A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Eric Dinerstein David M. Olson Douglas ). Graham Avis L. Webster Steven A. Primm Marnie P. Bookbinder George Ledec r Published in association with The World Wildlife Fund The World Bank WWF Washington, D.C. A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Eric Dinerstein David M. Olson Douglas J. Graham Avis L. Webster Steven A. Primm Marnie P. Bookbinder George Ledec Published in association with The World Wildlife Fund The World Bank Washington, D.C. © 1995 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W., Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing September 1995 The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organiza- tions, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list and indexes of subjects, authors, and coun- tries and regions. The latest edition is available free of charge from Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publica- tions, The World Bank, 66 avenue d'I6na, 75116 Paris, France. ISBN 0-8213-3295-3 Library of Congress Cataloging-in-Publication Data A conservation assessment of the terrestrial ecoregions of Latin America and the Caribbean / Eric Dinerstein . .. [et al.]. p. cm. Includes bibliographical references (p. ). ISBN 0-8213-3295-3 1. Biological diversity conservation-Latin America-Evaluation. 2. Biological diversity conservation-Caribbean Area-Evaluation. 3. Biotic communities-Latin America-Evaluation. 4. Biotic communities-Caribbean Area-Evaluation. 5. Ecology-Latin America- Evaluation. 6. Ecology-Caribbean Area-Evaluation. I. Dinerstein, Eric, 1952- QH77.L25C66 1995 333.9516'098-dc2O 95-227 CIP Contents Foreword viii Acknowledgments ix Authors xi Contributors xii Acronyms and Abbreviations xv Executive Summary xvi Introduction 1 1 Approach 4 Fundamental Goals Underlying the Approach 4 Snapshot Conservation Status 7 Final Conservation Status 8 Biological Distinctiveness 8 Biodiversity Conservation Priority 10 2 Major Ecosystem Types, Major Habitat Types, and Ecoregions of LAC 12 Major Ecosystem Types (METs) 12 Major Habitat Types (MHTs) 12 Ecoregions 14 3 Conservation Status of Terrestrial Ecoregions of LAC 16 Results 17 4 Biological Distinctiveness of Terrestrial Ecoregions of LAC at Different Biogeographic Scales 20 Results 20 5 Integrating Biological Distinctiveness and Conservation Status 22 Results 22 Major Trends 24 6 Conservation Assessment of Mangrove Ecosystems 36 Definition 36 Ecological Attributes of Mangrove Ecosystems 36 Delineation of Mangrove Complexes and Units 37 Conservation Status 39 Biological Distinctiveness and Conservation Activities 39 7 Conclusions and Recommendations 42 Comparisons with Other Priority-Setting Frameworks for LAC 46 Application of the Methodology to Finer Geographic Scales 46 V vi A Consemation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Appendixes A Methods Used for Assessing the Conservation Status of Terrestrial Ecoregions 49 B Methods Used for Assessing the Conservation Status of Mangrove Units 59 C Definitions of Major Ecosystem Types and Major Habitat Types 63 D Hierarchical Classification Scheme of LAC Ecoregions 66 E Results of Assessments of Landscape-Level Criteria, Conservation Status, and Biological Distinctiveness of Non-Mangrove Ecoregions 71 F Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 84 G Sources for Remaining Natural Habitat and Protected Area Assessments 117 Glossary 123 References 127 Figures 1-1 Analytical Steps Used to Derive Biodiversity Conservation Priorities 5 1-2 Hierarchical Classification Scheme of METs, MHTs, Ecoregions, and Bioregions 6 2-1 Number of Ecoregions by Size Categories 14 Tables 1-1 Matrix for Integrating Biological Distinctiveness and Conservation Status to Assign Priorities for Biodiversity Conservation 11 2-1 Important Attributes of the Major Habitat Types of Latin America and the Caribbean 13 3-1 Snapshot Conservation Status of Ecoregions by Major Habitat Type 17 3-2 Final Conservation Status of Ecoregions by Major Habitat Type 18 3-3 Final Conservation Status by Size of Ecoregion 18 3-4 Final Conservation Status by Bioregion 19 4-1 Biological Distinctiveness of Ecoregions by Major Habitat Type 21 5-1 Conservation Importance of Ecoregions by Major Habitat Type 23 5-2 Final Conservation Status and Biological Distinctiveness of All Non-Mangrove Ecoregions 24 5-3 Tropical Moist Broadleaf Forests: Integration Matrix of Biological Distinctiveness and Conservation Status 25 5-4 Tropical Dry Broadleaf Forests: Integration Matrix of Biological Distinctiveness and Conservation Status 27 5-5 Temperate Forests: Integration Matrix of Biological Distinctiveness and Conservation Status 28 5-6 Tropical and Subtropical Coniferous Forests: Integration Matrix of Biological Distinctiveness and Conservation Status 29 5-7 Grasslands, Savannas, and Shrublands: Integration Matrix of Biological Distinctiveness and Conservation Status 30 5-8 Flooded Grasslands: Integration Matrix of Biological Distinctiveness and Conservation Status 31 5-9 Montane Grasslands: Integration Matrix of Biological Distinctiveness and Conservation Status 32 Contents vii 5-10 Mediterranean Scrub: Integration Matrix of Biological Distinctiveness and Conservation Status 33 5-11 Deserts and Xeric Shrublands: Integration Matrix of Biological Distinctiveness and Conservation Status 34 5-12 Restingas: Integration Matrix of Biological Distinctiveness and Conservation Status 35 6-1 Conservation Status of Mangrove Units of LAC 38 6-2 Most Appropriate Conservation Activities for Mangrove Units of LAC 40 7-1 Ecoregions of Highest Priority at Regional Scale by Bioregion and Major Habitat Type 43 A-1 Habitat Block Analysis for Tropical Broadleaf Forest MET 54 A-2 Habitat Block Analysis for Conifer/Temperate Broadleaf Forest MET 54 A-3 Habitat Block Analysis for Grassland/Savanna/Shrubland MET 54 A-4 Habitat Block Analysis for Xeric Formation MET 54 A-5 Degree of Protection Analysis for Broadleaf and Conifer Forest METs 56 A-6 Degree of Protection Analysis for Grassland/Savanna/Shrubland and Xeric Formation METs 56 A-7 Degree of Protection Analysis Suggested for Large Ecoregions 56 B-1 Habitat Block Analysis for Mangrove Units 60 B-2 Degree of Protection Analysis for Mangrove Units 60 Maps 1 Bioregions of Latin America and the Caribbean 2a Major Habitat Types of Mexico and Central America 2b Major Habitat Types of the Caribbean 2c Major Habitat Types of South America 3 Ecoregions of Latin America and the Caribbean 4 Mangrove Complexes and Units of Latin America and the Caribbean 5 Snapshot Conservation Status of Ecoregions of Latin America and the Caribbean 6 Final Conservation Status of Ecoregions of Latin America and the Caribbean (Snapshot Conservation Status Modified by Threat) 7 Biological Distinctiveness of Ecoregions of Latin America and the Caribbean 8 Biodiversity Conservation Priority of Ecoregions of Latin America and the Caribbean 9 Biodiversity Conservation Priority of Ecoregions of Latin America and the Caribbean (Incorporating Consideration of Bioregional Representation) Large-format map (insert): Ecoregions of Latin America and the Caribbean Foreword As leading financiers of biodiversity conservation in savannas, shrublands, and drylands are among the Latin America and the Caribbean, the World Bank highest priorities for conservation action in Latin and the Global Environment Facility (GEF) have a America and the Caribbean. responsibility to target conservation funds to the The information and orientation this report pro- areas of greatest concern and need. The present vides will support Bank operations that involve the study represents a step toward this goal by helping protection or management of natural habitats. For us to understand more clearly the conservation the GEF, the results of this study will be useful at the status of the terrestrial ecoregions of Latin America regional planning level. and the Caribbean. This study is also noteworthy because of the collab- Donors and conservation planners have tradi- oration between the Bank and the World Wildlife tionally focused their attention on the lowland Fund, a leading nongovernmental organization for rainforests, extremely species-rich habitats. The conservation. The convergence of priorities between present study emphasizes that many other habitat our two organizations suggests an evolving collabo- types are also very important for biodiversity and ration on conservation issues. I extend my thanks to merit close attention in conservation strategies. the World Wildlife Fund for contributing its exper- Tropical dry broadleaf forests, montane forests, tise and for working with Bank specialists to make coniferous forests, temperate forests, grasslands, this report possible. Ismail Serageldin Vice President Environmentally Sustainable Development The World Bank viii Acknowledgments This study has benefited from the contributions of assisted with data research. Eric Dinerstein was many biodiversity specialists, representing an array supported by an Armand G. Erph Conservation of academic institutions, conservation organizations, Fellowship. and government agencies from Latin America and We would also like to thank the Biodiversity Sup- the Caribbean (LAC) or with strong interests in the port Program (BSP), supported by the United States region. These individuals and the organizations they Agency for International Development (USAID), for represent are listed in the Contributors section. indirectly facilitating our work through their funding Within LATEN, the World Bank's Environment of a LAC conservation priority-setting exercise. The Unit for Latin America and the Caribbean, we single BSP exercise included a major workshop, attended out the role of Michel de Nevers, who originally by over 60 regional biodiversity specialists, which proposed the idea for this study. Dean Gibson, provided a significant amount of data for this study. Guillermo Paz y Mifio, and Manuel Bonifaz all Some groups helped by supplying extensive spa- worked as LATEN research assistants on this proj- tial databases at no cost to this project. We are grate- ect. Dennis Mahar and later William Partridge su- ful for their generosity and their commitment to pervised this project. Peter Brandriss prepared the sharing conservation data. We acknowledge Chuck report for publication. Kathy Dahl and Tracey Smith Carr, John Robinson, the Wildlife Conservation from the Bank's Office of the Publisher provided Society, and colleagues for the use of their data on very valuable editing and publishing assistance. Luis remaining vegetation and protected areas of Central A. Sol6rzano translated the report into Spanish. America. Frank Reichenbacher and David Brown of Substantial funding for this study was provided Southwestern Field Biologists provided a map of by the Global Environment Coordination Division of biotic communities of Central America and the the World Bank, responsible for administering the Caribbean. Kent Redford and Roger Sayre of The Bank's Global Environment Facility (GEF) portfolio. Nature Conservancy provided data on several We would like to highlight the support and contri- countries of LAC. The Colombian Centro Interna- butions of Kenneth Newcombe and Kathy Mac- cional para la Agricultura Tropical (CIAT) provided kinnon. data on protected areas of LAC. Jorge Sober6n and The World Wildlife Fund (WWF) contributed his colleagues at CONABIO (the Mexican Comisi6n substantial funds of its own for the completion of Nacional para el Conocimiento y Uso de la Biodi- this study. WWF, through a grant from the Ford versidad) provided data on the remaining vegeta- Foundation, provided funding for the mangrove tion of Mexico. Tom Stone and his colleagues at conservation assessment workshop. We extend our Woods Hole provided advanced very high resolu- thanks to all of the mangrove experts who partici- tion radiometer (AVHRR) data on remaining vege- pated in the workshop. The LAC program of WWF tation of South America. Gonzalo Castro and provided valuable expertise in assessing the conser- colleagues at Wetlands for the Americas provided vation status of ecoregions and their biological dis- maps of LAC wetlands. We also received valuable tinctiveness, and helped develop the decision rules unpublished data from the national park agencies of for prioritizing among ecoregions. Josh Podowski, Argentina and Bolivia. The Environmental Systems Stuart Sheppard, and Ian Heywood helped with the Research Institute, and in particular Jack Danger- map digitizing and other related tasks. Pia lolster, mond, Lance Shipman, and Charles Convis, provided Emily Aikenhead, and Pedro Lopez-Valencia technical support and made software available for ix x A Conscroation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean GIS applications. Hewlett-Packard, Inc., and Apple included James Quinn, Gordon Orians, John Computer, Inc., provided technical support through Terborgh, Leonardo Lacerda, Kent Redford, Allan hardware and software donations. The World Putney, John Mackinnon, and Javier Simonetti. Sev- Conservation Monitoring Centre provided data eral individuals read the final draft report and made on remaining vegetation and protected areas for valuable comments, including Annie Brunholzi, selected countries. Kent Redford, Margaret Symington, Steve Cornelius, A number of individuals reviewed early drafts Garo Batmanian, Miguel Pellerano, Luis Sol6rzano, of the methodology and shared their ideas. These and Kathy Mackinnon. Authors Eric Dinerstein, Ph.D., Director, Conservation Science Program, World Wildlife Fund-US, Washington, D.C. David M. Olson, Ph.D., Conservation Scientist, Conservation Science Program, World Wildlife Fund-US, Washington, D.C. Douglas J. Graham, M.S., Environmental Specialist, Environment Unit, Technical Department, Latin America and the Caribbean Region, The World Bank, Washington, D.C. Avis L. Webster, M.S., Spatial Technologies Specialist, Conservation Science Program, World Wildlife Fund-US, Washington, D.C. Steven A. Primm, M.S., Research Associate, Conservation Science Program, World Wildlife Fund-US, Washington, D.C. Marnie P. Bookbinder, M.S., Conservation Science Fellow, Conservation Science Program, World Wildlife Fund-US, Washington, D.C. George Ledec, Ph.D., Ecologist, Environment Unit, Technical Department, Latin America and the Caribbean Region, The World Bank, Washington, D.C. xi Contributors E. Aikenhead (World Wildlife Fund-US) G. Cintron-Molero (USFWS, USA) J. Alcorn (Biodiversity Support Program, USA) M. Collins (World Conservation Monitoring C. Alderman (World Bank, USA) Centre, UK) C. Alho (World Wildlife Fund-BRAZIL) J. E. Conde (Laboratorio de Biologia Marina, Centro M. Ayres (Universidade Federal do Pard, BRAZIL) de Ecologia y Ciencias Ambientales, P. Bacon (University of the West Indies, TRINIDAD VENEZUELA) AND TOBAGO) S. Contreras Balderas (Universidad Aut6noma de J. Barborak (Wildlife Conservation Society, USA) Nuevo Le6n, MEXICO) E. Barriga (USAID, COLOMBIA) C. Convis (Environmental Systems Research G. Batmanian (World Wildlife Fund-US) Institute, USA) M. Baudoin (Secretaria Nacional del Medio S. Cornelius (World Wildlife Fund-US) Ambiente, BOLIVIA) L. Co to (Universidade Federal de Vihosa, BRAZIL) M. Bonifaz (World Bank, USA) J. Dangeriond (Environmental Systems Research P. Brandriss (World Bank, USA) Institute, USA) J. Brokaw (USAID, USA) P. Dansercan (Universit( du Qu6bec A Montr&al, D. Brown (Southwestern Field Biologists, USA) CANADA) K. Brown (Universidade de Campinas [UNICAMP], 1. Davidson (Wetlands for the Americas, CANADA) BRAZIL) R. de la Maza (Conservation International, MEXICO) D. Bryant (World Resources Institute, USA) P. DeVries (Harvard University, USA) E. Bucher (Universidad Nacional de Cordoba, B. Dias (Ministerio do Meio Ambiente e da ARGENTINA) Amaz6nia Legal, BRAZIL) G. Burgess (University of Florida, USA) A. Dickie (USAID, USA/GUATEMALA) B. Buschbacher (Bel6m Project Office, World Wildlife L. Diego GCiez (Organization for Tropical Studies, Fund-US) COSTA RICA) J. Cajal (Fundaci6n para la Conservacion de las R. Dirzo (Universidad Nacional Aut6nomo de Especies y Medio Ambiente [FUCEMAI, M6xico [UNAMI, MEXICO) ARGENTINA) W. Duelnan (University of Kansas, USA) P. Canevari (Wetlands for the Americas, J. Echevarria (Universidad Nacional de Tumbes, ARGENTINA) PERU) M. Canevari (Administraci6n de Parques F. Erize (ARGENTINA) Nacionales, ARGENTINA) C. Espinosa (Instituto Nicaragiiense de Recursos C. Carr (Wildlife Conservation Society, USA) Naturales y del Ambiente [IRENA], G. Castilleja (World Wildlife Fund-US) NICARAGUA) G. Castro (Wetlands for the Americas, USA) E. Ezcirra (Secretaria de Desarrollo Social R. Cavalcanti (Universidade de Brasilia, BRAZIL) [SEDESOL], MEXICO) B. Chernoff (Field Museum of Natural History, R. Ford Smith (VENEZUELA) USA) F. Flores Verdngo (Centro Unidad Mazatlan en K. Chomitz (World Bank, USA) Agricultura y Manejo Ambiental, MEXICO) M. Cifuentes (World Wildlife Fund-US) G. Fonseca (Conservation International, BRAZIL) xii Contributors xiii E. Forno G. (Centro de Datos para la Conservaci6n, J. Madill (CANADA) BOLIVIA) 5. Malone (Conservation International, SURINAME) R. Foster (Field Museum of Natural History, USA) F. Marcoux (CANADA) P. Foster-Turley (USAID, USA) J. Mariaca P. (Direcci6n Nacional de Conservaci6n de P. Freeman (Environment and Natural Resources In- la Biodiversidad, BOLIVIA) formation Center [ENRIC], USA) P. Marquet (Pontificia Universidad Cat6lica de Chile, C. Freese (World Wildlife Fund-US) CHILE) E. Fuentes (CHILE) J. McKenna (World Bank, USA) M. Garcia Donayre (Instituto Nacional de Recursos G. Medina (World Wildlife Fund-US) Naturales [INRENA1, PERU) N. Menczes (Universidade de Sio Paulo, BRAZIL) R. Gauto (Fundacion Moises Bertoni, PARAGUAY) F. Mereles (Universidad de Asunci6n, PARAGUAY) D. Gibson (World Bank, USA) R. Mitterieier (Conservation International, USA) W. M. Graham (CANADA) B. Moffat (World Wildlife Fund-US) A. Grajal (Wildlife Conservation Society, USA) A. Moluar (World Bank, USA) G. Gran (World Wide Fund for Nature-UK) A. Moreira (USAID, BRAZIL) D. Gross (World Bank, USA) 1. Morello (Harvard University, USA) D. Heesen (USAID, COSTA RICA) D. Neill (Fundaci6n Jatun Sacha, ECUADOR) A. Henderson (New York Botanical Gardens, USA) D. Nepstad (Woods Hole Research Center, USA) 0. Herrera-MacBryde (IUCN/Smithsonian, USA) K. Newcombe (World Bank, USA) 1. Heywood (World Wildlife Fund-US) S. Oliver (World Bank, USA) S. Higman (Guyana Forest Commission, GUYANA) S. Olivieri (Conservation International, USA) R. Howard (Arnold Arboretum, USA) G. Orians (University of Washington, USA) 0. Huber (Fundaci6n Botanico, VENEZUELA) J. Ottenw1alder (United Nations Development Pro- R. Huber (World Bank, USA) gramme [UNDP], DOMINICAN REPUBLIC) E. lIigo-Elias (University of Florida, USA) G. Paz y Mifio (World Bank, USA) P. lolster (World Wildlife Fund-US) M. Pellerano (World Wildlife Fund-US) A. Iriarte Walton (Servicio Agricola y Ganadero, C. A. Peres (Universidade de Sio Paulo, BRAZIL) Ministerio de Agricultura, CHILE) F. Pierce (USAID, GUATEMALA) J. lzquierdo (FAO, CHILE) J. Pitt (University of Grenada, GRENADA) D. Janzen (University of Pennsylvania, USA/COSTA C. Plaza (World Bank, USA) RICA) J. Podowski (World Wildlife Fund-US) N. Johnson (World Resources Institute [WRI], USA) B. Potter (Island Resources Foundation, USA) P. Jones (CIAT, COLOMBIA) G. Powell (Centro Cientifico Tropical, COSTA RICA) M. Kalin de Arroyo (Universidad de Chile, CHILE) S. Price (World Wildlife Fund-CANADA) C. Kane (World Wildlife Fund-US) G. Prickett (USAID, USA) S. Keel (The Nature Conservancy, USA) A. Putney (IUCN, USA) S. Kelleher (Biodiversity Support Program, USA) L. Quevedo (World Wildlife Fund-BOLIVIA) R. Kerr (The Hope Zoo, JAMAICA) J. Quinn (University of California, USA) M. Kiernan (World Wildlife Fund-US) K. Redford (The Nature Conservancy, USA) B. Kirise (World Bank, USA) H. Reichart (World Wide Fund for Nature- P. Koohajkan (FAO, ITALY) NETHERLANDS) M. Kux (USAID, USA) F. Reichienbacher (Southwestern Field Biologists, USA) L. Lacerda (World Wide Fund for Nature-Interna- M. Ribera (Centro de Datos para la Conservaci6n, tional) BOLIVIA) G. Lamas (Universidad Nacional Mayor San Marcos, J. Robinson (Wildlife Conservation Society, USA) PERU) M. Rodrigues (Universidade de Sao Paulo, BRAZIL) P. Lopez-Valencia (World Wildlife Fund-US) C. Rodstroi (Conservation International, USA) R. Luxmoore (World Conservation Monitoring P. Rosabal (IUCN, PUERTO RICO) Centre, UK) J. Ruiz (ECOFONDO, COLOMBIA) N. Maceira (Instituto Nacional de Tecnologia E. Santana (Universidad de Guadalajara, MEXICO) Agropecuaria [INTA], ARGENTINA) C. Saravia T. (Centro de Investigaciones Ecol6gicas R. Machado (Conservation International, USA) del Chaco, ARGENTINA) J. Mackinnon (Asian Bureau of Conservation, HONG F. Sarmiento (University of Georgia, USA) KONG) K. Saterson (Biodiversity Support Program, USA) K. Mackinnon (World Bank, USA) R. Sayre (The Nature Conservancy, USA) xiv A Conservation Assessment of the Terrestrial Ecoregions of Latin Amenrca and the Caribbean Y. Schaeffer-Novelli (Universidade de Sao Paulo, R. Tuazon (Inter-American Development Bank, USA) BRAZIL) R. Twilley (University of Southwestern Louisiana, S. Schonberger (World Bank, USA) USA) F. Seymour (World Wildlife Fund-US) M. Vales G. (CUBA) S. Sheppard (World Wildlife Fund-US) B. Watson (The Nature Conservancy, L. Shipnan (Environmental Systems Research NICARAGUA/USA) Institute, USA) M. Webb Records (World Bank, USA) J. Simonetti (Universidad de Chile, CHILE) R. Welch (SOS Mata Atlantica, BRAZIL) 1. Sober6n (CONABIO, MEXICO) C. Wicks (World Wide Fund for Nature-UK) C. Sobrevila (World Bank, USA) B. Wilcox (Institute for Sustainable Development, USA) 0. Solbrig (Harvard University, USA/ARGENTINA) E. Wilcox (World Wildlife Fund-US) L. Sol6rzano (Princeton University, N. Windevoxhel (IUCN Regional Wetlands USA/COLOMBIA) Programme for Mesoamerica, COSTA RICA) T. Stone (Woods Hole Research Center, USA) D. Wood (World Wildlife Fund-US) D. Stotz (Field Museum of Natural History, USA) M. Yates (USAID, BOLIVIA) M. Symington (Biodiversity Support Program, USA) E. Yerena 0. (Instituto Nacional de Parques J. Terborgh (Duke University, USA) [INPARQUES], VENEZUELA) K. Thelen (FAO, CHILE) F. Zadroga (USAID, MEXICO) J. Tosi (Centro de Ciencias Tropicales, COSTA RICA) L. Zeitlin-Hale (University of Rhode Island, USA) Acronyms and Abbreviations AVHRR Advanced very high resolution radiometer BSP Biodiversity Support Program CI Conservation International CIAT Centro Internacional para la Agricultura Tropical (Colombia) CONABIO Comisi6n Nacional para el Conocimiento y Uso de la Biodiversidad (Mexico) FAO Food and Agriculture Organization of the United Nations GEF Global Environment Facility GIS Geographic information system IUCN The World Conservation Union LAC Latin America and the Caribbean LATEN Environment Unit (Technical Department) for LAC (The World Bank) MET Major ecosystem type (see also Glossary) MHT Major habitat type (see also Glossary) MSS Multispectral scanner TNC The Nature Conservancy USAID United States Agency for International Development USFWS United States Fish and Wildlife Service WRI World Resources Institute WWF World Wildlife Fund xv Executive Summary This priority-setting study elevates, as a first princi- gion boundaries recognized by conservation plan- ple, maintaining the representation of all ecosystem ners and biogeographers. The classification scheme and habitat types in regional investment portfolios. helps maintain the goal of representing all habitat Second, it recognizes landscape-level features as an and ecosystem types in an investment portfolio. The essential guide for effective conservation planning. scheme also allows us to assess conservation status Without an objective framework to assess the con- using criteria tailored to the distinct dynamics and servation status and biological distinctiveness of spatial patterns of biodiversity of different ecosys- geographic areas, donors run the risk of overlooking tem types. areas that are seriously threatened and of greatest The conservation assessment integrates two fun- biodiversity value. damental data layers: conservation status and biologi- The lack of such an objective regional framework cal distinctiveness. To assess the conservation status of prompted this study, whose goals were (a) to replace ecoregions, we adapted the Red Data Book approach the relatively ad hoc decisionmaking process of do- used by the World Conservation Union (IUCN) for nors investing in biodiversity conservation with a species. We conducted a snapshot conservation more transparent and scientific approach; (b) to assessment of ecoregions to classify them as either move beyond evaluations based largely on species Extinct (completely converted), Critical, Endangered, lists to a new framework that also incorporates Vulnerable, Relatively Stable, or Relatively Intact. maintaining ecosystem and habitat diversity; (c) to The classifications are based on five indicators of better integrate the principles of conservation biol- landscape integrity: total loss of original habitat, ogy and landscape ecology into decisionmaking; and number and size of blocks of intact habitat, rate of (d) to ensure that proportionately more funding be habitat conversion, degree of fragmentation or deg- channeled to areas that are of high biological value radation, and degree of protection. We assumed that and under serious threat. unfavorable changes in these indicators lowered the The study was financed by the World Bank, the probability of ecological processes and major com- Global Environment Facility, and the World Wildlife ponents of biodiversity being maintained within a Fund (WWF), and it was carried out by WWF's Con- given ecoregion. The snapshot conservation status is servation Science Program in close collaboration further modified to a final conservation status after with the World Bank's Environment Unit for Latin considering potential threats over the next 20 years America and the Caribbean. to ecoregions based on their type, timeframe, spatial The study's biogeographic approach to setting scale, and intensity. conservation priorities begins by dividing LAC into The biological distinctiveness of an ecoregion is 5 major ecosystem types (METs), 11 major habitat types assessed within its MHT, thus ensuring that we (MHTs), and 191 ecoregions. This last number includes compare tropical moist broadleaf forests only to 13 mangrove complexes but otherwise excludes other tropical moist broadleaf forests and not to marine areas and freshwater habitats (except large, montane grasslands with major differences in geographically contiguous freshwater ecosystems). patterns of biodiversity. We classify ecoregions as The classification scheme for ecoregions builds on Globally Outstanding, Regionally Outstanding, existing work and, wherever possible, uses ecore- Bioregionally Outstanding, or Locally Important. xvi Executive Sumnary xvii To undertake these analyses, we solicited help The integration of the conservation status of from a wide range of biodiversity specialists and ecoregions with their biological distinctiveness helps conservation planners from the LAC region. to identify biodiversity conservation priorities. We Regional experts reviewed the landscape level used a simple matrix of these two variables to iden- criteria and threats in order to determine the con- tify the ecoregions of highest conservation impor- servation status. They also provided data and tance. Each MHT was analyzed separately in order assessments of the key variables used to determine to maintain representation of all major habitat types. the biological distinctiveness of ecoregions: beta Fifty-five out of 178 ecoregions distributed among diversity, species richness and endemism for several the 10 non-mangrove MHTs were designated as of major taxa, unique ecological communities or proc- Highest Priority at the Regional Scale (level I). These esses, and rarity or distinctiveness of ecosystem or include 23 ecoregions in tropical moist broadleaf habitat types at varying biogeographic scales. forests; 5 in tropical dry broadleaf forests; 2 in tem- As indicated by the final conservation status of the perate forests; 5 in tropical and subtropical conifer- 178 non-mangrove ecoregions, we identified 31 ous forests; 2 in grasslands, savannas, and shrub- ecoregions as Critical, 51 as Endangered, 55 as Vul- lands; 4 in flooded grasslands; 8 in montane grass- nerable, 27 as Relatively Stable, 8 as Relatively lands; 2 in Mediterranean scrub; 2 in deserts and Intact, and 6 as Unclassified. Thirty-nine ecoregions xeric shrublands; and 2 in restingas. were considered more threatened after the assess- We also created and applied a set of decision rules ment of threat was applied to the snapshot conser- designed to achieve better bioregional representa- vation status. The highest number of Critical and tion. Applying these decision rules added 19 ecore- Endangered ecoregions occurred in tropical moist gions to the highest priority list. broadleaf and tropical dry broadleaf forests. How- Mangrove units were assessed separately; we ever, only 3 percent of the tropical dry broadleaf assigned priority rankings to the 40 units making up forest ecoregions were Relatively Stable or Relatively the 13 mangrove complexes. Rankings were assigned Intact, whereas 27 percent of the tropical moist to investments for restoration, conservation with broadleaf forest ecoregions were Relatively Stable or restricted use, or conservation for sustainable use. Relatively Intact, indicating that tropical dry broad- Patterns of biodiversity and landscape integrity leaf forests are on average more threatened than any of whole ecoregions are given precedence over other forest type in the LAC region. Most (83 per- other analytical layers because changes in these cent) xeric ecoregions were either Critical, Endan- two layers are essentially irreversible. The integra- gered, or Vulnerable. tion model based on these two layers identifies Thirty-four ecoregions in LAC were deemed priorities and suggests the timing and sequence of Globally Outstanding and 32 others were considered investment and the level of effort required to Regionally Outstanding. There were 59 Bioregionally conserve biodiversity. Outstanding ecoregions and 47 Locally Important The matrix and decision rules presented can help ecoregions. Numerically, tropical moist broadleaf guide governments and donors in preventing com- forests contain most of the Globally Outstanding plete degradation and conversion in the most threat- ecoregions and 47 percent of all the Regionally Out- ened regions. Such investments may be quite standing ecoregions. They are concentrated in the expensive. The matrix also identifies the most intact western portion of Amazonia and the tropical north- examples of biologically outstanding ecoregions ern Andes, with others scattered among northern with the best chance for long-term persistence and Mexico, the Atlantic forests of Brazil, and southern the highest potential for cost-effective investments. Chile. Montane grasslands have the highest propor- We stop short of identifying specific investment pri- tion of Globally or Regionally Outstanding ecore- orities within ecoregions, which should be done only gions, followed by tropical moist broadleaf forests. after assessing political, social, and economic factors. Three MHTs (tropical dry broadleaf forests; grass- These secondary factors are more fluid than biologi- lands, savannas, and shrublands; and deserts and cal variables and, in our view, are best applied in xeric shrublands) had a higher proportion of Biore- intra-ecoregion analyses. Application at finer geo- gionally Outstanding and Locally Important ecore- graphic scales will better direct scarce funds to the gions than did the other MHTs. All MHTs were sites within ecoregions that are the most intact, bio- represented by at least one ecoregion classified as logically important, and most likely to preserve either Globally or Regionally Outstanding. biodiversity over the long term. Introduction The tropical, subtropical, and temperate habitats of Fund (WWF) to carry out a study to identify geo- Latin America and the Caribbean (LAC) contain graphic priorities for biodiversity conservation some of the most important areas in the world for based on the integration of the biological distinctive- biodiversity conservation. Many of these areas face ness and conservation status of ecoregions. severe threats, and financiers and donors have A second objective was to create an approach that responded to these pressures by financing biodi- would be rigorous enough to be incorporated into versity conservation programs across the region. The the conservation planning exercises of LAC countries World Bank and the Global Environment Facility and other donors active in the region. The approach (GEF) have invested heavily in projects that support developed for this study has indeed influenced sev- the creation or strengthening of protected areas, eral priority-setting exercises underway in LAC: institutional strengthening of parks departments, protection of endangered species, and management * A priority-setting exercise financed by USAID, of natural resources.' The total portfolio of conser- coordinated by the Biodiversity Support Program vation investments of the World Bank and the GEF, (BSP), and conducted by the WWF Conservation particularly as new GEF projects for LAC are Science Program, the Wildlife Conservation approved over the next few years, probably exceeds Society, The Nature Conservancy, World Resources the financial contributions of all other conservation Institute, Conservation International, and a large donors combined. group of regional experts. This exercise was If targeted properly, the resources of the GEF, the designed to assist USAID in identifying geo- World Bank, and other major financiers and donors graphic areas of high priority for biodiversity can have enormous impact in stemming the destruc- investments in LAC (BSP 1995). tion and degradation of biologically important areas * A priority-setting exercise conducted by the World in LAC. Specialists within the GEF and LATEN, the Wildlife Fund's LAC program as part of its strate- World Bank's Environment Unit for Latin America gic planning activities to design the next three-year and the Caribbean, recognized three years ago the funding cycle (Wood and Cornelius 1994). need for better information to help target these * Argentina's National Parks Department's analysis resources. This study was one way of addressing of the conservation status of Argentine ecoregions these concerns. The World Bank contracted the Con- using methods based in part on a preliminary servation Science Program of the World Wildlife draft of our method (Olson and Dinerstein 1994). The analysis of Cajal (1994) is similarly based on 1. As of July 1995, the World Bank had an active portfolio of 21 an early draft of the method. approved loans, to most of the major countries of LAC, that specifically included resources targeted for biodiversity con- The ecoregion base map developed for this study servation. These loans represent a total investment in biodi- is being widely used, even prior to publication, in a versity of about $470 million of which about $320 million is number of studies and planning activities. These direct Bank lending. The World Bank is also one of the imple- menting agencies of the GEF, and in LAC, five World Bank include: an analysis of indigenous areas by Wilcox GEF biodiversity projects had been approved by July 1995 for and Duin (1995); The Nature Conservancy planning a total of $72 million. A sum of about $290 million has been exercise for their Parks in Peril program; and a pri- pledged to the G-7 Pilot Program to Conserve the Brazilian Rain Forest, coordinated by the World Bank, which has many ority-setting exercise for LAC wetlands conducted projects now in the planning or implementation stages. by Wetlands for the Americas. 1 2 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean This exercise, and other similar efforts, seek to systems is currently being undertaken by Wetlands advance biodiversity conservation planning beyond for the Americas. The World Bank, in collaboration previous approaches, such as "Hotspots" (Myers with the Great Barrier Reef Marine Park Authority 1988) and the "Megadiversity-country" approach and the The World Conservation Union (IUCN) (Mittermeier and Werner 1990). In these approaches, have recently published a four volume series on geographic priorities were selected largely on the global priorities for marine ecosystems (Kelleher et basis of the single indicator of species richness. al. 1995). These exercises complement this effort. "Hotspots" and "Megadiversity" approaches did not This report is divided into seven chapters. The directly address what is now widely accepted as a first chapter (Approach) provides an overview of the central goal of regional and national biodiversity methodologies used in this study. A more detailed conservation strategies: to maintain representation description of the methods can be found in Appen- of all ecosystem and habitat types in order to con- dixes A and B. The second chapter provides a serve the distinct communities of organisms they description of the METs and MHTs found in LAC, contain (Noss 1992; Scott et al. 1993; Noss and Coop- their geographic locations, and the ecoregions they errider 1994; Dinerstein et al. 1994). This study, and contain. See more detailed definition of MHTs in others like it, signals a shift in priority-setting focus, Appendix C and the hierarchical classification away from species richness as the sole or major dis- scheme of ecoregions in Appendix D. Brief descrip- criminator, and towards conserving ecosystem and tions of the ecoregions and key sources consulted for habitat diversity. their delineation, classification and assessment are in Although building on excellent previous work, we Appendix F. highlight eight specific advances of this study: The third chapter presents the results of the analysis of conservation status by MHT and across * A hierarchical classification scheme that divides MHTs. The raw data for each ecoregion used to LAC into major ecosystem types (METs), major compute conservation status are in Appendix E and habitat types (MHTs), and 191 ecoregions to detailed information on sources consulted for ensure that representation is maintained at an remaining natural habitat and protected areas is in appropriate biogeographic scale for regional con- Appendix G. The fourth chapter presents an assess- servation planning ment of the biological distinctiveness of ecoregions * A transparent method that incorporates the prin- within their MHTs. The fifth chapter explains how to ciples of landscape ecology to assess the snapshot use the methodology as a tool in conservation plan- conservation status of the ecoregions of LAC ning. The sixth chapter addresses the conservation * A classification of the conservation status of eco- status of mangroves of LAC which we assessed regions in the tradition of the IUCN Red Data Books using an approach adapted from that developed for * An assessment of conservation status and threats the terrestrial ecoregions. tailored to the particular dynamic features and The last chapter provides our conclusions and threats specific to different types of ecosystems recommendations. After summarizing the results of * An approach for assessing the biological distinct- the study, we discuss the possibility of using our iveness of ecoregions at varying biogeographic approach at an ecoregional or intra-ecoregional scales scale. Such analyses at finer geographic scales are an * Consideration of mangrove ecosystems, ecologi- essential follow-up to this study. Without them, cally and economically valuable formations that donors run the risk of financing biodiversity conser- have received inadequate attention from conser- vation in the most important ecoregions within a vation donors major habitat type, but conserving some of the less * Well-documented GIS databases that can be used important habitat blocks within those ecoregions, or by other individuals to reanalyze and update the at an insufficient level to conserve their biodiversity. data layers in this analysis We offer some guidelines that highlight maintaining * A method for integrating biological distinctive- landscape integrity and better conserving areas ness of ecoregions with their conservation status characterized by high rates of species turnover designed to identify regional biodiversity priori- (beta diversity), and we give prominence to unique ties for conservation in each MET and to promote habitats. bioregional representation. It is essential that biodiversity be conserved every- where, but clearly some areas have higher value than We limit our analyses to terrestrial ecoregions. A others and degrees of threat vary. The geographic regional priority-setting initiative for freshwater eco- biodiversity priorities identified by this study take Introduction 3 these considerations into account. They also serve as * Douglas J. Graham or George Ledec a precursor for guiding investment priorities, which Environment Unit, Technical Department can only be determined after economic, social, insti- Latin America and the Caribbean Region tutional and political factors are considered within The World Bank priority ecoregions. This study stops short of de- 1818 H Street, NW termining site-specific investment priorities, which, Washington, DC 20433 USA in any event, should be determined by comprehen- Fax: (202) 676-9373 sive planning efforts conducted at the ecoregion or Internet: DGraham@Worldbank.org, or country level. Nonetheless, we stress that biodiver- GLedec@Worldbank.org sity assessments and analyses of landscape integrity should be considered the foundations of priority- * Eric Dinerstein setting exercises at any scale. Conservation Science Program World Wildlife Fund Comments on this report would be welcomed by the 1250 24th Street, NW World Bank or the World Wildlife Fund at the fol- Washington, DC 20037-1175 USA lowing addresses: Fax: (202) 861-8377 1 Approach Fundamental Goals Underlying the Approach of representative examples of the most diverse array of natural habitats in a region is the most cost- The terrestrial ecoregions of Latin America and the effective way to prevent species extinction. Previous Caribbean (LAC) cover the spectrum -from some of efforts at priority-setting that were largely deter- the driest deserts, such as the Atacama of Chile and mined by species lists fail to capture the diversity of Peru, to some of the wettest forests, such as the ecosystems and the biota they contain, to the detri- Choc6/DariCn forests of Panama, Colombia, and ment of xeric and non-forested formations in par- Ecuador. A few of these ecoregions are almost com- ticular (Scott et al. 1993; Csuti in prep). pletely converted or degraded, whereas others To achieve the goal of representation, we devel- contain vast, intact blocks of original habitat. The oped a hierarchical classification scheme that degree of protection of remaining habitat blocks also divides LAC into 5 major ecosystem types (METs), varies widely from one ecosystem type to another. In 11 major habitat types (MHTs), and 191 ecoregions some cases there is nearly complete protection, while (Figure 1-2). The hierarchy helps achieve represen- in others there is little or no protection either now or tation by assessing priorities at biologically mean- in the foreseeable future. Most of the ecoregions in ingful levels. Chapter 2 presents in more detail LAC fall between these extremes. LAC's METs, MHTs, and ecoregions. This study provides a biologically based frame- An ecoregion is defined as a geographically distinct work for a conservation assessment of LAC's assemblage of natural communities that share a large diverse ecoregions. The design framework, based on majority of their species, ecological dynamics, and similar widely accepted goals in conservation biology and environmental conditions and whose ecological interac- landscape ecology (Noss 1992; Krever et al. 1994), tions are critical for their long-term persistence. We incorporates a number of analytical steps to help believe that the ecoregion unit is the minimum level determine biodiversity conservation priorities of resolution required for achieving regional repre- (Figure 1-1). The fundamental goals underlying our sentation and effective conservation planning.' In approach include: this manner, ecoregions that fall under tropical moist broadleaf forests, for example, are compared to * Representation of all distinct natural communities other such forests rather than to coniferous or grass- * Maintenance of ecological and evolutionary pro- land ecoregions. cesses that create and sustain biodiversity Ecoregions within the same major habitat type can * Maintenance of viable populations of species be similar in their structure and ecological processes * Conservation of blocks of natural habitat large but share few species due to biogeographic barriers enough to be responsive to large-scale periodic or species turnover with distance. Representation of disturbances and long-term changes (e.g., global biodiversity at the level of species-as well as at warming). higher hierarchical levels of ecosystems and habitat Goal 1: Representation types-is critical in any comprehensive regional Maintaining representation is critical to conserving 1. The boundaries of an ecoregion should identify an area ecosystem and habitat diversity. Also, conservation over which a single conservation strategy could be effectively biologists are in virtual consensus that conservation applied. 