaldi, M., Wyzga, B., Zawiejska, J. 2005. Public perception as a barrier to intro-ducing wood in rivers for restoration purposes. Environmental Management 36: 665–674.

Pielke, R.A. 2005. Land use and climate change. Science 310: 1625–1626.

Poeppig, E. 1960. Un testigo en la alborada de Chile (1826–1829). Santiago, Chile. Editorial Zigzag.

Pontius Jr, R.G., Shusasand, E., McEachern, M. 2004. Detecting important categorical land changes while accounting for persistence. Agriculture, Ecosystems and the Environment 101: 251–268.

Prieto, M., Villagra, P., Lana, N., Abraham, E. 2003. Utilización de documentos históricos en la reconstrucción de la vegetación de la Llanura de la Travesía (Argentina) a principios del siglo XIX. Revista Chilena de Historia Natural 76: 613–622.

Rey Benayas, J.M., Bullock, J., Newton, A.C. 2008. Creating woodland islets to reconcile eco-logical restoration, conservation, and agricultural land use. Frontiers in Ecology and the Environment 6: 329–336.

Rey Benayas, J.M., Schulz, J., Cayuela, L., Echeverría, C., Salas, J., Kitzberger, T., Cristóbal, T., Man-son, R., López-Barrera, F., Vaca, R., Golicher, D., Rivera, R., del Castillo, R. 2010a. Synthetic Re-port on GIS Analysis and associated regression models of project Restoration of forest land-scapes for biodiversity conservation and rural development in the drylands of Latin America (REFORLAN EU INCO PROGRAMME INCO-CT–2006–032132). Unpublished material.

Rey Benayas, J.M., Schulz, J., Echeverría, C., Salas, J., Kitzberger, T., Cristóbal, T., Manson, R., López-Barrera, F., Vaca, R., Golicher, D., Rivera, R., del Castillo, R. 2010b. Synthetic Report on Maps of current forest cover and forest loss, produced in report form of project Res-toration of forest landscapes for biodiversity conservation and rural development in the drylands of Latin America (REFORLAN EU INCO PROGRAMME INCO-CT–2006–032132). Unpublished material.

Rzedowski, J. 1990. Vegetación Potencial. Atlas Nacional de México. Vol. 2. Scale 1:4,000,000. Instituto de Geografía. Universidad Nacional Autónoma de México (UNAM).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (88)

62

Principles and Practice of Forest Landscape Restoration

Schulz, J., Cayuela, L., Echeverria, C., Salas, J., Rey Benayas, J.M. 2010. Land-cover dynamics of the dryland forest landscape of Central Chile. Applied Geography 30: 436–447.

Serra, P., Pons, X., Saurí, D. 2008. Land-cover and land-use change in a Mediterranean land-scape: A spatial analysis of driving forces integrating biophysical and human factors. Ap-plied Geography 28: 189–209.

Silva, E. 2004. The political economy of forest policy in Mexico and Chile. Singapore Journal of Tropical Geography 25: 261–280.

Shao, G., Wu, J. 2008. On the accuracy of landscape pattern analysis using remote sensing data. Landscape Ecology 23: 505–511.

Schmidtmeyer, M. 1947. Viaje a Chile a través de los Andes. En los años 1820 y 1821. Buenos Aires, Argentina. Editorial Claridad.

Teillet, P.M., Guindon, B., Goodeonugh, D.G. 1982. On the slope-aspect correction of multi-spectral scanner data. Canadian Journal of Remote Sensing 8: 84–106.

Therneau, T.M., Atkinson, B. 2009. R port by Brian Ripley. rpart: Recursive Partitioning. R pack-age version 3.1–45 (2009) S-PLUS 6.x original at http://mayoresearch.mayo.edu/mayo/research/biostat/splusfunctions.cfm.

Trejo, I., Dirzo, R. 2000. Deforestation of seasonally dry tropical forest: a national and local analysis in Mexico. Biological Conservation 94: 133–142.

Turner, B.L., Moss, R.H., Skole, D.L. 1993. Relating land use and global land-cover change: A proposal for an IGBP-HDP core project. Report from the IGBP-HDP Working Group on Land-Use/Land-Cover Change. Joint publication of the International Geosphere-Bio-sphere Programme (Report No. 24) and the Human Dimensions of Global Environmen-tal Change Programme (Report No. 5). Royal Swedish Academy of Sciences, Stockholm, Sweden.

Turner, B.L.II., Lambin, E.F., Reenberg, A. 2007. The emergence of land change science for global environmental change and sustainability. Proceedings of the National Academy of Science USA 104, 20666–20671.

Universidad de Chile. 2007. Profundización de la línea de base ambiental y ecológica del sec-tor de mayor valor ecológico del Cordón de Cantillana. Environmetal National Committee (CONAMA). Report. 260pp.

Vaca, R., Cayuela, L., Golicher, D., in preparation. A quantitative analysis of land-cover change and degradation in Chiapas, Mexico (1990–2006). Biotropica.

Valdés, A., Foster, W. 2005. Externalidades de la Agricultura Chilena. Ediciones Universidad Católica de Chile, Santiago de Chile, Chile.

Vallejo, R., Aronson, J., Pausas, J.G., Cortina, J. 2006. Restoration of Mediterranean woodlands. In: van Andel, J., Aronson, J. (eds.), Restoration Ecology: The New Frontier. Blackwell Pub-lishing, Malden, USA: pp. 193–207.

van Etten, E.J.B. 1998. Mapping vegetation in an arid, mountainous region of Western Aus-tralia. Applied Vegetation Science 1(2): 189–200.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (89)

63

Assessing the current extent and recent loss of dryland forest ecosystems

Veblen, T.T., Lorenz, D.C. 1988. Recent vegetation changes along the forest/steppe ecotone of northern Patagonia. Annals of the Association of American Geographers 78: 93–111.

Velázquez, A., Durán, E., Ramírez, I., Mas, J.F., Bocco, G., Ramírez, G., Palacio, J.L. 2003. Land use-cover change processes in highly biodiversity areas: the case of Oaxaca, Mexico. Global Environmental Change 13: 175–184.

Villagrán, C. 1995. El Cuaternario en Chile: evidencias de cambio climático. In: Argollo, J., Mour-guiart, P.H. (eds.), Cambios cuaternarios en América del Sur: pp. 191–214. ORSTOM, La Paz.

Villalba, R. 1995. Estudios dendrocronológicos en la selva Subtropical de Montaña, implicaciones para su conservación y desarrollo. In: Investigación, conservación y desarrollo en las selvas subtropicales de montaña. In Brown, A.D., Grau, H.R. (eds.), Laboratorio de Investigaciones Ecológicas de las Yungas, Universidad Nacional de Tucumán, Tucumán, Argentina: pp. 59–68.

Vitousek, P.M. 1994. Beyond global warming: ecology and global change. Ecology 75: 1861–1876.

Wassenaar, T., Gerber, P., Verburg, P.H., Rosales, M., Ibrahim, M., Steinfeld, H. 2007. Projecting land use changes in the Neotropics: the geography of pasture expansion into forest. Global Environmental Change 17: 86–104.

Wood, S. 2004. mgcv: GAMs with GCV smoothness estimation and GAMMs by REML/PQL. R package version 1: 1–8.

Woodward, F. 1987. Climate and plant distribution. Cambridge University Press, Cambridge. 158pp.

Yuan F., Sawaya, K.E., Loeffelholz, B.C., Bauer, M.E. 2005. Land-cover classification and change analysis of the Twin Cities (Minnesota) Metropolitan Area by multitemporal Landsat re-mote sensing. Remote Sensing of the Environment 98: 317–328.

Zerbe, S. 1998. Potential natural vegetation: validity and applicability in landscape planning and nature conservation. Applied Vegetation Science 1(2): 165–172.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (90)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (91)

65

3 ASSESSING FRAGMENTATION AND

DEGRADATION OF DRYLAND FOREST

ECOSYSTEMS

C. Echeverría, T. Kitzberger, R. Rivera, R. Manson, R. Vaca, L. Cristóbal, G. Machuca, D. González, R. Fuentes

Introduction

Spatial patterns of forest cover can be understood as the spatial arrangement or configura-tion of forested ecosystems across a landscape (Forman and Godron, 1986). The importance of studying spatial patterns of forest cover is now widely appreciated, owing to the complex link between pattern and process in a landscape (Nagendra et al., 2004), and the widely documented effects of habitat fragmentation on biodiversity. As a result, diverse studies have sought to develop measures of landscape pattern that may be used to monitor changes in forest cover (Sano et al., 2009; Shuangcheng et al., 2009; Zeng and Wu, 2005).

According to the driving factors that operate in a given landscape, spatial pattern can present a variety of different behaviours over time. For instance, loss and fragmentation of forest cover are among the most important transformations of landscape configuration oc-curring in many parts of the world (Carvalho et al., 2009; Fialkowski and Bitner, 2008). On the other hand, pattern change associated with forest recovery or regeneration may lead to an increase of forest cover and connectivity (Baptista 2010; Box 3.1).

Box 3.1 Landscape features associated with the passive recovery of Mediterranean sclerophyllous woodlands of central Chile

A. Rivera-Hutinel, A. Miranda, T. Fuentes-Castillo, C. Smith-Ramirez, M. Holmgren

Although Mediterranean ecosystems are considered global hotspots of biodiversity and priority targets for conservation (Myers et al., 2000), they are among the most severely degraded and fragmented ecosystems in the world. In central Chile, land-cover of Mediterranean sclerophyllous woodlands (Chilean matorral) has been significantly reduced and transformed by a combination of human activities including logging, firewood extraction, vegetation burning, agriculture, livestock grazing, and the spread of exotic species of herbivores (Fuentes and Hajek, 1979; Holmgren, 2002; see Chapter 2). Ecosystems that have been highly degraded and extirpated from large areas, such as Chilean Mediterranean forests, are difficult and expensive to restore, especially because of the extremely dry and long summer period and strong impact of herbivory. Both factors, in addition to recurrent fire, can stop or retard successional processes (Fuentes et al., 1984). Frequently, severely degraded dryland ecosystems cannot be returned to their pre-disturbance condition without expensive management. The less costly strategies to restore vegetation cover in these ecosystems is to combine the passive regeneration of relatively less impacted areas, resulting from relatively slow natural processes, with active restoration activities that stimulate vegetation change from early successional stages to more mature and diverse forest.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (92)

66

Principles and Practice of Forest Landscape Restoration

We assessed the regeneration potential of sclerophyllous woodlands of central Chile (33º S) over 50 years at three sites in the foothills of the Andes and two sites in the Coastal Range, and related the rates of vegetation change to specific landscapes features. Each study site (Fig. 1) was a mosaic of sclerophyllous forest and open pastures, with an average 40% of woodland cover and an extension of 700 ha on average (range: 631–911 ha) and had not been burned for at least two decades (1985–2008). Vegetation change was determined by comparison of aerial photographs taken in 1955 and 2007 over a regular grid of 250 m of points using standard supervised classification methods. We considered as evidence of woodland regeneration (1) a change in land-cover for a given point in the grid from bare soil or artificial grassland to forest cover. Persistence of the open cover condition was considered as a lack of forest regeneration (0). Any other observed changes in the vegetation or the maintenance of forest cover were excluded from the analyses. We related the recovery of forest cover to topographic variables (slope, orientation, altitude and exposure to solar radiation), as well as to spatial location of the regenerating patch (distance to the closest forest patch present in 1955, and distance to the nearest ravine). We used spatial regression models to control for spatial autocorrelation among sampling points.

We found an average rate of increase in land-cover of sclerophyllous forest from 0.4–1.0 ha/year. The probability of recovery of forest cover increased significantly at shorter distances from remnant (1955) forest patches, especially on south-facing slopes. This effect may be related to fact that patches can be a source of propagules but their environmental conditions may also facilitate tree seed germination and seedling survival (Fuentes et al., 1984, 1986; Holmgren et al., 2000). The spatial regression models also suggest that regeneration occurs in patches (at a 250 m scale), which could be related to local differences in grazing pressure, resource availability (nutrients and water) and micro-climatic conditions (temperature and air relative moisture).

Our work shows that Chilean sclerophyllous forest, which is considered strongly resistant to passive recovery from severe disturbance, can grow back in unburned sites under certain conditions. The proximity to existing forest patches or seed sources, slope aspect, and the aggregated patch structure of the vegetation are key features to be considered in the design of successful long-term restoration strategies to promote the passive restoration of Mediterranean sclerophyllous woodlands. The removal of herbivores, if possible, could accelerate the passive recovery of woodland vegetation cover (see also Chapter 8).

Figure 1

Box 3.1 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (93)

67

Assessing fragmentation and degradation of dryland forest ecosystems

Central Valley of Chiapas, Mexico; tropical deciduous forest. Photo: R. Vaca

Deforestation of seasonal dry premontane forest in northwestern Argentina. Photo: L. Malizia

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (94)

68

Principles and Practice of Forest Landscape Restoration

Progressive deforestation typically results in an increase in the spatial heterogeneity, fragmentation, and edge characteristics of a forested landscape (Trani and Giles, 1999). In particular, fragmentation refers to the division of spatially continuous forest areas into isolated patches, which are separated by some other type of land-cover (such as agricultural land), commonly referred to as the landscape matrix (Forman and Godron, 1986). At the patch level, fragmentation causes an increase in patch isolation and edge, and a reduction of patch size (Echeverria et al., 2006). In turn, this can increase the isolation of populations of individual species (Echeverría et al., 2007), and can reduce population viability through its effects on key ecological processes such as dispersal, migration and gene flow (Giriraj et al., 2010; Vergara and Armesto, 2009). As a result, for-est fragmentation is now considered to be one of the principal causes of biodiversity loss (Baillie et al., 2004). As forest loss takes place in a landscape, certain changes in the spatial configuration of the landscape can be observed (Cayuela et al., 2006; Geri et al., 2009). The analysis of the spatial attributes through landscape indices is a suitable ap-proach to demonstrate the process of forest fragmentation at the landscape level (Zeng and Wu, 2005). Additionally, information on landscape structure can be used to inform forest management (Sano et al., 2009).

Dryland systems are recognized as being of high biodiversity value, while representing the largest terrestrial biome on the planet (MEA, 2005; Schimel, 2010). Throughout the ar-eas where they occur, dryland forests have been rapidly degrading and declining owing to anthropogenic disturbance (Hill et al., 2008; Ravi et al., 2010; Reynolds et al., 2007). Loss of dryland forests has had a significant impact on carbon sequestration and temperature at the global scale (Rotenberg and Yakir, 2010). In Latin America, this ecosystem has been as-sociated with human poverty, unhealthy living conditions and environmental degradation (Altieri and Masera, 1993). Management for the conservation and sustainable use of dry-land forests should consider restoration approaches and habitat modification (McIntyre and Hobbs, 1999) with the aim of enhancing both biodiversity and human livelihoods. Few stud-ies on the spatial pattern of dryland landscapes have been performed to examine the effects of human activity on dryland forests (Wang et al., 2010), particularly in the context of their restoration.

In this chapter we present the results of research that assessed the trends in landscape patterns of forest cover in seven dryland study areas in Mexico, Argentina and Chile, by analyzing the dynamics of selected landscape metrics over the last four decades. Through a comparative analysis of these study areas, we identified both the high variability of landscape metrics and common trends in the spatial patterns of dryland forest. The aim of this research was to inform the development of plans for forest landscape restoration, one of the objec-tives of which is to restore connectivity in forest areas that have been fragmented (Box 3.2). Examination of the patterns and processes influencing forest fragmentation are therefore of direct relevance to the development of restoration approaches to be implemented at the landscape scale. The study landscapes included Veracruz, Oaxaca and Chiapas in México; Salta in northern Argentina and Bariloche in southern Argentina; and central Chile (Fig 3.1). A set of Landsat satellite images was classified and analyzed in each study area (see Chapter 2). Most of the study intervals spanned the past four decades, except for Chiapas where the period from 1990 to 2005 was analyzed.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (95)

69

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 3.1

Figure 3.2

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (96)

70

Principles and Practice of Forest Landscape Restoration

Box 3.2 Landscape connectivity in the highly fragmented drylands of the Central Valley of Chiapas

R. Vaca, J.D. Golicher, L. Cayuela

The current pattern of forest cover observed in the Central Valley of Chiapas is the result of historic deforestation. We found that 68% of the original hypothesized area of dry forest (as defined by Olson et al., 2001) was lost by 1990. The remaining forest (32% of the original putative area) is either retained by local landowners for its utility as a source of fuelwood and timber, or located in nature reserves or sites with steep slopes and low accessibility. Most of the forest in this region is highly fragmented, and only 19% is found in core areas, i.e. forest with a minimum distance of 110 m from the nearest patch edge. This landscape, dominated by human land use, presents a significant challenge for maintaining and conserving biodiversity.

Although most of the original forest cover had already been lost, the forest has not been completely cleared and replaced with an inhospitable matrix, as has occurred in other agricultural croplands or suburban environments of the world. The spatial concept of fragmentation in this case does not necessarily imply that habitat remnants are isolated by areas that function as hostile environments to the organisms within the remnants (Cayuela, 2009). The agricultural landscape still retains large isolated trees, woodlots, scattered groups of trees, secondary regrowth, hedgerows and living fences, amongst forest and shrubland patches of varying size, disturbance and management history. Together, they provide the habitats upon which the conservation of much of the flora and fauna in developed landscapes ultimately depends (Bennett, 1998; 2003). Even though regenerating or degraded forests and isolated trees may not provide all of the resources that a particular species need to survive, they may pose little resistance to the movement of many animals between patches and protected areas of forest where these resources are available (Bennett, 1998; 2003). In this context, an important priority for biodiversity conservation is to maintain a mosaic of semi-natural connected habitats within the agricultural land.

To investigate this issue, we measured dry forest connectivity in the study area (Fig. 1a) based on two different approaches, and we identified barriers to movement and priority actions for the region. The first approach focused on forest specialist species, i.e. species that have strict forest requirements. These species therefore require core areas for their survival over the long term. We focused on core areas of continuous forest cover larger than 5 ha. We developed a connectivity analysis based on distance matrices. Cores were considered neighbours if the smallest distance between their edges was less than 4 km (Figure 1b). The second approach focused on species that are less restricted in their forest habitat requirements, and can use isolated trees or small woodlots as well as core areas, and disperse easily through the matrix (e.g. some birds, insects, and many pioneer plant species). For these species, a highly fragmented landscape becomes more permeable to dispersion. In this case, we developed a connectivity analysis buffering away from any pixel classified as tree cover, using different buffer distances (100 m, 200 m, 300 m, 500 m, 1000 m, 5000 m, etc.). The map developed through this analysis shows the distance zones (proportion of area) between pixels classified as tree cover (Figure 1c). This analysis thus allowed the recognition of areas of decreasing permeability to movement.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (97)

71

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 1a

Figure 1b

Box 3.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (98)

72

Principles and Practice of Forest Landscape Restoration

Barriers to movement tended to coincide spatially using both approaches. Results suggest that core areas were generally not well connected, especially in the centre of the study area. Nevertheless, isolated trees and small patches may enhance connectivity considerably for mobile organisms. The distance between any form of tree cover was generally below 200 m. The lowest connectivity was found in the area around El Parral (pointed out in Fig. 1a). But even in this area, trees were still present (Fig. 2). Biodiversity conservation can be achieved by maintaining the diffuse mosaic of forest, open woodland and scattered trees, but also through restoration of habitats focused both on linking core areas and increasing permeability. The forest can be conserved by working with landholders in order to minimize human impacts in the remaining forest patches, many of which are greatly disturbed and degraded as a result of livestock grazing. Future actions for increasing connectivity and permeability should target the restoration of degraded pastures, the development of fuelwood plantations, and the expansion of living fences, shade and forage trees within the landscape. Finally, further actions should be focused on protecting and managing major links between conservation reserves to assist their long term viability.

Figure 2

Box 3.2 (cont.)

Figure 1c

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (99)

73

Assessing fragmentation and degradation of dryland forest ecosystems

Cropland in dry forest areas of Chile. Photo: C. Echeverria

Dry forest in central Veracruz, Mexico. Photo: C. Alvarez

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (100)

74

Principles and Practice of Forest Landscape Restoration

Techniques to quantify spatial patterns of forest cover

Analysis of forest fragmentation was conducted using the following set of selected landscape metrics: (a) patch area (ha), (b) proximity index, (c) patch density (n/100 ha), (d) total edge length (km), and (e) largest patch index (LPI, %). All of these metrics reflect the different effects of fragmentation on the spatial attribute of forest patches. Index proximity was calculated for a radius of 1 km and core area for an edge depth of 50 m. In addition, we estimated aggregation index and adjacency index between forest cover and the major land-cover types. It is expected that the aggregation of forest patches decreases as a result of fragmentation and increases with forest contiguity. Similarly, the adjacency between native forest and human-induced land-cover types should increase with changes in the matrix.

A minimum mapping unit of greater than 5 pixels was used for the spatial analyses. This enabled differences in data quality produced by the resampling of the MSS images to be min-imized. Map preparation was performed using ARC MAP (version 3.3; ESRI, 2009). Landscape metrics were computed by FRAGSTATS (version 3) (Mcgarigal et al., 2002) to compare the spatial patterns of forest cover for each time interval and study area.

Mapping spatial patterns of forest cover

Maps of forest cover based on patch size were generated for each study area and for each study year (Fig. 3.3). Most of these maps provide evidence of typical patterns of anthropogenic land-scape change. The patterns are comparable to those observed in many other parts of the world (Abdullah and Nakagoshi, 2006; Wang et al., 2010). Deforestation and fragmentation of dryland forests have occurred in most of the study areas, except in Bariloche where some forest frag-ments moved to upper classes of size during the study interval (Fig. 3.3c), and in Chiapas, where forest fragments did not appear to change in size over time (Fig. 3.3b). Maps of forest fragmentation showed a considerable increase in the number of smaller patches over time in Xalapa, Oaxaca, central Chile and Salta (Figs. 3.3a, d, e and f respectively).

Burned stand of Austrocedrus chilensis in Southern Argentina. Photo: J. Birch

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (101)

75

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 3.3

Figure 3a

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (102)

76

Principles and Practice of Forest Landscape Restoration

Figure 3c

Figure 3b

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (103)

77

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 3d

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (104)

78

Principles and Practice of Forest Landscape Restoration

Figure 3e

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (105)

79

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 3f

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (106)

80

Principles and Practice of Forest Landscape Restoration

Analysis of spatial patterns of dryland forests

Most of the study areas exhibited a decline in patch size between the earliest and the most recent maps. In particular, Veracruz, Oaxaca, central Chile and Salta showed a decline in patch size, while Chiapas remained practically constant and Bariloche exhibited an increase in this variable (Table 3.1). In Oaxaca and central Chile, the total edge length of forest patches increased and then declined over time (Table 3.1). In Veracruz and Salta there was an in-crease in the total of edges. In contrast, Bariloche was the only study area that exhibited a permanent decline in the number of fragment edges, whereas Chiapas did not demonstrate changes in this variable over time. With respect to the core areas of forest fragments, all of the study areas exhibited a decline in this index through time, except for Bariloche, which showed an increase (Table 3.1). The greatest declines occurred in central Chile and in Ver-acruz, where 66% and 51% of the core area was lost during the study periods respectively. In contrast, Chiapas did not present a substantial change in this index (1.3%), while in Bariloche the core area increased 16% (Table 3.1). Index of proximity (which provides a measure of the degree of isolation) decreased in Chiapas, central Chile and Salta (Table 3.1). In Veracruz and Oaxaca, this index varied during the study period, without providing a clear trend. In Bariloche this index presented an increase between 1973 and 1997 and then it declined dur-ing the last time interval.

These trends in landscape indices were associated with variation in patch density (Fig. 3.4). Owing to the fact that the number of patches may increase as a result of the creation of new patches by fragmentation, a further decline can be observed either by the loss of the new forest patches or the union of patches as a result of forest regeneration. This trend ena-bled different stages in the spatial dynamics of forest to be identified. In Veracruz and Salta, a gradual increment in patch density (Fig. 3.4) and edge length, and a decline in patch size and core areas (Table 3.1) characterized a landscape affected by progressive fragmentation during the study periods.

Oaxaca is the only study area where the patch density and edge length were curvilin-ear, with metrics changing direction at the half-way point of the study period (Fig. 3.4). This reflects a rapid division of forest patches that were later eliminated by high rates of deforestation. In central Chile, the number of patches gradually decreased owing to the conversion of forest patches (Fig. 3.4). This pattern was associated with a decrease in core area and an increase in patch isolation, and with a continuous loss of forest frag-ments over time (Table 3.1). In contrast to this situation, in Bariloche the increase in patch size, core area and proximity index and a decline in patch density and edge length showed that the forest cover was recovering, showing an opposite trend to forest frag-mentation (Fig. 3.4 and Table 3.1). In Chiapas, the slight decline in patch density (Fig. 3.4), and the almost constant values of patch size, edge length and core area, revealed a low level of forest fragmentation in this landscape and the stabilization of forest cover (Table 3.1).

During the study periods, the largest forest patch occupied no more than 2% of the whole landscape in each of Chiapas, central Chile and Bariloche. On the other hand, in Salta this index reached a higher value, ranging from 52% to 32% between 1977 and 2006. In Oaxaca values varied slightly from 24% to 23%, and in Veracruz, from 6% to 2%.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (107)

81

Assessing fragmentation and degradation of dryland forest ecosystems

Table 3.1

Veracruz, Mexico

Landscape indices 1973 1990 1999

Mean patch size (ha) 139.7 73.9 27.9

Total edge length 3,091,320 6,334,110 6,345,420

Total core area (ha) 44,160.84 41,404,14 21,771.7

Mean proximity 1,235.1 16,712.5 1,229.35

Oaxaca, Mexico

Landscape indices 1979 1989 1999 2005

Mean patch size (ha) 99.9 22.9 41.2 46.7

Total edge length 64,069.8 105,900.9 106,688.9 89,516.0

Total core area (ha) 514,323.3 2,246,893.7 386,258.9 428,649.66

Mean proximity 332,727.6 28,168.4 366,210.5 388,341.7

Chiapas, Mexico

Landscape indices 1990 2000 2005

Mean patch size (ha) 13.1 14.1 14.4

Total edge length 113,509.0 110,960.3 111,196.6

Total core area (ha) 277,821.0 275,821.0 274,287.0

Mean proximity 4,432.03 4,174.92 4,164.15

Central Chile

Landscape indices 1975 1985 1999 2008

Mean patch size (ha) 8.8 6.3 6.2 6.0

Total edge length 44,400.1 49,837.8 50,768.7 41,897.8

Total core area (ha) 76,901.2 29,922.8 23,500.2 26,149.2

Mean proximity 1,028.0 454.9 380.9 427.1

Salta, Argentina

Landscape indices 1977 1987 1993 2006

Mean patch size (ha) 1,074.7 757.9 528.6 330.2

Total edge length 8,540.1 15,826.2 15,455.0 14,872.7

Total core area (ha) 682,693.0 614,457.0 581,090.0 506,464.0

Mean proximity 299,131.0 522,803.0 217,447.0 174,601.0

Bariloche, Argentina

Landscape indices 1973 1985 1997 2003

Mean patch size (ha) 8.85 14.3 12.2 13.9

Total edge length 79,359.68 52,586.80 62,149.86 52,021.08

Total core area (ha) 115,080 135,654 145,918 132,901

Mean proximity 1854 2226 2336 2059

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (108)

82

Principles and Practice of Forest Landscape Restoration

Figure 3.4

As forest loss continues, it is expected that the largest patch index (LPI) will decline owing to a division of large patches by fragmentation (Trani and Giles, 1999). By graphing the forest loss versus LPI for the study areas, it was observed that in Oaxaca, Veracruz and Salta (Fig. 3.5) a continuous fragmentation has led to a division of large forest patches, causing a decline in the LPI. However, in central Chile and Bariloche, there was a slight increase in LPI as forest loss increased (Fig. 3.6). This opposite trend was the result of the union of large forest fragments despite the loss of others (Fig. 3.3). In Chiapas the LPI did not show variation, owing to the fact that forest area remained almost constant during the study period (Fig. 3.6).

Years

Central Chile

Pat

ch d

ensi

ty n

/ 100

ha

19730,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

�� �

1983 1993 2003

Years

Chiapas

Pat

ch d

ensi

ty n

/ 100

ha

19730,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1983 1993 2003

� �

Years

Bariloche

Pat

ch d

ensi

ty n

/ 100

ha

19730,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1983 1993 2003

Years

Salta

Pat

ch d

ensi

ty n

/ 100

ha

19730,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

� � �

1983 1993 2003

Years

Veracruz

Pat

ch d

ensi

ty n

/ 100

ha

19730,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1983 1993 2003

Years

Oaxaca

Pat

ch d

ensi

ty n

/ 100

ha

19730,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

��

1983 1993 2003

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (109)

83

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 3.5

Figure 3.6

Salta

(%) Forest Loss

Larg

est p

atch

inde

x

00 10 20 30 40

10

20

30

40

50�

��

Veracruz

(%) Forest Loss

Larg

est p

atch

inde

x

00 10 20 30 40

10

20

30

40

50

Oaxaca

(%) Forest Loss

Larg

est p

atch

inde

x

00 10 20 30 40

10

20

30

40

50

central Chile

(%) Forest Loss

Larg

est p

atch

inde

x

00 10 20 30 40

1

2

3

4

5

� ��

Chiapas

(%) Forest Loss

Larg

est p

atch

inde

x

00 0,5 1 1,5

1

2

3

4

5

��

Bariloche

(%) Forest Loss

Larg

est p

atch

inde

x

09 10 11 12

1

2

3

4

5

� �

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (110)

84

Principles and Practice of Forest Landscape Restoration

Results also showed different rates of forest disaggregation over time, illustrated by the degree of forest loss and fragmentation (Figs. 3.7 and 3.8). The greatest decline in forest ag-gregation was observed in central Chile, where this index decreased from 82% to 65% across the study period, most rapidly during the first time interval (Fig. 3.7). However, in the earliest study years, Salta and Veracruz exhibited the highest levels of forest aggregation or spatial integrity, with 99% and 96% respectively (Figs. 3.7 and 3.8). In central Chile the disaggrega-tion of forest cover was accompanied by a loss of forest patches rather than by a division of forest patches, as demonstrated by values of patch density (Fig. 3.4). On the other hand, in Veracruz and Salta the number of patches increased (Fig. 3.4) while patch size declined (Table 3.1), reflecting a gradual decline in the level of forest aggregation or an increase in forest fragmentation. Chiapas remained constant with 85% forest aggregation, indicating no change in spatial patterns (Fig. 3.7).

Oaxaca was the only study area that exhibited a decline and further increase in the aggregation index over time (Fig. 3.12). By comparing this result with the values gener-ated for the other metrics, it can be observed that during the first time interval, forest cover was disaggregated by the division of a large patch, which resulted in an increase in the number of patches (Fig. 3.8). Between 1989 and 2005, the forest became more aggregated, increasing the patch size (Table 3.1). In Bariloche, forest cover showed a gradual increase in aggregation (Fig. 3.12), indicating the recovery of new patches and an increase in patch size (Table 3.1)

Figure 3.7

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (111)

85

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 3.8

Changes in the spatial patterns of the dryland forests are explained largely by chang-es in neighbouring land-cover. Interestingly, in three study areas forest fragments were primarily surrounded by croplands while in the other three areas, the fragments were adjacent to degraded forest (in the case of Salta and central Chile) or to shrubland (in Bariloche) (Figs. 3.13 and 3.14). In most of the study areas the dryland forest fragments were surrounded by more than 70% croplands or by degraded forest and shrubland. This high percentage of adjacency to human-induced land uses indicates that most for-est edges may be subjected to anthropogenic activities that could potentially affect the survival of many species.

Oaxaca and Veracruz showed greater dynamics in the percentage of adjacency be-tween croplands and forest fragments (Fig. 3.9) than in study areas surrounded by de-graded forest or shrubland (Fig. 3.10). The greater variation in Oaxaca and Veracruz is related to changes in the matrix composition represented by a replacement of grass-land and bare ground by cropland, particularly during the 1970s and 1980s. Later in the 1990s, the adjacency to cropland tended to decline in Veracruz owing to the expansion and replacement of cropland by grassland. In Chiapas and in Oaxaca, more than 90% of forest fragments were by cropland during the last decade (Fig. 3.9). On the other hand, in all the South American study areas, forest patches were largely surrounded by degraded forest or shrubland (Fig. 3.10). In Salta a gradual decline in the adjacency between forest and degraded forest was observed after the mid–1980s. In central Chile, which recorded the greatest adjacency to degraded forest, a kind of ‘pseudo-savannah’ named espinal was reached by 2000, which then declined by 2008 (Fig. 3.10). This variation was related to the dynamics of the espinal that was converted to cropland, and which originated from the degradation of sclerophyllous forest.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (112)

86

Principles and Practice of Forest Landscape Restoration

Figure 3.9

Figure 3.10

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (113)

87

Assessing fragmentation and degradation of dryland forest ecosystems

Landscape states and habitat destruction

McIntyre and Hobbs (1999) discuss the process of habitat destruction and habitat modifica-tion, which can be conceptualized as a continuum, associated with the influence of human dis-turbance. They identified four types of landscape states along a gradient of destruction: intact (<10% habitat destroyed), variegated (10–40%), fragmented (40–90%) and relictual (>90%). In variegated landscapes, the habitat still forms the matrix, whereas in fragmented landscapes, the matrix is composed of destroyed habitat. During the study period, the study areas exhibited different degrees of forest loss (Fig. 3.3 and Table 3.2). In Veracruz, the percentage of remaining forest cover (see Table 2.1 in Chapter 2 on recent loss) was 9% in 2000, which corresponds to a relictual landscape. In this area the effect of intensive agricultural development has led to a progressive fragmentation of forest habitat and to a high rate of forest loss (Fig. 3.5).

On the other hand, the percentages of forest cover detected were 16% in Bariloche, 32% in Chiapas, 35% in central Chile and 60% in Oaxaca. These study areas correspond to frag-mented landscapes characterized by more than 40% of habitat destroyed. In Oaxaca and central Chile, most of the forest fragments are under high pressure from an intensively used matrix, which has led to a progressive transformation towards agriculture and degraded for-est respectively (Fig. 3.10 and Table 3.2). Central Chile suffered the largest reduction of forest habitat from 43% to 34% of forest cover and has been severely fragmented (Fig. 3.12). In con-trast, in Bariloche and Chiapas the native forest fragments remain relatively unmodified and the forest persisted and even defragmented after long periods of grazing by livestock, expan-sion of crops and fires. Salta was found to be in an intact state in 1977 (94% forest cover) and then changed to a variegated state in 2006 (73%). In this study area, the dryland forest still forms the landscape matrix and is represented by large patches (Fig. 3.9). However, forestry operations in these forests impose slight but continuous changes in the spatial patterns of forest cover, reflected by a reduction in the degree of aggregation (Fig. 3.7). The identifica-tion of the level of modification is an important consideration for management planning, as this can assist in deciding where and when to allocate greater and lesser protection to the landscape (Hobbs, 2002).

Trends in spatial patterns in the dryland forest in Latin America

The analysis of the landscape indices enabled the spatial patterns of dryland forests in Latin America to be assessed. Our results show that the spatial patterns of change of dryland for-est were dynamic and did not necessarily represent a unidirectional process of forest loss and fragmentation (Table 3.2). This is consistent with what has been described for some other landscapes, emphasizing that there is no single correct way in which to think about spatial patterns in modified landscapes (Lindenmayer and Fischer, 2006). By recognizing the diversity of landscape trends, conservation strategies could be better focused (McIntyre and Hobbs, 1999).

In central Chile, the spatial patterns are related to a unidirectional landscape alteration, with continued alteration assumed to reduce both the size of individual patches (known as shrink-age) and the overall number of patches (known as attrition) (Forman, 1995b) (Table 3.1). Ve-racruz and Salta also showed a unidirectional alteration, but with an opposite trend in the number of patches that increased over time, defined as fragmentation (Forman, 1995a) (Table 3.2). On the other hand, Oaxaca showed a bidirectional alteration characterized by a rapid

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (114)

88

Principles and Practice of Forest Landscape Restoration

fragmentation in the earliest years followed by a loss of forest patches in the last time interval (Table 3.2). This trend was more evident owing to a rapid increase in the number of patches during the first time interval (Table 3.1). However, there are many cases where trends in land-scape change have been reversed (Metcalfe and Bradford, 2008; Vellend, 2003; Wittenberg et al., 2007). Bariloche showed an increase in patch size and increase in patch proximity over time (Table 3.3). In this area, the spatial patterns changed during the study period because of forest regeneration in logged areas. Chiapas showed a more stable pattern of change, increas-ing slightly the size of forest fragments and reducing the number of patches (Table 3.1).

Our results demonstrate that the spatial patterns of dryland forests were highly dynamic over the last four decades. While most of the study areas experienced a reduction of forest habitats others showed an increase or stability in forest cover. Understanding the trends in spatial patterns of dryland forest is important for the conservation of its biodiversity and the provision of diverse ecosystem services. Despite this importance, many forest assessments and international initiatives still focus on the extent of forest loss without concern for its spatial pattern (Kupfer, 2006). This work confirms further the advantages of using landscape metrics to describe pattern change, as has been demonstrated in other parts of the world (Bhattarai et al., 2009; Cayuela et al., 2006; Martínez et al., 2009; Peng et al., 2010; Trani and Giles, 1999; Zeng and Wu, 2005).

Table 3.2

Study area Description Trend in spatial patterns Landscape state

Salta

Veracruz

Division of large forest patches, increasing number of patches, decrease in forest aggregation

Progressive fragmentation Intact to variegated

Relictual

Oaxaca Loss of forest cover; decrease and increase in forest aggregation; substantial changes in matrix composition

Fragmentation followed by deforestation

Fragmented

Central Chile Loss of forest patches; and forest continuity

Progressive deforestation Fragmented

Chiapas No spatial changes in forest cover

Forest persistence Fragmented

Bariloche Union of forest patches; increase in forest aggregation

Forest patch coalescence Fragmented

Mapping forest degradation

Human disturbances not only alter the spatial patterns of forest cover but can also lead to a modification or degradation of the remaining habitat (McIntyre and Hobbs, 1999). Modi-fications include changes to the structure, biotic composition, or ecosystem functioning of habitat (McIntyre and Hobbs, 1999; Ravi et al., 2010). Degradation of dryland forest habitat is associated with diverse human activities such as livestock grazing, tree harvesting and changed fire regimes (Reynolds et al., 2007).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (115)

89

Assessing fragmentation and degradation of dryland forest ecosystems

MODIS data were used to map forest degradation in central Chile between 2002 and 2009 (Box 3.3). Owing to the low spatial resolution of MODIS images (250 m), a threshold of 3m2/m2 in Leaf Area Index (LAI) was used to select only dense forests. This enabled other land-cover types such as pasture or shrubland to be excluded in order to monitor the degradation of forest cover. Then, changes in red and infrared band values extracted from MOD 13 Q1 be-tween 2002 and 2009 were analyzed to detect pixels with a degree of degradation. Essential-ly, it was assumed that dense forest pixels with an increase in reflectance values for the red band and a decline in the infrared band over time correspond to pixels affected by degrada-tion. Pixels without changes in reflectance values for both bands have not been affected by degradation. Degraded forest pixels were selected to determine the degree of degradation. This was conducted applying the NDVI (Normalized Difference Vegetation Index) for which three levels of degradation were defined: NDVI>0.71: low degradation; 0.57<NDVI<0.71: intermediate degradation; 0.57<NDVI: high degradation. These levels were validated in the field. Finally, an assessment of forests degraded during each time interval (2002–2005 and 2005–2009) was obtained by overlapping the corresponding binary maps of degraded for-est/non-degraded forest cover in a Geographic Information System (GIS).

A logistic regression modelling approach was used to determine the immediate drivers of degradation. Most of the environmental and socio-economic explanatory variables that are potentially related to forest degradation were identified and mapped. These are: property size, slope, elevation, distance to roads, distance to rivers, distance to towns and distance to agricultural land in the earliest image.

In Chile, the forests degraded between each time interval (2002–2005 and 2005–2009) were identified by overlapping the corresponding binary maps of degraded forest/non-de-graded forest cover in a GIS. In Argentina, degradation analysis was conducted for eroded pre-montane forest for each of the following time intervals: 1977–1987, 1987–1983, 1983–2006 and 1977–2006. The binary response variable, degraded habitat vs. non-degraded habitat, was analyzed using a logistic regression model in the statistical package R (Echeverria et al., 2008). In Salta, Argentina, degraded forest was derived from the Landsat image classification. This corresponded to forest areas with less than 50% of tree cover classified as eroded pre-montane forest with limiting edaphic factors. This included two types of savannah: (a) with edaphic restrictions and fire controlled, dominated by tusca blanca (Acacia albicorticata) and urundel (Astronium urundeuva), (b) tusca blanca and pasto cubano savannah, small patches of secondary forest (previously used for agriculture) and low stature riparian forests. The following were also included: eroded premontane forest, secondary forest and riparian forest. Degradation analysis was conducted for eroded premontane forest for each of the fol-lowing time intervals: 1977–1987, 1987–1983, 1983–2006 and 1977–2006.

A total of 27,831 ha, equivalent to 28% of the total dense dryland forest in 2002, was de-graded by 2009 in the Chilean study area (Fig. 3.11). Regarding the degree of degradation, the results demonstrated that the proportion of highly degraded forest increased in the second time interval. In 2005, 64% of the degraded forests were categorized as highly degraded, 23% as a moderately degraded and 14% as slightly degraded (Figs. 3.11 and 3.12). In 2009, 74% of the degraded forest was highly degraded, 24% moderately degraded and 2% slightly degraded (Figs. 3.11 and 3.12). This increase in the area of highly degraded forest can be explained by the fact that local people continued logging the dense forest for firewood and other forest products. These processes have caused a modification of the structure and composition of the remaining forests.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (116)

90

Principles and Practice of Forest Landscape Restoration

In Salta, 4.8% of the forest cover was degraded in 1977 (Fig. 3.13). This value remained almost constant during the following years, reaching 4.6% in 2006 (Fig. 3.14). A higher pro-portion of degraded forest occurred in 1987 with 5.3% of the forest cover. Between 1977 and 1987, forest degradation was mainly associated with livestock grazing, and forest logging for firewood and timber. At the beginning of the 1990s the rapid conversion of degraded forest to soya crops led to a decline in the area of degraded forest. During the last decade, the pro-portion of degraded forest remained almost constant, which reflects the current presence of degrading activities such as forest logging and browsing by livestock.

Figure 3.12

Figure 3.11

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (117)

91

Assessing fragmentation and degradation of dryland forest ecosystems

Figure 3.13

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (118)

92

Principles and Practice of Forest Landscape Restoration

Figure 3.14

Box 3.3 Estimating forest degradation in dryland landscapes in central Chile using MODIS products

D. González, R. Fuentes, C. Echeverría

Forest degradation may occur owing to natural (forest fire, earthquakes, volcanism, etc.) and anthropogenic perturbations (urban and agriculture surface expansion, forest use, etc.) (Stuart et al., 2002; Pickett and White, 1985; Hüttl and Schneider, 1998). Nutritional disturbances can also lead to decreasing stand stability and productivity (Stolpe et al., 2008). Although dry forests are being subjected to a range of different disturbances, a major factor responsible for their loss and degradation is the recent expansion of industrial agriculture, resulting from increasing global food demand (Grau et al., 2009).

Remote sensing imagery becomes a powerful tool to evaluate the threats to forest ecosystems (Luque, 2000; Armenteras et al., 2003; Echeverría et al., 2007). The research described here focused on quantifying the degradation of a dryland forest in an area of 1,250,000 ha using MODIS satellite products. The study area is located in one of the most populated regions of dryland forest in Chile, between 33º and 38º S latitude, located between the Central Valley and Coastal Range (Fig. 1).

Using the MOD15A2 product at 1000 m spatial resolution, for three years (2002 (t0), 2005 (t1) and 2009 (t2)), we selected those forest patches whose pixels were equal to or greater than 3 m2/m2 of leaf area index (LAI). This threshold enabled dense forest patches that may exhibit degradation to be distinguished. Those patches in which disturbances may have caused a removal of the forest cover were discarded as they represented deforestation instead of degradation.

Further, we applied the Near Infrared (NIR) and Red (R) reflectance responses from MOD13Q1 products at the pixel level to quantify the degradation over time. In a given forest pixel, when the NIR reflectance increases and the R reflectance decreases between two measurements (at the same daily time), the forest is increasing its canopy cover and, therefore, is becoming more dense. On the other hand, when the NIR reflectance goes down and R reflectance goes up, the forest is decreasing its canopy cover, which indicates that the forest patch is being degraded through time (Chuvieco, 1996). These responses in reflectances were modelled in ARC GIS applying a decision tree procedure (Fig. 2).

To visualize the levels of forest degradation, we related the Red and NIR behaviour with Normalized Difference Vegetation Index (NDVI) (Eq. 1). This was conducted applying the NDVI where three levels of degradation were defined: NDVI>0.71: low degradation; 0.57<NDVI<0.71: intermediate degradation; 0.57<NDVI: high degradation. All these levels were validated in the field.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (119)

93

Assessing fragmentation and degradation of dryland forest ecosystems

Eq. 1. NDVI=(�NIR-�Red) / (�NIR+�Red), where:

�NIR: Near Infrared reflectance

�Red: Red reflectance

Figure 1

Figure 2

A logistic regression modelling approach was applied to determine the immediate drivers of degradation. Most of the environmental and socioeconomic explanatory variables that are potentially related to forest degradation were identified and mapped. These include: property size, distance to roads, distance to rivers, distance to towns, distance to agricultural land in the earliest image and slope and elevation. Our results revealed that in 2005, 64% of the degraded forest surface was categorized as highly degraded, 23% as a moderately degraded and 14% as slightly degraded. In 2009, 74% of the degraded forest was highly degraded, 24% moderately degraded and 2% slightly degraded.

Box 3.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (120)

94

Principles and Practice of Forest Landscape Restoration

Table 1

Municipality Degradation (%)

Municipality Degradation (%)

Municipality Degradation (%)

Municipality Degradation (%)

Olmué 11.8 Navidad 25.6 Doñihue 28.7 Graneros 33.7

Limache 15.3 Coinco 26.7 San Antonio 29.8 Melipilla 34.6

Quilpue 19.4 Alhué 26.8 Buín 30.1 San Pedro 37.9

Casablanca 20.6 Valparaíso 27.8 Talagante 33.0 María Pinto 38.1

Santo Domingo

23.2 Paine 28.1 Curacaví 33.2 El Monte 42.5

Litueche 24.4 Rancagua 28.2 Las Cabras 33.5 Cartagena 46.9

The municipality with highest forest degradation percentage was Cartagena with 47% (Table

1). The municipalities that showed the lowest forest degradation percentages were Olmué (12 %), and Limache (15%) (Table 1). The multivariate logistic regression model, used to identify the main drivers of the landscape change process, indicated that the probability of an area being degraded is highly significant and positively related to the distance to streams. In contrast, the distance to urban areas and the distance to agricultural land in 2008 were negatively related to forest degradation. These results showed that the probability of degradation increases in forests located near urban areas, near agricultural land and far away from streams. In these areas, forest logging for fuelwood and browsing by cattle in dense forest are more intense causing a decline in canopy cover and tree density. Human access to dense forest appeared to be one of the main drivers of forest degradation.

The temporal analysis of the forest canopy changes based on NIR and red reflectance behaviour appears to be a suitable procedure to evaluate the degradation of dense dry forests. The main limitation of this procedure relates to the spatial resolution of the MODIS products and the size of dense forest areas.

Box 3.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (121)

95

Assessing fragmentation and degradation of dryland forest ecosystems

Causes of forest degradation

In central Chile, the multivariate logistic regression models indicated that the probability of an area being degraded is highly significant and positively correlated with distance to streams (p<0.001; Table 3.2). In contrast, distance to urban areas and distance to agricultural land in 2008 were negatively related to forest degradation (p<0.01; Table 3.2). These results showed that the probability of degradation increases in forests located near urban areas, agricultural land and far away from streams. In these areas, forest logging for fuelwood and browsing by cattle in dense forest are more intense, causing a decline in canopy cover and tree density. Human access to dense forest appeared to be one of the main drivers of forest degradation.

Similarly, the probability of degradation in Salta was positively related to distance to urban areas in all the time intervals and for the whole study period (p<0.001; Table 3.2). Eleva-tion also was positively related to forest degradation in all of the study periods. Distance to streams was marginally significant in the first and the third time interval as well as during the overall study period. This is because most of streams are located in areas that are less acces-sible for people. The probability of forest degradation was positively explained by distance to villages in the first time interval as well as during the whole study period (1977–2006). Before the 1990s, the distance to secondary roads was significantly related to forest degrada-tion. Since the 1990s this variable has not been significant as most of the degraded forests near secondary roads were converted to agricultural land. During the 1990s, the distance to agricultural land was associated with the presence of degraded forests owing to the use of fire to expand the agricultural frontier.

In both study areas, the results revealed that accessibility to forest areas is one of the main drivers of forest degradation. The probability of a forest area being degraded is higher when a forest is located near urban and agricultural lands, and in lowlands. This trend reflects that the remaining dryland forests in central Chile and in Salta have been undergoing continuous degradation over recent decades (Box 3.4). This is consistent with the results obtained by recent assessments (MEA, 2005; Ravi et al. 2010), which demonstrate that dryland ecosys-tems around the world are undergoing rapid land degradation as a result of anthropogenic disturbances. Diverse studies emphasize that modifications of dryland forest habitats may lead to changes in ecosystem processes (Jafari et al., 2008; Smet and Ward, 2006; Stolpe et al., 2008), which may affect the productivity of the landscape, with important environmental and socioeconomic implications.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (122)

96

Principles and Practice of Forest Landscape Restoration

Box 3.4 Human-caused forest fires in Mediterranean ecosystems of Chile: modelling landscape spatial patterns of forest fire occurrence

A. Altamirano, C. Salas, V. Yaitul, A. Miranda, C. Smith-Ramírez

Fire disturbance is recognized as an important problem because it can devastate natural resources and human property, and threaten human lives. Forest fires result in enormous economic losses because they affect environmental, recreational, and amenity values as well as consume timber, degrade real estate, and generate high costs of suppression. Modelling forest fire occurrence (i.e., where and when a forest fire starts) has recently been conducted in the northern hemisphere (Calef et al., 2008; Lozano et al., 2007; Ryu et al., 2007; Vega-Garcia and Chuvieco, 2006), however, efforts on the subject are lacking for the southern hemisphere, in particular for Chilean ecosystems. Some studies in Chile have focused on post-fire effects on vegetation dynamics (Navarro et al., 2008; Litton and Santelices, 2003), but studies on predicting forest fires occurrence are lacking. In Chile it has been reported that fire can encourage exotic plant invasions (see Chapter 8) and cause significant losses of local biodiversity. Forest fire occurrence has increased in recent years in Chile, with a mean frequency of about five thousand forest fires per year. These fires have affected a mean area of about 500 km2 per year (Navarro et al., 2008; CONAF, 2009), human activity being the main cause of fire ignition (CONAF, 2009). The extensive fires produced by human activity in central Chile need to be addressed in order for forest restoration approaches to be effective.

It is important to understand the impact of processes such as fire ignition and spread on landscape patterns in order for land management practices to be effective (Foster et al., 1997). At the landscape scale (i.e. extents >100,000 ha), the probability of a large fire is associated with multiple factors including: forest type, physiographic characteristics, climate, and human activities. In this study we developed models to investigate the relationship between forest fire occurrence and landscape heterogeneity spatial patterns in Mediterranean ecosystems of Chile. The study area extends over 892 km2 and is located in eastern central Chile covering parts of the Valparaíso and Metropolitan administrative regions (Fig. 1). We selected a landscape with temporal stability in composition operating at the landscape scale. We used georeferenced forest fire data from a 5-year period of fire occurrence from 2004 to 2008. A distance of 25 x 25 pixels (750 m) was used to compute the co-occurrence matrices, since small windows result in very sparse matrices. Our data on landscape spatial patterns were obtained at multiple spatial scales, including climatic, topographic, human-related, and land-cover variables from satellite imagery. We fit a logistic model in order to predict forest fire occurrence as a function of our potential predictor variables. The relationship modeled was that between the binary response variable (one = burned, zero = not burned) and the predictor variables. In order analyze forest fire occurrence we produced categorized maps of the predicted forest fire occurrence probability into four levels: very high (0.75 ≤ p ≤ 1), high (0.5 ≤ p ≤ 0.75), low (0.25 ≤ p ≤ 0.5) and low (0 ≤ p ≤ 0.25).

