Tropical Homegardens - library.uniteddiversity.coop

Tropical Homegardens

Advances in Agroforestry 

Volume 3 

Series Editor: 

P.K.R. Nair 

School of Forest Resources and Conservation, 

University of Florida, Gainesville, Florida, U.S.A. 

Aims and Scope 

Agroforestry, the purposeful growing of trees and crops in interacting combinations, began to attain 

prominence in the late 1970s, when the international scientific community embraced its potentials in the 

tropics and recognized it as a practice in search of science. During the 1990s, the relevance of agroforestry 

for solving problems related to deterioration of family farms, increased soil erosion, surface and ground 

water pollution, and decreased biodiversity was recognized in the industrialized nations too. Thus, 

agroforestry is now receiving increasing attention as a sustainable land-management option the world over 

because of its ecological, economic, and social attributes. Consequently, the knowledge-base of 

agroforestry is being expanded at a rapid rate as illustrated by the increasing number and quality of 

scientific publications of various forms on different aspects of agroforestry. 

Making full and efficient use of this upsurge in scientific agroforestry is both a challenge and an 

opportunity to the agroforestry scientific community. In order to help prepare themselves better for facing 

the challenge and seizing the opportunity, agoroforestry scientists need access to synthesized information 

on multi-dimensional aspects of scientific agroforesty. 

The aim of this new book-series, Advances in Agroforestry, is to offer state-of-the art synthesis of research 

results and evaluations relating to different aspects of agroforestry. Its scope is broad enough to encompass 

any and all aspects of agroforestry research and development. Contributions are welcome as well as 

solicited from competent authors on any aspect of agroforestry. Volumes in the series will consist of 

reference books, subject-specific monographs, peer-reviewed publications out of conferences, 

comprehensive evaluations of specific projects, and other book-length compilations of scientific and 

professional merit and relevance to the science and practice of agroforestry worldwide. 

The titles published in this series are listed at the end of this volume.

Tropical Homegardens 

A Time-Tested Example of 

Sustainable Agroforestry 

Edited by 

B.M. Kumar 

Kerala Agricultural University, India 

and 

P.K.R. Nair 

University of Florida, Gainesville, FL, U.S.A.

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

ISBN-10 1-4020-4947-1 (HB) 

ISBN-13 978-1-4020-4947-7 (HB) 

ISBN-10 1-4020-4948-X (e-book) 

ISBN-13 978-1-4020-4948-4 (e-book) 

Published by Springer, 

P.O. Box 17, 3300 AA Dordrecht, The Netherlands. 

www.springer.com 

Printed on acid-free paper 

All Rights Reserved 

© 2006 Springer 

No part of this work may be reproduced, stored in a retrieval system, or transmitted 

in any form or by any means, electronic, mechanical, photocopying, microfilming, recording 

or otherwise, without written permission from the Publisher, with the exception 

of any material supplied specifically for the purpose of being entered 

and executed on a computer system, for exclusive use by the purchaser of the work. 

Printed in the Netherlands.

List of Contributors 

Chapter Reviewers 

Preface 

Introduction 

P.K.R. Nair and B.M. Kumar 

CONTENTS 

Section 1: Historical and Regional Perspectives 

Diversity and change in homegarden cultivation in Indonesia 

K.F. Wiersum 

Urban and homegarden agroforestry in the Pacific islands: Current status 

and future prospects 

R.R. Thaman, C.R. Elevitch, and J. Kennedy 

Amazonian homegardens: Their ethnohistory and potential contribution 

to agroforestry development 

R.P. Miller, J.W. Penn, Jr., and J. van Leeuwen 

Homegardens of Mesoamerica: Biodiversity, food security, and nutrient 

management 

F. Montagnini 

Section 2: Structure, Function, and Dynamics of Homegardens 

Homegarden dynamics in Kerala, India 

A. Peyre, A. Guidal, K.F. Wiersum, and F. Bongers 

Structure and dynamics of coconut-based agroforestry systems in Melanesia: 

A case study from the Vanuatu archipelago 

N. Lamanda, E. Malézieux, and P. Martin 

Diversity and dynamics in homegardens of southern Ethiopia 

Tesfaye Abebe, K.F. Wiersum, F. Bongers, and F. Sterck 

Homegarden plant diversity in relation to remoteness from urban centers: 

A case study from the Peruvian Amazon region 

A. Wezel and J. Ohl 

Gender and social dynamics in swidden and homegardens in Latin America 

P.L. Howard 

vii 

xi 

xiii 

1 

13 

25 

43 

61 

87 

105 

123 

143 

159

vi 

Section 3: Some New Thrust Areas 

Carbon sequestration potential of tropical homegardens 

B.M. Kumar 

Medicinal plants in tropical homegardens 

M.R. Rao and B.R. Rajeswara Rao 

Commercialization of homegardens in an Indonesian village: Vegetation 

composition and functional changes 

O.S. Abdoellah, H.Y. Hadikusumah, K. Takeuchi, S. Okubo, and Parikesit 

Transpiration characteristics of some homegarden tree species in Central 

Sri Lanka 

W.A.J.M. de Costa, K.S.P. Amaratunga, and R.S. Udumullage 

Ecology versus economics in tropical multistrata agroforests 

E. Torquebiau and E. Penot 

Financial analysis of homegardens: A case study from Kerala state, India 

S. Mohan, J.R.R. Alavalapati, and P.K.R. Nair 

Section 4: Future of Homegardens 

The role of homegardens in agroforestry development: Lessons 

from Tomé-Açu, a Japanese-Brazilian settlement in the Amazon 

M. Yamada and H.M.L. Osaqui 

Urban homegardens and allotment gardens for sustainable livelihoods: 

Management strategies and institutional environments 

A.W. Drescher, R.J. Holmer, and D.L. Iaquinta 

Are tropical homegardens sustainable? Some evidence from Central Sulawesi, 

Indonesia 

K. Kehlenbeck and B.L. Maass 

Whither Homegardens? 

P.K.R. Nair 

Subject Index 

CONTENTS 

185 

205 

233 

251 

269 

283 

299 

317 

339 

355 

371

LIST OF CONTRIBUTORS 

Abdoellah O.S. 

Institute of Ecology and Department of Anthropology, Padjadjaran University, 

Bandung, Indonesia; E-mail or 

Alavalapati J.R.R. 

School of Forest Resources and Conservation, Institute of Food and Agricultural 

Sciences, University of Florida, Gainesville, FL 32611, USA; E-mail 

Amaratunga K.S.P. 

Department of Crop Science, Faculty of Agriculture, University of Peradeniya, 

Peredeniya 20400, Sri Lanka; E-mail 

Bongers F. 

Forest Ecology and Management group, Wageningen University, The Netherlands; 

E-mail 

De Costa W.A.J.M. 


Peredeniya 20400, Sri Lanka; E-mail 

Drescher A.W. 

Albert-Ludwigs-Universität, Freiburg, Germany; E-mail 

Elevitch C.R. 

Agroforestry Net Inc., Holualoa, Hawai‘i 96725, USA; E-mail 

Guidal A. 

Forest and Nature Conservation Policy group, Wageningen University, The 

Netherlands (present address: GERES-CFSP #45 St.606, Toulkok, PO Box 2528, 

Phnom Penh-3, Cambodia);E-mail 

Hadikusumah H.Y. 

Institute of Ecology and Department of Biology, Padjadjaran University, Bandung, 

Indonesia; E-mail 

Holmer R.J. 

Xavier University College of Agriculture, Cagayan de Oro, The Philippines; E-mail 

 

Howard P.L. 

Department of Social Sciences, Wageningen University, Hollandseweg 1, 6706 KN 

Wageningen, the Netherlands; E-mail

viii 

LIST L OF CONTRIBUTORS 

Iaquinta D.L., 

Nebraska Wesleyan University, Lincoln, Nebraska, USA; E-mail 

Kehlenbeck K. 

Institute for Crop and Animal Production in the Tropics, Georg-August-University, 

Grisebachstr. 6, D-37077 Göttingen, Germany; E-mail 

Kennedy J. 

Research School of Pacific and Asian Studies, Australian National University, 

Canberra, Australia; E-mail 

Kumar B.M. 

College of Forestry, Kerala Agricultural University, Thrissur 680656, Kerala, India; 

E-mail 

Lamanda N. 

CIRAD UMR SYSTEM, TA 80/ 01, Avenue Agropolis, 34 398 Montpellier Cedex 

5, France; E-mail or 

Maass B.L. 

Institute for Crop and Animal Production in the Tropics, Georg-August-University, 

Göttingen, Grisebachstr. 6, D-37077 Göttingen, Germany; E-mail 

Malézieux E. 

CIRAD UMR SYSTEM, TA 80/ 01, Avenue Agropolis, 34 398 Montpellier Cedex 

5, France; E-mail 

Martin P. 

INA P-G département AGER, bâtiment EGER BP 01 78850 Thiverval-Grignon, 

France; E-mail 

Miller R.P. 

Instituto Olhar Etnográfico, SHIN CA 5 Conj. J Bl. B, Sala 105, Brasília-DF 71505, 

Brazil; E-mail 

Mohan S. 

CREST-RESSACA, Texas A&M University, MSC 213, 700 University Blvd, 

Kingsville, TX 78363, USA; E-mail 

Montagnini F. 

Yale University, School of Forestry and Environmental Studies, 370 Prospect St., 

New Haven, CT 06511, USA; E-mail

LIST L OF CONTRIBUTORS ix 

Nair P.K.R. 

School of Forest Resources and Conservation, Institute of Food and Agricultural 

Sciences, University of Florida, Gainesville, FL 32611, USA; E-mail 

Ohl J. 

School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, United 

Kingdom; E-mail 

Okubo S. 

Department of Ecosystem Studies, Graduate School of Agricultural and Life 

Sciences, University of Tokyo, Japan; E-mail 

Osaqui H.M.L. 

Division of International Environmental and Agricultural Science, Graduate School 

of Agriculture, Tokyo University of Agriculture and Technology, 3-5-8 Saiwaich, 

Fuch-shi, Tky 183-8509 Japan; E-mail 

Parikesit 

Institute of Ecology and Department of Biology, Padjadjaran University, Bandung, 

Indonesia; E-mail 

Penn J. W. Jr. 

Grand Valley State University, 1155 Au Sable Hall, Allendale, MI, 49401, USA; 

e-mail 

Penot E. 

CIRAD TERA, TA 60/ 15 – 34398 Montpellier CX5 – France; E-mail 

Peyre A. 


Netherlands (present address: 50 Avenue Henri GINOUX 92 120 Mont Rouge, 

France); E-mail or r 

Rajeswara Rao B.R. 

Central Institute of Medicinal and Aromatic Plants (CIMAP) Resource Centre, 

Boduppal, Uppal P.O. Hyderabad 500 039, India; E-mail . 

Rao M.R. 

Plot No. 11, ICRISAT Colony (Phase-I), Brig. Syed Road, Manovikasnagar (P.O.), 

Secunderabad–500 009, India; E-mail 

Sterck F. 

Forest Ecology and Management group, Wageningen University, The Netherlands; 

E-mail

x 

LIST L OF CONTRIBUTORS 

Takeuchi K. 

Department of Ecosystem Studies, Graduate School of Agricultural and Life 

Sciences, University of Tokyo, Japan 

Tesfaye Abebe 

Debub University, Awassa College of Agriculture, Ethiopia; E-mail 

Thaman R.R. 

The University of the South Pacific, Suva, Fiji i Islands; 1487; E-mail 

Torquebiau E., 

CIRAD TERA, TA 60/15 – 34398 Montpellier CX5, France; E-mail or 

Udumullage R.S. 


Peredeniya 20400, Sri Lanka 

van Leeuwen J. 

Instituto Nacional de Pesquisas da Amazônia – INPA, Manaus, Amazonas, Brazil; 

E-mail 

Wezel A. 

Institute of Landscape and Plant Ecology (320), University of Hohenheim, 70593 

Stuttgart, Germany; E-mail 

Wiersum K.F. 


Netherlands; E-mail 

Yamada M. 


of Agriculture, Tokyo University of Agriculture and Technology, 3-5-8 Saiwaich, 

Fuch-shi, Tky 183-8509 Japan; E-mail

CHAPTER REVIEWERS 

Abdoellah, O.S, Padjadjaran University, Bandung, Indonesia 

Allen, S.C., University of Florida, USA 

Becker, Brian, University of Florida, USA 

Bellow, John G., COAPS-Florida State University, USA 

Bourdeix, Roland, CIRAD, France 

Campilan, Dindo, CIP-Users’ Perspectives with Agricultural Research and Development, 

The Philippines 

Clement, C.R., National Research Institute for the Amazon (INPA), Brazil 

De Costa, W.A.J.M., University of Peradeniya, Sri Lanka 

De Zoysa, Mangala, University of Ruhuna, Kamburupitiya, Sri Lanka 

Depommier, Denis, CIRAD-Forêt, Montpellier, France 

Doolittle, A.A., Yale University, USA 

Fleischman, Forrest, Forest Service Employees for Environmental Ethics, Eugene, 

USA 

Geethakutty, P.S., Kerala Agricultural University, Thrissur, India 

Jose, S., University of Florida, USA 

Kallarackal, J., Kerala Forest Research Institute, India 

Maass, B.L., Georg-August-University, Goettingen, Germany 

Miller, R.P., Instituto Olhar Etnográfico, Brazil 

Mohan, S., Texas A&M University, USA 

Montagnini, Florencia, Yale University, USA 

Muraleedharan, P.K., Kerala Forest Research Institute, India 

Nair, V.D., University of Florida, USA 

Palada, Manuel C., Asian Vegetable Research and Development Centre, Taiwan 

Penot, Eric, CIRAD TERA, Montpellier, France 

Puri, S., Indira Gandhi Agricultural University, Raipur, India 

Rao, J.M., University of Florida, USA 

Rao, M.R., Secunderabad, India 

Russell, A.E., Iowa State University, USA 

Schroth, G., Conservation International, Washington, USA 

Torquebiau, E., CIRAD TERA, Montpellier, France 

Wiersum, K.F., Wageningen University, The Netherlands 

Yamada, Masaaki, Tokyo University of Agriculture and Technology, Japan

PREFACE 

Tropical homegardens are a topic of discussion in most agroforestry conferences 

especially those covering humid tropical lowlands, but publications on this topic are 

scattered in the literature; comprehensive books and reports focused on it are rare. 

The motivation for this book was the desire to address that deficiency, following a 

session on Tropical Homegardens at the 1st World Congress of Agroforestry, 

Orlando, Florida, USA in June – July 2004 (http://conference.ifas.ufl.edu/wca). The 

initial idea was to bring out a publication based on the presentations at the Congress 

session; but consequent to enthusiastic responses from the professional community, 

the scope of the book was broadened to make it more comprehensive than a 

conference publication. 

As it turned out, only five chapters out of the total 20 in the book are based on 

presentations at the above Congress session. Three chapters are adaptations from 

papers that have recently been published (or have been accepted for publication) in 

Agroforestry Systems journal on issues that are important from the point of 

comprehensiveness of the book. Seven of these eight chapters are research articles and 

are presented in the conventional research-publication format (Introduction, Materials 

and Methods, Results, and Discussion); they present a glimpse of the nature of current 

research in homegardens. All other chapters are review and synthesis of current state 

of knowledge on homegarden issues from all three developing continents (Africa, 

Asia, and Latin America & the Caribbean). The chapters are organized into five 

sections (Historical and Regional Perspectives; Structure, Function, and Dynamics; 

Some New Thrust Areas; and Future of Homegardens); each section contains a mix of 

research and review articles. We believe that these 20 chapters represent the state-ofthe-art 

of tropical homegardens today. 

The expeditious publication of the book would not have been possible without the 

cooperation and dedication of the authors and reviewers. All chapters were 

rigorously peer-reviewed. We thank the reviewers (see the list attached) for their 

insightful comments and critical suggestions, which helped to enhance the quality of 

the chapters. The authors too have been a very pleasant and professional group to 

work with; we greatly appreciate their cooperation and understanding in putting up 

with our requests for repeated revisions within very short and strict time schedules. 

Once again, we sincerely thank all the authors and reviewers for their splendid 

cooperation. Special thanks go to Dr. Michael Bannister, who did an excellent job of 

reading through the manuscripts and scrutinizing the literature citations. 

B. Mohan Kumar, Thrissur, Kerala, India 

February 2006 P. K. R. Nair, Gainesville, Florida, USA

CHAPTER 1 

INTRODUCTION 

P.K.R. NAIR 1 AND B.M. KUMAR 2 

1 School of Forest Resources and Conservation, University of Florida, Gainesville, 

FL 32611, USA; E-mail: . 2 College of Forestry, Kerala 

Agricultural University, Thrissur 680656, Kerala, India; 

E-mail: 

1. THE CONCEPT OF HOMEGARDEN 

It is rather customary that any writing on homegardens starts with a “definition” of 

the term. The first drafts of several chapters in this book were no exception. This 

indicates that there is no universally accepted “definition” of the term and therefore 

the authors feel compelled to make their perception clear. An examination of the 

various “definitions” used or suggested by various authors (of chapters of this book 

as well as other recent homegarden literature) shows that they all revolve around the 

basic concept that has been around for at least the past 20 years, i.e., since the “early 

literature” on the subject (Wiersum, 1982; Brownrigg, 1985; Fernandes and Nair, 

1986; Soemarwoto, 1987): homegardens represent intimate, multistory combinations 

of various trees and crops, sometimes in association with domestic animals, around 

the homestead. This concept has been developed around the rural settings and 

subsistence economy under which most homegardens exist(ed). But, as some 

chapters in this book describe, the practice of homegardening is now being extended 

to urban settings (Drescher et al., 2006; Thaman et al., 2006) as well as with a 

commercial orientation (Abdoellah et al., 2006; Yamada and Osaqui, 2006). 

Even before the advent of such new trends as urban and commercial homegardens, 

the lack of clear-cut distinctions between various stages in the continuum 

from shifting cultivation to high-intensity multistrata systems and the various terms 

used in different parts of the world to denote the different systems has often 

created confusion in the use of the term homegarden and its underlying concept. 

The confusion is compounded by the fact that in many parts of the world, especially 

1 

B.M. Kumar and P.K.R. Nair (eds.), Tropical Homegardens: A Time-Tested Example of 

Sustainable Agroforestry, 1-10. 

© 2006 Springer. Printed in the Netherlands.

2 P.K.R. NNAIR AND B.M. KUMAR 

in the New World, swidden farming such as the milpa of Mesoamerica evolve over a 

period of time into full-fledged homegardens consisting of mature fruit trees and 

various other types of woody perennials and the typical multistrata canopy 

configurations. In such situations, it is unclear where the swidden ends and 

homegarden begins – and often they co-exist. Yet another cause of confusion is the 

term itself: homegarden. Even for most agricultural professionals who are either not 

familiar with or are not appreciative of agroforestry practices, what we write as one 

word ‘homegarden’ sounds as two words ‘home’ and ‘garden’ sending the signal 

that the reference is to ornamental gardening around homes. While ornamentals are 

very much a part of homegardens in many societies, homegardens, in our concept, 

are not just home gardens of strictly ornamental nature. 

As we explained in our recent paper (Kumar and Nair, 2004), we use the term 

homegardens (and homegardening) to refer to farming systems variously described 

in English language as agroforestry homegardens, household or homestead farms, 

compound farms, backyard gardens, village forest gardens, dooryard gardens and 

house gardens. Some local names such as Talun-Kebun and Pekarangan that are 

used for various types of homegarden systems of Java (Indonesia), Shamba and 

Chagga in East Africa, and Huertos Familiares of Central America, have also 

attained international popularity because of the excellent examples of the systems 

they represent (Nair, 1993). In spite of the emergence of homegardening as a 

practice outside their “traditional” habitat into urban and commercial settings, the 

underlying concept of homegardens remains the same as before “intimate, 

multistory combinations of various trees and crops, sometimes in association with 

domestic animals, around homesteads.” Intimate plant associations of trees and 

crops and consequent multistory canopy configuration are essential to this concept. 

Equally important in this concept is the home around which most homegardens are 

maintained; but in some situations, multistory tree gardens (such as the Talun or 

Kebun of Indonesia: Wiersum, 1982) that are not in physical proximity to homes but 

receive the same level of constant attention from the owners’ household and have 

similar structural and functional attributes as other homegarden units located near 

homes are also considered as homegardens. 

2. GENESIS AND GLOBAL DISTRIBUTION OF HOMEGARDENS 

Tracing the history of homegardening, Kumar and Nair (2004) describe it as the oldest 

land use activity next only to shifting cultivation that has evolved through generations of 

gradual intensification of cropping in response to increasing human pressure and the 

corresponding shortage of arable lands. The Javanese homegardens of Indonesia and the 

Kerala homegardens of India – the two oft-cited examples – have reportedly evolved 

over centuries of cultural and biological transformations and they represent the 

accrued wisdom and insights of farmers who have interacted with environment, 

without access to exogenous inputs, capital, or scientific skills. Wiersum (2006) 

mentions that the origin of homegardening in Southeast Asia has been associated 

with fishing communities living in the moist tropical regions ca 13 000 to 9000 B.C. 

Implying the predominance of homegardens in ancient India, Vatsyayana in his 

great book of Hindu aesthetics – Kamasutra, written ca 300 to 400 AD, describe

INTRODUCTION 

I 

house gardens as a source of green vegetables, fig trees (Ficus spp.), mustard 

(Brassica spp.) and many other vegetables (c.f. Randhawa, 1980). Ibn Battuta in his 

travelogue (1325 – 1354) also wrote that the densely populated and intensively 

cultivated landscape with coconut (Cocos nucifera), black pepper (Piper nigrum), 

ginger (Zingiber officinale), sugarcane (Saccharum officinarum), pulses (grain 

legumes) and the like surrounding the houses formed a distinctive feature of the 

Malabar coast of Kerala (Randhawa, 1980). In both Java and Kerala, homegardening 

has been a way of life for centuries and is still critical to the local subsistence 

economy and food security (Kumar and Nair, 2004). This is true of several other Old 

World homegardens as well (e.g., the Chagga of Mt. Kilimanjaro in East Africa: 

Fernandes et al., 1984; Soini, 2005). 

In spite, or perhaps because, of the pre-historic origin of the practice, accurate 

data on the extent of area under homegardens are not available. Estimating the area 

of homegardens is beset with several problems (Kumar, 2006). A major one is 

the lack of distinct boundaries or demarcation between homegardens and other 

cultivated agricultural fields. As Tesfaye Abebe et al. (2006) point out; most 

homegardens studies are focused on gardens that constitute a component of a 

farming system consisting of cultivated fields away from homes complemented by 

the homegardens surrounding residential houses. In those situations, it is difficult to 

determine where homegardens end and other cultivated fields begin. Added to this 

problem is the “commodity-centric” approach to recording land use statistics: 

statistics are prepared and presented for specific (single) crops and commodities. In 

most cases, the area is listed under the most conspicuous or visible crop (e.g., fruit 

trees, coconut palms, and other trees that occupy the upper stratum of multistoried 

homegarden system) and the lower-story crops are seldom reported – and, often the 

reporting forms do not allow entries to be made of such mixed stands. Thus, 

homegardens are a “non-entity” for agricultural statistics and land revenue records. 

In spite of these difficulties, some efforts have been made in compiling statistics 

on the spread of homegardens. Such estimates include 5.13 million ha of land under 

pekarangans in Indonesia, 0.54 million ha under homesteads in Bangladesh, 1.05 

million ha in Sri Lanka, and 1.44 million ha in Kerala, India (Kumar, 2006). 

Christanty (1990) reported that more than 70% of all households in the Philippines 

maintained homegardens; but the extent of area occupied by them was not reported. 

Area statistics of homegardens are also not available from a number of other parts of 

the world although the prevalence of the practice – indeed predominance in many 

situations – has been reported from various parts of the tropics as several chapters in 

this volume also attest to. In an attempt to present a global distribution of 

homegardens, we selected 135 entries from the CABI Abstracts for the period from 

1990 to 2003 for which geographical locations are either mentioned or can be 

deduced; these included: Africa 21, Europe (Catalonia, Austria, etc.) 10, Central and 

South America 23, South Asia 45, Southeast Asia 30, other parts of Asia 2, Pacific 

islands 4. Based on these reports, supplemented with available statistics from other 

sources (e.g., reports on agricultural censuses) as well as personal experiences and 

observations of the authors, we have attempted a “Homegarden Map of the World” 

as presented in Fig. 1. The presentation only means that homegardens are present in 

3

4 

P.K.R. NNAIR AND B.M. KUMAR 

Explanation of Figure 1. 

The global distribution of homegardens. This attempt is based on the geographical 

distribution of 135 selected studies (the specific geographical locations of which are reported 

or can be deduced) from the CABI abstracts for the period from 1990 to 2003, including 

Africa (21 studies), Europe: Catalonia, Austria, and others (10), Central and South America 

(23), South Asia (45), Southeast Asia (30), other parts of Asia (2), and Pacific Islands (4), 

supplemented with available statistics from other sources (e.g., reports on agricultural 

censuses) and authors’ experiences/observations. Differing shade intensities in the figure 

represent high, moderate, and low frequency of occurrence of homegardens. We have used 

‘High’ for areas where the frequency of occurrence in the CABI abstracts is more than 20 

and/or if other databases (Statistical Yearbook 2000, Bangladesh Bureau of Statistics; 

Statistical Yearbook of Indonesia 2000, Badan Pusat Statistik; Census of Agriculture – Sri 

Lanka 2002. Agricultural holdings, extent under major crops and livestock statistics by 

district and DS/AGA division—based on operator’s residence: small holding sector, 

Colombo; Land Resources of Kerala State 1995, Kerala State Land Use Board; see Kumar, 

2006 for full citations) report that more than 50% of all households maintain homegardens, 

‘Medium’ for 10 to 20 mentions in CABI abstracts or 25 to 50% of the households maintain 

homegardens according to the other reports listed above, and ‘Low’ for all those cases where 

presence of homegardens has been reported in one or more ways but at levels below the above 

limits. “Apparently present” is the term used to denote regions where homegardens are said to 

be abundant based on the authors’ personal observations and/or communications from other 

sources, but on which published (accessible) information, especially on their area statistics, is 

limited or absent; such regions include tropical and subtropical parts of China, and some such 

other regions in Asia and Africa. The presentation only means that homegardens are present 

in the regions as indicated; it does not imply that homegardens are the only or the major land 

use system in any of these regions.

Figure 1. The global distribution of homegardens (see description on the left hand side, p. 4). 

INTRODUCTION 

II 

5

6 


the regions as indicated; it does not imply that homegardens are the only or major 

land use system in any of these regions. 

Based on the above, it is reasonable to assume that homegardens are most 

popular in the tropics, but can also be found between 40 o N and 30 o S latitudes. 

South- and Southeast Asia, the Pacific islands, East- and West Africa, and 

Mesoamerica are the regions where largest concentrations of homegardens can be 

found. Homegardens are also reportedly very popular in tropical and subtropical 

parts of China; however, other than general descriptions of the systems (e.g., 

Zhaohua et al., 1991; Wenhua, 2001), practically no information could be gathered 

on their area statistics. The Mediterranean region of Catalonia (Agelet et al., 2000) 

and southern Africa (High and Shackleton, 2000) also are reported to have 

homegardens. In terms of ecological distribution, the highest concentrations of 

homegardens are in the humid and subhumid tropics, but they are also common in 

other ecological regions, especially the tropical highlands of Asia, Africa, and 

Mesoamerica (Nair, 1989). Clearly, our understanding about the spread of homegardens 

is incomplete; more efforts are needed to compile these statistics at local, 

regional, national, and global levels. 

Although homegardens are known as a predominantly tropical ‘phenomenon’, 

homegardening – or, conceptually similar practices – exist outside the tropical zone 

as well. For instance, Gold and Hannover (1987) and Herzog (1998) describe fruittree 

based agroforestry systems in North America and Europe, respectively. Vogl 

and Vogl-Lukasser (2003) reported that homegardens were typical elements of the 

mosaic of agroecosystems in the mountainous Alpine region of Austria. Streuobst 

(fruit trees grown on agricultural lands with crops or pasture as understorey), a 

traditional practice in Europe that has been on the decline since around 1930s, is 

now receiving increasing attention and acceptance among the general public and 

promoted by nongovernmental and conservation agencies. Although the fruit-tree 

based agroforestry systems are strictly not homegardening, such systems occasionally 

involve homegardening, and their socio-cultural, ecological, and aesthetic values often 

exceed their economic values. Based on an extensive survey and interview with 

practitioners of African-American gardening traditions in the rural southern United 

States, Westmacott (1992) traced the principal functions and features of African- 

American yards and gardens. During slavery, the gardens were used primarily to 

grow life-sustaining crops and vegetables, and the yard of a crowded cabin was 

often the only place where the slave family could assert some measure of 

independence and perhaps find some degree of spiritual refreshment. Since slavery, 

working the garden for the survival of the family has become less urgent, but there 

seems to be a revival of appreciation of their recreational, social, and other uses. 

For example, the gardeners are now finding pleasure in growing flowers and 

produce and deriving satisfaction from agrarian life-style, self-reliance, and private 

ownership. Through historical research, field observations, and oral interviews, 

Westmacott (1992) traces the West African roots of this gardening tradition and 

elucidates how the African-American community manipulated the garden space to 

their best advantage – something very similar to the motivations of subsistence 

gardeners in well-established homegardens in other parts of the world (Fig. 1).

INTRODUCTION 

I 

Related to the above-mentioned “African-American Yards and Gardens” of the 

southern United States is the increasing interest in hobby farming and weekend 

gardening that is getting popular in many urban and rapidly urbanizing societies in 

both industrialized and developing nations. Drescher et al. (2006) describe the urban 

homegardens and some of the operational and institutional issues related to them 

from a number of locations around the world. In a survey of agroforestry practices 

and opportunities in southeastern United States, Workman et al. (2003) identified 

several “special applications” of agroforestry such as use of fruit trees combined 

with gardens, ponds, and as bee forage and so-called patio gardens as an increasingly 

popular activity especially among immigrant Latin American communities. Thus, 

although homegardening as a major land use practice is most widespread in thickly 

populated tropical regions, the concept is being adopted in other geographical 

regions as well to a limited extent. 

3. COMPLEXITY OF HOMEGARDENS 

Species diversity is one factor that is common to all homegardens, and this point has 

been well brought out in homegarden literature time and again. Indeed, authors tend 

to get nostalgic about describing how diverse the plant communities in homegardens 

are and rather adamant about including elaborate species lists in their papers on 

homegardens to the extent that many seem to consider that a paper on any aspect of 

homegarden is incomplete without a species list! Interestingly, most of the plants 

that are listed in most such publications are the same irrespective of the geographical 

regions from where they are reported (see Nair, 2006). As various analyses and 

summary reports have repeatedly indicated (e.g., Kumar and Nair, 2004), food 

plants (food crops and fruit trees) are the most common species in most homegardens 

throughout the world. This underscores the fact that food- and nutritional 

security is the primary role of homegardens – again, a point well recognized in 

homegarden literature right from the “early” years (e.g., Brownrigg, 1985; Fernandes 

and Nair, 1986). Next in importance to food crops are cash crops, and with 

increasing trend toward commercialization, the interest in such crops is likely to 

only increase. 

We recognize that complexity by itself may not be a desirable attribute in land 

use systems that are (also) expected to fulfill production objectives. Being located 

on the “prime land” around homesteads and receiving utmost managerial attention 

of the homeowners all the time, farmers have high expectations of productivity from 

homegardens. After all, farmers decide on the species to be planted and retained in 

the homegardens based on the utilitarian value of the species. Species complexity in 

homegardens is therefore not a natural phenomenon, but a result of deliberate 

attempts and meticulous selection and management by farmers to provide the 

products they consider are important for their subsistence and livelihood. Species 

complexity in homegardens is thus a manmade feature, unlike in natural systems. 

This distinction is seldom recognized in comparisons involving ecological indices of 

species diversity of homegardens, several of which have lately been reported (see 

Nair, 2006). 

7

8 


Furthermore, it is likely that the extreme structural complexity and diversity may 

be a “bane” of the homegardens in a sense. Each homegarden is a unique land use 

entity in terms of component arrangement, organization, and management, and it 

reflects the personal preferences of its owner. This frustrates the development 

community that seeks out “replicable models”; this is presumably the main reason 

why homegardens have not received adequate attention in the development paradigms 

around the world. 

4. HOMEGARDENS IN THE CONTEXT OF CONTEMPORARY LAND USE 

ISSUES 

Today land use systems are challenged as never before with mounting concerns of 

environment and ethics on the one hand and pressures of economic development on 

the other. Production and economic issues that reigned supreme as ultimate goals in 

agricultural and forestry development activities during the past few decades are 

slowly yielding to environmental, societal, and social issues. Sustainability – 

meeting today’s needs without compromising the ability of future generations to 

satisfy their needs – is a key issue in all land use activities today. Central to this 

concept is the urge to achieve a balance between ecological preservation, economic 

vitality, and social justice. Land use systems today are thus evaluated based not only 

on their ability to fulfill any single objective such as production of a preferred 

commodity, but also on how best they fulfill the sustainability criteria. Contemporary 

issues that dominate the discussions in this context include natural-resource use in 

perpetuity, biodiversity conservation, gender equity, social justice, environmental 

integrity, appreciation of indigenous knowledge, preservation of cultural heritage, 

and so on. 

While systematic studies on the role of homegardens in many of these 

contemporary issues have not been done, there is a long-held belief and intuition that 

homegardens score very high on most – perhaps all – of these so-called “intangible” 

benefits. Logic, circumstantial evidences, and limited empirical results that are 

available support these conjectures; but certainly more convincing evidence based 

on rigorous research is needed. Several chapters in this book point in this direction 

and provide the framework for formulating future research plans. 

REFERENCES 

Abdoellah O.S., Hadikusumah H.Y., Takeuchi K., Okubo S. and Parikesit. 2006. 

Commercialization of homegardens in an Indonesian village: vegetation composition 

and functional changes. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: 

A time-tested example of sustainable agroforestry, pp 233 – 250. Springer Science, 

Dordrecht. 

Agelet A., Angels B.M. and Valles J. 2000. Homegardens and their role as a main source of 

medicinal plants in mountain regions of Catalonia (Iberian Peninsula). Econ Bot 54: 

295 – 309. 

Brownrigg L. 1985. Home Gardening in International Development: What the literature 

shows. The League for International Food Education, Washington, DC, 330p.

INTRODUCTION 

I 

Christanty L. 1990. Homegardens in tropical Asia with special reference to Indonesia. 

In: Landauer K. and Brazil M. (eds), Tropical home gardens, pp 9 – 20. United Nations 

University Press, Tokyo. 

Drescher A.W., Holmer R.J. and Iaquinta D.L. 2006. Urban homegardens and allotment 

gardens for sustainable livelihoods: management strategies and institutional environments. 

In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 

example of sustainable agroforestry, pp 317 – 338. Springer Science, Dordrecht. 

Fernandes E.C.M. and Nair P.K.R. 1986. An evaluation of the structure and function of 

tropical homegardens. Agric Syst 21: 279 – 310. 

Fernandes E.C.M., O’Kting’ati A. and Maghembe J. 1984. Chagga homegardens: a multistory 

agroforestry cropping system on Mt. Kilimanjaro, northern Tanzania. Agroforest Syst 2: 

73 – 86. 

Gold M.A. and Hanover J.W. 1987. Agroforestry systems of the temperate zone. Agroforest 

Syst 5: 109 – 21. 

Herzog F. 1998. Streuobst: a traditional agroforestry system as a model for agroforestry 

development in temperate Europe. Agroforest Syst 42: 61 – 80. 

High C. and Shackleton C.M. 2000. The comparative value of wild and domestic plants 

in homegardens of a South African rural village. Agroforest Syst 48: 141 – 156. 

Kumar B.M. 2006. Carbon sequestration potential of tropical homegardens. In: Kumar B.M. 

and Nair P.K.R. (eds), Tropical homegardens: A time-tested example of sustainable 

agroforestry, pp 185 – 204. Springer Science, Dordrecht. 

Kumar B.M. and Nair P.K.R. 2004. The enigma of tropical homegardens. Agroforest Syst 61: 

135 – 152. 

Nair P.K.R. (ed.). 1989. Agroforestry systems in the tropics. Kluwer, Dordrecht, 664p. 

Nair P.K.R. 1993. An introduction to agroforestry. Kluwer, Dordrecht, 499p. 

Nair P.K.R. 2006. Wither homegardens? In: Kumar B.M. and Nair P.K.R. (eds), Tropical 

homegardens: A time-tested example of sustainable agroforestry, pp 355 – 370. Springer 

Science, Dordrecht. 

Randhawa M.S. 1980. The history of Indian agriculture, vol. 2, pp 67 – 68 and 414 – 415. 

Indian Council of Agricultural Research, New Delhi. 

Soemarwoto O. 1987. Homegardens: a traditional agroforestry system with a promising 

future. In: Steppler H.A. and Nair P.K.R. (eds), Agroforestry: A decade of development, 

pp 157 – 170. ICRAF, Nairobi. 

Soini E. 2005. Changing livelihoods on the slopes of Mt. Kilimanjaro, Tanzania: challenges 

and opportunities in the Chagga homegarden system. Agroforest Syst 64: 157 – 167. 