4 Approach 5 Figure 1-1. Analytical Steps Used to Derive Biodiversity Conservation Priorities Develop hierarchical classification scheme (Figure 1-2 and Appendix D) Determine snapshot conservation status of ecoregions (Map 5) using following weighting of landscape features: * Habitat loss 40% * Habitat blocks 20% * Fragmentation/ Assess the scale-dependent biological degradation 20% distinctiveness of ecoregions (Map 7) based * Conversion rate 10% on species richness, endemism, beta * Degree of protection 10% diversity, biological phenomena, and relative rarity of habitat type Determine final conservation status (Map 6) by modifying snapshot conservation status with a threat analysis Apply integration matrix (Table 1-1) Identify biodiversity conservation priorities (Map 8) Fine-tune priorities by incorportn consideration of bioregional representation (Map 9) cQnselvatioh'(e.g., analyze "Iksa'~ej p~~ ot1 prot&ctd areas, land use, and&tflittI1 otmc oil M Figure 1-2. Hierarchical Classification Scheme of METs, MHTs, Ecoregions, and Bioregions Major Ecosystem Types Major Habitat Types Selected Ecoregions Selected Bioregions (METs) (MHTs) (of 191) (of 9) Tropical broadleaf No Tropical moist broadleaf 135. Mexican alpine tundra forests forests Central America 136. Costa Rican paramo Tropical dry broadleaf forests 137. Santa Marta paramo Conifer/ temperate No Temperate forests broadleaf forests 138. Cordillera de Mdrida paramo Tropical and subtropical Northern Andes coniferous forests 1ep 139. Northern Andean paramo Grasslands/savannas/ 00 Grasslands, savannas, and 140. Cordillera Central paramo shrublands shrublands 141. Central Andean puna Flooded grasslands Central Andes Montane grasslands -- No 142. Central Andean wet puna Xeric formations - p Mediterranean scrub 143. Central Andean drypuna Deserts and xeric shrublands 144. Southern Andean steppe Restingas 145. Patagonian steppe Southern South America Mangroves m - Mangroves 146. Patagonian grasslands Definition: An MET is a set of ecoregions Definition: An MHT is a set of eco- Definition: An ecoregion is a geo- Definition: A bioregion is a that (a) share comparable ecosystem regions that (a) exrience comparable graphically distinct assemblage of geographically related dynaics; (b) have similar response climatic regimes; have similar vegeta- natural communities that (a) share a assemblage o ecoregions that characteristics to disturbance; (c) exhibit tion structure; (c) display similar spatial lre majority of their es and share a similar biogeographic similar degrees of beta diversity; and ptterns of biodiversit (eg,lvlso clgical dynamics; (b) har similar history and thus have stron& (d) requnre an ecosystem- c beta diversity () contain flora and environmental conditions; and (c) affinities at higher taxonomic conservation approach. fauna with similar guild structures and interact ecologically in ways that are levels (e.g., genera, families). life histories. critical for their long-term persistence. Approach 7 biodiversity strategy. For these reasons, representa- Snapshot Conservation Status tion must be achieved among ecoregions that belong to the same MHT but occur in different geographic The conservation status of ecoregions was deter- areas of LAC. mined by summing the numerical values assigned to We thus defined nine discrete biogeographic the five key landscape-level variables mentioned areas that we call bioregions. As explained in more above: loss of original habitat, number and size of detail in subsequent chapters, these bioregions are large blocks of original habitat, degree of fragmen- used both in the definition of scale-dependent bio- tation and degradation, rate of conversion of remain- logical distinctiveness and to fine-tune the final ing habitat, and degree of protection. To assess results of the priority-setting exercise to achieve the landscape-level features, we relied on expert opinion goal of better biogeographic representation. The nine and on spatial databases and maps (where such data bioregions, which reflect the biogeographic distinct- were available and of reasonable quality). More de- iveness of major areas of LAC (Map 1), are Carib- tailed information on assessment of landscape-level bean, Northern Mexico, Central America, Orinoco, features by ecoregion is presented in Appendix A, Amazonia, Northern Andes, Central Andes, Eastern and information on remaining vegetation and pro- South America, and Southern South America. tected areas is in Appendix G. To what extent are the ecoregions that support Goals 2 to 4: Maintenance of Ecological Processes populations, communities, and unique ecosystems and Viable Species Populations threatened? To answer this question, we assessed conservation status of ecoregions in the tradition of Directly assessing the status of ecological processes the IUCN Red Data Books. The IUCN Red Data and viability of species populations (Goals 2 and 3) Book categories of threat have gained widespread is often logistically difficult, but estimates are critical acceptance as a framework for determining the for a sound priority-setting framework. We achieve conservation status of species and populations. this goal with an index of conservation status using Such assessments have been codified into various landscape-level features. The indicators are the loss Red Data Books to call attention to species and of original habitat, the presence of large blocks of populations considered to be on the verge of contiguous habitat, degree of fragmentation and extinction (IUCN 1988; Collar et al. 1992). This degradation, rates of habitat conversion, and the study marks the first time that ecoregions have degree of protection within an ecoregion. A key been classified this way. assumption of our approach is that the indicators Mace and Lande (1991) reviewed IUCN categories selected, and the manner in which they are of threat and suggested three objectively defined weighted, serve as robust predictors of (a) the extent categories: Critical, Endangered and Vulnerable. to which important components of biodiversity (e.g., Inspired by this review, we classified ecoregions as rare keystone species or the presence of viable Extinct (completely converted), Critical, Endan- populations of top predators) will likely persist over gered, Vulnerable, Relatively Stable, and Relatively time; and (b) the maintenance of important ecologi- Intact. Classification was determined by weighting cal processes. The method to determine conservation the numerical values defined for each of the five status is discussed further in this chapter and in landscape variables listed above. In weighting these Appendix A. variables, we gave much greater prominence to loss Maintaining areas large enough to withstand peri- of original habitat and number of large blocks of in- odic, large-scale disturbances and long-term changes tact habitat. We believe that these variables are the (Goal 4) is also best assessed using landscape-level best indicators of the probability of persistence of features such as the presence of large blocks of ecological processes within ecoregions. The ranges of original habitat and total loss of original habitat. values used to classify the ecoregions were derived Degradation, indicated by these landscape features, from the conservation biology, theoretical ecology, would have major impacts on regional biodiversity and landscape ecology literature (see Appendix A). conservation because extensive loss of original In each of the definitions below, we first describe habitat and large blocks of natural habitats is essen- the ecoregion's landscape integrity as assessed by tially irreversible. With few exceptions - such as some the more quantifiable criteria listed in Appendix A. mangrove units and some savannas -restoration of The remaining sentences describe qualitatively the converted or highly degraded habitats is too slow predicted ecological and biological impacts of loss and costly to be an effective conservation tool. Thus, of landscape integrity. They reflect how with achieving Goal 4 depends on sufficiently large, rela- increasing habitat loss, degradation, and fragmen- tively intact units that still function naturally. tation, ecological processes cease to function 8 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean naturally, or at all, and major components of sive, but are still patchily distributed relative to biodiversity are steadily eroded. The categories of the area of intact habitats. Ecological linkages conservation status are: among intact habitat blocks are still largely func- tional. Guilds of species that are sensitive to human Extinct. No natural communities resembling the activities, such as top predators, the larger pri- original ecosystems remain. Some of the original mates, and ground-dwelling birds, are present but biota are still present but persist only within at densities below the natural range of variation. highly modified communities and landscapes. No * Relatively intact. Natural communities within an opportunities for restoration of the original natu- ecoregion are largely intact with species, popula- ral communities exist due to permanent alteration tions, and ecosystem processes occurring within of physical conditions, loss of source pools of their natural ranges of variation. Guilds of species native species, substantive alteration of natural that are sensitive to human activities, such as ecological processes, or an inability to eradicate or top predators, the larger primates, and ground- control aggressive alien species. dwelling birds, occur at densities within the natu- * Critical. The remaining intact habitat is restricted ral range of variation. Biota move and disperse to isolated small fragments with low probabilities naturally within the ecoregion. Ecological pro- of persistence over the next 5-10 years without cesses fluctuate naturally throughout largely con- immediate or continuing protection and restora- tiguous natural habitats. tion. Many species are already extirpated or extinct due to the loss of viable habitat. Remaining habi- The criteria for classifying ecoregions were tailored to tat fragments do not meet the minimum area re- reflect biological and ecological differences among quirements for maintaining viable populations of METs. To produce the snapshot assessment, we many species and ecological processes. Land use in relied upon regional experts at the BSP workshop areas between remaining fragments is often in- (BSP/CI/TNC/WRI/WWF 1995), the mangrove compatible with maintaining most native species workshop, the WWF LAC program biogeographic and communities. Spread of alien species may be a priorities workshop (Wood and Cornelius 1994), the serious ecological problem, particularly on islands. World Bank, and elsewhere to assess the landscape- * Endangered. The remaining intact habitat is level criteria. Where data were up to date, we used restricted to isolated fragments of varying size overlays of remaining vegetation to assess land- (a few larger blocks may be present) with scape-level features for forested MHTs. We look medium to low probabilities of persistence over forward to other specialists updating and refining the next 10-15 years without immediate or con- this analysis as higher quality data become available. tinuing protection or restoration. Some species are already extirpated due to the loss of viable habi- Final Conservation Status tat. Remaining habitat fragments do not meet minimum area requirements for most species The "snapshot assessment" of conservation status populations and large-scale ecological processes. incorporated an estimation of threat in the sense that Land use in areas between remaining fragments is rate of conversion and degree of degradation and largely incompatible with maintaining most fragmentation were considered. However, one can native species and communities. Top predators easily identify other severe threats that are likely to are almost exterminated. affect the longer-term trajectory of conservation * Vulnerable. The remaining intact habitat occurs in efforts in a particular ecoregion. The type, scale, habitat blocks ranging from large to small; many intensity, and timeframe of threats (e.g., deforesta- intact clusters will likely persist over the next tion, mining, overgrazing, pollution, overharvesting 15-20 years, especially if given adequate protec- of wildlife) need to be assessed to make conservation tion and moderate restoration. In many areas, status a more reliable planning tool (refer to Appen- some sensitive or exploited species have been dix A for more details). Snapshot conservation status extirpated or are declining, particularly top and threat assessments-to give a final conservation predators and game species. Land use in areas status for each ecoregion-were analyzed by the between remaining fragments is sometimes com- same organizations and individuals. patible with maintaining most native species and communities. Biological Distinctiveness * Relatively stable. Natural communities have been altered in certain areas, causing local declines in Assessing the relative biological importance of exploited populations and disruption of ecosys- ecoregions must be part of any comprehensive tem processes. These disturbed areas can be exten- priority-setting exercise. For this study, we interpret Approach 9 the biological importance of ecoregions as the degree in the world (e.g., Mediterranean scrubs), to which its biodiversity (both components and Regionally Outstanding if less than three processes) is distinctive at different biogeographic occur regionally, and Bioregionally Outstand- scales. We define this scale-dependent assessment as ing if it is the only example of its MHT in its the biological distinctiveness of an ecoregion. All bioregion (e.g., ParaguanS restingas, which are ecoregions are biologically distinct to some degree, the only example of coastal dune formations particularly at the level of species and species in northern South America). assemblages and at increasingly broader biogeo- 6. Size of the ecoregion. The largest example(s) graphic scales. However, some ecoregions are so of an MHT at global and regional scales are exceptionally rich, complex, or unusual that they characterized as outstanding because they merit extra attention from conservation planners. maintain biodiversity processes and compo- In order to ensure appropriate biodiversity com- nents characteristic of that habitat type, but are parisons, the biological distinctiveness of an ecore- not always represented in smaller units. For gion is only assessed within its MHT (e.g., montane example, the Pantanal is rated Globally Out- grasslands were compared only to other ecoregions standing because it is one of the world's larg- of this MHT). Four categories of biological distinct- est seasonally flooded freshwater complexes. iveness at different biogeographic scales are used in this study: Globally Outstanding, Regionally Out- The biological distinctiveness assessments used in standing, Bioregionally Outstanding, and Locally this study represent a first attempt to categorize Important. Ecoregions are classified as outstanding ecoregions in this way. Rather than give quantitative if- for the biogeographic scale being considered - weightings to the criteria described above, we they have extraordinary levels of the attributes sought a consensus of opinion from experts.3 We described in either criteria 1, 2, 3, or 4, or are evalu- recognize that measuring and assigning relative val- ated as meeting either criteria 5 or 6 described below: ues to such a complex ecoregion attribute requires a number of subjective assessments, a task made even 1. Species richness with an emphasis on the more challenging by the incompleteness of biodi- following taxa: plants, mammals, birds, rep- versity data for many regions and taxa, by the fact tiles, amphibians, and butterflies.2 For exam- that available data are not now systematically and ple, the Napo region of western Amazonia is comprehensively organized, and by the lack of considered a Globally Outstanding ecoregion global ecoregion maps of comparable scale and because of having the world's highest known classification (like the one developed for LAC in this alpha diversity for many taxa. study). 2. Endemism (i.e., the number and proportion of However, we believe that the conservation com- species occurring only in that ecoregion) with munity has access to sufficient information on conti- an emphasis on the same taxa as species nental patterns of biodiversity (through expert richness. opinion and the technical literature) to identify 3. Complexity of species distributions within the ecoregions that are exceptionally distinctive at ecoregion (e.g., beta diversity, gamma diver- global, regional, and bioregional scales. As new data sity at larger scales, and patterns of local become available, some ecoregions might shift up or endemism). down one level. In general, we have most confidence 4. Uniqueness and rarity of certain ecological in our classification of Globally Outstanding or Re- phenomena in terms of their structure or dyna- gionally Outstanding ecoregions. mic properties. For example, on a global scale, Several MHTs were particularly challenging to as- Varzea forests, with their remarkable seasonal sess because detailed biodiversity information or migrations of fish to flooded forests, and the regional experts were unavailable during this study. Galapagos Islands, a xeric scrub ecoregion re- Specifically, some tropical dry broadleaf forests, nowned for extraordinary adaptive radiations, flooded grasslands, montane grasslands, and deserts are classified as Globally Outstanding. 5. Number of ecoregions in the same MHT. An 3. At the BSP workshop (BSP/CI/TNC/WRI/WWF 1995), ecoregion is considered Globally Outstanding regional habitat units-essentially aggregations of our if less than seven ecoregions in its MHT exist ecoregions-were ranked as being of high, medium, or low biological importance, based primarily on species richness and endemism but also, depending on the taxa, taking into account phyletic diversity and rare and endangered species. 2. These taxa are the ones primarily assessed by the experts at To determine our own biological distinctiveness rankings, the BSP workshop, but many other taxa were considered we reanalyzed the BSP workshop data with input from through further consultation with experts and review of the experts at the WWF-US LAC workshop (Wood and Cornelius technical literature. 1994) and at the World Bank. 10 A Conservation Assessment of the Terrestnal Ecoregions of Latin America and the Caribbean and xeric shrublands were difficult to evaluate. We logical distinctiveness. To achieve representation expect and look forward to biogeographers revising among all habitat types, a separate matrix is created ecoregion rankings, particularly as new biodiversity for each of the MHTs. The matrix allows us to information becomes available. classify each ecoregion into four biodiversity conser- All ecoregions not classified as outstanding are of vation priority categories (levels I-IV). local importance even though they are of average or The following three considerations determine the less than average biological distinctiveness when selection of the seven "level I" cells: compared with other ecoregions of the same MHT and bioregion. Nevertheless, it should be stressed * The relative rarity worldwide or the extraordinary that even Locally Important ecoregions harbor biodiversity found in ecoregions designated as globally unique biodiversity, maintain important Globally Outstanding warrant their inclusion as ecosystem services, and are useful to local human "level I." Exceptions are ecoregions classified as communities (Woodwell 1995). Such ecoregions Relatively Intact, which are ranked as "level II." should figure prominently in national or subregional These ecoregions are assumed to be under much biodiversity strategies. less threat over the next several decades and do not warrant proportionately greater attention at Biodiversity Conservation Priority this juncture. The certainty of severe and immediate loss of Conservation status and biological distinctiveness biodiversity in Critical and Endangered eco- are two essential discriminators for conservation regions classified as Globally Outstanding or priority-setting at regional scales (ecological func- Regionally Outstanding. Critical and Endangered tion, a third important discriminator, is most effec- ecoregions that are Bioregionally Outstanding or tively applied in finer-scale analyses). As a practical Locally Important are certainly important and tool for conservation planning, biological distinctive- contain unique species and communities; ness assesses the relative rarity of different natural however, the integration model proposed here communities and phenomena that, in conjunction emphasizes that proportionately more attention with other parameters, can be used to estimate the should be given to ecoregions with the same extent of opportunity for their conservation. Conser- conservation status but with higher biological vation status represents an estimate of the current distinctiveness. ability of an ecoregion to maintain viable species * The opportunity to conserve biologically out- populations, sustain ecological processes, and be standing ecoregions that are on the precipice of responsive to short- and long-term change-three major decline calls for inclusion of the top two basic goals of biodiversity conservation. Moreover, Vulnerable cells. They are included to help guard patterns of biodiversity are fixed and large-scale against Vulnerable ecoregions slipping into Criti- landscape changes are largely irreversible within the cal or Endangered status. timeframe of current conservation efforts. With the exception of some mangrove systems and a few We recognize that alternative integration models other habitats, restoration of ecosystems is either may prioritize ecoregion attributes that are different impossible, too costly, or too slow to offer much than those used in our model (e.g., all Critical scope for conservation investment. Similarly, exter- ecoregions) or may introduce new variables or minated species can never be recreated. The popula- modifiers such as the importance of different ecore- tions they comprise, or the assemblages, communi- gions in maintaining large-scale ecological functions ties, and ecological processes they are a part of, are such as global carbon sequestration or watershed also difficult and costly to replace. In contrast, insti- capacity. The relative importance of different para- tutional, economic, social, and political discrimina- meters may vary depending upon the spatial and tors show much more fluidity (e.g. reversibility) and temporal scale of conservation investments (e.g., are more subject to change within relevant time urgent measures versus strategic conservation plan- scales. ning) and the kinds of activities under consideration. The Integration Matrix Ensuring Bioregional Representation We offer a simple integration matrix to help identify The matrix presented on the following page provides, priority ecoregions for biodiversity conservation by MHT, a set of Highest Priority at Regional Scale (Table 1-1). Along the horizontal axis, we arrange ecoregions (level I) for all of LAC. However, this list ecoregions by their final conservation status. Along within each MHT does not necessarily ensure the vertical axis, we classify ecoregions by their bio- bioregional representation. Approach 11 Table 1-1. Matrix for Integrating Biological Distinctiveness and Conservation Status to Assign Priorities for Biodiversity Conservation Final Conservation Status Biological Relatively Relatively Distinctiveness Critical Endangered Vulnerable Stable Intact Globally II Outstanding Regionally III Outstanding Bioregionally 1 II III III IV Outstanding Locally LoalyIII III I V IV IV Important Note: The roman numerals indicate biodiversity conservation priority classes: Level I = Highest Priority at Regional Scale (shaded area) Level II = High Priority at Regional Scale Level III Moderate Priority at Regional Scale Level IV Important at National Scale Bioregional representation is important because of Priority at Regional Scale to reflect the importance, the dissimilarities in the species and natural at a LAC scale, of the biodiversity of that bioregion. communities of ecoregions within the same MHT We propose that within each bioregion each but from different biogeographic zones of LAC; constituent MHT be represented -subject to certain geographic replacement of species occurs as range conditions-by at least one ecoregion classed as boundaries, environmental gradients, and physical Highest Priority at Regional Scale. We attain this barriers are crossed at large geographic scales. By goal by "upgrading," when necessary, a single level including bioregional representation as a goal in a II or level III ecoregion (but never a level IV) to regional biodiversity strategy, problems of scale are "level la." We choose a level II over a level III also dealt with in a more objective fashion. For ecoregion and a worse conservation status over a example, many Caribbean ecoregions that share the better conservation status. Only level III ecoregions same MHT with continental ecoregions come out that are Critical, Endangered, or Vulnerable are poorly in priority rankings. Typically, small island considered eligible. If there is a tie between two or faunas and floras are depauperate compared to more ecoregions, we choose the ecoregion with the continental biotas because of size and isolation. highest biological distinctiveness (either as ranked in Ideally however, at least one example of each MI-IT Appendix E or, if ranked similarly, according to our in the Caribbean should be considered as Highest judgment). 2 Major Ecosystem Types, Major Habitat Types, and Ecoregions of LAC One of the major stumbling blocks to creating a rig- in shades of green. Mangrove units are pink. Ecore- orous framework for setting priorities at the regional gions on the maps represent estimates of the level has been the absence of a widely accepted "original" pre-Colombian distribution of habitats classification scheme of biogeographic units. In its and thus cover the entire LAC region. absence, donors have often set priorities using politi- cal rather than biogeographic boundaries, even Major Ecosystem Types (METs) though the two bear little overlap. When donors use country boundaries they seldom consider the goal of The first level of the hierarchy is the major ecosys- maintaining representation of all ecosystem types tem type (MET). We identify five METs: tropical and instead largely rely on methods based on spe- broadleaf forests; conifer/temperate broadleaf forests; cies richness by country or region. The emphasis on grasslands/savannas/shrublands; xeric formations; species richness as an indicator of priority ecore- and mangroves (Figure 1-2). Ecoregions in a given gions has skewed interest to tropical moist broadleaf MET tend to share (a) minimum area requirements forests and caused us to neglect the diverse ecosys- for conservation of ecological processes and other tems and biota found in the drier, non-forested or components of biodiversity; (b) response character- semi-forested ecoregions of LAC (Redford et al. istics to major disturbance; and (c) similar levels of 1990). beta diversity (the rate of turnover of species along Ignoring the goal of representation has also elevational or environmental gradients). Thus, METs prompted some biologically misleading questions, are defined primarily on dynamic properties and such as, Which are more important: the paramo of spatial patterns of biodiversity and not wholly on Colombia, the Atlantic forests of Brazil, or the dry vegetation structure. forests of Costa Rica? The question is inappropriate Tropical broadleaf forests cover the largest areas because it compares ecoregions representing of LAC (9.29 million kM2), followed by grass- markedly different habitat types. More appropriate lands/ savannas/ shrublands (7.13 million kM2), questions would be, Among the montane grasslands xeric formations (1.85 million kM2), conifer/ of Central and South America, which are the most temperate broadleaf forests (1.09 million kM2), and threatened and biologically important?, or Among mangroves (40,623 km2). We do not dwell on METs the tropical moist broadleaf forests of South in this report because they are too coarse for regional America, which are the most intact or threatened? conservation planning. The next hierarchical level is To answer such questions requires a hierarchy that of the major habitat type (MHT). For each MET based on habitat types. This chapter and the maps at and MHT, defining characteristics, sensitivity to the back of this book present such a hierarchy. On disturbance, and general conservation guidelines are the maps, we associate colors with habitat types. For described in Appendix C. example, xeric formations are coded in shades of red, grasslands/savannas/shrublands in shades of Major Habitat Types (MHTs) yellows and orange, tropical dry broadleaf forests in earth tones, conifer/temperate broadleaf forests in MHTs represent habitat types that are similar in shades of blue, and tropical moist broadleaf forests terms of their general structure, climatic regimes, 12 Major Ecosystem Types, Major Habitat Types, and Ecoregions of LAC 13 Table 2-1. Important Attributes of the Major Habitat Types of Latin America and the Caribbean Mean Total Size Percent of Number of Ecoregion Major Ecosystem Type Major Habitat Type (km12) LAC Ecoregions Size (km) Tropical Tropical moist broadleaf forests broadleaf forests 8,214,285 38.0 55 149,351 Tropical dry broadleaf forests 1,043,449 4.8 31 33,660 Conifer/temperate Temperate forests 332,305 1.5 3 110,768 broadleaf forests Tropical and subtropical coniferous forests 770,894 3.6 16 48,181 Grasslands/savannas/ Grasslands, savannas, shrublands and shrublands 7,058,529 32.7 16 441,158 Flooded grasslands 285,530 1.3 13 21,964 Montane grasslands 1,416,682 6.6 12 118,057 Xeric formations Mediterranean scrub 168,746 0.8 2 84,373 Deserts and xeric shrublands 2,276,136 10.5 27 84,301 Restingas 34,975 0.2 3 11,658 Mangroves Mangroves 40,623 0.2 [See Chapter 6] major ecological processes, and level of species turn- South America. Tropical dry broadleaf forests are over with distance (beta diversity), with flora and the second most numerous set of ecoregions, some fauna showing similar guild structures and life his- of which are among the smallest. tories. Nested within the 5 METs are 11 MHTs * Temperate forests are restricted to the Southern (Table 2-1); tropical moist broadleaf forests are most cone of South America and include one of the extensive and several xeric formation MHTs (e.g., world's seven temperate rain forests. Temperate restingas) are the smallest (Maps 2a-c). forest ecoregions are rare in LAC but quite large General distribution of these MHTs is as follows: in mean size. Tropical and subtropical coniferous forests are * Tropical moist broadleaf forests dominate in much widely dispersed. The greatest concentrations of southern Mexico and Central America, north- occur in central and southern Mexico, northern ern and northwestern South America, along the Central America, southern Brazil, and the Greater middle and lower slopes of the northern and cen- Antilles. tral Andes, and along a narrow belt of eastern * Grasslands, savannas, and shrublands are most Brazil (the Atlantic moist forests). The largest conspicuous in the southern cone region of South areas of tropical moist broadleaf forests in the America, Mexico, and the Llanos of Venezuela Caribbean occur in the Greater Antilles, particularly and Colombia. This is the second largest MHT in on Hispaniola and Cuba. Some experts further LAC, largely because we classified the Cerrado subdivide this MHT into lowland and montane and the Chaco as savannas rather than as dry forests. forests with a demarcation at about 1,500 m. This * Flooded grasslands are most prominent in Ama- is by far the largest MHT in original extent, and it zonia and central South America. contains the highest number of ecoregions as well * Montane grasslands are found primarily along the as some of the largest. Andes. From a global perspective, the LAC region * Tropical dry broadleaf forests predominate in contains the vast majority of the world's paramo- southern Mexico, parts of Pacific Central America, type vegetation - a rare ecoregion type. This cate- the Greater Antilles, central and eastern Brazil, gory also includes cold temperate grasslands such eastern Bolivia, and along parts of northwestern as the Patagonian steppe. 14 A Conservation Assessment of the Terrestrial Ecoregions of Latin Amenrca and the Canbbcan . The Mediterranean scrub MHT shows a disjunct shown in the ecoregion maps provided (Map 3 and distribution, with one unit in northwestern large-format map) represent the estimated original Mexico/southwestern USA and the other in Chile. extent of each unit of MHT. The ecoregion is the Although small in overall area, these units principal unit we use to analyze priorities. The deri- are of global importance because only five vation of these units is discussed in greater detail Mediterranean-type ecoregions occur in the below. world, and two are in LAC. An ecoregion represents a geographically distinct * Deserts and xeric shrublands are most extensive assemblage of natural communities that share a in Northern Mexico, parts of the Greater and large majority of their species, ecological dynamics, Lesser Antilles, Colombia, Venezuela, and Pacific and similar environmental conditions, and whose coastal South America (Peru, Chile). ecological interactions are critical for their long-term * Restingas and coastal dune formations are largely persistence. On the basis of this definition, some restricted to Brazil and Argentina, though to a ecoregions that contain a mosaic of distinct habitat lesser degree they are also found in Venezuela types (e.g., the Cerrado of Brazil with its grasslands, and Mexico. gallery forests, dry forests, and open woodland) are * Mangroves are limited to the tropical zone along considered as a single ecological unit. We assume the coasts of continental landmasses or on islands that conversion or degradation of any major portion of the Caribbean and Pacific (Map 4). Mangroves of an ecoregion would alter the ecological processes were classified into 13 major mangrove complexes and species population dynamics of remaining areas, and further subdivided into 40 mangrove units. but have relatively less impact outside the ecore- Mangroves were classified into these units based gion's boundaries. on hydrographic, geomorphic, and ecological Delineating the terrestrial ecoregions of LAC features. proved to be an ambitious two-year undertaking. The original extent of LAC ecoregions was deline- Ecoregions ated on a base map using the Digital Chart of the World, and then digitized and organized in a geo- Ecoregions represent geographically discrete units graphic information system (GIS) database using of major habitat types. The ecoregion boundaries ARC/INFO software. For more details on delinea- Figure 2-1: Number of Ecoregions by Size Categories 35- 32 3028 25 20 0 20- 18 o 15 14 S- 13 13 10- 5 31-1,000 1,00- 3 10,001- 30,001- 50,001- 100,001- 200,001- 400,001- 3,000 10,000 30,000 50,000 100,000 200,000 400,000 2,000,000 Ecoregion Size Category (kin2) Major Ecosystem Types, Major Habitat Types, and Ecoregions of LAC 15 tion of ecoregions, descriptions, and references, see ecoregions we emphasized the habitat that plays Appendix F. the most important role in maintaining ecological In most cases, the ecoregions are based on classifi- processes. cations from existing ecological studies in order to Ecoregions range in size from the Cerrado of (a) enable this conservation assessment to benefit Brazil at nearly 2 million km2 to the Costa Rican from the wealth of scholarly work that has been car- paramo of only 31 km2 (Appendix E). A breakdown ried out in the region, and (b) to enhance the efficacy by size class (Figure 2-1) shows that ecoregions vary of conservation investments by working with classi- enormously in size. Among the non-mangrove fication systems that are widely used within the re- ecoregions, only 6.7 percent are less than 1,000 km2, gion. However, differences among regional systems most of which are island units in the Caribbean are common, and we highlight here some of the deci- bioregion. Ecoregions in the two smallest size classes sion rules we established for delineating boundaries. are about equal in number to those in the two largest Some ecoregions are composed of several disjunct size classes. Among the four forested MHTs (tropical areas because of bioclimatic factors, soils, hydro- moist broadleaf forests, tropical dry broadleaf for- graphic conditions, or isolation by physical barriers ests, temperate forests, and tropical and subtropical such as large rivers or mountain ranges. Such for- coniferous forests), tropical dry broadleaf forests are mations are considered as a single unit for the pur- clustered around the smallest size categories and poses of this analysis if they are geographically temperate forests-except for the Chilean winter- clustered (e.g., Tepuis) or maintain some level of rain forests-among the larger units. Grass- biotic interaction among the habitat blocks. Some lands/savannas/shrublands units are widely scat- ecoregions and habitat types contain a continuum of tered. Flooded grasslands include the Pantanal-at ecological communities and classifying some of these nearly 141,000 km2 one of the world's largest wet- systems was sometimes problematic. Examples include lands-and some much smaller units elsewhere in several partially wooded systems such as grasslands LAC. Other wetland units have been identified by that contain smaller amounts of gallery forests, savan- Wetlands for the Americas and will be described in a nas, woodlands, and scrublands. In classifying these forthcoming publication. 