Our best model suggests that the probability of forest fire occurrence is related to both high temperature and precipitation, and lower distance to cities. Our predictions suggest that 46% (410 km2) of the study area has high probability of forest fire occurrence, being concentrated in the eastern locations of the study area (Fig. 1). Our model correctly classified about 73% of our validation dataset. The information from this study may be useful for hazard reduction, indicating risk of forest fire occurrence (Ryu et al., 2007; Vega-Garcia and Chuvieco, 2006). The study area is one of the most populated regions of Chile. Therefore, our findings can be used to inform decision making regarding land and urban planning. If climate determines patterns of forest fire occurrence, then when the climatic variables change, forest fire occurrence may also change. This might have important consequences for long-term land and urban planning, since prioritization of areas with high probability of forest fire occurrence today might not be effective in the face of climate change. Exploring a new statistical model approach would allow to improve the predictive capability of the models. Therefore, part of our future research will focus on this subject.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (123)

97

Assessing fragmentation and degradation of dryland forest ecosystems

Conclusions

Our results indicate that dryland forests exhibit a progressive fragmentation and degradation in most of the Latin American landscapes studied during the research. In central Chile and in Salta, dryland forests have been simultaneously affected by forest loss (Chapter 2), fragmenta-tion and degradation. In Veracruz and Oaxaca, the landscape has experienced a continuous fragmentation and loss of forest habitats. On the other hand, Chiapas and Bariloche show different trends, towards forest persistence and coalescence respectively. Results presented here clearly show that dryland forest is under considerable human pressure from economic development and imply policy challenges for the countries involved. Owing to the impor-tance of dryland forest for providing different ecosystem services for human well-being, diverse actions should be undertaken to minimize or reverse the human impacts of fragmen-tation and degradation on dryland forests. Ecological restoration actions have the potential to address both the fragmentation and degradation of forest that has been documented here in multiple study areas. Such interventions should be planned and implemented at the landscape scale, to ensure they are effective in increasing connectivity among forest patches. Recent advances emphasize the development of integrative approaches to counter land deg-radation, poverty, safeguard biodiversity and protect the culture of the 2.5 billion people who live in dryland systems (Reynolds et al., 2007). Forest landscape restoration actions should constitute an element of such approaches. Urgent and comprehensive reframing of rural development strategies in Latin America should be undertaken to achieve this goal.

Figure 1

Box 3.4 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (124)

98

Principles and Practice of Forest Landscape Restoration

References

Abdullah, S.A., Nakagoshi, N. 2006. Changes in landscape spatial pattern in the highly devel-oping state of Selangor, peninsular Malaysia. Landscape and Urban Planning 77: 263–275.

Altieri, M.A., Masera, O. 1993. Sustainable rural development in Latin America: building from the bottom-up. Ecological Economics 7: 93–121.

Armenteras, D., Gast, F., Villareal, H. 2003. Andean forest fragmentation and the representative-ness of protected natural areas in the eastern Andes, Colombia. Biological Conservation 113: 245–256.

Baillie, J.E.M., Hilton-Taylor, C., Stuart, S.N. 2004. IUCN Red List of threatened species. A global species assessment. IUCN, Gland, Switzerland and Cambridge, UK.

Baptista, S.R. 2010. Metropolitan land-change science: A framework for research on tropical and subtropical forest recovery in city-regions. Land Use Policy 27: 139–147.

Bennett, A.F. 2003. Linkages in the Landscape: The role of corridors and connectivity in wild-life conservation. IUCN, Gland, Switzerland and Cambridge, UK. Xiv+254pp.

Bhattarai, K., Conway, D., Yousef, M. 2009. Determinants of deforestation in Nepal’s Central Development Region. Journal of Environmental Management 91: 471–488.

Calef, M.P., McGuire, A.D., Chapin, F.S. 2008. Human Influences on Wildfire in Alaska from 1988 through 2005: An Analysis of the Spatial Patterns of Human Impacts. Earth Interac-tions 12 (1): 1–17.

Carvalho, F.M.V., De Marco Júnior, P., Ferreira, L.G. 2009. The Cerrado into-pieces: Habitat fragmentation as a function of landscape use in the savannas of central Brazil. Biological Conservation 142: 1392–1403.

Cayuela, L. 2009. Fragmentation. In: Gillespie, R., Clague, D. (eds.), Encyclopedia of Islands. University of California Press, California: pp. 328–330.

Cayuela, L., Benayas, J.M.R., Echeverría, C. 2006. Clearance and fragmentation of tropical mon-tane forests in the Highlands of Chiapas, Mexico (1975–2000). Forest Ecology and Manage-ment 226: 208–218.

Chuvieco, E. 1996. Fundamentos de teledetección espacial. Ediciones RIALP, S.A., Third ed., Madrid, Spain. 568pp.

CONAF. 2009. Corporación Nacional Forestal. Recursos Forestales. Protección contra incen-dios forestales. Consultado 9 Jun. 2008. <http://www.conaf.cl>.

Echeverría, C., Newton, A., Lara, A., Rey-Benayas, J.M., Coomes, D. 2007. Impacts of forest frag-mentation on species composition and forest structure in the temperate landscape of southern Chile. Global Ecology and Biogeography 16: 426–439.

Echeverria, C., Coomes, D., Salas, J., Rey-Benayas, J.M., Lara, A., Newton, A. 2006. Rapid deforesta-tion and fragmentation of Chilean temperate forests. Biological Conservation 130: 481–494.

Echeverria, C., Coomes, D.A., Hall, M., Newton, A.C. 2008. Spatially explicit models to analyze forest loss and fragmentation between 1976 and 2020 in southern Chile. Ecological Mod-elling 212: 439–449.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (125)

99

Assessing fragmentation and degradation of dryland forest ecosystems

Echeverría, C., Newton, A.C., Lara, A., Benayas, J.M.R., Coomes, D.A. 2007. Impacts of forest fragmentation on species composition and forest structure in the temperate landscape of southern Chile. Global Ecology and Biogeography 16: 426–439.

Fialkowski, M., Bitner, A. 2008. Universal rules for fragmentation of land by humans. Land-scape Ecology 23: 1013–1022.

Forman, R.T.T. 1995a. Land Mosaics. The ecology of landscapes and regions. Cambridge Uni-versity Press, New York.

Forman, R.T.T. 1995b. Some general principles of landscape and regional ecology. Landscape Ecology 10: 133–142.

Forman, R.T.T., Godron, M. 1986. Landscape Ecology. John Wiley and Sons, New York, NY.

Foster, D.R., Aber, J.D., Melillo, J.M., Bowden, R.D., Bazzaz, F.A. 1997. Forest response to distur-bance and anthropogenic stress. Bioscience 47: 437–445.

Fuentes, E.R., Hajek, E. 1979. Patterns of landscape modifications in relation to agricultural practice in central Chile. Environmental Conservation 6: 265–271.

Fuentes, E.R., Hoffmann, A., Poiani. A., Alliende, M.C. 1986. Vegetation change in large clear-ings: patterns in the Chilean matorral. Oecologia 68: 358–366.

Fuentes E.R., Otaiza, R.D., Alliende, M.C., Hoffmann, A., Poiani, A. 1984. Shrub clumps of the Chilean matorral vegetation: structure and possible maintenance mechanisms. Oecologia 62: 405–411.

Geri, F., Amici, V., Rocchini, D. 2009. Human activity impact on the heterogeneity of a Mediter-ranean landscape. Applied Geography. In Press.

Giriraj, A., Murthy, M.S.R., Beierkuhnlein, C. 2010. Evaluating forest fragmentation and its tree community composition in the tropical rain forest of Southern Western Ghats (India) from 1973 to 2004. Springer, Heidelberg.

Grau, H.R., Gasparri, N.I., Aide, T.H. 2009. Agriculture expansion and deforestation in season-ally dry forests of north-west Argentina. Environmental Conservation 32(2): 140–148.

Hill, J., Stellmes, M., Udelhoven, T., R der, A., Sommer, S. 2008. Mediterranean desertification and land degradation: Mapping related land use change syndromes based on satellite ob-servations. Global and Planetary Change 64: 146–157.

Hobbs, R. 2002. Habitat networks and biological conservation. In: Gutzwiller, K., (ed.), Appyling landscape ecology in biological conservation. Springer, New York, USA: pp. 150–170.

Holmgren, M. 2002. Exotic herbivores as drivers of plant invasion and switch to ecosystem alternative states. Biological Invasions 4: 25–33.

Holmgren, M., Segura, A.M., Fuentes, E.R. 2000. Limiting mechanisms in the regeneration of the Chilean matorral: Experiments on seedling establishment in burned and cleared mesic sites. Plant Ecology 147: 49–57.

Hüttl, R.F., Schneider, B.U. 1998. Forest ecosystem degradation and rehabilitation. Ecological Engineering 10: 19–31.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (126)

100

Principles and Practice of Forest Landscape Restoration

Jafari, R., Lewis, M.M., Ostendorf, B. 2008. An image-based diversity index for assessing land degradation in an arid environment in South Australia. Journal of Arid Environments 72: 1282–1293.

Kupfer, J.A. 2006. National assessments of forest fragmentation in the US. Global Environmen-tal Change 16: 73–82.

Lindenmayer, D.B., Fischer, J. 2006. Habitat fragmentation and landscape change. An ecologi-cal and conservation synthesis. Island Press, USA.

Lozano, F.J., Suárez-Seoane, S., de Luis, E. 2007. Assessment of several spectral indices derived from multi-temporal Landsat data for fire occurrence probability modelling. Remote Sens-ing of Environment 107: 533–544.

Luque, S. 2000. Evaluating temporal changes using Multi-spectral Scanner and Thematic Map-per data on the landscape of a natural reserve: the New Jersey pine barrens, a case study. International Journal of Remote Sensing 21: 2589–2611.

Martínez, M.L., Pérez-Maqueo, O., Vazquez, G., Castillo-Campos, G., Garcìa-Franco, J., Mehl-treter, K., Equihua, M., Landgrave, R. 2009. Effects of land use change on biodiversity and ecosystem services in tropical montane cloud forests of Mexico. Forest Ecology and Man-agement 258: 1856–1863.

McGarigal, K., Cushman, S.A., M.C., N., Ene, E. 2002. Fragstats: spatial pattern analysis program for categorical maps. Retrieved January 20, 2009. Landcape Ecology Program web site: <www.unmass.edu/landeco/researach/fragstats/fragstat.html>

McIntyre, S., Hobbs, R. 1999. A framework for conceptualizing human effects on landscapes and its relevance to management and research models. Conservation Biology 13: 1282–1292.

Metcalfe, D.J., Bradford, M.G. 2008. Rain forest recovery from dieback, Queensland, Australia. Forest Ecology and Management 256: 2073–2077.

Millennium Ecosystem Assessment (MEA). 2005. Ecosystems and human well-Being: Deserti-fication synthesis. World Resources Institute, Washington DC.

Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A., Kent, J. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853–858.

Nagendra, H., Munroe, D.K., Southworth, J. 2004. From pattern to process: landscape frag-mentation and the analysis of land use/land-cover change. Agriculture, Ecosystems and Environment 101: 111–115.

Navarro, R.M., Hayas, A., García-Ferrer, A., Hernández, R., Duhalde, P., González, L. 2008. Cara-cterización de la situación posincendio en el área afectada por el incendio de 2005 en el Parque Nacional de Torres del Paine (Chile) a partir de imágenes multiespectrales. Revista Chilena de Historia Natural 81: 95–110.

Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V.N., Under-wood, E.C., D’amico, J.A., Itoua, I., Strand, H.E., Morrison, J.C., Loucks, C.J., Allnutt, T.F., Ricketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W.W., Hedao, P., Kassem, K.R. 2001. Terrestrial ecoregions of the world: a new map of life on Earth. BioScience 51(11): 933–938.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (127)

101

Assessing fragmentation and degradation of dryland forest ecosystems

Peng, J., Wang, Y., Zhang, Y., Wu, J., Li, W., Li, Y. 2010. Evaluating the effectiveness of landscape metrics in quantifying spatial patterns. Ecological Indicators 10: 217–223.

Pickett, S., White, P. 1985. The ecology of natural disturbance and patch dynamics. First ed. Academic Press, San Diego, USA. 472pp.

Ravi, S., Breshears, D.D., Huxman, T.E., D’Odorico, P. 2010. Land degradation in drylands: In-teractions among hydrologic-aeolian erosion and vegetation dynamics. Geomorphology 116: 236–245.

Reynolds, J.F., Smith, D.M.S., Lambin, E.F., Turner, B.L., II, Mortimore, M., Batterbury, S.P.J., Downing, T.E., Dowlatabadi, H., Fernandez, R.J., Herrick, J.E., Huber-Sannwald, E., Jiang, H., Leemans, R., Lynam, T., Maestre, F.T., Ayarza, M., Walker, B. 2007. Global desertification: build-ing a science for dryland development. Science 316: 847–851.

Rotenberg, E., Yakir, D. 2010. Contribution of semi-arid forests to the climate system. Science 327: 451–454.

Ryu, S., Chen,J., Zheng, D., Lacroix, J.J. 2007. Relating surface fire spread to landscape struc-ture: An application of FARSITE in a managed forest landscape. Landscape and Urban Plan-ning 83: 275–283.

Sano, M., Miyamoto, A., Furuya, N., Kogi, K., 2009. Using landscape metrics and topographic analysis to examine forest management in a mixed forest, Hokkaido, Japan: Guidelines for management interventions and evaluation of cover changes. Forest Ecology and Manage-ment 257: 1208–1218.

Schimel, D.S. 2010. Drylands in the Earth system. Science 327: 418–419.

Shuangcheng, L., Qing, C., Jian, P., Yanglin, W. 2009. Indicating landscape fragmentation using L-Z complexity. Ecological Indicators 9: 780–790.

Smet, M., Ward, D. 2006. Soil quality gradients around water-points under different manage-ment systems in a semi-arid savanna, South Africa. Journal of Arid Environments 64: 251–269.

Stolpe, N., Munoz, C., Zagal E., Ovalle, C. 2008. Modelling soil carbon storage in the “Espinal” agroecosystem of central Chile. Arid Land Research and Management 22: 148–158.

Stuart, F.C., Matson, P.A., Mooney A.H. 2002. Principles of terrestrial ecosystem ecology. Springer, New York, USA.

Trani, M.K., Giles, J.R.H. 1999. An analysis of deforestation: metrics used to describe pattern change. Forest Ecology and Management 114: 459–470.

Vega-García, C., Chuvieco, E. 2006. Applying local measures of spatial heterogeneity to Land-sat-TM images for predicting wildfire occurrence in Mediterranean landscapes. Landscape Ecology 21: 595–605.

Vellend, M. 2003. Habitat loss inhibits recovery of plant diversity as forests regrow. Ecology 84: 1158–1164.

Vergara, P.M., Armesto, J.J. 2009. Responses of Chilean forest birds to anthropogenic habitat fragmentation across spatial scales. Landscape Ecology 24: 25–38.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (128)

102

Principles and Practice of Forest Landscape Restoration

Wang, S.Y., Liu, J.S., Ma, T.B. 2010. Dynamics and changes in spatial patterns of land use in Yel-low River Basin, China. Land Use Policy 27: 313–323.

Wittenberg, L., Malkinson, D., Beeri, O., Halutzy, A., Tesler, N. 2007. Spatial and temporal pat-terns of vegetation recovery following sequences of forest fires in a Mediterranean land-scape, Mt. Carmel Israel. CATENA 71: 76–83.

Zeng, H., Wu, X.B. 2005. Utilities of edge-based metrics for studying landscape fragmentation. Computers, Environment and Urban Systems 29: 159–178.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (129)

103

4 FRAGMENTATION AND ALTITUDINAL

EFFECTS ON TREE DIVERSITY IN

SEASONALLY DRY FORESTS OF

MEXICO AND CHILE

C. Smith-Ramírez, G. Williams-Linera, R. F. del Castillo, N. Ramírez-Marcial, R. Aguilar, N. Taylor-Aquino, D. Golicher, P. Becerra, C. Echeverría, J.L. Celis-Diez, J.J. Armesto

Introduction

Changes in the number of plant or animal species in relation to area of habitat patch (Harris, 1984), latitude and, its mirror image elevation (Rahbek, 1997), provide the best-known relation-ships describing natural patterns of species richness for a number of diverse taxa (Connor and McCoy, 1979; Rohde, 1992). In areas historically subjected to strong and persistent anthropo-genic disturbance, such as tropical dry forests (TDF) and Mediterranean climate woodlands, variables associated with human impact, such as fire, cattle grazing and logging, can greatly in-fluence species diversity patterns to TDF (e.g. Bullock et al., 1995; Gentry, 1995; Trejo and Dirzo, 2002; Segura et al., 2003; Gordon et al., 2004; White and Hood, 2004; Balvanera and Aguirre, 2006; Williams-Linera and Lorea, 2009) and Mediterranean climate woodland (e.g. Bond,1983; Armesto and Martínez, 1978). Other physical variables expected to be associated with pat-terns of tree species richness in seasonally dry forests are the length of the dry season and the amount and timing of precipitation (Richerson and Lum, 1980).

Some authors have shown that tree species richness is not always significantly correlated with the quantity and seasonality of precipitation in TDF (Lott et al., 1987; Gentry, 1995; Gillespie et al., 2000; Trejo and Dirzo, 2002). Instead, patterns of plant species diversity have been found to be associated with variation in potential evapo-transpiration (Trejo and Dirzo, 2002), and with differ-ences in soil moisture availability in relation to elevation, insolation, slope, and soil water-holding capacity (Balvanera et al., 2002; Segura et al., 2003; Balvanera and Aguirre, 2006). Based on re-search undertaken in Mexico, Balvanera and Aguirre (2006) reported that different tree species occupied different parts of the soil moisture gradient, and that many species were excluded from the driest sites, where productivity was lowest. In another Mexican study, Segura et al. (2003) reported that tree species richness declined as soil water availability decreased along a 1 km-long watershed, showing that drier conditions tend to support lower plant species richness. In these forests, live stem densities increased substantially with water availability, while the proportion of dead stems increased towards the drier end of the gradient.

Many of the remnants of dry forests have been subjected to anthropogenic fragmenta-tion, and hence abiotic changes that result from reductions in forest patch area are likely to produce declines in local diversity and density of native trees and other species (e.g. Ben-nett, 2003; Cadenasso and Pickett, 2001; Echeverría et al., 2006; Drinnan, 2005; Hersperger and Forman, 2003; Hobbs, 2001; Holt et al., 1995; Honnay et al., 1999; Laurance et al., 1998a, 1998b, 2001; Matlack, 1994; Quinn and Harrison, 1988; Simonetti et al., 2001; Soulé et al., 1992; Tabarelli et al., 1999; Willson et al., 2001). However, the typical effects of patch area on

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (130)

104

Principles and Practice of Forest Landscape Restoration

species richness have not always been found, for example in the case of South African Medi-terranean woodlands (Kemper et al., 1999). This is due to the fact that tree species richness is not only influenced by the area of remnant forest patches, but also by other patch-related variables, such as edge effects, distance to roads or history of human impact. Variables such as the proportion of core area, perimeter, shape, and connectivity of fragments can also have important consequences for biodiversity (Forman and Godron, 1986; Drinnan, 2005). Recent studies indicate that the spatial configuration of remnant patches may also influence species richness of herbaceous plants (Petit et al., 2004). No studies have explicitly examined the influence of remnant patch attributes on tree diversity patterns in dry tropical and Mediter-ranean forests (but see Kemper et al., 1999).

This study describes and compares the combined effects of elevation and forest patch attributes on tree species richness and composition in four representative, seasonally dry forests in the Americas (see also Boxes 4.1 – 4.8 for associated studies). All landscapes stud-ied have been greatly transformed by human activities. The study areas were located in the dryland tropical forests of southern Mexico (three areas) and the Mediterranean woodlands of central Chile. We present statistical models of environmental and patch variables that ac-count for the tree diversity patterns observed. The aim of this research was to examine the factors influencing patterns of species richness in fragmented dryland landscapes, with the aim of informing approaches to forest landscape restoration. To be effective in restoring bio-diversity, such approaches will need to be based on a firm understanding of the processes influencing species richness patterns.

Study sites, sampling and statistics

Veracruz, Mexico

The study area was central Veracruz in the adjacent municipalities of Comapa and Paso de Ovejas (19° 17’ N and 96° 26’ W, between 97 and 420 m elevation), covering an area of 300 km². Mean minimum and maximum temperatures are 19.8°C and 30.7°C, respectively. Mean annual pre-cipitation is 966 mm (range: 502–1466 mm), which is unevenly distributed during the year. The dry season extends from October to May (station at Loma Fina; 7 to 28 km from the study sites). Land use in this region is dominated by small-scale cattle ranching by private landowners, but communal tenants (ejidatarios) practice more diverse land uses, mainly maize farming (Gallardo-López et al., 2002). Some dominant tree species are Bursera cinerea, Calyptranthes schiediana, Comocladia engleriana, Ipomoea wolcottiana, Leucaena lanceolata, Luehea candida, Savia sessiliflora, Spondias purpurea, Tabebuia chrysantha and Thouinidium decandrum. Ten for-est fragments were selected to characterize the forest of Paso de Ovejas (Williams-Linera and Lorea, 2009). The study sites were located 0.5 to 22 km away from each other (mean = 10.7 km); the fragments are believed to be remnants of a once continuous forest cover.

Oaxaca, Mexico

In Oaxaca, study sites were located in the municipalities of Santiago Apoala (17°33’ N–97°5’W), Santiago Huauclilla (17°25’N, 97°1’ W) and Santiago Tilantongo (17°3’ N, 97°17’W) (see also Box 4.4). The climate is sub-humid to moderately semi-dry, annual rainfall ranges from 600–700 mm (Santiago Apoala) to 800–1000 mm (Huauclilla; Tilantongo). In each area, we selected four forest fragments with a relatively hom*ogeneous vegetation cover. Fragment selection

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (131)

105

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

was aimed at obtaining ecologically contrasting areas. The fragments were selected by using 2005 SPOT satellite images, using ArcView® to detect masses of hom*ogeneous vegetation, separated by at least five pixels (150 m distance) from similar vegetation patches (Fig. 4.1). Fragment metrics including area, perimeter, and perimeter/area ratios were quantified using FRAGSTAT® (McGarigal and Marks, 1994). A total of 216 circular plots (5.7 m radius, 102.1 m2) were sampled. The plots were randomly distributed and separated from each other by a minimum distance of 70 m. Each tree (>2 m in height and >2.5 cm stem diameter) was positioned by means of polar coordinates. When trees could not be identified taxonomically, they were assigned to morphospecies categories.

Figure 4.1

Chiapas, Mexico

The study area was located in the Central Depression of Chiapas between 17° 59’–14° 32’ N, and 90° 22’–94° 14’ W. The region is completely surrounded by warm, moist mountain areas in transition zones with colder and drier woodlands that provide a complete isolation from other regions with TDF. The vegetation was originally dominated by tropical deciduous vegitation, but much of this has been replaced by secondary vegetation owing to grazing, fire, construction of dams and cutting. Only in certain protected areas such as canyons and upper slope areas are strongholds of vegetation present that exhibit relatively little evidence of human influence. At upper elevations, the climate is driest with a mean annual temperature of 22.8°–25.8°C and an average rainfall varying between 660 and 1051 mm annually. At the middle and upper elevations (>800 m a.s.l.) the climate is semi-warm and dry, with the rainfall varying between 1110 and 1267 mm and a mean annual temperature below 23°C. Site selection was preferential (Matteucci and Colma, 1982), and was performed using high resolution satellite images in Google Earth 4.3 supported by field surveys, to identify the present vegetation physiognomy in the study area. All individual trees >10 cm dbh were counted in 1000 m2 circular plots, and individuals between 5–10 cm dbh (diameter at breast height) were counted in an inner 100 m2 circular plot. The number of plots surveyed in each forest fragment was determined by the structural complexity (the degree of hom*ogeneity, taking into account the dominant species) and spatial extent.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (132)

106

Principles and Practice of Forest Landscape Restoration

Central Chile

The study was conducted on the coastal range of central Chile (between 32º–34ºS), between 300 and 1200 m elevation, covering an area of about 7000 km2. Mean minimum and maxi-mum temperatures are 4.5°C and 33.4°C, respectively (di Castri and Hajek, 1976). The cli-mate is characterized by a dry season of 5 to 7 months during the austral summer (October to April), with a mean annual air temperature of 12.7°C (di Castri and Hajek, 1976). Mean annual precipitation ranges from 350 mm to 800 mm, increasing from north to south (di Cas-tri and Hajek, 1976), and rain is concentrated in relatively few events that occur during the austral winter (May to August). Remnant patches of sclerophyllous woodlands in this region are often surrounded by exotic conifer, or eucalypt plantations, highly degraded shrublands, or anthropogenic grasslands. Dominant tree species in remnant woodlands include shade-tolerant species such as Cryptocarya alba, Peumus boldus, Dasyphyllum excelsum, Beils-chmiedia miersii, and markedly shade-intolerant tree species such as Quillaja saponaria, Schinus latifolius, and Acacia caven. The Mediterranean-climate region of central Chile is characterized by strong abiotic and vegetational heterogeneity, especially between contrast-ing slopes of north vs. south aspect (Armesto and Martínez, 1978).

Selection and digitization of woodland patches was carried out from Google Earth images. Fragments were classified into four size classes (0.5–10 ha, 10–100 ha, 100–1,000 ha, and >1,000 ha), representative of the full range of remnant woodlands in the landscape. The number of patches sampled varied for each size class: 22, 9, 7 and 3 patches, for each of the above size classes respectively. Fragments above 1200 m elevation were not sampled, to exclude non-sclerophyllous forests. We controlled for slope aspect effects by sampling on slopes with south, southwest and/or southeast exposure. Plots were predominantly located on slopes between 15 and 35 degrees. In each fragment we recorded the number of tree species, densities of stems >5 cm, diameter at breast height (dbh) and number of seedlings (0.1 cm–2 m tall) present within one (size classes 1–3) or two (size classes 4–5) 10 x 10 m plots. This number of plots provides a reliable estimate of tree species richness in the entire patch, according to species/area curves (Becerra et al., unpublished data).

Statistical analysis

Dependent variables were tree and seedling species richness and abundance. In Chiapas, seedling abundance was not measured. Independent variables were elevation, fragment area, perimeter and perimeter/area ratio. Patch areas and perimeters were significatively corre-lated in Veracruz (r2 = 0.75, p = 0.08), Oaxaca (r2 = 0.94, p <0.001) and Chile (r2 = 0.96, p <0.001), and hence only area was analyzed as a patch variable and elevation as an environ-mental variable. Statistical analyses were conducted using area and log area, but since no differences were found between these approaches, only linear relationships are reported. To assess the effect of patch area and elevation on tree and seedling species richness and abundances, a GLIM model selection was conducted using R project software (version 2.7.1, 2008) (R Development Core Team, 2005), with log transformed data to correct for departure from normality, and including the effect of location and country as co-variates. We selected models with and without interactions among factors, based on the lowest AIC (� AIC >2). Otherwise models were considered statistically identical, in which case we selected the model with the smallest number of parameters, assuming a parsimonious criterion (Burn-ham and Anderson, 2002).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (133)

107

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Cleared dry forest in Veracruz, Mexico. Photo: C. Alvarez

Acacia spp., Chile. Photo: C. Echeverria

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (134)

108

Principles and Practice of Forest Landscape Restoration

Results and discussion

Patterns of tree species richness

In Veracruz, Mexico, a total of 175 species of adult trees and 60 species of tree seedlings were recorded in a total of 11 remnant fragments of tropical dry forest that were sampled (Box 4.8). In Oaxaca, southern Mexico, a total of 52 species of adult trees and 140 species of tree seed-lings were recorded in 12 tropical dryland forest fragments. For TDF in Chiapas, a total of 263 tree species were recorded in all forest fragments surveyed (see also Box 4.2). In contrast, for Mediterranean sclerophyllous forests in central Chile a much lower total of 14 species of adult trees and 15 species of tree seedlings were recorded in a total of 41 remnant forest fragments.

Pearson correlations were used to analyze the relationships between elevation above sea level and adult tree species richness in all regions, which was positive and significant only for Veracruz (Table. 4.1; see also Box 4.1). Both in Veracruz and Oaxaca elevation was also related to the over-all abundance of adult trees. Elevation was significantly and positively related to tree seedling abundance in remnant forest patches in Oaxaca only (Fig. 4.1). Forest fragment area was related to species richness of adult trees only in Oaxaca, and to tree seedling abundance only in Chile. In Chiapas, there was no relationship between forest patch area and adult tree species richness.

Figure 4.2

Table 4.1

p p

Veracruz Oaxaca Chiapas Chile

Elevation and S 0.83*, –0.10 ns, –0.12 ns, 0.26 ns,

Area and S –0.2 ns 0.77* –0.16 ns 0.16 ns

Elevation and D 0.85* 0.62* — —

Area and D 0.30 ns –0.38 ns — 0.04 ns

Elevation and d –0.22 ns 0.67* — 0.20 ns

Area and d –0.44 ns 0.44 ns — 0.36*

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (135)

109

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Box 4.1 Altitudinal variation in vegetation structure and diversity of tree species in the tropical dry forest region of central Veracruz

M. Toledo Garibaldi, G. Williams-Linera

In central Veracruz, from the coast of the Gulf of Mexico to the summit of Cofre de Perote (4282 m a.s.l.) located at the end of the Neovolcanic Transversal belt of Mexico, vegetation types range from coastal vegetation to dry forest, cloud forest, oak forest, coniferous forest, high elevation grassland, and alpine vegetation. This continuous change in plant communities along altitudinal gradients has led to several hypotheses – including many related to topography, soil, and climate – to explain the changes in biodiversity and structural patterns of some forests. In this study, we investigated whether the tropical dry forest displayed patterns of variation along the altitudinal gradient. The objective was to determine changes in the vegetation structure and tree species composition of dry forest as they relate to climatic variation.

The study area is located in the tropical dry forest (TDF) region of central Veracruz, Mexico, in the municipalities of Paso de Ovejas and Emiliano Zapata. In the study area, eight sites were selected between 100 and 1000 m altitude in the dry and sub-humid climatic zone. The criteria were that forest fragments be relatively undisturbed and located along the 19th century Royal Road from Veracruz to Mexico City.

In each forest fragment, trees >5 cm dbh (diameter at 1.3 m) were measured and identified to species level on ten 10 x 10 m plots. Precipitation and temperature were obtained from the nearest meteorological stations. Alpha diversity of richness (number of species) and Shannon’s diversity index were calculated for each site. Beta diversity was analyzed using the complementarity of fragment richness, which is the proportion of all species at two sites that occur in only one or the other; it varies from zero (when the lists are identical) to one (when the lists are completely different). Basal area, density, and vegetation height were analyzed using ANOVA with post hoc Tukey’s significant difference test. The correlation between pairs of variables was determined using Pearson correlation coefficients.

Precipitation and temperature were positively (r = 0.67, p = 0.07) and negatively related to elevation (r = –0.97, p <0.0001), and both are therefore important environmental factors influencing plant species distribution in the altitudinal gradient. The vegetation structure of TDF sites was heterogeneous (Table

1). Changes in tree density were significant among sites (F = 3.40, p = 0.004), and the site at the lowest altitude showed the lowest density (Table 1). Vegetation height also changed among sites (F = 5.75, p <0.0001), with a trend for lower tree heights at sites with lower elevations (Table 1). Tree density was positively related to elevation (r = 0.63, p = 0.09) whereas the height of the trees was lower at higher altitudes (r = –0.63, p = 0.09) and tree height and temperature were also positively correlated (r = 0.69, p = 0.06). In contrast, changes in basal area were significant among sites (F = 4.18, p = 0.0007) but were not correlated to altitude or to precipitation or temperature (Table 1).

A total of 136 overstory tree species were found at the study sites. According to importance values indices (IVI), some of the dominant tree species were Bernardia mexicana, Bursera simaruba, Caesalpinia cacalaco, Ceiba aesculifolia, Comocladia engleriana, Croton reflexifolius, Ipomoea wolcottiana, Lysiloma acapulcense, Leucaena lanceolata, and Piscidia piscipula. Richness varied between 17 and 34 tree species, and Shannon’s diversity index varied between 2.26 and 3.13 (Table

1) without a clear altitudinal trend, although they were significantly correlated (r = 0.86, p = 0.006).

Beta diversity as a measure of complementarity indicated a high species turnover among sites. Nearby sites shared only a few species (76 to 94%). Bursera simaruba was the only species distributed along the entire altitudinal gradient; some species appeared to be restricted to one site, but no clear altitudinal trend was detected. The lowest similarity in tree species was between site 8 and the other sites (96 to 100%). The site at the highest altitude was different since it was not TDF: species such as Quercus sapotifolia and Clethra macrophylla became dominant. TDF and hot sub-humid climate reached its limit at this elevation of 986 m, where the ecotone between TDF and the upper vegetation type – tropical montane cloud forest – occurs.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (136)

110

Principles and Practice of Forest Landscape Restoration

TDF had a very heterogeneous vegetation structure; a consistent pattern throughout the altitudinal gradient was not found, although at a lower elevation, the temperature was higher, tree density was lower, and trees were taller. Species composition, on the other hand, did not follow a trend; few species were present at all sites, and others were found only at a few. We concluded that the altitudinal range occupied by TDF is wide in central Veracruz. Within this altitudinal range, changes in structure and tree species dominance from one site to another may be related to factors such as topography, slope, soil type, or anthropogenic disturbance (see Box 4.9).

Table 1

1 2 3 4 5 6 7 8

Puente Nacional

Don Tirzo La Virgen Plan del Río

Dos Caminos

Cerro Gordo

Corral Falso

Lencero

Altitude 140 204 227 335 376 501 780 986

Precipitation 1186 890 890 912 1045 892 1112 1421

T°C 27.2 24.9 24.9 25.1 24.4 23 21.1 19.6

Basal area 29.34ab 20.87b 21.94b 17.54bc 30.32ab 39.19a 22.62b 28.05ab

Density 844b 1100a 1020a 1487a 1320a 1378a 1244a 1470a

Height 10.0a 8.7a 8.5ab 8.0b 9.5a 8.7a 6.78bc 8.1b

Richness 17 34 22 26 33 31 21 29

Shannon Index 2.26 3.13 2.49 2.68 3.01 2.79 2.34 2.37

Box 4.1 (cont.)

Box 4.2 Diversity of woody vegetation in the Central Depression of Chiapas, Mexico

N.E. Taylor-Aquino, N. Ramírez-Marcial, R. Vaca

The analysis of the patterns, causes and maintenance of tropical biodiversity are issues that have generated considerable attention for many years among biologists and ecologists (Bullock et al., 1995). Understanding the spatial variation of plant species diversity is particularly relevant in the tropics because of their high diversity and threatened status (Lawton et al., 1998). Generating more biological and ecological information is necessary for developing such an understanding. This information is still scarce for most tropical regions, but is needed to support land-use planning, monitoring, and the development of restoration plans in degraded areas (Lindenmayer and Franklin, 2002; Huston, 2004).

Chiapas is the second most floristically diverse state of Mexico, with a high number of endemic plants (Miranda et al., 1963; Rzedowski, 2006; Breedlove, 1981). Its latitudinal range, topography and geological history determine a high spatial environmental heterogeneity and create a large variety of ecological conditions (Breedlove, 1981). Climate is considered as the major determinant

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (137)

111

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

of vegetation distribution (Woodward, 1987), and the most influential factor on the climate of this region is the topography. Local micro-climates are very fine-grained.

The Central Valley of Chiapas, also known as Central Depression of Chiapas, is located in the central portion of the state. The dry forests of the Central Valley are completely surrounded by moist forested mountain areas, providing relative isolation from other areas of dry vegetation. In this context, dry forests extend over altitudinal gradients, ranging from lowland deciduous and sub-deciduous tropical forests up to oak and pine-oak forests (Breedlove, 1981). Extensive cultivation and grazing has led to large tracts of thorn woodland and savannah (Breedlove, 1981; Challenger, 1998). Following these altitudinal gradients and according to the WorldClim database, the mean annual temperature ranges from 22 to 25.2ºC, and annual precipitation ranges from 750 to 1500 mm. In this work we aimed to identify and describe tree-species associations of dry forest occurring along altitudinal gradients in the Central Valley of Chiapas.

We evaluated the floristic composition and structure of woody vegetation by 131 circular plots (13.1 ha total) in different localities of the Central Depression of Chiapas. The sample was stratified into two diameter-size categories: (1) small trees (individuals 5–10 cm dbh, in plots of 0.01 ha), and (2) large trees (individuals with dbh >10 cm in plots of 0.1 ha). Although this geographical region has a long history of land-use activities, there are still some remnants of woody vegetation in varying stages of successional development. The plots were located in these remnants, over a wide range of environmental variation following altitudinal gradients (440–1740 m).

We recorded a total of 263 tree species distributed in 161 genera and 66 families. Through a Cluster Analysis we identified a total of four tree species associations (Fig. 1), based on species dominance: (1) Matayba oppositifolia-Ternstroemia tepezapote-Tapirira mexicana, related to tropical sub-deciduous forest (other dominant species: Nectandra salicifolia and Bursera simaruba); (2) Bursera simaruba-Cochlospermum vitifolium, related to tropical deciduous forest (other dominant species: Heliocarpus reticulates, Leucaena shannonii and Bursera excelsa); (3) Quercus segoviensis-Quercus crispipilis, related to oak forest (other dominant species: Ternstroemia tepezapote, Rhus schiedeana and Quercus polymorpha); and (4) Quercus peduncuaris-Quercus acutifolia-Pinus oocarpa, related to pine-oak forest (other dominant species: Byrsonima crassifolia, Quercus castanea and Quercus conspersa). These groups were arranged along altitudinal and geographical gradients. Bursera simaruba, Cochlospermum vitifolium, Leucaena shannonii, Heliocarpus reticulatus, Calycophyllum candidissimum, and Bursera bipinnata were the most abundant species in the lowest elevations (under 900 m a.s.l.). In the highest elevations (above 800 m a.s.l.) the most abundant species were Quercus (mostly Quercus peduncularis) and Pinus oocarpa.

We identified four more groups for which distribution was better explained by human influence than climate variables; 85% of their species had less than 10 cm DBH and were pioneer trees: Ficus pertusa, Mimosa tenuiflora, Heliocarpus reticulatus, Acacia cornigera, Diphysa robinioides, Guazuma ulmifolia, Bursera simaruba, Ficus cotinifolia, Luehea candida, Genipa americana, Casearia corymbosa, Alibertia edulis and Stemmadenia obovata. Most of these species are found within open pasturelands or may be used as a protein supply for cattle and live fences.

Although regional and beta diversity are relatively high, local diversity is relatively low and is represented by initial and secondary pioneer species, which pre-supposes a long history of changes associated with extreme weather events and interaction with anthropogenic factors.

Box 4.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (138)

112

Principles and Practice of Forest Landscape Restoration

Figure 1 Matayba oppositifolia-Ternstro-emia tepezapote-Tapirira mexicana Bursera simaruba-Cochlospermum vitifolium Quercus segov-iensis-Quercus crispipilis Quercus peduncularis-Quercus acutifolia-Pinus oocarpa

Caesalpinia eriostachys Quercus polymorpha

Box 4.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (139)

113

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Deforestation of dry forest in central Veracruz, Mexico. Photo: C. Alvarez

Dry forest in Chiapas, Mexico. Photo: N. Ramírez-Marcial

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (140)

114

Principles and Practice of Forest Landscape Restoration

Box 4.3 Tree species diversity and forest structure in subtropical dry forest of northwestern Argentina

C. Blundo, L. R. Malizia

Subtropical seasonally dry forests (SSDF) in northwestern Argentina (22º–24ºS and 63.5º–65ºW) include Chaco forest (CF) from 300–400 m a.s.l. and premontane forest (PF) from 400–900 m a.s.l. PF has usually been called ‘transitional forest’ (Hueck, 1972) between the neighbouring dry CF and the more humid Yungas forest (i.e. Selvas de Montaña, Cabrera 1976). However, Prado (2000) recognized PF as a vegetation unit more closely related to other tropical seasonal forests of South America, owing to its characteristic flora and physiognomy. SSDF covers approximately 10,000 km2 and shows a broad range of rainfall. Annual rainfall averages 625 mm (range: 450–700 mm) at CF and 820 mm (range: 550–1400 mm) at PF, concentrated during the summer (November to March) (Bianchi and Yañez, 1992). The mean annual temperature 21.5º C is relatively hom*ogeneous, although thermal amplitude is variable in the study area (Arias and Bianchi, 1996). Based on this climatic variability and the common origin of these forests, our main objectives were to identify the tree-species compositional gradient across SSDF and to describe forest structure in different tree communities at the regional scale.

We established 23 1-ha permanent plots, three plots in CF and 20 plots in PF. All plots were 20 m x 500 m corrected for slope, to actually cover 1 ha. A full inventory was made of all trees ≥10 cm in diameter at breast height (dbh). Trees were marked, measured for dbh and height, and identified to species level. For data analysis we performed a Detrended Correspondence Analysis (DCA). DCA is an unconstrained ordination analysis that provides the basic overview of the compositional gradients in species-abundance data (Lepš and Šmilauer, 2003). We calculated basal area, species richness and canopy height in each plot to compare forest structure between sample plots.

We identified 10,029 trees belonging to 116 species, 93 genera and 43 families. The length of the first axis provided an estimate of the high beta diversity in tree species composition in SSDF (5.3 SD units). First and second axes explained about 30% of total species variability (axis 1, 19.8; axis 2, 9.8), whereas the remaining axes explained much less. Distribution of samples and species in a bi-plot suggested that there are three groups or community assemblies at the regional scale. First, near zero on axis 1, were plots located in CF with Ziziphus mistol, Ruprechtia triflora and Geoffrea decorticans as exclusive species. Then, on the other side, upward, were PF plots located toward the west of the study region, and downward were PF plots located toward the east. Rainfall is higher toward the east of the study area, where species such as Pisonia ambigua, Chrysophyllum gonocarpum and Diatenopteryx sorbifolia were abundant. These species are common at higher elevation (i.e. Yungas forest), whereas they are poorly represented in PF plots located in the west. On the other hand, species such as Ceiba insignis, Phyllostilon rhamnoides, Calycophyllum multiflorum and Astronium urundeuva were abundant in PF plots located in the west, where moisture stress could be higher because temperatures reach more than 40ºC in the summer (Brown et al., 2001). When we compared forest structure, mean CF basal area (15.3±2.7 m2/ha) and canopy height (13.3±3.2 m) yielded lower than average PF values (west PF plots: 22.1±1.7 m2/ha and 20.6±1 m; east PF plots: 21.3±0.8 m2/ha and 19.7±0.7 m). Species richness per 1 ha plot varied between all groups: 20 species (range: 18–24) at CF, 35.2 species (range: 22–48) in the west PF, and 39.3 species (range: 32–45) in the east PF.

Our results showed important gradients in species diversity and forest structure in the SSDF, particularly in terms of species richness. Distribution of tree species across and within Yungas forest is strongly affected by climatic factors (Malizia, 2004). We believe that variability in rainfall and temperature could be playing a mayor role in tree species distribution at SSDF in northwestern Argentina. Therefore, an important goal for ecologists in the future is to predict how these forests will respond to climate change.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (141)

115

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Figure 1 Achatocarpus praecox Agonandra excelsa Anadenanthera

colubrina Astronium urundeuva Calycophyllum multiflorumCedrela balansae Ceiba insignis Chrysophyllum gonocarpumCordia trichotoma Cupania vernalis Diatenopterix sorbifoliaEugenia uniflora Geoffrea decorticans Gleditzia amorphoidesMyroxylon peruiferum Myrcianthes pungens Myriocarpa stipitataOcotea puberula Parapiptadenia excelsa Patagonula americanaPhyllostilon rhamnoides Pisonia ambigua Pogonopus tubulosusRuprechtia apetala Ruprechtia triflora Saccelium lanceolatumTabebuia impetiginosa Terminalia triflora Trichilia hieronymi Urera baccifera Urera caracasana Zizipus mistol

Box 4.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (142)

116

Principles and Practice of Forest Landscape Restoration

Box 4.4 Patterns of diversity of fungi in an altitudinal gradientS. Vázquez Mendosa, R.F. del Castillo, R, Valenzuela Garza

Macromycetes are one of the most diverse group of organisms on Earth. However, very little is known about their ecology and species composition, particularly in the tropics. We estimated macromycete richness and productivity in four localities ranging from 2100 to 3120 m elevation in Santa Catarina Ixtepeji, Sierra Madre de Oaxaca, Mexico. In each locality an area of 875 m2 distributed in 125 plots was surveyed. A total of 1762 specimens were collected, from which 669 morphospecies could be identified. Productivity and species richness were positively correlated with altitude. This pattern was also found when individual groups of fungi were analyzed separately, namely mycorrhizal, saprobic and coprophilous. As ground moisture was also positively correlated with elevation, humidity – more than temperature – appears to be a major limiting factor for macromycetes at the altitudinal range studied. Lignicolous fungi, however, displayed a monotonic relationship with altitude showing a peak of diversity and productivity near 2250 m. This group had the highest biomass in all the studied localities and, with some exceptions, appears to be less affected by humidity than other groups of fungi. Species similarity calculated with Sorensen (presence/absence) and Renkonnen (abundance) indices were very low among localities, suggesting a high species turnover. Conservation and restoration efforts should pay particular attention to middle and high altitude areas in this mountain range, given its high species richness and the vulnerability of its biota because of climate change.

Box 4.5 Species diversity in northwestern Patagonian dryland forests: implications for restoration

C. P. Souto, K. Heinemann, T. Kitzberger, A. C. Premoli

Drylands comprise 30% of the Earth’s surface; in South America 94% of Patagonia is exposed to some degree of desertification risk. Particularly in northwestern Patagonia, Argentina, treeless areas have been traditionally viewed by foresters and land managers as barren lands unable to support native forest and have been used for extensive sheep and cattle ranching, or the establishment of exotic (mostly pine) plantations, which drastically impact on dry native forests. Nevertheless, at the dry eastern edge of temperate forest, the endemic conifer Austrocedrus chilensis (D. Don) Florin & Boutelje (Cupressaceae) occurs in patches as almost the only tree, while towards the west, this species forms mixed continuous forests with Nothofa*gus species. Austrocedrus is considered the most drought-tolerant tree species in the Patagonian region. Particularly towards its driest distributional range, seedling establishment is dependent on shrub presence, since the limiting factors for seedling establishment in open woodland habitats appear to be the desiccating effects of open sites (Kitzberger, 1995; Kitzberger et al., 2000), a phenomenon known as ‘nurse syndrome’. This syndrome is characterized by the amelioration of micro-environmental variables under a plant – the nurse – which enhances survivorship and/or growth of other species growing in association with it (Raffaele and Veblen, 1998).

We analyzed plant diversity at the landscape scale under heterogeneous and disturbed environments inhabited by Austrocedrus forests. The aim of the research was to examine stand structure and species composition, and thus available nurse shrubs along Austrocedrus’ range, with the aim of informing restoration practices. Austrocedrus is a timber species of high economic value and international conservation concern (status: Vulnerable). It is the most conspicuous tree species in the Patagonian steppe. Austrocedrus is a dioecious species of a monotypic

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (143)

117

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

genus, with wind dispersed pollen and winged seeds. In Argentina, it occurs discontinuously from 36º30’ to 39º30’S and more extensively from 39º30’ to 43º35’S latitude (Seibert, 1982). Over the natural range of the species, most of the precipitation occurs during autumn and winter generating a period of drought in summer. Toward the more humid environments of the south and the west, Austrocedrus forms pure stands that tend to be continuous, with a dense understorey of shrubs and other sub-canopy trees. At the centre of its range, the rain shadow effect of the Andes depicts a more evident west to east natural fragmentation gradient. In northern and eastern dry areas, precipitation declines, aridity increases, the understorey of shrubs and small trees becomes less dense, and Austrocedrus stands open up into sparse woodlands adjacent to the Patagonian steppe (Seibert, 1982). Finally, scattered trees, typically on rocky outcrops, occur sparsely as ingressions into the Patagonian steppe, surrounded by a matrix of bunch grasses and low shrubs.

To identify the broad-scale trends of differentiation throughout the species range in Argentina, we subdivided sampled stands into three regions representing north (N), centre (C) and south (S), according to their geographical proximity, and environmental envelope. We assessed the presence of all plant species and scored 15 Austrocedrus trees along an approximately 20 x 100 m long strip following the major axis in 67 Austrocedrus-dominated forest patches on the eastern slopes of the Patagonian Andes in Argentina. Tree ages are significantly higher in fragmented woodlands adjacent to the steppe in the northern and central region of the species range (mean ages are 106 ± 69 and 98 ± 46 years respectively) than in continuous forests occurring in the central and southern regions (52 ±27 and 48 ±23 years respectively). On the other hand, the annual radial growth rate increases significantly from north to south (1.9, 2.3 and 3 mm/year on average in northern, central and southern forest stands respectively), without significant differences in tree diameter (mean tree diameter sizes decrease from 41 in northern forest patches to 38 and 32 cm in central and southern stands). Comparing both latitudinal range extremes, trees from the northern region are older but smaller in size and thus demonstrate a relatively low growth rate.

A total of 89 understory species were recorded in Austrocedrus stands (Table 1). Only eight species were introduced weeds while naturally occurring plants consisted of 38 herb, 37 shrub, and eight tree species. In northern Austrocedrus stands a total of 55 different species were scored in 23 sites (average = 9.39; SD = 7.71), including seven exotic and 14 nurse species. In the central region, 63 species were scored in 25 sites (average = 12.48; SD = 7.39), including three exotic species and 14 nurse species. Meanwhile, in the south 67 species were recorded in 19 sites (average = 15.11; SD = 6.97), including five exotic species and 18 nurse species (Table

1). Species richness differed among sampled regions (F(1,63) = 3.205, p = 0.047). Specifically, tree species richness increased southward. Along the range of Austrocedrus in Argentina the three regions shared almost 35% of the understory species, but more than 30% of them were exclusive (i.e. only present in one region). Thus, the C region shared with N and S more than 10% of the species but, N and S only shared 4%. Consistently, the C region has only 8% of exclusive species, meanwhile N and S have almost 20% of exclusive species, respectively. In terms of nurse species, a total of 24 were scored in Austrocedrus-dominated communities, which differed across Austrocedrus’ range in Argentina. The three regions shared 40 % of nurse species. In contrast, 12.5%, 8.33% and 20.8% of them were exclusive from N, C, and S region, respectively.

In summary, species richness and stand structure of Austrocedrus vary at the landscape scale, probably in response to climatic and disturbance gradients. Considering scenarios of increasing inter-annual climatic variability and global warming trends, it is possible that climate changes in Patagonia will affect Austrocedrus forests. As a consequence, entire regions may change in terms of landscape and forest patch configuration. For successful restoration actions in Patagonian dry lands with Austrocedrus, special concern should be given to the presence of nurse species (Chapter 5). The significant genetic structure of Austrocedrus (Chapter 7) along with the heterogeneity in community structure and composition reported here should also be taken into account.

