84 Volume 84 November 2017 ISSN 1389-9341
Volume 84 , November 2017 oetPlc n cnmc Vol. Forest Policy and Economics
CONTENTS
Abstracted / indexed in: Biological Abstracts, Biological & Agricultural Index, Current Advances in Ecological Science, Current Awareness in Biological Sciences, Current Contents AB & ES, Ecological Abstracts, EMBiology, Environment Abstracts, Environmental Bibliography, Forestry Abstracts, Geo Abstracts, GEOBASE, Referativnyi Zhurnal. Also covered in the abstract and citation database Scopus®. Full text available on ScienceDirect®.
Special Issue: Forest, Food, and Livelihoods Guest Editors: Laura V. Rasmussen, Cristy Watkins and Arun Agrawal
Forest contributions to livelihoods in changing Forest ecosystem services derived by smallholder agriculture-forest landscapes farmers in northwestern Madagascar: Storm hazard L.V. Rasmussen , C. Watkins and A. Agrawal (USA) 1 mitigation and participation in forest management An editorial from the handling editor R. Dave , E.L. Tompkins and K. Schreckenberg (UK) 72
S.J. Chang (United States) 9 A methodological approach for assessing cross-site 84 (
Opportunities for making the invisible visible: Towards landscape change: Understanding socio-ecological 2017 an improved understanding of the economic systems )
contributions of NTFPs T. Sunderland (Indonesia, Australia), R. Abdoulaye 1–120 C.B. Wahlén (Uganda) 11 (Indonesia), R. Ahammad (Australia), S. Asaha Measuring forest and wild product contributions to (Cameroon), F. Baudron (Ethiopia), E. Deakin household welfare: Testing a scalable household (New Zealand), J.-Y. Duriaux (Ethiopia), I. Eddy survey instrument in Indonesia (Canada), S. Foli (Indonesia, The Netherlands), R.K. Bakkegaard (Denmark), N.J. Hogarth (Finland), D. Gumbo (Indonesia), K. Khatun (Spain), I.W. Bong (Indonesia), A.S. Bosselmann (Denmark) M. Kondwani (Indonesia), M. Kshatriya (Kenya), and S. Wunder (Indonesia) 20 L. Leonald (Indonesia), D. Rowland (Indonesia, UK), Challenges to governing sustainable forest food: N. Stacey (Australia), S. Tomscha , K. Yang , S. Gergel Irvingia spp. from southern Cameroon (Canada) and J. Van Vianen (Indonesia) 83 V. Ingram (Cameroon, The Netherlands), M. Ewane Wild food collection and nutrition under commercial (USA) , L.N. Ndumbe and A. Awono (Cameroon) 29 agriculture expansion in agriculture-forest Origin products from African forests: A Kenyan landscapes pathway to prosperity and green inclusive growth? R.B. Broegaard (Denmark), L.V. Rasmussen (USA), H. Egelyng , A.S. Bosselmann (Denmark), M. Warui , N. Dawson (UK), O. Mertz (Denmark), T. Vongvisouk F. Maina , J. Mburu and A. Gyau (Kenya) 38 (Denmark, Laos) and K. Grogan (Denmark) 92 Prevalence, economic contribution, and determinants Forest edges in western Uganda: From refuge for the of trees on farms across Sub-Saharan Africa poor to zone of investment D.C. Miller (United States), J.C. Muñoz-Mora (Spain) J. L'Roe and L. Naughton-Treves (USA) 102 A companion journal to Forest Ecology and Management and L. Christiaensen (USA) 47 Trees, soils, and warthogs – Distribution of services Trees for life: The ecosystem service contribution and disservices from reforestation areas in southern of trees to food production and livelihoods in the Ethiopia tropics A. Byg , P. Novo (UK), M. Dinato , A. Moges , T. Tefera Special Issue J. Reed (Indonesia, UK), J. van Vianen , S. Foli , (Ethiopia), B. Balana (Ghana), T. Woldeamanuel Forest, Food, and Livelihoods J. Clendenning (Indonesia), K. Yang (Canada), (Ethiopia) and H. Black (UK) 112 M. MacDonald (Indonesia), G. Petrokofsky (UK), C. Padoch (Indonesia) and T. Sunderland (Indonesia, Guest Editors Australia) 62 Laura V. Rasmussen, Cristy Watkins and Arun Agrawal ELSEVIER
Available online at www.sciencedirect.com
1389-9341(201711)84:C;1-5 Forest Policy and Economics EDITOR-IN-CHIEF Lukas Giessen, European Forest Institute (EFI), Bonn, Germany Elsevier’s Science & Technology content is available online via ScienceDirect—providing easy, ASSOICIATE EDITORS instant, DRM-free access to searchable books, journals and citation tools all in one place. Learn more Sarah L. Burns, Georg-August Universität Göttingen, Göttingen, Germany and Faculty of Agriculture and Forestry, at http://info.ScienceDirect.com/books National University of La Plata, Argentina Ahmad Maryudi, Gadjah Mada University, Yogyakarta, Indonesia Ting Zhu, School of Economics at Shanghai University, Shanghai, China EDITORS Hemant Ojha, The University of New South Wales, Sydney, New South Wales, Australia Roland Olschewski, Economics and Social Sciences Unit, Swiss Federal Research Institute WSL Birmensdorf, Switzerland Jeremy Rayner, University of Saskatchewan, Johnson-Shoyama Graduate School of Public Policy, Saskatoon, Saskatchewan, Canada Runsheng Yin, Michigan State University, East Lansing, Michigan, USA
FOUNDING EDITOR Books Content on ScienceDirect Max Krott, Institute for Forest Policy and Nature Conservation, Georg-August-University of Goettingen, Offering comprehensive coverage of the full range of scientific disciplines, including those Germany published under the renowned Woodhead Publishing and Gulf Professional Publishing imprints, EDITORIAL BOARD over 17,000 eBooks, 59,000 Major Reference Works chapters, and 200,500 Book & Handbook Series Arun Agrawal, University of Michigan, Ann Arbor, Michigan, USA Ivan Kolenka, Technical University of Zvolen, Zvolen, Slovakia chapters are now fully integrated with journals on ScienceDirect. Mubariq Ahmad, University of Indonesia, Jakarta, Indonesia Jinlong Liu, Renmin University of China, Beijing, China Janaki Rami Reddy Alavalapati, School of Forest Resources & Constance McDermott, University of Oxford, Oxford, UK We’ll provide the resources to help you maximize awareness and usage of Conservation, Gainesville, Florida, USA Mitsuhiro Minowa, University of Tokyo, Tokyo, Japan Bas Arts, Wageningen Universiteit, Wageningen, Netherlands the titles in your collection: Steve M.J. Bass, International Institute for Environment and Martin Moog, Technische Universität München, Freising, Germany Development, London, UK Iben Nathan, University of Copenhagen, Frederiksborg C, Denmark Ɣ tutorials and guides Michael Böcher, FernUniversität in Hagen, Hagen, Germany Dodik Nurrochmat, Bogor Agricultural University, Bogor, Indonesia Ɣ local “lunch and learn” training sessions Maria Brockhaus, Center for International Forestry Research, Bogor, Jay O'Laughlin, Idaho State University, Moscow, Idaho, USA Indonesia Andreas Ottitsch, University of Cumbria, Penrith, England, UK Ɣ customized marketing toolkits Vilis Brukas, Sveriges lantbruksuniversitet (SLU), Alnarp, Sweden Anatoly P. Petrov, Education and Training Center for Forest Ɣ webinars Gérard Buttoud, Università degli Studi della Tuscia, Viterbo, Italy Specialists, Pushchino Moscow District, Russian Federation Peter Csoka, FAO of the United Nations, Rome, Italy Ɣ free MARC records NEW Fred W. Cubbage, North Carolina State University, Raleigh, North Davide Pettenella, Università degli Studi di Padova, Legnaro PD, Italy TITLES Carolina, USA Eduardo Rojas-Briales, FAO of the United Nations, Rome, Italy &RQWDFW XV WR ȈQG RXW KRZ ZH FDQ KHOS \RX PDNH WKH PRVW RXW for 2014! Wil de Jong, Kyoto University, Kyoto, Japan Nazmuz Sadath, Khulna University, Forestry and Wood Technology J. Edward de Steiguer, University of Arizona, Tucson, Arizona, USA Discipline, Bangladesh of your ScienceDirect Books investment. Katarina Eckerberg, University of Umeå, Umeå, Sweden Jin Sato, University of Tokyo, Tokyo, Japan Paul V. Ellefson, University of Minnesota, St. Paul, Minnesota, USA Birger Solberg, Norwegian University of Life Sciences, Aas, Norway Ian S. Ferguson, University of Melbourne, Melbourne, Victoria, Australia Mirjana Stevanov, Georg-August Universität Göttingen, Göttingen, Christian Pilegaard Hansen, University of Copenhagen, Germany Copenhagen, Denmark CHOICE George Hoberg, University of British Columbia, Vancouver, British Niels Strange, University of Copenhagen, Fredriksberg C, Denmark Outstanding Columbia, Canada Luca Tacconi, Australian National University, Canberra, Australian Academic Karl Hogl, Universitat für Bodenkultur Wien (BOKU), Wien, Austria Capital Territory, Australia Title Purchasing Models Michael Howlett, Simon Fraser University, Canada and National Larry Teeter, Auburn University, Auburn, Alabama, USA There are many f lexible purchase options available for the ScienceDirect University of Singapore, Sigapore Jussi Uusivuori, The Finnish Forest Research Institute, Helsinki, Finland books collection, which includes eBooks, Book Series, Handbooks, Major David Humphreys, The Open University, Milton Keynes, UK G. Cornelis van Kooten, University of Victoria, Victoria, British Makoto Inoue, University of Tokyo, Tokyo, Japan Reference Works, and Reference Modules. Columbia, Canada Yeon-Su Kim, Ecological Economics School of Forestry Northern 2013 Arizona University Flagstaff, USA Norbert Weber, Technische Universität Dresden, Tharandt, Germany Pick and Collections Subscription One time Daniela Kleinschmit, Sveriges lantbruksuniversitet (SLU), Uppsala, Yeo-Chang Youn, Seoul National University (SNU), Suwon, PROSE Choose (perpetual) Sweden Kyeonggi-do, South Korea R.R. Hawkins Award eBooks 9999 Book Series 9999 Handbooks 999 Reference Works 999 Reference Modules 99
For more information and to see a complete list of titles by subject collection, talk to your Elsevier Sales Representative today, or visit: info.sciencedirect.com/books Forest Policy and Economics
Volume 84 ( 2017 )
Special Issue:
Forest, Food, and Livelihoods
Guest Editors:
Laura V. Rasmussen, Cristy Watkins and Arun Agrawal
Amsterdam • Boston • London • New York • Oxford • Paris • Philadelphia • San Diego • St. Louis Aims and scope Oxford: Elsevier Customer Service Department, The Boulevard, Langford Forest Policy and Economics is a leading scientific journal that publishes peer- Lane, Kidlington, Oxford OX5 1GB, UK; phone: (+44) (1865) 843434; fax: reviewed policy and economics research relating to forests, forested landscapes, (+44) (1865) 843970; e-mail: [email protected] forest-related industries, and other forest-relevant land uses. It also welcomes Tokyo: Elsevier Customer Service Department, 4F Higashi-Azabu, contributions from other social sciences and humanities perspectives that make 1-Chome Bldg, 1-9-15 Higashi-Azabu, Minato-ku, Tokyo 106-0044, clear theoretical, conceptual and methodological contributions to the existing Japan; phone: (+81) (3) 5561 5037; fax: (+81) (3) 5561 5047; e-mail: state-of-the-art literature on forests and related land use systems. These dis- [email protected] ciplines include, but are not limited to, sociology, anthropology, human geog- The Philippines: Elsevier Customer Service Department, 2nd Floor, Building raphy, history, jurisprudence, planning, development studies, and psychology H, UP-Ayalaland Technohub, Commonwealth Avenue, Diliman, Quezon City, research on forests. Forest Policy and Economics is global in scope and publishes Philippines 1101; phone: (+65) 63490222; fax: (+63) 2 352 1394; e-mail: multiple article types of high scientific standard. Acceptance for publication is [email protected] subject to a double-blind peer-review process.To find out if their work reso- © 2017 Elsevier B.V. nates well with previous research published in our journal and to easily refer to these works for stipulating academic debate in the field, authors should con- This journal and the individual contributions contained in it are protected sult with previous research published in our journal, e.g. by using the search under copyright, and the following terms and conditions apply to their use functions at the top of HYPERLINK "http://www.sciencedirect.com/science/ in addition to the terms of any Creative Commons or other user license journal/13899341?sdc=1"the journal's ScienceDirect page. that has been applied by the publisher to an individual article: The journal publishes the following, peer-reviewed, citable article types: Photocopying Single photocopies of single articles may be made for personal use Research articles are full-length original scientific publications based on as allowed by national copyright laws. Permission is not required for clearly defined methods and adequate data photocopying of articles published under the CC BY license nor for Review articles provide a systematic, analytical overview of a specific photocopying for non-commercial purposes in accordance with any other field of scientific literature based on the analysis of existing international user license applied by the publisher. Permission of the publisher and publications payment of a fee is required for all other photocopying, including multiple Special Issues consist of a collection of articles resulting from previous or systematic copying, copying for advertising or promotional purposes, scientific exchange among a group of writers. Contributions to Special resale, and all forms of document delivery. Special rates are available for Issues are invited by the Guest Editor(s). Potential Guest Editors are invited educational institutions that wish to make photocopies for non-profit to submit brief proposals for Special Issues, including envisaged contribu- educational classroom use. tions, to any of our Editors at any time. Commentaries aare science-based, peer-reviewed, short communications Derivative Works (approx. 1000-3000 words) formulated as one of the following types: Users may reproduce tables of contents or prepare lists of articles includ- ○ Science Critiques: Critically discuss previous research published in our ing abstracts for internal circulation within their institutions or companies. journal or in other high-impact outlets with direct relevance to our aims Other than for articles published under the CC BY license, permission of and scope. Critiques ideally address other authors' theories, hypotheses, the publisher is required for resale or distribution outside the subscribing methodologies, or conclusions, or a mix thereof, in a critical and straight- institution or company. forward yet constructive manner. Authors embed their line of reasoning For any subscribed articles or articles published under a CC BY-NC-ND into current scientific literature; critiques are not based on opinion. license, permission of the publisher is required for all other derivative works, Through Science Critiques, authors quickly advance scholarly debate. including compilations and translations. ○ Research Trends: These articles identify emerging empirical phenomena Electronic Storage or Usage and issues of importance that should be addressed by future research. Permission of the Publisher is required to store or use electronically any Ideally, such contributions propose research agendas, raise analytical material contained in this journal, including anyarticle or part of an article research questions, and propose methods for the future study of the phe- (please consult www.elsevier.com/permissions). nomenon. The commentaries should be embedded in existing scientific Except as outlined above, no part of this publication may be reproduced, knowledge and related literature. stored in a retrieval system or transmitted in any form or by any means, Policy Forum: ○ These are short commentary pieces on contemporary, electronic, mechanical, photocopying, recording or otherwise, without internationally relevant forest or forest-related policy issues that enable prior written permission of the Publisher. researchers, policy makers, and practitioners to make timely contribu- tions to policy debates. Contributions are based on research, expert Notice analysis, literature review, or practitioner reflections regarding concrete No responsibility is assumed by the Publisher for any injury and/or damage policy issues. Pure opinion pieces will not be considered. Forum articles to persons or property as a matter of products liability, negligence or oth- should be written in an accessible style and supported by real world erwise, or from any use or operation of any methods, products, instructions examples and/or referenced scientific evidence. or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and Publication information: Forest Policy and Economics (ISSN 1389-9341). drug dosages should be made. For 2017, volumes 74–85 (12 issues) are scheduled for publication. Although all advertising material is expected to conform to ethical (medical) Subscription prices are available upon request from the Publisher or standards, inclusion in this publication does not constitute a guarantee or from the Elsevier Customer Service Department nearest you or from this endorsement of the quality or value of such product or of the claims made of it journal’s website (http://www.elsevier.com/locate/forpol). Further infor- by its manufacturer. mation is available on this journal and other Elsevier products through Elsevier’s website (http://www.elsevier.com). Subscriptions are accept- Language (usage and editing services): Please write your text in ed on a prepaid basis only and are entered on a calendar year basis. good English (American or British usage is accepted, but not a mixture Issues are sent by standard mail (surface within Europe, air delivery of these). Authors who feel their English language manuscript may outside Europe). Priority rates are available upon request. Claims for missing require editing to eliminate possible grammatical or spelling errors issues should be made within six months of the date of dispatch. and to conform to correct scientific English may wish to use the English Language Editing service available from Elsevier’s WebShop Author inquiries: You can track your submitted article at http:// http://webshop.elsevier.com/languageediting/ or visit our customer support www.elsevier.com/track-submission. You can track your accept- site http://support.elsevier.com for more information. ed article at http://www.elsevier.com/trackarticle. You are also welcome to contact Customer Support via http://support.elsevier.com. Illustration services: Elsevier’s WebShop (http://webshop.elsevier.com/ illustrationservices) offers Illustration Services to authors preparing to submit Funding body agreements and policies: Elsevier has established agreements a manuscript but concerned about the quality of the images accompanying and developed policies to allow authors whose articles appear in journals pub- their article. Elsevier’s expert illustrators can produce scientific, technical and lished by Elsevier, to comply with potential manuscript archiving requirements medical-style images, as well as a full range of charts, tables and graphs. Image as specified as conditions of their grant awards. To learn more about existing ‘polishing’ is also available, where our illustrators take your image(s) and improve agreements and policies please visit http://www.elsevier.com/fundingbodies them to a professional standard. Please visit the website to find out more. Advertising information: If you are interested in advertising or other com- mercial opportunities please e-mail [email protected] and your The paper used in this publication meets the requirements of ANSI/NISO inquiry will be passed to the correct person who will respond to you within Z39.48-1992 (Permanence of Paper). 48 hours. Printed by Polestar Wheatons Ltd, Exeter, UK. Orders, claims, and journal inquiries: please contact the Elsevier Customer Service Department nearest you: St. Louis: Elsevier Customer Service Department, 3251 Riverport Lane, Maryland Heights, MO 63043, USA; phone: (877) 8397126 [toll free For a full and complete Guide for authors, please refer to the within the USA]; (+1) (314) 4478878 [outside the USA]; fax: (+1) (314) Journal’s Website: http://www.elsevier.com/locate/forpol 4478077; e-mail: [email protected] Forest Policy and Economics 84 (2017) 1–8
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
☆ Forest contributions to livelihoods in changing agriculture-forest landscapes MARK ⁎ Laura Vang Rasmussen , Cristy Watkins, Arun Agrawal
Forests and Livelihoods: Assessment, Research and Engagement (FLARE) Network, School of Natural Resources & Environment, University of Michigan, MI 48109, USA
ARTICLE INFO ABSTRACT
Keywords: Forests support the livelihoods of a vast number of people through subsistence use of products, such as food, fodder, Subsistence and commercial contributions and medicinal plants; cash income obtained from sale of products; and more indirect ecological benefits such as the Agriculture-forest landscapes contributions of forests and trees to agricultural productivity. It is widely acknowledged that these contributions can Landscape transitions be paramount to local livelihoods, yet country- and region-wide data on their linkages remains sparse and limited Poverty attention has been devoted to understanding synergies and trade-offs between, for example, subsistence and cash Conservation policies exchange-based contributions. And because many forest landscapes are now transitioning towards patchworks of land uses owing to agricultural expansion, conservation interventions, urbanization, and other drivers, the ways in which forests support livelihoods are in flux leaving questions about potential shifts in their importance relatively unexplored. This editorial as well as the papers collected in this special issue on Forests, food, and livelihoods,discuss the ways in which forests contribute to livelihoods, including interactions between them, and how they change as landscapes transition. By doing so, we point to the need to move beyond single-year data collection to comparable temporal points and panel data as well as the importance of accounting for a) subsistence use values, b) commercial use values, and c) ecological forest contributions in poverty alleviation policies.
1. Introduction cial goods and non-marketed goods, and non-consumptive uses, such as recreation, and 2) indirect use values, e.g. watershed protection. This Forests are often portrayed as central to the poverty alleviation approach is also in line with more recent attempts to conceptualize how efforts of millions of rural smallholders across the developing world forests contribute to livelihoods, such as Newton et al. (2016) conceiv- (Sunderlin et al., 2005; Sunderlin, 2006; Shackleton et al., 2007; ing the livelihood dimension as being split into subsistence and Wunder et al., 2014a). Even when only the officially reported monetary commercial means. Also, it aligns with scholarly efforts on the contributions of forests to developing world economies are taken into valuation of ecosystem services asserting direct use values, or sub- account, they exceed US$ 250B – more than twice the value of total sistence and commercial contributions, as usually related to provision- development assistance and more than the annual global output of gold ing or cultural ecosystem services whereas indirect use values or and silver combined (Agrawal et al., 2013). Yet, these direct cash ecological contributions primarily are associated with regulating eco- exchange-based contributions are only one type of contributions that system services (Pascual et al., 2010). Although these conceptualiza- forests make to livelihoods. The various ways in which forests tions seem promising, analyses often focus on the various types of forest contribute to livelihoods have been investigated in a number of contributions as being separate from each other despite expectations of manners, for example according to three different functions: safety synergies and trade-offs between, for example, the subsistence and nets, support of current consumption, and a pathway out of poverty commercial outcomes generated by forests. Attention to these different (Vedeld et al., 2007). While such categorization might be useful to contributions, as well as their interactions, is particularly needed grasp forest reliance in certain time periods, the empirical examination because of ongoing landscape changes. For example, conservation of the actual contributions from forests makes it clear that there are interventions and agricultural expansions affect, both positively and substantial overlaps among those categories and that they are not negatively, which forest products local people may derive, from where, collectively exhaustive (e.g. Dokken and Angelsen, 2015) nor mutually when, and at what quantities (FAO, 2016), leaving questions about exclusive. Here, our approach to forest contributions rests on Gregersen potential shifts in their importance relatively unexplored. Thus, the (1995) who perceived the contributions as divided between 1) direct ways in which forests contribute to livelihoods, the interaction between use values associated with consumptive uses, including both commer- them, and the influence exercised upon the contributions by ongoing
☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” published at the journal Forest Policy and Economics 84C, 2017. ⁎ Corresponding author. E-mail address: [email protected] (L.V. Rasmussen). http://dx.doi.org/10.1016/j.forpol.2017.04.010 Received 16 May 2016; Received in revised form 13 April 2017; Accepted 26 April 2017 Available online 08 May 2017 1389-9341/ © 2017 Elsevier B.V. All rights reserved. L.V. Rasmussen et al. Forest Policy and Economics 84 (2017) 1–8 landscape transformations are the key focus of this introductory article where forests and agricultural land use coexist in mosaic patterns and the articles that follow. (Agrawal et al., 2014). One key challenge to further our understanding of how forests With our point of departure as the findings presented within the contribute to livelihoods is the gap in reliable data. Country- and eleven papers in this special issue, our editorial addresses the role of region-specificefforts indicate that where data is available, the non- forests in socioeconomic development in a changing agriculture-forest cash economic contributions of forests to household and national landscape. As presented by our contributors, this special issue helps economies range between three and five times the formally recognized advance our understanding and quantification of forest-livelihood cash contributions (Agrawal et al., 2013). Quantifying the value of linkages which is key to researchers, as well government agencies, products not obtained at markets, for example water, any ecosystem private sector actors, and NGOs, who fund and implement conservation services, or cultural products, is particularly challenging and can result and development programs in and around forests. Such interest is in substantial underestimations of the total forest benefits (Babulo rising, especially in the context of forest-based climate change mitiga- et al., 2009; Ojea et al., 2016; Wahlen, this issue). Also, even tion mechanisms (e.g. REDD+) that aim to achieve both forest commercial, formally recognized monetary contributions of forests conservation and socioeconomic development (Kanowski et al., tend to be underestimated. Reasons include that national accounting 2011). We outline first the general contours of the key ways in which systems often lump forestry under agriculture (FAO, 2008) or do not forests contribute to livelihoods as well as the interaction between address forest products and contributions at all (Lescuyer et al., 2016; these. We then turn to the broader issue of land use change in Wunder et al., 2014a). For example, in population-representative agriculture-forest landscapes and describe how such changes affect household surveys, such as the Living Standards Measurement Surveys the outlined ways and their interaction. Subsequently, we synthesize (LSMS), information on forest products is often limited, at best the findings from the articles in our special issue. These contributions including only questions on fodder, building materials, or fuel, while were first presented at the annual FLARE (Forests & Livelihoods: the link between forests and food intake remains a black box (Powell Assessment, Research, and Engagement) network meeting held in Paris et al., 2015). in November 2015. Because the subsistence use and ecological values of forests in many cases are ‘invisible’ and the commercial contributions tend to be 2. The role of forests in changing agriculture-forest landscapes underestimated, natural resource accounting methods also often fail to acknowledge the role of forest products for local livelihoods 2.1. Subsistence contributions from forests (PROFOR, 2008). As a result, many resource and poverty reduction strategies have been based on inadequate evidence and insufficiently Wild foods support food security and nutrition in a number of direct considered forest products. For example, when the subsistence use ways (Hickey et al., 2016). First, people living near forest have greater value of forests is ignored in land use decisions, governments and other access to forest foods such as wild fruits, leafy greens, grubs, snails, and agencies may choose to promote agriculture expansion over forest bush meat (Arnold et al., 2011; Ickowitz et al., 2014; Powell et al., conservation, not recognizing the full livelihood impacts of these 2013a, 2015; Pingali, 2015). For example, Shackleton et al. (2002) choices. To help overcome such challenges, important efforts have demonstrate how wild herbs and wild fruits were among the top forest recently been made to systematize data collection and estimate both resources extracted from communal woodlands in South Africa by 91% subsistence and commercial forests contributions. For example, and 82%, respectively, of surveyed households. Also, the amount of Angelsen et al. (2011) provide a guide to improve the quality of data wild edible herbs consumed has been estimated to range from 12 to collection efforts based on the Poverty Environment Network (PEN) over 130 kg per household per year (Dovie et al., 2002; Shackleton project (CIFOR, 2007) which led to a global dataset with information on et al., 2002; Twine et al., 2003). Likewise, Christensen (2002) shows forest contributions from 333 villages in 24 countries. Using this how communities in Borneo consumed as many as 700 different wild dataset, Angelsen et al. (2014) suggest that natural forests on average and semi-wild plant species, and Bharucha and Pretty (2010) summar- provide 21% of total household incomes but with large regional ize evidence from 36 studies in 22 countries in Asia and Africa and variations. The same dataset also shows (Hickey et al., 2016) that estimate that an average of 90–100 wild plant food species were used or although wild foods contribute just 4.2% to household incomes, 77% of available per location. That wild plants contribute both to diversity and the respondents report collection for subsistence purposes. quantity of food consumption is thus well recognized. This also holds These examples are valuable and important steps towards the for wild meat. For example, Pangau-Adam et al. (2012) demonstrate, acknowledgement of both commercial and subsistence forest contribu- based on 546 records of meals in Indonesia, that the percentage of tions. But challenges persist. For example, it remains difficult to value meals containing wild meat (51%) was much greater than those certain products, such as water and cultural products, impeding their containing fish (17%), domestic animals (14%), and vegetables inclusion in many national and global accounting systems. Also, there is (16%). Likewise, in the Congo Basin the daily per capita wild meat a continued need to move beyond single-year snapshots to comparable consumption by rural dwellers was ten times greater than that of urban temporal points and panel data. Recently, scholars have made advances dwellers (Wilkie and Carpenter, 1999) and more than 85 animal species in such efforts: one example is Walelign et al. (2016) who use house- were consumed (Ingram et al., 2012). In a more recent study from the hold total income panel data sets from Nepal and find that total Amazon, Van Vliet et al. (2015) found that about 11% of all meals absolute environmental income was highest in 2006 and lowest in consumed in rural communities contained bush meat as compared to 2009 with average environmental reliance decreasing from 19.3% in only 2% in urban areas. 2006 to 12.7% in 2009 and 12.5% in 2012. Better use of panel data is Second, people that include trees or fallow areas in their agricultur- critical for an improved understanding of the degree of volatility al systems may benefit from harvesting wild plant material and animals inherent to both commercial and subsistence forest contributions and (Cruz Garcia and Price, 2012). Recently, scholars have demonstrated the impacts over time of certain livelihood-related interventions. that the collection of wild foods (e.g., nuts, fruits, and leaves) might in Longitudinal studies are also particularly needed because the very fact be much higher in such non-forest environments as compared to ways along which forests benefit livelihoods are in flux. That is, the the forest (Pouliot and Treue, 2013; Rasmussen et al., 2016). rapid expansion of agriculture is causing forest-landscapes to transition Although the scholarly efforts described above clearly portray wild towards new patchworks of land uses (Newton et al., 2013) and with foods as ubiquitous and habitually used by rural people around the these ongoing changes, the boundary between forest and agriculture is world, their actual contribution to nutrition is dependent not only on becoming fuzzier in both space and time. In this editorial, we therefore availability but also on the number of people consuming them, use the term ‘agriculture-forest landscapes’ which we define as places frequency of use, and quantities consumed relative to other foods –
2 L.V. Rasmussen et al. Forest Policy and Economics 84 (2017) 1–8 all factors which vary greatly from one socio-ecological setting to point is also well illustrated in a more recent case study from the another (Powell et al., 2015). Yet, the extant research often lacks Himalayas by Shrestha and Bawa (2014) who illuminate how the empirical specifications of these actual contributions of wild foods to caterpillar fungus (Ophiocordyceps sinensis), the world's highest-priced nutrition (Vira et al., 2015). This is worrying as, for example, the Food biological commodity, accounted for 53% of the sampled household's and Agriculture Organization asserts that although forests may not be a cash income on average. Another example is from Burkina Faso, where major source of global food supply when measured in calories, they are Pouliot (2012) shows that income generated from the collection of critical in terms of overall nutrition (FAO, 2014). This view is supported environmental resources accounted on average for 28% of household by Powell et al. (2013b) who demonstrate that in Tanzania, wild foods income across the sampled population. In this case, shea (Vitellaria were harvested by the majority of the sampled population but paradoxa) contributed on average 7% of total household income. contributed only 2% of total energy. Yet, wild foods contributed 31% Further, even higher estimates were reported in, for example, the of vitamin A (RAE), 20% of vitamin C, and almost 20% of iron Peruvian Amazon, where products derived from the forest accounted consumed (Powell et al., 2013b). One of the other few notable estimates for more than 60% of average total incomes (L'Roe and Naughton- of the actual contribution of wild foods to nutrition is Fungo et al. Treves, 2014). (2015). They illuminate how a daily intake of one readily available Overall, numerous case studies indicate that the contribution of forest food (fruits or nuts), widely consumed among local populations forest products to household budgets is often substantial, but that the in Cameroon, could supply 100% of iron and zinc recommended dietary scale of the contribution varies widely, depending on context, land use allowances for children aged 1–3 years and furthermore provide 100% types from which the products are derived (see also Miller et al., this daily vitamins C and E requirements for both children and adults. How issue), and wealth group, with often higher proportional contributions wild meat may contribute to nutrition has been addressed by Nasi et al. to poorer households (Angelsen et al., 2014; FAO, 2014; Vira et al., (2011) who found that in rural areas of the Congo Basin, five to six 2015). In their review of 21 Asian cases, Belcher and Kusters (2004) million tons of bush meat were harvested each year and contributed up find that NTFPs contributed more substantially (50% or more) to to 80% of fats and proteins consumed. Also, Golden et al. (2011) households that specifically specialized in the collection of particular suggest that in Madagascar, the loss of access to wild bush meat may products. Pouliot (2012) also highlights how environmental income result in a 29% increase in the number of children with anemia. varied with wealth group across the sampled households in Burkina In addition to food, subsistence contributions from forests include Faso, with 43% of poorer households' income and 18% of richer fodder, medicines, construction materials, fuelwood, and other non- households' income coming from the forest. Likewise, Chhetri et al. cash material goods to households – all often referred to as the ‘hidden (2015) demonstrate how reliance on environmental income decreased harvest’ (Scoones et al., 1992). Especially studies of fuelwood con- with rising income across sampled households in Nepal; average sumption have been widespread and considerable scholarship has environmental income was estimated at 16% of total household explored fuelwood consumption patterns using weight surveys (Singh income, ranging from 30% for the lowest income quintile to 11% in et al., 2010). Estimates from rural villages in India range from 20 to the highest. 25 kg fuelwood/household/day (Singh et al., 2010) and a recent study The studies mentioned above also indicate that externalities, such as from Nepal has demonstrated how more than 80% of the sampled changing global demands and technological change, drive the profit- population was engaged in fuelwood collection for own consumption ability of NTFPs to a great extent (for a case on the profitability of the (Meilby et al., 2014). Since fuelwood is not only consumed by house- bush mango value chain, see Ingram et al., this issue). Whereas those holds but also traded locally, market-based and contingent valuation NTFPs which can be traded beyond the immediate area are more techniques have been widely employed to estimate the value of vulnerable to changes in accessibility and transportation, there is also fuelwood consumption (Campbell and Luckert, 2002). greater potential to upscale economic prospects of such forest products, for example via new political and institutional conditions that promote 2.2. Commercial contributions from forests and value origin specific products in the larger market (see Egelyng et al., this issue). Although locally traded forest products are more Apart from the subsistence contributions, forests support livelihoods immune to such externalities, these products may be overtaken by new through income earning opportunities. Despite difficulties in quantifi- domestic items (such as plastic brooms) as these become more available cation, overall estimates of the value exist. For example, based on a in local rural markets and substitute products traditionally derived from meta-analysis of 51 studies from 19 countries Vedeld et al. (2007) the forests (Belcher and Kusters, 2004). found that forest income represented on average 22% of the total Moreover, the importance of environmental income is assumed to income in the population sampled. Using the PEN dataset from 24 vary depending on market access (World Health Organization, 2015). countries, Angelsen et al. (2014) have demonstrated that, for the In communities isolated from markets, NTFPs appear to account for a surveyed communities, environmental income constituted 28% of total greater share of household consumption and a lower share of household household income, around three-quarters of which came from natural income as compared to communities with access to robust commodity forests. Yet, large regional variations were found – i.e. the environ- markets (Ferraro et al., 2012). Thus, it has been asserted that to mental income contribution was estimated at 32% across the Latin increase the contribution of forest products to household income, American sites and 22% across the African sites. Across all sampled regional markets need to be developed, and processors linked to communities, the major products and their contributions to forest domestic and international markets to further improve value chain income were fuelwood (35%) and food (30%), albeit also with variation opportunities, albeit this might translate into unprecedented demands between geographic regions, with food for example, being more placing pressure on certain species (World Health Organization, 2015). important in Latin America than in Africa, and the reverse being true In sum, any attempt to generalize about income contributions of for fuelwood. forests faces inherent challenges related to the diverse nature of A wide range of other studies has also indicated an important role available data by site and region, and variations in income based on for non-timber forest products (NTFPs) in supporting rural peoples' the type of products extracted, where the products are extracted from, incomes. For example, Belcher and Kusters (2004), Sunderland et al. by whom, and why they are extracted (see also Bakkegaard et al., this (2004), and Alexiades and Shanley (2005) collated a series of case issue). Moreover, only limited information is available on how cash studies on the role of forest products for rural incomes in Africa, Asia, incomes from these resources are spent with regard to food security and and Latin-America, respectively. Each of these three papers address a nutrition. range of forest products and the authors show how diverse different products were in their potential to yield an income contribution. This
3 L.V. Rasmussen et al. Forest Policy and Economics 84 (2017) 1–8
2.3. Ecological contributions from forests products reliance) are not well understood. In particular, questions about how commercial contributions from forests are used to support Whereas Gregersen (1995) noted that forests also play an indirect food purchase and dietary diversity and how they supplement or role for livelihoods, more recent work has framed certain ecosystem replace subsistence products are left relatively unexplored. For exam- services (or ecological contributions) as indirect benefits of forests (Vira ple, it remains unknown when and why certain products fall out of et al., 2015). That is, many supporting and regulating ecosystem consumption patterns as incomes increase due to sale of other forest services associated with forests are critical to agricultural production. products. By contrast, it is well recognized that harvesting of wild For example, forests, woodlands, and trees found or integrated into products, foods in particular, tends to dwindle as people gain more other landscapes control water flows, and prevent soil erosion and access to purchased goods, as improved supplies of food crops diminish nutrient leaching, which are essential for food production (Vira et al., the need to depend on forest foods, or as the opportunity cost of 2015). The presence of trees may in fact even increase yields of gathering wild products rather than purchasing them, becomes higher cultivated crops (see Makumba et al., 2006; Reed et al., this issue) (Arnold and Periz, 2001). thereby contributing to food and nutrient intake (Mbow et al., 2014). 2.5. Changing agriculture-forest landscapes 2.4. The interrelation between subsistence and commercial products Because agricultural expansion often happens alongside the imple- Many forest products are used both for household consumption and mentation of conservation policies, competing demands are placed on as an income source (Senaratne et al., 2003; Dave et al., this issue). land, penetrating many agriculture-forest landscapes (Newton et al., Food products, like spices, meat, and honey, as well as fuelwood are 2013; Sayer et al., 2013; L'Roe and Naughton-Treves, this issue). Yet, examples of products that commonly are classified as ‘subsistence the consequences for the ways in which forests contribute to livelihoods products’, but quickly can become commercial (Vira et al., 2015). are poorly understood and not easily resolvable. For example, it Because household consumption and sale might happen both simulta- remains a fundamental controversy whether conservation interventions neously and at different times, the situations in which sale takes such as protected areas create or reinforce poverty traps by restricting precedence over consumption or vice versa is likely related to the very forest product use, displacing people and livelihoods, and not managing reason why rural households trade NTFPs. Shackleton et al. (2011) ecosystem disservices, or if they instead offer mechanisms to improve identify four reasons for trading NTFPs, all assumed to influence the the livelihoods of rural people (Ferraro et al., 2011; Byg et al., this type and amount of products collected. NTFPs might be traded because: issue). 1) it can be an insurance to bridge income gaps or to manage specific One obvious way, in which the ongoing landscape changes affect income needs in emergency situations; 2) it is a recurrent livelihood forest product extraction is by curtailing the forest area where people diversification and risk reduction strategy serving as a complement to derive resources with extensive and increasingly adverse effects on a agriculture or income smoothing; 3) they can be a primary or regular variety of interlinking aspects, including biodiversity and wild food source of income that can act as a stepping stone out of poverty; or 4) provisioning (Broegaard et al., this issue; Sunderland et al., this issue). there is a lack of alternatives, an approach which may turn into a long- These ‘pressures’ may also influence the very type of forest products term livelihood source. people derive. Restrictions on hunting and support for the use and Because NTFPs allow households to respond to shocks and fill gaps marketing of certain products such as wild honey are examples hereof. during seasonal shortfalls (through both subsistence reliance and sale), Also, Pangau-Adam et al. (2012) show that hunting of bush meat in NTFPs are also referred to as ‘safety nets’ (Colfer, 2008; Powell et al., Indonesia has shifted from a purely subsistence activity towards a more 2015; Vira et al., 2015; Wunder et al., 2014b). For example, a number commercial form due to the emergence of markets enabled by rapid of studies have documented higher dependence on wild foods in the development and improved infrastructure. As we would expect from lean or food insecure season (de Merode et al., 2004; Powell et al., related economic theory (Homma, 1996), increasing integration into 2013b). There is also strong evidence that fruit trees, in particular, have market systems, and the resulting opportunities to a) meet some needs a critical role in responding to shocks and filling seasonal gaps given from imports rather than local production and b) generate income by their resistance to drought and availability during pre-harvest food exporting other products previously consumed locally, can thus rapidly shortages (Kehlenbeck et al., 2013). However, recent studies challenge alter patterns of forest product use. While the reliance on low value, such views. Using the PEN dataset, Wunder et al. (2014b) find that labor-intensive forest products may decline, the extraction of higher households ranked wild product-extraction as a response to shocks value products should increase. For example, Senaratne et al. (2003) lower than most common alternatives. On average, only 10% of show how the relative importance of subsistence forest products households in the full sample used environmental resources as their dropped over time among rural households in Sri Lanka, because forest primary shock response. Likewise, not much evidence was found in management restricted villagers' access to forest areas and their ability support of the popular view that forest income served as a primary to hunt wild animals. seasonal gap filler to make up for intra-annual income shortfalls, e.g., in That it is paramount to consider this changing nature of agriculture- agricultural off-seasons. Rather, households were more likely to seek forest landscapes when addressing forest-livelihood linkages is begin- temporary employment than to exploit forests. ning to gain attraction in recent scholarly work. For example, Jiao et al. How subsistence and commercial contributions from forests interact (2015) find that land concessions, such as establishments of mono- is also shaped by the number and type of NTFPs households collect: i.e. culture plantations of rubber and sugarcane, have had negative impacts whether households rely on a myriad of forest products or a single high- on environmental income across 600 sampled households in Cambodia. value product. Belcher et al. (2005) suggest that maintaining product While environmental income on average made up 33% of household diversity is critical to stable, albeit low-level economic benefits, but as income, this proportion decreased by 24% for households residing in a markets develop, products are prone to exploitation and depletion, land concession village, relative to what they would have secured if effectively limiting any substantial profit. Conversely, one-product they had not experienced a land concession event. Such research efforts incomes (e.g. shea butter in Africa (Schreckenberg, 2004), and the present important steps to better understandings of how forest-liveli- song rong mushroom in China (Chen, 2004)) are perhaps riskier, but hood linkages are in flux, yet they are far from enough. For example, have the potential for greater profit. While both strategies present changes in agriculture-forest landscapes are also about how changes in potential avenues for increasing income or consuming directly (thereby governance arrangements affect livelihoods and wellbeing - the social off-setting money to purchase assets), the interlinkages and trade-offs and economic relationship of people to forests is inevitably structured between sale and consumption (and between single- and multiple by the institutions that govern what happens in these landscapes (Lund
4 L.V. Rasmussen et al. Forest Policy and Economics 84 (2017) 1–8 et al., 2017). ranging from 28 to 44% of harvester's total household incomes. For those stakeholders located further away from the forest, the income 3. Key results from this special issue gains were even more pronounced. That is, bush mango contributed an average of 48% of exporter's annual income and 57% of retailer's. The eleven ensuing articles in this special issue examine the Because bush mango made such substantial livelihood contributions nuanced dynamics of resource extraction in agriculture-forest land- throughout the value chain, the authors call for attention to the scapes. Together, the papers seek to fill some of the empirical- apparent unsustainable nature of the chain indicated by a) declining quantitative knowledge gaps related to forest-livelihoods linkages. As wild resources evidenced by harvesters travelling longer distances, b) such, they provide a more situated analysis of how these levers work continuing high demands, c) clearance of the species' natural forest than is feasible through a general examination of existing literature. In habitat, d) low levels of cultivation, and e) lack of consistent regulatory presenting results, we start with a review of a set of key studies control and enforcement. (Wahlen), focused on the first type of forest contributions outlined Egelyng et al. bring a larger theoretical view into the discussion of above: subsistence contributions from the agriculture-forest landscape. commercial contributions from forests. They focus on the potential to We then turn to case studies oriented towards the commercial upscale economic prospects of forest products via new political and contributions of forests to livelihoods. Bakkegaard et al. present a case institutional conditions that promote and value origin specific products from Southeast Asia, whereas Ingram et al., Egelyng et al., and Miller in the larger market. Based on interviews with key stakeholders in the et al. focus on African cases. Next, we present the results of a systematic production of Mwingi Honey and Kakamega Silk in Kenya, they find review (Reed et al.) on the ecological contributions of forests: the effect that the environmental conditions under which a product is grown is of tree presence on food yields. The interlinkages and trade-offs one of the most important factors in upscaling these products to the between subsistence and commercial contributions are addressed by standard required for Geographic Indication (GI) certification. The Dave et al. Finally, Sunderland et al., Broegaard et al., L'Roe and authors argue that GIs can protect public goods and collective values Naughton-Treves, and Byg et al. offer insight on the ways in which the and improve smallholders' production of these products with corre- changing nature of agriculture-forest landscapes affect how forests sponding increases in forest-derived incomes. Therefore, the time is ripe support livelihoods. for developing countries, particularly in Africa, to pursue this pathway of economic development. 3.1. Studies on subsistence contributions from forest The paper by Miller et al. considers, like that by Reed et al. (this issue), the effect of tree presence on farms. While Reed et al. devote The first article by Wahlen presents an overview of the continuing specific attention to the effects on food yields; Miller et al. address the failure of governments and international agencies to take stock of and contribution of trees to household income. Using nationally represen- evaluate the various local level and non-monetized uses of forests in tative data from the Living Standards Measurement Study - Integrated designing policies and programs. Providing a review of several key Surveys on Agriculture (LSMS-ISA) conducted in five African countries: studies, she argues that a focus on goods with high commercial value Ethiopia, Malawi, Nigeria, Uganda and, Tanzania, they provide the first and a lack of rigorous measurement techniques account for this national-scale evidence on the contribution of trees outside forests to undervaluation. For example, few studies have provided reliable household incomes in Africa. They find that trees were widespread on estimates of the economic value of low value goods, albeit many farmlands as more than 30% of all rural households reported having studies have documented the number of people who rely on such trees for productive use on their farms. Most importantly, they find that products. Because values of forests are not properly and fully deter- the production from these trees accounted for 6% of annual gross mined, forests remain undervalued and hence governments do not income on average for all rural households, while the contribution invest in them or make sound decisions. The paper concludes with a call increased to 17% of annual gross income for those households with for more consistent and rigorous data collection efforts that are also trees on-farm. The results demonstrate that income contributions from accessible to policy makers. trees outside forests deserve greater attention in debates on linkages between changing agriculture-forest landscapes and poverty alleviation 3.2. Studies on income contributions from forests in Sub-Saharan Africa.
With the goal of collecting more precise data on the contributions of forests to household livelihoods, Bakkegaard et al. present a new 3.3. Studies on ecological contributions from forests Forestry Module to be used in conjunction with LSMS instruments. They pilot the module across four villages in Indonesia, where national Reed et al. present the results of a systematic review of 74 studies and regional surveys of income and wellbeing do not yet consider the addressing the contribution of forests and trees to agricultural produc- role of forest and wild products. Their pilot assessment reveals that tion and livelihoods in the tropics. They find a net positive effect of tree mean per capita forest and wild product income shifts according to a presence on food yields in the majority of studies (47%), albeit with geographical “forest gradient” represented by the four villages. In the some regional variation. For example, in the Americas and Africa, tree forest-rich upstream village, mean forest and wild product income and presence was more likely to enhance food yields, while in Asia the mean forest-related wage and business incomes far exceeded mean opposite was the case with the majority of studies (48%) reporting agricultural income. Also, consumption of forest products and the decreased food yields because of tree presence. However, when importance of forest products as a coping strategy in times of shock comparing the effects of trees on crop yields with the overall livelihood was most pronounced in the forest-rich upstream village. Overall, the effects of trees across all 74 studies, they find a reduction in the total results indicate that the forest-related income is not adequately negative effects from 36% for yield to 16% for overall livelihood effects, reflected in national socioeconomic surveys, suggesting that a more suggesting that a reduction in yields might be compensated by other systematic collection of such data may lead to a change in percentages benefits provided by trees to the farm system. In sum, the review of households that are actually under the poverty line. highlights that while trees can provide important livelihood benefits Next, Ingram et al. apply a value chain approach to study how bush that go beyond the products extracted from the trees, there is little clear mangoes (Irvingia spp.) contribute to the livelihoods of different evidence of the effect of multiple interacting ecosystem services flowing stakeholders in the chain, from harvesters in three major production from forest fragments to food systems. areas in Cameroon to traders in Cameroon, Nigeria, and Equatorial Guinea. They find that bush mango contributed on average 31%,
5 L.V. Rasmussen et al. Forest Policy and Economics 84 (2017) 1–8
3.4. Studies on the interrelation between subsistence and commercial use changes. products L'Roe and Naughton-Treves use longitudinal field data collected in 1993 and 2015 to describe a transition in ownership and use of forest- The paper by Dave et al. suggests that even though we distinguish at edge agricultural lands around a protected forest in Uganda. They the individual product level between subsistence and commercial describe how the land has risen in value and attracted wealthier land contributions, some forest products are extracted for both purposes. owners. They find that those households with more land, wealth, and Focusing on 240 households across two study sites in Madagascar, the education were more likely to have planted commodity and non-food authors show how more than 57% of the respondents engaged in honey investment crops, such as tea, eucalyptus, and coffee, while poorer production, and of those more than 45% sold anywhere from half to all residents lost access to the forest edge as a safety net and were being of their honey. Further, the authors find that 44% of the respondents squeezed onto ever smaller parcels. Although there were economic depended upon three or four dietary items from the forest and benefits to be gained from the commodity crops, their work portends of surrounding habitats while over 82% depended upon three or four future scenarios where non-food crops dominate agriculture-forest raw materials. All respondents who collected wild vegetables and fruits landscapes and where marginalized or disadvantaged populations' (75%) did so for household consumption only, whereas both subsis- access to forest and forest-adjacent land decreases or, at worst, tence and commercial purposes were fueling extraction of raw materi- disappears entirely. als. For example, of those who produced charcoal (26% of respondents), Finally, Byg et al. bring to light the ways in which conservation and 70% were selling half or more of the production. Also, hazard development projects impact livelihoods, both positively and nega- mitigation benefits from forests in the form of reduced flooding and a tively. Focusing on Ethiopia, they explore the factors that affect the lower amount of sediment flowing onto fields were valued by more creation and distribution of ecosystem services and disservices of than 58% of the respondents. Yet, no association was found between reforestation exclosures and they then describe how these were extraction of food or raw materials and viewing forests as important for unequally distributed across households. For example, households that hazard reduction. The findings presented raise the fundamental insight were closer to the exclosures experienced more crop-raiding than those that in the development of methods to derive estimates of subsistence farther away, yet households that were wealthy had the means to contributions, one must account for the dual purposes associated with protect themselves against such risks, for example by planting less certain products and special attention should be devoted to those vulnerable crops in plots near the forest, and more vulnerable crops in benefits, such as hazard mitigation, which remain difficult to value. plots farther away. The paper points to an urgent need to acknowledge the trade-offs inherent in these kinds of conservation and development 3.5. Studies on changing agriculture-forest landscapes projects, wherein environmental benefits for humans (e.g. ecosystem services) are the explicit focus, yet unintended dis-benefits can occur. Sunderland et al. present preliminary results from a multi-disciplin- ary research initiative aiming to assess the socio-economic and 4. New perspectives and future directions environmental outcomes of agrarian change processes across six countries: Ethiopia, Cameroon, Indonesia, Bangladesh, Zambia, and The purpose of this special issue is to encourage, develop, and Burkina Faso. In each country, data collection was stratified across a expand discussions regarding subsistence, commercial, and ecological gradient of indicators of agrarian change and agricultural modification. contributions from forests to livelihoods, their interlinkages, and the The preliminary results show that the agrarian trajectory from diverse ways in which such contributions are altered by rapid landscape smallholder agricultural production towards more intensive agribusi- transformations. As such, the issue brings together a number of studies ness appears to impact fuelwood use and local people's dietary that add to the empirical and quantitative knowledge of forest–liveli- diversity, among other factors. For example, in Ethiopia rural house- hood relations. Attempts such as those undertaken by the papers in this holds closest to the forest used significantly more fuelwood from the special issue to develop assessments of forest contributions face, forest than households further away from it. Also, farms closest to the however, persistent problems related to the varying definitions of forest had more livestock and thus more manure, which they concen- forests and the lack of comparable data on forest cover and reliance trated on home gardens producing a variety of nutrient-dense food on and use of forest products. Indeed, forest-livelihood relations is a items. The proportion of farms having a home garden was 86%, 62%, field of inquiry characterized by sparse country- and region-wide data and 39% in the near, intermediate and distant zones, respectively. on forests, poverty, health, and other institutional factors. Owing to the Accordingly, the dietary diversity score was significantly higher for limited data on forest product extractions, particularly subsistence rural household closest to the forest. The authors suggest in line with products, comparisons across countries and regions remain difficult Broegaard et al. (this issue) that a ‘nutrition transition’ has occurred, and existing knowledge is often deemed insufficient as a firm basis for whereby diverse forest foods have been replaced by processed food. policy action. Yet, the findings presented in this special issue are The article by Broegaard et al. scrutinizes how wild food provision- sufficient to infer that forests continue to provide substantial subsis- ing and local people's diet is changing in landscapes transitioning tence and commercial contributions as well as ecological benefits that towards increasing cash crop production and more conservation inter- should be considered in the context of forestry, natural resource, and ventions. The authors question whether common assumptions about poverty alleviation policies. rising incomes leading to reduced hunger or improved nutrition are true The studies collected in this special issue also direct attention in the shifting cultivation systems of Laos. Based on a dataset collected towards areas of research urgently in need of additional investigation across three villages representing a gradient of landscape transitioning, and more rigorous data tools to conduct such investigation. First, the they find that although the expansion of commercial agriculture role of on-farm trees for both subsistence purposes and cash income appeared to improve rice sufficiency through rising household incomes, must be acknowledged. Second, we need a better understanding of it failed to secure a nutritionally-balanced diet. In fact, cash crop opportunities for enhancing commercial values of forest products for expansions appeared to threaten it. That is, the protein intake from wild local producers. Third, more refined data collection tools and better food decreased with expanding cash crop cultivation and if households estimation methods are necessary to accurately value certain products, failed to replace the wild foods with other protein sources, it could such as water and cultural products, which continue to be overlooked in result in protein deficiencies in the order of 22–36%. The study thereby many national accounting systems (Rasmussen et al., 2017). Fourth, highlights the importance of devoting attention to wild foods in poverty there is a need to move beyond single-year snapshots to comparable alleviation and conservation programs, particularly because the wild temporal points and panel data. Directing attention to these issues is food contributions to local diets are changing in the face of rapid land- important to further the understanding of the ways in which forests
6 L.V. Rasmussen et al. Forest Policy and Economics 84 (2017) 1–8 contribute to livelihoods, the interaction between those contributions, CIFOR, 2007. PEN Technical Guidelines Version 4. Center for International Forestry fl Research, Bogor, Indonesia. http://www.cifor.org/pen/research-tools/the-pen- and the in uence exercised upon the forest contributions by ongoing technical-guidelines.html. landscape transformations, including transformations in governance Colfer, C.J.P., 2008. Human Health and Forests: Global Overview of Issues, Practice and arrangements (Agrawal et al., 2008). This is paramount, not only to Policy. Earthscan, London, UK. Cruz Garcia, G.S., Price, L.L., 2012. Weeds as important vegetables for farmers. Acta Soc. push the boundaries of current research, but also to provide policy- Bot. Pol. 81, 397–403. relevant information for more effective planning. The paradox of forest Dave, R., Tompkins, E.L., Schreckenberg, K., 2017;al., this issue. Forest ecosystem products is that despite increasing evidence on their importance for services derived by smallholder farmers in northwestern Madagascar: storm hazard both commercial and subsistence purposes, data gaps hinder the ability mitigation and participation in forest management. Forest Policy Econ (this issue). de Merode, E., Homewood, K., Cowlishaw, G., 2004. The value of bushmeat and other to monitor, regulate and manage them sustainably (Ingram and wild foods to rural households living in extreme poverty in Democratic Republic of Bongers, 2009). That is, when values are uncaptured in statistics and Congo. Biol. Conserv. 118 (5), 573–581. Dokken, T., Angelsen, A., 2015. Forest reliance across poverty groups in Tanzania. Ecol. policy, their contributions to livelihoods and national economies are – ffi Econ. 117, 203 211. di cult to evaluate and policies for economic development based on Dovie, D.B.K., Shackleton, C.M., Witkowski, E.T.F., 2002. Direct-use values of woodland their trade and consumption are consequently difficult to develop, resources consumed and traded in a South African village. Int. J. Sustain. Dev. World albeit essential to ensure sustainable management of forests. Ecol. 9, 269–283. Egelyng, H., Bosselmann, A.S., Warui, M., Maina, F., Mburu, J., Gyau, A., 2017;al., this issue. Origin products from African forests: a Kenyan pathway to prosperity and Acknowledgments green inclusive growth? Forest Policy Econ (this issue). Food and Agriculture Organization (FAO), 2008. Links between national forest programmes and poverty reduction strategies. In: Forestry Policy and Institutions. This research was supported by the Department for International Working Paper 22. Food and Agriculture Organisation of the United Nations, Rome. Development (DFID) (Accountable Grant Agreement PO #40099095), Food and Agriculture Organization (FAO), 2014. State of the World's Forests. Food and UK. We thank all authors contributing to this special issue and two Agriculture Organization of the United Nations (FAO), Rome. Food and Agriculture Organization (FAO), 2016. State of the World's Forests 2016. In: anonymous reviewers for useful comments on an earlier version of this Forests and Agriculture: Land-use Challenges and Opportunities, (Rome). editorial. Ferraro, P.J., Hanauer, M.M., Sims, K., 2011. Conditions associated with protected area success in conservation and poverty reduction. Proc. Natl Acad. Sci. USA 108, – References 13913 13918. Ferraro, P.J., Lawlor, K., Mullan, K., Pattanayak, S., 2012. Forest figures: ecosystem services valuation and policy evaluation in developing countries. Rev. Environ. Econ. Alexiades, M.N., Shanley, P., 2005. Forest Products, Livelihoods and Conservation. Case Policy 6 (1), 20–44. studies of Non-Timber Forest Product Systems. Vol. 3 - Latin America. CIFOR, Bogor. Fungo, R., Muyonga, J., Kaaya, A., Okia, C., Tieguhong, J.C., Baidu-Forson, J.J., 2015. Agrawal, A., Cashore, B., Hardin, R., Shepherd, G., Benson, C., Miller, D., 2013. Economic Nutrients and bioactive compounds content of Baillonella toxisperma, Trichoscypha Contributions of Forest. Background Paper Prepared for the United Nations Forum on abut and Pentaclethra macrophylla from Cameroon. Food Sci. Nutr. 3 (4), 292–301. Forests. http://www.un.org/esa/forests/pdf/session_documents/unff10/ Golden, C.D., Fernald, L.C.H., Brashares, J.S., Rasolofoniaina, B.J.R., Kremen, C., 2011. EcoContrForests.pdf, (Istanbul, 8-19 April). Benefits of wildlife consumption to child nutrition in a biodiversity hotspot. Proc. Agrawal, A., Chhatre, A., Hardin, R., 2008. Changing governance of the world's forests. Natl. Acad. Sci. 108 (49), 19653–19656. Science 320 (5882), 1460–1462. Gregersen, H.M., 1995. Etude FAO, Forêts. Food & Agriculture Org. Agrawal, A., Wollenberg, E., Persha, L., 2014. Governing agriculture-forest landscapes to Hickey, G.M., Poiliot, M., Smith-Hall, C., Wunder, S., Nielsen, M.R., 2016. Quantifying achieve climate change mitigation. Glob. Environ. Chang. 28, 270–280. the economic contribution of wild food harvests to rural livelihoods: a global- Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N., Bauch, S., et al., 2014. comparative analysis. Food Policy 62, 122–132. Environmental income and rural livelihoods: a global-comparative analysis. World Homma, A.K.O., 1996. Modernisation and technological dualism in the extractive Dev. 64, 12–28. economy in Amazonia. In: Pérez, M.R., Arnold, J.E.M. (Eds.), Current issues in Non- Angelsen, A., Larsen, H.O., Lund, J.F., Smith-Hall, C., Wunder, S., 2011. Measuring Timber Forest Products Research. Proceedings of the Workshop ‘Research on NTFP’, Livelihoods and Environmental Dependence: Methods for Research and Fieldwork. Hot Springs, Zimbabwe, 1995. Centre for International Forestry Research (CIFOR), CIFOR, Bogor, Indonesia. Bogor. Arnold, M., Periz, M.R., 2001. Can non-timber forest products match tropical forest Ickowitz, A., Powell, B., Salim, M.A., Sunderland, T.C.H., 2014. Dietary quality and tree conservation and development objectives? Ecol. Econ. 39, 437–447. cover in Africa. Glob. Environ. Chang. 24, 287–294. Arnold, M., Powell, B., Shanley, P., Sunderland, T.C.H., 2011. Editorial: forests, Ingram, V., Ewane, M., Ndumbe, L.N., Awono, A., 2017;al., this issue. Challenges to biodiversity and food security. Int. For. Rev. 13 (3, SI), 259–264. governing sustainable forest food: Irvingia spp. from southern Cameroon. Forest Babulo, B., Muys, B., Nega, F., Tollens, E., Nyssen, J., Deckers, J., Mathijs, E., 2009. The Policy Econ (this issue). economic contribution of forest resource use to rural livelihoods in Tigray, Northern Ingram, V., Ndoye, O., Iponga, D.M., Tieguhong, J.C., Nasi, R., 2012. Non-Timber Forest Ethiopia. Forest Policy Econ. 11 (2), 109–117. Products: Contribution to National Economy and Strategies for Sustainable Bakkegaard, R.K., Hogarth, N.J., Bong, I.W., Bosselmann, A.S., Wunder, S., 2017;al., this Management. In: Wasseige, C, Marcken, P, Bayol, N, Hiol, F, Mayaux, P, Desclée, B, issue. Measuring forest and wild product contributions to household welfare: testing Nasi, R, Billand, A, Defourny, P, Eba’a, R. (Eds.), The Forests of the Congo Basin: State a scalable household survey instrument in Indonesia. Forest Policy Econ (this issue). of the Forest 2010. Office des publications de l’Union Européenne, Luxembourg, pp. Bharucha, Z., Pretty, J., 2010. The roles and values of wild foods in agricultural systems. 137–154. Phil. Trans. R. Soc. B Biol. Sci. 365 (1554), 2913–2929. Ingram, V., Bongers, G., 2009. Valuation of Non-Timber Forest Product Chains in the Belcher, B., Ruiz-Pérez, M., Achdiawan, R., 2005. Global patterns and trends in the use Congo Basin: A Methodology for Valuation, CIFOR. FAO-CIFOR-SNV-World and management of commercial NTFPs: implications for livelihoods and Agroforestry Center-COMIFAC, Yaounde, Cameroon. conservation. World Dev. 33 (9), 1435–1452. Jiao, X., Smith-Hall, C., Theilade, I., 2015. Rural household incomes and land grabbing in Belcher, B., Kusters, K., 2004. Non-timber forest product commercialisation: development Cambodia. Land Use Policy 48, 317–328. and conservation lessons. In: Kusters, K., Belcher, B. (Eds.), Forest Products, Kanowski, P.J., McDermott, C.L., Cashore, B.W., 2011. Implementing REDD+: lessons Livelihoods and Conservation. Case studies of Non-Timber Forest Product Systems. from analysis of forest governance. Environ. Sci. Pol. 14, 111–117. Volume 1 - Asia. CIFOR, Bogor. Kehlenbeck, K., Asaah, E., Jamnadass, R., 2013. Diversity of indigenous fruit trees and Broegaard, R.B., Rasmussen, L.V., Dawson, N., Mertz, O., Vongvisouk, T., Grogan, K., their contribution to nutrition and livelihoods in sub-Saharan Africa: examples from 2017;al., this issue. Wild food collection and nutrition under commercial agriculture Kenya and Cameroon. In: Fanzo, J., Hunter, D., Borelli, T., Mattei, F. (Eds.), expansion in agriculture-forest landscapes. Forest Policy Econ (this issue). Diversifying Food and Diets: Using Agricultural Biodiversity to Improve Nutrition and Byg, A., Novo, P., Dinato, M., Moges, A., Tefera, T., Balana, B., Woldeamanuel, T., Black, Health Issues in Agricultural Biodiversity. Earthscan Routledge, pp. 257–269. H., 2017;al., this issue. Trees, soils, and warthogs – distribution of services and Lescuyer, G., Atyi, R.E., Nasi, R., Nembot, T.F., 2016. Valuing the Cameroonian Forest. disservices from reforestation areas in southern Ethiopia. Forest Policy Econ (this Int. For. Rev. 18 (s1), 1–3. issue). L'Roe, J., Naughton-Treves, L., 2014. Effects of a policy-induced income shock on forest- Campbell, B., Luckert, M. (Eds.), 2002. Uncovering the Hidden Harvest: Valuation dependent households in the Peruvian Amazon. Ecol. Econ. 97, 1–9. Methods for Woodland and Forest Resources. Earthscan Ltd.. L'Roe, J., Naughton-Treves, L., 2017L'Roe and Naughton-Treves, this issue. Forest edges Chen, Y.L., 2004. Song rong (Tricholoma matsutake), a valuable forest mushroom from in western Uganda: from refuge for the poor to zone of investment. Forest Policy Econ China: consumption, development, and sustainability. In: Kusters, K., Belcher, B. (this issue). (Eds.), Forest Products, Livelihoods and Conservation: Case Studies of Non-Timber Lund, J.F., Sungusia, E., Mabele, M.B., Scheba, A., 2017. Promising change, delivering Forest Product Systems. Jakarta, CIFOR. continuity: REDD+ as conservation fad. World Dev. 89, 124–139. Chhetri, B.B.K., Larsen, H.O., Smith-Hall, C., 2015. Environmental resources reduce Makumba, W., Janssen, B., Oenema, O., Akinnifesi, F.K., Mweta, D., Kwesiga, F., 2006. income inequality and the prevalence, severity and depth of poverty in rural Nepal. The long-term effects of a Gliricidia-maize intercropping system in southern Malawi, Environ. Dev. Sustain. 17, 513–530. on Gliricidia and maize yields, and soil properties. Agric. Ecosyst. Environ. 116, Christensen, H., 2002. Ethnobotany of the Iban and the Kelabit. In: Kuching and Aarhus, 85–92. Forest Department, Sarawak. NEPCon and University of Aarhus. Mbow, C., Van Noordwijk, M., Luedeling, E., Neufeldt, H., Minang, P.A., Kowero, G.,
7 L.V. Rasmussen et al. Forest Policy and Economics 84 (2017) 1–8
2014. Agroforestry solutions to address food security and climate change challenges Product Systems. Vol. 2 CIFOR, Africa, Bogor. in Africa. Curr. Opin. Environ. Sustain. 6, 61–67. Scoones, I., Melnyk, M., Pretty, J.N., 1992. The Hidden Harvest: Wild Foods and Meilby, H., Smith-Hall, C., Byg, A., Overgaard, Larsen H., Juul, Nielsen Ø., Puri, L., Agricultural Systems. A Literature Review and Annotated Bibliography. CAB & IIED, Rayamajhi, S., 2014. Are forest incomes sustainable? Firewood and timber extraction London. and productivity in community managed forests in Nepal. World Dev. 64, 113–S124. Senaratne, A., Abeygunawardena, P., Jayatilake, W., 2003. Changing role of NTFPs in Miller, D.C., Muñoz-Mora, J.C., Christiaensen, L., 2017;al., this issue. Prevalence, rural household economy. The case of Sinharaja World Heritage Site in Sri Lanka. economic contribution, and determinants of trees on farms across Sub-Saharan Environ. Manag. 32 (5), 559–571. Africa. Forest Policy Econ (this issue). Shackleton, S.E., Shackleton, C.M., Netshiluvhi, T.R., Geach, B.S., Ballance, A., Fairbanks, Nasi, R., Taber, A., Van Vliet, N., 2011. Empty forests, empty stomachs? Bushmeat and D.H.K., 2002. Use patterns and value of savanna resources in three rural villages in livelihoods in the Congo and Amazon basins. Int. For. Rev. 13 (3), 355–368. South Africa. Econ. Bot. 56 (2), 130–146. Newton, P., Miller, D., Byenkya, M., Agrawal, A., 2016. Who are forest-dependent people? Shackleton, C., Shackleton, S.E., Buiten, E., Bird, N., 2007. The importance of dry A taxonomy to aid livelihood and land use decision-making in forested regions. Land woodlands and forests in rural livelihoods and poverty alleviation in South Africa. Use Policy 57, 388–395. Forest Policy Econ. 9 (5), 558–577. Newton, P., Agrawal, A., Wollenberg, L., 2013. Enhancing the sustainability of Shackleton, C., Shackleton, S., Shanley, P., 2011. Non-timber forest products in the global commodity supply chains in tropical forest and agricultural landscapes. Glob. context. In: Tropical Forestry. Springer-Verlag, Berlin. Environ. Chang. 23 (6), 1761–1772. Shrestha, U.B., Bawa, K.S., 2014. Economic contribution of Chinese caterpillar fungus to Ojea, E., Loureiro, M.L., Alló, M., Barrio, M., 2016. Ecosystem services and REDD: the livelihoods of mountain communities in Nepal. Biol. Conserv. 177, 194–202. estimating the benefits of non-carbon services in worldwide forests. World Dev. 78, Singh, G., Rawat, G.S., Verma, D., 2010. Comparative study of fuelwood consumption by 246–261. villagers and seasonal “Dhaba owners” in the tourist affected regions of Garhwal Pangau-Adam, M., Noske, R., Muehlenberg, M., 2012. Wildmeat or Bushmeat? Himalaya, India. Energ Policy 38, 1895–1899. Subsistence hunting and commercial harvesting in Papua (West New Guinea), Sunderland, T., Harrison, S., Ndoye, O., 2004. Commercialisation of non-timber forest Indonesia. Hum. Ecol. 40, 611–621. products in Africa: history, context and prospects. In: Sunderland, T., Ndoye, O. Pascual, U., Muradian, R., Brander, L., Gómez-Baggethun, E., Martín-López, B., Verma, (Eds.), Forest Products, Livelihoods and Conservation. Case studies of Non-Timber M., Armsworth, P., Christie, M., Cornelissen, H., Eppink, F., Farley, J., 2010. The Forest Product Systems. Volume 2 - Africa. CIFOR, Bogor. Economics of Valuing Ecosystem Services and Biodiversity. In: Kumar, P. (Ed.), The Sunderland, T., Abdoulaye, R., Ahammad, R., Asaha, S., Baudron, F., Deakin, E., Duriaux, Economics of Ecosystems and Biodiversity Ecological and Economic Foundations. J.Y., Eddy, I., Foli, S., Gumbo, D., Khatun, K., Kondwani, M., Kshatriya, M., Leonald, Earthscan, London, pp. 183–256. L., Rowland, D., Stacey, N., Tomsha, S., Yang, K., Gergel, S., Vianen, J.V., 2017;al., Pingali, P., 2015. Agricultural policy and nutrition outcomes – getting beyond the this issue. A methodological approach for assessing cross-site landscape change: preoccupation with staple grains. Food Sec. 7 (3), 583–591. understanding socio-ecological systems. Forest Policy Econ (this issue). Pouliot, M., Treue, T.J., 2013. Rural people's reliance on forests and the non-forest Sunderlin, W.D., Angelsen, A., Belcher, B., Burgers, P., Nasi, R., Santoso, L., et al., 2005. environment in West Africa: evidence from Ghana and Burkina Faso. World Dev. 43, Livelihoods, forests, and conservation in developing countries. An overview. World 180–193. Dev. 33 (9), 1383–1402. Pouliot, M., 2012. Contribution of “women's gold” to West African livelihoods: the case of Sunderlin, W.D., 2006. Poverty alleviation through community forestry in Cambodia, shea (Vitellaria paradoxa) in Burkina Faso. Econ. Bot. 66 (3), 237–248. Laos, and Vietnam: an assessment of the potential. Forest Policy Econ. 8 (4), Powell, B., Iockowitz, A., Mcmullin, S., Jamnadass, R., Padoch, C., Pinedo-Vasque, M., 386–396. Sunderland, T., 2013a. The role of forests, trees and wild biodiversity for nutrition- Twine, W., Moshe, D., Netshiluvhi, T.R., Shipugu, V., 2003. Consumption and direct-use sensitive food systems and landscapes. In: Expert Background Paper for the values of savanna bio-resources used by rural households in Mametja, a semi-arid International Conference on Nutrition (ICN 2). FAO, Rome. area of Limpopo Province, South Africa. S. Afr. J. Sci. 99, 467–473. Powell, B., Maundu, P., Kuhnlein, H.V., Johns, T., 2013b. Wild foods from farm and forest Van Vliet, N., Quiceno-Mesa, M.P., Cruz-Antia, D., Tellez, L., Martins, C., Haiden, E., in the east Usambara Mountains, Tanzania. Ecol. Food Nutr. 52, 451–478. et al., 2015. From fish and bushmeat to chicken Nuggets: the nutrition transition in a Powell, B., Thilsted, S., Ickowitz, A., Termote, C., Sunderland, T., Herforth, A., 2015. continuum from rural to urban settings in the Colombian Amazon region. Ethnobiol. Improving diets with wild and cultivated biodiversity from across the landscape. Conserv. 4 (6), 1–12. Food Sec. 7 (3), 535–554. Vedeld, P., Angelsen, A., Bojo, J., Sjaastad, E., Berg, G., 2007. Forest environmental PROFOR, 2008. Poverty and Forest Linkages: A Synthesis and Six Case Studies. Report incomes and the rural poor. Forest Policy Econ. 9 (7), 869–879. The Program on Forests (PROFOR) at the World Bank, Washington, DC. Vira, B., Mansourian, S., Wildburger, C. (Eds.), 2015. Forests, trees and landscapes for Rasmussen, L.V., Mertz, O., Christensen, A.E., Danielsen, F., Dawson, N., Xaydongvanh, food security and nutrition: a global assessment report. IUFRO World Series Volume P., 2016. A combination of methods needed to assess the actual use of provisioning 33 International Union of Forest Research Organizations, Vienna. Available from: ecosystem services. Ecosyst. Serv. 17, 75–86. http://www.iufro.org/science/gfep/. Rasmussen, L.V., Bierbaum, R., Oldekop, J.A., Agrawal, A., 2017. Bridging the Wahlen, C.B., 2017Wahlen, this issue. Opportunities for making the invisible visible: practitioner-researcher divide: indicators to track environmental, economic, and towards an improved understanding of the economic contributions of NTFPs. Forest sociocultural sustainability of agricultural commodity production. Glob. Environ. Policy Econ (this issue). Chang. 42, 33–46. Walelign, S.Z., Charlery, L., Smith-Hall, C., Chhetri, B.B.K., Larsen, H.O., 2016. Reed, J., van Vianen, J., Foli, S., Clendenning, J., Yang, K., MacDonald, M., Petrokofsky, Environmental income improves household-level poverty assessments and dynamics. G., Padoch, C., Sunderland, T., 2017;al., this issue. Trees for life: the ecosystem Forest Policy Econ. 71, 23–35. service contribution of tress to food production and livelihoods in the tropics. Forest World Health Organization, 2015. Connecting Global Priorities: Biodiversity and Human Policy Econ (this issue). Health. A State of Knowledge Review, Geneva, Switzerland. Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.-L., Sheil, D., Meijaard, E., Venter, M., Wilkie, D.S., Carpenter, J.F., 1999. Bushmeat hunting in the Congo Basin: an assessment Boedhihartono, A.K., Day, M., Garcia, C., Van Oosten, C., Buck, L.E., 2013. Ten of impacts and options for mitigation. Biodivers. Conserv. 8, 927–955. principles for a landscape approach to reconciling agriculture, conservation, and Wunder, S., Angelsen, A., Belcher, B., 2014a. Forests, livelihoods, and conservation: other competing land uses. Proc. Natl. Acad. Sci. 110 (21), 8349–8356. broadening the empirical base. World Dev. 64, 1–11. Schreckenberg, K., 2004. The contribution of Shea Butter (Vitellaria paradoxa C.F. Wunder, S., Börner, J., Shively, G., Wyman, M., 2014b. Safety nets, gap filling and forests: Gaertner) to local livelihoods in Benin. In: Sunderland, T.C.H., Ndoye, O. (Eds.), a global-comparative perspective. World Dev. 64, 29–42. Forest Products, Livelihoods and Conservation: Case-studies of Non-Timber Forest
8 Forest Policy and Economics 84 (2017) 9–10
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Guest Editorial ☆ An editorial from the handling editor MARK
As the handling editor of this special issue, I had the privilege to read every one of the articles in this special issue and want to congratulate the guest editors and the authors for a job well done. I would also like to offer the following observations. First, development and conservation projects for whom? The article by Byg et al. (in this issue) demonstrates that conservation programs intended to promote biodiversity, enhance soil fertility, protect watershed and prevent soil erosions often cause hardships for the rural poor by limiting their ability to collect from the forest for their livelihoods. These projects do further damages to their livelihoods through losses causedby wildlife such as warthogs attracted to the conservation area. Not surprisingly, conservation programs/areas are frequently despised by the rural poor. At the same time, the elites, including central government officials in the recipient countries and NGOs from developed countries bath in the glories of establishing such conservation areas. Without the support of the rural poor, would these conservation projects ultimately survive when the foreign assistance money dries up? Instead of imposing the preferences and values of the elites from the top down, were there bottom-up attempts to formulate projects based on what the rural residents really want in the first place? Shouldn't the declaration of “of the people, by the people, and for the people” by President Lincoln be honored and followed? One of the key themes of the papers is the importance of forest's contributions to the wellbeing of rural residents. Given that the title of this special issue is Forest, Food, and Livelihoods, it is not surprising that the importance and contributions of timber products are not covered in this issue. However, as Rasmussen et al. (in this issue) pointed out the official monetary contributions of forest to developing world exceed $250 billion. The unofficial contributions should, most likely, be even larger. Surely, income from timber products contributes to the livelihoods of rural residents. These articles seem to imply that the rural residents do not benefit significantly from the monetary contributions of timber and other marketed products. The findings of Ingram et al. (in this issue) shows that while bush mango collectors earned 31% of their income from bush mango collection, the exporters and retailers earned 48% and 57% respectively. Economists have long recognized that the ones closest to the market always take the biggest cut along the supply chain. They also are the least affected by the shrinking supply of a collected item. If bush mango is no longer available, they will simply move on to something else. What they get or don't get matter less as far as the development and conservation projects are concerned. As the economies of developing countries become more market oriented, what matters is which groups would realize the benefits now accounted for by the subsistence contribution of forests most? Would rural poor be better off or worse off as a result? If it is the latter, how development and conservation projects justify themselves? In the agriculture-forestry landscape in transition, something intriguing is happening. Most, if not all, of the case studies imply that the forests are owned by the governments or community or commonly owned or managed. On the other hand, when the forests are converted to farmlands, there is, at least, an implicit understanding that the farmland belongs to the farmer. Such a dichotomous system is simply non-sustainable. The incentive to convert common forests to private farmlands is simply too compelling. Not surprisingly, forest acreages decline over time in most recipient countries. Those remaining forests survive either because they are too remote or their site productivity is too poor. At the same time, collecting food, firewood, and etc. from these remaining forests almost surely leads to the tragedy of the commons (Hardin, 1968) as demonstrated by Ingram et al. (in this issue). That is to say under common ownership, inevitably the quality of the forest declined, productivity decimated, and resource depleted. Given a priori this well-known consequence, it is therefore puzzling that one rarely sees the NGOs and development agencies from the developed countries attempt to promote some forms of private ownership. It must be pointed out that in these donor-developed countries private ownership is the dominant type of property ownership and it is this very type of ownership which creates the wealth that enables the development and conservation projects in the first place. Do these development and conservation agencies and NGOs not believe in private ownerships or is it implied that they do not believe the rural poor are equipped to handle private ownership despite the fact that they handle their private farms just fine? Why can't the subsistence contributions be internalized through private ownership? The papers by Bakkegaard et al. (in this issue) and Wahlen (in this issue) again demonstrated that the forests are under-valued and under- appreciated by the government to the detriment of the poor rural residents when they are not privatized. But, once the users of the forests become the owners of the forests, this valuation problem is effectively internalized as clearly demonstrated by Miller et al. (in this issue) and Reed et al. (in this issue) with the prevalent presence of trees on farms. Apparently, farmers not only protect trees on their farms, they also actively plant and cultivate them. Without a doubt, these farmers fully appreciate the value of the trees on their farms. When a forest or a farm with trees is being sold, do we not believe the seller would fully price all the things he or she gets from the forest so as not to sell for less than what they are worth? Conversely, the
☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” published at the journal Forest Policy and Economics 84C, 2017. http://dx.doi.org/10.1016/j.forpol.2017.08.001
Available online 19 August 2017 1389-9341/ © 2017 Elsevier B.V. All rights reserved. Guest Editorial Forest Policy and Economics 84 (2017) 9–10 buyer must also incorporate all these items in the purchase price if they want to acquire the property. The product origin attempts reported by Egelyng et al. (in this isssue) to increase the income of rural residents are fascinating. The issues here are whether the Mwingi honey and Kakamega silk are, in what way, better quality products. Geographic indication (GI) is necessary but not sufficient as a marketing tool. In addition, are they produced at a lower cost than the competitors but not necessarily sold for less; with faster delivery and superior services? In the modern market economy, consistent product quality and reliable supply quantity are a must. Last but not the least, can they fend off the knock-offs? Indeed, as Egelyng et al. (in this issue) point out capacity building at local level, development of quality standard bodies, and a wider institutional environment supporting GI products and markets are needed. As a practical matter, would it be wise to implement the idea first at the regional markets to test the idea and work out all the kinks mentioned above before pursing national and eventually international markets? Lastly, as the economies of developing countries grow and new opportunities arise for rural residents, the opportunity cost of collecting wild food and items rises over time. It is inevitable that such collections will dwindle. Purchased food and other things will replace wild ones, for better or for worse. First, in what sense is wild collected food necessarily better or is the purchased food just more convenient. Second, understanding how rural residents allocate their time over time may also become important. It is surprising that after over half a century of development and conservation assistance efforts, few longitudinal studies exist in the literature. Given the rapid changes in developing countries and agriculture – forest landscape, a longitudinal record of how things have changed provides far more insights than a one-time snap shot. In this special issue only one article by L'Roe and Naughton-Treves (in this issue) presents such a study. In the future, we might need to be more selective in accepting manuscripts that provide only a one time look at a particular issue.
References
Bakkegaard, R.K., Hogarth, N.J., Bong, I.W., Bosselmann, A.S., Wunder, S., 2017. Measuring forest and wild product contributions to household welfare: testing a scalable household survey instrument in Indonesia. Forest Policy Econ (in this issue). Byg, A., Novo, P., Dinato, M., Moges, A., Tefera, T., Balana, B., Woldeamanuel, T., Black, H., 2017. This issue. Trees, soils, and warthogs — distribution of services and disservices from reforestation areas in southern Ethiopia. Forest Policy Econ (in this issue). Egelyng, H., Bosselmann, A.S., Warui, M., Maina, F., Mburu, J., Gyau, A., 2017. This issue. Origin products from African forests: a Kenyan pathway to prosperity and green inclusive growth? Forest Policy Econ (in this issue). Hardin, G., 1968. The tragedy of the commons. Science 162 (3859), 1243–1248. Ingram, V., Ewane, M., Ndumbe, L.N., Awono, A., 2017. Challenges to governing sustainable forest food: Irvingia spp. from southern Cameroon. Forest Policy Econ (in this issue). L'Roe, J., Naughton-Treves, L., 2017. Forest edges in western Uganda: from refuge for the poor to zone of investment. Forest Policy Econ (in this issue). Miller, D.C., Muñoz-Mora, J.C., Christiaensen, L., 2017. Prevalence, economic contribution, and determinants of trees on farms across Sub-Saharan Africa. Forest Policy Econ (in this issue). Rasmussen, L.V., Watkins, C., Agrawal, A., 2017. Forest contribution in changing agriculture-forest landscapes. Forest Policy Econ (in this issue). Reed, J., van Vianen, J., Foli, S., Clendenning, J., Yang, K., MacDonald, M., Petrokofsky, G., Padoch, C., Sunderland, T., 2017. Trees for life: the ecosystem service contribution of tress to food production and livelihoods in the tropics. Forest Policy Econ (in this issue). Wahlen, C.B., 2017. Opportunities for making the invisible visible: towards an improved understanding of the economic contributions of NTFPs. Forest Policy Econ (in this issue).
Sun Joseph Chang School of Renewable Natural Resources, Louisiana State University Agricultural Center, Baton Rouge, LA 70803, United States
10 Forest Policy and Economics 84 (2017) 11–19
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Opportunities for making the invisible visible: Towards an improved ☆ MARK understanding of the economic contributions of NTFPs
Catherine Benson Wahlén1
Uganda
ARTICLE INFO ABSTRACT
Keywords: Forests around the world remain under-valued because governments, policymakers, and other key actors do not NTFPs consider the global or national contributions of non-timber forest products (NTFPs) when estimating the value of Forests forests. Where available, existing estimates suggest the non-cash benefits of forests are substantial, in many cases Forest valuation three to ten times higher than those for which systematic national and global data are collected. Part of this Non-cash value under-valuation stems from a general focus by governments on forest resources that are commercially valued as SDGs well as from government failure to include estimates of NTFP contributions in their national accounts. Beyond NWFPs these reasons, however, lay methodological challenges in measurement techniques and comparability across studies, countries, and regions, both of which result in limited data on NTFPs. This article reviews NTFP studies at the global and national levels to provide estimates of the non-cash contributions of NTFPs and to shed light on challenges related to the absence of systematic, reliable data on the economic contributions of forests. The article then considers the implications for forest governance, management, and policy, arguing that the Sustainable Development Goals (SDGs) offer an opportunity to increase attention on the non-cash contributions of forests and turn this invisible contribution into a visible one.
1. Introduction households (Shackleton and Pandey, 2014). Given the scarcity of systematic, reliable data on these benefits, non- The formal forestry sector makes a significant contribution to gross cash contributions of forests are typically unaccounted for in national domestic product (GDP), with an estimated value of US$468 billion and international economic planning. Further, it is difficult to dis- (Agrawal et al., 2013). The Food and Agriculture Organization of the aggregate the cash and non-cash uses of forests at the household and UN (FAO) estimates that forest industries contribute nearly 1% of the community levels (Agrawal et al., 2013). For instance, many households global GDP and provide formal employment to 0.4% of the global labor that use NTFPs for subsistence may also sell NTFPs for cash income. force (FAO, 2012). This figure is a significant under-estimation of the Another reason forests are typically unaccounted for is that governments true value of forests because it excludes the vast informal and un- do not view them as drivers of economic growth because many forest reported contributions of forests to international and national econo- goods and services are not traded in formal markets (PROFOR, 2008). mies, including consumption values of non-timber forest products In cases where studies do consider the value of NTFPs, such studies are (NTFPs) and non-wood forest products (NWFPs), ecosystem service more likely to consider NTFPs that are exported, an approach that hinders values, and cultural, medicinal, and tourism benefits.2 The value of a fuller understanding of the broader picture of domestic trade and con- forest products consumed at home, which are not recorded, are often sumption, given that a large percentage of collected NTFPs are assumed to more than two to three times greater than the formal, cash-based sector be consumed rather than sold. NTFPs that are primarily traded in local (Agrawal et al., 2013). At the household level, estimates suggest that markets are also likely to be overlooked in estimates on the value of NTFPs the contribution of NTFPs ranges from 10 to 60% of household income, (Shackleton et al., 2007, 2008). Consequently, even figures that do esti- with the variation attributable to different national, economic, and mate the value of NTFPs tend to only consider their cash value and ex- cultural contexts as well as differences between poor and rich clude their non-cash value, resulting in a significant under-valuation.
☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” published at the journal Forest Policy and Economics 84C, 2017. E-mail address: [email protected]. 1 Independent Consultant. 2 The term NTFPs, non-timber forest products, refers to everything taken from the forest for household use or sale and includes fuelwood, light poles and other house construction material, and fiber. FAO uses the term NWFPs, non-wood forest products, so that all wood products, such as fuelwood and timber, can be grouped together. The term NTFPs is the most commonly used in the literature, in part because it allows for differentiating forest products into two distinct forest user groups: local people and loggers. http://dx.doi.org/10.1016/j.forpol.2017.04.006 Received 16 May 2016; Received in revised form 10 April 2017; Accepted 11 April 2017 Available online 11 May 2017 1389-9341/ © 2017 Elsevier B.V. All rights reserved. C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19
Some of the most important contributions of forests to economic understanding of forestry is critical for governments and policymakers development are in the form of poverty alleviation and the benefits to design appropriate policies and interventions. forests provide to vulnerable forest-dependent populations.3 In rural Following this very brief introduction, this paper reviews a few key areas, for instance, the benefits from forests for livelihood and sub- studies of NTFPs, with a focus on how to better understand and capture sistence benefits are more critical for local populations than the value of the value of forest benefits. The third section highlights the main data timber. An extensive body of literature illustrates the contribution of challenges related to estimating the contribution of NTFPs. The fourth forest products as social safety nets that can provide food security and section considers how to make the role of forests more visible to gov- cash income in times of crisis, scarcity, or environmental and economic ernments and policymakers, suggesting the current emphasis on SDG shocks (e.g., Angelsen and Wunder, 2003; Shackleton and Shackleton, implementation offers one avenue for increasing forests' prominence in 2004; Sunderland et al., 2004; Kar and Jacobson, 2012; Wunder et al., the development agenda. The paper concludes with recommendations 2014). A related literature considers how forests can also provide a on how to strengthen the recognition and assessment of the full range of pathway out of poverty (Angelsen and Wunder, 2003; Shackleton et al., forest values in national and international economies as well as ways to 2008; Babulo et al., 2009; Shepherd et al., 2013). Others underscore the improve capacities for data collection and analysis. importance of forest and tree-based systems for food and nutrition (Powell et al., 2011; Foli et al., 2014; Ickowitz et al., 2014; Vira et al., 2015a; Vira et al., 2015b; Rowland et al., 2016). 2. Studies on the valuation of NTFPs Studies on the non-cash contributions of forests often focus on the role of forests in meeting wood energy needs, which is hardly surprising This section reviews the major data collection efforts at the global given that 33% of the global population relies on woodfuel as their level, such as the FAO's Global Forest Resources Assessment (FRA) and primary source of energy (FAO, 2014a). In lower-middle-income and the State of the World's Forests (SOFO) and studies by the Center for low-income countries, this figure is even higher, with 66 and 93% of International Forestry Research (CIFOR) and the Poverty Environment populations, respectively, relying on woodfuel (FAO, 2015). Although Network (PEN), to assess the current state of knowledge on the non- data on the consumption of wood products is important in illustrating cash contributions of NTFPs. It then considers national level case stu- the importance of charcoal, woodfuel, and similar products for house- dies that showcase a few promising approaches for capturing the con- hold cooking and heating needs, this data provides only a limited pic- tributions of NTFPs at the national level. ture of the consumption benefits of forests. If governments and pol- At the global level, the 2015 FRA ‘Economics and livelihoods’ sec- icymakers only use data on wood product consumption, they are likely tion presents national data on how much money governments collect to continue to underestimate the value of forests to their national from and spend on forests, employment in the forest sector, and the economies. contribution of forestry to GDP, as shown through the value-added More generally, given the diversity of NTFPs used by households measurement (FAO, 2015). Out of 234 countries and territories, 149, or 4 and traded both informally and formally, any study that focuses pri- 64%, provide data on value added from forestry. India ranks the marily on one type of NTFP will under-estimate the contribution of highest, with US$25,388.81 in total value added from forestry, fol- NTFPs and forests to local, national, regional, and international lowed by Brazil, with US$10,258, and Chile, with US$5229.26. One economies. For example, even studies on commercially value products challenge in comparing data across countries is uncertainty and varia- like mushrooms demonstrate both high-value of these NTFPs for bility in the reporting year; this data is all coded 9999, indicating that fi households (Fu et al., 2009a; Fu et al., 2009b; Huber et al., 2010) and the data year is not available (FAO, 2015). Further, the gures in this the simultaneous collection of fruits, spices, and other vegetables from section present an extremely limited perspective of the economic and the forests (Fu et al., 2009b). livelihood contributions of forests, resulting in an under-estimation of Another growing set of literature seeks to document the value of forests' contributions to economics and livelihoods. forest ecosystem services and promotes national accounting of eco- The FRA uses data from the Collaborative Forests Resource system services, also known as natural capital accounting (NCA). NCA Questionnaire (CFRQ), which compiles data provided by countries. This recognizes that measures of a country's economic performance using data collection is both a strength and a weakness. By using country indicators such as GDP can be misleading because they fail to in- data, the CFRQ relies on national agencies to properly implement fi corporate the contribution of natural capital, including forests, in na- methodologies and accurately report on ndings. One advantage of this tional accounts. For instance, timber resources are included in national approach is that it allows for collection of a large amount of data by fi accounts but other forest services and NTFPs are excluded from such national correspondents, which promotes the use of common de ni- accounts. Consequently, initiatives such as the World Bank's Wealth tions, the sharing of forests statistics, and reduction of data collection ff Accounting and the Valuation of Ecosystem Services (WAVES) and the e orts. Conversely, in countries where data collection and analysis UN Statistical Commission of the System for Environmental and capacity is weak, data reliability may be lower. Economic Accounts (SEEA) aim to better capture these values and re- Although the FRA itself does not recognize or analyze the cash or flect the contribution of forests and other environmental services and non-cash contributions of NTFPs or NWFPs to national economies, the benefits (WAVES, n.d.; UN SEEA, n.d.; PROFOR, 2015). National ac- supplementary dataset includes a variable on NWFP value. Seventy-four counting that incorporates data on forest products related to environ- countries provided data on this variable. China (US$10,494,093), India mental, cultural, recreational, and tourism services and construction (US$2,496,123), the US (US$2,254,514), and the Republic of Korea (US material, fodder, food, fuel, and medicinal values could facilitate better $10,42,914) report the highest NWFP values (FAO, 2015). documentation of the full value of forests. This literature can also In 2014, FAO released an updated study on the contribution of the contribute to an improved understanding of the contributions of forests forestry sector to national economies (FAO, 2014b), which considers to sectors such as agriculture, health, tourism, and water. Such linkages the global, regional, and country levels. The study only analyzes three are increasingly important as governments implement the Sustainable variables: employment in the forestry sector; value-added, or the sec- Development Goals (SDGs), which will require a transformation away tor's contribution to GDP; and export earnings. This analysis does not from silo policies and interventions towards coherent, integrated in- include the value that the non-cash sector makes to national economies terventions (e.g., see Van Vianen et al., 2015). Such an improved and therefore represents a gross underestimation of the true
4 The value for four of these countries and territories (Faroe Islands, Gibraltar, 3 For typologies forest dependence, see Byron and Arnold (1997) and Newton et al. Greenland, and Malta) is reported as 0.Consequently, it may be more accurate to say that (2016). 145 countries and territories reported data.
12 C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19 contribution of the forestry sector to national economies by only con- CIFOR, and considers the role of forests in poverty alleviation and how sidering the formal forestry sector and direct economic impacts. The to enhance this role through improved policy formulation and im- authors briefly acknowledge this limitation, observing that NTFP ac- plementation. PEN brings together a network of researchers to explore tivities “are likely to be significant in many developing countries,” but the forest-poverty link among different groups, forest environments, saying, “it is hard to observe and measure these activities at a national institutional, and market contexts, including data collection on the role or regional scale” (FAO, 2014b: 3). of forests in consumption. FAO's SOFO 2014 highlights the non-cash contributions of forests Using PEN data from 7978 households in 24 developing countries, for energy, food, health, and shelter, suggesting that billions of people Angelsen et al. (2014) suggest forest income ranges from 6% to 44% of rely on forest outputs to meet food, energy, and shelter needs. These total household income, with an average annual household income of estimates focus on the number of people who use forest products for US$440 at the study's sampled sites, a figure that represents 22.2% of these services, rather than on the values of these services to households total household income. In addition, they report that non-forest en- or national economies. For instance, SOFO 2014 calculates the total vironmental income contributes an additional US$86. They also found consumption and per capita consumption of food from forests for both that 28% of household income comes from environmental sources, out animal-based and plant-based NWFPs as well as the contribution of of which 77% is from natural forests. edible NWFPs to food supply. Similarly, on wood fuel, SOFO presents Angelsen et al. (2014) conclude that the failure to include the data on the contribution of wood energy to total primary energy supply, contribution of forests and other environmental areas and products the numbers of people using wood energy, and the proportion of gives a “misleading picture of rural livelihoods and … an inadequate households cooking with woodfuel (FAO, 2014a). Although this in- basis for policy design.” This dataset suggests the importance of forest formation is helpful in terms of understanding benefits provided by and environmental income for poverty reduction and rural households. forests, such information merely presents an illustration of the numbers However, the PEN dataset focuses on estimates of the total and per- of people who rely on forest products for consumption benefits and does centage contribution to household incomes, rather than presenting an not contribute to generating a value for the non-cash benefits of forests overall figure that can be used to estimate forests' contribution to na- that governments and policymakers could use to make a case for in- tional economies. Consequently, the data is not presented in numbers creased investments in sustainable forest management (SFM). SOFO that are accessible for policymakers or government officials who wish (2014) itself recognizes that clear evidence on the “role of forests in to demonstrate the contribution of forests to national economies. providing employment, energy, food, and a wide range of other goods Other global comparative efforts include the International Forestry and ecosystem services …. has been lacking” and that such evidence is Resources and Institutions (IFRI) and Forests & Livelihoods: needed to inform policy decisions and to reflect the socioeconomic Assessment, Research, Engagement (FLARE), a Community of Practice benefits of forests in policy (FAO, 2014a: xi). on forests and livelihoods. Although both FRA and SOFO rightly argue that data on NTFP uses At the regional level, several bodies collect information about the and values is difficult to obtain, FAO and its partners could use com- non-cash benefits of forests, including the Amazon Cooperation Treaty plimentary data from the vast body of published literature on NTFPs to Organization Agenda, the Economic Community of West African States cross-check their own figures. Such studies could also provide illus- and the related Forest Convergence Plan for West Africa, FOREST trative case studies that show how data on economics and livelihoods is EUROPE, the International Tropical Timber Organization (ITTO), the an under-estimate by comparing data reported to FAO with other Montréal Process for Temperate and Boreal Forests, and the South published studies. Jamnadass et al. (2015) point out, for instance, that African Development Community (SADC) Protocol on Forestry. On the the FRA's estimates of the value of woodfuel in Africa is US$1.4 billion whole, however, this information is patchy and not considered reliable annually while the World Bank places the value of sub-Saharan Africa's enough to present an accurate picture of the non-cash contributions of charcoal industry at US$8 billion annually. forests. For instance, ITTO and the Montréal Process collect data on the CIFOR carried out three major studies on the contributions of NTFPs number of people who depend on forests for subsistence and the area of to household incomes (Belcher and Kusters, 2004; Sunderland et al., forests used for subsistence, but very few countries provide this in- 2004; and Alexiades and Shanley, 2005). The global comparison un- formation. Such information provides information about forest area and dertaken as part of the CIFOR study found clear differences in reliance use but does not help to illustrate the value of forests benefits for on NTFPs among African, Asian, and Latin American forest products people. Consequently, such data is unlikely to be useful in demon- (Sunderland et al., 2004). In Africa, 14 out of the 17 products studied strating the value of forests for policymakers. Countries also use a range contribute less than 50% of household income and 9 of the 17 con- of methodologies to calculate the number of forest dependent people, tribute less than 25% of household income. Only three of the NTFPs making the validity and comparability of the data questionable (FAO, contributed more than 70% of household income, from craft markets 2014a). Another challenge with this measurement is the lack of an such as woodcarvings, rattan baskets, furniture, and chew sticks, which agreed definition of forest dependence, such as on the level and type of have a large, organized market. Sunderland et al. (2004) identify these dependence (FAO, 2014a). products as an exception, concluding that most African household in- At the national level, a few studies have estimated the value of come from NTFPs represents a coping strategy. For instance, they found NTFPs. Most studies focus on the percentage contribution of NTFPs or that even small NTFP contributions are important for households NWFPs to household income (e.g., Mahaptra and Shackleton 2012; Kar during times of need or emergencies. In Latin America and Asia, and Jacobson, 2012; Hussain and Badola, 2010; Babulo et al., 2009; households derived a higher level of income from the production of Belcher et al., 2005; Mahapatra et al., 2005; Dovie, 2003; and NTFPs. Gunatilake et al., 1993), the total household income (e.g., Gram et al. One major limitation in the CIFOR studies in terms of generating 2001), or, in a few cases, the cash equivalent of NTFP consumption systematic data on the subsistence and economic value of forest pro- (e.g., Delang, 2006; Appasamy, 1993). These studies illustrate the value ducts is that the included studies do not all examine non-cash value. A of NTFPs and/or NWFPs to household economies, suggesting the im- second limitation is that the studies use the product as their unit of portance of forests for livelihoods. However, because the studies do not analysis rather than the household, making comparative analysis more provide estimates of the overall economic contribution of NTFP con- difficult. These studies are valuable in presenting case study evidence sumption, it is not possible to compare non-cash economic values with on the role of NTFPs for the poor, but they do not illustrate the roster of cash values for products such as timber, which is often necessary for NTFPs used for livelihood support or help in compiling global or na- policymakers to make the case for NTFPs' contribution to national tional estimates on the non-cash value of NTFPs. economies. Another global comparative effort is PEN, which is coordinated by In Swaziland, Dlamini and Geldenhuys (2009) found a median value
13 C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19 of NTFPs of US$49.38 million, including a contribution of $32.1 incomes and forest contributions. At the global level, this review un- million from medicinal plants and $13.5 million from fuelwood. Using derscores a need to focus on the values of NTFPs, rather than on the a national accounting approach, they suggest that the contribution of number or percentage of people who depend on them to generate re- natural forests, including NTFPs, was equal to 2.2% of the country's levant data for policymakers. This review further highlights the im- total GDP and that NTFPs represent 439% of the contribution of for- portance of selecting the right variables for analysis. Data included in estry reported in national accounts for 2000. Their findings support the FRA's supplementary material is a promising step but it must be the point that the non-economic contribution of forests are highly promoted more broadly and highlighted, with textual commentary in undervalued (Dlamini and Geldenhuys, 2009; Dlamini and future FRA reports if it is to be of significant value for governments and Geldenhuys, 2011). policymakers, a point further discussed in the recommendations. At the Other national level studies compare the value of NTFPs to cash national level, examples support the argument that the value of non- benefits from timber. In India, one study estimated the value of NTFPs cash benefits of forests is much greater than recognized by governments at US$27 billion annually, a figure higher than the estimated US$17 or included in national and international figures and reports. The fol- billion from timber products (ITTO 2007 in Shackleton and Pandy lowing section highlights some of the reasons these types of national 2014). A study from Turkey found that the export value of NWFPs, studies are not more common. including pine nuts, thyme, laurel leaves, mushrooms, and carob beans, exceeded the value of roundwood exports (Karayilmazlar, 2005). 3. NTFP data challenges and implications NWFPs contribute 98% of Turkey's foreign trade of forest products, for a total US$100 million annually, compared to roundwood exports, Despite the importance of NTFPs to household economies around which contributed only 2%. In Bangladesh, overall household income the world and their contribution to national GDP, data on the non-cash from NTFPs is significantly higher than income from timber, poles, or contributions of forests tends to be ad hoc and case-study based or only firewood. However, Kar and Jacobson (2012) point out that the reason available in country and organizational reports, resulting in unreliable NTFPs contribute a higher percentage of household income in Bangla- and un-aggregated data at global and national scales (Barik and Mishra, desh may be because the forest is heavily degraded, a point that un- 2008; Delang, 2006; Agrawal et al., 2013; Shackleton and Pandey, derscores the importance of contextual information when considering 2014). The absence of reliable information and aggregated data on the the value of various products. They conclude that attention to site- economic contributions of NTFPs and their value to national economies specific factors is critical when generalizing findings on the value of hinders a better understanding of the contributions of the forest sector NTFPs. In the Czech Republic, approximately 75% of surveyed house- and represents a fundamental challenge in convincing policymakers of holds reported collecting NTFPs, with an average annual household the benefits of NTFPs. This section highlights a few of the key chal- collection of 10.6 kg of mushrooms and berries, most of which is con- lenges in developing more systematic data collection on the non-cash sumed rather than sold (Sisak et al., 2016). Sisak et al. (2016) explain economic contributions of forests. that the monetary value of this NTFP collection is equivalent to 3100 First, the majority of NTFPs are used for subsistence purposes and million Czech Crowns (over 121 million US dollars), or one-sixth of the therefore never enter the market. These NTFPs remain un-quantified or value of the country's timber harvest in an average year. Further, Sisak unrecorded as part of regulation, market records, or taxation (Agrawal et al. (2016) stress that the actual value of NTFPs in the Czech Republic et al., 2013; Vira et al., 2015b). In other words, household consumption is even higher because their study only includes mushrooms and five of NTFPs is “invisible” to outsiders. The seasonality of some products types of berries and excludes other NTFPs like medicinal and orna- further increases their invisible nature (Guariguata et al., 2011; mental plants. Shackleton and Pandey, 2014; Bakkegaard et al., 2016). For instance, if Findings from Uganda also suggest that forests' contribution to na- a survey occurs while a particular species is not being harvested or tional GDP is much greater than current assessments indicate. Shepherd consumed, a researcher could miss the consumption of such a product et al. (2013) observe that the commonly cited figure for forests' con- completely. Likewise, a survey designed to capture that species might tribution to Uganda's GDP is 6.1% but explain this figure does not in- miss another species with a different seasonal maturation. clude domestic consumption of forest products, which they estimate are In some cases, subsistence users themselves may resist the quanti- worth an additional US$2.9 billion. fication or recording of their NTFP benefits. For instance, users may be This analysis suggests the value of forest products for rural reluctant to share information with outsiders about their NTFP use for Ugandans is approximately $4.019 billion, which translates into ap- fear of having their rights contested or taken away or because they do proximately US$730 annually per household. Out of this amount, US not want to interact with government bodies (Agrawal et al., 2013; $2.882 billion, or 72%, is from consumption and use of forest products Bakkegaard et al., 2016). Some studies caution that commercialization at home while $1.137 billion, or 28%, comes from the sale of forest of NTFPs can undermine livelihood security for traditional users as a products (Shepherd et al., 2013: 56). In order of importance, forest result of elite capture (e.g., Lybbert et al., 2002; Mahapatra and products collected for generating cash are: fuelwood and charcoal, Shackleton, 2012). Another related reason may be a reluctance among which generates 36% of cash sales; building materials; forest foods; researchers and others to recommend formalizing an informal sector, fiber; medicine; and timber.5 Shepherd et al. conclude that forests' ac- with resulting taxation and other implications that could dis- tual contribution to Uganda's economy may be as high as 15% of GDP proportionately hurt poorer forest users. (2013: xv). One consequence of this under-estimation of the contribu- An additional reason that many governments are uninterested in tion of Uganda's forests is that government decision makers fail to NTFPs is because NTFPs, in contrast to timber, do not generate hard consider the value of Uganda's forests when making land use decisions currency. For instance, although enormous volumes of fuelwood and (Shepherd et al., 2013) and budget allocations (Wahlén and Ielmini, charcoal are traded annually, especially in Africa (e.g., Falconer, 1990), 2015). governments rarely collect data on their consumption and trade. Gov- In summary, the studies included in this brief review suggest a lack ernments and others may also assume that the use and trade of NTFPs of systematic data on the subsistence and economic value of forest has very little real value; consequently, NTFPs are not viewed as products, which contributes to an underestimation of both household something that can contribute to development or poverty alleviation and are therefore not the focus of national forest data collection efforts. Another government-related challenge relates to coordination
5 This analysis includes two districts that were undergoing post-conflict reconstruction across agencies. Government activities on forests and forestry may during the study period. Shepherd et al. (2013) suggest the value of forests to rural people occur across multiple government agencies, presenting territorial con- will decline to US$3.1 billion once reconstruction is complete. cerns and additional data collection challenges. The FRA, for instance,
14 C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19 identifies lack of coordination across agencies as one difficulty in col- underestimated and under-recognized, including by key policymakers lecting data on public sector expenditure on forestry and income from in finance and planning ministries and other national agencies charged forestry (Whiteman et al., 2015). with poverty reduction (FAO, 2010; Agrawal et al., 2013; Shepherd Fourth, a substantial percentage of NTFP consumption occurs in et al., 2013).6 Consequently, NTFPs are under-represented in national rural areas, which are typically further away from government offices development plans, forest strategies, and other development documents and may be more difficult or time consuming for researchers to access (FAO, 2014a), with budgets and focal points for NTFPs largely absent. (e.g., Agrawal et al., 2013; Shackleton and Pandey, 2014). Conse- Even in cases where regional or national policies reference NTFPs, on quently, rural or marginalized areas tend to be de-prioritized or ignored the ground implementation is largely weak or absent. in both research efforts and policies. In Uganda, a country that often features in forestry case studies, the Fifth, researchers appear to have a preference for collecting NTFP National Forest Authority (NFA) has one employee responsible for data at the case study level rather than carrying out studies that ag- managing community forestry activities, including any NTFP activities. gregate data up to the national level (Shackleton and Shackleton, 2004; This employee is mandated with designing and implementing com- Agrawal et al., 2013). Others highlight an emphasis on studies in sub- munity forest management agreements and NTFP activities with all Saharan Africa (Hogarth et al., 2013), noting there are comparatively communities throughout Uganda, using a nominal budget, a virtually fewer studies from Latin America (Angelsen et al., 2014). Further, some impossible task (Wahlén and Ielmini, 2015). Further, the NFA has never NTFPs are excluded from rural household surveys altogether carried out a national study of demand for, use of, and values of NTFP (Shackleton et al., 2011; Vira et al., 2015b). products and does not collect any annual data on NTFPs. Consequently, A related challenge is that different studies rely on different units of the NFA is unable to demonstrate the true value of Uganda's forests to analysis (household versus product) or measure the number or per- other government ministries, which means that the sector is de-prior- centage of people that use the NTFP rather than the value of the NTFP, itized. Such limited knowledge and budget allocation on NTFP re- making national, regional, and international comparisons across studies sources is likely more representative of developing country situations impossible. One reason for this preference may be because of metho- than not. dological challenges in aggregating NTFP data at higher levels. Another important consideration is the degree to which countries Further, assessing the non-cash value of forests at the national level actually address NTFPs in their national forest policies. FAO's State of isadifficult and expensive task (Agrawal et al., 2013). For instance, the World's Forests (2014a) argues that there is an increasing emphasis Shepherd et al. (2013) explain that it would be impossibly costly to on socioeconomic benefits from forests in forestry policies and mea- undertake inventories of all species used and sold and even more sures around the world, pointing to general adoption of the concept of complex to try and undertake a survey of all the NTFPs that are com- SFM in many countries, increased attention to participation in forest monly collected but rarely sold. Even the SOFO (2014), which focused management and policy processes, and countries in Latin America that exclusively on the socioeconomic benefits of forests states: “Due to the are developing and using payments for ecosystem services (PES) difficulty in obtaining data and the wide scope of actual and potential schemes. Burkina Faso has created a government agency focused on benefits, SOFO 2014 does not present a comprehensive picture of the NWFPs, the national ‘Agence des produits forestiers non ligneux’ socioeconomic contributions of forests” (FAO, 2014a: 2). This state- (Agency for the Promotion of NWFPs, APFNL), and developed a na- ment underscores the massive challenge of such a data collection effort. tional strategy and action plan for the sustainable use of NWFPs (SOFO, Recognizing this immense challenge, some researchers have re- 2014). The APFNL includes a pilot unit for the demonstration and commended gathering data on a sub-set of the most important NTFPs promotion of NWFPs (FAO, n.d.). It is not clear, however, to what ex- sold in a particular location and then generalizing based on that data. tent the recognition of the role of forests in policy documents and For instance, Laird et al. (2010) and SOFO (2011) recommend sur- country reports has translated into any real action or implementation, a veying the six most important species used and sold in any location. point also noted by Shackleton and Pandey (2014). They observe, for Shackleton and Pandey (2014) propose beginning with 10 to 20 key example, that India and South Africa both mention NTFPs in their na- species and building from there. Although such an approach would only tional forestry policies but neither country conducts a national in- demonstrate a percentage of the total value of NTFPs, this approach ventory of NTFPs or estimates the market values of NTFPs. would still represent a step forward from the fragmented efforts that One way to examine whether increased recognition of the role of the currently exist and could serve as the building block for more robust non-cash and informal benefits from forests in policy documents has studies that demonstrate the importance of documenting the value of contributed to change on the ground is to compare language on NTFPs NTFPs. in national forest plans with actual budget allocations. Such analysis Additional challenges include limited financial and technical re- could assess if governments merely pay lipservice to the importance of sources for collecting data on environmental and social indicators, lack such products or if they actually devote their resources and attention to of political commitment, and confusion over concepts. For instance, managing and promoting NTFPs. Again, Uganda serves as an illus- many researchers point out that there is not an internationally accepted trative example. For instance, SOFO, (2014) references Uganda several definition or classification system for NTFPs (e.g., Dlamini and times, categorizing it as one of the few countries to recognize NWFPs as Geldenhuys, 2011; Angelsen et al., 2014; Shackleton and Pandey, a contributor to food security and income generation and observing that 2014). Others observe challenges related to collecting quality data Uganda's National Development Plan recognizes forestry as a primary (Lund et al., 2011) and find that different data collection methods yield growth sector in the country that contributes to economic growth and different estimates of forest income and expenditure (Jagger et al., income (FAO, 2014a). 2012). A closer examination of Uganda's actions in the forestry sector, In summary, significant methodological challenges exist in esti- however, paints a somewhat different picture. In Uganda's NFA, 2015 mating the economic contributions of non-cash forest products. The Business Plan, the NFA states its intention to ensure an equitable supply conclusion and recommendations section presents a few solutions to of forest and non-forest products and services as one of four strategic some of these challenges. The next section considers the implications of objectives (NFA, 2015). Despite this recognition, the NFA budget is the lack of data for forest governance, management, and policy. devoted almost entirely to generating revenue for its operations through plantation focused activities and timber sales. According to 3.1. The implications of a lack of data
The lack of evidence on the non-cash values of forests means that 6 For a discussion of some of the reasons why NTFPs have remained low on the de- the contributions of forests to poverty alleviation and development are velopment agenda see Sunderland et al., 2011.
15 C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19
NFA's community staff member, “there is no budget” for implementing policies that are necessary in recognizing the non-cash benefits of for- community forest management plans, including activities related to ests: namely, policies that incorporate economic and livelihood objec- NTFPs (Wahlén and Ielmini, 2015). tives and also encourage and promote SFM. For instance, a national The lack of evidence on the non-cash values of forests also has development plan that sets out a goal of increasing access to renewable significant implications for development, particularly efforts to reduce sources of energy (SDG 7) would help alleviate pressure on forests (SDG poverty and improve livelihoods and well-being (see also World Bank, 15) and contribute to improved food security (SDG 2) and health (SDG 2008). A large body of work examines the role of forests and trees in 3) and decreased greenhouse gas (GHG) emissions (SDG 13). Similarly, poverty alleviation (e.g., Cavendish, 2000; Angelsen et al., 2014). These given the well-documented contribution of NTFPs to rural diets and studies often tease out differences among NTFP use between poorer and nutrition, policymakers can make a case for promoting SFM through wealthier households (e.g., Paumgarten and Shackleton, 2009; SDG 15 (life on land) to support progress on goals related to zero Thondhlana et al., 2012). Ambrose-Oji (2003) argues NTFPs can play a hunger (SDG 2) and good health and well-being (SDG 3). These types of role in poverty alleviation strategies and forest conservation, based on integrated, multi-sectoral approaches will be critical in achieving the her findings that households in Cameroon derive value from NTFPs in 2030 Agenda. Conversely, achieving the SDG targets will also require secondary forest and forest fallow areas rather than the forest areas managing trade-offs among SDGs, such as by ensuring that progress on prioritized by conservationists for protection. Hogarth et al. (2013) find one SDG does not come at the expense of achieving another SDG (e.g., that forest-related income, in contrast to off-farm income, provided van Vianen et al., 2015). benefits for households across all income levels. Based on these find- By promoting integrated approaches, forest ministries and others ings, Hogarth et al. (2013) suggest poverty alleviation targeting could who care about forests have a rare opportunity to position the forest help reach the lower-income households in their study that were “left sector to receive increased attention and support. Currently, in many behind”, in part due to the elite capture of forest resources. Such developing countries, governments often place low priority on the en- findings are even more relevant in the context of the SDGs' emphasis on vironmental sector, including forests and other natural resources, with leaving no one behind. The section below elaborates on the role of greater attention and budget allocations to “priority” sectors such as forests in the SDGs and opportunities to better position forests and trees education, health, and infrastructure. The Government of Uganda's in the SDG agenda. budget allocations provide an illustrative example. In 2013–2014, for instance, 3.5% of the Government of Uganda's budget allocation went to water and environment in comparison with 13% for education and 4. Integrating the non-cash value of forests into the development 19.2% for works and transport (Wahlén and Ielmini, 2015). Forest and agenda environment ministries can use the current focus on SDG implementa- tion to demonstrate how NTFPs in particular and forests in general can As the world begins implementing the 2030 Agenda on Sustainable contribute to achieving a more sustainable world by 2030 and to argue Development and its SDGs, forest researchers and others have an op- for increased prioritization of their sectors, including increased portunity to increase the visibility of forests and trees in the develop- budget allocations. ment agenda and to position the role of forests and trees as critical in Finally, inter-governmental organizations, governments, and other supporting the livelihoods and subsistence of the poor. This section il- stakeholders have all stressed the importance of additional data col- lustrates key areas in which forests offer transformative potential to lection efforts to support SDG implementation, monitoring, follow-up, increase attention on the non-cash benefits of forests. and review (UN, 2016). As the UN's Inter-Agency and Expert Group on Forests are primarily addressed in SDG 15, which aims to “protect, SDG indicators (IAEG-SDGs) and others devise new data collection ef- restore, and promote sustainable use of terrestrial ecosystems, sus- forts to support SDG implementation, those interested in forests should tainably manage forests, combat desertification, and halt and reverse position forests more prominently within this data collection effort and land degradation, and halt biodiversity loss” (UN SDKP, n.d.). Target suggest specific areas for targeted data collection and enhanced statis- 15.2 specifically focuses on forests: “By 2020, promote the im- tical capacity. Such areas could include national level indicators on the plementation of sustainable management of all types of forests, halt non-cash and informal benefits forests generate through NTFPs and deforestation, restore degraded forests, and substantially increase af- NWFPs and similar questions in the CRFQ on NTFP consumption and forestation and reforestation globally” (UN SDKP, n.d.). Perhaps even use. Stakeholders should also use this opportunity to argue for im- more importantly, Target 15.9 calls for action to, “integrate ecosystem provements in data that is already collected by FAO to enhance its and biodiversity values into national and local planning, development accuracy and accessibility for policy. processes, poverty reduction strategies and accounts,” by 2020, a target in line with the recommendations on encouraging NCA in this review. 5. Conclusions and recommendations NCA can also help countries ensure that progress on one SDG does not negatively impact another SDG. Can forests every be truly valued? As this review illustrates, esti- SDG 15 and other SDGs present an opportunity for transformative mating the non-cash value of NTFPs for direct use benefits is a complex approaches across government agencies to achieve the Goals.7 It is undertaking, with most researchers preferring to focus at the case study widely acknowledged that government agencies will need to work level or at least at the sub-national level. Such preferences and limita- across previous boundaries to achieve the SDGs and leverage financial tions result in data collection that is insufficient in its representation of resources. The forest sector is no exception: government agencies can forest activities and benefits that are not formally captured through take advantage of the strategic opportunity offered by the SDGs to market records, taxation, or other formal means. The systematic col- develop joint datasets on NTFPs, which could help to create a more lection of such data is necessary to quantify and recognize the economic representative picture of the contributions of forests and trees and en- contributions from forests and trees and can serve as a critical first step able policymakers to illustrate the role of NTFPs in poverty alleviation. in documenting the enormous contribution of NTFPs to economies In turn, policymakers could use this data to increase awareness on the around the world. Once governments, international organizations, and role NTFPs could play in helping to achieve the SDGs. other stakeholders are better able to put a number on the value of Similarly, the SDGs offer a chance to promote cross-agency and NTFPs, it will then be possible to design and implement actions to cross-sectoral collaboration, as governments are increasingly called improve the generation and capture of such benefits, thereby increasing upon to adopt the types of cross-sectoral strategies and development their contribution to poverty alleviation and other development aims. As this brief review suggests, better information on the con- 7 See CIFOR (n.d.) for an elaboration on how forests relate to the SDGs. tributionsofforestsintheformofNTFPsispossible.Theinclusion
16 C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19 of data on NWFP value in the FRA supplementary section is an harvesting of these forests are potential areas for government inter- important next step in the evolution of the FRA and understanding vention. informal benefits more broadly. Inclusion of questions on non-cash Although global efforts to collect and analyze NTFPs are necessary, income in the CFRQ, with the aim of providing close to universal individual researchers are likely to continue using case study ap- coverage of this variable, would help to improve recognition of the proaches. Initiatives like PEN, IFRI, and FLARE can raise awareness on value of forests at the national level and eventually contribute to their aims and efforts, with the goal of ensuring that such case studies improved understanding of the non-cash contributions of forests at can contribute to better comparative studies with consistent meth- both national and international levels. odologies that can be used to better assess the impact of land-use A few studies suggest improved methodological approaches for changes on forests, the cash and non-cash contributions of forests to collecting such data (e.g., Agrawal et al., 2013; Shackleton and Pandey, economic development, and their cross-sectoral linkages. Other options 2014). PEN promotes a common methodology among researchers and include drawing on and adapting the wide range of case studies on across geographic areas and aggregates case study data, an approach different NTFPs, including using such studies to develop a handbook for that could be scaled up to include other international forestry efforts. national agencies that presents sample techniques and methodologies FAO has increased attention to the socioeconomic benefits of forestry for use by national data collection agencies (Shackleton and Pandey, through SOFO (2014), supplementary information in the FRA, and 2014). The growing use of information and communications technol- guidance on national socioeconomic surveys in forestry (Bakkegaard ogies (ICTs) around the world also offer an opportunity to develop easy- et al., 2016). to-use tools that local communities can use to collect data and then To address challenges related to seasonality, governments or re- report using smart phones (see Liebenberg, 2003 in Shackleton and searchers can carry out a more thorough initial inventory to determine Pandey, 2014). harvest times and select appropriate seasons for data collection, with Finally, any form of data collection must be accessible, practical, multiple surveys throughout a year, if necessary. The PEN project, for and relevant for use by policymakers. The true test of such data will be instance, uses surveys that cover a 12-month period, which enables whether it is used by policymakers to reform forest policy to change collection of data on seasonal species. Still, additional methodological management practices or drive investments that result in positive im- research is needed on simple and inexpensive techniques for capturing provements in forest management. the contributions of NTFP, which could help convince cash-strapped In conclusion, there is an urgent need for better documentation of governments and multilateral organizations to include indicators on the non-cash benefits provided by forests that can be used to inform NTFPs in their data collection efforts. governments and policymakers on the true value of forest resources. In The recognition and evaluation of the full range of forest values is turn, such data could enable governments and policymakers to update important not just to demonstrate the value of forests for policymakers and reform forest policies to more explicitly recognize the non-cash or to contribute to poverty alleviation but also to preserve and protect benefits provided by forests. the world's forests and trees. Over the long-term, failure to prioritize the environment and natural resources sector will affect countries devel- Funding opment. The MDG Report for Uganda 2013 cautions, for instance, that depletion of Uganda's natural wealth costs the country four to 12% of Initial background research for this paper was supported through a its national income annually (GoU and UNDP, 2013: 31). Such figures consultancy with the UN Forum on Forests (UNFF) as part of back- underscore the vast, and unrecognized, loses that the world will face if ground research for Agrawal et al. (2013). it continues to fail to prioritize its forests. The forestry sector will continue to be under-prioritized and under-resourced until govern- Acknowledgements ments provide more than nominal recognition of the true value of their country's natural capital, including the non-cash value of forest pro- My deep appreciation goes to Arun Agrawal and Gill Shepherd, with ducts and the ecosystem services provided by its natural forests. whom I had the pleasure of collaborating on an in-depth examination of Although true valuation of NTFPs is unlikely ever to be realized, this the cash and non-cash value of NTFPs (Agrawal et al., 2013), which review presents some avenues for moving towards an improved un- generated many of the initial ideas for this paper. Thank you to Arun, derstanding of the non-cash and informal benefits generated by forests Cristy Watkins, and Laura Rasmussen for the invitation to contribute to and trees, including by taking advantage of the transformative potential this issue and for comments on an earlier version of this paper. I am also of the 2030 Agenda to re-position forests more prominently within the grateful to two anonymous reviewers whose comments helped me to development agenda and within government ministries. On the SDGs, develop more convincing evidence for several parts of this paper. this article highlights opportunities for promoting cross-agency and cross-sectoral collaboration on forests, increasing prioritization of the References forestry sector, including through increased budget allocation, and improving data collection on forests. Agrawal, A., Cashore, B., Hardin, R., Shephard, G., Benson, C., Miller, D., 2013. The Another option for improved data collection is collaboration among Economic Contributions of Forests. (Paper prepared for the UN Forest Forum). Alexiades, M., Shanley, P. (Eds.), 2005. Forest Products, Livelihoods and Conservation. international organizations, such as the FAO, the World Bank, and Case studies of Non-Timber Forest Product Systems. Volume 3 - Latin America Center CIFOR, to design an instrument or toolkit to collect, compile, and for International Forestry Research (CIFOR), Bogor, Indonesia. harmonize data on forest resources, including NTFPs and NWFPs, and Ambrose-Oji, B., 2003. The contribution of NTFPs to the livelihoods of the ‘forest poor’: ff evidence from the tropical forest zone of south-west Cameroon. Int. For. Rev. 5 (2), pilot test this instrument in a few key countries. Such an e ort should 106–117. build on existing efforts, including FAO's CFRQ and the World Bank's Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N.J., Bauch, S., Borner, J., WAVES partnership. Such efforts should also work to improve capa- Smith-Hall, C., Wunder, S., 2014. Environmental income and rural livelihoods: a – cities for data and information collection and analysis and support to global-comparative analysis. World Dev. 64 (S1), S12 S28. Angelsen, A., Wunder, S., 2003. Exploring the forest–poverty link: Key concepts, issues collect, store, and analyze this new data. and research implications. Occasional Paper No. 40. CIFOR, Bogor, Indonesia. In the absence of a globally coordinated effort to collect such data, Appasamy, P., 1993. Role of non-timber forest products in a subsistence economy: the – national forest agencies and decision makers can improve their efforts case of a joint forestry project in India. Econ. Bot. 47 (3), 258 267. Babulo, B., Muys, B., Nega, F., Tollens, E., Nyssen, J., Deckens, J., Mathijs, E., et al., 2009. to address the needs of forest-dependent peoples and recognize the true The economic contribution of forest resource use to rural livelihoods in Tigray, contribution of forests and trees to livelihoods and economies. National Northern Ethiopia. Forest Policy Econ. 11, 109–117. and local efforts to clarify institutional arrangements for forests and Bakkegaard, R.K., Agrawal, A., Animon, I., Hogarth, N., Miller, D., Persha, L., Rametsteiner, E., Wunder, S., Zezza, A., 2016. National Socioeconomic Surveys in strengthen management and governance around the allocation and
17 C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19
forestry: Guidance and Survey Modules for Measuring the Multiple Roles of Forests in related socio-economic factors: lessons from Bangladesh. Forest Policy Econ. 14, Household Welfare and Livelihoods. FAO Forestry Paper No. 179. FAO, CIFOR, IFRI 136–142. and World Bank. Karayilmazlar, S., 2005. Analysis of Turkey's foreign trade in non-wood forest products. Barik, S., Mishra, S., 2008. Assessment of the contribution of forests to the economy of the For. Prod. J. 55 (4), 78–87. northeastern states of India. Int. For. Rev. 10 (2), 349–361. Laird, S., McLain, R., Wynberg, R., 2010. Wild Product Governance: Finding Policies that Belcher, B., Kusters, K., 2004. Non-timber forest product commercialisation: development Work for non-timber Forest Products. Earthscan, London. and conservation lessons. In: Kusters, K., Belcher, B. (Eds.), Forest Products, Liebenberg, L., 2003. A new environmental monitoring methodology. Available online: Livelihoods and Conservation. Case studies of Non-Timber Forest Product Systems. http://www.cybertracker.org. Volume 1- Asia CIFOR, Bogor. Lund, J.F., Shackleton, S., Luckert, M., 2011. Getting quality data. In: Angelsen, A., Belcher, B., Ruiz Perez, M., Achdiawan, R., 2005. Global Patterns and Trends in the use Larsen, H.O., Lund, J.F., Smith-Hall, C., Wunder, S. (Eds.), Measuring Livelihoods and and Management of Commercial NTFPs: Implications for Livelihoods and Environmental Dependence: Methods for Research and Fieldwork. CIFOR Available Conservation. CIFOR, Bogor, Indonesia. at: http://www.cifor.org/publications/pdf_files/Books/BAngelsen110211.pdf. Byron, N., Arnold, M., 1997. What Futures for the People of Tropical Forests?’ CIFOR Lybbert, T.J., Barrett, C.B., Narjisse, H., 2002. Market based conservation and local Working Paper 19. CIFOR, Bogor. benefits: the case of argan oil in Morocco. Ecol. Econ. 4, 125–144. Cavendish, W., 2000. Empirical regularities in the poverty–environment relationship in Mahapatra, A., Shackleton, C., 2012. Exploring the relationships between trade in natural rural households: evidence from Zimbabwe. World Dev. 28 (11), 1979–2003. products, cash income and livelihoods in tropical Forest regions of eastern India. Int. Delang, C.O., 2006. Not just minor forest products: the economic rationale for the con- For. Rev. 14 (1), 62–73. sumption of wild food plants by subsistence farmers. Ecol. Econ. 59 (1), 64–73. Mahapatra, A., Albers, H., Robinson, E., 2005. The impact of NTFP sales on rural Dlamini, C., Geldenhuys, C., 2011. Towards a theoretical framework for the management households' cash income in India's dry deciduous forest. Environ. Manag. 35 (3), of non-timber forest products (NTFPs) in Swaziland: a review. J. Geogr. Reg. Plann. 4 258–265. (15), 715–730. NFA, 2015. NFA 2015-2019 Business Plan. Draft copy August 2015. Dlamini, C., Geldenhuys, C., 2009. The socioeconomic status of the non-timber forest Newton, P., Miller, D., Byenkya, M., Agrawal, A., 2016. Who are forest-dependent people? product subsector in Swaziland. South. For. 71 (4), 311–318. A taxonomy to aid livelihood and land use decision-making in forested regions. Land Dovie, D., 2003. Rural economy and livelihoods from the non-timber forest products Use Policy 57, 388–395. trade. Compromising sustainability in southern Africa? Int. J. Sustain. Dev. World Paumgarten, F., Shackleton, C.M., 2009. Wealth differentiation in household use and Ecol. 10 (3), 247–262. trade in non-timber forest products in South Africa. Ecol. Econ. 68 (12), 2950–2959. Falconer, J., 1990. The Major Significance of ‘Minor’ Forest Products: The Local use and Powell, B., Hall, J., Johns, T., 2011. Forest cover, use and dietary intake in the East Value of Forests in the West African Humid Forest Zone. FAO, Rome, Italy. Usambara Mountains. Tanzania 13 (3), 305–317. FAO, 2015. Global Forest Resources Assessment 2015. Main report. FAO, Rome, Italy. The Program on Forests (PROFOR) at the World Bank, 2015. Global Assessment of FAO, 2014a. SOFO: Enhancing the Socioeconomic Benefits from Forests. FAO, Rome, Nonwood Forest Ecosystem Services: Spatially Explicit Meta-Analysis and Benefit Italy. Transfer to Improve the World Bank's Forest Wealth Database. The World Bank, FAO. 2014b. Contribution of the Forestry Sector to National Economies, 1990–2011, by Washington, DC. A. Lebedys and Y. Li. Forest Finance Working Paper FSFM/ACC/09. Rome, Italy: PROFOR, 2008. Poverty and Forest Linkages: A Synthesis and Six Case Studies. The World FAO. Bank, Washington, DC. FAO, 2010. Global Forest Resources Assessment 2010. Main report. FAO, Rome, Italy. Rowland, D., Ickowitz, A., Powell, B., Nasi, R., Sunderland, T., 2016. Forest foods and FAO, 2012. State of the World's Forests 2012. FAO, Rome, Italy. healthy diets: quantifying the contributions. Environ. Conserv. 1–13. http://dx.doi. FAO n.d. Burkina Faso Factsheet. Available at: http://www.fao.org/fileadmin/user_ org/10.1017/S0376892916000151. upload/wa_workshop/Factsheet/13.FSNL_Fact_sheet_NWFP_Burkina.pdf (Accessed Shackleton, C.M., Shackleton, S.E., Buiten, E., Bird, N., 2007. The importance of dry on 12 May 2016). woodlands and forests in rural livelihoods and poverty alleviation in South Africa. Foli, S., Reed, J., Clendenning, J., Petrokofsky, G., Padoch, C., Sunderland, T., 2014. To Forest Policy Econ. 9, 558–577. what extent does the presence of forests and trees contribute to food production in Shackleton, S., Campbell, B., Lotz-Sisitka, H., Shackleton, C., 2008. Links between the humid and dry forest landscapes? A systematic review protocol. Environ. Evid. 3 local trade in natural products, livelihoods and poverty alleviation in a semi-arid (15), 1–8. region of South Africa. World Dev. 36 (3), 505–526. Fu, Y., Chen, J., Guo, H., Chen, A., Cui, J., Hu, H., 2009a. The role of non-timber forest Shackleton, S., Paumgarten, F., Kassa, H., Husselman, M., Zida, M., 2011. Opportunities products during agroecosystem shift in Xishuangbanna, southwestern China. Forest for enhancing women's economic empowerment in the value chains of three African Policy Econ. 11 (2009), 18–25. non timber forest products. Int. For. Rev. 13, 136–151. Fu, Y., Chen, J., Guo, H., Hu, H., Chen, A., Cui, J., 2009b. Rain forest dwellers' liveli- Shackleton, C., Pandey, A., 2014. Positioning non-timber forest products on the devel- hoods: income generation, household wealth and NTFP sales: a case study from opment agenda. Forest Policy Econ. 38, 1–7. Xishuangbanna, S W China. Int. J. Sustain. Dev. World Ecol. 16 (5), 332–338. Shackleton, C.M., Shackleton, S.E., 2004. The importance of non-timber forest products in Government of Uganda (GoU) and UN Development Programme (UNDP), rural livelihood security and as safety nets: a review of evidence from South Africa. S. 2013Government of Uganda (GoU) and UN Development Programme (UN. Afr. J. Sci. 100 (11−12), 658–664. Millennium Development Goals Report for Uganda 2013. Ministry of Finance, Shepherd, G., Kazoora, C., Müller, D., 2013. Forests, Livelihoods and Poverty Alleviation: Planning and Economic Development, Kampala, Uganda Available at: http://www. The Case of Uganda. FAO, Rome, Italy. ug.undp.org/content/dam/uganda/docs/UNDPUg-2013MDGProgress%20Report- Sisak, L., Riedl, M., Dudik, R., 2016. Non-market non-timber forest products in the Czech Oct%202013.pdf. Republic: their socio-economic effects and trends in forest land use. Land Use Policy Guariguata, M.R., García-Fernández, C., Nasi, R., Sheil, D., Herrero-Jáuregui, C., 50, 390–398. Cronkleton, P., Ndoye, O., Ingram, V., 2011. Timber and non-timber forest product Sunderland, T., Harrison, S., Ndoye, O., 2004. Commercialisation of non-timber forest extraction and management in the tropics: towards compatibility? In: Shackleton, products in Africa: history, context and prospects. In: Sunderland, T., Ndoye, O. S.E., Shackleton, C.M., Shanley, P. (Eds.), Non-timber Forest Products in the Global (Eds.), Forest Products, Livelihoods and Conservation. Case Studies of Non-Timber Context. Springer, Heidelberg, pp. 171–188. Forest Product Systems. Volume 2 - Africa CIFOR. Gunatilake, D., Senaratne, A., Abeygunawardena, P., 1993. Role of non-timber forest Sunderland, T.C.H., Ndoye, O., Harrison-Sanchez, S., 2011. Non-timber forest products products in the economy of peripheral communities of knuckles National Wilderness and conservation: what prospects? In: Shackleton, S., Shackleton, C., Shanley, P. Area of Sri Lanka: a farming systems approach. Econ. Bot. 47 (3), 275–281. (Eds.), Non-timber Forest Products in the Global Context. Heidelberg, Springer, pp. Hogarth, N., Belcher, B., Campbell, B., Stacey, N., 2013. The role of forest-related income 209–224. in household economies and rural livelihoods in the border-region of Southern China. Thondhlana, G., Vedeld, P., Shackleton, S., 2012. Natural resource use, income nad de- World Dev. 43, 111–123. pendence among San and Mier communities bordering Kgalagadi Transfrontier Park, Huber, F., Ineichen, R., Yang, Y., Weckerle, C.S., 2010. Livelihood and conservation as- southern Kalahari, South Africa. Int. J. Sustain. Dev. World Ecol. 19 (5), 460–470. pects of non-wood Forest product collection in the Shaxi Valley, Southwest China. UN Statistics Division (UNSD). n.d. System of Environmental-Economic Accounting Econ. Bot. 64 (3), 189–204. (SEEA). Available at: http://unstats.un.org/unsd/envaccounting/seea.asp (accessed Hussain, S., Badola, R., 2010. Valuing mangrove benefits: contribution of mangrove on 5 May 2016). forests to local livelihoods in Bhitarkanika Conservation Area, East Coast of India. UN, 2016. The SDGs Report. Available at: https://unstats.un.org/sdgs/report/2016/The Wetl. Ecol. Manag. 18, 321–331. %20Sustainable%20Development%20Goals%20Report%202016.pdf. Ickowitz, A., Powell, B., Salim, M., Sunderland, T., 2014. Dietary quality and tree cover in UN Sustainable Development Knowledge Platform (SDKP). Available at: https:// Africa. Glob. Environ. Chang. 24, 287–294. sustainabledevelopment.un.org/sdgs (Accessed on 5 May 2016). Jagger, P., Luckert, M.K., Banana, A., Bahati, J., 2012. Asking questions to understand van Vianen, J., Reed, J., Sunderland, T., 2015. From global complexity to local reality: rural livelihoods: comparing disaggregated vs. aggregated approaches to household aligning implementation frameworks with SDGs and landscape approaches. In: livelihood questionnaires. World Dev. 40 (9), 1810–1823. CIFOR Policy Brief no. 1. Jamnadass, R., McMullin, S., Iiyama, M., Dawson, I., Powell, B., Termote, C., Ickowitz, A., Vira, B., Wildburger, C., Mansourian, S. (Eds.), 2015. Forests and Food: Addressing Kehlenbeck, K., Vinceti, B., van Vliet, N., Keding, G., Stadlmayr, B., Van Damme, P., Hunger and Nutrition Across Sustainable Landscapes. Carsan, S., Sunderland, T., Njenga, M., Gyau, A., Cerutti, P., Schure, J., Kouame, C., Vira, B., Wildburger, C., Mansourian, S. (Eds.), 2015. Forests, Trees and Landscapes for Obiri, B., Ofori, D., Agarwal, B., Neufeldt, H., Degrande, A., Serban, A., 2015. Food Security and Nutrition: A Global Assessment Report, IUFRO World Series Understanding the roles of forests and tree-based systems in food provision. In: Vira, Volume 33. International Union of Forest Research Organizations (IUFRO), Vienna, B., Wildburger, C., Mansourian, S. (Eds.), Forests, Trees and Landscapes for Food pp. 172. Security and Nutrition: A Global Assessment Report, IUFRO World Series Volume 33. WAVES n.d. Wealth Accounting and the Valuation of Ecosystem Services. Available at: International Union of Forest Research Organizations (IUFRO), Vienna, pp. 172. https://www.wavespartnership.org/ (accessed on 5 May 2016). Kar, S.P., Jacobson, M.G., 2012. NTFP income contribution to household economy and Wahlén, C. B. and M. Ielmini. 2015. Assessment of Uganda's National Forestry Authority.
18 C.B. Wahlén Forest Policy and Economics 84 (2017) 11–19
Paper prepared for the US Forest Service (USFS) and the US Agency for International World Bank, 2008. Forests Sourcebook: Practical Guidance for Sustaining Forests in Development (USAID). Development Cooperation. The World Bank, Washington DC. Whiteman, A., Wickramasinghe, A., Pina, L., 2015. Global trends in forest ownership, Wunder, S., Angelsen, A., Belcher, B., 2014. Forests, livelihoods and conservation: public income and expenditure on forestry and forestry employment. For. Ecol. broadening the empirical base. World Dev. 64 (S1), S1–S11. Manag. 352 (2015), 99–108.
19 Forest Policy and Economics 84 (2017) 20–28
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Measuring forest and wild product contributions to household welfare: Testing a scalable household survey instrument in Indonesia☆
Riyong Kim Bakkegaard a,⁎, Nicholas J. Hogarth b, Indah Waty Bong c, Aske S. Bosselmann d, Sven Wunder c a Caroline Amalie Vej 78, Kongens Lyngby 2800, Denmark b Viikki Tropical Resources Institute (VITRI), University of Helsinki, Department of Forest Sciences, P.O. Box 27, (Latokartanonkaari 7), FI-00014 University of Helsinki, Finland c Center for International Forestry Research (CIFOR), Jalan CIFOR, Situ Gede, Sindang Barang, Bogor (Barat) 16115, Indonesia d University of Copenhagen, Rolighedsvej 25, 1258 Frederiksberg, Denmark article info abstract
Article history: Systematic comparisons of human dependence on forests and environmental resources have been challenging, as Received 25 April 2016 a result of heterogeneous methodologies. Specialized Forestry Modules have been developed, with the goal of fill- Received in revised form 12 September 2016 ing current information gaps concerning the economic importance of forest and wild products in household wel- Accepted 5 October 2016 fare and rural livelihoods. Results from a pilot assessment of the Forestry Modules in West Kalimantan, Indonesia, Available online 15 October 2016 are presented, showing that the Forestry Modules perform well in extracting the expected information: mean per capita forest and wild product income shifts according to the geographical “forest gradient”.Significantly, in the Keywords: Socioeconomic survey forest-rich upstream village, mean forest and wild product income and mean forest-related wage and business Methodology incomes exceeds current mean agricultural income statistics for West Kalimantan and mean non-agricultural Forest-based livelihoods rural household incomes in the lowest bracket. Consumption of forest products and importance as a coping strat- Household income egy was higher in the most upstream village, where sale of forest products in times of shock was more marked in West Kalimantan the most downstream village (where forest coping strategies were also least important). The Forestry Modules' detailed and systematic approach can help ensure that contributions of forest and wild products are not underestimated in national figures. © 2016 Elsevier B.V. All rights reserved.
1. Introduction to date has been comprised primarily of case studies using heterogeneous methodologies. In the 2005 World Development Special Issue on ‘Liveli- Forests usually play important provisioning and supporting roles in hoods, Forests and Conservation’, one of the main conclusions was that the livelihoods of rural households (Byron and Arnold, 1999; more worldwide studies, or synthesis of case studies, were needed in fu- Sunderlin et al., 2005). Some figures estimate that as much as 90% of ture research (Sunderlin et al., 2005). This call led to a global meta-study those who live in extreme rural poverty are to some degree reliant on by Vedeld et al. (2007), synthesizing 54 case studies with an estimated forests for their livelihoods (Chao, 2012). Beginning with seminal stud- average forest income contribution of 22%. The Center for International ies nearly two decades ago (e.g. Cavendish, 2000), a growing body of Forestry Research (CIFOR) initiated the Poverty and Environment Net- case-studies from a range of contexts showed that products and services work (PEN), a pan-tropical comparative study with cases in 24 countries, from non-cultivated ecosystems (such as natural forests, woodlands, where household (including forest-related) income was scrutinized wetlands, lakes, rivers and grasslands) can be significant sources of in- using best-practice standardized methods, such as quarterly household come for rural households, providing energy, food, construction mate- surveys (www1.cifor.org/pen). PEN results showed an average contribu- rials and medicines, both for subsistence and cash uses (e.g. tion of 27.5% forest and environmental income to households living in or Bakkegaard et al., 2016a; Fisher, 2004; McSweeney, 2004; Mamo et near forests; a figure that was only marginally lower than that of crop in- al., 2007; Appiah et al., 2009; Rayamajhi et al., 2012). come (Angelsen et al., 2014). Other studies found that even people living However, systematic comparisons of human dependence on forests in areas of lower tree densities may still rely substantially on the extrac- and other environmental resources have been challenging, as research tion of surrounding wild resources (Shepherd, 2012). Given these indications of the importance of forests to the well- being of rural populations in many contexts around the world, there is a strong case to routinely include an adequate set of questions regarding ☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” households' reliance on forest and wild products in household welfare published at the journal Forest Policy and Economics 84C, 2017. ⁎ Corresponding author. surveys that are used for policy development and evaluation. However, E-mail address: [email protected] (R.K. Bakkegaard). at present there is a systematic failure by the world's key household-
http://dx.doi.org/10.1016/j.forpol.2016.10.005 1389-9341/© 2016 Elsevier B.V. All rights reserved. R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28 21 level socioeconomic surveys to capture the full contribution of forest products derived from e.g. trees on farms. Wild products refer to prod- and environmental income in rural livelihoods (GTZ, 2004; FAO, 2008; ucts originating from “non-forest” and “wild” systems (e.g. other wood- World Bank, 2008). ed lands, savannahs, miombo, fallows, scrub-, grass- and rangelands). A set of standard methodologies that consistently measure the wel- Encompassing non-forest wild products is important, as their combined fare contribution of forests and the environment to household income harvest in some environments can exceed the value derived from for- and poverty alleviation could eventually ensure that forests and other ests (e.g. Pouliot and Treue, 2013). Excluded from the module are prod- environmental products are more reliably captured in local livelihood ucts grown in agricultural lands (cropland, pastures, agroforestry, metrics, regional poverty measures, and national gross domestic prod- silvipasture, fallow areas) and cultivated and captured resources from uct (GDP). Nevertheless, several measurement and data collection chal- aquatic environments, which are already covered in the LSMS under lenges are associated with this goal. For instance, forest product the Agricultural (World Bank, 2015a) and Fisheries Modules (World extraction may be illegal, so that respondents may be uncomfortable Bank, 2015b), respectively. reporting it in a household survey. Forests may provide essential subsis- The objective of this paper is to present the results from a pilot as- tence-oriented products, but lacking a market price makes it difficult to sessment of the Forestry Modules in Indonesia, and scrutinize their ef- value accurately (PROFOR, 2008; Wunder et al., 2011). Furthermore, ex- fectiveness in capturing key socioeconomic data related to forest and traction of many forest products is markedly more seasonal and some- wild products. We do this by first presenting an existing official tool times related to specific events, such as household shocks, than that measures socio-economic data in forest areas, namely the Indone- average household income, for both forest supply and demand reasons sian Forestry Survey, and then we turn to a description of the pilot (Byron and Arnold, 1999). site, the Forestry Modules and the main results of the assessment of Despite these challenges, work towards a standardized data-collec- the survey tool. We present the results of forest and wild product in- tion process for the contribution of forests to household welfare has come across the village sites, which we predicted would reflect the gra- been progressing in recent years (Angelsen et al., 2011). Yet, developing dient in forest cover and forest types, if the Forestry Modules were nationally representative data on the role of forest and wild products in effective in collecting the data they were designed for. From the most the household economy requires a more systematic approach across important sections of the Forestry Module (in welfare terms), the “In- forest types and ecoregions that considers how to deal with background come” and “Shocks and Crises” modules, we present some in-depth sub- factors determining the levels of resource use (e.g. population density, stance findings from the pilot test. We then conclude with insights into ethnicity, forest cover, or proximity to roads). further areas for methodological development, as well as on the contri- In response to this challenge, FAO along with CIFOR, IFRI (Interna- bution of forest-related data to national-level planning processes. tional Forestry Resources and Institutions), PROFOR (Program on For- ests, World Bank), and the LSMS-ISA team of the World Bank (Living 2. Pre-existing Indonesian sources of forest-related socioeconomic Standards Measurement Study – Integrated Surveys on Agriculture) data have joined forces to develop specialized modules on forest and wild products (hereafter referred to as Forestry Modules), with the goal of There are several pre-existing Indonesian data instruments that aim filling current information gaps concerning the economic importance to collect socioeconomic data on households. A few years apart, differ- of forest and wild products. The work involved two phases. In phase ent national household surveys have been carried out across Indonesia, one, three reports were produced: (1) a review of the coverage of for- including national socioeconomic household surveys since 1976, the est-related socioeconomic issues in selected surveys (Russo, 2014); Family Life Surveys since 1993, and agricultural censuses every decade (2) a micro-data analysis of selected socioeconomic surveys (Riggott, starting in 1963. Following the 2003 Agricultural Census, several sub- 2014); and (3), an analysis of CIFOR's Poverty Environment Network surveys were developed and carried out in 2004, including the Indone- (PEN) survey (Bakkegaard, 2013). Phase two included: (1) the develop- sian Forestry Survey1, which collected data on households living within, ment of standard and expanded survey modules on forest and wild or on the fringes of forest areas. The Indonesian Forestry Survey was products; (2) field testing of modules in three different country contexts Indonesia's first attempt at gathering comprehensive data on house- (including testing of a tablet version); and (3) producing a sourcebook holds' use of different types of non-private forests, including conserva- to guide potential users (http://foris.fao.org/preview/90390/en/). The tion areas and protected forest areas, according to reviews of national primary goal is for national statistical offices to integrate this module socioeconomic surveys back to 1990 (www.rand.org/labor/bps/ into national-level household socioeconomic surveys, thus providing susenas). At least as far back as 1990, the national socioeconomic sur- more complete information on national income, welfare, and veys collected data on products gathered in the forest, but were limited livelihoods. to rough estimates of yearly collection, consumption, and sales based on The Forestry Modules include household-level and community-level retrospective questions. The Forestry Survey was repeated in 2014, fol- instruments to collect data on the welfare contribution of forest and lowing the 2013 Agricultural Census. According to Statistics Indonesia wild products (and forest services) to rural households. They cover 13 (2014), the primary aim of the Indonesian Forestry Survey was to collect different themes including aspects such as direct income, wage-related data on shifting cultivation, harvesting of forest products, and the socio- income, business-related income, health, construction and energy con- economic condition of the households residing within, or in close prox- tributions, among other themes, as well as qualitative data on gover- imity to, forests, primarily to allow the government to establish effective nance of forests and its resources, and their importance in crisis or plans and policies to develop communities within or near forests. The coping responses. In the modules, forests are defined according to the survey components record a yes/no participation in, or occurrence of, FAO (2006, p. 169) definition as: an activity in a checklist form, rather than documenting the actual Land spanning more than 0.5 hectares with trees higher than 5 meters value or quantity. and a canopy cover of more than 10 percent, or trees able to reach these Given the intended aim of the Indonesian Forestry Survey, the re- thresholds in situ. It does not include land that is predominantly under ag- sults highlight some of the difficulties in obtaining quality data on forest ricultural or urban land use. income. Though the survey includes questions regarding different types This definition encompasses old-growth natural forest, secondary of product quantities extracted from forest areas, the lack of price or and regenerating natural forest, and managed plantations. Forest prod- ucts are therefore products originating from forests as defined above, 1 https://sirusa.bps.go.id/webadmin/kuesioner/2014_3352_ques_ST2013-SKH.S.pdf.A and include timber and a wide range of non-timber forest products guidance for agricultural and forestry survey has also been developed. https://sirusa.bps. (NTFPs), including tree-based products (e.g. fruits and nuts), plants go.id/webadmin/pedoman/2014_3352_ped_Pedoman%20Pencacah%20ST2013-SKH.PCS. (e.g. tubers), and animals (e.g. bush-meat), and including other wood pdf. 22 R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28
Fig. 1. Map of sample village locations with vegetation cover; Kalis sub-district, Kapuas Hulu district, West Kalimantan, Indonesia. value data impedes the accounting of (gross) income from forest and seasonal shortfalls. Having such information would provide holistic wild products collected by the households, thereby contributing to un- data on the nature and level of reliance of households on forest areas derestimation of this income source to total household incomes. The around them, thereby helping to shape more effective conservation section on forest products collects data on groups of products, the quan- and sustainable development policies. The Forestry Modules tested in tity collected, and the proportion consumed, sold or other (e.g. given, Indonesia and presented in this paper, aim to provide exactly this kind lost or not yet consumed) over the past year. This is followed by a of information: by collecting detailed data using a comprehensive ap- ‘top-down’ approach of asking the household to determine the percent- proach to forest-people dynamics. age of the total household income that is derived from collection of for- est products and capture of wild animals during the last year. This 3. Site description approach can lead to issues of imprecision, as the concept of percent- ages can be elusive and difficult to comprehend by some respondents The Forestry Modules were pilot-tested in February 2015 as a stand- as opposed to ‘bottom-up’ approaches of deriving income proportions alone survey in the Kalis sub-district of Kapuas Hulu district, West Kali- from various, disaggregated sources (e.g. CIFOR, 2008a,b). Furthermore, mantan Province, Indonesia (also known as ‘the heart of Borneo’)(Fig. 1). the survey also collects data on ownership of household assets with a Kapuas Hulu district was selected for its broad variety of land covers and limit of 10 household items, including a chainsaw, but their value or socio-economic conditions, which were favourable for pilot-testing the quantity is not recorded. This both limits the use of the data to assess Forestry Module under a range of conditions.2 In 2014, the National Sta- the capacity of the households to exploit the forest resources, and also tistics Agency calculated the percentage of the population under the impedes any kind of meaningful wealth ranking based on asset holdings poverty line for the Kapuas Hulu district as 10.03% and the poverty (which is needed for many common forest-livelihood analyses, e.g. see line as IDR 323,786 (USD 27.85)3 per capita per month (BPS, 2016). Nielsen et al. (2012)). This region is well known for its tropical rainforest and rich biodiversity. The Indonesian Forestry Survey also contains other interesting data, The landscape in the southern part of the district - where the pilot test- such as on people's knowledge of and involvement in activities in the ing was conducted - is relatively remote, hilly and mountainous. forest area, awareness of forest boundaries, existence of permits for for- Swidden cultivation is the main focus of livelihoods for the majority est product extraction, as well as perceived causes of any detrimental of households in all of the pilot villages, whereby a small area of natural changes in the forest condition. While such data is useful to understand how people are involved with the forest areas around them, it would be 2 Kalis sub-district has also had little previous research or NGO interventions, and was more useful if it was complemented with quantitative approaches to deemed a good site so as to avoid research fatigue. 3 how households are actually relying on their forests (in terms of contri- Exchange rate of 1 USD = 11,628 IDR as of 28 February 2012. (http://www. exchangerates.org.uk/USD-IDR-28_02_2014-exchange-rate-history.html). For West Kali- butions to current consumption), the degree of reliance on forest and mantan Province the rural poverty line was lower and at IDR 294,044 per capita per month wild products as a safety net (in case of shocks and crises), and patterns (IDR 237,928 in food and IDR 56,115 in non-food) compared to IDR 307,789 in urban areas of reliance when households use forest products for gap-filling during (IDR 230,730 in food and IDR 77,058 in non-food) (BPS Kalimantan Barat, 2015). R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28 23
Table 1 Basic characteristics of pilot testing village/hamlet sites.
No. Hamlet (village) Land cover description No. of Hh in the Hamlet/village Approx. distance to nearest Approx. dist. to Ethnicity name hamlet (based on population market (to the sub-district district capital of our definition) capital of Nanga Kalis) Putussibau (km)
1 Rantau Bumbun - Lowland and hill forests 40 162/394 41 51 Majority (Rantau Bumbun) - Mosaics of old and young fallow secondary Dayak forests Orung - Logged-over hill and lowland forests Da'an - Heath forest - Swidden fields 2 Tilung (Nanga - Mosaics of young fallow secondary forest 133 815/1196 34 44 Majority Raun) - Shrub and low fallow regrowth Dayak - Swidden fields Orung - Logged-over lowland forests Da'an - Grassland/fern-land 3 Lebangan Luar - Smallholder rubber plantation mixed with 48 236/456 20 30 Majority (Lebangan) secondary forest Dayak - Shrub and low fallow regrowth Orung - Swidden fields Da'an - Mosaics of old fallow/secondary forest 4 Nanga Jeniung - Secondary regrowth swamp forest 107 740/590 11 21 Majority (Semerantau) - Smallholder rubber plantation mixed with Malay secondary forest Kapuas - Shrub and low fallow regrowth Hulu - Swidden fields forest is cleared to grow ‘dry-rice’ and other crops for one or two years, and other forestry-related income sources; forestry-related assets and then the land is left to fallow and re-grow into secondary forests be- (Module B); forests in energy, health and construction (Module C); fore repeating the cycle after a period of up to 20 years. Being located on and forest and wild product use in food shortage and crises (Module the Mandai River, villagers also rely heavily on fishing for both subsis- D). These modules used quantitative income accounting to collect infor- tence and cash income. mation on the monetary contributions of forests and wild products to Four sample villages were selected along the Mandai River using a households (but without accounting for non-forest income sources, detailed vegetation cover map to represent a gradient of development, such as agriculture or fisheries, as this should already be collected by forest cover, and village accessibility, from the most upstream village other household socioeconomic surveys; e.g. LSMS). The main respon- of Rantau Bumbun to the downstream village of Semerantau. Rantau dent of the household modules was the household head or spouse. In Bumbun has high levels of natural forest cover, traditional swidden ag- many instances, both household head and spouse were present during ricultural systems, and poor accessibility. Conversely, Semerantau has the interview and completed or clarified each other responses. In fewer little natural forest, predominantly cultivated landscapes (including instances, another household member (e.g. adult child/child in law, par- smallholder rubber plantations), and was relatively easy to access ent/parent in law) also participated in the survey and provided informa- (close to the district capital) (see Fig. 1). The two middle villages, tion on forest extraction activities carried out exclusively by them. Nanga Raun and Lebangan, have characteristics somewhere in-between 2) The standard community modules: to provide contextual infor- that of the most upstream and downstream villages. Nanga Raun (sec- mation about the site as well as overarching data on use and access to ond most upstream) has predominantly natural forest with patches of resources; several modules were implemented. Through community swidden fields and smallholder rubber plantations. Lebangan (further focus group discussions (FGDs), most important4 forest and wild prod- downstream) is more accessible (being closer to the main road and to ucts for cash and for subsistence (Module A) and a seasonal calendar the nearest town) than Nanga Raun, having more natural forest than (Module B) were discussed. Indeed, as ‘importance’ differs whether it Semerantau, and the landscape is dominated by smallholder rubber is for cash income generation or subsistence, this division is necessary plantations and swidden fields. to appropriately capture the varying roles of forest and wild products. Table 1 presents basic characteristics of the four sample villages. Each community FGD comprised around 10–15 people of both genders, Each village has two hamlets that are physically distant from each who were proposed by other community members at an introduction other, and basically organized as two different settlements; therefore community meeting for their knowledge on the discussion topics. the pilot testing was conducted at the hamlet, instead of the village Through key informant interviews, information on units and pricing level. Only one hamlet - the hamlet that serves as the centre of village (Module C) and community benefits (Module D) were derived. Infor- administration - was selected in each village. The most important prod- mants were village officials and other long-term residents who were ucts vary by village between cash and subsistence, as shown in Table 2. knowledgeable of village events. Thirty households were randomly selected from each of the four 3) The extended household modules: these modules are extra sec- purposefully selected villages (i.e. a total of 120 households), to test tions that can be appended to the standard modules, and comprise de- the Forestry Modules under a range of conditions along the previously tailed questions about forest cover changes and clearance (Module E), mentioned development/forest-use intensity/accessibility gradient. participation in environmental service programs and perceptions of cli- The selected sample represented an average of 47% of the total house- mate change (Module F). holds in the hamlets. 4) The extended community modules: using FGDs, information on the forest institutions governing resource use (Module E) was derived, 4. Assessment of the Forestry Modules survey tool as well as community participation in environmental service programs (Module F). The Forestry Modules in their entirety were pilot-tested: 1) The standard household modules: implemented as stand-alone surveys, these surveys included modules on income (Module A) from 4 Importance was defined by the community, whether for income, consumption or forest and wild products, forest-related wage, forest-related business other. 24 R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28
Table 2 Most important forest and wild products for cash income and subsistence (based on an FGD result with approx. 10–16 village representatives in each site).
Type/Importance Village 1 (upstream) Village 2 Village 3 Village 4 (downstream)
Cash 1 Ironwood Rubber Puri leaf Rubber Cash 2 Semah fish Ironwood Tekam wood Fern Cash 3 Tekam wood Tekam wood Fish Meranti wood Subsistence 1 Ironwood Fish Ironwood Fern Subsistence 2 Animal (boar, deer, mousedeer, roe) Boar Fish Fish Subsistence 3 Rattan Rattan Fish fern Meranti wood
When conducting the standard and extended household-level sur- figures derived from the national statistics. These included the annual veys, the enumerators used a five-level Likert scale to systematically re- per capita poverty line for Kapuas Hulu district in West Kalimantan in cord their observations and impressions about the individual survey 2014 (i.e. USD 334.14 or IDR 3,885,432), the mean per capita agricultur- questions. The results were then analyzed to quantitatively evaluate al income per annum (USD 649.94 or IDR 7,557,456) for agricultural the structure and flow of the interview, the time taken to complete in- households in West Kalimantan in 2013 (SPP, 2013), as well as 2008 fig- dividual survey modules (and total interview length), and to identify ures for total income for rural households grouped as agricultural work- questions that were problematic for the enumerators to deliver or for er households (USD 514.83, IDR 5,986,392), agricultural business the respondents to understand. General observations and timing of households (USD 982.12, IDR 11,420,100), and non-agricultural rural the community modules and the key informant interviews were also households (lowest income brackets5; USD 1154.27, IDR 13,421,796; recorded. BPS, 2016). Since 59.5% of the working age population (above 15 years The main results from the enumerator evaluation of the perceived of age) are engaged in agriculture in West Kalimantan, per capita agri- adequacy of the survey tools are presented in Fig. 2. Generally, the Indo- cultural annual income is a relevant comparison to judge the size of for- nesian pilot tests suggest that rapport with the respondent scored the est-related income earned (SPP, 2013). Income from forest and wild highest in terms of having ‘very good’/‘good’ Likert scale scores. Other products includes quantities and cash or cash-equivalent value of prod- ‘good’ responses included the attentiveness and seriousness of the re- ucts collected for sale or subsistence, using own-reported values for the spondent and the flow of interview. For the ‘fair’ score, length of the in- past 12 months. The analysis of absolute income from forest and wild terview and level of understanding of the concept was scored the products per capita shows marked variation between the sample vil- highest. The more problematic areas (i.e. those receiving ‘bad’ scores) lages (Fig. 3), which as expected correlates with the variation in proxim- were on the language and translation of complex concepts, and the ity, abundance, and types of forest from upstream to downstream structure and sequencing of questions. villages (see Fig. 1 for details of the forest types and land use classifica- The category “level of understanding of concepts” was assessed to tion in the study sites). be mainly ‘fair’ or ‘good’, yet the majority of cases where questions The most upstream village (V1), Rantau Bumbun, is located at the were difficult to answer were a result of concepts that respondents ‘forest frontier’; being surrounded by dense hilly and lowland natural found hard to grasp. Out of a total of 156 recorded cases of difficult forests, with patches of swidden fields and swidden fallows that have al- questions, 104 (67%) were related to respondents not understanding ready grown into varying ages of secondary forests. About 10 km down- the concepts asked about. The majority of these cases were specifi- stream is the second village site of Nanga Raun (V2), which has similar cally related to environmental services and related terms, which types of forest as Rantau Bumbun, except that shrub and grasslands are were new to most of the respondents (as there had never been any located close to the settlement area. Households in these two most up- environmental service related projects in their community before). stream villages rely on the natural forest as an important source of food This resulted in enumerators spending a substantial amount of and timber. Hunting and logging of lucrative Bornean Ironwood time carefully explaining concepts including pollination services (Eusideroxylon zwageri) are common livelihood activities. Villagers from insects in the forest, control of agricultural pests by proximity also collect some NTFPs, such as forest fruits, rattan, and tubers, to forest, and climate regulation by forests, but with little compre- reflected by the proportion of collection occurring in old-growth natural hension. This demonstrates the complexity of transferring concepts forests, which is over 50% for V1 and V2 (Table 3). The most important that may be commonly used in academic and policy discourse to forest and wild products for cash and subsistence identified during the household level, in various forest contexts. focus group discussion in the community Module A (Most Important Other complex concepts that were difficult for respondents to Forest and Wild Products; MIP), also reflect ironwood and tekam comprehend related to adaptation strategies towards climate wood (Shorea spp.) as most important cash products, and various wild- change. In order to include these concepts in the Forestry Module, life and rattan as most important for subsistence. more field testing with new approaches to questioning will be re- For Village 1, the mean per capita annual income from forests and quired. Besides the difficulties in explaining the concept of environ- wild products amounted to USD 1031.42 (IDR 11,993,372). Mean for- mental services, there were only a few other cases of difficult est-related business income per capita was USD 314.65 (IDR questions, which related to respondents' ability to attach prices to 3,658,726) and consisted of mainly trade in forest products (60%; n = forest products or attach a value to assets. Yet, most rural households 10), with logging, traditional medicine and other forest-based activities are capable of giving price and value estimates of collected forest and (NTFP collection) making up (10%, 10% and 20% respectively). For mean wild products. The methodological lessons learnt from these results, forest-related wage income (n = 15), this amounted to USD 158.34 as well as subsequent pilot tests from Tanzania and Nepal, were used (IDR 1,841,230) per capita with 53% of households engaged in forestry to develop the final version of the Forestry Modules (Bakkegaard et transport, 27% in forestry logging, 13% in ‘other forestry’ and 7% in car- al., 2016b). pentry. In total for Village 1, forest-related income per capita far exceeds the mean agricultural income for West Kalimantan and even the mean 5. Forest and wild product contributions non-agricultural rural household income (lowest income bracket) for Indonesia. This fact suggests that forest-related income in national so- 5.1. Income from forests and wild products cioeconomic surveys is not adequately reflected, suggesting that a
Data on the income from forest and wild products, as well as forest- 5 As reference, the highest income bracket annual income for non-agricultural rural related business, wage, and other income were compared to several households in Indonesia amounts to USD 2511.51 per capita (i.e. IDR 29,203,800). R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28 25
Fig. 2. Summary of enumerator assessment of the survey tool. Each of the eight aspects were assessed after every household interview (n = 120); the bars indicate the percentage of interview counts for each category. systematic collection of forest-related income in national surveys may and 30% of forest products collected by Village 4 households were lead to a change in percentages of households that are actually under from secondary/regenerating forests (Table 3). Tapping rubber used to the poverty line. be one of the most important livelihoods in the villages and most of In Village 2, mean per capita annual income from forests and wild the households had small rubber plantations. However many house- products amounted to USD 392.61 (IDR 4,565,294), with mean for- holds left their rubber untapped because of the low rubber price (rubber est-related wage comprising USD 118.03 (IDR 1,372,493), and price had been declining for the past two years from 1 USD per kg to half mean forest-related business less than USD 10 (IDR 110,833; n = that price). For cash income, villagers harvest puri leaf (Kratom Borneo, 2). Forest-related wage occupations (n = 15) were mainly in forest- Mitragyna speciosa) leaves, used for medicinal purposes as an anti-de- ry transport (47%), forest-other (13%), forest logging and processing pressant and pain relief, as well as for recreational purposes. In (33%) and carpentry (7%). Forest-related business was recorded in Semerantau, many men go to Malaysia, illegally, for months to look only two households engaged in carpentry and rubber production. for highly valuable gaharu (agarwood), a dark resinous heartwood Forest-related income here lies below the mean agricultural income used for perfume and incense (the best quality gaharu, called ‘super per capita for West Kalimantan (Fig. 3), but again has the potential to king’, fetches up to USD 3000 per kg). increase the mean household income figures for rural households, if In Village 3, mean per capita annual income from forests and wild forest-related incomes are incorporated into total household income products amounted to USD 120.61 (IDR 1,402,460), with mean forest- figures. related wage comprising USD 87.72 (IDR 1,019,975) per capita annually, The two most downstream village sites, Lebangan (V3) and and mean forest-related business USD 240.10 (IDR 2,791,927) per Semerantau (V4), are located in the lowlands, which compared to up- capita annually. The majority of forest-related wage occupations were stream areas are characterized by forests that are predominantly a in forestry transport (7 out of 8 households, 88%), while the remaining mix of rubber plantations and secondary forests with shrubland and one household concentrated on forestry processing (12%). In this vil- secondary swamp forests. Hunting and consumption of wild animals lage, forest-related business was more common (9 households), with is rare because of the villages' close proximity to other villages and the four households (44%) concentrating on NTFP collection of agarwood, limited natural forest. This was also reflected in the locations of collec- puri leaf and second grade agarwood, another four (44%) involved in tion of forest and wild products, which could occur in more than one rubber process and trade, one household (12%) in carpentry, and one area; 41% of the forest products collected by households in Village 3 household (12%) in trade of other forest product. Compared to the levels
Fig. 3. Annual per capita forest and environmental incomes by village (in bars), compared to means of agricultural income and total income for agricultural households (in USD). 26 R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28
Table 3 Table 4 The percentage of forest and wild products collected in different forest categories by Count of shocks by village and action taken with forest or wild product in shock coping, by households in each village. village.
Old-growth Secondary/regenerating Managed Other non-forest Village Type of shock or crises faced by household in last 12 Grand natural plantation environmental months 1234total V1 52.0 15.0 8.5 24.5 Crop disease or crop pests 8 16 13 18 55 V2 50.2 24.5 5.4 19.9 Large rise in price of food 6 13 17 13 49 V3 25.2 41.3 11.2 22.4 Drought, severe water shortage or floods 3 10 18 14 45 V4 23.1 30.3 24.2 22.4 Livestock died or were stolen 5 8 9 2 24 Large rise in agricultural input prices or large fall in 2119931 sale price for crops of other incomes in Fig. 3, forest-related income contributions would Chronic/severe illness, accident or death of household 9612734 likely be a supplement to the household economy. member In Village 4 mean per-capita annual income from forests and wild Other shocks incl. job loss, business failure, property 6185837 products amounted to USD 184.16 (IDR 2,141,410), with mean forest- loss, damage, theft related wages comprising USD 87.21 (IDR 1,014,127), and again a neg- Total shock count 39 82 83 71 275 Action taken ligible involvement in forest-related business (n = 3) in NTFP collection Missing response 9 41 50 45 145 (66%) and trade of forest products (33%) (mean income being less than 1Sell 1825222489 USD 4 or IDR 45,381). The forest-related occupations (n = 7) were in Timber/wood 14 24 9 17 64 forestry processing (29%), forestry transport (43%), and 14% each in log- Animals 4 1 0 5 10 Leaves, ferns, other 0 0 13 2 15 ging and carpentry. 1.2 Both 8 3 1 12 In module design, the data derived from household modules and the Timber/wood 8 1 1 10 community modules were assumed to resonate with each other well - Animals 2 2 that is, in each site the most important products for cash and subsis- 2 Consume 13 8 7 1 29 tence should have the highest proportion of household engagement Timber/wood 9 8 4 0 21 Animals 1 0 1 0 2 and involvement. However, results showed the opposite: across the Leaves, ferns, other 3 0 2 1 6 four sites, the most important cash products (V1 = ironwood, V2 = Total action count 40 82 82 71 275 rubber, V3 = puri leaf, and V4 = rubber) mentioned during the com- munity module FGD were obtained by less than 50% of households in the standard household module on Income from Forest and Wild Prod- respectively) was also prominent, reflecting the large engagement in ucts. This suggests that participants in the FGD on Most Important For- rubber production and ready markets for their sale: these products est and Wild Products (MIP) module may have defined ‘importance’ may be easily liquidated for cash, perhaps due to proximity to markets. based primarily on the monetary value of products. Moreover, such From this it is clear that forest income plays some role as a safety net cash products, like many other lucrative forest products, may be collect- (Angelsen et al., 2014) although it is not always clear how important ed by a few households trading in that product, as a result of limited ac- forest income is in relation to other coping strategies (Wunder et al., cess or lack of availability of the product. The cumulative value of 2014). To help fill this gap in knowledge, in the Forestry Modules house- mentioned MIPs may therefore also be relatively low, due to the limited holds were asked to rank how important forest or wild products are in participation in product collection. coping strategies, compared to the other coping strategies used in times of shock and crises. Table 5 indicates that over 30% of households 5.2. Forest and wild product role in shocks and crises across the four villages responded that forest and wild products were the most important coping strategy when faced with shocks or crises, The types and amount of shocks faced by households are presented yet correspondingly almost 50% said it was the least important com- in Table 4. The data are split according to the sample villages, which, pared to other strategies. Ranking of forest and wild products differed given the variation between the villages, shows that the Forestry Mod- however along the forest gradient - in Village 1 the large majority of re- ules are reflecting the context-specific patterns of forest use in the dif- sponses stated that forest and wild products were the most important ferent villages. For example, droughts and floods were more compared to other coping strategies, where in Village 3 and Village 4 prominent in the lowland Villages 3 and 4, which is consistent with it was almost exactly the opposite (56% and 58% respectively stating the increased flood occurrence in lowland areas. Another example is that forest and wild products were least important). the difference in frequency of crop disease and crop pests between the most upstream village (Village 1, which has more traditional, low inten- 6. Discussion and conclusion sity agricultural systems) and the furthest downstream village (Village 4, which has larger-scale more intensive agricultural systems). Out of With a more detailed focus on specific forest and wild products, e.g. the 275 counts of shocks experienced among all the surveyed house- recording absolute quantities collected and household resources spent holds, almost 50% of the responses regarding corresponding coping ac- on forest and wild product collection, as well modules on supporting in- tions involved an increased use of forest products, either for own formation such as units and pricing, we made the case that the Forestry consumption, for sale, or both. General patterns show that consumption Modules presented in the above can collect in-depth information, and of forest and wild products was significantly higher in Village 1 (up- do so at scale. The Forestry Modules also enable a closer examination stream, 13 counts) than Village 4 (downstream, 1 count). In Village 1, of the role of forest income in household economies, when the tool is belian (or Borneo ironwood) consumption was often for the construc- used in conjunction with standard LSMS modules. Hence, it can poten- tion of coffins in responses to deaths in the family. Conversely, sale of tially make regional or national income accounts more accurate, and forest products was least common in V1 compared to the other villages thus provide a better quantitative basis for development plans, poverty (18 counts). In Village 3, sale of puri leaf (Kratom Borneo) was frequent alleviation strategies, comprehensive conservation initiatives, and other (13 counts), reflecting the fact that puri grows well in lowland, swampy policies and interventions. areas alongside rivers - an ecosystem only found in Village 3. Other Operationally, we looked at the tool itself, evaluating the various as- wood species are used in response to chronic/severe illness or accident pects of the survey, questions, and sequencing. During the pilot testing, of household member for their medicinal and healing properties. In Vil- the sequence of questions, which attempted to maximize the flow of the lage 4, the sale of rubber and wild animals (ten and five counts, interview through connecting different sections, was found to hamper R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28 27
Table 5 Ranking of importance of forest and wild products in coping with shocks (in percentage).