3 Conservation Status of Terrestrial Ecoregions of LAC Conservation of biodiversity is important every- The original extent of ecoregions map (Map 3 and where. However, in those ecoregions where human large format map) illustrates what LAC probably activities have caused widespread habitat destruc- looked like in pre-Colombian time (recognizing tion, the costs and level of effort to conserve biodi- that some pre-Colombian landscapes were already versity will be far higher than in more intact anthropogenically modified to some degree). The ecoregions. Some ecoregions are going to require snapshot conservation status assessment (Map 5) the immediate implementation of recovery plans if (i.e., an assessment based on current habitat configu- most of the plant and animal communities they rations) provides insights into how seriously some contain are to persist into the next decade. Other ecoregions have been degraded. The presence of ecoregions will remain relatively stable over the large blocks of original habitat, percent of next few decades. remaining habitat, and degree of protection high- Determining which ecoregions belong on a critical light opportunities for conservation within the list requires weighing a number of factors, including ecoregion. Combined with rate of conversion and their natural resiliency to major disturbance, their degree of degradation and fragmentation, these degree of beta diversity, the configuration of the variables also help predict, from a biological per- remaining habitat, the degree of protection, and the spective, the maintenance of ecological processes type, intensity, and timeframe of major threats to (e.g., predation, pollination and seed dispersal sys- biodiversity. For example, in most situations man- tems, nutrient cycling, migration, dispersal, and groves regenerate far more quickly from habitat gene flow) that ultimately determine how much disturbance than do tropical moist forests. Highly biodiversity will persist over the long term. fragmented landscapes will normally be more The final conservation status-the snapshot threatened than those that contain large blocks of assessment of conservation status modified by threat original habitat. Intensive grazing may be a less (Map 6)-forecasts an ecoregion's trajectory given severe threat in grasslands than in dry forests. current trends over the next 5 to 20 years. The threat Ecoregions characterized by high beta diversity assessment is essential because the snapshot assess- may require more protected areas that are well- ment, although valuable, only shows what presently distributed across the landscape to conserve the remains. The threat assessment takes into account full complement of endemic plants and animals. factors that we know or assume will impact biodi- We provide an analytical framework for assessing versity conservation in an ecoregion. For example, the conservation status of ecoregions that is tailored using our methodology, the snapshot assessment of to the particular dynamics of their respective major the Guianan moist forests (tropical moist broadleaf ecosystem type. A more detailed account of the forest MHT) indicates it is Relatively Intact. How- methods we employed to evaluate conservation ever, foreign timber concessions have already staked status is found in Appendix A, which also includes a claim to large tracts of this ecoregion. With wide- more thorough discussion of the landscape-level spread logging and road building into pristine areas features assessed in this methodology and a ration- about to begin-and likely resulting in extensive ale for the weighting scheme. habitat destruction and degradation-the snapshot 16 Conservation Status of Terrestrial Ecoregions of LAC 17 assessment is changed to Relatively Stable. The final Critical and Endangered ecoregions are overrepre- conservation status is used for integration with sented in the tropical dry broadleaf forests, and in biological distinctiveness to determine conservation deserts and xeric shrublands (Table 3-1). By contrast, priorities (Chapter 5). tropical moist broadleaf forests spanned a wide We begin by presenting the results of the snapshot range but contained significantly more Relatively conservation status of ecoregions. We then show Stable and Relatively Intact ecoregions than did how a threat analysis modifies the status, and we tropical dry broadleaf forests. The Relatively Stable make comparisons within and across MHTs. Within and Relatively Intact tropical moist broadleaf forests an MHT, we identify the most critical ecoregions. include many of the largest ecoregions. From a Comparisons across MHTs allow us to determine if broad bioregional perspective, Amazonia contains one MHT is more threatened than another. Finally, some of the most intact ecoregions. The other eight we highlight major trends in relationships between bioregions show about the same distribution of the size of ecoregions, the threats they face, and their status classes. biogeographic location. Three ecoregions (Tehuantepec savannas, Veracruz palm savannas, and Veracruz pine-oak forests), Results either partly or entirely in Mexico, were classified as Extinct by some workshop experts but as Critical by Snapshot Conservation Status others. We chose to designate all of these ecoregions as Critical for the final assessment. The most striking observation from the snapshot conservation status map of ecoregions (Map 5) is Final Conservation Status how few ecoregions are designated as either Rela- tively Stable (n=37; 22 percent) or as Relatively Intact Final conservation status was derived by applying a (n=14; 8 percent). Essentially, much of the area in threat analysis to the snapshot conservation status these two categories is located in the Amazon basin, (half of the 12 previously unclassified ecoregions the moist forests of the Pet6n (Mexico, Guatemala), were also classified at this stage). Thirty-nine ecore- and northwestern Mexican xeric systems. Twenty- gions were considered more threatened after the three ecoregions (14 percent) are Critical, 47 (28 per- assessment of threat was applied to the snapshot con- cent) are Endangered, and 45 (27 percent) are Vulner- servation status. When analyzed by major habitat able. Twelve ecoregions were not classified for lack of type, these break down to 23 (42 percent of the reliable information. MHT) tropical moist broadleaf forest ecoregions; Table 3-1. Snapshot Conservation Status of Ecoregions by Major Habitat Type Snapshot Conservation Status Relatively Relatively Major Habitat Type Critical Endangered Vulnerable Stable Intact Unclassified Tropical moist broad- leaf forests 3 13 16 13 10 Tropical dry broadleaf forests 10 14 4 2 1 Temperate forests 1 1 1 Tropical and subtropical coniferous forests 1 5 5 4 1 Grasslands, savannas, and shrublands 2 1 6 5 2 Flooded grasslands 2 3 1 3 1 3 Montane grasslands 5 6 1 Mediterranean scrub 2 Deserts and xeric shrublands 3 7 7 3 2 5 Restingas 2 1 TOTAL 23 47 45 37 14 12 18 A Conservation Assessment of the Terrestnal Ecoregions of Latin America and the Canbbean five (16 percent) tropical dry broadleaf forest ecore- The final conservation status of ecoregions gions; one (33 percent) temperate forest ecoregion; revealed that 48 percent of all ecoregions are either two (13 percent) tropical and subtropical coniferous Critical (18 percent) or Endangered (30 percent). forest ecoregions; two (13 percent) grassland, savanna, Thirty-two percent are Vulnerable, 16 percent are and shrubland ecoregions; one (8 percent) flooded Relatively Stable, and 5 percent are Relatively Intact grassland ecoregion; three (25 percent) montane (Table 3-2; Map 6). There are six ecoregions for grassland ecoregions; one (50 percent) Mediterra- which a final assessment could not be made because nean scrub ecoregion; and one (4 percent) desert and data were lacking. These are primarily xeric ecore- xeric shrubland ecoregion. Only one ecoregion gions of northern Mexico, which are currently being (Puerto Rican moist forests) was considered less assessed under the Mexico Country Study for the threatened after the assessment of threat was applied Biodiversity Convention. to the snapshot conservation status because of ex- Numerically, the highest number of Critical and pected regeneration of natural forest. Endangered ecoregions occurred in tropical moist Table 3-2. Final Conservation Status of Ecoregions by Major Habitat Type Final Conservation Status Relatively Relatively Major Habitat Type Critical Endangered Vulnerable Stable Intact Unclassified Tropical moist broad- leaf forests 6 15 19 11 4 Tropical dry broadleaf forests 11 17 2 1 Temperate forests 1 2 Tropical and subtropical coniferous forests 3 3 5 4 1 Grasslands, savannas, and shrublands 2 2 6 4 2 Flooded grasslands 3 4 3 2 1 Montane grasslands 9 3 Mediterranean scrub 1 1 Deserts and xeric shrublands 3 7 9 2 2 4 Restingas 2 1 TOTAL 31 51 55 27 8 6 Table 3-3. Final Conservation Status by Size of Ecoregion Final Conservation Status Size of Ecoregion Relatively Relatively (n=number of ecoregions) Critical Endangered Vulnerable Stable Intact Unclassified 31-1,000 km2 (n=15) 3 5 3 3 1 1,001-3,000 km2 (n=13) 3 3 4 1 2 3,001-10,000 km2 (n=32) 8 9 10 4 1 10,001-30,000 km2 (n=28) 6 11 5 3 3 30,001-50,000 km2 (n=18) 6 7 3 1 1 50,001-100,000 km2 (n=20) 2 7 9 1 1 100,001-200,000 km2 (n=25) 1 5 13 5 1 200,001-400,000 km2 (n=14) 2 2 4 4 2 400,001-2,000,000 km2 (n=13) 2 7 3 1 Conservation Status of Terrestrial Ecoregions of LAC 19 broadleaf forests and tropical dry broadleaf for- original extent of an ecoregion and its final con- ests (Table 3-2). However, only 3 percent of the servation status (Table 3-3). tropical dry broadleaf forest ecoregions were As a means to test the results of the threat- Relatively Stable or Relatively Intact, whereas modified assessment, we asked the participants in 27 percent of the tropical moist broadleaf forests the WWF LAC strategic planning workshop to were Relatively Stable or Relatively Intact. From repeat the threat-modified assessment without these data, we can conclude that tropical dry having access to the BSP assessments. Overall the broadleaf forests are on average more threatened two workshops yielded similar rankings, with the than either tropical moist broadleaf forests or WWF LAC team reaching different conclusions from conifer/temperate broadleaf forests in the LAC the BSP workshop for only 26 of the 178 non- region. The restingas MHT contains the most mangrove ecoregions (14 percent). threatened ecoregions among MHTs, followed by Among the nine bioregions, the Northern Andes the Mediterranean scrub MHT. Most xeric forma- had the highest percentage (69 percent) of ecore- tion ecoregions were either Critical, Endangered, gions clustered in the two most threatened catego- or Vulnerable. Grassland, savanna, and shrubland ries (Table 3-4). Central America (67 percent), the ecoregions illustrated the widest spread in con- Caribbean (57 percent), Eastern South America servation status. In general, montane grasslands (50 percent), and Northern Mexico (39 percent) also are less threatened than are lowland grasslands. had significant proportions of their ecoregions in the There was no clear relationship between size of most threatened categories. Table 3-4. Final Conservation Status by Bioregion Final Conservation Status Relatively Relatively Bioregion Critical Endangered Vulnerable Stable Intact Unclassified Northern Mexico 3 4 4 4 3 6 Central America 11 11 5 5 1 Caribbean 2 11 9 1 Orinoco 4 3 3 2 Amazonia 1 3 7 11 2 Northern Andes 9 11 8 1 Central Andes 1 3 7 Eastern South America 4 2 6 Southern South America 2 6 2 4 Biological Distinctiveness of Terrestrial Ecoregions of LAC at Different Biogeographic Scales Ecoregions vary by the number of species they ecoregions to one of the four distinctiveness classes contain, by the level of endemism, and by the (see Chapter 1). The specific reasons for designating uniqueness of the assemblages, natural communi- ecoregions as either Globally or Regionally Out- ties, ecological interactions, and biological phenom- standing are presented in Appendix F. The biodi- ena found within them. In defining categories of versity of every ecoregion has unique elements and biological distinctiveness, we included the relative is ecologically important and this approach high- rarity of certain MHTs besides the more commonly lights the value of protecting each ecoregion in a used variables of species richness and endemism. national biodiversity strategy. From a regional per- For example, some MHTs are represented by only a spective, some ecoregions have a higher level of few ecoregions worldwide. Mediterranean scrub, biological distinctiveness than others, and they merit temperate rain forests, and paramo are three greater attention from conservation planners. such types that are rare globally but prominent in LAC. Results We also pay greater attention to a less-used but important criterion, the degree of beta diversity of Thirty-four ecoregions in LAC were deemed Glob- an ecoregion. Simply put, beta diversity is a measure ally Outstanding and 31 ecoregions were considered of the turnover of species with distance or along Regionally Outstanding (Map 7; Table 4-1). There environmental gradients. Several areas of LAC, such are 62 Bioregionally Outstanding ecoregions and 50 as the tropical Andes, contain areas of extraordinary Locally Important ecoregions. beta diversity, where major turnover of species as- semblages is common over a small elevational Globally and Regionally Outstanding Ecoregions change or across nearby mountain ranges. By con- trast, one can travel for hundreds of kilometers An analysis of the biological distinctiveness ratings within other ecoregions and find similar assem- of ecoregions by MHT shows that, numerically, blages of plants and animals. Globally Outstanding ecoregions are most com- The conservation of areas high in beta diversity, and monly found in tropical moist broadleaf forests how to predict their occurrence at finer geographic (Table 4-1). These ecoregions are concentrated in the scales is now a major new thrust of the science of western portion of Amazonia and the tropical conservation biology and, we predict, will soon Andes (Map 7). become an essential component in future priority- If this MHT were further subdivided into lowland setting exercises. and montane forests, the Globally Outstanding units We confine comparisons of biological distinctive- would be divided almost equally. The other exam- ness to ecoregions within the same MHT, link bio- ples are scattered among northwestern Mexico, the logical distinctiveness to a biogeographic spatial Atlantic forest in Brazil, and central and southern scale, and state explicitly the rules for assigning Chile. 20 Biological Distinctiveness of Terrestrial Ecoregions of LAC at Different BiogeographiC Scales 21 Table 4-1. Biological Distinctiveness of Ecoregions by Major Habitat Type Biological Distinctiveness Globally Regionally Bioregionally Locally Major Habitat Type Outstanding Outstanding Outstanding Important Tropical moist broadleaf forests 16 13 18 8 Tropical dry broadleaf forests 3 2 7 19 Temperate forests 1 1 1 Tropical and subtropical coniferous forests 2 6 4 4 Grasslands, savannas, and shrublands 1 1 9 5 Flooded grasslands 1 3 8 1 Montane grasslands 4 4 3 1 Mediterranean scrub 2 Deserts and xeric shrublands 2 1 12 12 Restingas 2 1 TOTAL 34 31 63 50 In addition, tropical moist broadleaf forests contain MHT). Tropical dry broadleaf forests; grasslands, 42 percent of the Regionally Outstanding ecoregions. savannas, and shrublands; and deserts and xeric Among the other MHTs, montane grasslands have the shrublands had the lowest proportion of Globally or highest proportion (67 percent of the MHT) of Glob- Regionally Outstanding ecoregions, but all MHTs ally or Regionally Outstanding ecoregions, followed were represented by at least one ecoregion classified by tropical moist broadleaf forests (56 percent of the as either Globally or Regionally Outstanding. 5 Integrating Biological Distinctiveness and Conservation Status The integration of the conservation status of ecore- * Bolivian montane dry forests (II, Critical) gions with data on biological distinctiveness offers a * Brazilian Araucaria forests (II, Critical) powerful tool for determining the relative conserva- * Tabasco/Veracruz savannas (III, Critical) tion importance of different ecoregions within a * Pampas (II, Endangered) major habitat type. In this chapter, we present the * Beni savannas (II, Endangered) results of the integration exercise; the model and * Jalisco palm savannas (II, Critical) decision rules are described in Chapter 1. * Guayaquil flooded grasslands (II, Endangered) * Eastern Amazonian flooded grasslands (III, Results Vulnerable)2 * Mexican alpine tundra (III, Vulnerable) The classification of ecoregions as level I, II, III, or IV * Leeward Islands xeric scrub (III, Critical) is presented visually in Map 8 and is summarized in * Pueblan xeric scrub (II, Critical) Tables 5-1 and 5-2. Note that the red areas in Map 8 . Araya and Paria xeric scrub (II, Endangered) (level I) cover large areas that may contain very lim- * Sechura desert (III, Vulnerable) ited intact habitats. With the exception of some . Caatinga (III, Vulnerable) western Amazonian units, remaining natural habi- * Paraguana restingas (II, Endangered) tats in most ecoregions classified as Highest Priority at Regional Scale are only a fraction of the area Ecoregions of Highest Biodiversity Conservation shown in red. Tables 5-3 to 5-12 present the results Priority by MHT by MHT. Following the rules presented in Chapter 1, Among the 55 Tropical moist broadleaf forest 19 ecoregions were upgraded to the Highest Priority ecoregions, 23 are classified as Highest Priority at at Regional Scale ranking to ensure bioregional rep- Regional Scale, numerically the most among any resentation. We designate these ecoregions as level Ia MHT (Tables 5-1 and 5-3). One additional level P ecoregions (Map 9). In the following list, the former ecoregion was selected in Central America. Atten- biodiversity conservation priority level and final tion should be given to two prominent level II conservation status are given in parentheses:' ecoregions, the Tepuis of the Orinoco bioregion and the Japura/Negro moist forests of Amazonia, which * Tehuantepec moist forests (II, Endangered) are classified as Globally Outstanding and Relatively * Cuban dry forests (II, Endangered) Intact. These are the only ecoregions in Latin America * Tamaulipas/Veracruz dry forests (III, Endangered) and the Caribbean with this combination of features. * Llanos dry forests (III, Endangered) 2. According to the suggested rules for upgrading ecoregions, 1. Either the Eastern Mexican grasslands or the Central Mexican the Siio Luis flooded grasslands should have been chosen in grasslands should probably be upgraded as a representative place of the Eastern Amazonian flooded grasslands. We sub- of the grasslands/savannas/shrublands MHT in the Northern stitute the latter, however, because of some doubts as to Mexico bioregion. Both are, however, unclassified because of whether or not the former area should even be designated as lack of information. a distinct ecoregion. 22 Integrating Biological Distinctiveness and Conservation Status 23 Table 5-1. Conservation Importance of Ecoregions by Major Habitat Type Number of Ecoregions in Each Conservation Priority Class Major Habitat Type Level I Level 11 Level III Level IV Unclassified Tropical moist broadleaf forests 23 10 18 4 Tropical dry broadleaf forests 5 6 18 2 Temperate forests 2 1 Tropical and subtropical coniferous forests 5 5 3 3 Grasslands, savannas, and shrublands 2 2 8 2 2 Flooded grasslands 4 4 4 1 Montane grasslands 8 3 1 Mediterranean scrub 2 Deserts and xeric shrublands 2 6 9 6 4 Restingas 2 1 TOTAL 55 34 64 19 6 Tropical dry broadleaf forests contain only three ing the Pantanal. To achieve better bioregional rep- Globally Outstanding ecoregions (Table 5-4). Among resentation requires adding three other ecoregions the 31 ecoregions, five are level I's, two of which (Table 5-8). occur in the Central American bioregion. To achieve Montane grasslands contain 12 ecoregions, four of better bioregional representation, four other ecore- them Globally Outstanding (Table 5-9). All four are gions are upgraded to a la status. disjunct examples of paramo, a globally rare montane There are only three ecoregions in the temperate for- community that is most extensive in South America. est MHT, all in the Southern South America bioregion Eight montane grassland ecoregions are Highest Pri- (Table 5-5). Two of the three are considered Highest ority at Regional Scale. The Mexican alpine tundra Priority at Regional Scale. No further ecoregions are receives high consideration as a la ecoregion. needed to achieve bioregional representation. Only two Mediterranean scrub ecoregions exist in Tropical and subtropical coniferous forests contain LAC, the Californian coastal sage-chaparral and the two Globally Outstanding ecoregions, the Sierra Chilean matorral (Table 5-10). Both are considered Madre del Sur pine-oak forests and the Sierra Madre Globally Outstanding and highly threatened and are Occidental pine-oak forests (Table 5-6). Five of the 16 therefore Highest Priority at Regional Scale. ecoregions are classified as level I, and all but one of Only two of the 27 desert and xeric shrubland the bioregions are represented in this category. ecoregions were selected as Highest Priority at Brazilian Araucaria forests are classed as P to achieve Regional Scale: Northern Sonoran cactus scrub and bioregional representation. The Miskito pine forests the Galapagos Islands xerics (Table 5-11). We were are noteworthy as a relatively rare large unit of unable to determine the conservation status of four tropical lowland pine forest. ecoregions in this MHT; all are northern or central Grasslands, savannas, and shrublands contain one Mexican xerics. To achieve better bioregional repre- Globally Outstanding ecoregion, the Cerrado sentation, five additional ecoregions are selected as (Table 5-7). Among the 16 ecoregions, only two are I ecoregions. Highest Priority at Regional Scale: the Cerrado and There are only three ecoregions in the restingas the Chaco savannas. To achieve better bioregional MHT; both of the Brazilian representatives are representation, three are added as P ecoregions: Globally Outstanding and Critical and thus Highest Tabasco/Veracruz savannas, Beni savannas (Ama- Priority at Regional Scale (Table 5-12). Paraguani zonia) and the Pampas. Two grassland ecoregions in restingas are added as a la ecoregion to ensure Mexico are unclassified. bioregional representation. Flooded grasslands contain four ecoregions To summarize, only 55 out of 178 ecoregions classed as Highest Priority at Regional Scale, includ- distributed among the 10 terrestrial MHTs were 24 A Conservation Assessment of the Terrestnal Ecoregions of Latin America and the Caribbean Table 5-2. Final Conservation Status and Biological Distinctiveness of All Non-Mangrove Ecoregions Final Conservation Status Relatively Relatively Biological Critical Endangered Vulnerable Stable Intact Unclassified Distinctiveness (31) (51) (55) (27) (8) (6) Globally I I I I II Outstanding (34) (9) (6) (12) (5) (2) Regionally I I I 1l III Outstanding (31) (2) (9) 114) | S) (2) Bioregionally II II III III IV Important (63) (7) (19) (19) (13) (2) (3) Locally III III IV IV IV Important (50) (13) (17) (11) (4) (2) (3) Note: The values in parentheses are the number of ecoregions in that category. The roman numerals indicate biodiversity conservation priority classes: Level I = Highest Priority at Regional Scale (shaded area) Level II= High Priority at Regional Scale Level III = Moderate Priority at Regional Scale Level IV = Important at National Scale designated as level I. Application of our method to Amazonia, were classified as Relatively Intact and achieve bioregional representation adds 19 addi- Globally Outstanding. The relative scarcity of tional ecoregions designated as level Ia. Globally or Regionally Outstanding ecoregions clas- Combining these 19 ecoregions with the ecore- sified as Relatively Stable or Relatively Intact is of gions classed as Highest Priority at Regional Scale regional concern (Table 5-2). These fourteen ecore- gives a total of 74 units, 42 percent of all of the non- gions include some Amazonian units that are ex- mangrove ecoregions. tremely large, but represent only 8 percent of the total number of ecoregions analyzed. Conversely, Major Trends the number of ecoregions that are either Globally or Regionally Outstanding and Critical (11) or Globally Only two ecoregions, the Tepuis of the Orinoco Outstanding and Endangered (6), account for only bioregion and the Japura/Negro moist forests of 10 percent of the total number of ecoregions. Table 5-3. Tropical Moist Broadleaf Forests: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * Cauca Valley montane * Northwestern Andean * Ucayali moist forests - * Napo moist forests -Peru, * Tepuis- Venezuela, Brazil, forests-Colombia montane forests- Brazil, Peru (Amazonia) Ecuador, Colombia Guyana, Suriname, Colombia Outstanding (N.Andes) Colombia, Ecuador * Varzea forests-Brazil, (Amazonia) (Amazonia) Magdalena Valley montane (N.Andes) Peru, Colombia * Western Amazonian swamp * Japura/Negro moist forests- Colombia * Venezuelan Andes (Amazonia) forests -Peru, Colombia forests - Colombia, (N.Andes) montane forests - * Choc6/DariCn moist (Amazonia) Venezuela, Brazil, Peru * Brazilian Coastal Atlantic Venezuela, Colombia forests - Colombia, * Southwestern Amazonian (Amazonia) forests- Brazil (N.Andes) Panama (N.Andes) moist forests-Brazil, Peru, (E.S.America) * Peruvian Yungas-Peru * Cordillera Oriental Bolivia (Amazonia) (C.Andes) montane forests - Colombia, Venezuela (N.Andes) * Eastern Cordillera Real montane forests -Ecuador, Colombia, Peru (N.Andes) Regionally * Western Ecuador moist * Hispaniolan moist forests - * Cuban moist forests - * Talamancan montane * Guianan Highlands moist forests- Ecuador, Colombia Haiti, Dominican Republic Cuba (Carib.) forests -Costa Rica, Panama forests -Venezuela, Brazil, Outstanding (N.Andes) (Carib.) * Cordillera La Costa (C.America) Guyana (Orinoco) * Jamaican moist forests- montane forests- * Jurna moist forests- Brazil Jamaica (Carib.) Venezuela (Orinoco) (Amazonia) * Bolivian Yungas -Bolivia, * Macarena montane Argentina (C.Andes) forests-Colombia * Brazilian Interior Atlantic (Amazonia) forests - Brazil, Argentina, * Rond6niafMato Grosso Paraguay, (ES.America) moist forests - Brazil, Bolivia (Amazonia) * Santa Marta montane forests-Colombia (N.Andes) (Table continues on the following page) 0l Table 5-3. Tropical Moist Broadleaf Forests: Integration Matrix of Biological Distinctiveness and Conservation Status (Continued) Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Bioregionally * Oaxacan moist forests- * Puerto Rican moist * Leeward Islands moist Mexico (C. America) forests -Puerto Rico forests -Leeward Islands OutstandingTehuantepec moist 1 (Carib.) (Carib.) forests - Mexico, I * Windward Islands moist * Orinoco Delta swamp Guatemala, BelizeI forests- Windward forests-Venezuela, Guyana (C.America) I Islands (Carib.) (Orinoco) * Sierra Madre moist * YucatAn moist forests - * Uatama moist forests - forests-Mexico, Mexico (C.America) Brazil, Venezuela, Guyana Guatemala, El Salvador * Central American Atlantic (Amazonia) (C.America) moist forests-Guatemala, * Amapa moist forests- * Central American montane Belize, Honduras, Brazil, Suriname (Amazonia) forests -Mexico, El Nicaragua, Costa Rica, * Guianan moist forests - Salvador, Guatemala, Panama (C. America) Venezuela, Guyana, Honduras (C.America) * Paramaribo swamp Suriname, Brazil, French * Isthmian-Pacific moist forests - Suriname Guiana (Amazonia) forests -Costa Rica, Panama (Amazonia) (C.America) * Eastern Panamanian * Magdalena/Uraba moist montane forests -Panama, forests -Colombia Colombia (N.Andes) (N.Andes) * Andean Yungas- Argentina, Bolivia (C.Andes) Locally * Costa Rican seasonal moist * Belizean swamp forests - * Trinidad & Tobago moist * Purus/Madeira moist forests-Costa Rica, Belize (C.America) forests-Trinidad & forests-Brazil (Amazonia) Nicaragua (C.America) * Tocantins moist forests - Tobago (Orinoco) * Beni swamp and gallery * Catatumbo moist forests - Brazil (Amazonia) * Tapaj6s/Xingu moist forests - Bolivia, Brazil Venezuela, Colombia forests- Brazil (Amazonia) (Amazonia) (N-Andes) Level 1: Highest Priority at Regional Scale for biodiversity conservation I I Level 1a: Highest Priority at Regional Scale (added to ensure bioregional representation) F-1 Table 5-4. Tropical Dry Broadleaf Forests: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * Bolivian lowland dry * Tumbe*Piura dry forests - Outstanding forests-Bolivia (Amazonia) Ecuador, Peru (N.Andes) * Ecuadorian dry forests- Ecuador (N.Andes) Regionally * Jalisco dry forests-Mexico Outstanding (C.America) * Balas" dry forests -Mexico (C.America) y * Central American Pacific uba y forests-Cuba I * Sinaloan dry forests- Outstandin dry forests -Mexico, (Carib.) Mexico (N.Mexico) O Guatemala, El Salvador, - - - --- - - Honduras, Nicaragua, Costa * Hispaniolan dry forests- Rica (C. America) Haiti, Dominican Republic - ------------- (Carib.) BBolivian montane dry forests-Bolivia (C.Andes) I * Oaxacan dry forests- - -------- -- -- Mexico (C.America) * MaraFi6n dry forests -Peru (N.Andes) Locally * Leeward Islands dzy * Jamaican dry forests- * Bahamian dry forests- * Baja California dry forests-Leeward Islands Jamaica (Carib.) Bahamas, Turks & Caicos forests-Mexico (N.Mexico) Important (Carib.) * Puerto Rican dry forests- Islands (Carib.) * Veracruz dry forests - Puerto Rico(Carib.) Mexico (C. America) * Cayman Islands dry * Panamanian dry forests - forests -Cayman Islands o Panama (C. America) (Carib.) 2 * Cauca Valley dry forests - * Windward Islands dry Colombia (N.Andes) forests- Windward Islands (Carb.) * Magdalena Valley dry an forests - Colombia TamaulipaqfVeracruzuy (N.Andes) forests-Mexico (.Mexico) * Patia Valley dry forests - Mand oe - Colombia (N.Andes) Mexico (C.America) * Sinii Valley dry forests- - -- Colombia (N.Andes) I* Ll-anos dry forests ----- Colobea(N-des-' Level I: Highest Priority at Regional Scale for Q * Trinidad & Tobago dry biodiversity conservation forests - Trinidad & Tobago (Orinoco)---- ( oc ' Level la: Highest Priority at Regional Scale * Maracaibo dry forests - ; (added to ensure bioregional representation) Venezuela (N.Andes) * Lara/Falc6n dry forests - Venezuela (N.Andes) N Table 5-5. Temperate Forests: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * Valdivian temperate Outstanding forests-Chile, Argentina (S.S.America) Regionally Chilean winter-rain forests - Chile Outstanding (S.S.America) Bioregionally * Subpolar Nothofagus forests -Chile, Argentina (S.S.America) Locally Important Li Level I: Highest Priority at Regional Scale for biodiversity conservation 10 Qr Table 5-6. Tropical and Subtropical Coniferous Forests: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Global * Sierra Madre del Sur pine- * Sierra Madre Occidental oak forests -Mexico pine-oak forests- Mexico, Outstanding (C.America) USA (N.Mexico) * Mexican transvolcanic * Cuban pine forests -Cuba * Sierra Madre Oriental pine-oak forests-Mexico (Carib.) pine-oak forests-Mexico Outstanding (C.Arnerica) (N.Mexico) * Belizean pine forests- Belize (C.America) * Miskito pine forests- Nicaragua, Honduras (C.America) * Brazilian Araucaria * Central Mexican pine-oak * Hispaniolan pine forests- * Veracruz montane forests- oregionay forests- Brazil, Argentina forests - Mexico (N.Mexico) Haiti, Dominican Republic Mexico (C.America) Outstanding I (Carib.) * Central American pine-oak forests - Mexico, Guatemala, Honduras, El Salvador, Nicaragua (C.America) Lol * Veracruz pine-oak * Bahamian pine forests- * San Lucan pine-oak ocaly forests- Mexico (N.Mexico) Bahamas, Turks & Caicos forests-Mexico (N.Mexico) Important Islands (Carib.) * Sierra Juarez pine-oak forests -Mexico, USA (N.Mexico) Level 1: Highest Priority at Regional Scale for biodiversity conservation I I Level la: Highest Priority at Regional Scale (added to ensure bioregional representation) FI-1- C 0 Table 5-7. Grasslands, Savannas, and Shrublands: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * Cerrado -Brazil, Paraguay, Outstanding Bolivia (ES.America) Regionally * Chaco savannas- Outstanding Argentina, Paraguay, Bolivia, Brazil (E.S.America) Bioregionally I Beni savannas- Bolivia * Argentine Espinal- * Llanos-Venezuela, Outstanding (Amazonia) Argentina (S.S.America) Colombia (Orinoco) * Pampas-Argentina 1 * Uruguayan savannas- * Guianan savannas- L {S.S.America) Uruguay, Brazil, Argentina Suriname, Guyana, (S.S.America) Venezuela, Brazil (Amazonia) * Amazonian savannas- Brazil, Colombia, Venezuela (Amazonia) * Argentine Monte- Argentina (S.S.America) Locally - Tabasco/Veracruz - * Humid Chaco-Argentina, Important savannas-Mexico Paraguay, Bolivia, Brazil (C.America) (E.S.America) * Tehuantepec savannas * C6rdoba montane Mexico (C.America) savannas- Argentina (E.S.America) Note: The final conservation status of two ecoregions in this MHT are unclassified: Central Mexican grasslands (Bioregionally Outstanding) and Eastern Mexican grasslands (Locally Important). Level1: Highest Priority at Regional Scale for biodiversity conservation I Level la: Highest Priority at Regional Scale (added to ensure bioregional representation) Table 5-8. Flooded Grasslands: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * Pantanal-Brazil, Bolivia, Outstanding Paraguay (E.S.America) Regionally * Central Mexican * Cuban wetlands -Cuba * Enriquillo wetlands -Haiti, wetlands - Mexico (Carib.) Dominican Republic (Carib.) Outstanding (.eio S(N.Mexico) Bioregionall Jalisco palm savannas - * Guayaquil flooded* Eastern Amazonian de * Orinoco wetlands- * Quintana Roo wetlands- Ygr 1 slnd 1 VcuadorOinco Mexico (C.America) Outstanding , Mexico (C.America) - _ grasslands-Ecuador grasslands - Brazil Venezuela (Orinoco) Outstndin 4_c.B _Ainrica (Amazonia) * Veracruz palm savannas- N _ - - _ _ _ _ Western Amazonian Mexico (C.America) * Parani flooded savannas - flooded grasslands -Peru, Argentina (E.S.America) Bolivia (Amazonia) Locally * Sio Luis flooded Important grasslands- Brazil (Amazonia) Level I: Highest Priority at Regional Scale for biodiversity conservation I Level la: Highest Priority at Regional Scale (added to ensure bioregional representation) 1.- -C- Table 5-9. Montane Grasslands: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * Santa Marta paramo - * Cordillera de Mrida Outstanding Colombia (N.Andes) paramo -Venezuela * Northern Andean (N.Andes) paramo - Colombia, Ecuador (N. Andes) * Cordillera Central paramo- Ecuador, Peru (C.Andes) Regionally * Central Andean puna- Outstadin Bolivia, Argentina, Peru, Chile (C.Andes) Z * Central Andean wet puna - Peru, Bolivia, Chile (C.Andes) * Central Andean dry puna - Argentina, Bolivia, Chile (C.Andes) * Patagonian steppe - Argentina, Chile (S.S.America) Bioregionally * Mexican alpine tundra-1 * Costa Rican paramo-Costa Outstand Mexico (C.America) Rica (C.America) Ou mstanding * Patagonian grasslands - Argentina, Chile (S.S.America) Locally * Southern Andean steppe- Argentina, Chile Important (S.S.America) Level I: 1-ighest Priority at Regional Scale for biodiversity conservation I 1 Level la: Highest Priority at Regional Scale (added to ensure bioregional representation) LF. -- Table 5-10. Mediterranean Scrub: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * California coastal sage- * Chilean matorral -Chile chaparral-Mexico, USA (C.Andes) Outstanding (.eio S(N.Mexico) Regionally Outstanding Bioregionally Outstanding Locally Important Level 1: Highest Priority at Regional Scale for biodiversity conservation Cn) Table 5-11. Deserts and Xeric Shrublands: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Citical Endangered Vulnerable Relatively Stable Relatively Intact Globally Galapagos Islands xeric * Northern Sonoran cactus scrub-Ecuador (N.Andes) scrub-Mexico, USA Outstanding (N.Mexico) Regionally * Sonoran xeric scrub- Mexico, USA (N.Mexico) Outstanding r- - - Gurea catu sc-b Ba- Ca-ori -xerici ---- ---- Bioregionally 9 Pueblan xeric scrub - . Araya & Paria xeric scrub - Guerreran cactus scrub- Baja California xenc Mexicoenezuela (Orinoco} Mexico (C.America) scrub-Mexico (N.Mexico) Outstanding Mexico a r - s--b - M - * Motagua Valley * Guajira/Barranquilla xeric * Sechura desert -Ecuador, thornscrub - Guatemala scrub-Colombia, Chile, Peru (C.Andes (C.America) Venezuela (N.Andes) * Atacama desert-Chile * Paraguan xeric scrub- (C.Andes) Venezuela (N.Andes) - * Caatinga-Brazil (E.S.America) Locally r Leeward Islands xeric * Cayman Islands xeric * Cuban cactus scrub- Cuba * San Lucan mezquital - scrub-Leeward Islands scrub-Cayman Islands (Carib.) Mexico (N.Mexico) (Ca(Carib(Cab.) * Chihuahuan xeric scrub- * Windward Islands xeric Mexico, USA (N.Mexico) scrub -Windward Islands * Eastern Mexican (Canb.) mezquital -Mexico, USA * Central Mexican (N.Mexico) mezquital - Mexico Aruba/CuragaqfBonaire (N.Mexico) cactus scrub -Netherlands * La Costa xeric shrublands - Antilles (Orinoco) Venezuela (Orinoco) Note: The final conservation status of four ecoregions in this MHT is unclassified: Western Mexican mezquital and Mexican Interior chaparral (both Locally Important) and Eastern Mexican matorral and Central Mexican cactus scrub (both Bioregionally Outstanding). Level 1: Highest Priority at Regional Scale for biodiversity conservation I I Level 1-: Highest Priority at Regional Scale (added to ensure bioregional representation) I -I Table 5-12. Restingas: Integration Matrix of Biological Distinctiveness and Conservation Status Biological Final Conservation Status Distinctiveness Critical Endangered Vulnerable Relatively Stable Relatively Intact Globally * Northeastern Brazil restingas- Brazil Outstanding (Amazonia) * Brazilian Atlantic Coast restingas- Brazil (E.S.America) Regionally Outstanding Bioregionally r* Paragana restingas- OtstaninllyVenezuela (N.Andes) Ottndi ng- Locally Important Level 1: Highest Priority at Regional Scale for biodiversity conservation I i Level la: Highest Priority at Regional Scale (added to ensure bioregional representation) LFCA W 6 Conservation Assessment of Mangrove Ecosystems Among the five major ecosystem types (METs), ecoregions and a list of priorities for investments in mangroves probably receive the least attention from restoration, biodiversity conservation, and sustain- conservation donors and agencies. Perhaps the able use. major reason that mangroves are underrepresented in conservation programs is that they are of superfi- Definition cially similar appearance across the entire region since mangroves, as aforest type, consist of only a few The word "mangrove" describes a salt-tolerant for- tree species. However, a mangrove ecosystem contains est ecosystem that occupies sheltered tropical and a wide diversity of aquatic and marine species and subtropical coastal estuarine environments. The in some cases mangrove ecosystems can rival other constituent plant species (also called mangroves) are habitat types in alpha diversity when both terrestrial not closely related, but they share morphological, and aquatic species are included in species lists. It is physiological, and reproductive adaptations that important not to confuse the definition of a man- allow survival in very saline, waterlogged, and re- grove forest with that of a mangrove ecosystem. duced substrates. These substrates are often subject This chapter summarizes the results of a conser- to rapid changes. Mangrove forest ecosystems are vation assessment exercise targeted specifically at different from other forests in that they receive in- mangrove ecosystems and draws heavily from a puts of matter and energy from both land and sea. synopsis prepared by Dr. Gilberto Cintr6n (USFWS) These inputs include fresh water, sediments and and other contributors. The exercise was undertaken nutrients from land and tidal flushing, and saline by the WWF Conservation Science Program in paral- intrusions from the sea. These inputs act as energy lel with the WWF/World Bank study of LAC ecore- subsidies that increase productivity and help main- gions and included major contributions from the tain ecosystem processes and high rates of organic leading experts on mangrove ecosystems in LAC matter fixation. The efficient allotment of organic (see the Acknowledgments section). production allows mangroves to develop complex We begin by first defining a mangrove ecosystem forests in areas subject to active geomorphic proc- and then briefly summarizing some important esses and change (Thom 1%7, 1984), climatic stresses attributes necessary to better understand appropri- like storms (Craighead and Gilbert 1%2; Tabb and ate conservation approaches. We then delineate the Jones 1962), hypersalinity (Cintr6n et al. 1978), and mangroves of LAC into a series of smaller units that even occasional frosts (Lugo and Patterson-Zucca 1977). permits prioritizing investments as in the non- mangrove ecoregions. We then define and apply the Ecological Attributes of Mangrove Ecosystems method for the mangrove analysis, adapted by the workshop participants from the non-mangrove The relatively high plant productivity and the active approach used elsewhere in this report (a more biological processes characteristic of mangrove eco- detailed description of the method can be found in systems yield many goods and services of direct or Appendix B). We then present the results of an indirect benefit. In the LAC region mangroves are assessment of conservation status of mangrove exploited for their timber, fuelwood, and charcoal. 