Box 4.5 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (144)

118

Principles and Practice of Forest Landscape Restoration

Tabl

e1

Austr

oced

rus c

hilen

sis

Lati

tude

Lon

gitu

de

RSp

ecie

s

37°5

8’55

.5”

to 39°3

1’26

.5”

70°4

7’19

.1”

to 70°5

8’09

.6”

N

Aca

ena o

valifo

lia

; Aca

ena p

inn

ati

fida

; Aca

ena s

plen

den

s; A

desm

ia b

oron

ioid

es; A

lstr

oem

eria

au

rea

; An

emon

e m

ulti

fida

; Ara

uca

ria

ara

uca

na

; Arm

eria

mari

tim

a; B

acc

hari

s sp

.; B

alb

isia

gra

cilis;

Ber

beri

s bu

xif

oli

a; B

erbe

ris

empe

trifol

ia; B

owle

sia tro

paeo

lifo

lia

; Cer

ast

ium

arv

ense

; Chu

squ

ea c

oule

ou; C

olli

gua

ja i

nte

gerr

ima

; Cor

tade

ria a

rau

can

a; D

isca

ria

art

icu

lata

; Ech

ium

vu

lgare

(ex

ot);

Ep

hed

ra

brea

na

; Ep

hed

ra f

rust

illa

ta; E

scallo

nia

vir

gata

; Fa

bia

na

im

bric

ata

; Gau

lther

ia s

p.; G

eran

ium

mage

llan

icu

m; H

apl

opapp

us

glu

tin

osu

s;

Loasa

ber

gii;

Lom

ati

a h

irsu

ta; M

ayte

nu

s bo

ari

a; M

ayte

nu

s ch

ubu

ten

sis;

Mel

ilot

us

alb

a (

exot

); M

ulin

um

ech

inu

s; M

uli

nu

m s

pin

osu

m;

Mu

tisi

a s

p.; N

othof

agu

s an

tarc

tica

; Osm

orhiz

a c

hilen

sis;

Oxa

lis

ade

nop

hylla

; Per

ezia

rec

urv

ata

; Phace

lia s

ecu

nda

; Pla

nta

go la

nce

olata

(e

xot)

; Qu

inch

am

aliu

m c

hilen

se; R

ibes

cu

culla

tum

; Rib

es m

age

llan

icu

m; R

hod

ophia

la m

endo

cin

a; R

osa r

ubi

gin

osa (

exot

); R

um

ex

ace

tose

lla (

exot

); Sc

hin

us

od

onel

lii; S

chin

us

pa

tago

nic

us;

Sen

ecio

sp.

; Sis

yrin

chiu

m v

ulg

are

; Tara

xacu

m m

edic

inale

(ex

ot);

Trop

aeo

lum

in

cisu

m; V

erba

scu

m thaps

us

(exo

t); V

icia

nig

rica

ns;

Vio

la s

p.

40°4

3’18

.6”

to 41º1

3’25

.9”

71°0

8’27

.6”

to 70º4

2’00

.8”

C

Aca

ena o

valifo

lia

; Aca

ena p

inn

ati

fida

; Aca

ena s

plen

den

s; A

desm

ia b

oron

ioid

es; A

desm

ia a

fin v

olck

man

ni;

Als

troe

mer

ia a

ure

a; A

nem

one

mu

ltifid

a; A

rist

ote

lia

ch

ilen

sis;

Arm

eria

mari

tim

a; B

acc

ha

ris

sp.;

Balb

isia

gra

cilis;

Ber

beri

s bu

xif

oli

a; C

aio

phor

a s

p.; C

alc

eola

ria s

p.;

Cer

ast

ium

arv

ense

; Cyn

an

chu

m d

esco

lei;

Co

lleti

a h

ystr

ix; D

isca

ria

art

icu

lata

; Em

both

riu

m c

occi

neu

m; E

scallo

nia

ru

bra

; Ep

hed

ra

chil

ensi

s; E

rin

giu

m p

an

icu

latu

m; E

uph

orbi

a s

p.; F

abi

ana

im

bric

ata

; Fra

gari

a c

hilen

sis;

Galiu

m h

ypoc

arp

ium

; Gau

lther

ia s

p.; G

eran

ium

m

age

llan

icu

m; G

rise

bach

iella

hie

ron

ymi;

Hapl

opapp

us

glu

tin

osu

s ; L

ath

yru

s sp

.; Lo

asa

ber

gii;

Lom

ati

a h

irsu

ta; M

ayte

nu

s bo

ari

a; M

ayte

nu

s ch

ubu

ten

sis;

Mu

ehle

nbe

ckia

hast

ula

ta; M

ulin

um

ech

inu

s; M

uli

nu

m s

po

no

sum

; Mu

tisi

a s

p.; M

yosc

hilos

obl

ongu

m; N

ard

ophy

llum

ob

tusi

foliu

m; N

ass

au

via

glo

mer

ulo

sa; O

rqu

idea

sp.

; Osm

orhiz

a c

hilen

sis;

Oxa

lis

ade

nop

hylla

; Per

ezia

rec

urv

ata

; Phace

lia s

ecu

nda

; Q

uin

cham

aliu

m c

hilen

se; R

ibes

cu

culla

tum

; Rib

es m

age

llan

icu

m; R

osa r

ubi

gin

osa (

exot

); R

um

ex a

ceto

sella

(ex

ot);

Ru

moh

ra a

dian

tifo

rmis

; Sc

hin

us

od

onel

lii;

Sch

inu

s p

ata

gonic

us;

Sen

ecio

sp.

; Sis

yrin

chiu

m v

ulg

are

; Sol

idago

chilen

sis;

Tro

paeo

lum

in

cisu

m; V

ale

rian

a s

p.;

Verb

asc

um

thaps

us

(exo

t); V

icia

nig

rica

ns;

Vio

la s

p.

41°4

7’48

.2”

to 43°4

3’24

.9”

71°2

5’51

.7”

to 71°2

4’16

.4”

S

Aca

ena o

valifo

lia

; Aca

ena p

inn

ati

fida

; Aca

ena s

plen

den

s; A

desm

ia s

p.; A

na

rtrh

op

hyl

lum

str

igu

lip

eta

lum

; An

emon

e m

ulti

fida

; Ari

stot

elia

ch

ilen

sis;

Arm

eria

mari

tim

a; A

zore

lla m

onan

tha

; Ba

cch

ari

s sp

.; B

alb

isia

gra

cilis;

Ber

beri

s bu

xif

oli

a; B

erbe

ris

empe

trifol

ia; B

lech

nu

m

mage

llan

icu

m; B

owle

sia tro

paeo

lifo

lia

; Caio

phor

a s

p.; C

alc

eola

ria s

p.; C

erast

ium

arv

ense

; Cir

siu

m v

ulg

are

; Co

lleti

a h

ystr

ix; D

iost

ea

junce

a; D

isca

ria

art

icu

lata

; Dis

cari

a c

ha

caye

; Dis

cari

a t

riner

vis;

Em

both

riu

m c

occi

neu

m; E

ph

edra

ch

ilen

sis;

Eri

ngi

um

pan

icu

latu

m;

Esc

allo

nia

ru

bra

; Eu

phor

bia s

p.; F

abi

ana

im

bric

ata

; Fra

gari

a c

hilen

sis;

Galiu

m h

ypoc

arp

ium

; Gau

lther

ia s

p.; G

eran

ium

mage

llan

icu

m;

Gri

seba

chie

lla h

iero

nym

i; H

apl

opapp

us

glu

tin

osu

s; J

un

iper

us

com

mu

nis

(ex

ot);

Lath

yru

s sp

.; Lo

mati

a h

irsu

ta; M

ayte

nu

s bo

ari

a; M

ayte

nu

s ch

ubu

ten

sis;

Mu

linu

m s

pin

osu

m; M

uti

sia s

p.; M

yosc

hilos

obl

ongu

m; N

ard

ophy

llum

obt

usi

foliu

m; N

othof

agu

s an

tarc

tica

; Not

hof

agu

s pu

milio

; Orq

uid

ea s

p.; O

smor

hiz

a c

hilen

sis;

Ovi

dia a

ndi

na

; Oxa

lis

sp.;

Pere

zia r

ecu

rvata

; Phace

lia s

ecu

nda

; Pla

nta

go la

nce

olata

(ex

ot);

Po

lyst

ichu

m p

lica

tum

; Qu

inch

am

aliu

m c

hilen

se; R

ibes

cu

culla

tum

; Rib

es m

age

llan

icu

m; R

osa r

ubi

gin

osa (

exot

); R

um

ex a

ceto

sella

(ex

ot);

R

um

ohra

adi

an

tifo

rmis

; Sch

inu

s o

do

nel

lii;

Sch

inu

s p

ata

gonic

us;

Sen

ecio

sp.

; Vale

rian

a s

p.; V

erba

scu

m thaps

us

(exo

t); V

iola

sp.

Bo

x 4

.5 (

co

nt.

)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (145)

119

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Box 4.6 Avian-generated seed rain and germination in the patchy shrubland of central Chile

S. Reid, C. Christophers, J.L. Allendes, J.J. Armesto

The pattern of seed rain across a landscape can determine the distribution of potential recruitment rates, influencing the spatial processes of colonization, range expansion and gene flow. Likewise, the quantification of propagule input and germination response in different micro-sites is of fundamental importance to the restoration of degraded plant communities (e.g. Méndez et al., 2008). This information is key for understanding the limiting factors for natural regeneration and succession, and for applying information about seed dispersal and germination in restoration programmes.

In this study, we provide evidence of the positive contribution of avian-frugivores to seed dispersal and the outcome of seeds in dispersed sites in the patchy Mediterranean-type shrubland in central Chile. Here, the regeneration of woody species is limited by seed inputs because soil or aerial seed banks of woody species are extremely poor or entirely absent (Fuentes et al., 1984). In addition, regeneration is severely limited to wet micro-sites under bushes in a mosaic of sparse shrub clumps separated by open areas exposed to drought and herbivory by feral rabbits (introduced from Europe), cattle and horses. This shrubland presents a high incidence of avian-dispersed woody species where 50% of the species are dependent mostly on birds for seed dispersal and germination. In addition, seed germination of woody species is not responsive to fire or smoke stimulation. Given the patch structure of Chilean semi-arid ecosystem, we assessed the effect of the present patch structure on avian-generated seed rain and germination.

We quantified avian-generated seed rain patterns directly by collecting defecated seeds from the ground. We searched for avian droppings along five 90 m linear transects separated by 10 m, covering a total area of 450 m2. Avian-generated seed deposition patterns were based on a total sample of 370 seeds of Schinus polygamus (Anacardiaceae) from 95 bird droppings, which made up 80% of the seeds present in droppings. Avian-generated seed rain was compared among different patch-types defined as: (1) ‘open’, bare ground between shrub clumps; (2) ‘low’, beneath Baccharis sp. and Retanilla trinervia (25% shrub cover, 1.5 m mean height); (3) ‘midheight’, including Lithrea caustica, Schinus polygamus, Azara dentata and Colliguaja odorifera (77.8% mean shrub cover and 3.4 m mean height); and (4) ‘tall’ (64.6% mean cover and 5.8 m mean height), including Maytenus boaria and Quillaja saponaria. Germination trials were conducted with Lithrea caustica (Anacardiaceae) in the same four contrasting patch types characteristic of this shrubland. In each of the four patches, 60 seeds were placed in open-ended plastic cups (10 seeds per cup) and these were covered with a 5.8 mm mesh cage (50 x 50 cm) to keep out vertebrate herbivores. Six replicates were set up in each patch type, except in the open ground patches where we set up three replicates. The experiment began in the austral winter (August 2006) and germination (emergence of the radicle) was recorded weekly until December (14 records), when germination ceased because of lack of rainfall.

The avian-generated seed rain differed among patch types and differences did not correspond to those expected from the frequency of different patch types in the study area (c2 = 28.1, df = 3, p <0.001; Fig. 1). Dispersed seeds were highly concentrated under ‘tall’ patches (under Q. saponaria and M. boaria trees). Tall patches received more seeds than expected based on their ground cover percentage, in contrast to ‘open’ patches that received significantly lower seed rain than expected. There was also a significant effect of patch type on the cumulative percentage of germination of L. caustica (�2 = 25.94, p <0.001; Fig. 2). Germination was higher under tall and midheight patches and significantly lower under low Baccharis patches (Z = 4.95, p <0.001 between low and midheight patches, and Z = 4.12, p <0.001 between low and tall patches). Seed germination was zero in open ground between patches.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (146)

120

Principles and Practice of Forest Landscape Restoration

Figure 1 Schinus polygamus

Figure 2 Lithrea caustica� � � �

In this study, we showed that the majority of seeds (62% of 370 seeds) transported by birds are dispersed to sites were germination and seedling survival have higher probabilities of success, although a large percent is still also dispersed to open areas. Considering seed germination rates were higher under tall and midheight patches of Q. saponaria and L. caustica respectively, we suggest that dispersal to these ‘nurse’ patches is positive for plant recruitment, in agreement with previous studies in this shrubland (e.g. Del Pozo et al., 1989) and Mediterranean shrublands in southern France (Debussche and Isenmann, 1994). Consequently, avian-frugivores are contributing ‘direct’ seeds to favourable sites for germination. This disperser-mediated facilitation which leads to the aggregation of seedlings of woody plants around pre-existing shrub ‘nurses’, offers a tool for restoration, as it can accelerate succession, by mostly driving the early appearance of late-successional species (e.g. M. boaria, L. caustica, and Schinus sp. in this shrubland). Most likely succession could slowly lead to the coalescence of patches from seedling recruitment along patch edges, but this requires decreased herbivory and high moisture conditions.

We recommend the conservation of tall patches (e.g. Q. saponaria and M. boaria) as important to facilitate succession in shrublands of central Chile. This guideline can be applied in restoration practices in degraded shrubland or dry forest areas, by planting or protecting remnant adult Q. saponaria and M. boaria to facilitate progagule input. Protection from herbivory should also be necessary to accelerate the process of seedling establishment under patches.

Box 4.6 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (147)

121

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Box 4.7 Effect of fragmentation on plant communities of central Chile

P. Becerra, C. Smith-Ramírez, C. Echeverría, J.J. Armesto

We assessed the consequences of fragmentation for plant species diversity and population densities of remnant sclerophyllous forests of central Chile. We selected 41 woodland fragments of different patch sizes (1 to 2000 ha). Each fragment was characterized by several fragmentation variables related to area and shape (area, core area, shape index, fractal dimension). In addition, we quantified the distance from each forest patch sampled to the nearest urban centre and recorded fire, and also the elevation of the sites where each fragment was sampled. We tested the hypotheses that different communities and population variables vary along gradient of fragmentation, distance to urban centres and fires, and elevation.

We found that larger fragments were also more complex in shape. Only some community and population variables were significantly related to fragmentation. Tree, shrub and epiphyte, but not herb species richness, and diversity increased in larger and more complex fragments. In turn, only tree species richness was significantly correlated to the distance to urban centres and no community variable to distance to fires. Also, density of different size classes of particular tree species were significantly correlated with fragmentation variables, although with different patterns observed for different species. Density of more shade-tolerant species was positively correlated to patch area and complexity, while density of more shade-intolerant tree species was not. Also, density of old individuals of two tree species were positively correlated with distance to fires suggesting that fires could have negatively affected density of some tree species. Finally, elevation was significantly positively correlated with richness, diversity and density of some species suggesting that climatic variation with elevation affects plant communities and populations. In conclusion, patterns detected in this study suggest that fragmentation and elevation are relevant factors affecting communities and populations of plant species in this Mediterranean-climate ecosystem.

Details of methods

Field surveys were undertaken in forest fragments classified into four size classes, to ensure that a broad range of patch sizes, representative of remnant woodlands in the landscape, was sampled. Four classes were: 0.5–10 ha, 10–100 ha, 100–1000 ha, >1000 ha. From each size area we selected 22, 9, 7 and 3 patches respectively. Thus we worked with 41 fragments distributed between 32ºS and 34ºS, located only on the coastal range of central Chile. Surveys were undertaken by establishing one 10 x 10 m plot per patch in the two smaller patch size classes, and two plots of the same dimension at least 300 m apart in fragments belonging to the two larger patch size classes. Within each sample plot we recorded the dbh of each individual >5 cm of all tree species. In addition, we sampled 15, 1 x 1 m sub-plots located systematically within the larger plot. In each sub-plot we recorded the presence of each vascular plant species, including herbs and ferns, and the regeneration (seedlings and juveniles) of tree species. Regeneration was subdivided in two classes: saplings including individuals <5 cm dbh higher than 0.5 m, and seedlings including individuals <0.5 m height. Additionally, evidence of recent fires (yes or no), recent logging (yes or no) and/or recent trampling or browsing by livestock (faeces) (yes or no) was noted for each plot and the 20 m around it.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (148)

122

Principles and Practice of Forest Landscape Restoration

Box 4.8 Tree species diversity driven by environmental and anthropogenic factors in tropical dry forest fragments of central Veracruz, Mexico

G. Williams-Linera and F. Lorea

We examined vegetation structure and woody species diversity in relation to 14 environmental and anthropogenic factors in ten tropical dry forest (TDF) fragments in central Veracruz, Mexico. The basal area of the canopy (30.2 ± 2.11 m2/ha) and understory (1.96 ± 0.12 m2/ha) trees was similar, but density (1014 ± 104 and 2532 ± 227 individuals/ha, respectively) differed among sites. We recorded 98 canopy, 77 understory, and 60 seedling species. Richness was 24–45 species per site, Fisher’s alpha and Shannon’s indices increased with site altitude. Chao Jaccard indices revealed high species turnover, and a consistently higher similarity within the sites at the lowest and within the highest elevation sites. Ordination identified altitude, aspect, slope, water proximity, cattle and trails as significant explanatory variables of species patterns, and showed that sites at lower elevations were clearly separated from the other sites. Environmental heterogeneity alone did not control species diversity distribution, but species were affected by environmental filters at different stages in their life cycle, e.g. water proximity was significant for saplings and seedlings but not for adults. Anthropogenic disturbances act synergistically, e.g. trails played a key role in determining structure and tree diversity patterns. An important finding is that human disturbance diminishes species diversity in this TDF, but sites at lower elevations were more disturbed and less diverse. There is therefore a need to study how environmental factors would act if there were no anthropogenic disturbance. Full details of this study are presented by Williams-Linera and Lorea (2009).

Box 4.7 (cont.)

Table 1

p

et al.

Distance to urban centres

Distance to fires

Area Shape index

Fractal index

Core area Elevation

Cover 1–2 m (%) 0.04 0.24 0.45 0.60 0.53 0.38 0.23

Richness tree species 0.31 –0.08 0.17 0.37 0.42 0.12 0.28

Diversity tree species 0.27 –0.18 0.19 0.36 0.38 0.13 0.28

Richness shrub species –0.24 –0.10 0.17 0.33 0.33 0.11 0.31

Richness epiphyte species 0.29 –0.17 0.33 0.41 0.46 0.29 0.52

Diversity epiphyte species 0.29 –0.23 0.23 0.31 0.36 0.21 0.42

Richness seedlings 0.26 –0.17 0.10 0.37 0.50 0.05 0.38

Richness saplings 0.13 0.06 0.27 0.49 0.53 0.20 0.44

Richness class 5–15 0.20 –0.24 0.25 0.34 0.31 0.20 0.10

Density saplings 0.12 –0.02 0.60 0.61 0.51 0.54 0.18

Density class 5–15 0.40 –0.05 0.30 0.15 0.16 0.34 0.13

Density class 15–30 0.17 0.32 0.31 0.16 0.17 0.39 –0.09

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (149)

123

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

We found no differences between GLIMs with and without interactions among patch attributes and environmental variables as predictors of tree species richness, and there-fore we selected models without interactions, and the lower number of parameters for all the analyses presented. We found a significant effect of the study sites and the country on tree species richness patterns, but no effects of patch area and elevation (Table 4.2). We also found a significant effect of elevation and site, but not of patch area, on adult tree densities in remnant forest patches from all sites (Table 4.3). With regard to species richness of tree seedlings, we found significant effects of elevation, site and country, but not of patch area (Table 4.4). Finally, for the variation in tree seedling densities among forest patches, we found no significant effects of environmental or patch variables, and no effects of site or country (Table 4.5).

Table 4.2

p

Estimate Standard error T value Probability – ItI

Elevation –3.11 1.74 –1.78 0.07

Fragment area 0.85 0.47 1.81 0.073

Country 22.6 2.05 11.0 <0.001*

Sampling site –6.30 0.78 –8.04 <0.001*

Table 4.3

Estimate Standard error T value Probability -ItI

Altitude 0.0004920 0.0001830 2.688 0.009*

Area –0.0001411 0.0001286 –1.097 0.277

Country - - - -

Site 0.5517715 0.1830766 3.014 0.004*

Table 4.4

Estimate Standard error T value Probability -ItI

Altitude 9.224e–03 2.329e–03 3.960 <0.001*

Area 4.737e–04 6.211e–04 0.763 0.448

Country 4.587e+01 5.591e+00 8.205 <0.001*

Site 3.329e+01 4.881e+00 –6.820 <0.001*

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (150)

124

Principles and Practice of Forest Landscape Restoration

Table 4.5

Estimate Standard error T value Probability -ItI

Altitude 0.0017392 0.0018324 0.949 0.346

Area 0.0007402 0.0004885 1.515 0.135

Country 6.8388384 4.3976176 1.515 0.125

Site –7.0301651 3.8390390 –1.831 0.072

All of the analyses presented showed a significant or marginally significant (p = 0.05 to 0.1) effect of the study sites on tree species richness. In other words, results were dependent on the specific region of Mexico or Chile where dry forests occurred. Further, elevation was positively and significantly related to tree densities and species richness of tree seedlings in all sites. This means that patches occurring at higher elevations generally presented greater densities of trees and more abundant regeneration, but no significant trends in species rich-ness of adult trees were recorded. At the same time, patches at increased elevation had more species of tree seedlings, but not higher densities of tree seedlings, than lower elevation patches. In all of the analyses in both countries, the area of remnant patches had no effect on the species richness of trees or tree seedlings.

Studies in other forest types have attributed the lower species richness found at higher elevation to environmental stress factors such as harsher climate and infertile soils (Rahbek, 1995; 1997; Bachman et al., 2004; Smith-Ramírez et al., 2007). However, we are not aware of previous reports that documented increasing tree diversity with elevation. Potentially, chang-es in the duration of the dry season and moisture availability at different elevations could influence the species richness of seedling and adult trees. In Chilean coastal hills, oceanic fog frequently covers the mountaintops above 500–600 m elevation, which consequently receive significantly more precipitation than lowland areas (del-Val et al., 2006). This could positively affect the recruitment and survival of tree species, by increasing habitat and re-source heterogeneity. Furthermore, the patterns found for tree species in relation to altitude are not the same as those for herbs and shrubs. In Chilean dry-sclerophyllous forests, greater species richness of bulbs and herbs occurs in lowland areas, and hence, if all plant species (woody and non-woody) are included in the analyses, the highest number of species is found at low elevations (U Chile, 2007).

Because greater anthropogenic impact affects the vegetation of lowland areas in Chile (Armesto et al., 2010), Veracruz (Williams and Lorea, 2009; Box 4.8), and probably Chiapas (Neptalí Ramírez-Marcial, personal communication), it is likely that two important factors, al-titude and antropogenic impact, are acting together, or one may be masking the effects of the other. For example, in central Chile, the distance from forest fragments to cities and towns is negatively correlated with woody species richness (Becerra et al., unpublished manuscript). Future studies should elucidate the relative importance of both factors.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (151)

125

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Deforested dry forest landscape in Chiapas, Mexico. Photo: R. Vaca

Conclusions

One of the most important decisions with regard to restoration programmes is where restoration actions should be undertaken to obtain the best results, in terms of recovery rates of native vegetation cover and species richness. Key questions include: is the tree species richness greater in large than in medium or small forest fragments? What is the threshold patch size of remnant fragments that could sustain the highest tree species richness or density of regeneration? What physical factors of an individual site are im-portant to consider when selecting sites for restoration? Our research did not identify a standard answer to these questions that was valid for all dryland forests analyzed. Rather, the effect of site (country and province) was stronger than other effects, highlighting the importance of local context when identifying restoration priorities. Results suggest that restoration is likely to be more successful in terms of impact on species richness when restoration activities are conducted at higher elevation that in lowland areas, and when the size of the remnant fragments is relatively large. However, the conditions of each site must be analyzed separately. We conclude that environmental factors related to variation in altitude and other specific variables associated with disturbance history in each study area are important determinants of the diversity of adult trees and tree seed-lings in American dryland forests. Effects derived from local anthropogenic impact must be analyzed separately to fully understand the processes that account for the present patterns of tree species richness in each region.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (152)

126

Principles and Practice of Forest Landscape Restoration

References

Arias, M., Bianchi, A. 1996. Estadísticas climatológicas de la Provincia de Salta. INTA, Salta, Argentina.

Armesto, J.J., Martínez, J.A. 1978. Relations between vegetation structure and slope aspect in the mediterranean region of Chile. Journal of Ecology 66: 881–889.

Bachman, S., Baker, W. J., Brummitt, N., Dransfield, J., Moat, J. 2004. Elevational gradients, area and tropical island diversity: an example from the palms of New Guinea. Ecography 27: 299–310.

Balvanera, P., Aguirre, E. 2006. Tree diversity, environmental heterogeneity, and productivity in a Mexican tropical dry forest. Journal Biotropica. 38: 479–491.

Balvanera, P., Lott, E., Segura, G., Siebe, C., Islas, A. 2002. Patterns of beta-diversity in a Mexican tropical dry forest. Journal of Vegetation Science 13: 145–158.

Becerra, P., Smith-Ramírez, C., Echeverría, C., Armesto, J. 2010. Effect of landscape fragmenta-tion on plant communities in Central Chile. Unpublished manuscript.

Bennet, A.F. 2003. Linkages in the landscape. The role of the corridor and connectivity in wildlife conservation. IUCN, Gland, Cambridge.

Bianchi, A., Yáñez, C. 1992. Las precipitaciones en el noroeste argentino, Second edition. INTA, Salta, Argentina.

Bond, W.J. 1983. On alpha diversity and the richness of the Cape flora: a study in the southern Cape fynbos. In: Kruger, F.J., Mitchel D.T., Jarvis, J.U.M. (eds.), Mediterranean type ecosys-tems: the role of nutrients Inand. Springer-Verlag. New York: pp. 337–356.

Breedlove, D.E. 1981. Flora of Chiapas. Part I: Introduction to the Flora of Chiapas. California Academy of Sciences. San Francisco, USA.

Brown, A.D., Grau, H.R., Malizia, L.R., Grau, A. 2001. Argentina. In: Kappelle, M., Brown, A.D. (eds.), Bosques nublados del Neotrópico. Instituto Nacional de Biodiversidad, San José, Costa Rica: pp. 623–659.

Bullock, S.H., Mooney, H.A., Medina, E. 1995. Seasonally dry tropical forests. Cambridge Uni-versity Press, Cambridge, UK.

Burnham, K.P., Anderson, D.R. 2002. Model selection and multimodel inference: a practical-theoretic approach, Second edn. Springer,Verlag, New York.

Cabrera, A. 1976. Regiones fitogeográficas argentinas. Enciclopedia Argentina de Agricultura y Jardineria. Editorial Acme, Buenos Aires, Argentina.

Cadenasso, M.L., Pickett, S.T.A. 2001. Effect of edge structure on the flux of species into forest interiors. Conservation Biology 15: 91–97.

Connor, E.F., McCoy, E.D. 1979. The statistics and biology of the species–area relationship. American Naturalist 113: 791–833.

Debussche, M., Isenmann, P. 1994. Bird-dispersed seed rain and seedling establishment in patchy Mediterranean vegetation. Oikos 69: 414–426.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (153)

127

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Del Pozo, A.H., Fuentes, E.R., Hajek, E.R., Molina J.D. 1989. Zonación microclimática por efecto de los manchones de arbustos en el matorral de Chile central. Revista Chilena de Historia Natural 62: 85–94.

del-Val, E., Armesto J., Barbosa O., Chriestie D., Gutierrez A., Clive J., Marquet, P., Weathers, K. 2006. Rain forest islands in the Chilean semi-arid region: fog-dependency, ecosystem per-sistence and tree regeneration. Ecosystems 9: 598–608.

Di Castri, F., Hajek, E. 1976. Bioclimatología de Chile. P. Universidad Católica de Chile, Santiago.

Drinnan, I.N. 2005. The search for fragmentation thresholds in a southern Sydney suburb. Biological Conservation 124: 339–349.

Echeverría, C., Coomes, D., Salas, J., Rey-Benayas, J.M., Lara A., Newton, A. 2006. Rapid de-forestation and fragmentation of Chilean temperate forests. Biological Conservation 130: 481–494.

Forman, R.T.T., Gordon, M. 1986. Landscape ecology. John Wiley and Sons, New York.

Fuentes, E.R., Otaíza, R.D., Alliende, M.C., Hoffmann, A.J., Poiani, A. 1984. Shrub clumps of the Chilean matorral vegetation: structure and possible maintenance mechanisms. Oecologia 62: 405–411.

Gallardo-López, F., Riestra-Díaz, D., Aluja-Schunemann, A., Martínez-Dávila, J.P. 2002. Fac-tores que determinan la diversidad agrícola y los propósitos de producción en los Agr-oecosistemas del municipio de Paso de Ovejas, Veracruz, México. Agrociencia 36(4): 495–502.

Gentry, A.H. 1995. Diversity and floristic composition of neotropical dry forests. In: Bullock, S.H., Mooney, H.A., Medina, E. (eds.), Seasonally dry tropical forests, Cambridge University Press, Cambridge, UK: pp. 146–194.

Gillespie, T.W., Grijalva, A., Farris, C.N. 2000. Diversity, composition, and structure of tropical dry forests in Central America. Plant Ecology 147: 37–47.

Gordon, J.E., Hawthorne, W.D., Reyes-Garcia, A., Sandoval, G., Barrance, A.J., 2004. Assessing landscapes: a case study of tree and shrub diversity in the seasonally dry tropical forests of Oaxaca, Mexico and southern Honduras. Biological Conservation 117: 429–442.

Harris L. 1984. The fragmented forest: island biogeography theory and the preservation of biotic diversity. University of Chicago Press, Chicago.

Hersperger, A.M., Forman R.T.T. 2003. Adjancency arrangement effects on plant diversity and composition in woodland patches. Oikos 1001: 279–290.

Hobbs, R.J. 2001. Synergisms among habitat fragmentation, livestock grazing, and biotic inva-sions in southwestern Australia. Conservation Biology 15: 1522–1528.

Holt, R.D., Robinson, G.R., Gaines, M.S. 1995. Vegetation dynamics in an experimentally frag-mented landscape. Ecology 76: 1610–1624.

Honnay, O., Hermy, M., Choppin. P. 1999. Effects of area, age, and diversity of forest patches in Belgium on plant species richness, and implications for conservation and reforestation. Biological Conservation 87: 73–84.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (154)

128

Principles and Practice of Forest Landscape Restoration

Hueck, K. 1972. As florestas da América do Sul. Ecología, composiçaõ e importancia econom-ica. Universidade de Brasilia and Editora Poligono S.A. São Paulo, Brazil.

Huston, M.A. 2004. Management strategies for plant invasions: manipulating productivity, disturbance, and competition. Diversity and Distribution 10: 167–178.

Kemper, J., Cowling, R.M., Richardson, D.M. 1999. Fragmentation of South African renos-terveld shrublands: effects on plant community structure and conservation implications. Biological Conservation 90: 103–111.

Kitzberger, T. 1995. Fire regime variation along a northern Patagonian forest-steppe gradient: stand and landscape responses, Ph.D. thesis, Department of Geography, Univ. Colorado, Colorado, USA: pp. 1–203.

Kitzberger, T., Pérez, A., Iglesias, G., Premoli, A., Veblen, T. 2000. Distribución y Estado de con-servación del alerce (Fitzroya cupressoides (Mol.) Johnst.) en Argentina. Bosque 21: 79–89.

Laurance, W.F., Ferreira, L.V., Rankin de Merona, J.M., Laurance S.G. 1998b. Rain forest fragmen-tation and the dynamics of Amazonian tree communities. Ecology 79: 2032–2040.

Laurance, W.F., Gascon, C. Rankin de Merona, J.M. 1998a. Predicting effects of habitat destruc-tion on plant communities: a test of a model using Amazonian trees. Ecological Applica-tions 9: 548–554.

Laurance, W.F., Perez-Salicrup, D., Delamonica, P., Fearnside, P.M., D’Angelo, S., Jerozolinski, A., Pohl L., Lovejoy, T.E. 2001. Rain forest fragmentation and the structure of Amazonian liana communities. Ecology 82: 105–116.

Lawton, J.H., Bignell, D.E., Bolton, B., Bloemers, G.F., Eggleton, P., Hammond, P.M., Hodda, M., Holt, R.D., Larsen, T.B., Mawdsley, N.A., Stork, N.E., Srivastava, D.S., Watt, A.D. 1998. Biodiver-sity inventories, indicator taxa and effects of habitat modification in tropical forest. Nature 391: 72–76.

Lepš, S., Šmilauer, P., 2003. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, Cambridge, UK.

Lindenmayer, D.B., Franklin, J.F. 2002. Conserving forest biodiversity: a comprehensive multi-scaled approach. Island Press.

Lott, E.J., Bullock, S.H., Solís-Magallanes, J.A. 1987. Floristic diversity and structure of Upland and Arroyo forests of coastal Jalisco. Biotropica 19: 228–235.

Malizia, L.R. 2004. Diversity and distribution of tree species in subtropical Andean forest. Doctoral thesis. University of Missouri, St. Louis, USA.

Matlack, G.R. 1994. Plant species migration in a mixed-history forest landscape eastern North America. Ecology 75: 1491–1502.

Matteucci, S.D., Colma, A. 1982. Metodología para el estudio de la vegetación, Serie Biología, Monografía 22. Secretaría General de la Organización de los Estados Americanos. Programa Regional de Desarrollo Científico y Tecnológico. Washington, D.C.

Mcgarigal, K. and Marks, B. 1994, FRAGSTATS: Spatial Pattern Analysis Program for Quantify-ing Landscape Structure. USDA Forest Service, Pacific Northwest Research Station, Port-land, Oregon.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (155)

129

Fragmentation and altitudinal effects on tree diversity in seasonally dry forests of Mexico and Chile

Méndez, M., García, D., Maestre, F.T., Escudero, A. 2008. More ecology is needed to restore Mediterranean ecosystems: a reply to Valladares and Gianoli. Restoration Ecology 16: 210–216.

Miranda, F., Hernández-X., E. 1963. Los tipos de vegetación de México y su clasificación. Bo-letín de la Sociedad Botánica de México 28: 29–179.

Petit, R.J., Bialozyt, R., Garnier-Géré P., Hampe A. 2004. Ecology and genetics of tree invasions: from recent introductions to Quaternary migrations. Forest Ecology and Management 197: 117–137.

Prado, D.E. 2000. Seasonally dry forest of tropical South America: from forgotten ecosystems to a new phytogeographic unit. Edinburgh Journal of Botany 57 (3): 437–461.

Quinn, J.F., Harrison, S.P. 1988. Effects of habitat fragmentation and isolation on species rich-ness: evidence from biogeographic patterns. Oecologia 75: 132–140.

R Development Core Team, 2005. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3–900051–07–0. < http://www.R-project.org>.

Raffaele, E., Veblen, T. 1998. Facilitation by nurse shrubs on resprouting behavior in a postfire regeneration of matorral in northwest Patagonia, Argentina. Journal of Vegetation Science 9: 693–698.

Rahbek, C. 1995. The elevational gradient of species richness: a uniform pattern? Ecography 18: 200–205.

Rahbek, C. 1997. The relationship among area, elevation, and regional species richness in Neotropical birds. The American Naturalist 149: 875–902.

Richerson, P.J., Lum, K. 1980. Patterns of plant species diversity in California: relation to weather and topography. The American Naturalist 116(4): 504–536.

Rohde, K. 1992. Latitudinal gradients in species diversity: the search for the primary cause. Oikos 65: 514–527.

Rohde, K., Heap, M., Heap, D. 1993. Rapoport’s rule does not apply to marine teleosts and cannot explain latitudinal gradients in species richness. American Naturalist 142:1–16.

Rzedowski, J., 2006. Vegetación de México. 1ra. Edición digital, Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, México. 504pp.

Segura, G., Balvanera, P., Durán, E., Pérez, A. 2003. Tree community structure and stem mortal-ity along a water availability gradient in a Mexican tropical dry forest. Plant Ecology 169: 259–271.

Seibert, P. 1982. Carta de vegetación de la región de El Bolsón y su aplicación a la planifi-cación del uso de la tierra. Fundación para la Educación, la Ciencia y la Cultura, Buenos Aires, Argentina. 120 pp.

Simonetti, J.A., Moraes, M., Bustamante R.O., Grez, A.A. 2001. Regeneración de bosques tropi-cales fragmentados del Beni. In: Mostasedo, B., Fredericksen, T.S. (eds.), Regeneración natu-ral y silvicultura en los bosques tropicales de Bolivia. Bolfor, Santa Cruz: pp. 139–155.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (156)

130

Principles and Practice of Forest Landscape Restoration

Smith-Ramírez, C., Pliscoff, P., Díaz, D., Valdovinos, C., Méndez, M., Samaniego, H. 2007. Dis-tribution patterns and biodiversity hotspots of flora and fauna in the Coastal Range of Southern Chile: integrating natural history and GIS. Biodiversity and Conservation 16: 2627–2634.

Soulé, M. E., Alberts, A.C., Bolger, D.T. 1992. The effects of habitat fragmentation on chaparral plants and vertebrates. Oikos 63: 39–47.

Tabarelli, M., Mantovani W., Peres. C.A. 1999. Effects of habitat fragmentation on plant guild struture in the montane Atlantic forest of southeastern Brazil. Biological Conservation 91: 119–127.

Tejedor, N., 2007. Preliminary assessment of the structure and composition of Chiapas’ dry forest in response to anthropogenic disturbance. MSc thesis. School of Conservation Sci-ences, Bournemouth University, Poole.

Trejo, I., Dirzo, R. 2002. Floristic diversity of Mexican seasonally dry tropical forests. Biodiver-sity and Conservation 11: 2063–2048.

U Chile, 2007. Profundización de la línea de base ambiental y ecológica del sector de mayor valor ecológico del cordón de Cantillana. Elaborado para CONAMA-GEF-PNUD.

White, D.A. Hood, C.S. 2004. Vegetation patterns and environmental gradients in tropical dry forests of the northern Yucatan Peninsula. Journal of Vegetation Science 15(2): 151–161.

Williams-Linera, G., Lorea, F. 2009. Tree species diversity driven by environmental and anthro-pogenic factors in tropical dry forest fragments of central Veracruz, Mexico. Biodiversity and Conservation 18: 3269–3293.

Willson, M. F., Morrison, K., Sieving, K.E., De Santo, T.L., Díaz I., Santisteban, L. 2001. Predation risk to bird nests: patterns in a Chilean agricultural landscape. Conservation Biology 15: 447–456.

Woodward, F. 1987. Climate and plant distribution. Cambridge University Press, Cambridge. 158 pp.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (157)

131

5 EXPERIMENTAL ANALYSIS OF DRYLAND

FOREST RESTORATION TECHNIQUES

G. Williams-Linera, C. Alvarez-Aquino, A. Suárez, C. Blundo, C. Smith-Ramírez, C. Echeverria, E. Cruz-Cruz, G. Bolados, J.J. Armesto, K. Heinemann, L. Malizia, P. Becerra, R.F. del Castillo, R. Urrutia

Introduction

Forest landscape restoration in dry forest in the Americas is an urgent priority. Tropical, subtropical, and temperate seasonally dry forests in Mexico, Chile, and Argentina are rep-resented by a wide range of different forest types from Mexican tropical dry deciduous forest, Chilean sclerophyllous and deciduous dry forest, Argentinean subtropical seasonally dry forests – including Andean premontane forest – and the transition to dry Chaco forest, to the forest-steppe ecotone on the eastern slopes of the Patagonian Andes. What all these forests have in common is that they are largely threatened by deforestation (Chapter 2), the establishment of plantations of exotic tree species, overharvesting (particularly wood for fuel), conversion to cropland, and cattle-raising. Worldwide, dryland forest ecosystems have been degraded by unsustainable land-use practices that may alter the structure and composition of forest stands, as well as reduce tree density and the extent of canopy cover. These effects can have serious negative impacts on the ecological processes influenc-ing forest dynamics, including seed dispersal, seed germination, seedling establishment, and growth. As a result, the ability of forests to regenerate naturally can be significantly impaired, reducing the viability of forest patches and threatening the provision of envi-ronmental services to local communities as well as the biodiversity associated with dry-land forest ecosystems. The development of successful restoration approaches depends on understanding the capacity of forest regeneration and how different human activities influence it.

Ecological restoration is increasingly adopted as an approach to land use in areas that have suffered ecological degradation as a result of human impact (e.g. Lindenmayer and Franklin, 2002; Lamb and Gilmour, 2003; Mansourian et al., 2005). Rey Benayas et al. (2009) sum-marized those types of human activity resulting in degraded ecosystems and the forms of restoration action that are currently being undertaken to address this problem. Restoration actions typically focus on the reduction or removal of factors causing environmental degra-dation and/or the re-establishment of key ecosystem components to influence the rate and direction of recovery. The simplest approach is passive restoration or cessation of the causal action, accompanied by recovery through natural processes, while other active measures include activities such as tree planting (Rey Benayas et al., 2009).

The development of successful restoration approaches depends in part upon understand-ing the capacity of forest stands to regenerate naturally and how this process is influenced by different human activities. The main approaches used for forest restoration include en-couraging natural regeneration and artificially establishing trees within or around degraded

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (158)

132

Principles and Practice of Forest Landscape Restoration

forest stands. In situations where forests are so degraded that natural regeneration is in-adequate, such artificial establishment methods may be preferable. A successful outcome from both approaches depends on the understanding of successional processes in dryland forest communities (Quesada et al., 2009) and the factors influencing the establishment and growth of tree seedlings in degraded sites.

Recent research on drylands has focused on investigating restoration from natural regen-eration (Aronson et al., 2005). The factors that influence succession must be understood in order to capitalize on natural regeneration mechanisms (Walker et al., 2007). Griscom et al. (2009) included in their recommendations for the early stages of forest succession the exclusion of cattle, making site-specific decisions about herbicide application, and the active conservation and protection of riparian zones that function as a critical source of diverse propagules. Vieira and Scariot (2006a) considered the ecology of tropical dry forest regeneration as a tool to restore disturbed lands. Seed collection, planting time, growth of established seedlings, and re-sprouting ability as a prominent mechanism of regeneration in dry forests are factors that must be taken into consideration. With respect to natural regen-eration, studies have examined soil seed banks as a source of propagules (Uasuf et al., 2009), seed rain (Ceccon and Hernández, 2009), and seed fate by desiccation or insect predation in abandoned dry forest pastures (Vieira and Scariot, 2006a). However, the number of seed bank species and the quantity of soil-stored seeds are typically relatively low, requiring the natural regeneration process to be assisted through direct seeding, seedling plantation, and the manipulation of a site to improve the environmental conditions for seedling establish-ment and growth (Uasuf et al., 2009).

Enrichment planting has been assessed as a potential reforestation tool that could comple-ment natural regeneration (Griscom et al., 2005) along with seedling plantation in logged forests (Vieira et al., 2007). Forest succession may be accelerated and planting techniques improved if the use of fertilizer, mycorrhizal inoculation, irrigation, and herbicides is consid-ered. Obviously, cattle and other livestock must be removed prior to enrichment planting (Gricom et al., 2005; Montagnini, 2005).

Other restoration efforts have studied the growth and productivity of native species in pure and mixed plantations, comparing them with the exotic species that are widely used, such as Tectona grandis in Mexico and Central America. There is a consensus that the use of mixed plantations with native species is always preferable because of its contribution to sustainable management; single-species plantations are generally of lower biodiversity value and do not provide as great a range of goods and services as natural forest, although mixed plantations are likely to increase this range of benefits (Piotto et al., 2004; Montagnini, 2005).

Several different approaches have been taken for the restoration of dryland systems. The use of living fence species as a restoration tool has the advantage of planting tree species vegetatively; species can act as seed recruitment foci by attracting seed dispersers and pro-vide shade to improve micro-climatic conditions for seedling establishment (Zahawi, 2005). The use of precipitation pulses is essential for dryland regeneration. The increasing ability to predict El Niño-Southern Oscillation (ENSO) effects can be used to enhance management strategies for the restoration of degraded ecosystems (Holmgren et al., 2001). Since dryland ecosystems are dependent on rainfall pulses for their regeneration, understanding the com-plex effect of ocean conditions may be critical for their management and ecological restora-tion (Caso et al., 2007).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (159)

133

Experimental analysis of dryland forest restoration techniques

Planted Maytenus spp. seedling in Chile. Photo: C. Echeverria

Nursery tree in Chile. Photo: C. Echeverria

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (160)

134

Principles and Practice of Forest Landscape Restoration

Selection of species for restoration activities is an issue of vital importance. The preferenc-es of local people should be taken into account when planning restoration (see Chapter 5; Montagnini, 2005; Suarez et al., submitted). Garibaldi and Turner (2004) suggested that we restore not only landscapes but also the diversity-enhancing capabilities of the human com-munities inhabiting those landscapes. They proposed the identification of cultural keystone species that play more than one role; often these roles are supported and enabled by other non-keystone species. Cultural keystone species may play a paramount role in restoration. In addition, restoring forest remnants may increase their valorization by local land users. A non-random pattern of forest degradation has been identified, and there is a risk of potential loss of the most degraded forest remnants unless active forest restoration plans are applied (Tarrasón et al., 2010).

This chapter describes a series of field experiments and surveys that were undertaken in order to identify the main constraints to the establishment and growth of tree species in degraded forest stands, and to determine how these constraints may be overcome through practical management interventions (see also Boxes 5.1–5.11 for associated investigations). The experiments examined natural and artificial establishment of tree species subjected to a variety of different management approaches, with emphasis on tree species of conservation and/or socioeconomic importance to improve the value of forest resources for local com-munities. The objectives of this research were as follows: (1) to test forest restoration and land reclamation techniques for reversing degradation and loss of dryland forest ecosystems through a programme of field surveys and experiments established in the study areas, (2) to identify the key ecological processes limiting establishment and growth of threatened and/or socioeconomically important native tree species on degraded forest sites, and (3) to identify appropriate methods of restoring dryland forest ecosystems that contribute both to the conservation and restoration of biodiversity and the economic development of local communities. This chapter presents the results of the programme of field experiments that were established in each study area, and identifies the key ecological processes limiting establishment and growth of threatened and/or socioeconomically important native tree species on degraded forest sites. A discussion is then presented of the implications of these results for the identification of appropriate methods for restoring dryland forest ecosystems that contribute to the conservation and restoration of biodiversity as well as to the economic development of local communities.

In the ReForLan project, two practical approaches to forest restoration were stud-ied: (i) encouragement of natural regeneration, and (ii) artificial establishment of tree species within degraded/deforested stands. Various complementary restoration experi-ments were conducted in six dryland regions in Latin America, which permit a broad overview of the challenges that practical ecological restoration must confront. Other studies permit exploration of aspects related to active and passive restoration in dry-lands. Experimental approaches differed between the study areas in accordance with contrasting local circ*mstances. Together, the different approaches presented here are key to identifying solutions to the problem of degraded drylands in Latin America. The experimental areas were located in three countries: Chile (Central Valley and Coastal Range), Argentina (northwestern and Patagonia), and Mexico (Veracruz and Oaxaca). Other studies relating to restoration activities (not on plantations) were carried out in some of the areas mentioned above and in Chiapas, Mexico. General characteristics of each region are presented in Chapter 1.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (161)

135

Experimental analysis of dryland forest restoration techniques

Box 5.1 Holistic ranching and landscape restoration in Chiapas, Mexico

B. G. Ferguson and R. Alfaro Arguello

The widespread conversion of tropical forest to pasture was perhaps the most significant land use change in Latin America during the second half of the twentieth century (Kaimowitz, 1996). Forest restoration therefore typically takes place within landscapes dominated by cattle ranching and its success depends upon its compatibility with ranch management. However cattle ranching in the American tropics has been based largely on grass monocultures, a production model that threatens tree cover at the pasture and landscape scales through its inefficient, extensive nature and dependence upon fire and herbicides (Villafuerte et al., 1997; Savory et al., 1999; Sánchez et al., 2003; Szott et al., 2000; Roman-Cuesta et al., 2003; Vieira and Scariot, 2006b). Soil and pasture degradation and biodiversity loss increase dependence on herbicides, pesticides, fertilizers and feed supplements, reducing profit margins and driving pasture expansion.

To break this vicious circle, a small group of ranchers in the Central Valley of Chiapas turned to the holistic management (HM) decision-making framework (Savory et al., 1999). HM focuses upon relationships among the land, people and their communities and is designed to confront the challenges of land management where humidity is markedly uneven throughout the year, as in areas of tropical dry forest. In 1994, these ten ranchers formed an ‘Intensive, Technical Grazing’ club (‘PIT Las Villas’). In 2007 we documented the advances of the seven remaining members by comparing their management, plant communities and soils to those of 14 neighbouring ‘conventional’ ranches (Alfaro Arguello, 2008; Ferguson et al., in review).

Key management elements we observed on each of the holistic ranches include:

Numerous pasture divisions created using electric and living fences;

Frequent rotation (at least daily) of cattle among pasture divisions, maintaining high stocking density and adequate recovery periods;

Complete or almost complete elimination of use of fire, herbicides, pesticides and chemical fertilizers;

Manual weeding eliminating only plants that cattle do not consume or that might scratch their udders;

Maintenance of forest reserves several ha in extent;

Diversification of forage resources with trees and shrubs such as Guazuma ulmifolia, Pithe-cellobium dulce and Enterolobium cyclocarpum, grasses and herbaceous legumes;

Diversification of ranch products (e.g. pigs, sheep, poultry and honey, breeding stock and sem*n, and timber from rotational harvests);

Good treatment and adequate spaces and installations for livestock;

Commitment to constant experimentation, learning and rancher-to-rancher capacity build-ing; and

A planning process oriented toward quality of life and community well-being as well as to productivity and profits.

In comparison with their conventional neighbours holistic ranches have achieved:

Greater milk productivity;

Reduced cow and calf mortality;

Diminished dependence on purchased inputs, including agrochemicals, feed and hay;

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (162)

136

Principles and Practice of Forest Landscape Restoration

Better grass cover;

Deeper topsoil; and

Higher soil microbial respiration rates and increased earthworm presence.

We also observed (but did not measure) forest regeneration on the hills surrounding holistic pastures as well as a (non-significant) tendency toward more trees in holistic than conventional pastures. Thus HM boosts ranch productivity at the same time that it fosters tree cover on ranches and surrounding landscapes. HM and other production models based upon agroecological principles are essential elements of ecological restoration of tropical dry forest landscapes. Reforestation efforts that do not address the underlying drivers of deforestation will not succeed beyond the short term. We hope that the experience of PIT Las Villas will encourage other ranchers, large and small, to try HM, and will encourage government support for such strategies. Further investigation will be necessary to identify appropriate technology, training and support mechanisms.

Box 5.1 (cont.)

Box 5.2 The role of cattle in tropical dry forest regeneration in Chiapas, Mexico

B. G. Ferguson, M. Rueda Pérez, G. Pascacio Damián, L. Domínguez Morales, P. Bichier

Lack of seed dispersal is one of the most common barriers to regeneration of neotropical forests (Holl, 1999). Intriguingly, cattle, a familiar feature of many neotropical landscapes, can move among vegetation types, consuming and dispersing seeds of woody plants (Janzen and Martin, 1982; Miceli-Méndez et al., 2008). Miceli-Méndez and colleagues (2008) identified 13 cattle-dispersed tree and shrub species in Chiapas, mostly in the seasonally dry tropics. They proposed that seed dispersal by livestock holds potential for management as a forest restoration tool, but emphasized the need for a better understanding of the ecology of the phenomenon. We explored three aspects of the role of livestock in dry forest succession: the seed banks present in cattle dung, the role of cattle-dispersed species as nurse trees, and the population structure of one cattle-dispersed species in active pastures.

For the seed bank study, we sieved cattle dung and soil samples from 14 ranches in the Villaflores and Villacorzo municipalities. We found greater diversity and density of tree seeds in manure than in pasture soils. Tree species dominating the manure seed bank included Guazuma ulmifolia (Sterculiaceae), Ficus spp. (Moraceae) and the legumes Enterolobium cyclocarpum and Acacia spp. Density of herbaceous seeds, however, was higher in soil than in manure. We suspect that these smaller, softer seeds do not easily survive passage through the bovine digestive tract.