1 (most important) to 10 (least important) Village 1 Village 2 Village 3 Village 4 Total
Rank % Number % Number % Number % Number % Number
1 most important 57% 21 36% 28 26% 21 17% 12 31% 82 2 14% 5 6% 5 4% 3 1% 1 5% 14 3 3%1 3%2 1%3 43%1 0% 1 5 8% 3 8% 6 23% 16 10% 25 7 1% 1 0% 1 8 9% 7 3% 7 9 1% 1 0% 1 10 least important 16% 6 49% 38 56% 45 58% 40 49% 129 −9 (don't know) – 2 – 5 – 3 – 2 – 12 Total 39 82 83 71 275
the interview process. While most questions were readily understood, consume more forest and wild products in response to shocks com- some involving complex concepts such as environmental services and pared to other villages, and forests also featured in over half of the climate change were difficult to explain and time-consuming to gather household responses as the most important coping strategy compared information for, furthermore questioning the validity of the responses to other coping strategies. This was almost opposite to the situation in given. This led to modifications in the final version of the Forestry Mod- the downstream villages (V3 and V4), where over half of the responses ules, and acknowledgement on the need to further work separately on ranked forests and wild products as being the least important in their methods for valuing environmental services (Bakkegaard et al., 2016b). choice of coping strategies. Similarly, consumption of forest products To scrutinize the effectiveness of the Forestry Modules in captur- in times of shock and crises reduces and forest products were predom- ing forest and wild product income, we analyzed the levels of income inantly sold in the downstream village (V4), reflecting their proximity from forest and wild products, as well as the wide array of forest uses to markets and ability to realize cash values of such products. and benefits. While these questions are time consuming and, in cer- In general the analysis has shown that the Forestry Modules do per- tain sections, difficult, they also capture important income contribu- form well in extracting the expected information according to the “for- tions. The general trend shows that per capita annual income derived est gradient”. In addition, the pilot testing underlines the significance of from forest and wild products follows the “forest gradient”,suchthat differentiating data collection at different spatial and time scales, e.g. the upstream villages are collecting more in terms of value than community vs household and cash vs subsistence, capture of seasonal downstream villages, and also collecting more frequently in the products vs accuracy of a twelve month recall - considerations that old-growth forests. may go unmarked in other sectors. Importantly, the Forestry Modules Moreover, we demonstrated how mean forest and wild product in- needed to consider the balance between survey implementation at the come and other forest-related incomes compared to readily available national scale, and the importance of capturing detail at various scales. information on the poverty line, mean per capita agricultural income, Therefore, the Forestry Modules aim to ensure that the specificities of and mean per capita rural household incomes. Importantly, we demon- forest-related activities and contributions can be captured while strated that, comparatively speaking, mean forest-related incomes ex- guaranteeing their applicability across many forest contexts. Their sys- ceed mean per capita agricultural incomes and even mean per capita tematic implementation will maximize (under constraints of survey non-agricultural household incomes in the lowest income brackets for harmonization and costs) the effectiveness and representation of forest Village 1 (the village closest to the forest). This may indeed demonstrate and wild product environments in national socioeconomic data and that forest-proximate household incomes are being significantly contribute to shaping appropriate national policy that reflects the situa- underestimated: many market-remote, forest-near villages may not tions of households in forest areas. be quite as radically poor as the official statistics so far have told us, when we manage to better take into account their privileged access to good-quality forests and their extractive resources. Furthermore, Acknowledgements mean forest and wild product incomes and forest-related incomes cur- rently exceed the poverty line in Village 1, Village 2, and Village 3. If, as The field-work, analysis and write up of this report were supported suggested above, these incomes are not adequately reflected in house- by the Center for International Forestry Research (CIFOR) as part of the hold statistics, absolute numbers of households below the poverty line Poverty and Environment Network (PEN) project, using funds from the may change. At the very least the forest-related and direct forest and UK's Department for International Development (DFID) KNOWFOR Pro- wild product income are a demonstrably significant proportion of gramme. The development of the Forestry Modules was coordinated by total household incomes, and comparable to other significant house- FAO, specifically Illias Animon under the overall guidance of Eva Muller, hold income sources e.g. agriculture, in cases like Village 1. Combined Director, and Senior Forestry Officers Thaís Linhares-Juvenal and Ewald with earlier scrutiny of the existing national forest surveys, such as the Rametsteiner of the FAO Forestry Policy and Resources Division. The au- Indonesian Forestry Survey, we may conclude that current national thors would like to acknowledge other (current and former) co-mem- tools will not sufficiently reflect the economic contribution of this in- bers in the Steering Group for their guidance: FAO - David Morales, come source in national figures. This has implications for the compre- Anssi Pekkarinen and Adrian Whiteman, IFRI - Arun Agrawal, Heather hensiveness of information that is used to inform national policy McGee, Pete Newton and Lauren Persha, University of Copenhagen - development. Carsten Smith-Hall and Thorsten Treue, and World Bank - Gero Carletto To illustrate the depth of information that could be derived from the and Alberto Zezza (Living Standards Measurement Study) and Daniel Forestry Modules, we presented how households use forest and wild Miller (Program on Forests). We would like to express our gratitude to products in coping with shocks and crises, as well as the diversity of the provincial, district and sub-district forestry officials, to the leaders products used in their coping strategies. Interestingly, the ranking of of the sample villages, and all the households in the villages who so forest and wild products relative to other coping strategies also reflected kindly hosted us in their homes, and willingly cooperated in their par- the geographical “forest gradient”. Households in Village 1 tended to ticipation of the surveys. 28 R.K. Bakkegaard et al. / Forest Policy and Economics 84 (2017) 20–28
References Mamo, G., Sjaastad, E., Vedeld, P., 2007. Economic dependence on forest resources: a case from Dendi District, Ethiopia. Forest Policy Econ. 9, 916–927. Angelsen, A., Larsen, H.O., Lund, J.F., Smith-Hall, C., Wunder, S., 2011. Measuring Liveli- McSweeney, K., 2004. Forest product sale as natural insurance: the effects of household hoods and Environmental Dependence: Methods for Research and Field Work. characteristics and the nature of shock in eastern Honduras. Soc. Nat. Resour. 17 – Earthscan, Edinburgh (UK) (240 pp). (1), 39 56. Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N.J., Bauch, S., Börner, J., Nielsen, M.R., Pouliot, M., Bakkegaard, R.K., 2012. Combining income and assets measures Smith-Hall, C., Wunder, S., 2014. Environmental income and rural livelihoods: a glob- to include the transitory nature of poverty in assessments of forest dependence: ev- – al-comparative analysis. World Dev. 64 (1), S12–S28. idence from the Democratic Republic of Congo. Ecol. Econ. 78, 37 46. Appiah, M., Blay, D., Damnyag, L., Dwomoh, F.K., Pappinene, A., Luukkanene, O., 2009. De- Pouliot, M., Treue, T., 2013. Rural people's reliance on forests and the non-forest environ- pendence on forest resources and tropical deforestation in Ghana. Environ. Dev. Sus- ment in West Africa: evidence from Ghana and Burkina Faso. World Dev. 43, – tain. 11, 471–487. 180 193. – Bakkegaard, R.K., 2013. CIFOR PEN Survey Analysis Consultancy Report (December 2013). PROFOR, 2008. Poverty and Forests Linkages A Synthesis and Six Case Studies. World Bakkegaard, R.K., Agrawal, A., Animon, I., Bosselmann, A., Hogarth, N.J., Miller, D., Persha, Bank, Washington, DC. L., Rametsteiner, E., Wunder, S., Zezza, A., 2016b. National Socioeconomic Surveys: Rayamajhi, S., Smith-Hall, C., Helles, F., 2012. Empirical evidence of the economic impor- – Guidance and Survey Modules for Measuring Multiple Roles of Forests in Household tance of Central Himalayan forests to rural households. Forest Policy Econ. 20, 25 35. Welfare and Livelihoods. FAO, World Bank, CIFOR, IFRI, Rome (2016). Riggott, M., 2014. Micro-data analysis of selected socioeconomic surveys. FAO Consultan- Bakkegaard, R.K., Nielsen, M.R., Thorsen, B.J., 2016a. Household determinants of harvest- cy Report. ing of bushmeat and eru (Gnetum africanum) for cash in the Democratic Republic of Russo, L., 2014. Review of the coverage of forest-related socio-economic issues in selected Congo. Environ. Dev. Sustain. http://dx.doi.org/10.1007/s10668-016-9812-9. surveys. Consultancy Report (March 2014). BPS (Badan Pusat Statistik), 2016. Rata-rata Jumlah Pendapatan perkapita Menurut Shepherd, G., 2012. Rethinking forest reliance. Findings about Poverty, Livelihood Resil- “ ” Golongan Rumah Tangga (ribu rupiah), 2000, 2005, 2008. (Online at: http://www. ience and Forests from IUCN's Livelihoods and Landscapes Strategy. International bps.go.id/linkTabelStatis/view/id/953. Accessed April 2016). Union for Conservation of Nature and Natural Resources, Gland, Switzerland. BPS Kalimantan Barat, 2015. Berita Resmi Statistik Provinsi Kalimantan Barat No. 05/01/ SPP, 2013. Analisis Sosial Ekonomi Petani di Kalimantan Barat: Hasil Survei Pendapatan 61/Th. XVIII (2 Januari 2015). (Online at: http://kalbar.bps.go.id/website/brs_ind/ Rumah Tangga Usaha Pertanian Sensus Pertanian 2013. Badan Pusat Statistik Provinsi brsInd-20150115094925.pdf. Accessed April 2016). Kalimantan Barat (Online at: http://kalbar.bps.go.id/website/pdf_publikasi/Analisis- – Byron, N., Arnold, M., 1999. What futures for the people of the tropical forests? World Sosial-Ekonomi-Petani-Di-Kalimantan-Barat Hasil-Survei-Pendapatan-Rumah- Dev. 27 (5), 789–805. Tangga-Usaha-Pertanian-Sensus-Pertanian-2013-.pdf.AccessedApril2016). Cavendish, W., 2000. Empirical regularities in the poverty-environment relationship of Sunderlin, W.D., Angelsen, A., Belcher, B., Burgers, P., Nasi, R., Santoso, L., Wunder, S., 2005. rural households: evidence from Zimbabwe. World Dev. 28 (11), 25. Livelihoods, forests, and conservation in developing countries: an overview. World – Chao, S., 2012. Forest Peoples: Numbers Across the World. Forest Peoples Programme, Dev. 33 (9), 1383 1402. ӧ Moreton-in-Marsh, U.K. Vedeld, P., Angelsen, A., Boj , J., Sjaastad, E., Berg, G.K., 2007. Forest environmental in- – CIFOR, 2008a. PEN Prototype Questionnaire. (Online at: http://www.cifor.org/pen/ comes and the rural poor. Forest Policy Econ. 9 (7), 869 879. research-tools/the-pen-prototype-questionnaire.html. Accessed February 2016). World Bank, 2008. Forests Sourcebook: Practical Guidance for Sustaining Forests in Devel- CIFOR, 2008b. PEN Prototype Questionnaire. (available online at http://www.cifor.org/ opment Cooperation. World Bank, Washington, D.C. – pen/research-tools/the-pen-prototype-questionnaire.html). World Bank, 2015a. National panel survey 2010 2011. Agricultural Questionnaire Tanza- FAO, 2006. Global Forest Resources Assessment 2005 (Rome). nia (Online at: http://siteresources.worldbank.org/INTLSMS/Resources/3358986- FAO, 2008. Links between national forest programmes and poverty reduction strategies. 1233781970982/5800988-1286190918867/NPS_Agriculture_Qx_English_(Year2)_ Forestry Policy and Institutions Working Paper No. 22. Forestry Department, Food v2.pdf. Accessed March 2016). – and Agriculture Organization of the United Nations, Rome. World Bank, 2015b. National panel survey 2010 2011. Fisheries Questionnaire Tanzania Fisher, M., 2004. Household welfare and forest dependence in Southern Malawi. Environ. (Online at: http://siteresources.worldbank.org/INTSURAGRI/Resources/7420178- Dev. Econ. 9, 135–154. 1294154345427/NPS_Fishery_Qx_English_Year_2.pdf. Accessed March 2016). GTZ, 2004. National Monitoring of Strategies for Sustainable Poverty Reduction/PRSPs. Wunder, S., Luckert, M., Smith-Hall, C., 2011. Valuing the priceless: what are non- vol. 1: Main Report. Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ), marketed products worth? In: Angelsen, A., Larsen, H.O., Lund, J.F., Smith-Hall, C., Bonn, Germany. Wunder, S. (Eds.), Measuring Livelihoods and Environmental Dependence: Methods – Statistics Indonesia, 2014. Census of agriculture 2013. National Figures of for Research and Fieldwork. Earthscan, London, pp. 127 145 fi Households around Forest Areas, 2014. Results of ST2013 Survey XVIII. BPS-Statistics Wunder, S., Börner, J., Shively, G., Wyman, M., 2014. Safety nets, gap lling and forests: a – Indonesia, p. 248 (2 Januari 2015 http://kalbar.bps.go.id/website/brs_ind/brsInd- global-comparative perspective. World Dev. 64 (1), S12 S28. 20150115094925.pdf). Forest Policy and Economics 84 (2017) 29–37
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Challenges to governing sustainable forest food: Irvingia spp. from southern Cameroon☆
Verina Ingram a,b,⁎,MarcusEwanec, Louis Njie Ndumbe d, Abdon Awono a a Center for International Forestry Research, BP 2008, Messa, Yaoundé, Cameroon b Forest and Nature Conservation Policy Group, Wageningen UR, P.O. Box 47, 6708 PB Wageningen, The Netherlands c Lake Erie College of Osteopathic Medicine School of Pharmacy, 1858 West Grandview Boulevard, Erie, PA 16509 – 1025, USA d University of Dschang, Faculty of Agronomy and Agricultural Sciences, BP 222, Dschang, Cameroon article info abstract
Article history: Across the Congo Basin, bush mango (Irvingia spp.) nuts have been harvested from forest landscapes for consump- Received 16 May 2016 tion, sold as a foodstuff and for medicine for centuries. Data on this trade however are sparse. A value chain approach Received in revised form 21 December 2016 was used to gather information on stakeholders in the chain from the harvesters in three major production areas in Accepted 22 December 2016 Cameroon to traders in Cameroon, Nigeria, and Equatorial Guinea, the socio-economic values, environmental Available online 5 January 2017 sustainability and governance. Around 5190 people work in the complex chain in Cameroon with an estimated 4109 tons harvested on average annually in the period 2007 to 2010. Bush mango incomes contribute on average Keywords: Irvingia spp. to 31% of harvester's annual incomes and dependence increases for those further from the forest. Customary rules Value chain governance govern access to resources. Although regulations exist, most trade is illegal, with corruption and collective action Non-timber forest products governing access to markets. The majority of nuts harvested are sustainably collected. Although 51% of the harvest Forest policy is sourced from the forest, trees are also managed on cultivated land. Forest degradation and deforestation threaten Rural livelihoods the species. Policy measures such as linking stakeholders, promoting cultivation, pragmatic regulation, and supporting processer groups may make trade in this forest food more sustainable. © 2016 Elsevier B.V. All rights reserved.
1. Introduction reduction, conservation, and food security (Sunderland et al., 2013). Achieving and balancing these objectives is however extremely difficult Cameroon has persistently had a low level of development and gen- (Kusters, 2009). Growing demand has led to a number of high volume der equality (UNDP, 2013). Almost half of the population lives in rural and value NTFP markets. However wild harvests can intensify stress areas and around 40% of land area is covered by dense humid lowland on wild populations, increasing the possibility of over-exploitation forest, which covers southern Cameroon (de Wasseige et al., 2012)In and possible local extinction (Clark and Sunderland, 2004). Wild popu- this context, non-timber forest products (NTFPs) have been used for lations are also threatened by continued high rates of deforestation and subsistence and trade for centuries (Reader, 1998), and their commerce degradation (0.1% and 0.6% respectively per annum for the decade to appears to be increasing (Lescuyer et al., 2011). In 1997 over 1100 2010) (de Wasseige et al., 2016). traders were involved in the trade of NTFPs from the humid forest Such pressures are illustrated by the use and trade of products zone valued US$ 1.75 million (Ndoye, Pérez and Eyebe, 1997/98). In known collectively as bush mango in the Southwest region, mangue 1998, NTFPs in Southwest and Northwest of Cameroon were valued at sauvage, ndo'o,andandok in Centre, South and Littoral regions, and US$ 19 million (CERUT-AIDEnvironment, 1999) and in 2009, five peké in the East Region of Cameroon. These products originate from NTFPs traded in the humid zone were worth US$ 54 million, employing two species: Irvingia gabonensis, a tree bearing fruits with fragrant, 45,000 people (Ingram et al., 2010; Ingram, 2014a; Awono et al., 2016). juicy flesh and sweet juice, and Irvingia wombolu (also known as dry sea- More people are engaged in NTFP trade than in the formal and informal son mango), a similar tree with smaller, bitter fruit (Tchoundjeu and timber sector (Lescuyer et al., 2011). Although harvesting and trading Atangana, 2007; Oyen, 2007), Both species grow to between 25 m to NTFPs in Cameroon is largely informal and small scale (Eba'a Atyi et 40 m tall and co-exist in the lowland tropical humid forests across Cen- al., 2013), interest in NTFPs has increased due to their role in poverty tral Africa, with the range of Irvingia wombolu extending further east and west (Clark and Sunderland, 2004). The trees are also semi-cultivat- ed, maintained in farms and fallows. Bush mango has ranked among the ☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” 10 most economically important NTFPs in Congo Basin countries, and has published at the journal Forest Policy and Economics 84C, 2017. long been one of the most used and valued NTFPs in Cameroon (Ingram ⁎ Corresponding author at: Center for International Forestry Research, BP 2008, Messa, Yaoundé, Cameroon. et al., 2010; Clark and Sunderland, 2004). Irvingia gabonensis is IUCN Red E-mail addresses: [email protected] (V. Ingram), [email protected] List classified as lower risk/near threatened (needing updating) and (M. Ewane), [email protected] (L.N. Ndumbe). Irvingia wombolu is not listed (IUCN, 2013). Although no range-wide
http://dx.doi.org/10.1016/j.forpol.2016.12.014 1389-9341/© 2016 Elsevier B.V. All rights reserved. 30 V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37 inventories have been carried out, the 1998 IUCN Red List risk assess- Table 1 ment is based on a perception of declining populations due to logging Villages and bush mango harvesters interviewed. operations, the expansion of human settlements and poor natural regen- Region Sub division Village Number of Sample as % of eration (IUCN, 2013). Across Central Africa, products from Irvingia spp. harvesters households in have multiple uses. The oil-rich nuts are used as a popular condiment interviewed village and sauce thickener. Cooking oil is also extracted from the nut, the East Messock Moange le 19 21 juice is used in cooking and wine, the pulp as a dye, the bark and kernels Bosquet have multiple medicinal uses, and the timber is used for construction. Koungoulou 18 26 fi Ngoyla Nkolndong 7 39 The kernels have been traded pro tably since at least the 1970s in Cam- Messok – 525 eroon, when it was estimated to be worth 50 million US$, and have been Messok exported, notably to Nigeria, Gabon, and Equatorial Guinea, and onwards Yokadouma Bandekok 626 in West and Central Africa (Ingram et al., 2010; Clark and Sunderland, Central Polidort Landjoue I 7 23 2004), In the last decade, the nuts have been increasingly processed in Subtotal 6 62 26 the USA and Europe as the active ingredient in herbal, weight-loss health Southwest Ekondo Titi Massore – 625 supplements and cosmetics (Sun and Chen, 2012). Balue A value chain denotes how often economic and financial values Funge 5 25 change and increase with the activities involved in bringing a product Mundemba Meka Ngolo 5 8 Central Besingi 6 15 from the forest, through processing and production, to delivery to Bamusso Ekombe 330 final consumers (Kaplinsky and Morris, 2000). When NTFPs move Moffako from subsistence use to commercialization, the livelihoods of rural Ekombe 520 and urban, Central African and global stakeholders involved, such as Liongo Mamfe Central Besongabang 7 10 harvesters, processors, traders and consumers, become interlinked Egbekaw 11 8 through demand and supply value chain interactions. Concerns have Akwaya Takpwe 3 25 been raised about the sustainability of the specie's trade and need for Bache 3 15 species and ecosystem conservation, particularly to conserve genetic Matene 4 4 variation (Ainge and Brown, 2004). Mbilishie 5 9 Basho 1 8 25 Despite the ubiquitous use of bush mango products in Cameroon Obonyi 1 13 19 and the Central Africa region, much of the information on Irvingia spp. Assam 8 24 harvest, use, and trade concerns only parts of the value chain. Data fo- Kajifu 2 2 cuses on specific geographical areas, with in particular, gaps concerning Bodam 2 2 the main harvest regions (Ndoye et al., 1998). Most data on the species Nyang 7 6 Mukonyong 6 8 and trade is now over two decades old. The importance of Irvingia spp. Eyumojock Ayaoke 6 6 to the Cameroonian economy and its environmental and social value Kembong 7 6 for all stakeholders in the chain has not been fully elaborated. This pau- Upper Bachuo 34 city of information hinders sustainable governance of the chain, partic- Banyang Akagbe Etuko 3 5 ularly given that poverty reduction, sustainable livelihoods, and food Nguti Mungondor 3 8 security are key elements in Cameroon's forestry and poverty reduction Ekenge 3 6 policies. This study aims to fill these gaps by identifying the activities Subtotal 25 134 13 and stakeholders in the bush mango chain from the humid lowland for- South Meyomessala Nkolenyeng 8 9 est zone of southern Cameroon; analysing the economic, social and en- Mengon 4 5 Djoum Minko'o 4 5 vironmental aspects of the chain; and examining how governance of the Endengue 4 5 chain affects its sustainability. Melan 3 3 Miatta 3 3 2. Methods Minkom 3 3 Bengbis Bengbis 4 5 Yemedang 3 3 To contribute to the understanding of the bush mango value chain, a Biwong Bulu Endameyos 13 15 review of the literature, as well as of government trade and permit data, Nkengou 13 15 and internet trading websites, were conducted. A situation analysis was Mvangan Mvangan 22 conducted based on meetings with research organisations, government Ville Zoe Befam 7 8 agents, an NGO, and two NTFP trader associations. This led to 47 villages Biwong Bane Biwong-Bane 33 being selected in the Southwest, South, and East regions as the three Centre main regions where bush mango is harvested for commercialisation in Ngoulemakong Mvanda 6 7 Cameroon. Fieldwork was conducted in between September 2007 and Zoétélé Ebamina I 6 7 fi Subtotal 16 86 6 July 2010. First, rapid eld assessments were conducted in each village Total 47 282 15 to determine if there were harvester organisations in the village and the number of members, of which a proportion were randomly selected for interview. An equivalent number of harvesters who were not mem- used to verify interview data. One tree nursery manager was bers of a group were also interviewed, or if no organisations were pres- interviewed in the Southwest. ent, a sample aiming at 15% of the harvester population in that village The chain was then traced further to markets in the Southwest, was interviewed. Due to the combination of rapid assessment and Littoral, Centre, and Eastern regions of Cameroon, and border snowballing technique used, sometimes the total sample was revised markets in Gabon, Nigeria, and Equatorial Guinea. Adapting Ruiz and was smaller or larger depending upon availability of harvesters in Pérez et al.'s (2000) market typology (Wiersum et al., 2014), 31 mar- the time period. This strategy resulted in 282 harvesters being kets were identified and selected, shown in Table 2. These represent interviewed, shown Table 1. Nineteen focus group meetings were held a mix of small, local markets, close to supply zone (type I), medium in selected harvesting villages, to generate additional socio-economic sized markets of regional importance (type II), large urban markets information. Harvesting and processing activities were observed and with national projection (type III), frontier markets (type IV), and V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37 31
Table 2 Bush mango retailers and wholesalers interviewed.
Country Region Market Market type Wholesalers interviewed Retailers interviewed Total interviewed
Cameroon Southwest Kembong I 5 5 Fiango II 1 5 6 Muea, Buea II 8 8 Limbe II 3 9 12 Tiko/Idenau IV 2 2 Mutengene I 1 3 4 Mamfe/Ekok II/IV 9 9 Kumba II 7 7 East Messock II 6 6 Lomié II 1 1 Yokadouma II 6 3 9 South Lac, Ebolowa II 4 4 8 Central, Ebolowa II 25 13 38 Oyenga, Ebolowa II 4 2 6 Sangmélima II 4 3 7 Biwong-Bane II 3 3 Abang Minko IV 2 3 5 Kye-Ossi IV 5 5 Ménguikom IV 6 3 9 Littoral New Bell, Douala III 3 3 Sanaga, Douala III 3 13 16 Centre Mfoundi, Yaounde III 3 6 9 Etoudi, Yaounde III 2 2 Mbalmayo II 5 2 7 Subtotal 5 24 97 90 187 Nigeria Ebonyi Abakaliki IV 16 16 Cross River Ikom III 5 5 Watt Market, Calabar III 4 4 Anambra Onitsha II 5 5 Akwa Ibom Oron III, IV 5 5 Equatorial Guinea Bitam Ebébeyin IV 5 5 Gabon Estuaire Libreville III 2 5 7 Subtotal 6 7 42 5 47 Total 11 32 139 95 234
export specialised, high transaction value international markets multiplied by the total number of stakeholders estimated as active (type V). An internet search was conducted to identify exporters at the relevant stage in the chain. through trading sites in 2010. In Cameroon, Nigeria, Equatorial Guinea, and Gabon, meetings were held with local authorities 3. Results and discussion responsible for forestry, trade, phytosanitary issues, and customs, allowing a rapid assessment of the number of traders. Meetings 3.1. The bush mango chain from southern Cameroon: activities and with market managers, unions, and officials, and snowballing during stakeholders interviews were used to samples between 10 and 25% of traders for interviews in each market, resulting in 139 wholesalers and 95 re- At least 5190 people are directly employed in the bush mango value tailers being interviewed, shown in Table 2. chain originating from southern Cameroon (shown in Fig. 1). These in- Interviews were conducted in English, pidgin, French, and local clude harvesters and various types of traders (known as buying agents dialects as appropriate, and were guided by structured question- or intermediaries, wholesalers, exporters, and retailers in markets). naires tailored to the chain activities. This enabled data to be obtain- The number of associated indirect stakeholders which include trans- ed on the stakeholder's household demographic and value chain porters (head load, porters, taxis, carts, trucks, and canoes), informal fi- characteristics, seasonal activity calendars, and qualitative and nancial organisations, exporting agents, retail market organisations, quantitative data on economic, social, governance, and environmen- non-government support organisations, nurseries, regulatory officials, tal aspects of the stakeholders' involvement in the bush mango chain and consumers is unknown. Fig. 1 also shows how the chain is com- in the period 2007 to 2010. Local field units for measuring bush prised of five main channels, flowing from the harvesting areas in the mango (such as cups, kombos – a two litre container, five litre Southwest, East and South regions. One route flows to local, rural buckets, and sacks) were converted into metric units and the finan- small volume (Type I) and medium sized, domestic markets (Type II). 1 cial values were standardised. Data were quantitatively and statisti- Higher volumes are channelled to urban markets (Type III). The fourth cally analysed using Excel and SPSS (v.16), and qualitatively route major volume exports route extends via transit markets such as analysed using content analysis. Profit margins were calculated Ebébeyin (Equatorial Guinea), Abang Minko, Kye-Ossi and Ménguikom using mean total revenues less actual total costs provided by respon- (on the Cameroon, Gabon, Equatorial Guinea border), Tiko and Idenau dents and do not include labour unless hired labour was paid for. Av- ports (Cameroon), Abakaliki (the largest market encountered in the re- erage prices and costs are based on reported quantities, costs, and gion) and Oron (Nigeria) (Type IV) and subsequently to other smaller prices. The standard deviation (SD) from the average is reported. Es- urban markets (Types III and II) in the neighbouring countries. The timates of total values and volumes at stages in the chain are calcu- fifth route is via exporters to companies processing into health products lated based on averages of the quantity harvested and traded for sale in shops and via internet globally, particularly via American and European sites (Type V). The structure of the chain and the gender of the 1 The exchange rate as of February 2010 was 1 Nigerian Naira to 2.94 Central African stakeholders involved is similar to reports in the 1980s (Falconer, 1990) Franc (FCFA) and 1 US$ = 500 FCFA. and 1990s (Ruiz Pérez et al., 2000). 32 V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37
Fig. 1. Stakeholders and channels in the Irvingia spp. chain from southern Cameroon.
An estimated 2673 harvesters (52% of total stakeholders) in the 47 household of seven, primary school educated, travelling on average villages gather fallen bush mango fruits. Either the fruits are piled into 4 km (5.3 km in East, 3 km in the Southwest, and 3.4 km in the heaps and left to rot, and after several days the nut is extracted from South), into the forest to harvest, and has been harvesting for the fleshy mesocarp and cracked using (stone, stick, etc.), and left to 18 years. Many ethnic groups collect, of three groups in the East, the dry for a day. Or the fruit is split using a machete or simple cutting ma- most dominant group were Baka (52%); in the Southwest of 9 ethnic chines to reduce injuries and halve processing time, and the nut extract- groups Banyangi (30%) predominated, and of 10 groups in the South, ed. The nuts are then transported, generally manually by head load or Bulu (67%) were most numerous. On average, 29% of harvesters were canoe, to a forest-edge point of sale or to the village for storage and pro- members of a group (24% in Southwest, 59% in South and 3% in East). cessing (on average 20% also process the nuts into paste, noting that a In the South, 64 common initiative groups, 6 enterprises, 12 coopera- lack of knowledge and technology is the main constraint to adding tives, and 2 NGOs were encountered, with a total of 1099 members value). Eleven storage methods were used, including sun drying, who generally sell collectively. Harvesters tended to sell individually smoking, and then storing in dry sacks, on open racks, pressed into in the other regions, with three harvester cooperatives found with mud on the walls of houses, in glass jars and by adding insecticide. around 60 members in the East, and two cooperatives with around 40 The methods have different levels of effectiveness at enabling the nuts members in the Southwest. to be kept from three to twelve months and facilitate long distance About 943 traders are estimated to be active in the chain, 18% of total trade. Despite this, between 5 and 8% of total harvest was reported as stakeholders. The term trader includes individuals known as agents or lost due to mould, insects, and vermin. Mainly men transport the ker- intermediaries, who visit villagers to buy from harvesters or their nels to markets, using bikes, motorbikes, or trucks. Women were largely groups. They travel on average 20 km in East, 24 km in Southwest, reluctant or unable to travel the often long distances to markets (aver- and 18 km in the South to sell to buyers, who often provide them age 9.4 in East, 4.3 km in the Southwest and 1.5 km in South) due to with capital for the purchase and transport. An estimated 199 whole- farm, home, and family obligations. Collection is seasonal, with Irvingia salers, 4% of total stakeholders, are mostly based in urban centres wombolu harvest peaking in the dry season (December to March) and where they operate storage facilities. They buy in large quantities Irvingia gabonensis harvest peaking in the rainy season (June to Septem- from harvesters and intermediaries and sell to retailers and exporters. ber), with a large variability in amounts harvested per season and for in- Traders are average 25 (SD 4) years of age, male (92%), either bachelors dividual trees from year to year. Overall, 65% of harvesters are women, or married, have a household of eight people and secondary schooling. with variations per region: harvesting in the Southwest is conducted In the Southwest, 83% and 55% in the East were members of trader's by both men (49%) and women (51%), although in the East, 84% and unions and cooperatives, through which they share information on in the South 29% of harvesters are men. When harvesting occurs during sources, prices and markets and can obtain credit. Retailers sell in school vacations, families camp in the forest for several weeks to harvest bulk, but more often in smaller quantities directly to household con- the nuts, particularly in the Southwest and in Baka villages in South re- sumers. There are an estimated 1375 retailers, 26% of total stakeholders, gion. A typical harvester is middle-aged (38 SD10), married with a the majority (81%) are female, older and live nearby the urban areas V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37 33 they work in. About 50 exporters operate in the Cameroon chain, whom powder. A bucket (3.5 kg) of kernels generally produces eight 400 g were almost as likely to be male (58%) as female (42%), and were on av- balls, sold at prices varying between 500 and 1000 FCFA each. Based erage older (38 SD 80), secondary educated (27% with university), mar- on the average selling village price of kernels, the value of trade for har- ried, had worked for 2 years in the trade and travelled on average 70 km vesters from the three regions in Cameroon is estimated at for their work. In Ikom and Oron markets in Nigeria, all the traders 1,175,121,208 FCFA (2,335,242 US$). Strong seasonal price variations (Cameroonian and Nigerians) were members of an association which were found, with on average 24% higher selling prices after the peak specified the villages where members can approach Cameroonian har- season. There were also large price variations per village, from 168 to vesters and traders, provide credit, tax members, specify common mea- 1714 FCFA per kg, reflecting differences in harvester's bargaining suring units, and allocate market spaces. power, knowledge of market prices, distance from markets, and mem- bership of groups. Harvesters in the South who were members of 3.2. Economic aspects of the bush mango value chain groups earned on average 856 FCFA per kg (1.71 US$) compared to 826 FCFA (1.65 US$) per kg for those working individually. In markets An estimated 345 tons of bush mango kernels were harvested annu- near borders, prices are influenced by demand and supply as well as ally from the 47 villages studied in East, Southwest and South regions in the unions and associations of buyers operating in the markets. Price- southern Cameroon (Table 3). Table 3 highlights the large variance in setting buyer's unions were challenged by the ‘Market Information Sys- quantities produced per village, which is attributed to access to major tem’ in 2008 which enabled new entrants to the trade and higher selling roads and markets, with 82% of the high producing villages correlating prices and margins for individual traders, especially at remote locations. positively with having easier access to major roads. Access to higher In the markets, similar price variations were found both seasonally densities of trees in the forest also plays a role, with the average annual (with a 30% difference between the peak season and after the peak), quantity of bush mango collected per harvester in the East over three and between markets. This price variation is similar to the situation times higher than that collected in the Southwest. found in the same markets in the 1990s (Ndoye et al., 1997/98). Differ- On average 4109 tons of bush mango are estimated to have been ences between markets reflect also distance to the main production lo- traded annually in the 24 markets studied in southern Cameroon in cations, proximity to main consumer markets in Cameroon and the the period 2007 to 2010 (Table 4). This confirms the market type classi- demand based on the size of their urban populations, and proximity to fication, with large urban markets of Douala and Yaoundé being retailer markets and export routes regionally, with the border markets of Kye consumer hubs, the largest volumes are sold wholesale in type IV fron- Ossi, Douala and Yaoundé having the highest average prices. The aver- tier markets in the Southwest and East regions. The average quantity age wholesaler market price was 1696 FCFA per kg (3.4 US$) and aver- sold per trader again varies widely, shown by the large variation, with age retail price was 1938 FCFA per kg (3.8 US$). Based on the average a small number of large scale traders (over 1 ton) in the Southwest market prices per region, the market value of bush mango is estimated and East. The difference of 3764 tons between the quantity traded in at 4,801,062,134 FCFA (9,602,124 US$). Southern markets and the amount harvested in the villages shown in In the Littoral, Centre and South regions, at least 13 small-scale en- Table 3, is assumed to be the result of bush mango sourced from other terprises and groups process kernels by grilling and then grinding into villages in the harvest regions in southern Cameroon. paste, which is then moulded into a ‘cake’ or ‘balls’, and then ground An estimated 3125 tons of kernels were exported annually from into powder. When crushed the kernels produce oil or a ‘butter’, Cameroon to Nigeria 431 tons via Bamenda (in the Northwest), which is also processed into soap and alcohol. These products are Kumba and Mamfe and into Idenau, Limbe markets in the Southwest. retailed in markets and from shops. On average, a 75% increase in Approximately 2391 tons flowed from the South via Mbalmayo, value was created when kernels were transformed into cakes (selling Sangmélima, Ebolowa markets and on to Equatorial Guinea and Gabon for an average of 1250 FCFA per 400 g cake or ball in the South and Cen- (302 tons) via exporters in Messock, Akom, Kye-Ossi and Abang tre), and 83% value addition when powder was sold in 100 g sachets. Re- Minko. The search of internet based traders indicated 33 sellers of tailers in Cameroon reported that consumers mainly purchased bush bush mango kernel powder and extract from Cameroon, offering quan- mango for use in cooking sauces, with the quantity purchased varying tities varying from kilograms to tons. Exports through the port of Douala between 3.2 and 14.1 kg. and airports were not registered by customs authorities. The Ministry of Forestry and Wildlife's annual lists of special forestry products report 3.3. Socio-economic aspects of the bush mango value chain only that one permit for 100 tons was given in 2008 and 54.6 tons were exported in 2009. The Ministry's export trade database On average harvesters have been collecting bush mango for 10 years (COMCAM) did not report that bush mango was exported in the study (SD 8) with the longest 40 years, and 39% within the last five years, period. Historical reports also reflect regional and international trade: confirming the long history of use and trade indicated in literature. Falconer (1990) reported that 78,000 tons was exported to Nigeria in Bush mango sales were ranked by 37% of harvesters as their primary 1986 and Tabuna (1999) that an estimated 5 tons of dried kernels source of household income (39% in Southwest, 44% in the South and were exported to French and Belgian retail outlets in 1998, mainly 28% in East). Incomes from bush mango increased further along the from Cameroon. Ingram (2014b) reported that the European export chain towards retailing, as did dependence upon bush mango incomes trade continued and appeared to have grown since the late 1990s. as a proportion of total income for traders and exporters (Table 5). Harvesters sold bush mango kernels at between 458 and 916 FCFA The average annual income for harvesters from bush mango was per kg (0.9 to 1.8 US$), with an average selling price in harvesting vil- 145,945 FCFA (SD 59,756) (US$286). Regional income variations were lages of 989 per kg. This is equivalent to 30 to 50% of the average retail evident: in the South, individuals earnt on average 213,206 FCFA (426 buying price. Value is increased by processing into ‘cake’ or ‘balls’ and US$) whilst members of groups earnt on average 98,433 FCFA (196
Table 3 Annual quantity of bush mango harvested in southern Cameroon 2007–2010.
Annual quantity harvested for period 2007–2010 (kg) East Southwest South Total
Average quantity per harvester household 199 (SD 109) 74 (SD 67) 50 (SD 85) 108 Average quantity per organization ––1442 (SD 753) 1442 Average quantity per village 8896 4532 6662 6697 Total harvested in villages studied 71,010 113,304 161,021 345,335 34 V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37
Table 4 Annual quantity of bush mango sold in markets in southern Cameroon 2007–2010.
Average annual quantity harvested (tons) for period 2007–2010 East Southwest South Littoral Centre Total
Average per trader 13.65 (SD 52) 718.11 (SD16) 76.8 (92 SD) 2.47 (1 SD) 62.13 (0.2SD) 873.13 Estimated total for markets sampled 1570 51 1986 9 492 4109
US$). The difference is attributed to some of the largest harvesters Nkwatoh and Weh (1998) for harvesters in Manyu, in the Southwest, working alone, and groups being comprised of very small scale har- suggesting the trade is now more profitable. Wholesalers also made a vesters, and in areas with seemingly low densities of bush mango significant profit, ranging from 1086 in the East, to 942 FCFA per kg in trees. The groups however generally earn above average selling prices. the Southwest. Traders in Tiko market registered a loss in 2007 due to Harvester households had an average of six sources of income: with ag- high losses caused by boat capsizing and delays, leading to higher pay- riculture the most important (33% of incomes), followed by harvesting ments for storage, and profits of 1665 FCFA per kg in 2009. This finding between 9 and 20 other NTFPs (25%), trade (13%), livestock (11%), emphasises the high risk nature of wholesaling and transport, due to and hunting (10%) (Table 5). Bush mango contributed on average 31%, long distances, poor transport infrastructure, and high levels of corrup- ranging from 28 to 44% of harvester's total household incomes. This tion. Retailers in the Southwest had a profit margin of 852 FCFA per kg, finding signifies a dependence upon forest resources, and highlights di- in the South 1215 FCFA and Centre 1004 FCFA per kg. Ndoye et al., versified household income generating strategies, and is similar to other (1998) calculated weekly profit margins from bush mango in 28 mar- studies of NTFP harvesters in Cameroon (Nkwatoh, 2005; Sunderland et kets in the South, Centre and Littoral regions in 1995/1996 at 3800 al., 1998; Guedje et al., 1998; Walter and Mbala, 2006; FAO, 1999). FCFA, indicating that the retail trade is also now more profitable. Whole- Wholesale traders earned on average 4,602,764 FCFA (US$ 9205). Re- salers and importers in Nigeria enjoyed higher margins of 2840 FCFA tailers earned on average 273,436 FCFA (US$ 546) from bush mango, per kg and into Equatorial Guinea and Gabon 2790 FCFA per kg. MISPEG, which contributed to 57% of their income on average, and like traders an enterprise in the Southwest processing kernels into balls and power, wide variations were found between regions (Table 5). Exporters reported a profit margin of 3780 FCFA per kg. Twelve other organisa- earned the most, on average 28,582,986 (US$ 57,165), and were the tions were unable to provide cost data to calculate profits, but indicated most dependent, having on average two sources of income. that processing was profitable. Bush mango was as likely to be used for direct household consump- tion, as it was to generate cash income, with Bantu harvesters in the 3.4. Environmental aspects of the bush mango value chain south selling the highest proportion of their harvest (Fig. 2). On average, three main uses of two species were recorded: the kernel for use in On average, 51% of the bush mango collected in southern Cameroon cooking, as a condiment and for oil, and the bark for various medicinal was harvested from wild sources (29% from primary, dense forest and uses. A larger number (six) of medicinal uses of different tree parts 22% from secondary forest), and almost half (49%) from cultivated were noted by Baka families in South. Although the tree is also used lands (farms 28%, fallows 19%, and home gardens 2%). Regional differ- for timber, known as ‘andok’ (Ayuk et al., 1999), none of the respon- ences are evident, shown in Fig. 3. These differences are attributed to dents mentioned harvesting the timber for commercial or own use. several factors. One is the variation in population density, as the highest The kernels are also valued as gifts and for barter in all areas. Income densities of Irvingia spp. occur in primary forest, followed by secondary from bush mango was used similarly by harvesters in all regions, mainly forest, cocoa farms, and swampy areas (van Dijk and Wiersum, 1999). on education (35%), basic household needs including agriculture (33%), Second is a cultural dimension, particularly for ethnic groups in the healthcare (20%), food (19%), clothing (14%), and saving (2%). Traders East, where domestication is uncommon, particularly the semi-nomadic used income similarly on education (32%), household needs (12%), Baka. In the southwest in contrast, 85% of harvesters from the Banyangi health (6%), and food (8%), but spend more on purchasing transport ethnic group had planted bush mango. Third, the finding reflects the (22%) and land (11%) (Fig. 2). greater extent of domestication in the Southwest, where harvesters Bartering bush mango occurred particularly in the East by the Baka have a longer history of maintaining the trees as part of agroforestry fal- ethnic group. The kernels were exchanged for household basics such low systems which were previously forested (Ainge and Brown, 2004). as rice, kerosene, batteries, locally distilled wine, soap, and cloth. How- In the Centre and South most bush mango is wild, reflecting work by ever, the exchange rate is lower than the equivalent cash price, shown Leakey et al. (2004). NGOs and research organisations have run domes- in Table 6. tication trials of Irvingia wombolu and set up tree nurseries, with 10% of The bush mango trade is generally profitable, although margins vary harvesters in the Southwest and 4% in the East reporting planting bush widely per region and the stage of the chain, from an average of 1198 mango trees. Reasons for not cultivating bush mango include not know- per kg in the Southwest to 638 per kg in the East. Losses of between ing how (35%), a lack of land (11%), and seeing no need as a sufficient 493 FCFA and 1033 per kg also occurred for harvesters in three villages amount could be found in the forest (51%). Tabot (2008) also found (Bandekok Polidort, Landjoue I and Moange le Bosquet), due to the high that most (93%) bush mango was harvested on farmland in Manyu Di- cost of transport and the low bartering exchange rate. Harvester's vision in the Southwest. Fourthly, regional differences are also attribut- profits in the Southwest are higher than the 230 FCFA/kg reported by ed to the slightly higher levels of forest degradation in the South and
Table 5 Stakeholder's annual average incomes from bush mango (FCFA) and number of household income sources 2007–2010.
Region Income Harvester Wholesaler Exporter Retailer Average
Southwest Average 98,950 1,703,480 11,455,400 80,203 10,045,146 No sources 6 4 3 5 4 South Average 213,206 9,496,175 57,137,000 176,754 10,934,698 No sources 3.84 2.8 0 3.5 2 East Average 125,723 2,608,640 17,156,559 563,530 6,757,457 No sources 6 6 4 5 5 Average Average 145,960 4,602,765 28,582,986 273,495 8,401,30 No sources 6 4 2 5 4 V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37 35
Fig. 2. Harvester's use and sale of bush mango in southern Cameroon.
East (Robiglio et al., 2010). In the 1990s Irvingia spp. were reported as commerce and quarantine officers. Wholesalers generally did not have being mainly sourced from the then lesser disturbed forests in southern permits, as even if they did possess one, corruption was prevalent. The Cameroon (Ntamag, 1997). Reports of increased distances travelled to so called ‘informal taxes’ amounted to 4% of trader's costs, averaging the forest to harvest (10% of harvesters in the Southwest and South, 257 FCFA per sack per journey and 24% of exporter's costs to Nigeria, and 15% in the East), indicate that forest degradation is an increasing on average 2998 FCFA per sack, and 15% of exporter's costs to Guinea issue and that increasing forest clearance poses a risk at species popula- and Gabon at around 6000 FCFA per truck. tion level. Customary tenure and ownership rules dominate governance ar- Harvest methods were generally regarded by harvesters as sustain- rangements concerning access to the species in practice. Bush mango able, as only fallen fruits are gathered and trees are usually left or active- trees within forests are not owned by individuals or families and access ly managed on farmland. Harvesters noted that fruiting is highly is generally on a first-come, first-serve basis. About 99% of harvesters in- variable from year to year and demand is generally higher than supply, dicated they did not require prior authorization from any authority be- presenting a risk that insufficient fruits are left in forested areas for fore harvesting. However, particularly in the Southwest, families tend to regeneration. harvest in the same area each year, constructing ‘bush houses’ for the harvest season, indicating a tacit acknowledgement of ownership and 3.5. Governance of the bush mango chain access rights within communities. Trees planted or maintained on farm- land are owned by the landowner, with access restricted without per- The trade in bush mango is partly governed by regulation. Irvingia mission. As bush mango was reported to have increased in value, spp. are not classed as a ‘Special Forestry Product’ according to Decision some people have cleared land around these trees in the forest. This ex- No 336 of 2006, which sets annual exploitation quotas for products of tension of tenure through clearance usually relates to farmland, but re- economic and/or environmental interest, granted by the Ministry of sources from retained trees are also considered to be owned by the Forestry and Wildlife. The legal basis for the 1000 FCFA per sack ‘tax’ family that cleared the land. ‘Outsiders’ and large scale buyers generally that many traders reported paying at checkpoints to both Forestry and register with village authorities, pay a fee to traditional authorities, and Wildlife and Agricultural sanitary health officials during transport sometimes to the family landowner. Rates differ, from 2393 FCFA (US routes was not established, indicating that it is a form of corruption. $4.7) plus an additional gift of palm wine per harvest, to 5000 FCFA From 2005 to 2010, quotas were granted for 120 and 100 tons to one (US $10), to 5600 FCFA (US $11.2) for the whole season in the South- company in each year and agreements (gré à gré) were granted by Min- west. These fees have increased by up to 50% since 2001 (Sunderland isterial orders to seven companies for a total of 111 tons in 2004 and for et al., 2002). The encroachment of customary forested land by Nigerian six companies for 60 tons in total in 2005. These permits are demand communities was a considerable cause of conflict in Obonyi I, Basho, based and do not reflect the supply of Irvingia spp., as despite the re- Matene and Mobilise villages in the Southwest. Such conflicts led the quirement of the 1994 Forest Law there has never been an inventory Mbilishe people to plant both Irvingia species. Other traditional controls of available stock. Way bills, the administrative paperwork used to include prohibitions on felling individual bush mango trees, climbing track Special Forestry Products and the permit allocation, did not trees to harvest, and collecting unripe fruit. match the volumes found in this study nor the quotas allocated, indicat- Market-based, collective action was the most common arrangement ing a large discrepancy between practice, the regulatory framework, governing access to markets, with conditions for exchange, prices and and its enforcement. Interviews with traders and exporters indicated trade terms negotiated, particularly through traders and exporters widespread corruption during transport nationally and particularly at unions. For example, the Technical Operations Unit was set up for the border posts by police officers, forestry officials, customs, council, Takamanda area in the Southwest to decentralize decision-making
Table 6 Cash value (FCFA) of barters of bush mango in Baka villages, South region.
Village Average selling price per kombo Average selling price by barter per kombo Loss incurred per kombo by barter Loss per kg
Moange le Bosquet 1250 431 819 493 Landjoue I 1050 100 950 572 Bandekok Polidort 2000 285 1715 1033 Average 1576 272 1233 699 36 V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37
Fig. 3. Source of Irvingia spp. harvests.
and establish forestry and village committees. However financial and Brown, 2004) and the highly variable extent of cultivation in different administrative power and authority were not transferred in practice. regions in Cameroon (Degrande et al., 2006). The IUCN red list status Projects such as Government-GIZ Programme for the Sustainable of Irvingia spp. merits further research and updating, given concerns Management of Natural Resources in the Southwest Province that a higher level is protection is necessary, and policy and practical ac- (PSMNR-SWP) set up development committees to provide an interface tions if its status is confirmed as threatened. between the population and other natural resource users and guardians. Governance issues are critical in the bush mango value chain, The village management committees and community forests (adjoining reflecting that of other NTFPs in Cameroon and Central Africa (Ingram, Takamanda national park) introduced new co-management arrange- 2014a, 2014b). Weak formal regulations and fragmented customary ments, aiming to strengthen governance arrangements and devolve governance and the largely informal nature of trade are exacerbated power and authority. These however also increased trading and gover- by a policy framework that undervalues the contribution of NTFPs to nance complexity, caused coordination problems, and created new op- livelihoods and creates a disenabling environment for sustainable ex- portunities for (mis)appropriation of power, authority, and financial ploitation. The lucrative commerce and high demand, provide benefits and forest resources. Rules were introduced that prohibited harvesting to over 5000 people directly in Cameroon, and additional indirect stake- in the newly-gazetted Takamanda National Park, which were feared to holders associated with the trade, such as transporters. The chain em- negatively impact livelihoods. bodies a classic trade-off between livelihoods and conservation As traders reported obtaining information primarily through col- (Kusters, 2009). The substantial reliance upon bush mango incomes leagues and from trading in different markets, and harvesters reported by those engaged in the chain is threatened by deforestation and degra- obtaining information mainly from buyers, an FAO funded project (see dation. Although bush mango trees are often maintained in farms and Acknowledgements) aimed to influence chain governance by providing fallows, forest clearance for non-agricultural use reduces the species information on market prices and demand. The perceived advantages of population, reducing the gene pool. The trade is further weakened by accessing market information were increasing prices, profits, and vol- formal and customary regulations which inadequately ensure sustain- umes, particularly for remote communities and marginalized ethnic able access to wild resources, do not sufficiently promote cultivation, groups. Traders, NGOs (such as the Fondation Fritz Jakob (Fonjak), the do not protect the most vulnerable stakeholders at the beginning of Centre pour la Protection Durable de l'environnement (CEREP) and the chain, and provide perverse incentives in the form of quotas IUCN) and the media (Network of Community Radios of Cameroon which are demand based, rather being based on any understanding of and Radio Environment) set up a Market Information System in 2008, the species and available supply. Stakeholders in the chain, due to its which enabled new entrants and higher selling prices, especially at re- fragmented nature, also have little national level knowledge and com- mote locations (Fon, 2010; SNV, 2010). Although this floundered after municate little information on resource availability, as the emerging the project ended, remnants of a market information system continued market information system concerns only prices. to be present in 2010, aided by increasing mobile phone coverage and Recommendations for policies to address these problems include (1) use. revising the regulatory framework to better capture and monitor bush mango permits, particularly at major markets and border crossings, 4. Conclusions (2) government promotion and guidance on sustainable harvesting techniques, and (3) collaboration with customary authorities who The sale of kernels and processed products from wild and semi could help enforce regulations in forest areas. Together with private sec- cultivated bush mango (Irvingia spp.) makes a substantial contribution tor and development partners, other ways to implement a more sus- to the livelihoods of actors throughout the value chain in southern tainable policy include furthering collective action, particularly at the Cameroon. Because the majority of stakeholders use revenues to meet harvester level, improved market information, disseminating more effi- basic household needs, the development implications of this trade are cient and safe processing and storage techniques, promoting domestica- significant. Increasingly, however, this commerce appears unsustain- tion and cultivation, and support to strengthen producer associations able. The unsustainable nature of the value chain is indicated by declin- and enterprises, with a focus on vulnerable ethnic groups and remoter ing wild resources reported by experienced harvesters travelling longer communities. Pilots have shown such measures to be effective in the distances, combined with continuing high demand, clearance of the bush mango chain in the study area (Tieguhong et al., 2009; Awono, species' natural forest habitat, low levels of cultivation, and a lack of con- 2005; Awono et al., 2013; van der Goes and Ngueko, 2009; Leakey, sistent regulatory control and enforcement. This situation has arisen de- 2011; Degrande et al., 2006). Improved coordination between the Min- spite long running concerns of over-exploitation (World Conservation istries of Agriculture, Forestry and Wildlife, Commerce and Customs Monitoring Centre, 1998; Clark and Sunderland, 2004; Ainge and with direct stakeholders in the chain, and indirect actors such as V. Ingram et al. / Forest Policy and Economics 84 (2017) 29–37 37 research and development NGOs concerning trade and distinguishing IUCN, 2013. IUCN Red List of Threatened Species. IUCN. Kaplinsky, R., Morris, M., 2000. A Handbook for Value Chain Research. IDRC, Canada. cultivated from wild Irvingia spp. in permitting could further enhance Kusters, K., 2009. Non-timber forest product trade. A trade-off between conservation and sustainable governance. development. Assessing the outcomes of non-timber forest product trade on liveli- hoods and the environment, with special emphasis on the damar agroforests in Sumatra, Indonesia. Faculty of Social and Behavioural Sciences 233. University of Acknowledgements Amsterdam, Amsterdam. Leakey, R.R.B., 2011. NWFPs: cultivation - the key to better agriculture and improved live- We thank all the stakeholders interviewed, also Yves Minlo, Martin, lihoods. Non-Wood News 22, 3–4. Michel, Hilaire, Daniel, Bruno Mvondo, Léopold Aimé Ndongo, Leakey, R.R.B., Tchoundjeu, Z., Smith, R.I., Munro, R.C., Fondoun, J.-M., Kengue, J., Anegbeh, P.O., Atangana, A.R., Waruhiu, A.N., Asaah, E., Usoro, C., Ukafor, V., 2004. Evidence that Ousseynou Ndoye, Henri Owono, Athanasius Nkwatoh Fuashi, Arend subsistence farmers have domesticated indigenous fruits (Dacryodes edulis and van der Goes, Maurice Schill, Honoré Tabuna, Raoul Ngueko, Francis Irvingia gabonensis) in Cameroon and Nigeria. Journal Agroforestry Systems 60 (2), – Sangwa, Norbert Sonné and Fidele Mekongo for their collaboration. 101 111. Lescuyer, G., Cerutti, P.O., Mendoula, E.E., Ebaa-Atyi, R., Nasi, R., 2011. Chainsaw milling in The study was funded by CIFOR as part of the European Commission the Congo Basin. European Tropical Forest Research Network News 52, 121–129. funded project GCP/RAF/408/EC Mobilisation et renforcement des Ndoye, O., Pérez, M.R., Eyebe, A., 1997/98. The Markets of Non-timber Forest Products in capacités des petites et moyennes entreprises impliquées dans les the Humid Forest Zone of Cameroon. Rural Development Forestry Network, no. fi Network Paper 22c. lières des produits forestiers non-ligneux en Afrique Centrale. We are Ndoye, O., Ruiz Pérez, M., Eyebe, A., 1998. The Markets of Non-timber Forest Products in grateful for anonymous reviewer's valuable comments. the Humid Forest Zone of Cameroon. Rural Development Forestry Network, no. Net- work Paper 22c. 25. Nkwatoh, A.F., 2005. Production and marketing of Non Timber Forest Products NTFPs, a References major source of rural and national income in the South West Province of Cameroon. 17th Commonwelath Forestry Conference (Colombo, Sri Lanka). Ainge, L., Brown, N., 2004. Bush mango Irvingia gabonensis and I. wombolu. In: Clark, L.E., Nkwatoh, A., Weh, F., 1998. Final Report for the NTFPs Market Survey for Zone 3 ed. M.C. Sunderland, T.C.H. (Eds.), The Key Non-timber Forest Products of Centre Africa. State Project,40.MCP,Buea. of the Knowledge. Central African Regional Program for the Environment, Oxford, Ntamag, C., 1997. Spatial distribution of non-timber forest product collection. A case – pp. 15 28. study of south Cameroon. Department of Forestry Wageningen. Wageningen UR. Awono, A., 2005. Women and ntfp trade in cameroon: reconciling research and capacity Oyen, L.P.A., 2007. Irvingia wombolu (Vermoesen). Wageningen, Netherlands: PROTA building for development. Livelihood Program Cameroon 28. CIFOR, Yaounde, (Plant Resources of Tropical Africa/Ressources végétales de l'Afrique tropicale). Cameroon. Reader, J., 1998. Africa: A Biography of the Continent (Penguin UK). Awono, A., Ingram, V., Schure, J., Levang, P., 2013. Guide for Small and Medium Enter- Robiglio, V., Ngendakumana, S., Gockowski, J., Yemefack, M., Tchienkoua, M., Mbile, P., prises in the Sustainable Non-timber Forest Product Trade in Central Africa Bogor. Tchawa, P., Tchoundjeu, Z., Bolognesi, M., 2010. Reducing Emissions From all Land CIFOR, Indonesia. uses in Cameroon. 110. ASB Partnership for the Tropical Forest Margins, Yaounde. Awono, A., Eba'a Atyi, R., Foundjem-Tita, D., Levang, P., 2016. Vegetal Non-timber Forest Ruiz Pérez, M., Ndoye, O., Eyebe, A., Puntodewo, A., 2000. Spatial characterisation of non- Products in Cameroon, Contribution to the National Economy. timber forest products markets in the humid forest zone of Cameroon. Int. For. Rev. 2 Ayuk, E.T., Duguma, B., Franzel, S., Kengue, J., Mollet, M., Tiki-Manga, T., Zenkeng, P., 1999. (2), 71–83 (147, 148). Uses, management and economic potential of Irvingia gabonensis in the humid low- SNV, 2010. Système d'Information sur les Marches (SIM) des Produits Forestiers Non – lands of Cameroon. For. Ecol. Manag. 113, 1 9. Ligneux au Cameroun; Association des exploitants des Produits Spéciaux au Ebolowa. CERUT-AIDEnvironment, 1999. The wealth of forests in Cameroon. Results of Field Testing Project GCP/RAF/408/EC Mobilisation et Renforcement des Capacités des Petites et a Methodology for the Valuation of Non-timber Forest Products (NTFPs) in North- Moyennes Entreprises impliquées dans les Filières des Produits Forestiers Non West and South-West Cameroon 137. CERUT and AIDEvironment. Ligneux en Afrique Centrale 1. FAO-CIFOR-SNV-World Agroforestry Center-COMIFAC, Clark, L.E., Sunderland, T.C., 2004. The key non-timber forest products of Central Africa: Yaoundé. state of the knowledge. SD Publication Series Technical Paper No. 122 185. U.S. Agen- Sun, J., Chen, P., 2012. Ultra high-performance liquid chromatography with high-resolu- fi cy for International Development, Of ce of Sustainable Development, Bureau for tion mass spectrometry analysis of African mango (Irvingia gabonensis) seeds, extract, Africa, Washington D.C. and related dietary supplements. J. Agric. Food Chem. 60 (35), 8703–8709. de Wasseige, C., de Marcken, P., Bayol, N., Hiol Hiol, F., Mayaux, P., Desclée, B., Nasi, R., Sunderland, T.C., Clark, L.E., Vantomme, P., 1998. Non-wood Forest Products in Central – Billand, A., Defourny, P., Eba'a Atyi, R., 2012. Les Forêts du Bassin du Congo Etat Africa; Current research issues and prospects for conservation development. In: fi des Forêts 2010. 276. Of ce des Publications de l'Union Européenne, Luxembourg. Sunderland, T.C., Clark, L.E., Vantomme, P. (Eds.), International Expert Meeting on de Wasseige, C., Tadoum, M., Eba'a Atyi, R., Doumenge, C., 2016. Les Forêts du Bassin du Non-wood Forest Products in Central Africa 288. FAO, Limbe, Cameroon. Congo. Etat des Forêts 2015 - Forêts et changements climatiques en ligne. Weyrich, Sunderland, T.C.H., Besong, S., Ayeni, J.S.O., 2002. Distribution, utilization and sustainabil- Belgique. ity of non-timber forest products from Takamanda forest reserve, Cameroon. Consul- Degrande, A., Schreckenberg, K., Mbosso, C., Anegbeh, P., Okafor, V., Kanmegne, J., 2006. tancy Report, ed. PROFA, 37. Protection of the Forests Around Akwaya (PROFA). Farmers' fruit tree-growing strategies in the humid forest zone of Cameroon and Sunderland, T., Powell, B., Ickowitz, A., Foli, S., Pinedo-Vasquez, M., Nasi, R., Padoch, C., – Nigeria. Agrofor. Syst. 67 (2), 159 175. 2013. Food security and nutrition. The role of forests. Discussion Paper. Center for In- Eba'a Atyi, R., Lescuyer, G., Poufoun, J.N., Fouda, T.M., 2013. In: CIFOR (Ed.), Étude de ternational Forestry Research (CIFOR), Bogor. l'importance économique et sociale du secteur forestier et faunique au Cameroun Tabot, P., 2008. Carbon Sequestration in Agro Forestry Systems and the Socio-Economic 315. CIFOR, Yaounde. Evaluation of Some Woody Non-timber Forest Products in Southern Cameroon fi ‘ ’ Falconer, J., 1990. The Major Signi cance of Minor Forest Products: The Local Use and Buea. University of Buea. Value of Forests in the West African Humid Forest Zone. Food and Agriculture Tabuna, H., 1999. 'Le marché des produits forestiers non ligneux de l'Afrique Centrale en Organization Of The United Nations (FAO), Rome. France et en Belgique: produits, acteurs, circuits de distribution et debouches actuels' FAO, 1999. Données statistiques des produits forestiers non-ligneux du Cameroun. 32p. – Programme de partenariat CE-FAO (1998 2001) Ligne budgétaire forêt tropicale Tchoundjeu, Z., Atangana, A., 2007. Irvingia gabonensis (Aubry-Lecomte ex O'Rorke). B7-6201/97-15/VIII/FOR PROJET GCP/INT/679/EC 36. FAO/European Commission PROTA (Plant Resources of Tropical Africa/Ressources végétales de l'Afrique Directorate-General Viii Development. tropicale), Wageningen, Netherlands. Fon, J., 2010. Pilot Phase Market Information System for NTFPs in North West Region (Oc- Tieguhong, J.C., Ndoye, O., Tchatat, M., Chikamai, B., 2009. Processing and marketing of tober to December 2009). Project GCP/RAF/408/EC Mobilisation et Renforcement des non-wood forest products: potential impacts and challenges in Africa. Discov. Capacités des Petites et Moyennes Entreprises impliquées dans les Filières des Innov. 21, 60–65. Produits Forestiers Non Ligneux en Afrique Centrale 13. SNV North West, FAO- UNDP, 2013. In: UNDP (Ed.), Human Development Report 2013 Cameroon 5. UN, New CIFOR-SNV-World Agroforestry Center-COMIFAC, Bamenda. York. Guedje, N.M., van Dijk, H., Nkongmeneck, B.A., 1998. Ecologie et exploitation de quelques van der Goes, A., Ngueko, R., 2009. Market Information System as a contribution for social – produits forestiers non ligneux (PFNL) de la forêt humide du sud Cameroun. inclusion in NTFP value chain. In: SNV (Ed.), XIII Congrès forestier mondial 8. FAO, Séminaire FORAFRI. Libreville CIRAD. Buenos Aires, Argentina. Ingram, V., 2014a. Win-wins in Forest Product Value Chains? How Governance Impacts van Dijk, J.F.W., Wiersum, K.F., 1999. Non-timber forest product resources: abundance, the Sustainability of Livelihoods Based on Non-timber Forest Products From distribution, and the impact of timber exploitation. In: T. International (Ed.), Sustain- Cameroon. African Studies Centre, Leiden. able Management of African Rainforests. Tropenbos, Kribi, pp. 200–210. Ingram, V., 2014b. Win-wins in forest product value chains? How governance impacts the Walter, S., Mbala, S.M., 2006. Etat des lieux du secteur ‘Produits Forestiers Non Ligneux’ en sustainability of livelihoods based on non-timber forest products from Cameroon. Afrique Centrale et analyse des priorites politiques. Note d'Information 34. European Amsterdam Institute for Social Science Research, Governance and Inclusive Develop- Commission (EC). Food and Agricultural Organisation of the United Nations (FAO), ment Group 361. University of Amsterdam, Amsterdam. COMIFAC, Malabo, Guinée Equatoriale. Ingram, V.J., Ousseynou, N., Midoko Iponga, D., Chupezi Tieguhong, J., Nasi, R., 2010. Non- Wiersum, K.F., Ingram, V.J., Ros-Tonen, M.A.F., 2014. Governing access to resources and timber forest products: contribution to national economy and strategies for sustain- markets in non-timber forest product chains. Forest, Trees and Livelihoods 23 (1– able management. In: de Wasseige, C., de Marcken, P., Bayol, N., Hiol Hiol, F., Mayaux, 2), 6–18. P., Desclée, B., Nasi, R., Billand, A., Defourny, P., Eba'a Atyi, R. (Eds.), The Forests of the World Conservation Monitoring Centre, 1998. Irvingia gabonensis. The IUCN Red List of fi Congo Basin - State of the Forest 2010. Publications Of ce of the European Union, Threatened Species. 1998. Luxembourg, pp. 137–154. Forest Policy and Economics 84 (2017) 38–46
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Origin products from African forests: A Kenyan pathway to prosperity and green inclusive growth?☆
Henrik Egelyng a,⁎,AskeS.Bosselmannb, Mary Warui c, Fredah Maina d,JohnMburud, Amos Gyau e a University of Copenhagen, Department of Food and Resource Economics, Global Development Section, Denmark, Rolighedsvej 25, 1958 Frb. C, Denmark b University of Copenhagen, Department of Food and Resource Economics, Denmark, Rolighedsvej 25, 1958 Frb. C, Denmark c University of Nairobi, Land Resource Management and Agricultural Technology, Box 29053-00620, Nairobi, Kenya d University of Nairobi, Department of Agricultural Economics, Box 29053-00620, Nairobi, Kenya e World Agroforestry Centre, Kenya, United Nations Avenue, Gigiri, Box 30677, Nairobi, Kenya article info abstract
Article history: Many tropical countries have potential for adding market value to unique forest origin products similarly to how Received 10 May 2016 EU gain billions of Euro's annually from registering agricultural origin products, with Protected Denomination of Accepted 2 September 2016 Origin or Protected Geographical Indication. Following analysis of the renaissance for the global Geographical In- Available online 21 September 2016 dication (GI) regime, this article provides case-studies from Kenya – on Mwingi Honey, Kakamega Silk and insti- tutional conditions under which producers may incorporate territory specific cultural, environmental, and social Keywords: qualities of their unique products. We investigate prospects for Kenyan producers to create and capture addition- Geographical indications Forest products al monetary value for their forest related origin products, allowing smallholders to build livelihood, while Honey stewarding natural environments. The origin products are investigated for their potential for protection with a Policies GI, within five different dimensions of and links with the social and natural world. Our study shows that Mwingi Wild silk Honey and Kakamega Silk have potential for registration under a GI regime based mainly on close links between local environment, flora and product quality, and product specificity. The institutional environment presents major challenges for the development of GI products and markets, exemplified by the Kenyan GI bill which is not yet enacted after almost a decade in the making. © 2016 Elsevier B.V. All rights reserved.
1. Introduction products include Parmigiano-Reggiano cheese, Parma ham, Darjeeling tea, Champagne and Tequila. The GI regime in the EU has added billions Originproducts,i.e.productswithspecific quality traits attributed to of Euro worth of monetary value to an exclusive selection of high quality their geographical origin, from forests and other unmanaged areas can European food products (EC, 2012). A renaissance for GI products create substantial monetary values to households in rural areas in de- (Egelyng et al., 2015) provides the global south with both challenges veloping countries, provided the origin qualities can be protected and and opportunities, translating into a second chance for nations not al- marketed under registered product labels. One opportunity for protec- ready benefitting from GIs as economic policy instruments and institu- tion is afforded by Geographical Indication (GI), which originates from tional vehicle for rural development through value addition to small the French concept terroir and is based on the link between a product's scale food producers. Grounded in development studies and drawing quality and the geographical and human environment in which it has on recent studies of the global regime for geographical indications, been produced. GI is a special type of intellectual property right, recog- this paper investigates two forest products in Kenya - Kakamega Forest nized by the World Trade Organization in 1994, which reflects the mon- Silk and Mwingi Honey – and draw out options for countries and stake- etary value of cultural and social qualities of a particular production holders to use the international GI regime to pursue among them a vir- landscape, including it's shaping by producers, i.e. local landscape custo- tuous circle of sustainable development of forest communities and dians (Douguet and O'Connor, 2003). Some of the most famous GI livelihoods, of the kind theoretically envisioned by proponents of GIs, e.g. Teshager (2015;2), seeing GI's as instruments to pursue ´economic, biodiversity, cultural and food security objectives´. The case studies aim ☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” to illustrate why and how the introduction of GIs may be expected to published at the journal Forest Policy and Economics 84C, 2017. positively impact producer livelihoods as well as biodiversity in ⁎ Corresponding author. E-mail addresses: [email protected] (H. Egelyng), [email protected] (A.S. Bosselmann), Kenya. The paper concludes the analysis by summing up the challenges [email protected] (M. Warui), [email protected] (F. Maina), of globalization and presenting an agenda for development researchers [email protected] (J. Mburu), [email protected] (A. Gyau). and agencies to identify and overcome institutional, technical, and other
http://dx.doi.org/10.1016/j.forpol.2016.09.001 1389-9341/© 2016 Elsevier B.V. All rights reserved. H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46 39 barriers for stakeholders wishing to adopt GIs in their struggle for local livelihoods and sustainable rural development. Natural link to Cultural link to quality quality 2. Theoretical framework
Oftenconfusedwithbrandsandtrademarks,GIsarequiteopposite; GI potential where brands and trademarks provide for individual legal entities, GIs are collectively owned and aim to protect collective values and public of product Social ties goods (Augustin-Jean et al., 2012). Where, from an ecological economics Reputation and specificity among perspective, application of the Fordist industrial model to agriculture producers has perhaps de-valorized many a cultural and natural landscape, GIs rep- of product resent a theoretical promise of re-valorizing such landscapes, the realiza- tion of which is however challenged by many institutional factors. Institutional According to Izac et al. (2009), GIs can be seen as economic policy instru- environment ments for sustainable development. The wider implementation of GIs in the global South is expected to impact conservation of culturally and en- vironmentally important landscapes and biodiversity, traditional knowl- Fig. 1. A simplified representation of the five major elements of GI potential. edge and handicrafts, consumer and producer alliances favoring shorter value chains, redistribution of value added, and food sovereignty1 fi (Dagne, 2015; Ilbert, 2012). proliferation of instruments (fair trade, organic certi cation and GIs) fi The development of institutional and policy environments for valo- aiming for redistribution of value in existing value chains, diversi ca- rization of origin products via GIs can be theoretically explained by tion of production and markets, and generation of shorter value chains. the special nature of food markets (Steiner, 2012), the multidimension- The assessment of the institutional component of GI's may be structured ality of food quality (Allaire, 2012) and arrival of the modern ´quality as an analysis of the broader institutional environment, presenting bar- turn´ consumer, for which food origin provides identity and ´consumer riers, challenges and opportunities for a GI to succeed in a given nation power´ articulated as ´market preference´ for specialty products; all fac- or case. Institutional theory is core to such an approach. While devel- fi tors which in some cases have existed for centuries and in some cases oped in the different context of organic product certi cation, an exam- have emerged or re-emerged recently. ple of such a broader institutional approach can be found in Egelyng et An analysis of the concrete potential of any origin product to be reg- al. (2013), providing an analytical framework for analyzing the institu- fi fi istered under a GI regime, such as the EU quality regime, may be struc- tional environment of certi ed organic production displaying ve di- – tured based on the requirements listed in existing legislation and GI mensions: i) overall policies, ii) regulation in particular conformity registrations of products. Fig. 1 shows the major elements of GI poten- assessment systems; iii) research, education and extension that targets fi tial, identified through a review of products that are registered in the certi ed organic agriculture; iv) agency and roles of the private sector EU DOOR database with Denomination of Origin (PDO) or Protected and Civil Society Organizations (CSOs); and v) a broader contextual Geographical Indication (PGI), case studies of GI products outside the analysis, i.e. the overall agrarian and rural development strategy. For EU, and the FAO/SINER-GI report ‘Linking People, Places and Products’ this paper, the focus is mainly on i) and iv). (Vandecandelaere et al., 2010). In Fig. 1, GI potential is made up of i) the natural link, i.e. the natural 3. The GI renaissance - a new global regime for geographical setting, environmental and climatic conditions etc. of the area of produc- indications has emerged tion which is held to affect the quality attributes of an origin product; ii) the human link, i.e. the cultural environment, cultural heritage, traditions The milestone for the new international GI regime is after 1992, and history, and local know-how that affects the product, e.g. though cer- when EU created its mechanisms for registration of Protected Designa- tain production and processing activities; iii) social ties, such as trust and tion of Origin (PDO) and Protected Geographical Indication (PGI) within cooperation among producers, and collective efforts found in coopera- the European Union and after 1995, when the WTO gave its 154 mem- fi tive or other types of producer associations and groups; iv) reputation ber states a single de nition of GIs (article 22.1). Geographical indica- and specificity of the product, linking consumer awareness of the prod- tion, as a concept, traces much further back. The academic literature uct to its specific quality and characteristics, which is an important pre- often traces origin products back in history to ancient Egypt and Greece, requisite for GI success; and v) institutions, which refer to formal and before also mentioning France and China; a 1915 international exposi- informal rules governing the production and marketing of the product, tion gold medal winning product originating from Zhejiang province and is affected by the presence of local NGOs, state authorities issuing in South East China, tracing its history a thousand years back regulations, extension staff, research bodies etc. Together, the natural (Guihong, 2012). Similarly, an international GI regime goes back to and human links align with ‘place and people’ or the concept of terroir, the Paris Convention for the Protection of Industrial Property (1883) resulting in tangible and intangible product quality attributes that cannot and Madrid agreement for the Repression of False or Deceptive Indica- be reproduced outside the original area (Mancini, 2013). tions of Source on Goods (1891). As shown by a wealth of European studies, and recently by Filoche & Agricultural development is at a global crossroads and choosing ag- Pinton (2014) in the case of Brazilian guanará, the conditions for appro- ricultural institutional pathways is an imperative (Izac et al., 2009). It is priation and use of an origin product is a very complex matter institu- not only that achieving food sovereignty without seriously compromis- tionally, with many stakeholders, legal plurality and political and ing critically important ecosystem services remains a global concern economic circumstances often favoring actors within global value (Grafton et al., 2015); and not only that the World Bank already called “ ” “ ” chains who are committed to strategies of so-called agricultural mod- for a visible hand to create markets for environmental services ernization and industrial processing. These actors stand in contrast to (World Bank, 2007). It is also that African countries are called upon to actors committed to ´ecologization´ of agriculture/forestry and exploit markets for origin products (AU-EC 2012). In November 2012, the African Regional Intellectual Property Organization (ARIPO) and the European Commission signed a cooperation agreement to improve 1 The concept of food sovereignty ´speaks to the right of states to maintain and develop their own capacity to produce their basic foods, respecting cultural and productive diver- the legal protection of traditional agricultural products in Africa sity´ and is a ´precondition to genuine food security´ (Dagne 2015,81) (ARIPO and EU, 2012). In 2013, Oku White Honey and Penja pepper 40 H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46 from Cameroon and Ziama-Macenta coffee from Guinea were the first marketed products derived from forests and unmanaged areas. Howev- three African food products to be awarded PGI by the African Organiza- er, in the pursuit for inclusive green economic growth in rural areas, a tion of Intellectual Property (OAPI), displayed at the Paris International new opportunity may be found in creating awareness of GI qualities Agricultural Show the same year. This renaissance is a major reason why and building local capacities and an institutional framework to support additional and renewed ´global´ research on GIs is highly relevant, also GI product and market development. This is for example being promot- in the context of forest communities. ed among women producers of Shea butter in Benin (IP Watch, 2013). The basis for a renaissance of the international GI regime was laid Collection of Shea nuts in open woodlands and processing into Shea when the rift between the new world (Americas, Australia & New butter is a traditional income generating activity among women who Zealand) and the old world (EU) culminated and led the EU to adopt a often do not own their own land. A GI registration of the Shea butter, new council regulation (510/2006 and 1898/2006). This move opened based on i) the local texture, color and aroma of the butter, ii) the im- up the EU GI system to all countries: a move making the European GI re- portant traditional role of Shea butter in the local culture, as well as gime far more compatible with WTO rules, consequently strengthening iii) on the traditional collection and processing of the nuts, would be a its position as an international regime, sensu Krasner (1983), and re- means to valorize women's agricultural efforts, bring them closer to placing Josling's (2006) ‘transatlantic war on terroir’ with a new inter- economic autonomy and help to conserve the resource (IP Watch, national GI regime of more constructive competition. A remarkable 2013). result is the increasing number of products primarily from Asia with Though the majority of GI products are agricultural products, there PGI registered in Europe since 2012. Outside Asia, only a few products are number of forest and wild products2 that have been registered or have so far been registered by the EU. While the EU only recently started proposed as GI products, ranging from hard wood timber species and the discussion on whether to extend GI protection beyond food and forest vegetables to alcoholic beverages and handicrafts made from for- drinks, and include products like Scottish tartans (Press release Brus- est and wild products, see Table 1. In Europe, several GI honey products sels/Strasbourg, 15 July 2014), national GI regimes in countries such as are produced and collected from bees foraging in forest and other un- China, India, South Korea and Colombia also recognize handicrafts, gar- managed habitats and some of these have obtained GI status based on ments and other non-foodstuffs; hats, potteries, woodcarvings, orna- arguments for quality derived from wild habitats (DOOR database). In mental flowers, traditional medicine, tobacco etc. If non-foodstuffs are Korea, no less than 28 forest products, such as different kinds of mush- also included in the EU GI regime, it will open up for even more EU reg- rooms, wild vegetables and medicinal plants, were registered with GI in istrations of GI products from the Global South. Another result of the 2010. In Ethiopia, while the country has no specific GI legislation (Hirko, new international GI regime is the institution of AU-EU collaboration 2014), several forest products such as wild forest coffee, wild pepper, on promoting GIs in Africa (AU-EC 2012), including initiatives such as and other spices from the forest have been suggested as possible GI the ´Power of Origin´ conference in Uganda 2011. As it already happened products (Roussel & Verdeaux, 2007). In Morocco, several GI products with the globalization of the market for certified organic products are derived from natural areas with limited management, such as desert (Egelyng et al., 2013 and Egelyng, 2007), African producers of origin euphorbia honey, wild rosemary, and argan oil from the Argan forest re- products can be expected to catch up with global trends, following the gion, an UNESCO Biosphere reserve (Le Maroc Vert, 2015). The specific general trend of an increasing number of countries with affluent con- case of GI products from forest and other primarily unmanaged areas sumers demanding high quality food products with specific attributes is that often many small producers and rural communities are involved in terms of culture and means of production. However, so far African in the collection and marketing of the products. The market premium products are underrepresented among the world's GI products, even associated with the GI label of the products may thus be an important though many African countries have had legal regimes for GI protection addition to household economies and local livelihoods. of products for a number of years and many being active in either OAPI Two of the products in Table 1 are among the first three African GI or ARIPO. The Quakers UN Office lists 47 African countries with GI re- products registered by OAPI, which governs a common GI framework gimes in 2007, 23 of which have specific GI protection laws while the re- for 17 African countries. Oku White honey, Penja Pepper and Ziama- maining operates under trademark laws (Musungu, 2008). With the Macenta coffee were PGI registered in 2013 after efforts by the French legislative frameworks being further developed across Africa and the research institute CIRAD and funding from the French Development Second Decade of CAADP (2015–2025) as a possible influential force Agency (AFD). One of the PGI products, Oku White Honey, is produced on this development, opportunities arise for tapping into the GI product by groups of small-scale honey producers and marketed through coop- potential that is ‘hidden’ in the diverse cultural and natural environ- eratives. Oku White honey is produced in a unique ecosystem (distinc- ments of Africa. This can lead to increases in numbers of African GI prod- tive vegetation and favorable climate) in the nationally protected forest ucts and growth in regional GI markets, which is currently spearheaded of Kilum Ijim near Mount Oku in Cameroon (Coulet & Mahop, 2012). The regionally by countries like Morocco, which has 35 products registered pollination by both wild and domesticated bees of the unique floral di- under the national GI legislation, in place since 2008 (Le Maroc Vert, versity, dominated by two white flower plants, Schefflera abyssinica and 2015). In other African countries, several products have already been Nuxia congesta, yield a high quality honey with a characteristic white proposed as potential GI products, such as Kalahari Melon Seed Oil color and slightly acidic flavor (WIPO, 2014). The Oku honey has been (Southern Africa, incl. Namibia), Madagascan Vanilla, Mauritius Deme- known in Cameroon for a long time (Blakeney et al., 2012), but after rara Sugar, Shea butter from the Massi plateau (Burkina Faso), and Zan- PGI registration in 2013, the honey is now facing export markets in Eu- zibar Cloves (Tanzania). rope and North America. The marketing is on the rise and expanded in- come streams are already benefitting a substantial number of local 4. Forest origin products and new opportunities for rural livelihoods beekeepers. For sustainable production and maintenance of the quality of the Oku honey, the producer group members in collaboration with Products collected in forests and unmanaged areas often play an im- non-governmental organizations are making efforts to conserve the for- portant role for the livelihood of households in rural areas. These prod- est, create awareness of ecosystem among local communities, and have ucts not only support the households' own consumption and provide a reforested degraded areas (Bainkong, 2014). The GI registration of the safety net during crisis and as a response to shocks, but may also be a Oku White Honey includes a GI standard with rules governing the spe- means to accumulate assets and provide a path out of poverty cific honey chain from production to retail. Since the GI registration, (Angelsen et al., 2014). The latter role of forest products may be strengthened for those products that have qualities linked to origin. As 2 By forest and wild products we refer to products collected in natural forest and other shown in Fig. 1 there are a number of important elements of GI poten- unmanaged areas. This includes timber and non-timber forest products such as fruits, tial, some of which are often not currently documented in the case of plant parts (tubers), fauna and products derived local fauna, such as honey and wild silk. H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46 41
Table 1 List of products from forests or other unmanaged areas, proposed or registered with GI protection. The products have been identified through online searches and review of literature and documents; therefore the list should not be seen as exhaustive.
Country Product GI registration Reference
Benin Shea Butter Proposed IP Watch, 2013 Cameroon Penja pepper OAPI1 Chabrol et al. 2015 Oku honey Colombia Aguadeño hat & other handicrafts Appellation of Origin laws Pérez Villareal, 2012 Ethiopia Wild Forest Coffee, e.g. Heranna Proposed Mengistie et al., 2012 Forest pepper Proposed Roussel & Verdeaux, 2007 EU Honey from wild habitats EU GI law DOOR Database Gabon Okoumé wood Proposed Coulet & Mahop, 2012 India Nilambur Teak In process Vasu 2015 A number of handicrafts Indian GI of Goods Act Gautam & Bahl, 2011 Indonesia Javanese Teak Proposed Dwi et al., 2006 Mexico Mezcal from wild Agave cupreata Appellation of Origin laws Granich et al. 2010 Morocco Argane oil Law no. 25–06 Distinctive Signs of Origin and Quality Le Maroc Vert, 2015 Taliouine saffron Desert Euphorbia honey Wild rosemary South Korea 5 kinds of mushroom Agricultural Quality Control Act Italian Intellectual Property Rights, 2010 5 kinds of nuts 10 kinds of fruits and berries 5 kinds of medicinal plants 3 kinds of wild vegetables USA Natural wild rice Trademark Babcock & Clemens, 2004
1 OAPI: African Organization of Intellectual Property. prices have increased from 1500 to 4000 FCFA (2.8–7.5 USD) for a liter The interviews and product evaluations aimed to collect data re- of processed honey and several new SMEs have been created (WIPO, garding the five major elements of GI potential as depicted in Fig. 1. 2014). The higher prices are injecting a substantial monetary value The natural and human links were assessed in terms of how strong into the local economy and improving livelihoods of beekeepers and the links were between specific product properties and qualities on processors. one hand and origin on the other. The institutional environment and so- cial ties were assessed according to their supportive and positive influ- ence on the producers' opportunities and capabilities to strengthen and 5. Kenyan cases of forest products with GI potential benefit from the origin-based product qualities. Lastly, the element of reputation and specificity was broadly assessed based on interviews, Among African countries with an interest in developing GI products observations and local researchers' own insights into each sector, as and markets, and yet lacking a fully operational legislative and institu- well product evaluations. Section 5.1 and 5.2 present the two cases tional framework, is Kenya. A GI bill, drafted in 2007, is yet to be enacted based on collected information if no specific reference is mentioned, (Blakeney et al., 2012). The Kenyan Intellectual Property Institute (KIPI) and in section 5.3 the GI potential is presented in Fig. 2. carried out a project in 2009 describing a number of pilot GI products, including products from agriculture (dairy products), agroforestry (cof- 5.1. Mwingi honey –“linking forest biodiversity to sustainable livelihoods” fee) and forest areas (honey and wild silk) (KIPI, 2009). The following sections present new case studies of two forest products, honey and Beekeeping has been practiced in Kenya since the hunting age and wild silk. The studies were carried out in the period 2014–2015 and in- today is a widespread activity in rural areas, contributing to household cluded interviews with a number of actors in the local honey and silk diets and economy. There are an estimated 145,000 beekeepers in sectors and representatives from various organizations, as well as liter- Kenya, most of whom are small scale with less than 10 hives (Carroll ature and document review and preliminary product evaluations. In & Kinsella, 2013). The most suitable areas of beekeeping in Kenya are Mwingi, respondents included ten producers, three processors, five arid and semi-arid areas where rain-fed agriculture is difficult and for- consumers, five marketing agents, one NGO representative, two repre- ests and shrubs dominate. In these areas, honey production is often a sentatives of research institutions and six key informants from official particularly important contributor to livelihoods of rural communities agencies. Also in Kakamega producer and processor interviews (six) (Shiluli et al., 2012). Mwingi sub-county in Eastern Kenya is one such were combined with focus group discussions (see Table 2). area, known for its production of high quality honey with unique floral
Table 2 Data collection methods and informants in each of the two case studies.