36 Conservation Assessment of Mangrove Ecosystems 37 They are also known to be important for estuarine control which mangroves exert on nutrient and fisheries because of the contribution of detritus and sediment fluxes and shoreline erosion can determine dissolved organic carbon to estuarine food webs and the extent of the dominance of mud or coral reef the shelter their roots provide for juveniles (Heald environments within a landscape. Extensive coastal 1969; Odum 1971; Twilley 1982). Often unrecognized erosion has occurred in some areas where man- is the fact that mangrove ecosystems support groves have been removed, thus indirectly influenc- organic matter fixation by other primary producers, ing coastal biological communities. and that the importance of mangroves also resides in the role they play in the maintenance of the geomor- Delineation of Mangrove Complexes phic structure of these environments (e.g., estuaries, and Units lagoons, and reefs). Mangroves grow over landforms created and Mangrove forests vary greatly in structural and shaped by local geomorphic processes, creating a functional characteristics (Pool et al. 1977; Lugo and complex woody structure (trunks, branches, aerial Snedaker 1974; Cintr6n et al. 1985). The estuarine roots, pneumatophores) that varies in degree of distribution of mangroves also gives them a rather development and architecture as a response to linear appearance in contrast to other ecoregions. physiographic and climatic conditions. The hetero- However, at the landscape level, geomorphic proc- geneity of landforms and forest architecture gives esses provide a range of discrete landform types rise to a variety of habitats that provide shelter, for- over which mangroves become established (Hayes aging grounds, and nursery areas to many marine 1975; Galloway 1975; Thom 1984). For this study, and terrestrial animals. mangroves were categorized on the basis of four The production of large amounts of litterfall leads major landscape types colonized by mangroves: to large exports of particulate and dissolved organic deltas, estuaries, lagoons, and carbonate platforms carbon. High levels of secondary productivity are (adapted from eight major types recognized by supported by the production of phytoplankton and Thom 1967, 1984). benthic autotrophs (sea grasses and algae), and by In order to begin delineating mangroves, we made detritus inputs from upland sources. This high pro- the assumption that if a coast is divided into seg- ductivity allows the establishment of a complex and ments having comparable environmental conditions diverse food web that may support large resident and physiography, each unit would be characterized and migratory populations of mammals, reptiles, by distinctive landforms (see above) and processes. birds, fish, crustaceans, mollusks, and other associ- Mangroves within each segment would (a) occupy ated animals. the suite of suitable landform types typical of that The relative importance of various ecological energy regime; (b) develop ecosystems with similar processes in a given mangrove ecosystem is shaped attributes and outputs (levels of development and by the geomorphic, hydrologic, and climatic charac- productivity); (c) exhibit similar vulnerabilities and teristics of the area. Mangroves are also strongly responses to disturbance regimes; and (d) be most coupled to adjacent coastal and terrestrial areas responsive to a similar set of conservation activities. through ecologically important processes such as the This approach was used by Schaeffer-Novelli et al. movement of migratory species of fish and shrimp to (1990) to characterize the mangrove environments of coastal shelves and through bird and mammal the Brazilian coast. We adopted this concept since it populations moving between mangroves and inland provides a bioregional view with great potential for areas. There are also ecological linkages, in terms of the development of conservation priorities and goods and services, binding mangroves, seagrass management strategies for mangroves. beds, and coral reefs- where these systems coexist. Thirteen major biogeographic units, designated as Mangrove ecosystems perform other services as complexes, were identified along the coastlines of well. They play an important role in water storage Latin America and the Caribbean (Map 4; Table 6-1). and the trapping of sediments and carbon, contribut- Each complex was further subdivided to give a total ing to the control of the quality and quantity of wa- of 40 smaller units or segments. Individual sites, cor- ter, particulates, and solutes discharged to the ocean. responding to major landscape types, can be recog- The intricate network of roots that binds the sub- nized within each segment. In this study we only strate, the trunks and the root breathing organs present results at the complex and unit level. (pneumatophores) dissipate water energy and pro- A detailed report on the results of the workshop, mote the deposition of materials and reduces ero- including descriptions of high priority sites and sion. The underground root network contributes to updated maps of mangrove localities, will be avail- bank and sediment stabilization. The degree of able in 1995 from WWF. 38 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Table 6-1. Conservation Status of Mangrove Units of LAC Snapshot Final Conservation Conservation Mangrove Complex Mangrove Unit Status Threat Status Atlantic Gulf of Mexico Alvarado Vulnerable Medium Vulnerable Usumacinta Relatively Stable Medium Relatively Stable Yucatan Petenes Relatively Stable Low Relatively Stable Rio Lagartos Relatively Stable Low Relatively Stable Mayan Corridor Vulnerable High Endangered Belizean Coast Vulnerable Medium Vulnerable Belizean Reef Vulnerable Medium Vulnerable Atlantic Central America Northern Honduras Vulnerable High Vulnerable Mosquitia/ Nicaraguan Caribbean Coast Relatively Stable Low Relatively Stable Rio Negro/Rio San Sun Vulnerable Medium Endangered Bocas del Toro/Bastimentos Island/San Bias Relatively Stable Medium Vulnerable West Indies Bahamas Vulnerable Medium Vulnerable Greater Antilles Endangered Medium Endangered Lesser Antilles Endangered High Critical Continental Caribbean Coastal Venezuela Vulnerable Medium Vulnerable Magdalena/Santa Marta Endangered High Endangered Amazon-Orinoco-Maranh5o Trinidad Vulnerable High Endangered Segment 0: Orinoco-Cabo Orange Relatively Stable Medium Relatively Stable Segment 1: Cabo Orange- Cabo Norte Relatively Stable Medium Relatively Stable Segment II: Cabo Norte-Ponta Curura Relatively Stable Medium Relatively Stable Segment Ill: Ponta Curuga- Parnaiba Delta Relatively Stable Medium Vulnerable NE Brazil Segment IV: Parnaiba Delta- Cabo Calcanhar Vulnerable Medium Vulnerable Segment V: Cabo Calcanhar- Rec6ncavo Baiano Vulnerable Medium Endangered Segment VI: Rec6ncavo Baiano-Cabo Frio Vulnerable Medium Endangered SE Brazil Segment VII: Cabo Frio-Laguna Vulnerable Medium Endangered Conservation Assessment of Mangrove Ecosystems 39 Snapshot Final Conservation Conservation Mangrove Complex Mangrove Unit Status Threat Status Pacific Sea of Cortez NW Mexican Coast Vulnerable High Endangered Marismas Nacionales/ San Bias Endangered High Endangered Southern Mexico Southern Pacific Coast of Mexico Vulnerable Medium Vulnerable Pacific Central America Tehuantepec/El Manch6n Relatively Stable Low Relatively Stable Northern Dry Pacific Coast Endangered High Endangered Gulf of Fonseca Vulnerable High Endangered Southern Dry Pacific Coast Endangered High Critical Moist Pacific Coast Vulnerable Medium Vulnerable Panama Dry Pacific Coast Relatively Stable Low Relatively Stable Pacific South America Gulf of Panama Relatively Stable Medium Relatively Stable Esmeraldas/ Pacific Colombia Relatively Stable Medium Relatively Stable Manabi Critical High Critical Gulf of Guayaquil/Tumbes Vulnerable High Endangered Piura Relatively Stable High Vulnerable Galapagos Galapagos Islands Unclassified Unclassified Unclassified Conservation Status El Manch6n region of Pacific Central America; and the South American Atlantic coast from just south of We estimated the conservation status of mangrove the Amazon Delta to the Orinoco Delta. No Rela- units through assessments of the following criteria: tively Intact mangrove units were recognized. Fif- loss of habitat; large blocks of intact habitat; water teen (38 percent) mangrove units were either Critical quality and hydrographic integrity; rate of habitat or Endangered. conversion; degree of protection; degree of fragmen- tation; and the degree of alteration of the catchment Biological Distinctiveness and basin (refer to Appendix B for a more detailed dis- Conservation Activities cussion of the methods). As was done for the terres- trial ecoregions, the snapshot conservation status of Although the workshop identified individual man- the mangrove units was modified by assessments of grove sites that had outstanding ecological or bio- current and projected threats (e.g., clearing for devel- logical features (e.g., unusual species assemblages, opment, shrimp farming, and agriculture, pollution, critical habitat for migratory birds), no attempt was wildlife exploitation and overharvesting of fisheries, made to rank mangrove units on the basis of their and alteration of hydrologic flows through road biological distinctiveness because (a) discrimination building or channelization; Table 6-1). among mangroves on the basis of community di- Mangrove units were Relatively Stable in only five versity is difficult as mangroves throughout the LAC areas (13 percent): portions of the Yucatan Penin- region share a high proportion of species and have sula; the Pacific mangroves of Panama; the Mos- relatively similar high levels of alpha diversity and quitia region of Central America; the Tehuantepec/ low endemism; and (b) mangrove ecosystems are 40 A Conservation Assessment of thc Terrestrial Ecoregions of Latin America and the Carbbean Table 6-2. Most Appropriate Conservation Activities for Mangrove Units of LAC Conservation with Conservation for Mangrove Complex Restoration Restricted Access Sustainable Use Gulf of Mexico Usumacinta 1 Alvarado 3; Usamacinta 1 Yucatan Belizean Reef 5; Belizean Coast 1; Mayan Corridor 4; Belizean Rio Lagartos 3 Coast 2; Petenes 3 Atlantic Central America Mosquitia/Nicaraguan Northern Honduras 3; Rio Ne- Caribbean Coast 1; Northern gro/Rio San Sun 3; Bocas del Honduras 3 Toro/Bastimentos Island/San Bias 4 West Indies Lesser Antilles 2 Greater Antilles (Cuba only) 2; Bahamas 5; Greater Antilles Lesser Antilles (Guadalupe) 2 (excluding Cuba) 2 Continental Caribbean Magdalena/ Coastal Venezuela 1; Mag- Coastal Venezuela 3 Santa Marta 2 dalena/Santa Marta 4 Amazon-Orinoco-Maranh5o Orinoco 1; Segment 0 1; Trinidad 2; Segment III T; Seg- Segment 1111 ment 1 1; Segment 111 NE Brazil Segment lV 5; Segment V 2; Segment VI 2 Segment V 4; Segment VI 3 SE Brazil Segment Vil 4 Segment VII 4 Sea of Cortez Marismas Na- Marismas Nacionales/San Bias 1 NW Mexican Coast 5; Marismas cionales/ Nacionales/San Bias 1 San Bias 3 Southern Mexico Southern Pacific Coast of Mexico 4 Pacific Central America Southern Dry Moist Pacific Coast (between Moist Pacific Coast (between Pacific Coast I Panama and Costa Rica) 2; Pan- Panama and Costa Rica) 1; Pan- ama Dry Pacific Coast 4; Tehuan- ama Dry Pacific Coast 3; Gulf of tepec/El Manch6n 1 Fonseca 2; Northern Dry Pacific Coast 3; Tehuantepec/ El Manch6n 2 Pacific South America Manabi 1 Esmeraldas/Pacific Colombia 3; Esmeraldas/ Pacific Colombia Piura 3; Gulf of Panama 2 (southern part) 2; Gulf of Guayaquil/Tumbes 1; Gulf of Panama 2 Galapagos Galapagos Islands 3 Total Units 9 21 29 Note: Mangrove units are placed in the conservation activity category deemed most appropriate (some are placed in more than one activity column) according to the following criteria: (a) Restoration (ecological feasibility of restoration; relative beneficial ecological impact from restoration); (b) Conservation with restricted access (distinctiveness of biota and wildlife value; relative importance of ecosystem services assured through strict protection); and (c) Conservation for sustainable use (diversity of resources; capacity of system for recovery or resiliency). Each unit receives a priority ranking (1 being the highest and 5 the lowest) reflecting the urgency with which the unit should be considered. Conservation Assessment of Mangrove Ecosystems 41 exceptionally important in maintaining populations and the capacity for recovery were used to rank man- and ecological processes in surrounding marine, grove units within the conservation for sustainable freshwater, and terrestrial ecosystems. Thus, all use category. These analyses are intended to help mangroves were recognized as "keystone" ecosys- guide donors in choosing appropriate conservation tems, whose persistence is critical for the functioning activities for each mangrove unit and to provide a of diverse and extensive ecosystems well beyond the preliminary ecological cost-benefit analysis for the boundaries of an individual mangrove forest. units within each activity category. These rankings Because of the important ecological role of man- should guide the spatial and temporal sequence of groves, the workshop team stressed that the conser- conservation activities, and not be used to justify un- vation of all mangroves should be a priority and sustainable uses of particular mangrove systems. chose to categorize mangrove units on the basis of Restoration was identified as the most appropriate the conservation activity most appropriate for each: conservation activity for nine mangrove units, con- restoration; conservation with restricted access (e.g., servation with restricted access for twenty units, and protected areas); and conservation with an emphasis conservation with sustainable use for twenty-nine on sustainable use (Table 6-2). Within each category, units. Seventeen units were categorized under two mangrove units were ranked on the basis of how or more activities. important and feasible each of the assigned activities The results of the workshop are presently under would be for the habitat and surrounding ecosys- final review by mangrove experts and the updated tems. For example, in the restoration category, man- maps are being input into a geographic information grove units were ranked according to the relative system (GIS). We look forward to integrating the ecological impact and ecological feasibility of results of the mangrove workshop with those for the successful restoration. terrestrial ecoregions and upcoming analyses of fresh- For conservation with restricted access, different water and marine ecosystems. Ecosystem dynamics units were assessed in terms of the distinctiveness of require that linkages among terrestrial, freshwater, their associated biota and value for wildlife, and the mangrove, and marine systems are increasingly level of ecosystem services that would be assured emphasized if we are to develop effective priority- through strict protection. The diversity of resources setting methods and conservation strategies. 7 Conclusions and Recommendations It is important to conserve biodiversity in every for major donors interested in investing in biodi- ecoregion. This study is not intended to diminish or versity conservation in the region. The list and the discourage conservation programs in any ecoregion, integration exercises for each MHT (Tables 5-3 to particularly ecoregions identified as lower priority, 5-12) can also help donors to: but rather to assist conservation donors in planning the timing, sequence, and level of funding of current * Intervene quickly to ward off complete degrada- and future conservation efforts. tion and conversion in Critical ecoregions, particu- This study identifies the ecoregions of highest larly where threats to biodiversity are projected to biodiversity conservation priority based on final remain intense over the next five years conservation status and biological distinctiveness. * Identify ecoregions that, due to their conservation All high-priority ecoregions merit immediate atten- status, cannot absorb further intensive develop- tion from donors. Ultimately, investment priorities ment projects can be derived only after passing the list of priority * Flag those high beta diversity ecoregions that run sites through other political, social, institutional, and a high risk of species extinctions if threatened by economic filters. We feel that these filters are most extensive habitat conversion associated with major accurate and useful at the intra-ecoregion scale development projects rather than in comparisons among ecoregions. * Identify ecoregions best suited for environmen- In this chapter we review the results of the inte- tally sound sustainable development projects: gration model and make several broad recommen- those ecoregions that are largely intact, those with dations. We also discuss the results with respect to average (i.e., locally important) biodiversity, and several other efforts to use biodiversity indicators to those falling in an MHT characterized by low beta set regional priorities and explore the next steps diversity needed to refine biodiversity priorities into invest- * Conduct a much needed overlay analysis of the ment priorities. Finally, we examine the application allocation of donor funding by ecoregion, within a of this method to select priority sites at subregional MHT or bioregion. and national scales. The integration exercise identifies 55 ecoregions as Seventeen ecoregions are classified as (a) Globally being of Highest Priority at Regional Scale (level l). The Outstanding and either Critical or Endangered, or distribution of these ecoregions by bioregion shows (b) Regionally Outstanding and Critical (the three a relatively wide scatter (Table 7-1). The table also cells in the upper left corner of the integration includes those 19 ecoregions of elevated rank added matrix shown in Table 1-1). These ecoregions (shown to achieve better bioregional representation (level ,). in italics on Table 7-1) require immediate attention The Northern Andes bioregion has the greatest because of (a) the extraordinary or very unusual number of ecoregions classed Highest Priority at types of biodiversity found within them, and (b) the Regional Scale. With the exception of Northern possibility that many of the remaining habitat blocks Mexico, Orinoco, and Southern South America, in these ecoregions will be completely degraded ecoregions of Highest Priority at Regional Scale within 5 to 20 years, resulting in a loss of many occur in roughly equal numbers in the other six species endemic to these units, particularly plants bioregions. Table 7-1 could serve as a starting point and invertebrates. 42 Conclusions and Recommendations 43 Table 7-1. Ecoregions of Highest Priority at Regional Scale by Bioregion and Major Habitat Type Major habitat type Level I = Highest Priority at Regional Scale Level I (in italics) = Critical and Globally/Regionally Outstanding or Bioregion Endangered and Globally Outstanding (Countries partially or Level a = Ecoregion considered of Highest Priority at Regional Scale to entirely within the achieve bioregional representation (all of these were originally classified bioregion) as level II or III) Caribbean Tropical Moist Broadleaf Forests (Bahamas, Greater Antilles, I-Cuban moist forests-Cuba Lesser Antilles, I-Hispaniolan moist forests - Haiti, Dominican Republic Netherlands Antilles, I-Jamaican moist forests-Jamaica Turks and Caicos Islands, Tropical Dry Broadleaf Forests Cayman Islands) Id-Cuban dry forests - Cuba Tropical and Subtropical Coniferous Forests I-Cuban pine forests - Cuba I-Hispaniolan pine forests -Haiti, Dominican Republic Flooded Grasslands I-Cuban wetlands- Cuba I-Enriquillo wetlands -Haiti, Dominican Republic Deserts and Xeric Shrublands la-Leeward Islands xeric scrub-Leeward Islands Northern Mexico Tropical Dry Broadleaf Forests (Mexico, USA) I,-Tamaulipas/ Veracruz dry forests -Mexico Tropical and Subtropical Coniferous Forests I-Sierra Madre Occidental pine-oak forests - Mexico, USA Flooded Grasslands I-Cen tral Mexican wetlands - Mexico Mediterranean Scrub I-California coastal sage-chaparral - Mexico, USA Deserts and Xeric Shrublands I-Northern Sonoran cactus scrub -Mexico, USA Central America Tropical Moist Broadleaf Forests (Mexico, Belize, Ia-Tehuantepec moist forests - Mexico, Guatemala, Belize Guatemala, Honduras, Tropical Dry Broadleaf Forests Nicaragua, Costa Rica, I-Jalisco dry forests -Mexico Panama, El Salvador) I-Balsas dry forests- Mexico Tropical and Subtropical Coniferous Forests I-Mexican transvolcanic pine-oak forests -Mexico I-Sierra Madre del Sur pine-oak forests - Mexico Grasslands, Savannas, and Shrublands la-Tabasco/ Veracruz savannas - Mexico Flooded Grasslands Ia-Jalisco palm savannas- Mexico Montane Grasslands Ia-Mexican alpine tundra - Mexico Deserts and Xeric Shrublands la-Pueblan xeric scrub - Mexico (Table continues on the following page) 44 A Consemation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Table 7-1. Ecoregions of Highest Priority at Regional Scale by Bioregion and Major Habitat Type (continued) Major habitat type Level I = Highest Priority at Regional Scale Level I (in italics) = Critical and Globally/Regionally Outstanding or Bioregion Endangered and Globally Outstanding (Countries partially or Level Ia = Ecoregion considered of Highest Priority at Regional Scale to entirely within the achieve bioregional representation (all of these were originally classified bioregion) as level lI or III) Orinoco Tropical Moist Broadleaf Forests (Trinidad and Tobago, 1-Cordillera La Costa montane forests-Venezuela Venezuela, Colombia, Tropical Dry Broadleaf Forests Brazil, Guyana) 1-Llanos dry forests-Venezuela Deserts and Xeric Shrublands 1--Araya and Paria xeric scrub-Venezuela Amazonia Tropical Moist Broadleaf Forests (Brazil, Colombia, 1-Napo moist forests -Peru, Ecuador, Colombia Venezuela, Guyana, French I-Macarena montane forests-Colombia Guiana, Suriname, Bolivia, I-Ucayali moist forests - Brazil, Peru Peru, Ecuador) I-Western Amazonian swamp forests-Peru, Colombia I-Southwestern Amazonian moist forests-Brazil, Peru, Bolivia I-Varzea forests -Brazil, Peru, Colombia Tropical Dry Broadleaf Forests i-Bolivian lowland dry forests - Bolivia, Brazil Grasslands, Savannas, and Shrublands la-Beni savannas -Bolivia Flooded Grasslands la-Eastern Amazonian flooded grasslands -Brazil Northern Andes Tropical Moist Broadleaf Forests (Venezuela, Colombia, I-Choc6/ Dari6n moist forests -Colombia, Panama Panama, Ecuador, Peru) I-Northwestern Andean non tane forests- Colombia, Ecuador I-Western Ecuador moist forests - Ecuador, Colombia I-Cauca Valley mon tane forests - Colombia 1-Magdalena Valley montane forests - Colombia 1-Cordillera Oriental montane forests - Colombia, Venezuela I-Eastern Cordillera Real montane forests- Ecuador, Colombia, Peru I-Santa Marta montane forests -Colombia I-Venezuelan Andes mon tane forests - Venezuela, Colombia Tropical Dry Broadleaf Forests 1-Ecuadorian dry forests - Ecuador I-Tumbes/Piura dry forests - Ecuador, Peru Flooded Grasslands la-Guayaquil flooded grasslands -Ecuador Montane Grasslands I-Santa Marta paramo-Colombia I-Cordillera de M6rida paramo-Venezuela I-Northern Andean paramo-Colombia, Ecuador Deserts and Xeric Shrublands I-Galapagos Islands xeric scrub -Ecuador Restingas la-Paraguan, restingas- Venezuela Conclustons and Reconimendations 45 Table 7-1. Ecoregions of Highest Priority at Regional Scale by Bioregion and Major Habitat Type (continued) Major habitat type Level I = Highest Priority at Regional Scale Level I (in italics) = Critical and Globally/Regionally Outstanding or Bioregion Endangered and Globally Outstanding (Countries partially or Level Ia = Ecoregion considered of Highest Priority at Regional Scale to entirely within the achieve bioregional representation (all of these were originally classified bioregion) as level II or III) Central Andes Tropical Moist Broadleaf Forests (Peru, Bolivia, Chile, I-Pertvian Yungas - Peru Argentina) I-Bolivian Yungas-Bolivia, Argentina Tropical Dry Broadleaf Forests Id-Bolivian montane dry forests-Bolivia Montane Grasslands 1-Cordillera Central paramo-Ecuador, Peru I-Central Andean puna- Bolivia, Argentina, Peru, Chile I-Central Andean wet puna-Peru, Bolivia, Chile I-Central Andean dry puna - Argentina, Bolivia, Chile Mediterranean Scrub I-Chilean inatorral - Chile Deserts and Xeric Shrublands la-Sechura desert-Peru, Chile Eastern South America Tropical Moist Broadleaf Forests (Brazil, Bolivia, Paraguay, I-Brazilian Coastal Atlantic forests - Brazil Argentina) I-Brazilian Interior Atlantic forests -Brazil, Argentina, Paraguay Tropical and Subtropical Coniferous Forests P-Brazilian Araucaria forests -Brazil, Argentina Grasslands, Savannas, and Shrublands I-Cerrado-Brazil, Paraguay, Bolivia I-Chaco savannas- Argentina, Paraguay, Bolivia, Brazil Flooded Grasslands I-Pantanal -Brazil, Bolivia, Paraguay Deserts and Xeric Shrublands I^-Caatinga - Brazil Restingas I-Northeastern Brazil restingas - Brazil I-Brazilian Atlantic Coast restingas - Brazil Southern South America Temperate Forests (Brazil, Uruguay, I-Chilean winter-rain forests-Chile Argentina, Chile) 1-Valdivian forests -Chile, Argentina Grasslands, Savannas, and Shrublands la-Pampas- Argentina Montane Grasslands 1-Patagonian steppe- Argentina, Chile 46 A Conservation Assessment of the Terrestrial Ecoregions of Latin Amcrica and the Caribbean At the same time, the few ecoregions that contain * Conducting the analysis at a finer geographic relatively intact ecosystems deserve immediate con- scale by using ecoregions rather than the more servation investment because ecosystem processes coarse-grained regional habitat units. and species have the best chance for long-term per- . Using additional criteria to achieve representation sistence within these rare intact landscapes. It would among MHTs and bioregional representation be unwise to miss the window of opportunity they within MHTs. offer for strategic conservation planning. Some of the * Linking biological distinctiveness to a biogeo- investments in these ecoregions may be limited to graphic spatial scale and using four categories of monitoring environmental conditions, lobbying biological distinctiveness rather than the three against poorly conceived development projects, or classes used in the BSP workshop. other low-cost activities. At a minimum, planning and establishment of effective protected area sys- Both studies identified as highest priority the tems should be initiated within each of these impor- Atlantic moist forests and the Cerrado of Brazil, the tant areas.' Patagonian steppes, the submontane and montane The purpose of this study is to provide a regional regions of the northern Andes, the montane grass- overview and an input to national biodiversity lands of the central and northern Andes, the mon- strategies. We offer here a few observations of inter- tane mixed conifer forests of Mexico, and the est to conservation planners at national levels. Sonoran desert ecoregion. Among the countries of LAC, Ecuador is notable for There were also some notable differences. I his being entirely covered by ecoregions of Highest Pri- study recognized as Highest Priority at Regional ority at Regional Scale (i.e., red or purple on Map 9). Scale the Valdivian temperate forest, the Chilean Peru, Bolivia, Cuba, Haiti, Dominican Republic, and winter-rain habitats (forest and scrub), the drier Jamaica also are largely covered by such ecoregions, portion of the Chaco savannas, ecoregions compos- and Colombia, Chile, southern Brazil, and southern ing the western arc of Amazonia, varzea moist for- Mexico are not far behind. ests of Amazonia, the Choc6/DariCn moist forests, the moist forests of the Greater Antilles, and the dry Comparisons with Other Priority-Setting forests of southern Mexico. These regions were con- Frameworks for LAC sidered of secondary or lower priority by the BSP exercise. In contrast, the BSP exercise classified One of the more interesting opportunities offered many ecoregions among the Mexican xeric forma- by this study is to overlay results from other re- tions and the Beni savannas as highest priorities. gional and subregional priority-setting exercises. The results of the Birdlife and Centers of Plant The three most comprehensive efforts to date are Diversity studies were of value to us in the delimita- the priority-setting exercise coordinated by the tion of ecoregions. It would be very instructive to Biodiversity Support Program on behalf of USAID closely compare the results of these two studies with (BSP/CI/TNC/WRI/WWF 1995), an analysis of the high-priority ecoregions we have identified. endemic bird areas in LAC by Birdlife Interna- These two studies will also be useful for setting pri- tional (Long et al. 1994; Wege and Long in press), orities at finer geographic scales (see below). and the database on Centers of Plant Diversity (WWF and IUCN 1994). Application of the Methodology to Finer A number of the data layers and analyses con- Geographic Scales ducted during the BSP exercise are also part of this study. Not surprisingly, there is considerable overlap The integration of biological distinctiveness and con- in the results. The hierarchical classification scheme servation status helps target the highest priority outlined in Chapter 1 served as the template for ecoregions for urgent attention by donors interested maintaining representation in the BSP exercise, and in biodiversity conservation. By giving increased their map of "regional habitat units" was derived, in funding to these ecoregions, donors and governments part, from the ecoregions map. The method to assess will increase the likelihood of maintaining repre- conservation status was virtually the same. The major sentation of all ecosystem and habitat types in a differences in method from the BSP study include: regional investment portfolio. However, regional priority-setting exercises-by virtue of their coarse scale-can seldom identify 1. Some ecoregions conceivably may be of high priority but where the most important investments should be stable with negligible conversion rate and good protection of made within ecoregions nor what to do at those sites to remaining habitat. Since these factors tend to vary widely . . across an ecoregion, they should be evaluated at finer geo- best conserve biodwersity. Prioritizing among 191 graphic scales. ecoregions at a regional scale is difficult enough. Conclusions and Recommendations 47 Prioritizing among the tens, hundreds, or thousands * A persistence value analysis of remaining habitat of blocks of remaining habitat within an ecoregion blocks or clusters of blocks to determine their is even more difficult. Yet, it is often at the site relative probability of maintaining biodiversity level that decisions must be made for conservation over the long term. We use landscape features action. such as minimum size of blocks to conserve bio- Thus, the development of frameworks at finer diversity and processes, configuration of remain- geographic scales, such as intra-ecoregional analyses, ing blocks, and analysis of intervening habitats to are an essential second step of the priority-setting assess landscape integrity. process. Without this framework, donors risk financ- * A biodiversity and critical habitats analysis to overlay ing biodiversity conservation in the most important patterns of species richness hotspots, areas of high ecoregions, but conserving some of the less impor- endemism or beta diversity, rare habitat types, tant, unique, or intact habitat blocks within them. populations of rare species, critical habitat for For example, even Globally Outstanding ecore- migrants and resident species, and core habitats gions contain some blocks of remaining habitat within an ecoregion. that are of relatively low biological or persistence . An analysis of existing protected areas and corridors value compared with other blocks in the same overlaid on the first two layers. This step helps to unit. Alternatively, even ecoregions designated as determine where investments should go first and Locally Important may contain a few blocks larger which areas should receive proportionately and more intact than those found in other ecore- greater attention. gions within the same MHT and so would qualify . A land use analysis to determine the feasibility of as sites of high conservation interest. A frame- creating new reserves and managing habitats work that also interprets spatial patterns of biodi- adjacent to protected areas in a manner more versity within an ecoregion can also help to compatible with conservation of biodiversity. This determine the level of effort (e.g., number and step assesses other uses or demands for resources size of protected areas) needed to conserve the by stakeholders who have access to or heavily biodiversity present. utilize blocks with high conservation potential. As illustrated by this study, a biologically based framework can help to set priorities among ecore- Integration of these layers can also identify prior- gions within an MHT. The same landscape ecology ity areas that may never receive formal protection or and conservation biology principles can help guide were not classified as intact natural habitats. Because donors to the most biologically important blocks or of their biodiversity value or their potential for clusters of habitats within ecoregions with the high- enhancing the persistence of adjacent natural habi- est probabilities of persistence. tats, these areas may require restoration or man- Several methods for determining priority sites for agement plans that emphasize conservation of conservation, specifically protected areas, have been biodiversity and natural resources. created (e.g., Margules et al. 1988; Pressey et al. 1993). These overlays incorporate the fundamental bio- All of these methods, however, demand extensive logical data that should serve as the foundation for data-specifically, detailed information on the dis- setting investment priorities within ecoregions. After tribution of species representing many different these overlays are performed, non-biological data taxa. Such data sets are largely incomplete for most can be used to further refine the list of options for ecoregions in LAC. They also pay less attention to investing in biodiversity. The methods described landscape features that will influence the effective- above can be used for all terrestrial ecoregions. By ness of conservation of unique taxa at those sites. applying landscape-level criteria to their own maps To address this need, WWF is developing a of remaining habitat, as planners in Argentina are method to set priorities among habitat blocks within now doing, conservationists in LAC can better iden- ecoregions. The method consists of a series of analy- tify and prioritize among (a) the large blocks of ses, the results of which can be overlaid using a GIS remaining natural habitat with high potential for to determine intra-ecoregion conservation priorities. long-term biodiversity conservation, and (b) the Briefly, four core analyses are needed to prepare biologically rich but highly fragmented habitats overlays: requiring urgent attention. Appendix A Methods Used for Assessing the Conservation Status of Terrestrial Ecoregions An assessment of the conservation status of each and analyze relationships among variables. By selec- ecoregion is intended to (a) identify major habitat tively removing or aggregating criteria, it is possible types and ecoregions that are most threatened and to test for correlation among variables, identify rela- thus help prioritize interventions to prevent their tionships among factors that are unique to particular complete degradation or conversion; (b) create pro- METs or MHTs, and determine strong predictors for grams to conserve the most intact examples of major conservation status assessments. habitat types; and (c) help define appropriate con- servation activities for different landscape scenarios. GIS-Based Analyses Here we define the conservation status of ecoregions in the tradition of the IUCN Red Data Book catego- The parameters used in the conservation status ries for threatened and endangered species. The con- assessment can be accurately measured using GIS servation status categories we use for ecoregions are technology when adequate digital data bases are Extinct, Critical, Endangered, Vulnerable, Relatively available. Although many digital databases of Stable, and Relatively Intact. These categories repre- landscape parameters were consulted in the con- sent different degrees of alteration and of spatial servation status assessments (e.g., remaining patterns of remaining habitats across landscapes. habitat data based on MSS and AVHRR satellite They reflect how with increasing habitat loss, degra- imagery), final values for the conservation status dation, and fragmentation, ecological processes cease criteria were arrived at through assessments by to function naturally, or at all, and major compo- regional experts. nents of biodiversity are steadily eroded. Landscape In the vast majority of cases, the heightened accu- features are used as indicators for the ecological racy available from a strictly GIS-based landscape integrity of ecosystems. analysis would be unlikely to change the conserva- tion status assigned to ecoregions in this study Comments on the Approach because the categories used for each criterion are sufficiently broad. We also decided that there was The conservation status assessment used here no net benefit from applying rigid habitat classifica- evolved out of an earlier approach proposed by tion standards to digital databases of uncertain WWF (Olson and Dinerstein 1994) that separately quality and incomplete coverage. Currently, the assessed the conservation potential and threat status available digital data on remaining natural habitats of ecoregions. Having subsequently benefited from varies widely in terms of quality and availability numerous insightful critiques, we feel the current throughout the LAC region. Moreover, our ability to approach now contains significant advances over the effectively interpret habitat information based on earlier method. Incorporating conservation potential remote sensing is currently poor for several major and threat into the single conservation status assess- habitat types (e.g., deserts, grasslands). ment is the most prominent change. We urge future researchers to apply GIS-based The transparent and flexible design of this method analysis of remaining habitats to test, update, and should facilitate future efforts to reexamine results refine the results of this study, particularly as new 49 50 A Conservation Assessrnent of the Terrestnal Ecoregions of Latin America and the Caribbean and better-quality data become available. However, species loss in different ecosystems, both theoreti- we are confident that the snapshot conservation cal and empirical studies support the general corre- status assessments made here have accurately lation between habitat loss and species loss. Loss of placed the vast majority of ecoregions in the appro- habitat reduces biodiversity (a) by eliminating priate category. species or communities limited to particular geo- graphic localities, (b) by decreasing the area of Explanation of Criteria Used in This Analysis available original habitat below the minimum size needed to maintain critical, large-scale ecosystem Because the loss of biodiversity and alteration of dynamics; and (c) through the degradation and ecological processes, both current and projected, fragmentation of remaining habitats such that they are difficult to measure directly, conservation bio- become too small or isolated to individually or logists are increasingly relying on landscape-level collectively support viable populations or maintain parameters as indicators. In this study, we employ important ecological processes. the percent of original habitat lost, the presence of The second and third effects of habitat loss are large blocks of intact original habitat, the degree reflected in large part in the criteria discussed of habitat fragmentation and degradation, the rate below of habitat blocks and habitat fragmentation. of conversion, and the degree of protection. Based Total habitat loss, measured at an ecoregional scale, on principles of landscape ecology and conserva- reflects all these consequences but underscores the tion biology (for detailed discussions see Noss first and second. The loss of species due to elimina- 1992, Primack 1993, Meffe and Carroll 1994, Noss tion or truncation of habitats within their ranges is and Cooperrider 1994, Kareiva and Wennergren particularly important in ecoregions with high beta 1995), we assume that these variables help predict or gamma diversity. (a) an ecosystem's ability to maintain ecological processes (e.g., population and predator-prey dyna- Habitat Blocks mics within natural ranges of variation, pollination and seed dispersal, nutrient cycling, migration, A critical parameter for assessing conservation dispersal, and gene flow); and (b) components of status is the number and extent of blocks of contigu- biodiversity (e.g., top predators or other keystone ous habitat large enough for populations and ecosys- species) that will strongly influence how much tern dynamics, each with different minimum area and what kind of biodiversity will persist over the requirements, to function naturally. Large blocks of long term. habitat sustain larger and more viable species Although the first three criteria-habitat loss, habitat populations, and they permit a broader range of blocks, and fragmentation-can be highly correlated species and ecosystem dynamics to persist. The geo- in some landscapes (e.g., continuing habitat frag- graphic coverage of multiple large blocks also con- mentation can reduce both the total area of habitat serves a wider range of habitats, environmental available and the size of habitat blocks), each of gradients, and species. these criteria is analyzed separately because their The number of large blocks of habitat in different relationships can be quite variable and different size categories is an important component of this combinations can have substantially different eco- criterion. Redundancy theory suggests that the logical effects. For example, an ecoregion with presence of three or more examples of an ecosystem 50 percent of its original habitat in two large blocks significantly increases its probability of long-term would have a higher probability of persistence than persistence. Factors such as fire, disease, pollution, a similar habitat area that is highly fragmented into deforestation or degradation can eliminate species multiple small patches. or natural habitats within blocks. The presence of several blocks with similar communities allows Total Habitat Loss recolonization and persistence of particular habitat types and species. Multiple habitat blocks that are Habitat loss has been widely recognized as one of well-distributed across the landscape are particu- the primary factors contributing to the reduction larly important for conserving species and habitats and loss of terrestrial populations, species, and in ecoregions that are characterized by a high degree ecosystems. This criterion underscores the rapid of beta diversity (species turnover along environ- loss of species predicted to occur in ecosystems mental gradients). when the total area of remaining habitat falls below The threshold size for viable blocks of habitat is minimum critical levels. Although there is no con- broadly tailored to the scale of important ecosys- sensus on the mechanisms or exact thresholds for tem dynamics for different METs. In order to avoid Methods Used for Assessing the Conservation Status of Terrestrial Ecoregions 51 misleading conclusions by applying continental remote sensing imagery. Moreover, quantifying size thresholds to island ecoregions (or very small habitat degradation at this scale is problematic continental or naturally disjunct systems), different because (a) habitat degradation is often patchy; sets of size thresholds are employed for each (b) degraded states form a continuum and are not ecoregion size category. Scale problems are simi- easily classified; and (c) the ecological effects of larly addressed in the protected area criterion (see different forms of degradation are unclear and may below). occur on the scale of weeks to centuries. For these reasons, habitat degradation was not used as a cri- Habitat Fragmentation terion for most MHTs, although we suggest it be incorporated into future analyses when better data Persistent small population sizes are widely per- and interpretative methods become available. ceived as a major threat to conservation of terres- Regional experts felt that habitat degradation trial species. Habitat fragmentation places many would be particularly important in contributing to low-density species in demographic jeopardy the conservation status of the grassland/savanna/ (Berger 1990; Laurance 1991; Newmark 1991; Wil- shrubland and xeric formation METs, but were only cove et al. 1986). Fragmented ecosystems are comfortable in evaluating degradation for two MHTs: stressed over a relatively large percentage of their grasslands, savannas, and shrublands, and flooded intact habitat area by hunting pressure, fires from grasslands. Therefore, for these two MHTs, the surrounding human activity, changes in micro- index value attributed to fragmentation was split climates, and invasion of exotic species (Lovejoy between the fragmentation index and a degradation 1980; Saunders et al. 1991; Skole and Tucker index. 