Tesfaye Abebe, Wiersum, K.F., Bongers, F. and Sterck, F. 2006. Diversity and dynamics 

in homegardens of southern Ethiopia. In: Kumar B.M. and Nair P.K.R. (eds), Tropical 



Thaman R.R., Elevitch C.R. and Kennedy J. 2006. Urban and homegarden agroforestry in 

the Pacific islands: current status and future prospects. In: Kumar B.M. and Nair P.K.R. 

(eds), Tropical homegardens: A time-tested example of sustainable agroforestry, pp 25 – 

41. Springer Science, Dordrecht. 

Vogl C.R. and Vogl-Lukasser B. 2003. Tradition, dynamics and sustainability of plant species 

composition and management in homegardens on organic and non-organic small scale 

farms in Alpine Eastern Tyrol, Austria. Biol Agric Hortic 21: 349 – 366. 

Wenhua L. (ed.). 2001. Integrated farming systems at different scales. In: Agro-ecological 

farming systems in China, Chapter 12, pp 201 – 252. UNESCO Man and Biosphere Series 

26, Partheon Publishing, New York. 

Westmacott R.N. 1992. African-American gardens and yards in the rural south. University 

of Tennessee Press, Knoxville, TN, 198p. 

9

10 


Wiersum K.F. 1982. Tree gardening and taungya in Java: Examples of agroforestry 

techniques in the humid tropics. Agroforest Syst 1: 53 – 70. 

Wiersum K.F. 2006. Diversity and change in homegarden cultivation in Indonesia. In: Kumar 

B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested example of sustainable 


Workman S.W., Bannister M.E. and Nair P.K.R. 2003. Agroforestry potential in the southeastern 

United States: Perceptions of landowners and extension professionals. Agroforest 

Syst 59: 73 – 83. 

Yamada M. and Osaqui H.M.L. 2006. The role of homegardens for agroforestry development: 

Lessons from Tomé-Açu, a Japanese-Brazilian settlement in the Amazon. In: Kumar B.M. 



Zhaohua Z., Mantang C., Shiji W. and Youxu J. (eds). 1991. Agroforestry systems in China. 

Chinese Academy of Forestry, Beijing, and International Development Research Centre, 

Singapore, 216p.

SECTION 1 

HISTORICAL AND REGIONAL 

PERSPECTIVES

CHAPTER 2 

DIVERSITY AND CHANGE 

IN HOMEGARDEN CULTIVATION 

IN INDONESIA 

K.F. WIERSUM 

Forest and Nature Conservation Policy group, Department of Environmental 

Sciences, Wageningen University, The Netherlands; 

E-mail: 

Keywords: Homegarden dynamics, Rural transformations, Social sustainability. 

Abstract. Homegardens have been described as traditional agroforestry systems that are 

ecologically and socially sustainable. The concept of social sustainability has two dimensions: 

positive role to present livelihood conditions and ability to respond to socioeconomic 

changes. The dynamics of homegardens and its repercussions on social sustainability have 

received relatively little research attention. On the basis of results of extensive studies in Java 

and other parts of Indonesia, this article summarizes the historic and recent developments in 

the homegardening context. The structure and composition of homegardens depend both on 

their position in the overall farming system and on livelihood strategies of the managers. 

Rural transformations result in changes in livelihoods and farming systems, and have impacts 

on homegarden function and composition. The opinions of various authors on homegarden 

dynamics range from positive to negative; the former consider that changes in homegarden 

features are associated with socio-professional changes of villagers and the rural-urban 

interface, while the latter view these changes as indicative of the demise of a traditional 

system and argue for its revitalization. These different opinions represent different norms in 

assessing social sustainability of homegardens and differences in value judgments on the ideal 

structure of homegardens. 

1. INTRODUCTION 

Homegardening has been hypothesized as being the oldest form of agriculture in 

Southeast Asia. Its origin has been associated with fishing communities living in the 

13 


Sustainable Agroforestry, 13–24. 


14 K.F. WWIERSUM 

moist tropical region of Southeast Asia during 13 000 to 9000 B.C. In these regions 

an assured supply of fish and shells allowed fixed settlements and a relatively high 

population density, while the fertile soils along rivers and coasts favored cultivation 

(Sauer, 1969). As happened also in other regions (Miller et al., 2006), homegardening 

probably started as a spontaneous growth of plants from leftovers of 

products brought to the camps of the hunter/gatherers. Gradually, the accidental 

propagation became more deliberate with valuable species being planted to facilitate 

their use. At first such cultivation probably involved vegetative propagation techniques 

and only later seeding was introduced (Sauer, 1969). The earliest evidence of garden 

cultivation dates back to at least 3000 B.C. (Soemarwoto, 1987). 

From these pre-historic and probably scattered origins, homegardens has 

gradually spread to many humid regions in South- and Southeast Asia including 

Java (Indonesia), the Philippines, Thailand, Sri Lanka, India and Bangladesh. For 

instance, according to Randhawa (1980), travelers already described homegardens 

with coconut (Cocos nucifera), black pepper (Piper nigrum), ginger (Zingiber 

officinale), sugarcane (Saccharum officinarum) and pulses (grain legumes) in 

Kerala, India, in the early 14th century, while Michon (1983) mentions that tree 

gardening systems were already common on the Indonesian island of Java in the 

tenth century AD. In all these regions, homegardening is almost always practiced in 

combination with other types of land use. The original association with gathering 

and fishing was gradually extended to shifting cultivation and permanent cropping. 

In the most widely studied homegarden systems in South- and Southeast Asia such 

as in Java (Soemarwoto, 1987), Kerala (Nair and Sreedharan, 1986; Kumar et al., 

1994), and Sri Lanka (Jacob and Alles, 1987; McConnell, 1992), gardening is 

combined with permanent field cultivation often in the form of wetland rice (Oryza 

sativa) production. These regions with good farming conditions and relatively high 

population densities contributed to optimal development of the complementary 

system of staple food cultivation in open fields and supplementary diversified 

homegarden production for self-sufficiency and trade. 

Since the recognition of agroforestry as a type of land use worthy of research and 

development, homegardens have been considered as an excellent example of a 

traditionally developed agroforestry system with good promise for the future 

(Soemarwoto, 1984; Hochegger, 1998; Gajaseni and Gajaseni, 1999). Much attention 

has been given to analyzing the structure and function of tropical homegardens and 

describing their features in respect to both ecological and socioeconomic sustainability 

(Torquebiau, 1992; Kumar and Nair, 2004). Regarding socioeconomic 

sustainability, these studies focused specifically on the roles of homegardens within 

the livelihood systems of rural producers. A commonly perceived indicator of 

homegardens’ socioeconomic sustainability is the fact that homegardens typically 

contribute towards nutritional security, energy needs and income generation even 

under conditions of high population densities (Kumar and Nair, 2004). Recently it 

has been remarked, however, that the concept of socioeconomic sustainability 

should not only be related to the homegardens’ function in the present livelihood 

conditions, but also to their ability to adjust to socioeconomic changes (Peyre et al., 

2006). At present, many rural areas are undergoing major transformations involving 

diversification of rural livelihood strategies (Ellis, 1998; Ashley and Maxwell,

DIVERSITY AND CHANGE C IN HOMEGARDEN CULTIVATION IN INDONESIA 

I 

2001). Due to commercialization, cultivation systems are becoming more specialized 

on the one hand, and rural people are increasingly employed in non-primary production 

activities on the other. As a result, in many rural areas, farming systems in 

general, and homegardens in particular, are changing. Kumar and Nair (2004) have 

even posed the question as to whether homegardens are becoming extinct. This 

illustrates that the notion of socioeconomic sustainability of homegardens should be 

interpreted as referring not only to their ability to contribute towards the livelihood 

needs of traditional rural dwellers, but also to their ability to adjust to the process of 

rural change. 

In contrast to studies on homegarden diversity, relatively little attention has been 

given to assessing the dynamics of homegardens. It seems that, since many studies 

in the past have been focused on ascertaining factors that explain the ecological 

stability of homegardens (Kumar and Nair, 2004), the concept of sustainability has 

mainly been attributed as referring to stability in an ecological sense, and that the 

concept of socioeconomic sustainability was by association interpreted as referring 

to livelihood stability. Only recently have the dynamics of homegardens been 

receiving some attention. In some studies, the traditional homegarden structure and 

composition is taken as ideal, and changes such as loss in some of the traditional 

species and structure are discussed in terms of homegardens becoming extinct 

(Kumar and Nair, 2004) and needing revitalization (Parikesit et al., 2004), while 

some other studies have tried to relate the various types of dynamics in homegarden 

structure and composition to the process of rural transformations (Michon and Mary, 

1994; Peyre et al., 2006). 

This review will assess the dynamics of homegarden development in Indonesia, 

focusing specifically on Java. First, it will describe the historic developments of 

homegardens on Java. Next, using data from both Java and Sulawesi, it will 

summarize the factors that impact on the structure and composition of homegardens 

and describe how under the influence of these factors different types of homegardens 

have evolved. On the basis of these data, the main trends in changing homegarden 

structure and composition will be summarized. 

2. THE DEVELOPMENT OF HOMEGARDENS IN JAVA 

The first studies on tropical homegardens in Southeast Asia that were started in the 

late 1940s in Java, Indonesia (Terra, 1953a; 1953b) remained relatively unnoticed 

for several years. For example, even in the 1970s it was noted that, in contrast to the 

open-field land use systems, homegardens had hardly yet been subject to detailed 

study (Stoler, 1978). This situation changed in the late 1970s when a series of new 

homegarden studies were initiated in Java (Soemarwoto, 1987; Soemarwoto and 

Conway, 1991). The Javanese experiences formed an important source of information 

when in the 1980s the potential of homegardens to contribute towards 

increasing food production and reducing malnutrition in tropical countries received 

greater international interest (Niñez, 1984; Brownrigg, 1985). This international 

interest in homegardens was further stimulated by the recognition of homegardens 

as a typical example of a multistoried agroforestry system (Nair and Sreedharan, 

1986; Jacob and Alles, 1987). The first international conference on tropical 

15

16 

K.F. WIERSUM W 

homegardens organized in Java in 1985 (Landauer and Brazil, 1990) is a testament 

to the leading role of the homegarden research in Java during that period. 

The extensive research on Javanese homegardens has contributed significantly to 

the present understanding of the structure and function of tropical homegardens. The 

Javanese homegardens demonstrate the typical functions of homegardens as 

summarized by Kumar and Nair (2004): they yield products with high nutritional 

value (proteins, vitamins, and minerals), medicinal plants and spices, firewood, and 

sometimes a1so forage crops and construction wood; all these products are used to 

supplement the staple food crops that are usually produced in open-field cultivation 

systems. Normally, the homegarden products provide a small, continuous flow of 

these supplementary products for subsistence and a possible small surplus for sale 

through local markets. In times of sudden necessities (unfavorable climatic conditions or 

social necessities like marriage), higher production and marketing levels may be 

attained (Wiersum, 1982). 

In many homegarden studies (Kumar and Nair, 2004), these gardens have been 

described as a distinct agroforestry system with a set of generic features. Relatively 

little attention has been given to studying the diversity within homegardens as well 

as their relation to the surrounding land use systems. Moreover, in addition to home- 

gardens, other types of tree gardening systems consisting of a mixture of several 

cultivated fruit- and other trees and crops exist (Wiersum, 2004), and the distinction 

between homegardens and other types of tree gardening systems is not straightforward. 

In Java, Terra (1953a; 1953b) originally differentiated three different types 

(see also Wiersum, 1982; Soemarwoto, 1984; Christanty et al., 1986): 

• The homegarden (pekarangan): fenced-in gardens, surrounding individual 

houses, planted with fruit- and other trees, vegetable herbs and annual crops. 

Historically they are associated with wetland rice fields and more recently also 

with dry fields. They occurred in regions with individual land-ownership. 

Typically these homegardens occur in Central Java and are inhabited by the 

Javanese people. 

• The tree garden (kebun or talun): mixed tree plantations on communal lands 

surrounding villages with dense clusters of houses, sometimes also at some 

distance from the villages. These plots are not inhabited and they are historically 

associated with shifting cultivation. They occur in regions with communally 

owned land. Mostly they are found in West Java and are inhabited by the 

Sundanese people. These tree gardens are much less tended than homegardens 

and often include more wild trees than present in the homegardens. 

• Clumps of fruit- or other trees planted on abandoned shifting cultivation sites. 

Such plantings could denote a right of priority of these lands for the people who 

planted the trees in an area of otherwise communal land ownership. 

As demonstrated by the characterizations, the tree gardening systems in Java 

normally forms a sub-set of an integrated farming system (Terra, 1958), which also 

comprises annually cultivated fields used for the production of staple, high calorific 

foods such as rice, maize (Zea mays) and cassava (Manihot esculenta). Consequently, 

the structure and function of homegardens significantly depends on the nature of the 

overall farming system.


I 

Over the ages, gradual changes have taken place in these systems (Soemarwoto, 

1984). The most important was perhaps the extension of the Javanese culture and 

subsequent spread of homegardens. For instance, in the eighteenth century, the 

pekarangan system was already practiced in West Java, where it partly replaced the 

talun system of the Sundanese (Michon, 1983). Also, gradually communal lands 

were divided among individual landowners, who by building houses in such individual 

tree gardens, converted them to homegardens. In other tree gardens, annual 

crops were introduced and management became more intensive. Also shifting 

cultivation virtually disappeared and in areas with clumps of planted trees on 

fallow lands, a conversion to tree gardens took place. According to Wiersum (1982), 

in the early 1980s it was possible to distinguish the following three types of tree 

gardening: 

• Homegardens (pekarangan): a land use form on private lands surrounding individual 

houses with a definite fence, in which several tree species are cultivated 

together with annual and perennial crops, often including small livestock. 

• Mixed gardens (kebun campuran): a land use form on private lands outside the 

village, which is dominated by planted perennial crops, mostly trees, under 

which annual crops are cultivated. 

• Forest gardens (talun, kebun): a land use form on private lands outside the 

village in which planted and sometimes spontaneously grown trees and sometimes 

additional perennial crops occur. 

The pekarangan is often considered as a typical prototype for homegardens. But 

as illustrated by the diversity of tree gardening system in Java, the distinction 

between homegardens and other types of tree-gardening systems is often diffuse and 

may be related more to location than to vegetation structure 1 . Moreover, homegarden 

structure may gradually change with time. 

3. DIVERSITY IN HOMEGARDEN STRUCTURE AND COMPOSITION 

The diversity in tropical homegardens types is not only illustrated by the historic 

developments in tree gardening systems, but also by the existing variation in 

homegarden structure and composition. Several homegarden studies in Java have 

assessed what factors impact on the homegarden structure and composition as well 

as function. Karyono (1990) demonstrated that homegarden composition was 

affected both by geographic conditions and their role in the farming systems. 

Compared to lowlands, homegardens in highland areas have lower plant diversity 

and simpler species composition. Also a different pattern of species composition 

exists in homegardens associated with irrigated rice production as opposed to those 

associated with dry-land agriculture: fruit species are dominant in the former, and 

food crops in the latter. Stoler (1978) also emphasized the relation between garden 

composition (as well as management intensity) and other components of the farming 

system. Households with sufficient croplands to produce rice to cover basic staple 

food requirements cultivated more commercial fruit trees than households who 

could not meet staple food requirements from croplands and hence had to cultivate 

more subsistence crops in the homegardens. Christanty (1990) differentiated urban 

17

18 


and rural homegardens, and mentioned that these could be further classified 

depending on: 

• The dominant plant species grown, e.g., fruit, vegetable, or flower species, and 

• The main function of the homegarden, e.g., subsistence garden, kitchen garden, 

market garden, plant nursery garden, and aesthetic garden. 

Soemarwoto (1984) added that in rural areas homegardens have important social 

functions through the provision of gifts in the form of fruits, leaves or products for 

religious or medicinal purposes. In urban areas this social function diminishes 

whereas their aesthetic function increases with ornamentals replacing food crops. 

Michon and Mary (1994) and Abdoellah et al. (2006) described that, in addition to 

urbanization, the rise of a market economy profoundly influences the homegarden 

function resulting in an increase in commercial crops. Abdoellah (1990) reported 

that the effect of various cultures (Javanese or Sundanese) was often still reflected in 

the structure of homegardens: for example, vegetables and ornamentals were often 

more common in Sundanese homegardens. 

Also in the Indonesian island of Sulawesi different types of homegardens have 

been reported. For example, Kehlenbeck and Maass (2004) described four homegarden 

types distinguished by differences in garden age and size, and the level of 

diversity: 

1. Small, moderately old, species- and tree-poor spice gardens 

2. Medium-sized, old, species-rich fruit tree gardens 

3. Large, rather young, species- and tree-poor gardens of transmigrant 

families 

4. Diverse assemblages of rather old, individual gardens with very high crop 

diversity. 

According to Terra (1958), the typical Javanese landscape with irrigated rice 

fields, dry croplands and mixed gardens was already common in this region in the 

1950s. The types 2 and 4 mentioned above may reflect this traditional situation. But 

as illustrated by type 3, recently the area is becoming further settled by transmigrants 

from Java. These transmigrants do not only open up new agricultural lands, 

but also establish homegardens around their new settlements. Such homegarden 

development takes time. Often, at first essential food crops are grown and only 

gradually supplementary crops are introduced. Other factors influencing homegarden 

structure are related to differences in access to markets and availability of 

garden products in the market. Moreover, the composition is found to be influenced 

by official homegarden development programs (Kehlenbeck and Maass, 2006). 

In other studies on Asian homegardens too, several geographic and socioeconomic 

factors have been found to influence the homegarden structure and 

composition (e.g., Kumar et al., 1994; John and Nair, 1999; Peyre et al., 2006). 

Table 1 summarizes the various factors that have been reported to impact on 

homegarden composition. As illustrated in this table, notably livelihood conditions 

are an important factor influencing the structure and composition of homegardens. 

Livelihood conditions are reflected in both the farming system and the 

socioeconomic status of households. For poor people, homegardens may form the 

only land available to them for primary production, and consequently they are likely

to serve partly for production of essential staple foods rather than only for supplementary 

crop production. On the other hand, for affluent people living in urbanized 

areas and having access to non-farm incomes, homegardens may not any longer 

form a part of a farming system, but function only as an ornamental area around the 

living quarters. Thus, not only the overall livelihood conditions, but also specific 

socioeconomic variables such as access to land or off-farm labor opportunities 

impact the homegarden structure and composition. Generally, a decrease in the 

availability of land results in intensification of cultivation and the inclusion of more 

annual crops. Also, when alternative income opportunities are present, cultivation is 

“extensified” (and more ornamentals are included near urban areas). Where better 

marketing opportunities exist (near cities), specialization in fruit production may 

take place. 

Table 1. Factors impacting structure and composition of homegardens with special reference 

to Indonesian homegardens. 

Factors Conditions Examples and remarks 

Geographic 

location 

Environmental 

conditions 

Role in farming 

systems 

Socioeconomic 

conditions of the 

household 


I 

Urban versus rural Urban homegardens often smaller and more 

location 

aesthetic oriented 

Climate conditions Variation in annual crops cultivated only in 

favorable climatic seasons is mostly less pronounnced 

than in permanent crops that have to be 

adapted to variable climatic conditions over much 

larger periods 

Soil conditions With decreasing soil fertility crop diversity tends 

to decrease and the effect of competition by trees 

on understorey becomes more pronounced. Dense 

tree gardens occur mostly on volcanic soils, while 

on tertiary soils tree gardens are more open 

Degree of complementarity 

to open field 

cultivation systems 

Established versus 

incipient farming system 

19 

If homegardens are the only land asset more 

inclusion of staple food crops 

Incipient gardens first dominated by annual crops, 

with time increased incorporation tree crops 

Wealth status With increased wealth increased importance of 

commercial and aesthetic plants 

Access to markets Commercial crops stimulated by good market 

access 

Access to off-farm In case of access to financially lucrative employ- 

employment 

ment decreased importance commercial crops 

Gender-related issues Gardens of female-headed households often more 

household use oriented 

Cultural factors Food preferences Cultural preferences in respect to consumption of 

vegetables and spices

20 

Up to a certain level, the cultivation of homegardens can respond well to changes 

in socioeconomic conditions by means of intensification of cultivation, shifting the 

ratio of perennials to annuals and sometimes domestic animals, and a certain degree 

of specialization in crops. But major differences in socioeconomic status are 

reflected in homegardens having a clearly different composition. It is possible to 

differentiate various types of homegardens in respect to their role in the household 

economics (Table 2). 

Table 2. Different types of homegardens in relation of household economics. 

Homegarden type Characteristics 

Survival gardens Gardens form single component farming system of 

otherwise landless rural people 

Combined production of staple food crops and comple- 

mentary crops 

Subsistence gardens Part of multi-component farming system in conjuncttion 

with permanent or shifting field production 

Complementary system to open-field staple food cultivation 

systems 

Provision of daily supply of vegetables, herbs, spices 

and fruits for household needs and occasional sale 

Market gardens Specialized farming system or part of multi-component 

farming system 

Cultivation of cash crops with possible complementary 

production of household products 

Budget gardens Gardens of households with economic bases in rural or 

urban employment; family needs are mostly purchased 

from the market 

Cultivation of ‘hobby’ products for household consumption 

and ornamentals 

Source: Adapted from Niñez (1984). 


4. HOMEGARDEN DYNAMICS 

Many of the factors that impinge on homegarden structure and composition change 

with time, and it is therefore logical to infer that the homegarden structure and 

composition change whenever socioeconomic factors change (e.g., Peyre et al., 

2006; Abdoellah et al., 2006). Such changes often reflect the general processes of 

rural changes and may involve several aspects of rural transformations. Areas that 

used to be remote are increasingly being incorporated into the national economy 

with traditional land use systems such as shifting cultivation gradually becoming 

transferred to more permanent cropping systems. Remote areas may also be actively 

opened up for migrants. In Indonesia the transmigration from the densely populated


I 

island of Java to other islands is actively stimulated, and as a result the typical 

Javanese homegarden is being introduced in new regions. Moreover, in many rural 

areas the (semi)subsistence household economies of former times are increasingly 

becoming more commercially oriented. In others, urban life-styles are developing 

and the household dependence on primary production is changing to include 

activities in the manufacturing or service and trade sectors. In some places, these 

dynamics are intensive; in others they take place more gradually. Depending on the 

nature and intensity of rural changes, the developments in structure and composition 

as well as functions of homegardens may show different trends (Table 3). 

Table 3. Main trends in homegarden development. 

Main trends Consequences 

Extension in area 

Extension of housing due to population 

growth 

Extension to new areas due to change in 

farming systems, e.g., from shifting cultivation 

to permanent cultivation 

Extension to new areas due to migration 

Changing structure and composition 

Adaptation of gardens to new food habits 

and changing household needs or new 

agronomic practices 

Increasing commercialization 

• Decreasing importance subsistence 

production 

• Increasing commercial production 

Increasing role of aesthetic function garden 

Prevalence of (bi)annual food crops in 

newly established gardens 

Young homegardens have not yet 

reached full diversity 

Extension of homegarden to new areas 

with adaptation to different land use 

systems than in area of origin 

Gradual change in structure and composition 

including incorporation of new 

species 

Decreasing importance of supplementary 

production 

Increasing specialization on either 

vegetable, spices or fruit tree production 

Increase in ornamental plants 

5. CONCLUSIONS 

Homegardens have often been described as a sustainable agroforestry system with 

positive ecological and socioeconomic features. While several studies have 

explicitly highlighted homegardens as traditional systems, relatively little attention 

has been given to studies on the dynamics of homegardens under influence of rural 

transformations. Nonetheless, several studies have demonstrated that homegarden 

function and composition depends greatly on socioeconomic conditions as well as 

household livelihood strategies. In this context, not only different types of 

21

22 


homegarden systems can be recognized, but also different pathways of homegarden 

development can be identified. The changes in homegarden function and 

composition have been interpreted differently by various authors. Some argue that 

although the traditional homegardens have gradually lost their original ecological 

and economic features, they still are a major asset for the modernization of village 

economy and society. The changes in homegarden features are associated with 

socio-professional changes of villagers and reflect a search for a new balance in the 

relationship between cities and villages. Other authors take a more negative point of 

view of the dynamics in tree gardening systems; they view the changes under 

influences of rural dynamics as the disappearance of a traditional system and 

propose measures to revitalize such traditional tree gardening systems. These 

different and somewhat opposing views on the trends in homegarden function 

and composition represent different norms in assessing social sustainability 

of homegardens and differences in value judgments on the ideal structure of 

homegardens. 

ENDNOTE 

1. The international literature on tropical homegardens is often ambivalent on 

whether homegardens are characterized by structure or location. 

REFERENCES 

Abdoellah O.S. 1990. Homegardens in West Java and their future development. In: Landauer 

K. and Brazil M. (eds), Tropical home gardens, pp 69 – 79. United Nations University 

Press, Tokyo. 

Abdoellah O.S., Hadikusumah H.Y., Takeuchi K., Okubo S. and Parikesit 2006. 

Commercialization of homegardens in an Indonesian village: vegetation composition and 

functional changes. Agroforest Syst (in press). 

Ashley C. and Maxwell S. 2001. Rethinking rural development. Dev Policy Rev 19: 

395 – 425. 

Brownrigg L. 1985. Home gardening in international development: What literature shows? 

The League for International Food Education, Washington, DC, 330p. 

Christanty L. 1990. Homegardens in tropical Asia with special reference to Indonesia. In: 

Landauer K. and Brazil M. (eds), Tropical home gardens, pp 9 – 20. United Nations 


Christanty L., Abdoellah O.S., Marten G.G. and Iskander J. 1986. Traditional agroforestry in 

West Java: the pekarangan (homegarden) and kebun-talun (annual-perennial rotation) 

cropping system. In: Marten G.G. (ed.), Traditional agriculture in South East Asia, pp 

132 – 158. Westview Press, Boulder, CO. 

Ellis F. 1998. Household strategies and rural livelihood diversification. J Dev Stud 35: 1 – 38. 

Gajaseni J. and Gajaseni N. 1999. Ecological rationalities of the traditional homegarden 

system in the Chao Phraya Basin, Thailand. Agroforest Syst 46: 3 – 23. 

Hochegger K. 1998. Farming like the forest: traditional homegarden systems in Sri Lanka. 

Tropical agroecology. Margraf Verlag, Weikersheim, 203p. 

Jacob V.J. and Alles W.S. 1987. The Kandyan gardens of Sri Lanka. Agroforest Syst 5: 

123 – 137.


I 

John J. and Nair M.A. 1999. Socio-economic characteristics of homestead farming in south 

Kerala. J. Trop Agric 37: 107 – 109. 

Karyono 1990. Homegardens in Java: their structure and function. In: Landauer K. and Brazil 

M. (eds), Tropical home gardens, pp 138-146. United Nations University Press, Tokyo. 

Kehlenbeck K. and Maass B.L. 2004. Crop diversity and classification of homegardens in 

Central Sulawesi, Indonesia. Agroforest Syst 63: 53 – 62. 

Kehlenbeck K. and Maass B.L. 2006. Are tropical homegardens sustainable? Some evidence 

from Central Sulawesi, Indonesia. In: Kumar B.M. and Nair P.K.R. (eds), Tropical 




135 – 152. 

Kumar B.M., George S.J. and Chinnamani S. 1994. Diversity, structure and standing stock of 

wood in homegardens of Kerala in peninsular India. Agroforest Syst 25: 243 – 262. 

Landauer K. and Brazil M. (eds) 1990. Tropical home gardens. United Nations University 

Press, Tokyo, 257p. 

McConnell D.J. 1992. The forest-garden farms of Kandy, Sri Lanka. Farm systems management 

Series No. 3, FAO, Rome, 117p. 

Michon G. 1983. Village-forest-gardens in West Java. In: Huxley P.A. (ed.), Plant research 

and agroforestry, pp 13 – 24. ICRAF, Nairobi. 

Michon G. and Mary F. 1994. Conversion of traditional village gardens and new economic 

strategies of rural households in the area of Bogor, Indonesia. Agroforest Syst 

25: 31 – 58. 

Miller R.P., Penn, Jr. J.W. and Leeuwen J. van. 2006. Amazonian homegardens: their 

ethnohistory and potential contribution to agroforestry development In: Kumar B.M. 



Nair M.A. and Sreedharan C. 1986. Agroforestry farming systems in the homesteads of 

Kerala, southern India. Agroforest Syst 4: 339 – 363. 

Niñez V.K. 1984. Household gardens: theoretical considerations on an old survival strategy. 

Potatoes in Food Systems Research Series Report No. 1. International Potato Center, 

Lima, Peru, 41p. 

Parikesit, Takeuchi K., Tsunekawa A. and Abdoellah O.S. 2004. Kebon tatangkalan: a 

disappearing agroforest in the Upper Citarum watershed, West Java, Indonesia. 

Agroforest Syst 63: 171 – 182. 

Peyre A., Guidal A., Wiersum K.F. and Bongers F. 2006. Dynamics of homegarden structure 

and function in Kerala, India. Agroforest Syst 66: 101 – 115. 

Randhawa M.S. 1980. The history of Indian agriculture. Vol. 2, pp 414 – 415. Indian Council 

of Agricultural Research, New Delhi. 

Sauer C.O. 1969. Agricultural origins and dispersals (2 nd ed.). M.I.T. Press, New York. 175p. 

Soemarwoto O. 1984. The Javanese rural ecosystem. In: Rambo A.T. and Sajise P.E. (eds), 

An introduction to human ecology research on agricultural systems in Southeast Asia, 

pp 254 – 287. University of Philippines, College, Los Banos. 




Soemarowoto O. and Conway G.R. 1991. The Javanese homegarden. J Farming Syst Res 

Extn 2: 95 – 118. 

Stoler A. 1978. Garden use and household economy in rural Java. Bull Indonesian Econ Stud 

14: 85 – 101. 

23

24 


Terra G.J.A. 1953a. Mixed-garden horticulture in Java. Malaysian J Trop Geogr 1: 33 – 43. 

Terra G.J.A. 1953b. The distribution of mixed gardening on Java. Landbouw (Indonesia) 

25: 163 – 203. 

Terra G.J.A. 1958. Farm systems in South-east Asia. Neth J Agric Sci 6: 157 – 182. 

Torquebiau E. 1992. Are tropical agroforestry homegardens sustainable? Agric Ecosyst 

Environ 41: 189 – 207. 

Wiersum K.F. 1982. Tree gardening and taungya on Java: examples of agroforestry 

techniques in the humid tropics. Agroforest Syst 1: 53 – 70. 

Wiersum K.F. 2004. Forest gardens as an ‘intermediate’ land use system in the nature-culture 

continuum: characteristics and future potential. Agroforest Syst 61: 123 – 134.

CHAPTER 3 

URBAN AND HOMEGARDEN 

AGROFORESTRY IN THE PACIFIC 

ISLANDS: CURRENT STATUS AND 

FUTURE PROSPECTS 

R.R. THAMAN 1 , C.R. ELEVITCH 2 , AND J. KENNEDY 3 

1 

The University of the South Pacific, Suva, Fiji Islands; 148; E-mail: 

. 2 3 

Agroforestry.net, Holualoa, Hawai’i. Research School of 

Pacific and Asian Studies, Australian National University, Canberra, Australia 

Keywords: Agrodeforestation, Food security, Nutrition, Sustainability, Traditional agroforestry. 

Abstract. Pacific islanders traditionally had abundant, predominantly rural, agroforestry 

systems that provided a wide array of products for meeting the necessities of life, and 

conducive environments for the rich Pacific island cultures. In recent years, however, 

increasing urbanization and accompanying removal of trees and perennial agroforests 

(“agrodeforestation”) have resulted in the breakdown of these traditional agroforestry 

systems, accompanied by increasing economic, cultural, nutritional, and environmental 

problems, particularly in the urban areas. A critical analysis of the nature and future prospects 

of the urban and homegarden agroforestry systems in these rapidly urbanizing islands 

suggests that intensification and enrichment of these systems could serve as an important 

foundation for sustainable development. In addition to addressing the nutrition-related health 

problems, food security, poverty alleviation, and trade deficits, these systems also help protect 

and enrich the cultural traditions of Pacific peoples who are increasingly out-migrating from 

rural areas and embracing urban living. 


Pacific island countries have historically been resource self-reliant because of their 

relative geographic isolation. Their traditional land- and sea-based economies, 

25 




26 R.R. THAMAN T 

ET AL. 

cultures, and isolation from major markets forced island communities to develop 

sustainable land use systems such as agroforestry and freshwater and marine fishery 

production systems. Today, however, the small island states of the Pacific Ocean are 

rapidly urbanizing and increasingly large populations no longer have access to 

traditional agricultural and wildland holdings. Moreover, trees are disappearing from 

both rural and urban areas, a process referred to as “agrodeforestation” (Thaman, 

1992). Studies in the Ha’ppai Islands of Tonga, for example, showed that about 100 

traditionally important trees or shrubs, all integral components of the traditional 

Tongan bush-fallow agroforestry, were reported to be endangered or in short supply. 

These included 28 large wild trees, 32 large cultivated trees, 19 wild small trees or 

shrubs, and 24 planted small trees and shrubs (Thaman et al., 2001). Species most 

commonly mentioned were multipurpose trees valued for medicine, fruit, firewood, 

construction purposes, and fragrant plants used for body ornamentation and 

perfume. Foremost among the reasons for the loss of tree cover was the failure to 

plant or replant trees in general, as well as changing land-management practices that 

included indiscriminate felling, clearing and burning, and the increasing use of the 

plow. These practices have caused a gradual shift away from the traditional mixed 

agroforestry systems in which fruit trees and other culturally useful trees, such as 

coconut (Cocos nucifera), breadfruit (Artocarpus ( altilis), 

traditional banana and 

plantain clones (Musa spp.), citrus (Citruss spp.), Malay apple ( Syzygium malaccense) 

and Polynesian vi-apple (Spondias dulcis) were dominant, to monocultural 

production of commodities. Spread of tall grass species such as guinea grass 

(Panicum maximum) and the impact of grazing animals and other domesticated 

livestock made regeneration of trees particularly difficult (Thaman et al., 2001). 

Today, most Pacific island countries are increasingly dependent on imported 

food, fossil fuels, medicines, and other industrial products to satisfy the basic needs, 

with the result that they experience negative balance of payment situations. Food 

security has become a major concern in all of the independent island states 1 . 

Disappearance of forests and agroforests, which traditionally provided these 

products and services, may partially explain this increasing dependence on imported 

foods and fuels (Thaman, 1988). The people of the Pacific islands also have high 

rates of nutrition-related, non-communicable diseases, such as diabetes, cardiovascular 

disease, hypertension, hyperuricemia and gout, obesity, iron deficiency 

anemia, and dental diseases (Parkinson, 1982; Thaman, 1982; Coyne, 2000)—with 

rates of obesity as high as 75% reported in Nauru, Samoa, the Cook Islands, Tonga, 

and French Polynesia. The main factor for most of these diseases has been the shift 

in the food habits from traditional “healthy” foods to processed, imported “unhealthy” 

foods (Curtis, 2004). 

Experience from the region as well as elsewhere suggests that the promotion of 

urban and homegarden agroforestry may provide culturally appropriate and costeffective 

means of addressing both urban agrodeforestation, and many of the economic, 

cultural, nutritional and environmental problems arising out of urbanization and 

globalization (Thaman and Clarke, 1993; Thaman, 2002; Lamanda et al., 2006). It 

can also help to address indirectly many emerging environmental problems, such as 

climate change and associated sea-level rise, coastal erosion, pollution, and loss of

native trees and biodiversity. In this scenario, this chapter assesses the status and 

future prospects of urban and homegarden agroforestry in the Pacific islands. 

This chapter is based on the authors’ extensive experience in the region as well 

as inventories of urban and homegardens in Papua New Guinea (PNG), Fiji, Tonga, 

Kiribati, and Nauru during the 1970s and 1980s (Thaman, 1983; 1987), together 

with subsequent studies in these countries, and in New Caledonia, Solomon Islands, 

Vanuatu, Samoa, Niue, the Cook Islands, Tuvalu, French Polynesia, Hawai’i, the 

Marshall Islands, and Palau (Thaman, 1995). 

1.1. Homegardens 

Homegardens are a ubiquitous feature of the Pacific island landscapes, from the very 

densely populated urban areas in atoll microstates, such as South Tarawa, Kiribati, 

Fogafale Islet on Funafuti Atoll, Tuvalu, and RETA in northeast Majuro Atoll in the 

Marshall Islands to rural villages and plantations in areas of low population density 

in Fiji, Vanuatu, and Papua New Guinea. Even in areas not known for agricultural 

diversity, such as Kiribati, Tuvalu, the Marshall Islands, and Nauru, homegardens 

contain a wide range of food trees, non-tree staple and supplementary food plants, 

medicinal plants, and other non-food trees and plants of cultural and economic 

importance (Table 1). Homegarden surveys in these localities have indicated the 

cultivation of 33 to 114 species or distinct types of food plants in these localities 

(Thaman, 1995). In Palau, where women are responsible for most gardening, homegardens 

with a diversity of food trees and extensive multispecies, taro-dominated 

agroforests are found throughout the main towns and in areas surrounding villages. 

Table 1. Number of species and distinct varieties of food plants found in surveys of 

homegardens systems in different Pacific islands. 