We worked on seven ranches in the same area to quantify seed dispersal and micro-climate beneath two cattle-dispersed tree species (Guazuma ulmifolia and Pithecellobium dulce) and in open pasture. During a full year of seed trapping with 36 m2 of traps we recovered 55,832 seeds belonging to 173 morphospecies. For both species, seed rain was significantly greater beneath the canopy than in the open. These differences were particularly marked for tree seeds and for seeds of species dispersed by animals that consume their fruit. Seed rain of half a crown radius outside of tree crowns was not significantly greater than that detected in the open. Micro-climate was significantly cooler and darker under tree crowns. A complementary bird survey recorded 104 species using the ranch landscapes (including pastures, riparian vegetation, fence lines and forest patches), including 30 species observed visiting isolated G. ulmifolia and/or P. dulce.

We documented the population structure of trees in five pastures near the village of Ocuilapa in the Ocozocoautla municipality. Acacia pennatula, a cattle-dispersed species, was present in all five pastures and was by far the most abundant of the 25 tree species we encountered. The

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (163)

137

Experimental analysis of dryland forest restoration techniques

species accounted for 54% of the 55+33 trees/ha (diameter at breast height >10cm) in our pasture census. In our seedling and sapling plots (diameter at breast height <5 cm), we found 0.93+0.45 individuals/m2, of which 99% were A. pennatula. Ranchers reported at least one use for 80% of the tree species in their pastures. Acacia pennatula is used for fodder and shade for livestock, post and construction wood and, most importantly, for fuelwood.

Taken as a whole, these three studies demonstrate that: cattle are effective dispersers of several dry forest tree species that are common in pastures; birds frequently visit these trees within the agricultural landscape; cattle-dispersed trees of modest size act as foci for seed dispersal and as nuclei for tree seedling establishment; and ranchers value and protect some cattle-dispersed trees. In effect, these are autochthonous silvopastoral systems that arise from local rangeland management practices. Understanding the functioning of these systems will contribute to novel strategies for dry forest restoration through livestock management. Furthermore, recognition of these agroforestry systems as such may justify changes in natural resource management policies that sanction and promote these systems as elements of sustainable cultural landscapes.

Figure 1 Acacia pennatula

Box 5.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (164)

138

Principles and Practice of Forest Landscape Restoration

Seedling monitoring in central Veracruz, Mexico. Photo: C. Alvarez

Prehispanic mound in Acazonica, Mexico. The tropical dry forest of central Veracruz has numerous re-mains of Prehispanic settlements (600 to 1500 AD). Photo: G. Williams-Linera

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (165)

139

Experimental analysis of dryland forest restoration techniques

Box 5.3 Evaluation of commercial plantations and their management in the tropical dry forest region of Paso de Ovejas, Mexico

R.A. Pedraza Pérez

In Mexico, there are two programmes administered by the National Forestry Commission (CONAFOR) to gain access to government incentives and to establish trees on plantations. One of these programmes is ‘Pro Árbol’ previously known as PRONARE (the Spanish acronym for National Programme of Reforestation 1992–2001), which distributes trees that are propagated in nurseries and sometimes provides other support services relating to reforestation and the maintenance and protection of reforested areas. The other programme is known as PRODEPLAN (the Spanish acronym for Programme for Commercial Plantation Development) that began operations in 1997 and was redesigned in 2001. The main objective of PRODEPLAN is to encourage production of inputs to provide the forest industry with competitive prices in addition to generating jobs and decreasing the stress on natural forests. Such support includes the following aspects of management programmes: establishment, maintenance, insurance premiums, technical support, and development. The plan also finds sources of financing for the cultivation and management of forest species in agricultural lands that have lost native vegetation, leading to the production of timber and non-timber raw materials for commercialization or industrialization. The native species from tropical zones that are most heavily used are mahogany (Swietenia macrophylla) and cedar (Cedrela odorata); the exotic species are teak (Tectona grandis), white beech wood (Gmelina arborea), and pink cedar (Acrocarpus frainifolius), among others.

The main objective of this study was to determine the number and kind of plantations established in the tropical dry forest of central Veracruz, Mexico based on the official census list. Ten cases were selected in order to ascertain the farmers’ purposes; the performance of selected species and established trees was recorded in terms of survival, height, and diameter as well as the commercial wood volume. Ten plantation cases were selected on the basis of the following criteria: (a) tree life form, (b) plantation of at least one hectare, and (c) individual trees planted close together. All trees found on ten 10 x 10 m plots were measured at each site. Empty spaces and stumps were counted and additional information was requested from the landowners. Age and density of plantations were obtained. Survival, total and commercial height, and mean annual increment (MAI) among sites and species were compared.

From 119 recorded cases between the years 2000 and 2006, only 39 met the selection criteria, but they represented 70% of the total planted surface (around 1100 ha). This means that few or scattered trees were planted in most cases, the preferred species being Casuarina equisetifolia, Cedrela odorata, and Swetenia macrophylla. Insufficient tree plantations resulted, forcing us to use cases data from other years to complete ten study cases. The main support from CONAFOR was seedlings grown in their official nurseries, while PRODEPLAN gave 1.5 million pesos to seven plantations established on 208 ha. Some 75% of the total trees studied (684) belonged to Cedrela odorata, which is renowned for its natural beauty and outstanding physical properties. That is the reason for their removal from tropical forest and for the species’ present endangered status. Their establishment in tree plantations is difficult because of Hypsipyla grandella (Zeller) larval attack (Mayhew and Newton, 1998). Tree individuals of Tabebuia rosea (13%), T. donnell-smithii (9.4%), and Swetenia macrophylla (3.5%) were found on three of the plantations studied.

The ten selected sites were very different from each other due to variation in plantation age, tree density, and owners’ objectives. Tree density per hectare reflected landowners’ objectives upon establishment of the plantation and provided useful criteria for classifying study sites: (1) agro-systems (185 trees/ha) on flat terrain combined with lemon trees (Citrus limon); watering and fertilization were permanent on Don Rene (10 ha) and Casa Blanca (16 ha), established three and six years ago, respectively; (2) secondary vegetation under low density tree plantation (216 trees/ha) on the slope of Guaje Mocho (1 ha), established seven years ago; (3) multipurpose mixed plantation with middle and high densities, irrigation frequent, Paso de Varas (556 trees/ha), La Guadalupe (625 trees/ha) and Loma Coyotes (1666 trees/ha), 3, 30, and 4 ha, established three, five, and 12 years

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (166)

140

Principles and Practice of Forest Landscape Restoration

Box 5.4 Terraces to reforest degraded lands in Oaxaca, MexicoE. Cruz-Cruz

In Mexico, dry farming began around 5000 B.C. in the Oaxaca-Puebla region (Flannery, 1983). Pressure from the growing population led to vast forested areas, from hills and slopes, to be cleared for farming purposes, accelerating the rate of erosion. This was a major phenomenon until A.D. 1000–1530. In the Mixteca region in Oaxaca state, farmers developed the lama bordo system for growing annual crops, which fell into disuse after the Spanish conquest (Flannery, 1983). Since that period, the rate of erosion has been 10 mm year–1 (Kirkby, 1973). Soil erosion and degradation were accelerated owing to changes in the agricultural systems, abandonment of cropland (maintenance for the traditional lama bordo terraces was stopped), and introduction of goat and sheep herds by Spanish colonists. Populations of sheep and goats steadily increased after 1540, but the rate of increase was higher in the middle of the seventeenth century. This tendency changed because of the following factors: (a) Spain demanded more food; (b) the population of native people declined to its lowest level, and croplands were abandoned while shrub communities grew; and (c) the Spaniards were more interested in raising goats and sheep on those abandoned lands (Romero, 1990).

Disturbances such as deforestation, farming, and grazing, changed plant communities, ranging from a matrix of wide completely denuded areas to occasional small pristine patches. Some of the latter plant communities still exist in remote areas of the Mixteca (Fig. 1). Currently, rangelands in Oaxaca are overgrazed due to continuous and heavy grazing, intensive herd management (everyday household-rangeland-household) and a high stocking rate, which is about three times higher than the rangeland’s capacity. Such extreme overgrazing constitutes one of the most important worldwide examples of land degradation that has become a permanent and almost irreversible process. Because of social and economic factors the problem is highly complex.

ago, respectively; (4) conventional density (1111 trees/ha), La Covadonga (six years old) with 62 ha where natural forest was cut before, and Cascajal, La Gloria and Palo Verde (ten years old with 1–5 ha), the three sites were used for agriculture and cattle husbandry. La Covadonga was the only plantation that received financing for its establishment; today it is for sale.

Mean percent survival was high (70%); agro-systems had the highest (100%) survival rate because they replaced dead trees in the first year. Plantations that were ten or more years old had the lowest (16–43%); farmers delayed tree cutting to reduce competition among them. The total height of Mexican cedar increased with plantation age, from 0.89±0.85 m (3 yr) to 6.70±2.4 m (6 yr) and up to 13.3±1.44 m (12 yrs), respectively. Don Rene (3 yr) was 8.3±1.53 m, the product of intense management. In this area, trees did not show evidence of Hypsipyla attack. However, all other plantations did, resulting in reduced commercial height. The diameter followed the same tendency as height: the largest measurements were for Don Rene (14±1.3 cm), Casa Blanca (14±2.4 cm), and Guaje Mocho (18±2.8 cm), all with low plantation density. The highest mean annual increment (MAI) for height and diameter was measured on Don Rene (2.8 m and 4.8 cm annually) and Guaje Mocho (1.58 m and 2.6 cm). The commercial volume of those plantations was greater than that estimated by the business plan for Mexican cedar on plantations established in wet forest zones.

In conclusion, landowners preferred tree forest plantations used for commercial purposes. Most landowners selected Mexican cedar, a native with relatively fast growth, which is facilitated by watering, low densities, and the growth of secondary vegetation, as seen in Guaje Mocho. We recommend technical support for improved management as well as thinning; thus, the wood trade could yield more products and benefits. It is important to emphasize that producers in this region have been farmers for many years and that forestry is an activity that requires up-to-date technical and complex knowledge in order for it to continue developing.

Box 5.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (167)

141

Experimental analysis of dryland forest restoration techniques

Figure 1

The land degradation gradient commonly looks like a pattern of successive concentric circles. Soil erosion is severe near the villages and reduces gradually farther way. Local inhabitants believe that some native plants are disappearing in those areas because of goat grazing. However, other factors are likely to be more important in driving this process, such as: high animal population, mixed herds (improved forage utilization), sedentary pastoral system, age of herders (children and elders travel shorter grazing routes), non-favourable physiographic characteristics of the grazing areas and continuous extraction of firewood.

Several soil factors including low moisture retention capacity, poor content of nutrients and organic matter, common presence of a cemented top layer and low aggregate stability, together with a scarce and low precipitation regime, an extended dry season (7 months) and the steep slopes, constrict the success of reforestation. Soils have lost their potential for production and it is probable that a state change has occurred. A threshold has been crossed in these areas. Thus, the integrity of the ecological processes has changed, and the natural stands have different potential sets of plant communities (Strigham et al., 2003). The upper and most fertile soil layers have been removed and the infertile and cemented C horizon is exposed. On the hillsides, the gullies dominate the landscape and the steep conditions reduce the possibilities for plant growing and establishment (Fig. 1).

Reforesting degraded lands

Restoration programmes in the Mixteca region were started by the Comisión del Balsas in 1947 and the Comisión del Papaloapan in 1960 (Ruiz, 1996), and were based on watershed management. Both agencies applied widespread conservation practices, these included compressed reforestation (mainly with exotic species), terracing (bank type, narrow base), exclusion of domestic animals (cattle, goat, sheep and horse), and social organization. Success was very limited because reforestation was not adequately planned to reduce soil erosion, the planted species were exotic and ecologically not adapted, and the plantation method was inappropriate and depleted plant survival.

Peasants in the Mixteca region are used to planting forest trees and they do so to create ‘green zones’, but without a particular conservation aim (i.e. rather their objectives are erosion control, landscape improvement, springs protection and recharge, forage production). In order to delineate a collective goal, local people must define: (a) the purpose of reforesting; (b) what

Box 5.4 (cont.)

A B

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (168)

142

Principles and Practice of Forest Landscape Restoration

species to choose from those appropriate according to the objectives; (b) choice of plantation method to ensure plant survival and growth; (c) social and financial organization for watching and treating disturbance agents; and (d) the future use of the plantation. There is also a need to diversify species used and to favour native ones, which have more probability of success in the local harsh and marginal conditions than exotic species.

In order to assure a high percentage of plant survival when reforesting degraded soils, it is necessary to increase moisture availability and soil retention. In Oaxaca, two plantation systems (trench-bank, common hole) resulted in an increase in moisture around the root system. If reforestation with terraces is selected, expensive costs must be prevented and there is a need for ease of construction and low labour requirements for maintenance. Terraces are useful to control soil erosion, increase infiltration and reduce the slope length. The runoff can be intercepted and kept at the bottom of the channel (Fig. 2). According to field evaluations, the levelled curve terraces increase the moisture content in the soil between 3 to 4%.

Terraces are built following the contour curves and the distance among them must be determined on the basis of the land form and slope. To facilitate the construction of terraces, users can use a deep plough with an agricultural tractor following a line, which was traced previously. The furrow depth should be of 50 to 60 cm. Plants are to be planted at the slope of the bank. The distance between plants depends on the plant life form: 2 m for shrubs and 4 m for trees. Both groups of species may be mixed by including a shrub between two individual trees.

Figure 2

Box 5.4 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (169)

143

Experimental analysis of dryland forest restoration techniques

Box 5.5 Replacement of a forest stand of exotic species by native plants in the dryland landscape in central Chile

C. Echeverria and G. Bolados

The Chiletabacos factory, a British American Tobacco operator, is located within the Casablanca valley, an area increasingly famed for its vineyards and white wine production in the coastal zone in central Chile. As a result of urban and agriculture expansion there is very little dryland forest remaining in the valley. Chiletabacos (CT) owns 70 ha of land that was planted with Eucalyptus globulus about 20 years ago. The trees have no commercial value but serve as a useful buffer that screens the factory from the local community. This community uses the Eucalypt plantation as an area for grazing cattle and horses. At present, CT relies on three wells situated within the plantation for its water consumption. The company has faced serious water shortfalls in recent years and future problems with the existing wells could prove to be very costly. The restoration to native forest will better secure this ecosystem service for the future as well as reducing the risk of fire. On the other hand, the replacement of Eucalyptus by native species will favour biodiversity conservation and will enhance forest connectivity in this highly fragmented forest landscape.

A long-term restoration experiment of 1.6 ha was initiated in 2006 to investigate the site conditions required to restore the eucalypt plantations to native species in a dryland landscape. Within the Eucalyptus plantation, three different site conditions exist: (i) Acacia caven shrubland (Acacia site), (ii) open site dominated only by herb species (open site), and (iii) site cover >50% by Eucalyptus trees (Eucalypt site). On each site, we established six plots of 10 seedlings for each species selected: Quillaja saponaria, Colliguaja odorifera, Maytenus boaria and Baccharis linearis. On the Acacia site, 20 Eucalyptus trees that were growing between the Acacia trees were cut down and transported out of the area. Similarly, some trees were cut down in the interior of the Eucalyptus stand in order to create gaps for the establishment of the native plants. Plants were established in July 2006. Following the planting, a monitoring programme was established to measure survival every month. Also, air temperature, soil temperature and relative humidity were measured every two weeks to determinate the interespecific variation and relate the survival of the species to the climate of the area. The survival of the species was analyzed by fitting survival curves for each site condition. This was conducted using the Survival module of the statistical package R project and the Kaplan Meir distribution. The survival curves were fitted considering the species as the explanatory variable and the month of death as the response variable. Significant differences in survival were analyzed using a statistical test at 95% of significance with census data (living plants were recorded at the last sampling). A total of 32 months of survival monitoring until February 2009 was analyzed in this study.

The Acacia site presented the highest rate of survival with 52%, followed by the Eucalypt site with 47%. The survival rate at the open site reached up to 17% (Fig. 1). All of the sites exhibited a decline in the survival curves during the first summer, which was characterized by a long dry season (between months 5 and 8 in Fig. 1). In particular, the survival at the open site decreased from 90% to 57% (Fig. 1). From month 11 onwards, survival was almost constant until the last measure. In the case of Acacia and Eucalypt sites, the mortality rates were of 13% and 18% respectively during the whole study period.

At the species level, B. linearis and Q. saponaria registered the highest survival rates on the Acacia site, while on the Eucalypt site C. odorifera was the species that exhibited the highest survival along with B. linearis (Table 1). On the other hand, M. boaria presented the lowest percentages of survival for all the sites (Table 1), especially on the open site where any plant survived. The high rate of mortality on the open site was associated with the highest maximum temperature and solar radiation. In the Eucalypt site, the shade favoured the survival of the plants by decreasing the maximum temperature; however, it is probable that the allelopathy of the Eucalypts may have had negative effects on plant survival. On the Acacia site, the low level of mortality for all the species might be related to the positive effects of nurse plants provided by the A. caven that provides shade, humidity and nutrients to the seedlings.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (170)

144

Principles and Practice of Forest Landscape Restoration

Our results reveal the potential for restoration of Acacia caven shrubland, which covers a large area in central Chile. Most of this vegetation is severely degraded and restoration efforts should be undertaken in the near future. B. linearis and Q. saponaria are key species that may play an important role in recovering ecosystem functions. In case of restoring deforested sites in drylands, it is necessary to adopt some silvicultural techniques (both in the nursery and in the field) to reduce the mortality of plants in summer.

Figure 1 Eucalyptus

Table 1

Box 5.5 (cont.)

Species Acacia site Open site Eucalyptus site

Baccharis linearis 66.7 46.7 57.5

Colliguaja adorifera 46.7 3.3 40.0

Maytebus boaria 36.7 0.0 12.5

Quillaja saponaria 73.3 10.0 37.5

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (171)

145

Experimental analysis of dryland forest restoration techniques

Box 5.6 Early secondary succession as passive restoration in initial stages of ecological restoration of tropical dry forest

G. Williams-Linera and E. Ascension Hernández

Although the processes of tropical dry forest recovery (TDF) are still largely unknown, there has been a recent growth in interest in secondary succession for ecological restoration. Secondary succession research would provide information on key ecological procedures and species able to enhance forest recovery. Passive restoration of TDF, i.e. the regeneration of natural forest after agricultural land is abandoned, may be used in an initial stage before active restoration takes place. Whether from an ecological or a conservation perspective, assessing the recovery of TDF needs to include an evaluation of changes in functional groups of species according to their successional status and seed dispersal modes. The objective of this study was to assess vegetation structure, woody species composition, diversity, and successional status as well as the dispersal mode of the woody species at very early successional sites.

The study area is located in the tropical dry forest region of central Veracruz, Mexico. In this area, fallow periods are short, in general no longer than 7–10 years, and old successional sites are scarce or absent. We learned that productive activity resumes on fallow land after eight years, making it practically impossible to find older successional stages. We selected five early secondary successional sites between seven to 72 months after abandonment (Table 1). Fallow age and land-use history were determined based on information from local inhabitants. Although the sites were not continuously used or burned every year, they have gone through a series of repeated intermediate short fallow periods that were not recorded.

At each successional site, all woody vegetation was surveyed in sixteen 5 x 5 m plots adjacent to the restoration experiments described in this chapter. Total basal area and density at early successional sites was 0.40 to 3.88 m²/ha, with 900 to 5450 individuals/ha. Mean height varied from 1.0 to 2.1 m (Table 1). Community diversity was evaluated as richness (number of species, S), Fisher’s alpha, and Shannon’s diversity index (H). Forty-five woody species were recorded at the early successional sites. Richness varied from eight to 21 woody species per site (Table 1). The relative ecological importance of each species at each site was expressed as an importance value index (IVI) calculated by averaging the values of relative dominance, relative density, and relative frequency. Species with the highest IVI values are shown in Fig.1.

Tree species were classified according to the light growth conditions required by their juveniles. Successional status was primary species or shade intolerant (growing exclusively or preferentially in TDF), shade intolerant species as pioneer and secondary tree species (establishing in cleared areas and only persisting under high light conditions), and intermediate species. The 45 species were classified in those broad categories: 23 were primary species, while 13 were intermediate and occur naturally in dry forest but also grow at early and intermediate successional sites. Only six species were clearly shade intolerant or pioneers (Table 1). Tree species were grouped according to seed dispersal syndrome as either zoochorous (seeds dispersed by animals), anemochorous (seeds dispersed by wind), or autochorous (seeds self-dispersed by gravity and ballistics). Woody plants were predominantly zoochorous (31 animal-dispersed species), followed by autochorous (six self-dispersed) and anemochorous (five wind -dispersed species) (Table 1).

In our early successional stages, differences in vegetation structure and species dominance patterns may be attributable to time since abandonment but possibly depend on the particular land-use history and disturbance level of each site prior to abandonment. Additionally, forest recovery potential may be affected by environmental factors (distance to nearby forest and seed accessibility, water proximity, topography) or anthropogenic ones (proximity to dirt roads and human settlements or the presence of trees in the pastures and living fences) related to regional TDF heterogeneity and high beta diversity (Williams-Linera and Lorea, 2009). The few dominant species in most early successional sites were Acacia cochliacantha, A. cornigera, and Guazuma ulmifolia (Fig. 1). Dominance by the Fabaceae family and the importance of Guazuma ulmifolia in early successional sites has been reported for most TDF successional studies in Central America.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (172)

146

Principles and Practice of Forest Landscape Restoration

Our early successional sites had already recruited several of the intermediate and primary species, indicating that forest species were entering the successional process at very early stages. Some of the species recorded at early successional sites and also found in regional forests were Brosimum alicastrum, Bursera simaruba, Croton reflexifolius, Ipomoea wolcottiana, Leucaena lanceolada, Maclura tinctoria, Malpighia glabra,Pisonia aculeata Randia aculeata, Spondias purpurea, Tabebuia chrysantha, Thouinidium decandrum, and Trichilia trifolia (Williams-Linera and Lorea, 2009).

At early TDF successional sites, the recruitment of individuals from re-sprouting is very high and therefore a significant mode of forest regeneration. Tree species that sprouted at our successional sites were Caesalpinia cacalaco, Diphysa carthagenensis, Guazuma ulmifolia, Ipomoea wolcottiana, Leucaena lanceolata, Maclura tinctoria, Malpighia glabra, Pisonia aculeata, Tabebuia chrysantha, Thouinidium decandrum, and Xylosma velutina. Interestingly, these tree species have the potential to colonize disturbed areas and can therefore be considered potentially important in ecological restoration trials in the region. The entry of mature forest species into the successional process at very early stages and the recruitment of individuals from sprouting may facilitate the recovery of the dry forest in Veracruz. Overall, there is evidence to suggest that early successional stages can lead to the recovery of TDF in central Veracruz.

Figure 1

Box 5.6 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (173)

147

Experimental analysis of dryland forest restoration techniques

Table 1

S1 S2 S3 S4 S5

Site Hato Marines Rinconada Don Tirso Xocotitla Dios Tigre

Fallow age 7 8 36 48 72

Basal area 1.97 (0.52) 0.40 (0.21) 2.46 (1.77) 3.88 (0.45) 2.74 (1.13)

Density 2225 (505) 900 (336) 4750 (859) 5450 (565) 2925 (436)

Height 2.07 (0.23) 1.8 (0.16) 0.96 (0.27) 1.79 (0.06) 1.80 (0.31)

Richness 11 8 19 21 10

Individuals 87 36 189 218 116

Fisher’s alpha 3.33 3.19 5.27 5.73 2.62

Shannon Index 1.88 1.37 2.28 2.05 1.63

Successional status

Primary 3 3 7 10 5

Intermediate 2 2 6 4 2

Pioneer 5 2 2 5 3

Seed dispersal mode

Animal 7 4 13 12 6

Wind 1 0 0 5 0

Self 2 3 2 2 4

Box 5.6 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (174)

148

Principles and Practice of Forest Landscape Restoration

Box 5.7 Tropical dry forest restoration in Chiapas, Mexico, and basic knowledge for native tree species: phenology, seed germination and seedling growth

Á.G. Becerra Vázquez, N. Ramírez-Marcial, S.C. Holz

Forest restoration is an option that must be reconciled with other social demands, such as production systems for home consumption and the market and the allocation of extensive areas for conservation (González-Espinosa et al., 2007). One of its strategies is the acceleration of natural forest regeneration in tandem with forest resources to provide economic and social value (Parrota et al., 1997). This has resulted in the identified need to develop strategies and methodologies for sustainable management of forests through the use of native species of economic value for the recovery of functionality and productivity of forests. The scarce information about initial stages in the biology of native tree species is one of the limitations for dry forest recovery. Some required studies deal with phenological stages, seed germination and seedling growth for selected species that are useful for local people, mainly for fuelwood and poles (see Chapter 6). We evaluated these variables in nine deciduous tree species from forest remnants located in the Central Depression of Chiapas, Mexico. The study site is located at the buffer zone of the El Ocote Biosphere Reserve (16° 53’ 52” and 16º 50’ 47” N and 93° 27’ 28” and 93º 24’17” W). The climate is warm to sub-humid with a mean annual precipitation of 1100 mm with summer rains. The dry season extends from November to May. Average annual temperature is 22°C. The elevation ranges between 820 and 980 m. Soils are rendzina and fine textured lithosols. Populations are mainly from the Zoque indigenous group. The main economic activities include traditional agriculture, with maize, beans, pineapple, banana, and coffee, as well as cattle ranching. There are also other activities that demand large quantities of firewood, such as pottery, cooking, and the toasting of coffee.

Flowering of most species in TDF occurred between January–May. In general, fruiting in most studied species was observed during several months, particularly of Acacia pennatula, Bursera simaruba, Magnolia mexicana, Ternstroemia tepezapote, and Trichospermum mexicanum. The ripening of fruits and seed dispersal was higher from the end of the dry season. Seed germination and seedling survival were significantly different among studied species (p <0.05). Acacia and Erythrina started germination earlier and completed 95% of germination in ca. 35 days, whereas other species delayed germination between 30–70 days and presented higher variation in germination percentages among species. Final seed germination (SG) was greater than 40% for seven out of nine species (except Trichospermum mexicanum and Leucaena leucocephala, which had SG below 24%), which is considered adequate for purposes of propagation in nurseries, but the high variation in SG denoting the speed and synchrony in germination (Mean Time of Germination, MTG and Germination Value GV) were deployed in Acacia pennatula and Erythrina goldmanii (MTG = 11.5 and 13.3 days, and GV = 34.7 and 22.7, respectively ), so this could suggest the presence of some mechanism of seed dormancy in other species (MTG which presented over 57 days and VG under 1).

Seedling survival was relatively high (90–100%) for most species (except for T. mexicanum, 76%). Maximum height and stem diameter at ground level were measured for 50 seedlings obtained previously for germinated seeds. Marked differences in RGR were found among species (Fig. 1). The highest relative growth rates, RGR were presented by the pioneer Trichospermum mexicanum that displayed twice the growth of late-successional species (such as Magnolia mexicana and Sideroxylum sp). These results suggested that all studied species are easy to propagate and therefore have a high potential to be used for purposes of forest restoration. For the implementation of specific restoration projects it is necessary to consider the usefulness of the species assigned by the local human population and above all it is necessary to incorporate these and other species in different forest management models based on the biological characteristics of each one of them.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (175)

149

Experimental analysis of dryland forest restoration techniques

Figure 1

Box 5.8 Soil seed bank, seed removal, and germination in early secondary succession of a tropical dry forest region in central Veracruz

L.P. Barradas-Sánchez, O.O. Ponce-González, C. Alvarez-Aquino

Artificial establishment of tree species and encouragement of natural regeneration are the main practical approaches to tropical forest restoration. Artificial establishment requires a considerable financial investment, whereas natural regeneration or passive restoration might be a slow but relatively inexpensive method that represents an option when large areas previously used for grazing or agriculture have been abandoned. However, a successful approach depends on an understanding of forest regeneration capacity and the key process that facilitates or inhibits it. Seed density and soil composition in the seed bank, predation, and germination capacity are important aspects of forest recovery in abandoned areas.

Seed density and floristic composition were evaluated in the soil seed bank in order to assess the potential for vegetation recovery. Ten soil samples (30 cm x 30 cm and 5 cm depth) were randomly collected at five abandoned sites (previously occupied by cattle or agriculture) and in two forest fragments. Samples were collected at the end of the dry season and transported in black plastic bags to the laboratory prior to the germination assessment. Soil samples were spread over 70 plastic trays (53 × 26 cm) in a thin layer and were watered every 1–2 days. Seed density and composition were determined by observing seedling emergence twice a week over 7 months. Emerging seedlings were counted and identified in terms of genus or species. Species were identified through field observations, and comparison with herbarium specimens. A total of 3946 seeds germinated in trays, most of the seeds corresponding to grass and herbaceous

Box 5.7 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (176)

150

Principles and Practice of Forest Landscape Restoration

species; woody species were scarce, with only the following species present: Acacia spp., Piscidia piscipula, Bursera cinerea, Cesalpinea cacalaco, Croton reflexifolius, Gliricidia sepium, and Ipomoea wolcottiana. The highest density was registered at the site previously used for agriculture (1303 seeds/m2, mostly grasses) followed by those used for cattle (701, 707, 704, and 411 seeds/m2), whereas the lowest densities were recorded in the forest fragments (458 and 101 seeds/m2, without woody species). Germinated seeds corresponded to 69 species, 49 genera, and 22 families. The best represented families with the highest number of species were Euphorbiaceae and Asteraceae. Cyperaceae was the family with the highest number of individuals.

Percentage and germination rate were evaluated in the field and in the laboratory. Seeds were tested with and without mechanical scarification with sandpaper. The selected species were native and common in the study area (Acacia cochliacantha, Caesalpinia cacalaco, Ipomoea wolcottiana, and Senna atomaria). In field experiments, they became established in tree environments typical of tropical dry forest in Veracruz: pasture, secondary vegetation, and forest fragment. Seeds were protected from predation with metallic netting. In the laboratory, germination was performed in controlled light and temperature conditions similar to those of the field (25–35ºC and 12 hours/day). Seed removal was evaluated in the field under the mother tree. Tested species were the same used for germination evaluation. Exclusions made from metallic netting of different sizes were used to test seed removal; treatments included total access, rodent exclusion, and insect exclusion. Germination percentage, both in the field and the laboratory, was higher when seeds were mechanically scarified. In the field, Ipomoea, Cesalpinia and Acacia yielded a percentage of ca. 50% (97, 57, and 47%, respectively). The exception was Senna, with a low value (10%). There were no differences among environments. Under controlled conditions, germination percentages were elevated (Ipomoea 99%, Cesalpinia 98%, Acacia 81%, and Senna 18%). In general, seed removal was below 30%, the exception being for Senna, with over 50%. As expected, the highest value was recorded for total access and the lowest for insect exclusion. However, any of the values represent a significant seed loss.

Results suggest that the contribution of the seed bank to natural regeneration is not significant, at least in early succession. Seed removal was not a constraint and woody species might germinate in the field, although the scattered periods of rain complicate early seedling establishment. On the basis of germination percentage in the field and in the laboratory as well as the seed removal rate, all of the selected species can be used in restoration plans in the area. Seeds can be germinated ex situ and grown for some months in nurseries prior to seedling transplant – or they can be planted directly during the rainy season. Of the selected species, Ipomoea appears to be particularly promising for restoration plans because it is present in the seed bank, exhibits high in situ and ex situ germination, and has a low removal percentage. Field observations during the dry season suggest that seeds of this species are removed by ants near nests where they can germinate in high numbers during the first summer rains. In contrast, Senna is recommended for germination in controlled conditions and presents a low potential for direct seeding because of the high seed removal rate.

Box 5.9 Effects of avian ingestion on seed germination in central ChileS. Reid and J.J. Armesto

A study of the effects of avian ingestion on seed germination of Mediterranean species in central Chile provides evidence of the positive effects of birds on seed germination with prospects to facilitate the regeneration of sub-Andean shrublands. Given that a mean of 14 plant species, i.e. 34.3% (range 10.5 to 53.1%) of the total woody flora in a gradient from dry to wet sites in central Chile bear fleshy fruits dispersed primarily by birds (Hoffmann and Armesto, 1995), we evaluated the effect of avian ingestion on seed germination of five woody species whose seeds are commonly consumed by a few species of birds in central Chile. We compared the responses of bird-defecated seeds to manually-extracted seeds and seeds surrounded by intact pulp in controlled laboratory conditions.

Box 5.8 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (177)

151

Experimental analysis of dryland forest restoration techniques

Collection of seeds for germination assays was conducted in the Estación de Investigaciones Ecológicas Mediterráneas (EDIEM) which lies in the Andean foothills, 20 km east of Santiago, between 1050 and 1915 m surrounded by evergreen sclerophyllous woodland. Seeds were collected from five avian-dispersed woody species, Azara dentata (Flacourtiaceae), Schinus polygamus, S. molle (Anacardiaceae), Cestrum parqui (Solanaceae), and Maytenus boaria (Celastraceae). These shrub species conform to 90% of the plant species found in bird droppings and are frequently found on the outskirts of the large city of Santiago (Reid, 2008). During the summer of 2006 (January to March), we collected bird droppings containing shrub seeds in the matorral of the EDIEM. Fresh fruits were also collected from a minimum of five individuals of each of the five shrub species found in bird droppings. Fruit pulp was manually removed from a number of seeds corresponding to the number of seeds extracted from bird droppings. Germination assays were conducted in glasshouse conditions set for a spring ratio of 14 hours light to 10 hours darkness. Light intensities were 350 to 500 μmol m–1 s–2 from a metal halide light measured at the outer surface of containers. Temperatures ranged from 16ºC to 28ºC. For each species treatments were: seeds collected from bird droppings (ingested seeds), seeds manually removed from the pulp (extracted seeds), and seeds sowed with the pulp (intact fruits). Seeds were placed on filter paper in Petri dishes, 9 cm diameter, and watered every 2–3 days with distilled water. Germination was monitored every three days by recording the emergence of the radicle during three months in the austral winter and spring (July to October 2006). A total of 774 seeds were tested. We assessed germinability (the final percentage of sown seeds germinated after three months) and germination rate defined as the number of seeds germinated per time interval.

Avian ingestion significantly increased seed germinability (ANOVA F2,45 = 12.1, p <0.001; Fig.

1). Fifty percent of the seeds that were avian ingested germinated, while 30.5% germinated in the extracted seed treatment and 12% of the seeds germinated within intact fruits. Species-specific level analyses show that for A. dentata germinability did not differ between ingested and extracted seeds (Mann-Whitney U-test: Z = 0.09, p = 0.93) and no seeds germinated within intact fruits. For S. molle, germinability was significantly higher for avian ingested seeds than for seeds within intact fruits (ANOVA F2,9 = 4.94, p = 0.04; Tukey’s test, p = 0.03; Fig. 2). For C. parqui, germinability was significantly higher for avian ingested seeds than for manually-extracted seeds (ANOVA F2,3 = 10.25, p = 0.05; Tukey’s test, p = 0.04). In M. boaria, germinability was equal for ingested and extracted seeds (85% germinability) and half the seeds sowed with the pulp germinated. In contrast, only one out of 100 seeds of S. polygamus germinated in each of the ingested and the intact fruit treatments, and no extracted seeds germinated.

Figure 1

Box 5.9 (cont.)

20

40

60

80

100

120

Ingested seeds Extracted seeds Intact fruit

% G

erm

inab

ility

S. polygamus

A.dentata

S. molle

C. parqui

M. boaria

All species

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (178)

152

Principles and Practice of Forest Landscape Restoration

Avian ingestion accelerated the rate of seed germination. Species-specific analysis showed similar rates for ingested and extracted A. dentata seeds (RMANOVA F1,16 = 1.22, p = 0.28 and HS, p = 0.49). For S. molle, a significant increase in germination rate was observed for avian ingested seeds compared to those within intact fruits (RMANOVA F1,6 = 18.91, p = 0.01 and HS, p = 0.01; Fig. 2). In C. parqui, the germination rate of ingested seeds was significantly faster than extracted seeds and those within intact fruits (RMANOVA F1,2 = 110.45, p = 0.01 and HS, p = 0.01; and F1,2

= 108.97, p = 0.01 and HS, p = 0.02, respectively). For M. boaria, avian ingested seeds had a significantly faster germination rate than seeds within intact fruits (GB = 8.88, d.f. = 2, p = 0.01).

Figure 2 Schinus molle

We concluded that avian ingestion increased seed germinability in four of the five shrub species that are commonly consumed by birds in this landscape. Although higher germinability is positively related to plant fitness (Traveset et al., 2007), the dependency of plant species on avian ingestion for the completion of their reproductive cycle was variable and strongest for A. dentata. Given that seeds within intact fruits of this species did not germinate at all, we suggest a dependency of this species on avian frugivory to regenerate in vegetated or open areas. Alternatively, the lack of seed germination in S. polygamus may have been due to low seed viability (A. Sandoval, unpubl. results).

In contrast, the advantages of faster germination for seeds are less well understood. Izhaki and Safriel (1990) suggest avian frugivores add temporal heterogeneity to seed germination responses, by enhancing germination rates and shortening seed dormancy. The advantage of increased temporal germination heterogeneity in variable environments is spreading the risk of seed and seedling mortality over time. Traveset et al. (2001) suggest that faster seed germination allows dispersed seeds to have a shorter exposure to seed predators. Paulsen and Högstedt (2002) suggest that early emerging seedlings in spring have an advantage because they develop deeper root systems making them more resistant to drought.

This study shows that avian frugivores in this Mediterranean ecosystem do not harm seeds after ingestion and tend to speed up germination.

Box 5.9 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (179)

153

Experimental analysis of dryland forest restoration techniques

Box 5.10 Tree-seedling establishment in fragmented Mediterranean forests of central Chile

M. Holmgren, J.L. Celis-Diez, J.J. Armesto

Mediterranean-type ecosystems are global biodiversity hotspots on all continents. Vegetation cover is extremely fragmented by the combination of natural and man-induced disturbances. Regeneration of these semi-arid ecosystems has proved difficult but can be strongly facilitated by plant-nurse interactions. We studied whether the interaction of micro-climate and herbivores under remnant vegetation patches defines a patch-size threshold for tree seedling recruitment and whether that threshold is dependent on overall climate conditions and plant functional types.

We used a combination of correlational and experimental approaches to investigate this problem. Field observations and experiments were conducted in Andean and coastal Chilean shrublands, representing a gradient in rainfall (350 and 500 mm annual precipitation, respectively). We planted one-year old seedlings of relatively drought-tolerant (Quillaja saponaria) and drought intolerant (Cryptocarya alba) species under open and shaded conditions considering a gradient of shrub patch sizes (1, 5, 10–15, >30 m in diameter, n = 10 per patch size) and at increasing distances from the canopy edge (5 and 0.5 m outside of the border of the patch, 0.5, 2, 5, 15 m inside of the patch). Half of the seedlings were protected against mammal herbivores (mostly rabbits and horses).

We found no naturally established tree seedlings in the drier foothills of the Andes. In the moister coastal range, seedlings were frequently under the canopy of small and medium-sized shrub patches (5–15 m diameter). Seedlings in open areas were found only at the edge of large shrub patches (>30 m). Herbivore pressure by rabbits and hares is enormous in the Andean foothills where no seedlings survived in the experimental non-protected plots. At the moister coastal range site, herbivore pressure was lower and was further reduced under larger patches. Seedling mortality due to drought stress was reduced under the shrub canopy and significantly decreased with shrub patch size, particularly in the drier Andean site. Large shrub patches are cooler and moister which ameliorates plant thermal and water stress particularly at the drier site. Seedling survival was strongly linked to physiological performance under different water and irradiance conditions.

Our results indicate that shrub fragmentation might be irreversible at the drier end of semi-arid ecosystems as shown in the Andean foothills of central Chile. Conservation of large remnant shrub patches in the landscape, wherever possible, is essential here to facilitate ecological restoration by combining herbivore exclusion and shrub shade. Under moister conditions (coastal sites and wet years), herbivore protection may be sufficient to enhance tree regeneration. In such areas, seedling establishment will also increase along large patch edges and under existing small shrub patches. ENSO events in central Chile increase precipitation to levels comparable to those found in the coastal region, and could potentially increase the probability of seedling establishments in Andean foothills.

Figure 1a, b

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (180)

154

Principles and Practice of Forest Landscape Restoration

Box 5.11 Post-fire restoration of native tree species: effects of wood shaving application

M. Lallement, C. Tognetti, M.E. Gobbi

Fires are the most devastating anthropogenic disturbances in forests of the Andean-Patagonian region. They severely affect surface soil physical and chemical characteristics, as well as vegetation and fauna, thus increasing the risk of erosion. These forests are very important for the conservation of biodiversity, climate and watershed regulation, and soil stability. Therefore, it is relevant to implement strategies that favour ecological restoration after a disturbance has occurred.

Reforestation with native species is one of the most used restoration strategies in the Andean-Patagonian forests. This strategy initially depends on the availability of plants, both in terms of quantity and quality. The adequate growth and development of these plants will determine the success of a restoration project. Aspects such as the type of substrate used for seedling production and the strategies used to plant these seedlings in the field should be carefully considered when planning restoration projects. The general objective of the work carried out here was to evaluate strategies that facilitate the recovery of burned forest areas of the Nahuel Huapi National Park (northwestern Patagonia), specifically by means of reforestation with native tree species. For this, we evaluated:

i. The success of a reforestation project that included the aid of volunteers to plant three native tree species (Austrocedrus chilensis, Nothofa*gus pumilio and Lomatia hirsuta) in a post-fire area of the Nahuel Huapi National Park.

ii. The effect of wood shavings, either applied as mulch or incorporated into the soil, on the survival and growth of A. chilensis, N. pumilio and L. hirsuta: (i) in a nursery, and (ii) in the field.

iii The effects of wood shavings, either applied as mulch or incorporated into the soil, on the water dynamic of a burned volcanic soil.

The three tree species selected for this study are characteristic of the xeric forests in which most of the fires in the region occur. They have different life-forms and reproductive strategies (Table 1). Regarding L. hirsuta, there are no previous records of this species being used in restoration projects. The study was carried out in a post-fire shrubland in the Challhuaco Valley, where mean annual precipitation is 1000 mm (mostly rain and snow in autumn-winter). Seedlings were obtained at local nurseries, and their ages varied as follows: 3–5 years old (Objective I), 1-year old (A. chilensis and L. hirsuta, Objective II), and 3-years old (N. pumilio, Objective II). In the field study, seedlings were planted under shrubs (nurse plants).

Growth and survival of plants in the reforestation project aided by volunteers (Fig. 1) was satisfactory; values obtained were within the range of those obtained in previous studies by experienced personnel and under similar climatic conditions. The highest survival after the first year was for A. chilensis (51%), followed by L. hirsuta (43%), and N. pumilio (29%). In conclusion, reforestation project aided by volunteers proved to be a good strategy to recover degraded areas, and provided an opportunity to educate citizens on environmental problems.

To evaluate the effects of wood shavings on the survival and growth of the three species, pine (Pseudotsuga menziesii) wood shavings were either incorporated into the soil at a 1:3 shavings to soil ratio (v/v) or mulched over the soil forming a 2 cm deep layer. As controls, soil (burned for the field assay, unburned for the nursery assay) with no wood shavings was used. In the field, seedlings were only watered when they were planted; while in the nursery they were watered tree times a week. In the nursery assay, mixing wood shavings in with the soil increased survival of the three species and improved some of the growth indicators (Fig. 2). Apparently, this is a promising strategy to improve the nursery production of native species. However, mulching soil with wood shavings had no effects on survival or growth (Fig. 2). In the field assay, wood shavings had no effects on the measured parameter, regardless of whether it was mulched or was incorporated into the soil (Fig. 2). However, mulching did reduce and stabilize soil surface temperature.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (181)

155

Experimental analysis of dryland forest restoration techniques

This is the first report of the use of L. hirsuta in restoration projects. The results outlined so far indicate that this species has good survival and growth rates, and should be considered in future reforestations.

For the assays on the dynamic of soil water (assays without plants) surface soil from a burned shrubland was used. The soil was either mixed with wood shavings (at three soils to shavings rates) or mulched with wood shavings (two mulch depths). In both cases, a burned soil without wood shavings was used as a control. Mixing wood shavings in the soil increased water draining speed and field capacity; this effect was more marked as the proportion of wood shavings increased. Applying a layer of wood shaving mulch on the soil delayed evaporation, although final soil moisture values were not affected. The effects of wood shavings on the dynamics of soil water could prove to be positive for plants, provided that the water remains available near the root system.

The application of new strategies aimed at increasing survival and growth of native plants, especially those that use low-cost and easily available resources, stimulates the development and implementation of restoration projects. Ultimately, it accelerates the succession process in deteriorated natural habitats, and improves recovery of degraded areas.

Table 1

Austrocedrus chilensis Nothofa*gus pumilio Lomathia hirsuta

Family Cupresaceae fa*gaceae Proteaceae

Order Coniferales fa*gales Proteales

Common name Ciprés Lenga Radal

Foliage Perennial Caduca Perennial

Life-form Tree Tree Shrub/Tree

Re-sprouting ability No No Yes

Masting No Yes No

Figure 1

Box 5.11 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (182)

156

Principles and Practice of Forest Landscape Restoration

Figure 2 A. chilensis N.

pumilio L. hirsuta

Box 5.11 (cont.)

Case Studies

Central Valley, Chile

The study area is located in the Central Valley between the Andes and coastal mountain ranges, in the transition zone between Mediterranean sclerophyllous forest and deciduous dry for-est (35°–38°S; Fig. 5.1). The climate is semi-arid Mediterranean. The annual average (30 year) precipitation is 330 mm during winter, with 6–7 dry months. The annual mean temperature is 15°C. Vegetation is composed of tree, shrub, and herbaceous patches that are dominated by native species (Pasithea coerulea, Bromus berteroanus, Clarkia tenella, Amsinckia calycina, Moscharia pinnatifida and Helenium aromaticum) and exotic species (Conium maculatum, Centaurea melitensis, Fumaria capreolata, Carduus pycnocephalus, Erodium cicutarium and Brassica rapa). The selected sites for field experiments are representative of different types of vegetation, soils, and human disturbance history present in the region. They were located on private lands (Los Maitenes, Las Tórtolas, and Pirque), experimental stations (Quebrada de la Plata, Calán, and San Carlos de Apoquindo) and private reserves (San Ramón watershed and Mahuida Park) (Fig. 5.1). The watershed has been protected from logging, cattle grazing, and woodland fires for the past 10 years by the Chilean forest service. Experiments were also conducted in the dry woodlands of Fray Jorge National Park to assess the response of species to irrigation simulating an increased frequency of wet (ENSO) years.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (183)

157

Experimental analysis of dryland forest restoration techniques

Figure 5.1

Experimental approaches focused on the effect of herbaceous cover, precipitation, nurse shrub species (exotic, N-fixing, and non N-fixing), herbivory (combining their effect with shrub patch cover and precipitation), the potential for restoration during rainy years, and natu-ral recolonization of trees and birds. Experiments were established in the study area to test the effect of seven factors on woody seedling establishment of three native species (Table 5.1).

Table 5.1.

Sites Species traits Selected species used in restoration

Central Valley, Chile

Colliguaya odorifera, Cryptocarya alba, Lithrea caustica, Quillaja saponaria Shinus polygamus

Coastal Range, Chile Casablanca Valley

Shade tolerant Cryptocarya alba, Peumus boldus

Intermedia Lithraea caustica, Maytenus boaria, Schinus latifolius, Quillaja saponaria

Shade intolerant Acacia caven, Baccharis linearis, Colliguaja odorifera

Lago Peñuelas National Reserve and Colliguay ValleyShade tolerant Beilschmiedia miersii, Cryptocaria alba, Peumus boldus

Intermedia Maytenus boaria, Quillaja saponaria, Schinus latifolious

Shade intolerant Acacia caven, Senna candolleanaNorthwestern Argentina

Natives

Amburana cearensis, Anadenanthera colubrina, Astronium urundeuva,Caesalpinia paraguariensis, Calycophylum multiflorum, Cedrela balansae, Coccoloba tiliacea, Cordia trichotoma, Diatenopteryx sorbifolia, Enterolobium contortisiliquum, Gleditsia amorphoides, Inga saltensis, Jacaranda mimosifolia, Myroxylon peruiferum, Pithecelobium scalare, Phyllostylon rhamnoides, Pterogyne nitens, Saccelium lanceolatum, Tabebuia impetiginosa, Tipuana tipu

ExoticsToona ciliata, Grevillea robusta, Eucalyptus grandis, E. teretricornis, Corymbia torelliana, C. citriodora, C. maculate, Flindersia xanthoxila, F. australis, Khaya senegalensis, Pautownia fortune

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (184)

158

Principles and Practice of Forest Landscape Restoration

Sites Species traits Selected species used in restoration

Southwestern Argentina

Austrocedrus chilensis

Oaxaca, MexicoAcacia angustissima, Amelanchier denticulata, Cercocarpus fothergilloides, Desmodium orbiculare, Dodonaea viscosa, Eysenhardtia polystachya

Central Veracruz, Mexico

NativeCedrela odorata, Ceiba aescutifolia, Guazuma ulmifolia, Ipomoea wolcottiana, Luhea candida, Tabebuia rosea

Selected by local people

Chloroleucon mangense, Diphysa carthagenensis, Leucaena lanceolata, Lysiloma acapulcense, Lysiloma divaricatum, Maclura tinctorea, Cordia alliodora, Cedrela odorata

The treatments were as follows:

(1) Herbaceous cover and precipitation. The hypothesis was that herbaceous cover in the Chilean scrubland has a negative effect on the establishment of woody seedlings, and this effect would be stronger with heavier rainfall owing to the increase in herbaceous cover and height. Results showed that the effect of herbaceous cover on woody seedling es-tablishment differed between years. During the year corresponding to ‘La Niña’, an event involving extremely low precipitation, the herbaceous cover negatively affected seedling survival. In contrast, during the year with more precipitation, the effect of herbaceous cover on seedling survival was positive (Fig. 5.2).

Table 5.1.

Figure 5.2 �

� ��

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (185)

159

Experimental analysis of dryland forest restoration techniques

(2) N-fixing and non N-fixing nurse shrub species. The hypothesis was that N-fixing shrub species have a stronger positive effect on the establishment of woody seedlings under their cover than non N-fixing shrub species. In the Chilean scrubland, N-fixing species correspond to Leguminosae and Rhamnaceae shrubs. Results showed that the effect was positive owing to an increase in soil nutrients; however, the effect of N-fixing species was greater than that of the non-fixing species.

(3) Exotic tree species. The hypothesis was that exotic trees (introduced in many areas of central Chile) have a negative effect on seedling establishment. However, the results showed a significantly positive effect; but the effect of Eucalyptus globulus was mi-nor in comparison to the other exotic species.

(4) Shrub patch cover and herbivory. The hypothesis was that the effect of herbivory by native and exotic mammals is weaker within remnant shrub patches than in open areas and that this impact will decrease with patch size. Results showed that the effect of forest fragmentation and herbivory was negative. In smaller forest fragments and in the area nearer to fragment edge, the herbivory effect was greater.

(5) Herbivory and precipitation. The hypothesis was that the effect of herbivory on seedling establishment is negative, being weaker in years with higher rainfall. Results showed that the effect of herbivory and high precipitation (simulated by artificial ir-rigation) on seedling survival was negative.

(6) The potential use of El Niño (ENSO) rainy years for restoration of native vegetation in central Chile. An experiment was designed to test whether rainy years represent an opportunity for restoration. Shrub species with contrasting growth forms were subjected to a range of simulated ENSO conditions and partial herbivory by rabbits and rodents (herbivory was simulated by ‘clipping’). Results showed no effect of wa-ter on survival and a strong herbivory effect. Survival was 0% and extremely low both in plots with and without herbivores. Results suggest that extending the water pulses into spring-summer, when the drought commences, will increase survival rates.