1993). As fragmentation increases, the amount of critical core habitat area decreases. Fragments Habitat Conversion under 100 km2 are inadequate for maintaining viable populations of most large vertebrates. Conversion rates are less powerful estimators of Some species of birds, trees, and butterflies that conservation status than large-scale landscape typically occur in very low densities, or have ex- features because (a) the actual ecological effects tremely patchy distributions, may also be lost in associated with conversion rates varies consid- small fragments.' Ecoregions that still maintain erably depending on the original size of the large blocks (e.g., >1000 kM2) of intact original ecoregion, the amount of remaining habitat when habitat will at least have some core areas where the rates were estimated, and spatial patterns of large-scale ecological processes function natu- conversion; (b) the major uncertainties associated rally. The point values associated with different with current conversion rate estimates (see Whit- categories of this criterion reflect the greater more and Sayer 1992); (c) the sensitivity of con- severity of ecosystem disruption in landscapes version rates to relatively minor changes in human where habitat fragmentation is more advanced behavior; and (d) the actual loss of habitat asso- (see Groom and Shumaker 1993). ciated with recent estimates of habitat conversion, even for high rates, is typically small relative to Habitat Degradation the extensive landscape alteration of past centu- ries that is best reflected in the first three criteria. Habitat degradation resulting from human activi- However, recent conversion rates do provide ties such as selective logging, pesticide exposure, some information on the short-term trajectory of burning, and overgrazing can have profound im- future habitat loss and fragmentation and are pacts on the long-term viability of ecosystems. used to improve the accuracy of conservation However, degraded systems can be difficult or status assessments rather than for estimating impossible to distinguish from pristine areas in long-term threats. Our final conservation status analysis projects trends for habitat loss, fragmen- tation, and patch size into the next two decades 1. However, we also recognize that small fragments can be and accounts for predicted or proposed large- valuable for conserving representative communities and scale conversion events (e.g., pending agricultural species (Shafer 1995), particularly in regions that are charac- expansion projects or roads). After priority terized by high levels of beta diversity. Many invertebrates, ecoregions have been identified, habitat conver- plants, fungi, and small vertebrates can be effectively conserved within small blocks of original habitat. Small sion rates could be used to indicate which ecore- fragments can also act as important stepping-stones for gions would benefit most from certain kinds of movement and dispersal of species. conservation investments. 52 A Conservahon Assessment of the Terrestrial Ecoregions of Latin Arica and the Canbbean Degree of Protection latter three are best considered in more detailed analyses at the intra-ecoregion level. The degree of protection criterion assesses how well humans have conserved sufficiently large blocks of Determination of Snapshot Conservation intact habitat. Protected areas managed primarily for Status: Weighting and Categories the conservation of biodiversity, or which otherwise effectively protect intact habitat, are emphasized in The parameters described in more detail in the this criterion. Protected areas are not used as pri- following section were used to assess a snapshot mary indicators for the conservation status of conservation status index for each ecoregion. ecoregions because (a) the distribution of protected A threat analysis subsequently modifies the results, areas do not necessarily reflect the current extent if necessary, to produce the final conservation status and configuration of original habitat or the integrity for each ecoregion. of ecosystems over the entire landscape; (b) many The conservation status index has a point range protected areas encompass habitats that would not from 0 to 100, with higher values denoting a higher be considered intact; and (c) most protected areas level of endangerment. The range from 0 to 100 was are currently too few and small to encompass com- deemed appropriate because previous experience plete ecosystems and will only be effective if the sur- with prioritizing and ranking large numbers of con- rounding landscape is well managed for biodiversity servation areas has shown that a narrower index conservation. scale (e.g., from 0 to 30 as tested in the WWF Russia One could potentially emphasize a lack of for- Biodiversity Project) yields little discrimination mally protected areas in the threat analysis (leading among units. to the final conservation status) rather than consider- We feel that several landscape parameters should ing their presence as a predictor of the snapshot con- be given greater weight in the determination of the servation status. However, a lack of protected areas snapshot conservation status index: total habitat may not threaten some ecoregions due to their inac- loss, habitat blocks, and degree of habitat fragmen- cessibility or harsh environments. Assessing threats tation. The weighting of the different parameters in using negative criteria (i.e., absence of protected the index is: areas) increases the probability of making a poor decision compared to basing conclusions on existing Weight parameters. (percent) P a r a me te r Several important aspects should be considered in 40 Total habitat loss a comprehensive analysis of protected areas: 20 Habitat blocks 20 Habitat fragmentation (or, for grassland, * The degree to which large remaining blocks of savanna, and shrubland, and flooded habitat are adequately protected within a system grassland MHTs: 10 percent fragmenta of protected areas tion and 10 percent habitat degradation) * The level of redundancy of protected areas that 10 Habitat conversion is needed to help ensure the long-term persistence 10 Degree of protection of habitat types, communities, endangered spe- cies, or critical habitats for species or ecological The point thresholds for different categories of processes conservation status are listed below (classification as * The degree to which representative habitat types, "Extinct" is based on expert assessment): communities, ecological gradients, endangered species, or critical habitats for resident or migra- Points Conservation Status tory species or ecological processes are contained 0-6 Relatively Intact within a system of protected areas 7-36 Relatively Stable * The degree of connectivity among reserves for the 37-64 Vulnerable dispersal of species and contiguity of large-scale 65-88 Endangered ecosystem processes 89-100 Critical * The effectiveness of management of protected areas and the ability of managers to defend protected Determination of Point Values for Each areas based on their landscape configurations. Criterion The first two considerations are addressed in the Within each criterion, the determination of point degree of protection criterion used here, while the values is intended to reflect real biological processes Methiods Used for Assessing the Conservation Status of Terrestrial Ecoregions 53 or the relative contribution of a particular situa- * Grassland/savanna/shrubland and xeric forma- tion to long-term biodiversity conservation. For tion METs: habitat has not been plowed or example, both empirical evidence and theoretical affected by major changes in flooding or surface ecology suggest that species loss and secondary water patterns. The vast majority of native plant extinctions increase dramatically with extensive species are still present in abundances within habitat loss and fragmentation (Simberloff 1992; their natural range of variation and succession Terborgh 1992). In some cases, the assigned point patterns follow natural cycles (e.g., grazing by values closely reflect the relationship between the domestic livestock and human-caused fires have causal factor and the ecological response (i.e., no significant impact on native biota). Although plotting of the points will approximate the curve large mammals and birds may presently be of the general relationship). More subjective cri- absent from some blocks of habitat due to exploi- teria are divided into broad categories that facili- tation or insufficient area, such blocks may still tate classification. sustain many native plant, invertebrate, and Certainly, alternative classification systems are vertebrate species and their associated ecological possible. The original databases are provided in this processes. report for those who may wish to reanalyze the data using different indices, criteria, weightings, or index Points Original Habitat Lost ranges. 0 0-10 percent For most of the criteria, some point values other 10 10-24 percent than those defined for the various categories were 20 24-49 percent allowed in instances when regional experts deemed 32 50-89 percent them useful and unavoidable. 40 > 90 percent Total Habitat Loss Habitat Blocks This and the following criteria require a definition In addition to using a different set of criteria for dif- of what constitutes intact habitat. Although it ferent MHTs, variables for this and the protected would be preferable to classify habitats into several area criterion differ by ecoregion size category. In categories reflecting different levels of habitat the tables for these two criteria, the smallest ecore- alteration and degradation, constraints imposed by gion size category (<100 kM2) is also used for ecore- data availability and time and resource limitation gions, such as those on some Caribbean islands, that necessitated using two broad classes-intact and originally occurred in naturally disjunct patches of altered this size. We propose that intact, or remaining, habitat rep- In the following tables for habitat block analysis resents relatively undisturbed areas that are charac- (Tables A-1 to A-4), the information in the cells, un- terized by the maintenance of most original ecologi- less stated otherwise, refers to the minimal require- cal processes and by communities with most of their ment for at least one block of intact habitat. For an original suite of native species. The following are ecoregion within any given column, corresponding some further clarifications of our definition for two to its size, the table should be read from top to bot- broad types of habitat: tom until a statement true of the ecoregion is reached. The text ">500" should be interpreted as: * Tropical broadleaf forest and conifer/temper- "the ecoregion contains at least one block of intact ate broadleaf forest METs: Canopy disturbance habitat greater than 500 km2." Percentage values through human activities such as logging or refer to the portion of the original ecoregion size that small-scale agriculture is restricted to less than is still considered to be intact habitat. 10 percent of defined habitat block. Understory At the BSP workshop, some regional experts felt largely undisturbed by timber extraction, graz- that the minimum size and number of habitat blocks ing, agriculture, or human caused fires. Although should be increased for very large ecoregions to large mammals and birds may presently be better address the greater proportional amount of absent from some blocks of habitat due to exploi- habitat necessary to conserve the range of biodiver- tation or insufficient area, such blocks may still sity in extensive ecoregions. We acknowledge that sustain many native plant, invertebrate, and some of the values above may represent minimally vertebrate species and their associated ecologi- adequate conditions for some conservation status cal processes. categories. 54 A Conservation Assessment of te Terrestrial Ecoregions of Latin America and the Caribbean Table A-1. Habitat Block Analysis for Tropical Broadleaf Forest MET Ecoregion Size Point Value >3,000 km2 1,000-3,000 kmn2 100-1,000 km2 <100 km2 2 >3,000 or >1,000 or >500 80-100 percent intact 3 blocks >1000 3 blocks >500 5 >1,000 >500 >250 40-80 percent intact 10 >500 3 blocks >250 3 blocks >100 10-40 percent intact 15 >250 >250 >100 1-10 percent intact 20 None >250 None >250 None >100 <1 percent intact Note: The value ">1,000" means "the ecoregion contains at least one habitat block greater than 1,000 km2." For an ecoregion of any given size, this and the following tables should be read from top to bottom until a statement is reached that is true of the ecoregion. Table A-2. Habitat Block Analysis for Conifer/Temperate Broadleaf Forest MET Ecoregion Size Point Value >3,000 kn,2 1,000-3,000 km 100-1,000 kmn2 <100 km2 2 >2,000 or >800 or >500 80-100 percent intact 3 blocks >800 3 blocks >500 5 >800 >500 >250 40-80 percent intact 10 3 blocks >250 3 blocks >250 3 blocks >100 10-40 percent intact 15 >250 >250 >100 1-10 percent intact 20 None >250 None >250 None >100 <1 percent intact Table A-3. Habitat Block Analysis for Grassland/SavannafShrubland MET Ecoregion Size Point Value >3,000 km2 1,000-3,000 km2 <1,000 km2 2 >750 or 3 blocks >500 >500 or 3 blocks >250 80-100 percent intact 5 >500 >250 40-80 percent intact 10 >250 22 blocks >100 10-40 percent intact 15 >100 >100 1-10 percent intact 20 None >100 None >100 <1 percent intact Table A-4. Habitat Block Analysis for Xeric Formation MET Ecoregion Size Point Value >3,000 km2 1,000-3,000 km2 <1,000 km 2 2 2 blocks >500 >500 80-100 percent intact or 3 blocks >200 or 2 blocks >200 5 >500 >200 40-80 percent intact 10 >200 2 blocks >100 10-40 percent intact 15 >100 >100 1-10 percent intact 20 None >100 None >100 <1 percent intact Methods Used for Assessing the Conservation Status of Terrestrial Ecoregions 55 Habitat Fragmentation Habitat Degradation WWF has developed several fragmentation indices Grassland, savanna, and shrubland, and flooded (Olson et al. 1994) that can be used to assess prop- grassland MHTs were assessed for both tragmen- erties such as (a) the proportion of core habitat tation and degradation. For these MHTs, frag- (habitat unaffected by edge effects using a prede- mentation receives half the points allotted above termined effect distance) in a fragmented land- for each category, and habitat degradation is scape; (b) the relative isolation of local fragments rated as follows: (multi-directional) based on inter-fragment distances relevant for most species and ecological processes; Points Degree of Habitat Degradation (c) the relative isolation of fragment clusters based 0 Low: populations of native plant species, on long distance dispersal abilities of vagile species; successional processes, and disturbance and (d) the influence of land use in areas between regimes relatively unaffected by anthro- the fragments and type of habitat on the ecological pogenic grazing and burning. isolation of fragments or fragment clusters. These 5 Medium: population of native plant GIS-based fragmentation analyses were not used in species persist in reduced numbers; this study for the reasons discussed earlier in this succession and disturbance processes appendix. Instead, we relied on regional experts' modified. classification of ecoregions into broad fragmen- 10 High: few native plant species persist; tation scenarios (humans are actually quite good large native herbivores eliminated; at assessing fragmentation scenarios relative to succession and disturbance processes computers). significantly altered. Points Degree of Habitat Fragmentation Habitat Conversion 0 Relatively contiguous: high connectivity; fragmentation low; long-distance dis- For each ecoregion, we estimated the recent rate persal along elevational and climatic of habitat conversion (i.e., proportion of the re- gradients still possible. maining habitat in an ecoregion being converted 5 Low: higher connectivity; more than half from intact to altered habitat per year). We tried of all fragments clustered to some de- to estimate rates relevant for the previous five- gree (i.e., have some degree of interac- year period. tion with other intact habitat blocks). 12 Medium: intermediate connectivity; frag- Points Conversion per Annum ments somewhat clustered; intervening 0 <0.5 percent landscape allows for dispersal of many 6 0.5-2.0 percent taxa through some parts of ecoregion. 8 2.1-3 percent 16 Advanced: low connectivity; more larger 9 3.1-4 percent fragments than in High category; frag- 1 >4 percent ments highly isolated; intervening land- scape precludes dispersal for most Future large-scale conversions predicted on the 20 High: most fragments small and or non- basis of impending development projects, exploitation circular; little core habitat due to edge plans, colonization, or other factors are addressed in effects (e.g., extending for 0.75-1.0 km the threat analyses. for physical edge effects and for 40 km for hunting pressure); most individual Degree of Protection fragments and clusters of fragments highly isolated; intervening landscapes As for the habitat block criterion, Tables A-5 to A-7 preclude dispersal for most taxa. should be read from top to bottom, within any given size category, until a statement is reached that is true In ecoregions that are comprised of several for the ecoregion. A definition such as ">500" naturally disjunct areas, fragmentation should be should be interpreted as "the ecoregion contains at assessed within individual remaining blocks (the least one protected area which includes a block of largest blocks contribute the most to the assess- intact habitat greater than 500 km2." Percentage val- ment) and not on the basis of connectivity between ues refer to the portion of remaining intact habitat blocks. incorporated into a protected area system. 56 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean At the BSP workshop, some regional experts felt ecoregions. Future analyses should probably incor- that only protected areas with a staff of qualified porate such a refinement; Table A-7 suggests a pos- managers, transportation, and a budget should be sible set of criteria that were suggested by regional used in this analysis. Some experts also felt that a experts but which, however, were not used to obtain different set of criteria should be used for very large the final values in this study. Table A-5. Degree of Protection Analysis for Broadleaf and Conifer Forest METs Ecoregion Size Point Value >3,000 kn2 1,000-3,000 km2 100-1,000 kn2 <100 kin2 1 2 areas >1,000 2 areas >750 3 areas >250 >50 percent protected 4 >500 >500 >250 40-50 percent protected 6 >250 >250 >100 20-40 percent protected 8 >100 >100 Areas exist but 1-20 percent protected none >100 10 None >100 None >100 No areas No areas Note: The value ">1,000" means "ecoregion contains at least one protected area containing a block of intact habitat greater than 1,000 km2." For an ecoregion of any given size, this and the following tables should be read from top to bottom until a statement is reached that is true of the ecoregion. Table A-6. Degree of Protection Analysis for Grassland/Savanna/Shrubland and Xeric Formation METs Ecoregion Size Point Value >3,000 kn2 1,000-3,000 km2 100-1,000 km2 <100 kn2 1 2 areas >750 2 areas >500 2 areas >250 >50 percent protected 4 >500 >500 >250 40-50 percent protected 6 >250 >250 >100 20-40 percent protected 8 >100 >100 Areas exist but 1-20 percent protected none >100 10 None >100 None >100 No areas No areas Table A-7. Degree of Protection Analysis Suggested for Large Ecoregions Ecoregion Size Point Value >500,000 kn2 200,000-500,000 kn2 50,000-200,000 km2 1 >8,000 >4,000 >2,000 4 3 areas >4,000 3 areas >2,000 3 areas >1,000 6 >2,000 >1,000 >500 8 >1,000 >500 >250 10 None >1,000 None >500 None >250 Methods Used for Assessing the Conservation Status of Terrestrial Ecoregions 57 Assessing Final Conservation Status The following table describes how points were al- located within each threat category. The snapshot conservation status of ecoregions was revised on the basis of a threat analysis to develop final conservation status assessments. For example, Poits Degree of Threat some forests may be largely intact and would Conversion Threats receive a snapshot status of Relatively Stable, but the 0 No conversion threats recognized for imminence of extensive logging concessions might ecoregion. warrant shifting the ecoregion's final conservation 10 Threats may significantly alter between status to Vulnerable. The final conservation status 5 percent and 9 percent of remaining assessments should reflect the urgency of con- habitat within 20 years. servation action as well as the ecological integrity of 20 Threats may significantly alter between ecoregions. 10 percent and 24 percent of remain- To evaluate as objectively as possible the threats ing habitat within 20 years. facing an ecoregion over about the next 20 years, we 50 Threats may significantly alter 25 percent estimated a threat value for each ecoregion of high, or more of remaining habitat within medium or low. Ecoregions with high threat 20 years. estimates were moved up one class in their conservation status ranking (e.g., from Vulnerable to Degradation Threats Endangered). Medium threat estimates influenced 0 No degradation threats recognized for the rankings depending upon our judgment. Low ecoregion. threat estimates were not used to modify snapshot 15 Populations of native plant species conservation status rankings. experiencing significant mortality and At the BSP Workshop, experts qualitatively poor recruitment due to degradation determined high, medium and low evaluations. factors. Succession and disturbance A more quantitative method, using points, was processes modified. Some abandonment generally used at the subsequent WWF LAC and underuse of seasonal, migratory, Program Workshop. This point system is outlined and breeding movements by species. below. For the most part, the latter determinations Pollutants and linked effects commonly are those that are included in this report. found in target species or assemblages. 30 Many populations of native plant species Method of Threat Analysis experiencing high mortality and low recruitment due to degradation factors. The major threats to each ecoregion in terms of their Succession and disturbance processes type, intensity, and timeframe are identified here. significantly altered. Low habitat qual- Threat analyses are inherently complex since factors ity for sensitive species. Abandonment may effect ecosystems directly or indirectly, and and disruption of seasonal, migratory, there are numerous and not well understood syner- or breeding movements. Pollutants and gistic interactions among factors. Thus, the formal linked effects widespread in ecosystem analysis of threats used here is based on the overall (e.g., recorded in several trophic levels). effect of threats on habitat modification and ecosys- tem degradation, regardless of the number or type of Wildlife Exploitation threats identified. 0 No wildlife exploitation recognized We used an index of 0-100 points to determine for ecoregion. pending threats to the ecoregion. Points were attri- 10 Moderate levels of wildlife exploitation; buted to three major types of threat: conversion populations of game or trade species threats (maximum of 50 points), degradation threats persisting but in reduced numbers. (maximum of 30), and wildlife exploitation threats 20 High intensity of wildlife exploitation in (maximum of 20 points). High threat was considered region with elimination of local equivalent to 70-100 total points, medium threat populations of most target species equivalent to 20-69, and low threat to 0-19 points. imminent or complete. 58 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Canbbean The bulleted list that follows describes the type of * firewood extraction threats in each category. * grazing * unsustainable extraction of non-timber products Conversion threats * road building and associated erosion and land- * intensive logging and associated road building slide damage * intensive grazing * off-road vehicle damage * agricultural expansion, plantations, and clearing * selective logging for development * excessive recreational impacts * permanent alteration from burning Wildlife exploitation Degradation threats * hunting and poaching * pollution (e.g., oil, pesticides, herbicides, mercury, * unsustainable extraction of wildlife and plants as heavy metals, defoliants) commercial products * burning * harassment and displacement by commercial and * introduced species recreational users Appendix B Methods Used for Assessing the Conservation Status of Mangrove Units The assessment of mangroves took place at the Points Conservation Status LAC Mangroves Workshop held by WWF Novem- 0-6 Relatively Intact ber 2-4, 1994. The method presented here is 7-36 Relatively Stable much abbreviated given that the general approach 37-64 Vulnerable is discussed in detail for terrestrial ecoregions. 65-88 Endangered Unless stated otherwise, refer to Appendix A for 89-100 Critical details on the justification and application of the method. A full report on the results of the man- grove workshop will be eventually published Determination of Point Values for Each by WWF. Criterion Determination of Snapshot Conservation Total Habitat Loss Status: Weighting and Categories The following definition of intact mangrove habitat Two landscape parameters were given greatest is used. Canopy opening through logging or other weight in the conservation status index: loss of human disturbance is restricted to less than original habitat and habitat blocks. The relative 10 percent of defined habitat block. No major contribution of parameters used in the index is: changes in hydrology or severe pollution immedi- ately threaten persistence of habitat block. Under- Weight story largely undisturbed from draining, timber (percent) Paraeter extraction, grazing, agriculture, bark harvesting, or 40 Total habitat loss human-caused fires. Although large mammals, birds, 20 Habitat blocks and reptiles may presently be absent from some 10 Degree of protection blocks of habitat due to exploitation or insufficient 10 Habitat conversion area, such blocks may still sustain many native 10 Water quality and hydrographic plant, invertebrate, and vertebrate species and their integrity associated ecological processes. 5 Habitat fragmentation 5 Degree of alteration of catchment Points Original Habitat Lost basin 0 0-10 percent 10 10-24 percent The point thresholds for different categories of 20 24-49 percent conservation status are listed below (classification 32 50-89 percent as "Extinct" is based on expert assessment): 40 > 90 percent 59 60 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Habitat Blocks Table B-1 presents the method used to determine point values for the habitat blocks criterion. Table B-1. Habitat Block Analysis for Mangrove Units Unit Size Point Value >3,000 kn2 1,000-3,000 kn2 <1,000 k1n2 2 >1000 or 3 blocks >500 >750 or 3 blocks >500 90-100 percent 5 >500 >500 70-90 percent 10 >250 3 blocks >200 40-70 percent 15 >100 >75 10-40 percent 20 None >100 None >75 <10 blocks Note: The value ">500" means "the unit contains at least one habitat block greater than 500 kM2." The value "90 percent" means "the unit contains at least one habitat block that is 90 percent the size of the largest original unit." For a unit of any given size, the table should be read from top to bottom until a statement is reached that is true of the unit. Degree of Protection The degree of protection analysis used for mangroves is presented in Table B-2. Table B-2. Degree of Protection Analysis for Mangrove Units Unit Size Point Value >1,000 km2 250-1,000 knm2 <250 knm2 1 3 areas >200 3 areas >100 >50 percent protected 4 >200 2 areas >100 40-50 percent protected 6 2 areas >100 >100 20-40 percent protected 8 >100 Areas exist but none >100 <20 percent protected 10 None >100 No areas No areas Note: The value ">200" means "the unit contains at least one protected area containing a block of intact habitat greater than 200 kM2." For a unit of any given size, this table should be read from top to bottom until a statement is reached that is true of the unit. Habitat Conversion Future large-scale conversions predicted on the basis of impending development projects, exploitation For each ecoregion, the recent rate of direct habitat plans, colonization, or other factors were addressed in conversion was estimated (i.e., proportion of the re- the threat analyses. maining original habitat within an ecoregion being converted from intact to altered habitat per year). Water Quality and Integrity of Hydrographic We tried to estimate rates that are relevant for about Processes the last five years. Mangrove ecosystems are known to be particularly Points Conversion per Annum sensitive to changes in water quality and hydro- 0 <0.5 percent graphic processes. The purpose of this criterion is to 6 0.5-2.0 percent provide a means to reassess the amount of available 8 2.1-3.0 percent high quality habitat estimated from the habitat loss 9 3.1-4.0 percent and large habitat blocks assessments. Even standing 10 >4.0 percent mangrove forests may soon represent inviable Methods Used for Assessrng the Conservation Status of Mangrove Units 61 habitat due to significant changes in water quality or Fragmentation was assessed within individual hydrographic processes. Important parameters are blocks and not on the basis of connectivity between pH, turbidity, dissolved oxygen, pesticides, heavy blocks. The state of the largest blocks contributes metals, suspended solids, hydrocarbons (oil), and relatively more to the assessment. alteration of tidal cycles and freshwater discharge. The following categories were used: Degree of Alteration of tHie Catchnent Basin Points Remaining Habitat Affected by Changes Changes in catchment basins that can affect man- 2 0-20 percent groves include alteration of the freshwater inflow 4 21-40 percent from rivers and runoff, increased sedimentation, 6 41-60 percent increased pollution, increased access to hunters and 8 61-80 percent loggers, and loss of surrounding wildlife populations 10 81-100 percent (i.e., assumed loss of man- and resource areas for mangrove associated species. grove without significant restoration) Points Catchment Basin Altered Habitat Fragmentation 0 0-19 percent 2 20-39 percent Mangrove ecoregions are comprised of naturally 3 40-59 percent disjunct habitats occurring along the coastlines of 4 60-89 percent continents and islands. Many habitat units them- 5 81-100 percent selves are highly fragmented due to channels and the patchy and dynamic nature of suitable substrate. Assessing Final Conservation Status Mangroves also disperse well and are able to colo- nize new areas relatively rapidly. For these reasons, The point totals for each of the landscape parameters mangrove forests and their associated biota are, with were summed and a snapshot conservation status some exceptions, not particularly sensitive to habitat assigned to each mangrove unit (as in Appendix A). fragmentation as long as natural hydrographic The snapshot conservation status assessments were conditions persist. subsequently modified by a threat analyses to better reflect the long-term trajectory of the mangrove unit Points Degree of Fragmentation based on significant, large-scale threats. 0 Relatively contiguous: high connectivity; fragmentation low; long-distance disper- Method of Threat Analysis sal of mangrove specialists still possible. 1 Low: higher connectivity; more than half of Suggested categories for threats to mangroves include: all fragments clustered to some degree (i.e., have some degree of interaction Conversion threats with other intact habitat blocks). * logging and charcoal production 3 Medium: intermediate connectivity; frag- * agricultural expansion and clearing for development ments somewhat clustered; intervening * salt extraction landscape allows for dispersal of many * shrimp farming taxa through some parts of ecoregion. * bark harvesting 4 Advanced: low connectivity; more larger * draining or channelization (alteration of tidal or fragments than in High category; frag- freshwater regimes) ments highly isolated; intervening land scape inhibits dispersal for most taxa. Degradation threats 5 High: most fragments small and or non- * pollution (e.g., oil, pesticides, heavy metals, circular; little core habitat due to edge defoliants) effects (e.g., extending for 0.75-1.0 km * burning for physical edge effects and for 40 km * introduced species for hunting pressure); most individual * firewood extraction fragments and clusters of fragments highly isolated; intervening landscapes Wildlife exploitation preclude dispersal for most taxa. * hunting, wildlife trade, and overfishing 62 A Conservation Assessment of the Terrestnal Ecoregions of Latin America and the Caribbean As in Appendix A, we used an index of 0-100 points to Modification of the snapshot conservation status of determine pending threats to the unit. High threat was mangrove units followed the same decision rules used considered equivalent to 70-100 points, medium threat for terrestrial ecoregions. The table below describes equivalent to 20-69, and low threat to 0-19 points. how points were allocated within each threat category. Points Description Points Description Conversion Threats Degradation Threats (continued) 0 No conversion threats recognized for 30 Many populations of native plant ecoregion. species experiencing high mortality 10 Threats may significantly alter between and low recruitment due to degra- 5 percent and 9 percent of remaining dation factors. Succession and distur- habitat within 20 years. bance processes significantly altered. 20 Threats may significantly alter between Pollutants and linked effects wide- 10 percent and 24 percent of remain- spread in ecosystem (e.g., recorded in ing habitat within 20 years. several trophic levels). 50 Threats may significantly alter 25 percent or more of remaining habitat within Wildlife Exploitation 20 years. 