Crop types 

URBAN U AND HOMEGARDEN AGROFORESTRY IN THE PACIFIC P ISLANDS I 

Number of species/varieties in different Pacific islands 

PNG Fiji Tonga Kiribati Nauru Location 1 

Non-tree staples 7 10 8 6 5 8 

Non-tree supplementary 48 65 44 35 14 41 

Food trees 2 30 39 27 20 14 16 

Total 85 114 79 61 33 65 

Source: Thaman (1995); 1 Location, a contract worker settlement in Nauru; 2 The totals for 

Papua New Guinea and Nauru, where banana clones and Citrus spp., respectively, were not 

differentiated, would have been slightly higher for tree crops if these differentiations had been 

made). 

In addition to the food plants, many useful non-food plants were also found in 

the homegardens. Examples include important “handicraft plants” such as Pandanus, 

the leaves of which are processed to make mats, thatching, baskets, hats and a wide 

range of other plaited ware; paper mulberry (Broussonetia papyrifera), the treated 

27

28 

bast fiber of which is used for bark cloth (tapa); and sources of dyes such as annatto 

(Bixa orellana) and Java cedar (Bischofia javanica). There may also be 

multipurpose trees such as Leucaena leucocephala, a wide range of medicinal 

plants, plus countless other plants of considerable technological, economic, social, 

ecological, and ornamental values (Table 2; Fig. 1). 

Table 2. Significant plant species of urban and rural Pacific island homegardens and 

underdeveloped urban and periurban open areas. 

a. Homegardens 

Category Species 

Staple root 

crops 

Supplementary 

food crops 

R.R. THAMAN T ET AL. 

Alocasia macrorrhiza (giant taro) 

Colocasia esculenta (taro) 

Cyrtosperma chamissonis (giant swamp taro) 

Dioscorea alata (greater yam) 

Dioscorea esculenta (sweet yam) 

Ipomoea batatas (sweet potato) 

Manihot esculenta (cassava) 

Xanthosoma spp. (tannia or cocoyam) 

Abelmoschus esculentus (okra) 

Abelmoschus manihot t (hibiscus spinach) 

Allium spp. (bunching onions) 

Amaranthus spp. (amaranth spinach) 

Ananas comosus (pineapple) 

Arachis hypogaea (peanuts) 

Brassica chinensis, B. juncea and B. oleracea vars. (cabbages, 

including Chinese, mustard and English or head cabbages) 

Cajanus cajan (pigeon pea) 

Citrullus lanatus (watermelon) 

Coccinea grandis (ivy gourd) 

Colocasia esculenta (taro leaf spinach) 

Cucumis melo var. cantalupensis (cantaloupe or rock melon) 

Cucumis sativus (cucumber) 

Cucurbita pepo (pumpkin) 

Luffa acutangula (angled loofah or ridgegourd) 

Luffa cylindrica (sponge gourd) 

Momordica charantia (bitter gourd) 

Passiflora edulis (passionfruit) 

Phaseolus, Psophocarpus and Vigna spp. (beans and other 

legumes) 

Saccharum officinarum (sugarcane) 

Solanum lycopersicon (tomato) 

Solanum melongena (eggplant) 

Xanthosoma spp. (taro leaf spinach) 

Zea mays (corn)

Fruit and nut 

yielding trees 

Spice plants 

and social 

beverage and 

stimulant plants 

Non-food 

plants 

b. Undeveloped open areas 

Staple root 

crops 


Annona spp. (soursop and sweetsop) 

Artocarpus altilis and A. camansi (breadfruit and breadnut) 

Carica papaya (papaya) 

Citrus aurantifolia (limes), C. aurantium (sour orange), C. limon 

and C. medica x limon (lemon), C. maxima (pummelo), 

C. mitis (calamondin), C. reticulata (tangerine and mandarin 

orange), and C. sinensis (orange) 

Cocos nucifera (coconut) 

Ficus spp. (fig trees) 

Inocarpus fagifer r (Tahitian chestnut) 

Mangifera indica (mango) 

Musa cultivars (banana and plantains) 

Pandanus spp. (edible pandanus) 

Persea americana (avocado) 

Pometia pinnata (oceanic litchi) 

Psidium guajava (guava) 

Spondias dulcis (vi apple) 

Syzygium aqueum (water apple) 

Syzygium malaccense (Malay apple) 

Terminalia catappa (beach almond) 

Areca catechu (betel nut palms) 

Capsicum frutescens and C. annum cvs (chili) 

Coriandrum sativum (coriander) 

Cymbopogon citratus (lemon grass) 

Mentha spp. (mint) 

Piper betle (betel vine) 

Piper methysticum (kava) 

Zingiber officinale (ginger) 

Cananga odorata (ylang-ylang) 

Fagraea berteroana (pua ( ) 

Gardenia taitensis (Tahitian gardenia) 

Guettarda speciosa (guettarda) 

Pandanus spp. (pandanus cultivars) 

Pimenta racemosa (bay rum) 

Plumeria obtusa and P. rubra ( frangipani) 

Colocasia esculenta (taro) 

Dioscorea esculenta (sweet yam) 

Ipomoea batatas (sweet potato) 

Manihot esculenta (cassava) 

Xanthosoma spp. (tannia or cocoyam) 

29 

Table 2 (cont.)

30 

Category Species 


Fruit and nut Artocarpus altilis (breadfruit) 

yielding trees 

Carica papaya (papaya) 

Citrus spp. (citrus) 

Cocos nucifera (coconut) 

Mangifera indica (mango) 

Musa cultivars (banana and plantains) 

Pometia pinnata (oceanic litchi) 

Psidium guajava (guava) 

Syzygium spp. 

Terminalia catappa (beach almond) 

Non-food plants Bischofia javanica (koka) 

Cassia and Senna spp. (shower trees) 

Casuarina spp. 

Delonix regia (flamboyant) 

Erythrina variegata (coral tree) 

Eucalyptus spp. 

Ficus spp. (banyans) 

Gliricidia sepium (madre de cacao) 

Hibiscus rosa-sinensis 

Hibiscus tiliaceus (beach hibiscus) 

Lagerstroemia speciosa (pride of India) 

Leucaena leucocephala 

Macaranga spp. 

Morinda citrifolia (noni) 

Plumeria obtuse and P. rubra ( frangipani) 

Polyscias spp. (hedge panax) 

Samanea saman (rain tree) 

Spathodea campanulata (African tulip tree) 

Source: Based on Thaman (1983, 1987, 1995, 2002); Levett (1992, 1996); Levett and Uvano 

(1992). Categories such as ‘supplementary food crops’ and ‘spice plants and social beverage 

and stimulant plants’ were clearly absent in the undeveloped open areas. 

Homegardens also contain a great diversity of cultivars of important food and 

handicraft plants. As stressed by Soemarwoto et al. (1985) in their study of Javanese 

homegardens, true plant diversity is far greater than indicated by the numbers of 

species, since many species are represented by numerous cultivars. In Tonga, for 

example, there are numerous distinct breadfruit cultivars, the most common of 

which include ma’ofala, maopo, puou, loutoko, kea and ’aveloloa. There is similarly 

great cultivar diversity among other tree crops such as coconuts, banana, mango 

(Mangifera indica), pandanus (Pandanus spp.), papaya (Carica papaya), and 

especially among the traditional staple root crops such as yams (Dioscorea spp.), 

taros (Colocasia esculenta, Alocasia macrorrhiza and Xanthosoma spp.), and sweet 

potato (Ipomoea batatas), all of which add economic, ecological, and nutritional


stability to the urban gardening systems. “Tree gardens” in the settlements in Yap, in 

the Federated States of Micronesia, for example, had 21 coconut cultivars, 28 

breadfruit cultivars, and 37 banana cultivars (Falanruw, 1995). Similar cultivar 

diversity is found in the taro (Colocasia esculenta)-dominated agroforestry gardens 

in and around the main town of Koror and villages in Palau. Throughout Papua New 

Guinea, tree crops continue to provide a crucial component of the diverse 

subsistence agricultural systems of the rural population, with high cultivar diversity 

of many species such as bananas, breadfruit, pandanus, and many indigenous fruit 

and nut trees. This diversity is retained despite the addition of newly introduced 

high-yielding cultivars 2 . 

Figure 1. Homegarden in downtown Apia, Samoa includes numerous useful tree species 

including the fast-growing timber tree poumuli ( Flueggea flexuosa), fruit trees such as 

coconut (Cocos nucifera) and breadfruit (Artocarpus altilis), as well as many ornamentals 

(Photo: R. R. Thaman). 

Countless species, commonly overlooked as “weeds,” are important components 

of homegardens (Soemarwoto et al., 1985). Homegardeners have many uses for 

spontaneously propagating plants as medicines, fuel, fodder, mulch, roofing, fish 

poisons, toothbrushes, and food. “Weeds” such as Amaranthus spp., black 

nightshade (Solanum americanum), purslane (Portulaca oleracea), water spinach 

(Ipomoea aquatica), and fetid sea holly (Eryngium foetidum), for example, are 

important potherbs in Fiji and are often sold in the municipal market of Suva 

31

32 

R.R. THAMAN T 

ET AL. 

(Thaman, 1976/77), and almost all grass species are used for fodder if domestic 

animals are kept. 

1.2. Urban agroforestry gardens apart from homegardens 

Cultivation outside homegardens on undeveloped land (i.e., land without residences, 

buildings, or for other uses such as playing fields, parks, etc.) is very widespread in 

the Pacific island urban areas. These urban and periurban gardens also develop into 

agroforestry systems, and are important sources of food (including leaves, fruits, and 

nuts) and other products such as timber, fence posts, fuelwood, handicraft and light 

construction materials, medicines, and flowers (Table 2). Such areas include road 

frontages, empty lots, riverbanks and valleys, rights-of-way for proposed or existing 

paths and roads, and open land in general including hillsides and swamplands. Both 

subsistence and limited commercial production are attempted in these urban and 

periurban agroforestry gardens (Fig. 2). 

Figure 2. Periurban mixed planting with fruits, timber, medicinal, and staple crops on ‘Upolu 

island’, Samoa. Species include coconut (Cocos nucifera), breadfruit (Artocarpus altilis), 

poumuli ( Flueggea flexuosa), bananas and plantains ( Musa a spp.), and noni ( Morinda 

citrifolia). Note yam vine ( Dioscorea sp.) trellised onto breadfruit f tree on left (Photo: 

C. Elevitch). 

In the suburbs of Port Moresby, PNG, sampled in the 1970s, more than one-third 

of all households had “gardens” on other lands in addition to their homegardens. The


distinction here is in the location of these gardens with respect to homes: while 

homegardens are located surrounding homes, these “other gardens” are not 

physically close to the homes. Kilakila villagers, who were then largely original 

inhabitants of the area, had particularly large tracts of urban savanna lands, and all 

households had, in addition to their homegardens, up to four “bush” gardens 

averaging 1135 m 2 located on urban lands within 3 km of the center of Kilakila. 

With the expansion of the Port Moresby population from 124 000 in 1980 to over 

250 000 in 2000 (National Statistical Office of Papua New Guinea, 2000; Allen 

et al., 2002), such large urban gardens can no longer exist, although no detailed 

follow-up study has been undertaken. 

Open hillsides within Port Moresby support a distinctive system of agriculture 

based on wet season plantings dominated by sweet potato, along with cassava 

(Manihot esculenta), banana, taro (Colocasia and Xanthosoma spp.), hibiscus 

spinach ( (Abelmoschus manihot), 

Chinese cabbage (Brassica chinensis), corn (Zea 

mays), cucumber (Cucumis sativus), passionfruit (Passiflora spp.), peanut ( (Arachis 

hypogaea), pineapple ( (Ananas comosus), 

pumpkin (Cucurbita pepo), snake or long 

bean (Vigna unguiculata subsp. sesquipedalis), bunching onion ( (Allium spp.), 

sugarcane (Saccharum officinarum), tomato (Solanum lycopersicon) and watermelon 

(Citrullus lanatus). Tree crops include breadfruit, coconut, mandarin and 

sweet oranges (Citrus reticulata and C. sinensis), mango and papaya. Practiced by 

urban migrants who rent the gardened land from local traditional landowners, this 

system differs from the surrounding agriculture of rural people in that the grassland 

fallow period is shorter, and drains are dug across the slope, with soil heaped behind 

them into long beds. These gardens are even less well studied than Port Moresby’s 

homegardens (Allen et al., 2002), and they periodically attract public criticism in 

local newspapers, as a cause of erosion and smoke pollution. 

In Suva, Fiji, about 20% of all households surveyed in the late 1970s cultivated 

“unused” open lands. It has been estimated that in the 30 km 2 Suva Peninsula, 

approximately 5 km 2 which represents more than 70% of the area not under swamp 

or mangrove – is still under this type of cultivation (Thaman, 1995). Planting is done 

along road frontage in about 20% of all households despite the City Council 

regulations forbidding such practices, and the practice seems to have intensified 

recently in parts of Suva. 

In Tonga, Kiribati, and Nauru, there is little undeveloped “urban” land. However, 

in a number of cases, the Tongans planted entire adjacent unoccupied “town allotments” 

with sweet potato, taro, tannia (Xanthosoma 

( spp.), and a mixture of trees, or with 

traditional mixed yam gardens, where yams, giant taro (Cyrtosperma chamissonis), 

plantains (Musa spp.), and taro are intercropped, usually among or under coconuts 

and other trees (Thaman, 1978). There is virtually no open land in urban Kiribati, 

but in Nauru, some Chinese, Tuvaluan and I-Kiribati (nationals of Kiribati) contract 

laborers plant food gardens near the Nauru Phosphate Corporation’s workshops on 

the phosphate-rich central plateau and in the swampy areas surrounding landlocked 

Buada Lagoon. In Tuvaluan and I-Kiribati gardens, coconuts and banana clones 

were dominant. 

33

34 

1.3. Animal husbandry 

R.R. THAMAN T 

ET AL. 

Small-scale animal husbandry, although playing a minor role compared with plant 

food production, is also an important activity in urban and rural homegardens. In 

Port Moresby suburbs studied in the 1970s, animal keeping was minimal, with 11 of 

79 households keeping pigs, chickens, or ducks and a few households keeping 

tethered cows or goats. More recently, there are a few reports of urban household 

pigs, and of raising pigs on food wastes at the city dump (Hide, 2003), but there has 

been no recent detailed study. Pigs were not kept in Suva, but in Tonga over half of 

all sample households kept tethered or penned pigs, and almost two-thirds kept 

chickens or ducks. In most cases, poultry were penned or tethered at night and 

allowed to roam around during the day, and pigs and other larger animals were 

generally tethered or penned at all times. In Kiribati, Tuvalu and Nauru, pigs and 

chickens are also kept on home allotments. In Nauru, there was a large communal 

pig rearing area along the beach in Denigomodu District. In Betio, the most heavily 

populated area of South Tarawa, there was a large communal pig rearing area with 

individualized pens, established by the local town council, under coconuts, 

breadfruit, and other trees. In Tuvalu, pigs are kept near the main urban village 

along the airport runway on the seaside of the main Fogafale Islet, where they are 

fed with kitchen wastes and mangrove leaves. In general, homegardens in rural areas 

also have animals which are penned, tethered, or sometimes free ranging – 

particularly chickens around houses, which also serve to control cockroaches and 

other insects. 

Apart from kitchen waste, the main feed for pigs and chickens in most areas is 

coconut kernel. In Tonga, goats and pigs are commonly fed the leaves of Leucaena 

leucocephala, Pisonia grandis, and Erythrina variegata, while “living edible pens” (pens 

with edible living fencing) for poultry and pigs are made of these same species, plus 

others such as Hibiscus tiliaceus s and Polyscias s spp., all of which are easily pruned or 

pollarded to provide fodder. On open lands, horses, cattle, and goats are commonly 

tethered to trees, which also give them shade. Small animal pens that are commonly 

constructed of coconut logs, bamboo, Leucaena, or other local timber are also 

found occasionally. In rural homegardens, pigs, goats, and even cattle in Fiji, are 

stall-fed, or rotationally tethered to trees or fence posts where they can graze or 

browse. 

On the negative side, grazing animals and pigs seem to accelerate agrodeforestation 

in urban areas through browsing or trampling effects. Once established, 

however, trees and animals co-exist well, except where browsing goats eat the bark 

of trees. Cattle, in fact, seem to enhance the establishment and spread of guava 

(Psidium guajava), which although an important fruit, medicinal, and fuelwood 

source, has become a noxious pasture weed in many areas. Another serious problem 

related to pig keeping in urban areas is the effect of high-nutrient waste runoff on 

the nearby shore coral reefs. Nutrient-enriched water favors the growth of algae and 

phytoplankton over the growth and maintenance of coral reefs, which require clear, 

nutrient-poor waters. In the rural outer islands of Ha’ppai in Tonga, free-ranging 

pigs were seen as one of the major constraints to expanded homegardening and the 

planting of trees in rural villages (Thaman et al., 2001).


1.4. Ethnic basis of garden composition 

The most common plants of Pacific island homegardens tend to be traditionally 

important native plants or pre-European (aboriginal) introductions, except where the 

gardeners are from immigrant populations. For example, the Indian population of 

Fiji prefers species such as eggplant (Solanum melongena), okra ( (Abelmoschus 

esculentus), Amaranthus spp., pulses and cucurbits, and tree crops such as jackfruit 

(Artocarpus ( heterophyllus), 

tamarind (Tamarindus indicus), mango, Citrus spp., 

curry leaf (Murraya koenigii), Sebesten plum (Cordia dichotoma), horseradish or 

drumstick tree (Moringa oleifera), and the spiritually and medicinally important 

neem tree ( (Azadirachta indica). 

Similarly in a study of 150 urban lots in Hawai’i, where the native population is 

small relative to the immigrant population of Japanese, Chinese, Filipino, European 

and North American origin, plants introduced after European contact dominated the 

homegardens (Ikagawa, 1994). Of the 42 genera present in more than 10% of 

Honolulu gardens, only two were introduced by Hawaiians, ti (Cordyline fruticosa) 

and Musa spp. A strong preference for ornamental landscapes and the strong moneybased 

economy and culture presumably explain the relative lack of edible, culinary, 

and medicinal plants in Hawaiian homegardens. 

In Port Moresby and most other PNG urban areas and plantation or mining 

settlements, where there are high percentages of immigrants from other areas of 

PNG, homegardens reflect the great diversity of species, cultivars and cultivation 

practices arising from the cultural and ecological diversity for which the country is 

famous. This diversity is evident in the gardens of settlers on the oil palm (Elaeis 

guineensis) projects of West New Britain (Benjamin, 1985) and Milne Bay 

Province. In Port Moresby, urban migrants often have preferences to traditional 

crops of their native habitats that may be unsuited to the local soils or climate. 

Examples include struggling sago palms (Metroxylon sagu), and the small potherb 

Rungia klossii, lovingly nurtured to coax a second crop of leaves from cuttings 

brought from the highlands. Similarly, Trobriand islanders, attached to the social 

values of their crops, have transferred competitive yam growing to Port Moresby 

(Battaglia, 1985). 

Despite the dominance of these traditional crops, there is also a great range of 

more recently introduced crops, such as temperate vegetables, pineapple, papaya, 

avocado (Persea americana), guava, and improved citrus varieties and banana 

clones, as well as cassava, which is a ubiquitous staple in most Pacific island towns 

(Thaman and Thomas, 1985). In fact, Pacific homegardens seem to have been, and 

will probably continue to be, one of the most effective avenues for the introduction 

and acceptance of new plant species. The introduction of chaya or tree spinach 

(Cnidoscolus chayamansa) into homegardens in urban South Tarawa and elsewhere, 

mentioned above, is an excellent example. 

1.5. Spatial arrangement of components in the homegardens 

There is great diversity in the spatial distribution of food crops and their area. 

Whereas some households have only a few scattered fruit trees and vegetables, 

35

36 

many cultivate food crops on over 50% of the total area of their property. In Port 

Moresby, for example, in Morata and Gerehu suburbs, recently settled in the mid- 

1970s, an average of approximately 40% of 450 m 2 allotments were then under food 

crops. Similarly, in some cases in Nuku‘alofa, up to 75% of 500 to 1000 m 2 

allotments were under food cultivation, primarily root crops (such as yam, taro, 

tannia, cassava, and sweet potato) and banana among scattered trees (Thaman, 

1995). Trees gradually become dominant in long-settled areas as cash incomes increase, 

and tree seedlings mature and increasingly shade garden areas. Nevertheless, in suburbs 

such as Gerehu, where trees have matured, socioeconomic status has risen. Although 

and the contribution to household economies that homegardens provide has 

declined, gardening continues to be important (Levett, 1996). 

Ornamentals are commonly planted closest to the home, often in front yards, as 

well as in containers on front porches. Medicinal plants, sacred or fragrant plants, 

and other culturally valuable, common multipurpose plants, are scattered amongst 

the food plants and ornamentals. In gardens of the indigenous Nauruans (who as a 

result of phosphate mining royalties, have historically had high per capita incomes), 

ornamental, aromatic and medicinal plants dominate, along with the ubiquitous 

coconut, edible pandanus, some bananas, and breadfruit. At the Location contract 

worker settlement in Nauru, where people live in multistory tenements, and where 

family gardening is limited to no more than 15 to 30 m 2 , most families have only a 

few plants. The gardens of Tuvaluans and I-Kiribati who live as contract workers in 

Nauru often consist of juvenile tree seedlings, staple root crops, or a single coconut 

palm or stand of bananas. In the gardens of Chinese (mostly recruited from Hong 

Kong) and Filipinos, the emphasis is on intensive vegetable gardening, often in 

containers, reflecting a more intensive system than that was practiced by most 

indigenous Pacific island peoples. In Kiribati and Tonga, however, recent emphasis 

has been placed by the government and non-governmental organizations on more 

intensive types of gardening: in Kiribati, using hydroponic and deep mulching 

techniques because of the highly infertile calcareous and sandy soils there. In 

Kiribati, where vitamin A deficiency-induced night blindness and xerophthalmia 

have become problems, the planting and consumption of the vitamin-rich leaves of 

two native tree species: noni (Morinda citrifolia) and Pisonia grandis, and more 

recently chaya have been encouraged in urban areas (Thaman, 1995). 

1.6. Trends toward agrodeforestation 

R.R. THAMAN T 

ET AL. 

Despite the current importance of gardening on open urban and periurban land, these 

areas have been severely affected by deforestation and agrodeforestation (Thaman, 

1992). Increasing population, poverty, and need for firewood, expansion of squatter 

settlements, lack of legislation controlling tree removal, increasing dependence on 

root crops such as cassava and sweet potatoes, and the loss of knowledge on the 

importance of trees in the context of a rapidly urbanizing Pacific have led to the 

increasing elimination of trees from urban landscapes throughout the islands 

(Thaman, 2002). In rural areas, promotion of a wide range of export cash crops (e.g., 

coconut, banana, cacao [Theobroma cacao], sugarcane, coffee [Coffea spp.], ginger 

[Zingiber officinale], and butter pumpkin [Curcurbita maxima]) has led to the


clearing of diverse agroforests. This has been particularly serious in Tonga, where 

rapid expansion in the export of pumpkins to Japan has led to increasing use of the 

plow and clearance of multipurpose trees from agricultural holdings 3 . The Southeast 

Asian homegardens also experience a similar situation with varying degrees of 

commercialization (Abdoellah et al., 2006). When clearing land for short-term 

crops, trees in traditional agroforests used to be severely pruned or pollarded, but not 

killed, so they would regenerate after the crops have been harvested. However, in 

recent times they are commonly bulldozed, ploughed, deliberately killed with 

herbicides, or burned to make way for cash crops or for urban expansion. 

1.7. Constraints and limitations to homegardening 

Homegardeners face a number of problems in maintaining their traditional agroforests. 

These include poor soils, cost, and availability of land and water, insufficient 

time and labor, agricultural thefts, lack of planting materials, and lack of government 

assistance (Thaman, 1995). For example, drought is a major problem in Port 

Moresby, which has a 7-month dry season and has suffered prolonged droughts 

during the recent El Niño events. Gardeners must contend with the increasing 

unreliability of the overstretched, reticulated, water supply system and the failure of 

community faucets, regulations against the use of water for gardening purposes, and 

lack of alternative water supplies (Vasey, 1990). Restrictions on the use of water in 

gardens are also imposed during periods of extended droughts in Fiji. The atolls are 

also periodically affected by prolonged droughts, which commonly lead to the death 

of a significant proportion of the breadfruit, citrus, and other trees and food plants 

that are only marginally suited to the atoll environment 4 . 

Urban gardeners commonly have to contend with infertile, poor soils, such as the 

rocky or stony Lithosols of Port Moresby, the shallow soils that overlie a marl 

substrate in Suva, hydromorphic soils in low-lying areas, and the notoriously 

infertile, calcimorphic atoll soils of Kiribati, Tuvalu, and the Marshall Islands. 

Continual cropping on small urban plots also leads to declining fertility and loss of 

soil structure, unless ameliorative measures are taken (Thaman, 1995). Both water 

shortage and poor soils, however, often make trees a more attractive proposition 

than short-term ground crops, which require water and higher soil fertility. 

Insufficient land and insecurity of tenure are problems in most areas. More than 

half of all households in Suva, Fiji, said land shortage was a problem (Thaman, 

1983). Insecurity of tenure, especially in Suva, where a number of people had shortterm 

leases or were squatters, seems to be a major problem and a strong disincentive 

to homegardens and the protection of trees. City Council regulations, although not 

strictly upheld, have also been considered a disincentive; and have discouraged 

cultivation of ground crops and trees along road frontages, and the keeping of pigs, 

goats, cows, and horses within the city limits. Other problems for gardeners include: 

plant diseases, insects, snails, birds, rats, dogs, mongooses, and noxious weeds; theft 

of produce, especially of banana bunches and tree fruits; insufficient time; high costs 

of poultry feed and fertilizer; predation of firewood and deforestation on 

undeveloped urban and periurban lands where most low-income families still 

depend on firewood to cook their meals; boundary problems with respect to 

37

38 

R.R. THAMAN T 

ET AL. 

ownership of crops; and neighbors’ unfavorable response to gardening or livestock 

rearing (Thaman, 1995). 

In Port Moresby, hillside gardening has once again become the focus of 

criticism, on the grounds that it causes environmental damage, to the point that, in 

2005, the Prime Minister promised publicly a legislation to ban it (Quartermain, 

2005). Constraints to expanded homegardening are the greatest in Kiribati, Tuvalu, 

the Marshall Island, and Nauru, where extremely poor soils, limited water 

availability, and very high population densities, especially in South Tarawa and at 

Location, Nauru, are serious problems. Among the indigenous Nauruans, who are 

considered to be 100% urbanized, extremely high per capita incomes from 

phosphate royalties in the past and a resulting overdependence on imported foods 

seem to be the major disincentive to urban food gardening. The problem is 

complicated in Funafuti, where the soil from over half of the highest quality land on 

the main urban islet of Fogafale was excavated during World War II to build a 

runway, leaving only soil-less “borrow pits” of no agricultural utility. 

1.8. Future prospects of urban and homegarden agroforestry 

The importance of urban and homegarden agroforestry and its implications for 

planning are not clearly understood by most planners and policymakers in the 

Pacific islands because of a lack of quantitative data on its nature, extent, and 

cultural and ecological significance. There is little sign of a continuation of the 

interest once shown by some city planners and administrators. For example, the Port 

Moresby Housing Commission’s survey of urban gardening in the early 1970s and 

the studies by the Committee on Food Supplies of the Solomon Islands (1974) of the 

production of major staple crops (primarily sweet potato) in Honiara stressed the 

need to increase production per capita in both rural and urban areas. Fitzroy (1981) 

pointed out the correlation between vitamin deficiencies in “urbanized” people 

without garden plots in Honiara. Although further studies stressing the importance 

of urban and homegardens have been conducted since the mid-1970s, there is still a 

need for more information on the problems faced by gardeners, such as crops that do 

best under conditions of increasing pressure on land and deteriorating soils, best 

practices in terms of soil conservation and improvement, successful models for 

promoting urban and homegardening, and, models for the propagation and 

distribution of desirable cultivars of particularly useful plants. 

Nevertheless, there are some hopeful signs in favor of urban and homegarden 

agroforestry in the region. Among these are the continued efforts supporting the 

spread of kitchen gardening (“supsup” gardens) in Solomon Islands (Jansen et al., 

2001), recognition by the National Agricultural Research Institute and other bodies 

in Papua New Guinea of the continuing importance of urban gardening and the need 

for remediation of erosion problems (Quartermain, 2005), and the international Slow 

Food movement 5 , which promotes the appreciation of locally-grown food, and is 

gaining ground in Hawai’i.


It has been recognized that urban and homegarden garden agroforestry could 

help to prevent and alleviate poverty, reduce the alarming incidence of nutritional 

disorders and nutrition-related, non-communicable diseases, promote greater food 

security, reduce dependence on inferior imported medicines, fuels, ornamentation, 

handicrafts and other products and address environmental problems such as coastal 

erosion and pollution, loss of biodiversity and urban agrodeforestation (Thaman, 

1988). These practices can also stem the loss of traditional ethnobiodiversity (e.g., 

the uses, knowledge, beliefs, management systems, and languages; Thaman, 2004) 

of which trees, forests and tree-rich agroforestry systems constitutes a dominant 

component. Particular emphasis must be placed on the protection and enhancement 

of existing urban and homegarden agroforestry systems. Preserving and improving 

existing systems is an appropriate and cost-effective means of fostering the use of 

trees within the fabric of a rapidly urbanizing and homegarden-oriented Pacific 

island landscape. 

ACKNOWLEDGEMENTS 

We thank the countless Pacific island urban and homegarden agroforesters and 

gardeners, both male and female, who have shared knowledge of their plants and 

technologies and allowed us to spend time with them in their agroforests over the 

past decades. Many thanks also to Robin Hide for his generous help in writing this 

article. 

ENDNOTES 

1. Bourke R.M., Allen M.G. and Salisbury J.G. 2001. Food security for Papua 

New Guinea. Proceedings of the Papua New Guinea Food and Nutrition 2000 

Conference, ACIAR Proceedings 99. ACIAR, Canberra, Australia, 892p. 

2. Kennedy J. and Clarke W.C. 2004. Cultivated landscapes of the southwest 

Pacific. RMAP Working Paper 50. Resource Management in Asia–Pacific 

Program, Research School of Pacific and Asian Studies, Australian National 

University, 47p. 

3. Thaman R.R. 2004. Akau, A poto mo e ofa fonua: Ko e makatu unga 

, 

tukufakaholo (Trees, arboreal diversity and ethnobiodiversity: An agroforestry 

action plan for sustainable development in the Kingdom of Tonga). Technical 

Report 2004/04. Institute of Applied Science, University of the South Pacific, 

Suva, Fiji, 128p. 

4. Thaman R.R. 2004. Cool spots under threat: The conservation status of atoll 

biodiversity and ethno-biodiversity in the Pacific Islands. In: Lee K.J. and Tsai 

H.-M. (eds), Changing islands–changing worlds: Proceedings of Islands of the 

World VIII, pp 60–64. International Small Islands Studies Association (ISISA), 

Sydney, Australia. 

5. last accessed: January 2006. 

, 

, 

, 

39 

o e

40 


REFERENCES 

Abdoellah O.S., Hadikusumah H.Y., Takeuchi K., Okubo and Parikesit, 2006. Commercialization 

of homegardens in an Indonesian village: vegetation composition and functional 

changes. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 


Allen B.J., Nen T., Bourke R.M., Hide R.L., Fritsch D., Grau R., Hobsbawn P, and Lyon S. 

2002. Central Province: Text Summaries, maps, code lists and village identification. 

Agricultural Systems of Papua New Guinea. Working Paper 15. Revised edition. 

Department of Human Geography, Research School of Pacific and Asian Studies, The 

Australian National University, Canberra, Australia, 152p. 

Battaglia D. 1985. Bringing home to Moresby: Urban gardening and ethnic pride among 

Trobriand Islanders in the national capital. IASER Special Publication 11. Institute of 

Applied Social and Economic Research, Port Moresby, Papua New Guinea, 53p. 

Benjamin C. 1985. Some food market influences of a large-scale smallholder development in 

the West New Britain area of Papua New Guinea. Papua New Guinea J Agric Forest 

Fisher 33(3–4): 133 –141. 

Committee on Food Supplies (BSIP). 1974. Food and self-reliance: report of the Committee 

on Food Supplies. Government House, Honiara, Solomon Islands, 68p. 

Coyne T. 2000. Lifestyle diseases in Pacific communities. Technical Paper 219. Secretariat of 

the Pacific Community, Noumea, New Caledonia, 331p. 

Curtis M. 2004. The obesity epidemic in the Pacific Islands. J Dev Social Transfor 1: 37 – 42. 

Falanruw M.C. 1995. The Yapese agricultural system. PhD thesis. University of the South 

Pacific, Suva, Fiji, 386p. 

Fitzroy G.J. 1981. Commentary: Influence of development factors on nutritional patterns in 

the Solomon Islands. Ecol Food Nutr 10: 87 – 119. 

Hide R. 2003. Pig husbandry in New Guinea. A literature review and bibliography. ACIAR 

Monograph 108. ACIAR, Canberra, Australia, 291p. 

Ikagawa T. 1994. Residential gardens in urban Honolulu, Hawaii: neighborhood, ethnicity 

and ornamental plants. PhD thesis. University of Hawai i at Manoa, Honolulu, 208p. 

Jansen T., Kotali C. and Pitavavini G. 2001. Improving household food security in Lauru, 

Solomon Islands, through grass roots extension, kitchen gardens and nutrition education. 

In: Bourke R.M., Allen M.G., and Salisbury J.G. (eds), Food security for Papua New 

Guinea. Proceedings of the Papua New Guinea Food and Nutrition 2000 Conference, pp 

509–515. ACIAR Proceedings 99, ACIAR, Canberra, Australia. 

Lamanda N., Malézieux E. and Martin P. 2006. Structure and dynamics of coconut-based 

agroforestry systems in Melanesia: a case study from Vanuatu archipelago. In: Kumar 



Levett M.P. 1992. Urban gardening in Port Moresby: A survey of the suburb of Gerehu. Yagl- 

Ambu, Papua New Guinea J Soc Sci Hum 16(3): 47 – 68. 

Levett M.P. 1996. Fresh food production and marketing: national and Port Moresby 

perspectives. In: Gladman J., Mowbray D., and Duguman J. (eds), From Rio to Rai: 

Environment and development in Papua New Guinea up to 2000 and beyond: A quarter of 

next to nothing. Papers from the 20th Waigani Seminar Vol. 3, University of Papua New 

Guinea, Port Moresby, 302p. 

Levett M.P. and Uvano M. 1992. Urban gardening in Port Moresby: A survey of the suburbs 

of Morata and Waigani. Yagl-Ambu, Papua New Guinea J Soc Sci Hum 16(3): 69 – 91. 

National Statistical Office of Papua New Guinea. http://www.nso.gov.pg/2000_Census (last 

accessed: January 2006).


Parkinson S.V. 1982. Nutrition in the south Pacific—past and present. J Food Nutr 39(3): 

121 –125. 

Quartermain A. 2005. Food for thought. Nari Nius, Newsletter of the Papua New Guinea 

National Agric Res Inst 8(1): 9. 

Soemarwoto O.M., Karyono I., Soekar, Tadiredia W.M. and Raman A. 1985. The Javanese 

homegarden as an integrated agro-ecosystem. Food Nutr. Bull. 7(3): 44 – 47. 

Thaman R.R. 1976/77. Plant resources of the Suva Municipal Market, Fiji. Ethnomedicine IV 

1/2 (1976/77): 23 – 62. 

Thaman R.R. 1978. Cooperative yam gardens: An adaptation of a traditional agricultural 

system to serve the needs of the developing Tongan market economy. In: Fisk E.K. (ed.), 

The adaptation of traditional agriculture, pp 116–130. Development Studies Center 

Monograph No. 11. Australian National University, Canberra. 

Thaman R.R. 1982. Deterioration of traditional food systems, increasing malnutrition and 

food dependency in the Pacific Islands. J Food Nutr 39(3): 109 – 121. 

Thaman R.R. 1983. Urban gardening in Fiji: A direct means for sustainable national 

development. Proc Fiji Soc 13: 1 – 27. 

Thaman R.R. 1987. Urban agroforestry: The Pacific islands and beyond. Unasylva 39(55): 

2 – 13. 

Thaman R.R. 1988. Environmental issues in the Pacific Islands: Constraints to sustainable 

island development. Pacific Issues 1, pp 1 – 77. Pacific Circle Consortium, Woden, 

Canberra, Australia. 

Thaman R.R. 1992. Agrodeforestation as a major threat to sustainable development. In: 

Thistlethwaite R. and Votaw G. (eds), Environment and development: A Pacific Island 

perspective, pp 194–195. Asian Development Bank, Manila and South Pacific Regional 

Environment Programme, Apia, Samoa. 

Thaman R.R. 1995. Urban food gardening in the Pacific Islands: A basis for food security in 

rapidly urbanising small-island states. Habitat Int 19: 209 – 224. 

Thaman R.R. 2002. Trees outside forests as a foundation for sustainable development in the 

small island developing states of the Pacific Ocean. Int For Rev 4: 268 – 276. 

Thaman R.R. 2004. Sustaining culture and biodiversity in Pacific Islands with local and 

indigenous knowledge. Pac Ecologist (7&8): 43 – 48. 

Thaman R.R. and Clarke W.C. 1993. Pacific island agroforestry: Functional and utilitarian 

diversity. In: Clarke W.C. and Thaman R.R. (eds), Pacific Island agroforestry: Systems 

for sustainability, pp 17–33. United Nations University Press, Tokyo. 