(7) Restoration and the natural recolonization of native trees and birds in disturbed sites. The goal was to evaluate performance of different native woody species used in forestry plantations established by private owners and to assess natural recolonization of native birds and shrubs. Results showed that natural colonization by birds and shrubs was dif-ferent between planted versus non-planted adjacent sites. Within each plantation, survival and growth rates differed significantly between species, with Quillaja saponaria show-ing the highest performance. In general, the observed natural recolonization by woody species was low, whereas natural recolonization by birds was much greater. Difference in bird abundance between reforested and open adjacent sites was significantly greater in older reforestation areas, indicating that bird abundance is facilitated by reforestation.

Conclusions

According to these results, the following recommendations can be identified for restoration programmes: (1) in plantations, it is not always necessary to eliminate herbaceous cover, since only during drier years would herbs have a negative effect; (2) when plantations are established in open sites, irrigation is recommended during summer even when the precipitation of the year has been close to average; (3) plantation with successional native species should be

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (186)

160

Principles and Practice of Forest Landscape Restoration

undertaken underneath nurse species (exotic, N-fixing or non N-fixing), because they may facilitate seedling establishment. Exotic species are useful as nurses; however, in some cases (if they are aggressive) it is necessary to eradicate them after native seedlings are established. In general, nitrogen-fixing species are better nurses. (4) Despite the lower herbivory effect within fragments and on larger ones, herbivory is high enough to kill most of the seedlings, and therefore herbivore exclusion must be considered under any environmental conditions; and (5) reforestation programmes may be highly effective not only for planted species but also for other organisms such as birds, even when plantations are still relatively young.

Coastal Range, Chile

The study area is located in the Coastal Range of Valparaiso (33°S, 71°30’ W; Fig. 5.3). Annual precipitation ranges from 100 to 800 mm. Vegetation corresponds to sclerophyllous forest, characterized by high endemism and the presence of some xerophytic plant species such as Cactus spp. The main land uses are agriculture, urban areas, fruit cultivation, and mining. The soil is highly degraded as a result of erosion caused by forest removal. This region is character-ized by the presence of large wine exporters, leading to substantial land use changes. Also, large areas have been converted to plantations of exotic tree species. Restoration experiments on semi-arid sites were conducted in the Casablanca Valley (land owned by the Chile Tobacco Company), Lago Peñuelas National Reserve (LPNR), and the Colliguay Valley. Additionally, some plots were established in more humid micro-sites in the Colliguay Valley and the LPNR.

Figure 5.3

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (187)

161

Experimental analysis of dryland forest restoration techniques

Field experiments were established under three site conditions: presence of Acacia caven, presence of Eucalyptus trees, and an open site. The treatments were fertilizer, Eucalyptus bark, and drip irrigation. The tree species established in the restoration assays included shade-tolerant, intermediate, and shade-intolerant species (Table 5.1). The experiments were monitored for plant health conditions every month, and growth in height and diameter was monitored every six months.

(1) Casablanca Valley. In Casablanca Valley, a total of 2320 seedlings have been planted since 2006. In general, the areas restored with Acacia and Eucalyptus, where soil temperature and humidity were less extreme, showed statistically higher survival and growth than the open site. The Acacia site presented higher rates of survival and growth for most species, suggesting that factors such as nitrogen fixation by Acacia could improve plant development at this site. In the presence of Eucalyptus, the al-lelopathic effects of this species could influence seedling development.

Baccharis linearis performed well under the three site conditions. Quillaja sapon-aria also had a high survival rate, especially when planted at the Acacia site. In contrast, Cryptocarya alba and Peumus boldus did not perform well under these conditions. Irrigation in early stages of development had a positive effect on survival and growth, especially at the Acacia site. With the exception of the open site, results suggested a positive effect of the combined application of irrigation, fertilizers, and/or bark on the survival of most plants. However, increased mortality of Senna can-dolleana was observed when Eucalyptus bark was used (Fig. 5.4).

Figure 5.4 Acacia Eucalyptus

(2) Colliguay Valley and Lago Peñuelas National Reserve (LPNR). In Colliguay and LPNR in 2008, a total of 710 seedlings were established on Acacia sites. Additionally, four plots were established and a total of 120 individuals of species suitable for shade and humid conditions in riparian environments were planted in these areas’ more humid micro-environments.

At the Acacia site, Quillaja saponaria and Schinus latifolius had the highest sur-vival rates for all experiments (ca. 50 %), while Maytenus boaria had the lowest survival (5%). When the different study sites were compared, the highest species survival was observed at the site with conditions associated with almost no slope and the greatest influence of coastal fog. The effect of fertilizer application was

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (188)

162

Principles and Practice of Forest Landscape Restoration

uncertain; there were no clear effects on either species survival or growth rates in any experiment. However, at the more humid micro-sites, experiments without fertilizer had the lowest survival rates. Beilschmiedia miersii had a 40% survival rate in the Colliguay Valley, and P. boldus and C. alba around 10% in LPNR. Overall higher survival rates in Colliguay may be explained by its more humid conditions than LPNR. Results suggest that fertilizer application on humid sites increases sur-vival; however, there is no way to ensure that seedling performance improves as a result. In Colliguay, the highest growth rates at the Acacia site were recorded for Q. saponaria, S. latifolius, and M. boaria.

Conclusions

In Casablanca Valley, the best species performance was recorded at the Acacia sites. The establishment of A. caven, B. linearis, and S. candolleana facilitated the establishment of other species on open sites while irrigation facilitated higher native plant cover at Eucalyp-tus sites. However, a conclusive evaluation of fertilizer and bark experiments would require more than a year of analysis.

In Colliguay and LPNR, the results suggested that under Acacia shade, the species that establish have high survival and growth rates, such as Q. saponaria and S. latifolius. At open sites, A. caven displayed a higher survival rate than S. candolleana, although the growth rate of the latter was higher. One recommendation is to use irrigation in cases where dry condi-tions are more restrictive. Fertilization experiment results do not indicate whether plant sur-vival and growth were improved or not. B. miersii, C. alba, and P. boldus are recommended for shaded and more humid conditions.

Northwestern Argentina

Restoration activities focused on subtropical seasonally dry forests (SSDF) located in Salta (San Martin and Oran Departments) and Jujuy (Santa Barbara and Ledesma Departments) provinces in northwestern Argentina (22º–24ºS, 63º–65ºW; 350 and 750 m a.s.l.; Fig. 5.5). These forests include Andean premontane forest and a transition to dry Chaco forest. Chaco forest is the largest unit of tropical dry forest in the neotropics, extending through Argentina, Bolivia, and Paraguay. SSDF are largely threatened by deforestation and transformation to ag-riculture, mainly for sugarcane and soybean production. Timber activities are also economi-cally important in the area and are largely restricted to selective logging of about a dozen valuable native species. The study area covers approximately 10,000 km2 and combines highly profitable agricultural activities with local and indigenous communities living in ex-treme poverty. The study area harbours the highest concentrations of ethnic groups (nine) of Argentina, including groups of Andean, Amazonian, and Chaco origin. Forest transformation for agriculture is disrupting the historical forest continuity between Andean premontane forest and dry Chaco forest, creating an agricultural gap 5 to 25 km wide. In the study area, SSDF covers approximately 7,500 km2, most of which is highly disturbed and susceptible to transformation.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (189)

163

Experimental analysis of dryland forest restoration techniques

Figure 5.5

In 2001, a 50 ha experimental plantation was established in Valle Morado (Salta Province) that includes 20 native and 11 exotic tree species (Table 5.1). A plantation was established with 20 native species planted at random to simulate natural regeneration in order to gener-ate information about the restoration process in degraded or deforested areas. In this assay, no interventions (e.g. pruning) were conducted in order to avoid modifying natural establishment and growth patterns. In the plantation, pure and mixed assays with exotic and native tree spe-cies were established to evaluate growth rates, growth form, health, and species interactions (e.g. competition, nurse effect). Treatments included variations in tree density, mixtures of dif-ferent native species, use of exotics as nurse trees. Tree performance was evaluated in terms of diameter and height increments and vulnerability to pathogens under different treatments.

Surveys of the experimental plantation were performed in 2003, 2005, and 2007. In Oc-tober 2005, the experimental plantation was damaged by fire (ca. 12 ha); then, during the summer of 2007, 5 ha were replanted with native and exotic species, but unusually low temperatures recorded during the first winter after replanting resulted in the massive loss of seedlings. In March 2008, a 2.6-ha experiment of native and exotic species was established in Valle Morado to take the place of the experimental area damaged by fire and frost. One native species (Cedrela balansae) and two exotic species (Toona ciliata, Tectona grandis) were used in this experiment. In the 50-ha experimental plantation, after eight years of establish-ment total density was 744 individuals/ha and mortality was 18.5%. Mean diameter for all species was 11.4 cm, mean height was 8 m, and mean basal area was 10 m2/ha. A multispe-

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (190)

164

Principles and Practice of Forest Landscape Restoration

cific block with high canopy cover was reached after eight years of experimentation with C. balansae, Pterogyne nitens, and Tabebuia impetiginosa as dominant species in the block.

Pure and mixed plantations of native trees displayed different diameter, mortality, and biomass than mixed plantations with exotic species. The native species with the largest diameters were Enterolobium contortisiliquum, Tipuana tipu, and C. balansae, which reached diameters of more than 10 cm in the first five years; whereas the species with the smallest diameters (6–8 cm) were Cordia trichotoma, Jacaranda mimosifolia, P. nitens, Astronium urundeuva, and T. impe-tiginosa after eight years (Table 5.7). In mixed plantations of native and exotic trees, the diameter and height of native species were smaller to those of exotic species. In contrast, the mortality of native species was lower (between 5–30%) than that of exotic species (>30%). The biomass estimation for exotic tree assays (40 to >70 tonne/ha) tended to be greater than for native planta-tion assays. Pure plantations of the native tree C. balansae (80 tonne/ha) or T. tipu (50–60 tonne/ha) reached greater biomass than any other native species during the eight years of evaluation.

Table 5.7

Species dbh (cm) IMA (cm/yr) height (m) IMA (m/yr)

Cordia trichotoma 5.78 1.54 3.39 0.91

Cedrela balansae 6.85 1.83 5.54 1.48

Tipuana tipu 12.77 3.41 8.91 2.37

Pterogyne nitens 5.46 1.46 4.76 1.27

Jacaranda mimosifolia

6.88 1.83 4.45 1.19

Tabebuia impetiginosa

6.22 1.65 3.33 0.89

Astronium urundeuva

7.69 2.05 4.66 1.24

Conclusions

(1) Exotic tree species displayed a higher mortality than native tree species, particularly dur-ing the dry season (June–October); (2) native species tended to form more ramifications than exotic species; it is therefore advisable to favour high density plantations of native tree species; (3) mixing exotic (Grevillea robusta) and native species reduced ramification; (4) Phylostilon rhamnoides (native species) displayed low survival and growth in open plantation, owing to its shade requirements for recruitment; (5) some exotic species (Flindersia xanthoxila, F. australis, Khaya senegalensis, and Paulownia fortune) proved to be poorly adapted to local conditions; (6) C. balansae (native species) suffered severe damage from a shoot boring moth (Hypsipyla grandella); chemical pest control and pruning for stem conduction is therefore recommended; (7) T. tipu, C. trichotoma (native species) and T. ciliata (exotic species) dis-played the best performance; (8) T. ciliata (‘cedro australiano’) is considered equivalent to C. balansae (‘cedro Oran’). ‘Cedro australiano’ is not attacked by native pests and its genetic se-lection confers certain advantages from a forestry viewpoint, but some assays with C. balansae achieved higher biomass accumulation rates than in other secondary forests.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (191)

165

Experimental analysis of dryland forest restoration techniques

Southwestern Argentina

The study was conducted in the forest-steppe ecotone on the eastern slopes of the Patagonian Andes, Argentina (Fig. 5.6). The region is characterized by an abrupt west to east decrease in precipitation owing to the orographic effect of the Andes. Droughts in concert with natural and anthropogenic ignition sources make fire the main driver of ecological change in the re-gion. The high fire frequency during aboriginal and European settlement eras has led to forest fragmentation (Chapter 3); local extinctions of fire-sensitive arboreal taxa and/or the retraction of these taxa into fire-free rocky refuges with low fuel loads have permitted the survival of scattered remnant trees. Treeless areas have been traditionally viewed by foresters and land managers as barren lands unable to support native forest and have been used for extensive sheep, cattle, and goat ranching. In addition, other introduced herbivores such as European hares, rabbits, and exotic deer have a negative impact on native arboreal vegetation. Another source of dry native forest degradation is the current increasing trend of exotic plantation es-tablishment (mostly pine), which drastically changes fire regimes by increasing the extent and severity of fires. Restoration trials were established at two sites located in the central region of the Austrocedrus chilensis distribution area, in Arroyo del Medio Valley and in Estancia San Ramón, both privately-owned lands on the outskirts of Nahuel Huapi National Park.

Figure 5.6

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (192)

166

Principles and Practice of Forest Landscape Restoration

Restoration experiments focused on the economically important tree species Austrocedrus chilensis. Experiments were designed to test the following hypotheses: (1) Austrocedrus is ab-sent from areas where it has become locally extinct owing to higher fire frequency resulting from anthropogenic activities, and (2) Austrocedrus establishment is regulated by micro-site availability (nurse plants), herbivory, and traits relating to local adaptation (origin). Experimental trials emphasize the effect of nurse plants (with and without protection), herbivory (with and without exclosure), location on the slope (the highest part of the hill and mid-slope), and proxim-ity to shrubs and grasses (underneath a nurse plant and far from all neighbouring plants).

The first experimental trial was established in Arroyo del Medio during the austral autumn (humid season) in May 2008. Austrocedrus seeds were stratified and sown in pots in a green-house, and seedlings were transferred to experimental plots. The experiment consisted of: (1) using nurse plants (seedlings planted underneath a shrub or without shrub protection), and (2) herbivore exclosure (with and without). Seedlings with nurse plants and herbivore exclosures suffered very high seedling mortality during the first winter of 2008. Mortality was over 90% in open plots for both seedling cohorts, probably caused by root uplift by frost owing to the small size of seedlings, especially those less than one year old. Nevertheless, survival was significantly higher on treatment plots with nurse shrubs (between 35 and 75% for the very young ones and the 3-year-old seedlings, respectively; Fig. 5.7). Herbivores had no effect on survival rates during winter. Owing to slow Austrocedrus growth rates, monitoring was only based on survival rates.

Figure 5.7 Austrocedrus

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (193)

167

Experimental analysis of dryland forest restoration techniques

Under the same treatments, survival of 3-year-old seedlings was significantly higher than that of 1-year-old seedlings. After the following growing season, survival was low (between 10% and 30%), probably caused by the lack of precipitation during a particularly dry summer (2008–2009). Only plots with nurse shrubs had surviving seedlings at the beginning of the summer, and there was no significant difference among those with and without protection against herbivores. Again, the 3-year old seedlings had a significantly higher survival rate at the end of the summer: ca. 30% survival during the growing season compared to 10% for very young seedlings (Fig. 5.7).

The second experiment was established on San Ramón Ranch during spring 2008 (Oc-tober), on a small fragment of Austrocedrus woodland on a rocky outcrop surrounded by steppe vegetation. Three-year-old seedlings from a commercial nursery were planted in the experimental plots. All seedlings were transplanted underneath a shrub or in artificial shade. Treatments were herbivore exclosure (with and without), location on the slope (on a rocky substrate on the highest part of the hill and in soil on the mid slope), and proximity to shrubs and grasses (underneath a nurse plant and far from all neighbouring plants, covered by a shade mesh). Survival during the growing season was very good (ca. 80%) on the mid slope, in the matrix of grasses and shrubs, in treatments under nurse shrubs, as well as in those with artificial shade. No effects from neighbouring grasses and herbs were detected, at least after one growing season. From November to January, there were no significant differences in survival among treatments. Conversely, survival on the ridge and on a rocky substrate was extremely low in February at the end of summer (Fig. 5.8). This may be attributable to a lack of sufficient organic soil to retain humidity; it was also a very dry summer, and the nurse plants used were smaller and less dense shrubs than those on slope plots.

Figure 5.8

Conclusions

The main limiting factor for establishment of Austrocedrus is drought, especially when it is adjacent to the steppe, where some scattered, isolated woodlands of this species occur. For restoration programmes, it is recommended to use nurse shrubs (to protect against ex-

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (194)

168

Principles and Practice of Forest Landscape Restoration

tremely cold weather and soil frost during the winter) or artificial shade (to lessen direct sunshine and wind desiccation during the summer), to improve seedling survival. Seedlings should be planted when they are over two years old, since in their first year they are vulner-able and mortality is very high.

Central Veracruz, Mexico

In central Veracruz, Mexico, the study area was located in Paso de Ovejas and Emiliano Zapata (19°17’N, 96°26’W, 100–250 m a.s.l.; Fig. 5.9). The climate is characterized as hot and dry. Mean minimum and maximum temperatures are 20 and 31°C, respectively. Mean annual precipita-tion is ca. 900 mm and is unevenly distributed throughout the year. The dry season extends from October to May. Soils are mainly cambisols and vertisols with considerable extensions of exposed rock. The land is mainly used for cattle ranching, typically on relatively small-scale private farms. Activities on communal lands are more diversified, the main activity being maize production. Other crops are papaya, bean, green chili, watermelon, sorghum, sugarcane, and mango. Selection of the study area was based on ecological criteria and historical factors. This area has numerous remains of pre-Hispanic settlements (600 to 1500 AD) and played an impor-tant role in Mexican independence (19th century). Human occupation of the region has been considerable, as documented by the presence of more than 100 archaeological sites; paintings on cave walls and ceilings indicate pre-Hispanic settlements as well as remnants of the Royal Road, bridges, and estates (haciendas) from the 19th century. In this region, five early second-ary successional sites with differences in land-use history and time since abandonment were selected to establish the restoration experiments.

Figure 5.9

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (195)

169

Experimental analysis of dryland forest restoration techniques

(1) Land-use history and site effect on mixed plantations. Restoration assays were designed to evaluate the potential of selected tree species to establish and grow in fallows with different degrees of disturbance. Four restoration assays were established using six native tree species (Table 5.1). In September 2007, a total of 960 seedlings (60 per site, 10 of each selected species) were transplanted to each site. Plant survival and growth in basal diameter and height were then monitored every four months until October 2009 to include two dry and two rainy seasons.

Results indicated that seedling survival was statistically similar among species, with the exception of Cedrela odorata. After the first dry season (five months without precipitation), survival of all species was greater than 55%. However, after the second dry-season, survival was the lowest for C. odorata and the highest for Ceiba aescutifolia and Guazuma ulmifolia (Fig. 5.10). Following drought periods, Tabebuia rosea, Ipomoea wolcottiana, and Luhea candida featured an average survival rate lower than 50%. The relative growth rate in height (RGRh) was statis-tically similar among species, whereas growth in diameter (RGRd) differed (Table 5.2). C. odorata had the highest RGRd, followed by G. ulmifolia, T. rosea, and I. wol-cottiana. C. odorata, in spite of its excellent growth performance, appeared unable to adapt to drought conditions. G. ulmifolia maintained a high survival percentage owing to re-sprouting of apparently dead individuals during the rainy season. C. aesculifolia showed a slow growth because its stem top frequently broke during the dry season, recovering when the rainy season began. For all species, there was a trend toward better performance in old fields with less disturbance and greater age since abandonment. Site differences are the result of previous land uses (cropland and cattle pastures) and time since abandonment (between 1 month and 5 years before experiments started).

Table 5.2

Species RGRh (cm/cm/mo) RGRd (mm/mm/mo)

Rainy Dry Rainy Dry

Cedrela odorata 0.24 (0.03) –0.02 (0.02) 0.13 (0.027) 0.02 (0.007)

Ceiba aesculifolia 0.05 (0.007) –0.006 (0.01) 0.05 (0.004) 0.002 (0.003)

Guazuma ulmifolia 0.18 (0.03) –0.03 (0.01) 0.10 (0.01) 0.01 (0.003)

Ipomoea wolcottiana 0.18 (0.02) –0.02 (0.005) 0.11 (0.01) 0.008 (0.005)

Luehea candida 0.04 (0.01) –0.016 (0.007) 0.04 (0.01) 0.007 (0.003)

Tabebuia rosea 0.09 (0.02) 0.01 (0.007) 0.06 (0.01) 0.023 (0.007)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (196)

170

Principles and Practice of Forest Landscape Restoration

Figure 5.10

(2) Performance of species selected by local people. An experimental mixed plantation with eight woody native species selected by local people (Table 5.1) was established in the 2008 rainy season for evaluation of initial seedling performance. C. odorata was also evaluated in an enrichment plantation. Seedlings of these species were pro-duced in local backyards and transplanted to the field (Xocotitla, 48 months fallow). Survival, growth (height and diameter), re-sprouting, and herbivory (leaf and stem) were evaluated every four months until October 2009. Stem girdling in C. odorata and leaf necrosis in Maclura tinctorea were also recorded.

The mixed-species plantation displayed a survival rate of 95–100% in five months; how-ever, following the 2009 dry season, survival decreased to 83–97% for most species except C. odorata. C. odorata had survival rates of 14 and 28% in mixed and enrichment plantations, respectively. In 2008, RGRd was statistically different among species; the highest growth was exhibited by Lysiloma acapulcense, Chloroleucon mangense, and C. odorata. However, for the next dry season (2009) there were no differences among species. In both periods, there were no differences in RGRh among species.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (197)

171

Experimental analysis of dryland forest restoration techniques

Cedrela odorata performance between mixed and enrichment plantations was similar in terms of RGRd, but RGRh was higher for the former (Table 5.3). There, after initial trans-planting, girdling stem was the main cause of mortality; during the 2008 rainy season, this type of damage was observed in 11% of individuals that died during 2009. In the enrichment plantation, girdling was present in only 2.3% of the individuals. Re-sprouting was observed in four species (C. mangense, L. acapulcense, Leucaena lanceolata and C. odorata). Damage by herbivores was low in Cordia alliodora, L. acapulcense, C. odorata, Diphysa carthagen-ensis, and Lysiloma divaricatum. The results suggest that C. odorata can be used to enrich secondary vegetation, while the other species have the potential for use in mixed planta-tions to restore disturbed areas.

Table 5.3

RGRd) RGRh)(mm mm–1 month–1) (cm cm–1 month–1)

Species Norte 2008 Dry 2009 Norte 2008 Dry 2009

Chloroleucon mangense 0.122 (0.02)a –0.008(0.01)a 0.051 (0.02)a 0.044 (0.01)a

Diphysa carthagenensis 0.085 (0.03)ab –0.016(0.03)a 0.028 (0.03)a –0.016 (0.05)a

Lysiloma acapulcense 0.130 (0.05)a 0.001(0.01)a 0.004 (0.03)a 0.037 (0.01)a

Lysiloma divaricatum 0.075 (0.04)ab 0.001(0.05)a 0.050 (0.06)a 0.021 (0.03)a

Leucaena lanceolata 0.059 (0.03)ab 0.010(0.01)a 0.021 0.02)a 0.011 (0.02)a

Cordia alliodora 0.021 (0.03)b 0.015(0.03)a –0.007 (0.0)a –0.017 (0.07)a

Maclura tinctorea 0.073 (0.03)ab 0.016(0.01)a 0.042 (0.02)a 0.002 (0.02)a

Cedrela odorata 0.110 (0.04)a 0.003(0.02)a 0.015 (0.00)a 0.046 (0.02)a

Cedrela odorata (E) 0.151 (0.02) –0.023(0.03) –0.002 (0.00) 0.054 (0.030)

Conclusions

All selected species can potentially be used for restoration; however, an appropriate transplant-ing site must be chosen for each particular species. Timber species (C. odorata and T. rosea), which are the species most valued by local people, are severely drought intolerant. Therefore, they have to be planted under nurse trees that help to maintain humidity. In contrast, C. aescu-tifolia and G. ulmifolia are drought tolerant and may be used in disturbed areas with no woody vegetation; their re-sprouting ability is an advantage that allows establishment and survival in extremely dry conditions. I. wolcottiana and L. candida can be used in disturbed areas.

Local people are not enthusiastic about non-timber species, but these are important for restoration because they can change the microenvironment, providing suitable conditions for establishment of other species of economic importance. Enrichment planting is recom-mended in early successional sites that lack both non re-sprouting and key primary tree spe-cies (e.g. C. odorata or other multipurpose, foraging, and fruit species). In mixed plantations, C. odorata is one of the most valuable tree species in the area; it can be selected if irrigation is available and the density of transplanted seedlings is low.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (198)

172

Principles and Practice of Forest Landscape Restoration

It is clear that restoration success is directly related to site conditions, since the best spe-cies performance was recorded in the least disturbed sites with isolated shrubs or trees that act as nurse individuals. Transplantation of seedlings to disturbed sites is an important technique to accelerate restoration. Selection of tree species for restoration should take into account local knowledge and site characteristics.

Oaxaca, Mexico

The study area is located in the Mixteca Oaxaqueña and Central Valley regions. Experiments were established in the municipalities of Asunción Nochixtlán, Santo Domingo Yanhuitlán, San Pedro and San Pablo Teposcolula in the Mixteca region (Fig. 5.11). The area is divided into three precipitation zones (annual precipitation of more than 900 mm, between 600 and 700 mm, and less than 550 mm). The vegetation types are oak forest, pine forest, grassland, and shrubland. In the Mixteca region, plant cover is less than 25%, with dispersed patches of shrubs growing on hilltops (Cruz-Cruz, 2005). Soil erosion is present in 59% of the area with a soil loss rate of over 50 tonne ha/yr (Romero et al., 1986). Indigenous groups (Mixtec) inhabit this region; their livelihood is primarily based on traditional forms of agriculture and the use of natural forest resources. However, this region represents one of the most extreme cases of environmental degradation of drylands in Mexico. The main causes of degradation include high livestock density, agricultural expansion, unsustainable rangeland management practices, and high deforestation rates.

Figure 5.11

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (199)

173

Experimental analysis of dryland forest restoration techniques

(1) Native shrubs in four substrate types. The field experiments focused on the performance of native shrubs in four substrate types. The studied species appear in Table 5.1. The treatments differed in the soil characteristics of the four study sites. The substrate types were as follows:

sedimentary red soil (Yanhuitlán Formation, Montmorillonite clay), pH >8.0, organic matter (OM) 1.2%, carbonates 31.9%, and high erosionsedimentary limestone soil (white soil), pH >8.0, OM 1.3%, cemented layer by carbonates, and shallow soilvolcanic tuff soil, pH 6.8, OM 1.9%, low carbonate content, and shallow soilblack soil (volcanic soil from andesitic material), pH 6.8, OM 6.1%

Field experiments were established in July 2008 (rainy season) and were monitored dur-ing August and November 2008 and May and November 2009. The recorded variables were survival, height, cover, and biomass growth rate.

The survival of native shrubs decreased during the 15-month evaluation. Desmodium gra-hamii showed over 95% survival in all substrates and Acacia angustissima over 90% survival; whereas over 80% of Eysenhardtia polystachya and Dodonaea viscosa survived. Amelanchier denticulata and Cercocarpus fothergilloides displayed lower survival in black soils and volcanic tuff soil. Both collections of A. denticulata had only 37% plant survival in black soil. C. fothergil-loides survived at a rate of 75% in black soil and 70% in volcanic tuff soil. A general trend was that species had lower survival in black soil due to the type and content of clay (20%) and in volcanic tuff possibly due to the high sand content (>55%). During the dry season (November to May), the black soil cracked and the root system was damaged by the expansion and contraction of the soil.

In the four substrates, plant height for D. orbiculare, E. polystachya, and D. viscosa was superi-or to that of A. denticulata, A. angustissima, and C. fothergilloides. Among substrates, in volcanic tuff and black soil plants grew more than in Yanhuitlán formation and limestone soil. This differ-ence relates to the organic matter and nutrient content. The species included in the experiment are capable of re-growth, as observed at the four study sites. Plant growth rate per week differed among substrates. C. fothergilloides reached the highest value in poor nutrient content limestone soil yet showed no nutrient deficiency; consequently, C. fothergilloides can be recommended for this soil type. D. orbiculare, A. angustissima and E. polystachya showed high plant growth rates in volcanic tuff and D. orbiculare, A. denticulata, and C. fothergilloides in the Yanhuitlán forma-tion (Table 5.4). The results suggested that the species showed different responses in each soil type and probably displayed certain adaptations to these respective environments.

Table 5.4

Species Limestone soil Growth rate (cm wk–1)

n Volcanic tuff Growth rate (cm wk–1)

n Black soil Growth rate (cm wk–1)

n Yanhuitlán formation Growth rate (cm wk–1)

n

Desmodium orbiculare

0.14(0.04) 12 0.62(0.09) 27 0.30(0.06) 18 0.76(0.50) 6

Eysenhardtia polystachya

0.25(0.08) 15 0.33(0.08) 22 0.16(0.06) 13 0.20(0.05) 23

Dodonaea viscosa

0.34(0.04) 28 0.12(0.01) 35 0.25(0.02) 38 0.26(0.05) 25

Amelanchier denticulata Y

0.24(0.02) 30 0.15(0.02) 26 0.07(0.01) 12 0.24(0.02) 36

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (200)

174

Principles and Practice of Forest Landscape Restoration

Species Limestone soil Growth rate (cm wk–1)

n Volcanic tuff Growth rate (cm wk–1)

n Black soil Growth rate (cm wk–1)

n Yanhuitlán formation Growth rate (cm wk–1)

n

Amelanchier denticulata T

0.37(0.04) 34 0.15(0.02) 23 0.15(0.03) 12 0.48(0.02) 32

Acacia angustissima

0.23(0.06) 7 0.49(0.13) 8 0.13(0.05) 5 0.29(0.11) 8

Cercocarpus fothergilloides

0.65(0.07) 24 0.27(0.04) 22 0.15(0.03) 21 0.43(0.04) 12

(2) Nursery conditions. The same species and the four substrates mentioned above (plus a control) were grown in nursery conditions. On March 2008, seeds were sown in plastic bags. The recorded variables were days to emergence, number of stems, plant height, plant cover, and dry matter per plant. Native shrubs showed significant differences in several variables (Ta-ble 5.5). The number of days to emergence was different among species. Three species emerged in less than 17 days, while A. denticulata emerged after 60 days. D. orbiculare developed more stems per plant (2.4) than C. fothergilloides, D. viscosa, and A. denticulata (1.0). The multi-stem production of species at the soil surface level is an advantage for soil erosion control and rain water infiltration. D. orbiculare and A. angustissima reached the highest plant cover (>270 cm2/plant), while both collections of A. denticulata had the lowest values (<13 cm2/plant). Plant cover was related to dry matter production per plant. D. orbiculare (0.53 g/day) and A. angustissima (0.27 g/day) showed the highest growth rates, while the lowest values were recorded for the collections of A. denticulata and C. fothergilloides (<0.045 g/day) (Table 5.5).

Table 5.5

SpeciesEmergence (days)

Stems (number)

Cover (cm2/plant)

Dry matter (gr/plant)

Plant growth (gr/wk)

Desmodium orbiculare 8.90a 2.41a 272.77a 7.23a 0.53a

Acacia angustissima 10.15a 1.11bc 272.63a 4.67b 0.27b

Eysenhardtia polystachya 17.05ab 1.26b 129.38b 2.16c 0.22c

Cercocarpus fothergilloides 18.40ab 1.00c 17.26c 0.35d 0.05d

Dodonaea viscosa 21.20b 1.00c 108.45b 2.87c 0.19c

Amelanchier denticulata T 61.90c 1.00c 11.99c 0.32d 0.04d

Amelanchier denticulata Y 64.55c 1.01c 12.74c 0.41d 0.05d

The number of days to emergence varied for native shrubs planted in different substrates. Seedling appearance (24 days) was faster in black soil and in the control. In limestone and Yanhuitlán formation soils, it took 33 days (Table 5.6). Black soil and the control were associ-ated with the production of more than 1.3 stems per plant. For plant cover, dry biomass, and plant growth rate, limestone soil and Yanhuitlán formation soil types limited plant growth. The lowest values of all recorded variables corresponded to Yanhuitlán formation soil (Table 5.6). Dry matter production per plant was closely related to plant cover and growth. Dry mat-ter values were higher in black soil and in the control than in limestone and the Yanhuitlán formation, a pattern similar to plant growth rates.

Table 5.4

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (201)

175

Experimental analysis of dryland forest restoration techniques

Table 5.6

Substrate Emergency (days)

Stems (number)

Cover (cm2/plant)

Dry matter (g/plant)

Growth (cm/wk)

Black soil 24.60b 1.33a 228.25a 5.65a 0.36a

Control 24.75b 1.34a 219.54a 5.37ª 0.35a

Volcanic tuff soil 27.89ab 1.24ab 67.15b 0.88b 0.13b

Limestone soil 33.32ª 1.22ab 42.18bc 0.55b 0.06c

Yanhuitlán Formation 33.82ª 1.13b 32.31c 0.41b 0.06c

(3) Vegetative reproduction. A vegetative reproduction experiment was established in a greenhouse with two shrub species, Cercocarpus fotergilloides and Amelanchier denticulata. Each stem was treated with Indol–3-butiric acid 10,000 ppm (Radix 10,000). Sprouting and rooting were harvested in March and evaluated in November 2009. The vegetative reproduction experiments showed that C. fothergilloides cuttings did not develop roots in any of the treatments. In contrast, A. denticulata cuttings developed roots with the following treatments: 17–33 % in Yanhuitlán collection, from 2.0 cm cutting diameter, and 17% in Yanhuitlán collection, 1.0 cm cutting diameter.

Conclusions

Since the long dry season affects seedling survival, particularly in black and volcanic tuff soils, it is advisable to transplant at the beginning of the rainy season in order to allow the development of a better root system. More research is recommended on seed germination, growth of native tree species, and vegetative reproduction from cuttings.

Conclusions

In view of the wide variety of regional differences in the drylands of America, and the relatively small proportion of area covered in this project, no attempt has been made to produce general recommendations. Nevertheless, the different dryland restoration experiences evaluated in this chapter allowed us to discern the factors influencing successful forest restoration and to produce some specific recommendations. The results of field experiments established in each study area, along with other studies, led to the identification of some key ecological processes that limit the establishment and growth of threatened and/or socioeconomically important native tree species found in dry forests. Furthermore, these experiences permitted the identification of restoration techniques that overcome some of those constraints. Thus, based on the factors that influenced successful restoration, the major findings have been grouped and presented below.

In all of the study areas from Mexico to Patagonia, the main limiting factor for establish-ment was reported to be drought. This finding may be obvious but is not trivial. Air tempera-ture is not a limiting factor except in south western Argentina during the winter, but the irregularity of precipitation is a problem. Restoration efforts in the drylands of America have to confront a common environmental constraint, namely the long dry season affecting seed-ling survival during any transplanting effort. In order to permit the development of a better root system, transplantation should take place at the beginning of the rainy season – should

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (202)

176

Principles and Practice of Forest Landscape Restoration

be planned in accordance with the incidence of El Niño rainy years. Results in different study areas suggest the use of supplemental irrigation for plantations established on open sites during the summer, even when precipitation has been normal that year.

The use of nurse species proved to be important for protecting seedlings from desicca-tion, thus improving seedling survival and initial growth. In Patagonia, nurse plants provide protection against extremely cold weather and soil frost during winter.

Another technique suggested was the use of terraces to facilitate plant survival. This in-creases rainfall collection and concentrates moisture at the bottom of the furrow. However, in Oaxaca few communities use them because of their high cost (Box 5.4).

Since the effect of herbivores is sufficient to kill most seedlings, the results of experi-ments in all study areas and under any environmental condition suggest that it is essential to exclude cattle and small mammals before starting a restoration effort. Exclosures were not implemented in only one case – when extended plantations were established.

The use of native tree species was preferred in all study cases. However, employing exot-ic species is recognized as important in some circ*mstances, such as in very degraded sites or mixed plantations. It is noteworthy that native tree species displayed higher survival rates than exotic species, especially during the dry season. Preferably, the choice of native species should be based on their drought tolerance as well as on the site disturbance level. In some forests, native species need to be selected because of their re-sprouting ability. On occasions, particular species must be chosen for a given transplanting site; locally impor-tant site factors include soil type (e.g. Oaxaca, Mexico) or location on the slope (Patagonia, south western Argentina).

Appropriate methods for restoring dryland forest ecosystems are those that contribute both to the conservation and restoration of biodiversity and to local economic development. It is suggested that seedling transplant restoration techniques should be applied in defor-ested/degraded areas and seedling transplant enrichment employed in logged forest and old successional sites. Results from most of these field experiments suggested that natural regeneration can be encouraged by protecting successional areas from cattle (exclusions for livestock/small mammals), fire, and selective cutting; enrichment planting is an appropriate method in early successional sites lacking non re-sprouting and key primary tree species; and mixed species plantations can be established on highly degraded sites.

When planning a restoration strategy, it is important to include species of economic and social value as well as endangered species. Local knowledge must be taken into account for the selection of tree species; local people must participate in the selection and be aware of the importance of forest recovery owing to the environmental services that it provides. In-formation on dry forest native tree species has to be made available to local people through workshops and guided walks. Also, demonstrative restoration assays should be displayed in school yards or communal land for educational purposes. Training in tree species propaga-tion is essential in order to increase the establishment and management of native species plantations.

A lack of knowledge of the biology of native tree species and secondary successional processes limits their implementation in management and conservation plans. Information about native trees is restricted to a few species; it is therefore paramount to carry out plant species research that addresses their phenology, seed dispersal, germination, growth, and

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (203)

177

Experimental analysis of dryland forest restoration techniques

Tabebuia rosea in central Veracruz, Mexico. Photo: C. Alvarez

vegetative reproduction. Along with biological studies on native tree species, another sug-gested approach to improving restoration techniques is to learn from studies of secondary succession.

Dryland restoration requires more time than the duration of the experiments described here. However, the experience that has been obtained, based on comparison between and within countries, studying specific cases and different potential solutions, represents a first step in identifying general approaches to the restoration of dry forest ecosystems. From these experiences, we recognize that dryland restoration requires more research since it is not possible to directly apply techniques learned from experiences in other forest types in more humid environments.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (204)

178

Principles and Practice of Forest Landscape Restoration

References

Alfaro Arguello, R. 2008. Sustentabilidad del manejo ganadero holístico y convencional en el trópico seco de Chiapas, México. Master’s thesis. El Colegio de la Frontera Sur, San Cristóbal de Las Casas, Chiapas, Mexico.

Aronson, J., Vallauri, D., Jaffre, T., Lowry P.P. 2005. Tropical dry forest restoration. In: Mansouri-an, S., Vallauri, D., Dudley N. (eds.), Forest restoration in landscapes: beyond planting trees. Springer, New York, USA: pp. 285–290.

Caso, M., Gonzalez-Abraham, C., Ezcurra, E. 2007. Divergent ecological effects of oceano-graphic anomalies on terrestrial ecosystems of the Mexican Pacific coast. Proceedings of the National Academy of Sciences of the United States of America 104: 10530–10535.

Ceccon, E., Hernández, P. 2009. Seed rain dynamics following disturbance exclusion in a secondary tropical dry forest in Morelos, Mexico. Revista de Biologia Tropical 57: 257–269.

Cruz-Cruz, E. 2005. Morphological variability and seed dormancy of Amelanchier denticu-lata (Rosaceae) grown in Oaxaca, Mexico. Ph.D Dissertation. Oregon State University. Cor-vallis, OR, USA. 200 pp.

Ferguson, B.G., Diemont, S.A.W., Alfaro Arguello, R., Martin, J.F., Nahed Toral, J., Álvarez Solís, J.D., Pinto Ruíz, R. Sustainability of holistic and conventional cattle ranching in the season-ally dry tropics of Chiapas, Mexico. Agriculture, Ecosystems and Environment. In review.

Flannery, K.V. 1983. Precolombian farming in the Valleys of Oaxaca, Nochixtlán, Tehuacán, and Cuicatlán: A comparative study. In: Flannery, K. V., Marcus, J. (eds.), The cloud People: divergent evolution of the Zapotec and Mixtec civilizations. Academic Press. New York, USA: pp. 323–339.

Garibaldi, A., Turner, N. 2004. Cultural keystone species: implications for ecological conservation and restoration. Ecology and Society 9: 1–18. www.ecologyandsociety.org/vol9/iss3/art1.

González-Espinosa, M., Ramírez-Marcial, N., Camacho-Cruz, A., Holz, S.C., Rey-Benayas, J.R., Parra-Vázquez. M.R. 2007. Restauración de bosques en territorios indígenas de Chiapas: Modelos ecológicos y estrategias de acción. Boletín de la Sociedad Botánica de México 80 (Supplement): 11–23.

Griscom, H.P., Ashton, P.M.S., Berlyn, G.P. 2005. Seedling survival and growth of native tree species in pastures: Implications for dry tropical forest rehabilitation in central Panama. Forest Ecology and Management 218: 306–318.

Griscom, H.P., Griscom, B.W., Ashton, M.S. 2009. Forest regeneration from pasture in the dry tropics of Panama: effects of cattle, exotic grass, and forested riparia. Restoration Ecology 17: 117–126.

Hoffmann, A.J., Armesto, J.J. 1995. Modes of seed dispersal in the Mediterranean regions in Chile, California and Australia. In: Arroyo, M.T.K., Zedler, P.H., Fox, M.D. (eds.), Ecology and biogeography of Mediterranean ecosystems in Chile, California and Australia. Springer-Ver-larg, New York, USA: pp. 289–310

Holl, K.D. 1999. Factors limiting tropical rain forest regeneration in abandoned pasture: seed rain, seed germination, microclimate, and soil. Biotropica 31: 229–242.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (205)

179

Experimental analysis of dryland forest restoration techniques

Holmgren, M., Scheffer, M., Ezcurra, E., Gutierrez, J.R., Mohren, G.M.J. 2001. El Niño effects on the dynamics of terrestrial ecosystems. Trends in Ecology and Evolution 16: 89–94.

Izhaki, I., Safriel, U.N. 1990. The effect of some Mediterranean scrubland frugivores upon germination patterns. Journal of Ecology 78: 56–65.

Janzen, D.H., Martin, P.S. 1982. Neotropical anachronisms – the fruits the gomphotheres ate. Science 215: 19–27.

Kaimowitz, D. 1996. Livestock and Deforestation. Central America in the 1980’s and 1990’s: a policy perspective. Center for International Forestry Research: Jakarta.

Kirkby, A.V.T. 1973. The use of land and water resources in the past and present valley of Oaxaca, Mexico. In: Flannery, V. (ed.), Prehistoric and human ecology of the Valley of Oax-aca. Michigan, USA. Memories of the Museum of Anthropology University of Michigan 1. 174 pp.

Lamb, D., Gilmour, D. 2003. Rehabilitation and restoration of degraded forests. IUCN, WWF, Gland, Switzerland and Cambridge, UK.

Lindenmayer, D.B., Franklin, J.F. 2002. Conserving forest biodiversity. Island Press, Washington.

Mansourian, S., Vallauri, D., Dudley, N. (eds.). 2005. Forest restoration in landscapes: beyond planting trees. Springer-WWF. New York, USA.

Mayhew, J., Newton, A.C. 1998. Silviculture of mahogany. CABI Bioscience, Oxford.

Miceli-Méndez, C.L., Ferguson, B.G., Ramírez-Marcial, N. 2008. Seed dispersal by cattle: natural history and applications to neotropical forest restoration and agroforestry. In: Myster, R. (ed.), Post-Agricultural Succession in the Neotropics. Springer, New York: pp. 165–191.

Montagnini, F. 2005. Selecting tree species for plantation. In: Mansourian, S., Vallauri, D., Dud-ley, N. (eds.), Forest restoration in landscapes: beyond planting trees. Springer-WWF. New York: pp. 262–268

Parrotta, J.A., Turnbull, J.W., Jones, N. 1997. Catalyzing native forest regeneration in degraded tropical lands. Forest Ecology and Management 73: 271–277.

Paulsen, T.R., Högstedt, G. 2002. Passage through bird guts increases germination rate and seedling growth in Sorbus aucuparia. Functional Ecology 16: 608–616.

Piotto, D., Viquez, E., Montagnini, F., Kanninen, M. 2004. Pure and mixed forest plantations with native species of the dry tropics of Costa Rica: a comparison of growth and produc-tivity. Forest Ecology and Management 190: 359–372.

Quesada M., Sanchez-Azofeifa G.A., Alvarez-Añorve, M., Stoner K.E., Avila-Cabadilla L., Calvo-Alvarado J., Castillo A., Espírito-Santo M.M., fa*gundes M., Fernandes G.W., Gamon J., Lope-zaraiza-Mikel M., Lawrence D., Cerdeira Morellato L.P., Powers J.S., Neves F. de S., Rosas-Guerrero V., Sayago R., Sánchez-Montoya G. 2009. Succession and management of tropical dry forests in the Americas: Review and new perspectives. Forest Ecology and Manage-ment 258: 1014–1024.

Reid, S. 2008. Interaction dynamics of avian frugivores and plants in a subandean sclerophyl-lous shrubland of central Chile: implications for seed dispersal and regeneration patterns. Ph.D. Thesis. Pontificia Universidad Católica de Chile, Santiago, Chile.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (206)

180

Principles and Practice of Forest Landscape Restoration

Rey Benayas, J.M., Newton, A.C., Diaz, A., Bullock, J.M. 2009. Enhancement of Biodiversity and Ecosystem Services by EcologicalRestoration: A Meta-Analysis. Science 325: 1121–1124.

Roman-Cuesta, R.M., Gracia, M., Retana, J. 2003. Enviromental and human factors influencing fire trends in ENSO and non-ENSO years in tropical Mexico. Ecological Applications 13: 1177–1192.

Romero, F.M.A. 1990. Economía y vida de los españoles en la Mixteca alta: 1519–1720. Colec-ción Regiones de México. Instituto Nacional de Antropología e Historia. México, D.F. Gobi-erno del Estado de Oaxaca. 636 pp.

Romero, P.J., García, L., Martínez, J.C., Ramírez, C., Valencia, R., Reyes, F., Ramos, M.T. 1986. Diagnóstico de la producción agrícola en las Mixtecas Oaxaqueñas Alta y Baja. UACH-CONACYT, Chapingo, México. 1006 pp.

Ruiz, M.M. 1996. Las plantaciones forestales en la Mixteca Oaxaqueña. Oaxaca, Mexico. Docu-mento Interno. CIRPS-INIFAP.

Sánchez, M.D. Rosales, M., Murgueitio, E., 2003. Agroforestería pecuaria en América Lati-na. In: Sánchez. M.D., Rosales Méndez, M. (eds.), Agroforestería para la Producción Ani-mal en América Latina. FAO-CIPAV: Rome.< http://www.fao.org/DOCREP/006/Y4435S/Y4435S00.HTM>.

Savory, A., Butterfield, J. 1999. Holistic Management. A New Framework for Decision Making. Island Press, Covelo, California, USA.

Stringham, T.K., Krueger, W.C., Shaver, P.L. 2003. State and transition modeling: an ecological process approach. Journal of Range Management 56: 106–113.

Suárez, A., Williams-Linera, G., Trejo, C., Valdez-Hernández, J.I., Cetina-Alcalá, V., and Vibrans, H. Local knowledge helps select species for forest restoration in a tropical dry forest of central Veracruz, Mexico. Agroforestry Systems. In review.

Szott, L., Ibrahim, M., Beer, J. 2000. The Hamburger connection hangover: cattle pasture, land degradation and alternative land use in Central America. Centro Agronómico Tropical de Investigación y Enseñanza. Turrialba, Costa Rica.

Tarrasón, D., Urrutia, J. T., Ravera, F., Herrera, E., Andrés, P., Espelta, J. M. 2010. Conservation status of tropical dry forest remnants in Nicaragua: Do ecological indicators and social perception tally? Biodiversity and Conservation 19: 813–827.

Traveset, A., Riera, N., Mas, R.E. 2001. Passage through bird guts causes interspecific differ-ences in seed germination characteristics. Functional Ecology 15: 669–675.

Traveset, A., Robertson, A.W., Rodríguez-Pérez, J., 2007. A review on the role of endozoochory in seed germination. In: Dennis, A.J., Schupp, E.W., Green, R.J., Westcott, D.W. (eds.), Seed dispersal: theory and its application in a changing world. CABI Publishing, Wallingford, UK: 78–101.

Uasuf, A. Tigabu, M., Oden, PC. 2009. Soil seed banks and regeneration of Neotropical dry de-ciduous and gallery forests in Nicaragua. Bois et Forets des Tropiques 299: 49–62.

Vieira, D.L.M., Scariot, A., Holl, K.D. 2007. Effects of gap, cattle and selective logging on seed-ling survival and growth in dry forests of Central Brazil. Biotropica 39: 269–274.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (207)

181

Experimental analysis of dryland forest restoration techniques

Vieira, D.L.M., Scariot, A. 2006a. Effects of logging, liana tangles and pasture on seed fate of dry forest tree species in Central Brazil. Forest Ecology and Management 230: 197–205.

Vieira, D.L.M., Scariot, A. 2006b. Principles of natural regeneration of tropical dry forests for restoration. Restoration Ecology 14: 11–20.

Villafuerte, D., García, M.C., Meza, S. 1997. La cuestión ganadera y la deforestación: viejos y nuevos problemas en el Trópico y Chiapas. Universidad de Ciencias y Artes del Estado de Chiapas. Tuxtla Gutiérrez, Mexico.

Walker, L.R., Walker, J., Hobbs, R.J. (eds.). 2007. Linking restoration and ecological succession. Springer, New York.

Williams-Linera, G., Lorea, F. 2009. Tree species diversity driven by environmental and anthro-pogenic factors in tropical dry forest fragments of central Veracruz, Mexico. Biodiversity and Conservation 18: 3269–3293.

Zahawi, R.A. 2005. Establishment and growth of living fence species: An overlooked tool for the restoration of degraded areas in the tropics. Restoration Ecology 13: 92–102.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (208)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (209)

183

6 SOCIOECONOMIC VALUATION OF

DRYLAND FOREST RESOURCES IN DRY

AREAS OF ARGENTINA, CHILE AND

MEXICO

R.F. del Castillo, R. Aguilar-Santelises, C. Echeverría, E. Ianni, M. Mattenet,G. Montoya Gómez, L. Nahuelhual, L.R. Malizia, N. Ramírez Marcial, I. Schiappacasse, C. Smith-Ramírez, A. Suárez, G. Williams-Linera

Introduction

Extensive areas in Latin America, arid or semi-arid, are the homeland of millions of people, and have been inhabited since pre-Columbian times (UNDP, 2004). Despite their harsh envi-ronmental conditions, these areas are extremely diverse and abundant in plants with actual or potential economic value (Newton, 2008). The value of these plants has been documented for centuries in several classical books such as those of Francisco Hernández (1659) or Maximino Martínez (1936) in México, just to mention a few examples. Despite the long tradition of collecting knowledge about native plants, habitat destruction and soil erosion prevail in many of these areas, constituting a severe threat not only to biodiversity but to the well-being of the local people. Documenting the socioeconomic relationships between local people and native plants is clearly essential for proposing any cogent plan for forest restoration and conservation in these areas. The present chapter summarizes the findings of socioeconomic studies on native dry forest in Latin America and provides recommendations to be taken into account for implementing restoration or conservation plans. The follow-ing text summarizes the results obtained from a variety of different socioeconomic surveys conducted in the different study areas as part of the ReForLan project. Further details of the individual studies, together with some information from other relevant research initiatives, are provided in Boxes 6.1–6.6.

Results of socioeconomic surveys

Patterns of plant use and indigenous knowledge

Native plant species of dry forests in Latin America can be an important source of products, which can contribute to improve the well-being of local human communities. These prod-ucts are used in general for subsistence and, to a lesser extent, for commercial purposes. However, many native species that are recognized as having potential uses are not currently being used. In fact, we found that, in some areas, local people are not always aware of all the benefits of native species found in dry forest areas of Latin America.