0 No wildlife exploitation recognized for ecoregion. Degradation Threats 10 Moderate levels of wildlife exploitation, 0 No degradation threats recognized for populations of game or trade species ecoregion. persisting but in reduced numbers 15 Populations of native plant species 20 High intensity of wildlife exploitation experiencing significant mortality and in region with elimination of local poor recruitment due to degradation populations of most target species factors. Succession and disturbance imminent or complete. processes modified. Pollutants and linked effects commonly found in target species or assemblages. Appendix C Definitions of Major Ecosystem Types and Major Habitat Types Tropical Broadleaf Forest MET Conservation Considerations: Tropical broadleaf forests typically require large protected areas to General Characteristics: These are closed canopy maintain viable populations, sustain ecological tropical forests dominated by broadleaf species. processes, and buffer core habitats from hunting Four distinguishing characteristics are (a) drip tips and other edge-related disturbances. High levels of on leaves; (b) smooth columnar boles of trees; beta diversity may warrant numerous, well- (c) presence of ant-plants; and (d) presence of true dispersed reserves in many cases. Biota very sensi- lianas. Other characteristics of these forests are: tive to edge effects and isolation-linkages among high species richness, particularly in wet, aseasonal habitat blocks should be of closed canopy forest. environments; propensity of many taxa toward Altitudinal linkages important for seasonal move- restricted geographic distributions (endemism) and ments of many bird, mammal, and invertebrate high habitat specificity; propensity toward a high species. degree of beta diversity, particularly in wetter, asea- sonal environments and in more complex environ- Tropical Moist Broadleaf Forest MHT ments; seed dispersal and pollination largely by animals, sometimes over large distances; high The general criterion used to define this MHT was numbers of obligate pollinators; high proportion of that less than 50 percent of canopy tree species be tightly linked ecological interactions (e.g., symbio- deciduous. ses); many tree, vertebrate, and invertebrate species occurring at relatively low densities (i.e., large areas Tropical Dry Broadleaf Forest MHT needed for viable populations); high area require- ments for some species due to patchy resource dis- The general criterion used to define this MHT was tribution (e.g., frugivores); and importance of key- that more than 50 percent of canopy tree species be stone predators and resource species (e.g., fig trees) deciduous. for maintaining ecosystem integrity. Conifer/Temperate Broadleaf Forest MET Sensitivity to Disturbance: Secondary extinctions and changes in species compositions are predict- General Characteristics: Forests are composed of able as a result of loss of keystone predators and conifers, mixed conifer/broadleafs, or temperate resource species. Many taxa, particularly under- broadleaf species. Characterized by lower levels of story species, are sensitive to changes in microcli- species richness, endemism, and beta diversity than mate. Movement of many species is restricted to tropical broadleaf forests, but regional and local intact, closed canopy forest. Widespread forest range-restricted species common in some regions. clearance can significantly alter soil structure and Narrow altitudinal zonation of communities com- successional patterns. Ecosystem processes and mon. Wind pollination common. Tree species occur populations significantly affected by habitat frag- at higher individual densities. Periodic distur- mentation due to sensitivities discussed above. bances can often affect large areas in some forest 63 64 A Conscration Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean types (e.g., fire, epizootics). Altitudinal linkages controlled. Riparian areas and sources of water are important for seasonally migrating species. Ripar- critical for many species. Introduced species can ian and bottonland communities often are critical have significant ecological impacts. climatic refuges and resource habitats for wildlife. Species are generally better adapted for dispersal Grassland, Savanna, and Shrubland MHT across non-forest habitats than tropical broadleaf species. Understory species sensitive to overgrazing Characteristics of this MHT are as given for the and frequent burning. MET above. Conservation Considerations: Large reserves are Flooded Grassland MHT required to maintain viable populations of larger predators and to help ensure that some undis- These areas are regularly inundated by water and turbed habitat remains after large disturbance are particularly sensitive to pollution and changes events to act as refuges for wildlife and source in patterns of flooding from draining, channeliza- pools for species and to maintain ecological pro- tion, or alteration of surrounding habitat. cesses. Habitat linkages along altitudinal gradients are critical as is conservation of riparian and bottom- Montane Grassland MHT land habitat. Grazing and anthropogenic fires should be strictly controlled. Some montane communities regenerate slowly and may have low resiliency to disturbances such as Temperate Forest MHT grazing or burning. Plant species may have patchy distributions due to altitudinal zonation or other The forests of the southern cone of South America specific habitat requirements. Larger vertebrates are the only representatives of this MHT in Latin may require intact altitudinal corridors for seasonal America and the Caribbean. movements or large reserves to maintain sufficient resources in patchy environments. Tropical and Subtropical Coniferous Forest MHT Xeric Formation MET Some taxa display pronounced endemism and narrow altitudinal zonation. Isolated montane General Characteristics and Conservation Consi- forests may require numerous, well-dispersed derations: Xeric formations can be characterized by reserves for conservation of their biodiversity. relatively rich floras and a propensity for ende- mism in some taxa. These communities are sensi- Grassland/Savanna/Shrubland MET tive to overgrazing, burning, invasion by exotics, and clearing. Many species sensitive to fragmenta- General Characteristics: This category encom- tion and the loss of critical habitats such as riparian passes a wide range of community types with vari- zones and waterholes. Populations of larger verte- ous combinations of forest, grassland, shrublands, brates often occur at low densities. and woodlands. Their shared features include: many species of plants and animals widely dis- Mediterranean Scrub MHT tributed within ecoregions; communities relatively resilient to short-term disturbances but not to pro- These scrublands with Mediterranean-type climates longed burning or overgrazing; and sensitivity to are found only in two areas of LAC: central Chile fragmentation (i.e., from habitat loss and fencing), and the coast of California. They are characterized which disrupts movement patterns of larger verte- by: high species richness, particularly for plants brates that rely on patchy or seasonal resources. and invertebrates; very high levels of beta diversity and endemism (both local and regional); sensitivity Conservation Considerations: Small reserves may to overgrazing and exotic species; and sensitivity of effectively conserve plant and invertebrate species many vertebrate species to fragmentation. Small but need to be large enough to avoid complete alter- reserves may be effective for conserving many ation by fire or edge effects. Larger reserves needed components of biodiversity, particularly plants and to conserve intact vertebrate faunas, particularly invertebrates, but they must be large enough to be for species tracking patchy or seasonal resources. resilient to disturbance (e.g., retain some undis- Hunting, burning, and grazing need to be strictly turbed habitat after periodic fire events). Definitions of Major Ecosystem Types and Major Habitat Types 65 Desert and Xeric Shrubland MHT reduced substrates. These substrates are often poorly consolidated and subject to rapid changes. Characteristics of this MHT are as given for the Mangrove forest ecosystems receive inputs of MET above. matter and energy from both land and sea. These inputs include fresh water, sediments and nutrients Restinga MHT from land and tidal flushing and saline intrusions from the sea. These inputs act as energy subsidies Restingas refer to habitats of xeromorphic coastal that increase the performance of the system and dune or sandy soil vegetation, ranging from low help maintain high rates of organic matter fixation herbaceous ground cover to open forests. They are and active ecosystem processes. extensive in some areas with rich floras and many Forests growing in areas periodically inundated endemics and are sensitive to overgrazing, burning, by brackish or marine water and comprised of and disturbance of dunes. trees in the genera Avicennia, Conocarpus, Laguncu- laria, Pelliciera, and Rhizophora; high species rich- Mangrove MET ness when associated marine and freshwater species are considered; low endemism and beta General Characteristics and Conservation Consi- diversity; habitat fragmentation not a major derations: The word mangrove is used to describe problem at local scales because of rapid regenera- a salt-tolerant forest ecosystem that occupies tion of mangroves (as long as necessary hydro- sheltered tropical and subtropical coastal estu- graphic conditions are still present) and adapta- arine, lagoon, deltaic, or carbonate platform envi- tion of biota to patchiness of habitat; mangroves ronments. Although the constituent plant species are particularly sensitive to changes in hydrogra- (also called mangroves) are not closely related, phy (e.g., draining or changes in tidal or river they have common morphological, physiological outflow patterns) and pollution; surrounding and reproductive adaptations that allow survival forests often contain important resources for and development in very saline, waterlogged and terrestrial wildlife. Appendix D Hierarchical Classification Scheme of LAC Ecoregions MAJOR ECOSYSTEM TYPE 14. Costa Rican seasonal moist forests-Costa Rica, Nicaragua Major Habitat Type 15. Isthmian-Pacific moist forests-Costa Rica, Panama Bioregion 16. Talamancan montane forests - Costa Rica, Ecoregion -countries in which ecoregion Panama occurs (ecoregion numbers correspond to those on the large-format map) Orinoco 17. Cordillera La Costa montane forests - TROPICAL BROADLEAF FORESTS Venezuela 18. Orinoco Delta swamp forests-Venezuela, Tropical Moist Broadleaf Forests Guyana 19. Trinidad & Tobago moist forests - Caribbean Trinidad & Tobago 1. Cuban moist forests-Cuba 20. Guianan Highlands moist forests- 2. Hispaniolan moist forests-Haiti, Venezuela, Brazil, Guyana Dominican Republic 21. Tepuis-Venezuela, Brazil, Guyana, 3. Jamaican moist forests - Jamaica Suriname, Colombia 4. Puerto Rican moist forests -Puerto Rico 5. Windward Islands moist forests- Amazonia Windward Islands 22. Napo moist forests-Peru, Ecuador, 6. Leeward Islands moist forests -Leeward Colombia Islands 23. Macarena montane forests-Colombia 24. Japura/ Negro moist forests-Colombia, Central America Venezuela, Brazil, Peru 7. Oaxacan moist forests - Mexico 25. Uatama moist forests- Brazil, Venezuela, 8. Tehuantepec moist forests -Mexico, Guyana Guatemala, Belize 26. Amapa moist forests-Brazil, Suriname 9. YucatAn moist forests - Mexico 27. Guianan moist forests- Venezuela, 10. Sierra Madre moist forests- Mexico, Guyana, Suriname, Brazil, French Guiana Guatemala, El Salvador 28. Paramaribo swamp forests -Suriname 11. Central American montane forests- 29. Ucayali moist forests-Brazil, Peru Mexico, Guatemala, El Salvador, Honduras 30. Western Amazonian swamp forests -Peru, 12. Belizean swamp forests-Belize Colombia 13. Central American Atlantic moist forests - 31. Southwestern Amazonian moist forests - Guatemala, Belize, Honduras, Nicaragua, Brazil, Peru, Bolivia Costa Rica, Panama 32. JuruA moist forests-Brazil 66 Hierarchical Classification Scheme of LAC Ecoregions 67 33. Varzea forests-Brazil, Peru, Colombia 62. Windward Islands dry forests-Windward 34. Purus/Madeira moist forests - Brazil Islands 35. Rond6nia/Mato Grosso moist forests- 63. Leeward Islands dry forests-Leeward Brazil, Bolivia Islands 36. Beni swamp and gallery forests - Bolivia, Brazil Northern Mexico 37. Tapaj6s/Xingu moist forests -Brazil 64. Baja California dry forests -Mexico 38. Tocantins moist forests-Brazil 65. Sinaloan dry forests-Mexico 66. Tamaulipas/ Veracruz dry forests -Mexico Northern Andes 39. Choc6/DariCn moist forests-Colombia, Central America Panama 67. Jalisco dry forests-Mexico 40. Eastern Panamanian montane forests - 68. Balsas dry forests -Mexico Panama, Colombia 69. Oaxacan dry forests -Mexico 41. Northwestern Andean montane forests- 70. Veracruz dry forests-Mexico Colombia, Ecuador 71. Yucatan dry forests -Mexico 42. Western Ecuador moist forests -Ecuador, 72. Central American Pacific dry forests - Colombia El Salvador, Honduras, Nicaragua, Costa 43. Cauca Valley montane forests-Colombia Rica, Guatemala 44. Magdalena Valley montane forests- 73. Panamanian dry forests-Panama Colombia 45. Magdalena/Urabd moist forests- Orinoco Colombia 74. Llanos dry forests - Venezuela 46. Cordillera Oriental montane forests- 75. Trinidad & Tobago dry forests-Trinidad Colombia, Venezuela & Tobago 47. Eastern Cordillera Real montane forests - Ecuador, Colombia, Peru Amazonia 48. Santa Marta montane forests -Colombia 76. Bolivian lowland dry forests -Bolivia, 49. Venezuelan Andes montane forests - Brazil Venezuela, Colombia 50. Catatumbo moist forests-Venezuela, Northern Andes Colombia 77. Cauca Valley dry forests-Colombia 78. Magdalena Valley dry forests-Colombia Central Andes 79. Patia Valley dry forests-Colombia 51. Peruvian Yungas-Peru 80. Sini Valley dry forests-Colombia 52. Bolivian Yungas-Bolivia, Argentina 81. Ecuadorian dry forests-Ecuador 53. Andean Yungas-Argentina, Bolivia 82. Tumbes/Piura dry forests-Ecuador, Peru 83. Marafi6n dry forests-Peru Eastern South America 84. Maracaibo dry forests -Venezuela 54. Brazilian Coastal Atlantic forests-Brazil 85. Lara/Falc6n dry forests-Venezuela 55. Brazilian Interior Atlantic forests -Brazil, Argentina, Paraguay Central Andes Tropical Dry Broadleaf Forests 86. Bolivian montane dry forests - Bolivia CONIFER/TrEMPERATE BROADLEAF Caribbean FORESTS 56. Cuban dry forests-Cuba 57. Hispaniolan dry forests-Haiti, Dominican Temperate Forests Republic 58. Jamaican dry forests-Jamaica Southern South America 59. Puerto Rican dry forests -Puerto Rico 87. Chilean winter-rain forests -Chile 60. Bahamian dry forests -Bahamas, Turks & 88. Valdivian temperate forests -Chile, Caicos Islands Argentina 61. Cayman Islands dry forests-Cayman 89. Subpolar Nothofagus forests-Chile, Islands Argentina 68 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Tropical and Subtropical Coniferous Forests 112. Amazonian savannas - Brazil, Colombia, Venezuela Caribbean 113. Beni savannas-Bolivia 90. Cuban pine forests-Cuba 91. Hispaniolan pine forests-Haiti, Eastern South America Dominican Republic 114. Cerrado-Brazil, Paraguay, Bolivia 92. Bahamian pine forests-Bahamas, Turks & 115. Chaco savannas-Argentina, Paraguay, Caicos Islands Bolivia, Brazil 116. Humid Chaco -Argentina, Paraguay, Northern Mexico Bolivia, Brazil 93. Sierra Juarez pine-oak forests- Mexico, 117. C6rdoba montane savannas-Argentina U.S. 94. San Lucan pine-oak forests-Mexico Southern South America 95. Sierra Madre Occidental pine-oak forests- 118. Argentine Monte - Argentina Mexico, U.S. 119, Argentine Espinal-Argentina 96. Central Mexican pine-oak forests- Mexico 120. Pampas - Argentina 97. Sierra Madre Oriental pine-oak forests- 121. Uruguayan savannas-Uruguay, Brazil, Mexico Argentina 98. Veracruz pine-oak forests -Mexico Flooded Grasslands Central America 99. Mexican transvolcanic pine-oak forests- Caribbean Mexico 122. Cuban wetlands-Cuba 100. Veracruz montane forests - Mexico 123. Enriquillo wetlands - Haiti, Dominican 101. Sierra Madre del Sur pine-oak forests- Republic Mexico 102. Central American pine-oak forests- Northern Mexico Mexico, Guatemala, El Salvador, Honduras, 124. Central Mexican wetlands-Mexico Nicaragua 103. Belizean pine forests -Belize Central America 104. Miskito pine forests-Nicaragua, 125. Jalisco palm savannas-Mexico Honduras 126. Veracruz palm savannas - Mexico 127. Quintana Roo wetlands-Mexico Eastern South America 105. Brazilian Araucaria forests-Brazil, Orinoco Argentina 128. Orinoco wetlands-Venezuela GRASSLANDS/SAVANNAS/SHRUBLANDS Amazonia 129. Western Amazonian flooded Grasslands, Savannas, and Shrublands grasslands-Peru, Bolivia 130. Eastern Amazonian flooded grasslands- Northern Mexico Brazil 106. Central Mexican grasslands - Mexico, U.S. 131. S5o Luis flooded grasslands-Brazil 107. Eastern Mexican grasslands - Mexico Northern Andes Central America 132. Guayaquil flooded grasslands-Ecuador 108. Tabasco/Veracruz savannas- Mexico 109. Tehuantepec savannas-Mexico Eastern South America 133. Pantanal -Brazil, Bolivia, Paraguay Orinoco 134. Parani flooded savannas - Argentina 110. LIanos - Venezuela, Colombia Montane Grasslands Amazonia 111. Guianan savannas-Suriname, Guyana, Central America Brazil, Venezuela 135. Mexican alpine tundra-Mexico Hierarchical Classification Scheme of LAC Ecoregions 69 136. Costa Rican paramo-Costa Rica 159. Chihuahuan xeric scrub-Mexico, U.S. 160. Central Mexican mezquital -Mexico Northern Andes 161. Eastern Mexican matorral-Mexico 137. Santa Marta paramo-Colombia 162. Eastern Mexican mezquital-Mexico, U.S. 138. Cordillera de Merida paramo- 163. Central Mexican cactus scrub-Mexico Venezuela 139. Northern Andean paramo-Colombia, Central America Ecuador 164. Pueblan xeric scrub-Mexico 165. Guerreran cactus scrub -Mexico Central Andes 166. Motagua Valley thornscrub-Guatemala 140. Cordillera Central paramo-Ecuador, Peru Orinoco 141. Central Andean puna-Bolivia, 167. Aruba/Curagao/Bonaire cactus scrub- Argentina, Peru, Chile Netherlands Antilles 142. Central Andean wet puna -Peru, Bolivia, 168. La Costa xeric shrublands -Venezuela Chile 169. Araya and Paria xeric scrub-Venezuela 143. Central Andean dry puna - Argentina, Bolivia, Chile Northern Andes 170. Galapagos Islands xeric scrub-Ecuador Southern South America 171. Guajira/Barranquilla xeric scrub- 144. Southern Andean steppe-Argentina, Colombia, Venezuela Chile 172. Paraguand xeric scrub-Venezuela 145. Patagonian steppe-Argentina, Chile 146. Patagonian grasslands-Argentina, Chile Central Andes 173. Sechura desert-Peru, Chile XERIC FORMATIONS 174. Atacama desert-Chile Mediterranean Scrub Eastern South America 175. Caatinga -Brazil Northern Mexico 147. California coastal sage-chaparral - Restingas Mexico, U.S. Northern Andes Central Andes 176. Paraguani restingas- Venezuela 148. Chilean matorral-Chile Eastern South America Deserts and Xeric Shrublands 177. Northeastern Brazil restingas-Brazil Caribbean 178. Brazilian Atlantic Coast restingas -Brazil 149. Cuban cactus scrub-Cuba MANGROVES 150. Cayman Islands xeric scrub-Cayman Islands Atlantic Mangrove Complexes and 151. Windward Islands xeric scrub- Mangrove Units Windward Islands 152. Leeward Islands xeric scrub - Leeward Gulf of Mexico Islands Alvarado-Mexico Usumacinta - Mexico Northern Mexico 153. Baja California xeric scrub- Mexico Yucatan 154. San Lucan mezquital - Mexico Petenes - Mexico 155. Western Mexican mezquital -Mexico, U.S. Rio Lagartos -Mexico 156. Sonoran xeric scrub -Mexico, U.S. Mayan Corridor -Mexico 157. Northern Sonoran cactus scrub-Mexico, Belizean Coast-Mexico, Belize, U.S. Guatemala 158. Mexican Interior chaparral - Mexico, U.S. Belizean Reef - Mexico, Belize 70 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Atlantic Central America Segment VI: Rec6ncavo Baiano-Cabo Northern Honduras -Honduras Frio - Brazil Mosquitia/ Nicaraguan Caribbean Coast - Honduras, Nicaragua SE Brazil Rio Negro/Rio San Sun-Nicaragua, Segment VII: Cabo Frio-Laguna-Brazil Costa Rica Bocas del Toro/Bastimentos Island/San Pacific Mangrove Complexes and Mangrove Blas-Panama Units West Indies Sea of Cortez Bahamas-Bahamas, Turks & Caicos NW Mexican Coast-Mexico Islands, U.S. Marismas Nacionales/ San Blas - Mexico Greater Antilles-Cuba, Jamaica, Hispaniola, Puerto Rico, Cayman Southern Mexico Islands Southern Pacific Coast of Mexico - Mexico Lesser Antilles - Lesser Antilles Pacific Central America Continental Caribbean Tehuantepec/ El Manch6n - Mexico, Coastal Venezuela -Venezuela, Guatemala Colombia, Netherlands Antilles Northern Dry Pacific Coast-Guatemala, Magdalena/Santa Marta- El Salvador Colombia Gulf of Fonseca - El Salvador, Honduras, Nicaragua Amazon-Orinoco-Maranhalo Southern Dry Pacific Coast- Nicaragua, Trinidad-Trinidad & Tobago Costa Rica Segment 0: Orinoco-Cabo Orange- Moist Pacific Coast-Costa Rica, Panama Venezuela, Guyana, French Guiana Panama Dry Pacific Coast-Panama Segment 1: Cabo Orange-Cabo Norte- Brazil Pacific South America Segment II: Cabo Norte-Ponta CurucA - Gulf of Panama -Panama, Colombia Brazil Esmeraldas/ Pacific Colombia - Segment III: Ponta Curuga-Parnaiba Colombia, Ecuador Delta - Brazil Manabi- Ecuador Gulf of Guayaquil/Tumbes- Ecuador, NE Brazil Peru Segment IV: Parnaiba Delta-Cabo Piura-Peru Calcanhar - Brazil Segment V: Cabo Calcanhar-Rec6ncavo Galapagos Baiano- Brazil Galapagos Islands - Ecuador Appendix E Results of Assessments of Landscape-Level Criteria, Conservation Status, and Biological Distinctiveness of Non-Mangrove Ecoregions MAJOR ECOSYSTEM TYPE Snapshot Filial Major Habitat Type Conser- Conser- Biological Bo- Bioregion Ecoregion Ecoregion Size Habitat Habitat Fragmen- Degra- Conver- Protec- vation vation Distinc- diversity Ecoregion Number (k12) Loss a Blocks b tation c dation 1 sion e tronl f Total g Status i Status trweness I Priorityk TROPICAL BROADLEAF FORESTS Tropical Moist Broadleaf Forests Caribbean Tropical Moist Forests .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . .. . . . . . . . . . ....................................................................................... 6 03 I.. . . .. . . . . Cuban moist forests-Cuba 1 20,069 32 10 8 n.a. 6 4 60 3 3 2 1 Hispaniolan moist forests - Haiti, Dominican Republic 2 43,136 40 10 12 n.a. 8 4 74 2 2 2 n Jamaican moist forests - Jamaica 3 7,849 40 15 16 n.a. 8 10 89 2 2 2 1 Puerto Rican moist forests - Puerto Rico 4 7,237 32 15 12 n.a. 0 6 65 2 3 3 Windward Islands moist forests-Windward Islands 5 1,914 20 5 0 n.a. 8 8 41 3 3 3 Leeward Islands moist forests -Leeward Islands 6 951 10 5 0 n.a. 0 1 16 4 4 3 111 Central American Tropical Moist Forests . ... . ....................................................I ......................................... .. ............ ..........- .. .. . .. . .. .. . .. . .. .. . .. . 2.. . . . . . . . . . . . . . . . . . . 2.. . . . . . . . . . . . . . . . . . 3 II.. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . .. . . . . ... . . . . . . . . . Oaxacan moist forests-Mexico 7 4,715 - - - n.a. - - -3 I Tehuantepec moist forests - Mexico, Guatemala 8 146,752 20 5 16 n.a. 6 6 53 3 2 3 11* YucatAn moist forests-Mexico 9 64,012 20 2 0 n.a. 6 4 32 4 3 3 III Sierra Madre moist forests - Mexico, Guatemala, El 10 9,137 32 20 5 n.a. 9 8 74 2 2 3 I1 Salvador Note: Footnotes explaining the symbols and numerical codes in this table are found at the end of the appendix. MAJOR ECOSYSTEM TYPE Majo Habtat ypeSnapshot Final Major Habitat Type Conser- Conser- Biological Bio- Bioregion Ecoregion Ecoregion Size Habitat Habitat Fragmen- Degra- Conver- Protec- vation vation Distinc- diversty Ecoregion Number (km) Loss a Blocks b tation c dation d sion e -ion Total g Status Status Z tiveiess I Priority Central American Tropical Moist Forests (continued) Central American montane forests-Mexico, El Salvador, 11 7,676 20 15 16 n.a. 6 6 63 3 2 3 II Guatemala, Honduras Belizean swamp forests-Belize 12 4,150 32 15 12 n.a. 6 10 75 2 2 4 III Central American Atlantic moist forests-Guatemala, 13 155,020 32 2 12 n.a. 10 1 57 3 3 3 III Belize Honduras, Nicaragua, Costa Rica, Panama Costa Rican seasonal moist forests-Costa Rica, 14 10,620 40 20 20 n.a. 8 10 98 1 1 4 III Nicaragua Isthmian-Pacific moist forests - Costa Rica, Panama 15 28,881 32 15 16 n.a. 6 6 75 2 2 3 II Talamancan montane forests- Costa Rica, Panama 16 15,991 20 5 0 n.a. 6 1 32 4 4 2 II Orinoco Tropical Moist Forests Cordillera La Costa montane forests-Venezuela 17 13,481 32 2 12 n.a. 0 1 47 3 3 2 1 Orinoco Delta swamp forests - Venezuela, Guyana 18 31,698 0 2 0 n.a. 0 1 3 5 4 3 III Trinidad & Tobago moist forests-Trinidad & Tobago 19 4,456 20 2 12 n.a. 6 4 44 3 3 4 IV Guianan highlands moist forests-Venezuela, Brazil, 20 248,018 - - - n.a. - - - 5 5 2 III Guyana Tepuis -Venezuela, Brazil, Guyana, Suriname, Colombia 21 49,157 0 2 0 n.a. 0 1 3 5 5 1 II Napo moist forests - Peru, Ecuador, Colombia 22 369,847 0 2 5 n.a. 6 1 14 4 4 1 1 Amazonian Tropical Moist Forests Macarena montane forests - Colombia 23 2,366 10 5 5 n.a. 6 10 36 4 3 2 I Japura/ Negro moist forests- Colombia, Venezuela, Brazil, 24 718,551 0 2 0 n.a. 0 1 3 5 5 1 II Peru Uatama moist forests - Brazil, Venezuela, Guyana 25 288,128 0 2 0 n.a. 0 4 6 5 4 3 III AmapA moist forests - Brazil, Suriname 26 195,120 0 2 0 n.a. 0 4 6 5 4 3 III Guianan moist forests - Venezuela, Guyana, 27 457,017 0 2 0 n.a. 0 1 3 5 4 3 Ill Suriname, Brazil, French Guiana Paramaribo swamp forests - Suriname 28 7,760 10 2 5 n.a. 8 6 31 4 3 3 III Ucayali moist forests-Brazil, Peru 29 173,527 10 2 0 n.a. 6 4 22 4 3 1 1 Western Amazoman swamp forests-Peru, Colombia 30 8,315 - - - n.a. - - - 4 4 1 I Southwestern Amazonian moist forests-Brazil, Peru, 31 534,316 0 2 0 n.a. 0 1 3 5 4 1 1 Bolivia Jurud moist forests-Brazil 32 361,055 0 2 0 n.a. 0 1 3 5 5 2 III Varzea forests-Brazil, Peru, Colombia 33 193,129 20 2 12 n.a. 6 4 44 3 3 1 I Purus/Madeira moist forests- Brazil 34 561,765 0 2 0 n.a. 0 6 8 4 4 4 IV Rond6nia/Mato Grosso moist forests-Brazil, Bolivia 35 645,089 10 2 5 n.a. 6 4 27 4 3 2 II Beni swamp and gallery forests-Bolivia 36 31,329 0 2 0 n.a. 0 4 6 5 4 4 IV Tapaj6s/Xingu moist forests - Brazil 37 630,905 10 2 5 n.a. 6 6 29 4 3 4 IV Tocantins moist forests-Brazil 38 279,419 32 5 16 n.a. 6 8 67 2 2 4 III Northern Andean Tropical Moist Forests Choc6/Dari6n moist forests- Colombia, Panama, Ecuador 39 82,079 10 2 0 n.a. 9 8 29 4 3 1 I Eastern Panamanian montane forests-Panama, Colombia 40 2,905 0 10 12 n.a. 6 8 36 4 3 3 III Northwestern Andean montane forests -Colombia, 41 52,937 20 10 12 n.a. 6 6 54 3 2 1 I Ecuador MAJOR ECOSYSTEM TYPE Snapshot Final Major Habitat Type Conser- Conser- Biological Bio- Bioregion Ecoregion Ecoregion Size Habitat Habitat Fragmen- Degra- Conver- Protec- vation vatzon Distinc- dversihj Ecoregion Number (kM2) Loss a Blocks b tation c dation d sion e to / Total 9 Status Status I tweness) Priority Northern Andean Tropical Moist Forests (continued) .. . . Western Ecuador moist forests-Ecuador, Colombia 42 40,218 40 20 12 n.a. 6 8 86 2 1 2 I Cauca Valle montane forests- Colombia 43 32,412 32 20 20 n.a. 6 10 88 2 1 1 I Magdalena Valley montane forests-Colombia 44 49,322 32 20 20 n.a. 6 10 88 2 1 1 I Magdalena/UrabA moist forests - Colombia 45 73,660 32 15 16 n.a. 6 10 79 2 2 3 II Cordillera Oriental montane forests-Colombia, 46 66,712 20 2 12 n.a. 9 4 47 3 3 1 I ri Venezuela Eastern Cordillera Real montane forests - Ecuador, 47 84,442 20 2 12 n.a. 8 8 50 3 3 1 I Colombia, Peru Santa Marta montane forests - Colombia 48 4,707 32 5 5 n.a. 6 4 52 3 3 2 I Venezuelan Andes montane forests-Venezuela, 49 16,638 10 15 16 n.a. 6 4 51 3 2 1 1 Colombia Catatumbo moist forests - Venezuela, Colombia 50 21,813 40 20 20 n.a. 0 10 90 1 1 4 III Central A ndean.Tropical M oist Forests .... ...... ... .. ................................ . ................. Peruvian Yungas - Peru 51 188,735 20 2 12 n.a. 8 8 50 3 2 1 I Bolivian Yungas - Bolivia, Argentina 52 72,517 20 2 12 n.a. 8 8 50 3 2 2 I Andean Yungas - Argentina, Bolivia 53 55,457 20 2 12 n.a. 6 1 41 3 3 3 III Eastern South A m erican Tropical M oist Forests . ........ ...................................... .................. Brazilian Coastal Atlantic forests-Brazil 54 233,266 40 15 20 n.a. 10 6 91 1 1 1 1 Brazilian Interior Atlantic forests-Brazil, Argentina, 55 803,908 32 5 20 n.a. 8 7 72 2 2 2 I Paraguay Tropical Dry Broadleaf Forests Caribbean Tropical Dr/ Forests Cuban dry forests-Cuba 56 61,466 40 10 16 n.a. 6 8 80 2 2 3 11* Hispaniolan dry forests - Haiti, Dominican Republic 57 14,610 32 5 10 n.a. 6 4 57 3 2 3 II Jamaican dry forests - Jamaica 58 2,189 20 20 5 n.a. 10 10 65 2 2 4 III Puerto Rican dry forests - Puerto Rico 59 1,295 32 20 16 n.a. 6 10 84 2 2 4 III Bahamian dry forests - Bahamas, Turks & Caicos 60 9,293 20 5 5 n.a. 8 8 46 4 3 4 IV Islands Cayman Islands dry forests - Cayman Islands 61 230 - - - n.a. - - - - 2 4 III Windward Islands dry forests-Windward Islands 62 431 32 10 5 n.a. 6 8 61 3 2 4 III Leeward Islands dry forests- Leeward Islands 63 182 40 15 20 n.a. 6 8 89 1 1 4 III Northern Mexican Tropical Dry Forests Baja California dry forests- Mexico 64 93 0 2 0 n.a. 0 10 12 4 4 4 IV Sinaloan dry forests-Mexico 65 119,184 20 5 5 n.a. 8 10 48 3 3 3 III Tamaulipas/Veracruz dry forests-Mexico 66 40,858 32 15 16 n.a. 6 10 79 2 2 4 111* Central American Tropical Dry Forests Jalisco dry forests-Mexico 67 19,973 32 5 12 n.a. 9 8 66 2 2 2 I Balsas dry forests-Mexico 68 161,098 20 2 12 n.a. 8 10 52 3 2 2 I Oaxacan dry forests-Mexico 69 10,566 32 15 16 n.a. 6 10 79 2 2 3 II Veracruz dry forests-Mexico 70 35,546 40 20 20 n.a. 3 10 93 1 1 4 III YucatAn dry forests-Mexico 71 45,554 32 15 16 n.a. 9 10 82 2 2 4 III Central American Pacific dry forests - El Salvador, 72 50,101 40 20 20 n.a. 6 4 90 1 1 3 II Honduras, Nicaragua, Costa Rica, Guatemala Panamanian dry forests - Panama 73 5,010 40 20 20 n.a. 6 8 94 1 1 4 III U1 MAJOR ECOSYSTEM TYPE Snapshot Final Major Habitat Type Conser- Conser- Biological Bio- Bioregion Ecoregion Ecoregion Size Habitat Habitat Fragmen- Degra- Conver- Protec- vation vation Distinc- diversity Ecoregion Number (kni2) Loss a Blocks tation C datzon d sion e tion f Total 8 Status I Status tivenessI Priority Orinoco Tropical Dry Forests .. .............................. .................. . ......................... . . . . . . . ......... . . .............. .. . . .. . . . .. . . . - . . . .. . . - . . . . . . . . . . . .. . . . . . . . . . .4.. . . . . . . . . . .. . . . ... . . . .II I*.. . . . . . . .. . . . . . . . . . .. . . . . . . . . . . .. . . . . . . . . . . Llanos dry forests -Venezuela 74 44,177 32 15 16 n.a. 6 10 79 2 2 4 111* Trinidad & Tobago dry forests -Trini ad & Tobago 75 251 32 15 12 n.a. 6 8 73 2 2 4 III Amnazonian Tropical Dril Forests.s Bolivian lowland dry forests- Bolivia, Brazil 76 156,814 40 15 20 n.a. 0 10 85 2 1 1 I Northern Andean Tropical Dn/ Forests Cauca Valley dry forests - Colombia 77 5,130 40 20 20 n.a. 6 10 96 1 1 4 III Magdalena Valley dry forests- Colombia 78 13,837 40 20 20 n.a. 6 10 96 1 1 4 III Patia Valley dry forests - Colombia 79 1,291 40 20 20 n.a. 6 10 96 1 1 4 III Sini Valley dry forests - Colombia 80 55,473 40 20 20 n.a. 6 10 96 1 1 4 III Ecuadorian dry forests- Ecuador 81 22,271 40 20 20 n.a. 6 10 96 1 1 1 1 Tumbes/ Piura dry forests - Ecuador, Peru 82 64,588 32 15 12 n.a. 6 6 71 2 2 1 1 Marafi6n dry forests-Peru 83 14,921 - - - n.a. - - - 2 2 3 II Maracaibo dry forests- Venezuela 84 31,471 32 15 12 n.a. 6 8 73 2 2 4 Ill Lara/Falc6n dry forests - Venezuela 85 16,178 32 15 12 n.a. 6 8 73 2 2 4 III Central Andean Tropica. Dn/....o...es..s Bolivian montane dry forests - Bolivia 86 39,368 40 20 20 n.a. 6 10 96 1 1 3 11* CONIFER/TEMPERATE BROADLEAF FORESTS Temperate Forests Southern South American Temperate Forests .. Chilean winter-rain forests - Chile 87 24,937 32 15 18 n.a. 6 8 79 2 2 2 1 Valdivian temperate forests - Chile, Argentina 88 166,248 10 10 12 n.a. 8 1 41 3 3 1 I Subpolar Nothofagus forests- Chile, Argentina 89 141,120 0 2 12 n.a. 6 1 21 4 3 3 III Tropical and Subtropical Coniferous Forests Caribbean Tropical and Subtropical Conife rous Forests Cuban pine forests-Cuba 90 6,017 32 10 5 n.a. 3 8 58 3 3 2 1 Hispaniolan pine forests - Haiti, Dominican Republic 91 10,833 32 5 12 n.a. 6 4 59 3 3 2 I Bahamian pine forests - Bahamas, Turks & Caicos 92 3,920 32 10 0 n.a. 0 10 52 3 3 4 IV Islands - Northern Mexican Tropical and Subtropical Coniferous Forests ... Sierra Juarez ine-oak forests - Mexico, U. 93 24,229 20 2 5 n.a. 6 8 41 3 3 4 IV San Lucan pine-oak forests - Mexico, U.S. 94 895 0 2 0 n.a. 0 1 3 5 5 4 IV Sierra Madre Occidental pine-oak forests-Mexico, 95 204,374 32 15 12 n.a. 9 10 78 2 2 1 1 U.S. Central Mexican pine-oak forests - Mexico 96 3,719 32 10 20 n.a. 0 10 72 2 2 3 II Sierra Madre Oriental pine-oak forests -Mexico, U.S. 97 38,200 10 2 0 n.a. 8 6 26 4 4 2 II Veracruz pine-oak forests - Mexico 98 5,979 - - - n.a. - - - 2 1 4 III Central American Tropical and Subtropical Coniferous Forests Mexican transvolcanic pine- oak forests-Mexico 99 72,802 32 20 16 n.a. 8 10 86 2 2 2 I Veracruz montane forests - Mexico 100 6,615 0 2 0 n.a. 8 6 16 4 4 3 III Sierra Madre del Sur pine-oak forests-Mexico 101 41,129 - - - n.a. - - - 2 1 1 I 00 MAJOR ECOSYSTEM TYPE Snapshot Final Major Habitat Type Conser- Conser- Biological Bzo- Bioregion Ecoregion Ecoregion Size Habitat Habitat Fragmen- Degra- Conver- Protec- zation vation Distinc- diversity Ecoregion Number (kni2) Loss a Blocks b tation dation d szon e tion f Total g Status Status i tveness I Priority k Central American Tropical and Subtropical Coniferous Forests (continued) Central American pine-oak forests-Guatemala, El 102 127,910 32 2 12 n.a. 9 1 56 3 3 3 III Salvador, Honduras, Mexico, Nicaragua Belizean pine forests-Belize 103 2,750 - - - n.a. - - - 4 4 2 II Miskito pine forests - Nicaragua, Honduras 104 15,064 10 5 5 n.a. 6 10 36 4 4 2 II Eastern South American Tropical and Subtropical Coniferous Forests Brazilian Araucaria forests - Brazil, Argentina 105 206,459 40 20 20 n.a. 10 10 100 1 1 3 II* GRASSLANDS/SAVANNAS/SHRUBLANDS Grasslands, Savannas, and Shrublands Northern Mexican Grasslands, Savannas, and Shrublands Central Mexican grasslands - Mexico, U.S. 106 199,919 - - - - - - - - - 3 - Eastern Mexican grasslands - Mexico 107 3,674 - - - - - - - - - 4 - Central American Grasslands, Savannas, and Shrublands Tabasco/ Veracruz savannas - Mexico 108 9,252 - - - - - - - 1 1 4 111* Tehuantepec savannas - Mexico 109 5,614 - - - - - - - 1 1 4 III Orinoco Grasslands, Savannas, and Shrublands Llanos -Venezuela, Colombia 110 355,112 20 2 0 5 6 1 34 4 4 3 III Amazonian Grasslands, Savannas, and Shrublands Guianan savannas -Suriname, Guyana, Brazil, Venezuela 111 128,375 10 2 4 0 6 6 28 4 4 3 III Amazonian savannas - Brazil, Colombia, Venezuela 112 120,124 10 2 4 0 6 6 28 4 4 3 III Beni savannas -Bolivia 113 165,445 32 2 8 0 0 8 50 3 2 3 11* Eastern South American Grasslands, Savannas, and Shrublands Cerrado -Brazil, Paraguay, Bolivia 114 1,982,249 20 2 7 8 8 1 46 3 3 1 1 Chaco savannas - Ar entina, Paraguay, Bolivia, Brazil 115 611,053 20 2 3 8 8 4 45 3 3 2 1 Humid Chaco- Argentina, Paraguay, Uruguay, Brazil 116 474,340 10 2 3 8 6 8 37 3 3 4 IV C6rdoba montane savannas - Argentina 117 55,798 10 2 0 6 4 10 32 4 3 4 IV Southern South American Grasslands, Savannas, and Shrublands ..-.-. Argentine Monte -Argentina 118 197,710 10 2 0 8 6 6 32 4 4 3 III Argentine Espinal-Argentina 119 207,054 10 2 6 8 8 6 40 3 3 3 III Pampas-Argentina 120 426,577 40 15 8 8 6 8 85 2 2 3 11* Uruguayan savannas - Uruguay, Brazil, Argentina 121 336,846 20 15 6 5 6 8 60 3 3 3 III Flooded Grasslands Caribbean Flooded Grasslands Cuban wetlands-Cuba 122 5,345- - - - - - - 2 2 I Enriquillo wetlands - Haiti, Dominican Republic 123 574 - - - - - - - 3 2 1 Northern Mexican Flooded Grasslands Central Mexican wetlands - Mexico 124 362 - - - - - - - 1 2 I Central American Flooded Grasslands .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......................................... ............................ . . . . . . . . . . . . . . . . ....... .......... . . . . . . . . . . . . . . - -...3.. .II* . .. Jalisco palm savannas -Mexico 125 555 - - - - - - - 1 1 Veracruz palm savannas - Mexico 126 7,518 - - - - - - - 1 1 3 II Quintana Roo wetlands - Mexico 127 2,452 0 2 0 0 0 1 3 5 5 3 IV Orinoco Flooded Grasslands .. Orinoco wetlands - Venezuela 128 6,403 - - - - - - - 4 4 3 III MAJOR ECOSYSTEM TYPE Snapshot Final Major Habitat Type Conser- Conser- Biological Bio- Bioregion Ecoregion Ecoregion Size Habitat Habitat Fragmen- Degra- Conver- Protec- vation vation Distinc- diversity Ecoregion Number (km2) Loss a Blocks b tation dation d sion e tion f Total g Status Status I tiveness i Priority k Amazonian Flooded Grasslands Western Amazonian flooded grasslands -Peru, Bolivia 129 10,111 - - - - - - - 4 4 3 III Eastern Amazonian flooded grasslands - Brazil 130 69,533 - - - - - - - 3 3 3 111* Sdo Luis flooded grasslands- Brazil 131 1,681 - - - - - - - 2 2 4 III Northern Andean Flooded Grasslands Guayaquil flooded grasslands-Ecuador 132 3,617 - - - - - - - 2 2 3 11* Eastern South American Flooded Grasslands Pantanal -Brazil, Bolivia, Paraguay 133 140,927 10 2 0 6 10 7 35 4 3 1 I Parand flooded savannas - Argentina 134 36,452 - - - - - - - 2 2 3 II Montane Grasslands Central American Montane Grasslands Mexican alpine tundra - Mexico 135 147 20 5 12 n.a. 6 4 47 3 3 3 111* Costa Rican paramo-Costa Rica 136 31 10 2 5 n.a. 0 1 18 4 4 3 III Northern Andean Montane Grasslands Santa Marta paramo - Colombia 137 1,329 - - - n.a. - - - - 3 1 I Cordillera de M6rida paramo - Venezuela 138 3,518 0 2 16 n.a. 0 1 19 4 4 1 I Northern Andean paramo- Colombia, Ecuador 139 58,806 0 10 16 n.a. 6 4 36 4 3 1 I Central Andean Montane Grasslands Cordillera Central paramo - Peru, Ecuador 140 14,128 0 10 16 n.a. 0 10 36 4 3 1 I Central Andean puna - Bolivia, Argentina, Peru, Chile 141 183,868 20 5 6 n.a. 6 4 41 3 3 2 1 Central Andean wet puna - Peru, Bolivia, Chile 142 188,911 20 2 8 n.a. 6 1 37 3 3 2 I Central Andean dry puna - Argentina, Bolivia, Chile 143 232,958 10 2 4 n.a. 6 10 32 4 3 2 1 Southern South American Montane Grasslands Southern Andean steppe - Argentina, Chile 144 198,643 10 2 0 n.a. 0 4 16 4 4 4 IV Patagonian steppe- Argentina, Chile 145 474,757 32 2 0 n.a. 2 6 42 3 3 2 1 Patagonian grasslands - Argentina, Chile 146 59,585 32 5 0 n.a. 2 6 45 3 3 3 III XERIC FORMATIONS Mediterranean Scrub Northern Mexican Mediterranean Scrub California coastal sage- chaparral- Mexico, U.S. 147 27,104 - - - n.a. - - - 2 1 1 I Central Andean Mediterranean Scrub Chilean matorral-Chile -148 141,643 - - - n.a. - - - 2 2 1 I Deserts and Xeric Shrublands Caribbean Deserts and Xeric Shrublands Cuban cactus scrub-Cuba 149 3,044 20 10 5 n.a. 3 8 46 3 3 4 IV Cayman Islands xeric scrub- Cayman Islands 150 32 32 15 5 n.a. 8 8 68 2 2 4 III Windward Islands xeric scrub-Windward Islands 151 944 32 15 5 n.a. 6 10 68 2 2 4 III Leeward Islands xeric scrub- Leeward Islands 152 1,521 40 20 20 n.a. 6 10 96 1 1 4 111* Northern Mexican Deserts and Xeric Shrublands Baja California xeric scrub - Mexico 153 72,377 0 2 0 n.a. 0 1 3 5 5 3 IV San Lucan mezquital-Mexico 154 2,226 - - - n.a. - - - 5 5 4 IV Western Mexican mezquital- Mexico, U.S. 155 22,894 - - - n.a. - - - - - 4 - 00 MAJOR ECOSYSTEM TYPE Snapshot Final Major Habitat Type Conser- Conser- Biological Bio- Bioregion Ecoregion Ecoregion Size Habitat Habitat Fragmen- Degra- Conver- Protec- vation vation Distinc- diversity Ecoregion Number (kM2) Loss a Blocks b tationc dation d sia, e tion f Total g Status StatusI tvenessI Priorihj Northern M exican Deserts and Xeric Shrublands (continued)........................................................................... .... Sonoran xeric scrub-Mexico, U.S. 156 232,340 20 2 5 n.a. 8 1 36 4 4 2 II Northern Sonoran cactus scrub-Mexico, U.S. 157 97,962 10 5 5 n.a. 0 10 30 4 4 1 1 Mexican Interior chaparral- 4 Mexico, U.S. 158 22,252 - - - n.a. - - - - - 4 Chihuahuan xeric scrub- Mexico, U.S. 159 399,446 20 2 12 n.a. 8 8 50 3 3 4 IV Central Mexican mezquital - Mexico 160 29,347 32 15 20 n.a. 6 8 81 2 2 4 III Eastern Mexican matorral- Mexico 161 26,684 - - - n.a. - - - - 3 - Eastern Mexican mezquital- Mexico, U.S. 162 138,696 32 5 5 n.a. 8 10 60 3 3 4 IV Central Mexican cactus scrub - Mexico 163 37,860 - - - n.a. - - - - - 3 Central A m erican D eserts and Xeric Shrublands ................. ............................................................ ......... Pueblan xeric scrub -Mexico 164 6,818 32 20 16 n.a. 6 10 84 1 1 3 11* Guerreran cactus scrub- Mexico 165 5,232 - - - n.a. - - - - 3 3 11 Motagua Valley thornscrub- Guatemala 166 2,363 40 20 20 n.a. 8 10 98 1 1 3 II O rin oco D esertsandXeric Shrublands ...................................................................................................................... Aruba/Curagao/ Bonaire cactus scrub-Netherlands 167 407 32 10 5 n.a. 0 8 55 3 3 4 IV Antilles La Costa xeric shrublands - Venezuela 168 64,379 32 15 12 n.a. 6 8 73 2 2 4 III Araya and Paria xeric scrub - Venezuela 169 5,424 32 15 12 n.a. 6 8 73 2 2 3 II Northern Andean Deserts and Xeric Shrublands Galapagos Islands xeric scrub-Ecuador 170 9,122 10 2 5 n.a. 0 1 18 4 3 1 I Guajira/Barranquilla xeric scrub-Colombia, Venezuela 171 32,404 32 15 12 n.a. 6 8 73 2 2 3 II Paraguand xeric scrub- Venezuela 172 15,987 32 15 12 n.a. 6 8 73 2 2 3 II Central Andean Deserts and Xeric Shrublands Sechura desert-Peru, Chile 173 189,928 10 10 5 n.a. 6 8 39 3 3 3 111* Atacama desert-Chile 174 103,841 - - - n.a. - - - 3 3 3 III Eastern South American Deserts and Xeric Shrublands K Caatinga-Brazil 175 752,606 32 2 0 n.a. 6 8 48 3 3 3 II* Restingas Northern Andean Restingas ParaguanA restingas- Venezuela 176 15,987 32 15 12 n.a. 6 8 73 2 2 3 11* Aniazonian Restings Northeastern Brazil restingas - Brazil 177 10,248 40 20 20 n.a. 10 10 100 1 1 1 I Eastern South American Restingas Brazilian Atlantic Coast restingas - Brazil 178 8,740 40 20 16 n.a. 10 10 96 1 1 1 I Explanation of Numerical Codes (see Appendix A for details) a. Habitat loss: Index from 0 (least loss) to 40 (most). b. Habitat blocks: Index from 2 (largest and most numerous blocks) to 20 (smallest and least numerous). c. Fragmentation: Index from 0 (least fragmentation) to 20 (most); index ranges from 0 to 10 if degradation value present. d. Degradation: Index from 0 (least degraded) to 10 (most) used for two MIiTs- grasslands, savannas, and shrublands, and flooded grasslands (otherwise indicated as n.a. -not applicable). e. Conversion: Index from 0 (lowest annual conversion rate of natural habitat) to 10 (highest rate). f. Protection: Index from 1 (best protection) to 10 (least). g. Total: Index for snapshot conservation status ranging from 0 (best) to 100 (worst). h. Snapshot conservation status: 1=Critical (89-100 points), 2=Endangered (65-88), 3=Vulnerable (37-64), 4=Relatively Stable (7-36), and 5=Relatively Intact (0-6). i. Final conservation status: Snapshot status modified by threat assessment; codes are the same as above. J. Biological distinctiveness: 1 =Globally Outstanding, 2=Regionally Outstanding, 3= Bioregionally Outstanding, and 4=Locally Important. k. Biodiversity priority: l=Highest Priority at Regional Scale, ll=High Priority at Regional Scale, III=Moderate Priority at Regional Scale, and IV=Important at National Scale. * Indicates an ecoregion elevated to a level P priority ranking to achieve better bioregional representation. - Value is missing because sufficient high quality data were not available. Appendix F Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment Ecoregion Delineation and Classification in delineating Amazonian ecoregions, we attempted to avoid splitting the areas identified at the Manaus The ecoregion maps developed for this study clas- Workshop as highest priority for conservation based sify all terrestrial ecoregions of LAC into standard- on biodiversity and endemism). Our delineations ized major ecosystem and major habitat type cate- also incorporate information from a series of pre- gories to facilitate continental-scale analyses and liminary maps of ecological units for Amazonia and comparisons at biogeographic scales appropriate South America developed by Luis Diego G6mez for for conservation. Development of the ecoregion The Nature Conservancy. These maps divide the maps was challenging because the quality of eco- region on the basis of biophysical parameters and logical information available varied widely through- patterns of plant diversity. out the LAC region and no standardized system Clinebell et al.'s (1995) analysis of soil and climatic existed at the desired level of biogeographic reso- data as predictors for neotropical tree and liana spe- lution. In order to facilitate future refinements, we cies richness represents an important advance in un- highlight here some particularly challenging issues derstanding how biophysical parameters can be used and areas. for mapping biodiversity. Building on the work of Currently, there is no clear consensus on ecoregion Diego G6mez, Clinebell et al., CIAT/EMBRAPA- delineation in Amazonia and the Guianas. Geographic CPAC (1985), and others, further analyses of bio- patterns of biodiversity in this vast region are not physical parameters such as soil type, rainfall, and well known. However, there is evidence of signifi- seasonality should be conducted for Amazonia and cant heterogeneity over a wide range of biogeo- other regions of LAC to better predict patterns of graphic scales. For this reason, the large ecoregion species richness, endemism, and vegetation types, units used in this study for Amazonia should be and to improve ecoregion maps. The conservation subdivided in future studies to more accurately analysis of Amazonian watersheds by Peres and reflect patterns of biodiversity and better address Terborgh (1995) provides a new perspective on the conservation needs. factors that contribute to effective conservation The ecoregion units generally follow divisions units. Future analyses should look closely at the con- proposed on phytogeographers' vegetation maps servation benefits of using watershed boundaries (e.g., Ducke and Black 1953; Rizzini 1963; Hueck 1966; and other defensibility considerations for delineating Prance 1973, 1977; Fundaqdo IBGE 1993)' for the conservation units. Amazon, and priority areas proposed on the conser- Tepui formations occur from Suriname westward to vation maps developed for Amazonia at the inter- the foothills of the Andes. Many of the more west- national workshop "Biological Priorities for Conser- erly lower elevation formations are not well mapped vation in Amazonia" in Manaus, Brazil in January, (Brown 1987) and do not appear on our maps 1990 (Conservation International 1990; Rylands 1990; because of the coarse scale. Huber and Alarcon's (1988) map of vegetation types of Venezuela provides 1. Sources cited in Appendix F are listed on pages 106-116, rather excellent information on tepuis and other vegetative than in the general References section at the back of this book. formations in that country. 