Thaman R.R. and Thomas P.M. 1985. Cassava and change in Pacific Island food systems. In: 

Cattle D.J. and Schwerin K.H. (eds), Food energy in tropical ecosystems, pp 191–228. 

Gordon & Breach, New York. 

Thaman R.R., Eritaia B. and Faka’ osi S. 2001. Community-based biodiversity surveys and 

conservation action plans as tools for nature conservation in the Pacific Islands: Lessons 

learned from Fiji, Tonga and Kiribati (Melanesia, Polynesia and Micronesia). In: Miller S. 

and Sim J. (eds), Tools for conservation: 6th South Pacific Conference on Nature 

Conservation & Protected Areas (29 September – 3 October 1997), Pohnpei, Federated 

States of Micronesia. Volume 3: Conference papers, pp 225 – 285. South Pacific Regional 

Environment Programme, Apia, Samoa. 

Vasey D.E. 1990. On estimating the net social and economic value of urban home gardens. 

In: Landauer K. and Brazil M. (eds), Tropical home gardens, pp 203–213. United Nations 


41

CHAPTER 4 

AMAZONIAN HOMEGARDENS: 

THEIR ETHNOHISTORY AND POTENTIAL 

CONTRIBUTION TO AGROFORESTRY 

DEVELOPMENT 

R.P. MILLER 1 , J.W. PENN, JR. 2 , AND J. VAN LEEUWEN 3 

1 Instituto Olhar Etnográfico, SHIN CA 5 Conj. J Bl. B, Sala 105, Brasília-DF 

71505, Brazil; E-mail: . 2 Grand Valley State 

University, 1155 Au Sable Hall, Allendale, MI, 49401, USA. 3 Instituto Nacional de 

Pesquisas da Amazônia – INPA, Manaus, Amazonas, Brazil 

Keywords: Caboclo, Fruit tree domestication, Indigenous knowledge, Ribereño. 

Abstract. This chapter reviews how homegardens and a number of other traditional 

agricultural practices survived the aftermath of European conquest of Amazonia. The 

historical development of homegardens in Amazonia began with the evolution of agriculture 

and domestication of trees in prehistoric times, followed by the development of cultural 

complexes along the Amazon River and its main tributaries. These traditional societies, 

characterized by rich material culture and well-developed agricultural systems, were 

decimated by the combination of epidemics, wars and slavery that accompanied the European 

conquest. Yet, the homegardens survived in Amazonia, and today they represent the 

reorganization of the original indigenous practices within the context of the upheaval and 

changes brought by colonization and market economies, including the incorporation of 

introduced Asian fruit trees. Although homegardens near urban centers may provide income, 

in rural areas they are important chiefly for household subsistence. They are often the focus of 

experimentation with new tree species and cultivation techniques, and thus have the potential 

to contribute to the development of other agroforestry systems, and to extension efforts that 

seek alternatives for agricultural development in Amazonia. 


The local and regional diversity of Amazonian homegardens is best understood by 

studying their origins and how they have been influenced by the socioeconomic and 

43 




44 R.P. MMILLER 

ET AL. 

cultural forces that have shaped social organization and subsistence practices in the 

region, from prehistoric times to the present. This historical development begins 

with the evolution of agriculture and the domestication of trees in prehistoric times, 

followed by the emergence of complex cultures or chiefdoms along the main rivers, 

described by the first European explorers as exhibiting elaborate material culture and 

agricultural systems (Carvajal, 1542; Acuña, 1639). Although European conquest subsequently 

decimated these societies through a combination of epidemics, wars and 

slavery, as this chapter will show, a number of their agricultural practices, including 

homegardens, survived. 

The traditional l (i.e., prior to any interventions by research/extension agencies) 

homegardens of Amazonia represent a dynamic equilibrium of these original 

indigenous practices with the new social order and scenario created by the process 

of colonization. Included in this process was the incorporation of many Asian fruit 

trees introduced by the Europeans. The culture of traditional river-edge inhabitants, 

known as caboclos (in Brazil) or ribereños (in Peru) represents the fusion and 

synthesis resulting from this historical process, and homegardens today are an 

integral part of life throughout Amazonia. 

Some of these homegardens and their ethnoecology have been formally 

described in many scientific f publications (Denevan and Padoch, 1987; Padoch and 

de Jong, 1991; Smith, 1996; 1999; Coomes and Burt, 1997; Lamont et al., 1999; 

Denevan, 2002; Coomes and Ban, 2004), including some dissertations (Bahri, 

1992), Annals of the Brazilian Agroforestry Congresses, and other such records 

(e.g., Miller, 1994; van Leeuwen and Gomes, 1995; Rosa et al., 1998a; 1998b; 

1998c). Although a portion of this literature limits its scope to descriptions or lists of 

species found in the homegardens, some of these evaluate the factors determining 

choice of species, their management, and how proximity of markets influence these 

(e.g., Lamont et al., 1999). Based on this body of literature, and the personal 

experience of the authors in Amazonia, this chapter will attempt to reach some 

general conclusions as to the historical and cultural importance of homegardens, and 

how this can be linked to the underlying processes of the relationship between 

humans and cultivated trees. An understanding of this relationship is essential for 

evaluating the potential contribution of homegardens to extension efforts that seek 

alternatives for agricultural development in Amazonia, and some suggestions will be 

made along this line. 

2. ETHNOHISTORY OF HOMEGARDENS IN AMAZONIA 

2.1. Pre-historical development of agriculture and homegardens in Amazonia 

Archeological evidence from the lowland neotropics in Colombia, Ecuador, Peru, 

and Mesoamerica indicates that between 10 000 and 8600 b.p. (before present) 

horticulture emphasizing both native tubers and seed plants was taking place outside 

Amazonia (Piperno and Pearsall, 1998; Piperno et al., 2000; Smith, 2001). However, 

in a site in Rondônia, in western Brazilian Amazonia, where human occupation by 

hunter-gatherers dates back to 9000 b.p., vestiges of agricultural activity, in the form

AMAZONIAN HOMEGARDENS 

of processing utensils, only begin to appear around 4500 b.p. (Miller, 1992). Lathrap 

(1977) argues that the earliest agriculture in Amazonia was probably adjacent to 

dwellings, along or near rivers in forests that did not require frequent clearing. At 

some moment, native fruit trees were domesticated and incorporated into these 

prehistoric agricultural systems. This process may have occurred initially through 

the ‘dump heap’ (sensu Anderson, 1952) or incidental route to domestication, when 

seeds of edible fruits collected in the forest were discarded near dwellings. Although 

little information is available on the sequence of domestication for neotropical tree 

crops, it is likely that this was concurrent with the domestication of root-crops, as 

the maintenance of gardens near dwellings would have provided an ideal location 

for the discarded seeds of useful tree species to germinate and grow. The recognition 

and management of such 'volunteers' would have been the first step along the road to 

their domestication. 

By 3000 to 2000 b.p., agricultural development made possible the existence of 

larger villages of many hectares on the middle and lower Orinoco River in 

Venezuela, and by 2000 years ago, large, socially stratified chiefdoms were thriving 

along the principal rivers of Amazonia. There is evidence of crop domestication and 

diffusion from this period of Amazonian history. For example, Salick (1992) has 

found that the domestication and exchange of cocona (Solanum sessiliflorum), 

common to Western Amazonian homegardens today, may have begun as long as 

2000 years before present. When the first European explorers arrived in Amazonia 

in the 16th century, large population complexes, exhibiting an elaborate material 

culture and ceremonial art, occupied the margins of the main rivers, with links to 

surrounding regions through extensive trade networks (Roosevelt, 1994). From the 

description by Jesuit friar Gaspar de Carvajal, in his account of the first European 

exploration of Amazon in 1541–`42, we know that part of this cultural development 

consisted of agricultural systems based on a great variety of cultivated plants, 

including fruit trees, and the storage of various foods such as cassava (Manihot 

esculenta), maize (Zea mays), dried fish, and penned river turtles (Carvajal, 1542). 

Although the existence of some sort of homegarden is clear in these historical 

accounts, little detail is provided on the nature of these indigenous agroforestry 

systems. Carvajal, for example, mentions only that “much fruit of all kinds” was 

found in one village, and that fruit trees were planted on either sides of the road 

leading to another village (Carvajal, 1542). In all, at least 138 species of plants are 

thought to have been under cultivation or management at the time of European 

arrival in Amazonia, of which 68% were trees or woody perennials (Clement, 

1999a). Besides the species mentioned in historical accounts, it is possible that in 

pre-Columbian times many more species were also cultivated, or were in a state of 

incipient domestication. A number of commonly cultivated Amazonian fruit trees 

have the characteristics of long periods of selection and genetic improvement. 

Clement (1989; 1999b) suggests the existence of a pre-Columbian center of crop 

diversity in Western Amazonia, based on the genetic diversity of fruit tree 

domesticates. In terms of their manipulation of plant resources, pre-Columbian 

cultures in Amazonia appear to have operated along a gradient of domestication, 

with plants fully domesticated and reliant on human care for their dispersal and 

survival at one extreme, as is the case of the peach palm (Bactris gasipaes). At the 

45

46 

R.P. MMILLER 

ET AL. 

other extreme of this gradient were those wild plants that may be found in greater 

than normal concentrations around ancient village sites, as a result of agricultural 

clearing and burning, with the possible favoring of their regeneration, but which do 

not exhibit any apparent genetic differentiation from their wild counterparts. 

Between these extremes are found a number of interesting and useful plants, 

suggesting that an active process of genetic selection and domestication was taking 

place in pre-Columbian Amazonia. An example of how this process may have 

occurred (and continues to occur) is described by Schroth et al. (2004), for the palm 

Astrocaryum tucuma in the Manaus region. Nevertheless, for the most part, the 

continuing domestication of wild species was truncated by the European conquest. 

In less than 200 years after the events described in Carvajal’s report (Carvajal, 

1542), the great chiefdoms along the Amazon had succumbed to epidemics of 

imported diseases such as smallpox and measles, wars, and enslavement. Their 

sophisticated culture and political and trade networks collapsed, and large stretches 

of the Amazon River and its tributaries were totally deserted (Daniel, 1776). 

Despite the decimation of native Amazonian populations that occurred during 

European conquest, with an ensuing loss of agrobiodiversity, many elements of their 

agricultural and agroforestry systems survived and can be seen among the modern 

tribal groups. The agroforestry practices of some of the tribal peoples in Amazonia, 

reviewed in Miller and Nair (2006), range from the cultivation of fruit trees and 

other useful plants around dwellings (homegardens), to the incorporation of trees 

in agricultural fields and fallows, which may involve practices such as actively 

planting or managing useful tree species or sparing seedlings that regenerate 

naturally. The homegarden of fruit trees, condiments and medicinal plants may 

grade into a belt of fruit trees surrounding a village, fruit trees interspersed with field 

crops, orchards of mixed fruit trees, and fallows of forest species enriched with fruit 

trees – these last mentioned configurations having been termed “swidden-fallow 

agroforestry” (Denevan and Padoch, 1987; Denevan, 2002). Although there are 

exceptions, as in the case of tribes with a very rudimentary agriculture, for the most 

part, homegardens can be considered as an important component of the subsistence 

technologies and cultural knowledge of Amazonian tribes. 

Whether the specific cultivation methods employed by contemporary indigenous 

groups are the same as those of their pre-colonial ancestors is a difficult question 

to answer. Nevertheless, it is probable that the agroforestry systems practiced by 

indigenous peoples as well as the caboclos and ribereños are direct descendants 

of the systems in existence prior to European arrival, with the addition of a number 

of exotic species of fruit trees. This contribution of exotic species introduced by 

Europeans is discussed in the context of the ethnohistory of caboclo and ribereño 

culture, the subject of the following section. 

2.2. Ethnohistory of caboclo and ribereño culture and homegardens in Amazonia 

Although the use of the term caboclo has been criticized due to its negative social 

connotations (Lima, 1999), it is difficult to substitute, as it encompasses both 

colloquial as well as academic meanings in Brazil, and is a broad descriptor of a 

regional form of life and natural resource use. While modern-day tribal groups of


Amazonia in most cases represent the fragments of populations and cultures that 

escaped to survive and regroup following the colonial holocaust, caboclo society in 

Brazil or ribereño society in Peru and their cultures are the result of the fusion of the 

remnants of the native populations, decimated during colonization, with European 

and African racial and cultural elements (Padoch and Pinedo-Vasquez, 2001; 

Ribeiro, 1997). In this process, agricultural, social, economic, and belief systems 

were reconfigured and reconstructed upon an existing knowledge base of ecological 

systems and subsistence practices, with the addition of new tools and technologies. 

Key players in this process were the Catholic missionaries in Amazonia. As allies to 

the colonial economic system, they had a major role in providing an ideology for 

the domination of the native populations and their transformation into a labor force. 

Along with the forts, missions were fundamental elements in guaranteeing the 

domination of the region by the Portuguese from 1650 – 1750, and allowing 

the functioning of commerce (Alves-Filho et al., 2005). 

Despite the superiority of Portuguese armaments, the native peoples did not 

submit easily to Portuguese attempts to enslave or otherwise conscript them as 

agricultural workers growing subsistence and commercial crops, collectors of forest 

products (such as cacao, Theobroma cacao), in the construction of public works, and 

other forms of labor, without which the colonial economy in Brazil would have 

collapsed (Alves-Filho et al., 2005). In response, they waged war, rebelled in 

villages and missions, deserted from royal services, massacred when possible their 

enemies, and even made peace treaties when convenient (Santos, 2002). Elsewhere 

in Amazonia, natives also put up fierce resistance, lasting well into the republican 

period of the former Spanish colonies, especially in Peru and Colombia (San 

Ramon, 1994; Stanfield, 1998; Rios, 2001). 

The search for cacao using Indian labor, primarily from stands of wild or feral 

trees, motivated the Portuguese to range far upriver, leading Portuguese incursions 

west into Spanish territory (now Peru) to kidnap Indians on the Marañon River 

during 1686 – 1723 (Edmundson, 1922). By 1730, cacao had become the region’s 

dominant export, remaining so for more than a century (Alden, 1976; Hemming, 

1987). Cacao gathering expeditions had ceased by 1750 and cacao was being 

cultivated in plantations along the Amazon. Farmers grew seedlings on raised beds 

for a year, and then transplanted them into their cassava fields, where banana plants 

(Musa sp.) had been previously planted to provide shade. Native fruit trees, along 

with introduced species, such as orange (Citrus sinensis) and avocado (Persea 

americana), were also interplanted with cacao, as it was known that cacao produced 

better in shade (Daniel, 1776). Cacao appears to have been an important, if not the 

principal, economic element of the agroforestry systems of that time. By the mid- 

1800s, another exotic species, coffee (Coffea arabica), was one of the main 

agricultural exports of the region, along with cotton (Gossypium sp.), cacao, guaraná 

(Paulinia cupana), and tobacco (Nicotiana tabacum) (Amazonas, 1852). 

By 1875, the rising demand for rubber, an important material for the Industrial 

Revolution, led to an economic boom in Amazonia. Rubber, extracted from the 

forest tree Hevea brasiliensis, had by 1880 become the third most important export 

in Brazil and Peru (Stanfield, 1998; Homma, 2003). The caboclo population, 

concentrated on the Amazon and Solimões Rivers, spread out through the entire 

47

48 

R.P. MMILLER 

ET AL. 

basin in search of rubber trees. A mixture of caboclo, mestizo, European, and 

indigenous (tribal) gatherers tapped the forests of Peru, Colombia, and Bolivia; and 

Manaus, Belém, and Iquitos grew into the principal commerce centers along the 

Amazon River. The boom attracted many migrants as well as absorbing the local 

labor force, with the result that agricultural production in Amazonia dropped sharply 

(Ribeiro, 1997; Stanfield, 1998). The rubber boom also brought disastrous 

consequences to the remaining forest tribes, as rubber tappers penetrated even the 

most distant headwaters. The atrocities committed against the Indians and their 

conscription as forced labor were so widespread that they attracted international 

attention (Renard-Casevitz, 1992; Stanfield, 1998). With the drop in agricultural 

production, food prices soared. Tribal societies involved in the trade could do little 

farming, suffered from severe hunger, and often lost their lands to rubber tappers 

(Stanfield, 1998). Where they survived, homegardens undoubtedly played a key role 

in providing food for rural inhabitants, regardless of their ethnicity. 

The crash in rubber prices returned Amazonia to the state of an economic 

backwater by the end of the First World War (Homma, 2003). Indigenous knowledge, 

so important to the European and mestizo efforts to cultivate and exploit the 

most economically lucrative resources of the region, lay dying in the form of 

abandoned fields across the wide swaths of Amazon basin. According to Denevan 

(2002), homegardens in Amazonia became less important and poorly developed 

after the arrival of Europeans, mostly because indigenous villages changed their 

locations much more frequently than they did in the past, yet another consequence of 

this tragic history. 

2.3. Transformation of traditional agriculture during colonial times 

Although the Portuguese introduced a number of new crops to Amazonia, such as 

sugarcane (Saccharum officinarum), indigo (Indigofera indica), and rice (Oryza 

sativa), as well as domestic animals, indigenous agricultural practices remained the 

basis for subsistence, and they were also adapted for the production of commercial 

crops such as cacao. At the same time that technology guaranteed Portuguese 

military superiority, agricultural technology in the form of steel tools resulted in the 

transformation of indigenous practices, with stone axes and digging sticks being 

substituted by steel axes, machetes, hoes and brush hooks. Where previously large 

trees were ringed with stone axes and left to dry slowly, and saplings were 

bludgeoned over (Daniel, 1776), steel tools greatly reduced the labor expended in 

agricultural clearing, with the result that what is considered today as “slash-andburn” 

agriculture probably is quite different from what was practiced in pre- 

European Amazonia. Pre-Columbian agriculture most likely had greater affinity 

with slash-mulch systems, as fires used to prepare fields would have been much less 

intense, and ringed trees would slowly drop a layer of leaves over the field. The 

initial difficulty in opening fields out of forest probably led to a longer use of 

cleared areas, through complex polycultures and crop sequences, including trees. 

A more extended use of fields may have been possible due to the input of organic 

matter from the slowly dying original vegetation.


Catholic missions were in part responsible for the introduction of new 

technologies and agricultural practices. The Jesuit missions in particular were 

generally well-managed enterprises that exported a part of their production. 

Persuading natives to leave their villages and move to these missions involved a 

number of strategies, besides force, including convincing them that epidemics of 

introduced European diseases were caused by the insalubrities of their village sites. 

In some cases, life in a mission was the only alternative to being attacked and 

enslaved by colonists. 

Life in the missions brought together individuals of separate tribes, with different 

languages and cultures, for the compulsory adoption of the body of beliefs and 

customs of the colonizer. The cultural result was a patchwork of beliefs, the 

syncretism of shamanism with a vague observance of Catholic saints and holidays, 

the base for a “folk Catholicism,” incorporating various native practices and beliefs 

and the colonial influences of the Portuguese, as well as African slaves (Ribeiro, 

1997; Maués, 2001). Some of these beliefs are associated with a variety of 

magical/medicinal plants (e.g., pião roxo, Jatropha gossypiifolia) often cultivated in 

modern homegardens, and which along with ornamentals, are often seen even in 

diminutive front yards in cities such as Manaus. 

A characteristic of European colonization of Amazonia was the introduction of a 

number of exotic fruit trees. In 1662, Mauricio Heriarte (in Huber, 1904) described 

Belém as cheerful and full of fruit trees such as oranges, limes (Citrus aurantifolia), 

sweet limes (Citrus limetta) and biribás (Rollinia mucosa). The introduction of 

mango (Mangifera indica) to Belém in 1780 is credited to the Genovese architect 

Antonio Landi, who brought seeds from Bahia, the capital of Brazil until 1763. The 

Portuguese Crown officially sponsored a number of plant introductions from its 

eastern colonies of Goa (India) and Macau (China) and the establishment of a 

botanical garden in Belém (Dean, 1995). In 1808, in retaliation for the invasion of 

Portugal by France, the Portuguese invaded French Guiana and were able to take 

advantage of the collection of useful plants cultivated in Cayenne’s botanical 

garden. By the time Cayenne was returned to the French in 1818, a number of 

tropical species had been sent to Belém, along with unspecified European fruit 

trees that had been acclimated in Cayenne (Holanda, 1965). Coffee was another 

introduced tree crop that soon proved lucrative for Brazil by the 1800s. Coffee 

germplasm was introduced to Belém in 1727 by Sargeant-Major Francisco de Mello 

Palheta, who transported five coffee seedlings and a handful of seeds from Cayenne. 

The first sample of coffee grown in Pará was sent to Lisbon in 1732, and two years 

later in 1734, 45 tons were shipped (Homma, 2003). 

By the mid-19th century, exotic fruit trees had been fully incorporated into 

homegardens along the Amazon. Traveling on the Amazon between Óbidos and 

Manaus in 1849, the British naturalist Henry Walter Bates described homegardens 

with banana, papaya (Carica papaya), mango, orange, lemon (Citrus sp.), guava 

(Psidium guajava), avocado (Persea americana), abiu (Pouteria caimito), genipap 

(Genipa americana), and biribá, as well as coffee shrubs growing under the shade of 

the fruit trees (Bates, 1863). Ten years later, French traveler Robert Avé-Lallemant 

recorded a variety of fruit trees growing near houses on the outskirts of Belém: 

banana, mango, jackfruit (Artocarpus ( heterophyllus), 

various Annonaceae, orange 

49

50 

R.P. MMILLER 

ET AL. 

trees, coffee, as well as the giant granadilla or maracujá-açu (Passiflora quadrangularis). 

Surrounding the dwellings of Indians near Cametá, Pará, he found native 

calabash trees (Crescentia cujete) and orange trees competing with mango, and the 

native açaí (Euterpe oleracea) and bacaba (Oenocarpus bacaba) palms. The 

presence of various Annonaceae, the bacuri (Platonia insignis) and brazilnut 

(Bertholletia excelsa) trees was also noted. Besides the homegarden, other tree 

species were cultivated as commercial crops, and income sources for these households 

came from “extensive stands of cacao” and rubber trees. Continuing up the 

Amazon to Santarém, he found many cacao and orange groves, as well as 

concentrations of the native tucumã palm ( (Astrocaryum vulgare), 

highly appreciated 

for the edible mesocarp of its fruits (Avé-Lallemant, 1859). 

In Peru, coffee, mango and avocado germplasm entered the Amazon Basin from 

both the east and west. Avocado entered Peru and the Peruvian Amazon well before 

the arrival of the Spaniards, while coffee and mango cultivars in Amazonia were 

introduced from either direction. Accounts from early explorers suggest most mango 

germplasm came from coastal Peru. Besides Asian species, the Spaniards also 

brought plant species from and via Central America and the Caribbean. Thus, we 

might expect common crops of the colonial era such as bananas, beans (Phaseolus 

vulgaris), citrus, or sugarcane in the Peruvian Amazon to have diverse origins even 

soon after their introduction to the region. Explorers such as Eduard Poeppig, who 

studied the upper Amazon in 1829-31, have found that much of the cassava 

germplasm in Peru came from downriver in Brazil, while banana germplasm as far 

downriver as Manaus, Brazil, often came from Peru (Poeppig, 2003). 

By no means, however, was the cultivation of trees limited to the traditional 

pattern of homegardens or commodity crops. Some homegarden species were 

creatively adapted to other uses, as is the case of the yellow mombin (Spondias 

mombin; Smith, 1999) and the calabash tree for live fences in the várzea (floodplain) 

region. Similarly, other species that were not previously cultivated, such as the 

munguba (Pseudobombax munguba), a common tree of the várzea, were enrolled to 

mark property boundaries on floodplain ranches. Species such as the rubber tree 

were added as economic elements, as a small rubber boom during World War II led 

to a renewed interest in this crop, and a low level of tapping continued even after 

the war. 

2.4. The caboclo and ribereño in the regional economy 

While colonization caused the demise and/or slow absorption of the indigenous 

tribal populations, a new hybrid society of non-tribal peoples was on the rise. The 

caboclos of Brazilian Amazonia are of mixed descent, as well as the remnants of the 

acculturated tribes. Similarly, the ribereños in Peru are of mixed European and 

Amerindian descent. Despite the persistent use of the term in the literature, these 

rural inhabitants do not actually call themselves “ribereños.” They most often refer 

to themselves in occupational or class terms such as pescador r (fisherman) or 

chacarero, as chacra is the common name for the plots of land they farm (Penn, 

2004). Researchers point to the Cocama-Cocamilla tribal origins of ribereños in 

Peru, but ribereños have diverse origins, and it is not advisable to generalize about


their ethnicity. The origin and ethnicity of the Cocama-Cocamilla themselves is still 

poorly understood (Cabral, 1995). 

Although very similar to the original native populations in terms of their 

ecological adaptations and subsistence practices, the caboclos in Brazil were very 

different socially (Ribeiro, 1997). Historically, they have been embedded in an 

agricultural and extractive economy, trading raw materials and products collected 

from the forests and rivers, or grown in their fields, for the manufactured items and 

tools necessary for their subsistence. For the most part, there was an ample supply 

of land for the harvest of extractive products and for fields, under communal tenure 

or belonging to absentee owners and defunct rubber estates. In recent decades, 

however, this situation has changed as development of a different form has reached 

Amazonia, with the construction of roads shifting the economic axes away from rivers 

and floodplains to the terra firme, where human occupation has been characterized by a 

moving frontier of logging, ranching, and agricultural colonization, that leaves in its 

wake a landscape dominated by pasture and to a lesser extent swidden agriculture. 

As rights to land have become more disputed, homegardens have taken on another 

socioeconomic function, with the presence of cultivated trees used as proof of land 

tenure and property rights. 

3. HOMEGARDENS IN PRESENT-DAY CABOCLO AND RIBEREÑO 

SOCIETIES 

Homegardens in Amazonia are variously referred to in folk denomination as 

“huertos” or “jardíns” (in Peru), and “quintais” (yards) or “sítios” (homesteads) 

in Brazil, as well as “pomares caseiros” (home orchards) or “miscelânea” by researchers. 

They combine native species with fruit trees introduced from other parts of 

the globe during European colonization, as well as more recent introductions. In a 

survey of 33 upland homegardens across the Brazilian Amazon, Smith (1996) found 

a total of 77 tree species, of which 46% are indigenous to Amazonia, and 27% are 

from the Old World. In a study of 51 homegardens in Peru (Lamont et al., 1999) at 

least 30 of the 161 species found were exotics, including nine tree species. In the 

three villages (two of the Yagua tribe and one ribereño), the two most common 

species in all 51 gardens were of Asian origin (i.e., mango and banana). 

The importance of homegardens is chiefly the domestic supply of fruits, 

condiments, medicines, craft materials, and shade. Near urban centers, however, 

they may become part of both subsistence and income-earning strategies through the 

production of marketable fruits. How farmers manage the composition of their 

homegardens in order to influence production and income generation has been little 

studied, but it appears that there is a ubiquitous stock of species valued for domestic 

consumption, while others are cultivated specifically as income-earners. Homegardens 

near Iquitos, Peru, may cultivate native palms for use in the handicraft 

business (Lamont et al., 1999), or exotic species such as taperibá (Spondias dulcis) 

for their prized fruits. In the Colombian Amazon, lulo (Solanum sessiliflorum) is 

common in homegardens to supply the markets of Leticia, while the market for fruit 

from the ocoró tree (Rheedia spp.) makes it popular in homegardens near Santa 

Cruz, Bolivia (J. Penn, pers. obs.). 

51

52 

R.P. MMILLER 

ET AL. 

Amazonian homegardens are very diverse in terms of size and number of 

species, both on a local level, with properties in the same community exhibiting very 

different assemblages, as well as on a regional level. While some of these differences 

can be explained, it becomes clear that there is no such thing as a “typical” 

homegarden, only trends or patterns. The 21 homegardens studied by Padoch and de 

Jong (1991) in the community of Santa Rosa, 150 km upstream from Iquitos, 

generally covered between 300 to 700 m 2 

m , the size of a usual house lot in that 

community. However, the range in size was from 67 to 7322 m 2 . Outlying houses 

had larger gardens, but this was not always the case. A typical pattern observed in 

many parts of Amazonia is for houses to be located in the central area of the 

community, where school, church, meeting hall, soccer field, and television are 

normally found. These hamlets can be part of planned “agrovilas” of colonization 

projects, or spontaneously formed communities (often based on kin ties) that group 

together in order to be attended by municipal services such as schools, health posts, 

or power generators. In these cases agricultural fields are located at a distance, and 

some sort of homegarden may be found surrounding the shelter used for processing 

the cassava crop. 

Homegardens in Amazonia also must be studied in the context of how dynamism 

and change affect the economic, social, and cultural aspects of caboclo and ribereño 

societies. A community of 60 households near Iquitos, Peru, whose homegardens 

were studied by Coomes and Burt (1997), for example, was originally founded as 

an agricultural estate for the production of sugarcane, rum, and fuelwood, and subsequently 

was divided up among the former workers in 1971 as an act of agrarian 

reform. In the community studied by Padoch and de Jong (1991), also near Iquitos, 

life histories of the adults were found to typically include several long economic 

migrations and many changes of residence. Lamont et al. (1999) found that the 

intermarriage of ribereños within families of the Yagua tribe was associated with 

declining use of homegardens in Peru, indicating that researchers need to examine 

the resilience of these agricultural systems to social and cultural change. 

Further study is needed to determine the extent to which differences in 

homegarden size and diversity are random, a product of local processes of sociocultural 

development and germplasm accession, or whether they reflect changes in 

management choice with regard to cash and energy flows and the perceived 

functions of the homegardens. In some cases, traditional homegardens may be 

eliminated to make place for more profitable plantations, if agricultural land (space) 

increases in value, as has been observed in the region near Manaus. If the farmer has 

the means to invest in a profitable crop, the homegarden can be eliminated to plant 

papaya (Carica papaya) or passionfruit (Passiflora edulis), or if still closer to 

Manaus, to plant horticultural crops (e.g., okra, Abelmoschus esculentus). This 

happens especially on better soils, such as anthropogenic black earths or the várzea 

alta, the higher part of the floodplain or natural levee that accompanies the Solimões 

and Amazonas rivers (J. van Leeuwen, pers. obs.). Penn (2004; 2006) found that 

homegardens in Peru were being planted with camu camu trees (Myrciaria dubia) 

by ribereños anxious to participate in a regional development program that promoted 

the cultivation of this species, extremely rich in vitamin C.


A category of Amazonian homegardens originating from rubber-cacao plantations, 

in which an upper stratum of rubber tree canopies is combined with a lower stratum 

of cacao, frequently is found on the várzea alta of the rivers Solimões, Amazonas, 

and Madeira. The cacao and rubber trees of this two-layer system are always quite 

old (J. van Leeuwen, pers. obs.). On the Ilha de Careiro, cacao and rubber were 

planted at the beginning of the twentieth century when production of these two 

commodities was much more profitable, but planting no longer occurs (Bahri, 1993). 

On the Ilha de Careiro and elsewhere many cases can be seen of the gradual 

substitution of cacao and rubber by other fruit trees, with the result that the plantation 

develops into a multispecies homegarden (Bahri, 1992; 1993). These examples indicate 

that homegardens can have a long history, in the sense that present day species 

composition does not necessarily closely reflect current economic scenarios. This is the 

case in Central Amazonia, where várzea homegardens may contain rubber trees 

that have not been tapped for many years. Although the presence of species that 

presently have little economic contribution may simply result from low levels of 

management, and not a conscious effort of conservation, their maintenance may also 

be part of risk-avoidance strategies. Poor farmers will generally refuse to cut a tree if 

it is thought that it might be useful at some moment in the future (J. van Leeuwen, 

pers. obs., based on work with small farmers in Mozambique and the Amazon). 

Differing time horizons and expectations of farmers with regard to local market 

demands, land tenure and property size all can influence the configuration of 

homegardens and other agroforestry systems. Access to the markets of larger urban 

centers represents an important economic factor that comes into play. Studies by 

Rosa et al. (1998a; 1998b) near the state capitals Macapá (Amapá), and Belém 

(Pará), Brazil, for example, found that small livestock can have considerable 

economic importance as components of the homegarden system. In properties 

averaging 90 ha near Macapá, although more than 50% of the chickens, ducks, and 

pigs raised was consumed by the household, weekly revenue from livestock 

averaged R$ 35, a value greater than that obtained from the sale of fruits such 

as açaí, bananas, mangos, limes, and cupuaçu (Theobroma grandiflorum), which 

averaged R$ 20/week [the real (R$) was approximately equal to the US dollar at that 

time and is now exchanged at R$ 2.3 per US$]. Nevertheless, a good portion of the 

feed for these animals was said to come from homegarden fruits. In a survey of 20 

households near Belém, where property size averaged 1.7 ha, it was found that 

families consumed 69% of the fruits, 100% of the medicinals, 85% of the 

vegetables, and 85% of the livestock, with the remainder being sold (Rosa et al., 

1998c). Conversely, livestock can destroy homegardens, and make it impossible to 

maintain or restart a homegarden. The introduction of water buffalo near Iquitos has 

greatly reduced the number of homegardens where these animals are present 

(J. Penn, pers. obs.). 

4. HOMEGARDEN MANAGEMENT IN AMAZONIA 

According to Lathrap (1977), the maintenance of homegardens and clean yards 

around the dwellings of indigenous communities creates a domesticated microcosm 

out of the surrounding wild forest, otherwise the abode of spirits and other dangers. 

53

54 

R.P. MMILLER 

ET AL. 

In Waimiri Atroari villages in Central Amazonia, this zone is used by small 

children, who both forage and play at activities such as shooting lizards with toy 

bows and arrows (R. Miller, pers. obs.). Although the extent to which Lathrap’s 

cosmological interpretation of the significance of homegardens can be applied to 

caboclo and ribereño societies may be limited, the maintenance of a terreiro, or 

patio (bare-earth yard) often swept daily, is a ubiquitous feature of rural homes in 

Amazonia, and serves to reduce hiding places for snakes and insects. The size of this 

yard is typically about 500 m 2 (20 x 25 m) and may often be larger. The exact limit 

of the terreiro, however, may depend on the time and labor available for weeding. 

Beyond the terreiro, the divide between the homegarden of planted trees and 

neighboring second growth may not be clearly distinguishable. These fluctuating 

boundaries between the bare earth yard, the homegarden, and encroaching second 

growth vegetation are important in permitting the establishment and recruitment of 

volunteer seedlings of useful trees. Discarded or fallen seeds will germinate in the 

shelter of leaf litter and undergrowth, and resulting seedlings may be spared by the 

observant farmer during periodic weeding. This process was noted by Huber (1904), 

who was probably the first to make specific mention of the ease with which even 

introduced species of fruit trees in Amazonia become sub-spontaneous, germinating 

from discarded seeds in the more fertile soil around dwellings. This “spontaneous” 

aspect of homegardens is in fact an important form of management. Near Iquitos, for 

example, Padoch and de Jong (1991) found homegardens to be a “combination of 

trees left from pre-existing fallows or forests, deliberately planted vegetation, 

spontaneously occurring useful forest plants, species transplanted from the forest, 

seeds germinating from the forest,” resulting in mosaics of different-age vegetation. 

They also found that 14% of the plants identified as “non-cultivated” were useful 

and had been selected for in previous weeding operations. This process, also 

important for outlying fields, fits into what Wiersum (1996) described as the 

“second stage of domestication,” and is suggestive of how trees may have been 

incorporated into agricultural systems in Amazonia during the past millennia. Some 

species will simply regenerate more easily than others in these environments. This is 

a major reason why Rheedia, Genipa, and Inga species are so common in homegardens 

along the Peruvian Amazon (Penn, 2006). 

Areas beyond the yard that are not kept “clean” provide a dumping ground for 

assorted household and garden wastes, which besides being important as sources of 

seeds and forage for domestic fowl, can also represent significant nutrient additions. 

Over millennial time scales in Amazonia, humans have generated patches of higher 

fertility around their dwellings by concentrating nutrients obtained from surrounding 

terrestrial and aquatic ecosystems, resulting in anthropogenic “black earths” 

(Lehmann et al., 2004). Data from a hunting study with the Waimiri Atroari tribe in 

Central Amazonia (Mazurek, 2001) indicates that an average-size village of 50 

people discards approximately 1.5 Mg of bones of game animals every year. Bones 

represent a significant contribution of calcium and phosphorus, which complement 

the other nutrient elements found in other forms of household wastes. Although 

redirecting nutrients can be a conscious practice, such as when farmers place 

cassava peelings at the foot of selected fruit trees as fertilizer, for the most part, the 

nutrient peak around dwellings that greatly benefits homegardens is an unconscious


practice. Nevertheless, in the case of Waimiri Atroari villages, the zone of greater 

fertility is explored for the initial establishment of a belt of fruit trees around the 

communal dwelling, which then expands outward concentrically (Miller, 1994). 

5. IMPORTANCE OF HOMEGARDENS FOR AGROFORESTRY 

DEVELOPMENT IN AMAZONIA 

Throughout history, Amazonian farmers were subjected to exploitation as forces of 

colonization and trade penetrated the region. They have suffered immensely and 

have often been dispossessed of their traditional lands, but have shown a remarkable 

ability to adapt to new environments and socioeconomic scenarios. During this 

period, their homegardens have changed in many ways. Asian species soon became 

common in homegardens after the Conquests, and are an increasingly common part 

of these cultural landscapes. Among the various configurations of agroforestry 

systems, such as tree/crop combinations in fields, orchards of mixed fruit trees, and 

enriched fallows, homegardens represent the most widespread agroforestry practice 

employed by farmers in Amazonia today. 