Case study Data collection Respondents
Mwingi Honey Interviews –10 producers, 3 processors, 5 consumers, 5 marketing agents, and representatives from the Kenyan Honey Council, International Centre of Insect Physiology and Ecology, National Beekeeping Institute, Ministry of agriculture and Livestock, Ministry of Trade, Industry and Enterprise, certification body Kenya Organic Agricultural Network, the Kenyan Intellectual Property Institute. –4 FGDs with producers –7 samples from Mwingi honey producers, stored in food grade containers at room temperature until shipment to Germany for physico-chemical and pollen analysis by Quality Services International GmbH, Bremen. Focus Group Discussions Honey sample analysis Kakamega silk Interviews –2 processors, 3 silk and clothe traders, 1 representative of the Sericulture Institute of the Kenya Agricultural and Livestock Research Organization. –FGDs with members of 2 wild silk worm producer groups in Kakamega county, and field visits in the B. micrantha growing areas. Focus Group Discussions 42 H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46
The natural The human link to quality /cultural link to quality
Mwingi honey
Reputation Social ties Kakamega and specificity among silk of product producers
Institutional environment
Fig. 2. Qualitative evaluation of the five elements of GI potential in the case of Mwingi honey and Kakamega wild silk. and aromatic tastes according to producers as well as consumers. Honey common challenge in rural Kenya (Muli et al., 2007) This shows that from this area was selected as a pilot GI product by the Kenyan Industri- there is a need for process standard setting, training and capacity build- al Property Institute (KIPI) in 2009; a short-lived initiative that did not ing in the local part of the supply chain in order to ensure all producers lead to further development of a GI framework. Mwingi honey is pro- and processors in the area can supply the quality that is essential from a GI duced by local Apies melifera subspecies that pollinate and forage on perspective. GI registration implies specification of product qualities, the Acacia woodlands dominating the region and which are known to which among others includes detailed requirements for certain process- produce mild and sweet tasting honeys. According to producers, tradi- ing and quality parameters. This is used to safeguard GI honeys from tional log hives, which were observed hanging from branches of the local inferior honey products that fail to meet the quality specified, and local Acacia trees throughout the area, still dominate most apiaries but GI registration of honey in Mwingi would require such specifications. modern beehives are also becoming widespread. Extraction of honey According to representatives from the agricultural ministry and from the combs and further processing is either carried out by bee- ICIPE, the typical woodlands in the area have been under pressure for keepers themselves or, more often, at centralized processing units man- many years from agricultural activities and charcoal production, but re- aged by associations of beekeeper groups. Besides beekeeping, many cently, with more emphasis on improving honey production, communi- people in the area also practice agro-pastoralism, which traditionally ties are starting to realize the importance of a sound environment for have complimented beekeeping activities well. Beekeeping members bees. Reforestation programmes have been underway to recover lost of the Mwingi Honey Market Place, which is a cooperative and the pri- forested areas with community involvement. Training undertaken by mary sales outlet of Mwingi honey, are certified as organic producer, beekeeper groups includes awareness of conserving forests and sur- under the Kenya Organic Agriculture Network, an achievement sup- rounding vegetation where bees forage, and especially areas surround- ported by the institute of Market Ecology in Switzerland in collaboration ing apiaries are protected against degradation. Just like producers of with Kenya's Ministry of Agriculture and Livestock. This has created a certified organic products justify premium prices with a reference to en- wider appreciation and demand for the honey (pers. Com. Honey Mar- vironmental qualities, some GI producers claim to provide ecological ket place Chairman). One type of honey, which is processed in Mwingi, services, such as biodiversity conservation (Allaire, 2012). Local honey is labelled as ´Eco-honey´. It links forest biodiversity to sustainable live- producers in Mwingi are marketing their products based on this same lihoods, indicating that producers are well aware of the role of their bee- idea, which may also be an important part of a GI registration. keeping in the area and use this in the marketing. During field work, In Mwingi, honey producer groups from different locations within several buyers explained how they had traveled from afar elsewhere the area have formed an umbrella association where the production, in Kenya, in order to purchase the honey in Mwingi, supporting local processing, and marketing issues of honeys are addressed. These groups statements of Mwingi honey being broadly perceived as of high quality. invest in equipment and also advocate for capacity building on proper Besides the main elements of a GI product described in the previous beekeeping and honey handling practices in order to improve the qual- section, all GI honey that are registered in the EU door database are de- ity of their produce. The group participates in honey market centers, scribed with floral sources based on pollen analysis, which provide hard where producers, processors, traders and consumers interact. In the evidence of the natural settings surrounding the apiaries, and physio- same area, the research institute ICIPE has technically and materially chemical properties, such as water and sugars content as well as more supported a group of smallholder farmers to undertake beekeeping. specific properties related to food safety. Seven samples of different Through this initiative the smallholders' productivity has been im- honeys was collected among beekeeper groups in Mwingi and analyzed, proved by 10–18% with a corresponding household farm increase of according to standards set by the EU directive for honey (2001/119/EC). 15% (Ayuya et al., 2015; Affognon et al., 2015). The collective efforts of The pollen grain analyses for all samples showed dominating pollen beekeeper groups and ICIPE have facilitated a value addition that has el- content from shrubs and trees typical to the local semi-arid natural set- evated the status of beekeeping in the area, and created incentives for ting, with Acacia sp. often most dominating. Of the seven samples ana- producers to conserve bee habitats to sustain production. The local lyzed, four were classified as table honey according to the EU directive, honey producers explained how communities or clans often have com- i.e. a physio-chemical quality fit for human consumption, while three munal lands, held under trust by county governments and with unre- were classified as industrial honey, i.e. a lower quality only fit for further stricted access to resources for the local people. Beekeepers within processing and as ingredient in the food manufacturing industry. The such local areas therefore often share apiary sites, where each producer lower classification was due to low diastase activity (i.e. low presence has a social obligation to protect the resource in the area occupied by of specific enzymes) and relatively high levels of hydroxymethylfurfural beehives. As such, informal rules have a role in the conservation of bee (HMF, though only one sample exceeded maximum allowed content) forage areas. Local informal institutions, such as taboos and norms, are which are both indications of excessive heating during processing, a also important as explained during the focus group discussion with H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46 43 producers. For instance, the elder men are obliged to preserve the indig- tree and produced the cocoons the tree is left bare, which caused the enous and ancestral trees and they do this by hanging traditional bee people to assume the tree was dead. Today, farming of wild silkworms hives on the branches. Destruction of these trees is regarded as a bad in Kakamega is being promoted as an enterprise that contributes to for- omen. est and biodiversity conservation as well as providing alternative sources of income (Mbahin et al., 2010). With increased sensitization 5.2. Kakamega forest silk – origin specific production for livelihoods and and training mainly from research organizations, sericulture groups forest conservation have been able to adopt wild silk production and now harvest the co- coons. With sleeve nets introduced by ICIPE, the collectors are able to Forest silk production is practiced in several regions in Kenya. How- net the worms and keep them on the tree during the spinning stage, ever, the silkworm species are different and adapted to different ecosys- which not only ensures that more cocoons are produced but also pro- tems and biodiversity. The wild silkworm, Anaphe panda,isendemicto tects the spinning colony on the tree. Once the spinning is complete the Kakamega forest and feeds specifically on the leaves of the Bridelia and the adult moth has left, the cocoons are harvested. The tree is micrantha tree that occurs naturally within the forest. Mbahin et al. then allowed to rejuvenate. More trees are being planted in order to in- (2012) observe that the wild silkworm produces silk of high quality crease the population of the wild silkworm. and it has a longer lifespan in the predominantly indigenous part of The 14 groups involved in wild silk moth cocoons sell their collec- the forest compared to the exotic part. The Kakamega forest silk is de- tions for processing at the Kakamega market place. The groups market scribed as being a shade of brown in color and uneven. Some producers their cocoons collectively, which gives them some bargaining power combine it with the finer mulberry silk fabric, to give the latter a unique in negotiating for prices. However, they are not involved in processing texture. The production of the wild silk is dependent on seasons and the of the silk and their part of the value chain stops once they are paid presence of the B. micrantha tree, the rareness of which limits produc- for the cocoons delivered. The chain is also challenged by general low tion. This uniqueness, coupled with the fact that the silk is soft, lustrous, production volumes as well as a monopsony market situation (emerged and with fibers that are stronger and more elastic compared to the mul- from single project without marketing component) that has so far not berry silk (Kioko, 1998), gives the wild silk place-specific qualities that created real market incentives for investments to increase production. may very well fit a case for GI registration. ICIPE has played an important role in setting up the production and Like many other fabrics from different parts of the world, including still maintain its role as the main buyer. Other existing buyers only pur- Kente cloth (Ghana), Tartans (Scotland) and Thai wild silk, the existence chase in small quantities and are not consistent hence not reliable. A real of the raw material for the wild silk in Kakamega dates back to historical market outside Kakamega to sustain and improve the production is times. However, unlike the above mentioned, which were discovered largely missing, as silk traders in Nairobi and other large urban areas earlier and their potential exploited hence becoming part of the culture, are not sufficiently aware of the production. JAICAF (2007) indicated the Kakamega wild silk has not been exploited commercially. Only re- that the Kenya silk market has no clear set goals on quality, price, vol- cently, with support from ICIPE, was a pilot market for wild silk produc- ume, taste, and designs for the industry. These important aspects need tion introduced as a new venture in the region, and as such there has to be in place in order to target specific markets with clearly defined been ´only´ a cultural link to its production in a non-market context. products. A GI registration, based on the existing and not yet wide- Today, a producer association covering Kakamega and Vihiga counties spread reputation, may be instrumental in delivering the marketing has 14 groups practicing both sericulture (rearing of silk caterpillars) needed for an origin product to succeed in the market. and apiculture (beekeeping). The groups monitor the wild silkworms and after harvesting the cocoons, they either sell them directly to ICIPE 5.3. The five elements of GI Mwingi honey and Kakamega forest silk or deliver them to a joint marketplace in Kakamega County for process- ing. This marketplace was established to provide a market for the co- The two case studies described above show that there are strengths coons from the producer groups, and it also has capacity to process the and weaknesses for each product in terms of GI potential. Fig. 2 qualita- wild silk. At the time of the field work, one kilogram of wild silk cocoon tively depicts and compares the GI potential of the honey and silk prod- of good quality and size was being sold locally at KES 200/kg (USD 2/kg), ucts as a spider web across the five major elements identified in Fig. 1.In which is lower than that offered for mulberry silk. The staff at the mar- both cases, the natural link to quality is the strongest. In Kenya, Acacia ketplace, responsible for the processing, indicated that the wild silk- honeys are generally in high demand, and Mwingi is one of three worm cocoons have more wastage and take a longer time to process areas in Kenya that are known for substantial production of quality aca- compared to mulberry silk. However, costs associated with harvesting cia honey. The flowering patterns of the Acacia woodlands of Mwingi of wild silkworms are lower, since they occur naturally. Thus, even provide very good conditions for quality acacia honey production, with the very low price (above) collectors have an incentive to collect. which was also the reason for KIPI's choice of Mwingi honey as a pilot The Kakamega forest is the only mid-altitude tropical rainforest in GI product in 2009 (KIPI; 2009). Among the currently 23 EU registered Kenya and it hosts biodiversity of global importance. The forest has PGI/PDO honeys, the main argument for GI registration is likewise the characteristics of the rainforests in Central and West Africa including link between organoleptic traits and the specific floral nectar sources. the high biodiversity value (Wagner et al., 2008). The B. micrantha In Kakamega, the silk moths' dependence on the local B. micrantha tree, on which the wild silkworms feed and spin their cocoons, has mul- tree, growing in the biodiversity hotspot and only mid-altitude tropical tiple uses both for timber and non-timber products. The wood from the rainforest in Kenya, gives the wild silk production a unique relationship tree is durable, fairly hard and resistant to termite attacks, and is also with the natural setting. However, in both Mwingi and Kakamega, the used to make charcoal and firewood of excellent quality while its natural environment also presents challenges. The capacity of local poles are used for construction of huts and granaries. The tree is also communities to handle climate change risks, which bring irregular sup- used for medicinal purposes, gums/resins and it has edible fruits. Due ply of the products due to changes in the weather patterns, is but one to its multipurpose nature, the tree is attractive to loggers, local subsis- limitation. In particular, the bee keeping ecosystem in Mwingi requires tence users, and traders of tree products, and today the tree is endan- long term investments for the sustenance of production technologies gered (Dharani, 2011). The majority of the surrounding communities that are adaptable and resilient to environmental and ecosystemic derive their livelihoods from small scale agricultural activities and har- changes. These would include provision of dams and other water vesting of forestry products (Müller & Mburu, 2009). As mentioned by sources for the bees, construction of processing and storage centers sericulture members during focus group discussions, people used to de- that meet prescribed GI standards, and afforestation with drought resis- stroy the silk moth cocoons and cut down the host tree for its multiple tant, locally adapted tree species that bees can forage on for extended purposes. Furthermore, once A. panda has fed on the leaves of the host seasons. In terms of the human link, the honey production has been 44 H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46 an important part of local livelihoods, both for consumption and sale to from below. With the expected premium prices and/or expanded mar- other areas of Kenya, for a longer period of time, while the silk produc- kets from the prospect GI honey and silk, many of the prospective pro- tion has been limited and revived in a non-market (project) fashion, ducer alliance actors will not be coordinated through local institutions mainly due to the initiative by ICIPE. The human link is about know- but through state regulations, formal market rules and top value chain how and capacity to process origin products, which in the case of silk actors (supermarket chains). This has happened with other initiatives is mainly limited to training and facilities introduced by ICIPE. Honey supported by local institutions, such as payments for ecosystem conser- producers are organized at several levels, from local beekeeper groups vation in wildlife areas in arid and semi-arid areas (Homewood et al., to umbrella associations that furthermore connect producers in central- 2012). There is also very little vertical integration of the local producers ized honey markets. The specific quality of the Kakamega wild silk is a with other main value chain actors. This raises the likelihood of oppor- rare strength of the local product; however, the reputation of the specif- tunistic behavior (side-selling, adulteration, etc.) among other value ic quality has not yet been established at silk markets in the larger urban chain actors which would, eventually, adversely affect gains made by areas. There seems to be a wider recognition of the Mwingi honey, as the local producers of the prospective GI honey and silk. observed during field work, even though Acacia honeys are available With the current challenges and constraints in the institutional envi- in several areas of Kenya. ronment, along with a large need for promotion of wild silk, also in local In both cases, the institutional environment represents the weakest markets, GI registration of honey and silk is not just around the corner. element, with several challenges and constraints for GI recognition of However, with greater emphasis on origin and local qualities in agricul- honey and silk. Both silk and honey will need policy support in Kenya tural, forest and trade policies at a national and regional (ARIPO) level, to be recognized nationally as GIs. Thus the initiatives of KIPI to have a changes in the institutional environment is expected. For the cases of GI policy debated and enacted in Kenya may require both in-county Mwingi honey and Kakamega silk, this represents an opportunity to and external support. This would provide an institutional framework market local produce and create value addition that will benefit small to pave the way for GI registration in Kenya. Further, for the case of scale forest product producers, just as it has for honey producers on honey, a National Beekeeping Policy and a Beekeeping Industry Bill Mount Oku in Cameroon. A GI regime may also help to enhance social are still in the process of enactment, having lingered in the political sys- cohesion and preserve common heritage among the rural communities, tem for some time. Once enacted, these instruments can guide the de- reduce rural exodus and promote local “savoir faire”. Furthermore, the velopment of a honey monitoring plan in Kenya to enhance the natural resources supporting Mwingi honey production and Kakamega products' traceability, quality and accreditation both at the national wild silk, the Acacia woodlands and the Kakamega Forest harboring the and local level. The country will also need a country-wide association B. micrantha tree, respectively, are also expected to benefit, as a GI pro- that can drive forward prospects of GI registration of these products tection partially based on the natural link will create an economic incen- through various discourses and different fora in order to gather more tive for their preservation. This view is supported by the Convention of political support. Currently, the civil society is practically not engaged Biological Diversity, which recognizes GIs as traditional lifestyles rele- in GI issues, including marketing of the two case products. Thus, produc- vant for the conservation and sustainable use of biological diversity er concerns and interests are currently not recognized beyond the local (Dominte, 2009). level. There are however, research institutions specifically focusing on sericulture that can help build the capacity of the producers to popular- 6. Concluding remarks ize their product beyond the local level. In the case of the Kakamega silk, an additional issue is that the product is not a food commodity. There- The case studies and our broad and overall analysis of the prospects fore, any recognition under the EU GI regime will in any case have to of origin forest products suggests that options do exist for product valo- await the expansion of the EU GI regime to also cover non-food com- rization and sustenance based on African taste and place. Underpinned modities. As much of the silk is harvested in the state-protected part by a critical mass of activities and initiatives by a range of international of the forest, redesigning of communal property rights that cover silk actors and agencies, including WTO, WIPO, FAO, EU, AU, ARIPO and moth cocoons and the trees designated for cocoon production and net- OAPI, as well as international civil society bodies, a renaissance for GIs ting will also be important for local communities to secure the benefits building a stronger international regime for geographical indications, from a GI registration of their product. is now maturing to plausibly overcome polarization and engage in con- The prospects of GI registration of honey and silk are also hampered structive co-evolution. by inadequacy of out-scaling resources as only a few national (e.g. KIPI) The energizing part of our conclusion is that new opportunities exist and international (e.g. ICIPE) actors are involved, and a critical mass of for countries in the global South to pursue development pathways organizational intervention is missing. In the case of the wild silk, the adapted to their own agro-environmental systems, food and clothing current production rests upon a limited number of households who cultures, and traditional production systems. These opportunities are al- have been able to join the few groups. Thus societal net benefits of GI ready being seized by some Asian and African countries, pioneering sui registration are likely to be low due to high transaction costs and dis- generis institutional frameworks for geographical indications and regis- economies of scale. Prospective value chains are therefore characterized tering some of their origin products as PGI and PDO with the EU. The di- by low production and varying quality associated with capacity limita- verse cultural and natural diversity of rural communities and their local tions and a lack of implementation of quality assurance regulations. In produce, including products derived from forests and wild habitats rep- the case of honey, these constraints means that the production struggle resent a wealth of origin products with potential for inclusion under de- to meet national demands of quality honey and the quality is highly var- velopment regimes acknowledging such values. Capacity building at iable, as indicated by the honey sample analyses. Similar conditions are local level, development of quality standard bodies, and a wider institu- found for other origin products in Kenya, Tanzania, and Ghana. Getting tional environment supporting GI products and markets are needed. the framework for out-scaling right can open up a valuable market for Getting these framework conditions right can open up a valuable mar- small-scale producers of origin products from forests. ket for small-scale producers of origin products from forests. New Both honey and wild silk value chains in the case studies have, thus knowledge in the field of institutional environments for origin food far and to a large extent, relied on local institutions (social capital, col- products and GI products in a diverse African context is needed to lective action, norms, taboos, etc.) to work and to govern the local nat- spur and facilitate further development of the GI framework within ural resource base. It is uncertain if this can sustain a pathway the African Regional Intellectual Property Organization (ARIPO), also towards GI protection, particularly when benefits of GI become widely to transform positive externalities such as pollination services from recognized. There is a risk that state interventions and elite capture honey production into market value, thus adding to the emerging inter- may foreclose GI producer alliances and local institutions emerging est for GIs in the African region. H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46 45
Research on the specific nature of the challenges and case studies of Egelyng, H., Bosselmann, A., Hansted, L., 2015. An Asian – African perspective on geo- graphical indications as institutional vehicles of development: a theoretical discus- experiences with regard to GIs in the global South are gaining momen- sion and a case of honey. Paper presented to Conference on Food, Feeding, and tum. This extends the scope of research on GIs from decades of eurocen- Eating in & out of Asia University of Copenhagen, 26th June 2015. trism to include research on institutional environments for origin Filoche, G., Pinton, F., 2014. Who owns Guaraná? Legal strategies, development policies and agricultural practices in Brazilian Amazonia. J. Agrar. Chang. 14, 380–399. products in the many African countries where sui generis legislation Gautam, K., Bahl, N., 2011. Geographical indications of India. Socio-economic and devel- for GIs has not yet come to pass. It also broadens the scope to investigate opment issues. Policy Brief #2. All India Artisans and Craftworkers Welfare Associa- conditions under which GIs can truly ´deliver´, as a ´stand-alone´ eco- tion (aiaca), New Delhi. nomic policy instrument not embedded in foreign economic develop- Grafton, R.Q., Daubjerg, C., Qureshi, M.E., 2015. Towards food security by 2050. Food Secu- rity 7, 179–183. ment programs, for a wider range of developmental dimensions – Overcoming barriers in collectively managed NTFPs in Mexico 2010 In: Granich, C.I., from biodiversity conservation and traditional knowledge to food sov- Purata, S.E., Edouard, F., Sánchez Pardo, M.F., Tovar, C., Laird, S.A., McLain, R., ereignty and green inclusive growth. Wynberg, R.P. (Eds.), Wild Product Governance: Finding Policies that Work for Non-Timber Forest Products. Earthscan, London, pp. 205–228. Guihong, W., 2012. In: Augustin-Jean, L., Ilbert, H., Saavedra-Rivano., N. (Eds.), Geograph- ical Indications and International Agricultural Trade: The Challenge for Asia. Palgrave- Acknowledgements Macmillan 2012. Hirko, S.B., 2014. The legal framework for the protection of geographical indications in The authors acknowledges the Danish Consultative Research Com- Ethiopia: a critical review. Journal of African Law 58, 210–230. mittee for Development Research (FFU) for supporting through Grant Homewood, K., Trench, P.C., Brockington, D., 2012. Pastoralist livelihoods and wildlife rev- enues in East Africa: a case for coexistence? Pastoralism: Research, Policy and Practice 13-02KU the research on which this article is based. 2(19)2012. Ilbert, H., 2012. Products with denominations of origin and IPR – the international bargaining process 2012 In: Augustin-Jean, L., Ilbert, H., Saavedra-Rivano, N. (Eds.), References Geographical Indications and International Agricultural Trade: The Challenge for Asia. Palgrave-Macmillan. Affognon, H.D., Kingori, W.S., Omondi, A.I., Diiro, M.G., Muriithi, B.W., Makau, S., Raina, IP Watch, 2013. Protection Of Local GI Products Can Benefit Women, Speakers Say. Intel- S.K., 2015. Adoption of modern beekeeping and its impact on honey production in lectual Property Watch, Geneva Accessed online 10.11.2015 http://www.ip-watch. the former Mwingi District of Kenya: assessment using theory-based impact evalua- org/2013/06/04/protection-of-local-gi-products-can-benefit-women-speakers-say/. tion approach. International Journal of Tropical Insect Science 35, 96–105. Izac, A.-M., Egelyng, H., Ferreira, G., Duthie, D., Hubert, B., Louwaars, N., Rosenthal, E., Allaire, G., 2012. The multidimensional definition of quality 2012 In: Augustin-Jean, L., Suppan, S., Wierup, M., McLean, M., Acheampong, E., Avato, P., de la Torre, D., Ilbert, H., Saavedra-Rivano, N. (Eds.), Geographical Indications and International Ag- Ferris, S., Gyasi, E., Koning, N., Murray, D., Raynolds, L.T., Robbins, P., Röling, N., ricultural Trade: The Challenge for Asia. Palgrave-Macmillan. 2009. Options for enabling policies and regulatory environments. In: McIntyre, B.D., Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N.J., Bauch, S., Børner, J., Herren, H., Watson, R.T. (Eds.), Agriculture at a Crossroads - Volume I: The Global Re- Smith-Hall, C., Wunder, S., 2014. Environmental income and rural livelihoods: a glob- port. IAASTD, Chapter 7. Island Press. al-comparative analysis. World Dev. 64, S12–S28. JAICAF, 2007. Sericulture in East Africa. Japan Association for International Collaboration ARIPO, EU, 2012. Administrative Memorandum of Understanding on co-operation be- of Agriculture and Forestry (JAICAF) Tokyo, Japan. tween the African Regional Intellectual Property Organization and the Directorate Josling, T., 2006. The war on terroir: GIs as a transatlantic trade conflict. J. Agric. Econ. 57 General for Agriculture and Rural Development of the European Commission. Zanzi- (3), 337–363. bar, Tanzania. Kioko, E.N., 1998. Biodiversity of Wild Silkmoths (Lepidoptera) and their Potential for Silk AU-EC, 2012. Creating Value through Geographical Indications: The Power of Origin Production in East Africa. Kenyatta University, Kenya. (Workshop on Geographical Indications, Kampala (Uganda), 10–11/11/2011. Joint KIPI. 2009. MoU Swiss - Kenya Project on Geographical Indications (SKGI). Kenya Intellec- African Union – European Union workshop, Onlinehttp://ec.europa.eu/agriculture/ tual Property Institute, Nairobi. Online: https://www.ige.ch/fileadmin/user_upload/ events/gi-africa-2011_en.htm. Juristische_Infos/e/MoU_swiss_kenya_e.pdf. (Accessed January 2016). Augustin-Jean, L., Ilbert, H., Saavedra-Rivano, H., 2012. Geographical Indications and Inter- Krasner S.D. (Ed),1983. International regimes. Ithaca, NY: Cornell University Press. national Agricultural Trade: The Challenge for Asia. Palgrave-Macmillan. Mancini, M.C., 2013. Geographical indications in Latin America value chains: a “branding Ayuya, O.I., Gido, E.O., Bett, H.K., Lagat, J.K., Kahi, A.K., Bauer, A., 2015. Effect of certified or- from below” strategy or a mechanism excluding the poorest? J. Rural. Stud. 32, ganic production systems on poverty among smallholder farmers: empirical evidence 295–306. from Kenya. World Dev. 67, 27–37. Mbahin, N., Raina, S.K., Kioko, E.N., Mueke, J.M., 2012. Biology of the wild Silkmoth Babcock, B.A., Clemens, R., 2004. Geographical indications and property rights: protecting Anaphe panda (Boisduval) in the Kakamega Forest of western Kenya. International value-added agricultural products. MATRIC Briefing Paper 04-MBP 7. Iowa State Uni- Journal of Forestry Research 2012, 7. http://dx.doi.org/10.1155/2012/186549. versity, Iowa. Mbahin, N., Raina, S.K., Kioko, E.N., Mueke, J.M., 2010. Use of sleeve nets to improve sur- Bainkong, G., 2014. Cameroon: Oku White Honey - the Fruits of Certification. http:// vival of the Boisduval silkworm, Anaphe panda, in the Kakamega Forest of western allafrica.com/stories/201409110549.html. Accessed in September 2015. Kenya. J. Insect Sci. 10, 1–10. Blakeney, M., Coulet, T., Mengistie, G., Mahop, M.T. (Eds.), 2012. Extending the Protection Ethiopia: Fine coffee 2012 In: Mengistie, G.,Blakeney,M.,Coulet,T.,Mengistie,G., of Geographical Indications. Case studies of Agricultural Products in Africa. Mahop, M.T. (Eds.), Extending the Protection of Geographical Indications. EARTHSCAN, Routledge, New York. EARTHSCAN, Routledge, New York, pp. 150–174 Case studies of Agricultural Carroll, T., Kinsella, J., 2013. Livelihood improvement and smallholder beekeeping in Products in Africa. Kenya: the unrealised potential. Dev. Pract. 23, 332–345. Muli, E., Munguti, A., Raina, S.K., 2007. Quality of honey harvested and processed using Chabrol, D., Mariani, M., Sautier, D., 2015. Establishing geographical indications without traditional methods in rural areas of Kenya. Acta Vet. Brno 76, 315–320. state involvement? Learning from Case Studies in Central and West Africa. World De- Müller, D., Mburu, J., 2009. Forecasting hotspots of forest clearing in Kakamega Forest velopment http://dx.doi.org/10.1016/j.worlddev.2015.11.023 Western Kenya. For. For. Ecol. Manag. 257, 968–977. Coulet, T., Mahop, M.T., 2012. Gabon: Okoumé wood. In: Blakeney, M., Coulet, T., Musungu, S.F., 2008. The Protection of Geographical Indications and the Doha Round: Mengistie, G., Mahop, M.T. (Eds.), Extending the Protection of Geographical Indica- Strategic and Policy Considerations for Africa. QUNO IP issue paper no. 8. Quakers tions. Case studies of Agricultural Products in Africa. EARTHSCAN, Routledge, New United Nations Office, Geneva. York, pp. 175–196. Pérez Villareal, J.S., 2012. La artesanía en el derecho. Vniversitas 125, 287–318. Dagne, T.W., 2015. Intellectual Property and Traditional Knowledge in the Global Econo- Rights, I.I.P., 2010. Protection of Geographical Indications in the Republic of Korea. Italian my: Translating Geographical Indications for Development. Routledge. Trade Commission & Italian Ministry of Economic Development. Dharani, N., 2011. Field Guide to Common Trees and Shrubs of East Africa. Struik nature, Roussel, B., Verdeaux, F., 2007. Natural Patrimony and Local Communities in Ethiopia: Ad- Cape Town, South Africa. vantages and Limitations of a System of Geographical Indications. Africa 77, 130–150 Dominte, N.R., 2009. Can geographical indications bring value to environmental protec- speciel issue. tion? IUCNAEL EJournal 5, 106–110. Shiluli, M., Magenya, O., E, V., Kanda, R., Gesicho, M., Kidula, N., Miruka, M., Simioni, E., Douguet, J.-M., Connor, M.O., 2003. Maintaining the integrity of the French terroir: a study 2012. Beekeeping in Trans Mara: A Baseline Study- 13nth KARI Biennial Scientific of critical natural capital in its cultural context. Ecol. Econ. 44, 233–254. Conference; Book of Abstracts. KARI, Nairobi. Dwi, M., Guizol, P., Roda, J.M., Purnomo, H., 2006. Geographic Indications for Javanese Steiner, P., 2012. Are food markets special markets? Chapter 2 2012 In: Augustin-Jean, L., teak: a constitutional change. CIFOR/CIRAD UR40 Working Paper/Document de tra- Ilbert, H., Saavedra-Rivano, H. (Eds.), Geographical Indications and International Ag- vail 40402. Online: https://core.ac.uk/download/files/153/6318660.pdf. (Accessed ricultural Trade: The Challenge for Asia. Palgrave-Macmillan. April 2016). Teshager, W.D., 2015. Intellectual Property and Traditional Knowledge in the Global Econ- Egelyng, H., Abreau, L.S.D., Li, L., Fonseca, M.F., 2013. Comparative institutional analysis of omy: Translating Geographical Indications for Development. Routledge. certified organic agriculture conditions in Brazil and China. In: Halberg, N., Muller, A. Vandecandelaere, E., Arfini, F., Bellett, G., Marescotti, A. (Eds.), 2010. Linking People, (Eds.), Organic Agriculture for Sustainable Livelihoods. Routledge, pp. 204–222. Places and Products: A Guide for Promoting Quality Linked to Geographical Origin Egelyng, H., 2007. Institutional Environments for Certified Organic Agriculture (COA): En- and Sustainable Geographical Indications, 2nd ed FAO and SINER-GI, Rome. abling Development, Smallholder Livelihoods and Public Goods for Southern Envi- Vasu, A.E., 2015. Towards securing geographical indications status for nilambur teak – the ronments? International Seminar on ”Pro-Poor Development in Low Income wood quality perspective. 3rd World Teak Conference. Guayaquil, Ecuador May 11–15, Countries: Food, Agriculture, Trade, and Environment”. Montpellier, France. 2015. 46 H. Egelyng et al. / Forest Policy and Economics 84 (2017) 38–46
Le Maroc Vert, 2015. Labelled Products in Morocco. Moroccan Ministry of Agriculture and WIPO, 2014. Bees, geographical indications, and development. Oku white honey, camer- Fishery. Online: http://www.agriculture.gov.ma/sites/default/files/siam15-produit_ oon. Case study by World Intellectual Property Organization, Geneva. Available on- terroir-vang.pdf. (Accessed April 2016). line: http://www.wipo.int/ipadvantage/en/details.jsp?id=5554. Wagner, P., Köhler, J., Schmitz, A., Böhme, W., 2008. The biogeographical assignment of a World Bank, 2007. World Development Report 2008. USA, Washington DC. west Kenyan rain forest remnant: further evidence from analysis of its reptile fauna. J. Biogeogr. 35, 1349–1361. Forest Policy and Economics 84 (2017) 47–61
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Prevalence, economic contribution, and determinants of trees on farms across Sub-Saharan Africa☆
Daniel C. Miller a,⁎, Juan Carlos Muñoz-Mora b, Luc Christiaensen c a Department of Natural Resources and Environmental Sciences, University of Illinois, Urbana, IL 61801, United States b Department of Economics and Management, Pompeu Fabra University, C/Ramon Trias Fargas 25-27, Barcelona 08005, Spain c Luc Christiaensen, The World Bank, 1818 H St NW, Washington, DC 20433, USA article info abstract
Article history: Trees on farms are often overlooked in agricultural and natural resource research and policy in Sub-Saharan Africa. Received 14 May 2016 This article addresses this gap using data from the Living Standards Measurement Study-Integrated Surveys on Received in revised form 28 September 2016 Agriculture in five countries: Ethiopia, Malawi, Nigeria, Tanzania, and Uganda. Trees on farms are widespread. On Accepted 7 December 2016 average, almost a third of rural smallholders grow trees. They account for an average of 17% of total annual gross in- Available online 20 January 2017 come for tree-growing households and 6% for all rural households. Gender, land and labor endowments, and espe- cially forest proximity and national context are key determinants of on-farm tree adoption and management. These Keywords: Trees on farms new, national-scale insights on the prevalence, economic contribution and determinants of trees on farms in Africa Agroforestry lay the basis for exploring the interaction of agriculture, on-farm tree cultivation, and forestry to gain a more com- Forestry plete picture of the dynamics of rural livelihoods across the continent and beyond. Landscape approaches © 2016 Published by Elsevier B.V. Climate smart agriculture Livelihoods Africa
1. Introduction animals, and greater structural connectivity (Manning et al., 2006; Place and Garrity, 2015) and serve as a key basis for biodiversity conservation In Africa, as in many other parts of the world, trees on farms are often (Bhagwat et al., 2008; Schroth et al., 2013) and climate change adaptation overlooked in research and policymaking. In forestry, the focus is mostly and mitigation (Mbow et al., 2014a). Sub-national case studies further on trees in forests rather than outside them (Barton, 2002; Fay and suggest that on-farm trees often also provide a sizeable source of income Michon, 2005). In agriculture and livelihood studies, the focus is typical- (from timber or non-timber products such as fruit) (Degrande et al., ly on annual crops and their effects on household income. When peren- 2006; Kalaba et al., 2010; Mbow et al., 2014b). Roughly, a third of the ag- nials (such as coffee trees) are considered, it is mostly from a value chain ricultural land in Sub-Saharan Africa is estimated to have had at least 10% perspective. The organization of extension and other services reflects tree cover during 2008–2010 (Zomer et al., 2014). Trees and agricultural this division, with agriculture and forestry typically separated in differ- activities therefore often co-exist not only in larger landscape contexts but ent institutions (de Foresta et al., 2013). As a result, trees on farms are also in single landowner holdings. often not included in forest-related, agricultural and livelihood statistics The available research on trees on farms has so far largely focused on and little remains known about their prevalence and economic contri- case studies within particular countries (e.g. Dewees, 1995b; Godoy, bution, particularly at the national scale. 1992; Pouliot and Treue, 2013). Region-wide aggregated approaches Yet trees on farms, which may range from sporadically occurring trees have also shed light on the prevalence of on-farm trees (Zomer et al., to areas dominated by a single tree crop through to large forest-like 2014), but because they are based on remotely sensed data such studies stands of trees, are often an integral component of broader agriculture- have not directly accounted for household perspectives and practices. forest landscapes. They perform important ecological functions, including Cross-national (e.g. Poverty and Environment Network (PEN) studies the provision of soil nutrients, prevention of soil erosion, habitat for highlighted in Wunder et al., 2014) and global (Agrawal et al., 2013) syntheses of forest and broader environmental income also exist, but ☆ This article is part of a special feature entitled: "Forest, Food, and Livelihoods” systematic comparative information on the prevalence and economic published at the journal Forest Policy and Economics 84C, 2017. contribution of trees on farms remains missing. This is especially prob- ⁎ Corresponding author at: Department of Natural Resources and Environmental lematic given intensifying competition for land in Africa (Peters, 2013) Sciences, University of Illinois, S-406 Turner Hall, 1102 S. Goodwin Ave., Urbana, IL and the challenge of simultaneously advancing human development 61801, United States. E-mail addresses: [email protected] (D.C. Miller), [email protected] and environmental protection goals. The lack of reliable national-scale (J.C. Muñoz-Mora), [email protected] (L. Christiaensen). estimates of the prevalence and contribution of trees on farms increases
http://dx.doi.org/10.1016/j.forpol.2016.12.005 1389-9341/© 2016 Published by Elsevier B.V. 48 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 the risk that they are left out of relevant policymaking processes, which In the absence of a standard classification of trees within crop data, could in turn result in greater priority for competing land uses that may potential trees were first identified from the LSMS-ISA agricultural undermine sustainability goals. crop production data, following the biological convention that to qualify This article addresses this gap using nationally representative, geo- as a tree a plant must be a woody perennial with a trunk or elongated referenced household survey data from five African countries collected stem that supports branches and leaves. With the help of several ex- under the Living Standards Measurement Study-Integrated Surveys on perts, the LSMS-ISA crop list was subsequently divided into five subcat- Agriculture (LSMS-ISA) initiative. Together, these countries (Ethiopia, egories: (1) fruit trees (e.g. mango, orange, etc.); (2) tree cash crops (e.g. Malawi, Nigeria, Tanzania, and Uganda) represent 41% of the population coffee, tea, etc.); (3) timber and fuelwood trees (e.g. Mahogany, bam- in Sub-Saharan Africa and cover many of its agro-ecological zones. In boo, etc.); (4) plant/herb/grass/roots (e.g. maize, banana, etc.); and addition to comprehensive household level information about con- (5) a series of unidentified crops (e.g. wechino, etc.) (Table A.1 includes sumption and income sources, these surveys also collected geo-refer- a detailed list of all the crops considered as trees and their further clas- enced plot level information on the different crops and trees grown on sification across these subcategories). The LSMS-ISA data included 230 each farm as well as the products harvested. These features of the data crops in total, of which about 30% (n = 68) were classified as trees. are exploited here to measure the prevalence and economic contribu- Nearly all the remaining crops, as expected, fell in the plant/herb/ tion of trees on farms in Sub-Saharan Africa, the first objective of this grass/roots category. study. Only the first three subcategories are considered here. While they The long time lag between planting and harvesting, insecure proper- contain all three perennials (with substantial lags between planting ty rights, small plots and landholdings, and remoteness, which often and harvesting which distinguishes them from other crops), they are characterize smallholder farming in Africa, would all seem to play nonetheless still quite distinct in their biological and economic features against the adoption of trees. Yet case study evidence from across Africa and support systems. Unlike fruit and timber trees, cash crops have been also shows that small farmers do plant and manage trees on their farms. extensively studied in the development literature, for example, but not So, what are the key drivers? We begin to explore this question by esti- in forestry, and they are usually politically important and part of well- mating theoretically informed models of the determinants of on-farm organized and integrated cooperatives and value chains. Unlike timber tree adoption, the second objective of this article. trees, fruit trees yield an annual return. This dramatically changes the The article proceeds as follows. The next section reviews the empir- parameters of the investment decision. For these reasons, we explore ical approach. Given the multitude of possible tree-like crops, it begins the three tree subcategories alongside each other. with a brief typology of the different trees considered. This is followed The stock of trees on farms identified in our study likely represents a by a description of the data and methodologies used to measure and an- lower bound. First, home gardens are plausibly underreported as plots alyze their prevalence and contribution to overall household income (and thus also trees in home gardens) and trees with no immediate pro- and welfare. An empirical model is subsequently presented to estimate ductive function (e.g. shade trees, living fences, or those retained for the key socio-economic and agro-ecological drivers of on-farm tree their aesthetic value) may have been left out of household question- adoption, drawing on key insights from the literature. Section 3 dis- naires. Second, respondents may not recall all trees on their lands or cusses the findings. Concluding remarks are offered in Section 4. may be hesitant to report them where, for example, colonial legacies of state control of tree resources persist (Leach and Scoones, 2013; 2. Materials, methods, and theoretical underpinnings Ribot, 1999; Sendzimir et al., 2011). Lastly, the study was unable to clas- sify a few species for which only the local name was available (Table 2.1. Identifying and counting trees on farms A.1). Yet, such omissions would especially affect the number of trees re- ported, and not so much their incidence or the share of land allocated to The nationally representative household surveys conducted under trees (for each plot it is recorded whether trees are present or not). Con- the LSMS-ISA initiative during 2010–12 from five African countries sequently, this study focuses on analyzing the prevalence (i.e. presence form the primary information base for this study.1 LSMS-ISA household or absence) of trees on farms and the share of land allocated to trees as surveys have been stratified to be representative for rural and urban opposed to the number of trees per se. areas. The surveys gather a wide range of socio-economic information on households and the communities of which they are a part, with de- 2.2. Contribution to household income and welfare tailed attention to their sources of income and geo-referenced, plot- level information on their agricultural activities and crops grown. To examine the contribution of trees on farms to farmers' liveli- Most importantly for this study, they also include detailed informa- hoods, three indicators are examined: 1) how tree products are used tion for all cultivated plots about the type of crop (including tree crops), (as a source of cash or mainly for own use or consumption); 2) their the harvest, and expenses incurred. For fallow or uncultivated plots, share in household crop and income portfolios (as an indication of farmers were explicitly asked whether they contained trees. In coun- their direct economic value) and 3) the consumption levels among tries where two seasons of agricultural data were collected (Malawi, farmers with and without trees on farms. Consumption levels also cap- Tanzania and Uganda), the average presence of trees across both sea- ture some of the more indirect contributions of trees on farms such as sons was taken. Across these countries a total of N20,000 rural house- soil conservation, nitrogen fixing, and water regulation (Booth and holds (and 47,000 plots) were surveyed and through application of Wickens, 1988; Nair, 2007; Place and Garrity, 2015)orasproviderofor- survey sampling weights a representative portrait of on-farm tree prev- ganic fertilizer or fodder for livestock (at least to the extent that they alence and their economic contribution to rural household incomes was raise and sustain agricultural income). None of these indirect aspects obtained. is typically addressed in the LSMS-ISA surveys (or household surveys more generally). Information on the quantity of tree products harvested, their value 1 The LSMS-ISA initiative is a collaboration between the World Bank and national statis- (i.e. price per unit/kg), and their different uses (sale, auto-consumption, tics offices in partner countries in Sub-Saharan Africa to design and implement multi-top- or other uses2) was directly obtained in all surveys, except in Tanzania. ic, nationally representative panel household surveys focusing on agriculture. This To assess their contribution to household income, we estimated the “ collaboration, funded by the Bill and Melinda Gates Foundation, seeks to foster innova- share of gross household income derived from trees on farms as part tion and efficiency in statistical research on the links between agriculture and poverty re- duction in the region” (World Bank, 2015). For details, see www.worldbank.org/lsms. Niger was excluded because, unlike other LSMS-ISA countries with available data for the 2 “Other uses” denotes use as inputs into another production process (e.g. fodder for study period, total income from trees or tree products was not recorded, only sales. livestock, fruits for jam, timber for own house construction or fencing). D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 49
Table 1 Descriptive statistics.
Ethiopia 2011–12 Malawi 2010–11 Nigeria 2010–11 Tanzania 2010–11 Uganda 2010–11
N Mean Std. N Mean Std. N Mean Std. N Mean Std. N Mean Std. dev. dev. dev. dev. dev.
Trees on farm Trees on farm (yes = 1) 3347 0.38 0.49 9936 0.22 0.42 2602 0.16 0.37 2621 0.55 0.49 1814 0.30 0.45 Fruit trees (yes = 1) 3347 0.17 0.36 9936 0.22 0.42 2602 0.06 0.23 2621 0.45 0.50 1814 0.05 0.21 Tree cash crops (yes = 1) 3347 0.33 0.48 9936 0.00 0.04 2602 0.15 0.36 2621 0.22 0.41 1814 0.27 0.43 Timber of fuel-wood (yes = 1) 3347 0.03 0.14 9936 0.00 0.04 2602 ––2621 0.18 0.42 1814 0.02 0.15 Share of farmland with presence of trees 3347 0.18 0.34 9936 0.00 0.04 2602 0.08 0.25 2621 0.41 0.42 1814 0.35 0.65 Share of farmland with presence of fruit trees 3347 0.05 0.19 9936 0.09 0.36 2602 0.04 0.17 2621 0.31 0.39 1814 0.05 0.26 Share of farmland with presence of tree cash crops 3347 0.17 0.33 9936 0.00 0.00 2602 0.07 0.23 2621 0.15 0.32 1814 0.32 0.63 Share of farmland with presence of trees for 3347 ––9936 0.00 0.04 2602 ––2621 0.15 0.32 1814 0.02 0.16 timber or fuel-wood
Household controls Household size 3347 4.89 2.349 9936 4.70 2.19 2602 6.17 3.17 2621 5.34 3.18 1814 5.75 2.850 Number of children (b14 years old) 3347 1.78 1.505 9936 2.26 1.65 2602 2.82 2.29 2621 2.38 2.09 1814 2.97 2.106 Head's age (years) 3346 44.33 16.050 9936 43.14 16.47 2596 50.24 15.11 2620 47.52 16.05 1811 47.71 14.968 Head female (yes = 1) 3347 0.24 0.432 9936 0.25 0.43 2602 0.11 0.31 2621 0.23 0.42 1814 0.30 0.460 Head education (years) 3307 1.67 3.090 9913 5.11 4.10 2546 4.51 4.97 2569 4.98 3.92 1625 4.68 3.461
Assets and land Land owned (area - ha) 2854 1.78 3.64 9936 0.80 0.85 2573 1.00 1.84 2212 2.61 4.02 1658 1.66 3.14 Tropical livestock units (TLU) 3347 2.51 2.78 9936 0.44 1.86 2602 3.12 17.17 2621 1.53 3.30 1814 1.24 2.95
Geo-climatic variables Population density around 20 km (people/km2) 3347 200.08 180.87 9936 183.86 159.76 2602 284.89 281.76 2586 93.14 132.65 1812 309.23 313.61 (2010) Tree cover % within 20 km (mean) (2010) 3347 51.45 20.44 9936 66.08 11.79 2602 34.86 23.68 2586 59.91 24.29 1812 65.90 15.67 Fertile soil % within 20 km (mean) (2010) 3347 0.73 0.27 9936 0.36 0.27 2602 0.53 0.31 2586 0.44 0.31 1812 0.20 0.23 Annual mean temperature (°C) 3347 18.42 3.18 9936 21.98 1.94 2602 26.36 0.99 2586 22.15 2.44 1812 21.79 1.86 Annual precipitation (mm) 3347 1175 340 9936 1094.36 258.43 2602 1382 617 2586 1068 326 1812 1234 187
Note: Statistics are for rural areas only and reported with sampling weight correction. “–” stands for missing information. of gross agricultural and gross overall income.3 Overall gross household with the consumption levels of those who do not grow trees on their income was calculated using the standardized definitions and methods farms, controlling for the characteristics of their environment. In partic- developed under the Rural Income Generating Activities (RIGA) Project ular, to do so real daily consumption per person (in 2011 purchasing (Davis et al., 2010).4 This approach facilitated comparison across power parity $) was regressed on whether the household had trees on countries. Production that had been consumed or stored was valued at farms or not and district fixed effects. The average difference in con- unit values either derived from reported sales or, when absent, from sumption is thus identified from the within-district difference in real median unit values at community or district levels (see Davis et al. daily consumption per person between tree and non-tree growing (2010) for details). Other in-kind income was valued at market prices. households A similar approach was followed to value (in-kind) income from on farm trees.5 2.3. Correlates of on-farm tree adoption Finally, as a broader, more encapsulating measure of the welfare ef- fects associated with tree adoption and management, we compare aver- At first glance, there appear few incentives for smallholder farmers age consumption levels among farmers growing trees on their farms to incorporate trees into their farming systems (Arnold and Dewees, 1997; Dewees, 1995a; Franzel, 1999; Godoy, 1992). There is a long time lag between planting and harvesting, while poor farmers are often liquidity constrained with a high discount rate. Access to (formal and informal) credit is limited, making it difficult to overcome the li- 3 A more refined measure would be to calculate the share of net income from trees on quidity gap and poorer farmers often have smaller landholdings, neces- farms over the total net income per household. However, complete expense data were sitating attempts to generate revenues annually on all the land available not collected in all study countries and when collected, they were not collected at the crop level making it difficult to attribute costs to a particular crop. Importantly, the gross in- to them. Insecure property rights further discourage investment in land come ratio used here remains unbiased, under the assumption that the share of net over improvement and trees, which only pay off over time. Outdated policies gross income is the same for income from trees on farms as for overall agricultural or total relating to state forest and tree management claims sometimes provide income. Put differently, to the extent that the share of expenditures on tree crop produc- more disincentives for farmers to invest in trees on their land (Scherr, tion to income from trees on farms is smaller than the share of expenditures on all agricul- 2004). The role that these factors play will differ depending on the tural production to income from agricultural production, the gross income from trees to gross agricultural income ratios reported here will be underestimates. A similar reasoning characteristics of the tree types. Timber trees only yield a benefitat holds for the ratio of gross income from trees to overall gross income. the end of their lives for example, while fruit and cash crop trees yield 4 Under this method, seven basic categories of household income are considered: (i) an annual return. The institutional support available also differs widely crop production; (ii) livestock production; (iii) agricultural wage employment; (iv) non- across tree type as does the purpose of production (home consumption agricultural wage employment; (v) non-agricultural self-employment; (vi) transfer; and (vii) when available, other income sources like rental income, fishing or saving accounts. or sales). See Davis et al. (2010) for details. Against this background, a number of hypotheses and empirical in- 5 In a recent study, Angelsen et al. (2014) show that environmental and forest incomes sights have been advanced in the literature. With respect to the effects can also be important in certain communities, especially those with closer access to for- of the demographic composition of the household, fruit trees, which ests. As in most standard household budget surveys, such income is not appropriately re- have been associated with better nutritional status of household mem- corded in the LSMS-ISA data and has not been accounted for here either. A forestry module for LSMS-ISA and other national surveys has recently been developed to address this infor- bers, have been found to be more prevalent in female-headed house- mation gap (Bakkegaard et al., 2016). holds (Ickowitz et al., 2014; Meijer et al., 2015). Larger households, 50 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61
Table 2 Share of landholders with trees on their farms by category of tree (%).
Country Percent of landholders with presence of Percent of landholders with Percent of landholders with Percent of landholders with presence of trees any trees on farms presence of fruit trees presence of tree cash crops for timber or fuelwood
Ethiopia 38% 17% 33% 3% (23.76% intercropped) (23.73% intercropped) (27.80% intercropped) Malawi 22% 22% 0.1% 0.1% (16.05% intercropped) (16.24% intercropped) (0% intercropped) Nigeria 16% 6% 15% Not Available (85.91% intercropped) (91.89% intercropped) (86.67% Intercropped) Tanzania 55% 45% 22% 18% (87.50% Intercropped) (91.89% Intercropped) (87.63% Intercropped) (82.28% Intercropped) Uganda 30% 5% 27% 2% (95.59% Intercropped) (99.66% Intercropped) (96.59% Intercropped) (77.89% Intercropped) Overall 30% 20% 12% 3% average (47.37% Intercropped) (43.78% Intercropped) (63.74% Intercropped)
Note: All descriptive statistics corrected by sampling weight.
with more labor available, are also more likely to adopt tree-based cul- (ii) A continuous measure of the share of landholdings with presence tivations, which is especially labor intensive in the early stages of tree Area of plots with presence of trees ðhaÞ of trees (i.e. Farm size ðhaÞ). planting and management (Dewees, 1994; Godoy, 1992). Better endowed households, on the other hand, are likely better placed to overcome liquidity and credit constraints and thus more likely to The former was estimated using a probit model, the latter using OLS. adopt agroforestry practices (Pattanayak et al., 2003). The amount of Furthermore, because key factors determining the differences in tree land owned is in this regard a well-established determinant of the pres- growing strategies may vary across tree types (Degrande et al., 2006), ence of on-farm trees (Cattaneo, 2001; Dewees, 1995a). The presence of the analysis was also replicated by type of tree (i.e. fruit trees, tree trees also interacts with livestock assets. Studies in different African cash crops and trees for timber or fuelwood). Exploiting similarity in countries suggest two different relationships: small livestock (e.g. the design of the questionnaires, the data were pooled across countries. goats and sheep) may be associated with greater presence of trees on This enabled us to identify those socio-economic and agro-ecological farms while cattle may be seen as a competitor for space (Place and factors that were generic across countries in affecting the adoption Garrity, 2015; Scherr, 1995). and extent of on-farm tree growing. Through the inclusion of country Geographic, climatic and biophysical conditions further affect the dummies a sense of the importance of country-specific factors (e.g. pol- degree of on-farm tree planting. Geographic location shapes the bio- icies and institutions) is also obtained.6 The country dummies also help physical endowments and a household's comparative advantage in control for measurement differences across countries. Shapley values, accessing markets, which in turn can influence incentives to adopt agro- which provide a decomposition of the explained variance of the depen- forestry practices (Pattanayak et al., 2003). Factors such as soil quality, dent variable (measured by R2)byeachgroupofcontrolvariables slope of farmland, proximity to forest, among others, create conditions (Shorrocks, 2013), are also reported. This approach helps to understand that are more or less conducive to grow and maintain trees. Generally, the mean contribution of each dimension or group of variables to the these biophysical factors have been less well studied in the agroforestry overall model (i.e. share of R2 explained by dimension). Standard errors adoption literature and evidence on their effects remains mixed (Meijer account for household sampling weights. et al., 2015; Mercer, 2004). Poor soil quality, for example, has been To explore the effects of the household's human capital endow- found to encourage adoption, but when soil quality is too poor, farmers ments (HH′ivcρ) the following variables were included: household may avoid planting trees (Pattanayak et al., 2003). The effect of the size, number of children (b14 years old), age of household head, a availability of nearby forest land is also likely to vary. Near forests, dummy variable indicating a female headed household, and the farmers may maintain certain valuable tree species on newly cleared levelofformaleducation(inyears) of the household head (Godoy, land or they may want to create a full separation from the forest on 1992; Pattanayak et al., 2003). To capture the effects of the such land. Alternatively, as availability of forest resources decreases, household's physical capital (Assets′ivcδ), we included: (i) the size they may plant more trees to cope (Arnold and Dewees, 1995). Proxim- of the land owned (in hectares), and (ii) the number of tropical live- ity to markets may also shape on-farm tree adoption by, for example, stock units (TLU, an international standard equivalence scale for dif- generating incentives to favor certain types of trees, especially those ferent types of livestock). yielding perishable products like fruit (Godoy, 1992; Mercer, 2004; The set of geographic and climatic controls, GeoClimate′ivcγ,includ- Pattanayak et al., 2003). ed human population density, average percentage of tree cover within To explore the importance of these different factors in determining 20 km of each household, soil fertility, annual mean temperature (°C), the presence and extent of trees on farms, the following regression and average annual precipitation. These control variables were con- model is estimated: structed based on household standardized geo-coordinates, which were collected in the LSMS-ISA survey data.7 Farm location was used ¼ ∝ þ 0 ρ þ 0 δ þ 0 γ TreesOnFarmsivc 1 HH ivc Assets ivc GeoClimate ivc as a centroid to construct several variables covering the area within 5 20 km. The average percent tree cover within a 20 km radius of each þ ∑ θkdTk þ εivc ð1Þ k¼1 where sub-index i refers to a household in village v in country c.Toex- 6 fi plore whether the factors affecting adoption and the factors affecting Alternative speci cation using models for each country separately were used as ro- bustness check. Results were qualitatively equivalent. These results are available upon the extent of tree planting differ, Eq. (1) was run separately using request. 7 fi fi (i) A binary measure of presence or absence of any trees on a given LSMS-ISA surveys provide a modi ed coordinate to protect household con dentiality, by introducing a random distortion of 0–5 km from the original location of the rural house- household's landholdings (i.e. Trees on farm (yes = 1)) as the de- hold. For more details on this type of method and its implications for statistical inference pendent variable; and see Perez-Heydrich et al. (2013). D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 51
Fig. 1. Spatial distribution of households with presence of on-farm trees by tree type. Note: This figure shows the spatial distribution of trees on farms across the five study countries. Each map has a different scale depending on the country size. The geographical unit of analysis is the household. All statistics were corrected by sampling weight. Data Source: Authors' elaboration based on World Bank (2015). household was derived using tree cover from MODIS Vegetation Contin- 2012)andNunn and Puga (2010) estimates of percentage of land sur- uous Fields (MOD44B) data (DiMiceli et al., 2011). An indicator of pop- face area with fertile soil.8 Finally, we included two control ulation density based on the number of people per km2 within 20 km of variables—household specific measures of annual mean temperature each household was created using data from the Global Rural-Urban (C) and average annual precipitation—which were created using a Mapping Population Project (Balk et al., 2006; CIESIN et al., 2011). To construct a variable on the average percentage of fertile soil within 8 Defined as soil that is not subject to severe constraints for growing rain-fed crops in 20 km of each household, we combined information from the FAO/ terms of soil fertility, depth, chemical and drainage properties, or moisture storage UNESCO Digital Soil Map of the World (FAO/IIASA/ISRIC/ISSCAS/JRC, capacity. 52 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 standard methodology by the World Bank LSMS-ISA team and made Given the importance of agro-ecological conditions for tree growing, available for all LSMS-ISA surveys (World Bank, 2015). some spatial clustering of on-farm tree growing was expected. Statisti- All models included a set of country fixed effects to account for coun- cally significant, though moderate spatial correlation among our obser- ð∑5 θ Þ vations was found (within country Moran's I was on average, 0.2) (Table try-level unobservables ¼ kdT k . They were also rerun with dis- k 1 11 trict fixed effects, purging the estimated coefficients from potential A.2). Clustering was especially clear for Tanzania and Ethiopia, two unobserved district level variables (such as relative price differences countries with high prevalence of trees on farms (Fig. 1; Table A.2). In across crops or market access). Table 1 summarizes the main descriptive Tanzania, households with tree cash crops (mainly cashew nut trees) statistics for all covariates. were highly clustered in the southwest, suggesting spatial correlation with the presence of larger urban centers and seaports such as in Lindi. Ethiopia presents a similar pattern. There are few trees in the far 3. Prevalence, economic contribution and determinants of trees on east, which includes lower elevation land near Somalia, while land- farms holders with fruit trees (46% of all households with trees on farms) were concentrated in the northwest and southwest of the country 3.1. About a third of smallholder farmers cultivated trees near some of Ethiopia's major population centers (e.g. Jimma and Bahir Dar). These patterns may be explained by the greater perishability Trees are clearly not marginal on the smallholder farms across of fruit and the constraints that farmers distant from ready access to N the African countries studied: 30% of all rural households, on aver- markets face when trying to sell the fruits of their trees. Tree cash age, reported having at least one tree on their land (Table 2). Prev- crops (e.g. coffee trees), on the other hand, were widespread and alence was highest in Tanzania (54%) and Ethiopia (38%) and planted primarily in a mono-cropping system. lowest in Nigeria (16%). Further disaggregation by type of tree shows that Tanzanian farmers emphasize fruit trees (with 45% growing fruit trees, primarily mango, pawpaw (papaya), and 3.2. Trees on farms especially prevalent near forests oranges).9 Tanzanian farmers also reported the highest prevalence of trees for timber (18%) and almost a quarter reported (23%) growing Households with trees on their farms are generally located close to tree cash crops. forests in the LSMS-ISA study countries. The majority of tree-owning By contrast, in Ethiopia, b3.5% of landowners reported having trees households were found within 10 km of forestland (at the 30% tree for timber and fuelwood on their land, but the country had the highest cover threshold) in all but one of the study countries (Nigeria; see proportion of farms with tree cash crops (32%), mainly coffee (65% of Table 3). In those countries with the highest share of their land area cov- N total tree cash crops) and chat10 (34% of total tree cash crops), and ered by forest (Tanzania and Uganda), this rose to 85% of the house- 17% farmers reported growing fruit trees. The small share of on-farm holds with trees on farms, on average. Even when using the more timber trees is likely an underestimate as the presence of Eucalyptus stringent forest cover threshold of 50%, these countries still had more was not captured in the questionnaires. Recent case studies (e.g. than half of the households with trees on farms within 10 km of the for- Bluffstone et al., 2015) found, for example, that 70% of households est areas (rising to 67% in Tanzania). Nigeria was at the lowest end of the grew eucalyptus in the six districts they surveyed. spectrum with only 36% of its households with trees on farms located Uganda, which features much less land area in plantation forests, near forest areas at the less stringent 30% threshold. By contrast, Ethio- followed a similar pattern as Ethiopia, with few trees for timber or fuel- pia had roughly the same overall forest cover (11% at the 30% threshold) wood reported and tree cash crops the most common type of tree (27%, as Nigeria, but more than half of its farms with trees were located within nearly all of which (97%) are coffee trees). Few farmers reported grow- 10 km of a forest. fi ing fruit trees (bananas, which are prevalent in Uganda, are not classi- These ndings suggest that farms with trees are important compo- fied as tree crops). In Malawi, fruit trees (mainly mango (56%) and nents of broader agriculture-forest landscapes in at least four of the fi pawpaw (12%)) were the most common category (22%). Information ve LSMS-ISA countries. Even in Nigeria more than a third of house- was incomplete on tree cash crops for Malawi, which may lead to an un- holds with trees on farms were near forests. Our analysis of the different derestimate of the prevalence of trees on farms in that country. Only factors affecting on-farm tree adoption also shows the importance of 0.01% of households reported having tree cash crops and among these distance to forest (see below). The proximity of on-farm trees to forests households information about area planted and number of trees was and the prevalence of intercropping across most of the study countries missing. In Nigeria, tree cash crops comprised the most frequent catego- suggest substantial opportunities for more holistic, landscape-level ap- ry of tree found on farms, though information on timber and fuelwood proaches in policy and other practical efforts seeking to reconcile biodi- trees was not available. versity conservation, climate change mitigation, and poverty reduction Almost 60% of trees on farms were reported in areas with other crops goals. Given the geographic and socio-economic diversity encompassed present (i.e. they are intercropped). However, this figure masks sub- in our study countries, our results also suggest that such approaches stantial variation (Table 2). Farmers in Malawi and Ethiopia may also be applicable in other countries across Africa as does the grow- intercropped much less than farmers in Nigeria, Tanzania, and Uganda. ing case study literature (e.g. Milder et al., 2014). Only 10% of trees on farms were reported as present in the same area as other crops in Malawi, for example. In contrast, N95% of on-farm trees 3.3. Contribution of trees on farms to rural livelihoods was non-negligible were reported as part of intercropped systems in Uganda. The common practice in that country and Tanzania of intercropping coffee with, for Products harvested from trees on farms in the study countries were example, bananas is only part of the story and more detailed country- used mainly for self-consumption or sale (Fig. 2). The relative mix of by-country study is needed to shed additional light on the dynamics these uses varied among the study countries. In Uganda, fruit tree prod- of intercropping with trees. Our results lay the groundwork for such in- ucts were used primarily for self-consumption, whereas in Malawi such quiry by highlighting the prevalence of agroforestry practices in the trees served solely as a source of cash income. That fruit trees were used study countries. 11 We employed the Moran's I Spatial Correlation index. This index provides an intuitive 9 Bananas were especially prevalent in Tanzania and Uganda, but they are classified un- measure of correlation among nearby households in space. However, its absolute value der the category of plant/herb/grass/roots based on their biological characteristics. may under-estimate the actual spatial correlation among households as it is based on geo- 10 Chat (Catha edulis (also known as khat or qat) is a is a slow-growing woody plant in- graphical distances that ignore the sampling design of our data. Nonetheless, its statistical digenous to the Horn of Africa and the Arabian Peninsula where it has been cultivated and significance indicates that households with trees on farms tend to be closer to each other used for centuries, primarily as a stimulant. in space (i.e. clustered) than the other households in the sample. D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 53
Table 3 Household distance from nearest forest defined as A) 30% tree cover threshold and B) 50% tree cover threshold.
Country Extent of tree cover (ha) by country Percent tree cover relative to country land area Households in our sample Share (%) of households with trees on farms (2000) (2000) (#) within
10 km of 20 km of 50 km of forest forest forest
A) 30% tree cover threshold Ethiopia 12,040,763 10.72 3347 55.81 73.91 93.3 Malawi 1,521,741 16.17 9936 85.87 100 100 Nigeria 10,033,216 11.13 2602 36.33 46.51 59.7 Tanzania 26,42,2567 29.85 2621 79.82 88.1 94.2 Uganda 7,768,069 37.83 1814 91.85 98.02 100 Overall 6,272,758 17.95 20,320 58.47 68.91 77.05
B) 50% tree cover threshold Ethiopia 5,426,282 4.83 3347 32.05 44.19 74.62 Malawi 313,115 3.23 9936 53.57 87.81 100 Nigeria 4,716,199 5.23 2602 20.17 29.27 42.53 Tanzania 9,702,599 10.96 2621 66.84 77.68 86.65 Uganda 3,271,840 15.94 1814 55.59 81.76 98.95 Overall 3,905,006 17.95 20,320 38.45 53.22 68.13
Note: To protect confidentiality household location coordinates in LSMS-ISA data are not exact, but rather based on a random distortion of 0–5 km. Data on extent of tree cover by country and percent tree cover relative to country land area derive from Hansen et al. (2013). Note that “tree cover” is not the same as “forest cover” in these data. “Tree cover” refers to the bio- physical presence of trees, which may be a part of natural forests or tree plantations. Information on household distance to forest are based on the authors' calculations from LSMS-ISA data sets (World Bank, 2015)and“MOD44B MODIS Vegetation Continuous Field Coll. 5–2000 through to 2010: Percent Tree Cover” (DiMiceli et al., 2011).
significantly for self-consumption in countries like Uganda (and also countries (Ethiopia, Nigeria and Uganda). However, no positive effect Ethiopia) suggests that such trees may play an important role in house- was found for timber tree growing. Households with timber and fuel- hold food security, as shown in a variety of contexts across Africa wood on their land appeared to be even worse off in Malawi. This result (Garrity et al., 2010). In the case of tree cash crops, production was may be due to characteristics particular to the tiny (0.18%) subset of mainly used for sale, as expected, though in Ethiopia a non-negligible households reporting having such timber or fuelwood trees. Generally, share was also used for own consumption (linked to coffee consump- it accords with other research that has found a significant overlap be- tion) and in Nigeria for other uses, such as for storage and gifts. tween areas of high poverty and high forest cover (sites for timber Turning to the contribution to total income, income from trees and fuelwood resources) across a large proportion of Malawi's territory on farms contributed on average 6% of overall annual gross house- (Sunderlin et al., 2008). hold income (i.e. taking farmers with and without trees on their farms together). The income share averaged 7% in Nigeria, 9% in 3.4. Gender, land and labor endowments, proximity to Forest and country Tanzania, 6% in Ethiopia and Uganda, but b4% in Malawi (Table 4). factors drove on-farm tree growing We note that estimates for Malawi may be low because information on tree cash crops was not provided in the LSMS-ISA survey data. Table 6 presents the estimated effects of different correlates on the For those households with trees on their farms the average contri- adoption of and land allocation to trees on farms. Tables A.3, A.4, and bution across the study countries was almost three times as A.5 zoom in on the results by tree type. The model had little power in much,i.e.17%andabout32%asashareofagriculturalgross explaining the adoption of or land allocation to timber trees 12 income. Surprisingly, the contribution of trees on farms to gross (Table A.5), but a number of clear generic patterns emerge when income among farmers with trees on farms, was highest in Nigeria looking at the correlates of tree cash crops and fruit trees. First, (36%), even though tree growing occurred least frequently there, adoption and land allocation to fruit trees and tree cash crops in- suggesting a high degree of specialization among tree growing creases with the education level of the household head. Second, households. At 18%, income from trees among households with the adoption and land allocation to tree cash crops was about 5 per- trees on their farms was also significant in Uganda. Clearly, income centage points less among female-headed households. This result is from tree growing can be quite important, with the larger share consistent with studies showing lower land tenure security among typically coming from tree cash crops (14% of gross income womeninAfrica(Berry, 1988; Schroeder, 1999) and the notion that among farmers with on-farm trees and 18% of gross agricultural in- women lack time for tree crop planting and management given come). Nonetheless, even though much less commented upon, in- their responsibilities for food crops and domestic chores (Berry, come among fruit trees still contributed 5% of gross income and 1988). It also accords with studies showing how the limited avail- 16% of gross agricultural income. ability of adult male labor can limit a given household's ability to Finally, taking a fully reduced form, we compared real per capita include high value trees in their livelihood portfolio (Fisher, consumption levels (2011 PPP) among tree growing households and 2004). The negative correlation of on farm tree growing with fe- non-tree growing households controlling for district level effects male headship also holds for fruit tree crops, but is substantially (Table 5). As expected, tree cash crop growers were on average substan- less pronounced, consistent with the higher nutritional value of tially better off (84% in Ethiopia, 19% in Nigeria, and 3% in Tanzania, fruit trees (Degrande et al., 2006; Mbow et al., 2014b). Third, tree though no difference was discerned in Uganda). Farmers with cash crops and fruit trees also tend to be more likely among house- fruit trees on their farms were also better off in three of the five case holds with older heads, consistent with other studies (Pattanayak et al., 2003) and the notion that it takes time for trees to grow and mature. 12 By comparison, a recent cross national study of households living in or near forests found that natural forests in Africa contributed 21% to household incomes and plantation Fourth, land (but also labor) endowments also mattered, with larger forests b1% (Angelsen et al., 2014). farms (but also larger households) tending to grow more trees. Land 54 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61
Somewhat surprisingly, however, land ownership was not signifi- cant in the overall adoption model, but it was positively correlated in the fruit tree model (Table A.3). Household size, an indicator of labor endowment, given that the number of children is controlled for, is positively associated with the adoption and allocation of farm land to trees (Table 6). The finding held separately for the fruit tree model, but not for the tree cash crops (Tables A.3 and A.4), suggesting that labor constraints may be more binding for fruit than tree cash crops. Adoption of fruit trees was also negatively correlated with ownership of livestock, a result consistent with studies showing that larger domestic animals (i.e. representing a greater number of TLUs) compete with trees on farmers' land (Place and Garrity, 2015; Scherr, 1995). Fifth, the finding of more on-farm tree planting in forest rich environments was confirmed in the multivariate setting. While it might be argued that farmers would increase the density of trees on their farms as forest resources nearby decrease (Arnold and Dewees, 1995), they could also continue agroforestry sys- tems which already have an integral relationship with nearby forests, including practices such as retaining key tree species during forest clearing (e.g. Brottem, 2011; Degrande et al., 2006). The empirical results here are more in line with the latter. On farm trees also tend to be more prevalent in environments where the temperature is higher. Overall, geo-climatic variables were important predictors, accounting for about 31% of the total variation in the adoption of trees on farms (Table 6). The strength of this result buttresses previous findings demonstrat- ing that biophysical conditions are key drivers of farmer deci- sions to adopt and maintain trees on-farm (Pattanayak et al., 2003; Place and Garrity, 2015). Surprisingly, no systematic corre- lation was found across countries with population density (or soil fertility). Finally, country fixed effects also explained an important share of the variation (N40% for adoption). This could partly reflect differ- ences in survey design and implementation across the study coun- tries. Although the World Bank has taken pains to ensure standard practice in the LSMS-ISA countries, some differences remain (e.g. omission of eucalyptus in Ethiopia). Other and plausibly more im- portant, country differences relate to differences in national level policies and institutions. Forest and tree-related policies have for example been shown to play a decisive role in shaping whether farmers decide to adopt trees on their farms or not. In many coun- tries, forest regulations create disincentives for on-farm tree man- agement (Place and Garrity, 2015; Ribot, 1999)andchangesto suchregulationscanspurchangesinpracticeasshowninthere- cent re-greening in Niger, which has been catalyzed by political openings and reforms to colonial-era forest and rural policies that allowed local innovation in land management (Sendzimir et al., 2011). Among the case countries studied here, Tanzania stands out. Compared to the reference country, Ethiopia, where forest pol- icy has remained more centralized (Ayana et al., 2013), policy relat- ing to forests and trees in Tanzania has been more decentralized, with arguably a stronger emphasis on clearly defined tree tenure rights (Petersen and Sandhövel, 2001). Overall country governance quality may also affect farmer decisions vis-à-vis trees. For exam- ple, higher incidence of corruption, which may disproportionately Fig. 2. Share of tree products by use, by country. Note: These figures show whether affect the sale of high value tree products, is likely to provide a dis- different categories of trees on farms were sold, used for self-consumption, or had other incentive to invest in tree crops. uses. Data were not available for Tanzania, as information on main uses for trees was not collected in that country. In Nigeria, the share of tree products for self-consumption 4. Concluding remarks was not collected. All statistics were corrected by sampling weight.
In this article we used nationally representative household- level data to explore and compare the prevalence and economic ownership was positively and significantly correlated with share of contribution of trees on farms across five African countries. Three farmland with trees as expected. This result makes sense given the main findings emerge. First, trees on farms are widespread across higher land intensity of tree crops. the continent and comprise a key component of agricultural-forest D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 55
Table 4 Contribution of trees on farms to annual gross household and agricultural income.
Annual gross household income Annual gross agricultural income
Contribution from Contribution from Contribution from tree Contribution from Contribution from Contribution from tree trees on farm (%) fruit trees (%) cash crops (%) trees on farm (%) fruit trees (%) cash crops (%)
Ethiopia All farmers 5.55 0.14 5.80 8.41 0.35 8.61 2011–12 Only farmers with 13.75 0.37 14.39 20.93 0.90 21.44 trees on farm Malawi All farmers 3.32 3.32 0.00 20.32 20.32 0.00 2010–11 Only farmers with 13.43 13.43 0.00 82.32 82.32 0.00 trees on farm Nigeria All farmers 6.90 1.40 6.40 6.76 0.94 5.82 2010–11 Only farmers with 36.20 7.92 33.31 36.14 5.28 30.86 trees on farm Tanzania All farmers 8.82 4.02 4.05 14.27 7.84 5.30 2010–11 Only farmers with 13.32 6.07 6.11 21.56 11.83 8.00 trees on farm Uganda All farmers 5.94 0.32 5.73 7.31 0.61 6.90 2010–11 Only farmers with 18.75 1.02 18.09 23.10 1.93 21.80 trees on farm Overall All farmers 5.98 1.80 5.33 11.05 5.51 6.81 Only farmers with 16.85 5.16 13.91 31.47 15.82 17.93 trees on farm
Table 5 landscapes. The East African countries of Tanzania, Uganda and Relationship of trees on farms and daily consumption per person. Ethiopia had especially high incidence of trees on agricultural Dependent variable = Log. real daily consumption per person (in 2011 PPP) lands, with about a third to more than half of rural households (I) (II) (III) (IV) reporting on-farm trees. Fruit trees and cash crop trees were the ⁎⁎⁎ two most popular types of trees while trees for timber and fuel- Ethiopia Trees on farm (yes = 1) 0.597 2011–12 [0.037] wood were much less prevalent (reported by 5% of respondents). ⁎⁎⁎ Fruit trees on farm 0.382 The proximity of most households with trees on farms to forests (yes = 1) [0.053] and the high incidence of intercropping across the study countries ⁎⁎⁎ Tree cash crops on farm 0.612 suggest that on-farm trees are also an integral part of larger rural (yes = 1) [0.039] agricultural-forest landscapes. As elsewhere in the developing Trees for timber or 0.132 fuelwood on farm [0.134] world (Sayer et al., 2013), policies and practices designed to im- (yes = 1) prove the management of such trees in Africa therefore hold signif- Malawi Trees on farm (yes = 1) 0.000 icant promise for helping to reconcile the sometimes conflicting – 2010 11 [0.031] goals of reducing rural poverty, conserving biodiversity, and miti- Fruit trees on farm −0.006 (yes = 1) [0.010] gating climate change. ⁎⁎⁎ Trees for timber or −0.323 Our second finding is that trees on farms deliver non-negligible eco- fuelwood on farm [0.103] nomic benefits to rural households. Across the rural population as a (yes = 1) ⁎⁎⁎ whole, production from trees on farms accounts on average only for Nigeria Trees on farm (yes = 1) 0.212 6% of total annual gross income. Yet, this increases to 17% on average 2010–11 [0.035] ⁎⁎⁎ Fruit trees on farm 0.252 for those households growing trees on their farms. By way of compari- (yes = 1) [0.046] son, these results are similar to available evidence on forest and envi- ⁎⁎⁎ Tree cash crops on farm 0.177 ronmental income. For example, a recent study using national-scale (yes = 1) [0.030] data from Mexico (López-Feldman, 2014) found that forest and other Tanzania Trees on farm (yes = 1) −0.002 2010–11 [0.030] environmental resources contributed 6.2% to total incomes for Mexican Fruit trees on farm 0.011 rural households. A global study focusing on those households living in (yes = 1) [0.010] or near tropical forests (Angelsen et al., 2014) concluded that income ⁎⁎⁎ Tree cash crops on farm 0.032 from natural forests and forest plantations accounted for 21% of total (yes = 1) [0.011] household income in Africa. Together, these studies suggest that Trees for timber or 0.010 — — fuelwood on farm [0.010] trees in forests and outside forests provide non-negligible income to (yes = 1) rural households in Africa, especially, but not only, to those living with Uganda Trees on farm (yes = 1) 0.010 trees nearby or on their land. 2010–11 [0.025] ⁎⁎⁎ Finally, results from the analysis of the determinants of the adop- Fruit trees on farm 0.102 (yes = 1) [0.032] tion of and land allocation to trees on farms suggest the importance Tree cash crops on farm 0.002 of national governance context and proximity to forests in under- (yes = 1) [0.010] standing differences in on-farm tree growing. Country fixed effects Trees for timber or 0.002 accounted for N40% of the explained variation in the models. Given fuelwood on farm [0.021] broadly similar sampling design and survey implementation in the (yes = 1) collection of LSMS-ISA data and the fact that we controlled for the ef- fi Note: Sampling weights and xed effect were used for all regressions. fect of other large-scale factors (e.g. climatic zone), governance and ⁎ p b 0.10. ⁎⁎ p b 0.05. institutions likely comprise a large share of this national-level varia- ⁎⁎⁎ p b 0.01. tion. Proximity to forests proved also an important predictor and 56 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61
Table 6 Multivariate analysis of adoption and management of trees on farms. Data source: Authors' calculations from LSMS-ISA data sets, World Bank (2015).
Adoption analysis (Probit) Determinants of share of farmland with trees
Dep. variable: trees on farms (yes = 1) Dep. variable: share of farmland with presence of trees
(I) (II) Shapley value (III) (IV) Shapley value
Household controls 0.011 (4.06%) 0.008 (2.76%) ⁎⁎ ⁎⁎ ⁎ Household size 0.008 0.012 0.016 0.012 [0.006] [0.005] [0.007] [0.007] Number of children (b14 years old) −0.002 −0.004 −0.010 −0.007 [0.007] [0.007] [0.010] [0.009] ⁎⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ Head's age (years) 0.002 0.002 0.003 0.004 [0.001] [0.001] [0.001] [0.001] ⁎⁎⁎ ⁎⁎⁎ Head female (yes = 1) −0.055 −0.060 0.006 −0.023 [0.012] [0.013] [0.046] [0.032] ⁎ ⁎ Head education (years) 0.003 0.004 0.010 0.009 [0.003] [0.003] [0.005] [0.005] Assets and land 0.004 (1.51%) 0.206 (64.46%) Tropical livestock units (TLU) −0.003 −0.002 −0.001 −0.001 [0.002] [0.002] [0.001] [0.001] ⁎⁎⁎ ⁎⁎⁎ Land owned (area - ha) 0.004 0.005 0.267 0.263 [0.004] [0.004] [0.094] [0.094] Geo- and climate variables 0.033 (11.38%) 0.004 (1.28%) ⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Log population density around 20 km (people/sqkm) (2010) 0.086 0.077 0.166 0.132 [0.035] [0.025] [0.055] [0.045] ⁎⁎⁎ ⁎⁎⁎ Tree cover % within 20 km (mean) (2010) 0.007 0.007 0.003 0.003 [0.002] [0.002] [0.003] [0.003] Fertile soil % within 20 km (mean) (2010) −0.004 −0.020 0.134 0.134 [0.072] [0.075] [0.151] [0.147] ⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎ Log. annual mean temperature (C) 0.027 0.033 0.045 0.043 [0.011] [0.012] [0.022] [0.022] Log. annual precipitation (mm) −0.000 0.000 −0.000 −0.000 [0.000] [0.000] [0.000] [0.000] Country fixed effects 0.099 (33.87%) 0.075 (23.56%) ⁎⁎⁎ ⁎⁎⁎ Malawi −0.273 −0.258 −0.150 0.026 [0.043] [0.026] [0.128] [0.135] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ Nigeria −0.398 −0.433 −0.306 −0.171 [0.061] [0.055] [0.131] [0.134] ⁎ ⁎⁎⁎ ⁎⁎⁎ Tanzania 0.124 0.105 0.820 0.715 [0.063] [0.069] [0.146] [0.118] ⁎⁎⁎ ⁎⁎⁎ ⁎ Uganda −0.262 −0.270 0.260 0.365 [0.054] [0.042] [0.214] [0.207] Mean dependent variable 0.290 0.290 0.243 0.243 (Pseudo) R-squared 0.207 0.258 0.306 0.320 Observations 18,907 18,907 18,907 18,907 District/regional fixed effect No Yes No Yes
Note: Baseline country is Ethiopia. Columns (I) and (II) present the point estimates for the adoption analysis (probit), where the dependent variable is the presence of any kind of trees on farm (yes = 1). Columns (III) and (IV) show results for determinants of share of farmland with trees (OLS), where the dependent variable is the share of farmland with presence of trees. District/sub-national regional fixed effects were included in one set of probit and OLS models, but not the other set. Robust standard errors in brackets, clustered at strata level. Sampling weights used for all regressions. ⁎ p b 0.10. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01.
formed part the broader set of geo-climatic variables, which fully account for non-crop trees (e.g. for shade or different kinds accounted for about 30% of the variation. These findings suggest of non-provisioning ecosystem services) and trees with no imme- that further analyses focusing on specific countries or geo-climatic diate productive function because they were not queried explicitly zones are needed to gain a deeper understanding of how such factors in the LSMS-surveys. The indirect effects of trees on farms on may drive farmer decision-making. The results from our cross-coun- crops, livestock, and other productive activities are also very diffi- try study further indicate that households with larger landholdings cult to account for (Wunder et al., 2014) and information on them tend to allocate more land to trees (both cash crop and fruit trees) was not directly collected in the LSMS-ISA surveys. Ethiopia also and that female-headed households tend to be less engaged in tree presented a rare case where information on a key productive growing, with the effect largest for tree cash crops. This latter finding tree species—eucalyptus—was not collected. may be linked to higher land tenure insecurity for female farmers Overall, the results suggest that trees on farms should be given more and is consistent with the higher nutritional value of fruit trees. attention in agriculture, food security and poverty-related policy de- Country case studies are needed to shed more light on these general bates in Africa, particularly as the need to tackle climate change be- patterns. comes more urgent. The data and analysis presented here provide a Despite the already non-negligible prevalence and economic contri- baseline for future benchmarking as well as building blocks for improv- bution of trees on farms that our study demonstrates, the numbers are ing the information base relating to privately owned trees in Africa, in- likely still underestimates. The household data are unlikely to cluding through improved data collection in future surveys. The open D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 57 access LSMS-ISA surveys and other national survey data provide an im- means for farmers to escape poverty or achieve more enduring portant opportunity to do so. They deserve full support and could be prosperity. Such panel data might also be fruitfully analyzed along strengthened in at least two ways: 1) by capturing the full range of rel- with information on forest and tree-related institutions and policies evant trees on farms, including those that may not have an immediate in individual countries to better understand how these country-level productive function, and 2) by including cost information in a way variables may affect farmer decision-making relating to trees. Final- that facilitates comparison of income across LSMS-ISA countries. We ly, we see significant scope for future research to collect and analyze were unable to include Niger in our analysis, for example, despite its information on the economic contribution not only of trees on farms, status as an LSMS-ISA country and success in farmer managed natural but also from forests and other wildlands to gain a more complete regeneration and agroforestry (Garrity et al., 2010; Sendzimir et al., picture of the dynamics of rural livelihoods in Africa over time at 2011) because income from trees on farms did not cover in-kind in- the national scale. come from these trees, only income from sales. The findings also point to several new avenues for exploring the Acknowledgements interaction of agriculture, trees, and forests to better understand the dynamics of rural livelihoods in Africa and beyond. One area This article was written as part of the “Agriculture in Africa - Telling ripe for further exploration is the relationship between trees on Facts from Myths” project, which revisits common wisdom about African farms and forest areas. Are trees on farms associated with more or agriculture and farmer livelihoods using household survey data collected less forest clearing? Why are households with trees on farms under the World Bank Living Standards Measurement Study-Integrated more likely to be located near forests? Panel data from LSMS-ISA Surveys on Agriculture (LSMS-ISA) initiative. Funding from the Program surveys combined with newly available, high-resolution forest on Forests (PROFOR), activity #TF017937, is gratefully acknowledged. coverdatamakeitpossibletoshed new light on these dynamics. The authors thank Karen Brooks, Frank Place, Laura Vang Rasmussen, Within specific countries it may also be possible to distinguish ex- Cristy Watkins, two anonymous reviewers, and participants at the otic from indigenous trees, which may be more likely to be retained “Myths and Facts” workshop at IFPRI in June 2015 and the Forests & when forest is cleared for agriculture. LSMS-ISA panel data also en- Livelihoods: Assessment, Research, and Engagement (FLARE) Network able study of the economic contribution of trees on farms over time Conference in Paris in November 2015 for helpful comments on earlier soastounderstandtheextenttowhichsuchtreescanprovidea versions of the manuscript.
Appendix A. Appendix
Table A.1 Crop/tree classification by type of tree in LSMS-ISA data. Data source: World Bank, 2015.
Crop Type of tree Crop Type of tree
Agbono (Oro Seed) Fruit tree Black pepper Tree cash crops Apple Fruit tree Cashew Tree cash crops Avocado Fruit tree Castor beans Tree cash crops Bilimbi Fruit tree Chat Tree cash crops Bread fruit Fruit tree Clove Tree cash crops Buya Fruit tree Cocoa Tree cash crops Cherry (Agbalumo) Fruit tree Coffee Tree cash crops Cinnamon Fruit tree Dry leaves (Kuka) Tree cash crops Coconut Fruit tree Gum arabic Tree cash crops Custard apple Fruit tree Iyere Tree cash crops Date palm Fruit tree Locust bean Tree cash crops Durian Fruit tree Macadamia Tree cash crops Fig Fruit tree Monkeybread Tree cash crops Gishita Fruit tree Moringa Tree cash crops God fruit Fruit tree Oil palm Tree cash crops Grape fruit Fruit tree Palm kernel Tree cash crops Guava Fruit tree Ronier Tree cash crops Jackfruit Fruit tree Rubber Tree cash crops Kolanut Fruit tree Shea nuts Tree cash crops Lemon Fruit tree Tea Tree cash crops Lime Fruit tree Three leave yam Tree cash crops Malay apple Fruit tree Bamboo Trees for timber and fuelwood Mandarin/tangerine Fruit tree Black wattle Trees for timber and fuelwood Mango Fruit tree Fence tree Trees for timber and fuelwood Masau Fruit tree Firewood/fodder Trees for timber and fuelwood Oranges Fruit tree Kapok Trees for timber and fuelwood Paw Fruit tree Mahogany Trees for timber and fuelwood Peaches Fruit tree Natural forest trees Trees for timber and fuelwood Pear Fruit tree Other forest trees Trees for timber and fuelwood Plum Fruit tree Plantation trees Trees for timber and fuelwood Pomegranate Fruit tree Timber Trees for timber and fuelwood Pomelo Fruit tree Pomme Du Sahel Fruit tree Rambutan Fruit tree
(continued on next page) 58 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61
Table A.1 (continued) Crop Type of tree Crop Type of tree
Star fruit Fruit tree Tamarind Fruit tree Walnut Fruit tree
Note: Crops included in LSMS-ISA surveys are not fully standardized, so there is variation across languages and local terminology for the same species in some cases. This table consolidates where possible such differences to present all tree types found in LSMS-ISA data used in this study.
Table A.2 Descriptive statistics on spatial distribution of households and plots with trees. Data source: Authors' calculations from LSMS-ISA datasets, World Bank (2015).
Country Average distance among Average distance to Spatial correlation index (Moran's I) households (km) nearest neighbor (km) Number of plots Number of plots Number of plots with Number of plots with trees for with trees with fruit tree tree cash crops timber or fuelwood ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Tanzania 572.16 [301.65] 21.87 [19.72] 0.035 0.025 0.149 0.105 ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Ethiopia 687.90 [2141.28] 25.51 [19.13] 0.017 0.026 0.042 0.025 ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Uganda 272.41 [484.53] 12.01 [7.16] 0.005 0.047 0.041 ⁎⁎⁎ ⁎⁎⁎ ⁎ Malawi 259.66 [176.05] 12.96 [7.48] 0.035 0.038 0.012 ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Nigeria 479.94 [250.42] 25.56 [16.97] 0.082 0.090 0.046 ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ All countries 2277.7 [1673.77] 21.01 [16.79] 0.200 0.279 0.167 0.124
Note: This table presents spatial information and correlation index for key variables of interest. Average distance among households is the distance among all households included in the sample in kilometers (km). Average distance to nearest neighbor refers to the average distance between households in the same geographic area (i.e. community). Since information on type of tree is not available for all countries, there are some missing values for the spatial correlation index. Standard deviation in brackets. Two-sided null hypothesis reported. ⁎ p b 0.10. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01.