84 Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 85 We had difficulty in locating detailed or complete tebrates; (c) how best to conserve large blocks of in- ecological maps for Nicaragua, Honduras, El Salvador, tact habitat that encompass portions of several Guatemala, Uruguay, Chile, Dominican Republic, Haiti, ecoregions and regions with high levels of ecoregion and several Caribbean Islands. The ecoregion maps complexity (i.e., beta and gamma diversity); and used in this study would benefit from revisions and (d) the hypothesis that large-scale physical or eco- updates in these areas. logical processes (e.g., rainfall and nutrient cycling) Detailed ecological maps of Peru by G. Lamas in are significantly altered due to habitat loss long be- Prance (1982) could be compared with the ecoregion fore changes in species or communities are evident. units used in this study to refine boundaries and Workshop participants recognized an urgent need classifications for that region. to conduct similar priority-setting analyses for fresh- Several Mexican xeric and grassland ecoregions were water (i.e., marshes, rivers, and lakes) and marine not fully analyzed in this study and many of these ecosystems. In this study we delineated as ecore- complex systems would benefit from further review gions only particularly large distinct freshwater of their classifications and boundaries. complexes such as the Pantanal or the Enriquillo We recognize that many ecoregions have strong wetlands in Hispaniola. Several freshwater biolo- ecological gradients or contain a wide variety of habi- gists at the BSP workshop developed a preliminary tat types. For example, extensive tracts of dry forests map of freshwater ecoregions and a method for occur in both the Cerrado and western portions of assessing their conservation status. WWF plans to the Guianan moist forests. The conservation needs of collaborate with Wetlands for the Americas and these distinct habitat types should be addressed at other conservation organizations and regional ex- the sub-ecoregional level. perts to build upon this work and help identify con- Small oceanic islands were generally not analyzed servation priorities for these important ecosystems in this study. Some support endemic species and by developing objective, transparent, and scientific distinct natural communities and are highly threat- methods to identify conservation priorities among ened. We recommend a conservation analysis of freshwater and marine ecoregions, and conducting a islands be conducted. Examples of islands not workshop with regional experts to apply these analyzed include: San Andr6s and Providencia methods to all aquatic ecosystems of Latin America Archipelago; Malpelo; Fernando de Noronha; Juan and the Caribbean. FernAndez Archipelago; Desventuradas; Cocos; Although patterns of biodiversity and ecological Trindade and Martin Vaz Islands; Revillagigedo dynamics of freshwater and marine ecosystems are Archipelago; Marias Islands; and Guadalupe Island. significantly different from those of terrestrial eco- Some offshore islands were specifically included in systems, we believe that the integration of informa- adjacent mainland ecoregions: Coiba and Perlas tion on the conservation status (i.e., the current state Islands of Panama, islands in the Gulf of California, or ecological integrity of an ecoregion modified by and Margarita and Tortuga Islands near Venezuela. projected threats) and biological distinctiveness of The Malvinas (Falkland Islands) are included in the ecoregions represents an ecologically-sound approach Patagonian grasslands2 but they may merit being for identifying freshwater and marine priorities. delineated as a separate ecoregion after further However, the particular set of criteria used to assess review of their biogeography. The Galapagos Islands conservation status and biological importance must were considered sufficiently distinct to be delineated be tailored to the specific ecological requirements, as a separate ecoregion. dynamics, patterns of diversity, and responses to Four critical landscape-level issues need to be addressed disturbance of freshwater and marine ecosystems. more fully in future conservation assessments: Given the substantial differences between marine (a) ecological linkages among adjacent ecoregions and freshwater ecosystems, separate methods would (e.g., seasonal or altitudinal migrations; movements be appropriate for each. Appropriate ecoregion due to large-scale disturbances; gene flow and classifications and maps for LAC freshwater and metapopulation interactions for widespread species; marine ecosystems must also be developed. downstream effects of ecological changes in upper watersheds; linkages between terrestrial and aquatic Ecoregion Profiles ecosystems); (b) critical habitats and corridors for continental-scale migrations of vertebrates and inver- The ecoregion profiles provide the following in- formation- 2. A dispute concerning sovereignity over the islands exists between Argentina, which claims this sovereignity, and the * The number of the ecoregion, the name of the U.K., which administers the islands. ecoregion, and the countries in which it occurs. 86 A Conservation Assessment of the Terrestrial Ecoregions of Laun America and the Caribbean * The final conservation status, the biological distinc- TROPICAL BROADLEAF FORESTS tiveness and the biodiversity conservation priority. Tropical Moist Broadleaf Forests * Size in km- of the ecoregion as calculated from Caribbean GIS measurements of our ecoregion maps. The values should be interpreted as approximate. 1. Cuban moist forests-Cuba (Vulnerable; Re- * Sources used for delineation of the ecoregion, for gionally Outstanding; Highest Priority at Regional classification of the major habitat type, and for in- Scale) formation used to assess the conservation status. * 20,069 km2 References are coded with numbers and are listed * Sources: 14, 15, 131, 146 at the end of the appendix. References in bold The moist forests of Cuba, and those of the Greater type are the principal references used to define Antilles in general, maintain exceptionally diverse and delineate the ecoregion. insular biotas with many regional and island en- * Notable biodiversity features of Globally and Re- demic species in a wide range of taxa. Cuba, in par- gionally Outstanding ecoregions and, in many ticular, has a rich moist forest flora. The Greater cases, biodiversity summaries for Bioregionally Antilles are notable for numerous unusual relict Outstanding and Locally Important ecoregions. species and higher taxa. Expansion of cacao, coffee, * Major threats to ecoregions as conveyed by and tobacco production are serious threats in some regional experts. areas. Further notes on threat information: The major threats 2. Hispaniolan moist forests-Haiti, Dominican to ecoregions summarized in the profiles were for Republic (Endangered; Regionally Outstanding; the most part conveyed by regional experts during Highest Priority at Regional Scale) the 1994 BSP and WWF LAC Program workshops. * 43,136 km2 Ecoregion-specific threat information was not * Sources: 89, 121, 129, 203, 205, 207, 208, 232, 241, obtained for a few ecoregions, particularly for sev- 256, 270, 284 eral xeric ecoregions of Mexico. The majority of the Biodiversity considerations are discussed under threats summarized here represent factors that Cuban moist forests. Logging and agricultural directly contribute to alteration of natural land- expansion severely threaten the ecoregion. Firewood scapes and ecosystems. Potential ultimate factors gathering, grazing, and exploitative hunting are also such as human population growth or land tenure are threats. not considered. It is important to recognize that exploitation of 3. Jamaican moist forests-Jamaica (Endangered; wildlife occurs to varying degrees in all LAC ecore- Regionally Outstanding; Highest Priority at Regional gions. Intensive exploitation, particularly hunting Scale) activity fueled by commercial enterprises, can * 7,849 km2 quickly reduce or extirpate populations of target * Sources: 17, 20, 21, 119, 190 species from a region. Populations of large predators Jamaican moist forests are notable for exception- (e.g., cats, canids, giant otters) and herbivores (e.g., ally high levels of endemism in a wide range of taxa. tapir, vicufta), primates, birds (e.g., guans, Agricultural expansion, particularly of coffee plan- curassows, macaws), and larger aquatic species (e.g., tations, threaten the ecoregion. Anthropogenic fires, caiman) are often targeted and can be rapidly extir- grazing, and the invasion of exotic grasses contribute pated from habitats, with profound and long-term to the degradation and conversion of native habitat. effects on the composition and function of native Unsustainable collecting of the endangered swallow- ecosystems (see Dirzo and Miranda 1990; litigo-Elias tail butterfly, Papilio homerus, occurs. and Ramos 1991; Redford 1992; Terborgh 1992). Conservation efforts must strive to conserve habitats 4. Puerto Rican moist forests-Puerto Rico that maintain relatively intact faunas, which are (Vulnerable; Bioregionally Outstanding; Moderate typically located in relatively stable and intact Priority at Regional Scale) ecoregions or in very large blocks of intact habitat, * 7,237 km2 and to reduce and eliminate ecologically unsustain- * Sources: 20,92,177,181 able commercial exploitation and international trade Biodiversity considerations are discussed under of wildlife. Many plant and invertebrate species are Cuban moist forests. Urban development pressures, also threatened through commercial harvesting or pollution, grazing, roadbuilding, and exotic inva- collecting (e.g., mahogany, cacti, and swallowtail sions threaten the ecoregion. Restoration of moist butterflies). forest has occured in some areas over the last several Ecorcgion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 87 decades, but it is unclear to what extent the full cattle ranching pose threats to this ecoregion. Expan- complement of biodiversity and ecological processes sion of citrus, banana, and small-scale agriculture are presently represented in these habitats. threaten forests in the Maya Mountains of Belize. 5. Windward Islands moist forests-Windward 9. YucatAn moist forests-Mexico (Vulnerable; Islands (Vulnerable; Bioregionally Outstanding; Bioregionally Outstanding; Moderate Regional Moderate Priority at Regional Scale) Priority) * 1,914 km2 * 64,012 km2 * Sources: 12, 13, 33, 41, 42, 44-52, 54-56, 92, 127, * Sources: 30, 99, 225, 236 175, 202, 220, 284 Ranching, logging, population growth, and reset- The moist forests of both the Windward and Lee- tlement pressure are low-intensity threats to this ward Islands are comprised of many disjunct forests ecoregion. Development related to tourism could on different islands. Because many species are en- also result in degradation in some areas. demic to forests on a single island, conservation strategies should emphasize adequate representation 10. Sierra Madre moist forests-Mexico, Guate- of each island. Urban development pressures, pollu- mala, El Salvador (Endangered; Bioregionally Out- tion, grazing, roadbuilding, and exotic invasions standing; High Priority at Regional Scale) threaten the ecoregion. The types and intensity of * 9,137 km2 threats vary among different islands. Sources: 30, 99, 112, 113, 120, 125, 258, 275 Habitat conversion at middle elevations for "sun 6. Leeward Islands moist forests-Leeward Islands coffee" production and firewood gathering threaten (Relatively Stable; Bioregionally Outstanding; Mod- the ecoregion. erate Priority at Regional Scale) * 951 km2 11. Central American montane forests-Mexico, * Sources: 9, 12, 13, 33-39, 40, 41, 42, 44-52, 54-56, Guatemala, El Salvador, Honduras (Endangered; 92, 127, 175, 202, 220, 284 Bioregionally Outstanding; High Priority at Regional Biodiversity considerations are discussed under Scale) Windward Islands moist forests. Urban development * 7,676 km2 pressures, pollution, grazing, roadbuilding, and exotic * Sources: 30, 59, 99, 112, 120, 125, 133, 169, 198, invasions threaten the ecoregion. The types and inten- 203, 216, 236, 244, 260, 265, 275, 282 sity of threats vary among different islands. Conversion of montane forests occurs largely through agricultural expansion and clearing for Central America pastures. 7. Oaxacan moist forests-Mexico (Endangered; 12. Belizean swamp forests-Belize (Endangered; Bioregionally Outstanding; Highest Priority at Regional Locally Important; Moderate Priority at Regional Scale) Scale) * 4,715 km2 * 4,150 km2 * Sources: 99, 236; R. de la Maza and J. Sober6n, * Sources: 30, 112, 113, 123 pers. comm. Although not mapped, examples of this ecoregion The biodiversity found in this region of Mexico is also occur in Mexico and Guatemala. This ecoregion noted for high levels of endemism, beta diversity, faces moderate degradation threats from recrea- and habitat complexity. Large-scale agricultural ex- tional use and hunting, and high conversion threats pansion of inland coffee plantations and coastal cit- from roadbuilding and agricultural expansion. rus groves, ranching, and logging threaten the ecoregion. 13. Central American Atlantic moist forests- Guatemala, Belize, Honduras, Nicaragua, Costa 8. Tehuantepec moist forests-Mexico, Guatemala Rica, Panama (Vulnerable; Bioregionally Outstand- (Endangered; Bioregionally Outstanding; High Pri- ing; Moderate Priority at Regional Scale) ority at Regional Scale) * 155,020 km2 * 146,752 km2 * Sources: 29, 59, 60, 67, 112, 113, 115, 124, 133, * Sources: 93, 99, 106, 107, 112, 113, 178, 213, 221, 198, 216, 254, 259, 261, 271, 273, 282 225, 234, 236 Banana plantation and cattle ranch expansion, Oil exploration and associated roadbuilding, na- logging, clearing, and refugee settlements in Nicara- tional security road construction, and expanding gua all pose severe threats to the ecoregion. 88 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Exploitation of parrots and other wildlife is a affinities with the biota of Trinidad and Tobago. Ur- further threat. Conversion and degradation, even in banization pressures and risk of destructive fires are designated protected areas, will likely increase in the high-intensity threats to the ecoregion. The forests of next decade. the Paria Peninsula are under great threat from ex- pansion of coffee plantations. 14. Costa Rican seasonal moist forests -Costa Rica, Nicaragua (Critical; Locally Important; Moderate 18. Orinoco Delta swamp forests-Venezuela, Priority at Regional Scale) Guyana (Relatively Stable; Bioregionally Outstand- * 10,620 km2 ing; Moderate Priority at Regional Scale) * Sources: 85, 117, 247, 254, 261 * 31,698 km2 Conversion of forests for agriculture and pasture, * Sources: 111, 137, 140, 141, 176, 272 anthropogenic fires, and degradation from grazing The forests of the Orinoco (Amacuro) Delta are and wildlife exploitation are high-intensity threats to important wildlife habitats and are known to sup- this ecoregion. port a number of endemic plant and invertebrate species. Oil extraction, water projects, and dam con- 15. Isthmian-Pacific moist forests-Costa Rica, struction represent intensive threats over the next Panama (Endangered; Bioregionally Outstanding; decade. Excessive palmito harvest also represents a High Priority at Regional Scale) threat. * 28,881 km2 * Sources: 30, 116, 117, 261, 271 19. Trinidad & Tobago moist forests-Trinidad & Logging, mining, burning, and habitat conversion Tobago (Vulnerable; Locally Important; Important for pasture and agriculture represent extreme and at National Scale) high-intensity threats to this ecoregion. Wildlife ex- * 4,456 km2 ploitation and pollution are additional threats. * Sources: 10, 11, 94, 229, 272 No detailed threat information obtained. 16. Talamancan montane forests-Costa Rica, Panama (Relatively Stable; Regionally Outstanding; 20. Guianan Highlands moist forests-Venezuela, High Priority at Regional Scale) Brazil, Guyana (Relatively Intact; Regionally Out- * 15,991 km2 standing; Moderate Priority at Regional Scale) * Sources: 30, 117, 247, 261, 271 * 248,018 km2 The Talamancan montane forests are notable for * Sources: 140,182 their rich biotas and high number of regional and The Guianan highlands are recognized as an evo- local endemic species. Burning, logging, and other lutionary center for plant taxa found in both Ama- conversion leading to intensive agricultural use are zonia and the Guianan lowland forests. The varied the major threats to these forests. geology and topography of the ecoregion have helped create a wide range of plant communities Orinoco containing many endemic species. Significant endem- ism is also seen in birds, reptiles and amphibians, 17. Cordillera La Costa montane forests-Vene- invertebrates, and other taxa. The Guianan high- zuela (Vulnerable; Regionally Outstanding; Highest lands typically surround many of the tepuis, emer- Priority at Regional Scale) gent plateaus that support unique biotas, and share * 13,481 km2 some species with them. Heavy poaching and com- * Sources: 140, 272; R. Ford Smith, pers. comm. mercial exploitation of wildlife are becoming serious The moist montane forests of the Cordillera La threats in some areas. Costa are found on several isolated mountains and ranges near the Venezuelan coast. These extend 21. Tepuis-Venezuela, Brazil, Guyana, Suriname, from the Paria Peninsula westward to the Sierra de Colombia (Relatively Intact; Globally Outstanding; San Luis of the State of Falc6n. The biotas of these High Priority at Regional Scale) peaks have been long isolated from one another and * 49,157 km2 larger blocks of moist forest towards the south (e.g., * Sources: 106, 140, 204, 272 Orinoco Basin, Venezuelan Andes) by drier habitats Tepuis are sandstone plateaus occurring in an in the surrounding lowlands. The flora and fauna east-west belt from Suriname to Peru and Colombia contain large numbers of regional and local endem- just east of the Andes. The height of the plateaus ics, and many species show disjunct distributions increases towards the east with many tepuis reach- with populations found in the distant Andes or ing several thousand meters above the surrounding Amazonia. The Paria forests have particularly strong lowlands. In Peru and Colombia, some formations Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 89 rise only tens to hundreds of meters above the sur- 24. Japura/Negro moist forests-Colombia, Vene- rounding lowlands and yet sustain communities zuela, Brazil, Peru (Relatively Intact; Globally Out- characteristic of tepuis. The biological communities standing; High Priority at Regional Scale) of tepuis are notable for their high levels of ende- * 718,551 km2 mism (even within single plateaus), examples of rel- * Sources: 68, 108, 114, 156, 199, 221-223, 224, 231, ict taxa, and for the many unusual adaptations of 272 species to the nutrient-poor, cool, and soggy envi- The Japura/Negro ecoregion contains a great ronments typical of tepuis summits. Because of their complexity of forest types including terra firme for- isolation, few tepuis have been effected by human ests, igap6 forests, varzea forests, and swamp for- activities. Changes in rainfall patterns from lowland ests. Some of the world's largest blackwater river deforestation or acidification from distant industrial ecosystems occur in this ecoregion. Deforestation, activity have the potential to degrade sensitive agricultural conversion, and colonization pose tepuis ecosystems in the future. High-impact tour- threats in the next five to ten years. Road construc- ism at certain sites, burning, and other degradation tion also represents a threat over the next two dec- impacts pose threats within the next five years. Pos- ades. Mining may threaten environmental quality in sible expansion of gold mining operations is also a the region. threat. 25. Uatama moist forests-Brazil, Venezuela, Guy- Amazonia ana (Relatively Stable; Bioregionally Outstanding; Moderate Priority at Regional Scale) 22. Napo moist forests-Peru, Ecuador, Colombia * 288,128 km2 (Relatively Stable; Globally Outstanding; Highest * Sources: 68, 221-223, 224, 231, 272 Priority at Regional Scale) Large monospecific tree plantations, mining, asso- * 369,847 km2 ciated human settlement and hunting pressures, and * Sources: 68, 156, 221-224, 231, 272 selective logging are major threats to this ecoregion. Biological surveys of different taxa in the Napo Paving of forest highways would predictably increase provide evidence that this ecoregion contains one of settlement pressure and habitat conversion. the richest biotas in the world. The entire western arc of the Amazon, particularly the areas near the 26. AmapA moist forests-Brazil, Suriname (Rela- foothills of the Andes are known for their extraordi- tively Stable; Bioregionally Outstanding; Moderate nary diversity that has been attributed to a number Priority at Regional Scale) of factors including the high and relatively aseasonal * 195,120 km2 rainfall; the complex topography and soils; vast me- * Sources: 68, 103, 221-223, 224, 231, 272 andering river systems that create habitat mosaics; No detailed threat information obtained. and complex biogeographic histories. In the Napo ecoregion, hydrocarbon extraction and associated 27. Guianan moist forests-Venezuela, Guyana, roadbuilding have caused degradation and fragmen- Suriname, Brazil, French Guiana (Relatively Stable; tation, and have accelerated these processes by facili- Bioregionally Outstanding; Moderate Priority at tating further colonization. Virtually all of the Ecua- Regional Scale) dorian portion of the Napo is open for oil leasing. * 457,017 km2 Border controversies between Ecuador and Peru * Sources: 68, 103, 118, 144, 221-224, 272 have spurred further colonization in attempts to This ecoregion is threatened by logging operations claim disputed territory. which may expand considerably in the next few years. Roadbuilding by international timber com- 23. Macarena montane forests-Colombia (Vul- panies will likely spur colonization. Gold mining nerable; Regionally Outstanding; Highest Priority at also poses habitat conversion, pollution, and road- Regional Scale) building threats. * 2,366 km2 * Sources: 272 28. Paramaribo swamp forests-Suriname (Vul- There was considerable debate as to whether this nerable; Bioregionally Outstanding; Moderate Prior- ecoregion should be incorporated into the surround- ity at Regional Scale) ing Napo moist forests ecoregion. Although much of * 7,760 km2 the ecoregion is legally protected, progress on a ma- * Sources: 140, 272 jor roadbuilding initiative could increase coloniza- Agricultural expansion, especially draining of tion, legal and illegal commercial activities, and swamp areas and runoff of agrochemicals, is an logging. urgent threat. 90 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean 29. Ucayali moist forests-Brazil, Peru (Vulnerable; 33. Varzea forests-Brazil, Peru, Colombia Globally Outstanding; Highest Priority at Regional (Vulnerable; Globally Outstanding; Highest Priority Scale) at Regional Scale) * 173,527 km2 * 193,129 km2 * Sources: 68, 156, 214, 221-224, 231, 272 * Sources: 103, 272 Biogeographic considerations for the Ucayali The Varzea forests of the Amazon Basin represent moist forests are similar to those for the Napo moist some of the world's most extensive seasonally forests. No detailed threat information obtained. inundated forests. The seasonal migration of fish and terrestrial animal populations into the flooded 30. Western Amazonian swamp forests-Peru, forests represent a globally outstanding ecological Colombia (Relatively Stable; Globally Outstanding; phenomenon. A number of endemic species, includ- Highest Priority at Regional Scale) ing birds and primates, occur in these forests. * 8,315 km2 Intensive logging and selective exploitation of the * Sources: 167, 272 kapok tree (Ceiba pentandra) are accelerating defores- These swamp forests are ranked as Globally Out- tation. The Varzea already contains extensive standing because their biotas are strongly associated industrial timber infrastructure, which will probably with the Globally Outstanding ecoregions of western spur further logging. Open floodplains are being Amazonia. converted for cattle ranching. The spread of intro- duced water buffalo is a threat. 31. Southwestern Amazonian moist forests- Brazil, Peru, Bolivia (Relatively Stable; Globally 34. Purus/Madeira moist forests-Brazil (Rela- Outstanding; Highest Priority at Regional Scale) tively Stable; Locally Important; Important at * 534,316 km2 National Scale) * Sources: 68,156, 221-223, 224, 231, 272 * 561,765 km2 The southern part of this ecoregion in Bolivia * Sources: 68, 74, 103 (southern boundary), 221- should probably be distinguished as a separate 223, 224, 231 ecoregion on biogeographic grounds according to Agricultural colonization, land clearing, logging, M. Ribera (pers. comm.). Biogeographic considera- associated roadbuilding, and hunting represent the tions for the Southwestern Amazonian moist for- most severe threats to this ecoregion. ests are similar to those for the Napo moist forests. Hydrocarbon extraction and associated roadbuild- 35. Rond6nia/Mato Grosso moist forests-Brazil, ing have caused degradation and fragmentation Bolivia (Vulnerable; Regionally Outstanding; High and have accelerated these processes by facilitating Priority at Regional Scale) further colonization. Logging and mining are also * 645,089 km2 threats. An evaluation of the conservation status of * Sources: 3, 103 (eastern and southwestern the "Beni moist forests" (M. Ribera and E. Forno, boundaries), 221-224, 231, 272 (southwestern pers. comm.) suggests that this portion of the boundary) ecoregion is quite threatened; the Relatively Stable This ecoregion supports a wide range of forest status of the ecoregion as a whole is largely due to types with many transitional formations located the more remote northern portion in Brazil and southward towards the Cerrado and Beni savannas. Peru. Some regions are reported to support highly diverse communities, particularly for butterflies and plants, 32. Jurua moist forests-Brazil (Relatively Intact; with many endemic species. During the workshops, Regionally Outstanding; Moderate Priority at Re- there was considerable debate over whether this gional Scale) ecoregion should be categorized as vulnerable or * 361,055 km2 relatively stable. The consequence of these discus- * Sources: 68, 221-223, 224, 231, 272 sions was that this ecoregion is categorized as a Rapid development of petroleum resources in the Level II priority even though it falls within the Vul- region of the Rio Jurud poses a major threat. Oil spill nerable/ Regionally Outstanding cell in the priority- hazards, pipeline construction, and settlement of setting matrix. Deforestation for agriculture and oilfield workers all are causing habitat conversion ranching, mining, and roadbuilding all pose major and degradation. Unregulated hunting and intensive threats over the next two decades. Small-scale log- commercial fishing represent immediate threats to ging, wildlife exploitation, introduction of exotic vertebrate faunas. species, and hydroelectric projects are also threats. Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 91 Increasing colonization will likely bring further habi- mining, wildlife exploitation, and coca cultivation all tat conversion and plant and wildlife exploitation. threaten the ecoregion. 36. Beni swamp and gallery forests-Bolivia 40. Eastern Panamanian montane forests-Panama, (Relatively Stable; Locally Important; Important at Colombia (Vulnerable; Bioregionally Outstanding; National Scale) Moderate Priority at Regional Scale) * 31,329 km2 * 2,905 km2 * Sources: 82, 272 * Sources: 112,113,124,190, 262, 271, 282 Expansion of livestock and agricultural land uses The higher peaks of the Serrania de San Blas, pose medium-intensity threats in the next 20 years. Dari6n, Maj6, and Pirre of central and eastern Pan- ama are covered in tropical cloud forest. Both the 37. Tapaj6s/Xingu moist forests-Brazil flora and fauna of these relatively isolated ranges (Vulnerable; Locally Important; Important at Na- contain numerous endemic species and represent an tional Scale) unusual assemblage of species with South American * 630,905 km2 and Central American affinities. Habitat and envi- Sources: 68, 74, 103, 221-223, 224, 231 ronmental quality are being degraded by mining, Land-clearing for cattle ranching and selective and numerous wildlife species are being exploited logging of mahogany represent severe threats to this by overhunting for subsistence and trade. ecoregion. 41. Northwestern Andean montane forests- 38. Tocantins moist forests-Brazil (Endangered; Colombia, Ecuador (Endangered; Globally Out- Locally Important; Moderate Priority at Regional standing; Highest Priority at Regional Scale) Scale) * 52,937 km2 * 279,419 km2 * Sources: 64, 150, 272 * Sources: 68, 74, 103 (southern, northern and The biotas of the submontane and montane forests eastern boundaries), 162-164, 221-223, 224, 231 of the northern Andes are exceptionally rich in spe- The Tocantins moist forests face threats from cies and have a high proportion of regional and local clearing and cattle grazing, which is facilitated by endemics (i.e., species found only in the region or expanded road access. High-intensity selective log- with very limited geographic ranges). The complex ging operations are converting habitat and associ- topography, climate, geology, and biogeographic ated operations are leading to forest fires. history of the northern Andes have helped create a high turnover in species (i.e., beta diversity) over Northern Andes distance and along steep environmental gradients. This region has some of the highest known levels of 39. Choc6/Darien moist forests-Colombia, Pan- beta diversity and local endemism for many taxa ama, Ecuador (Vulnerable; Globally Outstanding; (e.g., birds-Terborgh and Winter 1983; Hilty and Highest Priority at Regional Scale) Brown 1986), even to the point where eastern and * 82,079 km2 western slopes of some of the major inter-Andean * Sources: 61, 65, 73, 91, 97, 98, 102, 105, 110, 132, valleys in Colombia have substantially different bio- 149, 150, 151, 152, 253, 271, 272 tas (these patterns are particularly pronounced in The Choc6/Dari6n ecoregion is considered to have southwestern Colombia). For these reasons, separate one of the world's richest lowland biotas, with ex- ecoregions were delineated for several ecologically ceptional richness and endemism in a wide range of distinct slopes of the northern Andes and all were taxa including plants, birds, reptiles and amphibi- considered as Globally Outstanding in terms of their ans, and butterflies. Unplanned colonization follow- associated biodiversity. Many of the moist montane ing the completion of roads and massive logging forests of the Northern Andes are under intense concessions are major threats to the Choc6 forests. threat from conversion for agriculture and pasture, Since 1960, over 40 percent of the forests have been mining operations, and logging. cleared or heavily degraded and deforestation rates are accelerating (Salaman 1994). This ecoregion faces 42. Western Ecuador moist forests-Ecuador, Co- serious threats in the next five to ten years from na- lombia (Critical; Regionally Outstanding; Highest tional-level development projects, including dams, Priority at Regional Scale) roads, seaports, pipelines, and military installations. * 40,218 km2 Currently, intensive logging, human settlement, * Sources: 12a, 64, 104, 148, 153, 154, 164, 272 92 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean These forests are rich in species with high levels of * 4,707 km2 local and regional endemism. The biota shares strong * Sources: 66, 132, 150, 151, 152, 253, 272 affinities with that of the adjacent Choc6/Dari6n The Sierra Nevada de Santa Marta rises to nearly moist forests and the Northwestern Andean moist 5800 m from the surrounding Caribbean coastal forests, both globally outstanding ecoregions. Intensive lowlands of Colombia. This isolated massif was pre- logging of non-reserve areas, road construction, and sent before the Andes rose and accordingly sustains colonization all pose severe threats to this ecoregion. an unusual biota with complex biogeographic affini- ties (e.g., Amazonian, Mesoamerican, Cho-coan) and 43. Cauca Valley montane forests-Colombia high endemism in most taxa. The natural habitats of (Critical; Globally Outstanding; Highest Priority at this range are severely threatened from agricultural Regional Scale) expansion, logging, and burning. * 32,412 km2 * Sources: 132, 150, 253, 272 49. Venezuelan Andes montane forests-Venezuela, Biodiversity and threat considerations for this Colombia (Endangered; Globally Outstanding; ecoregion are covered under the Northwestern Highest Priority at Regional Scale) Andean montane forests description. * 16,638 km2 * Sources: 140, 142, 143, 272 44. Magdalena Valley montane forests-Colombia Refer to Northwestern Andean montane forests (Critical; Globally Outstanding; Highest Priority at for biodiversity description. This ecoregion is threat- Regional Scale) ened by logging which continues to expand to * 49,322 km2 higher areas. Agricultural colonization represents a * Sources: 66, 132, 150, 151, 152, 253 low-intensity threat. Biodiversity and threat considerations for the Magdalena Valley montane forests are covered 50. Catatumbo moist forests-Venezuela, Colom- under the Northwestern Andean montane forests bia (Critical; Locally Important; Moderate Priority at description. Regional Scale) * 21,813 km2 45. Magdalena/Urabd moist forests-Colombia * Sources: 272; R. Ford Smith, pers. comm. (Endangered; Bioregionally Outstanding; High The Catatumbo moist forests are at lower eleva- Regional Priority) tions (100-300 m) around the southern end of the * 73,660 km2 Maracaibo depression. These forests are noted by * Sources: 73, 150, 253, 272 Huber and Alarcon (1988) as having few endemic No detailed threat information obtained. species but interesting affinities with Amazonian floras. Portions of this ecoregion are regularly inun- 46. Cordillera Oriental montane forests-Colombia, dated and sustain swamp forests. Beef and dairy Venezuela (Vulnerable; Globally Outstanding; cattle ranching in particular are intense threats. Highest Priority at Regional Scale) Draining and channelization of wetlands and peri- * 66,712 km2 odic oil spills are further threats. * Sources: 132, 150, 272 Biodiversity and threat considerations for the Cor- Central Andes dillera Oriental montane forests are covered under the Northwestern Andean montane forests description. 51. Peruvian Yungas-Peru (Endangered; Globally Outstanding; Highest Priority at Regional Scale) 47. Eastern Cordillera Real montane forests- * 188,735 km2 Ecuador, Colombia, Peru (Vulnerable; Globally * Sources: 156, 214, 237, 272 Outstanding; Highest Priority at Regional Scale) The montane forests of the eastern slope of the * 84,442 km2 Andes in Peru support some of the world's richest * Sources: 64, 132, 150, 272 montane ecosystems. Regional and local endemism Biodiversity and threat considerations for this in a wide range of taxa is common. Extensive land ecoregion are covered under the Northwestern clearing, agricultural conversion, and logging Andean montane forests description. severely threaten the ecoregion. Roadbuilding and colonization amplify these threats. 48. Santa Marta montane forests-Colombia (Vulnerable; Regionally Outstanding; Highest 52. Bolivian Yungas-Bolivia, Argentina (Endangered; Regional Priority) Regionally Outstanding; Highest Regional Priority) Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 93 * 72,517 km2 * 61,466 km2 * Sources: 58, 80, 88, 90, 172, 207, 237, 265, 272 * Sources: 14, 15, 20, 32, 137, 145 Ongoing biological surveys suggest that the mon- Clearcutting and selective logging, charcoal pro- tane forests of the Bolivian Yungas maintain high duction, frequent burning, and slash-and-burn agri- levels of endemism, species richness, and beta di- culture pose threats to the ecoregion. versity. The Bolivian Yungas are being deforested for subsistence agriculture and crop production (e.g., 57. Hispaniolan dry forests-Haiti, Dominican Re- coca, coffee, and tea) (Parker 1990). public (Endangered; Bioregionally Outstanding; High Priority at Regional Scale) 53. Andean Yungas-Argentina, Bolivia (Vul- * 14,610 km2 nerable; Bioregionally Outstanding; Moderate Prior- * Sources: 89, 121, 129, 203, 205, 207, 208, 241, 256, ity at Regional Scale) 270, 284 * 55,457 km2 The ecoregion faces threats from clearing for devel- * Sources: 75, 233, 245, 272 opment, firewood gathering, and heavy recreational No detailed threat information obtained. use. Eastern South America 58. Jamaican dry forests-Jamaica (Endangered; Lo- cally Important; Moderate Priority at Regional Scale) 54. Brazilian Coastal Atlantic forests-Brazil * 2,189 km2 (Critical; Globally Outstanding; Highest Priority at * Sources: 20, 21, 119, 190 Regional Scale) The ecoregion faces threats from clearing for de- * 233,266 km2 velopment, firewood gathering, and heavy recrea- * Sources: 103 tional use. This ecoregion should likely be subdivided into two or more ecoregions to reflect substantial geo- 59. Puerto Rican dry forests-Puerto Rico (Endan- graphic differences in species assemblages. The At- gered; Locally Important; Moderate Priority at Re- lantic forests are characterized by extraordinarily gional Scale) species-rich biotas and very high levels of both re- a 1,295 km2 gional and local endemism. Urbanization, industri- * Sources: 20, 76, 92, 177 alization, agricultural expansion, and associated Urban expansion, tourism development, livestock roadbuilding are most severe threats within the next grazing, and predation on native birds by exotic decade. Logging and wildlife exploitation are also mammals all pose threats to the ecoregion. threats. 60. Bahamian dry forests -Bahamas, Turks & Caicos 55. Brazilian Interior Atlantic forests-Brazil, Islands (Vulnerable; Locally Important; Important at Argentina, Paraguay, (Endangered; Regionally National Scale) Outstanding; Highest Priority at Regional Scale) * 9,293 km2 * 803,908 km2 * Sources: 20, 284 * Sources: 103, 206, 226, 272 The ecoregion faces threats from tourism develop- The Interior Atlantic forests display much com- ment, heavy recreational use, and firewood gathering. plexity and geographic variation and probably should be subdivided into several distinct ecore- 61. Cayman Islands dry forests-Cayman Islands gions in future analyses (e.g., zone intergrading (Endangered; Locally Important; Moderate Regional with the Caatinga, Mata AtlAntica proper, forests of Priority) southern Bahia and Espirito Santo, and the forests * 230 km2 of Rio de Janeiro). These forests face threats from * Sources: 81, 248 agricultural expansion, colonization, logging, and The ecoregion faces threats from tourism develop- associated road construction within the next five ment, heavy recreational use, and firewood gathering. years. 62. Windward Islands dry forests-Windward Tropical Dry Broadleaf Forests Islands (Endangered; Locally Important; Moderate Caribbean Priority at Regional Scale) * 431 km2 56. Cuban dry forests-Cuba (Endangered; Biore- * Sources: 12, 13, 33, 40-42, 44-52, 54-56, 92, 127, gionally Outstanding; High Priority at Regional Scale) 175, 202, 220, 269 94 A Conservation Assessment of the Terrestrial Ecoregions of Latin Amenca and the Caribbean Agricultural expansion, grazing, firewood gather- 68. Balsas dry forests -Mexico (Endangered; Region- ing, and development are threats to the ecoregion. ally Outstanding; Highest Priority at Regional Scale) * 161,098 km2 63. Leeward Islands dry forests-Leeward Islands * Sources: 99, 225, 236 (Critical; Locally Important; Moderate Priority at Biodiversity considerations are discussed under Regional Scale) Jalisco dry forests. Agricultural expansion, intensive * 182 km2 cultivation for export crops and associated pollution * Sources: 12, 13, 33, 40-42, 44-52, 54-56, 92, 127, threaten the ecoregion. 