Although farmers near urban centers sell homegarden products (Lamont et al., 

1999) as well as livestock (principally fowl) raised in and around homegardens, their 

overall contribution for domestic consumption is probably more important. In this 

regard, homegardens represent a robust and time-tested technology, employed by the 

traditional inhabitants of Amazonia, whether indigenous tribes or caboclos and 

ribereños, and from the point of view of food security, they may be of great value on 

agricultural colonization frontiers, where farmers face a difficult struggle to 

establish themselves and their families. 

Originally managed for subsistence according to ethnic practices, homegardens 

are now increasingly important for farmer experimentation with commercial crops. 

As the locus of experimentation with new tree species and cultivation techniques, 

homegardens have the potential to contribute to the development of other agroforestry 

systems, and may expand into more commercial groves, as discussed by 

Penn (2004) on the new camu camu industry in Peru, and Yamada and Osaqui 

(2006) concerning the farmers of Japanese descent in Tomé-açu, Pará, Brazil. 

The homegarden can function as a “staging area” for testing new species and 

storing, safeguarding, and multiplying germplasm for transfer to and between fields 

(Coomes and Ban, 2004). In this manner, the homegarden can be an integral 

component of the larger agricultural system of the property as well as a key node in 

the local network of agrobiodiversity, if one considers the exchange of plant genetic 

resources between households in a community. 

The historical study of the course of development of homegardens as a basic unit 

of interaction between humans and trees holds lessons relevant to the present-day 

scenario of advancing deforestation, in which agroforestry is ascribed a potential 

role in developing more sustainable land use. While the technologies or practices 

involved in expanding agroforestry systems out to fields are not necessarily those 

employed in homegardens, they entail similar concepts such as tree culture, nutrient 

cycling, and permanent soil cover, among others, and in this respect, homegardens 

55

56 

R.P. MMILLER 

ET AL. 

could be considered as a conceptual core for agroforestry development. The basic 

units of information that farmers need to develop new models of agroforestry 

systems are in essence the knowledge of tree species, as to their behavior and 

interaction with other species. Homegardens, where trees can more easily be cared 

for and observed, offer optimal locations for the introduction and evaluation of new 

species. 

Nevertheless, in any given community, members will exhibit different levels of 

perception and relationship with plants, varying from the “green thumbs” to those 

whose interest in plants goes little beyond their daily needs. In the past, such plant 

lovers were most likely responsible for the domestication of useful species, and 

today, they are the experimenters and innovators who generate new technologies by 

acute observation and the ability to create heuristic models of the behavior, growth, 

and interactions of the various components of their agroforestry systems. This is a 

very personal and human process of plant management, which mixes personality 

traits and life histories, and cannot simply be replicated or substituted by research 

agencies! The complexity of this social/agronomic interface may explain why 

homegardens appear to elude science, as Nair (2001) remarked. 

Making the leap from growing fruit trees around houses for domestic 

consumption to planting trees in fields for production of fruit, timber, and other 

products, nonetheless, requires dealing with an entirely different set of constraints. 

The main constraints to further developing homegardens or expanding them out to 

fields for greater productivity and income generation are the lack of adequate 

germplasm, risk of accidental fires, survival of seedlings in the dry season and soil 

fertility (Smith et al., 1995; Smith et al., 1996; Smith et al., 1998; Miller, 2001). 

There may also be a need to modify the configuration of species and management 

practices observed in traditional systems to meet increased nutrient exports and 

labor requirements, as well as market demands. At present, commercial products 

obtained from early stages of agroforestry systems are mostly fruits, and marketing 

such products, especially processed pulps, requires facilities most farmers cannot 

afford to have by themselves, while farmers’ associations lack the entrepreneurial 

and managerial expertise to run such installations. This factor has led many 

innovative agroforestry projects dependent on pulp processing facilities down the 

path to failure (Penn, 2004). 

Despite the official interest in agroforestry, due to the immensity of the Amazon 

region, extension services have been unable to meet the growing demands for 

technical assistance. This scenario implies that if agroforestry is to fulfill its promise 

of providing an alternative and more sustainable form of land use in Amazonia, 

extension efforts need to break out of traditional paradigms and the mold of 

commodity-based systems to interact with farmers on a different level of knowledge. 

The traditional socio-cultural practices involved in acquiring and testing new 

germplasm, as seen in homegardens, must be included in rural development projects, 

and stimulated by creative new approaches, with farmers viewed as partners and 

experimenters in the development and domestication of new generations of tree 

crops. In this partnership, a major role for extension should be to help provide the 

necessary germplasm and information.


Surrogate homegardens, based at rural schools, where interesting germplasm can 

be tested and multiplied for access by frontier farmers, while at the same time 

improving nutrition for their children, are one suggestion to increase the spread and 

efficiency of extension services. With homegardens as a conceptual core, this form 

of agroforestry extension should be accompanied by other initiatives and small-scale 

experiments to improve the productivity of subsistence crops, through the use of 

green manures, polycultures, and management of organic matter, among other practices. 

Although this proposal appears to be simple, existing experiences in a similar vein 

must be identified and studied to know if it can work and how to make it work. 


Portions of this paper were adapted from Miller and Nair (2006). Charles Clement 

provided valuable comments on an early version of this chapter. We also thank the 

comments and suggestions of three anonymous reviewers. 

REFERENCES 

Acuña C. 1639. Novo descobrimento do Grande Rio das Amazonas. Agir, Rio de Janeiro 

[reprinted 1994], 179p. 

Alden D. 1976. The significance of cacao production in the Amazon Region during the late 

Colonial period: an essay in comparative economic history. Proc Am Philos Soc 120: 

103 – 135. 

Alves Filho A., Souza Jr. J.A. and Bezerra Neto J. M. 2005. Pontos de história da Amazônia 

Vol. 1 Paka-Tatu, Belém, 208p. 

Amazonas L.S.A. 1852. Dicionário Topográfico, Historico, Descritivo do Alto Amazonas. 

Facsimile ed., Grafima, Manaus [reprinted 1984], 208p. 

Anderson E. 1952. Plants, man and life. Little, Brown, Boston, 245p. 

Avé-Lallemant R. 1859. No Rio Amazonas. EDUSP, São Paulo [reprinted 1980], 288p. 

Bahri S. 1992. L’Agroforesterie, une alternative pour le développement de la plaine alluviale 

de l’Amazone – L’exemple de l’île de Careiro. PhD dissertation, Université de 

Montpellier II. Montpellier, 277p. 

Bahri S. 1993. Les systèmes agroforestiers de l’île de Careiro. Amazoniana 12: 551 – 563. 

Bates H.W. 1863. The naturalist on the river Amazons. University of California Press, 

Berkeley [reprinted 1988], 465p. 

Cabral A.S. 1995. Contact-induced language change in western Amazonia: The non-genetic 

origin of the Kokama language. PhD dissertation. University of Pittsburg, Pittsburg, 415p. 

Carvajal G. 1542. The Discovery of the Amazon. AMS Press, New York [reprinted 1970], 

467p. 

Clement C.R. 1989. A center of crop genetic diversity in western Amazonia: a new 

hypothesis of indigenous fruit-crop distribution. BioScience 39: 624 – 631. 

Clement C.R. 1999a. 1492 and the loss of Amazonian crop genetic resources. I. The relation 

between domestication and human population decline. Econ Bot 53: 188 – 202. 

Clement C.R. 1999b. 1492 and the loss of Amazonian crop genetic resources. II. Crop 

biogeography at contact. Econ Bot 53: 203 – 216. 

Coomes O.T. and Ban N. 2004. Cultivated plant species diversity in a home garden of an 

Amazonian peasant village in northeastern Peru. Econ Bot 58: 420 – 434. 

Coomes O.T. and Burt, G.J. 1997. Indigenous market-oriented agroforestry: dissecting local 

diversity in western Amazonia. Agroforest Syst 37: 27 – 44. 

57

58 

R.P. MILLER M 

ET AL. 

Daniel J. 1776. Tesouro descoberto no Rio Amazonas. Contraponto, Rio de Janeiro (2 vol.) 

[reprinted 2004], 1219p. 

Dean W. 1995. With broadax and firebrand: the destruction of the Brazilian Atlantic forest. 

University of California, Berkeley, 482p. 

Denevan W.M. 2002. Cultivated landscapes of native Amazonia and the Andes. Oxford 

University Press, Oxford, 395p. 

Denevan W.M. and Padoch C. 1987. Swidden-fallow agroforestry in the Peruvian Amazon. 

Advances in Economic Botany 5, New York Botanical Garden, Bronx, 112p. 

Edmundson G. 1922. Journal of the travels and labors of Father Samuel Fritz in the River 

Amazons between 1686 and 1723. Hakluyt Society, London, Second Series, No. 51, 

164p. 

Hemming J. 1987. Amazon frontier: The defeat of the Brazilian Indians. Harvard University 

Press, Cambridge, MA, 647p. 

Holanda S.B. 1965. Historia Geral da Civilização Brasileira. Tomo 2 – O Brasil Monárquico. 

DIFEL, São Paulo, 415p. 

Homma A.K.O. 2003. História da agricultura na Amazônia: da era pré-colombiana ao terceiro 

milênio. Embrapa Informação Tecnológica, Brasília, 274p. 

Huber J. 1904. Notas sobre a patria e distribuição geographica das arvores fructiferas do Pará, 

pp 375 – 406. Boletim Museu Goeldi Historia Natural e Ethnographia. Tomo IV. 

Lamont S.R., Eshbaugh W.H. and Greenberg A.M. 1999. Species composition, diversity, and 

use of homegardens among three Amazonian villages. Econ Bot 53: 312 – 326. 

Lathrap D.W. 1977. Our father the cayman, our mother the gourd: Spinden revisited, or a 

unitary model for the emergence of agriculture in the New World. In: Reed C.A. (ed.), 

Origins of agriculture, pp. 713 – 751. Mouton, The Hague. 

Lehmann J., Kern D.C., Glaser B. and Woods W.I. 2004. Amazonian dark earths: origin, 

properties and management. Kluwer, Dordrecht, 523p. 

Lima D.M. 1999. A construção histórica do termo caboclo. Novos Cadernos NAEA 2: 5 – 32. 

Maués R.H. 2001. Um aspecto da diversidade cultural do caboclo amazônico: a religião. In: 

Vieira I.C.G., Silva J.M.C., Oren D.C. and D’Incao M.A. (eds), Diversidade biológica e 

cultural da Amazônia, pp. 253-272. Museu Paraense Emílio Goeldi, Belém. 

Mazurek R.R.S. 2001. Kinja txi taka nukwa myrykwase: fishing and hunting among the 

Waimiri Atroari Indians from Central Amazonia. PhD dissertation, University of Illinois, 

Chicago, 115p. 

Miller E.T. 1992. Arqueologia nos empreendimentos hidrelétricos da Eletronorte; resultados 

preliminares. Eletronorte, Brasília, 93p. 

Miller R.P. 1994. Estudo da fruticultura tradicional dos índios Waimiri-Atroari: Base para a 

extensão agroflorestal. Anais I Congresso Brasileiro de Sistemas Agroflorestais 2: 

449 – 462, Embrapa-CNPF, Colombo. 

Miller R.P. 2001. Extractive forest products and agroforestry on an agricultural frontier: A 

case study with the Parakanã ã tribe of the trans-Amazon region, Pará, Brazil. PhD 

dissertation. University of Florida, Gainesville, 227p. 

Miller R.P. and Nair P.K.R. 2006. Indigenous agroforestry systems in Amazonia: from 

prehistory to today. Agroforest Syst 66: 151 – 164. 

Nair P.K.R. 2001. Do tropical homegardens elude science, or is it the other way around? 


Padoch C. and de Jong W. 1991. The house gardens of Santa Rosa: diversity and variability in 

an Amazonian agricultural system. Econ Bot 45: 166 – 175. 

Padoch C. and Pinedo-Vasquez M. 2001. Resource management in Amazonia: Caboclo and 

ribereño traditions. In: Maffi L. (ed.), On biocultural diversity: Linking language, 

knowledge, and the environment, pp. 364-378. Smithsonian Institution press, Washington, 

DC, 578p.


Penn J.W. Jr. 2004. Another boom for Amazonia? Socioeconomic and environmental 

implications of the new camu camu industry in Peru. PhD dissertation, University of 

Florida, Gainesville, 298p. 

Penn J.W. Jr. 2006. The cultivation of camu camu (Myrciaria dubia): A tree planting 

programme in the Peruvian Amazon. Forests Trees Livelihoods 16: 85 – 101. 

Piperno D.R. and Pearsall D.M. 1998. The origins of agriculture in the lowland Neotropics. 

Academic Press, San Diego, 400p. 

Piperno D.R., Ranere A.J., Holst I. and Hansell P. 2000. Starch grains reveal early root crop 

horticulture in the Panamanian tropical forest. Nature 407: 894 – 897. 

Poeppig E. 2003. Viaje al Perú y al Río Amazonas, 1827-1832. CETA, Iquitos, 435p. 

Renard-Casevitz F.-M. 1992. História Kampa, memória Ashaninca. In: Cunha M.C. (ed.), 

História dos índios no Brasil. Companhia das Letras, pp 197 – 212. Sec. Municipal de 

Cultura, São Paulo. 

Ribeiro D. 1997. O povo brasileiro: a formação e o sentido do Brasil. Companhia das Letras, 

São Paulo, 476p. 

Rios M.C. 2001. Historia de la Amazonía Peruana. Periodos: Independencia y república. 

Editora Selva, Iquitos, 322p. 

Roosevelt A.C. 1994. Amazonian anthropology: strategy for a new synthesis. In: Roosevelt 

A.C. (ed.), Amazonian Indians from prehistory to present: Anthropological perspectives, 

pp 1 – 29. University of Arizona Press, Tucson. 

Rosa L.S., Cruz H.d.S., Tourinho M.M. and Ramos C.A.P. 1998a. Aspectos estruturais e 

funcionais dos quintais agroflorestais localizados nas várzeas da costa amapaense. II 

Congresso Brasileiro em Sistemas Agroflorestais – Resumos Expandidos, pp 164 – 166. 

Embrapa, Belém. 

Rosa L.S., Cruz H.d.S., Tourinho M.M. and Ramos C.A.P. 1998b. Caracterização dos 

quintais agroflorestais localizados nas várzeas do estuário amazônico. II Congresso 

Brasileiro em Sistemas Agroflorestais – Resumos Expandidos, pp 161 – 163. Embrapa, 

Belém. 

Rosa L.S., Da Silva L.C.B., Melo A.C.G. and Cabral W.d.S. 1998c. Avaliação e 

diversificação de quintais agroflorestais na comunidade de Murinim – Benfica, Município 

de Benevides - Pará. II Congresso Brasileiro em Sistemas Agroflorestais – Resumos 

Expandidos, pp 167 – 169. Embrapa, Belém. 

Salick J. 1992. Crop domestication and the evolutionary ecology of cocona (Solanum 

sessiliflorum Dunal). Evol Biol 26: 247 – 285. 

San Ramon J.S. 1994. Perfiles históricos de la Amazonia Peruana. CETA-CAAAP-IIAP, 

Iquitos, Perú. 274p. 

Santos F.J. 2002. Além da conquista: guerras e rebeliões na Amazônia pombalina. Editora da 

Universidade do Amazonas, Manaus, 239p. 

Schroth G., Mota M.S.S., Lopes R. and Freitas A.F. 2004. Extractive use, management, and in 

situ domestication of a weedy palm, Astrocaryum tucuma, in the Central Amazon. For 

Ecol Manag 202: 161 – 179. 

Smith B.D. 2001. Documenting plant domestication: the consilience of biological and 

archeological approaches. Proc Nat Acad Sci 98: 1324 – 1326. 

Smith N.J.H. 1996. Home gardens as a springboard for agroforestry development in 

Amazonia. Int Tree Crops J 9:11 – 30. 

Smith N.J.H. 1999. The Amazon river forest: a natural history of plants, animals, and people. 

Oxford University Press, New York, 208p. 

Smith N.J.H., Dubois J., Current D.L.E. and Clement C. 1998. Agroforestry experiences in 

the Brazilian Amazon: constraints and opportunities. Pilot program to conserve the 

Brazilian rainforest. World Bank, Brasília, 80p. 

59

60 

R.P. MILLER M 

ET AL. 

Smith N.J.H., Falesi I.C., Alvim P.d.T. and Serrão E.A.S. 1996. Agroforestry trajectories 

among smallholders in the Brazilian Amazon: innovation and resiliency in pioneer and 

older settled areas. Ecol Econ 18: 15 – 27. 

Smith N.J.H., Fik T.J., Alvim P.d.T., Falesi I.C. and Serrão E.A.S. 1995. Agroforestry 

developments and potential in the Brazilian Amazon. Land Degrad Rehabil 6: 251 – 263. 

Stanfield M.E. 1998. Red rubber, bleeding trees: Violence, slavery, and empire in northwest 

Amazonia, 1850-1933. University of New Mexico Press, Albuquerque, 270p. 

Van Leeuwen J. and Gomes J.B. 1995. O pomar caseiro na região de Manaus, Amazonas, um 

importante sistema agroflorestal tradicional. Actas II Encontro da Sociedade Brasileira de 

Sistemas de Produção, Londrina, PR, 21-23/nov/95, pp 180 – 189. IAPAR, Londrina 

(http://www.inpa.gov.br/cpca/johannes/joha-pomar.html; last accessed: February 5, 2006). 

Wiersum K.F. 1996. Domestication of valuable tree species in agroforestry systems: evolutionary 

stages from gathering to breeding. In: Leakey R.B., Temu A.B., Melnyk M., and 

Vantomme P. (eds), Domestication and commercialization of non-timber forest products, 

pp 147 – 158. FAO, Rome. 

Yamada M. and Osaqui H.M.L. 2006. The role of homegardens in agroforestry development: 

lessons from Tomé-Açu, a Japanese-Brazilian settlement in the Amazon. In: Kumar B.M. 


agroforestry, pp 299 – 316. Springer Science, Dordrecht.

CHAPTER 5 

HOMEGARDENS OF MESOAMERICA: 

BIODIVERSITY, FOOD SECURITY, 

AND NUTRIENT MANAGEMENT 

F. MONTAGNINI 

Yale University, School of Forestry and Environmental Studies, 370 Prospect St., 

New Haven, CT 06511, USA; E-mail: 

Keywords: Mayan culture, Nutritional value, Rural development, Structural complexity. 

Abstract. The region of Mesoamerica is densely populated and it suffers from poverty and 

malnutrition both in urban and rural areas. It is home to the Mayan civilization that practiced 

sustainable agricultural systems, involving many native crops and soil conservation strategies, 

for centuries. The homegardens, which provide the household with a basic food source as well 

as high value products to generate cash income are important in Mesoamerica, and are often 

used as tools in development projects that promote food security, especially in the poorest 

areas of Mesoamerica. The Mesoamerican homegardens are quite diverse in vertical and 

horizontal structure and species composition. Both exotic and native plants are used, with 

emphasis on fruit trees. Domestic animals, especially chickens and pigs, add protein to a diet 

that is generally protein-deficient. Many indigenous communities (descendants of the ancient 

Maya) still manage these homegardens using techniques that include residue management and 

ash deposition, thus enhancing nutrient recycling and conservation. Carbon sequestration may 

be important due to the efficient capture of solar radiation in the multi-layered homegardens, 

although its global or regional importance is minimal due to the relatively small area under 

the homegarden system. Management strategies that promote nutrient recycling and maintain 

high species diversity should be encouraged to ensure sustainability of homegardens in the 

region. 


A vast area of what is known today as Mesoamerica was the home of the Mayan 

civilization. The remnants of Mayan culture are concentrated in southern Mexico, 

61 




62 F. MONTAGNINI 

Guatemala, and Belize. The Mayan people are known to have practiced sustainable 

agricultural systems for centuries, cultivating a wide variety of native crops and 

applying indigenous knowledge on nutrient cycling and soil conservation (De Clerck 

and Negreros-Castillo, 2000; Benjamin et al., 2001). In regions such as the 

Tehuacán-Cuicatlán Valley in Central Mexico, human cultures have a history of 

nearly 10 000 years and at present several indigenous ethnic groups continue to 

follow cultural traditions in plant gathering and cultivation (González-Soberanis and 

Casas, 2004). The long history of interactions between human cultures and plant 

diversity has created a substantial body of traditional knowledge on the myriad uses 

of plants. The existence of nearly 1200 plant species utilized by local peoples for 

different purposes, most of them native wild plants, has been documented; many of 

these species are obtained through gathering, but several species are also under 

silvicultural management (Casas et al., 2001; González-Soberanis and Casas, 2004). 

This rich tradition of sustainable agricultural practices in Mesoamerica justifies 

an extensive study of homegardens in the region. Several types of homegardens are 

practiced in the region by the descendants of Maya in present-day Mexico, 

Guatemala, Belize, and Honduras, by other indigenous groups, and by people of 

Hispanic descent in Nicaragua, El Salvador, Costa Rica, and Panama. Traditional 

agroecosystems, which include ‘forest gardens’ or ‘homegardens,’ contain combinations 

of trees with an understorey of annual and perennial crops and sometimes 

livestock. Villagers live within or adjacent to their gardens and maintain them over 

many generations. In present-day Mayan towns in the Yucatán Peninsula of Mexico, 

this type of forest gardens covers about 10% of the region’s forested area (Noble and 

Dirzo, 1997). Small, scattered forest or agroforests can provide local or regional 

environmental services such as conservation of biodiversity (Guindon, 1996; Harvey 

and Haber, 1999). Thus, the practice of homegardens can meet forest conservation 

needs in regions where deforestation and population growth are constant threats, as 

is the case in much of the Mesoamerican region. 

As in other regions of the neotropics, such as Amazonia (Miller et al., 2006), 

present-day homegardens of Mesoamerica represent the reorganization of original 

indigenous practices as a result of the changes brought by colonization, among 

which the most outstanding feature was the incorporation of non-native fruit trees 

and crops. Today, homegardens are of vital importance to the local subsistence 

economy and food security in the region, especially in regions that still carry the 

influence of Mayan culture (De Clerck and Negreros-Castillo, 2000; Méndez et al., 

2001; Zaldívar et al., 2002; Wezel and Bender, 2003; Blanckaert et al., 2004; 

González-Soberanis and Casas, 2004). 

This chapter describes the characteristics of homegardens in Mesoamerica, with 

emphasis on biodiversity, their importance in sustaining food security in rural areas, 

and their role in nutrient cycling. Information is presented on indigenous systems 

that have been practiced by descendants of the ancient Maya for many centuries in 

regions of Mexico, Belize, and Guatemala, as well as on systems currently being 

practiced in regions beyond the Mayan influence such as in Nicaragua, El Salvador, 

Honduras, Costa Rica, and Panama.

2. GENERAL ECOLOGICAL AND SOCIOECONOMIC CHARACTERISTICS 

OF MESOAMERICA 

Culturally Mesoamerica joins the present-day middle and south Mexico, Belize, 

Guatemala, parts of Honduras, and El Salvador (Fig. 1). Geographically, the other 

three countries of Central America (Nicaragua, Costa Rica, and Panama) are also 

included in the region. Most geographers consider Central America to be part of the 

North American continent; however, they do not consider Mexico to be a part of 

Central America. The Caribbean islands are often considered separately from 

Mesoamerica because they are culturally very diverse. For the purposes of this 

chapter an example is drawn from Cuba, a country with a Hispanic tradition as rich 

as many of the countries of Mesoamerica, and with similar ecological and economic 

conditions. 

Figure 1. Map of Mesoamerica (source: www.biodiversityhotspots/mesoamerica/; last 

accessed: January 2006). 

2.1. Ecological setting 

HOMEGARDENS OF MESOAMERICA 

M 

The climate, vegetation, and soils of Mesoamerica are very heterogeneous given the 

latitudinal expanse of the region and its montane relief. This results in high 

ecological heterogeneity that gives room to a whole variety of agricultural systems. 

The climate ranges from mild temperate-subtropical in north-central Mexico and 

Guatemala, to tropical in the rest of Central America (Richards, 1996). There is a 

general pattern of a more humid climate (annual precipitation 3000 to 5000 mm) 

along the eastern or Caribbean side due to the influence of the humidity brought by 

the trade winds. Winds reach the central volcanic mountain range in the region, 

63

64 

cause rains on the Atlantic side, and quickly lose much of their humidity. 

Consequently, most of the Pacific watersheds are drier with annual precipitation 

ranging from 300 to 2000 mm yr –1 and a marked dry season (November through 

April). The vegetation follows the climatic pattern ranging from subtropical and 

tropical rainforest in the Caribbean lowlands to dry forests and savannas in the 

Pacific watersheds (Richards, 1996). 

The soil types cover a whole array from older, less fertile Oxisols and Ultisols 

(US Soil Taxonomy System), to younger Andosols and Inceptisols, especially in 

areas affected by past or present volcanic activity and in alluvial zones. Several 

other types of soils can be found due to the varied climatic and topographic 

conditions of the region (Sanchez, 1976; De Las Salas, 1987). In general, due to the 

recent volcanic influence, the soils of the region are relatively younger and more 

fertile than many soils of other regions of Latin America, such as the Amazon Basin. 

The Mesoamerican region comprises an area with diverse plant and animal life. 

The varied topography, geology, vegetation, and drainage patterns within the region 

result in a rich array of vegetation types and animal communities. More than 24 000 

plant species, 521 mammal species, 1193 bird species, 685 reptile species, and 460 

amphibian species have been identified within Mesoamerica, many of which are 

endemic to the region 1 . Moreover, Mesoamerica is the third most biologically 

diverse region in the world; Myers et al. (2000) identified it as one of the world’s 25 

biological hotspots. Covering an estimated 0.5% of the world’s terrestrial surface, it 

is home to roughly 7 to 10% of the world’s plant and animal species (Harvey et al., 

2005). Several trees that are currently planted worldwide in agroforestry 

combinations, such as Leucaena leucocephala, Gliricidia sepium, and some species 

of Acacia and Mimosa, have their centers of origin and diversity in Mesoamerica 

(NAS, 1979; 1980; Dommergues, 1987). Mesoamerican homegardens, in addition to 

meeting the immediate alimentary and economic needs of the people, also act as 

repositories of local biodiversity as they include a dynamic mixture of native and 

useful species. 

2.2. Cultural setting 

F. MONTAGNINI 

The region of Mesoamerica is culturally and socioeconomically diverse, sharing 

certain characteristics such as a strong Spanish colonial influence (with the 

exception of Belize) and a strong dependence on agriculture and natural resources 

(Harvey et al., 2005). In parts of Mesoamerica (southern Mexico, northern 

Guatemala, and Belize) homegardens and other types of agriculture carry the 

influence of ancient traditions from the indigenous Mayan groups that lived in this 

region prior to the arrival of European conquerors. 

Several studies have reported and discussed sustainable land use practices that 

were used by the Maya, including terracing, using soil algae or wetland soil to 

enrich upland garden plots, and other soil conservation strategies (Barrera et al., 

1977; Turner and Harrison, 1981; Beach and Dunning, 1995; Fedick and Morrison, 

2004). Archaeological evidence of the use of homegardens by the ancient Maya 

include the location of residential sites within prime agricultural land, strategic 

placement of households to allow for gardening space, the addition of soil

amendments as indicated by nutrient enrichment within house lots, and the 

distribution of tools in the vicinity of residences (Fedick and Morrison, 2004). 

The decline of the Maya civilization (~700 BC to 800 AD) has been attributed in 

part to complex economic, political, and social changes that led people to change 

their traditional sustainable agricultural practices to less diverse agricultural systems 

(Barrera et al., 1977; Turner and Harrison, 1981; Atran, 1993; Atran et al., 1999). 

A set of political and ecological factors apparently led some Maya groups such as 

the lowland Maya of Petén, Guatemala, to reject a diverse swidden-fallow 

management strategy for a more simple ‘milpa’ or shifting agricultural system that 

provided fewer forest products. This led to a less diverse agricultural landscape and 

a less diverse biological landscape (Atran, t 1993; Steinberg, 1998; Atran et al., 

1999). The milpa system consisted of 2 to 5 ha plots that were cut and burnt, and 

cultivated mainly with maize (Zea mays). In the traditional system, after a few 

harvests the plots were left to regenerate with a long fallow cycle, leaving tree 

species time to mature and bear fruits (15 to 40 years). 

At present, many Maya groups such as the Mopan of Belize have shortened the 

fallow periods to about 5 years. With such a short fallow cycle, the vegetation 

regenerating in the milpas is much less diverse than in the traditional Mayan 

systems, with only a few useful species of shrubs and palms (Steinberg, 1998). 

However, some authors argue that only the most sophisticated and intensive type of 

Maya agriculture collapsed, while the oldest, simplest, and most ecologically stable 

type is still being practiced (Atran, 1993). The milpa system – as practiced today – 

with dispersed fruit trees and vegetable crops and livestock has the attributes of a 

productive homegarden. 

2.3. Socioeconomic conditions 


M 

With a total land area of almost 2.5 million km 2 and a total population size of almost 

140 million people, Mesoamerica is one of the most densely populated regions of 

not only Latin America but also the entire world (Harvey et al., 2005). The current 

population of Central America is approximately 38 million people, of which about 

20% are indigenous (Harvey et al., 2005). With a yearly growth rate of about 2.6%, 

the population is expected to double within the next 25 years. The overall population 

density of the region is 56 people per km 2 , with a range from 296 people per km 2 in 

El Salvador to just 11 in Belize (Table 1). 

Despite its recent economic growth, Central America remains one of the world’s 

poorest regions. About 50% of the population is poor (i.e., unable to cover basic 

needs such as nutrition and housing) and 23% is extremely poor (i.e., not able to 

cover even daily basic nutrition; Harvey et al., 2005). Particularly striking are the 

cases of Honduras and Guatemala with poverty levels of 74.5% and 78.5%, 

respectively (Harvey et al., 2005). The region’s poverty has led to the massive 

exploitation of its natural resources. Large areas of forest have been cut down and 

burnt for firewood, used in the production of paper, and cleared for agricultural uses. 

Despite an equal distribution of population between rural and urban areas, rural 

populations are considerably poorer than their urban counterparts (Harvey et al., 

2005). 

65

66 

Mesoamerica has diverse ecological, cultural, and socioeconomic conditions that 

have given origin to varied agricultural systems. The prevalent conditions of rural 

poverty and associated malnutrition call for the need of agricultural systems that can 

help fulfill urgent household needs. Homegardens, whose main function is to provide the 

household with a basic food source and marketable products, are extremely important— 

given the socioeconomic conditions prevalent today in Mesoamerica. 

Country 

F. MONTAGNINI 

Table 1. Area,population size and population density of Mesoamerica. 

Area 

(km²) 

Population 

size 

Population 

density 

(no./km²) 

Mexico 1,964,375 101,879,000 52 

Nicaragua 131,847 4,918,000 37 

Honduras 112,520 6,406,000 57 

Guatemala 108,917 12,974,000 119 

Panama 75,536 2,846,000 38 

Costa Rica 51,113 3,773,000 74 

Belize 22,965 256,000 11 

El Salvador 21,046 6,238,000 296 

Total 2,488,319 139,290,000 56 

Source: Data update 2001 estimates, http://www.globalgeografia.com/north_america/nam_ 

sup.htm (last accessed: January 2006). 

3. COMPOSITION AND STRUCTURE OF HOMEGARDENS IN 

MESOAMERICA 

As in other regions worldwide, the structure and composition of homegardens in 

Mesoamerica are quite complex. A full spectrum of homegarden practices can be 

found in different locations of Mesoamerica, ranging from near complete 

domination of woody perennials to homegardens where trees may account for less 

than 20% of the annual productivity. Plant composition in homegardens of 

Mesoamerica is influenced by access to water, owners’ economic activities, labor 

availability, traditional social organization, modernization processes, and economic 

development (Blanckaert et al., 2004). In general, plant species composition within 

the homegardens is the result of continuous selection in which the family usually 

favors the planting of fruit trees with high productivity (Caballero, 1992). 

Most homegardens of Mesoamerica consist of several vertical and horizontal 

strata in which plants are arranged according to their adaptability to the existing 

light conditions and nutrient resources (Fig. 2). The number of individual plants per 

stratum, however, varies among homegardens; older, more mature homegardens 

display more developed tree strata. Some homegardens resemble agricultural fields 

with an emphasis on herbaceous and low shrub strata, with a greater focus on 

agricultural crop production. Others have more trees, with architecture similar to

that of the native forests of the region (Gillespie et al., 1993; De Clerck and 

Negreros-Castillo, 2000; Méndez et al., 2001; Zaldívar et al., 2002; Blanckaert et al., 

2004). 

Figure 2. Most homegardens of Mesoamerica consist of several vertical and horizontal 

strata, with plants arranged according to their adaptability to light and nutrient resources. A 

homegarden in Siquirres, in the Caribbean lowlands of Costa Rica showing vertical 

stratification with peach palm ( Bactris gasipaes) in the top layer (right), coconut palms 

(Cocos nucifera) and plantains or bananas ( Musa a spp.) in the lower tree strata (left), 

sugarcane (Saccharum officinarum), and other herbaceous crops in the herb layer (Photo: R. 

González). 

3.1. Vertical stratification 


M 

The vertically stratified homegardens are potentially more productive on an area 

basis since they can capture more resources and exhibit tighter nutrient cycling, than 

those without a stratified arrangement. For example, in a study of four homegardens 

in the Petén, Guatemala, Gillespie et al. (1993) reported high structural complexity, 

with full canopy closure in the layers within the canopy. The garden architecture 

made efficient use of light and space, with intensive management for food and fuel 

production. The development of homegardens in the area utilized existing trees, 

leaving the most useful as residuals after thinning, and inserting other desirable trees 

and shrubs in the understorey and open space. This strategy seemed to maximize 

67

68 

light use, according to results of measurements of incident radiation at different 

canopy levels reported by Gillespie et al. (1993). 

In most Mesoamerican homegardens, each stratum contains plant species that 

belong to a characteristic life form, much like in a native forest of the same region. 

In homegardens of the Zona Maya of Quintana Roo, Yucatán Península, Mexico, 

there were six strata: low herbs, low shrubs, tall shrubs, fruit trees, timber trees, and 

a stratum with vines (De Clerck and Negreros-Castillo, 2000). These authors studied 

the species composition of each stratum, and concluded that in these systems the 

efficient use of space and resources maximized the production of food, timber, 

medicinal plants, and non-timber products to cover the farmers’ needs. They 

suggested that these systems (or analogs of these in terms of structure and 

composition) could be managed in a manner that protects the natural resource base 

of the region. 

3.2. Plant species composition 

F. MONTAGNINI 

The species composition of the homegardens in Quintana Roo analyzed by De 

Clerck and Negreros-Castillo (2000) was much like others in Mesoamerica and in 

other regions of Latin America as well, with a mixture of native and exotic species 

in each stratum fulfilling the farmers’ needs. The herbaceous stratum (0 to 0.5 m 

tall) was comprised of herbs and creepers such as basil (Ocimum basilicum), squash 

(Cucurbita spp.), and sweet potatoes (Ipomoea batatas), containing an average of 

14% of all species in the homegarden. The low shrub stratum (0.5 to 1.5 m tall) 

contained annual and perennial herbaceous plants such as tomatoes (Lycopersicum 

esculentum), maize or corn (Zea mays), ruda (Ruta chalapensis), and included 

several shade-tolerant species such as cassava (Manihot esculenta), ginger (Zingiber 

officinale), pineapple ( (Ananas comosus), 

and taro (Colocasia esculenta). The low 

shrub stratum contained 12% and the tall shrub stratum contained 15% of the total 

number of species of the homegardens. The low tree stratum was dominated by fruit 

trees, most frequently by Citrus spp., and contained 41% of the total number of 

species; this stratum was often dominant in the absence of the fifth stratum (tall 

trees). The presence of the tall tree stratum, with 15% of the species, was an 

indicator of the maturity of homegardens. It was composed of several species of 

palms, tall fruit trees such as mango (Mangifera indica) and avocado (Persea 

americana), and timber trees. The vine stratum started at ground level and rose up to 

the top of the canopy, with 4% of the total number of species, composed mainly of 

tuber-forming vines such as sweet potatoes and several species of yams (Dioscorea 

spp.). Many epiphytic species were found on trees and shrubs (De Clerck and 

Negreros-Castillo, 2000). 

This complex horizontal and vertical structure allows for a variety of agricultural 

crops and tree products that are consumed in the household and sold in the local 

markets. Multistrata agroforests combining agricultural crops with high-value timber 

species, as described in the above example, can provide farmers with long- and 

short-term revenue with harvest distributed throughout the year. 

In a study of homegardens located in eastern Cuba, Wezel and Bender (2003) 

found that species composition and structure were similar to “typical” homegardens

of other regions in Mesoamerica. The top layer (3 to 10 m) consisted mostly of trees 

such as avocado, coconut (Cocos nucifera), mango, and breadfruit ( (Artocarpus 

communis). In the middle layer (1 to 3 m), smaller trees like guava (Psidium 

guajava), soursop ( (Annona muricata), 

orange (Citrus sinensis), or papaya (Carica 

papaya) were found together with bananas and plantains (Musa spp.), sugarcane 

(Saccharum officinarum), pigeon pea (Cajanus cajan), and climber yam (Dioscorea 

spp.). In the ground layer (0 to 1 m), different vegetables, spices, and medicinal 

plants were cultivated while others grew spontaneously. 