Awareness of the importance of native plant species of dryland forests varies considerably among regions and even among people within the same region. In central Chile, for instance, very few of the sclerophyllous forest species traditionally known as sources of medicine, food,

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (210)

184

Principles and Practice of Forest Landscape Restoration

and fibre were cited in the interviews conducted with local people. These results show that the knowledge of traditional uses of sclerophyllous species by rural inhabitants of central Chile is very limited and has been gradually lost over time. By contrast, in Paso de Ovejas in central Veracruz, Mexico, a region with mainly secondary vegetation and some remnants of tropical dry forest, data compiled from different sources (workshops, in-depth interviews with key informants, field walks with informants, and botanical collections) documented 76 species in one or more categories of use, from primary, secondary, agroforestry, and riparian habitats. In northwestern Argentina, a quantification of the use of medicinal plants in Kolla communities revealed that 117 plant species belonging to 52 families and 98 genera are being used (Hilgert, 2001).

In the dry areas of the Upper Mixtec Region in Oaxaca southern Mexico, a survey of 322 people on the uses of a sample of 112 native local plant species revealed that all species were recognized as useful by at least some of the interviewees. However, only 13 species were judged to be useful by 60% of them; and, on average, only 33% of the species shown were identified as useful to the interviewees. In the sclerophyllous forests of central Chile, a total of 12 species were recognized by local people as being potentially valuable. Overall, the Quillay, Quillaja saponaria, is preferred by most of the people with 51.9% preference, and used as source of honey and as a shade plant.

Knowledge about native plants among local people is unevenly distributed among the population and is usually concentrated among very few people. This knowledge is also irregularly distributed even among different municipalities within the same district. In the Upper Mixtec Region in southern Mexico, the studied municipality with the lowest level of formal education, economic welfare, and health services displayed the highest use value of native plant species among respondents. As expected, prevalence of local traditions and culture was strongly related to knowledge of native plants. As formal education is becoming more common and the native language (Mixtec) is being lost, awareness of the importance of native plants species is also being lost. Thus, rescuing traditional knowledge can be an indirect way of preserving native plant species. At a global level, it has been estimated that 80% of the cultural knowledge of indigenous people will disappear in the next one hundred years (Inter-Commission Task Force on Indigenous Peoples, 1997). This cultural loss will certainly jeopardize native forest species since few people will recognize the value of such species. This loss is likely a major driver of species extinction.

It is important to recall that the richness of traditional heritage stems not only from knowledge of a particular use for a particular native plant species, but from the diversity of uses given to such a species. Thus, in surveys assessing traditional knowledge, it is essential to evaluate not only the number of species with a use value for the local people but also the diversity of uses the community gives to native species. For example, in the dry forest of the Upper Mixtec Region, Oaxaca, Mexico, the following uses attributed by local people to native plants have been identified, in declining order of importance: forage for domestic animals, medicine for both domestic animals and humans, a source of energy (firewood and charcoal), and food for humans. Suárez et al. (submitted) reported that in Paso de Ovejas, Veracruz, Mexico, farmers identified a total of 12 uses for 76 native species (mostly in the Fabaceae family). Workshop participants agreed that the most important uses in terms of quantity were fuel and fence posts.

Firewood is the main energy source for heating and cooking, and is one of the most com-mon uses of native dry forest (see Boxes 6.1 and 6.2). On communal lands, firewood is usu-

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (211)

185

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

ally free for local people, but the efforts to replenish firewood extraction in native forests or plantations are nil or insufficient. In the Colliguay Valley, Chile, firewood, charcoal and organic soil are the most commercialized products of dry forests. Yet they have the lowest added value and their extraction has the highest impact on forests. Indeed, they contribute only 6.7% to total family income (Schiappacasse et al., 2009).

Recently, in the Upper Mixtec region in municipalities such as Santiago Tilantongo, planta-tions of oaks have been established in degraded areas. These plantations can help to reduce the effects of harvesting for firewood, but more studies are needed to assess the amount of firewood needed per inhabitant and the capabilities of firewood production of native forests or plantations under different harvesting scenarios. Such studies are now being conducted for some localities such as Ocuilapa de Juárez, Chiapas, where firewood is also used for heat-ing ovens for pottery. Therefore, firewood demand can be very high and the supply of local forest can be insufficient. Clearly, establishment of plantations for energy and analysis of the production of firewood are urgently needed (Holz and Ramírez Marcial, in preparation).

In northwestern Argentina, seasonal dry forests (premontane Yungas and Chaco) provide firewood for most local communities living in or close to these forests. Collection and use are primarily performed on a per-family basis, accessing relatively small amounts each time but consistently throughout the year. For instance, in Los Naranjos, an indigenous Kolla com-munity settled in the premontane forests, a preliminary estimation showed that the com-munity (about 260 inhabitants) uses on an annual basis about 315 trees of different sizes for firewood, and about 500 for fence posts. Hardwood species from the Myrtaceae and Fabaceae families were preferred for firewood, while a wider range of species were used for fence posts, including Quina (Myroxylon peruiferum), Afata (Cordia trichotoma), Lapacho (Tabebuia spp.), Nogal (Juglans australis), Mato (Myrcianthes spp.), and Cebil (Anade-nanthera colubrina). A high intensity of firewood use was detected in the forest in a 2 km radius around the community, after which distance the intensity of use decreased sharply. We do not know whether the spatial distribution of firewood extraction in the forest has been stable over time. The community settled permanently in the mid 1980s, and no information relating intensity of use to forest growth is available.

Tree species with high potential for restoration

In most parts of central Chile, native dry forests are subjected to intense logging to satisfy the internal demands for forest products or to clear cutting for the expansion of agriculture (Newton et al., 2009; Schiappacasse et al., 2009). Boldo (Peumus boldus) and quillay (Quil-laja saponaria) have been identified as an attractive alternative for restoration programmes (Echeverría et al., 2010).

In the Upper Mixtec region, Juniperus flaccida was the most common native plant used, as 90% of the informants recognized this plant as useful. This conifer is a promising plant for restoration purposes, since it can grow in highly degraded areas, its wood is of high quality in particular for construction purposes, it is resistant to rotting, and it is very hard and durable. Furthermore, Juniperus is considered to be highly resistant to drought and can actively colo-nize degraded areas (Willson et al., 2008). Unfortunately, growth rate is slow. On the other hand oaks (Quercus spp.) are valuable alternative options for restoration as they are very useful as firewood. In particular, the yellow oak (Quercus liebmannii) is the most valued firewood species owing to its rapid ignition and its durability as a fuel.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (212)

186

Principles and Practice of Forest Landscape Restoration

Children carrying firewood to their homes in the central depression of Chiapas. Photo: N. Ramirez-Marcial

One of the productive activities of the inhabitants of Ocuilapa, Mexico, is clay pottery, which requires significant amounts of firewood for fuel. Photo: N. Ramirez-Marcial

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (213)

187

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

In central Veracruz, species such as Chloroleucon mangense, Lysiloma acapulcense, Leucaena lanceolata, Cedrela odorata, Caesalpinia cacalaco, Tabebuia chrysantha and T. rosea are used as living fences and for enrichment of native woody vegetation with linear plantations. Diphysa carthagenensis and Gliricidia sepium are also suitable plants for res-toration in this area owing to their multiple uses.

In northwestern Argentina, native species such as Cedrela spp., Tabebuia spp. and Tipua-na tipu have been recommended for restoration efforts, owing to their high growing rates and medium to high timber quality, with a direct economic return as timber over the mid-term (20–30 years). Such combinations of features make these species attractive for private and public investment in restoration.

In all of the studied regions, very few of the most important native species are available from nurseries. Therefore, establishment of nurseries for restoration activities is urgently needed.

Potential for forest restoration activities

Socioeconomic analyses conducted in northern Argentina in the Yungas region compared forest activities with other land-use alternatives, such as sugarcane and soy plantations. From an economic perspective, forest activities were much less profitable than the alternatives. Furthermore, of the potential forest species, native species were less profitable than exotics, and required a government subsidy. These analyses, however, did not consider the indirect environmental and social costs of crops, such as biodiversity reduction, water contamination and subsistence needs. Reforestation and forest-enrichment projects in Yungas could be an option to give more value to relatively unproductive lands. Such projects could also offer a way to diversify commercial activities.

Communal and private lands display different opportunities for restoration. In Chile, for example, people are eager to reforest but not on their own land. In Yanhuitlán in the Up-per Mixtec region, some severely degraded areas are privately owned, but the owners live abroad, and local authorities and government agencies although eager to reforest these areas cannot conduct any action without the consent of the absent landowners.

One of the most important benefits of forests to local communities is the ecosystem services that they provide. Recognition of these services could provide a powerful incentive for forest restoration actions. In Paso de Ovejas, Veracruz, the local population is highly aware of the varying functions of trees in the landscape, and values these species accordingly. In the Upper Mixtec region, owing to a chronic scarcity of water for the communities, local populations with the financial aid of federal government programmes are reforesting several areas with the aim of improving water infiltration and reducing runoff from hillsides. Reforestation can be identified even using remote sensing tools. Nevertheless, vast areas of the Upper Mixtec region remain highly eroded, and water scarcity still prevails in the region and is blamed in part on the high rates of forest destruction. Furthermore, reforestation actions have been commonly conducted using exotic species such as Casuarina sp. and Pinus spp., the long-term effects of which are unknown; local people do not recognize these species as useful. Frequently, reforestation activities are badly planned and lead to the expensive introduction of exotic species that far from helping, have questionable benefits for ecosystem function and conservation, and for the well-being of local communities.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (214)

188

Principles and Practice of Forest Landscape Restoration

ECOSUR conducted a socioeconomic valuation and market analysis of dryland forest re-sources in the southern state of Chiapas, Mexico, in the central region. This region does not have a formal market for forest products, although illegal extraction is common. Local stores and carpenters add value to forest products, but all the markets are small in scale. Demand is higher than supply and people in this region import forest products from other regions. Fire-wood is an important forest product, but it is vulnerable to overexploitation because of high demand from the craft-making sector. The long lead time required to obtain commercial ben-efits from forest products discourages forestry as a profitable activity for many landholders. In-stead, cattle and sheep ranching are commonly preferred. Different government programmes lead to contradictory impacts on native forest, supporting forestry activities on one hand and ranching and deforestation on the other. Such programmes have an important influence on the decisions made by landholders regarding their economic activities, resulting in a complex socioeconomic and political context for forest restoration in the central region of Chiapas.

Forest activities compete with cattle ranching and traditional cropping in virtually all sur-veyed areas. The agriculturally most valuable areas are usually used for economic activities other than forestry, which, in turn, is relegated to low quality sites such as steep sites or areas with poor soils. Good forestry practices in these low quality sites, however, can help improve the well-being of the local people by providing basic needs such as firewood, and ecosys-tems services such as enhanced water infiltration and soil retention to prevent soil erosion.

Conclusions

Dryland forest species in Argentina, Chile, and Mexico can be an important source of eco-nomic resources to local people and provide essential ecosystem services at the same time. Nevertheless, rates of forest loss and degradation are often high and jeopardize the existence of native forest species. The use value of many such species has been identified through interviews and workshops with local people. Although many native tree species of dryland forest in the study areas were recognized as useful, knowledge about the use value is being lost and, at least in some cases, it is unevenly distributed among local people. Furthermore, formal commercialization of such products is uncommon. Firewood is one of the most wide-spread uses in all the study areas, but, in general, actions aimed at replenishing the losses due to extraction are either non-existent or insufficient. We conclude that socioeconomic factors are major drivers of habitat destruction and deforestation in the studied drylands. These include the loss of awareness of the importance of native forest species among local people, disengagement of formal education from local knowledge and traditions, insufficient information on the potential economic or ecological importance of native plants, lack of commercialization channels for native forest products, conflicting governmental policies, the introduction of exotic species, and a lack of coordination among stakeholders involved in forest management and conservation. Based on a consideration of these problems, a series of recommendations are provided.

1. Awareness of the importance of dry forest as a source of goods and services for humans is being lost as traditional knowledge is supplanted by formal education. Thus, rescuing traditional knowledge is important for demonstrating the use value of native plants and to provide a motivation for preserving and restoring populations of native plant species

2. Formal education is disengaged from traditional knowledge and tends to ignore tradi-

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (215)

189

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

tional values, including the native language and the use value of native plants. Therefore, including local culture and traditions in formal education would indirectly support the conservation and restoration of native forest.

3. Native plants and forests are commonly overexploited, particularly for firewood (Oax-aca, Chiapas, Chile). Increased forest restoration actions are urgently needed in the study areas to counterbalance the current high rates of firewood exploitation.

4. Nurseries with local native plants are rare or absent in most cases. This is one of the reasons why reforestation programmes use exotic species in some regions of Latin America. Therefore, nurseries with local plants should be encouraged in all areas.

5. The lack of market channels for trading native forest goods and services is one of the reasons why native forest resources usually have little or no monetary value. In many cases local people do not receive a clear economic benefit for preserving or restoring native forests. Indeed, in Oaxaca and central Chile, formal commercialization of forest resources is uncommon. Commercial use of native forest plants, although less profitable than that of exotic species in some areas, can be an option for economic diversification, with lower economic and ecological costs than other economic activities. Exploitation of native dryland tree species can in some situations be profitable, such as the case of boldo in northern Chile. Similarly, timber extraction in northern Argentina is economi-cally important for many small to medium private companies and rural families. The current challenge is to identify, disseminate and adopt sustainable forest management practices and to develop local manufacturing to generate in situ welfare and working opportunities.

6. Government policies clearly play an important role in preserving, restoring or destroy-ing the forest. Conflicting governmental policies from different agencies reveal lack of coordination and hinder forest conservation and restoration actions.

7. Government reforestation and restoration programmes should attend to the needs of lo-cal people and preferably use native plants instead of exotic species, which may provide little or no economic benefit to the local communities.

8. Native species can have high international market demand as is the case of boldo and quillay in Chile. Clearly, more studies on native plants of dry forests in Mexico and in northern Argentina are needed to explore their potential economic benefits.

9. The importance of ecosystem services should be emphasized among local people and policy makers as an important incentive for conserving and restoring native forest areas in all regions. It is therefore crucial to promote the monetary valuation of ecosystem goods and services provided by dryland forests and to support rural livelihoods (see also Chapter 8).

10. Coordinated participation of stakeholders, namely federal and local authorities, academics, educators, and local people is essential for successful forest restoration programmes.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (216)

190

Principles and Practice of Forest Landscape Restoration

Cedrela odorata, Chiapas, Mexico. Photo: R. Vaca

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (217)

191

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

Box 6.1 Firewood consumption for pottery and projections for woody biomass production from Bursera simaruba

R. Hernandez, S.C. Holz and N. Ramirez-Marcial

Firewood is the main fuel source for many rural populations in developing countries. In Mexico, fuelwood accounts for 80% of the energy used in rural households. Chiapas is among the five states of Mexico in a critical situation at the junction between consumption and availability of woody energy. Given the importance of fuelwood in the rural Mexican context, there is a need to generate information to support sustainable forest management based on the many tree species used for firewood. Our research aimed to estimate: (i) the woody biomass production of B. simaruba, and (ii) the consumption of firewood used in local pottery.

Pottery is an activity practised throughout Mexico and represents an important avenue of artistic expression and cultural transmission in addition to being a means of supporting economic growth of the people involved. In the Central Depression of Chiapas one of the softwood species commonly used for baking pottery is Bursera simaruba (‘palo mulato’), a native tree species that is widely used in hedges. The work was undertaken in the community of Ocuilapa de Juarez, located in the Central Depression of Chiapas (16°53’52”–16º50’47”N, 93°27’28”–93º 24’17” W). To determine biomass production, we selected 40 individuals of B. simaruba of different sizes, and measured the total height, diameter at base (DAB) and diameter at breast height (dbh) of each individual, then determined their biomass through direct harvesting. We estimated firewood consumption during the burning (baked clay pieces) conducted in 13 pottery events.

The dry biomass production increased rapidly during the early years of growth, and the highest values were recorded from 20 years (up to 700 kg per tree). To predict woody biomass from the stem diameter of individuals, we identified a linear equation (r2 = 0.93, p <0.001). Firewood consumption varied widely among the different pottery burning events (39 to 295 kg), while the median consumption was 81.92 kg. Although many species were used during burning, the most commonly used was B. simaruba. From the biomass production data we calculated the number of individuals of B. simaruba needed to produce one tonne of dried wood. In a scenario where trees of B. simaruba have high biomass production, it would take about 35 trees of 10–20 cm of DAB (aged 4–10 years), while in a scenario with low production of biomass, about 70 trees of the same size categories would be needed to produce one tonne of dry wood. The information generated in this work has been disseminated and discussed with local producers through community workshops. To apply these results, there is a need to support the development of forest plantations of native species for the purposes of providing wood energy, which would require the involvement of local self-government and forest governmental authorities.

Box 6.2. Patterns of firewood use in a tropical dry forest landscape in central Veracruz

M.E. Ramos Vásquez and F. López-Barrera

Firewood plays a significant role in the energy requirements of many developing countries. However, patterns of fuelwood use are determined by a range of social and environmental factors. Moreover, the impacts of firewood use are not necessarily negative for all groups of factors and can lead to permanent environmental change. We explored the variation in patterns of firewood use within the municipality of Paso de Ovejas (see also Chapter 10). Firewood collection and acquisition were compared across two study areas that present different environmental and social conditions. In the upland area (320 m a.s.l.), the predominant land uses are rain-fed agriculture, grasslands and secondary forest, but also in this area the steep topography has allowed the preservation of the largest forest fragments within the municipality. The communities are mainly rural. On the other

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (218)

192

Principles and Practice of Forest Landscape Restoration

hand, the lowland area (20 m a.s.l.) is dominated by permanent and irrigated agriculture, grasslands and urban areas. The communities are mainly urban and a higher proportion of the inhabitants migrate to work in nearby cities such as Veracruz. Past surveys made by INEGI (Instituto Nacional de Estadistica, Geografía e Informática, 2000) showed that wood fuel dependence of the municipality is relatively low (38% of households). However, these surveys are biased as the presence of a stove in a house does not exclude the use of wood fuels by family members. Therefore, the present study evaluated in detail the patterns of firewood use through social surveys. By extrapolating our data, we estimated the sustainability of current extraction rates.

A total of 136 interviews were conducted in four study sites that were located in both the upland (Acazonica (33 interviews), Angostillo (23), Rancho Nuevo (8), and El Limon (7)) and lowland sections of the municipality (Carretas (12), El Mango (13), Loma Fina (15), and Cerro Guzman (27)). In addition, to explore the environmental perception of the rural communities, a community-wide workshop was divided into two sessions and held on the communal property of the ‘Ejido El Limón’. The first session explored how local habitants perceive dry forests in connection with environmental problems such as degradation, forest transformation to other land uses, and habitat fragmentation. In addition, information was obtained on the value of this type of forest to local inhabitants. The second session was designed to distinguish between the ecological structures of forests denominated ‘acahual’ (old-field), ‘monte’, and ‘mata’, as well as to determine the common tree species present in each type of forest and their potential utility for communities in the study area (including construction, live fences, shelving, firewood, medicine, fodder, and for use in the production of handcrafts). The two meetings lasted approximately two hours each and were initiated by projecting images of the local flora and fauna of dryland forests in an effort to set the stage for later discussions on conservation and sustainable use of this type of forest. Table 1 describes the pattern of firewood use in the eight localities surveyed. Three firewood units (‘cargas’) were weighted and their volume was obtained, resulting in an average of 83.7 kg and 0.117 m3, respectively. We calculated the volume (m3) of firewood per year consumed in the localities. We found that firewood consumption is significantly higher in the upland communities compared to the lowland communities (p <0.05, Table 1).

Using forest structure data (1 ha = 365 m3 of wood; Williams-Linera and Lorea, 2009), the estimated annual rates of firewood consumption in the municipality (20,787 m3/year) and the area of secondary and preserved forest recorded in 2007 (1857 and 8672 ha, respectively, see also Chapter 1), we estimated the period of time taken for the forest to disappear owing to this activity. The forest cover loss rate due to firewood extraction resulted in 69.3 ha/year, therefore the period of time that the preserved and secondary forest will last considering the calculated firewood extraction rate within the municipality is 28 and 125 years, respectively. It is important to note that this estimation does not consider forest regeneration (currently being examined; Manson et al., in prep.). It was also found that habitants of the upland area of Paso de Ovejas are willing to plant useful trees, however they do not have enough land available in their properties as these have an average size of 1 or 2 ha. These small pieces of land within ejidos are dedicated to agriculture for subsistence needs. We also documented that firewood extraction is related to forest conversion, as it forms part of a land management strategy for landowners who supply trees derived from land-use conversion. In the upland area conversion is mainly from secondary forest to agriculture and grasslands, whereas in the lowland area it is mainly from fruit tree plantations to sugarcane.

This study revealed that at current output levels of firewood consumption, sustainable production is possible, and outlined that the firewood extraction is not the main cause of recent forest loss in the municipality, which is mainly attributable to conversion to other land uses. In spite of this trend, past extraction rates of firewood are thought to be higher than current rates, therefore we have to be cautious in attributing the present forest area and condition in the municipality to the impact of firewood extraction.

Box 6.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (219)

193

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

Box 6.3 Taking local knowledge into consideration when selecting tree species for dry forest restoration in central Veracruz, MexicoA. Suárez and G. Williams-Linera

Tree species selected for tropical forest restoration are often restricted to very few well-known timber species that are seldom native species. Relatively little is known about tree species selection and appropriate techniques for restoring dryland forests in Latin America (Newton, 2008). The standard strategy is to plant mainly exotic timber species for which technical knowledge is readily available, without consideration of the numerous other uses of woody flora, many of which contribute substantially to rural livelihoods. The assessment of species selection for reforestation and restoration programmes is usually based on the experience of technicians working in government agencies as well as on plant availability in official nurseries. This choice overlooks local knowledge and the needs of local populations, and contributes to the failure of reforestation enterprises. Local knowledge turns out to be a useful instrument for increasing the acceptance and interest of local populations in woody species for restoration, as well as for formulating practical management recommendations. Tree species of interest to the local population are well known in the community and may provide a variety of goods and services.

The main objective of this study was to select tree species for tropical dry forest restoration beginning with a local knowledge-based inquiry. The methods included participatory workshops, interviews, and field walks in five rural communities in the municipalities of Paso de Ovejas and Comapa in central Veracruz, Mexico. The communities were located near forest fragments,

Table 1

Localities * No. of houses *Firewood use (%)

Firewood units/year

m3/year consumed

Tonnes/year

Uplands

Acazónica 294 84.0 12,034 1,408 1,007

Angostillo 184 77.7 5,285 619 443

Rancho Nuevo 72 100 3,456 405 289

Limón 76 89.5 3,607 422 302

Average 87.8 6,095 713 510

Lowlands

Loma Final 168 13.1 768 90 64

Cerro Guzmán 296 31.1 3,654 428 306

Carretas 89 15.7 267 32 22

El Mango 151 27.1 1,820 213 152

Average 21.8 1,627 190 136

Average of lowlands and uplands 54.8 3,861 452 323

Box 6.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (220)

194

Principles and Practice of Forest Landscape Restoration

where peasants use forest resources and have a better knowledge of trees. The workshops were conducted as an open group interview. Three questions were posed: Which native trees are useful? Which native trees are scarce? Which native trees are beneficial for wildlife? Cards of three different colours (one for each question) were handed out, and participants wrote down their name and age as well as the local names of all trees in the category. Open discussion was allowed. People who could not write indicated their answers to a research assistant.

A total of five workshops, two focus group meetings, 40 interviews, and 35 field walks were carried out from May 2007 to November 2008. In the workshops, 95 people ranging from 12 to 84 years participated. The 40 people who listed the most names were considered key informants. They were 36 men and four women, all of them over age 39. Key informants were questioned more systematically with semi-structured interviews on the use and management of woody plants and the wild animals associated with them; their help was requested for field walks. During these, we located the trees mentioned during the workshops and collected botanical specimens from them. Key informants were also asked to complement the preliminary lists with more information on the use, management, and ecological characteristics of the habitat in which the trees occur. Information from workshops and field walks was used to calculate three indices of relative importance for each species: the Cultural Importance Index (CII), Scarcity Perception Index (SPI), and Wildlife Importance Perception Index (WIPI).

Participants in the workshops named between 34 and 47 species in each event, for a total of 76 species for the categories useful (CII), scarce (SPI), and important for wildlife (WIPI). Also, informants classified trees as typical of mature forests (‘matas’), secondary forests (‘acahuales’), gallery forests, and agroforestry systems (around fields, living fences, and home gardens; these classifications were verified by direct observation). The species belonged to 29 botanical families. Fabaceae had the most species (18), followed by Bignoniaceae and Malvaceae with five each.

All 76 named species were classified as useful (CII), and the most important categories were rural construction, edible, fence posts, and firewood. The species with six uses, the highest number, were Chloroleucon mangense, Leucaena lanceolata, and Tabebuia chrysantha; Lysiloma acapulcense, an important fence post species, was in second place. The ten species with the highest CII value comprised 36% of this index; seven of these were legumes from secondary forests (acahuales). The primary use of most of these was as fuel, with the following species preferred: Acacia cochliacantha, Acacia pennatula, Diphysa carthagenensis, and Leucaena lanceolata.

Two-thirds of the species were considered scarce (SPI). In this category, five out of the 10 species with the highest CII were included, with Diphysa carthagenensis, Lysiloma acapulcense, and Chloroleucon mangense considered the most scarce. ‘Scarce’ does not necessarily mean that they are rare; some are relatively common but, owing to severe exploitation, they are insufficient to meet the demands of local populations. For example, farmers said that Diphysa carthagenensis has been over-used for firewood and fence posts. Even though the species re-sprouts from cut stems, there were hardly any trees with sufficient wood for exploitation. Lysiloma acapulcense was also over-used but had additional problems for natural and even artificial regeneration.

Two-thirds of the total species were considered important for wildlife (WIPI), particularly species of the Moraceae family. In this study, Ficus cotinifolia was perceived as the most important species for animal food and habitat. The 17 most important species with potential for TDF restoration activities were selected because they had the highest values within each index (usefulness, scarcity, and wildlife importance) and add up to one-third of the total index value (Box 6.5, Table

2). Those species can be recommended based on local knowledge, as they should be widely accepted by local people. They will plant them and encourage natural regeneration, since they are all managed in agroforestry systems and provide a variety of useful products and services. Six of them have particular conservation importance because they produce fruits for wildlife consumption. Thus, their use would mean strong support for conservation efforts.

Box 6.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (221)

195

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

If only the 17 species with the highest values of each index are recommended, about one-third of all species are covered. This is a convenient number, particularly as the local population is already familiar with successful – though rudimentary – propagation techniques, and will be concerned with the success of the planting. Only two of these (Cedrela and Tabebuia rosea) are currently available in the region’s nurseries. We strongly suggest promoting the other 15 species through systematic collection of seeds, nursery propagation, and use in forest restoration. Activities should include establishment of local plantations, agroforestry systems, and enrichment plantings of secondary vegetation that would be highly valued by landowners.

Box 6.3 (cont.)

Box 6.4 Willingness to reforest with native species in rural communities of central Chile

C. Smith-Ramírez, V. Maturana, J.J. Armesto

Native sclerophyllous forests in central Chile are now largely restricted to steep creeks and uplands. The loss of sclerophyllous forest cover and its degradation started more than 300 years ago, primarily owing to the combined impacts of firewood extraction, woodland burning, and livestock grazing (see Chapter 2). Today, a high proportion of rural habitants in central Chile are poor families that live on less than US $400 per month and subsist on highly degraded land with limited or no forest cover. We asked about the willingness of six rural communities to plant native trees, and discuss these results in the light of the new Chilean Forestry Law, which offers monetary incentives to private initiatives to reclaim non-forested land.

We found that 53.9% of the respondents (N = 217 interviews) were interested in planting native trees if the state pays for the full cost. If the landowner pays for the costs, the interest was 47.0%. The willingness to reforest with native species was related mainly to the use of native trees as sources of honey, firewood or because of their ornamental value, and the geographic proximity of the communities to the forests. Gender and economic status of the respondents, presence of native trees on their land, and the identity of the rural community were not significant variables in the analysis. But, only 23% of the 47.0% interested in planting native trees, were prepared to do so on their own land. The majority were willing to plant on nearby hills, riparian habitats and public land, where the reforestation incentives of the new Chilean Forestry Law are not possible to apply. We concluded that even if monetary incentives to plant native trees as part of restoration plans could be offered extensively to rural communities, it is still necessary first to educate rural inhabitants about the values of native tree plantations.

Box 6.5 Traditional knowledge in the drylands of central Mexico: an endangered resource?

R. Aguilar-Santelises and R. F. del Castillo

The vast traditional knowledge of indigenous people in Mexico has been considered endangered (Caballero and Cortés, 2001; Carlson and Maffi, 2003). An important component of this knowledge, accumulated during millennia, is the result of a close contact between the people and their native species. However, starting with the Spanish conquest, this contact began to dwindle, first at a slow pace and recently at a very fast pace (Arredondo et al., 1981; Pérez, 2006). Several factors are blamed for this problem, but all of them are related to two phenomena: one social and one biological. The first one is related to acculturation, a gradual process of change, in which traditional culture is supplanted by modern culture (Zent, 2001; Aguilar-Santelises, 2007). The continuous deterioration and eventual local extirpation of many native species and ecosystems comprise the biological problem (Cotler et al., 2007).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (222)

196

Principles and Practice of Forest Landscape Restoration

One of the best examples in Mexico, in which these two phenomena can be detected, is in the Upper Mixtec region, in the southern Mexican state of Oaxaca. This is a dryland territory with a great diversity of native species, many of which are endemic (García-Mendoza et al., 1994), but which also displays one of the highest rates of soil erosion in the world (ca. 200- 280 tonne y–1; Anonymous, 2007). The Mixtecs settled in this region ca. 1400 B.C. and initiated a rich culture with their own language. As an effort to rescue part of the traditional knowledge of native plants in this region and to increase our understanding of the process of acculturation, we recorded the uses of native woody species and compared the distribution of the traditional among the local people in three municipalities of the Upper Mixtec region. Of particular interest is that these municipalities display contrasting proportions of Mixtec-speaking people, formal education levels, and public medical services (Anonymous, 2005). This suggests different degrees of contact with the preponderant culture, allowing a test of the hypothesis that acculturation is a primary factor contributing to local extirpation of traditional knowledge.

We collected information about native plant uses by showing each informant a set of herbarium specimens of 112 species of native species randomly chosen from the locality, and asked for the local name of the plant and uses he or she gave to each of the shown species (Fig. 1). We evaluated the importance of each species by the proportion of informants that reported such species to be useful, and the distribution of knowledge according to the native language proficiency and education level of the informants. The recorded uses were classified into eleven categories. All of the species were reported useful by at least a few of the informants. However, the importance of the species to the community was unevenly distributed: a few species were described as highly useful, whereas most of the species were found to be useful only to a minority (Fig. 2). Juniperus flaccida was the most useful species, as more than 90% of the informants reported this species as useful. Leucaena, a legume tree, and six oak (Quercus spp.) species followed as the next most useful species (Table 1). Native plants were mostly used as livestock food and medicine, fuel (firewood and charcoal), and human food (Fig. 3).

Regarding the traditional knowledge among local people, we found that it was strongly rooted in the study area. However, few people were aware of the usefulness of a high proportion of the species shown, whereas most people found useful only a small fraction of those species (Fig. 4). Interestingly, the municipality with the lowest evidence of contact with the preponderant culture showed the highest level of traditional knowledge (Fig. 5).

This study has shown that traditional knowledge about plant uses in the Upper Mixtec region is very rich. However, it is unevenly distributed both among people and municipalities. Preserving this valuable knowledge requires supporting the dissemination of this knowledge among the population. This could be achieved by the implementation of workshops led by the most knowledgeable people, and the eventual incorporation of this knowledge into formal education.

Figure 1

Box 6.5 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (223)

197

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

Figure 2

Figure 3

Figure 4

Box 6.5 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (224)

198

Principles and Practice of Forest Landscape Restoration

Figure 5

Table 1

Species Utility index

Juniperus flaccida 0.92

Leucaena diversifolia 0.86

Quercus acutifolia 0.84

Quercus laurina 0.79

Quercus deserticola 0.74

Quercus castanea 0.74

Quercus liebmannii 0.73

Acacia pennatula 0.73

Quercus rugosa 0.72

Acacia farnesiana 0.69

Arctostaphylos pungens 0.65

Tecoma stans 0.63

Litsea glaucescens 0.61

Box 6.5 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (225)

199

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

Seasonal dry premontane forest in northwestern Argentina. Photo: L. Malizia

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (226)

200

Principles and Practice of Forest Landscape Restoration

Box 6.6 Assessing the value and commercial potential of non-timber forest products: the CEPFOR project

A.C. Newton

As noted elsewhere in this chapter, many rural communities are strongly dependent on a range of non-timber forest products (NTFPs) to support their livelihoods. While many of these are used for subsistence purposes, in recent years there has been growing interest in the commercialization of NTFP, as an appropriate means of developing forest resources. This reflects a growing recognition of the contribution made by many NTFPs to rural livelihoods, both in terms of supporting subsistence and as a means of generating financial income. At the same time, as harvesting of NTFPs is generally considered to be less damaging to forest resources than timber extraction, the production of NTFPs is widely believed to be relatively compatible with forest conservation (Arnold and Ruiz Pérez, 1998). Thus commercialization of NTFPs potentially offers a means of achieving both conservation and development goals concurrently, by increasing the value of forest resources to local communities.

Recent reviews suggest that approaches to NTFP commercialization have not, however, been universally successful, and that scope for improving rural livelihoods through NTFPs is in doubt (Sheil and Wunder, 2002). In a comprehensive review of NTFP commercialization, Neumann and Hirsch (2000) indicate that sale of NTFPs often tends to provide a low level of income for the poorest sections of communities, rather than providing a method of socioeconomic advancement. The NTFP trade may actually perpetuate poverty rather than alleviate it (Neumann and Hirsch, 2000). Given that NTFPs are highly diverse in terms of their ecological and socioeconomic characteristics, there is a need to define which NTFPs have particular potential for development, and under what conditions their use is likely to make a positive contribution to both human livelihoods and forest conservation. Such information would help reduce the misdirection of donor investments identified by Sheil and Wunder (2002).

This issue was recently addressed by an interdisciplinary research project (CEPFOR), which examined the factors influencing NTFP commercialization in 19 case studies from Mexico and Bolivia, through an intensive programme of participatory research conducted with local communities and other stakeholders. A number of these case studies were located in tropical dry forest areas. Results of the research are presented in detail by Marshall et al. (2003, 2006), and outputs of the project are freely available from this website: http://quin.unep-wcmc.org/forest/NTFP/. This includes guidance on appropriate methods for assessing the value of forest resources.

Key findings of the research included the following:

NTFPs are important in the lives of the rural poor, and incomes vary greatly even between households engaged in the same activity. NTFP activities were found to contribute between 7% and 95% of a household’s annual cash income; regularly provide a safety net for the poor to fall back on when other activities – such as subsistence agriculture or cash crops like cof-fee – fail to deliver as expected; and sometimes provide a stepping stone to a non-poor life, and never lead to an increase in poverty.

The importance of NTFPs in household livelihood strategies is closely linked to their season-ality and the way they may be combined with other income-generating activities.

The more months a product can be traded, the more favourably households view the activity. Conversely, households involved in seasonal products are more likely to switch from NTFP activities to other livelihood options, reflecting their desire for a more consistent and year-round source of income.

In the case of communally-owned resources, improved management of the natural resource and better harvesting practices are common. If land is held privately and the plant can be easily propagated, individuals begin to engage in small-scale domestication. Many of the communities expressed a strong interest in cultivating plant species of high value for NTFP production, highlighting the economic potential and community support for forest restora-tion efforts.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (227)

201

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

Acacia pennatula in a pasture that that has been pollarded for firewood and secondarily for forage, Chiapas, Mexico. Photo: B. Ferguson

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (228)

202

Principles and Practice of Forest Landscape Restoration

References

Aguilar-Santelises, R. 2007. Etnobotánica cuantitativa en una región de Bosque de Niebla de Sierra Norte, Oaxaca. Tesis de Maestría. Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional Unidad Oaxaca. Instituto Politécnico Nacional.

Anonymous. 2007. Desarrollo de una microcuenca en la comunidad de Tepelmeme, Villa de Morelos, Oaxaca. Comisión Nacional de Zonas Áridas (CONAZA).

Anonymous. 2005. II Conteo de Población y Vivienda 2005. Instituto Nacional de Estadística, Geografía e Informática. México. <http://www.inegi.org.mx/inegi/default.aspx>

Arredondo, C., Liedo, J., Zúñiga, R., Campos, S., Solórzano, G. 1981. Marco de referencia para la planeación y evaluación de la investigación agrícola en la Mixteca Oaxaqueña. Campo Agrícola Experimental de la Mixteca Oaxaqueña. 123pp.

Arnold, J.E.M., Ruiz Pérez, M. 1998. The role of non-timber forest products in conservation and development. In: Wollenberg, E. and Ingles, A. (eds.), Incomes from the forest: Methods for the development and conservation of forest products for local communities. Centre for International Forestry Research, Bogor, Indonesia: pp. 17–42.

Caballero, J., Cortés, L. 2001. Percepción, uso y manejo tradicional de los recursos vegetales en México. In: Rendón A., B., S. Rebollar D., J. Caballero N. y M.A. Martínez A. (eds.), Plantas, Cultura y Sociedad. Estudio sobre la relación entre seres humanos y plantas en los albores del siglo XXI. Universidad Autónoma Metropolitana-Secretaría del Medio Ambiente, Recur-sos Naturales y Pesca (SEMARNAP). México: pp. 79–100.

Carlson, J.S., Maffi, L. 2003. Ethnobotany and conservation of biocultural diversity. Advances in Economic Botany 15: 6–35.

Cotler, H., Sotelo, E., Domínguez, J., Zorrilla, M., Cortina, S., Quiñones L. 2007. La con-servación de suelos: un asunto de interés público. INE-SEMARNAT, México. Gaceta Ecológica 83: 71.

Echeverría, C., Schiappacasse, I., Urrutia, R., Cárcamo, M., Becerra, P., Smith, C. Holmgren M. 2010. Manual de restauración de ecosistemas degradados para la conservación de la biodi-versidad y el desarrollo rural en la zona semiárida de Chile central. Proyectos REFORLAN, CONYCIT Valdivia, Chile.

García-Mendoza, A., Tenorio, P., Reyes, J. 1994. El endemismo en la flora fanerogámica de la Mixteca Alta, Oaxaca-Puebla, México. Acta Botánica Mexicana 27: 53–74.

Hernández, F. 1659. Nova plantarum animalium et mineralium Mexicanorum historia. Mas-cardi, Roma.

Hilgert, N.I. 2001. Plants used in home medicine in the Zenta River basin, Northwest Argen-tina. Journal of Ethno-Pharmacology 76: 11–34.

INEGI (Instituto Nacional de Geografía, Estadística e Informática) 2000. Censo de Población y Vivienda.

Inter-Commission Task Force on Indigenous Peoples. 1997. Indigenous peoples and sustain-ability: cases and actions. IUCN and International Books, Utrecht.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (229)

203

Socioeconomic valuation of dryland forest resources in dry areas of Argentina, Chile and Mexico

Manson, R.H., López-Barrera F. Landgrave, R. in prep. Patterns and drivers of tropical decidu-ous dry forest transformation in central Veracruz, Mexico.

Martínez, M. 1936. Plantas útiles de México. Ediciones Botas, Mexico City.

Marshall, E., Newton, A.C., Schreckenberg, K. (eds.). 2003. Commercialization of non-timber forest products: first steps towards analysis of the factors influencing success. Interna-tional Forestry Review 5(2): 128–137.

Marshall, E., Schreckenberg, K., Newton, A.C. (eds.). 2006. Commercialization of non-timber forest products: factors influencing success. Lessons learned from Mexico and Bolivia and policy implications for decision-makers. UNEP World Conservation Monitoring Centre, Cambridge, UK.

Montero Solano, J.A., Manson, R.H., López Barrera, F., Ortiz, J., Callejas, J. in prep. Public policy and land use change in central Veracruz: an important factor in efforts to restore a tropical dry forest landscape.

Neumann, R.P., Hirsch, E. 2000. Commercialisation of non-timber forest products: review and analysis of research. Center for International Forestry Research, Bogor, Indonesia. 176 pp.

Newton, A. 2008. Restoration of dryland Forests in Latin America: The ReForLan project. Ecological Restoration 26: 10–13.

Newton, A.C., Cayuela, L., Echeverría, C., Armesto, J.J., Del Castillo, R.F., Golicher, D., Genel-etti, D., Gonzalez-Espinosa, M., Huth, A., López-Barrera, F., Malizia, L., Manson, R., Premoli, A., Ramírez-Marcial, N., Rey Benayas, J.M., Rüger, N., Smith-Ramírez, C., Williams-Linera, G. 2009. Toward integrated analysis of human impacts on forest biodiversity: lessons from Latin America. Ecology and Society 14(2): 2. <http://www.ecologyandsociety.org/vol14/iss2/art2/>.

Pérez, J. 2006. Proyecto de conservación de suelos y aguas, y reconversión productiva en la microcuenca “El Arenal”, Región Mixteca Oaxaqueña. Departamento de Fitotecnia. Univer-sidad Autónoma Chapingo. México D.F. 9 pp.

Schiappacasse, I., Nahuelhual, L., Vásquez, F., Echeverría. C. 2009. Valuing the benefits of dry-land forest restoration in central Chile. XIII World Forestry Congress. Buenos Aires, Argen-tina.

Sheil, D., Wunder, S. 2002. The value of tropical forest to local communities: complications, caveats, and cautions. Conservation Ecology 6(2), 9. [online] URL: http://www.consecol.org/vol6/iss2/art9.

Suárez, A., Williams-Linera, G., Trejo, C., Valdez-Hernández, J.I., Cetina-Alcalá, V.M., Vibrans, H. In review. Local knowledge helps select species for forest restoration in a tropical dry for-est of central Veracruz, Mexico. Agroforestry Systems.

United Nations Development Programme (UNDP). 2004. examples of the successful con-servation and sustainable use of dryland biodiversity. Sharing Innovative Experiences, Vol. 9. UNDP Special Unit for South-South Cooperation, GEF, UNEP, TWNSO, TWAS, New York.

Williams-Linera, G., Lorea, F. 2009. Tree species diversity driven by environmental and anthro-pogenic factors in tropical dry forest fragments of central Veracruz, Mexico. Biodiversity and Conservation 18: 3269–3293.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (230)

204

Principles and Practice of Forest Landscape Restoration

Willson, C.J., Manos, P.S., Jackson R.B. 2008. Hydraulic traits are influenced by phylogenetic history in the drought-resistant, invasive genus Juniperus (Cupressaceae). American Jour-nal of Botany 95: 299–314.

Zent, S. 2001. Acculturation and ethnobotanical knowledge loss among the Piaroa of Ven-ezuela: Demonstration of a quantitative method for the empirical study of TEK change. In: Maffi, L. (ed.), On biocultural diversity: linking language, knowledge, and the environment. Washington, D.C., Smithsonian Institution Press: pp. 190–211.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (231)

205

7 IMPACT OF FOREST FRAGMENTATION

AND DEGRADATION ON PATTERNS

OF GENETIC VARIATION AND ITS

IMPLICATION FOR FOREST RESTORATION

A.C. Premoli, C.P. Souto, S. Trujillo A., R.F. del Castillo, P. Quiroga, T. Kitzberger, Z. Gomez Ocampo, M. Arbetman, L. Malizia, A. Grau, R. Rivera, R. Rivera García, A.C. Newton

Introduction

Dry forests are currently the focus of increasing conservation and restoration efforts. This is because one billion people live in dry regions of the world that cover nearly 40% of the Earth’s surface. These regions all have in common a reliance on natural resources – including biodiversity, which is declining at a rate unprecedented in recorded history (UNDP, 2004). The objectives of the ReForLan project were to identify and promote approaches for restora-tion of arid and semi-arid forest ecosystems. The focus of this chapter is to assess the impact of forest loss, fragmentation, and degradation on genetic variability within socioeconomical-ly important tree species of conservation concern, in the context of restoring functional for-est landscapes. In addition, the chapter provides recommendations for restoration of dryland forest resources based on an understanding of the processes influencing genetic variation.

Patterns of genetic diversity in plants are the result of current and past evolutionary proc-esses that can be used to guide conservation efforts. Molecular markers can be of great value for investigating the effects of neutral processes such as genetic drift affecting small popula-tions and isolation owing to barriers for gene flow. Such events tend to erode genetic varia-tion in natural populations. Markers may evolve at distinct evolutionary rates and therefore can provide information about processes acting at different temporal scales. Mutations per generation of uniparentally inherited DNA markers such as those of the chloroplast occur at rates of about 10–9, whereas for nuclear microsatellites, mutation rates are between four and six orders of magnitude greater (Provan et al., 1999). While sequences of chloroplast DNA can be used to reconstruct historical genetic patterns, nuclear markers may elucidate the contemporary genetic structure of natural populations. Hence, the combination of the two markers gives the opportunity to understand past and present genetic patterns so as to guide conservation and restoration efforts for the long-term preservation of species.

Our aim was to analyze patterns of within- and between-population genetic variation in species of conservation concern, economic importance, and/or socioeconomic relevance in three areas of Latin America. Different markers were used for genetic analyses of natural populations. These included traditional isozyme methods, and novel molecular analyses such as Single Nucleotide Polymorphism (SNP), nuclear species-specific Simple Sequence Repeats (SSRs) known as microsatellites, and DNA sequences of non-recombinant regions of the chloroplast. Whereas the latter provides a historical signal, the former three mostly reflect contemporary genetic structure.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (232)

206

Principles and Practice of Forest Landscape Restoration

Mapache (Procyon lotor) found in one of the study sites in central Veracuz, Mexico. Photo: C. Alvarez

Open Austrocedrus chilensis stands at the forest-steppe ecotone in southern Argentina. Photo : T. Kitzberger

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (233)

207

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Genetic patterns were analyzed in three geographic areas: northern and southern Argen-tina and Oaxaca, Mexico (see also Boxes 7.1 and 7.2). In these study areas the following spe-cies were studied: three Cedrela species, C. balansae, C. lilloi, and C. saltensis, from season-ally dry subtropical forests of northern Argentina (Yungas); the dominant tree Austrocedrus chilensis (hereafter Austrocedrus) of the forest-steppe ecotone of the Patagonian Andes of southern Chile and Argentina; and Malacomeles denticulata and Catopsis berteroniana from arid and disturbed environments in the central state of Oaxaca, Mexico.

Box 7.1 Genetic hotspots: the quest for preservation of Chile’s evolutionary history

M. Henríquez, C. Echeverria, A. Premoli, G. Machuca, C. Souto and P. Quiroga

A biological hotspot is defined as a bio-geographic region with a significant amount of biodiversity that is under threat of destruction owing to human action. The importance of knowing and preserving these hotspots lies in the fact that they continue the evolutionary history of the world and play a role in the biological equilibrium of the biosphere. In order to achieve the goal of conserving a biological hotspot, it is necessary to prioritize threatened areas so that we can identify which are the most valuable spots to preserve. However, since the biodiversity hotspot concept was originated, genetic information has not been considered in their designation.

Owing to its geography, Chile has been called a bio-geographic island. Because of that, Chile provides a unique centre of global biodiversity: the Chilean Winter Rainfall Valdivian forests (Dinerstein et al., 1995). During the last 40 years, this biodiversity hotspot has been affected by deforestation and land-use change. To determine which areas are important for biodiversity conservation, we used three large databases on the presence of threatened species within the bio-geographic region: the Catastro y evaluación de recursos vegetacionales nativos de Chile (CONAF et al., 1999), a Darwin Initiative Project (Hechenleitner et al., 2005), and data collected by the ECOTONO laboratory of the Universidad del Comahue in Argentina. More specifically, we used data from 26 threatened species to study their distribution patterns, and of these we used genetic information collected from 12 species.

To integrate the data we used ArcGIS software (provided by ESRI). First, we constructed a 5 x 5 km raster map that showed cells where the species are found. The treatment of the cells included an overlay of the locations of the species, in order to construct a richness map showing the areas with more than one species. Then, we generated genetic raster maps consisting of the presence or absence of unique alleles and a heterozygosity scale. These genetic data were obtained through both isoenzyme and DNA techniques, based on the presence of molecular and biological markers. This information, when combined with the species richness map, allowed us to determine genetic hotspots. These hotspots were defined as those areas with presence of at least one species with unique alleles, at least one species with high heterozygosity and species richness at least equal to 3.

Our results revealed the presence of high richness areas along the coastal range in Chile and the Andean range both in Chile and Argentina (Fig. 1). Most richness areas were located in the coastal range in the Maule and Bio-bio Regions. The species that occur in these areas are highly threatened as they are located in one of the areas with the highest rates of deforestation and land-use change in Chile (Echeverria et al., 2006). Populations with unique alleles were distributed in four diagonal lines from the coastal range in Chile to the Andean range on the Argentinean side (Fig. 1). Four areas share populations of two species with unique alleles: one near Curacautin (Araucanía Region, Chile), one in Valdivia (Los Ríos Region, Chile), one to the north of Bariloche (Neuquén Province, Argentina) and one near Esquel (Chubut Province, Argentina). Populations with high heterozygosity are mainly concentrated in Argentina, from Neuquén to the Chubut Provinces (Fig. 1); only a few populations are found across the Chilean side both in the coast and in the Andes.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (234)

208

Principles and Practice of Forest Landscape Restoration

Five genetic hotspots were identified only in Chile in the zones of Cauquenes, Bullileo, Curacautín, Valdivia and Puyehue (Fig. 1). Of these hotspots, four are not protected by the Chilean state; only Puyehue is completely located within national parks. It is important to note that of the five hotspots identified, three of them (Curacautin, Valdivia and Puyehue) possess populations of two species with high heterozygosity (Fig. 1).

Figure 1a, b

The genetic hotspots identified in the present study are candidate areas for protection because of the valuable genetic information and high biodiversity value contained there. These genetic hotspots can be considered as biogeographic regions with a significant number of genetic resources threatened by human actions and under risk. The most important issue about these genetic hotspots is that they are the basis for species adaptive potential, which will determine their ability to survive environmental change such as climate change. Because of the heterogeneous nature of the study landscapes, the management of the area needs to be undertaken at different scales. The next step has to be the development of multi-scale management policies to ensure the preservation of those habitats that hold species with different phylogenies, and to protect the populations not strongly affected by human beings, because they retain unaltered phylogeographic characteristics. This kind of management would allow us to understand how ecological processes change according to scale.

Box 7.1 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (235)

209

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Seasonally dry subtropical forests of northern Argentina

Cedrela species are among the most valuable hardwoods of Argentina. Consequently, they have been heavily logged within the study area. Since recruitment of these species is distur-bance-dependent and they display high growth rates, they are ideal for forest restoration and stand enrichment purposes (see also Chapter 6).

In subtropical seasonally dry forests of Argentina, the distribution of dominant woody Cedrela species follows elevation gradients. C. balansae occurs in piedmont forest between 300 and 800 m in elevation attaining its southern limit at 24º30’S. C. lilloi is found in montane forest between 500 and 1350 m reaching the southern-most distributional range of the genus at 28º15’ S latitude. These two forest types, piedmont and montane, are the most economically important (Brown and Pacheco, 2006). More recently a third species, C. saltensis, has been identified. Apparently, the latter has a restricted range (<1000 km2) (Malizia et al., 2006) and occurs in an area of sympatry between the other two species at elevations of between 700 and 1100 m a.s.l. (Grau et al., 2006), reaching its southern limit at 24º 40’ S latitude.

We used molecular marker techniques to test the hypothesis that C. saltensis is a species of hybrid origin between C. balansae and C. lilloi. The main purpose was to use information on patterns of genetic differentiation between the two parent species and putative hybrids to define appropriate sources of germplasm to guide future restoration and conservation actions.