Table A.3 Multivariate analysis of fruit trees. Data source: Authors' calculations from LSMS-ISA data sets, World Bank (2015).
Adoption analysis (Probit) Determinants of share of farmland with trees
Dep. variable: fruit trees on farm (yes = 1) Dep. variable: share of farm area with presence of fruit trees
(I) (II) Shapley value (III) (IV) Shapley value
Household controls 0.010 (3.63%) 0.010 (3.32%) ⁎ ⁎⁎ Household size 0.001 0.003 0.013 0.012 [0.003] [0.003] [0.006] [0.006] Number of children (b14 years old) −0.001 −0.002 0.003 0.002 [0.004] [0.003] [0.010] [0.010] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Head's age (years) 0.002 0.002 0.003 0.003 [0.000] [0.000] [0.001] [0.001] ⁎⁎ ⁎⁎⁎ Head female (yes = 1) −0.020 −0.020 −0.047 −0.046 [0.009] [0.008] [0.028] [0.028] ⁎ ⁎ ⁎ ⁎ Head education (years) 0.004 0.005 0.006 0.006 [0.003] [0.003] [0.003] [0.003] Assets and land 0.003 (1.25%) 0.096 (59.30%) ⁎⁎⁎ ⁎⁎⁎ Tropical livestock units (TLU) −0.006 −0.005 −0.001 −0.001 [0.002] [0.001] [0.000] [0.000] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎ Land owned (area - ha) 0.005 0.004 0.088 0.086 [0.001] [0.001] [0.032] [0.032] Geo- and climate variables 0.027 (9.60%) 0.006 (4.35%) Log population density around 20 km (people/sqkm) (2010) −0.000 0.000 0.000 0.001 [0.001] [0.001] [0.001] [0.001] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎ Tree cover % around 20 km (mean) (2010) 0.003 0.003 0.003 0.003 [0.001] [0.001] [0.001] [0.001] Fertile soil % around 20 km (mean) (2010) 0.000 −0.000 −0.002 −0.003 [0.005] [0.005] [0.004] [0.004] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎⁎ Log. annual mean temperature (C) 0.014 0.018 0.020 0.021 [0.005] [0.003] [0.008] [0.006] Log. annual precipitation (mm) 0.000 0.000 −0.000 −0.000 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 59
Table A.3 (continued) Adoption analysis (Probit) Determinants of share of farmland with trees
Dep. variable: fruit trees on farm (yes = 1) Dep. variable: share of farm area with presence of fruit trees
(I) (II) Shapley value (III) (IV) Shapley value
[0.000] [0.000] [0.000] [0.000] Country fixed effects (baseline country: Ethiopia) 0.084 (29.65%) 0.059 (36.81%) Malawi −0.041 −0.038 −0.060 −0.036 [0.026] [0.023] [0.054] [0.073] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎ Nigeria −0.211 −0.245 −0.131 −0.114 [0.031] [0.017] [0.063] [0.064] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Tanzania 0.159 0.143 0.556 0.513 [0.051] [0.050] [0.075] [0.082] ⁎⁎⁎ ⁎⁎⁎ Uganda −0.133 −0.128 −0.101 −0.080 [0.021] [0.008] [0.084] [0.105] Mean dependent variable 0.207 0.207 0.159 0.159 (Pseudo) R-squared 0.257 0.296 0.162 0.176 Observations 18,907 18,907 18,907 18,907 District/zone fixed effect No Yes No Yes
Note: This table presents the multivariate results for fruit trees on farms. Robust standard errors in brackets, clustered at strata level. Sampling weights used for all regressions. ⁎ p b 0.10. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01.
Table A.4 Multivariate analysis of tree cash crops. Data source: Authors' calculations from LSMS-ISA data sets, World Bank (2015).
Adoption analysis (Probit) Determinants of share of farmland with trees
Dep. variable: tree cash crops on farm (yes = 1) Dep. variable: share of farm area with presence of tree cash crops
(I) (II) Shapley value (III) (IV) Shapley value
Household controls 0.007 (2.70%) 0.006 (4.29%) ⁎ ⁎⁎ Household size 0.013 0.012 −0.003 0.002 [0.006] [0.006] [0.013] [0.012] Number of children (b14 years old) 0.003 0.002 −0.012 −0.015 [0.010] [0.010] [0.009] [0.010] ⁎⁎⁎ ⁎⁎⁎ ⁎ Head's age (years) 0.003 0.003 0.002 0.001 [0.001] [0.001] [0.001] [0.001] ⁎⁎ ⁎⁎ Head female (yes = 1) −0.047 −0.046 −0.055 −0.052 [0.028] [0.028] [0.025] [0.023] ⁎ ⁎ ⁎⁎ ⁎⁎ Head education (years) 0.006 0.006 0.008 0.009 [0.003] [0.003] [0.004] [0.004] Assets and land 0.001 (0.37%) 0.061 (44.08%) Tropical livestock units (TLU) −0.002 −0.002 −0.001 −0.001 [0.002] [0.002] [0.001] [0.001] ⁎⁎ ⁎⁎ Land owned (area - ha) −0.002 0.000 0.111 0.112 [0.003] [0.003] [0.043] [0.044] Geo- and climate variables 0.084 (30.35%) 0.021 (15.24%) Log population density around 20 km (people/sqkm) (2010) −0.002 −0.002 −0.000 0.001 [0.001] [0.001] [0.001] [0.001] ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Tree cover % around 20 km (mean) (2010) 0.007 0.008 0.007 0.006 [0.002] [0.001] [0.002] [0.002] ⁎ Fertile soil % around 20 km (mean) (2010) 0.012 0.008 0.001 −0.003 [0.007] [0.006] [0.005] [0.004] ⁎⁎ ⁎⁎ Log. annual mean temperature (C) 0.019 0.020 0.021 0.014 [0.009] [0.009] [0.015] [0.014] ⁎⁎ Log. annual precipitation (mm) −0.000 0.000 −0.000 −0.000 [0.000] [0.000] [0.000] [0.000] Country fixed effects (baseline country: Ethiopia) 0.11 (42.02%) 0.015 (10.85%) Malawi ⁎⁎⁎ ⁎⁎⁎ Nigeria −0.287 −0.295 −0.133 −0.037 [0.067] [0.060] [0.112] [0.109] ⁎⁎⁎ ⁎⁎⁎ Tanzania −0.198 −0.205 −0.074 −0.005 [0.043] [0.028] [0.101] [0.083] ⁎⁎⁎ ⁎⁎⁎ Uganda −0.189 −0.191 0.185 0.246 [0.041] [0.025] [0.183] [0.168] Mean dependent variable 0.200 0.200 0.261 0.261 (Pseudo) R-squared 0.192 0.263 0.122 0.141 Observations 8994 8975 8994 8994 District/zone fixed effect No Yes No Yes
Note: This table presents the multivariate results for tree cash crops on farms. Robust standard errors in brackets, clustered at strata level. Sampling weights used for all regressions. ⁎ p b 0.10. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01. 60 D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61
Table A.5 Multivariate analysis of trees for timber or fuelwood. Data source: Authors' calculations from LSMS-ISA data sets, World Bank (2015).
Adoption analysis (Probit) Determinants of share of farmland with trees
Dep. variable: trees for timber or fuelwood on Dep. variable: share of farm area with presence farm (yes = 1) of trees for timber or fuelwood
(I) (II) Shapley value (III) (IV) Shapley value
Household controls 0.009 (3.54%) 0.003 (1.75%) Household size −0.000 −0.001 −0.002 −0.012 [0.002] [0.002] [0.015] [0.017] Number of children (b14 years old) 0.004 0.003 0.011 0.015 [0.003] [0.003] [0.013] [0.013] ⁎⁎ ⁎⁎⁎ Head's age (years) 0.001 0.001 0.001 0.002 [0.000] [0.000] [0.001] [0.001] Head female (yes = 1) −0.003 −0.004 0.077 0.048 [0.006] [0.005] [0.071] [0.055] Head education (years) 0.001 0.002 0.001 −0.002 [0.001] [0.001] [0.005] [0.005] Assets and land 0.001 (0.65%) 0.117 (52.21%) Tropical livestock units (TLU) 0.000 0.000 0.017 0.011 [0.001] [0.001] [0.018] [0.016] Land owned (area - ha) 0.001 0.001 0.143 0.144 [0.001] [0.001] [0.099] [0.100] Geo- and climate variables 0.058 (20.73%) 0.009 (4.07%) ⁎ Log population density around 20 km (people/sqkm) (2010) 0.009 −0.001 0.105 0.063 [0.005] [0.004] [0.066] [0.047] ⁎ ⁎ Tree cover % around 20 km (mean) (2010) −0.000 −0.000 −0.005 −0.005 [0.000] [0.000] [0.003] [0.003] ⁎ ⁎⁎⁎ Fertile soil % around 20 km (mean) (2010) 0.024 0.026 0.418 0.331 [0.013] [0.010] [0.250] [0.196] ⁎⁎⁎ ⁎⁎ Log. annual mean temperature (C) −0.004 −0.003 0.012 0.017 [0.001] [0.001] [0.015] [0.015] Log. annual precipitation (mm) 0.000 0.000 −0.000 0.000 [0.000] [0.000] [0.000] [0.000] Country fixed effects (baseline country: Ethiopia) 0.19 (70.22%) 0.024 (11.05%) ⁎⁎⁎ ⁎⁎⁎ ⁎ Malawi −0.036 −0.039 0.308 0.275 [0.005] [0.004] [0.168] [0.168] Nigeria ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Tanzania 0.273 0.178 0.668 0.473 [0.065] [0.050] [0.194] [0.152] Uganda 0.027 0.014 0.285* 0.201 [0.021] [0.014] [0.157] [0.125] Mean dependent variable 0.032 0.032 0.115 0.066 (Pseudo) R-squared 0.257 0.302 0.190 0.225 Observations 16,392 16,361 16,392 16,392 District/zone fixed effect No Yes No Yes
Note: This table presents the multivariate results for tree for timber or fuelwood on farms. Robust standard errors in brackets, clustered at strata level. Sampling weights used for all regressions. ⁎ p b 0.10. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01.
References Berry, S., 1988. Property rights and rural resource management: the case of tree crops in West Africa. Cah. Sci. Hum. 24 (1), 3–16. Bhagwat, S.A., Willis, K.J., Birks, H.J.B., Whittaker, R.J., 2008. Agroforestry: a refuge for trop- Agrawal, A., Cashore, B., Hardin, R., Shepherd, G., Benson, C.S., Miller, D.C., 2013. Economic ical biodiversity? Trends Ecol. Evol. 23, 261–267. Contributions of Forests, Commissioned Background Paper for a Report to the UN Bluffstone, R., Yesuf, M., Uehara, T., Bushie, B., Damite, D., 2015. Livestock and private tree Secretary General. United Nations Forum on Forests, New York, NY. Holdings in Rural Ethiopia: the effects of collective action institutions, tenure security Angelsen,A.,Jagger,P.,Babigumira,R.,Belcher,B.,Hogarth,N.J.,Bauch,S.,Börner,J., and market access. J. Dev. Stud. 51, 1193–1209. Smith-Hall, C., Wunder, S., 2014. Environmental income and rural livelihoods: a glob- Booth, F.E.M., Wickens, G.E., 1988. Non-timber uses of selected arid zone trees and shurbs al-comparative analysis. World Dev. 64 (Supplement 1), S12–S28. in Africa. FAO Conservation Guide. Food and Agriculture Organization of the United Arnold, J.E.M., Dewees, P.A., 1995. Tree Management in Farmer Strategies: Responses to Nations, Rome. Agricultural Intensification. Oxford University Press, Oxford. Brottem, L., 2011. Rediscovering “terroir” in West African agroforestry parklands. Soc. Nat. Arnold, J.E.M., Dewees, P.A., 1997. Farms, Trees and Farmers: Responses to Agricultural In- Resour. 24, 553–568. tensification. Earthscan, London. Cattaneo, A., 2001. Deforestation in the Brazilian Amazon: comparing the impacts of mac- Ayana, A.N., Arts, B., Wiersum, K.F., 2013. Historical development of forest policy in Ethi- roeconomic shocks, land tenure, and technological change. Land Econ. 77, 219–240. opia: trends of institutionalization and deinstitutionalization. Land Use Policy 32, Center for International Earth Science Information Network - CIESIN - Columbia 186–196. University, International Food Policy Research Institute - IFPRI, The World Bank, and Bakkegaard, R.K., Agrawal, A., Animon, I., Bosselmann, A., Hogarth, N., Miller, D.C., Persha, Centro Internacional de Agricultura Tropical - CIAT, 2011. Global Rural-Urban Map- L., Rametsteiner, E., Wunder, S., Zezza, A., Zheng, Y., 2016. National Socioeconomic ping Project, Version 1 (GRUMPv1): Urban Extents Grid. NASA Socioeconomic Data Surveys in Forestry: Guidance and Questionnaires for Measuring the Multiple Roles and Applications Center (SEDAC), Palisades, NY http://dx.doi.org/10.7927/ of Forests in Household Welfare and Livelihoods FAO Forestry Paper No. 179. FAO, H4GH9FVG. CIFOR, IFRI, and the World Bank, Rome. Davis, B., Winters, P., Carletto, G., Covarrubias, K., Quiñones, E.J., Zezza, A., Stamoulis, K., Balk, D.L., Deichmann, U., Yetman, G., Pozzi, F., Hay, S.I., Nelson, A., 2006. Determining Azzarri, C., DiGiuseppe, S., 2010. A cross-country comparison of rural income generat- global population distribution: methods, applications and data. Adv. Parasitol. 62, ing activities. World Dev. 38, 48–63. 119–156. de Foresta, H., Somarriba, E., Temu, A., Boulanger, D., Feuilly, H., Gauthier, M., 2013. To- Barton, G.A., 2002. Empire Forestry and the Origins of Environmentalism. Cambridge Uni- wards the assessment of trees outside forests. Resources Assessment Working versity Press, Cambridge. Paper. FAO, Rome. D.C. Miller et al. / Forest Policy and Economics 84 (2017) 47–61 61
Degrande, A., Schreckenberg, K., Mbosso, C., Anegbeh, P., Okafor, V., Kanmegne, J., 2006. Mercer, D.E., 2004. Adoption of agroforestry innovations in the tropics: a review. Agrofor. Farmers' fruit tree-growing strategies in the humid Forest zone of Cameroon and Ni- Syst. 61, 311–328. geria. Agrofor. Syst. 67, 159–175. Milder, J.C., Hart, A.K., Dobie, P., Minai, J., Zaleski, C., 2014. Integrated landscape initiatives Dewees, P.A., 1994. Trees, Land, and Labor. The World Bank, Washington, DC. for African agriculture, development, and conservation: a region-wide assessment. Dewees, P.A., 1995a. Trees and farm boundaries: farm forestry, land tenure and reform in World Dev. 54, 68–80. Kenya. Africa 65, 217–235. Nair, P.R., 2007. The coming of age of agroforestry. J. Sci. Food Agric. 87, 1613–1619. Dewees, P.A., 1995b. Trees on farms in Malawi: private investment, public policy, and Nunn, N., Puga, D., 2010. Ruggedness: the blessing of bad geography in Africa. Rev. Econ. farmer choice. World Dev. 23, 1085–1102. Stat. 20–36. DiMiceli, C.M., Carroll, M.L., Sohlberg, C., Huang, C., Hansen, M.C., Townshend, J.R.G., 2011. Pattanayak, S., Evan Mercer, D., Sills, E., Yang, J.-C., 2003. Taking stock of agroforestry Annual Global Automated MODIS Vegetation Continuous Fields (MOD44B) at 250 m adoption studies. Agrofor. Syst. 57, 173–186. spatial resolution for data years beginning day 65, 2000–2010. Collection 5 Percent Perez-Heydrich, C., Warren, J.L., Burgert, C.R., Emch, M.E., 2013. Guidelines on the Use of Tree Cover, Collection 5. University of Maryland, College Park, MD. DHS GPS Data, Spatial Analysis Reports No. 8. ICF International, Calverton, Maryland. FAO/IIASA/ISRIC/ISSCAS/JRC, 2012. Harmonized World Soil Database (version 1.2). FAO, Peters, P.E., 2013. Land appropriation, surplus people and a battle over visions of agrarian Rome, and IIASA, Laxenburg, Austria. futures in Africa. J. Peasant Stud. 40, 537–562. Fay, C., Michon, G., 2005. Redressing forestry hegemony when a forestry regulataory Petersen, L., Sandhövel, A., 2001. Forestry policy reform and the role of incentives in Tan- framework is best replaced by an agrarian one. Forests, Trees and Livelihoods. 15, zania. Forest Policy Econ. 2, 39–55. pp. 193–209. Place, F., Garrity, D., 2015. Tree-based Systems in Africa's Drylands. World Bank Group, Fisher, M., 2004. Household welfare and forest dependence in Southern Malawi. Environ. Washington, DC, pp. 1–84. Dev. Econ. 9, 135–154. Pouliot, M., Treue, T., 2013. Rural people's reliance on forests and the non-forest environ- Franzel, S.C., 1999. Socioeconomic factors affecting the adoption potential of improved ment in West Africa: evidence from Ghana and Burkina Faso. World Dev. 43, tree fallows in Africa. Agrofor. Syst. 47, 305–321. 180–193. Garrity, D.P., Akinnifesi, F.K., Ajayi, O.C., Weldesemayat, S.G., Mowo, J.G., Kalinganire, A., Ribot, J.C., 1999. A history of fear: imagining deforestation in the West African dryland for- Larwanou, M., Bayala, J., 2010. Evergreen agriculture: a robust approach to sustain- ests. Glob. Ecol. Biogeogr. 8, 291–300. able food security in Africa. Food Sec. 2, 197–214. Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.-L., Sheil, D., Meijaard, E., Venter, M., Godoy, R.A., 1992. Determinants of smallholder commercial tree cultivation. World Dev. Boedhihartono, A.K., Day, M., Garcia, C., van Oosten, C., Buck, L.E., 2013. Ten principles 20, 713–725. for a landscape approach to reconciling agriculture, conservation, and other compet- Hansen, M.C., Potapov, P.V., Moore, R., Hancher, M., Turubanova, S.A., Tyukavina, A., Thau, ing land uses. Proc. Natl. Acad. Sci. 110, 8349–8356. D., Stehman, S.V., Goetz, S.J., Loveland, T.R., Kommareddy, A., Egorov, A., Chini, L.P., Scherr, S.J., 1995. Economic factors in farmer adoption of agroforestry: patterns observed Justice, C.O., Townshend, J.R.G., 2013. High-resolution global maps of 21st-century in Western Kenya. World Dev. 23, 787–804. forest cover change. Science 342 (6160), 850–853. Scherr, S.J., 2004. Building opportunities for small-farm agroforestry to supply domestic Ickowitz, A., Powell, B., Salim, M.A., Sunderland, T.C.H., 2014. Dietary quality and tree wood markets in developing countries. Agrofor. Syst. 61–62, 357–370. cover in Africa. Glob. Environ. Chang. 24, 287–294. Schroeder, R.A., 1999. Shady Practices: Agroforestry and Gender Politics in The Gambia. Kalaba, K.F., Chirwa, P., Syampungani, S., Ajayi, C.O., 2010. Contribution of agroforestry to Univ of California Press. biodiversity and livelihoods improvement in rural communities of Southern African Schroth, G., da Fonseca, G.A., Harvey, C.A., Gascon, C., Vasconcelos, H.L., Izac, A.M.N., 2013. regions. In: Tscharntke, T., Leuschner, C., Veldkamp, E., Faust, H., Guhardja, E., Bidin, Agroforestry and Biodiversity Conservation in Tropical Landscapes. Island Press, A. (Eds.), Tropical Rainforests and Agroforests Under Global Change: Ecological and Washington, DC. Socio-economic Valuations. Springer, Berlin, pp. 461–476. Sendzimir, J., Reij, C.P., Magnuszewski, P., 2011. Rebuilding resilience in the Sahel: Leach, M., Scoones, I., 2013. Carbon forestry in West Africa: the politics of models, mea- regreening in the Maradi and Zinder regions of Niger. Ecol. Soc. 16. sures and verification processes. Glob. Environ. Chang. 23, 957–967. Shorrocks, A.F., 2013. Decomposition procedures for distributional analysis: a unified López-Feldman, A., 2014. Shocks, income and wealth: do they affect the extraction of nat- framework based on the Shapley value. J. Econ. Inequal. 99–126. ural resources by rural households? World Dev. 64 (Supplement 1), S91–S100. Sunderlin, W.D., Dewi, S., Puntodewo, A., Muller, D., Angelsen, A., Epprecht, M., 2008. Why Manning, A.D., Fischer, J., Lindenmayer, D.B., 2006. Scattered trees are keystone structures forests are important for global poverty alleviation: a spatial explanation. Ecol. Soc. – implications for conservation. Biol. Conserv. 132, 311–321. 1–21. Mbow, C., Smith, P., Skole, D., Duguma, L., Bustamante, M., 2014a. Achieving mitigation World Bank, 2015. The Living Standards Measurement Study - Integrated Surveys on Ag- and adaptation to climate change through sustainable agroforestry practices in Africa. riculture (LSMS-ISA) http://www.worldbank.org/lsms. Curr. Opin. Environ. Sustain. 6, 8–14. Wunder, S., Angelsen, A., Belcher, B., 2014. Forests, livelihoods, and conservation: broad- Mbow, C., Van Noordwijk, M., Luedeling, E., Neufeldt, H., Minang, P.A., Kowero, G., 2014b. ening the Empirical Base. World Dev. 64, S1–S11. Agroforestry solutions to address food security and climate change challenges in Af- Zomer, R.J., Trabucco, A., Coe, R., Place, F., Van Noordwijk, M., Xu, J.C., 2014. Trees on rica. Curr. Opin. Environ. Sustain. 6, 61–67. Farms: an Update and Reanalysis of Agroforestry's Global Extent and Socio-ecological Meijer, S.S., Sileshi, G.W., Kundhlande, G., Catacutan, D., Nieuwenhuis, M., 2015. The role Characteristics. ICRAF, Bogor, Indonesia. of gender and kinship structure in household decision-making for agriculture and tree planting in Malawi journal of gender, agriculture and. Food Sec. 1, 54–76. Forest Policy and Economics 84 (2017) 62–71
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Trees for life: The ecosystem service contribution of trees to food production and livelihoods in the tropics☆
James Reed a,b,⁎,JoshvanVianena, Samson Foli a, Jessica Clendenning a,KevinYangc, Margaret MacDonald a, Gillian Petrokofsky d, Christine Padoch a, Terry Sunderland a,e a Center for International Forestry Research, Bogor, Indonesia b Lancaster Environment Centre, University of Lancaster, Lancaster, LA1 4YQ, UK c University of British Columbia, Vancouver, BC V6T 1Z4, Canada d University of Oxford, Oxford, UK e Center for Tropical Environmental and Sustainable Science, School of Earth and Environmental Sciences, James Cook University, Cairns, Qld 4870, Australia article info abstract
Article history: Despite expanding interest in ecosystem service research over the past three decades, in-depth understanding of the Received 27 April 2016 contribution of forests and trees to food production and livelihoods remains limited. This review synthesizes the cur- Received in revised form 5 January 2017 rent evidence base examining the contribution of forest and trees to agricultural production and livelihoods in the Accepted 12 January 2017 tropics, where production often occurs within complex land use mosaics that are increasingly subjected to concom- Available online 25 January 2017 itant climatic and anthropogenic pressures. Using systematic review methodology we found 74 studies investigating
Keywords: the effect of forest or tree-based ecosystem service provision on a range of outcomes such as crop yield, biomass, soil fi Ecosystem services fertility,andincome.Our ndings suggest that when incorporating forests and trees within an appropriate and con- Forests textualized natural resource management strategy, there is potential to maintain, and in some cases, enhance yields Agroforestry comparable to solely monoculture systems. Furthermore, this review has illustrated the potential of achieving net Food security livelihood gains through integrating trees on farms, providing rural farmers with additional income sources, and Livelihood security greater resilience strategies to adapt to market or climatic shocks. However, we also identify significant gaps in Food production the current knowledge that demonstrate a need for larger-scale, longer term research to better understand the con- tribution of forest and trees within the broader landscape and their associated impacts on livelihoods and food pro- duction systems. © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
1. Introduction provisioning services through the production of fuel and fibre (Rojstaczer et al., 2001; Vitousek et al., 1986). Furthermore, they can Forests provide a range of ecosystem functions that are fundamental aid in regulating pest control (Bale et al., 2008; Karp et al., 2013; Klein to sustaining terrestrial systems (Abson et al., 2014; Chazdon et al., et al., 2006) and supporting pollinating services (Kremen et al., 2002; 2009; MEA, 2005). These functions are thought to contribute vital sup- Klein et al., 2007). Finally, in Africa at least, the links between tree port to the provisioning of ecosystem goods and services needed to cover, access to food and improved dietary diversity are also becoming maintain human populations (Foley et al., 2005; Matson, 1997; Mery increasingly evident (Ickowitz et al., 2014; Johnson et al., 2013). et al., 2005). The contribution of forests to nutrient cycling (Power, The literature on ecosystem services has increased considerably in the 2010), soil formation (Pimentel and Kounang, 1998), climate (Daily last three decades and yet the concept remains contentious (Barnaud and and Matson, 2008), and water regulation (De Groot et al., 2002)is Antona, 2014). Early proponents of the ecosystem service concept now well established. Forests are also well recognised as important hab- (Ehrlich and Mooney, 1983; Westman, 1977) used the term to illustrate itats for faunal and floral resources that directly provide vital the depletion of natural resources through anthropogenic activities that would impede the capacity of ecosystems to provide vital services.
☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” These authors and others (Daily, 1997, Chapin et al., 2000)assertthat published at the journal Forest Policy and Economics 84C, 2017. such services are provided by nature and significantly contribute to ⁎ Corresponding author at: Center for International Forestry Research, Bogor, Indonesia. human well-being in numerous ways. E-mail addresses: [email protected] (J. Reed), [email protected] (J. van Others contest that it is the environmentally sensitive actions of Vianen), [email protected] (S. Foli), [email protected] (J. Clendenning), humans that facilitate the provision of ecosystem services (Gordon et [email protected] (K. Yang), [email protected] (M. MacDonald), [email protected] (G. Petrokofsky), [email protected] (C. Padoch), al., 2011; Sunderlin et al., 2005; Wunder, 2005) - discourse that is con- [email protected] (T. Sunderland). gruent with the motivation for researchers to develop and apply an
http://dx.doi.org/10.1016/j.forpol.2017.01.012 1389-9341/© 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71 63 economic valuation of ecosystems and the services they provide et al., 2015). While we acknowledge the value of tropical forests for (Costanza et al., 1998; Woodward and Wui, 2001). Subsequent policy the direct provisioning of food (i.e. fruits, nuts, leafy vegetables etc.) instruments, such as payments for ecosystem services (Wunder, 2008, that contributes to local dietary and nutritional quality (Powell et al., 2005) have been developed to financially compensate land managers 2015), this review is concerned with the indirect non-provisioning eco- for preserving ecosystem services and refraining from destructive system service (i.e. regulating and supporting services) contribution of land-use practices. More recently, researchers have posited that ecosys- forests and trees, and the effect these have on food production. tem services are co-produced by socio-ecological processes—that is a This systematic review synthesizes the current evidence base by mixture of natural, financial, technological, and social capital—typically assessing the contribution of trees and forests to food production requiring some degree of human intervention to support appropriation through ecosystem services derived from both within agroecosystems (Biggs et al., 2015; Palomo et al., 2016). and extant natural forests. We anticipate this synthesis will contribute While there remains some disagreement as to how ecosystem func- towards efforts that address the current controversies of independently tioning translates into the delivery of tangible benefits in the form of addressing food production and forest/biodiversity conservation and ecosystem services (Cardinale et al., 2012), it is now well acknowledged highlight the potential of integrating land uses within multifunctional that the preservation of biological diversity and associated habitats can landscapes to deliver a diverse suite of ecosystem services (Foli et al., maintain or enhance ecosystem service provision (Hooper et al., 2005; 2014; Glamann et al., 2015). Isbell et al., 2011; Lefcheck et al., 2015). As such, landscape management is increasingly considered to be best conceived through a holistic lens 2. Methods that encourages multi-functionality (O'Farrell et al., 2010; Reed et al., 2016; Scherr and McNeely, 2008; Vandermeer et al., 1998). In this re- We followed standard systematic review methodology, detailed in gard, multi-functionality typically refers to either spatial or temporal Foli et al. (2014), to identify and screen literature from a number of spe- segregation, or functional integration (Brandt, 2003). cialist databases, grey literature sources, and key institutional websites This review is concerned with the latter—the integration of multiple (Foli et al., 2014). All searches were conducted in English and covered functions within the same landscape—in this case, the contribution of for- publication years from 1950 to July 2015. Preliminary searches were ests and trees, and their associated ecosystem functions, to food produc- conducted to test the search terms and strategy in Web of Knowledge tion in the tropics. Food production systems globally have been greatly only. This initially yielded 321 hits. After expanding the number of intensified throughout the past century. As a consequence, primary for- search terms, the number of hits increased to 63,253. A final search ests,trees,andtheassociatedprovision of ecosystem services have suf- strategy (see: Foli et al., 2014 for protocol including detail on search fered sustained and ongoing decline (Foley et al., 2005; Power, 2010). strings employed) was determined which yielded 9932, which Furthermore, as the social and environmental costs of industrial food pro- duction have become better understood, it is increasingly recognised that this model cannot continue to be pursued sustainably (Foley et al., 2011; Godfray et al., 2010). Therefore, alternative strategies that reconcile biodi- versity conservation and food production warrant further consideration (Minang et al., 2014; Sayer et al., 2013; Sunderland et al., 2008). This is particularly pertinent in the tropics, where the majority of global biodi- versity hotspots occur (Myers et al., 2000). Yet these hotspots are highly susceptible to the drivers and impacts of global environmental change such as forest conversion, high levels of poverty, and food insecurity (Gardner et al., 2009; Laurance, 1999). Agriculture and forestry have traditionally been managed as sectori- al, and sometimes antagonistic, entities, often contributing to social and environmental conflicts. However, the two are inextricably interlinked. While the drivers of deforestation and forest degradation are complex and vary by region (Lambin et al., 2001), on a global scale agriculture is estimated to be the primary driver of deforestation (Foley et al., 2005, Scherr and McNeely, 2008, Gibbs et al., 2010), responsible for ap- proximately 80% of forest loss (Kissinger and Herold, 2012). These −1 losses account for emissions of 4.3–5.5 Pg CO2 eq. yr (Smith et al., 2014), which represents approximately 11% of total global carbon emis- sions (Goodman and Herold, 2014), accelerating climate change, and in turn inhibiting forests capacity to provide essential ecosystem services (Laurance et al., 2014). As such, a better understanding of the interac- tions between forest ecosystem services and agricultural production is fundamental to the sustainable management of terrestrial resources. This review was conceived around the notion that, despite a rapidly growing body of literature on the role and value of ecosystem services, the contribution of forests and trees—via ecosystem service provision—to adjacent or embedded food production systems in the tro- pics remains poorly understood. Furthermore, we speculate that the contribution of forests, in terms of ecosystem services provision, to food production systems may often be based on anecdotal evidence or may not be well supported with robust evidence of the “true” functional value. As such, this review assesses the contribution of trees and forests to food production in the tropics, where production often occurs within complex land use mosaics that are increasingly subjected to concomi- tant climatic and anthropogenic pressures (Gibbs et al., 2010; Steffen Fig. 1. Flow diagram showing the systematic screening process. 64 J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71 constituted the set of documents we worked with (see Fig. 1). The initial positive, negative, neutral or mixed effect on any reported food produc- searches were conducted in January 2014. An updated search was per- tion or livelihood component. Unsurprisingly, given the review focus on formed in July 2015 to account for additional articles produced during food production, all included studies reported a direct measure of the ef- the period of the initial literature screening process. All articles were fect of tree or forest presence on crop production or farm yields—except screened sequentially for relevance at title, abstract, and full text stages. in three cases where sufficient proxy measures of yields were explicitly given. These include two pest control studies (Gidoin et al., 2014; Karp 2.1. Study inclusion and exclusion criteria et al., 2013) and one pollination study (Blanche et al., 2006). Further analysis of the system-wide effects of trees/forests was per- At the title and abstract stage, studies were screened for relevance formed by aggregating all recorded outcomes for the effects of trees/for- and accepted for the next stage of assessment if they were studies with- ests. These system-wide effects of tree presence were classified as in the tropics that measured forest or tree-based ecosystem services and representative of an overall effect on livelihood outcomes. For example, agricultural output. trees may have had no effect on yields when compared to non-tree con- At full text screening, final study inclusion was determined if studies trols yet had a positive effect on soil fertility within the system, thus met the following three criteria: having a net positive system-wide effect. This would result in the Relevant study method/design: studies showed a transparent and study being documented as an overall (system-wide) positive effect of repeatable research design. trees and thus a positive livelihood outcome. Similarly, a negative effect Relevant study comparator: studies presented comparisons be- on yield and a positive effect on primary production would result in an tween agricultural systems with and without tree presence (either rep- overall (system-wide or livelihood) mixed effect of tree presence. Ap- licated or longitudinal comparison). pendix 1 provides a full list of the variables assessed in this review. Relevant study outcomes: studies measured and reported outcomes that showed a clear positive, negative or neutral effect of tree or forest 3. Results presence on ecosystem functions in agricultural systems. Studies were excluded from the review if they met one or more of 3.1. Review statistics the following criteria:
- Studied ecosystem services only at global scales. The initial 9932 articles were reduced to 1054 after title screening, - Exploratory studies, conceptual frameworks, non-empirical, or 178 after abstract screening and finally 62 articles for critical appraisal methods papers. and data extraction after full text screening. Updated searches conduct- - General forestry and agricultural policy briefs. ed in July 2015 identified a further 2481 articles, of which 36 were - Studies solely on the economic evaluation and accounting of ecosys- retained after full text screening. Twenty four articles were eliminated tem services. during critical appraisal—screened by a second reviewer to assess con- - Studies outside the tropics. formity to the inclusion/exclusion criteria—resulting in a total of 74 ar- - Studies solely on the contribution of wind pollination to crop pro- ticles in the final review. Fig. 1 summarizes the screening process. All duction. articles included in this review were published in peer-reviewed - Studies with relevant results but without transparent methodology journals, with the earliest retrieved published in 1991. or findings. 3.2. Geographic distribution and research focus Those articles accepted at full text were then critically appraised be- fore data extraction. A peer-reviewed protocol provides a detailed ac- A broad range of tropical countries were represented in this review. count of the research design, methods, and inclusion criteria (Foli et However, research was predominantly located in East and West Africa, al., 2014). South Asia (Indian sub-continent) and South America (Fig. 2, Table 1). The final suite of 74 studies investigated the roles of trees and forests 2.2. Data extraction on crop yields across a total of nine ecosystem services. However, the majority (n = 58) investigated bundled ecosystem service effects Data extraction was performed by all authors. Due to differences in (see: Renard et al., 2015) of trees and forests, resulting in 138 data reporting and use of terminology across the final suite of articles, eco- points (distributed across the nine ecosystem services and 74 studies) system services derived from forests or trees were grouped according (Table 1). Cumulatively, the most commonly studied ecosystem ser- to nine simplified categories for analysis (see Table 1). Similarly, an ar- vices were primary production and nutrient cycling, accounting for ticle often examined multiple ecosystem services and therefore report- 29% and 25% of the ecosystem services studied, respectively. These pat- ed multiple study outcomes. For the analysis of this review, outcomes terns were consistent across the regions with the exception of Australia for each ecosystem service reported in each article were grouped in 13 where both studies focused on pollination. The third most commonly categories (see: Fig. 5) by the presence or absence of trees having a studied ecosystem service varied across the regions – in Africa, resource
Table 1 regional distribution of ecosystem services studied.
Africa (n = 39) Asia (n = 12) Americas (n = 21) Australia (n = 2) Total
Primary production 19 14 7 40 Nutrient cycling 22 9 4 35 Pollination 5 4 3 2 14 Microclimate 7 6 1 14 Resource competition 8 4 1 13 Water retention 4 4 8 Soil formation 3 1 2 6 Pest control 4 1 5 Carbon storage 2 1 3 Total services studied 74 43 19 2 138 J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71 65
Fig. 2. Frequency plot showing study country distribution. competition (dis-service) (n = 8), in Asia, microclimate (n = 9), and in trees—were conducted over short timescales (b3 years, n = 58) (4). the Americas, pollination (n =4). As such, of the 54 genera of tree species recorded, the most frequently The study system characteristics were largely dominated by agrofor- represented were the common agroforestry taxa of Acacia, Gliricidia, estry studies (Fig. 4). Of the total 74 studies, 58 were agroforestry stud- Leucaena and Sesbania (represented in 12%, 11%, 6%, and 4% of studies ies, and only 5 of these were agroforestry systems under the forest respectively—for a full list of tree and crop species studied, see supple- canopy – the remaining 53 were trees introduced to the farm (typically mentary material). Of the studies that evaluated the contribution of alley cropping). Only 12 studies investigated the effect of spatially dis- off-farm forests and trees, eleven were researching the impact of forest tinct natural forest patches on agroecosystems, namely off-farm forests distance or diversity on pollination or pest control services. While most and trees – mostly consisting of studies utilizing agroforestry gradients of these were also within agroforestry systems, these were the few stud- (investigating yield outputs from a range of land use types from canopy ies that investigated ecosystem service provision within or from natural agroforestry to monoculture full sun systems) (see Figs. 3 & 4). Further- or semi-natural forest systems—as opposed to food systems that incor- more, we found that most studies—particularly those with planted porated planted trees. We found only nine long-term studies (≥7years)
Fig. 3. Figure showing a forest transition curve and the position along which the reviewed studies are placed according to their study system characteristics (above the transition curve) and corresponding food produced (below x axis). 66 J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71
Fig. 4. Scatter plot of study durations and forest proximity for different study types. and these were all on-farm or within research station experimental as often articles that reported depressed crop yields due to resource plots. Whereas studies that assessed off-farm provision of forest ecosys- competition effects of tree presence also reported (in discussion and tem services were all short term (≤3years).Fig. 4 clearly illustrates the conclusions) overall livelihood gains due to other economic benefits de- lack of long term, landscape-scale evaluations of forest ecosystem ser- rived from trees, such as the provision or sale of fuelwood, mulch, or vice provisioning. fodder for example. Hence, when examining the overall livelihood ef- fects of tree presence across the 74 articles in this review, the majority 3.3. The effect of tree presence on food production in the tropics report a positive effect (46%) which closely mirrors the effects on yield (47%) (Fig. 6). The overall trend across the studies shows that in the majority of The main difference when comparing the effects of trees on yields cases (52%) there was a net positive (47%) or neutral (5%) effect of with the overall livelihood effects of trees across all studies is the reduc- tree presence on food yields or food yield proxies. However, when the tion in the total negative effects from 36% for yield to 16% for overall results are disaggregated by region, there is a degree of variability livelihood effects, suggesting that a reduction in yields may be compen- (Fig. 5). For example, in the Americas and Asia, tree presence was sated by other benefits provided by trees to the farm system (Fig. 6). more likely to enhance food yields with positive effects of trees on yields This cost/benefit relationship—where the cost of crop yield losses is reported in 58% and 54% of studies for these regions respectively; while compensated by the overall benefits of incorporating trees—is consis- in Asia the opposite is the case, with the majority of studies (48%) tent across the study regions with Africa, Asia and the Americas reporting decreased food yields as a result of tree presence (Fig. 5). reporting negative effects of trees on crop yields in 33%, 48%, and 33% of studies but negative effects of trees on overall livelihood outcomes 3.4. The “overall livelihood” effect of tree presence in the tropics in only 15%, 24%, and 8% respectively (Fig. 7). In studies where trees were shown to have a positive effect on food Studies often investigated multiple ecosystem services and reported yield, the overall livelihood effect was also positive (86%) and never on multiple outcomes – for example, one study may investigate nutrient returned a negative outcome, although 11% of studies showed a mixed cycling and primary production and measure effects on differences in effect i.e. some negative and some positive effects on overall livelihood crop yield and soil fertility. Consequently, the final set of 74 articles re- outcomes. However, in the studies where trees decreased food yields, corded 138 data entries for ecosystem services studied and 164 data en- the overall livelihood effect were varied: in 37% of studies that showed tries for measured effects of trees. Due to inconsistencies in terminology trees having a negative effect on yield, livelihoods were also reduced; used across the studies, we developed thirteen broad categories of effect 59% of studies showed either a mixed effect or no change in livelihoods, variables. Given the review's primary focus on food production, some while one study showed that negative yield outcomes were fully com- measurement of yield was a prerequisite for inclusion and hence had pensated by improved overall livelihood outcomes (Fig. 8). a recorded outcome for all 74 final studies. Any other effects directly linked to tree/forest presence were also recorded, with the most widely 4. Discussion reported effects of trees other than yield being soil fertility and income (Fig. 6). Despite a significant increase in ecosystem service-related research By combining the empirical data and the self-reported anecdotal ev- in the past two decades (Abson et al., 2014), this review illustrates idence within the articles, the review team was able to broadly establish that there are clear gaps in the literature with regard to the contribution overall livelihood effects for each of the articles—i.e. whether there is a of tropical forest and tree-based ecosystem services to food production. net positive or negative effect of tree cover on livelihoods. While it has Principal amongst these is the lack of evidence for the contribution of to be noted that this was largely a subjective process and not always off-farm tropical forest patches to agricultural systems. Of the few stud- supported by empirical data, it was felt that this was a useful exercise ies identified, the majority—such as those conducted by Blanche et al., J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71 67
Fig. 5. Frequency plot showing tree effects on crop yields by: (a) regional distribution and Fig. 7. The overall livelihood effects of trees (determined by the authors by summing (b) study system. multiple system wide effects of trees) categorised by region (a) and study system (b).
Fig. 6. Frequency plot of the effect of trees/forest on multiple system components across all studies. Non-yield effects were broadly categorised by the authors.*Livelihood effects were categorised by the authors by summing multiple system wide effects of trees. 68 J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71
Fig. 8. Frequency plot comparing the direct effects of trees on crop yield and the overall livelihood effects reported across the study system as a result of tree/forest presence.
2006, Klein 2009, and Sande et al. 2009—used a forest distance gradient is the testing of such relationships over time. We agree with Pattanayak to establish the effects and thresholds for pollinator success as a sole (2009) and others (Bauch et al., 2014; Renard et al., 2015), that studies focus. that monitor how forests function over periods beyond the traditional While such studies are useful and clearly illustrate the importance of project cycle of 1–3 years are vital to assess the contribution of forests trees and forests for the delivery of a single ecosystem service, it is well to food production, livelihoods, and the long-term sustainability of inte- acknowledged that ecosystem services do not act in isolation (Boreux et grated landscape approaches (Barlow et al., 2011; Reed et al., 2016). al., 2013a, 2013b; Renard et al., 2015) and therefore studies that exam- While many tropical countries are represented in this review, there ine the interactions of multiple ecosystem services within multiple land is a clear geographical research bias towards India and East Africa. It use configurations are much needed. As such, the key finding of this sys- may be the case that the climatic and natural resource conditions of tematic review is that there is little clear evidence of the effect of multi- these regions make them particularly pertinent for ecosystem service ple interacting ecosystem services flowing from forest fragments to research; it may reflect the interests of donors funding primary re- food systems. This paucity of studies significantly limits our ability to search; there may be greater political will or existing national policies draw conclusions as to the value of forests and trees within the land- that support agroforestry system research; it may be the presence of re- scape to proximate agricultural systems. Therefore, despite our original search organisations in the region (i.e. ICRAF); or it may be a result of objective of attempting to quantify the contribution of off-farm forests other factors of which we are not aware. An important limiting caveat and trees to food production, the results presented in this review prin- of this review is that searches were conducted only in English. Conse- cipally reflect the contribution of trees to food production and liveli- quently, it is likely that searches in other languages would reveal more hoods at the farm scale only. studies from non-English speaking countries, providing a more even The temporal and spatial scales of the studies identified in this re- geographic distribution. One recommendation would be for future re- view point to further gaps in the current understanding of the longer- views to be performed in non-English languages, to complement and term contributions of forest and trees to food production. Although spa- build upon these findings. Furthermore, a review of temperate systems tial information was not always provided in the studies, the large major- would also complement the findings we have presented here. ity were conducted in either smallholder agroforestry systems This review indicates that the presence of forest and trees has vary- (typically 0.5–3.0 ha.) or research station small-scale experimental ing effects on food production, but that the majority of studies showed a plots (for example 20 × 18 m plots), and over a study period of less direct net positive or neutral effect of tree presence on crop yields. than three years. Much of the evidence in this review therefore provides When other factors are considered such as environmental impacts or a snapshot in time of ecosystem processes, therefore failing to recognise additional income derived from trees through sale of fuelwood for ex- the changes that can occur over space and time (Renard et al., 2015). ample, the overall livelihood benefit to land managers can buffer costs The assessment of ecosystem services is not easy and complexity will accrued through crop yield reductions. Even in Asia where a large pro- be increased when transitioning from local to landscape scales (Swift portion of studies showed the presence of trees was negatively et al., 2004), but given the extensive dialogue on ecosystem service pro- impacting crop yields (a finding that warrants further investigation), visioning as a contribution to long-term sustainable systems (Jordan, the overall net livelihood effect suggested that farmers could reduce 2013; Scherr and McNeely, 2008; Tilman et al., 2002), it seems clear negative impacts and gain a long-term benefitfromincorporating that further evidence on the spatiotemporal dynamics of ecosystem ser- trees on their farms as the total negative effects were greatly reduced vice provision to support such claims is both necessary and timely. (48% to 24%). Given the short term nature of the studies examined We strongly recommend that future research efforts attempt to here, it could be speculated that when examined over longer time scales bridge these gaps by moving beyond the farm gate, as it were. Research the broader benefits of maintaining trees would become more evident. that investigates the effects that tropical forests and forest patches have While this is an encouraging result, the evidence presented here is on spatially distinct agroecosystems would increase our understanding not sufficient to suggest that tree presence or incorporation will always of complex systems. This level of research is essential in order to further be the optimal management strategy for food and livelihood outcomes, dissolve the dichotomies of biodiversity conservation and food produc- and land managers should be encouraged to pursue a more nuanced ap- tion which often remain viewed as entities to be addressed individually proach to managing complex socio-ecological systems. It is important to (Glamann et al., 2015). A further requirement to aid our understanding note that many studies examined the effects of multiple ecosystem J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71 69 services on multiple outcomes, often with contrasting results. For exam- researchers in countries where efforts to integrate food production, live- ple, a study may reveal environmental gains from trees planted on lihood enhancement, and tree conservation are already underway. farms via improved soil fertility, but also report associated production However, it has also identified a number of key knowledge gaps, en- losses in terms of crop yield due to resource competition (Kidanu et abling us to provide the following recommendations for future re- al., 2005; Siriri et al., 2010). Similarly, one study reported an overall neg- search: Investigate the effect of off-farm trees and forest patches on ative effect but suggested this may be attributed to the fact that the sur- proximate food production systems; further examine spatiotemporal rounding forest matrix was intact and healthy and therefore the greater forest ecosystem service dynamics; assess how these services interact abundance of floral resources inhibited pollination success in the agro- with other system functions; and further develop appropriate instru- forestry system of interest (Boreux et al., 2013a, 2013b). A further ments for measuring and comparing ecosystem services. study reported mixed success: non-intensive systems were optimal in Current evidence on the association between forests, trees and food terms of pollination, however, proximity to forest was not significant production systems in the tropics lack the necessary precision to fully (Frimpong et al., 2011). A number of studies showed that net losses in inform practice and policy. A future research agenda that attempts to crop yield may be compensated by the additional biomass produced elucidate the above recommendations would enhance our understand- from the planted trees, resulting in an overall net gain (Asase et al., ing, providing further support for more integrated approaches to land 2008; Chauhan et al., 2010; Fadl, 2010). Moreover, it is clear from this management that seek to sustainably utilize rather than deplete natural review that the provisioning of individual forest ecosystem services to resources. food production do not act in isolation. Consequently, the potential socio-environmental costs and benefits need to be contextualized and Acknowledgements considered over time and space, with land use management strategies applied and adapted accordingly. This study is part of the CGIAR Research Program on Forests, Trees and Agroforestry (CRP-FTA). This collaborative programme aims to en- 5. Conclusion hance the management and use of forests, agroforestry and tree genetic resources across the landscape from forests to farms. CIFOR leads CRP- The study of forest and tree-based ecosystem services in the tropics FTA in partnership with Bioversity International, CATIE, CIRAD and the suffers from both a geographic and research focus bias, and is further International Center for Tropical Agriculture and the World Agroforest- limited by the propensity for small-scale and short-term evaluations. ry Centre. Funding for this study was provided by the United Kingdom's The relative dearth of studies prevents us from providing a definitive Department for International Development (DfID) (MTO 069018) and answer to our original research question—to what extent do forests the United States Agency for International Development (USAID) and trees support food production? There is insufficient evidence—most (MTO 069018). We are grateful to the guest editors and anonymous re- of which is not directly comparable— to assess the contribution of eco- viewers whose comments improved this manuscript. system services derived from forests to agricultural systems. The find- ings of this review very much reflect the contribution of trees to food Appendix A. Supplementary data production at the farm scale rather than the broader contribution of for- ests and trees within the landscape. To this end, we have generated a Supplementary data to this article can be found online at http://dx. database of 74 articles that demonstrate both positive and negative ef- doi.org/10.1016/j.forpol.2017.01.012. fect of trees on food yields and broader livelihood outcomes. Our find- ings suggest that when incorporating forests and trees within an appropriate and contextualized natural resource management strategy, References yields can be maintained or enhanced comparable to intensive mono- Abson, D.J., von Wehrden, H., Baumgartner, S., Fischer, J., Hanspach, J., Hardtle, W., culture systems. Furthermore, this review has illustrated the potential Heinrichs, H., Klein, A.M., Lang, D.J., Martens, P., Walmsley, D., 2014. Ecosystem ser- of achieving net positive gains through integrating trees on farms, pro- vices as a boundary object for sustainability. Ecol. Econ. 103:29–37. http://dx.doi. viding practitioners with additional income sources and greater resil- org/10.1016/j.ecolecon.2014.04.012. Asase, A., Wade, S.A., Ofori-Frimpong, K., Hadley, P., Norris, K., 2008. Carbon storage and ience strategies to adapt to market or climatic shocks. the health of cocoa agroforestry ecosystems in south-eastern Ghana. Africa Carbon Despite this, contemporary development pathways—particularly Cycle 131. within the tropics—often tend towards “conventional” approaches Bale, J.S., van Lenteren, J.C., Bigler, F., 2008. Biological control and sustainable food produc- tion. Philos. Trans. R. Soc. L. B Biol. Sci. 363:761–776. http://dx.doi.org/10.1098/rstb. to agriculture and food security that deplete the natural resource 2007.2182. base (Gibbs et al., 2010; Gibson et al., 2011; Roe et al., 2014; Barlow, J., Ewers, R.M., Anderson, L., Aragao, L.E.O.C., Baker, T.R., Boyd, E., Feldpausch, T.R., Steffen et al., 2015). Forest conservation rhetoric largely refers to Gloor, E., Hall, A., Malhi, Y., Milliken, W., Mulligan, M., Parry, L., Pennington, T., Peres, fi C.A., Phillips, O.L., Roman-Cuesta, R.M., Tobias, J.A., Gardner, T.A., 2011. Using learning the bene ts for the global community. Meanwhile, conservation of networks to understand complex systems: a case study of biological, geophysical and forests and trees at the local scale is often sold as generating other social research in the Amazon. Biol. Rev. 86:457–474. http://dx.doi.org/10.1111/j. tangible benefits to farmers and rural people through the provision- 1469-185X.2010.00155.x. Barnaud, C., Antona, M., 2014. Deconstructing ecosystem services: uncertainties and con- ing of ecosystem services. However, this review has highlighted that troversies around a socially constructed concept. Geoforum 56:113–123. http://dx. the current evidence of the latter—particularly with regard to food doi.org/10.1016/j.geoforurn.2014.07.003. production outputs—remains unclear. Research efforts are urgently Bauch, S.C., Sills, E.O., Pattanayak, S.K., 2014. Have we managed to integrate conservation and development? ICDP impacts in the Brazilian Amazon. World Dev. 64:S135–S148. required to strengthen the evidence base and provide clear, robust http://dx.doi.org/10.1016/j.worlddev.2014.03.009. data in order to support the transitioning to “alternative” approaches Biggs, R., Schlüter, M., Schoon, M.L., 2015. Principles for building resilience: sustaining to land management. Without strong evidence linking forests de- ecosystem services in social-ecological systems. Cambridge University Press. rived ecosystem services to food production and livelihood benefits Blanche, K.R., Ludwig, J.A., Cunningham, S.A., 2006. Proximity to rainforest enhances pol- lination and fruit set in orchards. J. Appl. Ecol. 43:1182–1187. http://dx.doi.org/10. there remains little incentive for food producers to acknowledge 1111/j.1365-2664.2006.01230.x. the need for forest conservation at the local and landscape scale. Fur- Boreux, V., Kushalappa, C.G., Vaast, P., Ghazoul, J., 2013a. Interactive effects among ecosys- tem services and management practices on crop production: pollination in coffee ag- ther evidence is required if we are to illustrate the potential local so- – fi roforestry systems. Proc. Natl. Acad. Sci. 110:8387 8392. http://dx.doi.org/10.1073/ cial and environmental bene ts that can be achieved through both pnas.1210590110. conserving trees within the landscape and incorporating them with- Boreux, V., Smitha, K., Cheppudira, K.G., Jaboury, G., 2013b. Impact of forest fragments on in food production systems. bee visits and fruit set in rain-fed and irrigated coffee agro-forests. Agric. Ecosyst. En- viron. 172:42–48. http://dx.doi.org/10.1016/j.agee.2012.05.003. This systematic review (and the accessible accompanying database) Brandt, J., 2003. Multifunctional Landscapes–Motives, Concepts and Perspectives Jesper provides a valuable resource for policy makers, practitioners, and Brandt and Henrik Vejre. 70 J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71
Cardinale, B.J., Duffy, J.E., Gonzalez, A., Hooper, D.U., Perrings, C., Venail, P., Narwani, A., Jordan, C.F., 2013. An Ecosystem Approach to Sustainable Agriculture. first ed. Springer Mace, G.M., Tilman, D.A., Wardle, D., Kinzig, A.P., Daily, G.C., Loreau, M., Grace, J.B., http://dx.doi.org/10.1007/978-94-007-6790-4. Larigauderie, A., Srivastava, D.S., Naeem, S., 2012. Corrigendum: biodiversity loss Karp, D.S., Mendenhall, C.D., Sandí, R.F., Chaumont, N., Ehrlich, P.R., Hadly, E.A., Daily, G.C., and its impact on humanity. Nature 489:326. http://dx.doi.org/10.1038/nature11373. 2013. Forest bolsters bird abundance, pest control and coffee yield. Ecol. Lett. 16: Chapin III, F.S., Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek, P.M., Reynolds, H.L., 1339–1347. http://dx.doi.org/10.1111/ele.12173. Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie, S.E., Mack, M.C., Díaz, S., 2000. Conse- Kidanu, S., Mamo, T., Stroosnijder, L., 2005. Biomass production of Eucalyptus boundary quences of changing biodiversity. Nature 405, 234–242. plantations and their effect on crop productivity on Ethiopian highland Vertisols. Chauhan, S.K., Sharma, S.C., Beri, V., Yadav, S., Gupta, N., 2010. Yield and carbon sequestra- Agrofor. Syst. 63, 281–290. tion potential of wheat (Triticum aestivum)-poplar (Populus deltoides) based agri- Kissinger, G., Herold, M., 2012. Drivers of deforestation and forest degradation. A Synth. silvicultural system. Indian J. Agr. Sci. 80 (2), 129–135. Rep. REDD+ Policymakers. Chazdon, R.L., Harvey, C.A., Komar, O., Griffith, D.M., Ferguson, B.G., Martínez-Ramos, M., Klein, A.M., 2009. Nearby rainforest promotes coffee pollination by increasing spatio-tem- Morales, H., Nigh, R., Soto-Pinto, L., Breugel, M.v., Philpott, S.M., 2009. Beyond reserves: poral stability in bee species richness. For. Ecol. Manag. 258 (9), 1838–1845. a research agenda for conserving biodiversity in human-modified tropical landscapes. Klein, A.-M., Steffan-Dewenter, I., Tscharntke, T., 2006. Rain forest promotes trophic inter- Biotropica 41:142–153. http://dx.doi.org/10.1111/j.1744-7429.2008.00471.x. actions and diversity of trap-nesting Hymenoptera in adjacent agroforestry. J. Anim. Costanza, R., d'Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, Ecol. 75:315–323. http://dx.doi.org/10.1111/j.1365-2656.2006.01042.x. S., O' Neill, R.V., Paruelo, J., 1998. The value of the world's ecosystem services and nat- Klein, A.M., Vaissiere, B.E., Cane, J.H., Steffan-Dewenter, I., Cunningham, S.A., Kremen, C., ural capital. Ecol. Econ. 25, 3–15. Tscharntke, T., 2007. Importance of pollinators in changing landscapes for world Daily, G.C., 1997. Nature's Services: Societal Dependence on Natural Ecosystems. Island crops. Proceedings of the Royal Society of London B: Biological Sciences 274 (1608), Press, Washington DC. 303–313. Daily, G.C., Matson, P.A., 2008. Ecosystem services: from theory to implementation. Proc. Kremen, Claire, Williams, Neal M., Thorp, Robbin W., 2002. Crop pollination from native Natl. Acad. Sci. 105:9455–9456. http://dx.doi.org/10.1073/pnas.0804960105. bees at risk from agricultural intensification. Proceedings of the National Academy De Groot, R.S., Wilson, M.a., Boumans, R.M.J., 2002. A typology for the classification, de- of Sciences 99 (26), 16812–16816. scription and valuation of ecosystem functions, goods and services. Ecol. Econ. 41: Lambin, E.F., Turner, B.L., Geist, H.J., Agbola, S.B., Angelsen, A., Bruce, J.W., Coomes, O.T., 393–408. http://dx.doi.org/10.1016/S0921-8009(02)00089-7. Dirzo, R., Fischer, G., Folke, C., et al., 2001. The causes of land-use and land-cover Ehrlich, P.R., Mooney, H.A., 1983. Extinction, substitution, and ecosystem services. Biosci- change: moving beyond the myths. Glob. Environ. Chang. 11, 261–269. ence 33, 248–254. Laurance, W.F., 1999. Reflections on the tropical deforestation crisis. Biol. Conserv. 91: Fadl, K.E.M., 2010. Effect of Acacia senegal on growth and yield of groundnut, sesame and 109–117. http://dx.doi.org/10.1016/S0006-3207(99)00088-9. roselle in an agroforestry system in North Kordofan state. Sudan. Agrofor. Syst. 78 (3), Laurance, W.F., Sayer, J., Cassman, K.G., 2014. Agricultural expansion and its impacts on 243–252. tropical nature. Trends Ecol. Evol. 29:107–116. http://dx.doi.org/10.1016/j.tree. Foley, J.A., Defries, R., Asner, G.P., Barford, C., Bonan, G., Carpenter, S.R., Chapin, F.S., Coe, 2013.12.001. M.T., Daily, G.C., Gibbs, H.K., Helkowski, J.H., Holloway, T., Howard, E.A., Kucharik, Lefcheck, J.S., Byrnes, J.E.K., Isbell, F., Gamfeldt, L., Griffin, J.N., Eisenhauer, N., Hensel, M.J.S., C.J., Monfreda, C., Patz, J.A., Prentice, I.C., Ramankutty, N., Snyder, P.K., 2005. Global Hector, A., Cardinale, B.J., Duffy, J.E., 2015. Biodiversity enhances ecosystem consequences of land use. Science 309:570–574. http://dx.doi.org/10.1126/science. multifunctionality across trophic levels and habitats. Nat. Commun. 6:6936. http:// 1111772. dx.doi.org/10.1038/ncomms7936. Foley, J.A., Ramankutty, N., Brauman, K.A., Cassidy, E.S., Gerber, J.S., Johnston, M., Mueller, Matson, P., 1997. Agricultural intensification and ecosystem properties. Science 277 (80): N.D., O'Connell, C., Ray, D.K., West, P.C., Balzer, C., Bennett, E.M., Carpenter, S.R., Hill, J., 504–509. http://dx.doi.org/10.1126/science.277.5325.504. Monfreda, C., Polasky, S., Rockström, J., Sheehan, J., Siebert, S., Tilman, D., Zaks, D.P.M., MEA, 2005. Ecosystems and human well-being. Millenium Ecosystem Assessment: Biodi- 2011. Solutions for a cultivated planet. Nature 478:337–342. http://dx.doi.org/10. versity Synthesis. Island Press, Washington, DC. 1038/nature10452. Mery, G., Alfaro, R.I., Kanninen, M., Lobovikov, M., 2005. Changing Paradigms in Forestry: Foli, S., Reed, J., Clendenning, J., Petrokofsky, G., Padoch, C., Sunderland, T., 2014. To what Repercussions for People and Nature (Techreport). extent does the presence of forests and trees contribute to food production in humid Minang, P.A., van Noordwijk, M., Freeman, O.E., Mbow, C., de Leeuw, J., Catacutan, D., and dry forest landscapes?: a systematic review protocol. Environ. Evid. 3, 15. 2014. Climate-Smart Landscapes: Multifunctionality in Practice (ASB Partnership for Frimpong, E.A., Gemmill-Herren, B., Gordon, I., Kwapong, P.K., 2011. Dynamics of insect The Tropical Forest margins). pollinators as influenced by cocoa production systems in Ghana. J. Polit. Econ. 5, Myers, N., Mittermeier, R.A., Mittermeier, C.G., Fonseca, G.A.B.d., Kent, J., 2000. Biodiversi- 74–80. ty hotspots for conservation priorities. Nature 403:853–858. http://dx.doi.org/10. Gardner, T.A., Barlow, J., Chazdon, R., Ewers, R.M., Harvey, C.A., Peres, C.A., Sodhi, N.S., 1038/35002501. 2009. Prospects for tropical forest biodiversity in a human-modified world. Ecol. O'Farrell, P.J., Reyers, B., Maitre, D.C.l., Milton, S.J., Egoh, B., Maherry, A., Colvin, C., Lett. 12, 561–582. Atkinson, D., Lange, W.d., Blignaut, J.N., Cowling, R.M., 2010. Multi-functional land- Gibbs, H.K., Ruesch, A.S., Achard, F., Clayton, M.K., Holmgren, P., Ramankutty, N., Foley, J.A., scapes in semi arid environments: implications for biodiversity and ecosystem ser- 2010. Tropical forests were the primary sources of new agricultural land in the 1980s vices. Landsc. Ecol. 25:1231–1246. http://dx.doi.org/10.1007/s10980-010-9495-9. and 1990s. Proc. Natl. Acad. Sci. 107:16732–16737. http://dx.doi.org/10.1073/pnas. Palomo, I., Felipe-Lucia, M.R., Bennett, E.M., Martín-López, B., Pascual, U., 2016. Chapter 0910275107. six-disentangling the pathways and effects of ecosystem service co-production. Gibson, L., Lee, T.M., Koh, L.P., Brook, B.W., Gardner, T.A., Barlow, J., Peres, C.A., Bradshaw, Adv. Ecol. Res. 54, 245–283. C.J.A., Laurance, W.F., Lovejoy, T.E., Sodhi, N.S., 2011. Primary forests are irreplaceable Pattanayak, S., 2009. Rough Guide to Impact Evaluation of Environmental and Develop- for sustaining tropical biodiversity. Nature 478:378–381. http://dx.doi.org/10.1038/ ment Programs (Working Paper No. 40–9). South Asian Network for Development nature10425. and Environmental Economics. Gidoin, C., Babin, R., Beilhe, L.B., Cilas, C., Hoopen, G.M.T., Bieng, M.A.N., 2014. Tree spatial Pimentel, D., Kounang, N., 1998. Ecology of soil erosion in ecosystems. Ecosystems 1: structure, host composition and resource availability influence mirid density or black 416–426. http://dx.doi.org/10.1007/s100219900035. pod prevalence in cacao agroforests in Cameroon. PLoS One 9. http://dx.doi.org/10. Powell, B., Thilsted, S.H., Ickowitz, A., Termote, C., Sunderland, T., Herforth, A., 2015. Im- 1371/journal.pone.0109405. proving diets with wild and cultivated biodiversity from across the landscape. Food Glamann, J., Hanspach, J., Abson, D.J., Collier, N., Fischer, J., 2015. The intersection of food Secur. 7:535–554. http://dx.doi.org/10.1007/s12571-015-0466-5. security and biodiversity conservation: a review. Reg. Environ. Chang. http://dx.doi. Power, A.G., 2010. Ecosystem services and agriculture: tradeoffs and synergies. Philos. org/10.1007/s10113-015-0873-3. Trans. R. Soc. B Biol. Sci. 365:2959–2971. http://dx.doi.org/10.1098/rstb.2010.0143. Godfray, H.C.J., Beddington, J.R., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F., Pretty, J., Reed, J., Van Vianen, J., Deakin, E.L., Barlow, J., Sunderland, T., 2016. Integrated landscape Robinson, S., Thomas, S.M., Toulmin, C., 2010. Food security: the challenge of feeding approaches to managing social and environmental issues in the tropics: learning 9 billion people. Science 327:812–818. http://dx.doi.org/10.1126/science.1185383. from the past to guide the future. Glob. Chang. Biol. 22:2540–2554. http://dx.doi. Goodman, R.C., Herold, M., 2014. Why maintaining tropical forests is essential and urgent org/10.1111/gcb.13284. for a stable climate. Cent. Glob. Dev. Work. Pap. Renard, D., Rhemtulla, J.M., Bennett, E.M., 2015. Historical dynamics in ecosystem service Gordon, A., Langford, W.T., White, M.D., Todd, J.A., Bastin, L., 2011. Modelling trade offs be- bundles. Proc. Natl. Acad. Sci. U. S. A. 112:13411–13416. http://dx.doi.org/10.1073/ tween public and private conservation policies. Biol. Conserv. 144:558–566. http:// pnas.1502565112. dx.doi.org/10.1016/j.biocon.2010.10.011. Roe,D.,Day,M.,Booker,F.,Zhou,W.,Allebone-Webb,S.,Kümpel,N.,Hill,N.A.O., Hooper, D.U., Chapin III, F.S., Ewel, J.J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J.H., Wright, J., Rust, N., Sunderland, T.C., Redford, K., Petrokofsky, G., 2014. Are alter- Lodge, D.M., Loreau, M., Naeem, S., Schmid, B., Setälä, H., Symstad, A.J., Vandermeer, native livelihood projects effective at reducing local threats to specified ele- J., Wardle, D.A., 2005. Effects of biodiversity on ecosystem functioning: a consensus ments of biodiversity and/or improving or maintaining the conservation status of current knowledge. Ecol. Monogr. 75, 3–35. of those elements?: a systematic review protocol. Environ. Evid. 3:6. http://dx. Ickowitz, A., Powell, B., Salim, M.A., Sunderland, T.C.H., 2014. Dietary quality and tree doi.org/10.1186/2047-2382-3-6. cover in Africa. Glob. Environ. Chang. Policy Dimens. 24:287–294. http://dx.doi.org/ Rojstaczer, S., Sterling, S.M., Moore, N.J., 2001. Human appropriation of photosynthesis 10.1016/j.gloenvcha.2013.12.001. products. Science 294 (80):2549–2552. http://dx.doi.org/10.1126/science.1064375. Isbell, F., Calcagno, V., Hector, A., Connolly, J., Harpole, W.S., Reich, P.B., Scherer-Lorenzen, Sande, S.O., Crewe, R.M., Raina, S.K., Nicolson, S.W., Gordon, I., 2009. Proximity to a forest M., Schmid, B., Tilman, D., van Ruijven, J., Weigelt, A., Wilsey, B.J., Zavaleta, E.S., leads to higher honey yield: another reason to conserve. Biol. Conserv. 142 (11), Loreau, M., 2011. High plant diversity is needed to maintain ecosystem services. Na- 2703–2709. ture 477:199–202. http://dx.doi.org/10.1038/nature10282. Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.L., Sheil, D., Meijaard, E., Venter, M., Johnson, K.B., Jacob, A., Brown, M.E., 2013. Forest cover associated with improved child Boedhihartono, A.K., Day, M., Garcia, C., van Oosten, C., Buck, L.E., 2013. Ten principles health and nutrition: evidence from the Malawi demographic and health survey for a landscape approach to reconciling agriculture, conservation, and other compet- and satellite data. Glob. Heal. Sci. Pract. 1:237–248. http://dx.doi.org/10.9745/GHSP- ing land uses. Proc. Natl. Acad. Sci. U. S. A. 110:8349–8356. http://dx.doi.org/10.1073/ D-13-00055. pnas.1210595110. J. Reed et al. / Forest Policy and Economics 84 (2017) 62–71 71
Scherr, S.J., McNeely, J.A., 2008. Biodiversity conservation and agricultural sustainability: Swift, M.J., Izac, A.M., van Noordwijk, M., 2004. Biodiversity and ecosystem services in ag- towards a new paradigm of “ecoagriculture” landscapes. Philos. Trans. R Soc. L. B ricultural landscapes—are we asking the right questions? Agric. Ecosyst. Environ. 104 Biol. Sci. 363:477–494. http://dx.doi.org/10.1098/rstb.2007.2165. (1), 113–134. Siriri, D., Ong, C.K., Wilson, J., Boffa, J.M., Black, C.R., 2010. Tree species and pruning regime Tilman, D., Cassman, K.G., Matson, P.A., Naylor, R., Polasky, S., 2002. Agricultural sustain- affect crop yield on bench terraces in SW Uganda. Agrofor. Syst. 78, 65–77. ability and intensive production practices. Nature 418, 671–677. Smith, P., Bustamante, M., Ahammad, H., Clark, H., Dong, H., Elsiddig, E.A., Haberl, H., Vandermeer, J., van Noordwijk, M., Anderson, J., Ong, C., Perfecto, I., 1998. Global change Harper, R., House, J., Jafari, M., et al., 2014. Agriculture, forestry and other land use and multi-species agroecosystems: concepts and issues. Agric. Ecosyst. Environ. 67, (AFOLU). Climate Change 2014: Mitigation of Climate Change. Contribution of Work- 1–22. ing Group III to the Fifth Assessment Report of the Intergovernmental Panel on Cli- Vitousek, P.M., Ehrlich, P.R., Ehrlich, A.H., Matson, P.A., 1986. Human appropriation of the mate Change. Cambridge University Press. products of photosynthesis. Bioscience 36, 368–373. Steffen, W., Richardson, K., Rockström, J., Cornell, S.E., Fetzer, I., Bennett, E.M., Biggs, R., Westman, W.E., 1977. How much are nature's services worth. Science 197 (80):960–964. Carpenter, S.R., de Vries, W., de Wit, C.A., et al., 2015. Planetary boundaries: guiding http://dx.doi.org/10.1126/science.197.4307.960. human development on a changing planet. Science 347 (80), 1259855. Woodward, R.T., Wui, Y.-S., 2001. The economic value of wetland services: a meta-analy- Sunderland, T.C.H.C.H., Ehringhaus, C., Campbell, B.M.M., 2008. Conservation and devel- sis. Ecol. Econ. 37, 257–270. opment in tropical forest landscapes: a time to face the trade-offs? Environ. Conserv. Wunder, S., 2008. Payments for environmental services and the poor: concepts and pre- 34:276–279. http://dx.doi.org/10.1017/S0376892908004438. liminary evidence. Environ. Dev. Econ. 13, 279–297. Sunderlin, W.D., Angelsen, A., Belcher, B., Burgers, P., Nasi, R., Santoso, L., Wunder, S., 2005. Wunder, S., 2005. Payments for environmental services: some nuts and bolts. CIFOR Livelihoods, forests, and conservation in developing countries: an overview. World Occas. Pap. 42:32. http://dx.doi.org/10.1111/j.1523-1739.2006.00559.x. Dev. 33:1383–1402. http://dx.doi.org/10.1016/j.worlddev.2004.10.004. Forest Policy and Economics 84 (2017) 72–82
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Forest ecosystem services derived by smallholder farmers in northwestern Madagascar: Storm hazard mitigation and participation in forest management☆
Radhika Dave a,⁎,EmmaL.Tompkinsa, Kate Schreckenberg b a Geography and Environment, University of Southampton, University Road, Southampton SO17 1BJ, UK b Engineering and the Environment, University of Southampton, University Road, Southampton SO17 1BJ, UK article info abstract
Article history: Tropical dry deciduous forests provide numerous ecosystem services yet their contribution to agricultural Received 13 May 2016 production remains underexplored. We address this research gap by quantifying the broader suite of ecosystem Received in revised form 22 August 2016 services that support small holder farmers and identifying farmers' knowledge of storm hazard reduction Accepted 2 September 2016 benefits provided by forest fragments in Madagascar. We survey 240 households and interview eight key Available online 18 September 2016 informants to identify household and community responses in two communities with contrasting forest cover trajectories. Using multivariate statistics, results show a heavy dependence on forests for food and raw materials Keywords: Forest ecosystem benefits and a majority of the respondents holding a positive view of hazard mitigation services provided by forest Livelihood benefits fragments. Education levels, earning an income from forest based tourism and honey production are the only Storm hazard reduction predictors of participation in forest management. Positive view of the hazard reduction benefits derived from Community forest management: forest policy forests could be due to external influences or personal observations, and together with barriers to participation in Madagascar forest management need to be further investigated to better link forest management to reduced hazards risks. These findings are significant for forest management policy, as local knowledge and rationale for decisions are in- strumental in the success of decentralized forest management and maintenance of vital forest benefits to farmers. © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
1. Introduction livelihoods through provisioning services, less work has been done on assessing the importance of the hazard reduction functions of forests Tropical dry deciduous forests, one of the most threatened biomes (Howe et al., 2014). In this paper we address this research gap by pro- on the planet, hold a high density of mammalian biomass and provide viding a detailed case study of the hazard mitigation services generated essential ecosystem services to people (Lerdau et al., 1991; Maass and valued by smallholder farmers in a tropical deciduous forest mosaic et al., 2005). These forests provide water regulation and pollination in northwest Madagascar. services as well as food, timber, water for irrigation and non-timber Land cover change, particularly deforestation, is hypothesized to forest products (Maass et al., 2005). Ecosystem services, defined as the increase flood risk (Bradshaw et al., 2007) and is seen as a primary benefits people derive from nature (MA, 2005), provide an anthropo- driver of soil erosion and consequent siltation of irrigation channels centric motivation for sustaining nature to support human needs and and agricultural fields (Bakoariniaina et al., 2006; Minten and society (Fisher et al., 2014). Provisioning services such as timber, food Randrianarisoa, 2012). Inland forests have been argued to reduce the and non-timber forest products that can be used directly by people frequency and magnitude of floods, and there is some consensus that are more easily linked to human needs (Daily et al., 1997; Barbier compared with other land uses, tropical forests reduce peak flows et al., 2010; Wunder et al., 2014). However, forests, wetlands and coast- from small catchments during small to medium rainfall events al habitats also provide several regulating services that aid in disaster (Bruijnzeel, 2004; Alila et al., 2009; Tan-Soo et al., 2014). In the bioengi- risk reduction by decreasing the exposure of communities to hazards neering literature, forests, especially in mountainous areas, are consid- such as floods and storm surge (Sudmeier-Rieux et al., 2006; Brauman ered to provide protection for exposed communities from rockfalls, et al., 2007; Laurance, 2007; Martin and Watson, 2016). While much debris flows, erosion, floods and shallow landslides (Brang et al., 2001; attention has been paid to the role of forests in supporting rural Dorren et al., 2004; Alila et al., 2009). Flood risk is also mediated by human decisions about land use and land cover change including the type and location of farms, urban and semi-urban areas, plantations, ☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” and industrial areas (Wisner et al., 2004; Wheater and Evans, 2009). published at the journal Forest Policy and Economics 84C, 2017. fl ⁎ Corresponding author. Thus the exact relationship between forest cover and changes in ood E-mail address: [email protected] (R. Dave). frequency and magnitude and consequent impacts on people varies
http://dx.doi.org/10.1016/j.forpol.2016.09.002 1389-9341/© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82 73 between sites, with both biophysical and social elements influencing this relationship (Bruijnzeel, 2004; Blöschl et al., 2007). Forests Two of the approaches commonly relied upon in the biophysical assessment of the effects of forest loss on hydrological processes are 1 paired catchment studies and process based modelling (Wilk et al., 2001; Bruijnzeel, 2004; DeFries and Eshleman, 2004; Krishnaswamy Ecosystem Services et al., 2012; Kuraś et al., 2012). Despite a lack of application in the devel- Timber 1 oping world, studies using these approaches have yielded some consen- Honey sus on the role of forest cover in reducing flood hazards and different Fuel (e.g. firewood, forms of erosion and sediment yield in some situations. For instance, it charcoal) fl Farmers and is understood that total annual water yield ( ooding) increases with Food (e.g. wild yam) Ecosystem management Services the percentage of forest biomass lost after conversion and that dry decisions Water season flow can decrease with time as groundwater replenishment 2 decreases after a number of years (Bruijnzeel, 2004; Kuraś et al., Flood regulation Erosion control 2012). In general these findings are site specific, and remain difficult Debris flow control to extrapolate to other areas or larger scales as there is too much varia- fi tion in ndings (Bruijnzeel, 2004; DeFries and Eshleman, 2004). Thus, as Cultural services Van Dijk et al. (2009) and Calder and Aylward (2006) state, there are no Tourism income simple causal relationships between forest cover change and changes in 3 4 floods or erosional impacts. 4 Ecosystem services generated by seasonally dry tropical deciduous forests (TDF) are some of the most understudied set of socio- Landuses (non ecological interactions (Maass et al., 2005) particularly in the develop- forests) ing world. Seasonally dry forests have seen widespread transformation by people and are considered as the most threatened of tropical forests Fig. 1. The pathways of ES flows from forests (1) to farmers include the use of raw (Miles et al., 2006; Becknell et al., 2012). Madagascar's dry deciduous materials, e.g. timber, honey production, fuelwood, food, tourism income and forests form one of 200 ecoregions identified as ecosystems with high recognition of regulatory services such as flood and erosion control (2). Such uses and global conservation value that are also facing critical threats (Olson benefits have consequent land use impacts (3) which feed back into ES flows and fi fl and Dinerstein, 1998). Threats to these forests differ in different regions, bene ts to farm households and in uence forest management decisions (4). Our fi research questions focus on the uses and perceived regulatory services derived by with res and conversion for agriculture the most important direct farmers and linkages to forest management decisions. threats in Africa (Geist and Lambin, 2002; Lambin et al., 2003; Miles et al., 2006). We argue that if the role of forests in reducing excess sedimentation and debris flows and floods in rural areas is to be better • What is the relationship between the acknowledged livelihood understood, a more focused investigation of the environmental benefits of ecosystem services and the willingness of people to knowledge base of local communities in forest frontier regions is need- actively protect forests (using a proxy variable of engaging in forest ed. Unlike the use of hydrological models or paired catchment studies, management group activities)? we employ a local knowledge and perceptions’ approach to assess agricultural risk reduction benefits derived from regulating services provided by forests as it is these perceptions that will shape local land We then present the results on the frequency of use of various forest use decisions and the effectiveness of forest management policies. In ecosystem services, whether farmers’ acknowledge flood and sedimen- this paper, using an ecosystem services’ lens, we employ household tation hazard reduction benefits of forest fragments found locally across surveys in seven villages in Madagascar as a means to identify local the study area, and how these uses and attitudes translate into valuation benefits derived from seasonally dry TDFs landscapes, and how hazard of forests through participation in forest management activities for the mitigation is perceived as an ecosystem benefit valued by farmers in site with existing forest management groups. We conclude by forest frontier areas (Fig.1). We hypothesize that: discussing the implications of these findings for different aspects of forest management policies, including decentralized governance of fi 1) Farmers in seasonally dry forest mosaics derive livelihood bene ts forest resources, and implications for linking forest management to from forest patches, including hazard mitigation services. agricultural hazard reduction policies and livelihood benefits. 2) Farmers’ understanding of the forest-hydrological cycle linkages is associated positively with less exploitative uses of forests and with 2. Study area perceiving the hazard mitigation benefits of forests, especially with fl ood regulation. We study this problem in the Boeny region of northwestern fi 3) Deriving forest use bene ts and perceiving hazard mitigation Madagascar, which experiences a strong seasonal variation in precipita- fi bene ts of forests motivates participation in community forest tion that influences all aspects of the agropastoral rural lifestyle. The management groups. yearly average rainfall is 1700 mm, with a distinct rainy season during We first describe the study area, methods of data collection and November to March (Funk et al., 2015). The region is characterized by analyses performed, then focus on answering three research questions, small-scale farming and a population dependent upon the surrounding which address the hypotheses above: landscape for everyday needs. In addition to farming, land cover in- cludes seasonally dry broadleaf deciduous forest patches, lakes, raphia • What are the forest ecosystem services, including hazard mitigation wetland remnants, and grasslands dominated by the endemic palm, services, and benefits to livelihoods derived by small-holder farmers known locally as satrana. Additionally, fruit trees such as the jackfruit in seasonally dry deciduous forest zones? tree and papaya are found abundantly in areas of human habitation, • How do farmers understand the relationship between forest cover with much of the produce being transported to the city of Mahajanga and water regulation, and how is this associated with their attitude (nearest urban center). Rice farming dominates agricultural production, towards forest ecosystem benefits including hazard mitigation which can be typically characterized as small-holder subsistence, services? primarily rain-fed, farming. Maize and cassava form the secondary 74 R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82
Fig. 2. Map showing the location of the study area within northwestern Madagascar. The red dots indicate various sites at which households were surveyed, the bold boundaries encom- pass the fokontany of Antanambao and Mariarano, within the boundary of Mariarano commune. Source: Landsat-TM 5, Year: 2009.
staples in the region. Cattle ownership is an important aspect of the tropical storms and cyclones affecting Madagascar. Second, there need- cultural identity of the ethnic Sakalava, the largest ethnic group of the ed to be natural resource dependent farmers present, who by the nature region, and is a sign of wealth and status. However over the last decade of their primary occupation are exposed to rainfall variability. In cattle theft has become a major threat to this traditional source of in- Madagascar’s Mahajanga II district the predominant farming practice come security and many people have seen their cattle numbers decline is small scale, rain fed agriculture, much of it in low-lying areas and steeply. thus exposed to both variability in rainfall and hazards associated with Our study involved household surveys and key informant interviews heavy rains. Third, there needed to be forest users. The residents of in seven villages across two “fokontany” (three villages in Mariarano and the two fokontany, Mariarano and Antanambao, actively utilize natural four in Antanambao) within the commune of Mariarano in western resources from the surrounding landscape and forests. The two Madagascar’s Mahajanga II district. “Fokontany” is a local level adminis- fokontany fall in two separate sub-catchments. One further reason for trative unit comprised of villages, hamlets or neighborhoods and has selecting this commune was the opportunity to compare between elected officials. Antanambao and Mariarano (the latter being the local fokontany with and without community or other type of formalized “capital” of Mariarano Commune) occupy a surface area of 400 km2 management of forests. Mariarano fokontany has seen forests under (40,000 ha) with a total population of 3539 (Fig. 2). Out of the community management since 2000. Forests in Mariarano fokontany 492 km2 of forests originally present in the commune of Mariarano faced pressures from fire, and forests being cleared for maize farming over 90% is degraded or very degraded according to satellite imagery or cattle grazing, which led to the establishment of community forest and ground verifications, with only about 28 km2 of dense forest frag- management groups, and these pressures continue for forests in ment present on the Ankatsabe massif within Mariarano fokontany Antanambao fokontany where to date there are no formal management (PGM-E and GOM, 2013). Data were collected during two phases of efforts to reduce these threats. fieldwork: in September 2014, and a second period of five months from May to October 2015. 3.2. Data collection 3. Methods As key informants (KI) are valuable sources of information and 3.1. Village selection provide explanatory context to the study, they were chosen based upon consultation with the head of the fokontany, the recently retired To undertake research relating to the associations between forest school director who is well known in the fokontany and the technical ecosystem services, including hazard mitigation benefits, and farming director with the GIZ1 supported Programme Germano-Malagache pour livelihoods, three main factors were important in site selection. First, l’Environnment (PGM-E) project. Key informant interviewees included the study region needed to experience hazards linked to heavy rainfall officers of the fokontany administrative council, and senior officers of events of different magnitudes and type over the last fifteen years. In the study region, there is an annual rainy season that is influenced by 1 GIZ - Deutsche Gesellschaft für Internationale Zusammenarbeit R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82 75 the two community forest management groups in Mariarano. Subse- use another type of service. Chi square tests of independence were quently, a household survey with closed and open-ended questions also conducted to test whether the two fokontany differed in their was conducted to identify the nature of agricultural livelihoods, the ecosystem service uses. risks faced by farmers during the rainy season caused by heavy rains, To identify farmers’ knowledge of forest benefits for agricultural and the value for forest ecosystem goods and services. Specifically, production, specifically hydrological services, we derived a composite households were asked whether they used particular ecosystem score from a series of Likert statements (Table 1). These statements services, their perception of whether the hazards of flooding, sedimen- draw upon hypothesized relationships between forest cover and the tation and debris flow were reduced by nearby forests, and, using Likert hydrological cycle in the forest hydrology literature (Kuraś et al., scales, their understanding of the relationships between forests and 2012; Lima et al., 2014) and from studies of local perceptions (Wilk, rainfall, and forests and river flows. 2000; Meijaard et al., 2013). We refer to this composite score as the Data were collected from seven villages using household surveys “Water Regulation Indicator”. Internal consistency reliability of scale and key informant interviews. As there is no list of households living responses using Cronbach’s alpha test puts this composite indicator in each village, we used the electoral list obtained from the fokontany within the acceptable range for exploratory analyses with Cronbach’s chief to use as our sampling frame. Many questions in our survey deal α =0.66(Gliem and Gliem, 2003; Asano et al., 2006). with identification of hazard experiences or lean season over a period To assess whether people valued the role of forests we used mem- of five years, thus, we set 26 years as the minimum age of the household bership in the forest management group as the dependent variable in head, in order to ensure our sample had households who had been a logistic regression. We use membership in the local forest manage- independently farming for at least five years, 21 being the average age ment group as a proxy for taking action as, in this area, it is the most at which households farm land independently from their parents likely option for people to take collective action to manage the threats (KI 4). We set the upper limit at 70 to ensure we interviewed people to and uses of the forest patches. Madagascar adopted forest decentral- who were still actively farming. Within these constraints we selected ization legislation in the late 1990s subsequent to which there is a households randomly using a random number generator. A full house- strong emphasis on encouraging communities to establish community hold survey with 240 household heads was conducted after pilot testing forest management groups (Antona et al., 2004). Independent variables the survey instrument. 146 households from Mariarano and 94 from included: standard socio-economic indicators; positive perception of Antanambao were interviewed, representing approximately 22.5% the role of forests in mitigating hazards; and, benefits derived from of the adult population aged 26 to 65 in the two fokontany.In the use of other ecosystem services, such as food or raw materials. Madagascar, as in many traditional African societies household heads are usually male, thus the majority of our respondents were men, other 4. Results than in situations with single women or female-headed households. To identify prevalence of ecosystem service use, hereafter referred to 4.1. Household characteristics as ES use, we asked respondents about twelve ecosystem services that can be categorized into different types based upon whether they are Households in our study area (N = 240) are predominantly subsis- valued as: food and raw materials as provisioning services because of tence farmers, with a majority growing all three staple crops: rice, direct use; nature tourism as income generating options; and, cultural maize and cassava (59%), and 35% engage in some form of wage labour services because of spiritual or religious use options. This list of services to complement their income. Average household size is 5 individuals was adapted from literature and piloted during the test phase to ensure (std. dev. = 2.69). Over 76% of households have 1–6 children under it was complete (Sodhi et al., 2010; Fagerholm et al., 2012). To identify the age of fifteen and 56.7 households % have 1–2adults(15–65 years whether farmers recognized and valued forest regulating services, we of age). Almost 30% of the households surveyed own no cattle, 46.2% posed two categorical questions on the perceived links between local own 1–5 heads of cattle and 2.9% own over 50. On average 10.4 cups forest cover and hazard reduction benefits, eliciting responses on flood of rice are consumed each day. Literacy levels are generally low with and sedimentation reduction as regulating services as hazard reduction 28% of household heads reporting no formal education and 64% attend- benefits of forests; bringing the total number of ES discussed in this ing but not necessarily completing primary schools. study to fourteen. To assess participation in forest management, two methods had to be used as no forest management group exists in Antanambao 4.2. Forest ecosystem services, including hazard mitigation services, and fokontany.InMariaranofokontany, where minimum distance from livelihood benefits households to forests is 430 m and forests have been under community control since 2000, participants were simply asked whether they were Here we present results on the ecosystems services that are most members of either of the two forest management groups (known in and least valued by farmers based upon the proportion of respondents Malagasy by the abbreviation “VOI”). For Antanambao where there is who answered “yes” to using these, whether there are differences in no forest management group (and minimum distance to nearest forest ES use by sites, and how socio economic factors may affect respondents' is 1 km), we asked respondents if they would be willing to become use of various services. Table 2 presents the proportions of respondents members if there were such a group (variable “hypotheticalVOI”). A in the two sites that are dependent upon these services. significant proportion (90%, N = 95) of Antanambao participants said they would be willing to become VOI members. This contrasts with Mariarano, where only 40% of survey participants (N = 137) noted Table 1 that they were actually members. As the location of the forests used Likert scale statements used to create a composite indicator for water regulation services. – by Antanambao residents is farther and more scattered than in Responses for each statement ranged from 1 - do not agree at all to 5 agree a lot. Mariarano, 90% participation is likely to be an over estimate. Thus, we 1 The forest cover in this region plays an important role in bringing rains restricted our analysis of participation in forest management groups to 2 The rainy season here is generally sufficient for replenishing the water flow in the site where these currently exist, the Mariarano fokontany. the river each year 3 There is more stable supply of water in the river, streams and lakes because of the forest cover in this region 3.3. Data analysis 4 Forests are important for providing reliable supply/availability of water for irrigating your rice fields Correlation analyses were performed to identify whether respon- 5 Forest cover in this region plays an important role in supporting your agricultural production capacity dents who used specific ecosystem services were more or less likely to 76 R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82
Table 2 Proportions of survey respondents using different forest ecosystem services from two sites with and without formalized forest management regimes in Madagascar.