3.3. Horizontal structure 


M 

The horizontal structure of homegardens shows interesting patterns, governed by the 

uses/functions of the different plant species. For example, ornamental plants are often 

found in linear patterns around the house. They are also found along the roadside of the 

garden, reflecting their aesthetic purpose as well as their use for the delineation of 

property or sections thereof (Blanckaert et al., 2004). In general, edible plants are found a 

little farther away from the house, mostly in small groups to facilitate management such 

as weeding or pruning. In semiarid environments such as south-central Mexico 

(Blanckaert et al., 2004), central Nicaragua (Méndez et al., 2001), and in the Baitiriqui 

region of Cuba (Wezel and Bender, 2003), irrigation is frequently used. In these cases, 

edible plants are located downhill from the house and in close proximity to it so that 

they can be watered using the wastewater recycled from domestic uses. Medicinal 

plants are often found still farther away than ornamental or edible plants (Blanckaert 

et al., 2004). Homegardens are also important for providing additional living and 

working space to supplement small household structures (Lok, 1998). 

In an effort to organize and systematize the study of this very complex type of 

agroecosystem, many authors have used statistical procedures to group descriptive 

characteristics of homegardens. For example, cluster analysis, correspondence 

analysis, and diversity indices have been used by several authors to explain the 

patterns of variations in floristic composition of the homegardens (Méndez et al., 

2001; Zaldívar et al., 2002; Blanckaert ett al., 2004). These procedures help in the 

description of the characteristics of the specific homegardens under study such as 

explaining differences in species diversity among homegardens of different 

settlements or localities in a region. 

4. PLANT SPECIES DIVERSITY IN HOMEGARDENS 

OF MESOAMERICA 

Results of several studies indicate that homegardens of Mesoamerica are rich in 

biodiversity, and need to be considered for in situ conservation and development 

programs. Table 2 shows a summary of studies on plant biodiversity in homegardens 

of different geographic regions of Mesoamerica. Several of the studies shown in 

Table 2 emphasize tree and shrub species and their uses and relevance for forest 

conservation, while others focus on the variety of plant species of all life forms and 

69

Location Climate Number of Number of plant species Source 

homegardens 

studied 

Tehuacán-Cuicatlán Valley, semiarid to arid 30 233 (66% ornamental, 30% Blanckaert et al., 2004 

Puebla, south-central Mexico 

Tixpeual and Tixcacaltuyub, 

Yucatán, Mexico 

Tropical forests of nine states, 

south-southeast Mexico 

Totonac community in 

Coxquihui, Veracruz, Mexico 

Zona Maya of Quintana Roo, 

Yucatán Peninsula, Mexico 

Maya community of San Jose, 

Toledo district, Belize 

El Camalote, Copán, SW 

Honduras near the border with 

Guatemala 

Table 2. Examples of studies on plant diversity of homegardens in Mesoamerica. 

tropical humid 

lowland 

edible, 9% medicinal) 

not available 301 trees and shrubs (70% Rico-Gray et al., 1991 

medicinal, 40% apiculture, 30% 

edible, 17% fuel, 19% building, 

12% timber) 

not available 278 Toledo et al., 1995 

tropical humid 

lowland 

warm, subhumid, 40 223 Del Angel-Perez and 

low elevation 

Mendoza, 2004 

tropical humid 78 80 De Clerck and Negreros 

lowland 

Castillo, 2000 

tropical humid 18 164 Levasseur and Oliver, 

lowland 

2000 

montane wet 10 253 (91 trees, 42 shrubs 90 House and Ochoa, 1998 

herbs, 24 lianas, 2 palms, 2 

mushrooms) 

70 

F. MONTAGNINI

Nicoya, SW Costa Rica tropical seasonal 12 289 (63 varieties) Lok et al., 1998 

lowland wet 

Five life zones (sensu Holdridge, tropical subhumid 225 236 (excluding ornamentals) Price, 1989 

1987) of Costa Rica 

to humid 

Eastern Costa Rica wet tropical 45 133 Price, 1989 

Talamanca, S. Costa Rica wet tropical 83 46 cultivated species Zaldivar et al., 2002 

Coto Brus, S. Costa Rica wet tropical 55 27 cultivated species Zaldivar et al., 2002 

Masaya, Nicaragua semiarid to arid 20 334 Mendez et al., 2001 

Eastern Cuba semiarid 31 101 Wezel and Bender, 2003 


M 

71

72 

their role in sustaining local livelihood needs. In a region with such broad 

geographic diversity as Mesoamerica, diversity of plants found in homegardens is 

expected to vary according to latitude, elevation, and rainfall. These trends are not 

evident from the data shown in Table 2, as similar numbers of species are reported 

for wet and for semiarid to arid locations. The number of species reported by the 

authors depends on the number of homegardens studied, types of species that were 

emphasized, size of the homegardens studied, reliance of homegardens for 

subsistence needs, and the traditional uses of the plants, among other factors as 

discussed below. 

A number of the studies shown in Table 2 also emphasize plant uses and 

management. For example, in Yucatán, Mexico, Rico-Gray et al. (1991) reported the 

uses of trees and shrubs from the tropical deciduous forests by the Yucatecan Maya. 

Despite the lack of important timber species in these forests, the authors conclude 

that management could lead to sustainable production of honey, deer, and building 

materials for houses. In homegardens of the Tehuacán-Cuicatlán Valley in Puebla, 

Mexico (Table 2), plants were categorized into three main groups: cultivated (68%), 

protected (10%), or spared (22%) (Blanckaertt et al., 2004). Cultivated plants are 

those that are sown or planted by the owner. Protected plants are those that are 

encouraged by the farmer, whether they are transplanted from zones outside the 

garden or grow spontaneously in the garden. The farmer may choose to protect or 

encourage the plant, for example, by supporting it or attaching it to a solid structure, 

or by putting stones around the plant. Spared plants are those that spontaneously 

grow in the garden and are not removed (Blanckaert et al., 2004). 

The high diversity in plant species and uses reported by Blanckaert et al. (2004) 

were found at 1217 m above sea level with a climate classified as semiarid to arid 

(total annual precipitation 395 mm). Theoretically, these conditions would place the 

region at the low end of the spectrum of potential plant species diversity. The most 

represented plant families were Cactaceae, Araceae, Liliaceae, Solanaceae, and 

Crassulaceae, reflecting the climatic characteristics as well as the preferences of 

the local farmers. Members of both Cactaceae and Solanaceae families in the 

homegardens are important edible plants. For instance, nopal (Opuntia spp. and 

other species of Cactaceae), chilli (Capsicum spp.), and tomato (Lycopersicum 

esculentum) (Solanaceae) are important ingredients of the Mexican diet. 

A possible explanation for the relatively large diversity of plants found in dry 

locations was advanced by Price (1989), who studied the characteristics of 

homegardens in five different ecological regions of Costa Rica (Table 2). The author 

found that homegardens were most important in regions of dry tropical forests 

because socioeconomic conditions were more difficult than in other regions of the 

country, making people rely more on homegardens for self-sustenance. In a semiarid 

region in eastern Cuba, Wezel and Bender (2003) also reported the importance of 

homegardens and their high species diversity (Table 2), with about 50% of the 

species consisting of fruit trees. 

Locally, plant diversity of homegardens can also be influenced by size of the 

homegardens. For example, in Nicoya, Costa Rica, Lok et al. (1998) found that the 

size of homegardens ranged from 0.1 to 1.4 ha with an average of 0.5 ha (Table 2). 

The smallest homegardens considered in the study had the highest diversity, with 

. 

F. MONTAGNINI


M 

205 to 745 species and an average of 348 species per ha. In contrast, the larger 

homegardens had only an average of 96 species per ha, with less variability among 

gardens in comparison to the smaller homegardens that exhibited higher variability 

in species diversity. 

4.1. Importance for species domestication and conservation 

The high plant species diversity of homegardens in Mesoamerica makes them an 

important resource for ethnobotanical studies. Since many species in homegardens 

are encouraged or cultivated, the process of domestication of useful species has long 

taken place in homegardens. This is true for homegardens in other regions of 

the neotropics where they are intensely managed and crops are carefully selected for 

specific purposes. For example, the homegardens of Japanese emigrants in the 

Tomé-Açu settlement in Pará, in the eastern Amazon region of Brazil, have served 

as “banks” of potential crop species that had been gathered and closely observed by 

the family members (Yamada and Osaqui, 2006). The homegardens of Tomé-Açu 

functioned as individual validation facilities for farmers making decisions about 

planting new crops in their farms. Farmers also used homegardens for improvement 

and propagation of nursery stock. 

Several studies shown in Table 2 emphasize the role of homegardens as sites for 

domestication and preservation of useful species (Toledo et al., 1995; House and 

Ochoa, 1998; González-Soberanis and Casas, 2004, among others). In El Camalote, 

Copán (Honduras), House and Ochoa (1998) found several introduced species along 

with native species that belonged to natural forests of the region, and they stressed 

the importance of homegardens as genetic banks of ancient crops and as a research 

field for new varieties and cultivars. The diversity of traditional vegetables in the 

homegardens studied by these authors was outstanding, with many species that are 

also present in Guatemala and Mexico but that are absent in other parts of Honduras. 

They cite examples of several vegetables and fruits that today are almost exclusively 

found in the homegardens. Such is the case of the chayo (Cnidoscolus chaymansa), 

a popular green vegetable in Camalote (similar to spinach) but almost absent in the 

rest of Honduras. They also cite other unique species of vegetables and fruits that, 

again, are found only in the homegardens of Honduras and Guatemala. 

Other examples of domestication of crop species can be found in regions such as 

the Tehuacán-Cuicatlán valley in central Mexico, where the Maya cultures have a 

history of over 10 000 years (González-Soberanis and Casas, 2004). These authors 

studied the management and domestication of a fruit of the Sapotaceae family, the 

tempesquistle (Sideroxylon palmei). This fruit is consumed and commercialized in 

large quantities in the villages studied. Apparently, management of this species in 

homegardens has resulted in larger, better quality fruits than those of the wild 

populations, demonstrating the importance of domestication of plant species by the 

owners and managers of homegardens. This is a good example of a process of 

selection by local farmers that may be true for many other species in other homegarden 

settings too. 

Homegardens may have other positive effects on biodiversity, as they can serve 

as local refuges for plants and animals that otherwise may be threatened by human 

73

74 

or natural disturbances. For example, Griffith (2000) reported that during the 1998 

fires in Petén, Guatemala, homegardens and other agroforestry systems might have 

served as critical refuges for many forest species. Apparently, agroforestry farms 

attracted birds by virtue of their complex structure – similar to that of intact forest 

patches – they harbor insects, provide nesting sites, and offer protection from 

predators (Griffith, 2000). They were also attracted by the cultivated fruit trees, 

which may have provided some of the only food sources in the region after fire 

destroyed most of the surrounding vegetation. Homegardens, thus, can provide 

additional services as buffers for protecting local biodiversity in times of stress. 

5. SIGNIFICANCE FOR HOUSEHOLD FOOD SECURITY 

Homegardens can enhance food security in several ways, most importantly through: 

(1) direct access to a diversity of nutritionally rich foods, (2) increased purchasing 

power from savings on food bills and income from sale of garden products, and (3) 

fall-back food provision during periods of temporary food scarcity. In many parts of 

the world, homegardens supplement food supply for people, but in some cases, 

homegardens can yield basic staples, when they are large enough to plant sufficient 

quantities of tubers or cereals (Eibl et al., 2000; Wezel and Bender, 2003). In this 

regard, homegardens fulfill a very important social function, especially in a region 

like Mesoamerica where poverty and malnutrition co-exist. For example, in the 

Maya community of San Jose, Belize, traditional agroforestry systems including 

milpa, cacao (Theobroma cacao) under trees, and homegardens almost entirely meet 

the family needs for food and wood, and generate 62% of family income (Levasseur 

and Oliver, 2000). 

In contrast to other types of agroforestry and other productions systems, 

homegardens are very important for supplying the household with food products 

year-round (Budowski, 1990; Lok, 1998; Eibl et t al., 2000). Their principal 

goal is 

not to optimize production, as it could be in the rest of the farm, but to guarantee a 

minimum supply of different food products at all times of the year, functioning as a 

buffer in times of low income and food scarcity. Often, high value products from 

homegardens can be sold to purchase staple foods during periods of scarcity. In 

Central America, women play an important role in the management, maintenance, 

and sale of homegarden food products (Lok, 1998; Howard, 2006). 

5.1. Edible plant species 

F. MONTAGNINI 

As seen in the previous sections, homegardens in Mesoamerica are planted with a 

variety of species used for various purposes, including food, medicinal, ornamental, 

timber, construction, crafts, among others (Zaldívar et al., 2002) . In addition to their 

use for self-sustenance, many studies have indicated that the potential for cash sales 

from homegardens is highly important in their composition and management. 

Frequently, excess homegarden production is given away to relatives working in 

urban areas, thereby supporting food security in both urban and rural areas.


M 

The importance of homegardens for household food security becomes greater in 

more extreme situations of poverty and isolation. In present day Cuba, homegarden 

products have contributed additional food to the basic provision such as bread, oil, 

flour, meat, and other products sold cheaply in government stores. After 1989, when 

the Soviet Union collapsed and dropped aid to Cuba, the economic situation 

worsened and food distribution declined precipitously. As it was imperative to find 

alternative sources of food supplies, farmers intensified homegarden production in 

order to feed their families (Wezel and Bender, 2003). 

Similar situations of low income and little assistance by government programs 

are common in several countries of Mesoamerica. In Nicaragua, one of the poorest 

countries of Central America, Méndez et al. (2001) found that families in Masaya 

obtained at least 40 different plant products from their homegardens (Table 2), as 

well as the benefit of space for working on handicrafts (a major source of income in 

Masaya), and socializing. People enjoyed meeting their neighbors and visitors in the 

homegarden because it was a pleasant area of their homes. Although dependence on 

homegardens varied according to specific conditions, they seemed to be a consistent, 

flexible resource used to meet a diversity of needs, although their main function was 

always to provide edible products for household consumption. 

Although Costa Rica probably has the best conditions of Mesoamerica in terms 

of average per capita income and social welfare programs, rural poverty and 

malnutrition persist there, especially among some indigenous groups living in 

remote areas. Chibchan Amerindians (Bribris, Cabecares and Guaymis) who live in 

reserves located in Talamanca and Coto Brus, in the south-central part of Costa 

Rica, practice slash-and-burn agriculture, and maintain polyculture fields or 

homegardens adjacent to their dwellings with a high diversity of plants (Zaldívar 

et al., 2002; Table 2). Both Bribris and Cabecares have lived in territories within the 

Talamanca Reserve for centuries, while the Guaymi migrated about 60 years ago 

from their ancestral territories in Panama. Most edible crops common to all 

settlements studied by Zaldívar et al. (2002) were native to the region, with the 

exception of plantains and bananas, ‘manzana de agua’ (water apple, Syzygium 

malaccense), oranges and mangoes. In other regions of Costa Rica also, homegardens 

are important for supplying food; they also serve as a buffer in times of 

harvest failures or economic depressions (Price, 1989). 

In the Chiriquí province of Panama, Lok and Samaniego (1998) found that 

among the Ngöbe (or Guaymi) indigenous populations, the homegarden was the 

system that provided the largest cash income and number of edible products for 

household consumption when compared with other farm activities. They studied 10 

farms with an average size of 6.7 ha each, of which about half a hectare was 

dedicated to homegardens. The Ngöbe grow annual food crops in plots where they 

also grow “fire-hardy” trees. These plots provide the basic food needs of the family 

(rice, maize, and beans) during much of the year. In the homegardens they grow 

about 100 plant species, of which 75 are woody species (trees, shrubs, and palms). 

Among the woody species most of them are fruit trees, including oranges, guayabas, 

avocados, and coconuts. Fruits are harvested for household consumption, and often 

are the sole source of food for the family in times of scarcity. Fruits are also a source 

of food for wildlife, especially birds that the Ngöbe hunt for food. About 80% of 

75

76 

land inhabited by the Ngöbe is of low productivity and is not suitable for 

commercial production of basic grains, as soils are low in organic matter and high in 

aluminium content. The cultivation of homegardens is one alternative that the Ngöbe 

families have successfully used to offset such edaphic constraints and/or to alleviate 

the problem of food shortage. 

5.2. Domestic animals 

F. MONTAGNINI 

Domestic animals are frequently found in the homegardens of Mesoamerica and 

Cuba. For example, in the Maya community of San Jose, Belize, poultry and pigs 

were found in about 80% of the households (Levasseur and Olivier, 2000). 

Likewise, in the Totonac backyard homegardens of Veracruz, Mexico, pigs, 

chickens, and small livestock were common (Del Angel-Perez and Mendoza, 2004). 

In a survey of 80 homegardens in the dry and humid regions of Costa Rica, 

Nicaragua, and Honduras, Wieman and Leal (1998) also noted chickens in 79% of 

the homegardens, pigs in 49%, and ducks in 10% of the households. In Cuba, 

animals such as pigs, sheep, chickens, and to a lesser extent ducks, rabbits, and 

turkeys abound in the homegardens (Wezel and Bender, 2003). Larger farm animals 

such as sheep, goats, and cows are often kept tethered on the nearby roadsides to 

permit grazing, or sometimes kept in small-fenced paddocks next to the house 

(C. Munford, pers. comm., October 2005). 

Small animals, in particular, represent a source of production of low-cost protein 

in homegardens, especially for the low-income households (Wieman and Leal, 

1998). Several small animals such as chickens, ducks, and rabbits also provide 

B-complex vitamins and minerals such as iron, calcium, and phosphorus. The small 

sizes of these animals also make their care and management, besides meat 

preparation (slaughtering, skinning, and cooking) relatively easy. Yet another 

advantage is the ease of selling the animals and their products in the local markets 

and their year-round production, unlike the orchard plant products which can be 

seasonal (Del Angel Pérez and Mendoza, 2004). 

Chickens are particularly important in the homegardens of the developing 

countries worldwide, primarily for their ability to generate cash income from the 

production of eggs, meat, and chicken manure. They also contribute to biological 

pest control by preying on insects and grubs, and facilitating household waste 

recycling. In the Totonac backyard homegardens of Veracruz, Mexico, chickens 

roamed free in about half of homegardens surveyed, although they are often penned 

at night (Del Angel-Perez and Mendoza, 2004). The families in Central America 

also consumed most of the chicken meat and eggs produced by them. In contrast, 

duck meat is not as much appreciated, as ducks are often considered pets. Overall, 

the home-raised livestock has high priority among the Totonac farmers, presumably 

because of the high value of these animals in the open market. 

Similarly, pigs are an important source of meat, despite the seasonality of 

production, mostly coinciding with festivities or special occasions. In the homegardens 

studied by Wieman and Leal (1998), an average of seven pigs were


M 

found in the larger homegardens of Limón, Costa Rica and Paraíso, Honduras, 

and a smaller number in the smaller-sized homegardens of Masaya, Nicaragua. 

Ornamental plant nurseries, wherever present, deterred pig husbandry because of the 

potential damage to nursery plants. 

In general, local breeds of animals with high resistance to pests and diseases are 

used, and women take care of the animals (Lok and Samaniego, 1998). Whenever 

the domestic animals are likely to interfere with the cultivation of plants within the 

homegardens, they are enclosed or tied up. In the orchards dominated by trees, pigs 

and chickens roam freely, suggesting that the farmers disregard the understorey 

vegetation, while backyards in town often have animals in cages or in enclosed 

quarters to protect ornamental, medicinal, condiment, and ritual plants. 

5.3. Promotion of homegardens in food security and development projects 

Homegardens have long been used as a tool to promote household food security in 

many regions of the world, and especially as part of many educational and 

dissemination efforts by international aid agencies, local governments and nongovernment 

organizations (NGOs). For example, FAO has produced materials for 

their training package ‘Improving Nutrition through Homegardening’ (FAO 2001), 

featuring homegardens for food security in many regions of the world, including 

specific projects in Nicaragua, El Salvador, and Honduras. In Nicaragua, 

government subsidies, in combination with international aid, have been used for 

decades to promote homegardens as a means to guarantee basic household food 

security. For example, the Plan Alimentario Nacional (National Food Plan) with 

financial support from foreign-aid and local logistic and technical support from 

NGOs working in the region, has distributed seeds, working tools, and cooking 

utensils to families in need, mostly from the rural areas of semiarid regions 

(El Nuevo Diario, Managua, Nicaragua, April 3, 2002). Similar promotion of 

homegardens to alleviate poverty and ensure basic food supply in rural and urban 

areas is underway in Panama and El Salvador, again supported by local NGOs 

and international assistance (e.g., Food Safety Program in Tacuba, El Salvador, 

sponsored by World Vision, Canada). In Nicaragua, the Peace Corps of the USA 

established the Food Security Program after Hurricane Mitch in 1998, while other 

organizations such as the Red Cross integrated homegarden projects into larger ones 

directed to address the post-Mitch needs including natural disaster mitigation efforts 

(D. Craven, pers. comm., October 2005). 

In several locations of Mesoamerica, homegardens are often grown and managed 

as part of the communal development efforts. For example, in Diriamba, Nicaragua, 

community homegardens form part of a larger development program (POSAF, 

Program for Agroforestry Development and Environment) funded by the World 

Bank (Piotto et al., 2004). Similarly, in El Salvador and Nicaragua, homegardens are 

components of community development efforts in coffee cooperatives. They are 

assisted by local NGOs working on rural development and biodiversity conservation 

(Méndez and Bacon, 2005). 

77

78 

F. MONTAGNINI 

6. NUTRIENT CYCLING 

Efficient nutrient cycling is a key to the ecological sustainability of traditional 

homegardens, and species and structural diversity are critical to maintaining it 

through optimum use and transfer of carbon, water, and nutrients. Many traditional 

homegardens in Mesoamerica have survived for centuries despite many ecological, 

social, and political changes, justifying the claim that they are a sustainable land 

use system (e.g., the Maya homegardens of Yucatán Peninsula; Caballero, 1992). 

However, this cannot be generalized to all systems practiced by traditional peoples 

of the region. A comparison of land use and land clearing by Maya descendants and 

Hispanic populations in the Sierra de Lacandón National Park in Petén, Guatemala, 

is a case in point; not only agricultural land use by these two groups is very similar 

but also the impacts on land clearing are comparable (Carr, 2004). Population 

pressure and changes in other socioeconomic conditions thus strongly influence 

nutrient management and recycling, affecting the sustainability of homegardens. Yet 

many traditional societies retain the conventional wisdom on sustainable land 

management. 

In a study conducted in the northwestern and north-central regions of the 

Yucatán Peninsula of Mexico, Benjamin et al. (2001) hypothesized that Mayan 

farmers have been choosing tree associations and garden structures that maximize 

productivity and optimize nutrient cycling of the homegardens. At present, however, 

the Maya have ceased to use many of their earlier technologies that improved 

production, e.g., using raised beds and muck. Nevertheless, ‘modern’ Mayan 

homegardens still maintain relatively high yields using technologies of nutrient 

recycling such as mulching for residue management and fertilization (Benjamin 

et al., 2001). Soils in the region are very thin and contain rocks and calcium 

carbonates due to the shallow limestone bedrock. Low annual precipitation results in 

depleted surface and ground water resources, making large-scale irrigation a nonviable 

option. Benjamin et al. (2001) also noted that the Maya recognize appropriate 

tree species for such sites and know their growth characteristics; they also have 

the knowledge on appropriate nutrient management practices, which are applied in 

the design and management of homegardens. Irrigation timing, pruning, addition of 

ash to soils, and composting are some of the practices that Maya farmers use to 

enhance tree growth and survival, resulting in high fruit production with less 

investment in leaf biomass. 

Sweeping and burning of litter in homegardens results in the export of 

substantial amounts of nutrients, decreasing the effectiveness of nutrient cycling. 

Ash is recycled in the homegardens, although not uniformly. However, soils had 

high concentrations of organic matter. If litter were not removed, potential nitrogen 

contributions from litter to the homegardens would be very high (Benjamin et al., 

2001). 

Nutrient addition through the litter of nitrogen-fixing species is also a practice 

used in many homegardens in Mesoamerica. The Maya communities of San Jose, 

Belize, use the litter of Gliricidia sepium, a tree species native to Mesoamerica, to 

fertilize their homegardens (Levasseur and Oliver, 2000). In addition, practices that 

are aimed to controlling soil erosion also contribute to nutrient recycling through


M 

soil and nutrient conservation. The Totonacs in Coxquihui, Veracruz, Mexico, 

perceive soil loss as the most serious hazard to their traditional homegardens, and 

therefore, have sought to control erosion by retaining a continuous canopy cover, 

and using litter for mulching among other soil conservation practices (Del Angel- 

Pérez and Mendoza, 2004). 

The small size of homegardens allows for the application of intensive 

management practices that can improve nutrient recycling and lead to higher 

productivity. In Tacuba, El Salvador, farmers often open small trenches (about 30 

cm deep, few meters long, and set perpendicular to the direction of the slope) in 

their homegardens (Fig. 3). They drop household residues as well as prunings and 

other organic materials in the trenches. This avoids losses of residues that otherwise 

could be washed down the slope during the rains. They change the location of the 

trenches in the area of the homegarden so that eventually residues are recycled all 

over the homegarden area (pers. obs.). 

Figure 3. Recycling of household residue in homegardens in Tacuba, El Salvador. Farmers 

dig small trenches set perpendicular to the direction of the slope, where they deposit 

household residues, prunings and other organic material, to avoid losses of residues down the 

slope. 

79

80 

F. MONTAGNINI 

Manure of small animals is valuable as a nutrient source for the homegardens. 

This may be a localized effect as chickens or pigs often wander free in portions of 

the homegarden and their manure falls near cultivated plants. It can also be part of a 

specific management strategy, as chicken manure in regions of Costa Rica is used to 

fertilize small patches planted with corn that is used to feed the chickens as well as 

other animals of the homegarden (author’s pers. obs.). 

Vermiculture, or growing earthworms in worm boxes to use the castings for 

fertilizing homegarden soils, is used in many parts of Central America to increase 

the productivity of vegetable gardens and fruit trees. The high production of worm 

castings by certain earthworm species (e.g., the red Californian earthworm, 

Allophora caliginosa) is a source of cheap fertilizer for staple crops such as corn and 

sorghum (Sorghum bicolor). In rural areas of Nicaragua, some families sell the 

worm castings as organic manure (D. Craven, pers. comm., November 2005). 

As mentioned above, in Mesoamerica the traditional agricultural knowledge 

existing in many regions that still carry the influence of ancient Mayan indigenous 

peoples includes management practices that improve nutrient cycling. Some 

management practices can be redirected or improved to optimize plant productivity 

in homegardens. Improved litter management and knowledge of the relative nutrient 

content of the litter from different species when used as mulch or compost may be 

one avenue for improving both water and nutrient cycling and homegarden 

production. Composting of homegarden litter, instead of burning it, would augment 

the amounts of nutrients recycled in the system. Water retention, by adding organic 

matter via compost, would help to improve water availability for plants, especially 

important in homegardens located in subhumid and semiarid regions of Mesoamerica. 

Long-rotation production systems such as agroforests and homegardens can also 

sequester sizeable quantities of carbon in plant biomass and in long-lasting wood 

products (Albrecht and Kandji, 2003; Montagnini and Nair, 2004; Kumar, 2006). 

Many of the traditional homegardens already described share ecological characteristics 

and management practices that make them efficient in the use of solar 

radiation and carbon, and allow high levels of productivity. Soil carbon sequestration 

constitutes another realistic option achievable in homegardens. 


The region of Mesoamerica suffers from social and environmental problems due to 

overpopulation and rural poverty. Under such conditions, homegardens have 

traditionally fulfilled and still provide an important function in terms of ensuring a 

basic food supply for the family. This is especially important in the remote areas 

such as in indigenous reserves or in other rural settings in the relatively more 

impoverished countries of the region. 

Mesoamerican homegardens are quite diverse, with a complex vertical and 

horizontal structure that includes plants for food, ornamental, medicinal and other 

purposes. Mesoamerican homegardens are important reservoirs of local biodiversity 

and have a prominent role in the domestication of useful species.


M 

Domestic animals in homegardens of Mesoamerica contribute to increased food 

security. Animal manure also contributes to nutrient recycling. The inclusion of 

domestic animals in homegardens is vital to ensure a more sustained protein supply. 

However, they require an investment for the care and management of animals that 

would be relatively large for the poor, rural households. 

Mesoamerica was the home of the ancient Maya civilization, whose descendants 

still practice sustainable agriculture and manage homegardens in ways that increase 

the efficiency of the capture of solar radiation, increase productivity and improve 

nutrient cycling. Soil organic matter in homegardens can be increased by several 

practices of residue management. It is important that such sustainable management 

practices be retained to ensure homegarden sustainability. Homegardens of Mesoamerica 

also contribute to environmental services such as carbon sequestration, even 

though globally their role may be minimal due to their small land area. 


Dylan Craven, Philip Marshall, Quint Newcomer, and Tatsuhiro Ohkubo provided 

useful comments and insights to this manuscript. Dylan Craven also helped with the 

literature review and editing. 

ENDNOTE 

1. CCAD (Comisión Centroamericana de Ambiente y Desarrollo). 2003. Estado 

del Sistema Centroamericano de Areas Protegidas: informe de síntesis regional. 

Comisión Centroamericana de Ambiente y Desarrollo (Central American 

Comisión for Environment and Development). San José, Costa Rica. 33p. 

REFERENCES 

Albrecht A. and Kandji S.T. 2003. Carbon sequestration in tropical agroforestry systems. 

Agric Ecosyst and Environ 99: 15 – 27. 

Atran S. 1993. Itza Maya tropical agroforestry. Curr Anthropol 34: 633 – 700. 

Atran S., Medin D., Ross N., Lynch E., Coley J., Ucan Ek’ E., and Vapnarsky V. 1999. Folk 

ecology and commons management in Maya lowlands. Proc Natl Acad Sci USA 96: 

7598 – 7603. 

Barrera A., Gómez-Pompa A., and Vázquez-Yanes C. 1977. El manejo de las selvas por los 

Mayas: sus implicaciones silvícolas y agrícolas. Biotica 2(2): 47 – 61. 

Beach T. and Dunning N.P. 1995. Ancient Maya terracing and modern conservation in the 

Peten rain forest of Guatemala. J Soil Water Conservation 50: 138 – 145. 

Benjamin T.J., Montañez P.I., Jiménez J.J.M. and Gillespie A.R. 2001. Carbon, water and 

nutrient flux in Maya homegardens in the Yucatán Península of Mexico. Agroforest Syst 

53: 103 – 11. 

Blanckaert I., Swennen R.L., Paredes Flores M., Rosas López R., and Lira Saade R. 2004. 

Floristic composition, plant uses and management practices in homegardens of San Rafael 

Coxcatlán, Valley of Tehuacán, Mexico. J Arid Environ 57: 39 – 62. 

Budowski G. 1990. Home gardens in Tropical America: a review. In: Landauer K. and Brazil 

M. (eds), Tropical home gardens, pp 3 – 8. United Nations University, Tokyo. 

81

82 

F. MONTAGNINI 

Caballero J. 1992. Maya homegardens: past, present and future. Ethnoecologica 1(1): 35 – 54. 

Carr, D.L. 2004. Ladino and Q’eqchi Maya land use and land clearing in the Sierra de 

Lacandon National Park, Peten, Guatemala. Agric Hum Values 21: 67 – 76. 

Casas A., Valiente-Baunet A., Viveros J.L., Dávila P., Lira R., Caballero J., Cortés L., and 

Rodríguez I. 2001. Plant resources of the Tehuacán Valley, Mexico. Econ Bot 55: 

129 – 166. 

De Clerck F.A.J. and Negreros-Castillo P. 2000. Plant species of traditional homegardens of 

Mexico as analogs for multistrata agroforests. Agroforest Syst 48: 303 – 317. 

Del Angel-Perez A.L. and Mendoza M.A. 2004. Totonac homegardens and natural resources 

in Veracruz, Mexico. Agric Hum Values 21: 329 – 346. 

De Las Salas G. 1987. Suelos y ecosistemas forestales con énfasis en América tropical. 

Instituto Interamericano de Cooperación paraa la Agricultura (IICA), San José, 479p. 

Dommergues Y.R. 1987. The role of biological nitrogen fixation in agroforestry. In: Steppler 

H.A. and Nair P.K.R. (eds), Agroforestry: A decade of development, pp 245 – 272. 

ICRAF, Nairobi. 

Eibl B., Fernández R., Kozarik J.C., Lupi A., Montagnini F. and Nozzi D. 2000. Agroforestry 

systems with Ilex paraguariensis (American Holly or yerba mate) and native timber trees 

on small farms in Misiones, Argentina. Agroforest Syst 48: 1 – 8. 

Food and Agriculture Organization (FAO) 2001. Improving nutrition through home 

gardening. A training package for preparing field workers in Southeast Asia. Food and 

Agriculture Organization of the United Nations, Rome, 171p. 

Fedick S.L. and Morrison B.A. 2004. Ancient use and manipulation of landscape in the 

Yalahau region of the northern Maya lowlands. Agric Hum Values 21: 207 – 219. 

Gillespie A.R., Knudson D.M. and Geilfus F. 1993. The structure of four home gardens in the 

Peten, Guatemala. Agroforest Syst 24: 157 – 170. 

González-Soberanis C. and Casas A. 2004. Traditional management and domestication of 

tempesquistle, Sideroxylon palmeri (Sapotaceae) in the Tehuacan-Cuicatlán Valley, 

Central México. J Arid Environ 59: 245 – 258. 

Griffith D.M. 2000. Agroforestry: a refuge for tropical biodiversity after fire. Conserv Biol 

14: 325 – 326. 

Guindon C. 1996. The importance of forest fragments to the maintenance of regional 

biodiversity in Costa Rica. In: Schelhas J. and Greenberg R. (eds), Forest patches in 

tropical landscapes, pp 168 – 186. Island Press, Washington, D.C. 

Harvey C.A. and Haber W.H. 1999. Remnant trees and the conservation of biodiversity in 

Costa Rican pastures. Agroforest Syst 44: 37 – 68. 

Harvey C., Alpizar F., Chacón M. and Madrigal R. 2005. Assessing linkages between 

agriculture and biodiversity in Central America. Historical overview and future 

perspectives. Mesoamerican and Caribbean region, Conservation Science Program. The 

Nature Conservancy (TNC), San José, 138p. 

Holdridge L.R. 1987. Ecología basada en Zonas de Vida. Instituto Interamericano de 

Cooperación para la Agricultura (IICA), San José, 216p. 

House P.H. and Ochoa L. 1998. La diversidad de especies útiles en diez huertos en la aldea de 

Camalote, Honduras. In: Lok R. (ed.), Huertos caseros tradicionales de América Central: 

características, beneficios e importancia, desde un enfoque multidisciplinario, pp 61 – 84. 

Centro Agronómico Tropical de Investigación y Enseñanza (CATIE), Turrialba. 

Howard P.L. 2006. Gender and social dynamics in swidden and homegardens in Latin 

America In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 


Kumar B.M. 2006. Carbon sequestration potential of tropical homegardens In: Kumar B.M. 



Levasseur V. and Olivier A. 2000. The farming system and traditional agroforestry systems in 

the Maya community of San Jose, Belize. Agroforest Syst 49: 275 – 288. 

Lok R. 1998. El huerto casero tropical tradicional en América Central. In: Lok R. (ed.), 

Huertos caseros tradicionales de América Central: características, beneficios e importancia, 

desde un enfoque multidisciplinario, pp 7 – 28. Centro Agronómico Tropical de 

Investigación y Enseñanza (CATIE), Turrialba. 

Lok R. and Samaniego G. 1998. La valorización sociocultural del huerto y del café con 

árboles entre la población Ngöbe de Chiriquí, Panamá. In: Lok R. (ed.), Huertos caseros 

tradicionales de América Central: características, beneficios e importancia, desde un 

enfoque multidisciplinario, pp 185 – 213. Centro Agronómico Tropical de Investigación y 

Enseñanza (CATIE), Turrialba. 

Lok R., Wieman A. and Kass D. 1998. Influencia de las características del sitio y el acceso al 

agua en huertos de la Península de Nicoya, Costa Rica. In: Lok R. (ed.), Huertos caseros 

tradicionales de América Central: características, beneficios e importancia, desde un 

enfoque multidisciplinario, pp. 29 – 59. Centro Agronómico Tropical de Investigación y 

Enseñanza (CATIE), Turrialba. 

Méndez V.E. and Bacon C. 2005. Medios de vida y conservación de la biodiversidad arbórea: 

las experiencias de las cooperativas cafetaleras en El Salvador y Nicaragua. LEISA 

Revista de Agroecologia 20 (4): 27 – 30. 


M 

Méndez V.E., Lok R. and Somarriba E. 2001. Interdisciplinary analysis of homegardens in 

Nicaragua: micro-zonation, plant use and socioeconomic importance. Agroforest Syst 51: 

85 – 96. 

Miller R.P., Penn Jr. J.W. and van Leeuwen J. 2006. Amazonian homegardens: their 

ethnohistory and potential contribution to agroforestry development. In: Kumar B.M. and 

Nair P.K.R. (eds), Tropical homegardens: A time-tested example of sustainable agro- 

forestry, pp 43 – 60. Springer Science, Dordrecht. 

Montagnini F. and Nair P.K.R. 2004. Carbon sequestration: An under-exploited 

environmental benefit of agroforestry systems. Agroforest Syst 61: 281 – 295. 