175, 202, 220, 269 The ecoregion faces threats from clearing for de- 69. Oaxacan dry forests-Mexico (Endangered; velopment, seasonal burning, firewood gathering, Bioregionally Outstanding; High Priority at Regional and heavy recreational use. Scale) * 10,566 km2 Northern Mexico * Sources: 99, 180, 225, 236 No clear consensus was reached as to whether this 64. Baja California dry forests-Mexico (Relatively ecoregion should be ranked as Regionally Outstand- Stable; Locally Important; Important at National ing or Bioregionally Outstanding. Conversion for cat- Scale) tle ranching, coffee and citrus plantations, and mod- * 93 km2 erate wildlife exploitation threaten the ecoregion. * Sources: 19, 99, 225, 236 Agricultural expansion and grazing threaten the 70. Veracruz dry forests -Mexico (Critical; Locally ecoregion. Important; Moderate Priority at Regional Scale) * 35,546 km2 65. Sinaloan dry forests-Mexico (Vulnerable; * Sources: 99, 225, 236 Bioregionally Outstanding; Moderate Priority at Agricultural expansion, intensive grazing, fire- Regional Scale) wood gathering, and burning all severely threaten * 119,184 km2 the ecoregion. * Sources: 19, 99, 101, 225, 236 Coffee plantations, firewood gathering, wildlife 71. Yucatan dry forests-Mexico (Endangered; exploitation, and grazing are threats to the ecoregion. Locally Important; Moderate Priority at Regional Scale) 66. Tamaulipas/Veracruz dry forests-Mexico * 45,554 km2 (Endangered; Locally Important; Moderate Priority * Sources: 99, 225, 236 at Regional Scale) Agricultural expansion, expanding citrus planta- * 40,858 km2 tions, urbanization, firewood gathering, and grazing * Sources: 99, 225, 236 pose threats to the ecoregion. Agricultural expansion and grazing threaten the ecoregion. 72. Central American Pacific dry forests-El Salva- dor, Honduras, Nicaragua, Costa Rica, Guatemala Central America (Critical; Bioregionally Outstanding; High Priority at Regional Scale) 67. Jalisco dry forests-Mexico (Endangered; * 50,101 km2 Regionally Outstanding; Highest Priority at Regional * Sources: 60, 99, 112, 113, 116, 117, 120, 133, 136, Scale) 138, 158, 165, 188, 261, 273, 282 * 19,973 km2 The ecoregion is threatened by grazing, burning, * Sources: 99, 225, 236 agricultural expansion, and exploitative hunting. The dry forests of Southern Mexico (i.e., the Jalisco and Balsas dry forests) are noted for high 73. Panamanian dry forests-Panama (Critical; levels of regional and local endemism in a wide Locally Important; Moderate Priority at Regional range of taxa. Urbanization and increasing tourism Scale) are high-intensity threats to the ecoregion. Road * 5,010 km2 construction, fruit plantations, and ranching also * Sources: 112, 117, 200, 271 pose threats. Exploitation of wildlife is a high The ecoregion is threatened by grazing, burning, intensity threat as well. and exploitative hunting. Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 95 Orinoco * 1,291 km2 * Sources: 105, 132, 150, 151, 152, 272, 283 74. Llanos dry forests-Venezuela (Endangered; Lo- No detailed threat information obtained. cally Important; Moderate Priority at Regional Scale) * 44,177 km2 80. Sinfl Valley dry forests-Colombia (Critical; * Sources: 140, 142, 272 Locally Important; Moderate Regional Priority) No detailed threat information obtained. * 55,473 km2 * Sources: 105, 132, 150, 151, 152, 272, 283 75. Trinidad & Tobago dry forests-Trinidad & No detailed threat information obtained. Tobago (Endangered; Locally Important; Moderate Priority at Regional Scale) 81. Ecuadorian dry forests-Ecuador (Critical; Glo- * 251 km2 bally Outstanding; Highest Priority at Regional Scale) * Sources: 10, 11, 94, 229, 272 * 22,271 km2 The ecoregion faces threats from tourism develop- * Sources: 64, 147, 270 ments, heavy recreational use, and firewood gathering. The dry forests of the Pacific coast of South America (i.e., Ecuadorian dry forests and Tumbes/ Amazonia Piura dry forests) are known for high levels of both regional and local endemism. Logging and overgraz- 76. Bolivian lowland dry forests -Bolivia, Brazil ing present severe degradation threats within the (Critical; Globally Outstanding; Highest Priority at next five years, even within protected areas. Regional Scale) * 156,814 km2 82. Tumbes/Piura dry forests-Ecuador, Peru * Sources: 171, 191, 272 (Endangered; Globally Outstanding; Highest Prior- Parker et al. (1993) suggest that the dry forests of ity at Regional Scale) Bolivia may be among the richest dry forest ecosys- * 64,588 km2 tems in the world. The biota of the ecoregion has * Sources: 64,147,156, 272 affinities with Amazonia, the Chaco, and the Cer- Biodiversity considerations are discussed under rado as well as containing many endemic species. Ecuadorian dry forests. No detailed threat informa- Agricultural expansion, burning, and grazing pose tion obtained. major threats within the next decade. Increasing wildlife exploitation has the potential to extirpate 83. Marafi6n dry forests-Peru (Endangered; several target species. Pollution from agricultural Bioregionally Outstanding; High Priority at Regional and associated human settlements also pose degra- Scale) dation risks to the ecoregion. * 14,920 km2 * Sources: 156, 183, 272; David Neill, pers. comm. Northern Andes There are a number of endemic plant and bird species associated with this ecoregion, and other 77. Cauca Valley dry forests-Colombia (Critical; taxa are likely to display similar distribution pat- Locally Important; Moderate Priority at Regional terns. Oil palm plantation expansion, cattle ranching, Scale) and logging threaten the ecoregion severely in the * 5,130 km2 next five years. Oil exploration and extraction are * Sources: 105, 132, 150, 151, 152, 272, 283 also major threats. No detailed threat information obtained. 84. Maracaibo dry forests-Venezuela (Endan- 78. Magdalena Valley dry forests-Colombia gered; Locally Important; Moderate Priority at Re- (Critical; Locally Important; Moderate Regional gional Scale) Priority) * 31,471 km2 * 13,837 km2 * Sources: 272 * Sources: 105, 132, 150, 151, 152, 272, 283 No detailed threat information obtained. No detailed threat information obtained. 85. Lara/Falc6n dry forests-Venezuela (Endan- 79. Patia Valley dry forests-Colombia (Critical; Lo- gered; Locally Important; Moderate Priority at cally Important; Moderate Priority at Regional Scale) Regional Scale) 96 A Consenation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean * 16,178 km2 Cold deciduous forests and evergreen swamp for- * Sources: 140,243,283; R. Ford Smith, pers. comm. ests dominate this ecoregion where still intact. In- No detailed threat information obtained. tensive logging and timber plantations, exotic spe- cies invasions (e.g., rabbits removing ground cover), Central Andes grazing, and firewood gathering are severe threats to the ecoregion. 86. Bolivian montane dry forests-Bolivia (Critical; Bioregionally Outstanding; High Priority at Regional Tropical and Subtropical Coniferous Forests Scale) Caribbean * 39,368 km2 * Sources: 265, 272 90. Cuban pine forests-Cuba (Vulnerable; Re- Settlement and agricultural conversion have al- gionally Outstanding; Highest Priority at Regional ready had dramatic effects on the ecoregion; further Scale) expansion seriously threatens remaining fragments * 6,017 km2 of habitat. * Sources: 14, 15, 20, 32, 131 The pine forests of Cuba and Hispaniola support a CONIFER/TEMPERATE BROADLEAF FORESTS number of endemic plant and animal species Temperate Forests (IGACC 1989; Borhidi 1991). Mining, citrus planta- Southern South America tions, grazing, and logging severely threaten the ecoregion. Exploitation of threatened parrot popula- 87. Chilean winter-rain forests-Chile (Endan-gered; tions occurs in western portions of the ecoregion. Regionally Outstanding; Highest Regional Priority) * 24,937 km2 91. Hispaniolan pine forests-Haiti, Dominican * Sources: 109, 272; M.K. Arroyo, pers. comm. Republic (Vulnerable; Regionally Outstanding; These forests are adapted to a Mediterranean Highest Priority at Regional Scale) climate and share many characteristics with the * 10,833 km2 adjacent Chilean matorral, including its high levels * Sources: 89, 121, 129, 203, 205, 207, 232, 241, 256, of endemic species. Intensive logging and timber 270, 284 plantations, exotic species invasions, anthropogenic Biodiversity considerations are discussed under fires, grazing, and firewood gathering are severe Cuban pine forests. Degradation and destruction of threats to the ecoregion. these forests continues through clearing for agricul- ture, grazing, firewood collection, and anthropo- 88. Valdivian temperate forests-Chile, Argentina genic fires. (Vulnerable; Globally Outstanding; Highest Priority at Regional Scale) 92. Bahamian pine forests- Bahamas, Turks & * 166,248 km2 Caicos Islands (Vulnerable; Locally Important; * Sources: 109, 185, 272 Important at National Scale) The Valdivian temperate forests represent one * 3,920 km2 of the world's five major temperate rainforest * Sources: 20, 284 ecosystems (i.e., Valdivian in Chile, Pacific North- The ecoregion faces threats from tourism develop- west of North America, western Black Sea, New ment, heavy recreational use, and firewood gathering. Zealand, Tasmania, and some small areas in Japan, Norway, Ireland, and the United Kingdom; Kellogg Northern Mexico 1992). The forests of this ecoregion can support extraordinary standing biomass (i.e., stands of very 93. Sierra Juarez pine-oak forests-Mexico, U.S. large trees that are characteristic of temperate rain- (Vulnerable; Locally Important; Important at Na- forests) and are known to contain many unusual tional Scale) species and higher taxa. Intensive logging and tim- * 24,228 km2 ber plantations are severe threats to the ecoregion. * Sources: 6, 20, 30, 99, 112, 113, 174, 216, 225, 236 Off-road vehicle use, intensive recreational use, 89. Subpolar Nothofagus forests-Chile, Argentina settlement pressure, agricultural expansion, and (Vulnerable; Bioregionally Outstanding; Moderate grazing threaten the ecoregion. Priority at Regional Scale) * 141,120 km2 94. San Lucan pine-oak forests-Mexico (Relatively * Sources: 109, 188, 219, 272 Intact; Locally Important; Important at National Scale) Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 97 * 895 km2 This ecoregion supports rich forests with many * Sources: 20, 30, 99, 112, 113, 216, 225, 236 regional endemics. Logging, agricultural expansion, Agricultural expansion and grazing threaten the firewood gathering, and intensive wildlife exploita- ecoregion. tion are threats. Sheep ranching and overgrazing, as well as intensive urbanization, pose threats to the 95. Sierra Madre Occidental pine-oak forests- ecoregion. Mexico, U.S. (Endangered; Globally Outstanding; Highest Priority at Regional Scale) 100. Veracruz montane forests-Mexico (Relatively * 204,374 km2 Stable; Bioregionally Outstanding; Moderate Priority * Sources: 6, 20, 30, 99, 112, 113, 173, 216, 225, 236 at Regional Scale) The montane forests of the Sierra Madre Occiden- * 6,615 km2 tal represent some of the world's most extensive * Sources: 99, 236; R. de la Maza, J. Sober6n and subtropical coniferous forests. Many plant and ani- G. Castilleja, pers. comm. mal species are restricted to the diverse forests of No detailed biodiversity information was avail- this range. Commercial logging, land conversion for able for this ecoregion and the appropriateness of its cultivation, and overgrazing by livestock pose seri- MHT classification is still unclear. Agricultural ex- ous threats to the ecoregion. pansion, intensive grazing, burning, and firewood gathering threaten the ecoregion. 96. Central Mexican pine-oak forests (Endangered; Bioregionally Outstanding; High Priority at Regional 101. Sierra Madre del Sur pine-oak forests- Scale) Mexico (Critical; Globally Outstanding; Highest Pri- * 3,719 km2 ority at Regional Scale) * Sources: 20, 30, 99, 112, 113, 225, 236 * 41,129 km2 Sheep ranching and overgrazing, agricultural ex- * Sources: 20, 30, 99, 100, 216, 217, 236 pansion, firewood gathering, and intensive urbani- The montane forests of southern Mexico represent zation pose threats. Hunting poses a less severe some of the world's most diverse and complex sub- threat to the integrity of the ecoregion. tropical mixed hardwood-conifer forests (WWF and IUCN 1994). The biota of the region is noted for 97. Sierra Madre Oriental pine-oak forests- many regional and local endemic species. Roads and Mexico (Relatively Stable; Regionally Outstanding; associated tourist developments, overgrazing, and High Priority at Regional Scale) exploitive hunting are severe threats to the ecore- * 38,199 km2 gion. Urbanization pressures are amplifying these * Sources: 20, 30, 99, 112, 113, 128, 209, 210, 216, problems and causing further habitat loss. 225, 236 The disjunct peaks and ranges of this ecoregion 102. Central American pine-oak forests- support a number of regional and local endemic spe- Guatemala, El Salvador, Honduras, Mexico, Nica- cies, notably in the birds, conifers, and herpetofauna. ragua (Vulnerable; Bioregionally Outstanding; Mod- Logging, burning, roadbuilding, grazing, and settle- erate Priority at Regional Scale) ment pressures are severe threats to the ecoregion. * 127,910 km2 * Sources: 13, 20, 29, 30, 59, 99, 112, 113, 115, 120, 133, 98. Veracruz pine-oak forests-Mexico (Critical; Lo- 134,136,138,155,159,216,236,254,258,275,282 cally Important; Moderate Priority at Regional Scale) The pine-oak forests of the Central American * 5,979 km2 highlands are rapidly disappearing from logging, * Sources: 20, 30, 99, 112, 113, 225, 236 firewood gathering, wildfires and uncontrolled Agricultural expansion, intensive grazing, fire- burning, agricultural expansion, and bark-beetle wood gathering, and burning all severely threaten epidemics that are exacerbated by degradation from the ecoregion. logging, grazing, and burning. Road building con- tinues to open areas for exploitation and eventual Ceti tral A imerica destruction from expanding human activity. 99. Mexican transvolcanic pine-oak forests- 103. Belizean pine forests-Belize (Relatively Sta- Mexico (Endangered; Regionally Outstanding; ble; Regionally Outstanding; High Priority at Re- Highest Priority at Regional Scale) gional Scale) * 72,802 km2 * 2,750 km2 * Sources: 20, 30, 99, 112, 113, 216, 225, 236 * Sources: 20, 59, 112, 113, 123, 216 98 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean Belizean pine forests are dominated by Caribbean No detailed threat information obtained. The dark pine (Pinus carribaea) and require periodic low- blue area on the ecoregions maps just north of the intensity burns for their regeneration (Perry 1991). Eastern Mexican grasslands represent the Laguna This ecoregion represents one of the few examples of Madre wetland complex. Insufficent data were lowland pine forests in the Neotropics. Although not available to incorporate this area into the study. mapped, examples of this ecoregion occur in Mexico and Guatemala. Selective logging and expansion of Central America citrus and banana plantations threaten coastal areas of the ecoregion. 108. Tabasco/Veracruz savannas-Mexico (Critical; Locally Important; Moderate Priority at Regional 104. Miskito pine forests- Nicaragua, Honduras Scale) (Relatively Stable; Regionally Outstanding; High * 9,252 km2 Priority at Regional Scale) * Sources: 99, 236; G. Castilleja, pers. comm. * 15,064 km2 The origin and maintenance of the Tabasco/ * Sources: 20, 30, 133, 216, 254, 257 Veracruz savannas and the Tehuantepec savannas The Miskito pine-savannas represent the largest are considered by some biogeographers to be lowland tropical pine-savannas in the Neotropics. primarily due to human activities. Ranching, The pine-savannas are dominated by Pinus carribaea, overgrazing, settlement pressures, and associated a lowland pine dependent on periodic low-intensity habitat conversion threaten the few remnants of the burns for its regeneration (Perry 1991). Industrial ecoregion. logging, firewood gathering, wildfires, and uncon- trolled burning are threats to the ecoregion. 109. Tehuantepec savannas-Mexico (Critical; Locally Important; Moderate Priority at Regional Eastern South America Scale) * 5,614 km2 105. Brazilian Araucaria forests-Brazil, Argentina * Sources: 20, 30, 99, 112, 113, 225, 236 (Critical; Bioregionally Outstanding; High Priority at Refer to note under Tabasco/Veracruz savannas. Regional Scale) Agricultural expansion, intensive grazing, and * 206,459 km2 burning all severely threaten the ecoregion. * Sources: 77, 78, 103, 272 (outside of Brazil); M. Pellerano, N. Vlarty, pers. comm. Orinoco Araucaria are primitive conifers restricted to the Southern Hemisphere. Araucaria angustifolia forests 110. Llanos-Venezuela, Colombia (Relatively Sta- typically occur in stands associated with dense ble; Bioregionally Outstanding; Moderate Priority at tropical moist forests, although pure stands occur. A Regional Scale) great diversity of soil types and associated florisitic * 355,112 km2 communities, many without Araucaria, occur within * Sources: 140, 142, 255, 272, 283 the ecoregion. The remaining natural habitats of this The Llanos represent the largest savanna ecosys- ecoregion face severe threats from logging and agri- tern in northern South America. The ecoregion con- cultural expansion within the next five years. sists of a mosaic of moist gallery forests, dry forests, grasslands, and wetlands. Conversion of forested GRASSLANDS/SAVANNAS/SHRUBLANDS habitats for agriculture and pasture, draining and Grasslands, Savannas, and Shrublands channelization of wetlands and aquatic habitats, and Northern Mexico frequent burning of habitats during the dry season all threaten the natural habitats of the Llanos. 106. Central Mexican grasslands-Mexico, U.S. (Un- classified; Bioregionally Outstanding; Unclassified) Ainazonia * 199,919 km2 * Sources: 4, 6, 9, 20, 30, 99, 112, 113, 173, 174, 225, 111. Guianan savannas-Suriname, Guyana, 236 Venezuela, Brazil (Relatively Stable; Bioregionally No detailed threat information obtained. Outstanding; Moderate Priority at Regional Scale) * 128,375 km2 107. Eastern Mexican grasslands-Mexico * Sources: 103 (Brazil), 249, 280; 0. Huber, pers. (Unclassified; Locally Important; Unclassified) comm. * 3,674 km2 Cattle ranching, rice plantations, and other agri- * Sources: 11, 20, 30, 99, 112, 113, 209, 210, 225, 236 cultural expansion pose substantial threats within Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 99 the next decade. Frequent anthropogenic fires projects pose major threats over the next two increasingly degrade and destroy gallery forests in decades. Pollution and road construction represent many areas. Road building and the human activities additional threats over the next decade. that follow represent a threat to remote areas. 115. Chaco savannas- Argentina, Paraguay, Bo- 112. Amazonian savannas-Brazil, Colombia, livia, Brazil (Vulnerable; Regionally Outstanding; Venezuela (Relatively Stable; Bioregionally Out- Highest Priority at Regional Scale) standing; Moderate Priority at Regional Scale) * 611,053 km2 * 120,124 km2 * Sources: 1, 23, 24, 26, 27, 31, 72, 75, 78, 95, 187, * Sources: 27, 103, 130, 144, 189, 218, 249-252, 255, 192, 233, 239, 249-252, 272 272 The Chaco supports a diverse flora and fauna with Amazonian savannas are distributed over a wide many regional endemics, and a great complexity of geographic area and accordingly display much habitat types. Excessive grazing by domestic live- variation in their biotas and community structures. stock significantly alters community structure and Savannas with unusual edaphic conditions typically ecological processes and destroys critical aquatic maintain the highest levels of local endemism habitats. Wildfires and seasonal burning contribute (Whitmore and Prance 1987). Conservation efforts to degradation and conversion of native habitats. should emphasize geographic representation of each Agricultural expansion threatens some areas. distinct savanna community. Burning and extraction of white-sand silica are threats to the ecoregion as 116. Humid Chaco- Argentina, Paraguay, Uru- well as mining, cattle grazing, and frequent burning. guay, Brazil (Vulnerable; Locally Important; Im- portant at National Scale) 113. Beni savannas-Bolivia (Endangered; Biore- * 474,340 km2 gionally Outstanding; High Priority at Regional * Sources: 1, 23, 24, 26, 27, 37, 72, 75, 103, 233, 272 Scale) No detailed threat information obtained. * 165,445 km2 * Sources: 58, 272 117. C6rdoba montane savannas-Argentina The Beni savannas support a diverse grassland (Vulnerable; Locally Important; Important at Na- flora and many species of large mammals and birds tional Scale) characteristic of Southern Cone grasslands and * 55,798 km2 savannas. Degradation from development and * Sources: 75, 233 wildlife exploitation are serious threats to the No detailed threat information obtained. ecoregion in the next two decades. Overgrazing and frequent burning are continuing severe threats. Southern South America Eastern South America 118. Argentine Monte-Argentina (Relatively Stable; Bioregionally Outstanding; Moderate Priority 114. Cerrado-Brazil, Bolivia, Paraguay (Vul- at Regional Scale) nerable; Globally Outstanding; Highest Priority at * 197,710 km2 Regional Scale) * Sources: 75 (eastern boundary), 228, 239, 280 * 1,982,249 km2 (western boundary) * Sources: 3, 5, 71, 103, 160, 161, 186, 187, 249-252, No detailed threat information obtained. 263, 264, 268, 272 (outside of Brazil), 274, 277, 281 119. Argentine Espinal- Argentina (Vulnerable; The Cerrado constitutes one of the largest Bioregionally Outstanding; Moderate Priority at savanna-forest complexes in the world and contains Regional Scale) a diverse mosaic of habitat types and natural * 207,054 km2 communities. We chose to classify the Cerrado as a * Sources: 27, 75, 228, 246 savanna, rather than a dry forest, due to the mosaic No detailed threat information obtained. nature of the habitat. Patterns of biodiversity are complex and many regional and local endemic 120. Pampas- Argentina (Endangered; Bioregion- species are present. The scarcity of information on ally Outstanding; High Priority at Regional Scale) patterns of biodiversity prevented further sub- * 426,577 km2 division of the Cerrado into two or more eco- * Sources: 75, 228, 278 regions, a biologically justifiable revision. Agri- Conversion of natural habitats for agriculture and cultural expansion, charcoal production, and water degradation through excessive grazing are impor- 100 A Conservation Assessment of the Terrestrial Ecorcgions of Latin America and the Caribbean tant threats. Burning and draining also threaten Central America remaining natural communities. 125. Jalisco palm savannas-Mexico (Critical; 121. Uruguayan savannas-Uruguay, Brazil, Bioregionally Outstanding; High Priority at Regional Argentina (Vulnerable; Bioregionally Outstanding; Scale) Moderate Priority at Regional Scale) * 555 km2 * 336,846 km2 * Sources: 20, 30, 79, 99, 112, 113, 225, 236, 240 * Sources: 75, 86, 103, 201, 206, 272, 278 Palmar formations are dominated by various spe- No clear consensus emerged as to the most cies of palm and occur in many localities throughout appropriate conservation status for the Uruguayan Mexico, with much variation in community structure savannas. The degree of habitat degradation from and composition (Rzedowski 1978). The Jalisco and grazing, burning, draining, and exotic species was Veracruz ecoregions represent two of the largest subject to much debate. Although the ecoregion was palmar formations. Conservation efforts should eventually categorized as Vulnerable, several strive to preserve each of the many distinct types of authors of this report and other experts felt that an palmar communities found in Mexico. Some palmars Endangered or Critical categorization was more ap- do not flood (Rzedowski 1978) and would be more propriate. Excessive grazing by livestock and con- appropriately categorized as dry savannas (in the version of natural habitats for agriculture represent grasslands, savannas, and shrublands MHT). Agri- the primary threats to this ecoregion. Logging re- cultural expansion, cattle grazing, and burning duces forest cover in the western portion of the threaten the ecoregion. ecoregion. 126. Veracruz palm savannas-Mexico (Critical; Bio- Flooded Grasslands regionally Outstanding; High Priority at Regional Scale) Caribbean * 7,518 km2 * Sources: 20, 30, 79, 99, 112, 113, 225, 236, 240 122. Cuban wetlands-Cuba (Endangered; Re- Agricultural expansion, cattle grazing, and burn- gionally Outstanding; Highest Priority at Regional ing threaten the ecoregion. Scale) * 5,345 km2 127. Quintana Roo wetlands-Mexico (Relatively * Sources: 14-16, 32, 87, 131, 240 Intact; Bioregionally Outstanding; Important at Na- The Zapata Swamp on the southern coast of Cuba tional Scale) is noted for its large size and endemic species. * 2,452 km2 Draining and agricultural expansion, agricultural * Sources: 20, 69, 79, 87, 99, 179, 240 pollution, charcoal production, grazing, peat extrac- No detailed threat information obtained. tion, and exotic invasions all pose severe threats to the ecoregion. Orinoco 123. Enriquillo wetlands-Haiti, Dominican Re- 128. Orinoco wetlands-Venezuela (Relatively public (Vulnerable; Regionally Outstanding; High- Stable; Bioregionally Outstanding; Moderate Priority est Priority at Regional Scale) at Regional Scale) * 574 km2 * 6,403 km2 * Sources: 87, 89, 203, 205, 207, 240, 256, 284 * Sources: 87, 102, 113, 137, 140, 141, 176, 240, 272 Water diversion for irrigation, draining, grazing, Flooded grasslands occur in a habitat mosaic with firewood gathering, and crocodile poaching are swamp forests and mangroves in the Orinoco threats in the ecoregion. (Amacuro) Delta. Oil extraction, water projects, and dam construction represent intensive threats over Northern Mexico the next decade. 124. Central Mexican wetlands-Mexico (Critical; Anazonia Regionally Outstanding; Highest Priority at Regional Scale) 129. Western Amazonian flooded grasslands- * 362 km2 Peru, Bolivia (Relatively Stable; Bioregionally Out- * Sources: 20, 69, 87, 99, 240 standing; Moderate Priority at Regional Scale) Central Mexican wetlands are not mapped on any * 10,111 km2 of the ecoregion maps. Conversion, draining, graz- * Sources: 197, 218, 272 ing, burning, and pollution threaten these wetlands. No detailed threat information obtained. Ecoregion Profiles and Sources Consulted for Their Delmeation, Classification, and Assessment 101 130. Eastern Amazonian flooded grasslands- Expansion of agriculture and pasture degrade or Brazil (Vulnerable; Bioregionally Outstanding; destroy natural habitats. Channelization and drain- Moderate Priority at Regional Scale) ing alter critical flooding and nutrient cycles that * 69,533 km2 help maintain natural communities. * Sources: 103, 197, 218, 272 Extensive areas of these flooded grasslands are in- Montane Grasslands creasingly being converted to pasture, particularly in Central America the Amazon Delta region. 135. Mexican alpine tundra-Mexico (Vulnerable; 131. Sdo Luis flooded grasslands-Brazil Bioregionally Outstanding; Moderate Priority at (Endangered; Locally Important; Moderate Priority Regional Scale) at Regional Scale) * 147 km2 * 1,681 km2 * Sources: 20, 30, 99, 112, 113, 225, 236 * Sources: 103, 272 Alpine tundra, or zacotonal, occurs near the Some experts question whether this area deserves summits of large volcanoes along the trans- recognition as an ecoregion. No detailed threat in- volcanic range of central Mexico. Several bird, formation obtained. mammal, invertebrate, and plant species are restricted to this ecoregion. No detailed threat Northern Andes information obtained. 132. Guayaquil flooded grasslands-Ecuador 136. Costa Rican paramo-Costa Rica (Relatively (Endangered; Bioregionally Outstanding; High Pri- Stable; Bioregionally Outstanding; Moderate Priority ority at Regional Scale) at Regional Scale) * 3,617 km2 * 31 km2 * Sources: 240, 272 * Sources: 62, 139, 165, 261 Flooded grasslands are reported to occur east of Refer to Santa Marta paramo description for bio- the Rio Daule and west of Guayaquil (UNESCO diversity information. The Costa Rican paramo is the 1980). No detailed threat information obtained. northernmost example of this formation. Extensive and frequent anthropogenic fires threaten portions Eastern South America of the Costa Rican paramo. 133. Pantanal - Brazil, Bolivia, Paraguay Northern Andes (Vulnerable; Globally Outstanding; Highest Priority at Regional Scale) 137. Santa Marta paramo-Colombia (Vulnerable; * 140,927 km2 Globally Outstanding; Highest Priority at Regional * Sources: 8, 103, 240, 272 Scale) The Pantanal represents one of the world's largest * 1,329 km2 wetland complexes and supports abundant popula- * Sources: 110, 272, 276 tions of wildlife. The region is noted for huge sea- Paramo formations are restricted to high peaks sonal aggregations of water birds and caiman. The and mountain ranges of the tropics. Although Pantanal is comprised of a mosaic of flooded grass- paramo-like formations occur on isolated peaks lands and savannas, gallery forests, and dry forests. and ranges of eastern and central Africa and on Mt. During the rainy season over 80 percent of the re- Kinabalu in Borneo, this habitat type is most gion floods, a process that helps to modify the se- extensive in the Neotropics. Paramo plants and verity and frequency of floods downstream along animals display remarkable adaptations to the cold the Rio Paraguay. Agricultural expansion, charcoal and drying conditions of high elevations. Many production, and water projects pose severe threats paramo species are restricted to these habitats, and over the next two decades (Bucher et al. 1993). Pol- local endemism occurs in a wide range of taxa, lution and road construction are additional threats particularly on more isolated peaks. Paramo projected for the next decade. formations are threatened by frequent burning, grazing, and conversion for agriculture in some 134. ParanA flooded savannas- Argentina areas. The loss of downslope forests can contribute (Endangered; Bioregionally Outstanding; High Pri- to the expansion of paramo in some cases ority at Regional Scale) (UNESCO 1980). A discussion of threats to paramo * 36,452 km2 formations in the Neotropics can be found in * Sources: 75, 240 Balslev and Luteyn (1992). 102 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean 138. Cordillera de Mrida paramo-Venezuela Refer to Central Andean puna description for (Relatively Stable; Globally Outstanding; Highest biodiversity and threat information. Priority at Regional Scale) * 3,518 km2 143. Central Andean dry puna- Argentina, Bolivia, * Sources: 272, 276 Chile (Vulnerable; Regionally Outstanding; Highest Refer to Santa Marta paramo description for Priority at Regional Scale) biodiversity information. No detailed threat * 232,958 km2 information obtained. * Sources: 28, 57, 96, 272 Refer to Central Andean puna description for 139. Northern Andean paramo-Colombia, biodiversity and threat information. The dry puna Ecuador (Vulnerable; Globally Outstanding; contains saline and soda lakes that support Highest Priority at Regional Scale) populations of flamingo and other wildlife. Soda * 58,806 km2 lake ecosystems represent an unusual ecological * Sources: 97, 272, 276 phenomenon. Refer to Santa Marta paramo description for biodiversity information. No detailed threat Sonthern South America information obtained. 144. Southern Andean steppe-Argentina, Chile Central Andes (Relatively Stable; Locally Important; Important at National Scale) 140. Cordillera Central paramo-Ecuador, Peru * 198,643 km2 (Vulnerable; Globally Outstanding; Highest * Sources: 31, 57, 70, 75 (eastern boundary), 96, Priority at Regional Scale) 272 (western boundary) * 14,128 kM2 No detailed threat information obtained. * Sources: 28, 156, 272, 276 Refer to Santa Marta paramo description for 145. Patagonian steppe-Argentina, Chile biodiversity information. No detailed threat (Vulnerable; Regionally Outstanding; Highest information obtained. Priority at Regional Scale) * 474,757 km2 141. Central Andean puna-Bolivia, Argentina, * Sources: 31, 57, 70, 75 (eastern boundary), Chile, Peru (Vulnerable; Regionally Outstanding; 126, 233, 272 (western boundary) Highest Priority at Regional Scale) Both the Patagonian steppe and Patagonian * 183,868 km2 grasslands ecoregions are considered under mon- * Sources: 28, 75 (southeastern boundary), 96, tane grasslands because their ecological dynamics 207, 214, 242, 272 and conservation requirements most closely match Puna formations are montane grasslands of the those for that MHT. The Patagonian steppe support central and southern High Andes and consist of regionally distinctive communities of mammals and various communities of bunchgrasses, small birds, including many unusual higher taxa. Over- shrubs, trees, and herbaceous plants. Intact grazing and associated erosion, conversion for agri- vertebrate faunas are characterized by camelids culture, and burning are major threats. (e.g., vicufia, alpaca), condors, and a variety of high altitude rodents, marsupials, canids, and 146. Patagonian grasslands-Argentina, Chile birds. The puna formations of LAC represent one (Vulnerable; Bioregionally Outstanding; Moderate of the world's largest complexes of montane Priority at Regional Scale) grasslands, the other being in central Tibet. Puna * 59,585 km2 formations have been extensively altered for * Sources: 31, 57, 70, 75, 230, 233 agriculture and are degraded in many areas The Malvinas (Falkland) Islands are included through grazing of domestic livestock (e.g., within this ecoregion, although further bio- llamas, goats, sheep), burning, and the collection geographic analysis may prove their inclusion to of firewood. be inappropriate.3 See also note under Patagonian steppe. Excessive grazing by livestock and intro- 142. Central Andean wet puna-Peru, Bolivia, duced herbivores is a major threat. Chile (Vulnerable; Regionally Outstanding; High- est Priority at Regional Scale) * 188,911 kM2 3. A dispute concerning sovereignty over the islands exists between Argentina, which claims this sovereignty, and the * Sources: 28, 96, 207, 214, 272 U.K.. which administers the islands. Ecoregion Profiles and Sources Consultedfor Their Delineation, Classification, and Assessment 103 XERIC FORMATIONS * 3,044 km2 Mediterranean Scrub * Sources: 14, 15, 32, 131 Northern Mexico Grazing, woodcutting, and the conversion and resource exploitation associated with increased 147. California coastal sage-chaparral- Mexico, urbanization pose threats to the ecoregion for the U.S. (Critical; Globally Outstanding; Highest Prior- foreseeable future. ity at Regional Scale) * 27,104 km2 150. Cayman Islands xeric scrub-Cayman Islands * Sources: 4, 6, 9, 19, 20, 30, 99, 112, 113, 173, 174, (Endangered; Locally Important; Moderate Regional 209, 210, 225, 234 Priority) The coastal sage-chaparral communities of Califor- * 32 km2 nia are categorized as Globally Outstanding because * Sources: 81, 248 (a) Mediterranean scrub communities are rare in the Habitats are threatened by development for tour- world, occurring only in five relatively small coastal ism facilities and overgrazing. areas characterized by cool winter rains and warm, dry summers with abundant fog (i.e., the Mediterra- 151. Windward Islands xeric scrub-Windward nean, California coastal chaparral, Chilean matorral, Islands (Endangered; Locally Important; Moderate Fynbos of southern Africa, and the heathlands of Priority at Regional Scale) southwestern Australia); (b) the California coastal * 944 km2 chaparrals are the only example of Mediterranean * Sources: 12, 13, 33, 41-52, 54-56, 92, 122, 127, 175, scrub ecosystems in North America; and (c) these 202, 284 ecoregions are extraordinarily rich in species given the Grazing, woodcutting, and the conversion and re- relatively low rainfall of these regions, largely because source exploitation associated with increased urbani- of the high levels of beta diversity and associated local zation pose threats to the ecoregion for the foreseeable endemism found in these communities. Many plant future. and invertebrate groups display high diversity in these communities (e.g., the highest known species 152. Leeward Islands xeric scrub-Leeward Islands richness in bees in North America occurs in this (Critical; Locally Important; Moderate Priority at ecoregion) and there is a high degree of endemism in Regional Scale) many taxa, including vertebrates. The ecoregion is * 1,521 km2 severely threatened by rapidly expanding suburban * Sources: 12, 13, 33, 41, 42, 44-52, 54-56, 92, 122, sprawl in southern California, exotic species, and fre- 127, 175, 202, 220, 284 quent anthropogenic fires. Grazing, woodcutting, and the conversion and resource exploitation associated with increased Central Andes urbanization pose threats to the ecoregion for the foreseeable future. 148. Chilean matorral-Chile (Endangered; Glob- ally Outstanding; Highest Priority at Regional Scale) Northern Mexico * 141,643 km2 * Sources: 234, 266, 272; M.K. Arroyo, pers. comm. 153. Baja California xeric scrub-Mexico (Relatively The Chilean matorral represents the only Mediter- Intact; Bioregionally Outstanding; Important at Na- ranean scrub ecoregion in all of South America, and tional Scale) it is only one of five such ecosystems in the world * 72,377 km2 (see California coastal sage-chaparral). The biota is * Sources: 9, 20, 30, 99, 112, 113, 173, 210, 225, 236 characterized by high levels of species richness, re- Off-road vehicle use, exploitative hunting, and gional and local endemism (particularly in plants), illegal logging of boojum trees (Idria columnaris) and beta diversity in a wide range of taxa. The threaten the ecoregion. ecoregion is threatened by conversion for agricul- ture, pasture, and development, frequent anthro- 154. San Lucan mezquital-Mexico (Relatively Intact; pogenic fires, exotic species, and grazing. Locally Important; Important at National Scale) * 2,226 km2 Deserts and Xeric Shrublands * Sources: 99, 236 Caribbean No detailed threat information obtained. 149. Cuban cactus scrub-Cuba (Vulnerable; 155. Western Mexican mezquital -Mexico, U.S. Locally Important; Nationally Important Priority) (Unclassified; Locally Important; Unclassified) 104 A Conservation Assessment of the Terrestrial Ecoregions of Latin America and the Caribbean * 22,894 km2 * 29,347 km2 * Sources: 4, 6, 9, 18, 19, 20, 30, 99, 112, 113, 173, * Sources: 4, 6,9,19, 20,30, 99,112,113,173, 225, 236 174, 225, 236 Agricultural conversion and overgrazing have No detailed threat information obtained. heavily affected the ecoregion and will continue to threaten remaining fragments. 156. Sonoran xeric scrub-Mexico, U.S. (Relatively Stable; Regionally Outstanding; High Priority at 161. Eastern Mexican matorral-Mexico (Unclas- Regional Scale) sified; Bioregionally Outstanding; Unclassified) * 232,339 km2 * 26,684 km2 * Sources: 4, 6, 9, 19, 20, 30, 99, 112, 113, 173, 174, * Sources: 4, 6, 9, 20, 30, 99, 112, 113, 173, 174, 225, 209,216,225,236 236 This ecoregion is categorized as Regionally Out- No detailed threat information obtained. standing because of its exceptionally rich desert flora, subregional endemism in some taxa (e.g., 162. Eastern Mexican mezquital-Mexico, U.S. Cactaceae), and unusual floristic communities (Vulnerable; Locally Important; Important at Na- (e.g., boojum fldria columnaris] deserts of Baja tional Scale) California). Irrigation, cattle ranching, fuelwood * 138,696 km2 extraction, and hunting pose serious threats to the * Sources: 4, 6, 9, 20, 30, 99, 112, 113, 173, 225, 236 ecoregion. Cattle ranching and industrial development along the Mexico/U.S. border have already affected the 157. Northern Sonoran cactus scrub-Mexico, U.S. ecoregion heavily and will continue to pose threats. (Relatively Stable; Globally Outstanding; Highest Priority at Regional Scale) 163. Central Mexican cactus scrub-Mexico (Unclas- * 97,962 km2 sified; Bioregionally Outstanding; Unclassified) * Sources: 4, 6, 9, 19, 20, 30, 99, 112, 113, 173, 174, * 37,860 km2 210, 225, 236 * Sources: 2, 20, 30, 99, 112, 113, 225, 236 The cactus scrub communities of the northern So- No detailed threat information obtained. nora desert have some of the most diverse and un- usual desert biotas in the world. Forests of giant Central America cacti (Cereus spp.) are notable here and are associ- ated with a rich variety of both plant and animal 164. Pueblan xeric scrub-Mexico (Critical; Biore- species. Irrigation, exploitation of wildlife and plants gionally Outstanding; High Priority at Regional by collectors and dealers, and intensive recreational Scale) use threaten the ecoregion. * 6,818 km2 * Sources: 2, 20,30,99,112, 113, 225, 236 158. Mexican Interior chaparral -Mexico, U.S. Lime and mineral extraction and overgrazing by (Unclassified; Locally Important; Unclassified) sheep and goats pose threats to the ecoregion. * 22,252 km2 * Sources: 4, 6, 20, 30, 99, 112, 113, 174, 225, 236 165. Guerreran cactus scrub-Mexico (Vulnerable; No detailed threat information obtained. Bioregionally Outstanding; Moderate Priority at Regional Scale) 159. Chihuahuan xeric scrub-Mexico, U.S. * 5,232 km2 (Vulnerable; Locally Important; Important at Na- * Sources: 2, 20, 30, 99, 112, 113, 225, 236 tional Scale) Overgrazing and household firewood gathering * 399,446 km2 are threats to the ecoregion. * Sources: 4, 6, 19, 20, 99, 112, 113, 153, 173, 174, 209, 210, 225, 236 166. Motagua Valley thornscrub-Guatemala Grazing, extraction of salt and gypsum, and ex- (Critical; Bioregionally Outstanding; High Priority at ploitation of water resources are threats to the Regional Scale) ecoregion. * 2,363 km2 * Sources: 166; G. Hartshorn, pers. comm. 160. Central Mexican mezquital -Mexico (Endang- Similar dry habitats also occur in Honduras and ered; Locally Important; Moderate Priority at Re- should perhaps be included in this ecoregion. Goat gional Scale) grazing and burning threaten the ecoregion. Ecoregion Profiles and Sources Consulted for Their Delineation, Classification, and Assessment 105 Orinoco 172. ParaguanA xeric scrub-Venezuela (Endang- ered; Bioregionally Outstanding; High Priority at 167. Aruba/Cura