Site Antanambao Mariarano (No forest management group present) (Two forest management groups existing)
Yes No Yes No
i Food (%) (%) N (%) (%) N Honey 51.1 48.9 92 45.5 54.5 145 Wild vegetable/fruits 79.3 20.7 92 71.9 28.1 139 Fish 91.2 8.8 91 73.4 26.6 139 Tenrecs 30.4 69.6 92 37.1 62.9 143
ii Raw materials Fuelwood 96.7 3.3 92 97.8 2.2 137 Polewood 95.6 4.4 91 95.0 5.0 139 Wood for charcoal 19.6 80.4 92 32.2 67.8 143 Raphia, satrana 86.8 13.2 92 83.4 17.9 140
iii Cultural services Medicinal plants 80.4 19.6 92 83.3 16.7 138 Sacred spaces 74.7 25.3 91 44.8 55.2 143 Religious spaces 75.8 24.2 91 52.4 47.6 143 Tourism benefits 20.9 79.1 91 21.0 79.0 143
iv Hazard reduction Believe forests reduce flooding 46.7. 53.3 92 64.2 35.8 142 Forests reduce sediments/debris flow. 64.1 35.9 92 78.7 21.3 141
The household survey identified a much greater level of dependence charcoal (26.4% of total respondents) do so predominantly as an income upon the extraction of raw materials from the forest than that of food source, with 69.6% selling half or more of the charcoal produced. None of items, cultural services, or hazard mitigation benefits: on average the household characteristics examined such as years farming, total 91.6% and 93.1% of the respondents in Mariarano and Antanambao re- household size, cups of rice consumed per day, cattle owned, or educa- spectively extract fuelwood, timber and plants like raphia or satrana tion levels are associated with engaging in charcoal production. Char- from the forest and surrounding mosaic compared to 56.6% and 63.1% coal production is banned within the community-managed forests in respectively of the respondents who engage in honey production, catch- Mariarano, yet a significant association was found between living in ing fish or harvesting tenrecs, a small insectivorous mammal, and wild Mariarano and practicing charcoal production than in Antanambao vegetables. Residents of Antanambao are more likely to report fishing where there is no such rule in place (Chi Square = 4.5 and p = 0.034). than those of Mariarano (Chi square = 11.1, p =0.001).Asignificantly While 76.7% of all respondents fish, only 19.8% of those who practice smaller proportion of respondents acknowledge extracting wood to fishing sell half or more of their catch. In contrast, of those who depend produce charcoal (31.1% in Mariarano and 19.4% in Antanambao) upon the forests for honey production (N = 97), 59.7% sell anywhere (Table 2). Amongst the raw materials used, polewood is used by 95.3% from half to all of their honey production (Fig. A.1). Honey is seen as of the respondents, with all reporting that they extract polewood for both a commodity that can be consumed at home and sold to augment household needs only, and not for sale. In contrast, those who produce income, with a litre of raw honey selling for 4000–6000 Ariary (1.22 to
Table 3 Socio economic characteristics as determinants of different ecosystem service uses (N = 240).
Ecosystem services used Socioeconomic variables B S.E. Sig. Exp (B) 95%C.I.for EXP (B)
Lower Upper
Use of raphia, satrana Years farming 0.047 0.018 0.008 1.048 1.012 1.085 Education level −0.561 0.304 0.065 0.571 0.315 1.035 Adults 15–65 0.581 0.246 0.018 1.788 1.104 2.896 Constant 2.129 0.793 0.007 8.404 Charcoal production Years Farming −0.026 0.015 0.082 0.974 0.946 1.003 Number of staple crops grown 0.977 0.306 0.001 2.657 1.457 4.844 Constant −3.734 0.948 0 0.024 Seeing income benefits from tourism Education level 0.617 0.276 0.025 1.852 1.079 3.18 Wage income 0.522 0.172 0.002 1.685 1.204 2.359 Constant −2.96 0.89 0.001 0.052 Honey production Total household size 0.24 0.1 0.016 1.271 1.046 1.545 Cattle owned 0.346 0.13 0.008 1.41 1.1 1.82 Constant −1.309 0.613 0.033 0.27 Hunting tenrecs Wage income 0.493 0.163 0.002 1.637 1.19 2.251 Adults 15–65 0.284 0.136 0.037 1.329 1.017 1.736 Constant −2.748 0.808 0.001 0.064 R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82 77
Hazard mitigation benefits in the form of reduced flooding and reduction in the amount of sediment and debris flowing onto rice fields are valued by 58.1% and 73.0% of the respondents in our survey. We do not find any association between benefits from food, raw materials or tourism income and viewing forests as important for hazard reduction. Our survey finds that the fokontany to which an individual belongs influences the value for hazard mitigation services. In Mariarano, the site with forest management, people are significantly more likely to value these benefits, as compared to Antanambao, the site without forest management (Chi square = 8.07 and 6.01, df 1, p = 0.005 and 0.014 for flooding and sedimenta- tion respectively). A simple explanation for this is possibly the fact that while the villages that make up Mariarano fokontany are consid- erably closer to the small hills that are forested, in Antanambao, the landscape is described variously as a “bowl” or basin, which tends to collect water and is seen to flood each year (KI8, August 2015). In addition,asisevidentfromFig. 1, there is less forest cover around the sites where people farm and live in Antanambao. Income generated by being involved in tourism related activities is a benefit seen by only 20.9% of the respondents. The main sources of – Fig. 3. This figure shows the level of dependence upon the four different raw materials tourism income are the three research camps set up each July August. (fuelwood, charcoal, polewood and raphia/fiber plants) discussed here in comparison to Common sources of employment are tourist guides, camp guards, and the food items (honey, fish, wild vegetables/fruits and tenrecs). For example, through transporting baggage between sites, selling food items and lo- households who do not depend upon any of the food items, still rely on 1, 2 or 3 raw cally produced handicrafts such as baskets and mats woven from fiber materials (left hand panel). from satrana palm fronds. The small number of beneficiaries of tourism may be a function of the hiring capacity of the seasonal research based tourism existing in Mariarano. The secondary road leading to Mariarano 1.82 USD) in the dry season when the quality of honey produced is is impassable in the rainy season, and this would limit a longer tourist optimum. Greater total household size and to a lesser degree, financial season. 79% of household heads that benefit from local tourism have capital (proxied by the number of cattle owned) significantly influence at least some formal primary school education. Education is positively engagement in honey production (Table 3). Honey is also used for correlated with acquiring benefits from tourism, and interestingly our medicinal purposes, nutrition and for traditional rituals in Mariarano; results also show that respondents who benefit from tourism are also uses which have allowed members of the Mariarano VOI Tanteraka to likely to hunt tenrecs during the open season in Mariarano fokontany convince others in the fokontany to support forest protection and (r = 0.197, p = 0.003). regeneration in order maintain this benefit in the short and long term 81.9%, 55.2% and 60.9% of the household’s surveyed valued medici- (pers. comm. J.E.R, KI 3 and 4). To date there is no organized cooperative nal plants, sacred areas and areas important for religious ceremonies. through which honey producers can get a stable and fair price. Individ- Medicinal plants are collected locally and within the forests by a major- ual shopkeepers and small business owners with ties to the city of ity of those interviewed and used to treat common stomach ailments Mahajanga dominate the trading channels though there are efforts un- and fevers. The continuation of these uses and beliefs is balanced against derway by local development groups like the PAGE/GIZ program to sup- the need for land, timber, and other resources, and against the pressures port producers through setting up a cooperative, a honey processing of migratory groups who may not hold the same beliefs. Indeed we find unit and facilitating access to markets for its members (pers. comm., that respondents in Antanambao are significantly more likely to hold J.E.R).2 beliefs in sacred sites in the surrounding landscape than those in All of the respondents who hunt tenrecs (34.5% of respondents), and Mariarano (Chi-square = 21.674, df = 1, p b 0.001) collect wild vegetables and fruits (75.5%) report doing so for household Socio-economic characteristics of households such as cattle owner- consumption only. The number of adults (aged 15–65) in a household is ship as an indicator of relative wealth, the average cups of rice eaten a small but significant positive factor influencing the likelihood of per day as an indicator of food needs per family, the length of time hunting tenrecs (Table 3) indicating that those households with a spent farming in years, the total household size and education levels higher number of able-bodied adults tend to take part in this activity do not show any relationship with beliefs about hazard mitigation as a more than others. However, hunting may be an opportunistic activity, forest service, contrary to what may be hypothesized based upon results providing a source of protein during the open season for tenrec collec- from other studies (Sodhi et al., 2010). In Antanambao, there is signifi- tion rather than a regular dietary need. Those who hunt tenrecs are cant positive relationship between education and those who benefit also likely to harvest wood for charcoal from the surrounding landscape from tourism (Pearson’s correlation, r = 0.218, p = 0.038). Overall (Pearson’s r = 0.27, p b 0.001). Interestingly hunting tenrecs, collecting our results support our first hypothesis that seasonally dry tropical wild vegetable and fruits and relying upon raphia and other plants for forests are important for livelihood benefits and for hazard mitigation making mats and roofs, are seen as significant predictors of considering services as seen by farmers. forests as important safety net providers during times of stress such as illness or a bad harvest (Table A.1). Over 44% of the participants depend upon three or four dietary items from the forest and surrounding 4.3. Knowledge of the relationships between forests and generation of habitats while over 82% depend upon three or four of the raw materials ecosystem services (Fig. 3). Three important questions are analysed here: first, do farmers understand the relationship between forests and the hydrological cycle? Second, does an understanding of this relationship lead to less 2 J.E.R, Mr. Jean Eric Rajaobelinirina, Technical Advisor, Programme d'Appui à la Gestion de l'Environnement (PAGE/GIZ) – Antenne Boeny, Deutsche Gesellschaft für extractive forest uses? Third, does a good understanding of these Internationale Zusammenarbeit (GIZ) GmbH; personal communication, March 30, 2016. linkages lead to valuing the hazard mitigation benefits of forests? 78 R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82
4.3.1. Farmers’ understanding of the relationship between forests, the management group, the proxy used for taking action (Table 4). Adding hydrological cycle and agricultural production the two variables that measure hazard mitigation benefits does not Most participants score high on the water regulation composite indi- improve the model output. None of the ES used are significant cator created from a series of Likert statements (Table 1) with the medi- predictors for Antanambao, however as already stated we have a likely an score being 20 out of a maximum of 25 for the total sample. The over estimate of people willing to participate in forest management mean score for this composite indicator is 19.6, (n = 231) indicating which would influence the statistical tests. As seen from results in that respondents in both sites understand the significance of forests Section 4.2, collecting timber is not an economic activity, and all of the for their daily lives and livelihood through the regulation of water respondents who collect wild vegetables or tenrecs do so for household absorption, flow and availability for agricultural production. We do consumption. These results – that only direct income generating activi- not find education as a significant factor in scoring highly on this ties, specifically honey production and tourism that rely on good natural indicator, with age the only household characteristic seen as a predictor forests, are seen to influence participation - are explained to a certain of respondents’ knowledge of the links between forest cover and water degree by the socioeconomic attributes characterizing the household regulation, and that too only in Mariarano. (Table 3). While hazard experiences may be of a sufficient magnitude, the lack of a causal relationship between positive valuation of flood re- 4.3.2. Forest – hydrological cycle linkages and engaging in exploitative duction benefits from forests and taking action to sustain these benefits forest uses for the long term by protecting forests reflects to some degree the vari- We hypothesized that having a good understanding of the forest- ability in these services and the complexity of the linkages between for- hydrological cycle linkages should be associated positively with less ests and the water cycle as mentioned in Section 1 and the basic reality exploitative uses of forests. Producing charcoal is seen as an illicit that people are more willing to act on tangible benefits in such scenarios activity especially in Mariarano, though to what extent is questionable. where daily needs dominate decision making of the majority. Education Nevertheless it is one of the most important proximate causes of levels and cattle ownership are the only socioeconomic, household broader forest loss (Ahrends et al., 2010), thus we use engagement in characteristics that predict whether respondents participate in forest charcoal production as an indicator of exploitative uses of the forest. management in Mariarano (Table 4). Overall for the two sites together engaging in charcoal production is negatively associated with scoring higher on water regulation indicator 5. Discussion at the 90% significance level (r = −0.115, p = 0.08). Engaging in charcoal production is seen to be a significant negative predictor of 5.1. Forest ecosystem services, including hazard mitigation services, and understanding the linkages between forests and water regulation for re- livelihood benefits derived by smallholder farmers spondents in Mariarano (Table A.2). Our results support our assumption that a more positive understanding of the forest-water cycle linkages Our results indicate a strong dependence upon food, raw materials would result in less exploitative uses of forests. and cultural benefits of dry forest ecosystems in both our sites, reinforc- ing existing research findings that seasonally dry tropical forests play an 4.3.3. Forest – hydrological cycle linkages and valuing forests for hazard important role in the daily lives of forest frontier communities (Maass reduction benefits et al., 2005). We find that farmers in northwestern Madagascar recog- The third hypothesis that we test here is whether a greater level of nize the role of local forests in reducing sediment and debris flow, and understanding of the linkages between forest cover and water regula- in reducing the magnitude of floods in agricultural fields. Scientists, tion is associated with the likelihood of holding a positive attitude policymakers and communities in different regions of the world share towards the hazard mitigation benefits derived from forests. Our results these perceptions of the hazard mitigation benefits of forests to differing indicate a small, but significant relationship at the 90% significance level degrees (Chomitz and Kumari, 1998; Wilk, 2000; Balmford et al., 2002; between higher scores on the water regulation indicator and a positive Maass et al., 2005; Silvano et al., 2005; Hauck et al., 2013). Our study attitude of the flood hazard mitigation benefits of forests (r = 0.117, provides insights on how forest edge communities in small catchment p = 0.07, N = 226), however no such association is evident with the areas view the role of forests in reducing storm hazards, adding to the attitude towards sedimentation reduction by forests. Disaggregating sparse literature on storm hazard mitigation benefits of forests in the the two sites, we find that the water regulation indicator is a significant tropics (Maass et al., 2005; Meijaard et al., 2013). Significant perception predictor of perceiving flood reduction benefits from forests for respon- of forests as safety net predictors during times of stress correspond to dents in Mariarano fokontany (p = 0.09), but not in Antanambao results from other studies that demonstrate reliance by households (Table A.3). upon forest resources in the immediate aftermath of a natural disaster or other household income reducing events (Shackleton and 4.4. Do farmers value forest ecosystem benefits sufficiently to take action Shackleton, 2004; Völker and Waibel, 2010; Liswanti et al., 2011). Vari- and if not why not? ous ethnic groups in Madagascar place a spiritual value on nature as a link between the living and the ancestors; these cultural norms serve Honey production and benefiting from tourism are the only two for- as means to continuing the kinship with the departed, which in some est benefits’ variables that significantly predict participation in a forest cases are linked to resource conserving behaviour such as a ban on
Table 4 Determinants of participation in forest management for Mariarano fokontany (n = 138).
Determinants of participation B S.E. Sig. Exp (B) 95% C.I. for EXP (B)
Lower Upper
Honey production 0.877 0.414 0.034 2.403 1.068 5.408 Places that provide income benefits from tourism 1.290 0.530 0.015 3.634 1.287 10.260 Forests reduce floods belief 0.585 0.452 0.195 1.795 0.741 4.349 Forests reduce sedimentation/debris flow belief −0.402 0.525 0.444 0.669 0.239 1.873 Education level 0.934 0.381 0.014 2.544 1.207 5.365 Cattle owned 0.433 0.169 0.011 1.542 1.106 2.149
Nagelkerke R square = 0.22. R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82 79 hunting lemurs or keeping some forests intact, thus providing a refuge as important providers of nectar and pollen necessary for honey pro- for threatened species and habitats (Horning Rabesahala, 2004; Jones duction (Sande et al., 2009) thus benefiting honey production from et al., 2008; von Heland and Folke, 2014). These social and cultural the forest and satrana dominated grasslands as seen in our research norms may see some erosion with time, yet, amongst the predominant- (KI 3, pers. comm. JER). Valuing forests as important for honey produc- ly Sakalava ethnic group in Mariarano commune we can see the tion is a significant determinant of supporting community management ancestral norms in place and influencing many people value for cultural based as shown here and supported by studies in other regions (Amare services of deciduous forests. et al., 2016). By participating in forest management and being interested in the short and long term sustainability of accessing the joint benefits 5.2. Knowledge of the relationships between forests, generation of from honey production, these users reflect one of the attributes of ecosystem services and support for community forest management successful local organization made by Ostrom et al. (1999).Ourfindings also corroborate those of other studies that show the importance of Our results on the understanding held by farmers of the linkages financial and human capital as essential elements of livelihood diversi- between forest cover, rainfall, absorption of water and the supply of fication strategies (Chopra, 2002). water in the streams are corroborated by Wilk (2000) who find The second forest dependent factor that is linked to participation in similar locally held knowledge of these linkages in two watersheds forest management groups in our sites is earning income from tourism in rural India and Thailand. Local knowledge of forests and water linked activities and business opportunities. Income from tourism is regulation links as shown by our results and these other studies seen to influence participation in forest management in other regions, correspond to broader scientific understanding of the role of forests for example, in Nepal (Mehta and Heinen, 2001). We see that receiving in delivering hydrological services (Bruijnzeel, 2004; Brauman tourism benefits in Mariarano, the site with two forest management et al., 2007). However, a better understanding of these linkages is groups is a predictor also for hunting tenrecs in the open season, sug- related to the valuation of the floodregulationservicesofferedby gesting that those who benefit from tourism are also better informed forests, but not with reduction of excessive sedimentation or debris and more able to participate in collecting these resources during the flows. Nevertheless we find that an awareness of hazard reduction open collection season in March. Additionally, we find that education benefits does not necessarily translate into willingness to actively is an important determinant in benefiting from income opportunities support or participate in forest management. This runs counter to brought by tourism in the area. A minimal literacy level may be a strong recent research which suggests that farmers in developing countries indicator of the ability of an individual to participate in activities, e.g. who value soil erosion reduction services generated by forests are to set up small businesses, to seek out employment with others, to more willing to participate in forest management, for example, volunteer for trainings as forest guides and even a requirement as through demonstrating a higher willingness to pay for forest being a forest guide requires the ability to read and communicate with management (Danquah, 2015; Amare et al., 2016). outsiders, thus favouring those with higher education levels (Gezon, We suggest that factors that prevent people from taking action are as 2014), which may be correlated with higher wealth and other assets important now as they were when decentralization efforts began in (Coria and Calfucura, 2012). We have shown that simply using ecosys- Madagascar and elsewhere (1990s, early 2000 for Madagascar). These tem services to garner wider support for forest management is not likely factors include: marginalization of segments of the community, the eco- to succeed unless there are significant tangible benefits to forest depen- nomic costs of membership payments however small these fees may be dent communities in the vicinity of these forests. Furthermore, the within the local context; distrust of external actors who are seen as the relationships between the ability to obtain a forest based income and source of formalizing community forest management and instituting engagement in forest management groups are not always linear and regulations against the clearance of forests for charcoal or grazing land need to be further investigated. (Agarwal, 2001; Aymoz et al., 2013; Cullman, 2015); and a tension Education and relative wealth are two of the socioeconomic between customary norms to govern the use of forests and engaging characteristics of households that influence participation in forest in formal forest management (Cullman, 2015). Furthermore, it may management groups in our study, as in the case of Burkina Faso simply be too difficult for some members of the community to spare and Sri Lanka (Nuggehalli and Prokopy, 2009; Sodhi et al., 2010; the time required to attend meetings and participate in activities such Coulibaly-Lingani et al., 2011). Education levels are important not as reforestation programs as members of a forest management group, only as indicators of formal knowledge, but as channels for empow- thus leading to the so called “free rider” phenomenon in community erment to create the capacity to participate in group decision making forest management (Klooster, 2000). Further detailed empirical work and being heard, thus creating facilitating conditions for individuals is needed to identify, under what socio-economic or cultural circum- to (Nuggehalli and Prokopy, 2009) participate in forest management stances, local farmers’ knowledge of risk mitigating services generates groups. Relative wealth influences everything from power dynamics a positive or a negative engagement with formal forest management to the ability to invest in the financial or human capital as seen from groups. several studies around the world (Leach et al., 1999; Armitage, Indeed, the majority of studies that investigate factors influencing 2005). These factors ultimately influence an individual’s capacity to participation in community forest management groups in the tropics engage in business ventures to gain from tourism or other activities identify demographic variables or forest derived economic benefits like honey production. as important determinants (Lise, 2000; Dolisca et al., 2006; Coulibaly-Lingani et al., 2011; Méndez-López et al., 2015). While our 6. Conclusions study finds income benefits derived from forests to be significant determinants of support for forest management, we do not see this Through this research we contribute to the wider body of knowledge relationship between management support and heavy dependence on forest ecosystem services, expanding the understanding of storm upon the forests for raw materials, food or cultural uses, contrary to hazard mitigation services of dry, deciduous forests, the livelihood earlier studies in India, Burkina Faso and Kenya (Lise, 2000; benefits derived by smallholder farmers from these forests, and how Coulibaly-Lingani et al., 2011; Musyoki et al., 2013), which show that these influence local support for forest management initiatives. We high dependence upon forest resources influences individual choices show that not only do farmers derive livelihood benefits from seasonal- to participate in forest management groups. Instead our results indicate ly dry tropical forest fragments in Madagascar, but they also perceive a that the only two significant predictors of participation in forest positive role for forest cover in hydrological processes including management groups are: activities that generate income and need sedimentation and flood hazard control. However, this extensive use forests, namely honey production and tourism. Natural forest are seen of forest services does not overwhelmingly translate into a willingness 80 R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82 to take action to support forest management, and this is an important participation in forest management. Unless we are better able to implication for forest policy and management. We demonstrate the im- identify what type of ecosystem service benefits motivate involvement portance of factors such as securing complementary income from forest in forest management and what the barriers to participation may be, it based honey production and tourism, relative wealth and the education is unlikely that appropriate forest management institutions will emerge levels to participation in forest management. These results reflect the in the case study region and elsewhere. Local knowledge of ecosystem heterogeneous nature of different households comprising a community services and the rationale behind household decision making around and consequently the differing abilities of individuals to take advantage forest use is important for effective policy interventions in forest of institutions and structures established to manage forest resources management, the long-term sustainability of forest resource use and and derive benefits. We suggest that efforts to improve and broaden conservation and land use policy. support for forest management should focus attention on the beneficia- ries of forest-dependent income generating activities and identify steps Acknowledgements to broaden participation in these ventures. Based upon our results, we suggest two main research areas to We are very grateful to the residents of Mariarano and Antanambao investigate further. The first is the relationship between the use of for their participation and to the key informants who helped facilitate provisioning services and the acknowledged benefits of regulating this research by granting permission and supporting the research ecosystem services, While hazard reduction services are widely team's objectives. We thank Mr. Jean Eric Rajaobelinirina at the GIZ for perceived as benefits from forest cover we did not find a significant discussions that informed data collection and for his comments on an relationship (either positive or negative) between these variables and earlier draft of this manuscript. We thank Onisoa Rasoamanalina Rivo, other ecosystem uses/benefits or with supporting forest management. Tahiana Joelle Ravelomanantsoa, Grazella Andrinambinina and A second area for further investigation is whether the local understand- Arimanana Andriniaina, for their field assistance and enthusiastic ing of forests and provision of hazard mitigation services is obtained support. We are appreciative of the logistical support of the GIZ Boeny through observations and experiences or through exposure to external team. We are also grateful to Dr. Bruno Ramamonjisoa at the University projects and education programs. Further examination of how the of Antananarivo for his support in obtaining permission to carry out knowledge of forest-hydrological linkages is formed and how it is trans- research in Madagascar. We thank the special issue editors for their lated into decision making on forest and land use management by comments on an earlier draft of the paper. We thank two anonymous farmers is important for both forest policy and for considering how reviewers for their constructive comments and suggestions. This broader land use policy can integrate rainfall linked hazard mitigation research was supported by an Economics and Social Research Council services provided by forests in such settings. Finally, we found only of the UK Doctoral Award to R.D. (Grant number ES/J500161/1); with two out of the fourteen ecosystem services valued by study respondents additional support from the ESRC Doctoral Training Centre’s Overseas significantly influence participation in forest management groups, again Fieldwork Award and an Exploration Fund grant award by The bringing to bear the question of what motivates collective action and Explorer’sClub,NY.
Appendix A
Table A1 Ecosystem services uses of forests as predictor variables of perceiving a safety net function of forests.
Predictor variables Unstandardized Standardized coefficients t Sig. 95.0% confidence interval for B coefficients
B Std. error Beta Lower bound Upper bound
(Constant) 2.866 0.242 11.852 0 2.389 3.342 Collecting wild fruits and vegetables 0.842 0.193 0.294 4.371 0.00 0.462 1.223 Collecting raphia, satrana and other plants for artisanal products 0.486 0.238 0.137 2.044 0.042 0.017 0.954 Hunting tenrecs 0.332 0.168 0.131 1.975 0.05 0.001 0.664
Table A2 Engaging in charcoal production as determinant of respondents' knowledge of forest-water regulation linkages, by fokontany.
Fokontany Unstandardized Standardized coefficients t Sig. 95.0% confidence coefficients interval for B
B Std. error Beta Lower Upper
Antanambao (Constant) 19.507 0.547 35.669 0.000 18.420 20.594 Charcoal producer −0.562 1.230 −0.048 −0.457 0.649 −3.006 1.881 Mariarano (Constant) 20.053 0.437 45.927 0.000 19.190 20.917 Charcoal producer −1.495 0.779 −0.163 −1.918 0.057 −3.036 0.046
Table A3 Water regulation score as a determinant of belief in the flood and sedimentation reduction services of forests (N = 226).
Parameters B S.E. Sig. Exp (B) 95% C.I. for Exp (B)
Lower Upper
Flood reduction belief 0.054 0.031 0.080 1.056 0.993 1.122 Constant −0.700 0.618 0.258 0.497 Sedimentation reduction belief 0.041 0.033 0.212 1.042 0.977 1.112 Constant 0.174 0.652 0.789 1.191 R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82 81
Coulibaly-Lingani, P., Savadogo, P., Tigabu, M., Oden, P.-C., 2011. Factors influencing people's participation in the forest management program in Burkina Faso, West Africa. Forest Policy Econ. 13, 292–302. Cullman, G., 2015. Community forest management as virtualism in northeastern Madagascar. Hum. Ecol. 43, 29–41. Daily, G.C., Alexander, S., Ehrlich, P.R., Goulder, L., Lubchenco, J., Matson, P.A., Mooney, H.A., Postel, S., Schneider, S.H., Tilman, D., 1997. Ecosystem Services: Benefits Supplied to Human Societies by Natural Ecosystems. Ecological Society of America Washington (DC). Danquah, J.A., 2015. Analysis of factors influencing farmers' voluntary participation in reforestation programme in Ghana. Forests, Trees and Livelihoods 24, 176–189. DeFries, R., Eshleman, K.N., 2004. Land-use change and hydrologic processes: a major focus for the future. Hydrol. Process. 18, 2183–2186. Dolisca, F., Carter, D.R., McDaniel, J.M., Shannon, D.A., Jolly, C.M., 2006. Factors influencing farmers' participation in forestry management programs: a case study from Haiti. For. Ecol. Manag. 236, 324–331. Dorren, L.K.A., Berger, F., Imeson, A.C., Maier, B., Rey, F., 2004. Integrity, stability and man- agement of protection forests in the European Alps. For. Ecol. Manag. 195, 165–176. Fagerholm, N., Käyhkö, N., Ndumbaro, F., Khamis, M., 2012. Community stakeholders' knowledge in landscape assessments–mapping indicators for landscape services. Ecol. Indic. 18, 421–433. Fisher, J.A., Patenaude, G., Giri, K., Lewis, K., Meir, P., Pinho, P., Rounsevell, M.D., Williams, M., 2014. Understanding the relationships between ecosystem services and poverty alleviation: a conceptual framework. Ecosystem Services 7, 34–45. Fig. A1. Proportion of households producing honey only for home consumption, for a mix Funk, C., Peterson, P., Landsfeld, M., Pedreros, D., Verdin, J., Shukla, S., Husak, G., Rowland, of home and sale and for only selling for income. J., Harrison, L., Hoell, A., Michaelsen, J., 2015. The climate hazards infrared precipita- tion with stations—a new environmental record for monitoring extremes. Sci. Data 2, 150066. Geist, H.J., Lambin, E.F., 2002. Proximate causes and underlying driving forces of tropical References deforestation tropical forests are disappearing as the result of many pressures, both local and regional, acting in various combinations in different geographical locations. Agarwal, B., 2001. Participatory exclusions, community forestry, and gender: an analysis Bioscience 52, 143–150. for South Asia and a conceptual framework. World Dev. 29, 1623–1648. Gezon, L.L., 2014. Who wins and who loses? Unpacking the “local people” concept in Ahrends, A., Burgess, N.D., Milledge, S.A.H., Bulling, M.T., Fisher, B., Smart, J.C.R., Clarke, G.P., ecotourism: a longitudinal study of community equity in Ankarana, Madagascar. Mhoro, B.E., Lewis, S.L., 2010. Predictable waves of sequential forest degradation and J. Sustain. Tour. 22, 821–838. biodiversity loss spreading from an African city. Proc. Natl. Acad. Sci. 107, 14556–14561. Gliem, R.R., Gliem, J.A., 2003. Calculating, interpreting, and reporting Cronbach's alpha re- Alila, Y., Kuraś, P.K., Schnorbus, M., Hudson, R., 2009. Forests and floods: a new paradigm liability coefficient for Likert-type scales. Midwest Research-to-practice Conference in sheds light on age-old controversies. Water Resour. Res. 45, W08416. Adult, Continuing, and Community Education. Amare, D., Mekuria, W., T/wold, T., Belay, B., Teshome, A., Yitaferu, B., Tessema, T., Tegegn, Hauck, J., Görg, C., Varjopuro, R., Ratamäki, O., Jax, K., 2013. Benefits and limitations of the B., 2016. Perception of Local Community and the Willingness to Pay to Restore ecosystem services concept in environmental policy and decision making: some Church Forests: The Case of Dera District, Northwestern Ethiopia. Forests, Trees and stakeholder perspectives. Environ. Sci. Pol. 25, 13–21. Livelihoods, pp. 1–14. Horning Rabesahala, N., 2004. The Limits of Rule: When Rules Promote Forest Conserva- Antona, M., Bienabe, E.M., Salles, J.-M., Péchard, G., Aubert, S., Ratsimbarison, R., 2004. tion and When They Do Not–Insights from Bara Country. Madagascar Cornell Rights transfers in Madagascar biodiversity policies: achievements and significance. University. Environ. Dev. Econ. 9, 825–847. Howe, C., Suich, H., Vira, B., Mace, G.M., 2014. Creating win-wins from trade-offs? Armitage, D., 2005. Adaptive capacity and community-based natural resource manage- Ecosystem services for human well-being: a meta-analysis of ecosystem service ment. Environ. Manag. 35, 703–715. trade-offs and synergies in the real world. Glob. Environ. Chang. 28, 263–275. Asano, M., Sugiura, T., Miura, K., Torii, S., Ishiguro, A., 2006. Reliability and validity of the Jones, J.P.G., Andriamarovololona, M.M., Hockley, N., 2008. The importance of taboos and self-report quality of life questionnaire for Japanese school-aged children with social norms to conservation in Madagascar (La Importancia de los Tabús y las asthma (JSCA-QOL v.3). Allergol. Int. 55, 59–65. Normas Sociales para la Conservación en Madagascar). Conserv. Biol. 22, 976–986. Aymoz, B.G.P., Randrianjafy, V.R., Randrianjafy, Z.J.N., Khasa, D.P., 2013. Community Klooster, D., 2000. Institutional choice, community, and struggle: a case study of forest co- Management of Natural Resources: a case study from Ankarafantsika National Park, management in Mexico. World Dev. 28, 1–20. Madagascar. Ambio 42, 767–775. Krishnaswamy, J., Bonell, M., Venkatesh, B., Purandara, B.K., Lele, S., Kiran, M.C., Reddy, V., Bakoariniaina, L.N., Kusky, T., Raharimahefa, T., 2006. Disappearing Lake Alaotra: monitor- Badiger, S., Rakesh, K.N., 2012. The rain–runoff response of tropical humid forest eco- ing catastrophic erosion, waterway silting, and land degradation hazards in systems to use and reforestation in the Western Ghats of India. J. Hydrol. 472–473, Madagascar using Landsat imagery. J. Afr. Earth Sci. 44, 241–252. 216–237. Balmford, A., Bruner, A., Cooper, P., Costanza, R., Farber, S., Green, R.E., Jenkins, M., Jefferiss, Kuraś, P.K., Alila, Y., Weiler, M., 2012. Forest harvesting effects on the magnitude and P., Jessamy, V., Madden, J., Munro, K., Myers, N., Naeem, S., Paavola, J., Rayment, M., frequency of peak flows can increase with return period. Water Resour. Res. 48, Rosendo, S., Roughgarden, J., Trumper, K., Turner, R.K., 2002. Economic reasons for W01544. conserving wild nature. Science 297, 950–953. Lambin, E.F., Geist, H.J., Lepers, E., 2003. Dynamics of land-use and land-cover change in Barbier, E.B., Hacker, S.D., Kennedy, C., Koch, E.W., Stier, A.C., Silliman, B.R., 2010. The tropical regions. Annu. Rev. Environ. Resour. 28, 205–241. value of estuarine and coastal ecosystem services. Ecol. Monogr. 81, 169–193. Laurance, W.F., 2007. Environmental science: forests and floods.Nature449,409–410. Becknell, J.M., Kissing Kucek, L., Powers, J.S., 2012. Aboveground biomass in mature and Leach, M., Mearns, R., Scoones, I., 1999. Environmental entitlements: dynamics and insti- secondary seasonally dry tropical forests: a literature review and global synthesis. tutions in community-based natural resource management. World Dev. 27, 225–247. For. Ecol. Manag. 276, 88–95. Lerdau, M., Whitbeck, J., Holbrook, N.M., 1991. Tropical deciduous forest: death of a Blöschl, G., Ardoin-Bardin, S., Bonell, M., Dorninger, M., Goodrich, D., Gutknecht, D., biome. Trends Ecol. Evol. 6, 201–202. Matamoros, D., Merz, B., Shand, P., Szolgay, J., 2007. At what scales do climate Lima, L., Coe, M., Soares Filho, B., Cuadra, S., Dias, L.P., Costa, M., Lima, L., Rodrigues, H., variability and land cover change impact on flooding and low flows? Hydrol. Process. 2014. Feedbacks between deforestation, climate, and hydrology in the Southwestern 21, 1241–1247. Amazon: implications for the provision of ecosystem services. Landsc. Ecol. 29, Bradshaw, C.J.A., Sodhi, N.S., Peh, K.S.H., Brook, B.W., 2007. Global evidence that 261–274. deforestation amplifies flood risk and severity in the developing world. Glob. Lise, W., 2000. Factors influencing people's participation in forest management in India. Chang. Biol. 13, 2379–2395. Ecol. Econ. 34, 379–392. Brang, P., SCHONENBERGER, W., Ott, E., Gardner, B., 2001. Forests as Protection from Liswanti, N., Sheil, D., Basuki, I., Padmanaba, M., Mulcahy, G., 2011. Falling back on forests: Natural Hazards. how forest-dwelling people cope with catastrophe in a changing landscape. Int. For. Brauman, K.A., Daily, G.C., Duarte, T.K.e., Mooney, H.A., 2007. The nature and value of Rev. 13, 442–455. ecosystem services: an overview highlighting hydrologic services. Annu. Rev. MA, 2005. Millennium ecosystem assessment. Ecosystems and Human Well-being: Environ. Resour. 32, 67–98. Synthesis. Island Press, Washington DC. Bruijnzeel, L.A., 2004. Hydrological functions of tropical forests: not seeing the soil for the Maass, J.M., Balvanera, P., Castillo, A., Daily, G.C., Mooney, H.A., Ehrlich, P., Quesada, M., trees? Agric. Ecosyst. Environ. 104, 185–228. Miranda, A., Jaramillo, V.J., Garcia-Oliva, F., Martinez-Yrizar, A., Cotler, H., Lopez- Calder, I.R., Aylward, B., 2006. Forest and floods. Water Int. 31, 87–99. Blanco, J., Perez-Jimenez, A., Burquez, A., Tinoco, C., Ceballos, G., Barraza, L., Ayala, Chomitz, K.M., Kumari, K., 1998. The domestic benefits of tropical forests: a critical R., Sarukhan, J., 2005. Ecosystem services of tropical dry forests: insights from long- review. World Bank Res. Obs. 13, 13–35. term ecological and social research on the Pacific Coast of Mexico. Ecol. Soc. 10. Chopra, K., 2002. Social capital and development processes: role of formal and informal Martin, T.G., Watson, J.E.M., 2016. Intact ecosystems provide best defence against climate institutions. Econ. Polit. Wkly. 37, 2911–2916. change. Nat. Clim. Chang. 6, 122–124. Coria, J., Calfucura, E., 2012. Ecotourism and the development of indigenous communities: Mehta, N.J., Heinen, T.J., 2001. Does community-based conservation shape favorable atti- the good, the bad, and the ugly. Ecol. Econ. 73, 47–55. tudes among locals? An empirical study from Nepal. Environ. Manag. 28, 165–177. 82 R. Dave et al. / Forest Policy and Economics 84 (2017) 72–82
Meijaard, E., Abram, N.K., Wells, J.A., Pellier, A.-S., Ancrenaz, M., Gaveau, D.L., Runting, R.K., Sodhi, N., Lee, T., Sekercioglu, C., Webb, E., Prawiradilaga, D., Lohman, D., Pierce, N., Mengersen, K., 2013. People's perceptions about the importance of forests on Borneo. Diesmos, A., Rao, M., Ehrlich, P., 2010. Local people value environmental services PLoS One 8, e73008. provided by forested parks. Biodivers. Conserv. 19, 1175–1188. Méndez-López, M.E., García-Frapolli, E., Ruiz-Mallén, I., Porter-Bolland, L., Reyes-Garcia, Sudmeier-Rieux, K., Masundire, H.M., Rizvi, A.H., Rietbergen, S., 2006. Ecosystems, V., 2015. From paper to forest: local motives for participation in different conserva- Livelihoods and Disasters: An Integrated Approach to Disaster Risk Management. tion initiatives. Case studies in southeastern Mexico. Environ. Manag. 56, 695–708. World Conservation Union. Miles, L., Newton, A.C., DeFries, R.S., Ravilious, C., May, I., Blyth, S., Kapos, V., Gordon, J.E., Tan-Soo, J.-S., Adnan, N., Ahmad, I., Pattanayak, S., Vincent, J., 2014. Econometric evidence 2006. A global overview of the conservation status of tropical dry forests. J. Biogeogr. on Forest ecosystem services: deforestation and flooding in Malaysia. Environ. 33, 491–505. Resour. Econ. 1–20. Minten, B., Randrianarisoa, C., 2012. Forest Preservation, Flooding and Soil Fertility: Van Dijk, A.I.J.M., Van Noordwijk, M., Calder, I.R., Bruijnzeel, S.L.A., Schellekens, J., Evidence from Madagascar. INTECH Open Access Publisher. Chappell, N.A., 2009. Forest–flood relation still tenuous – comment on ‘global evi- Musyoki, J.K., Mugwe, J., Mutundu, K., Muchiri, M., 2013. Determinants of household dence that deforestation amplifies flood risk and severity in the developing world’ decision to join community forest associations: a case study of Kenya. ISRN Forestry by C. J. A. Bradshaw, N.S. Sodi, K. S.-H. Peh and B.W. Brook. Glob. Chang. Biol. 15, 2013. 110–115. Nuggehalli, R.K., Prokopy, L.S., 2009. Motivating factors and facilitating conditions Völker, M., Waibel, H., 2010. Do rural households extract more forest products in times of explaining women's participation in co-management of Sri Lankan forests. Forest crisis? Evidence from the mountainous uplands of Vietnam. Forest Policy Econ. 12, Policy Econ. 11, 288–293. 407–414. Olson, D.M., Dinerstein, E., 1998. The Global 200: a representation approach to conserving von Heland, J., Folke, C., 2014. A social contract with the ancestors—culture and ecosystem the Earth's most biologically valuable ecoregions. Conserv. Biol. 12, 502–515. services in southern Madagascar. Glob. Environ. Chang. 24, 251–264. Ostrom, E., Burger, J., Field, C.B., Norgaard, R.B., Policansky, D., 1999. Revisiting the Wheater, H., Evans, E., 2009. Land use, water management and future flood risk. Land Use commons: local lessons, global challenges. Science 284, 278–282. Policy 26 (Supplement 1), S251–S264. PGM-E, GOM, 2013. Programme Germano-Malagache pour l'environnment: Schéma Wilk, J., 2000. Local perceptions about forests and water in two tropical catchments. d'Aménagement Communal Mariarano (2010–2015). GeoJournal 50, 339–347. Sande, S.O., Crewe, R.M., Raina, S.K., Nicolson, S.W., Gordon, I., 2009. Proximity to a forest Wilk, J., Andersson, L., Plermkamon, V., 2001. Hydrological impacts of forest conversion to leads to higher honey yield: another reason to conserve. Biol. Conserv. 142, agriculture in a large river basin in northeast Thailand. Hydrol. Process. 15, 2703–2709. 2729–2748. Shackleton, C., Shackleton, S., 2004. The importance of non-timber forest products in rural Wisner, B., P. Blaikie, T. Cannon, and I. Davis. 2004. At risk: natural hazards, people's livelihood security and as safety nets: a review of evidence from South Africa. S. Afr. vulnerability and disasters. J. Sci. 100, 658–664. Wunder, S., Angelsen, A., Belcher, B., 2014. Forests, livelihoods, and conservation: Silvano, R.A.M., Udvardy, S., Ceroni, M., Farley, J., 2005. An ecological integrity assessment broadening the empirical base. World Dev. 64, S1–S11. of a Brazilian Atlantic Forest watershed based on surveys of stream health and local farmers' perceptions: implications for management. Ecol. Econ. 53, 369–385. Forest Policy and Economics 84 (2017) 83–91
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
A methodological approach for assessing cross-site landscape change: ☆ MARK Understanding socio-ecological systems
⁎ Terry Sunderlanda,b, , Rabdo Abdoulayea, Ronju Ahammadc, Stella Asahad, Frederic Baudrone, Elizabeth Deakink, Jean-Yves Duriauxe, Ian Eddyf, Samson Folia,g, Davison Gumboa, Kaysara Khatunl, Mumba Kondwania, Mrigesh Kshatriyah, Laurio Leonalda,i, Dominic Rowlanda,j, Natasha Staceyc, Stephanie Tomschaf, Kevin Yangf, Sarah Gergelf, Josh Van Vianena a Centre for International Forestry Research, Indonesia b James Cook University, Australia c Charles Darwin University, Australia d Forests, Resources and People, Cameroon e CIMMYT, Ethiopia f University of British Columbia, Canada g University of Amsterdam, The Netherlands h UNEP/MIKE Programme, Kenya i Bogor Agricultural University, Indonesia j School of Oriental and African Studies, UK k Opus International Consultants, New Zealand l Universitat Autonoma de Barcelona, Spain
ARTICLE INFO ABSTRACT
Keywords: The expansion of agriculture has resulted in large-scale habitat loss, the fragmentation of forests, significant losses in Agrarian change biological diversity and negative impacts on many ecosystem services. In this paper, we highlight the Agrarian Change Forests Project, a multi-disciplinary research initiative, that appliesdetailedsocio-ecologicalmethodologies in multi-functional Livelihoods landscapes, and assess the subsequent implications for conservation, livelihoods and food security. Specifically, the Diets researchfocusesonlanduseimpactsinlocationswhich exhibit various combinations of agricultural modification/ Poverty change across a forest transition gradient in six tropical landscapes, in Zambia, Burkina Faso, Cameroon, Ethiopia, Biodiversity Indonesia and Bangladesh. These methods include integrated assessments of the perceptions of ecosystem service provision, tree cover loss and gain, relative poverty, diets and agricultural patterns of change. Although numerous surveys on rural livelihoods are undertaken each year, often at great cost, many are hampered by weaknesses in methods and thus may not reflect rural realities. We attempt to highlight how integrating broader socio-ecological methods can be used to fill in those gaps and ensure such realities are indeed captured. Early findings suggest that the transition from a forested landscape to a more agrarian dominated system does not necessarily result in better livelihood outcomes and there may be unintended consequences of forest and tree cover removal. These include the loss of access to grazing land, loss of dietary diversity and the loss of ecosystem services/forest products.
1. Introduction agricultural production (Perrings et al., 2006; Bharucha and Pretty, 2010; Sunderland, 2011). Managing, and negotiating, trade-offs be- Historically, the trade-off between increasing food security/produc- tween biodiversity and agriculture involves maximising food security tion and the maintenance of natural systems has led to a perception that benefits while minimising damage to the wider environment. the two were mutually exclusive (Tscharntke et al., 2005; Brussaard Globally, the total area of cultivated land increased by 466% from et al., 2010). This perspective, however, has failed to account for the 1700 to 1980 (Meyer and Turner, 1992). Croplands and pastures have fact that certain levels of biodiversity exists within some agricultural now become one of the largest terrestrial biomes on the planet, landscapes which provide multiple contributions to food security and occupying ~40% of the land surface (Ellis et al., 2010). Between
☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” published at the journal Forest Policy and Economics 84C, 2017. ⁎ Corresponding author. E-mail address: [email protected] (T. Sunderland). http://dx.doi.org/10.1016/j.forpol.2017.04.013 Received 17 May 2016; Received in revised form 26 April 2017; Accepted 26 April 2017 Available online 11 May 2017 1389-9341/ © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91
1980 and 2000, more than half of new agricultural land across the change mitigation and biodiversity conservation. Meanwhile ‘land tropics was established at the expense of intact forests, while a further sharing’, is supported by the fact that many species are dependent on 28% was opened up to the detriment of disturbed or secondary forests farmland and other habitats maintained by humans (Wright et al., (Gibbs et al., 2010). This habitat loss is further compounded by land 2012; Deakin et al., 2016), and that farmlands that are often structu- degradation and competition from other land uses such as urbanisation rally similar to the original native vegetation can support biodiversity (Ellis et al., 2010). Although the overall rate of agricultural expansion often as effectively as native vegetation (Clough et al., 2011). has slowed considerably over the last three decades the global focus on The land sparing versus land sharing debate has become somewhat food production has ensured a rapid rate of increase in yield per unit polarised in the scientific literature (Law and Wilson, 2015) and, it has area (Gibbs et al., 2010). Technological and scientific advancements been argued, has actually stagnated (Bennett, 2017). There is increasing have provided access to cheaper chemical fertilisers and pesticides, opinion that a ‘black and white’ dichotomy over-simplifies issues that in high-yielding crop varieties, advanced irrigation technologies and more practice are highly complex2 (Adams, 2012; Fischer et al., 2014). efficient mechanisation (Matson et al., 1997; Motes, 2010), which have Baudron and Giller (2014) suggest that both options are equally all contributed to increased crop yields. Unsurprisingly, given the important and can be complementary strategies under different circum- dependency of this model on fossil fuels, concerns have been raised stances and some landscapes may exhibit elements of both. Small- over the long-term sustainability of increasingly intensified agriculture, holder farmers for example, who provide up to 40% of the world's food, particularly as food demands are projected to more than double by mostly fall somewhere on the continuum between land sharing and 2050 (Green et al., 2005; Fischer et al., 2008; Godfray et al., 2010). land sparing (Tscharntke et al., 2012). The land sharing/sparing debate While there has been significant progress towards meeting global also suggests there is some level of “grand design” at the landscape commitments to reduce hunger, levels of food insecurity remain scale which is simply not the case (Reed et al., 2017). Most landscapes unacceptably high. Approximately 842 million people worldwide are inherently dynamic and evolve through the influence and interac- remain hungry and undernourished (UNICEF, 2011; Black et al., tions of environment, society and economies (Sayer et al., 2016). 2013; FAO et al., 2013) and this can be attributed as the cause of one It has also been recognised that land use strategies aimed at third of child mortality figures in developing countries. This situation is balancing agriculture and biodiversity conservation must also consider further exacerbated by global population growth and changing dietary socio-economic outcomes and trade-offs(Fischer et al., 2014; Loos patterns with a predicted 50% increase in the demand of agricultural et al., 2014; Khatun et al., 2015). Landscapes should be viewed as products by 2030 (Bruinsma, 2003). In this context, the provisioning of complex social-ecological systems that consist of mosaics of natural food is increasingly couched within multiple objectives sought from and/or human-modified ecosystems (Bennett et al., 2006; Reed et al., multifunctional mosaic landscapes namely, biodiversity conservation, 2016). However, there is a distinct lack of information on the human maintenance of ecosystem services, food production, sustainable liveli- impacts of agrarian change in forested areas, particularly with regards hood provision, and climate change mitigation (Sayer et al., 2013; Reed to socio-economic effects of agricultural intensification, long-tesrm et al., 2015; Khatun et al., 2016). However, in many places, land dietary diversity and market integration processes (Byerlee et al., scarcity results in trade-offs between many of these components, 2014). Previous research within the land-sharing vs. land sparing particularly between the need for agricultural commodities and con- debate has focused heavily on the trade-offs between food security serving biodiversity (Law and Wilson, 2015). and biodiversity at a macro-level (Phalan et al., 2011a; Green et al., To this end, two contrasting landscape management approaches; 2005; Clough et al., 2011), while local scale effects upon livelihoods, ‘land sparing’ and ‘land sharing’ have been identified as potential poverty, food security and nutrition have tended to be overlooked. strategies to minimise the negative consequences of agriculture on Furthermore, it is also important to recognise that more food produc- biodiversity. These consider land use change in such a way that tion does not automatically lead to better local food security and competing demands for food, commodities and forest services can be improved livelihoods for rural communities (Powell et al., 2015). reconciled (e.g. Pirard and Treyer, 2010; Phalan et al., 2011a). ‘Land In this paper, we present the Agrarian Change Project, a multi- sparing’ aims at intensifying production and maximising agricultural disciplinary, research initiative led by the Center for International yields by trading off its negative consequences on the environment by Forestry Research with direct funding from USAID's Biodiversity Bureau ‘sparing’ areas of natural capital (often in the form of protected areas) and the UK's Department for International Development (Deakin et al., and therefore reducing the need for agricultural expansion into forest 2016). The project applies detailed socio-ecological methodologies to areas (Pirard and Treyer, 2010).1 ‘Land sharing’, on the other hand - examine the outcomes/impacts of land use and agrarian change where agricultural production takes place within complex multi-func- processes in multi-functional landscapes, and the subsequent implica- tional landscapes - is based on a land use model that integrates tions for conservation, livelihood, and food security. Specifically, the production and conservation within the same land units. It proposes research focuses on land use impacts in locations which exhibit various to minimise the use of external inputs and to retain patches of natural combinations of agricultural modification/change across a forest habitat within farmlands in a form of extensive agriculture. Under the transition gradient in six tropical landscapes in Zambia, Burkina Faso, latter management regime, landscapes consisting of low-intensity Cameroon, Ethiopia, Indonesia and Bangladesh. The study attempts to productive areas are combined with areas of natural biodiversity highlight how integrating broader socio-ecological methods, within a (Wright et al., 2012). Such strategies include agroforestry systems novel experimental design can be used to fill in gaps in assessing local and traditional swidden farming practices (Ziegler et al., 2009; Clough food security, dietary diversity and nutrition levels, tenure, local et al., 2011). poverty, biodiversity/forest conservation and integration with global Land sparing offers a convincing narrative for achieving desirable commodity markets. Thus the project seeks to explore these landscape agrarian change, particularly in the developing world (e.g. Phalan components by answering the following research questions et al., 2011a, 2011b), suggesting that efforts to emulate land sparing through the application of incentives, regulations, and land use 1. How is land use changing over time and what are the underlying planning could lead to optimal outcomes for food production, climate drivers behind these changes? Are there consistencies/differences between the case study landscapes/countries? 2. What are local people's perceptions of the outcomes of land use 1 Agricultural intensification does not necessarily mean increases in inputs such as fertilizer and capital (e.g. through mechanisation), but it can also include changes to the use of labour and environmental services. See discussion in Pirard and Treyer, 2010, p.6. Most commonly, however, intensification is understood as additional inputs to increase 2 See also: http://blog.cifor.org/8110/land-sharing-or-land-sparing-reconciling- productivity. agriculture-and-biodiversity-conservation?fnl=en
84 T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91
Fig. 2. Agrarian Change Project landscape level hierarchical research design. (Source Deakin et al., 2016).
Fig. 1. Map showing the location of the six landscapes used as case studies for the dependent on forests. Zone 2 represents areas with an intermediate or Agrarian Change research project. mixed farming system with less access to forests. Zone 3 represents locations that have been converted to monoculture agricultural systems change in each landscape in terms of their livelihoods, access to with very little access to forests. From a landscape perspective, final natural resources, land tenure and food security? zones were selected based on the discretion of the Principal Investiga- 3. What is the relationship between land use, local livelihoods, and tors for each region. For Ethiopia and Indonesia, zones were digitised in food security under different land use scenarios along an agricultur- ArcMap (ESRI, 2012). For all other study sites (Bangladesh, Burkina al modification gradient? Faso, Cameroon, and Zambia) three-km buffers were created around each household and merged into one polygon for analysis using R Here we present a methodological overview of the project and statistical software (R Core Team, 2015). highlight some of the preliminary findings, For each landscape we ensured that there was limited variability among zones within a given landscape in rainfall/climatic and agro- 2. Research design at the landscape scale ecological characteristics, elevation and biome type. The three zones were distinguished according to a set criteria for each focal landscape to 2.1. Selection of landscapes, zones and villages help identify the different practices and relative differences of key characteristics of each zone (Table 1). This was modified for each The selection of six landscapes in the Agrarian Change project was landscape and helped to clarify whether the landscapes selected fi largely based on prior-knowledge of eld teams in the focal countries; exhibited a gradient of agrarian change/agricultural modification. Ethiopia, Cameroon, Indonesia, Bangladesh, Zambia and Burkina Faso The type and number of criteria/variables differed between focal – fi (Fig. 1) (see: Deakin et al., 2016). Preliminary ndings presented in landscapes. ‘ ’ this paper are derived from thorough reconnaissance or scoping Within each landscape along a forest transition gradient, we then studies conducted in each landscape to ensure that these met project undertook research, to examine the effects of landscape configuration and experimental design criteria and also provided detailed background (including fragmentation and levels of patchiness), land sharing/ information regarding current and historical land management prac- sparing scenarios, and synergies and trade-offs between different land tices, for the regions selected. uses (crops, livestock rearing, swidden agriculture, agroforestry) with Attempts were also made to co-locate landscapes where there were forests and tree-based systems. A common set of qualitative and already research initiatives in place, such as the Sentinel Landscapes quantitative research methodologies were applied to enable a global framework of the CGIAR's Forests Trees and Agroforestry research comparative analysis, which included household surveys, focus groups, 3 programme. A continuum of changing land use practices was required and semi-structured interviews with key informants (Fig. 3) to gauge to be present within each landscape, mimicking historical land use information on variables such as relative poverty food security, dietary trajectories. Thus by looking at the transition through a gradient of diversity and nutrition, agricultural production, land tenure, migration ff fi change, the project o ers a unique insight to potential impacts/bene ts and biodiversity as well as stakeholder perceptions of ecosystem of a particular land use change, i.e. Zone 1 can potentially transform services, and their relative values. We aimed to capture as much into Zone 3, should the same trajectory (or gradient) be followed. It variation along a modification gradient in each focal landscape regard- works directly with communities within multi-functional landscapes to ing the following characteristics: understand the social, economic and ecological consequences of land use and agrarian change processes in their vicinity. The study utilised a • Agricultural modification (from low inputs -diversified, extensive, landscape-level approach through a three stage nested hierarchical subsistence orientated practices through to high input -market experimental design. Village/s or settlement/s nested within or in close orientated, intensive, simplified practices), through extensive farm ‘ ’ proximity to the dominant land use zones - representing a gradient/ surveys fi — continuum of agricultural modi cation and decreasing tree cover - • Forest cover loss and gain (where detectable) which were nested within each representative landscape exhibiting • Community dependency on forest resources various changing land use practices (Fig. 2). • Market access and infrastructure Zones within the landscapes represented different levels of agrofor- estry pressure and were paired with household surveys exploring the The information gauged from the combined methodologies were diets and livelihoods of local people. By comparing households within complemented by alternative sources, such as researcher's own ob- ff di erent zones the impacts of forest loss and fragmentation within servation/measurement e.g., “distance to the nearest road usable ff fi di ering farming systems can be quanti ed. At the local scale, zones during all seasons and GPS to measure distances and secondary data were selected based on a set of criteria including population density, from village records e.g. population (time series), access to public crop types, infrastructure, etc. (Table 1). Zone 1 represents landscapes services, land categories etc. Secondary data was also used to obtain where people are carrying out subsistence farming and are heavily information for some agricultural and forest products at the national or district level, e.g. Food and Agricultural Organization, International 3 http://www1.cifor.org/sentinel-landscapes/home.html. Tropical Timber Organization, World Resources Institute, and national
85 T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91
Table 1 Example criteria used to help distinguish characteristics of different ‘zones’ within a focal landscape.
Criteria Zone I Zone 2 Zone 3
1 Population density Sparse Medium Dense 2 Land tenure State/customary Customary Customary/title deeds 3 Proximity to major towns Distant Far Near 4 Level of dependence on forest resources High High-medium Medium-low 5 Proximity to protected areas (forest reserves, Game Management Areas, National parks) Near Far Distant 6 Level of in migration High - medium Medium-Low Low 7 Level of agric. inputs (fertiliser) None-Low Medium High 8 Market oriented crop production/presence of cash crops: Tobacco, maize, groundnuts and cotton Rare Occasional Common 9 Presence of subsistence farming High Medium Low 10 Levels of infrastructure development Low Moderate High
Fig. 3. Research design and data collection at the landscapes scale. (Deakin et al., 2016).
Table 2 Assessing forest cover for the Agrarian Change project.
Region Years Data type/approach Data source
Ethiopia 1988–2011, 2015 Sequence of Landsat imagery was used to classify the Scene selection used United States Geological Survey's (USGS) landscape into forest/non-forest cover using image GLOVIS Earth Explorer for images containing < 10% cloud cover differencing technique with a Normalized Burn Ratio (http://glovis.usgs.gov/). Once selected, the scene list was (NBR) (Jin et al., 2013) supplemented with high spatial uploaded to the USGS ESPA services online ordering system resolution imagery. Maximum likelihood supervised specifying the LEDAPS model (Masek et al., 2006) corrected classification of orthorectified Rapideye 3A imagery (5 m) products. from January 2015 was used to further distinguish grasslands, agricultural fields and bare soils. Cameroon 1987, 2000, 2002, 2014 Vegetation indices (e.g., NDVI Huete 1988) derived from Surface reflectance Imagery from Landsat 5 and 7 were Landsat imagery were used in an image differencing downloaded from the United States Geological Survey's (USGS) process to create an image times series. Radiometric ESPA interface and was partially processed using the USGS ESPA normalization was conducted (Lu et al., 2004). service to deliver imagery provided as surface reflectance using the LEDPAS model (Masek et al., 2006). Landsat 8 imagery was downloaded directly using the USGS GLOVIS earth explorer tool. Kapuas Hulu 2000–2013 Moderate resolution global maps Hansen/UMD/Google/USGS/NASA (2014) Hansen et al. (2013) Regency- Indonesia Bangladesh 1960–80 & 1990–2005 National spatial data obtained online Bangladesh Forest Department (BFD, 2015) Zambia 1990–2013 Land-use and land cover analysis, including land change CIFOR's Nyimba Forest Project, (Gumbo, 2015) detection Burkina Faso 1999–2013 High resolution satellite imagery, NDVI calculation/ High resolution imagery Google Earth & classifications from Wilson classifications and Sader, 2002 & Kerr and Ostrovsky, 2003 inventories, that were consolidated by the interviews, focus groups and use. We cross-checked for consistency, with more extensive semi- questionnaire pre-testing. Forest cover estimates were derived from a structured interviews conducted with key informants, specifically the number of sources depending on data availability (Table 2). (See settlement leaders and administration officers, identified by project Table 3.) staff in order to obtain general information about history, economic Focus group discussions were carried out in all 54 communities activities, agricultural seasons and presence of infrastructure in the across the project. These were primarily undertaken with the village regions. Selected key informants were interviewed per village to head and elders on the historical and cultural background, land tenure understand activities occurring in each village/settlement and provide and ownership, resource dependency and sale of forest products/ field team leaders with sufficient background knowledge needed to resources, and the regulations and taboos concerning land and forest properly conduct household and farm surveys.
86 T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91
Table 3 Spatial extent of zones in each country/landscape.
Country Zone 1 area Zone 2 area Zone 3 area Total area (km2) (km2) (km2) (km2)
Burkina Faso 97 103 105 305 Bangladesh 44 66 40 149 Cameroon 29 19 24 73 Ethiopia 20 23 19 63 Indonesia 84 37 89 210 Zambia 25 42 37 104
A representative sample of households was interviewed with both household and farm surveys in each zone. The household surveys were carried out either in one village with an approx. number of 100 households, or each zone, containing several villages, with an approx. number of 100 households combined (approx. 300 total per site). There were 9–10 farm surveys per zone, totaling approximately 30 per research site. The selection of households and farms within the chosen villages were undertaken by random sampling.
Fig. 5. Community perceptions of change in availability of forest products. 3. Preliminary results
The comparative study is still in in the early stages of data analysis. However, in this paper, we are able to present some preliminary assessments of agrarian drivers of change in the six diverse tropical landscapes described above. Some patterns and trends are clearly emerging, even at this nascent stage of the project. Figs. 4–6 show some initial comparative data. In addition, we present a brief insight of country level analyses, with some case study countries e.g. Cameroon and Zambia being less developed in terms of data exploration and analysis. The variables within each case study site differ slightly based on the expertise and current focus of the country level researchers involved in the Agrarian Change Project. Figs. 4–6 show people's perception of the availability of forest cover, forest products, and agricultural expansion respectively. These were collated from the household and farm surveys. The perceptions are based on the time community members have resided in the area, which ranges from 8 to 38 years. Fig. 5 shows that in half the sites studied, there is a perception among households that there are fewer forest products available than in the past. Unsurprisingly, this is especially notable in the study zones further away from forest. Households in Ethiopia and Cameroon, stated that there are also fewer forest products Fig. 6. Community perceptions of change in availability of agricultural land. available in Zone 2, while in Burkino Faso, households said that all zones have fewer forest products. These perceptions correlate well with the perception that there has been a decrease in forest cover across all the sites (Fig. 4).
4. Ethiopia
The Ethiopian site is focused on the Munessa forest region, along corresponding agricultural intensification and forest cover gradients. The results of the focus group discussions conducted suggest that food security increased with agricultural expansion from the 1970s to date (Baudron et al., in press). Data along an intensification gradient show that rural households closest to the forest use significantly more fuelwood from the forest than households away from it (3257 ± 1461, 407 ± 978, and 353 ± 530 kg per household/year in Zones 1 to 3, respectively). The same pattern applies for biomass from the forest used as feed (2876 ± 2860, 1935 ± 859, and 306 ± 471 kg per household/year in Zones 1 to 3, respectively). As a consequence, farms closest to the forest contain more livestock (3.61 ± 3.10, 2.27 ± 2.03, and 1.85 ± 2.42 TLU per household in the near, intermediate and distant zones, respectively) and thus more Fig. 4. Community perceptions of change in forest cover. manure, which they concentrate on home gardens producing a variety
87 T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91 of nutrient-dense food items. agriculture in Zone 1. More broadly this shift represents a transition Only residents from villages adjacent to the forest are allowed to away from a rotational system of shifting cultivation that requires large graze the forest (and harvest firewood) and this access is closely areas forest under different stages of the forest-field-fallow-forest controlled. The proportion of farms having a home garden is 85.5%, transition (land sharing). In Zone 1 where almost all agricultural land 62.4%, and 38.5% in the near, intermediate and distant zones, is swidden, the distinction between forest, fallow and agricultural land respectively) and the dietary diversity score is significantly higher for is blurred whereas in Zone 3, oil palm estates establishes a strict rural household closest to the forest (weekly HDDS of 6.58 ± 1.21, delineation between forest and agricultural land (land sparing). The 5.38 ± 1.02, and 4.41 ± 0.77 in the near, intermediate and distant transition is accompanied by a simultaneous shift from fluid customary zones, respectively, (Baudron et al., in press). In areas where agricul- land tenure arrangements (overlapping with state defined national park tural intensification is at its greatest (Zone 3), there has been a boundaries) to strict, state-regulated demarcation of land use as well as significant decrease in the contribution of forest resources (notably a reduction in forest cover. In all three zones, local people perceived fuelwood). Ownership of cattle, has also reduced, given the removal of forest cover to be declining along with reduced availability of forest appropriate grazing land. The long-term implications of this transition products (Figs. 4 and 5). Gaharu (Aquilaria spp.) is the most important are yet to be fully understood. The role of the State has also made forest product in Zones 1 & 2. It is an endangered species, and is significant impacts on this landscape with uncertain property rights and declining in most forests across Kalimantan (Soehartono and Newton, land annexation playing significant role in driving rapid deforestation, 2001). Thus concern over declining forest resources applies both to loss that needs to be further explored in the subsequent phases of the of forest cover and over-extraction of particular high-value forest project. products.
5. Cameroon 7. Burkina-Faso
The livelihoods of the majority of households in South West The parkland agroforestry landscapes of southern Burkina Faso Cameroon depend primarily on agricultural activities (Asaha, 2015). show an interesting and much more historically rooted integration of These include both shifting cultivation and cash crop production forests, farm and markets for achieving food security. From our (coffee, cocoa and oil palm) and this is exhibited strongly across the household surveys (n = 296), 56% were food secure for all 12 months three zones with Zone 1 being characterised by high forest cover and in the year, the remainder (44%) faced 1–3 months of reduced food. On shifting cultivation systems and Zone 3 representing a more defined average households had sufficient food to have 3 meals a day for mosaic of sedentary farming and cash crops. Despite the relative 10.6 months in the year. Foli and Abdoulaye (2016) found that annual importance of agriculture, there remains a strong reliance on forest household food provisioning from autonomous farms, forests or un- products for rural income and consumption. This is reflected in the cultivated sources and markets is at 61%, 23% and 16% respectively, strong market values of NTFP's across the zones. As such, their cultural across the three zones respectively. Reoccurring droughts, erratic importance underpins their widespread availability. rainfall and a short rainy season means rural households continuously Within the last two and a half decades, local communities have rely on forests and wild foods for up to three months in the year to experienced different land-use changes around them, especially from supplement diets (Foli and Abdoulaye, 2016; Koffi et al., 2016). Forests the conservation sector. Annexation of land by conservation actors has clearly play a role in buffering dietary diversity during lean agricultural led to a considerable proportion of South West Cameroon being months (Koffi et al., 2016). Obvious trajectories of intensification in the gazetted as protected areas. However, the lack of enforcement of these Burkina landscape are difficult to identify compared to the other forest areas means that encroachment is common and many of these landscapes. It shows “clusters” of intensification, rather than a linear reserves and parks are regarded as potential sources of new farmland, temporal pattern of clear transition. The protection of important tree particularly with the advance of monoculture plantations that have species that contribute to local food provisioning is historically particularly impacted communities in Zones 2 and 3 (see Figs. 4–6). embedded in local norms. Baobab (Adansonia digitata), Shea (Vitilaria Recent commercial oil palm development has annexed yet more paradoxa) and Néré (Parkia biglobosa) are examples of local tree species land in a manner that results in total loss of access to previously that are protected from felling under customary laws throughout of available agricultural land. This new development has been received Burkina Faso (Coulibaly-Lingani et al., 2009). As such, these landscapes with mixed feelings; some people see it as a great opportunity for are managed to be as resilient as possible for the vicissitudes of climate development of their area and employment for them and their children, and environment that persistently affect the region. while others see it as a threat to the future of the village and their children in terms of future land shortages, particularly for small-scale 8. Zambia agricultural expansion. The ever-increasing demand for agricultural land, especially for cocoa and more recently for oil palm, has also Although between 74 and 86% of the respondents in this land- attracted considerable migration to the area, focused mainly along the scape—depending on the zone—derive income from agriculture. The roads (especially Zone 3). However, this influx has also extended to main source of income for forested communities (Zone 1) is derived inaccessible areas (Zone 1) where land is considerably cheaper, land from farming, fishing and aquaculture (79%), while game ranches tenure systems are simple, and acquisition and ownership processes are provide both formal and informal employment for some. In Zones 2 and relatively straightforward. As such, this migration and the concomitant 3, the main source of income is less diverse and is primarily derived economic activities will result in yet more forest loss in the future. from farm labour (Zone 2 = 12%, Zone 3 = 18%) and small-scale food production (Zone 2 = 73%, Zone 3 = 57%). 6. Indonesia Interestingly, in Zone 1, aquaculture and fishing make up 30% of top three sources of income. This compares with 2% and 0% in Zones 2 In Kupuas Hulu, Indonesia, the study, shows that the agrarian and 3 respectively, which indicates that the forests in these villages are trajectory from diverse smallholder agricultural production towards providing significant water related regulating and provisioning ecosys- more intensive agribusiness is having a major impact on the livelihoods tem services, that are either uneconomical to utilise commercially as of local people. Interestingly, communities showed concern over the people transition away from forest based landscapes or are degraded decreased availability of agricultural lands in all three zones (Fig. 6) and lost as agrarian change progresses. The latter is likely is the case, as reflecting the expansion of oil palm estates in Zone 3, rubber agrofor- the value of forest resources is not reflected in the policy environment, estry in Zone 2 and reduction in forest land available for swidden which favours intensified agriculture over the forestry sector.
88 T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91
Ultimately this has resulted in a lack of understanding of the role of Zones 2 (15%) and Zone 3 (14%). The expansion of agriculture is due to forests play in rural livelihoods and resource conflict at the agricultural more households (86%) clearing lands to increase farm areas in new or frontier (Zone 1) where it is evident that there remains a strong reliance fallow land in Zone 1 near the forest compared with Zones 2 (60%) and on forest resources to supplement agricultural income. 3 (42%). As a result more households (> 90%) in Zone 1 near to forests As with all countries and all Zones—the exception being the have large numbers of livestock (1.77 ± 1.73 TLU) than Zones 2 forested zone in Kapuas Hulu, Indonesia—the perceptions of change (0.58 ± 0.97 TLU) and 3 (0.54 ± 0.88 TLU) respectively. Despite the in the Zambian landscape are collectively ones of decreasing forest expansion of farming area in Zone 1 yearly food security was found to cover (Fig. 4) with 91%, 75%, and 93% of households perceiving some be relatively low within households there (8 ± 3 months/year) than form of deforestation across Zones 1,2 and 3 respectively. Not surpris- Zones 2 (10 ± 2.7 months/year) and 3 (10 ± 2.46 months/year). ingly, perceived forest lost also resulted in a perceived reduction in the Conversely, a higher proportion of households (38%) possess stable availability of forest products 92%, 75%, and 90% across Zones 1, 2 and farm areas in Zone 2 than Zones 1 (20%) and 3 (24%). A significant 3 respectively (Fig. 5). This is congruent with collective perceptions on difference found on the farm size was higher in Zone 2 (5.15 ± 4.52 the availability of forest products among all other landscapes and zones acre) than Zones 1 (3.04 ± 2.70 acre) and 3 (3.70 ± 4.95 acre). In within the study —with the exception of Cameroon. As there is zero addition, more households (70%) also possess home garden in Zone 2, indication from the income data that forest products are important in as a compliment to their agricultural activities. the Miombo woodlands, we could deduce that this indicates NTFPs are important for subsistence only or have cultural or medical significance 10. Discussion and conclusion (hence the perceived decline). These perceptions of landscape change almost certainly have a basis The Agrarian Change project aims to move land use debates forward in actual/real land use and landscape change. Yet somewhat counter- from solely examining trade-offs between food production/security and intuitively, when asked about the perception of the changes in the biodiversity, to one of understanding potential synergies. The conse- extent of agricultural land within their landscape, the majority per- quences of different land use strategies can only be fully understood ceived it to have also decreased (Fig. 6) (Zone 1 = 45%, Zone 2 = 50%, within the wider context of local histories, culture, politics, and market Zone 3 = 77%). Indeed, if the perceptions in the community are that dynamics. For example, land use decisions at the household level often both the extent of forest cover and the extent of agricultural land uses influence what happens at the landscape scale, yet in a majority of cases have decreased? Then it is logical to assume that some land use changes such decisions are driven by strong externalities such as government are unaccounted for in our survey or the scale of change is negligible policies, technological capacity, agricultural extension, and markets. In and so local perceptions are not a significant indication of actual change addition, local scale effects upon livelihoods, poverty, food security and at the landscape scale. nutrition have tended to be overlooked, or solely through sectoral lens'. Thus the Agrarian Change Project, addresses this research gap and 9. Bangladesh attempts to advance our understanding of agricultural landscapes as integral social-ecological systems. It offers insights into the impacts of The Chittagong Hill Tracts (CHT) has unique types of landforms due the transition from forest to agriculture, through the different Zones, in the many hills and valleys and a network of natural streams and rivers that Zone 1 can potentially become Zone 3 if the gradient of change is compared to the rest of the country, which is primarily deltaic and low- followed through in each case, thus allowing for early intervention lying. As a result, shifting cultivation is only carried out in this region, based on findings per zone. Early indications suggest that the transfor- with > 80% of the area exhibiting variable relief (Islam et al., 2007). mation of wildlands, notably forests, for agriculture has much more Aside from government control, there is a sparse distribution of the varied and complex impacts on livelihoods, nutrition and health than small size of indigenous forest reserves ranging from 50 to 300 acres previously anticipated. across the landscape. However, natural forests are declining in all cases The study aims to challenge the perceived wisdom that suggests that for state-owned forests and community reserves based on the findings rural communities with better access to markets, transportation and from interviews with government officials and community members. intensive agricultural systems are better fed, and are fiscally better off Forest loss is higher in remote and less accessible areas due to shifting than those in the proximity of more isolated, forested landscapes (cf. cultivation, illegal felling and insecure tenure rights (Bala et al., 2013). Levang et al., 2005). Although historical evidence suggests the transi- Deforestation is relatively less problematic in the areas where mono- tion away from a forest-based economy leads to overwhelmingly better culture plantations have increased in the recent past years, land use outcomes for poverty and human well-being (Sunderlin et al., 2007). changes remain stable and land ownership is secure. From the early 1960s, pervasive growth based theories of agricultural It was observed that 90% households experienced a decline of the development based on technological change were promoted as a forest cover over the last 30 years. The loss of forest area has caused solution to persistent rural poverty (Mellor, 1967). Yet local observers shortage in the availability of forest products in 86% of households. and village field researchers noted that rural development wasn't Though the community discussion revealed an increase in tree cover working as intended; the numbers of poor grew and some non-poor during 1990–2000, a high proportion of the households (94%) per- smallholders, fisherman, pastoralists subsequently became poor ceived a loss of the forest products in the intensification zones through loss of assets or common property resources, and overall (intermediate and distant from the forest). A significant proportion of health and nutrition benefits remained elusive. Larger farmers appro- households in intermediate (86%) and distant (85%) zones have priated the land of smaller farmers, rural labourers were displaced by experienced an increased decline of fuel wood than households (60%) mechanisation and intensive farming depleted scarce water resources near the forest. More than one-third of households (40%) near to the and affected soils. In short, as agricultural transformation takes place, forest in Zone 1 reportedly collect wild foods compared with 28% and there are inevitable winners and losers. Thus it might be argued that the 27% in Zones 2 and 3 respectively. A higher dietary diversity was also assumptions that underpin notions of better livelihoods in terms of found at households near to forest than in the intermediate and distant food, health and wealth as land becomes increasingly dominated by zones (weekly HDDS in Zone 1, 2 and 3 were 7.58 ± 1.41; agrarian systems, rather than by natural vegetation could be chal- 6.79 ± 1.52 and 6.70 ± 1.40 respectively). lenged. For example, the loss of forest can ultimately cause increasingly Agriculture has been found to be equally dynamic along the forest negative livelihood impacts, such as described for Ethiopia where and land use gradient. Overall, 50% households experienced a decline traditional grazing land provided by forests has been converted to of their farm size across the landscape. Only 27% households have agricultural fields with a concomitant loss of income to herders increased farming lands in Zone 1 near the forest area compared to (Baudron et al., in press).