Myers N., Mittermeier R.A., Mittermeier C.G., da Fonseca G.A.B. and Kent J. 2000. 

Biodiversity hotspots for conservation priorities. Nature 403: 853 – 858. 

National Academy of Sciences (NAS). 1979. Tropical legumes: Resources for the future. 

Board of Science and Technology for International Development, National Research 

Council, Washington, DC, 332p. 

National Academy of Sciences (NAS). 1980. Firewood crops. Shrubs and tree species for 

energy production. Board of Science and Technology for International Development, 

National Research Council, Washington, DC, 236p. 

Noble I.R and Dirzo R. 1997. Forests as human-dominated ecosystems. Science 277: 

522 – 525. 

Piotto, D., Montagnini F., Kanninen M., Ugalde L. and Viquez E. 2004. Forest plantations in 

Costa Rica and Nicaragua: performance off species and preferences of farmers. 

J Sustainable For 18: 59 – 77. 

Price N.W. 1989. The tropical mixed garden in Costa Rica. PhD Thesis. University of British 

Columbia, Vancouver, 403p. 

Rico-Gray V., Chemas A. and Mandujano S. 1991. Uses of tropical deciduous forest species 

by the Yucatecan Maya. Agroforest Syst 14 (2): 149 – 161. 

Richards P.W. 1996. The tropical rainforest: An ecological study (2 nd ed.). Cambridge 

University Press, London, 450p. 

Sanchez P.A. 1976. Properties and management of soils in the tropics. John Wiley and Sons, 

New York, 618p. 

83

84 

F. MONTAGNINI 

Steinberg M.K. 1998. Political ecology and cultural change: impacts on swidden-fallow 

agroforestry practices among the Mopan Maya in southern Belize. Prof Geogr 50: 

407 – 417. 

Toledo V.M., Batis A.I., Becerra R. Martinez E. and Ramos C.H. 1995. The useful forest: 

quantitative ethnobotany of the indigenous groups of the humid tropics of Mexico. 

Interciencia 20: 177 – 187. 

Turner II B.L. and Harrison P.D. 1981. Prehistoric raised-field agriculture in the Maya 

lowlands. Science 213: 399 – 405. 

Wezel A. and Bender S. 2003. Plant species diversity of homegardens of Cuba and its 

significance for household food supply. Agroforest Syst 57: 39 – 49. 

Wieman A. and Leal D. 1998. La cría de animales menores en los huertos caseros. In: Lok R. 

(ed.), Huertos caseros tradicionales de América Central: características, beneficios e 

importancia, desde un enfoque multidisciplinario, pp 85 – 115. Centro Agronómico 

Tropical de Investigación y Enseñanza (CATIE), Turrialba. 

Yamada M. and Osaqui H.M.L. 2006. The role of homegardens for agroforestry development: 

lessons from a Japanese-Brazilian settlement in the Amazon. In: Kumar B.M. and Nair 

P.K.R. (eds), Tropical homegardens: A time-tested example of sustainable agroforestry, 

pp 299 – 316. Springer Science, Dordrecht. 

Zaldivar M.E., Rocha O.J., Castro E. and Barrantes R. 2002. Species diversity of edible 

plants grown in homegardens of Chibchan Amerindians from Costa Rica. Hum Ecol 30: 

301 – 316.

SECTION 2 

STRUCTURE, FUNCTION, AND DYNAMICS 

OF HOMEGARDENS

CHAPTER 6 

HOMEGARDEN DYNAMICS 

IN KERALA, INDIA 

A. PEYRE 1,2 , A. GUIDAL 1,2 , K.F. WIERSUM 2 *, AND 

F. BONGERS 1 

1 Forest Ecology and Forest Management group and 2 Forest and Nature 

Conservation Policy group, Wageningen University, The Netherlands; *E-mail: 

 

Note: Adapted from: Peyre A., Guidal A., Wiersum K.F. and Bongers F. 2006. Dynamics of 

homegarden structure and function in Kerala, India. Agroforestry Systems 66: 101 – 115. 

Keywords: Biodiversity, Functional differentiation, Multispecies systems, Social adaptability, 

Socioeconomic change. 

Abstract. Homegardens in Kerala, India, have long been important multipurpose agroforestry 

systems that combine ecological and socioeconomic sustainability. These traditional 

homegardens, however, are subject to changes consequent to various on-going socioeconomic 

transformations. The study of structural and functional dynamics of homegardens offers an 

opportunity to understand the trends in socioeconomic sustainability in relation to their 

ecological sustainability. These dynamics were studied in a survey of 30 homegardens. Based 

on a cluster analysis of tree/shrub species density and a subsequent grouping using 

homegarden size, six homegarden types were differentiated, and these were assessed for 

structural, functional, and managerial characteristics, besides their dynamics. Four 

development stages of homegardens were found along a gradient from traditional to modern 

homegardens. Fifty percent of the homegardens still displayed traditional features, whereas 

33% incorporated modern practices. The process of modernization includes a decrease of the 

tree/shrub diversity, a gradual concentration on a limited number of cash crop species, an 

increase of ornamental plants, a gradual homogenization of homegarden structure and an 

increased use of external inputs. A traditional homegarden combining multispecies 

composition and intensive management practices could, however, offer an alternative 

development path to modern homegardens in adapting homegardens to changing 

socioeconomic conditions. 

87 




88 

A. PEYRE ET AL. 


Homegardens are recognized worldwide as an epitome of a sustainable agroforestry 

system (Torquebiau, 1992; Kumar and Nair, 2004). From a system-dynamics point 

of view, the concept of sustainability includes two main dimensions (Wiersum, 

1995): ecological sustainability (in the sense of keeping within ecological stability 

domains) and social sustainability (in the sense of adjusting to social dynamics). 

Most studies on sustainability of homegardens have been focused on ecological 

sustainability, while social sustainability has been given much less systematic 

attention (Torquebiau, 1992; Kumar and Nair, 2004). Social sustainability may 

relate either to the social acceptability of homegardens within the livelihood systems 

of rural producers or to the ability of homegardens to adjust to socioeconomic 

changes. The structure and composition of homegardens can well be adjusted to 

various livelihood conditions such as size of landholdings, role of homegardens 

within the overall farming-system and degree of commercialization. However, 

homegardens are not static, but have evolved over centuries; thanks to adaptive 

abilities of farmers in responding to changing rural and livelihood conditions 

(Wiersum, 2006). Traditionally, the homegardens mainly served to produce 

vegetables, fruits and other crops, which supplemented the staple food crops 

produced on open croplands. With the advent of commercialization, often an 

increase in selected cash crops such as coconut (Cocos nucifera) or rubber (Hevea 

brasiliensis) has been observed. The shift from subsistence agriculture to market 

economy often implies drastic structural and functional modifications, including a 

homogenization of the homegarden structure and use of external inputs (Kumar and 

Nair, 2004). 

Several authors have voiced concern that these developments result in the loss of 

relevance of the homegardens and threaten their future development. Recently the 

question was even raised whether the homegardens are becoming irrelevant or even 

extinct (Kumar and Nair, 2004). The expressed fears that the traditional, diverse and 

ecologically sustainable homegardens will gradually dissolve into monospecific 

agricultural systems with uncertain sustainability are in stark contrast to the earlier 

ideas on homegardens as having a promising future (Soemarwoto, 1987). The 

maintenance of multispecies and multistrata agroforests is deemed worthwhile 

because of the growing interest in developing multifunctional land use systems, 

which contribute not only to production objectives, but also to the objectives of 

biodiversity and environmental conservation. In order to maintain the positive 

characteristics of the traditional homegardens it is therefore necessary to develop 

improved homegardens that counterbalance the ongoing trend of homogenization. 

In order to better understand whether there is scope for such an alternative 

development path, and whether it is possible to adapt homegarden systems to the 

changing rural conditions while still maintaining the positive features of the 

traditional homegardens, it is necessary to study the trends in homegarden dynamics 

in detail. Up until the present, most homegarden studies have focused mainly on 

species inventories or system description (Nair, 2001) and still little attention has 

been given to their structural and functional evolution. In the past, differences 

between homegardens were mostly described based on characteristics such as size,

structure (vertical stratification, diversity indices) or socioeconomic factors (level of 

inputs, subsistence/commercial production). Only recently, studies have been undertaken 

to systematically classify the structure of homegardens using analytical 

methods such as cluster analysis common to vegetation science (Leiva et al., 2002; 

Quiroz et al., 2002; Mendez et al., 2001; Tesfaye Abebe, 2005). These methods 

offer good opportunities for obtaining a systematic insight into different types of 

homegardens. The further evaluation of these different types in respect to socioeconomic 

conditions, under which they evolved, can provide useful insights into the 

development trends of homegardens. 

Homegardens have traditionally been managed and adopted by farmers rather 

than through agroforestry research (Nair, 2001). Consequently, an interesting 

question is whether all farmers are following similar homegarden development 

trends, or whether farmers are following different pathways in maintaining their 

homegardens. The recent advances in using statistical methods for classification of 

homegarden systems provides a good basis for assessing whether there exist 

differences in homegarden types and evaluating whether different types follow 

different development trends. 

Based on these considerations, the present study was undertaken with the 

objective of assessing the nature of dynamics of homegarden characteristics. It 

focused on the following questions: 

1. What different types of homegarden are present in the study area and what 

are their characteristics? 

2. What changes in homegarden structure, function, and management 

characteristics took place during the past decade? 

3. What conclusion can be drawn regarding the position of the different 

homegarden types on an evolutionary axis? 

2.1. Study site 

HOMEGARDEN DYNAMICS IN KERALA, K INDIA I 89 

2. MATERIALS AND METHODS 

The study was conducted in Palakkad district, Kerala, India, which is one of the 

tropical regions where concerns about the future of homegardens have been raised 1 . 

In this region, the value of homegardens as multipurpose production systems 

combining ecological and socioeconomic sustainability is well-recognized (Nair and 

Sreedharan, 1986; Jose and Shanmugaratnam, 1993; Kumar et al., 1994). Various 

authors have voiced concern that socioeconomic changes and related adoption of 

modern managerial systems cause a negative conversion process of homegardens in 

this region (Jose and Shanmugaratnam, 1993; John and Nair, 1999). Several 

government development programs want to assist the farmers to raise their cash 

incomes and therefore promote the conversion of homegardens towards cashcropping 

systems by providing loans and subsidies for rubber or other cash-crop 

cultivation. Moreover, government controls on timber production discourage the 

growing of timber in homegardens.

90 

In view of these developments, Kerala offers a good opportunity to study the 

development trends in agroforestry systems. The case study was carried out in two 

panchayaths (administrative units: Mundur and Puduparyaram) of Palakkad district 

in central Kerala. The region is characterized by a tropical humid climate with a 

monsoonal pattern of rainfall. The topography is rolling to hilly and main soil types 

are Dystric Nitosols (FAO, 1977). 

2.2. Data collection 

Within the two study panchayaths, a stratified sample of 30 farm households was 

selected. The sample was stratified according to total landholding size, i.e., small 

(< 1 ha; n = 10), medium (1 to 2 ha; n = 10) and large (> 2 ha; n = 10). The 

households were selected based on information from a local rural development 

organization (Integrated Rural Technology Centre, Mundur, Palakkad) and from 

referrals of initial respondents. 

In the homegardens of each household, a detailed survey of the composition and 

management practices was made. The survey consisted of an inventory of tree and 

shrub species and a count of all individuals per species. Only presence was recorded 

for herbs and (bi)-annuals. The species were classified according to their use into the 

categories: fruits and nuts, staple food, beverages and stimulants, spices, timber and 

firewood, medicinal products, religious plants, ornamentals, multipurpose species 

with more than four uses, and others. Rubber was the only species used exclusively 

as a cash crop and classified as such. During the survey, information was also 

collected on the management practices for individual species based on the approach 

developed by Wiersum and Slingerland 2 . In this approach, five main practices are 

distinguished: controlled utilization, protection, and maintenance, stimulation of 

desired products, regeneration, and interface management. The first four categories 

represent an increasing input of human energy per unit of land (Wiersum, 1997). 

Additional information concerning the homegarden size, the overall strategy in 

homegardens orientation (subsistence or commercial) and management inputs was 

collected using structured interviews. In these interviews, additional information 

was collected on changes in homegarden structure and management during the 

preceding ten years. This concerned both changes in homegarden composition and 

spatial arrangements (including homegarden size) as well as in management 

practices (changes in vegetation structure, production characteristics, and chemical 

input use). 

2.3. Data processing and analysis 

A.PEYRE ET AL. 

A hierarchical cluster analysis was applied for classification of the 30 homegardens 

using tree/shrub species density (number of f individuals per species per unit area) as


the main variable. Chi-square was used as distance or similarity measure and 

between-group average linkage method. Nine clusters were distinguished of which 

five consisted of only one homegarden each. Those five “single” clusters were 

reclassified based on homegarden size into two new types. Thus, a group of four 

small homegardens and a “group” of one big homegarden emerged. These six 

homegarden types were assessed with respect to their structural, functional and 

management characteristics as well as dynamics. 

Structural characteristics: Four parameters were used to assess the structural 

attributes: homegarden size (land area including the house), total density of trees per 

homegarden, and species richness and evenness (except for species that could not be 

counted)—computed using Shannon’s and Simpson’s diversity indices (following 

Huston, 1994). Since only three of the six types had a sufficient number of 

homegardens, statistical analysis was only applied to differences among these 

homegarden types. Differences were tested using ANOVA for all the parameters 

except for number of species and tree density, as populations were not normally 

distributed even after transformation. In this case, a non-parametric Kruskall and 

Wallis test was applied. 

Functional characteristics: Two parameters were used to assess the functional 

characteristics: the proportions of mean number of trees per use-category and the 

differentiation in home- or cash-orientation in production. Relative contribution of 

each use group was calculated and compared within each homegarden type. Annual 

staple food crops and ‘other’ crops were not included in this analysis; coconut and 

rubber were treated as separate categories in view of their high value according to 

both farmers’ opinions and actual situation. 

Management characteristics: Management was characterized in respect to 

management intensity, spatial arrangement, and use of inputs. Assessment of the 

management intensity was based on a detailed assessment of the management 

practices for the seven most common and preferred species: rubber, coconut, arecanut 

(Areca ( catechu), 

mango (Mangifera indica), jackfruit (Artocarpus ( heterophyllus), 

teak 

(Tectona grandis) and neem ( (Azadirachta indica). 

It was characterized on a 

comparative scale according to the technique of Wiersum and Slingerland 2 . The 

characterization of management inputs was based on an assessment of the internal 

and external inputs applied in cultivating the seven tree species. 

Homegarden dynamics: The dynamics of each homegarden type were qualitatively 

assessed according to the changes in homegarden size, vegetation structure (introduction 

of new species, changes in respect to ornamental and medicinal plants, and 

changes in spatial arrangements), and production characteristics (change of homegarden 

orientation and evolution of chemical input use). 

The data were analyzed using the statistical package SPSS 10.0 (SPSS Inc.). 

Based on structure, functions, management, and dynamics, the different homegarden 

types were arranged along a gradient from traditional l to modern homegardens.

92 

3.1. Distinction in homegarden types 


3. RESULTS 

Based on the cluster analysis using a dissimilarity index of 12.2 as a cut-off point, 

the 30 selected homegardens were categorized into nine clusters with different 

patterns of tree/shrub species density (Fig. 1). As five clusters consisted of only one 

Figure 1. Hierarchical classification of 30 homegardens in Kerala, India. 

homegarden each, a qualitative assessment was made to further delineate different 

homegarden types. Four clusters (clusters 4, 5, 6 and 7) which were similar in 

respect of their very small size (0.12 to 0.2 ha) were combined. Cluster 8 was

maintained as a specific type due to its large size (0.81 ha) and specific structure. 

Consequently, the nine clusters were regrouped into six homegarden types for 

further analysis (Fig. 1). 

3.2. Structural characteristics 

The different homegarden types showed important variations in all their structural 

characteristics except for the number of species (Table 1). Although the average 

number of species in the various homegarden types ranged from 17 to 51, in types 1, 

2, 4, and 6, the average number of species per homegarden were relatively similar. 

Types 2, 3, and 4 are small homegardens, whereas types 5, 6, and 1 are much larger. 

Types 1, 2, and 3 have a much lower tree density than types 4, 5, and 6. Types 1, 3, 

and 6 have lower species diversities (higher Simpson and lower Shannon indices) 

compared to types 2, 4, and 5. Finally, types 1 and 6 have lower evenness values 

than the other garden types. This indicates that in these homegarden types, 

production is oriented toward fewer species compared to types 2, 4, and 5. Further 

statistical tests were applied on types 1, 2, and 3 (Table 1). Type 1 is significantly 

larger than types 2 and 3. The three types are significantly different in respect to 

their number of species, but have similar tree/shrub densities per homegarden. 

Concerning the diversity indices, type 1 is statistically less diverse and has a lower 

evenness compared to type 2. Type 3 is intermediate. 

Attributes 

Homegarden 

size (ha) 

Number of 

species 

Density 

(No./ha) 

Simpson’s 

index 

Shannon’s 

index 

Evenness 

1 

(n = 8) 

0.72 a 

(0.117) 

27.1 

(3.47) 

555.5 a 

(57.99) 

0.35 a 

(0.077) 

0.79 a 


Table 1. The structural characteristics of six homegarden types, Kerala, India. 

2 

(n = 11) 

0.40 b 

(0.076) 

28.7 

(1.33) 

449.0 a 

(54.9) 

0.08 b 

(0.007) 

1.24 b 

Homegarden types Tests 

3 

(n = 4) 

0.24 b 

(0.057) 

17.7 

(3.09) 

621.3 a 

(128.9) 

0.20 ab 

(0.028) 

0.89 a 

(0.108) (0.020) (0.092) 

0.56 a 

0.86 

(0.060) 

b 

0.72 

(0.013) 

b 

(0.052) 

4 

(n = 4) 

0.14 

(0.020) 

27.5 

(3.77) 

1105.8 

(137.7) 

0.09 

(0.015) 

1.21 

(0.076) 

0.84 

(0.018) 

5 

(n = 1) 

0.81 

(na) 

51.0 

(na) 

1671.6 

(na) 

0.09 

(na) 

1.32 

(na) 

0.77 

(na) 

6 

(n = 2) 

1.01 

(0.200) 

24.0 

(3.00) 

1387.9 

(46.6) 

0.51 

(0.068) 

0.50 

(0.087) 

0.36 

(0.049) 

Type F/Chi2 

Values in parentheses represent the standard error (na = not applicable). 

Values with different letters among homegardens are significantly different. 

Tests: A = ANOVA, KW = Kruskall Wallis; F value for ANOVA, Chi-square values for 

Kruskall & Wallis tests and P = probability level of significance. 

P 

A 4.04 0.034 

KW 6.14 0.046 

KW 2.74 0.254 

A 9.33 0.001 

A 13.18 0.001 

A 16.65 0.001

94 

In general, there is a tendency that with an increase in size of homegardens [from 

type 4 (very small) to types 2 and 3 (small) to type 1 (medium) and type 6 (very 

big)], there is an increase in the Simpson’s diversity index, a decrease in the 

Shannon’s diversity, and a decrease in the evenness index. Only homegarden type 5 

does not fit into this pattern; this big homegarden has a low Simpson’s diversity 

index, a high Shannon’s diversity index, and a high evenness index. 

3.3. Functional characteristics 


The various homegarden types differ in functional characteristics (Fig. 2). A range 

from five to nine use-categories was present in the homegarden types. Fruits and 

nuts, spices, timber, and coconut, are present in all homegarden types. In two types, 

only one use group consists of more than 50% of all trees: rubber in type 1 and 

beverage in type 6. The relatively less important use categories are the ones of 

religious, medicinal, and multipurpose trees. 

Figure 2. Functional characteristics of six different homegarden types, Kerala, India (% of 

number of individuals in each functional use group). 

The different types of homegardens can be arranged along a gradient from 

predominately single commodity production to intensive multiple cropping. Type 1 

is predominantly focused on rubber production, types 3 and 4 are characterized by a 

mixture of fruit trees and coconuts, type 6 had a mixture of fruit trees and beverage 

crops with some additional spices and timber trees, type 2 was a mixture of coconut, 

fruit trees, and timber trees, and type 5 by an intensive mixture of timber trees, 

spices, fruit trees, and beverage crops. These characteristics are related to the 

differences in whether farmers are oriented toward cash income generation or home

consumption. Farmers managing type 1 and 6 are cash-oriented as reflected by the 

dominance of rubber and beverage producing crops or stimulant producing trees 

respectively. In these gardens, cash crops represent more than half of the total 

number of trees. Also the farmer managing homegarden 5 is cash-production 

oriented, but in this case there is no clear dominance of any crop. The managers of 

homegarden type 3 are focused on both cash income generation (coconuts) and 

subsistence production (fruits). The homegarden types 2 and 4 are both home 

consumption oriented. These homegardens are small in contrast to the cash-oriented 

types 5 and 6, which are larger in size. The garden type 1 with highest production 

specialization in rubber production is of medium size. 

3.4. Management characteristics 

Most management practices concern the manipulation of the tree environment rather 

than the tree itself. Sanitary pruning, rejuvenation pruning, canopy pruning to 

increase light penetration and cutting low branches are seldom, whereas weeding, 

fertilization, and crop spacing are more common (Table 2). In particular, cash crops 

are subjected to a variety of management practices. The most intensively managed 

species are coconut, rubber and arecanut; their cultivation includes use of chemical 

fertilizers and insecticides, systematic weeding, organic fertilization, and row 

arrangement of trees. They are also relatively often protected from competitors and 

are the only crops that receive watering. Fruit trees and neem receive less attention 

(selective weeding, some application of organic manures). The valuable timber 

species, teak receives no particular attention to increase its productivity. 

The different homegarden types can be arranged along a gradient of management 

intensity. The small-sized types 4 and 2 are on one end of this gradient characterized 

by low to medium management with a concentration on internal inputs and with 

random arrangement of trees. On the other extreme of the gradient, the medium- to 

big sized homegarden types 1, 5, and 6 are subject to a more intensive management 

with use of both internal and external inputs such as chemical fertilizers, insecticides 

and purchased seedlings. In this case, row planting is dominant. The small-sized 

type 3 has intermediate characteristics, with only a medium intensity of management, 

but with a dominant spatial arrangement in row. In summary, the smaller 

homegarden types are managed at a lower intensity than the larger ones, but their 

production is more diverse. Type 5 has the distinctive feature of being a large garden 

with very intensive management but low use of external inputs; this homegarden is 

oriented at multiple-production. 

3.5. Homegarden dynamics 


During the past decade, there has been hardly any change in homegarden size. 

However, several changes in the structure and function occurred; these varied for the 

different homegarden types (Table 3). Spices (black pepper or Piper nigrum) and 

ornamental species are the only use categories that have been introduced in all 

homegarden types except in types 5 and 3 where they were already present.

96 

Table 2. Management practices of seven common species in different homegarden types. 

Inputs 

Inputs 

Inputs 

Management 

Management 

Management 


Species/management HG types 

1 2 3 4 5 6 

presence 8/8 0/11 0/4 0/4 1/1 0/2 

latex tapping 8 - - - 1 - 

weeding 6 - - - 1 - 

removing competition 4 - - - 0 - 

sanitary pruning 0 - - - 1 - 

cut low branches 0 - - - 1 - 

spatial arrangement 2 - - - 2 - 

Int organic fertilization 8 - - - 1 - 

Ext 

bought seedlings 

chemicals 

7 

8 

- 

- 

- 

- 

- 

- 

1 

0 

- 

- 

presence 8/8 11/11 4/4 4/4 1/1 2/2 

nut harvesting 8 11 4 4 1 2 

weeding 8 6 3 0 1 2 

watering 0 4 2 2 0 2 

ringing 8 9 4 0 0 2 

sanitary pruning 0 0 0 0 1 0 

removing competition 4 6 2 0 0 0 

leaf harvesting 0 5 2 0 0 2 

canopy prunings 0 0 0 0 1 0 

spatial arrangement 2 2 2 2 2 2 

organic fertilization 7 10 4 3 1 2 

Int 

mulching 

seeding 

8 

0 

10 

0 

4 

0 

3 

0 

0 

0 

1 

2 

nursery 0 0 2 0 0 1 

Ext 


chemicals 

5 

5 

7 

0 

2 

2 

4 

0 

1 

1 

0 

0 

presence 3/8 1/11 4/4 0/4 0/1 2/2 

nut harvesting 3 1 4 - - 2 

weeding 3 1 0 - - 2 

watering 2 0 0 - - 2 

ringing 3 1 2 - - 1 

sanitary pruning 0 0 0 - - 1 

removing competition 0 0 0 - - 1 

spatial arrangement 2 2 1/2 - - 1/2 

organic fertilization 3 1 3 - - 2 

Int 

mulching 

seeding 

2 

2 

0 

0 

2 

0 

- 

- 

- 

- 

0 

1 

nursery 0 0 1 - - 1 

Ext 


chemicals 

0 

2 

1 

0 

0 

0 

- 

- 

- 

- 

1 

0

MANGO 

JACKFRUIT 

TEAK 

Inputs 

Inputs 

Inputs 

Management 

Management 

Management 


Species/management H G types 

1 2 3 4 5 6 

presence 7/8 11/11 4/4 4/4 1/1 2/2 

fruit harvesting 7 11 4 4 1 2 


rejuvenation pruning 0 0 0 0 1 0 

canopy pruning 0 0 0 0 1 0 

lopping 2 2 2 0 1 0 

weeding 0 0 0 0 1 0 

cutting low branches 0 0 0 0 1 0 



seeding 3 3 0 0 0 0 

Int 

nursery 0 3 3 0 1 0 

protecting natural regeneration 0 0 0 2 0 1 

plant cuttings 0 0 0 0 0 1 

Ext 


chemicals 

3 

0 

5 

3 

0 

0 

3 

0 

0 

0 

0 

0 

presence 8/8 10/11 4/4 3/4 1/1 2/2 

fruit harvesting 8 10 4 3 1 2 




lopping 0 2 0 0 1 0 

weeding 0 0 0 0 1 0 




Int 

seeding 

nursery 

4 

0 

5 

0 

0 

0 

2 

0 

0 

1 

1 

0 


Ext 


wildings 

0 

0 

0 

3 

0 

0 

0 

0 

0 

0 

0 

0 

presence 6/8 10/11 1/4 3/4 1/1 2/2 




lopping 3 2 0 0 1 2 

weeding 0 2 0 0 1 1 


coppicing 0 0 0 0 1 1 

spatial arrangement 4 4 1 4 4 1/4 


Int 

nursery 0 2 0 0 1 1 


Ext bought seedlings 4 3 1 0 0 1

98 

NEEM 

Inputs 

Management 


Species/management HG types 

1 2 3 4 5 6 

presence 5/8 5/11 0/4 2/4 1/1 0/2 

leaf harvesting 3 5 - 0 1 - 

sanitary pruning 0 2 - 0 1 - 

rejuvenation pruning 0 0 - 0 1 - 

lopping 0 0 - 0 1 - 

cutting low branches 0 0 - 0 1 - 

spatial arrangement 1 4 4 4 1/4 4 

organic fertilization 0 0 - 0 1 - 

Int 

nursery 0 0 - 0 1 - 

protecting natural regeneration 5 3 - 1 0 - 

Ext bought seedlings 0 0 - 1 0 - 

Presence: a/b a = number of homegardens studied; b = number of homegardens with species; 

(-) tree is not present; HG = homegardens, Int = internal, Ext = external; Legend for spatial 

arrangements: 1 = borders, 2 = rows, 3 = strips, 4 = scattered. 

Ornamentals are usually cultivated around and in front of the house and along the 

paths. Black pepper is usually associated with palm trees in order to benefit from 

their soil management and inputs. Some farmers reported difficulties to harvest the 

palm nuts without damaging the pepper vines. Also some other support trees such as 

Erythrina spp. were introduced. In five out of the six homegarden types, fruit trees 

have also been introduced. They are usually cultivated close to the house, except for 

big trees such as mango or jackfruit or when planted in a large-scale. Another 

change concerns the medicinal plants. In the homegarden types 1, 5, and 6, many 

farmers have partially removed the medicinal species. 

Few structural and functional changes have occurred in types 2 and 4, especially 

when compared to types 1 and 6. The large majority (93%) of the homegardens of 

types 2 and 4 are still subsistence-oriented, just as they were 10 years ago. Crop 

introductions do not concentrate on any specific species or use and are of low 

intensity (less than 50 individuals per species). These homegardens have preserved 

the traditional features; they still have a multistoried structure, high diversity and 

low dependency on external inputs. In contrast, 60% of farmers managing types 1 

and 6 have shifted to a cash strategy with a modernized management oriented 

toward a few cash crops such as rubber, arecanut, and coffee. The introduction of 

these commercial crops resulted in important structural and functional changes. The 

canopy became less stratified and species diversity was reduced, notably in respect 

to species producing fruits and nuts, timber, and medicines. This caused a reduction 

in the multiple functions of homegardens. This change was most dominant in case of 

increased rubber cultivation, as this species is always grown as a monoculture. 

Coconut and arecanut are often still intercropped. Moreover, 70% of the farmers 

increased their use of chemical inputs. 

No clear pattern could be deduced concerning the dynamics of the homegarden 

types 3 and 5. Although the production pattern of type 3 changed, its vegetation 

structure did not undergo any fundamental modification and the predominant spatial 

arrangement of trees remained in rows. The owner of type 5 follows a long-term 

cash strategy oriented toward timber production. The farmer has been able to follow

the market demand by introducing more rubber and arecanut trees. However, these 

introductions did not affect the structural characteristics and vegetation structure. 

Table 3. Species introductions and changes in spatial arrangements in the period 1993 – 2003 in 

six homegarden types, Kerala, India. 

Garden 

types 

Introduced 

uses/species 

Type 4 Spices (pepper) 

Fruits (Citrus, 

guava) 

Beverage (coffee) 

Ornamentals 


Fruits (Citrus, 

guava, Annona, 

papaya) 

Ornamentals 


Fruits (guava, jack, 

cashew, papaya) 

Type 5 Beverage (arecanut) 

Cash (rubber) 

Ornamentals 

Type 6 Timber (teak) 

Ornamentals 

Beverage (arecanut, 

coffee) 

Spices (pepper) 

Fruits (banana) 


Fruits (guava, 

Citrus) 

Ornamentals 

Cash (rubber) 


Rate of 

introduction 

low 

low 

low 

low 

low 

low 

low 

low 

low 

low 

low 

low 

low 

low 

high 

high 

high 

low 

low 

low 

high 

Spatial arrangements of trees 

10 years ago Nowadays 

random random 

random random 

row row 

partly 

random, 

partly rows 

random row 

random row 

partly 

random, 

partly rows 

Low = less than 50 individuals introduced in total; High = more than 50 individuals 

introduced in total. 

3.6. Classification of homegarden types on an evolutionary axis 

Based on their structural and functional characteristics and dynamics, the different 

homegarden types can be arranged along a gradient from traditionall to modernized 

homegardens (Table 4). The homegarden types 2 and 4 are relatively small and have 

a high diversity and a random arrangement of trees. Few changes occurred during 

the past decade and the traditional features of homegardens have been preserved

100 

(high diversity, multi-storied canopy, and multi-production). These homegardens are 

oriented toward subsistence production and few products are sold. The management 

practices are predominantly based on internal inputs, although in type 2 some 

external inputs are also used. Based on these characteristics, they can be 

characterized as traditional l homegardens. These traditional homegardens can be 

contrasted with homegarden types 1 and 6, which can be characterized as “modern.” 

In these modern homegarden types, farmers have adopted a cash-orientation and 

have introduced several new management practices. In these relatively big 

homegardens the production became oriented at a few cash crops which are 

systematically arranged in rows. In the case of rubber, part of the homegarden is 

even transformed into single species plantation. Also, the use of external inputs 

(purchased seedling, chemical fertilizers, and insecticides) has increased. 

Table 4. Ordination of homegarden types along a gradient from traditional to modernized 

homegarden, Kerala, India. 

HG 

categories 

Type (s) No. of 

HGs 

Size Orientation Nature of 

production 

Tree/shrub 

diversity 

Traditional 2,4 15 (very) small home multiple high 

Adapted 

traditional 

6 1 big cash multiple high 

Incipient 3 4 small home and multiple medium 

modern 

cash 

Modern 5,1 10 medium to 

very big 

cash mono low 

HG = homegardens. 


Type 3 can be considered as incipient modern type as it shares both traditional 

and modern characteristics. This homegarden type consists of small homegardens 

with medium diversity, and involves a low management intensity that depends 

predominantly on external inputs. Although type 5 is characterized by its cash 

orientation including introduction of new cash crops such as rubber and systematic 

spacing of trees, it still maintains the multispecies composition of the traditional 

homegardens. The garden is very intensively managed, but mostly with internal 

inputs by using organic fertilization and mulching for soil management and by 

regenerating trees by protecting natural regeneration, seeding, and using local plant 

material such as plant cuttings. Thus, although this homegarden was adapted to the 

modern cash economy, it maintained several characteristics of the traditional 

homegardens. 

4. DISCUSSION AND CONCLUSIONS 

This study shows that homegardens should not be considered as being static. Rather, 

their composition and management are gradually evolving in response to the


socioeconomic dynamics. Only 50% of all respondents still followed traditional 

homegarden management practices, whereas 33% of all respondents have adopted 

modern practices by increasingly moving towards concentrated cash crop production 

and use of external inputs. Traditional homegardens were mostly of small size, while 

modern homegardens are much larger. This parameter should not be interpreted as 

the only, or main, feature influencing the development path of homegardens. Other 

factors, such as the role of the homegarden in the overall farming system and the 

degree to which a household has access to off-farm employment and income 

(Wiersum, 2006) might be of more importance. Unfortunately, these factors could 

not be taken into account in the present study. 

Our data reinforce the general fears regarding the loss of traditional characteristics 

of homegardens and their gradual demise into cash crop production systems 

(Kumar and Nair, 2004). Because of the rise of market economy, agriculture in 

Palakkad region of Kerala is currently struggling to find new intensification 

strategies. Although traditional Kerala homegardens are reputed to be sustainable in 

both biophysical and socioeconomic terms, they do gradually change from a 

traditional type to a more modern one. This process of modernization often brings 

with it a decrease of the tree/shrub diversity, a gradual concentration on a limited 

number of cash-crop species, gradual homogenization of homegarden structure and 

increased use of external inputs. 

Interestingly, however, one farmer in our sample had combined an increased 

orientation at cash crop production with the maintenance of a high species diversity 

and use of internal rather than external inputs. This example shows that there is no 

single uniform trend towards the modernization of homegardens in Kerala, but that 

alternative pathways exist. Moreover, this example also shows that traditional 

ecological features ensuring ecological sustainability of homegardens could still be 

maintained in modernized homegardens. This suggests that it might be possible to 

identify new development policies that aim at optimal combination of ecological and 

productive features of the homegardens rather than optimizing only cash crop 

production. 

Although the study was focused on ascertaining trends in tree composition 

resulting from the process of commercialization, other trends influencing the composition 

of the homegarden vegetation were also observed (Wiersum, 2006). These 

included an increase in the use of ornamental plants and an increase in staple food 

production. The trend in gradual replacement of functional plants to ornamentals has 

also been observed in cases where people became richer. The gradual increase in 

staple food production was specifically found in cases where homegardens were the 

last remaining farming unit of poor households. Unfortunately, little attention has 

been given towards systematically studying under which set of conditions these 

different trends in homegarden development occur, to what extent they are 

interrelated, how they are related towards changes in livelihood conditions, and what 

their impact on biodiversity is. 

Our study further shows that it is incorrect to assume a uniform development 

pattern for all homegardens, rather different pathways in homegarden development 

may co-exist. At present rural areas are subject to many socioeconomic changes 

(Ashley and Maxwell, 2001). The notion of homegardens being sustainable needs

102 


therefore to be specified in respect to ecological and social sustainability. Whereas 

the concept of ecological sustainability is time-independent, the concept of social 

sustainability includes the notion of agroforestry systems adjusting in a timely 

fashion to changing rural conditions. With respect to the potential of traditional 

agroforestry systems such as homegardens, the focus in assessing social sustainability 

should not only be on the question of whether the system fits into the 

traditional farming and livelihood systems, but also on the question of whether these 

agroforestry systems can be adjusted to modern rural conditions while still maintaining 

their features of ecological sustainability. Our study shows that research based on 

detailed assessments of the actual dynamics in the features of traditional agroforestry 

systems is rewarding. Such studies may indicate that different developments 

trajectories are being followed. The understanding of these development pathways and 

the factors involved offers good scope for the identification of options for further 

modification of agroforestry systems. 


The fieldwork for this research was accommodated by the Integrated Rural 

Technology Center, Mundur, Palakkad, Kerala. Dr. Unnikrishnan provided valuable 

background information and important assistance to the organization of the survey. 

Additional information on the status of homegardens was provided by Dr. B. Mohan 

Kumar, Kerala Agricultural University, Dr. S. Sankar, Kerala Forest Research 

Institute and M/s. K.P. Ouseph and M. Shetty of the Kerala Forest Department. 

ENDNOTES 

1. The trends in homegarden structure and composition in the research area were 

discussed in an International workshop on agroforestry and natural resource 

management, organized in 2002 by the Centre for Rural Development and 

Appropriate Technology, Cochin University of Science and Technology, in 

association with the Integrated Rural Technology Center, Mundur, Palakkad, 

Kerala, India. 