Box 7.2 Testing forest connectivity using genetic distance: spider monkeys and dry forest restoration, Nicaragua

S. Hagell, S. Otterstrom, C. Chambers

Wildlife across Central America is under immediate threat from forest loss and fragmentation. These threats are particularly evident in the critically endangered Central American Dry Forest (WWF), in which it is estimated that less than 5% of native forest remains intact or protected (Miles et al., 2006). We are conducting research to understand how wildlife, particularly endangered or threatened species, negotiate these human-dominated landscapes in south-western Nicaragua. As a part of ongoing forest restoration initiatives, Paso Pacifico is working with wildlife biologists at Northern Arizona University to study landscape genetic patterns in the Central American spider monkey (Ateles geoffroyi). The spider monkey is a highly charismatic arboreal mammal that is also a key seed disperser for native trees (Pancheco and Simonetti, 2000). Our goal is to preserve the spider monkey population in this landscape by restoring dispersal corridors between forest fragments. However, because these animals are rare and elusive, there is limited behavioural information as to the dispersing animals’ area in the current landscape. For this project, we are using non-invasively collected genetic samples from faecal material to test our hypotheses as to how spider monkeys may be able to use secondary and non-forest matrices for dispersal. More specifically, we are comparing mean genetic distances between social groups to the spatial distance calculated from alternative landscape resistance models. These models are based on the cost of travel through mature forest, regenerating or secondary forest, and non-forest matrices. This is a relatively new approach to conduct spatial genetic analyses, because genetic data is used as a means to measure dispersal cost and test multiple hypotheses of landscape connectivity (as in Cushman et al., 2006; Epps et al., 2007). Furthermore, our analyses provide estimates of resistance for each landscape feature in the model, data that can be used directly to build a predictive model for the whole landscape. As an example, the ‘best’ models of landscape resistance can be combined with new circuit theoretic tools to identify ‘pinch points’ that are critical for connectivity (McRae et al., 2008). In this way, the patterns of genetic diversity in this landscape will reveal pathways and barriers to dispersal, and indicate how forest management can be used to preserve this species.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (236)

210

Principles and Practice of Forest Landscape Restoration

Interspecific hybridization is a common phenomenon in plants. The frequency of hybridiza-tion in trees is particularly related to their longevity and various reproductive systems that al-low interespecific gene flow. However, hybrid formation will depend upon the genetic compat-ibility between species, the flowering phenology, and their degree of range overlap. In addition, favourable micro-site conditions, known as ‘hybrid habitats’ (Anderson, 1948) may facilitate the establishment and survival of hybrid progeny. These are usually related to disturbed sites and/or locations where pure plants have relatively reduced competitive potential and fitness.

To study the possible hybrid origin of Cedrela saltensis from C. balansae and C. lilloi we sampled pure species and potential hybrids inhabiting seasonally dry forests of northern Argentina. A genetic study of three species of Cedrela was conducted on populations located between 23º 5’ and 24º 30’S latitude near to Calilegua National Park. We collected fresh leaf tissue from natural populations of each species (Table 7.1) which were analyzed using 12 isozyme loci. These were Glycerate 2 de-hydrogenase (G2d), Isocitrate dehydrogenase (Idh1, Idh2), Malic enzyme (Me1, Me2), Phosphoglu-coisomerase (Pgi1, Pgi2, Pgi3), Peroxidase (Per1, Per2, Per3), and Shikimate dehydrogenase (Skdh). We calculated standard population genetic parameters. These included estimates of diversity and inbreeding at the population and species level, respectively. We estimated genetic distance metrics and ran multivariate cluster analyses to portray genetic relationships among species. Heterogeneity in allelic frequencies across populations and species was analyzed using chi-square tests.

Results showed differences among the three analyzed Cedrela species in allelic frequencies and diagnostic alleles, i.e. those present exclusively in just one species, were found in all studied species. Cedrela balansae had seven, C. lilloi had one, and C. saltensis had six diagnostic alle-les, respectively (Table 7.2). All three species were genetically diverse although C. saltensis had greater heterozygosity (He = 0.330) and mean effective number of alleles (Ne = 1.63) than puta-tive parent species C. balansae (He = 0.229 and Ne = 1.39) and C. lilloi (He = 0.276 and Ne = 1.48). In addition, estimates of within-population inbreeding yielded Fis values of 0.135 (CI 0.016 – 0.211) for C. balansae, 0.308 (CI 0.050 – 0.640) for C. lilloi, and 0.275 (CI –0.061 – 0.642) for C. saltensis. The three species were significantly different in their allozymic profiles for 10 out of 11 tests. Genetic distance indices and multivariate cluster analysis by means of Nei’s genetic distance (1978) showed that C. saltensis is more similar to C. lilloi than to C. balansae (Fig. 7.1).

Table 7.1 Cedrela

Species Population N Sº Latitude Wº LongitudeElevation m a.s.l

C. balansae R34CB 27 24 17 46.1 64 54 33.7 730

C. balansae PNCCB 3 23 46 54.7 64 48 54.6 535

C. balansae SSCB 22 23 40 28 64 33 49.1 380

C. balansae LNCB 1 23 7 1.9 64 40 34.7 800

C. balansae CICCB 32 23 7 39.4 64 27 56 495

C. lilloi SSJCL 15 24 9 48.4 65 18 50.6 1390

C. lilloi PSCL 4 24 30 7.7 65 18 43.5 1400

C. lilloi FTCL 5 23 0.5 11.1 64 51 30.9 1710

C. saltensis PNCCS 15 23 41 31.9 64 52 44 1450

C. saltensis CRCS 3 23 5 38.4 64 44 30.4 1000

C. saltensis ACCS 5 23 5 14 64 47 28.6 1200

C. saltensis LNCS 5 23 6 59.9 64 40 52.1 830

N: number of sampled individuals

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (237)

211

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Soil erosion in dry forests in Oaxaca, Mexico. Photo: R.F. del Castillo

Ripening fruit of Cochlospermum vitifolium in Central Veracruz, Mexico. Photo: C. Alvarez

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (238)

212

Principles and Practice of Forest Landscape Restoration

Table 7.2 Cedrelabold

Locus C. balansae C. lilloi C. saltensis

G2d 1 0.3062 0.6943 0.0754 0.9255 1.000Idh11* 0.021 0.4462* 0.031 0.4643 0.0894 0.3465 0.6156 0.0387 0.0318 0.8769 0.041Idh21 0.969 0.962 1.0002 0.031 0.038Me11 0.0712 0.901 0.875 0.9253 0.099 0.054 0.075Me21 0.862 0.760 0.9292 0.138 0.240 0.071Pgi11* 0.138 0.690 0.7682* 0.862 0.310 0.232Pgi21 0.0212 0.155 0.086 0.0183 0.454 0.603 0.4464 0.325 0.224 0.0185 0.041 0.086 0.1796 0.005 0.0367 0.304Pgi31 0.042 0.0532* 0.958 0.8163* 0.887 0.1324 0.112Per11 0.051 0.1722* 0.020 0.724 0.7413* 0.929 0.103 0.259Per21 1.000 1.000 1.000Per31 0.116 0.065 0.2612 0.820 0.870 0.7393 0.064 0.065Skdh1 0.022 0.083 0.2042 0.789 0.854 0.5933 0.172 0.042 0.2044 0.017 0.021

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (239)

213

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Figure 7.1 Cedrela.

Genetic diversity was high in the three Cedrela species. In contrast, elevated inbreeding coef-ficients indicated by Fis in all species probably reflect biparental inbreeding as a result of local-ized pollen movement by insects. As expected, the putative hybrid C. saltensis presented high heterozygosity in comparison with C. balansae and C. lilloi. Nevertheless, this result per se would not be evidence of hybridization. C. saltensis was relatively more similar to C. lilloi, as shown by a smaller genetic distance between them (Fig. 7.1). In contrast, C. saltensis is more similar morpho-logically to C. balansae than to C. lilloi (Zapater et al., 2004). On the other hand, the presence of diagnostic alleles in C. saltensis suggests that this species is taxonomically distinct. A previous molecular phylogenetic analysis suggests that the possible ancestral taxa of C. saltensis would have had a separate evolutionary history (Muellner et al., 2009). While more studies are needed to understand the phylogenetic relationships among these taxa, our results strongly suggest that Cedrela saltensis is not the result of recent and/or recurrent hybridization between C. balansae and C. lilloi. Therefore Cedrela saltensis needs to be treated as a separate taxonomic entity.

Information on tree species diversity and structure of seasonally dry subtropical Yungas is limited (Malizia et al., 2006; Brown and Pacheco, 2006). Data on distribution patterns of genetic diversity on species of economic and/or conservation concern is also scarce (Quiroga and Premoli, 2007). This information is needed given that selective logging of Cedrela species is eroding genetic diversity because the better genotypes with optimal characteristics are con-tinuously being removed. Such processes have been documented in closely related species in Central America (Gillies et al., 1997; 1999). However, natural populations of C. lilloi are protect-ed in many different geographic areas, e.g. within National Parks. In contrast, piedmont mature C. balansae forests no longer exist in the wild and only young stands occur in marginally pro-tected areas such as nearby National and Provincial Parks. On the other hand, the distribution range of C. saltensis has been delimited only preliminarily (Malizia et al., 2006). According to ecological niche models developed for C. lilloi and C. balansae, the potential area of distribu-

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (240)

214

Principles and Practice of Forest Landscape Restoration

tion of C. saltensis in northwestern Argentina is about 600 km2 (Malizia et al. 2006). C. saltensis stands might occur from southern Bolivia (Tariquía Reserve), across Salta (Baritú National Park) and Jujuy (Calilegua National Park) provinces. Further studies on the distribution range and genetic patterns of this species are needed to fully evaluate its conservation status.

Arid central Oaxaca, Mexico

Malacomeles denticulata (formerly Amelanchier denticulata), is a shrub member of the Rosaceae family. Along its distribution from central México to Guatemala, M. denticulata grows mainly in disturbed shrublands and pine-oak forests (Rzedowski and Calderon, 2005). As a tolerant species of such arid and degraded environments, it is considered ecologically important for recovering degraded areas.

One of the most recently developed molecular tools to study genetic variation are single nu-cleotide polymorphisms (SNPs). These are the result of either transition or transversion mutation events. SNPs are single base pair positions in the genome of two or more individuals, at which different alternative sequences or alleles exist in populations (Weising et al., 2005). Once SNPs have been determined through sequencing, they can be identified relatively rapidly in any indi-vidual of the studied population using real-time polymerase chain reaction (RT-PCR) equipment. RT-PCR monitors in real time the progress of the PCR as it occurs. Data are collected throughout the PCR process rather than at the end of the PCR. hom*ogeneous detection of PCR products can be obtained by using double stranded DNA binding dyes or fluorogenic probes. Quantification of DNA or RNA can be more precise and reproducible because it relies on threshold cycle (CT) values determined during the exponential amplification phase rather than at the endpoint.

The main objective of this research was to develop and apply a novel DNA marker to access the impact of forest lost and degradation in the genetic variability of Malacomeles denticu-lata in Oaxaca, Mexico where degradation of dry forest is extreme. Leaves were collected for genetic analyses from eight localities of different kinds of habitats, namely tropical dry forest, pine forest, oak pine forest, and chaparral. Parameters of within-population genetic diversity and among-population divergence (Fst) were calculated (see also Box 7.3).

Box 7.3 Genetic variability in populations of Amelanchier denticulataJ. Ramírez Luis, R.F. del Castillo, E. Cruz Cruz

Amelanchier denticulata is a shrub in the Rosaceae family, which can grow in a wide variety of habitats, including severely degraded and eroded areas. It is a valuable plant as fodder for goats and sheep, and displays a large phenotypic variation among populations. We studied the nature of such variation by means of a common garden experiment including four populations, using a half-sib design in order to divide the total phenotypic variance into its components for three characters: leaf area, leaf indentation, and relative growth rate. Leaf area was the most variable character and had the highest genetic variance, followed by leaf indentation. Relative growth rate displayed little or no detectable genetic variance in the studied populations. Narrow sense heritabilities were also calculated. We found two populations with little or undetectable genetic variation and two genetically variable populations, in particular the San Pablo Huitzo population, which could be a good candidate for breeding and conservation programmes. Finally, we detected evidence of genetic differentiation among the four studied populations, in particular for leaf area and leaf indentation but not for relative growth rate. Different habitat selection regimes may explain the genetic differentiation encountered among populations for leaf area and indentation, whereas natural selection may be more intense in all the studied habitats for relative growth rate.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (241)

215

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Stands of Austrocedrus chilensis forest associated with rocky outcrops at the forest-steppe ecotone in southern Argentina, Nahuel Huapi, Argentina. Photo: J. Birch

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (242)

216

Principles and Practice of Forest Landscape Restoration

Positive results were obtained with the SNP design assay for M. denticulata, which yield-ed recognizable hom*ozygote and heterozygote genotypes (Fig. 7.2). However, heterozygous individuals were only detected in the population of Santo Domingo Yanhuitlan, which was the most degraded locality consisting of denuded soil and gullies. An average Fst = 0.198 provides evidence of genetic differentiation among the eight studied populations. The aver-age negative within-population inbreeding (Fis = –0.282) can be explained by the excess of heterozygosity observed in Santo Domingo Yanhuitlán (observed and expected frequency of heterozygote genotypes were 0.44 and 0.34, respectively).

Figure 7.2 Mala-comeles denticulata

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (243)

217

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

The population of Santo Domingo Yanhuitlan includes the only heterozygous individuals found in the eight studied populations of M. denticula. As a result, it is extremely important to protect this unique genotype. The locality of Yanhuitlan was the most fragmented and degraded of all, so heterosis is probably related to survival mechanisms in M. denticulata in this degraded area.

Catopsis berteroniana (Schult and Schult., f) Mez is an epiphyte bromeliad of the Tillandsioi-deae subfamily. Catopsis berteroniana was classified in the Norma Oficial Mexicana 059 (NOM–059-ECOL–2001) as a species that requires special protection. In Oaxaca, México, it is commonly sold in the local markets as an ornamental plant for several religious festivities. Plants are col-lected by settlers in the localities of El Cerezal and Reynoso in the Santa Catarina Ixtepeji County where the vegetation is a dry fragmented shrubland and disturbed oak forest, respectively. Using satellite images from 1979 to 2005, the rate of deforestation in this region was calculated as 9.1% per decade. Forest remains in fragments that are becoming smaller (Fig. 7.3). Owing to forest re-duction and fragmentation we expected to find impoverished genetic diversity in these localities.

Figure 7.3 Catopsis berteroniana

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (244)

218

Principles and Practice of Forest Landscape Restoration

We tested recent bottlenecks on this bromeliad based on patterns of genetic diversity using allozymes. Nine out of the ten analyzed loci were polymorphic. The mean number of alleles per locus was 3.4 and mean observed and expected heterozygosities were 0.402 and 0.292 in El Cerezal and Reynoso, respectively. We found an average within-population inbreeding Fis = 0.256 and a degree of among-population divergence Fst = –0.021 (Table 7.3).

Contrary to our expectations, C. berteroniana displayed one of the highest levels of genetic diversity recorded in the family. We did not find evidence of recent bottlenecks despite the high levels of forest fragmentation and population reduction (Fig. 7.4). In addition, populations in El Cerezal and Reynoso were not genetically differentiated. A demographic study in progress revealed that population size is increasing. Although the total forest area is being severely reduced, the perimeter of the forest fragments has increased, probably favouring gene flow and establishment, as this species appears to grow better on the forest edges. This result shows that some plant species may benefit to a certain extent from forest fragmentation. Nevertheless, increasing fragmentation will eventually reduce fragment edges and thus decrease the available habitat for this species.

Table 7.3 A Ap RtCatopsis berteroniana

Locality A Ap Rt Ho He

El Cerezal 3.100 3.333 3.079 0.279 0.387

Reynoso 3.333 3.560 3.267 0.305 0.416

F values: Fit = 0.271Fis = 0.256Fst = – 0.021

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (245)

219

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Figure 7.4

Catopsis berteroniana .

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (246)

220

Principles and Practice of Forest Landscape Restoration

Dry temperate forests of southern Argentina

We conducted genetic surveys on Austrocedrus chilensis, the dominant tree species of dry warm-temperate forests of Patagonia, Argentina. It is a conifer within the Cupressaceae family that inhabits the forest-steppe ecotone along steep precipitation gradients of more than 1000 mm. It grows in Argentina between 37° and 42º S latitude and from the V to X Region in Chile. Austrocedrus chilensis is a timber species of high economic value and international conserva-tion concern. It is listed by the International Union for the Conservation of Nature as Vulnerable (VU A2c; B2ab (iii)). Main threats affecting Austrocedrus are herbivory and fire (Hechenleitner et al., 2005). The main research objective was to investigate if the gene pool of Austrocedrus is geographically structured, and to identify areas with high genetic diversity and/or unique variants. Those areas can be considered to be of high evolutionary potential and/or containing evolutionary novelties that should deserve conservation actions of such valuable and domi-nant trees. Also, a significant genetic structure along its distribution may guide germplasm col-lection for restoration actions to be undertaken in degraded habitats. Studies were focused on the understanding of underlying historical and contemporary processes shaping genetic pat-terns. We combined molecular information from three genetic markers: slowly-evolving DNA sequences of non-recombinant regions of the chloroplast, nuclear isozyme loci with moderate polymorphism, and hypervariable species-specific nuclear markers as microsatellites.

Plastid regions in Austrocedrus

To investigate the historical patterns of genetic diversity and gene flow, we optimized sequences of non-coding regions of mitochondria and chloroplast DNA in an attempt to perform phylogeo-graphic analyses on Austrocedrus. Seven different DNA non-coding regions of both, the mito-chondria and the chloroplast of Austrocedrus were sequenced. These yielded no polymorphism along the Austrocedrus range, which impeded phylogeographic analysis. This low mutation rate of the plastid DNA would allow phylogenetic reconstructions within the family, extending to the origin of Austrocedrus as a genus. Furthermore, these sequences blasted in NCBI, show affinities higher than 84% with other Cupressaceae such as Cryptomeria japonica or Chamaecyparis sp.

Isozyme variation in Austrocedrus

A study of the restoration genetics of dry forests of northern Argentina was performed on the dominant tree Austrocedrus chilensis. The study was conducted in the forest-steppe ecotone on the eastern slopes of the Patagonian Andes, Argentina, between 37° and 42ºS. We sampled a total of 1853 individuals in 67 populations along three regions covering the entire latitudinal range of Austrocedrus in Argentina represented as north (N), center (C), and south (S). All sampled locations are moderately disturbed, in terms of logging and grazing, such disturbance being less intense in the central area within Nahuel Huapi National Park. Previous studies have shown that fire is a major driver of ecological constraints in studied locations (Kitzberger, 2003). Regions not sampled included Chilean populations where the climatic and disturbance history of the species are probably older and are highly different to those in Argentina.

To identify the broad-scale trends in genetic differentiation throughout the species’ range, the populations were combined into three regions (north, centre, south), according to their geographical proximity and environmental envelope. The north region consists of 23 popu-lations located between 37–39ºS latitude characterized by dry climatic conditions and scarce vegetation. The centre region includes 25 populations from 40–41ºS where the west-east

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (247)

221

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

natural fragmentation gradient is more evident. The south region was based on grouping 19 populations located between 41–43ºS where the size of forest patches increases and the mean tree age decreases (K. Heinemann, UNCOMA, unpublished data).

A total of 12 allozyme loci were scored from 1,853 trees distributed over 67 populations. Resolved loci were Glycerate 2 dehydrogenase (G2D), Isocitrate dehydrogenase (Idh), Malate dehydrogenase (Mdh1, Mdh2), Malic enzyme (Me1, Me2), cathodal Peroxidase (Percat), Phos-phoglucoisomerase (Pgi1, Pgi2), 6-Phosphogluco dehydrogenase (6Pgd1, 6Pgd2), and Shiki-mate dehydrogenase (Skdh). All analyzed loci were polymorphic sensu stricto, with 3–5 alleles per locus. Genetic diversity metrics of Austrocedrus populations show a significant decrease with increasing latitude. Northern populations yielded higher effective number of alleles, total genetic diversity, and allelic richness than southern ones (F(1,65), p <0.05 all tests) (Fig. 7.5).

Figure 7.5 Austrocedrus chilensis

Inbreeding within northern and eastern populations of the centre of the range was high, positive, and statistically significantly different from zero (Fis > 0.14). Southern and western populations of the centre of the distribution displayed low levels of inbreeding that do not differ from zero (Fis < 0.06). The degree of genetic divergence between populations was low (Fst <0.16) and similar for the different areas.

Genetic studies performed on Austrocedrus show that along its range, populations vary in their rates of diversity and inbreeding. These genetic patterns can be explained in terms of the disturbance history of Patagonia. We hypothesize that fragmented, smaller, and relatively isolated northern populations with higher inbreeding are the result of a long history of human disturbance by fire. In contrast, towards the south, fire history and human impact are more recent and Austrocedrus populations consist of more continuous forest stands with lower inbreeding. In addition, the combined effects of human activity and historical factors such as the last glaciations in Patagonia explain the reduced genetic diversity recorded towards the south. This relates to the fact that Austrocedrus is a cold-intolerant plant. As a result, it is hypothesized that during cold periods Austrocedrus remained at northern warmer latitudes. Therefore, lower genetic diversity found in the south is the result of founder effects suffered during postglacial long-distance dispersal as suggested by the pollen record (Whitlock et al., 2006).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (248)

222

Principles and Practice of Forest Landscape Restoration

Microsatellite variation in Austrocedrus

In an attempt to combine historical and contemporary signals detected with the previous two molecular markers, nine micro-satellite markers (eight di-, and one trinucleotide bases) were isolated by ATG Genetics, Canada, and characterized for Austrocedrus chilensis. Four markers were optimized and scored with confidence, which were tested in 398 individual samples from 43 locations. Study areas included a subsample of eastern populations from southern Argentina used in the isozyme study and seven locations from Chile. The number of scored alleles ranged between 4 and 42 for the analyzed markers. Mean observed and expected heterozygosities were 0.732 and 0.684, respectively. Populations were grouped according to their location in five regions: coastal and Andean populations from Chile and in Argentina represented as north (N), centre (C), and south (S). Results show a reduction in polymorphism towards the coast in Chile and the south of the species’ range in Argentina (Figs. 7.6 and 7.7). These results suggest that eastern populations have derived from most variable northern Andean sources by long-distance dispersal. A decrease of genetic diversity towards the southeast suggests that this area has been relatively recently colonized as sug-gested by isozyme data, pollen records, and ecological niche modelling (Fig. 7.8).

Figure 7.6

Austrocedrus chilensis.

Figure 7.7 Austrocedrus chilensis

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (249)

223

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Figure 7.8 Austrocedrus

(a) (b)

(c)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (250)

224

Principles and Practice of Forest Landscape Restoration

Conclusions

Forest fragmentation and degradation affect patterns of genetic variation at different scales. The strength of such impacts depends particularly on autoecological characteristics of the studied spe-cies, historical events that have shaped gene pools, landscape-use change, environmental gradients, and disturbance regimes that are key determinants of such shifts. Dry forests are currently the focus of increasing conservation and restoration efforts. Restoration recommendations should con-sider patterns of within- and between-population genetic variation which in turn will be key de-terminants of restoration success. Specific conclusions for each of the study areas are given below.

(i) Seasonally dry subtropical forests of northern Argentina

Recommendations can be drawn based on previous studies and genetic results on Cedrela.

1. Mature stands of Cedrela balansae are in urgent need of conservation actions, if they still exist. Additionally, efforts should be made to encourage passive restoration of forest stands in piedmont forests, which in turn are the most threatened due to the advance of agriculture in lowland areas.

2. Active restoration of degraded areas should use germplasm of nearby local populations of Cedrela to avoid outbreeding depression, particularly in species with elevated inbreeding.

3. Additional studies are needed to describe the distribution range of C. saltensis and its conservation status, i.e. population assessment inside and outside protected areas.

4. Geographic variation patterns of economically relevant quantitative traits and their herita-bility in Cedrela should be investigated. These could be the base for genetic improvement and restoration programmes aiming at developing a resource for sustainable utilization.

5. The combination of distribution patterns at neutral and adaptive traits in the three Ce-drela species can be used to inform the designation of conservation areas.

(ii) Arid central Oaxaca, México

The genetic effects of fragmentation and forest degradation can be complex and vary consider-ably among species. Some species can benefit under moderate levels of fragmentation (Ramírez-Luis and del Castillo, 2009; del Castillo et al., unpublished data). Medium and large sized fragments can sustain some species, such as Catopsis berteroniana, which only grows at fragment edges. In such conditions, this species can maintain a relatively high genetic diversity without evidence of recent bottlenecks, and demographically shows a nearly stable size distribution and a positive growth (del Castillo, unpublished). There is an urgent need to conduct population studies on fragmented landscapes of temperate dry forests and shrublands of central and southern Mexico.

Some populations of the same species may hold high genetic diversity, as is the case of Malacomeles denticulata (Ramírez-Luis and del Castillo, 2009). This implies that breeding programmes should first examine different populations and not generalize a priori the ge-netic characteristics of the species based on studies of one or few populations. Therefore, conservation and restoration programmes should focus at the population level.

(iii) Dry temperate forests of southern Argentina

In the long term, the presence of genetic variation will be the main factor determining the suc-cess of restoration actions (Rice and Emery, 2003). Unfortunately it is uncommon to have genetic data available in advance of reintroduction efforts. Patterns of within- population genetic diver-sity and divergence among different populations across the entire species’ range should guide restoration actions. Some recommendations can be drawn from genetic studies on Austrocedrus:

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (251)

225

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

Genetic results from nuclear markers show that more genetic diversity exists towards the northern range of Austrocedrus. In particular, hypervariable micro-satellite markers provide evidence that northwestern populations in Chile were probably the source of eastern popu-lations located in Argentina. This result is interpreted as postglacial colonization from popu-lations that survived the Last Glacial Maximum on western slopes of the Andes, according to pollen records. Also, ecological niche modelling (Fig. 7.8) based on 18 bioclimatic variables and elevation models are also consistent with this hypothesis of range contraction towards northern latitudes for such mesothermic species during LGM conditions. As a result, Austro-cedrus populations in Chile are of high conservation concern. Urgent conservation actions should be developed to protect remnant stands of Austrocedrus in those highly disturbed and genetically diverse areas and also to promote passive restoration.

Northern naturally fragmented populations in Argentina have elevated genetic diversity a result of early postglacial colonization and therefore are of great conservation value. Efforts should be devoted to facilitate expansion of those populations. We recommend the exclu-sion of exotic cattle to promote passive restoration within those fragments such that natural establishment occurs.

Genetic diversity of Austrocedrus is geographically structured. Recommendations based on suitability maps for Austrocedrus have identified eight potential areas for restoration (Fig. 7.9). Given that Austrocedrus populations show latitudinal differences in genetic traits, and that potential areas to be restored are concordant with genetic structure, the design of resto-ration practices should include local germplasm collection and propagation.

Figure 7.9 Austrocedrus

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (252)

226

Principles and Practice of Forest Landscape Restoration

In conclusion, genetic data on Austrocedrus highlight the importance of historical and current processes when implementing conservation actions and restoration procedures. Rel-evant guidelines for restoration need to be species-specific and should also consider genetic patterns such that local germplasm is used in restoration efforts in species with significant genetic structures.

References

Anderson, E. 1948. Hybridization of the habitat. Evolution 2: 1–9.

Brown, A., Pacheco, S. 2006. Importancia del género Cedrela en la conservación y desarrollo sustentable de las Yungas australes. In: Pacheco, S., Brown, A.D. (eds.), Ecología y produc-ción de cedro (género Cedrela) en las Yungas australes. Ediciones del Subtrópico, Tucumán, Argentina: pp. 9–18. 224pp.

CONAF, CONAMA, BIRF. 1999b. Catastro y evaluación de recursos vegetacionales nativos de Chile. Monitoreo de Cambios. Universidad Austral de Chile, Pontificia Universidad Católica de Chile, Universidad Católica de Temuco, Santiago, Chile.

Cushman, S.A., McKelvey, K.S., Hayden, J., Schwartz, M.K. 2006. Gene flow in complex land-scapes: testing multiple hypotheses with causal modeling. American Naturalist 168: 487–499.

Dinerstein, E., Olson, D., Graham, D., Webster, A., Primm, S., Bookbinder, M., Ledec, G. 1995. A conservation assessment of the terrestrial ecoregions of Latin America and the Caribbean. WWF – World Bank.

Echeverría, C., Coomes, D., Salas, J., Rey-Benayas, J.M., Lara, A., Newton, A. 2006. Rapid de-forestation and fragmentation of Chilean temperate forests. Biological Conservation 130: 481–494.

Epps, C.W., Wehausen, J.D., Bleich, V.C., Torres, S.G. 2007. Optimizing dispersal and corridor models using landscape genetics. Journal of Applied Ecology 44: 714–724.

Gillies, A.C.M., Cornelius, J.P., Newton, A.C., Navarro, C., Hernandez, M., Wilson, J. 1997. Ge-netic variation in Costa Rican populations of the tropical timber species Cedrela odorata L. assessed using RAPDs. Molecular Ecology 6: 1133–1145.

Gillies, A.C.M., Navarro, C., Lowe, A.J., Newton, A.C., Hernandez, M., Wilson, J., Cornelius, J.P. 1999. Genetic diversity in Mesoamerican populations of mahogany (Swietenia macro-phylla), assessed using RAPDs. Heredity 83: 722–732

Grau, A., Zapater, M.A., Neumann, R.A. 2006. Botánica y distribución del género Cedrela en el noroeste de Argentina. In: Pacheco, S., Brown, A.D. (eds.), Ecología y producción de cedro (género Cedrela) en las Yungas australes. Ediciones del Subtrópico, Tucumán, Argentina: pp. 19–30. 224pp.

Hechenleitner, P., Gardner, M., Thomas, P., Echeverria, C., Escobar, B., Brownles, B.S., Martínez, C. 2005. Plantas amenazadas del Centro-Sur de Chile. Universidad Austral de Chile and Real Jardín Botánico de Edimburgo, Valdivia, Chile. 188pp.

Kitzberger, T. 2003. Regímenes de fuego en el gradiente bosque-estepa del noroeste de Pat-agonia: variación espacial y tendencias temporales. In: Kunst, C.R., Bravo, S., Panigatti, J.L.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (253)

227

Impact of forest fragmentation and degradation on patterns of genetic variation and its implication for forest restoration

(eds.), Fuego en los escosistemas argentinos. Ediciones Instituto Nacional de Tecnología Agropecuaria, Santiago del Estero, Argentina: pp. 79–92. 330pp.

Luikart G., Cornuet J.M. 1998. Empirical evaluation of a test for identifying recently bottle-necked populations from allele frequency data. Conservation Biology 12: 228–237.

Malizia, L.R., Blundo, C., Pacheco, S. 2006. Diversidad, estructura y distribución de bosques con cedro en el noroeste de Argentina y sur de Bolivia. In: Pacheco, S., Brown, A.D. (eds.), Ecología y producción de cedro (género Cedrela) en las Yungas australes. Ediciones del Subtrópico, Tucumán, Argentina: pp. 83–103. 224pp.

McRae, B.H., Dickson, B., Keitt, T.H., Shah, V.B. 2008. Using circuit theory to model connectiv-ity in ecology, evolution, and conservation. Ecology, 89: 2712–2724.

Miles, L., Newton, A., Defries, R., Ravilious, C., May, I., Blyth, S. Kapos, V., Gordon, J. E. 2006. A global overview of the conservation status of tropical dry forests. Journal of Biogeography 33: 491–505.

Muellner, A.N., Pennington, T.D., Chase, M.W. 2009. Molecular phylogenetics of Neotropical Cedreleae (mahogany family, Meliaceae) based on nuclear and plastid DNA sequences reveal multiple origins of “Cedrela odorata”. Molecular Phylogenetics and Evolution 52: 461–469.

Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89: 583–590.

NOM–059-ECOL–2001. Norma Oficial Mexicana. Environmental protection of native species of plants and animals of México, categoric risk.

Provan, J., Soranzo, N., Wilson, N.J., Goldstein, D.B., Powell W. 1999. A low mutation rate for chloroplast microsatellites. Genetics 153: 943–947.

Quiroga, M.P., Premoli, A.C. 2007. Genetic patterns in Podocarpus parlatorei reveal the long-term persistence of cold-tolerant elements in the southern Yungas. Journal of Biogeogra-phy 34: 447–455.

Ramírez-Luis, J., del Castillo, R.F. 2009. Quantitative genetics of Malacomeles denticulata WP6 Technical report REFORLAN.

Rice, K.J., Emery, N.C, 2003. Managing microevolution: restoration in the face of global cli-mate change. Frontiers in Ecology and the Environment 1: 469–478.

Rzedowski, J. Calderón G. 2005. Rosaceae. Flora del Bajio y de regiones adyacentes. Fasciculo 1: 135–163.

United Nations Development Programme. 2004. Dryland conservation and development: striking a balance. United Nations Development Programme, Washington.

Weising, K., Nybom, H., Wolff, K., Kahl, G. 2005. DNA fingerprinting in plants, principles, meth-ods and applications. CRC Press Taylor and Francis Group, Boca Raton Fl, USA. 444 pp.

Whitlock, C., Bianchi, M.M., Bartlein, P.J., Markgraf, V., Marlon, J., Walsh, M., McCoy, N. 2006. Postglacial vegetation, climate, and fire history along the east side of the Andes (lat 41–42.5°S), Argentina. Quaternary Research 66: 187–201

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (254)

228

Principles and Practice of Forest Landscape Restoration

Zapater, M.A., del Castillo, E.M., Pennington, T.D., 2004. El genero Cedrela (Meliaceae) en la Argentina. Darwiniana 42, 347–356.

Pacheco L.F., Simonetti J.A. 2000. Genetic structure of a mimosoid tree deprived of its seed disperser, the spider monkey. Conservation Biology 14: 1766–1775.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (255)

229

8 LANDSCAPE-SCALE DYNAMICS AND

RESTORATION OF DRYLAND FOREST

ECOSYSTEMS

A.C. Newton, E. Cantarello, N. Tejedor, T. Kitzberger, C. Echeverría, G. Williams-Linera, D. Golicher, G. Bolados, L. Malizia, R.H. Manson, F. López-Barrera, N. Ramirez-Marcial, M. Martinez-Icó, G. Henriquez, R. Hill

Introduction

The restoration of forest landscapes is typically achieved either through passive restoration, involving the establishment of forest cover through natural regeneration, or some form of active restoration approach, involving the establishment of trees by artificial means. The ex-istence of different restoration options raises the question of how an appropriate restoration approach might best be identified for any individual location. Where feasible, passive restora-tion approaches are often likely to be preferred, because of the intrinsically lower costs of tree establishment. However, the potential for natural regeneration of forest cover may often be limited, particularly in landscapes that are highly degraded. Factors that can limit the process of natural regeneration include a lack of a source of propagules, perhaps because an individual site is isolated from remnant forest stands; adverse site characteristics for seed germination or seedling establishment, such as degraded or compacted soils (see Chapter 5); or an adverse disturbance regime that causes high mortality of juvenile trees. In areas where passive restoration approaches are associated with a high risk of failure, active restoration approaches may be preferred.

Restoration planning should therefore ideally be informed by an understanding of where natural regeneration is likely to occur on a landscape, within a given timescale. Such informa-tion might help identify those locations where natural regeneration is most likely to occur under different patterns of land use. In order to make such predictions, some form of forest modelling approach is required. Ideally, such a model should incorporate those ecological processes that influence natural regeneration, including dispersal, competition, and tree sur-vival and growth in locations with different site characteristics. In addition, it is essential that the model enables predictions or forecasts to be made that are spatially explicit, in order to be of practical value in restoration planning.

As noted by Newton (2007a), a wide variety of different forest modelling approaches have been developed, although only a small proportion of these are suitable for exploring forest dynamics at the landscape scale. Spatially explicit approaches to forest modelling have been greatly supported by recent developments in Geographic Information System (GIS) technologies, which now enable GIS to be linked to models of forest dynamics, either for pre-processing data for use in a non-spatial modelling, or for displaying model output. Closer linkages between the model and GIS can be achieved if they share the same data structures. Examples of models that have been developed specifically to operate at landscape scales are provided by Frelich et al. (1998), Frelich and Lorimer (1991) and Liu and Ashton (1998).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (256)

230

Principles and Practice of Forest Landscape Restoration

However, the use of forest models to support restoration planning, or conservation man-agement more generally, has been very limited to date (Newton, 2007a). Such models have enormous potential for supporting decision making relating to sustainable forest manage-ment, particularly if they can be linked with other analytical tools, such as GIS (Newton et al., 2009a).

In this chapter, we describe the application of a spatially explicit model to examine the dynamics of dryland forest landscapes, with the aim of informing forest restoration planning. Specifically, the objective of this investigation was to simulate the landscape-scale dynamics of dryland forest in order to assess the potential for natural recovery of forest landscapes under different disturbance regimes. Information on the rate of ecological recovery under different disturbance scenarios is required to evaluate the feasibility of passive restoration approaches (Vieira and Scariot, 2006). Ideally such information should be spatially explicit, given that forest restoration should be undertaken at the landscape scale in order to address the problem of forest fragmentation and to restore connectivity (Mansourian et al., 2005). Although a number of studies have analyzed dry forest recovery after disturbance (Guari-guata and Ostertag, 2001; Vieira and Scariot, 2006; Sampaio et al., 2007; Griscom et al., 2009), very little information is available regarding the processes of recovery of dry forest at the landscape scale.

This study employed the model LANDIS-II, which was designed to simulate the dynamics of forested landscapes through the incorporation of ecological processes including succes-sion, disturbances and seed dispersal over long time domains (Scheller et al., 2007). The LANDIS-II model is an elaboration of the LANDIS family of landscape disturbance and forest succession models. Although the architecture has changed since the initial version (Mlad-enoff et al., 1996) and new features have been added, LANDIS-II retains many principles from earlier versions that have been widely tested and applied in different parts of the world (He and Mladenoff, 1999; Mladenoff and He, 1999; Mladenoff, 2004; Scheller et al., 2005; Wang et al., 2006; Swanson, 2009). However, we are not aware of any previous attempt to apply LANDIS-II, or any other spatially explicit model of forest dynamics, to dry forest land-scapes in Latin America.

This chapter first provides a brief description of the LANDIS-II model, and then presents an overview of the results obtained from applying LANDIS-II to four different study areas in Latin America; two in Mexico, one in central Chile and one in southern Argentina. The impli-cations of the results obtained for planning the restoration of dryland forest landscapes are then explored.

The LANDIS-II model

The LANDIS model is described in more detail elsewhere (He and Mladenoff, 1999; Mladenoff and He, 1999; Mladenoff, 2004); <http://www.landis-ii.org>). In essence, LANDIS-II uses a cell-based data format; within each cell it tracks the presence/absence of tree species age cohorts at a time step specified by the user. Vegetation patches can aggregate and disaggregate in response to spatial patterns of stochastic rules of disturbance and succession. Tree species succession is a competitive process governed by species life history parameters (longevity, age of sexual maturity, shade and fire tolerance class, effective and maximum seed dispersal distance, vegetative reproduction probability, minimum and maximum age of vegetative

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (257)

231

Landscape-scale dynamics and restoration of dryland forest ecosystems

reproduction, and post-fire regeneration), and the probability of species establishment on different ecoregions (or landtypes). Tree succession interacts with several spatial processes (i.e. seeding, wind and fire disturbances, and harvesting). As disturbance in LANDIS-II is stochastic, calibration is required to ensure the output is fitted to the ecological values of the simulated area (Franklin et al., 2001; Lafon et al., 2007; Syphard et al., 2007).

Case studies in Mexico

Study areas

Research was undertaken in two study areas dominated by tropical dry forest (TDF), name-ly the Tablon, Chiapas, and central Veracruz, Veracruz, Mexico. Both study areas are global conservation priorities, having been identified as global biodiversity hotspots (Myers et al., 2000), and in recent decades both have been degraded at a high rate owing to the effects of human disturbances (Challenger and Dirzo, 2009). Both study areas cover similar areas, but they differ in the percentage of forest cover. While in central Veracruz only 27% of the land-cover of the entire area is represented by forests and shrubland, almost 90% of the land-cover of Tablon is represented by a forest type. According to CONABIO (2006) both areas have a high degree of marginalization; an average of 23 inhabitants km2 and 14 inhabitants km2 respectively were recorded in 2000 in Central Veracruz and Tablon.

The Tablon study area covers 24,735 ha and is situated between 675 and 1537 m altitude in the municipalities of Villaflores and Jiquipilas, state of Chiapas (16º11’38’’ and 16º22’29’’ N, and 93º31’57’’ and 93º44’31’’ W). The climate is defined as warm sub-humid, with an average annual rainfall between 1200 and 2800 mm concentrated from late May to early November (Aguilar-Jiménez, 2008). The natural vegetation of Tablon forms a gradient of forest types ranging from low-stature deciduous tropical forest in the lower elevations of the study area, through dry oak and pine-oak with increasing elevation, and with pine forests on the highest ridges. Tablon falls within the La Sepultura Biosphere Reserve, which was designated in 1995 for its high number of endemic species, high biodiversity value and species richness (Box 8.1).

Box 8.1 Anthropogenic impact on dry forests in Chiapas, MexicoN. Tejedor

Tropical dry forests, which host a large number of endemic species (Mooney et al., 1995), are among the least studied forests worldwide and the most threatened by human actions (Miles et al., 2006). La Sepultura Biosphere Reserve, which is located in the southwestern region of the state of Chiapas, Mexico, between 16°00’18’’ and 16°29’01’’ of north latitude and 93°24’34’’ and 94°07’35’’ west longitude, with an area of 167,309 ha, has been designated for its large amount of endemic species and its various examples of ecosystems scarcely represented in other protected areas of Mexico, including areas of dry forests (INE,1999).

Forest fires are an important problem in Chiapas, and in the reserve this situation is worsened by the fact that the region is extremely vulnerable during the dry season (November to April, and even May) owing to the strong winds from the coast (October to March), which dry the grass and increase the likelihood of fast propagation of fires (Hernandez-Lopez, 2005). The impact of fires on the existing vegetation is exacerbated by agricultural practices such as maize cultivation using the slash and burn technique, which increases soil fertility for a period of 3 or 4 years until the soil nutrients are exhausted, leading to further forest fires and deforestation. This is despite the fact

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (258)

232

Principles and Practice of Forest Landscape Restoration

Slash and burn agriculture in a tropical subdeciduous forest in Ocuilapa, Chiapas, México. Photo: N. Ramirez-Marcial

Reforestation of seasonal dry premontane forest with timber species. Photo: L. Malizia

Forest restoration in progress in Chile. Photo: C. Echeverria

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (259)

233

Landscape-scale dynamics and restoration of dryland forest ecosystems

that there are restrictions on the use of fire as a mechanism to convert the forest to pasture land or control of forest succession, as mentioned in the management plan (INE, 1999) (Figs. 1 and 2). According to the geographical database in LAIGE (2007), land-use cover changes have been identified between 1975 and 2000, alongside roads and around population centres, where forests have been converted to agriculture and pasture land.

Areas within La Sepultura, as with many other areas within Chiapas (e.g. Hammond, 1995), have highly compacted soils and the sources of propagules are depleted as a result of anthropogenic disturbance, allowing only the colonization of ruderal species. For example, in a survey carried out in 2007 (Tejedor, 2007) at 1065 m a.s.l within the reserve, where species such as Quercus sp., Phoebe chiapensis, Inga sp. and Manilkara zapota, are expected to be found, the only species present from those mentioned were Quercus sapotifolia and a species that grows favourably in disturbed areas and savannah type vegetation, Byrsonima crassifolia (Fig. 3). The dominant vegetation was the ground vegetation, which was mainly composed of grass species. The canopy cover was 43.84%. This suggests that the main disturbance in the area is the conversion of forest for grazing under mature trees, through fire and cutting. A natural disturbance is soil erosion, as the area is on a slope; but this is exacerbated by the deforestation which leads to lack of soil stability.

Figure 1

Box 8.1 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (260)

234

Principles and Practice of Forest Landscape Restoration

Figure 2

Figure 3

Box 8.1 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (261)

235

Landscape-scale dynamics and restoration of dryland forest ecosystems

The central Veracruz study region, with an area of 29,468 ha, is situated between 10 and 507 m altitude in the state of Veracruz, Mexico (19º07’45’’ and 19º21’18’’N, and 96º21’33’’ and 96º41’12’’W). The climate is defined as warm sub-humid (minimum and maximum average temperatures are 20° and 31°C, respectively) with rainfall of 800 to 1500 mm, occurring primarily from June to September followed by an extended dry season. Areas on the eastern side of central Veracruz have a warm-humid climate, whereas those on the western side are characterized by a warm-dry climate. The original vegetation was predominantly tropical dry forest. None of the remaining forest fragments are under protection. The primary land use is cattle ranching, which is generally undertaken on a relatively small scale by private landowners.

Model parameterization

The LANDIS-II model is designed to accept raster imagery as a spatially explicit input to simulate landscape dynamics. Full details of model parameterization for these study areas are presented by Cantarello et al. (2010). In Tablon, input raster data included a Digital Eleva-tion Model (DEM) and QuickBird satellite imagery, from which a series of secondary maps were derived. The DEM (50 m cell-size) was derived from the 30 m resolution national DEM, resampled to a 50 m grid using regularized spline with tension (Mitasova and Mitas, 1993). A direct beam solar radiation map (50 m cell-size) was calculated using the formulae proposed by Rigollier et al. (2000) and implemented in the GRASS module r.sun (Neteler and Mitasova, 2008). An ecoregions map was produced from the combination of the DEM with the beam solar radiation map.

Three QuickBird scenes acquired in November–December 2004 were obtained as a mosaic to cover the study area. A basic land-cover map (50 m cell-size), which identified forest, pasture, roads, urban areas and permanent agricultural areas was derived from the QuickBird imagery. The production of the basic land-cover map (Fig. 8.1) involved a combination of supervised classification of the multispectral mosaic to separate for-est from pasture with the manual digitizing of roads, urban areas and permanent agri-cultural areas on the pan mosaic. A forest stand types map (50 m cell-size) was derived from the QuickBird imagery and ecoregions map. The production of the forest stand types map involved performing an unsupervised classification of the forest component of the QuickBird imagery into twenty forest classes which were split by ecoregions into a 28-forest stand types layer. This layer was validated by manual comparison against the panchromatic mosaic and the field survey data describing species and stem diameters distributions.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (262)

236

Principles and Practice of Forest Landscape Restoration

Figure 8.1

In central Veracruz, input raster data included a DEM and SPOT satellite imagery, which were used to obtain a series of secondary maps. The DEM (cell-size 80 m) was derived from Shuttle Radar Topography Mission (SRTM) data with a resolution of three arc seconds per pixel. A soil types map (cell-size 80 m) was extracted from the national edaphology map (INIFAP and CONABIO, 1995). An ecoregions map was produced from the combination of the DEM with soil types map.

Three SPOT high-resolution visible and infrared (HRVIR) multispectral images (20 m reso-lution) from December 2007 and January 2008 were used to produce a land-cover map (80 m cell-size), which identified shrubland, forest, agriculture, urban and roads, water, planta-tion and pasture. Validation was conducted through field visits and visual inspection of high resolution Google Earth images (©2008 Google Inc., California, USA). A forest stand types map (80 m cell-size) was derived from the land-cover map and the field survey data (Fig. 8.2).

Figure 8.2

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (263)

237

Landscape-scale dynamics and restoration of dryland forest ecosystems

Bursera simaruba, one of the dominant tree species in the tropical dry forest of Paso de Ovejas, Veracruz. Photo taken in Hato Los Marines during the rainy season. Photo: G. Williams-Linera

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (264)

238

Principles and Practice of Forest Landscape Restoration

To provide field data, 36 circular plots (0.1 ha each) were established along an altitudinal gradient in Tablon and 100 survey plots (0.01 ha each) were established within ten TDF frag-ments in central Veracruz (Williams-Linera and Lorea, 2009). In each plot, tree species were identified and diameter at breast height (dbh) of each tree >5 cm was measured. Dbh values were converted into ages by consulting a group of local experts.

In LANDIS-II the landscape is stratified into ecoregions, which are ecologically hom*og-enous sub-areas characterized by the same habitat suitability (establishment probability) for each species to be modelled. Ecoregions can be active or non-active depending on whether they represent areas where forests can grow or not. In Tablon, one non-active and fifteen active ecoregions were considered; active ecoregions covered 24,354 ha, non-active 381 ha. The species establishment probabilities for each ecoregion were de-rived from the outputs of generalized additive models as interpreted by local experts. Input to the models included species occurrence data, obtained from the MOBOT Tropi-cos® data base (©2010 Missouri Botanical Garden, USA), and climate data extracted from the Worldclim database (Hijmans et al., 2005). In central Veracruz, one non-active and thirteen active ecoregions were considered; active ecoregions covered 20,788 ha, and non-active 8,680 ha. The species establishment probabilities for each ecoregion were derived from the outputs of GARP models (Stockwell and Peters, 1999), which were produced by using the software DesktopGarp (©2002 University of Kansas Center for Research, Inc., USA) and interpreted by local experts.

The inputs required by LANDIS-II include information about the distribution, composition and age structure of forest stands at year 0. In both study areas, to populate each cell of the landscape with species and age cohort we combined the forest cover map with the field survey data describing species and age distributions.

Species attributes

For each species to be modelled, LANDIS-II requires information about longevity, age of sex-ual maturity, shade and fire tolerance class, effective and maximum seed dispersal distance, vegetative reproduction probability, minimum and maximum age of vegetative reproduction, and post-fire regeneration. In central Veracruz, species attributes were extracted for the 22 most abundant species (i.e. species with >14 sampled individuals in the field survey). The subset was characterized by a range of shade and fire tolerances (Table 8.1). In Tablon, none of the 23 most abundant species (i.e. species with >5 sampled individuals in the field sur-vey) had high shade tolerance. Two relatively infrequent species (Sapindus saponaria and Gyrocarpus mocinnoi) were therefore included in the model, to ensure that the full range of shade tolerance was included, giving a total of 25 species included in the model. In both study areas species attributes were extracted from the scientific literature and by consulting local experts.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (265)

239

Landscape-scale dynamics and restoration of dryland forest ecosystems

Table 8.1

Species Species abbr. Long Mat ShT FiT EffD MaxSD VRP Min VRP

Max VRP

P-FiR

Acacia cochliacantha Acaccoch 50 2 1 5 5 100 1 2 50 re-sprout

Brosimum alicastrum Brosalic 150 20 5 1 5 10000 1 3 60 re-sprout

Bursera cinerea Burscine 80 5 4 1 5 10000 0 0 0 none

Bursera fa*garoides Bursfa*ga 80 5 4 3 5 10000 0 0 0 none

Bursera graveolens Bursgrav 80 5 4 2 5 10000 0 0 0 none

Bursera simaruba Burssima 80 5 3 3 5 10000 1 2 50 re-sprout

Calyptranthes schiedeana Calyschi 60 15 5 2 5 10000 0 0 0 none

Cochlospermum vitifolium Cochviti 60 15 3 5 3 100 1 1 30 re-sprout

Ceiba aesculifolia Ceibaesc 40 10 3 2 5 100 1 2 30 re-sprout

Comocladia engleriana Comoengl 70 25 3 2 200 1000 1 7 60 re-sprout

Eugenia hypargyrea Eugehypa 40 5 5 1 5 100 0 0 0 none

Guazuma ulmifolia Guazulmi 40 3 1 5 5 100 1 2 40 re-sprout

Heliocarpus donnell-smithii Helidonn 50 10 1 3 100 100 1 3 25 re-sprout

Ipomoea wolkottiana Ipomwolc 60 10 1 3 11 100 1 2 20 re-sprout

Leucaena lanceolata Leuclanc 20 2 1 2 5 100 0 0 0 serotiny

Luehea candida Luehcand 70 20 3 2 20 100 1 3 60 re-sprout

Lysiloma microphyllum Lysimicr 70 15 1 2 5 100 1 3 60 re-sprout

Savia sessiliflora Savisess 30 5 5 1 5 100 0 0 0 none

Senna atomaria Sennatom 30 5 1 4 5 100 1 2 10 re-sprout

Stemmadenia pubescens Stempube 30 5 3 1 5 1000 0 0 0 none

Tabebuia chrysantha Tabechry 120 15 3 3 20 100 1 2 50 re-sprout

Thouinidium decandrum Thoudeca 70 10 1 4 20 100 1 3 60 re-sprout

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (266)

240

Principles and Practice of Forest Landscape Restoration

Table 8.2

Species Species abbr. Long Mat ShT FiT EffD MaxSD VRP Min VRP

Max VRP

P-FiR

Acacia cornigera Acaccorn 30 3 1 4 50 10000 1 1 15 re-sprout

Acacia pennatula Acacpenn 40 3 1 4 100 10000 1 1 20 re-sprout

Bursera bipinnata Bursbipi 50 5 3 3 25 10000 1 3 25 re-sprout

Bursera excelsa Bursexce 50 5 3 3 20 10000 1 3 25 re-sprout

Bursera simaruba Burssima 80 3 3 4 20 10000 1 3 30 re-sprout

Byrsonimia crassifolia Byrscras 30 5 1 5 10 15000 1 3 15 re-sprout

Diphysa robiniodes Diphrobi 40 8 2 3 20 100 1 3 40 re-sprout

Erythrina chiapasana Erytchia 40 5 2 3 10 100 1 3 20 re-sprout

Erythrina folkersii Erytfolk 40 6 1 3 10 100 1 3 25 re-sprout

Eysenhardtia adenostylis Eyseaden 30 4 2 1 20 100 1 1 30 re-sprout

Guazuma ulmifolia Guazulmi 30 3 1 5 100 1000 1 2 25 re-sprout

Gyrocarpus mocinnoi Gyromoci 30 4 5 4 10 100 0 0 0 none

Heliocarpus reticulatus Helireti 40 10 1 1 20 100 0 0 0 none

Leucaena diversifolia Leucdive 25 3 1 2 10 10 1 1 20 re-sprout

Lonchocarpus rugosus Loncrugo 45 10 2 3 20 15000 1 5 25 re-sprout

Pinus maximinoi Pinumaxi 100 20 1 4 20 100 0 0 0 none

Pinus oocarpa Pinuooca 100 7 1 5 20 100 0 0 0 serotiny

Quercus acutifolia Queracut 120 20 2 4 20 10000 1 5 60 re-sprout

Quercus castanea Quercast 80 15 2 2 10 10000 1 5 40 re-sprout

Quercus conspersa Quercons 120 20 2 4 20 10000 1 5 60 re-sprout

Quercus elliptica Querelli 200 20 2 5 100 10000 1 1 80 re-sprout

Quercus penducularis Querpedu 200 15 2 5 100 10000 1 1 80 re-sprout

Quercus segoviensis Quersego 200 20 2 4 100 10000 1 1 80 re-sprout

Sapindus saponaria Sapisapo 80 15 4 1 10 10 0 0 0 none

Ternstroemia tepezapote Terntepe 40 5 1 3 100 10000 1 1 30 re-sprout

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (267)

241

Landscape-scale dynamics and restoration of dryland forest ecosystems

Table 8.3

Tablon Central Veracruz

_x

S.D. _x

S.D.