89 T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91
This is mirrored by the important contribution of forest resources in and advance our understanding of multi-functional landscapes as socio- Burkina Faso, Cameroon and Zambia, albeit in different ways. At the ecological systems. Only by understanding the connectedness of other end of the spectrum, In Indonesia, commercialised agrarian intricate and complex socio-ecological systems through such integrated systems, in this case characterised by rubber and oil palm, have had a research methods are we able to fully appreciate the subtlety and somewhat negative impact on local diets and nutrition with access to nuance of these findings and the relationships between varying land- the cash economy facilitating a nutrient transition to a diet of processed scape components. As this research unfolds we aim to further test the food (Ickowitz et al., 2016). The increase in tree cover through hypotheses and answer the questions listed earlier in more detail. agroforestry practices in Bangladesh provides an indication of the value Agrarian change transitions are taking place all over the globe, but of trees within agricultural systems, despite the underlying chronic clearly not everyone benefits. Who does, how and why, will be the focus poverty of the population in the Chittagong landscape (see also Rahman of further assessment. et al., 2014). Preliminary findings from the study support those by other authors Acknowledgments such as Agrawal et al. (2013), who estimate that over 1.3 billion people utilise forests and trees in some way and that forested landscapes The authors would like to express heartfelt thanks to the villagers generate significant income for those that reside in and around them. and other national stakeholders who took park in numerous focus group Smallholders across the developing world may still derive as much discussions and key informant interviews in each of the six case study income from foraging forests and wildlands as from cultivating crops landscapes (Ethiopia, Cameroon, Indonesia, Nicaragua, Bangladesh, (Wunder et al., 2014). The findings of the tropics-wide Poverty and Zambia and Burkina Faso) discussed here. We are also grateful to the Environment Network (PEN), also suggest that rural households rely far national, regional and local governments who embraced our research more on income and other services from their immediate natural objectives, authorised us to conduct research in their jurisdictions and environment than previously thought. The PEN project found that over gave consent for us to obtain research permits. 25% of household income is sourced from natural resources; this Funding for this project has been provided by the United States represents a greater annual household than that of agricultural Agency for International Development (USAID) and the UK's production (Angelsen et al., 2014). Ickowitz et al. (2014),ina Department for International Development (DFID) KnowFor through continent-wide study in Africa found a correlation between the grants to CIFOR. This publication is an output of the CGIAR Consortium presence of forests and trees and dietary diversity. Thus the evidence- Research Program on Forests, Trees and Agroforestry. base on the synergies between agriculture and the wider environment are being gradually, and increasingly understood, and it is those References synergies that this project is attempting to articulate. The results of the segregation of agriculture from forestry and other [BFD] Bangladesh Forest Department, 2015. Forest cover of Bangladesh. Accessed 20 land uses has led to critical reflection as to how these seemingly June 2015. http://fd.portal.gov.bd/. fl [FAO] Food and Agriculture Organization of the United Nations, [IFAD] con icting land uses can be better integrated for improved outcomes International.Fund for Agricultural Development and [WFP] World Food Programme, (Sayer et al., 2013). As this paper attempts to illustrate, agricultural 2013[FAO] Food and Agriculture Organization of the United Natio. The State of Food production in most tropical landscapes is not the linear process from the Insecurity in the World, 2013. The Multiple Dimensions of Food Security FAO, Rome. fi http://www.fao.org/docrep/018/i3434e/i3434e00.htm. direct transition of tree cover to agricultural elds but more based on Adams, W.M., 2012. Feeding the next billion: hunger and conservation. Oryx 46, complex landscape mosaics that are managed for multiple benefits and 157–158. a broad suite of goods and ecosystem services (Padoch and Sunderland, Agrawal, A., Shepherd, G., Bensen, C., Hardin, R., Cashore, B., Miller, D., 2013. Economic 2014). Although greatly under-estimated, the presence of forests and contributions of forests. In: Background Paper 1 for UNFF10. United Nations Forum on Forests, New York (132 pp.). http://www.un.org/esa/forests/pdf/session_ trees in these landscapes provides a framework for the integration of documents/unff10/EcoContrForests.pdf. diverse cropping systems (Reed et al., 2016). They are also immensely Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N., Bauch, S., Borner, J., valuable to the livelihoods and well-being of those that live in such Smith-Hall, C., Wunder, S., 2014. Environmental income and rural livelihoods: a Global-comparative analysis. World Dev. 64 (1), S12–S28. environments. Asaha, S., 2015. Land Use Change, Rural Livelihoods and Food Security in the Manyemen- At the end of 2015, the launch of the Sustainable Development Nguti Forest Landscape, South West Region, Cameroon. A thesis submitted to the fi Goals provided a unique opportunity to begin that process of integrat- department of Development Studies, in partial ful lment of the requirements for the award of a Master of Science (M.Sc) degree in Environment and Natural Resource ing previously siloed disciplines into a more cogent development Management Pan African Institute for Development–West Africa Buea, Cameroon. agenda, with a strong focus on landscapes as the convening factor Bala, B.K., Majumder, S., Altaf Hossain, S.M., Haque, M.A., Hossain, M.A., 2013. (Van Vianen et al., 2015; Reed et al., 2016). In addition, the preamble Exploring development strategies of agricultural systems of Hill Tracts of Chittagong in Bangladesh. Environ. Dev. Sustain. 15 (4), 949–966. of the historic COP 21 (UNFCCC, 2015) agreement also mentions food Baudron, F., Giller, K.E., 2014. Agriculture and nature: Trouble and strife? Biol. Conserv. security from a broader perspective by “safeguarding food security and 170, 232–245. ending hunger, and the particular vulnerabilities of food production systems Baudron F, Duriaux Chavarria J-Y, Remans R, Yang K, and Sunderland T. Indirect ” contributions of forests to dietary diversity in Southern Ethiopia. Ecol. Soc. (in press). to the adverse impacts of climate change , and also refers to human rights, Bennett, E., 2017. Changing the agriculture and environment conversation. Nature gender, ecosystems and biodiversity, all issues that are central to Ecology and Evolution 1, 10018. agriculture. Thus the policy environment has become far more con- Bennett, A.F., Radford, J.Q., Haslem, A., 2006. Properties of land mosaics: implications ducive to a more environmentally sensitive agriculture, with broader for nature conservation in agricultural environments. Biol. Conserv. 133 (2), 250–264. development aspirations such as improved nutrition etc. The develop- Bharucha, Z., Pretty, J., 2010. The roles and values of wild foods in agricultural systems. ment lexicon has certainly changed in recent years to reflect a broader Philos. Trans. R Soc. B: Biol. Sci. 365 (1554), 2913–2926. system (or landscape) approach to food production in the context of the Black, R.E., Victora, C.G., Walker, S.P., Bhutta, Z.A., Christian, P., de Onis, M., Ezzati, M., fi Grantham-McGregor, S., Katz, J., Martorell, R., et al., 2013. Maternal and child wider environmental bene ts and yet systematic and empirically undernutrition and overweight in low-income and middle-income countries. Lancet focused research to enable us to actually implement this has been 382 (9890), 427–451. largely absent from the discourse (Reed et al., 2015). Bruinsma, J. (Ed.), 2003. World Agriculture: Towards 2015/2030. An FAO Perspective. Earthscan Publications, London. Therefore, as the studies in each country continue and more data Brussaard, L., Caron, P., Campbell, B., Lipper, L., Mainka, S., Rabbinge, R., Babin, D., becomes available in the subsequent stages of the project, it will allow a Pulleman, M., 2010. Reconciling biodiversity conservation and food broader cross-site comparison of studies that will ultimately enable security:Scientific challenges for a new agriculture. Curr. Opin. Environ. Sustain. 2 – – fi (1 2), 34 42. signi cantly more compelling conclusions to be drawn from this Byerlee, D., Stevenson, J., Villoria, N., 2014. Does intensification slow land conversion or project. These intend to provide much needed insights into how encourage deforestation? Glob. Food Sec. 3, 92–98. landscape level land use trajectories manifest in local communities Clough, Y., Barkmann, J., Juhrbandt, J., Kessler, M., Wanger, T.C., Anshary, A., Buchori,
90 T. Sunderland et al. Forest Policy and Economics 84 (2017) 83–91
D., Cicuzza, D., Darras, K., Putra, D.D., et al., 2011. Combining high biodiversity with Ithaca (403 pp.). high yields in tropical agroforests. Proc. Natl. Acad. Sci. 108 (20), 8311–8316. Meyer, W.B., Turner, B.L., 1992. Human-population growth and global land use cover Coulibaly-Lingani, P., Tigabu, M., Savadogo, P., Oden, P.-C., Ouadba, J.-M., 2009. change. Annu. Rev. Ecol. Syst. 23, 39–61. Determinants of access to forest products in southern Burkina Faso. Forest Policy Motes, W., 2010. Modern Agriculture and Its Benefits–Trends, Implications and Outlook. Econ. 11 (7), 516–524. Global Harvest Initiative, Washington, D.C. Deakin, E., Kshatriya, M., Sunderland, T. (Eds.), 2016. Agrarian Change in Tropical Perrings, C., Jackson, L., Bawa, K., Brussaard, L., Brush, S., Gavin, T., Papa, R., Pascual, Landscapes. Centre for International Forestry Research, Bogor, Indonesia (323 pp.). U., De Ruiter, P., 2006. Biodiversity in agricultural landscapes: saving natural capital http://www.cifor.org/publications/pdf_files/Books/BCIFOR1601.pdf. without losing interest. Conserv. Biol. 20 (2), 263–264. Ellis, E.C., Goldewijk, K.K., Siebert, S., Lightman, D., Ramankutty, N., 2010. Padoch, C., Sunderland, T., 2014. Managing landscapes for food security and enhanced Anthropogenic transformation of the biomes, 1700 to 2000. Glob. Ecol. Biogeogr. 19 livelihoods: building upon a wealth of local experience. Unasylva 64 (241) 2013/2. (5), 589–606. http://www.fao.org/docrep/019/i3482e/i3482e01.pdf. ESRI, 2012. ArcGIS Desktop: Release 10.1. Phalan, B., Balmford, A., Green, R.E., Scharlemann, J.P.W., 2011a. Minimising the harm Fischer, J., Brosi, B., Daily, G.C., Ehrlich, P.R., Goldman, R., Goldstein, J., Lindenmayer, to biodiversity of producing more food globally. Food Policy 36, S62–S71. D.B., Manning, A.D., Mooney, H.A., Pejchar, L., et al., 2008. Should agricultural Phalan, B., Onial, M., Balmford, A., Green, R.E., 2011b. Reconciling food production and policies encourage land sparing or wildlife-friendly farming? Front. Ecol. Environ. 6 biodiversity conservation: land sharing and land sparing compared. Science. (7), 380–385. Pirard, R., Treyer, S., 2010. Agriculture and Deforestation: What Role Should REDD+ Fischer, J., Abson, D.J., Butsic, V., Chappell, M.J., Ekroos, J., Hanspach, J., Kuemmerle, and Public Support Policies Play? Institut du Developpment durable et des relations T., Smith, H.G., von Wehrden, H., 2014. Land sparing versus land sharing: moving internationles (IDDRI), Paris (No.10/2010 December). forward. Conserv. Lett. 7 (3), 149–157. Powell, B., Thilsted, S.H., Ickowitz, A., Termote, C., Sunderland, T., Herforth, A., 2015. Foli, S., Abdoulaye, R., 2016. Drivers and outcomes of changing land use in parkland Improving diets with wild and cultivated biodiversity from across the landscape. agroforestry systems of central Burkina Faso. In: Deakin, E.L., Sunderland, T.C.H., Food Security 7 (3), 535–554. Kshatriya, M. (Eds.), Agrarian Change in Tropical Landscapes. CIFOR, Bogor, R Core Team, 2015. R: A Language and Environment for Statistical Computing. R Indonesia, pp. 269–299. Foundation for Statistical Computing. Gibbs, H.K., Ruesch, A.S., Achard, F., Clayton, M.K., Holmgren, P., Ramankutty, N., Foley, Rahman, S.A., Rahman, M.F., Sunderland, T., 2014. Increasing tree cover in degrading J.A., 2010. Tropical forests were the primary sources of new agricultural land in the landscapes: ‘Integration’ and ‘intensification’ of smallholder forest culture in the 1980s and 1990s. Proc. Natl. Acad. Sci. 107 (38), 16732–16737. Alutilla valley, Matiranga, Bangladesh. Small-scale Forestry 13, 237–249. Godfray, H.C.J., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F., Nisbett, N., Pretty, J., Reed, J., Deakin, L., Sunderland, T., 2015. What are “integrated landscape approaches” Robinson, S., Toulmin, C., Whiteley, R., 2010. The future of the global food system. and how effectively have they been implemented in the tropics: a systematic map Philos. Trans. R. Soc. Lond. Ser. B Biol. Sci. 365 (1554), 2769–2777. protocol. Environmental Evidence 4 (1), 2. Green, R.E., Cornell, S.J., Scharlemann, J.P.W., Balmford, A., 2005. Farming and the fate Reed, J., Van Vianen, J., Deakin, E., Barlow, J., Sunderland, T., 2016. Integrated of wild nature. Science 307 (5709), 550–555. landscape approaches to managing social and environmental issues in the tropics: Gumbo, D., 2015. Nyimba Forest Project dataset. UNF:5:18pkbj/I/onhGlLNP0SI1g== learning from the past to guide the future. Glob. Chang. Biol. 22 (7), 2540–2554. Center for International Forestry Research [Distributor]. V2 [Version]. http://hdl. Reed, J., Van Vianen, J., Barlow, J., Sunderland, T., 2017. Have integrated landscape handle/.net/11463/10055. approaches to reconciled societal and environmental issues in the tropics? Land Use Hansen, M., Potapov, P., Moore, R., Hancher, M., Turubanova, S., Tyukavina, A., Thau, Policy 63, 481–492. D., Stehman, S., Goetz, S., Loveland, T., Kommareddy, A., Egorov, A., Chini, L., Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.L., Sheil, D., Meijaard, E., Venter, M., Justice, C., Townshend, J., 2013. High-resolution global maps of 21st-century forest Boedhihartono, A.K., Day, M., Garcia, C., et al., 2013. Ten principles for a landscape cover change. Science 342 (80), 850–853. approach to reconciling agriculture, conservation and other competing land uses. Hansen/UMD/Google/USGS/NASA, 2014. Forest Change Data Measure Tree Cover Loss, Proc. Natl. Acad. Sci. U. S. A. 110 (21), 8349–8356. Tree Cover Gain, or Forest Disturbance. Creative Commons CC BY 4.0. Global Forest Sayer, J., Buck, L., Langston, J., Margulies, C., Reed, J., Riggs, R., Sunderland, T., 2016. Watch, Washington, DC. http://50.18.182.188.6080/arcgis/rest/services/ Measuring the effectiveness of landscape approaches to conservation and ForestCover_lossyear/Imageserver. development. Sustain. Sci. http://dx.doi.org/10.1007/s11625-016-0415-z. Ickowitz, A., Powell, B., Salim, A., Sunderland, T., 2014. Dietary quality and tree cover in Soehartono, T., Newton, A., 2001. Conservation and sustainable use of tropical trees in Africa. Glob. Environ. Chang. 24, 287–294. the genus Aquilaria II. The impact of gaharu harvesting in Indonesia. Biol. Conserv. Ickowitz, A., Powell, B., Salim, A., Sunderland, T., 2016. The effect of tree cover on child 97 (1), 29–41. nutrition in Indonesia: examining the relationship between tree cover and Sunderland, T., 2011. Food security: Why is biodiversity important? Int. For. Rev. 13 (3, consumption of micronutrient-rich foods. PLoS One 11 (5), e0154139. SI), 265–274. Islam, M., Alam, M., Mantel, S., 2007. Land use planning and environmental control in Sunderlin, W., Dewi, S., Puntodewo, A., 2007. Poverty and Forests: Multi-country the Chittagong Hill Tracts. In: CHARM Project Report 3. International Soil Reference Analysis of Spatial Association and Proposed Policy Solutions. Occasional Paper 47 and Information Centre (ISRIC), The Netherlands (Dhaka, Bangladesh). CIFOR, Bogor, Indonesia. http://www.cifor.org/publications/pdf_files/OccPapers/ Jin, S., Yang, L., Danielson, P., Homer, C., Fry, J., Xian, G., 2013. A comprehensive OP-47.pdf. change detection method for updating the National Land Cover Database to circa Tscharntke, T., Klein, A.M., Kruess, A., Steffan-Dewenter, I., Thies, C., 2005. Landscape 2011. Remote Sens. Environ. 132, 159–175. perspectives on agricultural intensification and biodiversity-ecosystem service Kerr, J., Ostrovsky, M., 2003. From space to species: ecological applications for remote management. Ecol. Lett. 8 (8), 857–874. sensing. Trends Ecol. Evol. 18, 299–305. Tscharntke, T., Clough, Y., Wanger, T.C., Jackson, L., Motzke, I., Perfecto, I., Vandermeer, Khatun, K., Gross-Camp, N., Corbera, E., Martin, A., Ball, S., Massao, G., 2015. When J., Whitbread, A., 2012. Global food security, biodiversity conservation and the Participatory Forest Management makes money: insights from Tanzania on future of agricultural intensification. Biol. Conserv. 151 (1), 53–59. governance, benefit sharing, and implications for REDD+. Environment and [UNICEF] United Nations Children's Fund, 2011. Levels and trends in child mortality. In: Planning A 47 (10), 2097–2112. Estimates Developed by the UN Inter-agency Group for Child Mortality Estimation. Khatun, K., Corbera, N., Ball, S., 2016. Fire is REDD+: offsetting carbon through early United Nations Children's Fund, New York, NY. http://www.unicef.org/media/files/ burning activities in south-eastern Tanzania. Oryx 51 (1), 43–52. Child_Mortality_Report_2011_Final.pdf. Koffi, C.K., Djoudi, H., Gautier, D., 2016. Landscape diversity and associated coping UNFCCC, 2015. Adoption of the Paris Agreement. Proposal by the President. (FCCC/CP/ strategies during food shortage periods: evidence from the Sudano-Sahelian region of 2015/L.9/Rev.1). 12 December 2015. http://unfccc.int/resource/docs/2015/cop21/ Burkina Faso. Reg. Environ. Chang. 1–12. eng/l09.pdf. Law, E.A., Wilson, K.A., 2015. Providing context for the land-sharing and land-sparing Van Vianen, J., Reed, J., Sunderland, T., 2015. From Global Complexity to Local Reality: debate. Conserv. Lett. 8 (6), 404–413. Aligning Implementation Frameworks with Sustainable Development Goals and Levang, P., Dounias, E., Sitorus, S., 2005. Out of the forest, out of poverty? Forests, Trees Landscape Approaches. CIFOR Policy Brief No. 1. Centre for International Forestry and Livelihoods 15, 211–235. Research, Bogor. http://www.cifor.org/library/5864/from-global-complexity-to- Loos, J., Abson, D.J., Chappell, M.J., Hanspach, J., Mikulcak, F., Tichit, M., Fischer, J., local-reality-aligning-implementation-frameworks-with-sustainable-development- 2014. Putting meaning back into “sustainable intensification”. Front. Ecol. Environ. goals-and-landscape-approaches/. 12 (6), 356–361. Wilson, E., Sader, S., 2002. Detection of forest harvest type using multiple dates of Lu, D., Mausel, P., Brondizio, E., Moran, E., 2004. Change detection techniques. Int. J. Landsat TM imagery. Remote Sens. Environ. 80, 385–396. Remote Sens. 25, 2365–2407. Wright, H.L., Lake, I.R., Dolman, P.M., 2012. Agriculture – a key element for conservation Masek, J., Vermote, E., Saleous, N., Wolfe, R., Hall, F., Huemmrich, K., Gao, F., Kutler, J., in the developing world. Conserv. Lett. 5 (1), 11–19. Lim, T.K., 2006. A Landsat Surface Reflectance Dataset for North America, Wunder, S., Angelson, A., Belcher, B., 2014. Forests, livelihoods, and 1990–2000. IEEE Geosci. Remote Sens. Lett. 3 (1), 68–72. conservation:broadening the empirical base. World Dev. 64, S1–S11. Matson, P.A., Parton, W.J., Power, A.G., Swift, M.J., 1997. Agricultural intensification Ziegler, A.D., Fox, J.M., Xu, J., 2009. Agriculture: the rubber juggernaut. Science 324 and ecosystem properties. Science 277 (5325), 504–509. (5930), 1024–1025. Mellor, J., 1967. The Economics of Agricultural Development. Cornell University Press,
91 Forest Policy and Economics 84 (2017) 92–101
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Wild food collection and nutrition under commercial agriculture expansion in agriculture-forest landscapes☆
Rikke Brandt Broegaard a,⁎,LauraVangRasmussenb,NeilDawsonc, Ole Mertz a, Thoumthone Vongvisouk a,d, Kenneth Grogan a a Department of Geosciences and Natural Resource Management, University of Copenhagen, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark b International Forestry Resources and Institutions (IFRI), School of Natural Resources & Environment, University of Michigan, 440 Church Street, Ann Arbor, MI 48109-1041, USA c School of International Development, University of East Anglia, Norwich Research Park, Norwich, Norfolk NR4 7TJ, UK d Faculty of Forestry, National University of Laos, 7322, Dongdok Campus, Vientiane, Laos article info abstract
Article history: Wild food constitutes a substantial part of household food consumption around the world, but rapid land use Received 11 May 2016 changes influence the availability of wild foods, which has implications for smallholders' food and nutrient in- Received in revised form 3 December 2016 take. With increasing commercial agriculture and biodiversity conservation efforts in forested tropical regions, Accepted 19 December 2016 many shifting cultivation systems are being intensified and their extent restricted. Studies examining the conse- Available online 5 January 2017 quences of such pressures commonly overlook the diminishing role of wild food. Using a combination of collec- tion diaries, participant observation, remote sensing, and interviews, we examined the role of agriculture-forest Keywords: Nutrition-sensitive landscape landscapes in the provision of wild food in rapidly transforming shifting cultivation communities in northern Shifting cultivation Laos. We found that wild food contributed less to human diets in areas where pressure on land from commercial Cash crop expansion agriculture and conservation efforts was more intense. Our results demonstrate that increasing pressure on land Conservation policy creates changes in the shifting cultivation landscape and people's use thereof with negative effects on the quality Diet of nutrition, including protein deficiency, especially in communities adjacent to core conservation areas. Our Asia study shows the importance of adopting a more nutrition-sensitive approach to the linkages between commer- cial agriculture and biodiversity conservation (and the policies that promote them), wild food provisioning, and food security. © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
1. Introduction availability of wild food as forests, fallows, and agricultural fields are converted to more intensive agriculture (Padoch and Sunderland, Nutritional outcomes for rural inhabitants are determined not sim- 2013). Ickowitz et al. (2014) find a positive relationship between forest ply by food production in a landscape or even household incomes, but cover and dietary diversity in a multi-nation African study. Ironically, are highly influenced by access to and control over the resources however, conserving forests does not guarantee wild food availability. which make up a person's diet (Sen, 1983). Rapid change is occurring Global efforts to reduce deforestation and increase the proportion of ter- across rural landscapes in the world's developing countries in terms of restrial land in protected areas, for example through Reduced Emissions both land use and governance. Commercial agriculture is rapidly in- from Deforestation and Forest Degradation (REDD+) schemes and creasing in forested tropical regions, thereby transforming many subsis- Aichi Target 11 of the United Nations Convention on Biological Diversity, tence-oriented shifting cultivation systems towards more commercial often result in a recentralising of control and reduced access to these re- agriculture, often in accordance with national policies aiming at eco- sources for local populations (Ribot et al., 2006; Sandbrook et al., 2010; nomic growth (Hall, 2011; Hall et al., 2011; van Vliet et al., 2012). West et al., 2006). These land use changes not only influence local people's income levels Limited attention has been devoted to understanding this intersec- and possibly the amount of food purchased, but also affect the tion between land use change, forest governance, wild food availability, and nutrition (Foran et al., 2014; Sibathu et al., 2015; Vira et al., 2015). ☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” The lack of attention to this complex intersection between land use published at the journal Forest Policy and Economics 84C, 2017." change and adequate nutrition is problematic as it remains unknown ⁎ Corresponding author: Øster Voldgade 10, DK-1350 Copenhagen K, Denmark. how cash crop expansion and conservation efforts change local people's E-mail addresses: [email protected] (R.B. Broegaard), [email protected] (L.V. Rasmussen), [email protected] (N. Dawson), [email protected] (O. Mertz), collection of wild foods and what the implications are for diet quality. [email protected] (T. Vongvisouk), [email protected] (K. Grogan). The consequences may be severe in contexts where subsistence
http://dx.doi.org/10.1016/j.forpol.2016.12.012 1389-9341/© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101 93 agriculture and food collection are prevalent and where purchase of healthy diets, or what the Global Nutrition Report (2014) refer to as varied food items is limited (Shackleton and Pandey, 2014). Despite “nutrition-sensitive landscapes”. Analogously, Sayer et al. (2013) high- many scholarly efforts to assess the role of ‘bush meat’ in rural diets light food security aspects as an important outcome of integrative land- (e.g. Sarti et al., 2015; van Vliet et al., 2015), the need for research on scape and land use planning. the intersection between land use change and nutrition has only gained In this paper, we take up this challenge with specific attention to the attention recently. Further, it has been stressed that such research must shifting cultivation systems of Southeast Asia. Northern Laos was select- take contributions from the entire landscape into account (Padoch and ed as our study site as it provides a pertinent experimental area to ex- Sunderland, 2013; Sayer et al., 2013; Sibathu et al., 2015; Global amine how cash crop expansion and increased conservation efforts Nutrition Report, 2014). Advances have been made with regards to change people's use of landscape for food provisioning. These land- the nutritional contribution of forest foods (e.g. Ickowitz et al., 2014; scapes traditionally delivered a broad variety of wild foods, which Rowland et al., 2016); we contribute by also including wild foods from formed local populations' subsistence. However, over the past 5– non-forest habitats in our analysis, in part inspired by Powell et al. 10 years landscapes have experienced rapid land use changes from sub- (2013). sistence-oriented upland rice cultivation towards commercial based The core of the problem is that increasing incomes from intensified maize cultivation happening at a very large scale (as documented by agriculture alone will not necessarily lead to reduced hunger or im- Castella et al., 2013; Lestrelin et al., 2013; Hall, 2011; Schönweger et proved nutrition because many people, particularly those with limited al., 2012; Vongvisouk et al., 2014). Similarly, the shifting cultivation land, capital, and food market-access, may be unable to shift to reliance landscape has been influenced by conservation efforts (Moore et al., on markets for suitable and affordable food, and instead continue to rely 2012). The introduction of cash cropping alongside policies seeking to on access to a diversity of local resources (Pinstrup-Andersen, 2009; increase forest cover in Laos have been shown to have had some nega- Ickowitz et al., 2014, Powell et al., 2015). This means that large propor- tive impacts on rural inhabitants' livelihoods and ability to cope with tions of rural populations, despite widespread modernisation of farming shocks (Castella et al., 2013). Here we examine these political and land- practices, continue to rely on forests and other habitats in addition to scape changes (i.e. the combined influence of biodiversity conservation the agricultural crops to secure adequate food and nutritionally bal- and cash cropping), with regards to diet and nutrition. anced diets for their families. The diverse contribution of wild foods We pose two questions in the article: 1) how does increased pres- from forest-agriculture landscapes to local diets has been demonstrated sure on land through conservation efforts and cash crop expansion by many studies (for example Angelsen et al., 2014; Christensen, 2002; change local people's use of shifting cultivation landscapes for wild Ickowitz et al., 2016; Lykke et al., 2002; Paumgarten and Shackleton, food provisioning? And 2) how is diet quality influenced by changes 2011; Wunder et al., 2014), but it is often overlooked in development in the collection of wild food resulting from land pressures? Our main efforts, particularly when the contribution is diminishing in the face of argument is that increased pressures on land through commercial agri- widespread, rapid land-use changes and associated alterations in the ac- culture expansion and conservation efforts reduce the quality of nutri- cess and control over food resources (Shackleton et al., 2015; Vira et al., tion when local people rely less on wild food derived from the 2015). Wild food, especially meat and fish, has been shown to be impor- terrestrial landscape without having market-access to diverse, nutri- tant in terms of dietary diversity, even if consumption frequency may be tious food. We take a special look at protein, as protein deficient diets low (Golden et al., 2011; Sarti et al., 2015; Shackleton et al., 2015; van have been identified as one of the main risks for rural Laotian communi- Vliet et al., 2015). Several studies highlight the important contribution ties (Krahn, 2003, 2005). of wild food to dietary diversity that risks being lost in a ‘nutritional transition’ away from locally produced and collected food to purchased 2. Nutrient sensitive landscapes: The intersection between land use food because of modernization and globalization (Piperata et al., 2011; change, wild food collection and nutrition in Laos Remis and Jost Robinson, 2014; Sarti et al., 2015; van Vliet et al., 2015). The contribution of forests to income and diet is better thought The number of undernourished people in Southeast Asia has been of as “the supermarket of the wild” rather than as gap-filling (Wunder more than halved between 1990 and 2015 and this is largely attributed et al., 2014: S39), and loss and degradation of forest areas can therefore to economic growth (FAO, 2015). The trend is similar for Laos and a be expected to exacerbate food insecurity and nutrition (Krahn, 2003, study by the World Food Program (WFP) (2007) finds a strong, positive 2005; Krahn and Johnson, 2007; Van Noordwijk et al., 2014). In addi- effect of household wealth assets on food security. Many governments tion, the poor rely heavily on wild food harvested from natural areas place economic growth high on their agenda, and the Government of other than forests (Angelsen et al., 2014; Mertz et al., 2001). A recent Laos is no exception. It is firmly committed to lifting Laos out of the study from Tanzania found that wild foods from agricultural land ranks of the Least Developed Countries by 2020 and halving the levels made a larger dietary contribution than wild foods from forests of extreme poverty (World Bank, 2014). But the vision of economic (Powell et al., 2013). growth in Southeast Asia is heavily embedded in large-scale land use Achieving all the components of food security is thus highly complex changes promoting cash crop production. The Government of Laos, and while economic growth can be shown to reduce food insecurity and like many other governments of developing countries, actively pro- improve the average nutritional status of populations (FAO, 2015; WFP, motes the expansion of cash crop production (Castella et al., 2013; 2007), inclusive growth and attention to local needs and context are Vongvisouk et al., 2016), with the general expectation that moderniza- fundamental for guaranteeing food and nutrition improvements (FAO, tion and intensification of agriculture should transform the lives of 2015; Dawson et al., 2016). The focus of food security studies has smallholders through a green revolution. Such transformations of liveli- changed from a primary preoccupation with the sufficiency of staple hoods are purported to happen through increasing productivity and in- grains and calories, towards the importance of a balanced and safe comes, thereby benefitting livelihoods and increasing the consumption diet that includes protein, vitamins, and other micronutrients of marketed foods and other goods. The dependence on the immediate (Ickowitz et al., 2014; Pingali, 2015), with micronutrient deficiency or surroundings for subsistence livelihoods is thereby assumed to decline, inadequate nutrition being referred to as the “hidden hunger” (e.g. and perhaps eventually leading to a shift away from farming. Such Ickowitz et al., 2014, p. 287). This is especially relevant for the poorest large-scale land use changes have been well documented (Castella et part of the world's population (FAO, 2015)andPowell et al. (2015) con- al., 2013; Dwyer, 2011; Hall, 2011; Lestrelin et al., 2013; Schönweger clude that for developing countries “diversity within rural and agricul- et al., 2012). Yet, several studies have also identified negative livelihood tural landscapes may be an important part of a food environment that impacts that land use changes have had for much of the upland popula- supports healthy dietary choices” (p. 535). They call for more research tion such as decreased livelihood- and biodiversity which limits the ca- on how local communities manage their landscapes for supporting pacity to cope with unexpected events (among them Castella et al., 94 R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101
2013; Ducourtieux et al., 2005; Kenney-Lazar, 2012; Suhardiman et al., changes cause declining collection of wild animals and other wild 2015). foods, protein deficiencies and reduced quality of diet are a likely out- In Laos, the number and size of protected areas have expanded since come and they may outweigh beneficial impacts of increased purchas- the establishment of 18 National Protected Areas (NPA) in 1993 and ing power gained through expanding cash crop production. now covers N12% of the national territory (GoL, 2005; Moore et al., 2012). While these NPAs are sometimes presented as areas that pre- serve biodiversity and thus help secure livelihood of nearby villages, 3. Study area and methodology many villages have experienced restrictions and prohibitions on their historical uses of the flora and fauna in the areas that were converted 3.1. Study area into protected areas (Johnson et al., 2010; Moore et al., 2012). The establishment of protected areas and other land use regulations The study took place in three villages (Fig. 1), all located in northern further intensifies the demand for land and increases (formal) restric- Laos in Huaphan Province and bordering the Nam Et-Phou Louey Na- tions on the use of many parts of the landscape, whether for food provi- tional Protected Area (NEPL-NPA). Huaphan Province represents a par- sioning or for other livelihood activities (Broegaard et al., 2017; Castella ticularly interesting province, reportedly having one of the highest rates et al., 2013; see also Vedeld et al. (2012) for a study of livelihood impacts of households engaged in the collection of non-timber forest products in of a protected area in Tanzania). A “comprehensive food security and Laos (WFP, 2007). At the same time, 21% of the households in the prov- vulnerability analysis” of Laos by WFP (2007) commented that the pres- ince were characterized in 2007 as having either poor (2%), or border- sure to change from more subsistence-oriented shifting cultivation to line (19%) food consumption patterns (WFP, 2007). commercial farming, together with the large-scale establishment of The NEPL-NPA was created in 1993, but until 2000 it remained more plantations and contract farming, influence cropping area and biodiver- alessa‘paper park’ with little impact on villages in the area. However, sity, and reduce the contribution of food from the landscape, which has active NPA management began in 2000 with funding and technical sup- traditionally been important (Krahn and Johnson, 2007). Likewise, the port from international environmental NGOs (Moore et al., 2012). The WFP's Laos Country Strategy 2011–2015 (WFP, 2010:14) quotes a park's boundaries were redefined and expanded in 2010. The establish- study showing that economic growth in Laos has not led to a significant ment and expansion of boundaries has created ‘enclosed’ communities reduction in child malnutrition despite the downward trend in under- inside the park (adjacent to core areas of the NPA), as well as increased nourishment in Southeast Asia in general, as well as in Laos, over the pressure on land for those villages situated outside of the NPA area, but past decades. Similarly, a UN report on Millennium Development Goal having lost land to the NPA. Although agricultural production is the Progress in Laos finds that finds that hunger at household-level has in- main source of sustenance for the population in the area, the promotion creased since 2002 (http://la.one.un.org/millennium-development- of contract farming - initiated by foreign investors from China and Viet- goals/mdg-progress-in-lao-pdr). nam whose main interest is maize production - has also had profound Although rice sufficiency is both a national and local concern in Laos, impacts across the region (Broegaard et al., 2017; Castella et al., 2013; the main political focus is on increased economic growth to which the Messerli et al., 2009; Vongvisouk et al., 2014; Vongvisouk et al., 2016). agricultural sector is expected to contribute significantly, not leaving The three villages (Khorn Ngua, Son Koua, and Phon Song) were much attention to the importance of diverse and balanced diets (e.g. chosen to represent different degrees of proximity to the NPA as well World Bank, 2015). This lack of attention to diet quality issues is prob- as represent different positions in a transition process from traditional lematic for several reasons. Multiple studies have shown how wild shifting cultivation systems towards more continuous cultivation and food - especially wild animals (Scoones et al., 1992) – constitutes a sub- cash crop production. These sampling criteria were chosen in order to stantial part of household food consumption in shifting cultivation sys- yield insight on the cumulative influence of cash crop production and tems (Christensen, 2002; Cruz Garcia and Price, 2012; Fiedler, 1994). conservation on wild food collection and nutrition. Due to lack of longi- Krahn (2003) pinpoints that “many wild plant and animal species pro- tudinal data, we focus on differences between the villages as we argue vide foods with greater nutrient densities than are present in alternative that they represent different degrees of influence of cash crop produc- foods imported through market networks in remote mountain areas”(p. tion (Rasmussen et al., 2016b). There has been an increase in commer- 6) (for studies of nutrition and dietary diversity in the Amazon region cial production in all three study villages since maize was introduced in see also Piperata et al., 2011; van Vliet et al., 2015 and Sarti et al., 2010, although the increase in production has taken place in very differ- 2015; for Madagascar see Golden et al., 2011; for a multi-country ent ways. While wild food collection in principle can take place study in Africa see Ickowitz et al., 2014; and for Indonesia see Ickowitz throughout the landscape, we distinguish between the collections ac- et al., 2016). According to Krahn (2003) the traditional diet rich in cording to the habitat type where it took place, as we show in the results meat, vegetables, and fruit appears to be nutritionally adequate, and section. she identifies the increased rice consumption and reduced intake of Khorn Ngua is situated outside of the NPA, has 60 households, and is meat as the greatest nutritional challenge for the Lao upland population the village where the land use system has changed the least. It is still (Krahn, 2003;seealsoKrahn and Johnson, 2007). Similarly, Powell et al. predominantly traditional shifting cultivation, with most cultivation (2015) call for more attention to wild food contribution to overall diets concentrated on steep slopes. Farmers' main crop is rice, grown in the and several studies in Laos have attempted to map the composition of uplands for 1 to 2 years, after which the fields are left fallow. The culti- wild food in overall diets (Clendon, 2001; Foppes and Ketphanh, vated rice fields are the mainstay of the human diet, albeit wild animals 2004; Johnson et al., 2010; Kenichiro et al., 2004; Krahn, 2005). The and plants also are considered important sources of food. Main trapping comprehensive food security and vulnerability analysis in Laos by and catching techniques for rodents include snares, single-capture WFP (2007) concludes that traps, and pitfall traps. All members of the household are involved in wild food gathering, while able-bodied adults contribute to the cultiva- “wild meat and aquatic resources […] are the biggest sources of an- tion. No commercially produced fertilizers or pesticides are applied, and imal protein in rural Lao PDR. Consumption of domesticated animals maize is being more or less integrated in the shifting cultivation system, cannot currently compensate for a potential loss of access to and but with shorter rotation than shifting cultivation for upland rice. availability of wildlife” (p.1). Son Koua is the largest of the three villages, with 178 households. Analogous to Khorn Ngua, the expansion of commercial maize produc- The World Food Program (2007) also asserts that animal protein dif- tion has occurred in Son Koua through integrating maize into the rota- ferentiates households with acceptable food consumption from house- tional agricultural system. However, the village is located close to a holds with poor or borderline consumption. If the ongoing land use paved road (Fig. 2) implying that the community has better market R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101 95
Fig. 1. Map of the study sites, northern Laos. Landsat imagery (RGB = 3–2–1) from January 2015 was used as the base layer. access and is more integrated in a larger economic system. For example, decisions (Delang, 2006). Data were collected in each village during the village has been widely promoted as a site for ecotourism. four periods, corresponding to each of the four sub-periods of 1) slash- Phon Song is located within the NPA and in this village the ing and burning; 2) planting; 3) weeding; and 4) harvesting, in order to expanding cash crop production has in contrast to Khorn Ngua and ensure data-collection during different seasons. Each data collection pe- Son Koua occurred through a land use change from shifting cultivation riod lasted for one week per village. We differentiated between collec- of rice to continuously cultivated maize. Part of the reason for the con- tion from fields, fallows, and forest areas for terrestrial habitats. tinuous cropping is that the NEPL-NPA surrounds Phon Song, whether Diaries were chosen as a methodology, given that prior studies had as Controlled Use Zone or as Total Protection Zone (Fig.2), thereby lim- shown their usefulness and high level of detail to explore household iting access to arable land, and not allowing the increase in land under consumption of wild vegetables and other plants (Christensen, 1997; agriculture identified in other villages through interviews. Yet, with Lykke et al., 2002; Mertz et al., 2001; Rasmussen et al., 2016a). A re- the fallow period being omitted in the maize systems of Phon Song, it search assistant visited all eleven participating households every eve- is uncertain how long the continuous maize cultivation system can be ning in each of the villages during the recording weeks, and together sustained without causing land degradation. In Phon Song, the cultiva- they recorded the products collected, the quantity taken, the location tion involves commercial fertilizers and pesticides. The village has 50 of collection, and the final use of the product. Records were kept on a households. daily basis for one week in each of the four sub-periods, amounting to a total of 924 household-days of recordings for all three villages. To ac- 3.2. Data collection count for intra-household variations in collection patterns, all house- hold members were invited to participate in the evening sessions. If Three complementary methods - collection diaries, semi-structured some members were not available, they were asked to tell the parti- interviews and participant observation - were employed to examine cipating members prior to the session about their collection for that how the collection of wild food contributed to the diet. A spatial analysis day. of land cover changes was made for the delimited village lands between Daily, rather than weekly recordings were chosen in order to mini- 2006 and 2015, based on remotely sensed satellite imagery. mize a memory lapse and to account for daily variation and capture in- frequently collected items. Daily visits turned out to prompt a great 3.2.1. Collection diaries level of detail, as the products that had been collected during the day Collection diaries were used to estimate the amount and variety of often were shown and discussed with the research assistant. All hunted wild animals and plants collected by households from terrestrial and animals were identified by research assistants to taxonomic group, rath- aquatic habitats. The diaries did not include cultivated food (for exam- er than at the species level. ple from vegetable gardens and domestic animals) as they were de- signed with a specific focus on wild food. We specifically targeted 33 3.2.2. Semi-structured interviews households (11 from each of the three selected villages). A stratified During the first sample weeks of diary recordings (the slashing peri- sample of households was selected to ensure representation of house- od), semi-structured interviews were conducted with the same 33 holds with fields located at short, medium and long distances from the households who took part in the diary recordings in the three villages. village, as previous studies suggest that distance influences collection These interviews served to yield insight on cultivation practices and 96 R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101
Fig. 2. Land cover maps for 2006–2015 within the village boundaries of Khorn Ngua, Son Koua and Phon Song, based on Land Use Planning Reports 2012. Maps are based on classification and time series analysis of Landsat imagery. Landsat Imagery (RBG = 3–2–1) from 2006 to 2015 is used as a background base layer. the general collection of various items from various habitats. Addition- 3.3. Data analysis ally, these households were interviewed a second time after the harvest was completed in order to gather information on the agricultural pro- In order to estimate protein gaps, the quantities of collected rats (in duction, hunting (in the primary forest areas) and to validate the pat- grams) were used to estimate the percentage of required protein intake terns emerging from the collection diaries. To understand socio- that was satisfied by the wild food collection from the landscape under economic aspects of the agricultural production, we also interviewed a the different agricultural systems present in the three study sites. The bigger sample of households across the three villages (n = 100). This calculations were based on the following assumptions: (1) the per sample included the 33 households from the diary recordings and in- capita protein requirement is approximately 50 g per day (http://iom. cluded questions regarding all possible contributions to diet, including nationalacademies.org/Activities/Nutrition/SummaryDRIs), and (2) wild food, grown food, reared livestock and purchased food. protein levels of 7% and 18% in rice and rats, respectively. For the spatial analysis our aim was to map changes in land cover/ 3.2.3. Participant observation use for each of the three villages over a 10-year period from 2006 to In order to get a better understanding of the products households 2015. For a detailed description of the data and methodology used, see gathered on their way to and back from the fields, we accompanied Appendix A. For mapping, we used the rich archive of Landsat satellite each of the 33 households on these walks. This was done four times imagery freely available from the United States Geological Survey. We evenly distributed over the agricultural season, thereby providing a con- targeted four land cover/use categories: mature forest, forest fallow, ac- crete vantage point from which the respondent could describe the prod- tive cropland and other land. Of these classes forest fallow is by far the ucts he or she collected. Although the methods outlined above are well most difficult class to map with Landsat because after 3–4 years of re- suited to assess people's collection of wild foods, we note that partici- growth it becomes difficult to distinguish it from more mature forest pants may have (intentionally) failed to record, describe or showcase (Hett et al., 2012). To reliably map the fallow system, a dense time series all collected products, especially regarding illegal products. change detection algorithm was developed. This allowed the forest- R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101 97 cropland-regrowth patterns to be sufficiently identified. We first used a Random Forests classifier (Breiman, 2001) to classify a cloud free Landsat image for the chosen base year 2005 into 3 classes: forest, active cropland and other land. To map forest fallow for the base year 2005 we analyzed annual time series of Landsat imagery from 1989 to 2005. This analysis revealed forest that had been previously cleared for agriculture and regrown. We then continued the time series analysis from 2006 to 2015 to track how the fallow system and active cropland developed in the three villages. Manual corrections were made to the maps to remove any obvious errors introduced by the automated mapping procedure. For accuracy assessment, we used a stratified random validation sample of 120 points. Overall accuracy of the maps was 89% for 2006 and 91% for 2015.
4. Results
Comparison of the land cover maps for 2006 and 2015 for each vil- lage (Fig. 2) shows that the area of active agricultural fields in Khorn Fig. 3. Contribution of shifting cultivation landscape habitats to local people's collection of Ngua in 2015 is estimated to be smaller (280 ha) than in 2006 wild meat in three villages, Northern Laos. N = 924 household days and 258 collection (400 ha). Yet, when comparing spatial data for each year (Appendix events for the five most frequently collected wild meats. A), we see fluctuations in the yearly cleared area with peaks around 2010–2014. Yet, in Son Koua the land cover maps show a marked in- crease in active agricultural land, from approximately 420 ha in 2006 distances to these areas. However, it should be noted that households to 670 ha in 2015 (also with a peak in 2010 of approx. 890 ha).1 The ac- in all three villages hunted to some degree. tive agricultural area has been more stable in Phon Song, both with Rats were by far the most frequently gathered animal, accounting for regards to extent (from 250 to 320 ha between 2006 and 2015, with 125 out of the 336 collection events registered across the three villages. no marked peaks) and location. For maize-production to be economical- However, hunting of rats was much less common in Phon Song (22% of ly viable, good road infrastructure is a must, favouring this activity in all collection events) than in Khorn Ngua (36% of all collection events). Son Koua over Khorn Ngua. Meanwhile, the agricultural area is limited Across the three villages, households did also engage frequently in by the NPA regulations in the total protection zone surrounding Phon squirrel hunting (Fig. 3). Squirrels were primarily caught in the primary Song. The impact of the NPA regulations, although not entirely success- forest, but the majority of collection events yielded only one squirrel, ful, can also be seen in the other two villages, where agricultural activi- meaning that the meat intake per person was rather limited. ties are increasingly relocated to areas outside of the two NPA In Khorn Ngua, there were 70 rat collection events over four weeks, management zones. The land cover map also shows an increase in fal- equivalent to 1.6 events per household per week (Fig. 4a). In Phon Song, low areas especially in Son Koua. only 14 events were observed over the four weeks, equivalent to 0.3 events per household per week. The gathering of rats across all villages took primarily place in the agricultural fields, which accounted for 90% 4.1. Wild animal gathering from the terrestrial landscape of all recorded rat collection events (n = 125 events) in Khorn Ngua and 60% in both Son Koua and Phon Song (Fig. 4a). The same pattern Based on the household collection diaries, we found that rats, grass- was observed for the contribution of agricultural fields towards the fi hoppers, bamboo weevils, squirrels, and bamboo rats were the ve most total number of rats collected (n = 336 rats) with 92%, 50%, and 65% frequently gathered animals from the terrestrial landscape, accounting in Khorn Ngua, Son Koua, and Phon Song, respectively (Fig. 4b). A 2 for 77% of 336 reported collection events (Fig. 3). one-way ANOVA revealed a statistically significant difference A rather limited and irregular collection of other wild animals was (F(2121) = 4.6, p b 0.05) between the villages as to the number of observed across all villages, representing from 18% of all collection rats collected per collection event with a fairly limited number in fi events in Khorn Ngua to 38% in Phon Song. Interviews con rmed this Phon Song (M = 1.9, SD = 1.2) as compared to Khorn Ngua (M =2.6, contribution of other wild animals to household food consumption, al- SD = 1.4) and Son Koua (M = 3.3, SD = 2.0). A post hoc Tukey test though no longer a major component. Some hunting took place in all showed that Phon Song differed significantly at p b 0.05 from the three villages, and consisted primarily of legal hunting of birds and other villages. Focusing on rats as the main wild animal collected, and small mammals in forest areas within the village lands. One in four in- considering that these are primarily collected from the active fields terviewees reported that members of their household continued to (Fig. 4b), we contribute this difference mainly to the different degree hunt larger animals, particularly wild pigs and deer despite the regula- to which the agricultural system has changed in the villages, as a re- 3 tions of the NPA, though they were caught only a few times per year sponse to external pressures and opportunities. and represented a limited contribution to contemporary diets. In Son While all households who participated in the diaries hunted rats in Koua, the hunting of other wild animals was also limited because collec- Khorn Ngua, only 8 and 5 of the eleven participating households en- tion from forest areas was considered burdensome due to the longer gaged in rat hunting in Son Koua and Phon Song, respectively. A one- way ANOVA showed a statistically significant difference (F(230) = 1 The marked increase in active agricultural area in 2010 in Khorn Ngua and Son Koua 3.6, p b 0.05) between the villages as to the number of rats collected collided in time with maize traders investing heavily in the area. per household with the lowest collection in Phon Song (M =2.4, 2 A collection event yields typically more than one individual of the smaller animals. Av- erage yield per collection event was 2.6 for grasshoppers, 6.4 for bamboo weevils, 2.2 for SD = 3.7) as compared to Son Koua (M = 11.9, SD = 18.8) and bamboo rats and 1.3 for squirrels. For larger animals, one collection event typically yields Khorn Ngua (M = 16.5, SD = 10.6). A post hoc Tukey test showed only one animal. that Phon Song differed significantly at p b 0.05 from the other villages. 3 While some smaller animals were permitted objects for hunting, the NPA regulations The findings suggest that not only may the land use have changed made hunting of most larger mammals illegal. While previous large fines given for illegal the landscape and thereby affected the availability of wild foods, the dif- hunting were reported to reduce hunting activity, this is probably also subject to some under-reporting. However, data from interviews fit quite well with what was observed ferent land use practices and related tasks may also have changed during prolonged and repeated stays in the three villages. people's time allocated to wild food collection. Another way that the 98 R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101
Fig. 4. Contribution of shifting cultivation landscape habitats to local people's collection of rats and wild vegetables in three villages, Northern Laos. N = 924 household days, 125 rat collection events (A), 336 rats (B), and 874 wild vegetables collection events (C). changed land use may have influenced both the number of rats collect- having livestock and their contribution to diets showed that people ed and the frequency of rat collection events was through the introduc- commonly did not seek to eat them. Rather, the purpose of keeping live- tion and use of rodenticides, which made the traditional rat collection stock in these villages was to meet living costs as they arose and pay for unattractive due to fear of poisoning. Rodenticides were introduced in weddings etc., while domestic animals were primarily consumed at fes- Phon Song, but not in the other two villages, and according to inter- tivals. Other studies have likewise shown that livestock provide an im- views, the rodenticides and other pesticides were introduced by cash portant, although limited contribution to the diet (Clendon, 2001; crop investors. Krahn, 2005; WFP, 2007). Secondly, there was limited hunting of large wild animals, mainly due to the NPA regulations.5 Thirdly, the intuitive- ly appealing assumption that increased income would be translated into 4.2. Wild plants gathered from the terrestrial landscape improved food sufficiency and improved nutritional status makes it rel- evant to ask whether the identified ‘protein-gap’ was actually fulfilled Analogous to our findings on wild animal collection, the frequency of by purchase of other sources of protein. Our interviews showed that wild plant collection events was lowest in Phon Song and highest in while people were starting to consume more livestock and purchase a Khorn Ngua. Households collected wild plants for consumption almost little more meat in Son Koua and Khorn Ngua, it was not widespread, twice as frequently in Khorn Ngua (n = 387 events) as in Phon Song as it was primarily the wealthier households who could afford to do (n = 211 events) (Fig. 4c). The agricultural fields accounted for a sub- so. In Phon Song, which had the lowest rice sufficiency, highest protein stantial proportion (47% in Khorn Ngua, 36% in Son Koua, and 39% in gap, lowest collection of greens, and most difficult road access, meat Phon Song), but with similar quantity estimates per collection from purchase was almost non-existent. In general, there was limited food the different habitats.4 purchase in all three villages. Our interviews6 revealed that the With regards to wild plants, our data from the collection diaries re- household's agricultural activities in Khorn Ngua and Son Khua totaled vealed that households in Khorn Ngua on average collected wild plants a considerable higher value (reported in Million Kip (M = 8.81, SD = four times a week from the agricultural fields. The lowest collection fre- 4.41 and M = 8.76, SD =5.68, respectively), than in Phon Song (M = quency from the agricultural fields was observed in Phon Song (1.9 5.63, SD = 6.79). However, in all three villages, household rice sufficien- times per week per household). Based on the number of collection cy for all twelve months of the year, whether through own production events observed, our findings indicated that the wild plants contributed or through income and purchase, had decreased between 2004 and much more to a diverse diet in the villages with more traditional 2014, in Khorn Ngua from 93% to 73%, in Son Koua from 92% to 88%, shifting cultivation agricultural systems, and much less in the village and worst in Phon Song, where it had decreased from 83% to only 57% dominated by cash crop production. of households. Fourth, attention had to be devoted to the protein derived from 4.3. Protein-intake aquatic habitats, since fish generally are an important source of protein in Laos. Based on the collection diaries, we found that households in A one-way ANOVA was conducted to compare the three villages in Khorn Ngua and Son Koua on average had a fishing frequency of 2.5 terms of additional protein requirements after considering the two and 1.8 times per week, respectively, over a year. Similar to our findings main protein sources – rice and rats (Table 1). We found a statistically on rat collection events, the frequency was lowest in Phon Song: House- significant difference (F(230) = 3.4, p b 0.05) with the lowest additional holds' fishing efforts were restricted to once a week. The lower frequen- requirements in Khorn Ngua (M = 14.6%, SD = 12.6) as compared to cy in Phon Song was not outweighed by the amount of fish caught. On Phon Song (M = 35.8, SD = 10.9). A post hoc Tukey test showed that the contrary, the amounts caught in Khorn Ngua (820 g per fishing Phon Song differed significantly at p b 0.05 from the other villages. event) and Son Koua (780 g per fishing event) were almost twice the As to the question of whether households in Phon Song fulfilled the protein need by relying on other sources, we found that this was not the 5 case. Five points derived from the interview and collection diary data Restrictions were especially strongly enforced in Phon Song, but also felt in the other substantiate this conclusion. Firstly, interviews on the reasons for two communities. 6 Interviews were carried out with 30 households in Phon Song, 30 in Khorn Ngua and 40 in Son Khua. However, we exclude data on agricultural income from 6 households in 4 Most often one collection event would yield a handful or approximatively 30 g. Son Khua who did not cultivate land, as they held professional offices. R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101 99
Table 1 Protein derived from rice and rats in the terrestrial ecosystems of three villages, northern Laos.
Terrestrial protein sources Khorn Ngua Son Koua Phon Song
Ricea protein intake (g) per household per day 161 (SD = 22) 158 (SD = 21) 152 (SD = 18) Ratb protein intake (g) per household per day 53 (SD = 34) 38 (SD = 61) 8 (SD = 12) Protein need (%) after rice and rat meat intake 15 (SD = 13) 22 (SD = 28) 36 (SD = 11) n = 924 household-days and 336 rats. a Based on a protein content of 7%. b Based on a protein content of 18%. amount caught in Phon Song (441 g per fishing event). When these protein-rich foods and vegetables are rarely bought, partly as there is aquatic protein sources were added to the terrestrially derived protein, no easy access to a market, we conclude that even the increasing income we found that the protein intake in Khorn Ngua with the more tradi- reported in interviews obtained from cash crop production does not off- tional shifting cultivation system actually matched the protein need set the decrease in collection of wild food. Our empirical findings are (Table 2). In Son Koua, the need was reduced to b15% (which for the supported by Krahn (2005) who also found a food consumption pattern wealthiest households was probably offset by meat purchase), but in in rural Laos that is changing rapidly towards an almost exclusive rice the village of the Phon Song, a substantial protein need was still ob- diet, partly due to the change in access to and use of natural resources. served - on the order of 22–36%, and as mentioned above, was not offset Our findings are likewise in line with studies of the ‘nutritional transi- by purchase of food. tion’ from other parts of the world (Sarti et al., 2015; van Vliet et al., Finally, one could argue that protein derived from beans and wild 2015). These studies have found that the most ‘modern’ populations plants should be accounted for. Since our interviews did not yield reli- and villages in rural settings are those with the most inadequate nutri- able estimates on the amount of beans consumed (as these are grown tion. Increasing commercialization of agriculture causing large-scale in vegetable gardens, rather than collected, and therefore not included and rapid landscape and land use changes in the tropics, often in inter- in our diaries), we had to rely on a study from another community in action with conservation efforts, may thus have severe impacts on die- the NEPL-NPA (Johnson et al., 2010) which found that families had an tary outcomes. Yet, dietary outcomes of these changes have received intake of 32 g of beans per person per day. Using this figure and a pro- little attention. In addition to the lack of local diversified food markets, tein content of approximately 3%, we found that a need for protein possible explanations for the missing relation between rising income remained in Son Koua (0–13%) and Phon Song (20–34%). Since we iden- and improved diets include that cash income from maize production tified N120 different plant species and a collection event most often mainly are spent on household assets such as scooters, tractors and tele- entailed collection of a handful of different leaves, it was inherently dif- phones in addition to being invested in children's education and, for the ficult to translate those wild plants into their respective protein contri- wealthier households, placed as savings on bank accounts (Vongvisouk bution. However, it is well established that plants contribute an et al., 2016). important source of vitamins and micro-nutrients, in addition to some Dietary outcomes are strongly linked to the nature of agricultural de- protein (see studies in Powell et al., 2015), and our findings (Fig. 4c) in- velopment, conservation of forests, and the availability of and access to dicated that the contribution by wild plants, especially from the active wild food and our study advances existing knowledge by showing the upland fields, to a diverse diet, was much more pronounced in the vil- important contribution made from active agricultural fields in this tradi- lages with more traditional shifting cultivation agricultural systems, tionally shifting cultivation system, undergoing change. We find that and much less in the village dominated by continuous cash crop produc- the collection of wild food represents an important source of protein tion. Interviews showed that practically no purchase of vegetables and and other nutrients in more traditional shifting cultivation villages. other plants took place in any of the three villages, as there were no However, the wild food provision from the shifting cultivation land- local markets for these. scape appears to be diminishing with increasing pressures on land, whether from the NPA or the expanding cash crop production. Change 5. Discussion and concluding remarks from shifting cultivation to more continuous maize cultivation limits the amount of wild food that local people collect from the agricultural Our empirical focus on shifting cultivation systems in Laos has fields which represent by far the most important place of collection in shown that the studied villages have significantly different levels of di- our study area. Further, if households do not replace the wild food etary diversity, both with regards to protein and vegetable intake from with other protein sources, which is the situation in the villages furthest wild sources. Based on interviews as well as land cover change analysis away from good roads and markets, it may result in a lack of protein. we explain the observed difference in wild food collection between the Thus, we also show how the conversion towards more intensive agricul- villages with the degree to which they have undergone recent land use ture has different outcomes in different types of villages. Those villages changes related to cash crop expansion as well as conservation initia- with more agricultural land and less pressure from conservation initia- tives. These changes are in parts driven by agricultural policies that pur- tives may benefit from increasing incomes from commercial farming sue land sparing rather than land sharing, although they may obtain and still be able to maintain, and possibly even enhance diets whilst neither (Vongvisouk et al., 2016), yet may affect local peoples' nutrition continuing to access a range of wild foods. In contrast, villages with lim- negatively. As our interviews and observations show a situation where ited agricultural land and restrictive rules on cultivation and foraging may experience income increases happening in concert with a radically Table 2 declining availability of wild food. However, it is important to remem- Protein derived through fish from aquatic ecosystems in three villages, northern Laos. ber that households differ within villages and while some may have in- come and be somewhat resilient to the dietary effects, the poorest and Aquatic protein sources Khorn Son Phon Ngua Koua Song those with access to the least land, are likely to suffer far more. Getting sufficient animal protein is a general challenge in many rural Fisha protein intake (g) per capita per day 10 6.8 2.5 Protein need after rice, rat meat, and fish 0–3% 1–15% 22–36% areas and in particular in rural Laos (Johnson et al., 2010; Krahn and intake Johnson, 2007; Krahn, 2003, 2005)eventhoughfish and other animals from aquatic habitats provide an important source of protein. Thus, our n = 924 household-days and 240 fishing events. a Based on a fish intake of 59, 40, and 15 g per capita per day for each of the three study study aims to improve the understanding of how changes in agricultural sites, respectively, and an estimated protein content of 17% in fish. production system (from more subsistence-oriented shifting cultivation 100 R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101 to a system with considerable production of cash crops for regional or cross-sectoral analysis, as outcomes pursued in land use planning may foreign markets) affect the collection of protein-rich foods. Further- have unintended consequences on other aspects of rural livelihoods. more, we have demonstrated that collection of wild plants also dimin- ishes with increasing pressures on land and a reduction of the shifting Acknowledgements cultivation landscape. As no purchase of vegetables and other plants takes place, the reduced collection translates directly into a lower diet We thank Finn Danielsen, Thomas Sikor, Adrian Martin, Sithong quality. Our findings thus question the commonly assumed link be- Thongmanivong and Pheang Xaydongvanh for their contribution during tween cash income and dietary diversity, although it may hold for the research process and to Kenneth Grogan for the elaboration of maps. household food sufficiency. Failing to pay attention to the ‘hidden hun- The work was part of the project ‘Ecosystem Services, Wellbeing and ger’ as mentioned in the introduction may overlook potential protein Justice: Developing Tools for Research and Development Practice’, and nutrient deficiencies caused by the declining collection of wild funded with support from the Ecosystem Services for Poverty Allevia- food and the very limited food purchase in many rural villages. tion (ESPA) programme. The ESPA programme is funded by the Depart- In the presented study, we have taken a landscape approach to ana- ment for International Development (DFID), the Economic and Social lyse changes in collection of wild foods in that we distinguish between Research Council (ESRC) and the Natural Environment Research Council what is collected from different habitat categories throughout the land- (NERC), Grant no. NE/L001411/1. See more at www.espa.ac.uk/. Addi- scape. We found that rats were the most frequently collected source of tional funding was provided by the project ‘Property and Citizenship protein and these were primarily hunted in the agricultural fields. This in Developing Societies (ProCit)’ funded by the Danish Independent Re- finding are in line with recent findings of the importance of wild food search Council - Social Sciences, Grant no. 11-104613. The funding bod- from agricultural areas by Powell et al. (2013) and underscores the im- ies had no involvement in the design or any other part of the research portance of taking into consideration the changes in production on process. those fields, or the use of different inputs on these fields, in order to un- derstand the implications of land use changes for the provisioning ser- Appendix A. Supplementary data vices from terrestrial ecosystems with regards to wild food. We therefore propose that future studies dealing with the intersection be- Supplementary data to this article can be found online at http://dx. tween land use changes, nutrition, and food security devote specificat- doi.org/10.1016/j.forpol.2016.12.012. tention to the contribution of wild food from agricultural fields and threats to this contribution. Increased use of agro-chemicals (pesticides, References herbicides and rodenticides) has clearly limited the collection of, for ex- ample, animals for consumption. We expect the use of agro-chemicals Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N.J., Bauch, S., ... Wunder, S., 2014. Environmental income and rural livelihoods: a global-comparative analysis. to increase further, with greater contact to traders/investors who pro- World Dev. 64, S12–S28 (Supplement 1). mote cash crop production. This would be an important element to Breiman, L., 2001. Random forests. Mach. Learn. 45, 5–32. monitor in future research on nutrition-sensitive landscapes, i.e. Broegaard, R.B., Vongvisouk, T., Mertz, O., 2017. Contradictory land use plans and policies in Laos: tenure security and the threat of exclusion. World Dev. 89:170–189. http:// the impact of the changing land use systems on food security and dx.doi.org/10.1016/j.worlddev.2016.08.008. nutrition. Here, the impact of increased use of agro-chemicals on Castella, J.C., Lestrelin, G., Hett, C., Bourgoin, J., Fitriana, Y.R., Heinimann, A., Pfund, J.L., the aquatic environment and its food provision has to be taken into 2013. Effects of landscape segregation on livelihood vulnerability: moving from ex- tensive shifting cultivation to rotational agriculture and natural forests in northern account as well. Laos. Hum. Ecol. 41, 63–76. At a more general level, we call for attention to the concept of nutri- Clendon, K., 2001. The role of forest food resources in village livelihood systems. A study tion-sensitive landscapes (Global Nutrition Report, 2014) for two pur- of three villages in Salavan province, Lao PDR. IUCN/NTFP Project 29. International Union of Nature Conservation, Vientiane. poses: First, to acknowledge the intersection between nutrient and Christensen, H., 1997. Uses of ferns in two indigenous communities in Sarawak, Malaysia. diet studies and land use studies, and secondly, to provide a basis for In: Johns, R.J. (Ed.), Holttum Memorial Volume. Kew, Royal Botanic Gardens. updating theories of changes about the linkages between land use Christensen, H., 2002. Ethnobotany of the Iban and the Kelabit. Kuching and Aarhus, For- changes on the one hand, and food sufficiency and possible nutrient de- est Department, Sarawak; NEPCon and University of Aarhus. Cruz Garcia, G.S., Price, L.L., 2012. Weeds as important vegetables for farmers. Acta Soc. ficiencies on the other. This has implications for the SDGs and their at- Bot. Pol. 81, 397–403. tempt to integrate the previously distinct aspects of landscapes, Dawson, N., Martin, A., Sikor, T., 2016. Green revolution in Sub-Saharan Africa: implica- conservation and poverty, which should receive much greater focus as tions of imposed innovation for the wellbeing of rural smallholders. World Dev. 78, 204–218. part of the SDG development agenda (Messerli et al., 2015). Our study Delang, C.O., 2006. Not just minor forest products: the economic rationale for the con- supports findings that dietary outcomes are strongly linked to the na- sumption of wild food plants by subsistence farmers. Ecol. Econ. 59 (1):64–73. ture of agricultural development, conservation of forests, and the access http://dx.doi.org/10.1016/j.ecolecon.2005.10.006. Ducourtieux, O., Laffort, J.-R., Sacklokham, S., 2005. Land policy and farming practices in to wild foods, especially where road access is limited. In line with Laos. Dev. Chang. 36 (3), 499–526. Shackleton and Pandey (2014), we argue that the role of wild foods Dwyer, M., 2011. Territorial Affairs: Turning Battlefields Into Marketplaces in Postwar must be integrated into poverty alleviation programmes as well as Laos. PhD, University of California, Berkeley. FAO, 2015. The state of food insecurity in the world 2015. Meeting the 2015 International taken into account in protected area management. Our study contrib- Hunger Targets: Taking Stock of Uneven Progress. FAO, Rome. utes to Sibathu et al. (2015)s' call for research that helps provide a better Fiedler, L.A., 1994. Rodent Pest Management in Eastern Africa. FAO, Rome. understanding of how agriculture and food systems can be made more Foppes, J., Ketphanh, S., 2004. NTFP use and household food security in Lao PDR. Paper – Presented at the NAFRI/FAO EM-1093 Symposium on “Biodiversity for Food Security” nutrition-sensitive in particular situations in the particular situation of Septemper 14, 2004, in Vientiane. shifting cultivation systems under increasing pressures from different Foran, T., Butler, J.R.A., Williams, L.J., Wanjura, W.J., Hall, A., Carter, L., Carberry, P.S., 2014. land uses, including conservation and expanding cash crop production. Taking complexity in food systems seriously: an interdisciplinary analysis. World – The policy implications of our findings are that conservation and devel- Dev. 61, 85 101. Global Nutrition Report, 2014. United National System Standing Committee for Nutrition. opment policies need to devote attention to the linked sectors of health, Technical note 4. Towards Sustainable, Healthy and Profitable Food Systems: Nutri- poverty, and agriculture, and address how these linkages play out in the tion and the Sustainable Management of Natural Resources 5. Global Nutrition. face of large-scale land use changes such as concessions and contract- GoL, 2005. Forestry Strategy to the Year 2020 of Lao PDR. Government of Laos, Ministry of Agriculture and Forestry, Vientiane, Lao PDR. farming schemes. Similar calls for integration of sectorial policies are Golden, C.D., Fernald, L.C.H., Brashares, J.S., Rasolofoniaina, B.J.R., Kremen, C., 2011. Bene- made by Sarti et al. (2015); Shackleton et al. (2015) and van Vliet et fits of wildlife consumption to child nutrition in a biodiversity hotspot. Proc. Natl. al. (2015), among others. In Laos, the government's Land Use Plans are Acad.Sci.U.S.A.108(49):19653–19656. http://dx.doi.org/10.1073/pnas. 1112586108. a major implementation tool for governing land use at a landscape Hall, D., 2011. Land grabs, land control, and Southeast Asian crop booms. J. Peasant Stud. level, and our study clearly shows the need for the above mentioned 38 (4), 837–857. R.B. Broegaard et al. / Forest Policy and Economics 84 (2017) 92–101 101
Hall, D., Hirsch, P., Li, T.M., 2011. Powers of exclusion: Land dilemmas in Southeast Asia. Ribot, J.C., Agrawal, A., Larson, A.M., 2006. Recentralizing while decentralizing: how na- Honolulu, University of Hawaii Press. tional governments reappropriate forest resources. World Dev. 34 (11), 1864–1886. Hett, C., Castella, J.-C., Heinimann, A., Messerli, P., Pfund, J.-L., 2012. A landscape mosaics Rowland, D., Ickowitz, A., Powell, B., Nasi, R., Sunderland, T.C.H., 2016. Forest food and approach for characterizing swidden systems from a REDD+ perspective. Appl. healthy diets: quantifying the contributions. Environ. Conserv.:1–13 http://dx.doi. Geogr. 32 (2), 608–618. org/10.1017/S0376892916000151. Ickowitz, A., Powell, B., Salim, M.A., Sunderland, T.C.H., 2014. Dietary quality and tree Sandbrook, C., Nelson, F., Adams, W.M., Agrawal, A., 2010. Carbon, forests and the REDD cover in Africa. Glob. Environ. Chang. 24, 287–294. paradox. Oryx 44 (03), 330–334. Ickowitz, A., Rowland, D., Powell, B., Salim, M.A., Sunderland, T.C.H., 2016. Forests, trees, Sarti, F.M., Adams, C., Morsello, C., van Vlien, N., Schor, T., Yagüe, B., ... Cruz, D., 2015. Be- and micronutrient-rich food consumption in Indonesia. PLoS One 11 (5) (e = yond protein intake: bushmeat as a source of micronutrients in the Amazon. Ecol. 154139). 10.137/journal.pone.0154139. Soc. 20 (4), 22. Johnson, A., Krahn, J., Seateun, S., 2010. Finding the Linkages Between Wildlife Manage- Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.-L., Sheil, D., Meijaard, E., Venter, M., ment and Household Food Consumption in the Uplands of Lao People's Democratic Boedhihartono, A.K., Day, M., Garcia, C., Van Oosten, C., Buck, L.E., 2013. Ten principles Republic: A Case Study from the Nam Et-Phou Louey National Protected Area. WCS for a landscape approach to reconciling agriculture, conservation, and other compet- TransLinks Program. Wildlife Conservation Society, Vientiane. ing land uses. Proc. Natl. Acad. Sci. 110 (21), 8349–8356. Kenney-Lazar, M., 2012. Plantation rubber, land grabbing and social-property transforma- Schönweger, O., Heinimann, A., Epprecht, M., Lu, J., Thalongsengchanh, P., 2012. Conces- tion in southern Laos. The Journal of Peasant Studies 39 (3–4), 1017–1037. sions and Leases in the Lao PDR: Taking Stock of Land Investments. 96. Centre for De- Kenichiro, Y., Masayuki, Y., Yasuyuk, K., Eiji, N., 2004. Use of natural biological resources velopment and Environment (CDE), University of Bern, Bern and Vientiane. and their roles in household food security in Northwest Laos. Southeast Asian Stud. Scoones, I., Melnyk, M., Pretty, J.N., 1992. The Hidden Harvest: Wild Foods and Agricultur- 41 (4), 426–443. al Systems. A Literature Review and Annotated Bibliography. International Institute Krahn, J., 2003. Cooking up: dietary change in lao upland kitchens. Juth Pakai 1, 4–15. for Environment and Development, London. Krahn, J., 2005. The Dynamics of Dietary Change of Transitional Food Systems in Tropical Sen, A., 1983. Poverty and Famines: An Essay on Entitlement and Deprivation. Oxford Uni- Forest Areas of Southeast Asia: The Contemporary and Traditional Food System of the versity Press, Oxford. Katu in the Sekong Province, Lao PDR. PhD, Rheinischen Friedrich-Wilhelms- Shackleton, C.M., Pandey, A.K., 2014. Positioning non-timber forest products on the devel- Universität. opment agenda. Forest Policy Econ. 38, 1–7. Krahn, J., Johnson, A.A.H., 2007. Upland food security and wildlife management. Juth Pakai Shackleton, C.M., Hamer, N., Swallow, B., Ncube, K., 2015. Addressing local level food in- 9, 17–33. security amongst small-holder communities in transition. Policy Brief 12–2015. De- Lestrelin, G., Trockenbrodt, M., Phanvilay, K., Thongmanivong, S., Vongvisouk, T., Pham, partment of Environmental Science, Rhodes University . http://www.ru.ac.za/ T.T., Castella, J.C., 2013. The context of REDD+ in the Lao People's Democratic Repub- media/rhodesuniversity/content/environmentalscience/documents/Policy%20Brief% lic. Drivers, agents and institutions. Occas. Pap. 92 (Center for International Forestry 2012.pdf. Research, Bogor, Indonesia). Sibathu, K.T., Krishna, V., Qaim, M., 2015. Production diversity and dietary diversity in Lykke, A.M., Mertz, O., Ganaba, S., 2002. Food consumption in rural Burkina Faso. Ecology smallholder farm households. Proc. Natl. Acad. Sci. 112 (34), 10657–10662. of Food and Nutrition 41:119–153. http://dx.doi.org/10.1080/03670240214492. Suhardiman, D., Giordano, M., Keovilignavong, O., Sotoukee, T., 2015. Revealing the hid- Mertz, O., Lykke, A.M., Reenberg, A., 2001. Importance and seasonality of vegetable con- den effects of land grabbing through better understanding of farmers' strategies in sumption and marketing in Burkina Faso. Econ. Bot. 55:276–289. http://dx.doi.org/ dealing with land loss. Land Use Policy 49:195–202. http://dx.doi.org/10.1016/j. 10.1007/BF02864565. landusepol.2015.08.014. Messerli, P., Bader, C., Hett, C., Epprecht, M., Heinimann, A., 2015. Towards a spatial under- Van Noordwijk, M., Bizard, V., Wangpakapattanawong, P., Tata, H.L., Villamor, G.B., standing of trade-offs in sustainable development: a Meso-scale analysis of the nexus Leimona, B., 2014. Tree cover transitions and food security in Southeast Asia. Glob. between land use, poverty, and environment in the Lao PDR. PLoS One 10 (7), Food Sec. 3 (3–4), 200–208. e0133418. http://dx.doi.org/10.1371/journal.pone.0133418. van Vliet, N., Quiceno-Mesa, M.P., Cruz-Antia, D., Tellez, L., Martins, C., Haiden, E., ... Nasi, Messerli, P., Heinimann, A., Epprecht, M., 2009. Finding homogeneity in heterogeneity - a R., 2015. From fish and bushmeat to chicken Nuggets: the nutrition transition in a new approach to quantifying landscape mosaics developed for the Lao PDR. Hum. continuum from rural to urban settings in the Colombian Amazon region. Ethnobiol. Ecol. 37 (3), 291–304. and Conserv. 4 (6), 1–12. Moore, C., Hansel, T., Johnson, A., Naughton-Treves, L., Day, C., 2012. REDD+ in Lao PDR: van Vliet, N., Mertz, O., Heinimann, A., Langanke, T., Pascual, U., Schmook, B., ... Ziegler, Is it also a “plus” for forest-dependent communities? A case study from the Nam Et A.D., 2012. Trends, drivers and impacts of changes in swidden cultivation in tropical Phou Louey national protected area, Lao PDR. Lessons About Land Tenure, Forest Gov- forest-agriculture frontiers: a global assessment. Glob. Environ. Chang. 22 (2), ernance and redd+. Case Studies From Africa, Asia and Latin America. UW-Madison 418–429. Land Tenure Center, Madison, Wisconsin, pp. 83–91. Vedeld, P., Jumane, A., Wapalila, G., Songorwa, A., 2012. Protected areas, poverty and con- Padoch, C., Sunderland, T., 2013. Managing landscapes for greater food security and im- flicts: a livelihood case study of Mikumi National Park, Tanzania. Forest Policy Econ. proved livelihoods. Unasylva 214 (64), 3–13. 21, 20–31. Paumgarten, F., Shackleton, C.M., 2011. The role of non-timber forest products in house- Vira, B., Wildburger, C., Mansourian, S., Vira, B., Wildburger, C., Mansourian, S., 2015. For- hold coping strategies in South Africa: the influence of household wealth and gender. ests and Food: Addressing Hunger and Nutrition Across Sustainable Landscapes. Popul. Environ. 33 (1), 108–131. Open Books Publisher & IUFRO, Cambridge, UK. Pinstrup-Andersen, P., 2009. Food security: definition and measurement. Food Security 1 Vongvisouk, T., Mertz, O., Thongmanivong, S., Heinimann, A., Phanvilay, K., 2014. Shifting (1), 5–7. cultivation stability and change: contrasting pathways of land use and livelihood Pingali, P., 2015. Agricultural policy and nutrition outcomes – getting beyond the preoc- change in Laos. Appl. Geogr. 46, 1–10. cupation with staple grains. Food Security 7 (3), 583–591. Vongvisouk, T., Broegaard, R.B., Mertz, O., Thongmanivong, S., 2016. Rush for cash crops Piperata, B.A., Ivanova, S.A., Da-gloria, P., Gonçalo, V., Polsky, A., Spence, J.E., Murrieta, and forest protection: Neither land sparing nor land sharing. Land Use Policy 55: R.S.S., 2011. Nutrition in transition: dietary patterns of rural Amazonian women dur- 182–192. http://www.sciencedirect.com/science/article/pii/S0264837716303088. ing a period of economic change. Am. J. Hum. Biol. 23 (4):458–469. http://dx.doi.org/ West, P., Igoe, J., Brockington, D., 2006. Parks and peoples: the social impact of protected 10.1002/ajhb.21147. areas. Annu. Rev. Anthropol. 35, 251–277. Powell, B., Maundu, P., Kuhnlein, H.V., Johns, T., 2013. Wild foods from farm and forest in WFP, 2007. Lao PDR: Comprehensive Food Security and Vulnerability Analysis (CFSVA) the east Usambara mountains, Tanzania. Ecol. Food and Nutr. 52 (6), 451–478. 139. World Food Programme, Vientiane. Powell, B., Thilsted, S., Ickowitz, A., Termote, C., Sunderland, T., Herforth, A., 2015. Improv- WFP, 2010. WFP Lao PDR Country Strategy 2011–2015. 24. WFP, Vientiane. ing diets with wild and cultivated biodiversity from across the landscape. Food Sec. 7 World Bank Group, 2014. Country Partnership Strategy Progress Report for Lao People's (3), 535–554. Democratic Republic. Period FY12-FY16. Report no. 90281-LA. September 16, 2014. Rasmussen, L.V., Mertz, O., Christensen, A.E., Danielsen, F., Dawson, N., Xaydongvanh, P., Southeast Asia Country Management Unit, East Asia and Pacific Region. International 2016a. A combination of methods needed to assess the actual use of provisioning Finance Corporation, East Asia and Pacific Region. ecosystem services. Ecosyst. Serv. 17, 75–86. World Bank, 2015. Lao Upland Food Security Improvement Project, Implementation Com- Rasmussen, L.V., Christensen, A.E., Danielsen, F., Dawson, N., Martin, A., Mertz, O., Sikor, T., pletion and Results Report 49. World Bank, Vientiane. Thongmanivong, S., Xaydongvanh, P., 2016b. From food to pest: conversion factors Wunder, S., Börner, J., Shively, G., Wyman, M., 2014. Safety nets, gap filling and forests: a determine switches between ecosystem services and disservices. Ambio:1–11 global-comparative perspective. World Dev. 64, S29–S42 (Supplement 1). http://dx.doi.org/10.1007/s13280-016-0813-6. Remis, M.J., Jost Robinson, C.A., 2014. Examining short-term nutritional status among BaAka foragers in transitional economies. Am. J. Phys. Anthropol. 154 (3):365–375. http://dx.doi.org/10.1002/ajpa.22521. Forest Policy and Economics 84 (2017) 102–111
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Forest edges in western Uganda: From refuge for the poor to zone of investment☆
Jessica L'Roe a,⁎,LisaNaughton-Trevesb a Middlebury College, Department of Geography, McCardell Bicentennial Hall, 287 Bicentennial Way, Middlebury, VT 05753, USA b University of Wisconsin-Madison, Department of Geography, Science Hall, 550 N. Park St., Madison, WI 53706, USA article info abstract
Article history: Western Uganda is home to growing populations of smallholder agriculturalists, expanding commodity planta- Received 14 May 2016 tions, and protected forests. In this setting, we document a shift in who uses forest edge land and how it is Received in revised form 8 December 2016 used. In developing countries, protected forest edges are traditionally sites where marginalized people can sub- Accepted 19 December 2016 sist, but increasing land competition has the potential to change this scenario. We used longitudinal field data Available online 11 January 2017 spanning two decades to characterize the evolution of landownership and land use neighboring Kibale National Park. The number of households has more than doubled since 1993. Land values are rising, and people buying Keywords: fi Land competition land near the park in recent years are signi cantly wealthier and have more off-farm income than those who ac- Land use change quired land there in earlier periods. The reverse is true of renters. More people are growing inedible perennial Land values cash crops like eucalyptus, tea, and coffee, especially those with larger amounts of land and capital. Some long- Commodity crops term residents are prospering, while others are squeezed onto ever smaller pieces of land and opting for precar- Protected area ious rental arrangements as land competition increases. We discuss the implications of this transitioning park East Africa neighborhood, both for conservation and local livelihoods. © 2016 Elsevier B.V. All rights reserved.
1. Introduction when parks themselves cause nearby land to be marginal for agriculture, this makes land near parks cheaper and thus more accessible to the poor 1.1. Changing dynamics around protected forests whomaynotbeabletosecurelandelsewhere(Naughton-Treves, 1997). On the other hand, poorer people may be attracted to park edges because The boundaries of tropical forest parks often mark areas where poor they benefitfromforestresources(Byron and Arnold, 1999, Angelsen and people subsist on economically marginal land (Naughton-Treves et al., Wunder, 2003; Naughton-Treves et al., 2011)orincomefromintegrated 2005). However, as economies and populations grow and physical in- conservation and development projects, forestry, or tourism (Wittemyer frastructure more tightly links remote places to international and et al., 2008; Sims, 2010). Parks are also associated with poverty through urban demand, competition for historically marginal land increases more circumstantial relations - land that is remote or marginal for agricul- (Lambin and Meyfroidt, 2011). Land competition can lead to land user ture is less likely to be cleared already and is politically easier to protect change as some groups compete more successfully, and shifts in users' because it is less suitable for competing uses, the same characteristics capital and constraints may also spur land cover change. As we plan that make it easier for the poor to find room and maintain claims in for the future of forest protection, with high stakes for forests and for- these locations (Joppa et al., 2008; Zommers and MacDonald, 2012). est-dependent people, it is important to attend to processes changing The geographic coincidence of protected areas with poorer people in the social and ecological context around protected forests. the tropics, i.e. the “poor people - rich forests” scenario, is a well-docu- Protected forests are affected by exogenous socioeconomic and eco- mented pattern with competing explanations (Peluso, 1992; logical processes, and parks in turn shape local land use and livelihood Naughton-Treves et al., 2011). Whether they find room near parks strategies. Parks can create or exacerbate poverty when they displace when they cannot compete for better land elsewhere or are attracted by people or cut off access to resources (Adams and Hutton, 2007). Existence natural resources or community conservation projects, poor households of protected areas can depress suitability for agriculture on neighboring are often disproportionately represented at the edges of protected forests. land due to crop loss to wildlife, a particularly serious concern around Af- The role of the forest edge as a haven for the poor may not last. rican parks where elephants abound (e.g. Vedeld et al., 2012). Yet, even Where populations or economies are growing rapidly, the value of for- merly marginal land may rise faster than the incomes of the people cur- rently using it, especially as land becomes scarce. Classic frontier ☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” dynamics are characterized by expansion of poorer households into in- published at the journal Forest Policy and Economics 84C, 2017. creasingly remote or less suitable land as land settled earlier is devel- ⁎ Corresponding author. E-mail addresses: [email protected] (J. L'Roe), [email protected] oped, consolidated, and purchased by wealthier owners (Barbier, (L. Naughton-Treves). 1997), but as these waves of socioeconomic pressure reach areas that
http://dx.doi.org/10.1016/j.forpol.2016.12.011 1389-9341/© 2016 Elsevier B.V. All rights reserved. J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111 103 are protected and closed to expansion, the social and ecological currents (Hartter, 2010), and growing at a rate of over 3% per year (Naughton- and eddies in that zone of contact are harder to predict. They depend on Treves et al., 2007). Recent decades have been marked by rising house- the mechanisms by which particular parks affect surrounding land hold incomes and declining forest cover around Kibale (Naughton- values, and on the regional drivers of development pressure. The poten- Treves et al., 2011). Today, little closed canopy forest remains outside tial for market or institutional conditions to quickly change park edge the park boundaries (Southworth et al., 2010). These local trends mirror dynamics may be significant, depending on which combination of fac- the rapid economic change and deforestation found elsewhere in West- tors is making land around a protected forest relatively cheap. In cases ern Uganda and the East African highlands (Plumptre, 2002). where parks are refuges for both humans and nonhumans struggling fi to nd space in increasingly crowded landscapes, the buffer of cheap 1.3. Study objectives and less desirable land may be shrinking. For example, park-edge land values could rise if they had been traditionally depressed by risk of In this study, we track park-edge land use dynamics over a 22-year crop damage from park animals and regional markets get stronger for period in a region of rapid population growth and economic develop- crops unpalatable to wildlife (e.g. eucalyptus) (Foster and Rosenzweig, ment. We use information about land sales during this time period to 2003). This could represent a boon for current users, or a risk of dis- support qualitative claims for rising land competition. We document a placement by more capitalized land owners enticed by new or im- shift over time in the characteristics of people buying land near the 1 proved prospects for potential returns (Jayne et al., 2014). Pathways park, alongside changes in how people are using land against the park. like these have the potential to dramatically change composition of We subsequently explore factors predicting which households are park neighbors, with implications for who gets to use the land and the able to participate in the shift to alternative more profitable land uses. way that land around protected forests is used. We conclude by discussing social and ecological implications of poten- Land dynamics surrounding protected forests are important for bio- tial changes in forest-people relationships around protected areas in diversity conservation, in addition to livelihood concerns raised above. the East African highlands and elsewhere in the tropics where high bio- Protected areas are a key strategy for maintaining forests and the spe- diversity, historically cheap land, and rapid growth coincide. cies that depend on them, especially in regions with high potential for agriculture or other land uses (Naughton-Treves et al., 2005). However, 2. Methods deforestation in surrounding regions is leaving tropical parks more iso- lated and biodiversity more vulnerable to climate change or other This study consists of complimentary lines of inquiry to characterize threats (DeFries et al., 2005; Hall et al., 2009). Sharp park edges are es- change over time in the focal area. We use reported land transactions to pecially evident in the highlands of East Africa, where deforestation track change in land prices, transects to characterize crop cover change, rates have soared in past decades (Oluka, 2014; Twongyirwe et al., and household surveys to capture change in land user characteristics 2015). A shift among land users around protected forests may mean in- and circumstances. Our focal area is set in agricultural land bordering tensified land use at the park edge or a displacement of poorer land Kibale National Park where 60 original transects were established in users to land still unclaimed and unprotected. Alternatively, a shift in 1993 in 5 hamlets to study crop-raiding by wildlife.3 In each hamlet, land users toward investors growing perennial tree crops could mean transects were spaced 50 m apart extending perpendicularly from the less extraction from natural forests and less wildlife conflict. As has park boundary into neighboring agricultural land. The small hamlets been documented at broader scales, the impacts of increasing wealth are located in 3 wards: Kanyawara (containing the hamlets of on forests can work in multiple directions (Foster and Rosenzweig, Kanyasohera and Kabucikire), Rurama (containing Kyakiheka and 2003; Rudel et al., 2005; Alix-Garcia et al., 2013). Kijonjomi), and one hamlet in Nyabubale (Fig. 1). The transects extend a quarter to a half kilometer from the edge of the park. In 2011, we re- 1.2. Study site: Uganda's Kibale National Park established the 1993 transects using landowner maps, natural land- forms (e.g. streams), and roads for reference. Two members of the The area around Kibale National Park in Western Uganda is an apt 1993 research team helped to maintain consistency in location and microcosm for examining the dynamics associated with increasing methods. In both time periods, we measured crop types, mapped land- land competition around protected forests. The park is home to many holdings, and surveyed land users in the area under transects. threatened species and boasts one of the highest primate densities in Crops were recorded at the beginning of each of the year's two plant- the world (Chapman et al., 2005b), attracting both tourists and forest ing seasons, every 10 m, for the 25 m on either side of a transect. All crop research programs. Its tropical pre-montane climate and fertile volcanic types (e.g. maize, cassava, and 1 year-old fallow) were mapped along soils favor rain-fed agriculture and the park is neighbored by smallhold- transects in 2011, and transect cover maps were updated in July 2015 4 er farmers intermixed alongside commercial tea plantations. In the using broader categories of land cover: food crops, brewing bananas, 1960s and 1970s, a wave of agriculturalists from land-scarce regions im- fallow, pasture, natural forest, and the main perennial inedible crops - migrated to the area to work in the tea plantations, subsequently pur- eucalyptus, tea, and coffee. We calculate crop cover as presence or ab- chasing land nearby and resuming independent farming activities sence of a given crop type within resulting transect grid cells with a (Mulley and Unruh, 2004; Hartter et al., 2015). Newer and poorer resi- 10 × 25 m resolution, and we track changes in percent of cells contain- dents disproportionately settled immediately adjacent to Kibale forest ing each category of land cover in 1993 and 2015. fi as locals preferred to give or sell land in this border zone to buffer We mapped parcels in both time periods to aid identi cation of their communities' farms from wildlife (Naughton-Treves, 1997).2 Dur- people using land on the transects. For socioeconomic data, our unit of 5 ing the past decade, the Ugandan government, the tea industry, and analysis is the household. The inclusion criterion was owning or other aid and development groups have been making significant invest- 3 ments in the region's infrastructure. Like other East African highland When the transects were first established in 1993, they covered practically all loca- sites, population density around Kibale is high: ~300 people/km2 tions where agriculturalists were living immediately adjacent to the park in the park's northwestern quadrant; since that time, intervening areas have also been cleared and set- tled. Because these sites are located in the quadrant of the park nearest the town of Fort 1 Land is an especially attractive investment when imperfect financial institutions limit Portal, the land dynamics we document are more representative of situations where other opportunities for investment (Shackleton et al., 2001). protected land occurs in areas of development pressure- socioeconomic conditions vary 2 Informal land markets have been active in Western Uganda at least as early as the across regions of the 766 km2 park (Naughton-Treves, 1997). 1960s (Chimowu and Woodhouse, 2006; Hartter et al., 2015), and the Ugandan govern- 4 Includes annual food crops along with sweet and edible bananas. ment has promoted the transition from customary to market-based freehold tenure 5 We defined a household as a group of people sharing a compound, regardless of famil- (Deininger and Mpuga, 2002). ial relationships. 104 J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111
incomplete information, we remained with 64 land market transactions spanning the decades from 1985 to 2015.9 To correct for the effect of significant monetary inflation in Uganda during this period, prices were divided by a consumer price index produced for Uganda (IMF, 2014). This deflates nominal prices to real prices, normalized to a base year of 2005/2006. Prices were then converted to US Dollars using the 2005/2006 exchange rate of 1 USD = 1810 UGX. To test for signals of rising land values over time, we used a basic OLS regression of the log- transformed price per acre against transaction year, controlling for ward-specific characteristics like distance to market centers. We used a log-transformation to reduce influence of high-value outliers and we measured time both as a continuous variable and as a categorical vari- able separated into 5-year increments to accommodate a potential non- linear relationship between time and land values during the study period. The 1993 surveys and 2015 surveys were not identical, but shared questions that allowed us to track change in variables like household size and coarse measures of wealth, off-farm employment, and land- holdings in the cross-sections of land users from both periods. To per- form further quantitative analysis, we examined more detailed socioeconomic characteristics of land users in 2015. We used 1993 land- holder maps to assign 2015 land users into categories based on when the decision was made to acquire land near the park. We defined “re- cent buyers” as households who have purchased land on the transects since 1993 (even if they already owned other land on the transects). We defined “current renters” as those renting or using land on the tran- sects who do not own other parcels on the transects (though they may own other parcels elsewhere).10 “Legacy households” are those using land on the transects that they already owned in 1993 or that they have since inherited from their parents (or other relatives) - they haven't bought any additional land. By these definitions, there are 47 re- cent buyers, 23 current renters, and 83 legacy households.11 Our aim with these categories was to separate the people recently choosing to Fig. 1. Map of the Study Region. Location of Kibale National Park in western Uganda and park-edge sites where transects are located along the park's northwestern edge. invest in park edge land from those who made that decision in decades past under different economic circumstances. We compared socioeco- nomic characteristics (defined in Table 1) among these groups to test using land that was part of a parcel that intersected the transects in for significant differences between recent investors and legacy park 1993, whether or not the household was located on that land. This iden- neighbors. Variables of primary interest are total reported landholdings, tification was possible because our research assistants were members of wealth, off-farm employment, and education of head of household as the communities in question and aware of absentee owners and rental these are proxies for investment capital and human capital. To account arrangements, and they regularly encountered people working in their for patterns associated with generational turnover among existing land fields when they traversed farms weekly to monitor crop raiding (sep- users, we include household size and age of household head, which are arate study). The 1993 maps of land users allowed for backtracking often associated with land use decisions through household lifecycle and triangulation when creating the more crowded and complicated processes (De Sherbinin et al., 2008),12 along with a variable based on maps of 2011 and 2015 land users by providing a point of reference household head's place of birth to signal households migrating to the and credible authority when seeking village participation in updating area from other regions. the maps.6 Household surveys of associated land users were conducted In addition to testing whether the socioeconomic factors in Table 1 in each period to document socioeconomic characteristics of land users. differ across recent investors, renters, and legacy households, we use Surveys were conducted by a team of research assistants in respon- fi dents' local language. We surveyed all households possible in the ve 9 7 Of the 370 parcels that respondents reported using, 175 were purchased (vs. sites over a period of 6 months in 1993 and 2 months in 2015. In inherited, rented, or used under other arrangements). We dropped transactions affiliated total, we collected household surveys for 54 of 63 mapped land users with “commercial plots” (small roadside strips in commercial centers, n = 20) to focus in 1993 and 129 of 153 mapped land users in 2015. Non-response bias primarily on agricultural land, as well as transactions with incomplete information on leans toward under-sampling of wealthier absentee owners. price, year, or area (n = 87). Many respondents did not remember at least one of these pa- rameters or did not know because they were not party to the transaction. Of the 68 obser- Land values in the study area were calculated based on information vations with complete data, 95% of reported transactions yielded per-acre values collected from surveys in 2015. Respondents were asked how and when substantially lower than $5000. We excluded four implausible outliers where reported they acquired all land parcels that they used (including parcels not on transactions yielded per-acre values N $20,000. the transects but in the same ward), and in cases where land was pur- 10 We defined rental arrangements as households using land belonging to other house- chased, they were asked to report the year of transaction, the price holds, with no expectation of future ownership rights to that land. Often this does not mean that households actually pay rent - “using freely” on a temporary basis is common, they paid, the estimated area of the parcel, and the land cover on the as is payment in crops rather than cash. 8 parcel at the time they bought it. After excluding reports with 11 We were able to survey 32/47 recent buyers, 22/23 renters, and 75/83 legacy house- holds. The under-sampling of recent buyers may mean that some socio-economic charac- teristics are underestimated for this group, since it was especially difficult to sample 6 For example: “Who is using the land that Katuramu once owned next to Okwilo's wealthier absentee owners. place?” yields better information than “Who owns this land?” 12 In land use prediction models, age and age2 are often used to describe a nonlinear re- 7 Pilot survey work and initial parcel mapping were carried out over 4 months in 2011. lationship in which very young and very old households are at different stages than house- 8 This was almost always reported as fallow. holds in their prime working years. J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111 105
Table 1 Socioeconomic characteristics examined for 2015 land users.
Characteristic How measured
Area of land owned Continuous, transformed using inverse-hyperbolic sine function (similar to a log-transformation except it permits inclusion of households owning no land) to accommodate a skewed distribution in land ownership Wealth indicator A continuous indicator constructed from a principal components analysis of livestock (goats, pigs, cows), car ownership, and housing measures (brick walls, roofing material) - assets that represent common non-land-based investments Off farm employment Binary: used coded version of employment histories to indicate whether a member of the household had off-farm employment at some point during the previous 5 years. High off-farm income Binary: defined as off-farm income of N100$ a month (within the past 5 years) as reported in employment histories. Household size Number of people currently living in the respondent's primary house Household head educ. Max year of schooling reached by household head Household head age Age of head of household Household head local Binary: set to one if head of household was born in the same district
logistic models to explore whether these factors are associated with a arrangements,15 but it is a strong indicator of increasing land competi- household's likelihood of planting the crops that increased most notably tion. We also find evidence for intensifying land competition in increas- on the transects between 1993 and 2015 – eucalyptus, tea, and coffee. ing reported land prices. Fig. 2 shows inflation-adjusted land price The model structure is shown in Eq. (1), where the outcome is a binary calculations from respondents' reported transactions during the past indicator for whether a household (h) grows a given crop on any of three decades. Although there is wide variation due to differences in 13 its landholdings, β1–10 are coefficients associated with the vector land quality, thin markets, and many potential sources of measurement of household-level socioeconomic characteristics (X) described in error, there is a discernible upward trend, particularly noticeable in the
Table 1, β11–12 are coefficients associated with variables indicating decreasing incidence of low-value sales during the last few years. Basic which ward contains a household's landholdings to control for any po- regressions of price against transaction year also point to a significant 16 tentially omitted site-specific geo-historical factors, β0 is a constant, increase of price over time (Table 2). In 2015, residents described and ε is an error term using robust standard errors. the local land-market as ever more expensive and even subject to land speculation, saying, for example: “you can buy land at 5 million shillings ½ = = ð Þ ” Planted eucalyptus tea coffee h 1 today and turn around and sell it tomorrow at 6 million shillings ¼ β þ β þ β þ ε (recent buyer in Kanyawara). 1–10Xh 11–12Ward Indicatorh 0
We report separate models for each crop type to accommodate po- 3.2. Changing land users tential differences in their accessibility and prevalence. It was necessary to exclude renters from these models because no renter planted euca- As the composition of land users on Kibale's edge has changed, lyptus, tea, or coffee. A household-level model allows us to speak households' connections to park-edge land have become more complex. about factors that may promote or inhibit land users' ability to transition By 2015, the traditional situation of a household living and farming on a to these crops. However, it does not directly link socioeconomic explan- single parcel of land had become more the exception than the rule. Ab- atory variables to land cover outcomes on the transects because house- sentee owners represent a relatively small fraction of affiliated house- holds are not weighted by the amount of transect land they control; our holds, but they own a disproportionate amount of transect land. data resolution is not high enough to link social factors to each grid cell Absentee owners have become more common: about 3% of landowners in a grid cell-level analysis. Nevertheless, by showing change in socio- in 1993 did not reside in the same ward as the transect parcel, and in economic characteristics of people using the transects, showing change 2015 this doubled to 7%.17 An additional 20% of affiliated landowners in land cover on the transects, and showing that some socioeconomic in 2015 resided in the same ward as their transect parcels but lived in characteristics predict use of certain land cover types, we triangulate the trading centers farther from the park. In fact, many of the transect to suggest a relationship between changing composition of land users parcels didn't contain any sort of residential compound in 2015, and and changing land cover. this was particularly true for parcels immediately bordering the park. Rental arrangements between unrelated households increased from 3. Results 7% of transect households in 1993 to 16% in 2015. Use of multiple land parcels also became more common; in 1993, 25% of surveyed house- 3.1. Increasing land competition holds reported owning a parcel elsewhere, while in 2015, 40% report owning another parcel and 78% report either owning or using another Between 1993 and 2015, the number of households using land parcel (including rented and family land). intersected by study transects rose from 62 to 153 - an increase of about 250%. This does not translate directly to changes in population 15 One third of 2015 households are the grown children of 1993 respondents who have density due to concurrent changes in household size14 and land holding started their own households on their parents' land. Intra-family land sharing arrange- ments vary. In some families, family land is split among children and they have the right to sell the land or plant more permanent crops. In others, the older generation retains sell- ing and permanent cropping rights. Several older respondents reported that they did not 13 Buyer status is based on households' transect land, but the land cover outcome vari- want to “show boundaries” to their children because then the children might sell their able is based on whether households reported planting a given crop type in their house- parcel for quick cash and then move back on to the family land when they got into finan- hold survey and thus may sometimes include plantings on landholdings not on the cial trouble. Some even talked of forbidding that family land be split or sold after their transects but within the same wards. Because transects were originally laid out to measure death, on penalty of haunting children who defied their wishes. crop-raiding, they do not align perfectly with parcel boundaries. 16 Many factors affect the price of land, but due to data limitations it was only possible to 14 Average household size has decreased from mean of 6.8 to 4.6 members (p b 0.001 in include a control for the ward where the transaction occurred. Land cover at the time of an unpaired t-test with unequal variances). Some of the decrease may be related to house- purchase was almost always reported as fallow. Trends are robust to the exclusion of hold lifecycles. Many households settled the land near the park edge in waves in the 1970s the high-value sale in 2014. as young adult migrants seeking land and employment (Hartter et al., 2015). As children of 17 Most of these absentee owners lived in the regional or national capital cities, but a few these households begin to start their own families, household sizes decline. lived abroad. 106 J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111
Fig. 2. Reported land values around Kibale National Park, 1985–2015. Prices are divided by reported acreages, adjusted with an IMF consumer price index using a base year of 2005–2006 to account for inflation, then converted to USD based on the 2005–2006 exchange rate of 1 USD = 1810 UGX.
In the cross-section, the group of land users at the park edge has be- near the park in 1993 but these fields were decimated in the intervening come wealthier. In the 1993 household survey, having a metal roof in- years by a banana weevil. Natural forest that extended beyond the park stead of a grass roof was an indicator of wealth. Since then, the boundary in 1993 has also since been cleared. The park-edge land once proportion of households living under grass roofs dropped from 54% under brewing bananas or natural forest is now planted with perennial to 2.5%, while the proportion hiring others to work for them at least oc- investment crops: eucalyptus, tea, and coffee. Eucalyptus was present in casionally rose from 5% to 30%.18 Wealthier households in 2015 owned 1993 only as sporadic trees marking parcel boundaries, but by 2015 it brick buildings, motorcycles, and cars - assets that were rare or nonex- covered N20% of sampled park-edge land (Fig. 4). Tea was also istent in 1993. Yet, some of the changes between 1993 households and completely absent on the smallholder land in the study sites in 1993 2015 households are associated with turnover in land users rather than but now accounts for 10% of the transect land and has nearly doubled just changing circumstances for existing land users and their families. in the past five years alone.20 Coffee has also expanded on land neigh- The majority (67%) of households in 1993 have sold at least some of boring the park, and there are two private coffee plantations on park their land near the park,19 though they have often stayed in situ on edge land beyond our study transects. The prominence of different smaller plots. Meanwhile, those who have bought land near the park types of inedible perennial crops varied by ward, but in combination, since 1993 differ socioeconomically from legacy landowners (see these alternative crops increased on land near the park in all sites, Table 3). The direction of these differences flips, depending if new from essentially 0% to between 15% and 50% (Table S1). Changes were users have purchased or are renting the land near the park. Recent increasingly pronounced nearest the park boundary (Fig. 3). buyers are wealthier and own more land compared to legacy house- holds, while the inverse is true of renters. Recent buyers also have 3.4. Relationship between land cover and land users more years of education, are more likely to have had recent off-farm in- come (during the past 5 years), and much more likely to have had high What explains the rise in eucalyptus, tea, and coffee near the park off-farm income (over 100 USD per month) than legacy households. and who is planting these crops? At least half of 2015 respondents Though recent buyers are not all newcomers to the region (25% are chil- said they chose to plant one of these crops because they are not eaten dren of legacy owners and 60% had a head of household born within the by wildlife. People also talked about these crop options as though they sub-county), systematic differences between the types of households were interest bearing accounts. In the case of eucalyptus, many people buying land today and those who acquired it earlier indicate that appreciated the fact that it had several markets and flexibility in the constraints and agendas associated with park edge land use may be timing of sales: “I can use [eucalyptus] after 3 years for firewood at changing. home or sell to local traders, I can sell it at age 5 to brick makers or for poles for construction, between 5 and 10 years I can sell it to UMEME (Uganda government body in charge of rural electrification that uses 3.3. Changing land cover eucalyptus for utility poles), and after that I can sell for timber” (female legacy owner, Kabucikire). People also liked eucalyptus because it yields Land cover change was marked by an increase in eucalyptus, tea, large cash sums at once, and several landholders said they use eucalyp- and coffee, and a sharp decrease in brewing bananas and natural forest tus to pay school term fees or keep it as a buffer in case of a health crisis. (Fig. 3, Table S1). Bananas for brewing beer carpeted much of the land Tea has a different investment profile: it requires the most money to es- tablish and maintain, but it pays more regularly once established: “I 18 Livestock ownership was relatively constant, though the average number of goats de- chose tea because of animals, and I wanted to invest in tea because creased somewhat and was offset by an increase in pig ownership. when I stop working [at my off-farm job], tea will get me school fees. 19 The timing of sales and circumstances around selling were highly variable and the sample size of 1993 owners is small for testing the predictors associated with sales. There is no clear evidence that households that were poorer in 1993 were more likely to sell – 20 Tea area was also canvassed on the transects in 2011 - at that point it occurred in 5% of both poorer and wealthier households have sold. cells. J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111 107
Table 2 Changing land values from transactions reported by respondents since 1985.
OLS regression on ln(price per acre) Model 1 Model 2 Model 3 Full sample Since 2010 Categorical
Year [continuous] 0.03* 0.25* Year [category] 1985–1990 (Base) 0.00 1990–1994 0.49 1995–1999 0.68 2000–2004 0.59 2005–2009 0.70 2010–2015 0.86* Site Site Kanyawara Ward (Base) 0.00 0.00 Kanyawara Ward 0.00 Nyabubale Ward −0.15 −0.62** Nyabubale Ward −0.13 Rurama Ward −0.29 −0.17 Rurama Ward −0.26 Num obs 64 23 64
Price is log-transformed, so coefficients can be interpreted as a percent change in price, i.e. price per acre increases 3% per year over the full time range, but 25% per year since 2010. For the categorical version of year, the interpretation is the percent change compared to the reference category of “1985–1990”. Models use robust standard errors. ⁎ p b 0.10. ⁎⁎ p b 0.05.
Tea is a good crop to plant when you have money because you on the land he still shares with his family and on land that he rents a safe use money planting and when it's harvesting time, you don't use distance away from the park, and has planted eucalyptus and tea on the much money you just sit and eat your cash” (male recent buyer, parcels he purchased near the park. Another recent buyer moved to the Kanyasohera). area to work at the park's biological research station. He owns land im- Recent buyers and legacy households were similarly likely to plant mediately neighboring the park in several locations, which he prefers eucalyptus (40% vs. 44%), but recent buyers were more likely to plant because it is cheaper. Some of the land was purchased from earlier in- tea (25% vs. 13%) and coffee (26% vs. 16%) than legacy owners. None vestors, and more land was acquired bit by bit from families needing of the households renting-in transect land near the park had planted cash to pay for school fees, health care emergencies, and legal fines. tea, coffee, or eucalyptus on any of their holdings (including land they His park edge land is planted in tea and eucalyptus. Meanwhile, house- own elsewhere), corroborating the interpretation that renting-in park holds who no longer have enough land for crops (through splitting for edge land is a sign of economic duress, while planting these crops is a inheritance or cashing in) are now renting from others, often in clusters signal of relative economic prosperity. Landlords sometimes plant from the large landholders. Near one of our sites, there is a 25 ha parcel eucalyptus, coffee, or tea on land they simultaneously rent out; renters of land where at least 48 households rent plots for food crops. It is provided free labor clearing land and keeping weeds down while plan- owned by a retired government employee who plans to evict the tations seedlings are getting established, but once the landlord's seed- renters once his eucalyptus plantation is established, but collects rent lings are large enough to shade out food crops, renters have to find in the meantime from people farming between the young trees. In the another site to use. words of one wealthy respondent who had also recently bought land Those who did not plant any investment crops blamed land or cash near the park and planted eucalyptus: “the area is booming, but I feel shortages. Our quantitative survey data corroborates this: households bad for the people without cash – they are falling behind.” with more land, wealth, and education were more likely to have planted one or more of these crops (Table 4). Landholdings, wealth, and educa- 4. Discussion tion are the same factors that are significantly higher for recent buyers (Table 2). Controlling for these factors, being a recent buyer does not af- The changes at the boundaries of Kibale National Park illuminate po- fect the odds that a household plants tea or coffee, and it is significantly tential pathways transforming social and ecological context around negatively associated with odds of having eucalyptus. This suggests that protected forests- from a more familiar case of poor people abutting the shifts in households using park edge land are associated with shifts rich forests, to an area of increasing investment. We document a transi- in land cover primarily through shifts in land user capital. The unit of ob- tion toward wealthier park neighbors planting perennial investment servation for the models in Table 4 is the household – a model using an crops while subsistence agriculture persists on increasingly smaller par- area-based unit of analysis would likely show a much stronger link be- cels and on rented land. This shift has implications for landscape con- tweenshiftsinlandcoverandshiftsinwealthbecauseindividualswith nectivity around parks, socio-political relations between protected large landholdings would be weighted more heavily. Less representa- area managers and neighbors, and the fate of poorer households living tion in the survey from absentee owners also will lead to a tendency near protected forests. to underestimate the relationship of factors like land, wealth, education, and high off-farm earnings with eucalyptus, tea, and coffee. Neverthe- 4.1. Increasing land competition and changing land users less, for a given household, the models in Table 4 suggest that land and other assets condition who can access more profitable land use Land use dynamics around Kibale are conditioned by active land opportunities. markets, proximity to expanding infrastructure and drivers of economic Increasing land competition and new (differentiated) opportunities development (e.g. tea plantations, park facilities, and the town of Fort for park edge land has had mixed implications for park neighbors. One Portal), and presence of crop-raiding wildlife (Mackenzie, 2012). long-term resident spoke of plans to move to the opposite side of the Some of these factors are exogenous to the park and some are endoge- park where land was reportedly still cheap. She worried that her nous: for example, Uganda's rural electrification initiative and relatively present inherited parcel was too small to bequeath to her several teen- high rates of economic and population growth would occur even with- age children and prices too high to buy more local land. In contrast, her out the existence of the protected forest, while factors like crop loss to 30-year-old nephew who inherited a small parcel nearby has purchased wildlife and economic opportunities from conservation projects and five additional pieces of land from his relatives and neighbors using in- tourism are conditions that the park creates around itself. The combina- come from his job with the Ugandan government. He grows food crops tion of landscape-level forces and park-level forces interact to 108 J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111
Table 3 Comparing incoming households vs. legacy owners (including heirs).
Recent buyers Current renters Legacy HHs Overall 2−Way differencea
Buyers vs. legacy Renters vs. legacy
Average acres owned 4.7 0.9 2.7 2.9 2.0** −1.7*** Wealth score 0.7 −0.4 −0.2 0.0 0.9*** −0.2 Any off farm emp. past 5 years 75% 50% 49% 56% 26%** 1% High earner (from off−farm inc.) 53% 5% 12% 21% 41%*** −7% Household size 4.8 4.6 4.4 4.6 0.4 0.2 Household head education 6.7 4.0 5.3 5.4 1.4* −1.3 Household head age 39.7 40.1 43.9 42.2 −4.1 −3.8 Household head born locally 63% 73% 69% 68% 6% 4% Number of observations 32 22 75 129
a Significant differences for two-way means comparisons were calculated using Fisher's Exact test in the case of binary variables and unpaired t-tests for continuous variables. Proba- bilities that means are equal are indicated with the following symbols: *: p b 0.1, **: p b 0.05, ***: p b 0.01. determine pressures on the forest edge at a given point in time. During There is a polarized debate about whether land markets in Africa will the study period, for this particular site, market contexts have apparent- help alleviate poverty through market efficiency or further impoverish ly changed enough to flip the risk-zone around the park from marginal- those who are priced out (Colin and Woodhouse, 2010). In some African ity to an attractive investment zone, overwhelming park-level factors countries, land sales are banned because of concern that they will lead that have kept the land cheaper. Though this result is dependent on to conflict but in others they are promoted as a non-violent mechanism the particular combinations of factors at play in this site, mounting for achieving an optimal land distribution (Berry, 2009; Deininger and land competition and emerging land markets are familiar phenomena Mpuga, 2009). A study in eastern Uganda concludes that land markets in other sites in East Africa and elsewhere in the developing world can operate as an equalizing force that counteracts inequality in land in- (Jayne et al., 2003; Lambin and Meyfroidt, 2011; Holland et al., 2016). heritance practices when farms become too small to be split equally Other research points to growing land scarcity around Kibale among the next generation. However, the authors caution that if land (Hartter et al., 2015), and our study adds that land pressure has in- is used primarily as a savings mechanism, this can lead to increasing in- creased not only as a function of population growth and migration of equality in land distribution (Baland et al., 2007). If land around the poor, but also as a result of investment and speculation by wealthier protected forests has traditionally been relatively cheap, this concern buyers. Several respondents indicated they aim to accumulate land even about increasingly unequal land distribution may be particularly valid, if they are not yet profiting because they know land prices will rise. Em- especially if there is a dichotomy between those attracted to or pirical information about land transactions and land values is hard to persisting in the area because they can't afford land elsewhere, and come by, particularly in East Africa (Holden et al., 2009). Though we those drawn to high investment potential given cheap baselines. do not have enough data to control for many factors that affect land Scholars of land change in Sub-Saharan Africa have called for more price and our information based on recall is imprecise, we nevertheless long term studies following dynamics over generations (Guyer et al., see inflation-adjusted land values at Kibale's edge increase over time, 2007). A generation ago, the areas bordering Kibale's protected forest especially in recent years, and qualitative interview data corroborate received the region's more marginalized and land-poor households this trend. Landholdings are seen as a promising place to invest wages, (Naughton-Treves, 1997; Hartter et al., 2015). In recent years, the peo- especially since banks and financial markets are not easily accessible ple buying park edge land in this same area are primarily wealthier in- and considered risky. dividuals with money to invest. Most recent buyers enjoy relatively
Fig. 3. Changes in Land Use at the Park Edge. Coverage based on crop-mapping in 1993 and 2015 along transects in 5 hamlets in 3 wards: ~100 total ha of smallholder land within 0.5 km of the park boundary. Values represent the difference in the percent of transect grid cells in which that crop type occurred in 2015 compared to 1993. J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111 109
Fig. 4. Expanding eucalyptus. F. Mugurusi stands on a hillside property in 1993 and 2011 - the natural forest fragment and brewing bananas in the background have been replaced with eucalyptus. well-paying off-farm employment, for example, as teachers, clinicians, amount of land under annual crops (Southworth et al., 2010). It also ac- government workers, tea estate managers, or employees of the park it- cords with aerial photo analyses showing an increase in tea and euca- self. Some are sons and daughters of long-term owners, others live else- lyptus at the expense of natural forest (Majaliwa et al., 2010). Our where and see park-edge land as an investment opportunity. field-based methods allow for more accurate identification of eucalyp- Meanwhile, poorer farming households and lower wage earners (e.g. tus (vs. native trees) and identification of young perennial seedlings tea pluckers) can no longer afford to buy land and increasingly turn to mixed in intercropped fields, thus we are able to detected more change renting. For renters, the park edge continues to represent a place of mar- and earlier stages of change than that reported from remote sensing ginality, where land is cheaper but planting annual food crops is a risky studiesatthislocalscale. proposition where wildlife, including elephants, abound (Naughton- During the past three decades, natural forest has been largely elimi- Treves, 1997). nated outside Kibale Park with severe consequences for the region's pri- Though protected areas have been criticized for estranging local mate populations, particularly endangered species (Chapman et al., people from land and resources that they depend on (Adams and 2005a, 2005b). As forests and wetlands beyond the park boundaries Hutton, 2007), the process we are describing is one in which the park are disappearing, it is noteworthy that the land cover types that are in- has historically played the arguably pro-poor role of slowing land com- creasing fastest are not the annual subsistence food crops that one petition, albeit at the cost of increased risk of crop damage. This role is might expect from a growing agrarian population. This is due in part weakening in the face of changing profitability of alternative land use to the park's effect on the expected risks and returns for food crops, options. This shift is resulting in loss of access for some and economic but it is also related to increasing profitability of non-food crops. With benefits for others. Alternative land use options are differentiated with rising demand for fuelwood and other wood products at local and na- respect to accessibility for existing land users. We show that legacy tional scales, there are indications that eucalyptus is becoming relatively households are as likely as recent buyers to plant eucalyptus, and actu- lucrative for those with capacity to wait for returns on invested capital ally more likely when controlling for socio-economic characteristics (Heuler, 2012; Wasake, 2012). Eucalyptus is expanding along other correlated with recent buyer status. Eucalyptus requires less initial cap- Ugandan forest edges, stirring concern that its demands on water and ital than export shrubs like tea and coffee but still generates good soil nutrients can inhibit cultivation of food crops (Watkins, 2009). At returns (Jagger and Pender, 2003), making it possible that eucalyptus Kibale thus far, except where edge land was rented out, most inedible could be playing a role in increasing wealth or maintaining land crops have replaced fallow lands, forest fragments, and brewing banan- among legacy households. If existing land users are riding the wave of as rather than land that was used for annual food crops. In the case of economic development and becoming wealthier themselves, then brewing bananas, a major component of the 1993 landscape around there is at least the possibility that some pressure is taken off natural the park, there was no choice but to find a new crop due to a severe out- forests. However, if new, wealthier land users are competing with break of banana weevil (Kangire et al., 1996; Speijer et al., 1994). existing land users, there is the danger of adding to and displacing Farmers' ability to shift from brewing bananas to alternative crops has land pressure. Though we do not show evidence of significant displace- been shaped by socioeconomic factors - inedible perennial crops all re- ment as yet, we document heightened vulnerability for those whose quire longer lead-times before they yield profit and require more up- properties are getting smaller with no room for the next generation. front costs to establish. This accords with other research showing that assets of landowners strongly influence response to shocks and new op- 4.2. Expanding investment crops and changing land cover portunities, and wealthier land users can often access options out of reach to the poor (e.g. Takasaki et al., 2001; Ebanyat et al., 2010; L'Roe Near the park's edge, we documented a notable increase in inedible and Naughton-Treves, 2014). perennial crops, primarily at the expense of brewing bananas, fallow A transition toward perennial investment crops around parks has land, and natural forest fragments. The area devoted to food crops has mixed implications for biodiversity conservation. Planting unpalatable remained relatively constant. This is consistent with remote-sensing monoculture crops (like tea) reduces habitat connectivity if they replace based reports of land cover change around the park that note a reduc- a matrix of inter-cropped rotational fallow farming and forest frag- tion in forest fragments, an increase in tea, and no clear trend in the ments. However, lowering connectivity also reduces crop-raiding by 110 J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111
Table 4 Predictors of planting inedible perennial crops - (logistic regressions among 2015 land owners).a
Planted eucalyptus Planted tea Planted coffee
Basic Full Basic Full Basic Full
New buyer (reference is legacy HH) 0.88 0.12*** 2.87** 1.46 1.40 1.04 Acres owned (continuous, transformed) 3.38*** 9.64*** 1.30 Wealth index (livestock & housing based) 2.04** 1.25 1.16 Any off-farm employment in past 5 years 0.58 0.72 0.38 High earner (N100 USD per month) 3.00 0.21 4.37 Household size (# people) 0.87 2.02*** 1.05 Household head education (years) 1.12 1.24* 1.18* Household head age 1.19 1.16 1.09 Household head age2 0.998 0.998 0.999 Household head born in district (yes/no) 0.61 5.68 2.38 Nyabubale Ward (Kanyawara is reference) 0.89 0.14 1.99 0.73 7.78*** 11.87** Rurama Ward (Kanyawara is reference) 0.53 0.65 0.16** 0.02** 6.65*** 20.67*** Num obs 130 107 130 107 129 106 Model psuedo-r2 0.02 0.38 0.11 0.54 0.12 0.27
a Models are based on Eq. (1) and use robust standard errors. Models are only for land owners - renters are excluded because they predict failure in the outcome variable(s) perfectly. Cells contain coefficients expressed as odds ratios - for example, the coefficient of 0.88 in the top left cell can be interpreted as “holding the other factors constant, recent buyers are 0.88 times as likely to plant eucalyptus compared to legacy households.” Significance of coefficients are indicated with the following symbols: *: p b 0.1, **: p b 0.05, ***: p b 0.01. wildlife, a conflict serious enough to lead to de-gazettement of parks Kimambo, N. Etchart, L. Rausch, J. Alix-Garcia, M. Turner, and two anon- elsewhere in East Africa (Pienkowski et al., 1998). The replacement of ymous reviewers for helpful feed-back and suggestions on earlier ver- brewing bananas, a prime raiding target for non-human primates sions of this manuscript. We acknowledge the Scott Kloeck-Jenson (Naughton-Treves, 1998), with inedible crops may also decrease the Foundation and the UW-Nelson Institute's Center for Humans and the likelihood of disease transmission between human and non-human pri- Global Environment IGERT Program for their financial support for this mates (Chapman et al., 2005a, 2005b). Impacts also vary by crop or project. plantation type. Unlike tea, eucalyptus woodlots may serve as viable corridors for some species (Kimambo, 2016; DeFries et al., 2007). References Wood from eucalyptus may also relieve pressure on natural forests by providing a substitute source for wood products (Ainembabazi and Adams, W.M., Hutton, J., 2007. People, parks and poverty: political ecology and biodiver- Angelsen, 2014). Tea and eucalyptus plantations are biodiversity waste- sity conservation. Conserv. Soc. 5, 147. Ainembabazi, J.H., Angelsen, A., 2014. Do commercial forest plantations reduce pressure lands compared to natural forests, but compared to annual agriculture on natural forests? Evidence from forest policy reforms in Uganda. For. Policy Econ. these perennial crops slow soil erosion and store carbon (Rudel et al., 40, 48–56. 2005). Given that some organizations have invested great sums in cre- Alix-Garcia, J., McIntosh, C., Sims, K.R., Welch, J.R., 2013. The ecological footprint of pover- ty alleviation: evidence from Mexico's Oportunidades program. Rev. Econ. Stat. 95, ating tea buffers around other East African highland parks (Masozera 417–435. and Alavalapati, 2004), it is interesting that at Kibale, the transition Angelsen, A., Wunder, S., 2003. Exploring the Forest–Poverty Link: Key Concepts, Issues seems to be happening without formal intervention. and Research Implications. Centre for International Forestry Research (CIFOR), Bogor, Indonesia. Baland, J.M., Gaspart, F., Platteau, J.P., Place, F., 2007. The distributive impact of land mar- 5. Conclusion kets in Uganda. Econ. Dev. Cult. Chang. 55, 283–311. Barbier, E.B., 1997. The economic determinants of land degradation in developing coun- As increasing land competition puts greater pressure on formerly tries. Philosophical Transactions of the Royal Society of London B: Biological Sciences 352 (1356), 891–899. low-value land, the context changes for forest conservation and for Berry, S., 2009. Property, authority and citizenship: land claims, politics and the dynamics the people who have traditionally lived near forests. Expanding infra- of social division in West Africa. Dev. Chang. 40, 23–45. structure and opportunities for off-farm income, land markets, and ris- Byron, N., Arnold, M., 1999. What futures for the people of the tropical forests? World Dev. 27, 789–805. ing values for agricultural commodities (including plantation trees) all Chapman, C.A., Gillespie, T.R., Goldberg, T.L., 2005a. Primates and the ecology of their in- may facilitate this transition. The land neighboring Kibale National fectious diseases: how will anthropogenic change affect host-parasite interactions? Park is changing in character, from a place of marginalization to a Evol. Anthropol. 14, 134–144. Chapman, C.A., Struhsaker, T.T., Lambert, J.E., 2005b. Thirty years of research in Kibale Na- place of investment opportunity. As such, it is becoming less of a refuge tional Park, Uganda, reveals a complex picture for conservation. Int. J. Primatol. 26, for poorer households, but is also providing more opportunities for 539–555. households with means to ‘ride the wave’ to alternative types of produc- Chimhowu, A., Woodhouse, P., 2006. Customary vs Private Property Rights? Dynamics and Trajectories of Vernacular Land Markets in Sub‐Saharan Africa. J. Agrar. Chang. tion. Where forests with high conservation value occur in regions 6, 346–371. experiencing rapid development, we should consider the possible im- Colin, J.-P., Woodhouse, P., 2010. Interpreting land markets in Africa. Africa 80, 1–13. plications for forest management and pro-poor land use planning if for- De Sherbinin, A., VanWey, L.K., McSweeney, K., Aggarwal, R., Barbieri, A., Henry, S., Hunter, L.M., Twine, W., Walker, R., 2008. Rural household demographics, livelihoods est edges become sites attracting investment. It is important to attend and the environment. Glob. Environ. Chang. 18, 38–53. both to broader socioeconomic processes as well as endogenous effects DeFries, R., Hansen, A., Newton, A.C., Hansen, M.C., 2005. Increasing isolation of protected of parks in order to understand land and poverty dynamics around areas in tropical forests over the past twenty years. Ecol. Appl. 15, 19–26. protected forests and predict future trajectories. DeFries, R., Hansen, A., Turner, B., Reid, R., Liu, J., 2007. Land use change around protected areas: management to balance human needs and ecological function. Ecol. Appl. 17, Supplementary data to this article can be found online at http://dx. 1031–1038. doi.org/10.1016/j.forpol.2016.12.011. Deininger, K., Mpuga, P., 2002. Land Markets in Uganda: Incidence, Impact, and Evolution over Time. World Bank Discussion Paper. Washington DC, World Bank. Deininger, K., Mpuga, P., 2009. Land markets in Uganda: what is their impact and who Acknowledgements benefits? In: Holden, S.T., Otsuka, K., Place, F.M. (Eds.), The Emergence of Land Mar- kets in Africa: “Impacts on Poverty, Equity, and Efficiency”. Resources for the Future, The authors would like to thank R. Karamagi, M. Nyakoojo, D. Washington, DC, pp. 131–154 Ebanyat, P., de Ridder, N., De Jager, A., Delve, R.J., Bekunda, M.A., Giller, K.E., 2010. Drivers Kuzaara, J. Ogwang, D. Bagambisa, F. Mugurusi, P. Baguma, and A. of land use change and household determinants of sustainability in smallholder farm- L'Roe for invaluable assistance in the field. We are grateful to N. ing systems of Eastern Uganda. Popul. Environ. 31, 474–506. J. L'Roe, L. Naughton-Treves / Forest Policy and Economics 84 (2017) 102–111 111
Foster, A.D., Rosenzweig, M.R., 2003. Economic growth and the rise of forests. Quart. Naughton-Treves, L., 1998. Predicting patterns of crop damage by wildlife around Kibale J. Econ. 601–637. National Park, Uganda. Conserv. Biol. 12, 156–168. Guyer, J.I., Lambin, E.F., Cliggett, L., Walker, P., Amanor, K., Bassett, T., Colson, E., Hay, R., Naughton-Treves, L., Holland, M.B., Brandon, K., 2005. The role of protected areas in con- Homewood, K., Linares, O., 2007. Temporal heterogeneity in the study of African serving biodiversity and sustaining local livelihoods. Annu. Rev. Environ. Resour. 30, land use. Hum. Ecol. 35, 3–17. 219–252. Hall, J., Burgess, N.D., Lovett, J., Mbilinyi, B., Gereau, R.E., 2009. Conservation implications Naughton-Treves, L., Kammen, D.M., Chapman, C., 2007. Burning biodiversity: woody bio- of deforestation across an elevational gradient in the Eastern Arc Mountains, Tanza- mass use by commercial and subsistence groups in western Uganda's forests. Biol. nia. Biol. Conserv. 142, 2510–2521. Conserv. 134, 232–241. Hartter, J., 2010. Resource use and ecosystem services in a forest park landscape. Soc. Nat. Naughton-Treves, L., Alix-Garcia, J., Chapman, C.A., 2011. Lessons about parks and poverty Resour. 23, 207–223. from a decade of forest loss and economic growth around Kibale National Park, Ugan- Hartter, J., Ryan, S.J., MacKenzie, C.A., Goldman, A., Dowhaniuk, N., Palace, M., Diem, J.E., da. Proc. Natl. Acad. Sci. 108, 13919–13924. Chapman, C.A., 2015. Now there is no land: a story of ethnic migration in a protected Oluka, B.H., 2014. The last best place no more: massive deforestation destroying prime area landscape in western Uganda. Popul. Environ. 36, 452–479. chimp habitat in Uganda. Mongabay.Available at:. http://news.mongabay.com/ Heuler, H., 2012. Uganda ‘tree banks’ offer investors good returns, Market 2014/07/the-last-best-place-no-more-massive-deforestation-destroying-prime- Watch.Available at:. www.marketwatch.com/story/uganda-tree-banks-offer- chimp-habitat-in-uganda/ Accessed 9/15/2016. investors-good-returns-2012-05-02 Accessed 10 March 2016. Peluso, N.L., 1992. Rich Forests, Poor People: Resource Control and Resistance in Java. Univ Holden, S.T., Otsuka, K., Place, F., 2009. Understanding land markets: questions and hy- of California Press. pothesis. In: Holden, S.T., Otsuka, K., Place, F. (Eds.), The Emergence of Land Markets Pienkowski, M., Watkinson, A., Kerby, G., Naughton-Treves, L., Treves, A., Chapman, C., in Africa: “Impacts on Poverty, Equity, and Efficiency”. Resources for the Future, Wrangham, R., 1998. Temporal patterns of crop-raiding by primates: linking food Washington, DC, pp. 18–56. availability in croplands and adjacent forest. J. Appl. Ecol. 35, 596–606. Holland, T.G., Coomes, O.T., Robinson, B.E., 2016. Evolving frontier land markets and the Plumptre, A.J., 2002. Extent and status of the forests in the Ugandan Albertine Rift. Wild- opportunity cost of sparing forests in western Amazonia. Land Use Policy 58, life Conservation Society. Available at:. http://programs.wcs.org/portals/49/media/ 456–471. file/albertineriftforeststatusextent.pdf. I.M.F., 2014. World Economic Outlook Database, April 2014. International Monetary Fund. Rudel, T.K., Coomes, O.T., Moran, E., Achard, F., Angelsen, A., Xu, J., Lambin, E., 2005. Forest Jagger, P., Pender, J., 2003. The role of trees for sustainable management of less-favored transitions: towards a global understanding of land use change. Glob. Environ. Chang. lands: the case of eucalyptus in Ethiopia. For. Policy Econ. 5, 83–95. 15, 23–31. Jayne, T.S., Yamano, T., Weber, M.T., Tschirley, D., Benfica, R., Chapoto, A., Zulu, B., 2003. Shackleton, C.M., Shackleton, S.E., Cousins, B., 2001. The role of land-based strategies in Smallholder income and land distribution in Africa: implications for poverty reduc- rural livelihoods: the contribution of arable production, animal husbandry and natu- tion strategies. Food Policy 28, 253–275. ral resource harvesting in communal areas in South Africa. Dev. South. Afr. 18 (2001), Jayne, T.S., Chapoto, A., Sitko, N., Nkonde, C., Muyanga, M., Chamberlin, J., 2014. Is the 581–604. scramble for land in Africa foreclosing a smallholder agricultural expansion strategy? Sims, K.R., 2010. Conservation and development: evidence from Thai protected areas. J. Int. Aff. 67, 35. J. Environ. Econ. Manag. 60, 94–114. Joppa, L.N., Loarie, S.R., Pimm, S.L., 2008. On the protection of “protected areas”. Proc. Natl. Southworth, J., Hartter, J., Binford, M.W., Goldman, A., Chapman, C.A., Chapman, L.J., Acad. Sci. 105 (18), 6673–6678. Omeja, P., Binford, E., 2010. Parks, people and pixels: evaluating landscape effects of Kangire, A., Karamura, E.B., Gold, C., Rutherford, M.A., 1996. Fusarium wilt of banana in an East African national park on its surroundings. Trop. Conserv. Sci. 3, 122–142. Uganda, with special emphasis on wilt-like symptoms observed on East African high- Speijer, P.R., Gold, C.S., Kashaija, I.N., Karamura, E.B., 1994. Assessment of nematode dam- land cooking cultivars (Musa spp., AAA). In: al., K.C.e. (Ed.), I International Conference age in East African highland banana systems, Banana nematodes and weevil borers in on Banana and Plantain for Africa. Acta Horticulturae, pp. 343–353. Asia and the pacific. Proceedings of the Workshop Serdang, Selangor, Malaysia, Kimambo, N.E., 2016. More People, More Trees: Social and Ecological Factors for Tree pp. 18–22. Cover Distribution and Their Implications for Forest Connectivity in Southern Tanza- Takasaki, Y., Barham, B.L., Coomes, O.T., 2001. Amazonian peasants, rain forest use, and in- nia. (MSc. Thesis). UW Madison. come generation: the role of wealth and geographical factors. Soc. Nat. Resour. 14, Lambin, E.F., Meyfroidt, P., 2011. Global land use change, economic globalization, and the 291–308. looming land scarcity. Proc. Natl. Acad. Sci. 108, 3465–3472. Twongyirwe, R., Bithell, M., Richards, K.S., Rees, W., 2015. Three decades of forest cover L'Roe, J., Naughton-Treves, L., 2014. Effects of a policy-induced income shock on forest-de- change in Uganda's Northern Albertine Rift Landscape. Land Use Policy 49, 236–251. pendent households in the Peruvian Amazon. Ecol. Econ. 97, 1–9. Vedeld, P., Jumane, A., Wapalila, G., Songorwa, A., 2012. Protected areas, poverty and con- Mackenzie, C.A., 2012. Accruing benefit or loss from a protected area: location matters. flicts: a livelihood case study of Mikumi National Park, Tanzania. For. Policy Econ. 21, Ecol. Econ. 76, 119–129. 20–31. Majaliwa, J., Twongyirwe, R., Nyenje, R., Oluka, M., Ongom, B., Sirike, J., Mfitumukiza, D., Wasake, D.E., 2012. Investing in forestry in Uganda? Who said money can't grow on Azanga, E., Natumanya, R., Mwerera, R., Barasa, B., 2010. The effect of land cover trees?, Inachee.Available at:. http://inachee.com/investing-in-forestry-in-uganda- change on soil properties around Kibale National Park in South Western Uganda. who-said-money-cant-grow-on-trees/ Accessed 10 March 2016. Appl. Env. Soil Sci. 2010, 1–7. Watkins, C.A., 2009. Natural resource use strategies in a forest-adjacent Ugandan village. Masozera, M.K., Alavalapati, J.R., 2004. Forest dependency and its implications for Hum. Ecol. 37 (6), 723–731. protected areas management: a case study from the Nyungwe Forest Reserve, Rwan- Wittemyer, G., Elsen, P., Bean, W.T., Burton, A.C.O., Brashares, J.S., 2008. Accelerated da. Scand. J. For. Res. 19, 85–92. human population growth at protected area edges. Science 321, 123–126. Mulley, B.G., Unruh, J.D., 2004. The role of off-farm employment in tropical forest conser- Zommers, Z., MacDonald, D.W., 2012. Protected areas as frontiers for human migration. vation: labor, migration, and smallholder attitudes toward land in western Uganda. Conserv. Biol. 26, 547–556. J. Environ. Manag. 71, 193–205. Naughton-Treves, L., 1997. Farming the forest edge: vulnerable places and people around Kibale National Park, Uganda. Geogr. Rev. 87, 27–46. Forest Policy and Economics 84 (2017) 112–119
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
Trees, soils, and warthogs – Distribution of services and disservices from ☆ MARK reforestation areas in southern Ethiopia
⁎ Anja Byga, , Paula Novoa, Mengistu Dinatob, Awdenegest Mogesb, Tewodros Teferab, Bedru Balanac, Teshale Woldeamanuelb, Helaina Blackd a Social, Economic and Geographical Sciences, The James Hutton Institute, Craigiebuckler, AB15 8QH Aberdeen, UK b Hawassa University, P.O. Box 5, Hawassa, Sidama, Ethiopia c International Water Management Institute, PMB CT 112; Cantonments, Accra, Ghana d Ecological Sciences, The James Hutton Institute, Invergowrie, Dundee DD2 5DA, UK
ARTICLE INFO ABSTRACT
Keywords: Conservation projects have often been criticised for creating global benefits while causing negative impacts on Reforestation local livelihoods. Ecosystem services approaches have been seen as one way to change this by focussing explicitly Ecosystem services on maintaining ecosystems for human well-being of stakeholders at various scales. However, ecosystem services ff Trade-o s approaches have often ignored trade-offs between groups of people and issues of power and do not automatically Conservation lead to better outcomes in terms of human well-being. Here we report on a study on the impacts of reforestation projects with an explicit focus on human well-being in three communities in southern Ethiopia. We investigated the distribution of services and disservices from reforestation using qualitative methods. Results showed that the services and disservices from reforestation were distributed unequally across space and wealth groups resulting in widespread dissatisfaction with existing reforestation projects despite the explicit focus on human benefits. To improve outcomes of reforestation it is necessary to acknowledge and manage disservices adaptively, include issues of power and make trade-offs transparent.
1. Introduction usefulness to humans. With ecosystem services approaches came a shift in the arguments from conservation for its own sake to conservation for During the last few decades the concept of ecosystem services has the sake of all the goods and services that nature provides to humans risen to prominence not just within academia but also as a new ap- (Mace, 2014), that is, conservation of nature based on utilitarian ar- proach to environmental conservation, restoration, and development. guments. In contrast to pure conservation approaches, ecosystem ser- The rise of ecosystem services approaches can in part be seen as an vices approaches explicitly focus on the links between ecosystems and attempt to reconcile conservation goals and human well-being. human well-being. We here define well-being as a multi-dimensional Traditional conservation approaches such as reforestation in enclosures concept encompassing not only material wealth, but also aspects such have often been criticised for imposing substantial costs on local people as autonomy and freedom to act, physical and mental health, relations and their livelihoods while the benefits such as biodiversity conserva- with others and security (Agarwala et al., 2014). Focusing on ecosystem tion and climate change mitigation mostly are of a global nature services may therefore seem like a promising approach in terms of (Adams et al., 2004). As a consequence, often such initiatives attract avoiding conservation at the cost of local people, and development at very little local support (e.g. Bennett and Dearden, 2014; Maikhuri the cost of the environment. Inherent in ecosystem services approaches et al., 2001; Shrestha and Alavalapati, 2006). Even though later con- is an underlying assumption that explicit acknowledgement and focus servation projects have increasingly sought to incorporate benefits for on nature's goods and services to humans and the externalities asso- local people (e.g. through sharing of revenues from wildlife tourism) or ciated with many economic activities (e.g. in the form of unaccounted to compensate them for losses, the main impetus was still that of con- for pollution) will automatically result in conservation of the environ- servation for its own sake, that is, conservation based on the notion that ment (Turnhout et al., 2013). Many of the frameworks and models nature should be preserved for its intrinsic value independent of its developed since the Millennium Ecosystem Assessment (2005) reflect
☆ This article is part of a special feature entitled: “Forest, Food, and Livelihoods” published at the journal Forest Policy and Economics 84C, 2017. ⁎ Corresponding author. E-mail address: [email protected] (A. Byg). http://dx.doi.org/10.1016/j.forpol.2017.06.002 Received 13 May 2016; Received in revised form 1 May 2017; Accepted 2 June 2017 Available online 07 June 2017 1389-9341/ © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). A. Byg et al. Forest Policy and Economics 84 (2017) 112–119 on the link between ecosystem services and well-being (Fisher et al., 2014; Sikor et al., 2013) and provide the rationale for interventions aiming at both environmental conservation/restoration and human well-being. However, a focus on ecosystem services to humans does not automatically lead to better outcomes in terms of biodiversity and (different) people's well-being (Dawson and Martin, 2015; Turnhout et al., 2013) and often the focus remains on higher scales, such as in landscape-scale restoration. Despite the emphasis on ‘what nature does for people’ many ques- tions remain regarding the implications of using ecosystem services frameworks as a tool for nature conservation. Critical voices have pointed out that ecosystem services approaches tend to promote a shift towards a reconceptualization of nature in terms of its market value and often reinforce existing power relationships (Berbés-Blázquez et al., 2016; Gómez-Baggethun and Pérez, 2011; Turnhout et al., 2013). Other ‘blind spots’ which are often ignored include ecological as well as social complexity within as well as between scales, the co-produced nature of ecosystem services, and the existence not only of services but also of disservices (Berbés-Blázquez et al., 2016; Lyytimäki and Sipilä, 2009; Norgaard, 2010; Rasmussen et al., 2016; Reyers et al., 2013). Following Lyytimäki (2015) we here define disservices as “functions or properties Fig. 1. The case study areas are situated in the Halaba district (circle) in the Southern Nations, Nationalities and People's Provincial state of Ethiopia (the area marked in red) of ecosystems that cause effects that are perceived as harmful, un- (source: TUBS, Wikicommons). (For interpretation of the references to colour in this ” pleasant or unwanted . Examples include pathogens of humans, live- figure legend, the reader is referred to the web version of this article.) stock and crops, and ‘pest’ species which eat or damage crops. In ad- dition, ecosystem services approaches often focus on one particular distributed and mediated in the context of reforestation projects with an service or benefit which can have negative impacts on other services explicit focus on benefits to local populations. As our starting point we (Kull et al., 2015). Like earlier approaches to conservation, ecosystem take local perceptions of the services and disservices and how these are services approaches mostly ignore the awkward issue of trade-offs be- distributed amongst people using a grounded analysis based on focus tween services and disservices and how these are distributed amongst group discussions and participatory methods. In addition, we in- different groups of people (Bennett et al., 2015; Daw et al., 2011; Howe vestigated how decisions are made and implemented and how this re- et al., 2014; Lele, 2013; Suich et al., 2015). Instead, emphasis is on lates to people's experiences of the services and disservices from the politically palatable win-win situations, despite the fact that such si- restoration areas. By doing so, this study also aims to inform current tuations have been shown to be the exception rather than the rule large-scale restoration efforts in the context of poverty alleviation and (McShane et al., 2011). However, ignoring existing trade-offs not only sustainable development. amongst different ecosystem services but also amongst groups of people may seriously undermine ecosystem services approaches, especially where ecosystem services approaches are coupled with development/ 2. Study area poverty alleviation goals. Distribution of services and disservices is ultimately a question of power though this is often mediated through The study took place in three communities (Laygnaw Arsho, Assore factors, such as gender, ethnicity, age, geographical location, etc. As and Andegna Choroko) in Halaba district (‘woreda’ in Amharic), in analyses of ecosystem services based approaches to conservation and Ethiopia's Southern Nations, Nationalities and People's Regional State restoration projects often look at the aggregated level of services and (see Fig. 1). The majority of the population in this area are small scale benefits, power issues and distribution of services and disservices re- farmers with on average less than 2 ha of land per household (un- main invisible (Daw et al., 2011). A disaggregated ‘accounting’ of both published survey data collected for the ESPA ALTER project). Most of services and disservices and their distribution across people and space the agricultural production is for subsistence (mainly Maize) with sur- can therefore help to understand the outcomes of ecosystem services plus being sold at the market as well as some cash crops such as pepper, based approaches for both conservation and the livelihoods of people coffee, and khat. Ethnically, there are two major groups in the area, the living in or around these areas. Likewise, explicitly including power as Halaba (accounting for about 50% of the population) and the Silte, part of the analysis may help to provide a richer and more informative accounting for about a third, while the rest of the population belong to picture of ecosystem services and trade-offs(Berbés-Blázquez et al., a number of other groups. The large majority of the population is 2016). Analyses of power and environmental justice not only need to Muslim (around 94%) with the rest being Orthodox and Protestant look at issues of distribution and resource access but also at recognition Christians. Based on our fieldwork and research in the area, ethnicity is and participation (Schlosberg, 2004; Sikor, 2013). not a major factor influencing or mediating access to resources and li- Here we look at the example of forest regeneration projects in velihood means. southern Ethiopia which have been initiated with the explicit aim of Soil degradation and erosion are recognised as serious problems in benefitting local people, and ask whether such a focus on ecosystem the area and deep erosion gullies are a common sight. According to services provision does in fact guarantee better outcomes in terms of farmers, erosion and soil degradation have been caused by deforestation livelihoods and human well-being. In the study region, forest re- linked to increases in the human population and a shift from livestock to generation in enclosed restoration areas is seen as an important way of mainly crop based livelihoods from around the 1970s onwards (see combatting soil erosion and degradation while at the same time pro- Section 4.1, below). One initiative to reverse soil degradation and erosion viding additional services in the form of timber and non-timber pro- has been the reforestation of degraded, communal areas. Reforestation ducts to the local residents. Forest regeneration areas have been es- and afforestation policies were favoured by the government during the tablished since the 1980s, with the newest ones dating from the 2010s. socialist military council –‘Derg’ regime (1974–1987) – to both tackle These are long-term enclosed communal areas administered by local biomass shortages due to the increasing rural population and the oil price forest committees under local government's supervision. We use this crisis in the 70s and counteract the negative environmental effects of example to look at the ways in which services and disservices are deforestation that resulted from the agricultural modernization
113 A. Byg et al. Forest Policy and Economics 84 (2017) 112–119 programme championed during the Imperial period (Ayana et al., 2013; Table 1 Jagger and Pender, 2000). This policy focused on collective and state- Main characteristics of the restoration areas. owned plantations that focused primarily on wood products from fast Source: ALTER (2014). growing exotic species (Ayana et al., 2013; Jagger and Pender, 2000). Restoration Size (ha) Age (years) Main species composition Population The drought and famine in the 80s triggered a shift from a primary focus area on production to the production of multiple benefits (Ayana et al., 2013). Large-scale efforts at restoring degraded areas were very much im- Assore 112 33 Eucalyptus spp., Acacia 2479 seyal, Grevillea robusta plemented in a top-down fashion under food-for-work programmes Andegna 65 6–33 Eucalyptus spp., Acacia 3115 (Nedessa et al., 2005). In addition to government-led initiatives, devel- Choroko saligna, Acacia albida, opment agencies such as The World Bank, United Nations World Food Grevillea robusta, Croton Programme and international NGOs, provided funds to support these macrostachyus Laygnaw 270 3–33 Eucalyptus spp., Acacia 4500 reforestation projects (Jagger and Pender, 2000). Current ‘greening’ Arsho albida, Acacia seyal programmes, such as the Humbo Assisted Natural Regeneration Project1 developed by World Vision and the World Bank, have been established as a means to address climate change and alleviate poverty (Brown et al., Table 2 2011). While current reforestation projects have principles of participa- Summary of methods and participants per study site. tion embedded in them, the interpretation of what this means varies widely amongst projects (Nedessa et al., 2005). Assore Andegna Laygnaw Arsho Choroko In the three study sites, the reforested areas were of different ages and sizes, with the oldest ones dating back to the 1980s and the youngest Focus group: restoration 7 (3 women, 4 5 (2 women, 3 8 (4 women, 4 ones less than five years old, and ranging from around 60 ha to around areas men) men) men) 270 ha in size (see Table 1). The three communities were selected to Focus group: livelihoods 16 (8 women, 8 12 (6 women, 6 12 (6 women, 6 men) men) men) include restoration areas of a range of ages, sizes and tree species for the Participatory mapping 3 women 5 women 8 women study of services, disservices and trade-offs. These restoration areas were 8 men 5 men 5 men established on what has been heavily degraded communal grazing land. Participatory 3 women 4 women 14 women Livestock grazing is always excluded from the restoration areas. While photography 4 men 4 men 5 men limited tree cutting and collection of dead wood is allowed, the refor- ested areas are maintained as forest and not cleared again. In all three questions on how access to different services was regulated and who communities, the reforestation had been initiated by outsiders (govern- benefited or suffered from disservices. In the focus group we did not use ment and/or NGOs). The techniques and species employed have changed the language of ecosystem services (and disservices) but asked about over time from when the first restoration areas were established in the the good things and the bad things about the reforestation areas and early 80s using mainly exotic species to more focus on native species in how they were distributed across the population and spatially. more recent restoration efforts. In all three communities, the reforested In the other focus group discussions centred on people's livelihoods areas were governed by locally elected community committees, though and changes that had taken place in the area, in order to understand the according to advice from the district government agency (see Section 4.4, wider context around the establishment of the restoration areas. The below). Community members implemented management decisions in focus groups had between 5 and 16 participants with about equal compulsory, annual work campaigns (watershed programmes) and numbers of men and women, aimed to bring together a diverse group of through voluntary food for work programmes (safety net programmes) local villagers to try to capture different gender and wealth perspec- targeting the poorest households. In all three areas, local guards (paid tives. The discussions lasted 1–2 h and were conducted in the local with money generated from the reforestation areas) were responsible for language (Halaba) with a local community member acting as translator ensuring that rules were followed, and breaking of rules was sanctioned, between Halaba and the national language, Amharic. However, as most mostly through fines. of the male participants were able to speak Amharic as well, discussions would sometimes slide over into Amharic thereby excluding non- 3. Methods Amharic speakers (typically women). In addition, cultural norms meant that even Amharic speaking women contributed less frequently to the In order to elicit local perspectives on the restoration areas, in May discussions than men in mixed groups. In the subsequent exercises 2014 we conducted two focus group discussions per site followed by (participatory mapping and photography) we therefore had separate participatory mapping and participatory photography exercises in groups for women and men. January 2015 (see Table 2 and additional information below). Partici- In addition to the focus groups, in January 2015 we conducted two pants across different wealth categories and gender were selected by participatory exercises per site. The first exercise consisted of a parti- local facilitators. To do this, first, a list of households in the community cipatory mapping of services and disservices from the local area (again was elaborated jointly with the community leader and key informants. phrased as good and bad things). The second exercise was a partici- Then, households were categorised into three different wealth cate- patory photography exercise (also known as photovoice) where we gories (i.e. poor, medium, and rich) using a participatory wealth went for a walk with small groups of participants during which they ranking exercise. Wealth categories were defined by the participants in took pictures of the things from the environment they regarded as good terms of material and non-material assets. Material assets included for or bad. These images were then uploaded on a computer and shown to example livestock, cash crops and good quality housing while examples the groups who had taken the pictures and used to discuss what the of non-material assets are knowledge, health and empowerment of pictures showed, and why they thought these things were important. women. This list and categorisation of households was used for the Discussions were audio recorded and translated from Amharic into selection of participants in order to ensure a representation of diverse English by Ethiopian co-authors. Notes from all exercises were subse- views and capabilities. quently uploaded into qualitative data analysis software Nvivo where One of the focus groups explicitly addressed the services, dis- they were coded for services, disservices, and distributional issues. services, and governance of the restoration areas. This included Within each of these categories, more specific codes were created by repeatedly going over the material and identifying recurring issues and topics. 1 https://wbcarbonfinance.org/Router.cfm?Page=Projport&ProjID=9625
114 A. Byg et al. Forest Policy and Economics 84 (2017) 112–119
4. Results individual households. At the community level, timber and income from sale of forest product quotas had been used to construct and run 4.1. Changes in livelihoods and environment community facilities such as schools and health centres. For house- holds, having access to these products from the restoration areas was In all three focus groups on wider changes in the area and people's seen as something which saved households money. Even though people livelihoods all the participants agreed that there had been a trend of had to pay a fee to be allowed to collect a certain amount of product, deforestation from the Imperial period prior to the military Derg regime this was cheaper than buying the products from the market. This also (i.e. before 1974) to the present. As causes for this trend, they identified meant that those who did not use the products themselves could resell an increase in the human population combined with a shift from live- them at a profit in the market and thereby supplement their income. In stock centred livelihoods to crop cultivation focussed farming. The some reforestation areas beehives had been established, and the forest latter was linked to the political changes when the military Derg regime was therefore also seen as a source of bees (for those wishing to es- came to power. Under the Imperial regime in the preceding period land tablish their own beehives) and honey. Amongst the recognised reg- ownership in the region had generally taken a feudal structure where ulating and supporting services were cooler temperatures, and reduced local people were generally tenants living on the land of large scale flooding, erosion and gully formation. Some of the focus groups dis- landlords. Under this system, half of any crops cultivated on the land tinguished between erosion reduction and soil conservation. Although had to be handed over to the landlord. According to the participants, these overlap, soil conservation refers to broader practices aimed at this meant that people were generally disinclined to cultivate crops, but conserving and enhancing soil structure and fertility. focussed on livestock rearing instead. During this time, people used to Cultural services associated with restoration areas were also re- graze their animals freely in the forest areas. In one focus group par- cognised, particularly in Assore where people went to a dam con- ticipants also mentioned that during this time they used to derive fruits structed in connection with the restoration area to fetch water and let and bushmeat from the forest. With the advent of the military regime their livestock drink. In addition to the water itself, people mentioned came a redistribution of land under the slogan ‘land to the tiller’, which the shade provided by big trees while waiting with their animals and handed over land to tenant farmers (though without the right to sell the the opportunity this shade gave to discuss social issues and gather with land). At the same time, the military regime promoted crop cultivation people from the community. In some cases, trees were also used to mark in the region, and the population in the area started to grow. Taken the end of Ramadan and to celebrate religious ceremonies. Similar is- together, these developments meant that people shifted towards more sues were mentioned in all the other villages, but in relation to trees in crop cultivation and forest areas were gradually decimated to bring other public and private spaces such as the school and front yards of the more land under cultivation as well as to provide construction materials houses. In Assore, restoration areas were also seen as attracting visitors and fire wood. In all three areas, participants said that the deforestation and support from different organisations and as a way of generating and more intensive crop cultivation gradually led to a decrease in soil knowledge. fertility and to problems such as soil erosion. In one focus group, par- In all three communities, villagers saw a link between deforestation ticipants also linked deforestation to climatic change, saying that it had and problems such as erosion and flooding. Although perceptions about made the area drier and caused crop failures. In addition to the decline the causes of deforestation were shared across villages (see section 4.1, in forest cover, participants in all three areas stated that there had been above), perceptions of the services of reforestation with regard to re- a change in forest composition away from native species to introduced duced flooding and erosion differed within and amongst communities. species such as Eucalyptus and Grevillea. While some participants thought that the reforestation areas had sub- stantially reduced erosion, others saw them as less effi cient and only 4.2. Local perceptions on services and disservices from reforestation applying to a very small area. Some, but not all, participants also felt that having participated in the communal reforestation and restoration In all three communities, villagers recognised that the reforestation projects had provided them with knowledge which they could make use areas provided a number of services (see Table 3). In terms of ecosystem of on their own plots. services these included provisioning as well as regulating, supporting The main disservice from reforestation, which was brought up in all and cultural services. The most important provisioning services were three communities, was the damage inflicted on crops by wild animals timber for construction and fuel, and grasses for livestock fodder and which had moved back into the areas as forests had been regenerated. thatch. Sale of these products (and of quotas for their collection) con- Hunting is not allowed under current wildlife legislation in Ethiopia stituted an income source both at the level of the communities and for and therefore the only way farmers could prevent crop attacks was by patrolling and guarding plots especially during the night. Disservices Table 3 were particularly large for those households whose plots were adjacent Ecosystem services from restoration areas identified in the three communities. to the restoration areas. These households could lose most of their harvest in a single animal attack. The species perceived as most pro- Ecosystem services Assore Andegna Laygnaw blematic was warthog, but porcupines and various kinds of monkeys Choroko Arsho were also mentioned as causing substantial damage to farmers' crops. In Provisioning Timber x x x addition, hyenas were said to occasionally attack livestock. Malaria and Grass x x x water-borne diseases were also mentioned as one of the disservices Honey x x from the development of small ponds and other rainwater harvesting Seedlings x Water storage x x systems in the restoration areas. Thus, while water storage was men- Regulating Flood reduction x x tioned as a service from these areas, it also comes associated with some Erosion reduction x disservices. However, disservices associated to standing waters were Soil conservation x x seen less problematic than wildlife attacks. Cooler climate x As the reforested areas had previously been communal grazing Supporting Improved upland x areas areas, we had anticipated that loss of access to grazing would be seen as Green space x a major negative outcome from restoration. However, this was men- Cultural Tree shade x tioned only once, presumably because the areas selected for reforesta- Visits & support x tion were heavily degraded areas and had hence already lost most of Knowledge x generation their grazing value prior to the establishment of the reforestation pro- jects. In addition, although services outnumber disservices from
115 A. Byg et al. Forest Policy and Economics 84 (2017) 112–119 restoration areas, the discussions showed that many of the participants reforestation areas, in two of the areas this was done in consultation thought that the services did not outweigh the damage done by animals. with government agencies. Prices tended to be set at below market Participants indicated that, although restoration areas contribute to li- prices, but to follow general market trends. For two of the areas there velihoods, subsistence strategies depend mostly on the cultivation of were also quota for timber, though these would not be provided an- individual plots which are used for private consumption and, in case of nually but rather at longer intervals and prices and quota were de- surplus, for selling at the local markets. This was also linked to the termined by the government agency. Timber was also cut for communal limited amounts of materials, especially timber, which could be (leg- purposes (e.g. building a school) or given free of charge as construction ally) extracted from the reforestation areas (see section 4.4). It also material to poor households in emergency cases where their houses had meant that disservices such as crop raiding affect the main source of been damaged and needed rebuilding. There were large differences income for households, especially those with plots located in areas amongst the communities and reforestation areas with some commit- nearby the restoration area. tees assigning timber quotas at regular intervals and others never de- pending on the species composition and age of their reforested areas. 4.3. Local perceptions' on the factors that influence the creation of For example, in older reforestation areas which mainly consisted of ecosystem services and disservices eucalyptus, timber extraction was more frequently allowed compared to younger areas and those consisting of native species. Hence the Participants in the three communities mentioned a range of factors services available to the community members differed. influencing delivery of ecosystem services from the reforestation areas. In all three communities, bylaws regulating the access to products The most important factors related to the composition and age of the were determined by the elected committees and followed the principle restoration areas. The selection of species for replanting was based on of equal rights to everyone. This meant that all households had the recommendations from government agencies and had changed over same right to buy a fixed amount of the available products at the price time as different species and approaches were favoured at different determined by the respective community forest committee. This was times. In some of the earlier restoration areas, mainly exotic species meant to ensure that wealthier households did not benefit more than such as eucalyptus had been planted while more recently established poorer ones. However, according to the participants, these rules were areas typically consisted of a mix of native species and part of one area frequently circumvented by informal agreements between richer was left to regenerate naturally. Different species provided different households (who were keen to obtain more than their allotted quota) services and disservices. Eucalyptus, for example, was valued for its and poorer households (who did not have sufficient funds to buy the quick growth and its timber, but was seen to use a lot of water and to products they were entitled to). In these cases, poorer households deter other plants from growing in the vicinity. Native species gave less would get the money from the richer households and buy the products timber, but resulted in areas that were good at providing fodder and on their behalf against a small commissioning fee. In addition, some thatch. While some thought that naturally regenerated reforestation participants also indicated that there were examples of richer house- areas and those consisting of native species harboured more wild ani- holds obtaining larger quota or better quality materials directly from mals (and hence resulted in more crop damage) than areas which had the forest committees in direct circumvention of the rules. Participants been replanted with species such as eucalyptus, others thought that also reported that richer households were more likely to steal products, there was no difference in the amounts of wild animals and crop da- such as timber, as they could afford the fines imposed in case they were mage linked to different reforestation areas. Some participants ex- caught. Poorer households on the other hand were said to be too pressed the wish for more useful species in the form of timber or fruit frightened by the threat of fines, too large for them to pay, to attempt trees to be planted in the reforestation areas. violating the rules. While richer households thus were in a position to In addition to species composition, the age of the restoration areas take more advantage of services (either semi-legally or illegally) from also influenced what services they delivered and how people benefited the reforestation areas, they were at the same time less dependent on from these areas. Participants acknowledged that while some products these services. Typically, richer households had more land than poorer were obtained after a couple of years, tree products only appeared after households and were therefore more likely to allocate some of their more than a decade of enclosure. Younger restoration areas were seen own land to tree planting which provided them with their own, private as providing a lot of fodder and thatch. As restoration areas matured supply of timber, fuel wood, and leaves for compost making, in addition and the tree canopy closed in the provision of fodder and thatch grasses to grazing land. The results also point to some gender differences re- typically declined. In contrast, older reforestation areas contained more garding the services generated from restoration areas, which may re- timber than younger ones but yearly services, such as grass and thatch, flect gendered roles, responsibilities and resources. For example, timber were lower. While timber had higher market value, quotas for its ex- was mostly mentioned in the focus groups conducted with men, while traction were very limited, and mostly timber was only used for the others, such as grass, were recognised both by men and women. More construction of communal facilities and in emergency situations. valuable products are usually considered the responsibility and prop- Hence, grasses which had a lower monetary value, but which could be erty of men, while women are responsible for harvest and sale of lower extracted in higher amounts played a more important role in the live- value products. Differences in fodder distribution can be explained, at lihoods of most households. least in one community (Andegna Choroko), by the fact that two wo- men's cooperatives had been allocated special rights in relation to the 4.4. Factors which influence the distribution of services and disservices reforested areas, such as being able to collect more fodder grass for livestock owned by the cooperatives. In all three communities, the distribution of services from the re- In all communities, improvements in soil fertility and erosion and storation areas was governed by forest committees consisting of elected flood control were two of the environmental reasons for the initial es- community members. Communities elected representatives for the tablishment of reforestation areas. Not surprisingly, the participants committees, who usually served for a term of 3 to 4 years, though in one generally acknowledged that soil fertility had improved in the refor- of the communities (Andegna Choroko) focus group participants in- ested areas. However, this did not provide any direct services to the dicated that committee members would stay longer if there were no farmers as the areas outside of the reforestation areas generally con- complaints. The committees were composed of between 7 and 10 tinued to be degraded. Hence, farmers only benefitted indirectly members. At the time of the field work the committee in Andegna through the other services that depended on the soil inside the refor- Choroko was composed of 5 men and 5 women, in Assore of 4 men and ested areas such as the growth of grasses for fodder and thatch. Some 2 women and in Laygnaw Arsho of 7 men. While these committees set participants were hoping that in time they would be allowed to once annual quotas and prices for grasses for thatch and fodder from the again cut down the forest and turn these areas back into agricultural
116 A. Byg et al. Forest Policy and Economics 84 (2017) 112–119
fields which would then benefit from the restored soil fertility. 5. Discussion However, the oldest areas were more than 30 years old and there was no sign that these would be allowed to be converted to agricultural land Many studies of ecosystem services mainly focus on inventories of again. In terms of reduced erosion and flooding, these services were different types of services while ignoring aspects of justice and power, more clearly felt and appreciated by at least some of the participants including aspects of distribution, recognition and participation. In this though others complained that the reforested areas had not been ef- study of reforestation projects in southern Ethiopia, we found that fective in stopping erosion and flooding in the surrounding areas. Even services outnumber disservices. Based on a simple aggregate inventory where the reforested areas were seen to have reduced erosion and of ecosystem services it would therefore seem that these reforestation flooding, this was mainly in the areas downhill from the restoration projects have been very successful and this is indeed the way these areas. Some participants also thought that their own agricultural fields kinds of projects have often been portrayed (Nedessa et al., 2005). This had benefitted indirectly, as participation in the reforestation work had is similar to the way in which costs and benefits are evaluated in larger meant they had acquired new knowledge of different options that they scale initiatives such as landscape-scale conservation and restoration could use to restore fertility on their own land. projects meant to tackle both environmental degradation and poverty The distribution of services from the reforested areas was thus in- alleviation (e.g., Menz et al., 2013). Nevertheless, in the three study fluenced by a variety of factors, including household wealth in inter- sites there was widespread dissatisfaction with the reforestation action with the existing governance of the reforested areas, and topo- amongst the local participants. Our study thus shows that employing an graphy and location. In contrast, the distribution of disservices in the ecosystem services lens does not in itself guarantee that restoration form of crop damage from wild animals was dependent only on the projects will overcome the shortfalls of previous restoration and con- location of farmers' fields in relation to the reforested areas. Those who servation approaches. Such approaches may therefore suffer from si- had their fields close to the reforested areas thus suffered more damage milar pitfalls as earlier conservation approaches lacking local support. than those whose fields were located further away. According to the From the discussions with villagers it was clear that several factors participants there were no discernible patterns in relation to where the contributed to their divergent views of the reforestation areas. These fields of richer or poorer people were located. were related to a perceived imbalance between services and disservices, issues of distribution and issues of participation. Firstly, while the participants acknowledged services such as cooler 4.5. Existing strategies for dealing with disservices temperatures, reduced erosion and flooding risk downhill from re- storation areas, and goods in the form of wood and grass these did not At the time of the fieldwork, no compensation mechanisms were in have the same immediate impacts on their livelihoods as the disservices place to recompense farmers for the losses suffered from wild animal from wildlife. Partly this was due to the nature of these services and attacks. However, richer households often had several plots and gen- issues of scale. erally more land, and were less vulnerable to the loss of part of their The scales at which management interventions take place and are crops. Households whose fields were close to the reforestation areas planned, are typically of a higher order (e.g. regional, national and typically tried to minimise the risk of crop damage by having a person longer term), compared to the scales at which local people derive stay out in the field during night time to guard the crops. However, this benefits from ecosystem services (Kull et al., 2015). The benefits of soil was seen to have negative impacts on people's health (due to lack of restoration are thus long-term, partly very localised (e.g. improved soil sleep) and the availability of labour for other tasks. Some people had fertility within the reforested areas) and partly at the landscape scale also tried to build fences, but these were generally not sufficient to keep (e.g. reduced siltation of rivers) or even at higher scales such as in the the animals away. Some participants also mentioned switching crops in case of carbon sequestration. For the individual farmers, improvements response to animal attacks. However, this was a problematic response in soil fertility would only be realised if they were allowed to convert in that the crops most prone to wildlife damage (e.g. maize, sorghum, the forest back to agricultural land. However, this did not seem to be and teff) were also amongst the most important subsistence crops. part of the planned management for the reforested areas. Hence the Maize was especially important due to its multiple functions with dif- ‘side-benefits’ from the reforested areas in the form of construction ferent parts of the plant being used for different purposes. Switching to materials, fire wood, fodder and thatch constituted the main services less vulnerable crops was thus more feasible for richer farmers who had for the participants and were highly important in relation to the support several fields and who typically grew a greater variety of crops (cash or lack thereof for reforestation. The perceived imbalance also related crops as well as subsistence crops). They could therefore grow vulner- to the experienced disservices in the form of wildlife attacks on crops able crops such as maize where there was less danger of animal attacks with potentially devastating consequences for their livelihoods. The and grow other crops where attacks were more likely (depending on the negative impacts from wildlife attacks cannot be measured solely in location of their fields). Participants also reported that some people had terms of the magnitude of crops lost, but also need to take into account simply stopped growing crops as a result of the damage inflicted by wild the uncertainty and anxiety caused by the largely unpredictable nature animals. However, this should not be seen as a coping mechanism but of wildlife attacks, and the need to spend the nights out in the fields rather as one of the most severe manifestations of the disservices from attempting to protect the crops resulting in adverse health impacts the reforestation areas. In case farmers became destitute as a result of (Khumalo and Yung, 2015). Taken together, this had resulted in villa- the loss of their crops, they did become eligible for general government gers' experience that the reforestation mainly seemed to be for wildlife ‘food for work’ aid programmes. and not for humans. Our study thus indicates that employing ecosystem As hunting is not allowed under current wildlife legislation in services frameworks is not in itself enough to ensure that conservation Ethiopia, some participants mentioned that they had tried to block the will contribute to an increase of human well-being. While our study is exits of the warthogs' burrows in order to get rid of them. Ultimately, based on the example of specific reforestation projects in one part of according to participants, the only effective way of dealing with the Ethiopia, we suspect that similar imbalances can be found in other problem of crop-raiding would be to reduce the number of animals conservation and restoration projects focusing on supporting and reg- through hunting. Participants suggested that this could be done either ulating ecosystem services. For ecosystem services based approaches to by allowing the villagers to hunt the animals or by getting a govern- be successful in the long term more needs to be done to get the balance ment agency to regulate the number of the animals. Participants did not right between longer term supporting and regulating services and lo- discuss any strategies for dealing with disservices linked to standing cally important services and disservices. This includes taking serious waters (malaria, water-borne diseases). local experiences and values (Turner et al., 2008) and if necessary finding compromises between services which are important over the
117 A. Byg et al. Forest Policy and Economics 84 (2017) 112–119 long term and often at higher scales and in the form of public goods, the ground meant that there was widespread disappointment and dis- and more immediate, locally important services for individual house- satisfaction with the reforestation projects in the three communities. holds. This could for example be done by including more timber species This happened despite the explicit focus on reforestation as a means to or maintaining a thinner tree cover to promote the production of fodder improve human well-being. The mismatch between the professed aims and thatch grass. of reforestation, local expectations, and the realised services and dis- Secondly, the distribution of services and disservices was unequal services and their distribution contributed to undermining local support both within and amongst different areas. While rules aimed at ensuring to reforestation as the reforestation areas were by some seen to mainly a degree of equality in the distribution were in place, we nevertheless benefit wildlife rather than humans. found that wealthier households were in a better position to appro- In order to design ecosystem services approaches that can truly priate services legally or illegally. At the same time wealthier house- contribute to human well-being and poverty eradication, it is necessary holds were less dependent on these services and less vulnerable to the to dis-aggregate services and disservices, to acknowledge trade-offs disservices. This is similar to findings from other studies and highlights between different services and disservices and different groups of the relationship between livelihoods and services as well as vulner- people, and find the mechanisms to tackle them (Daw et al., 2011; abilities and power relationships (e.g. Berbés-Blázquez et al., 2016; Dawson and Martin, 2015; Lele, 2013). Ecosystem services manage- Dawson and Martin, 2015; Dickman, 2010). Local power dynamics and ment entails decisions that are inherently political as decisions about, access to resources therefore need to be taken into account both in for example, which services to focus on, and how and what to measure, relation to the study of ecosystem services and in projects seeking to all have consequences for who benefits and who loses out (Kull et al., promote environmental restoration using an ecosystem services fra- 2015). Therefore, power relations also need to be taken into account mework. This includes asking who benefits in what ways and at what when trying to design reforestation and other environmental manage- scales and who loses out, making trade-offs visible and explicitly in- ment approaches that are meant to contribute to the improvement of cluding them in the management and decision making. However, trade- people's livelihoods and well-being. However, far too often ecosystem offs and how they play out may not be predictable beforehand, due to services approaches have ignored them. In line with Howe et al. (2014), the complex and dynamic nature of social-ecological systems. Dealing a focus on trade-offs, rather than on win-wins, provide a stronger basis with trade-offs therefore entails monitoring how the system and the for the design of interventions that are more likely to achieve those win- balance between services and disservices and their distribution de- win outcomes. To do so requires meaningful participation of local velops over time, and in response employing flexible rules and adaptive people in evaluating outcomes, designing and adjusting management management approaches, which can be adjusted in response (Armitage rules and negotiating the distribution of services and disservices. This is et al., 2009; Chazdon and Uriarte, 2016). important not only in relation to restoration efforts aimed at the local This leads us to the third point which concerns participation and scale, but also in relation to larger scale initiatives such as landscape- governance. Although local participation is embedded as a guiding scale conservation and restoration projects which are increasingly principle in reforestation projects in Ethiopia, the meaning of ‘partici- gaining support to tackle global problems, but which risk failing by pation’ is interpreted in varying ways and may in reality not mean ignoring trade-offs at the local scale and between the local and higher much more than participation in implementation (Nedessa et al., 2005). scales. Local participation can be seen as an end in itself, which may help to foster greater feelings of ownership and acceptability of environmental Acknowledgements management such as reforestation initiatives (Reed, 2008). Many eco- system services based approaches mainly focus on the outcomes in We wish to thank the participants in the three study communities terms of services themselves while ignoring issues of governance and for their time and patience and for sharing their insights and opinions participation (Spangenberg et al., 2015). This is especially important in with us. We also owe thanks to Anke Fischer and the editors of this countries like Ethiopia where the early history of reforestation projects special issue and an anonymous reviewer for helpful comments on was marked by top-down approaches, which has generated a legacy of earlier versions of this article. This work is part of the project distrust and uncertainty (Nedessa et al., 2005). In addition, greater ‘Alternative Carbon Investments in Ecosystems for Poverty Alleviation – local participation in management decisions could potentially help to below ground versus above ground opportunities for restoration of modify the balance between services and disservices (e.g. by modifying ecosystem services’ (ALTER) and was funded with support from the the species and age composition and managing wildlife). In addition, Ecosystem Services for Poverty Alleviation (ESPA) programme (grant services and disservices are intimately linked to values as well as live- no. NE/K010441/1). The ESPA programme is funded by the lihoods, and to understand trade-offs between them therefore also re- Department for International Development (DFID), the Economic and quires the inclusion of local knowledges and values in the decision Social Research Council (ESRC) and the Natural Environment Research making process (Galafassi et al., 2017). However, greater local parti- Council (NERC). cipation is not a panacea that can solve all problems. Villagers were already involved in designing distributional rules, and while these were References following principles of equality, they were at the same time de facto favouring richer households. Often multiple interpretations co-exist Adams, W.M., Aveling, R., Brockington, D., Dickson, B., Elliott, J., Hutton, J., Roe, D., regarding what ‘just’ means in the context of reforestation or con- Vira, B., Wolmer, W., 2004. Biodiversity conservation and the eradication of poverty. Science 306, 1146–1149. servation projects (Martin et al., 2014). Local perceptions of what is just Agarwala, M., Atkinson, G., Fry, B., Homewood, K., Mourato, S., Rowcliffe, J., Wallace, need to be taken into account for reforestation projects to succeed (He G., Milner-Gulland, E., 2014. Assessing the relationship between human well-being and Sikor, 2015). This includes making different interpretations and and ecosystem services: a review of frameworks. Conserv. Soc. 12, 437–449. ff ff Armitage, D.R., Plummer, R., Berkes, F., Arthur, R.I., Charles, A.T., Davidson-Hunt, I.J., their consequences explicit, including trade-o s between di erent Diduck, A.P., Doubleday, N.C., Johnson, D.S., Marschke, M., McConney, P., groups of people or different sets of values (Daw et al., 2015; McShane Pinkerton, E.W., Wollenberg, E.K., 2009. Adaptive co-management for social–eco- et al., 2011). logical complexity. Front. Ecol. Environ. 7, 95–102. Ayana, A.N., Arts, B., Wiersum, K.F., 2013. Historical development of forest policy in Ethiopia: trends of institutionalization and deinstitutionalization. Land Use Policy 32, 6. Conclusion 186–196. Bennett, E.M., Cramer, W., Begossi, A., Cundill, G., Díaz, S., Egoh, B.N., Geijzendorffer, While the reforestation areas have been successful in relation to the I.R., Krug, C.B., Lavorel, S., Lazos, E., Lebel, L., Martín-López, B., Meyfroidt, P., Mooney, H.A., Nel, J.L., Pascual, U., Payet, K., Harguindeguy, N.P., Peterson, G.D., stated goals in that they have managed to improve soil fertility and Prieur-Richard, A.-H., Reyers, B., Roebeling, P., Seppelt, R., Solan, M., Tschakert, P., reduce erosion and flooding, the reality of the experienced trade-offson Tscharntke, T., Turner Ii, B.L., Verburg, P.H., Viglizzo, E.F., White, P.C.L., Woodward,
118 A. Byg et al. Forest Policy and Economics 84 (2017) 112–119
G., 2015. Linking biodiversity, ecosystem services, and human well-being: three Lyytimäki, J., Sipilä, M., 2009. Hopping on one leg – the challenge of ecosystem dis- challenges for designing research for sustainability. Curr. Opin. Environ. Sustain. 14, services for urban green management. Urban For. Urban Green. 8, 309–315. 76–85. Mace, G.M., 2014. Whose conservation? Science 345, 1558. Bennett, N.J., Dearden, P., 2014. Why local people do not support conservation: com- Maikhuri, R.K., Nautiyal, S., Rao, K.S., Saxena, K.G., 2001. Conservation policy–people munity perceptions of marine protected area livelihood impacts, governance and conflicts: a case study from Nanda Devi biosphere reserve (a world heritage site), management in Thailand. Mar. Policy 44, 107–116. India. Forest Policy Econ. 2, 355–365. Berbés-Blázquez, M., González, J.A., Pascual, U., 2016. Towards an ecosystem services Martin, A., Gross-Camp, N., Kebede, B., McGuire, S., Munyarukaza, J., 2014. Whose en- approach that addresses social power relations. Curr. Opin. Environ. Sustain. 19, vironmental justice? Exploring local and global perspectives in a payments for eco- 134–143. system services scheme in Rwanda. Geoforum 54, 167–177. Brown, D.R., Dettmann, P., Rinaudo, T., Tefera, H., Tofu, A., 2011. Poverty alleviation McShane, T.O., Hirsch, P.D., Trung, T.C., Songorwa, A.N., Kinzig, A., Monteferri, B., and environmental restoration using the clean development mechanism: a case study Mutekanga, D., Thang, H.V., Dammert, J.L., Pulgar-Vidal, M., Welch-Devine, M., from Humbo, Ethiopia. Environ. Manag. 48, 322–333. Peter Brosius, J., Coppolillo, P., O'Connor, S., 2011. Hard choices: making trade-offs Chazdon, R.L., Uriarte, M., 2016. Natural regeneration in the context of large-scale forest between biodiversity conservation and human well-being. Biol. Conserv. 144, and landscape restoration in the tropics. Biotropica 48, 709–715. 966–972. Daw, T., Brown, K., Rosendo, S., Pomeroy, R., 2011. Applying the ecosystem services Menz, M.H.M., Dixon, K.W., Hobbs, R.J., 2013. Hurdles and opportunities for landscape- concept to poverty alleviation: the need to disaggregate human well-being. Environ. scale restoration. Science 339, 526–527. Conserv. 38, 370–379. Millenium Ecosystem Assessment, 2005. Ecosystems and Human Well-being: Synthesis. Daw, T.M., Coulthard, S., Cheung, W.W.L., Brown, K., Abunge, C., Galafassi, D., Peterson, Millenium Ecosystem Assessment, Washington, DC., pp. 155. G.D., McClanahan, T.R., Omukoto, J.O., Munyi, L., 2015. Evaluating taboo trade-offs Nedessa, B., Ali, J., Nyborg, I., 2005. Exploring Ecological and Socio-economic Issues for in ecosystems services and human well-being. Proc. Natl. Acad. Sci. 112, 6949–6954. the Improvement of Area Enclosure Management - a Case Study from Ethiopia. Dawson, N., Martin, A., 2015. Assessing the contribution of ecosystem services to human Drylands Coordination Group report. Drylands Coordination Group, Oslo, pp. 55. wellbeing: a disaggregated study in western Rwanda. Ecol. Econ. 117, 62–72. Norgaard, R.B., 2010. Ecosystem services: from eye-opening metaphor to complexity Dickman, A.J., 2010. Complexities of conflict: the importance of considering social factors blinder. Ecol. Econ. 69, 1219–1227. for effectively resolving human–wildlife conflict. Anim. Conserv. 13, 458–466. Rasmussen, L.V., Christensen, A.E., Danielsen, F., Dawson, N., Martin, A., Mertz, O., Fisher, J.A., Patenaude, G., Giri, K., Lewis, K., Meir, P., Pinho, P., Rounsevell, M.D.A., Sikor, T., Thongmanivong, S., Xaydongvanh, P., 2016. From food to pest: conversion Williams, M., 2014. Understanding the relationships between ecosystem services and factors determine switches between ecosystem services and disservices. Ambio 1–11. poverty alleviation: a conceptual framework. Ecosyst. Serv. 7, 34–45. Reed, M.S., 2008. Stakeholder participation for environmental management: a literature Galafassi, D., Daw, T.M., Munyi, L., Brown, K., Barnaud, C., Fazey, I., 2017. Learning review. Biol. Conserv. 141, 2417–2431. about social-ecological trade-offs. Ecol. Soc. 22. Reyers, B., Biggs, R., Cumming, G.S., Elmqvist, T., Hejnowicz, A.P., Polasky, S., 2013. Gómez-Baggethun, E., Pérez, M.R., 2011. Economic valuation and the commodification of Getting the measure of ecosystem services: a social-ecological approach. Front. Ecol. ecosystem services. Prog. Phys. Geogr. 1–16. Environ. 11, 268–273. He, J., Sikor, T., 2015. Notions of justice in payments for ecosystem services: insights Schlosberg, D., 2004. Reconceiving environmental justice: global movements and poli- from China's Sloping Land Conversion Program in Yunnan Province. Land Use Policy tical theories. Environ. Polit. 13, 517–540. 43, 207–216. Shrestha, R.K., Alavalapati, J.R.R., 2006. Linking conservation and development: an Howe, C., Suich, H., Vira, B., Mace, G., 2014. Creating win-wins from trade-offs? analysis of local People's attitude towards Koshi Tappu wildlife reserve, Nepal. Ecosystem services for human well-being: a meta-analysis of ecosysetm service trade- Environ. Dev. Sustain. 8, 69–84. offs and synergies in the real world. Glob. Environ. Chang. 28, 263–275. Sikor, T., 2013. The Juistices and Injustices of Ecosystem Services. Routledge, Jagger, P., Pender, J., 2000. The Role of Trees for Sustainalbe Management of Less-fa- Abingdon, UK. vored Lands: The Case of Eucalyptus in Ethiopia. Environmnet and Production Sikor, T., Fisher, J., Few, R., Martin, A., Zeitoun, M., 2013. The justices and injustices of Technology Division Discussion Paper. International Food Policy Research Institute, ecosystem services. In: Sikor, T. (Ed.), The Juistices and Injustices of Ecosystem Washington, D.C., pp. 91. Services. Routledge, Abingdon, UK, pp. 187–200. Khumalo, K., Yung, L., 2015. Women, human-wildlife conflict, and CBNRM: hidden im- Spangenberg, J.H., Görg, C., Settele, J., 2015. Stakeholder involvement in ESS research pacts and vulnerabilities in Kwandu conservancy, Namibia. Conserv. Soc. 13, and governance: between conceptual ambition and practical experiences – risks, 232–243. challenges and tested tools. Ecosyst. Serv. 16, 201–211. Kull, C.A., Arnauld de Sartre, X., Castro-Larrañaga, M., 2015. The political ecology of Suich, H., Howe, C., Mace, G., 2015. Ecosystem services and poverty alleviation: a review ecosystem services. Geoforum 61, 122–134. of the empirical links. Ecosyst. Serv. 12, 137–147. Lele, S., 2013. Environmentalisms, justices and the limits of ecosystem services frame- Turner, N.J., Gregory, R., Brooks, C., Failing, L., Satterfield, T., 2008. From invisibility to works. In: Sikor, T. (Ed.), The Justices and Injustices of Ecosystem Services. transparency: identifying the implications. Ecol. Soc. 13, 7. Routledge, Abingdon, UK, pp. 119–139. Turnhout, E., Waterton, C., Neves, K., Buizer, M., 2013. Rethinking biodiversity: from Lyytimäki, J., 2015. Ecosystem disservices: embrace the catchword. Ecosyst. Serv. 12, goods and services to “living with”. Conserv. Lett. 6, 154–161. 136.
119 Research for a better world
Atlas shares research that can significantly impact people’s lives around the world. Read the award winning stories: ‘SuperAmma’ to the hand washing rescue The Lancet Global Health ‘Sustainable’ coffee: what does it mean for local supply chains in Indonesia? World Development Making construction safety social Automation in Construction www.elsevier.com/atlas Training. Advice. Discussion. Networking.
Free online lectures, training and advice covering: • Introduction to scholarly publishing • How to get published in a journal • Book publishing • Peer review • Publishing ethics • Grant writing • Getting your paper noticed
Helping you on your way to publishing a world-class book or journal article publishingcampus.com
Get recognition CAMPUS CERTIFICATES for courses taken