NO-DIST 91.24 6.59 99.89 0.14

GRAZ 80.69 10.2 99.14 0.36

SIF 88.67 3.72 92.91 0.40

SIF-GRAZ 81.16 5.31 92.36 0.48

LFF 87.00 3.13 59.74 4.19

LFF-GRAZ 76.28 6.04 21.79 4.95

Disturbance scenarios

Scenarios were developed to explore the impacts of both fire and grazing, as these distur-bances have caused the most serious degradation of dry forest in Mexico. Six different sce-narios were simulated: (i) no disturbance (NO-DIST); (ii) grazing without fire (GRAZ); (iii) small, low intensity, infrequent fires without grazing (SIF); (iv) small, low intensity, infrequent fires with grazing (SIF-GRAZ); (iv) large, intense, frequent fires without grazing (LFF); and (vi) large, intense, frequent fires with grazing (LFF-GRAZ). The Base Fire (v2.1) and Base Harvest (v1.2) extensions of LANDIS-II were used to generate the fire and grazing scenarios. LANDIS simulations were conducted for 400 years to allow fire and grazing to fully manifest their effects on forest succession. The time steps were set at 10 years for tree succession, 5 years for fire disturbance and 1 year for grazing.

Results

The LANDIS-II outputs consist of raster maps, each corresponding to a time step specified by the user (10 years in this study). Forest cover (defined here as percentage cover of trees >10 years) increased quite rapidly under NO-DIST scenario from 89.9% to 97.6% in Tablon and 6.26% to 99.6% in central Veracruz after only 50 years, producing a landscape dominated by forest cover after 400 years (Figs. 8.3 and 8.4). Under LFF, SIF, GRAZ and SIF-GRAZ scenarios, forest cover still increased, but values were lower than the NO-DIST scenario after 50 years, with values reaching between 90.1% and 96.2% in Tablon; and 50.8% and 99.5% in central Veracruz. Under LFF-GRAZ scenario the forest cover decreased after 50 years to 86.3% in Tablon, but slightly increased to 9.62% in central Veracruz. Differences in mean forest cover were recorded under the six scenarios in both study areas, with lowest values recorded un-der LFF-GRAZ in both cases.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (268)

242

Principles and Practice of Forest Landscape Restoration

Figure 8.3

Figure 8.4

Simulation results indicated that tree size class structure differed under the six scenarios in both study areas. Over 400 years, the highest percentage cover of the landscape was oc-cupied by: (i) trees >100 years old under NO-DIST, SIF, GRAZ and SIF-GRAZ scenario after 100 years (73.9%, 46.5%, 72.3%, and 41.1%, respectively); and (ii) trees between 31–60 years old under LFF and LFF-GRAZ scenarios after 50 years (45.0% and 35.1%, respectively). When only the fire and grazing disturbance scenarios are considered, the median percentage cover

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (269)

243

Landscape-scale dynamics and restoration of dryland forest ecosystems

of trees older than 60 years was the highest under the SIF scenario (53.0%), which differed statistically from the median percentage cover value under the LFF-GRAZ scenario (p<0.05; Mann-Whitney test). In central Veracruz, similar results to Tablon were obtained when me-dian percentage cover of trees > 60 years was compared. Values were the lowest under the LFF-GRAZ scenario (0.05%), which differed statistically from the median percentage cover values under all of the fire and grazing scenarios (p <0.05 in each case; Mann-Whitney test). When only the fire and grazing disturbance scenarios are considered, the median percentage cover of trees older than 60 years was the highest under the SIF scenario (64.6%), which dif-fered statistically from the median percentage cover values under the LFF, SIF, LFF-GRAZ and SIF-GRAZ scenarios (p <0.05 in each case; Mann-Whitney test).

Simulation results show that modelled species richness (25 modelled species out of 53 sampled species in Tablon, and 22 modelled species out of 97 recorded species in central Veracruz) differed under the six scenarios in both study areas. In Tablon the median percent-age area with �5 species was the highest under the SIF scenario (47.8%), which differed sta-tistically from the median percentage cover values under the NO-DIST and GRAZ scenarios (p <0.05 in each case; Mann-Whitney test). When only the fire and grazing disturbance sce-narios were considered, the median percentage area with �5 species was the lowest under the GRAZ scenario (30.4%), which differed statistically from the median percentage cover values under the LFF, SIF and LFF-GRAZ scenarios (p <0.05 in each case; Mann-Whitney test). In central Veracruz at year 0, similar to Tablon, the median percentage area with �5 species was the highest under the SIF scenario (31.5%), which differed statistically from the median percentage cover values under all of the fire and grazing scenarios (p <0.05 in each case; Mann-Whitney test). When only the fire and grazing disturbance scenarios were considered, the median percentage area with �5 species was the lowest under the LFF scenario (0.14%), which differed statistically from the median percentage cover values under the SIF, GRAZ and SIF-GRAZ scenarios (p <0.05 in each case; Mann-Whitney test).

Discussion

The LANDIS-II simulations performed in this study suggest that forest cover can increase quite rapidly if protected from grazing and fire, a finding that is supported by field observa-tions. For example, Sampaio et al. (2007) indicated that TDF in central Brazil was able to re-cover rapidly in the absence of anthropogenic disturbance, owing to the re-sprouting ability of the tree species present in that area. Griscom et al. (2009) noted that in Panama, relatively diverse second-growth dry forests were produced within three years following pasture aban-donment. Results obtained here also suggest that with time, an increasing proportion of the landscape becomes occupied by trees older than 60 years. This is in accordance with the observation made by Powers et al. (2009) indicating that in regenerating dry forests in Costa Rica, larger individuals increase asymptotically with stand age. These results highlight the po-tential scope for passive restoration of TDF, if forests can be adequately protected, and imply that such forests may be relatively resilient to human disturbance.

Simulations indicate opposite trends with regards to species richness in the two study areas considered here, with the number of species tending to increase over time in central Veracruz but to decline in Tablon, in the absence of disturbance. The majority of the species present initially were highly fire-tolerant but relatively intolerant of shade, and are therefore likely to disappear as a result of competitive processes as the canopy closes. Simulations also

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (270)

244

Principles and Practice of Forest Landscape Restoration

show that in both study areas in the absence of disturbance, the landscape is increasingly occupied by shade-tolerant and re-sprouting species with a relatively long seed dispersal distance; these species cover between 76–99% of the landscape after 200 years. These results are consistent with other studies that suggest shade-tolerant species tend to be dominant in secondary forests after 100–400 years since abandonment (e.g. Guariguata and Ostertag 2001).

The LANDIS-II simulations under both a scenario of small, surface, low intensity infrequent fires (SIF) and a scenario of large, crown, high-intensity frequent fires (LFF) suggest that for-est cover can recover quite quickly in terms of extent, owing to the high re-sprouting ability of the tree species. However under LFF, in Tablon, the forest cover started to decline gradually after 50 years; in central Veracruz, tree species older than 10 years only occupied half of the cover value under the LFF scenario than under SIF. In contrast with the SIF scenario, the LFF scenario explored here strongly modified the forest structure and composition in both Tab-lon and central Veracruz, highlighting an issue of concern regarding the potential impacts of increasing frequency of high intensity fires on dry forest conservation in this region.

The LANDIS-II projections reveal that when grazing is acting in combination with fire, the forest cover, structure and composition vary markedly depending on the intensity, fre-quency and extent of the fires. The forest cover under LFF-GRAZ was lower than under LFF alone in both study areas, suggesting that grazing significantly affects the forest cover under a frequent, high intensity fire regime. The combination of frequent, high intensity fires and grazing was also found to negatively affect the structure and composition of dry forest in both study areas. These results are consistent with other studies that indicate repeated fire and grazing as the most important types of anthropogenic disturbance impeding TDF regen-eration (Janzen, 1988).

These results suggest that passive restoration of dry forest is achievable at the landscape scale in both Tablon and central Veracruz if grazing were to be excluded, and fire were to be carefully managed to achieve a regime of infrequent, low intensity fires. Of the four distur-bance scenarios analyzed, SIF is the only scenario that does not negatively affect forest cover, composition and structure of TDF. Simulations revealed that a combination of frequent, in-tense fires and grazing has a major negative impact on TDF, leading to its destruction. These results highlight the importance of forest protection for both the conservation and restora-tion of TDF, which is a global priority (Miles et al., 2006) as well as an urgent requirement in Mexico (Gordon and Newton, 2006a, b).

Case study in central Chile

This research was undertaken in a landscape in the central Chilean Mediterranean-climate region, focusing on an area that is currently being proposed for designation as a Biosphere Reserve. As a result of its high floristic endemism, the region is a global conservation priority, as illustrated by its inclusion in the Global 200 ecoregions (Olson and Dinerstein, 2000) and forms part of the Chilean Winter Rainfall-Valdivian Forests biodiversity hotspot (Myers et al., 2000). LANDIS-II was used to simulate a range of different types of disturbance, including fire, stem cutting, herbivory and spread of invasive species, both individually and in different combinations. Analyses examined the impact of these different disturbance regimes on for-est structure and composition.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (271)

245

Landscape-scale dynamics and restoration of dryland forest ecosystems

Study area

The investigation was conducted in Quilpue, located in the Vth Region of Chile between 32º56’7’’ and 33º22’49’’S, and 70º59’14’’ and 71º39’53’’W. The study area is located in the coastal mountain range situated south of the city of Santiago, and covers an area of 170,897 ha, with an altitudinal range of 15 to 2129 m a.s.l. The entire study area has recently been proposed as a UNESCO Biosphere Reserve, but two national protected areas that have al-ready been designated lie within its boundary, namely the Lago Peñuelas National Reserve (LP) and La Campana National Park (LC). The climate of the area is characterized by hot dry summers and cool wet winters with strong inter-annual variability owing to the El Niño – Southern Oscillation (ENSO) phenomenon. Mean annual temperature is 13.2°C and mean annual precipitation is 531 mm.

Vegetation of the area is characterized by a spatially heterogeneous mosaic of different types of dry forest, including some xerophytic plant species such as Cactus spp. Lower slopes are dominated by shrubland dominated by Acacia caven, whereas evergreen scle-rophyllous forest occurs mostly on south-facing slopes and in sheltered ravines, where dominant tree species include Cryptocarya alba, Lithraea caustica, Peumus boldus and Schinus latifolia. Rapid population growth within the region in recent decades has led to profound changes in patterns of land use (Schultz et al., 2010), resulting from logging, agricultural intensification (particularly vineyard and fruit cultivation, as well as corn and wheat cropping), and establishment of exotic tree plantations (principally Eucalyptus globulus). Grazing of livestock is also widespread in the study area. Native forests continue to be an important source of fuelwood and other non-timber forest products (NTFPs) for local communities.

Field survey

Forest structure and composition was assessed in 50 field plots, following the methods described by Newton (2007a). The plots each measured 25 x 20 m and were distributed throughout the study area using a stratified random approach, with strata defined accord-ing to five classes of aspect (north-, south-, east- and west-facing, and flat) and five classes of elevation (0–400 m, 4–800 m, 8–1200 m and >1200 m). Each individual tree was identified and its diameter at breast height (dbh) measured using a measuring tape. Wood samples of a representative subset of trees of each species were taken using an increment borer, for subsequent determination of age-diameter relationships in the laboratory.

Model parameterization and calibration

The inputs required by the model include a landtype or ecoregion map, which describes the ecological conditions influencing tree establishment, and an initial communities map, which describes the distribution and age of cohorts of each species at year 0. A land-cover map was produced based on classification of an unprocessed Landsat TM image (path 233, rows 83 and 84) acquired for November 2008 (Schulz et al., 2010). The land-cover map produced featured eight land-cover classes (native forest, shrubland, cropland, urban, bare ground, water, pasture, plantation). The spatial resolution of the data was 90 m. All maps were produced and manipu-lated using ArcGIS 9.2 (© 1999–2006 ESRI Inc., Redlands, California) and Idrisi Andes (Clark Labs, Clark University, Worcester MA, USA), projected using WGS 1984 UTM Zone 19S.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (272)

246

Principles and Practice of Forest Landscape Restoration

A total of 20 ecoregions were defined according to the strata used in the field sur-vey (see above), and mapped using an overlay of the land-cover map and a digital el-evation model. The initial forest age structure and composition of each ecoregion was determined using the data obtained from the individual field plots. The establishment probability of each species in each ecoregion was also defined by referring to the field survey data. Those species found to be present in a particular ecoregion were assigned an establishment probability of 1, for the timestep of the simulation (10 years). Those species not present in a particular ecoregion, but present in another elevation class in the same aspect class, were accorded an establishment probability of 0.8. Lower prob-abilities were assigned to all species in north-facing ecoregions, and at higher elevations, reflecting the higher drought risk of such sites. In this way, the model was calibrated to simulate the influence of aspect and elevation on vegetation composition, to reflect the patterns observed in the field (Schulz et al., 2010).

The life history characteristics of the 21 native species encountered in the field survey plus A. dealbata were obtained from field observations, the scientific literature and a range of local experts (Table 8.4). In LANDIS-II, forest succession interacts with several spatial components (i.e. seed dispersal, fire and harvesting disturbances). In this investigation, the LANDIS-II Base Fire v2.1 extension was used to explore fire dynamics. LANDIS-II also requires a fire ecoregion map as an input. Analysis of spatial data for recent (1985–2007) fires in the area indicated that proximity to urban areas or roads is a major factor influenc-ing ignition probability. Consequently, a map was produced by buffering around each road and urban infrastructure, according to four distance classes (0–1 km, 1–4 km, 4–8 km and >8 km). This map was then overlaid on the land-cover map, enabling 30 fire ecoregions to be defined, representing different combinations of land-cover type and buffer distance classes. Fire characteristics for each fire ecoregion were derived from the spatial database of fire history.

The harvest module of LANDIS-II (Base Harvest extension v1.2) was used to simulate the impacts of stem cutting and browsing animals. Both of these disturbances were modelled by the removal of specific cohorts, with browsing removing any cohorts <10 years old, and cut-ting removing all cohorts except the youngest. Harvesting impacts were distributed accord-ing to the ecoregion map, with each ecoregion considered as a separate stand type, nested within an individual management area. The management areas were defined according to the core and buffer areas of the two reserves (LP and LC), and the remaining part of the study area, giving five areas in total.

Scenarios

Once model calibration was completed, final values of the model parameters were used in a series of modelled scenarios. LANDIS simulations were conducted for 200 years. In each case, trees were prevented from establishing on water, cultivated land, forest plan-tations and urban land-cover types. Five replicated simulations (with varying random number seed) were performed for each disturbance scenario to explore the variability of model predictions. The time steps were set at 10 years for tree succession, 10 years for fire disturbance and 1 year for grazing. The list of ages for each species was therefore grouped into age cohorts as follows: ages 1 to 10 years (10), 11 to 20 years (20), 21 to 30 years (30), etc.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (273)

247

Landscape-scale dynamics and restoration of dryland forest ecosystems

Seven different disturbance scenarios were simulated, all of which included the ex-otic species A. dealbata, except where stated: (1) no disturbance (excluding A. deal-bata); (2) no disturbance (but including A. dealbata); (3) fire only; (4) fire and brows-ing, where browsing was applied randomly to 10% of the entire study area, including the protected areas; (5) fire and cutting, where all except the youngest trees were harvested, with the cutting applied to 0.1% of the protected areas (including both core and buffer areas), and 3% of unprotected areas; (6) fire, browsing and cutting, where cohorts <10 years old of target species were browsed, and cohorts >80 years old of target species (Table 8.4) were cut; both browsing and cutting were applied to 10% of the entire study area, including protected areas; (7) browsing and cutting implemented as for scenario 6, but without fire.

Results

A brief summary of the results is provided here; full details are described by Newton et al. (2010) (see also Box 8.2 for additional scenarios). Forest area increased rapidly during the first four decades in all scenarios, reaching values of around 121,000 ha in each case (Fig. 8.5). In scenario 1, forest area remained fairly stable thereafter. In the other scenarios, total forest area tended to decline slightly over time, particularly in scenario 6 (fire, browsing and cutting). However, the differences recorded between the scenarios in terms of total forest cover were slight, in each case accounting for <3% of the forest area present at the outset. More pronounced differences between scenarios were recorded in forest structure (Fig. 8.5). By the final timestep of the simulations (200 years), forest structure in the absence of distur-bance (scenarios 1 and 2) was dominated by relatively ‘old growth’ forest stands, with 96% of forest area characterized by forest stands with trees >120 years old. All of the scenarios that included fire (scenarios 3, 4, 5 and 6) were characterized by high proportions of relatively young forest stands, each with 37–39% forest cover dominated by stands <40 years old. In scenarios with both browsing and cutting (6 and 7), either with or without fire, relatively old-growth forest stands (>120 years old) were virtually eliminated from the landscape, ac-counting for <11% of the forested area.

Contrasting results were also obtained between the different scenarios in patterns of species richness. In particular, species richness values tended to be lower in those scenarios without disturbance than those with burning, cutting and/or fire. For example, no forest areas were recorded in scenarios 1 and 2 with more than four tree species. In each of the other scenarios, substantial forest areas (>10,000 ha) were recorded with five or more species.

The dynamics of Acacia dealbata abundance differed from any of the native species considered. Cover values tended to increase rapidly in all scenarios during the first 40 years. Thereafter, in the absence of disturbance, the species declined rapidly such that it was largely eliminated from the landscape after 100 years. Declines were also recorded in scenarios 3 and 7, although in both cases values tended to remain stable after ap-proximately 100 years. In the other three scenarios, after a period of stability, values in-creased continuously after 100 years. These results indicate that this introduced species is projected to spread only in the presence of fire when combined with browsing and/or cutting.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (274)

248

Principles and Practice of Forest Landscape Restoration

Figure 8.5

Scenario 1

Scenario 2

Scenario 3

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (275)

249

Landscape-scale dynamics and restoration of dryland forest ecosystems

Scenario 4

Scenario 5

Scenario 6

Figure 8.5

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (276)

250

Principles and Practice of Forest Landscape Restoration

Discussion

Results indicated that anthropogenic disturbance has a major impact on forest structure; when subjected to cutting and browsing, either with or without fire, old-growth forest was largely eliminated from the landscape. These results closely accord with those obtained from field surveys and analyses of remote sensing imagery conducted in the moist forests situated further south in Chile, where multiple types of anthropogenic disturbance are chronic and widespread (Newton, 2007b; Newton et al., 2009a,b; Echeverria et al., 2006; 2007; 2008; Wilson et al., 2005).

The simulations also indicated that the characteristics of the disturbance regime had a major influence on patterns of species richness. In particular, species richness in the study landscape tended to be lower in scenarios with no disturbance than in those featuring fire, burning and/or cutting. These results are consistent with the ‘intermediate disturbance hy-pothesis’, which states that species richness is maximized at intermediate frequencies or intensities of disturbance (Connell, 1978). The hypothesis can be attributed to the role of disturbance in preventing the competitive exclusion of species.

The interactions between disturbance regime and ecological characteristics are most strikingly illustrated by the results obtained for Acacia dealbata in these simulations. In the absence of disturbance, this introduced species was largely eliminated from the landscape within a century, presumably as a result of competitive exclusion by native species. In the presence of disturbance, however, the species was able to maintain itself, or even spread through the landscape. This highlights the critical importance of effective management of the disturbance regime if invasive, alien species such as this are to be effectively controlled.

These results have a number of implications for conservation planning and management in this landscape. Currently, two national protected areas have been designated within the study area (LP and LC), and the entire area has recently been proposed as a UNESCO Bio-sphere Reserve. In this context, key planning challenges relate to how ecological restoration might be managed in different parts of the landscape, for example to increase connectivity between the two protected areas (see Box 8.2). While passive restoration might be encour-

Scenario 7

Figure 8.5

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (277)

251

Landscape-scale dynamics and restoration of dryland forest ecosystems

aged by reduction of fire, cutting and browsing, the simulations presented here highlight the potential risks of preventing disturbance altogether, which might lead to a decline in species richness. However, if anthropogenic disturbance is to be tolerated or even encouraged in parts of this landscape to maintain species richness, then this will need to consider the po-tential risks of encouraging the spread of the invasive alien species, Acacia dealbata. Such potential conflicts highlight the need to explicitly consider tradeoffs during the management process, which could potentially be explored using the modelling approach described here.

Table 8.4

Name Long Mat ShT FiT EffSD MaxSD VRP Min VRP

Max VRP

P-FiR Dist

Acacia caven 150 20 1 3 1 150 1 10 80 re-sprout B, C

Acacia dealbata 35 5 2 1 2 2000 1 10 35 re-sprout

Aextoxicon punctatum 260 20 3 2 3 1000 1 10 200 re-sprout B

Azara celastrina 100 20 3 1 10 100 0 0 0 none

Azara dentata 100 20 3 1 10 100 1 10 80 re-sprout

Beilschmiedia miersii 300 30 5 3 2 50 1 10 180 re-sprout B

Cestrum parqui 50 25 2 3 2 50 0 0 0 none B

Citronella mucronata 200 15 3 3 5 25 0 0 0 none

Crinodendron patagua 180 20 3 1 4 100 0 0 0 none B

Cryptocarya alba 150 20 2 4 8 1000 1 10 100 re-sprout B, C

Drimys winteri 250 10 5 2 4 25 1 10 180 re-sprout

Ephedra chilensis 100 20 2 3 8 100 1 10 80 re-sprout

Lithraea caustica 200 20 3 3 8 1000 1 10 170 re-sprout B, C

Lomatia hirsuta 100 20 4 2 5 100 1 10 70 re-sprout

Luma apiculata 200 5 4 1 4 100 0 0 0 none

Maytenus boaria 120 5 3 3 4 1000 1 10 80 re-sprout B

Persea lingue 250 25 4 2 4 1000 1 10 200 re-sprout

Peumus boldus 250 30 5 2 8 1000 1 10 200 re-sprout B, C

Quillaja saponaria 200 25 4 4 4 100 1 10 180 re-sprout B, C

Schinus latifolia 150 10 2 3 8 1000 1 10 100 re-sprout B, C

Senna candolleana 150 15 1 1 1 150 1 10 100 re-sprout B

Trevoa trinervis 150 20 3 1 4 50 1 10 90 re-sprout

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (278)

252

Principles and Practice of Forest Landscape Restoration

Box 8.2 Effects of fire on sclerophyllous forests in the Biosphere Reserve La Campana-Peñuelas in central Chile

A. Miranda, C. Echeverría, G. Bolados, E. Cantarello, A.C. Newton

Fire is one the main disturbance agents of the sclerophyllous forests in central Chile (Armesto and Gutierrez, 1978; Holmgren et al., 2000). Although some native plants can regrow after fire (Araya and Avila, 1981; Gómez-Gonzalez, 2009), the high frequency of fire in some areas has led to a considerable loss of forest (Miethke, 1993; Schulz et al., in press). The Valparaíso region concentrates the highest fire frequency in Chile. On average, 962 fires destroy 8600 ha per year (CONAF, 2009). Fire occurrence is closely related to human activities such as urban expansion and clearance for agriculture and livestock. The recently created Biosphere Reserve La Campana-Peñuelas aims to recover degraded zones by fire and livestock and convert commercial plantations of exotic species to native forests (CONAF, 2008).

In this study we assessed the effect of fire on the dynamics of the sclerophyllous forest in the Biosphere Reserve using LANDIS-II, base fire v2.1 (Sheller and Domingo, 2009). This version is able to simulate the spatial distribution of fire considering the ignition probability, fuel type, and fire rotation interval among others. Initial community maps included ‘espinales’, an anthropogenic pseudo-savannah, and sclerophyllous forest formed by evergreen, hard-leaved tree species. A fire ecoregion map of 12 categories was built based on vegetation types (forest, espinal and grassland) and distance to urban areas (0–1000 m, 1000–4000 m, 4000–8000 m and greater than 8000 m).

The effect of fire on the spatial distribution and composition of tree species was analyzed for the following two scenarios: (i) scenario 1 as usual, in which the fire regime observed over the last 25 years is maintained. In this case, simulation parameters are calibrated for 4500 fires affecting 22,000 ha of native forest (CONAF, 2009); (ii): scenario 2 of landscape management, in which forest plantations of exotic species are replaced by native species that have been successful in forest restoration such us Acacia caven, Maytenus boaria, Schinus latifolius and Quillaja saponaria. Also, this scenario included the restoration in some specific riparian zones in order to enhance forest connectivity and the reduction of fire frequency and burnt area in 50%.

Under scenario 1 all the species analyzed exhibited a gradual decline in their abundance over the next 100 years (Fig. 1). S. latifolius and Cryptocarya alba showed a higher decline in abundance of 43% and 30% respectively, while A. caven and Q. saponaria decreased 5% and 2% respectively. Under scenario 2, the species more resistant to fire reached a higher regeneration rate than the forest loss rate by fire (Fig. 2). For instance, A. caven and Lithrea caustica increased 0.5% and 3% in abundance after 100 years respectively. On the other hand, the more vulnerable species to fire such as S. latifolius and C. alba decreased by 22% and 28% respectively. Although these two last species were favoured in abundance by a reduction in the fire regime parameters, S. latifolius exhibited a greater decrease in the loss of area, from 43% (scenario 1) to 22% (scenario 2).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (279)

253

Landscape-scale dynamics and restoration of dryland forest ecosystems

Figure 1

Figure 2

Box 8.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (280)

254

Principles and Practice of Forest Landscape Restoration

Case study in southern Argentina

This research was undertaken in the forest-steppe ecotone situated on the eastern slopes of the northern Patagonian Andes, Argentina. Climatically, the region is characterized by an abrupt precipitation gradient, which declines from west to east owing to the orographic ef-fect of the Andes. As a result of droughts and the presence of both natural and anthropogenic ignition sources, fire is widespread in the region. Although the impacts of fire on forest stand structure and dynamics have been documented in detail by previous research (Kitzberger, 2002; Kitzberger and Veblen, 1997; Kitzberger et al., 1997, 2001, 2007; Mermoz et al., 2005; Veblen et al., 1999, 2008), the spatial dynamics of fire, and their interactions with other types of disturbance, have not been examined previously. Other forms of human disturbance af-fecting forests in the region include herbivory, as a result of extensive sheep, cattle and goat ranching. The objective of this study was to examine the potential for passive restoration of forest in the study area, under different disturbance regimes. This was addressed by us-ing the LANDIS-II model to explore the following hypotheses, based on previous research undertaken in the area: (i) high fire frequency during aboriginal and European settlement eras has led to increased forest fragmentation and restriction of forest into fire-free rocky refugia, where low fuel loads have permitted the survival of isolated forest patches; (ii) the fire regime prevailing at the present time would permit forest recovery and expansion in the absence of herbivory, which is currently restricting forest distribution to rocky refugia.

Study area

The investigation was conducted in an area located between 40º54’37’’ and 41º15’20’’S, and 70º41’4’’ and 71º24’1’’W, on the eastern slopes of the northern Patagonia Andes. The area is situated between 737 and 2195 m a.s.l., and is 228,289 ha in extent. The western region of the study area comprises the east-central part of the Nahuel Huapi National Park and the remaining land is divided into large privately owned estancias (large farms or ranches). Topographically, from west to east, the area includes the Andean cordillera, the lower foothills of which are intersected by glacial lakes and valleys, and the Patagonian plains at approx. 700 m a.s.l. An abrupt west to east decrease in the amount of rainfall has created a unique forest-steppe ecotone. Within the study area, precipitation declines from approximately 1800 mm

Simulations revealed that the restoration actions along with a reduction in 50% in fire frequency and area caused a decline in the loss rate of some tree species in the Biosphere Reserve. Despite these actions, some species still continued to decrease over the next 100 years. On the other hand, fire-resistant species exhibited a slight increase in abundance in the landscape. Our results also showed that the selection of species used in the restoration areas was suitable to increase forestland in the study area. The species were able to survive and grow in the restoration area and to expand their propagules to the neighbouring sites.

It can be concluded that restoration actions are not the only measure to be implemented for the persistence of dryland forests in the Biosphere Reserve. It is also highly necessary to implement landscape planning and conservation actions that lead to a reduction in the occurrence of fire disturbances in the study area. The absence of this type of actions will probably lead to a gradual decline in some of the principal tree species in the Biosphere Reserve.

Box 8.2 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (281)

255

Landscape-scale dynamics and restoration of dryland forest ecosystems

to less than 500 mm per year across this ecotone. The distribution of precipitation is highly seasonal, with approximately 60% falling during May to August. Mean annual temperature is approximately 8°C and mean monthly temperatures vary from 2–14°C.

Austrocedrus chilensis is the only tree species present in much of the study area, al-though it grows alongside Nothofa*gus pumilio in the wetter western area (above 1000 m a.s.l.) of the ecotone. With increasing aridity the understory becomes less dense as Chusquea is replaced by shrubs and small trees such as Maytenus boaria and Lomatia hirsuta. At the most easterly edge of the study area, A. chilensis stands become sparser un-til the only remaining trees are found associated with rocky outcrops. In these driest areas, trees are replaced by bunch grasses and low-growing shrubs such as Discaria articulata and Mulinum spinosum. Water balance is negative from October to March, creating condi-tions conducive to fire. Fire frequency prior to Caucasian settlement was extremely high, as the Native American population set fires to drive guanaco (Lama guanicoe) into open areas for hunting (Veblen et al., 1999, 2008). Forest stands are now highly fragmented, and are also currently under threat from heavy grazing by livestock (principally cattle and sheep). In addition, introduced herbivores including European hares, rabbits and exotic deer are negatively impacting native arboreal vegetation.

Model parameterization and calibration

LANDIS-II requires parameters to be defined for each tree species included in the model, relating to a range of different ecological characteristics. Here, a single tree species (Aus-trocedrus chilensis) was incorporated in the simulations, reflecting its monodominance of ecotonal forests in the area. Species attributes were derived from the scientific literature, supported by field observations, as follows: maximum longevity, 500 years; age of sexual maturity, 20 years; shade tolerance, 2 (on a scale of 1–5); fire tolerance, 2 (on a scale of 1–5); effective seed dispersal distance, 15 m; maximum seed dispersal distance, 200 m; vegetative reproduction probability, 0; minimum age of vegetative reproduction, 0 years; maximum age of vegetative reproduction, 0 years; post-fire regeneration form, none.

A land-cover map was produced based on classification of an unprocessed Landsat TM image acquired for February 2003. Classification was achieved using field points to train the spectral signature of the selected land-cover classes in a supervised classification scheme, using a maximum likelihood algorithm. The map produced featured the following land-cover classes: wet forest, dry forest (A. chilensis), shrubland, wet grassland, dry grassland, bare ground, exotic plantations, urban, burned areas, and water bodies. The spatial resolution of the data was 28.5 m. In addition, a detailed distribution of Austrocedrus chilensis distribution and rock outcrops was produced by digitizing each outcrop from the high resolution orthophotos. The outcrops were considered as a separate land-cover type in the land-cover map. All maps were produced and manipulated using IDRISI Andes (Clark Labs, Clark University, Worcester MA, USA), projected using National Grid Argentina Faja1. LANDIS-II also requires a fire ecoregion map as an input. This was derived from the land-cover map, by assigning different fire characteristics to each land-cover type, based on available data. In this way, each land-cover type was modelled as a separate fire ecoregion. The harvest module of LANDIS-II (Base Harvest extension v1.2; Gustafson et al., 2000) was used to simulate the impacts of browsing animals. This type of disturbance was modelled by the removal of a specific cohort (<10 years old). A map of potential forest distribution was also produced,

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (282)

256

Principles and Practice of Forest Landscape Restoration

to provide an indication of areas where forest could potentially be restored, and where forest might have been present prior to human disturbance. This map was derived from the analysis of current forest distribution in relation to climatic variables, using Mahalanobis typicalities implemented in IDRISI.

Scenarios

LANDIS simulations were conducted for 300 years. For all scenarios, water bodies, urban areas and wet forest were excluded from the simulations. The time steps were set at 10 years for tree succession, 10 years for fire disturbance and 1 year for grazing. The list of ages for A. chilensis was grouped into age cohorts as follows: ages 1 to 10 years (10), 21 to 30 years (20), 31 to 40 years (30), etc. A range of different scenarios were simulated but six of these are presented here by way of illustration: (1) no disturbance, (2) current fire regime, (3) histori-cal fire regime, (7) harvesting and browsing only, at high intensity, without fire, (8) harvesting and browsing at low intensity, with current fire regime, (9) browsing at low intensity with historical fire regime.

Results

In all scenarios, forest cover increased continually over time (Figs. 8.6, 8.7). The rate of increase in forest area was highest in scenario 1 (no disturbance), and lowest under scenario 9 (browsing and historical fire regime). Forest cover also increased relatively rapidly under the current fire regime, without grazing (scenario 2), but less rapidly under the historical fire regime (scenario 3). These results highlight the additive effects of browsing and fire in reducing the potential for increase in forest area, and also illustrate the potential impacts of different fire regimes on the rate and extent of forest recovery.

Discussion

The preliminary results presented here for this study area highlight the potential for forest recovery in southern Argentina, if the disturbance regime was managed appropriately. In all of the scenarios explored, an increase in forest area over time was observed, indicating that forest restoration might be achieved even under disturbance regimes featuring fire and browsing. However, rates of increase in forest area were found to be highest in situa-tions without any form of disturbance. In each case, forest expanded from the remaining forest fragments, but at a rather low rate. The projections presented here suggest that many centuries would be required to re-establish forest cover throughout its former distribu-tion, prior to human settlement. For this reason, practical restoration efforts may need to consider active restoration approaches (involving establishment of plantations of native tree species) in addition to passive approaches. Potentially, modelling approaches such as that presented here could be used to identify those locations within a landscape where passive restoration is likely to be effective or not over a given timescale. Locations where passive restoration is less likely to occur might usefully be targeted for active restoration. In addition, many target areas where passive restoration is feasible are currently being transformed into exotic pine plantations (Chapter 2). Modelling approaches such as the one presented here allow for a fine-tuned identification of areas of potential conflict be-tween restoration of native dryland forests and other land uses.

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (283)

257

Landscape-scale dynamics and restoration of dryland forest ecosystems

Figure 8.6

Scenario 1

Figure 8.7

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (284)

258

Principles and Practice of Forest Landscape Restoration

Scenario 2

Scenario 3

Figure 8.7

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (285)

259

Landscape-scale dynamics and restoration of dryland forest ecosystems

Scenario 7

Scenario 8

Figure 8.7

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (286)

260

Principles and Practice of Forest Landscape Restoration

Scenario 9

Discussion

As noted by Bugmann (2001), the value of forest models does not necessarily lie in their ability to make accurate predictions, but in their ability to help understanding of processes and patterns by allowing exploration of the consequences of a set of explicitly stated as-sumptions. The outputs can be viewed as hypotheses that require further testing. Although ecological models have been employed to explore forest dynamics in a wide range of differ-ent locations, they have rarely been used to support forest restoration planning and practice (Newton, 2007a). This partly reflects a traditional focus on modelling the dynamics of indi-vidual forest stands rather than landscapes. It is now widely appreciated that understanding the spatial dynamics of forested landscapes is essential to their effective management and conservation (Newton, 2007b; Gardner et al., 2009; Lindenmeyer and Franklin, 2002). In this context, recent progress in developing spatially explicit forest modelling approaches is of particular value, enabling spatial processes such as dispersal and colonization to be incorporated. As illustrated here, the LANDIS-II model provides a valuable tool for exploring the dynamic patterns of forested landscapes under different disturbance regimes, and can provide insights into the potential for restoration of forest landscapes. A further advantage of using a modelling approach such as LANDIS-II is that it can help identify knowledge gaps, and provide a framework for focusing future research efforts.

One of the main limitations common to all process-based models of forest dynamics, including LANDIS-II, is the difficulty of obtaining rigorous model validation, owing to the general lack of long-term data describing the ecological behaviour of forests (Newton, 2007a; Shugart, 1984; Shugart and West, 1980). The current examples are no exception; relatively little ecological research has been undertaken previously in the forests described here. This lack of information also hinders model parameterization. In particular, information is lacking on the dispersal ability of different tree species, a process that has a major influence on model

Figure 8.7

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (287)

261

Landscape-scale dynamics and restoration of dryland forest ecosystems

outputs. Other areas of uncertainty include the precise impacts of herbivory and fire on the mortality of different age cohorts of trees, and variation in the establishment probability of different tree species across the landscape. As a consequence, the results presented should be viewed as highly tentative. However, as demonstrated here, the use of high resolution remote sensing imagery can be of value providing detailed information about forest structure and distribution, especially in areas that are relatively inaccessible (for example in Chiapas). The integration of such imagery with spatial modelling approaches offers a powerful set of tools with which to explore the sustainable management of forest landscapes (Newton et al., 2009a,b). Further elements of an emerging toolkit could potentially include models that enable the potential impacts of climate change to be explored (Boxes 8.3 and 8.4).

Conclusions

The results presented here highlight the value of spatially explicit modelling tools for exploring the potential for restoration of forest landscapes. Specifically, the modelling approach used (LANDIS-II) enabled projections to be made regarding the pattern of regeneration and spread of native forest under different anthropogenic disturbance regimes, providing an insight into the potential for passive restoration approaches. The model also highlighted interactions between different forms of disturbance and their impacts on restoration processes, an area in which information is currently lacking. For example, modelling scenarios conducted in Chile indicated how spread of the invasive exotic species Acacia dealbata is dependent on other forms of disturbance such as grazing and fire. These examples demonstrate how spatial models can inform approaches to forest landscape restoration, by indicating those locations within a landscape where particular restoration approaches are most likely to be successful. In addition, spatially explicit modelling tools provide a means of visualizing the potential impact of restoration actions at the scale of entire landscapes.

Box 8.3 Effects of climate change on subtropical forests of South America

S. Pacheco, L. R. Malizia, L. Cayuela

Premontane forests in northern Argentina and southern Bolivia represent the lowest vegetation belt of the Yungas or subtropical montane forests. They are a conservation priority as they play a key ecological role (Brown and Malizia, 2004; Brown, 2009). These forests have been subjected to a long history of use (Malizia et al., 2009; Brown et al., 2006; Fundación ProYungas, 2007) and climatic variation (Prado, 1995). The objective of this research was to determine the future distribution of the premontane forest as a response to climate change, and to analyze the consequences of this distributional shift for its conservation and restoration.

To determine changes in forest distribution we developed statistical models (Scout et al., 2002) using Maxent (Phillips et al., 2006; Phyllips and Dudik, 2008). Location points indicating forest presence were obtained from: (1) the Subtropical Network of Permanent Plots (Fundación ProYungas, 2007; Blundo and Malizia, 2009), and (2) rapid assessment of tree species using 0.1 ha circular plots. The current and future variables of the CCM3 scenario, for the end of the century (Govindasamy et al., 2003) were obtained from the WorldClim database (Hijmans, 2005).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (288)

262

Principles and Practice of Forest Landscape Restoration

The variables used were: mean annual precipitation, precipitation of driest month, temperature seasonality, and maximum temperature of warmest month. To obtain more accurate models we calibrated the mean annual precipitation data with Bianchi’s local precipitation model (Bianchi et al., 2008).

Once fitted, present and future distribution models were overlaid to define stable forest areas and areas of potential change (expansion and retraction). We defined as stable forest areas those which currently correspond to premontane forests and are likely to continue as forests in the future. The expansion areas are those that do not correspond to premontane forests at the present time but, as a consequence of climate change, could be colonized by species typical of this environment. Retraction areas are those areas of premontane forest that will shift to other potential vegetation types in the future.

The results obtained indicate that according to the internal validation executed by Maxent, the current distribution model of premontane forests has a high overall accuracy (AUC 0.95 ± 0.013). Ten percent of the original area of premontane forest is now transformed into agricultural land, mainly concentrated in flat areas below a 5% slope. The model of future distribution predicts a 53% decrease in cover as compared to the current area occupied by premontane forest. Of the future potential area occupied by this forest, approximately 30% is represented by the original cover and the remaining corresponds to areas of potential expansion owing to suitable changes in climatic conditions. The future climate scenario shows increases of ca. 1 ºC in temperature and 80 mm in mean annual precipitation. According to our models, such changes will trigger an upwards altitudinal shift of premontane forests of about 300 m a.s.l. in the mountains. The largest retraction of the current forest area will occur in the northeastern range of its distribution, whereas stable areas will be concentrated mostly along the western sector and the southeastern range of its distribution. The latter is also a potential expansion area under climate change (Fig. 1).

Currently, 8% of the premontane forest is included in the protected areas of Acambuco, Piarfon, Pintascayo, Pizarro (Salta), and Calilegua (Jujuy) (Fig. 1). Under a climate change scenario, only 50% of this surface will remain as premontane forest. Premontane forests are projected to disappear in Acambuco, Piarfon and Pintascayo, and remain in Calilegua, Pizarro, Lancitas (Jujuy), and the Biosphere Reserve of the Yungas (Fig. 1).

To the best of our knowledge, our study is the first to evaluate the potential effects of climate change on premontane forest distribution in Argentina. Our results show that changes in climate conditions will markedly affect the distribution of premontane forest. A decrease in the area occupied and an upward migration of premontane forests is likely to occur in response to climate change, mostly as a result of an increase in temperature. Areas of potential expansion are currently covered by montane forest, which constitutes the immediate vegetation belt above premontane forests. The stable areas are mainly located in the western range of its distribution, which include the Yungas Biosphere Reserve. The northeastern range of its distribution, on the contrary, is predicted to suffer a sharp contraction. In relation to these shifts, protected areas located in the west and southeast of the study region are more likely to preserve premontane areas in the future, whereas the northeastern protected areas are predicted to lose their premontane forest in the future. The latter areas are likely to be colonized by species from the chaco forest and by premontane species tolerant of warmer weather conditions. Owing to the spatial and functional connection between the Chaco forest and the premontane forest, the current system of protected areas would probably help maintain their connection.

Box 8.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (289)

263

Landscape-scale dynamics and restoration of dryland forest ecosystems

Figure 1

Box 8.3 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (290)

264

Principles and Practice of Forest Landscape Restoration

Box 8.4 Effects of climate change on dryland ecosystems of central Chile

G. Henriquez Tapia, S. Pacheco, C. Echeverría

Worldwide, dryland ecosystems have been rapidly degrading. Dryland systems include all subhumid, arid, semi-arid and hyper-arid regions of the world where water is the main limiting factor of primary production and, consequently, other ecosystem services and human well being. Drylands occupy about 41% of the Earth’s land area, comprising a high biodiversity, and providing a home to about two billion people. The persistence of drylands as a consequence of overexploitation is highly uncertain, especially now climate changes make predictions on future ecosystem functioning highly speculative. Generally, climate changes are thought to be accompanied by stronger and more frequent weather extremes (Easterling et al., 2009; Meehl et al., 2000). Among the species associated with this forest type are Acacia caven, Beilschmiedia miersii, Colliguaja odonifera, Crinodendron patagua, Cryptocarya alba, Drimys winteri, caustic Lithraea, Maytenus boaria, Myrceugenia exsucca, Peumus boldus and Quillaja saponaria (Donoso, 1995). Central Chile has been considered as one of the 25 high-priority areas of biodiversity (Hechenleintner et al., 2005) and one of the 34 biodiversity hotspots at a global scale.

In this work we modelled the current and future potential distribution of three tree species of the dryland forest in central Chile. We compared changes in their distributions using a scenario of climate change and the software MAXENT. All the input data were generated into a GIS using ARC GIS 9.3. The three species selected were Acacia caven, Quillaja saponaria and Cryptocarya alba. These species are socially important for local people as they are regularly used as a non-timber forest product (NTFP). Also, these species are present in one of the parts in Chile that will be severely impacted by climate change. Data of occurrence were obtained from the National vegetation mapping in Chile (known as Catastro de Bosque nativo, CONAF et al., 1999).

Climatic variables were obtained from the regional pattern PRECIS (Regional Providing Climates for Impact Studies), which was developed by the Hadley Centre of the United Kingdom and has a horizontal resolution of 25 km². PRECIS includes (i) current climatic variables for 1961–1990, and (ii) two models of future climate for the period 2071–2100: one moderate considering the scenario MESSRS B2 of the IPCC and another severe considering the scenario MESSRS A2 of the IPCC. In our work, we used the models of potential distribution and that of climatic change MESSRS B2.

The models of potential distribution, for the three species, presented a high general efficiency. The value of AUC for the pattern developed in the current scenario was 0.904 (0.022), and for the scenario B2 was 0.917 (0.019). They can therefore be considered as good current models of distribution to generate hypothesis for change in scenarios of climatic variation.

The models generated for the potential distribution of the three species are very similar to the current distribution of these species. Although the climate variables used in this work corresponded to earlier years (1961–1990) than those used for the occurrence of the species (1994–1997), the models of potential distribution derived from the climate variables appear to be suitable.

Our results revealed that in a future scenario, changes would in the distributions of the modelled species. In A. caven a reduction in the total area was observed without changes in the distributional range. In the case of Q. saponaria, projections indicate a reduction in the northern part of its range and for C. alba, a decline is projected in the northern and southern parts of its distribution (see Figs. 1–3).

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (291)

265

Landscape-scale dynamics and restoration of dryland forest ecosystems

Figure 1. Quillaja saponaria

Figure 2 Acacia caven

Box 8.4 (cont.)

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (292)

266

Principles and Practice of Forest Landscape Restoration

Figure 3 Criptocarya alba

Box 8.4 (cont.)

Austrocedrus chilensis forest towards the western edge of its distribution in Nahuel Huapi, Argentina. Photo: A.C. Newton

Principles and Practice of Forest Landscape Restoration· Principles and Practice of Forest Landscape Restoration Case studies from the drylands of Latin America. About IUCN ... - [PDF Document] (293)

267

Landscape-scale dynamics and restoration of dryland forest ecosystems

References

Aguilar-Jiménez, J.R. 2008. Análisis de los sistemas de producción bovina en la Cuenca del Río El Tablón, en la zona de amortiguamiento de la Reserva de la Biosfera La Sepultura, Villaflores, Chiapas. MSc. thesis. Facultad de Medicina Veterinaria y Zootecnia, Universidad Autónoma de Chiapas, Tuxtla Gutiérrez, Chiapas.

Araujo, M.B., Rahbek, C. 2006. How does climate change affect biodiversity? Science 313 (5792); 1396–1397.

Araya, S. Ávila, G. 1981. Rebrote de arbustos afectados por el fuego en el “Matorral chileno”. Anales Museo de Historia Natural Valparaíso 14: 107–113.

Armesto, J.J., Gutierrez, J. 1978. El efecto del fuego en la estructura de la vegetación de Chile central. Anales Museo de Historia Natural Valparaíso 11: 43–48.

Bianchi, A.R., Elena, H., Volante, J. 2008. SIG climático del NOA. INTA-Salta.

Birch, J.C., Newton, A.C., Aquino, C.A., Cantarello, E., Echeverría, C., Kitzberger, T., Schiappac-asse, I., Garavito, N.T. 2010. Cost-effectiveness of dryland forest restoration evaluated by spatial analysis of ecosystem services. PNAS 107: 21925–21930.

Blundo, C., Malizia, L. 2009. Impacto del aprovechamiento forestal en la estructura y diversi-dad de la selva pedemontana. In: Brown, A.D., Blendinger, P.G., Lomáscolo, T., García Bes, P. (eds.), Selva pedemontana de las Yungas, historia natural ecología y manejo de un ecosis-tema en peligro.ProYungas, Argentina: pp. 387–406.

Brown, A.D. 2009. Las Selvas Pedemontanas de las Yungas: manejo sustentable y conser-vación de la biodiversidad de un ecosistema prioritario del noroeste argentino. In: Brown, A.D., Blendinger, P.G., Lomáscolo, T., García Bes, P. (eds.), Selva pedemontana de las Yungas, historia natural ecología y manejo de un ecosistema en peligro: pp. 13–36.

Brown, A.D., Malizia, L., 2004. Las Selvas Pedemontanas de las Yungas: en el umbral de la ex-tinción. Ciencia Hoy 14 (83), 52–63.

Brown, A.D., Pacheco, S., Lomáscolo, T., Malizia, L. 2006. Situación ambiental en los bosques andinos yungueños. In: Brown, A., Martinez Ortiz, U., Acerbi, M., Corcuera, J. (eds), La Sit-uación Ambiental Argentina 2005. Fundación Vida Silvestre Argentina: pp. 53–71.

Bugmann, H. (2001). A review of forest gap models. Climatic Change, 51(3–4): 259–305.

Cantarello, E., Newton, A.C., Hill, R.A., Tejedor-Garavito, N., Williams-Linera, G., López-Barrera, F., Manson, R.H., Golicher, D.J. 2011. Simulating the potential for ecological restoration of dryland forests in Mexico under different disturbance regimes. Ecological Modelling 222: 1112–1128.

Challenger, A., Dirzo, R. 2009. Tendencias de cambio de la biodiversidad. In: Sarukhán, J. (ed.), Capital natural de México, vol. II: Estado de conservación y tendencias de cambio. Comisión Nacional para el Conocimiento y Uso de la Biodiversidad. México. D.F.: pp. 37–73.

CONABIO. 2006. Grado de marginación a nivel localidad, 2000. Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO), México.

CONAF, CONAMA, BIRF, Universidad Austral de Chile, Pontificia Universidad Católica de Chile, Universidad Católica de Temuco. 1999. Catastro y Evaluación de los Recursos Veg-etacionales Nativos de Chile. Informe Nacional con Variables Ambientales. Santiago, Chile.

CONAF. 2008. Documento Base Reserva de la Biosfera “La Campana – Peñuelas” (Propuesta de Ampliación). Corporación Nacional Forestal. 188pp.