2. Wiersum K.F. and Slingerland M. 1996. Use and management of two 

multipurpose tree species (Parkia biglobosa and Detarium microcarpum) in 

agrisilvopastoral land use systems in Burkina Faso. Wageningen Agricultural 

University, Antenne sahelienne. Document de projet No. 41. 

REFERENCES 

Ashley C. and Maxwell S. 2001. Rethinking rural development. Dev Pol Rev 19: 395 – 425. 

FAO. 1977. FAO-UNESCO Soil Map of the World 1:5000000. Volume VII, South Asia. 

UNESCO, Paris. 

Huston M.A. 1994. Biological diversity. The coexistence of species on changing landscapes. 

Cambridge University Press. 681p. 

John J. and Nair M.A. 1999. Socio-economic characteristics of homestead farming in south 

Kerala. J Trop Agric 37: 107 – 109.


Jose D. and Shanmugaratnam N. 1993. Traditional homegardens of Kerala, a sustainable 

human ecosystem. Agroforest Syst 24: 203 –213. 


135 – 152. 


wood in the homegardens of Kerala in peninsular India. Agroforest Syst 25: 243 –262. 

Leiva J.M., Azurdia C., Ovanda W., Lopez E. and Ayala H. 2002. Contributions of 

homegardens to in situ conservation in traditional farming systems – Guatemalan 

component. In: Watson J.W. and Eyzaguirre P.B. (eds), Homegardens and in situ 

conservation of plant genetic resources in farming systems. Proceedings of the Second 

International homegarden workshop, Witzenhausen, Germany, pp 56 – 72. International 

Plant Genetic Resources Institute, Rome. 

Méndez V.E., Lok R., and Somarriba E. 2001. Interdisciplinairy analysis of homegardens in 


85 – 96. 





Quiroz C., Gutierrez M., Rodriguez D., Perez D., Ynfante J., Gamez J., Perez de Fernandez 

T., Marques A. and Pacheco W. 2002. Homegardens and in situ conservation of agrobiodiversity 

– Venezuelan component. In: Watson J.W. and Eyzaguirre P.B. (eds), 

Homegardens and in situ conservation of plant genetic resources in farming systems. 

Proceedings of the Second International homegarden workshop, Witzenhausen, Germany, 

pp 73 – 82. International Plant Genetic Resources Institute, Rome. 

Soemarwoto O., 1987. Homegardens: a traditional agroforestry system with a promising 

future. In: Steppler H.A. and Nair P.K.R. (eds), Agroforestry, a decade of development, 


Tesfaye Abebe 2005. Diversity in homegarden agroforestry systems of southern Ethiopia. 

Wageningen University, the Netherlands, Tropical Resource Management Paper No. 59, 

143p. 


Environ 41: 189 – 207. 

Wiersum K.F. 1995. 200 years of sustainability in forestry: lessons from history. Environ 

Manage 19: 321 – 329. 

Wiersum K.F. 1997. Indigenous exploitation and management of tropical forest resources: an 

evolutionary continuum in forest – people interactions. Agric Ecosyst Environ 63: 1 – 16. 




CHAPTER 7 

STRUCTURE AND DYNAMICS 

OF COCONUT-BASED AGROFORESTRY 

SYSTEMS IN MELANESIA: 

A CASE STUDY FROM 

THE VANUATU ARCHIPELAGO 

N. LAMANDA 1 *, E. MALÉZIEUX 1 , AND P. MARTIN 2 

1 CIRAD UMR SYSTEM, Agro M, Bâtiment 27, 2 place viala, 34060 Montpellier, 

France; *E-mail: . 2 INA P-G département AGER, 

bâtiment EGER BP 01, 78850 Thiverval-Grignon, France 

Keywords: Cocos nucifera, Cropping system dynamics, Melanesian agriculture, Vegetation 

structure analysis. 

Abstract. Coconut (Cocos nucifera)-based agroforestry systems hold promise as a sustainable 

land use activity in the Melanesian islands, where food dependency on foreign sources and 

land shortages are increasing dramatically. This chapter describes the dynamics of these 

smallholder production systems in the Malo Island of northern Vanuatu (Melanesia), where a 

dual economy operates in which resources are dedicated to both subsistence and commercial 

production. The floristic elements found in the coconut plantations were typical of those 

described in the humid tropical homegardens elsewhere, with an average of 12 tree species 

per plot. Mean Shannon Weaver index was 1.57 with the vertical profile of vegetation having 

one-to-five strata. Although the coconut palms dominate these production systems, in certain 

cases other trees may dominate it. Situations in coconut plots evolve throughout the 

development phase of the palms. Based on that, five types of smallholder coconut-based 

agroforestry systems were recognized, which falls into two main evolutionary patterns: (1) a 

perennial occupation of the cultivated land by coconut trees, because of coconut replanting, 

and (2) a gradual return to tree fallow in which the coconut palms gradually disappear 

because of changes in the complex multistrata vegetation. 


In the Melanesian archipelago of Vanuatu, about 80% of the estimated 0.2 million 

population lives in rural areas and are involved in agriculture (Labouisse, 2004). The 

105 




106 N. LLAMANDA 

ET AL. 

traditional farming systems are shifting cultivation with long fallows (food gardens) 

and cultivation of coconut palms (Cocos nucifera) with a mixture of other species. 

The coconut palm incidentally is known as the “tree of life” in the Pacific islands 

because of its multiple uses. The staple food crops in the multistory food gardens 

include root and tuber crops, for example, yam (Dioscorea spp.) and taro (Colocasia 

esculenta), often closely associated with other species such as banana (Musa spp.), 

island cabbage ( (Abelmoschus spp.), or cassava (Manihot esculenta) and numerous 

tree species ( (Artocarpus altilis, 

Barringtonia edulis, etc.). Some farmers also 

undertake pig breeding, mainly due to social considerations (Bonnemaison, 1996). 

During the 20th century, however, development of large “coconut estates” by 

the Europeans became a dominant land use activity and was rapidly followed by the 

evolution of a large number of smallholder plantations that substantially altered the 

indigenous farming systems in Vanuatu (Barrau, 1955; Clarke and Thaman, 1993; 

Bonnemaison, 1996). In particular, such coconut plantations became dominant in the 

northern islands of the archipelago where the agroclimatic conditions and market 

opportunities were ideally suited for coconut production. Furthermore, over a period 

of time, the intercropped, smallholder food gardens, with young coconut trees 

planted after bush or tree fallow clearing (Weightman, 1989) have evolved into 

complex farming systems in which coconut is associated with numerous other 

species and/or cattle grazing. Because of the development of coconut plantations, 

often on the best agricultural lands, food gardens were pushed farther from villages 

and onto the marginal lands (Clarke and Thaman, 1993; Bonnemaison, 1996). 

Concomitantly, forests or old tree-fallows were also converted into gardens and 

coconut plantations. 

At present about 60% of the cultivated area in Vanuatu is occupied by coconut 

plantations and copra production is still the major source of income for the northern 

Vanuatu’s rural population, despite the downward trend in copra prices worldwide 

during the past decade (Labouisse, 2004). With an increasing population that may 

double over the next 30 years, food dependency on external sources and pressure on 

natural resources may increase, and farmers and agricultural extension services are 

expressing concern about how to improve the current cropping systems. As a result, 

a diagnosis of the performance and sustainability of the existing situations in the 

coconut farms is needed in order to manage and prepare for the intensification of 

these systems. Moreover, the existing situations are the result of mixed species 

vegetation developing in plots managed with farmers’ practices, and are, in turn 

dependent on local agro-ecological conditions. This chapter analyzes the dynamics 

of these complex coconut-based land use systems and the evolution of its vegetation 

structure over time. 


2.1. Study area and selection of sample plots 

The study site was situated on Malo Island (15°40’S, 167°10’E) in northern Vanuatu 

that has a dual economy, in which resources are dedicated to both subsistence and

COCONUT-BASED AGROFORESTRY IN MELANESIA M 

107 

commercial production (Allen, 2000). Malo Island, which covers about 180 km², is 

located southeast of the Santo Island, 20 km from Luganville (Santo Island), the 

second-ranking urban center of the archipelago (Fig. 1). This island, with its highly 

fertile soils and an equatorial climate tempered by oceanic influence (Quantin, 

1982), offers optimum conditions for coconut cultivation (IRHO, 1969). Coincidentally, 

copra production is the main source of income to the inhabitants. 

Figure 1. Location of the study area, Malo Island, Vanuatu. 

The experimental approach involved characterizing the structure and dynamics 

of the smallholder coconut-based agroforestry systems. It consisted of four steps: (1) 

selection of sample plots representing the diverse situations existing in the study

108 

area; (2) description of the selected plots that included a farmer survey and plot 

observations; (3) classification and grouping of the monitored plots according to 

their vegetation structure; and (4) correlation studies to gain insights into the 

temporal dynamics of the cropping systems. 

Malo island was first stratified according to its biophysical and socioeconomic 

characteristics (soil type, climate, and market opportunity) using a participatory 

mapping technique (Caron, 1997). The bibliographical data (Quantin, 1982; 

Weightman, 1989; Allen, 2000) were supplemented with historical data gathered 

through interviews of the local chiefs, farmers, and extension officers, chosen for 

their knowledge of Malo and its agriculture. Soil and climate characteristics, 

population density and origin, roads and marketing areas were also mapped on the 

basis of these bibliographical and interview data. 

By linking the historical aspects and the biophysical characteristics of Malo 

(Lamanda et al., 2004), two distinct production areas were identified: the western 

area and the central eastern area. A native population and heterogeneous biophysical 

conditions characterize the western area. The predominant soil types are coral 

limestone in the coastal region and clay soils on the hills. In this area, the 

households are nested within villages and the farm fields of each family are 

scattered over large areas (i.e., the fields are often more than half an hour’s walk 

from the household). Infrastructures are also concentrated in the villages, including 

the main dispensary, markets and roads and a rapid sea-link with the second-ranking 

town of the archipelago (Luganville) where most of the harvested produce are 

marketed. The central eastern area is characterized by a population essentially 

composed of migrants from other islands, and with more homogeneous soil 

characteristics (clay soils with good agronomic potential). The habitat is scattered 

and the coconut plots are located around the houses. Subsistence food gardens are 

also cultivated, sometimes far from the household depending on land availability. 

The population density is lower than that in the western area with estimates 1 of 250 

inhabitants per km 2 in Avunatari, the main village of the western area, and 15 

inhabitants per km² in the central eastern area during 1997. 

In the main villages of each production area, farmers were interviewed in order 

to select the sample plots. The objective was to produce a sample that captures the 

diversity of situations existing in the coconut plots of Malo Island (i.e., different 

stages of development of coconut and/or different intercropping situations). We 

defined a plott in this study by evaluating the vegetation structure and stage of 

development of the coconut palms; accordingly, two or more plots could be 

distinguished within a farmer’s field. The fieldwork was carried out in 2002 – ‘03 

when 191 coconut plots were sampled and described. These plots (116 plots in the 

western area and 75 in the central eastern area) represented different stages of 

coconut development over an 85-year period and involved different species 

associations. 

2.2. Description of the selected plots 

N. LLAMANDA 

ET AL. 

A farmers’ survey was conducted in each selected plot in order to assess the 

management history of the plots from coconut pre-planting to the present. The


109 

survey covered aspects such as when and how the plantation was established 

(biological material, original planting patterns, how fallow was destroyed, or used: 

e.g., forest or tree fallow), what are the changes in associated vegetation types, e.g., 

presence of food gardens or cacao (Theobroma cacao) trees, when coconut palms 

began to produce, present management of the plots, and its evolution through time. 

The plots were located and their area was calculated using a geographical 

positioning system (GPS) when the canopy provided a clear signal; otherwise, direct 

measurements were used. All tree and crop species present in each plot were 

identified and grouped according to their nature and uses. Density was calculated by 

counting the individuals (i) on three sub-plots of size 900 m 2 each for species 

regularly planted such as coconut and cacao, and (ii) on the total plot area for other 

species. Species richness (number of species) and the Shannon Weaver index of 

species diversity (Krebs, 1985) were calculated collectively and separately for trees 

and other species. 

Horizontal distribution of species was assessed visually and the planting pattern 

of coconut and cacao evaluated using a 0 (no visible planting pattern) to +++ 

(systematic pattern) scale. The distance between rows and trees in a row was 

measured for 10 coconut and cacao trees each in order to assess the planting pattern. 

Location of the large trees was determined using GPS. The vertical differentiation of 

the vegetation profile was first assessed visually and then supplemented by height 

measurements obtained for all tree species with individuals taller than 1.5 m. 

2.3. Classification of the monitored plots 

Structural groups (or vegetation types) were constructed based on the responses to a 

hierarchical set of questions concerning vegetation structure in the plot: (1) major 

species in the plot, (2) extent of species diversity, (3) horizontal distribution of 

major species and species groups, (4) vertical differentiation of the vegetation 

(canopy) profile, (5) status of the major species in the vegetation profile, and (6) 

dominant species or species groups. Many of these structural parameters, however, 

showed a continuous gradation; consequently, the structural groups identified in the 

classification scheme also constituted a continuum. 

2.4. Dynamics of coconut-based agroforestry systems 

A matrix of ‘cropping situations’ that combined the structural groups vs. time was 

constructed in order to position the situations described in the coconut plots in a 

temporal scheme. Time was represented by the development stages of coconut trees. 

Based on literature reference and the information provided by the farmers, four 

distinct stages of development of the coconut palms were recognized: (1) juvenile 

stage (0 to 7 years), (2) low productive stage (8 to 15 years), (3) productive stage (16 

to about 60 years), and (4) senescent stage (over 60 years). 

All monitored plots were then positioned in the matrix according to the structural 

groups and the development stages of the palms, which was crosschecked through 

farmer survey, especially about the date of plantation. The cropping situations were

110 

then linked to the management histories of the plots, and expressed as a succession 

of cropping situations during coconut development representing its temporal dynamics. 

A ‘cropping situation’ defined by the intersection in the matrix of a structural group 

and a development stage could be represented by different plots. 

3.1. Vegetation structure of coconut plots 

N. LLAMANDA 

ET AL. 

3. RESULTS 

Vegetation characteristics of the experimental area are presented in Table 1. The 

major tree species found are coconut and cacao. Mean density of coconut palms in 

the smallholder plantations was 148 trees per ha, which is close to the 143 trees per 

ha recommended by local extension services. Density ranged from 11 to 744 palms 

per ha, and higher densities were common especially when two generations of 

coconut palms coexisted in the same plot. Mean density for cacao was 209 trees per 

ha, with a maximum of 1053, which was indeed less than the density recommended 

by the extension services (1111 cacao trees intercropped with 143 coconuts per ha). 

Mean size of the coconut plots was 1 ha (range 0.01 to 4 ha) and the smallest plots 

were mainly the food gardens associated with juvenile coconut palms. 

Table 1. Vegetation characteristics of coconut plots on Malo Island (Vanuatu). 

Vegetation characteristics Mean Min Max 

Number of species 15 4 40 

Number of tree species 12 0 28 

Number of semi-perennial herbs 3 0 12 

Shannon Weaver index (total) 1.57 0.14 2.81 

Shannon Weaver index (trees) 1.5 0 2.88 

Shannon Weaver index (semi-perennial herbs) 0.61 0 1.94 

Coconut planting density (all generations; no. per ha) 164 11 744 

Coconut planting density (number per ha) 148 0 457 

Number of tree species per ha 

(coconut and cocoa not included) 

223 3 2733 

Ratio of coconut palms-to-total tree species 0.5 0.05 0.99 

Cacao tree planting density (number per ha) 209 0 1053 

Age of first generation coconut palms (years) 35 planting 84 

Coconut field area (ha) 1 0.01 4 

In addition to the two main species mentioned, 90 other useful species were 

identified in the monitored plots (a list of conspicuous species with their local names 

and uses are given in Appendix I). According to their habit and uses, the species 

were grouped as ‘trees’ (49 species mainly with fruit trees such as Mangifera indica


or Barringtonia edulis and/or timber trees such as Hibiscus tiliaceus or Pometia 

pinnata) and ‘semi-perennial food crop species’ (41 species). The mean number of 

cultivated species per plot was 16, with a minimum of four and a maximum of 40 

species (Table 1). The most represented species were fruit trees such as Artocarpus 

altilis, B. edulis s and M. indica, found in 68, 60, and 55% of the monitored plots 

respectively. Semi-perennial food crops such as Musa spp., Carica papaya, Xanthosoma 

sagittifolium, and Dioscorea nummularia were also frequently intercropped along 

with coconuts (in about 40% of the monitored plots). 

The number of trees intercropped with coconut showed a wide range (3 to 2733 

trees per ha); consequently, the proportion of f coconut palms relative to other tree 

species was highly variable. Indeed, the ratio of the coconut tree density to the 

density of other tree species ranged from 0.05 to 0.99 (with a mean of 0.5), which 

illustrates the high floristic diversity of the coconut plots (Table 1). Consistent with 

this, the Shannon Weaver index collectively for tree and crop species ranged from 

0.14 to 2.81 (mean =1.57). Furthermore, species diversity was higher for trees than 

for the semi-perennial food crops (mean Shannon Weaver index of 1.50 for trees and 

0.60 for food crops). The extent of mixing food crops also varied substantially, with 

situations ranging from plots cropped with juvenile coconuts along with mixed food 

gardens, to plots in which only a few taro or banana plants were grown. 

3.2. Horizontal and vertical l structure of vegetation 

111 

Theoretically, the horizontal distribution of a species could be ‘systematic’ (with a 

repeated pattern), ‘distorted’ (if altered by the death or cutting down of certain 

individuals in the systematic pattern), ‘random’ (without any definite pattern), or 

‘patchy’ (presence of groups or clusters; Fig. 2). Our observations indicate that the 

horizontal distribution of coconut and cacao trees in the sample plots was mostly 

systematic or distorted. This is because the coconut trees were mostly planted in a 

square pattern with a mean distance of 7.7 m between trees (CV = 14.5%). Likewise, 

the cacao trees were interplanted between the coconut rows with a mean distance of 

5.3 m between cacao trees (extension recommendations are 9 m for coconuts and 3 

m for intercropped cacao). The horizontal distribution of other trees was mostly 

random, without any clear geometrical arrangement. Semi-perennial food crop 

species mainly had a patchy distribution pattern either on the boundaries or between 

tree species depending on species and their cultivation requirements. For instance, 

strong yam (Dioscorea nummularia) was often found close to large trees that 

provided supports for the vines, while banana plants were clustered in pure or mixed 

stands. 

As regards to vertical organization of the components, a multistrata arrangement 

with one-to-five strata depending on the number and architecture of the tree species 

was discernible. Coconut palms often formed the dominant component, but were 

sometimes dominated, especially when they were young and/or before the fallow 

clearing. Complex situations in which coconuts were both dominant and dominated 

were also noted when forest trees (dominant) were combined with food crops 

(dominated). Natural regeneration of tree/shrub species constituted the lower strata 

along with the food crop species. When cattle grazing under coconut and fruit trees

112 

was practiced, the naturally regenerating tree/shrub species of interest were 

protected by the farmers. 

The horizontal distribution of a species could be : 

Regular 

Distorted 

Randomised 

Patchy 

Horizontal distributions of each major species and species group 

are overlapped to obtain a map of the horizontal vegetation in the plot 

Figure 2. Different types of horizontal vegetation distribution on Malo Island, Vanuatu. 

3.3. Structural groups 

N. LLAMANDA 

ET AL. 

The monitored plots were finally classified into 14 structural groups (Fig. 3) and 

their principal attributes are summarized in Table 2. Similar structural groups were 

observed both in the western and central eastern areas and for the sole coconut and 

coconut+cacao systems. However, four structural groups not described on Malo but 

existing on other Vanuatu islands were also included in the classification scheme for 

the sake of comprehensiveness; these include the coconut estates where the palm 

was cultivated as a single species, possibly intercropped with cacao or associated 

with cattle grazing (I0A I , I0B I , II0A I , II0B I ; Fig. 3). In contrast to these, the smallholder 

plots were generally characterized by significant species diversity. 

3.4. Temporal dynamics of coconut-based cropping systems 

Five coconut-based agroforestry systems were identified and the dynamics of their 

vegetation structure during coconut development phases were reconstructed (Fig. 4). 

Different structural groups can be noted for a given stage in the coconut cycle. 

Productive coconut plots could also be classified into several structural groups, thus 

illustrating the profound variability in vegetation structure and farmers’ practices 

associated with coconut production in Malo Island (Fig. 5). 

The evolutionary pathway starts as a system involving fruit trees, coconut palms, 

and food gardens planted together in a tree fallow that has been selectively managed.

1 

2 

3 

4 

5 

6 

Coconut trees tr 

1 

2 

Coconut and d cocoa c 

trees 


3 

no vertical 

differentiation 

4 5 6 

regular ular 

no vertical 


4 5 6 

altered 

no specific cific 

diversity ersit 

significant significan specific 

diversity er 

regular ular 

5 

4 5 

vertical 


food foood 

crop s 

6 

youn you g 

coconuts coconut 

vertical 

differentiation on 

4 

vertical v 

dominant 

5 6 

cocoas 


diff 

regular ar 

5 

3 

4 

altered 

dominant 

no specific peci 

diversity divers 

6 

cocoas 

2 

5 6 

significant specific pe 

diversity 

no vertical al 

differentiation atio 

4 

dominant 

5 

cocoas 

6 

vertical ver 

Coconut trees 

differentiation ren 

regular gul 

Fruit trees 

Food crops 

Coconut trees 

replanting 

Cocoa trees 

What are the major species in the plot ? 

What species diversity is there ? 

What is the horizontal distribution of 

the major species and of the species group ? 

What is the vertical differentiation of the 

vegetation profile ? 

What is the status of the major tree species 

in the vertical profile ? 

What are the dominated species 

or species groups ? 

3 

4 

5 

no vertical tica 

differentiation enti 

no vertical tica 

differentiation ntia 

altered ered 

dominant nant 

dominated omina 

6 

6 

6 

dominant and a 

dominated ed 

cocoas ocoas 

altered ered 

no vertical 

cocoas coc and 


d 

6 

food crops 

dominant dom 

4 5 

6 

dominated omin 

food crops and 

young coconuts uts 

6 

dominant ant and 

dominated ed 

6 

Figure 3. Classification of coconut plots for Malo (Vanuatu) in structural groups. 

I0A 

I0B 

I2 

I1A 

I1B 

I1C 

I4A 

I4B 

I3A 

I3B 

II 0A 

II 0B 

II 2 

II 1A 

II 1B 

II 1C 

3A 

II 3B 

Illustrat ions C irad: N . Lam anda, M . D uport al 

113

Structural 

groups 

Monitored 

plots (no.) 

Table 2. Characteristics of structural groups for Malo Island (Vanuatu). 

Number of tree 

species per plot 

Number of semiperennial 

crop 

species per plot 

Structural attributes of vegetation 

Shannon 

Weaver index 

Coconut tree 

density 

(no. per ha) 

Cacao tree 

density 

(no. per ha) 

Tree density other 

than coconut and 

cocoa (no. per ha) 

mean CV% mean CV% mean CV% mean CV% mean CV% mean CV% 

I1A 22 13 33 1 251 1.13 48.8 165 29 0 469 81 163 

A 

I1B 41 12 29 0 - 1.35 42.6 126 30 1 384 120 134 

I1C 19 15 27 5 45 2.01 23.3 107 26 5 436 244 99 

I2 15 9 57 5 78 1.66 24.1 218 50 11 190 129 54 

I3A 17 9 51 7 42 1.58 26.6 210 57 3 141 206 84 

A 

I3B 10 7 23 1 218 1.36 33.5 332 56 0 0 1017 77 

I4A 4 13 13 2 141 1.31 39.6 240 100 0 0 108 60 

A 

I4B 4 14 38 8 29 1.86 27.7 198 81 1 200 60 76 

II1A 6 15 40 2 92 1.44 46.3 144 26 280 87 167 165 

A 

II1B 5 17 34 1 100 1.49 23.1 150 29 128 49 51 35 

II1C 12 17 30 5 64 2.03 22.9 134 32 249 90 172 55 

II2 8 15 45 3 77 1.62 39.1 197 33 125 187 198 52 

II3A 14 11 38 6 47 1.74 29.4 131 63 332 88 341 103 

A 

II3B 11 14 42 5 47 1.92 21.6 150 56 173 97 366 91 

114 

N. LLAMANDA 

ET AL.

115 

As the palms attain the bearing stage, cattle is introduced into the plantation and 

grazed until the coconut trees become senescent (cropping system I). I With the 

coconut palms becoming still older (~60 years) and that their planting pattern gets 

distorted, food crops are introduced and/or a new generation of coconut trees 

interplanted along with the first generation palms. Cacao might also be intercropped 

along with coconuts at the beginning. However, as the cacao trees die eventually, 

cattle might be introduced into the plantation, and this system (System III) I then 

evolves like the previous one; alternately food crops can be inter-planted (System 

IV). V 

When the tree fallows are not managed, however, the coconut palms would be 

dominated by other woody perennial components (systems III and V where cacao 

was intercropped with coconut). ‘Key situations’ of these systems, corresponding to 

the trajectories for the juvenile coconut stage are I3A I , I2 I , II3A I and II2 I (Fig. 4). 

Moreover, at the beginning of the senescent stage if gaps arise between coconut 

trees through altered planting patterns, it could be utilized in different ways (I1B I and 

I ). 

II1B 

Cropping system II 

Cropping system I 

Cropping system III 

Cropping system IV 

Cropping system V 


I3A I2 

I1A I1B 

II3A II2 

II1A II1B 

0 5 20 60 100 

I3B 

I1C 

I4A 

I4B 

II1C 

II3B 

Coconut trees 

Fruit trees 

Food crops 

Coconut trees 

replanting 

Cocoa trees 

years after coconut 

tree planting 

Illustrations Cirad: N. Lamanda, M. Duportal 

Figure 4. Temporal dynamics of coconut-based agroforestry systems on Malo Island, Vanuatu.

116 

N. LLAMANDA 

ET AL. 

4. DISCUSSION 

The coconut plots on Malo Island had species diversity levels close to those noted in 

the multistrata agroforestry systems of the humid tropics, i.e., an average number of 

12 tree species per plot with a mean Shannon Weaver index of 1.57 (for a complete 

review of the reported floristic elements in homegardens see Kumar and Nair, 2004). 

Figure 5. Some structural groups described on Malo Island (Vanuatu) [I I 1A: 

Systematic 

planting of coconut trees (++) and no vertical differentiation of the vegetation profile; 

mixtures of coconut palms (Cocos nucifera), Mangifera indica a and Hibiscus tilaceus can, 

however, be seen in the rear end. I 3B: 

Coconut plots with significant species diversity. 

Distorted planning pattern and a vertical differentiation of the vegetation profile where 

coconut trees are dominant (e.g., a first generation of coconut trees dominates a younger 

generation and Annonaa spp. at the first level of the picture) and dominated (mainly by forest 

trees). I 1C: 

Coconut trees with a significant species diversity. Distorted planning pattern and 

a vertical differentiation of the vegetation profile. Food crops such as Musaa spp. constitute 

the lower stratum of the vegetation profile]. 

A vertical stratification involving one to five strata is also characteristic of the 

structure of homegardens, one of the most common agroforestry systems in the 

tropics (Fernandes and Nair, 1986; Kumar and Nair, 2004). The vegetation structure


reported for the smallholder coconut plots on Malo Island is similar to that described 

for many smallholder coconut production systems in other locations, e.g., cattle 

grazing in the Pacific islands (Nair, 1983; Clarke and Thaman, 1993), multistory 

mixed species systems involving coconut and another cash crop such as cacao, or 

with food crops in South Asia, and in particular in India (for a comprehensive 

review of the agroforestry systems with coconut, see Nair, 1979; 1983; 1989). 

The situation of coconut plots was, however, not static and evolved throughout 

its development phases. We discerned five dynamic phases for the smallholder 

coconut-based agroforestry systems that corresponded to five major trajectories (Fig. 

4) from which two evolutionary patterns could be deduced. First, a perennial 

occupation of the cultivated land by the coconut palms, because of coconut 

replanting in cropping systems I, IIII and IV. V In these systems, the tree fallows 

gradually evolved into situations with one-to-three strata and a new coconut stand 

could be established in the original pattern after about 60 years of coconut 

cultivation. Second, a gradual return to tree fallows where coconut trees could 

gradually disappear because of the evolution of complex multistrata vegetation in 

which other tree species dominate (coconut-based agroforestry systems III and V). V 

This pattern of evolution could lead to a new cultivation cycle depending on the 

agroecological impacts (especially on soil fertility) and fallow duration (e.g., the 

food gardens). 

The smallholder coconut-based agroforestry systems have various economic/ 

social functions too: (1) generating cash flow by copra and/or cocoa production, 

which incidentally, is the main source of income for most Maloese, (2) contribution 

to food security by producing fruits, nuts, leaves, roots, etc., that are a substantial 

source of food supply – and some of which have high nutritional value, thus adding 

to the nutritional security, (3) an inheritance pattern with plantations being passed 

down to the children, (4) a social function with copra harvesting by a working group 

called ‘kompagny”, and (5) a cultural role with the production of decorative, 

medicinal/‘magic’ species. Cattle grazing in coconut plots is also associated with (1) 

generation of cash income, (2) food production that constituted an important source 

of animal proteins, (3) a social function with cattle slaughtered for marriage and 

funeral ceremonies, (4) weed control and (5) a nutrient recycling function, e.g., 

grazing and nutrient addition through dung and urine. 

In addition, the coconut-based agroforestry systems provide for ecological 

functions such as carbon sequestration (see Kumar, 2006); efforts to quantify this 

substantial potential of the plantations of Vanuatu are currently underway. Yet 

another advantage is in situ agrobiodiversity conservation, especially the high intraspecific 

variability and genetic diversity at plot level. For instance, seven named 

types of coconut trees were reported per ha of the smallholder plantations in Vanua 

Lava, another northern island of Vanuatu (Caillon, 2005). And a wide variety of 

breadfruit (Artocarpus ( altilis) 

is also cultivated in Vanuatu (~132 types; Walter, 

1989). 

Cultivating food crops in coconut gardens also might reduce the impact on tree 

fallows or forests by reducing the rate at which these are being cleared for food 

production. Furthermore, it represents a sustainable way of intensifying the current 

cropping practices. That is, with only 25% of the light and space being used by the 

117

118 

N. LLAMANDA 

ET AL. 

coconut palms, resources are often under-exploited in mature coconut plantations. 

Intercropping might be the best option for effectively utilizing these resources (Nair, 

1979; 1983). In the current context of Melanesian agriculture (with land shortages 

and a downward trend in the profitability of copra production), agroforestry, thus, 

appears to be a very attractive option for intensification of the smallholder coconut 

production systems, and in particular, the old plantations. Nowadays, in the western 

area of Malo, where there are land shortages due to human pressure, food crops such 

as banana, papaya, island cabbage, or strong yam are already being introduced on 

the farm boundaries, or in the distorted planting patterns of coconut trees in the older 

(~60 year-old) plantations (I1C I C in Figs. 3, 4 and 5). This situation, which is found at 

the senescent stage of coconut palms and in areas with high human population 

pressure, may constitute an innovation and a valid alternative to current land use 

problems. 

More intensive use of the older coconut plantations is possible by intercropping 

food crops or species with high economic value such as vanilla (Vanilla planifolia), 

that are adapted to the level of resources usually available under the canopy of the 

coconut palms. These species can provide a significant source of income or food. 

Moreover, food crops can be sold in the local markets that offer considerable 

potential for development and expansion due to the increase in urban populations. In 

isolated areas such as Melanesian islands, however, developing the production of 

species with high economic potential should be linked to niche-marketing 

opportunities and extension facilities to certify the quality attributes/organic origin 

of the produce. 

Existing smallholder copra production systems are also more complex than the 

large European coconut estates. Yet the development of copra production has led 

to a simplification of the pre-existing smallholder systems, a phenomenon called 

agrodeforestation (Clarke and Thaman, 1993), which had dramatic consequences on 

many Pacific islands. Therefore, it should be accompanied by another process of 

agroforestation, to avoid environmental disasters. 

6. FUTURE DIRECTIONS 

Characterizing the existing smallholder coconut-based agroforestry systems and 

their dynamics constitutes the first step towards evaluating their agro-ecological and 

agro-economical potentials, which is required to guide the future of these systems. 

Key situations, such as those where food crops are reintroduced into the coconut 

plots, are currently being studied to assess the possibility of more intensive use of 

the old coconut plantations. In particular, soil fertility levels (with organic matter 

indicators) and light availability in the vegetation profile and root occupation are 

being measured to estimate the degree to which various biophysical resources are 

used. Future studies should also take into account differences in soil fertility due to 

topographic differences (coral limestone in coastal area versus clay soils on hills), 

and economic evaluation of the coconut-based agroforestry systems to ensure that 

they match the farmers’ goals.



119 

We are grateful to the Vanuatu Research and Training Centre (VARTC) and the 

Vanuatu Agriculture Extension Services and the Farming Support Association (FSA) 

for their logistical support and cooperation. We would also like to express our 

gratitude to the Maloese farmers for their warm welcome and infinite patience. 

ENDNOTE 

1. Allen M.G. 2001. Change and continuity: Land use and agriculture on Malo 

Island, Vanuatu. Master of science. Australian National University, Canberra, 

201p. 

REFERENCES 

Allen M. 2000. Subsistence or Cash Cropping ? Food Security on Malo Island, Vanuatu. In: 

Bourke R., Salisbury M.G., and Allen J.C. (eds), Food security for Papua New Guinea - 

Proceedings of Papua New Guinea ffood 

and nutrition 2000 Conference, pp 100 – 111. 

ACIAR Proceeding. PNG University of Technology, Canberra. 

Barrau J. 1955. L’agriculture vivrière mélanésienne. Commission du Pacifique Sud. Nouméa, 

206p. 

Bonnemaison J. 1996. Gens de Pirogues et Gens de la Terre- Les fondements géographiques 

d’une identité, l’archipel du Vanuatu- Livre 1. Orstom (eds). Paris, 460p. 

Caillon S. 2005. Pour une conservation dynamique de l’agrobiodiversité: gestion locale de la 

diversité variétale d’un arbre “des Blancs” (cocotier, Cocos nucifera L.) et d’une plante 

“des ancêtres” (taro, Colocasia esculenta (L.) Schott) au Vanuatu. Doctorat, Géographie, 

Université d’Orléans, Orléans, 789p. 

Caron P. 1997. Le zonage régional à dire d’acteurs. Connaître, représenter, planifier, agir, une 

méthodologie expérimentée dans le Nordeste du Brésil. In: Quelle géographie au CIRAD? 

Montpellier 10: 145 – 156. 

Clarke W.C. and Thaman R.R. (eds). 1993. Agroforestry in the Pacific Islands: systems for 

sustainability. United Nation University, Tokyo, 297p. 

Fernandes E.C.M. and Nair P.K.R. 1986. An evaluation of the structure and function of 

tropical homegardens. Agroforest Syst 21: 279 – 310. 

Institut de Recherches des Huiles (IRHO) 1969. Accroissement de la production de coprah 

aux Nouvelles Hébrides. Institut de Recherches des Huiles et Oléagineux (IRHO), Paris, 

154p. 

Krebs C.J. 1985. Ecology: the experimental analysis of distribution and abundance. Harper 

and Row Publisher, New York, 686p. 





135 – 152. 

Labouisse J-P. 2004. Systèmes agraires et économie du cocotier au Vanuatu: historique et 

perspectives. Journal de la Société des Océanistes (CNRS-Musée de l’Homme), Paris, 

118: 11 – 33. 

Lamanda N., Malézieux E. et Martin P. 2004. Organisation spatiale et dynamique des 

systèmes de culture à base de cocotiers (Cocos nucifera L.) dans une île mélanésienne. 

Cahiers Agricultures 6: 459 – 467.

120 

N. LLAMANDA 

ET AL. 

Nair P.K.R. 1979. Intensive multiple cropping with coconuts in India. Paul Parey, 

Berlin/Hamburg, 148p. 

Nair P.K.R. 1983. Agroforestry with coconuts and other tropical plantation crops. In: Huxley 

P.A. (ed.), Plant research and agroforestry, pp 79 – 102. ICRAF, Nairobi. 

Nair P.K.R. (ed.). 1989. Agroforestry systems in the tropics. Kluwer, Dordrecht, 664p. 

Quantin P. 1982. Atlas des sols et de quelques données du milieu naturel: Atlas and 

exploratory notes. Orstom (eds), Paris, 37p. 

Walter A. 1989. Notes sur les cultivars d’arbres à pain à Vanuatu. Journal de la Société des 

Océanistes (CNRS-Musée de l’Homme), Paris, 1/ 2: 3 – 18. 

Weightman B. 1989. Agriculture in Vanuatu: An historical review. The British Friends of 

Vanuatu, Cheam, 320p. 

APPENDIX 1 

List of local, scientific names and uses of the conspicuous species reported in 

coconut plots on Malo Island, Vanuatu. 

List of conspicuous species Uses 

Scientific name Local name 

(bishlama) 

Abelmoschus manihot aeland 

kappish 

⊕ ⊕ 

Ananas comosus pineapple ⊕ ⊕ 

Sold Food Timber Fuel Used 

every 

day 

Eaten 

by 

animals 

Magic 

and 

medicinal 

uses 

Annona spp. korrosol ⊕ ⊕ ⊕ ⊕ 

Artocarpus altilis breadfruit ⊕ ⊕ ⊕ ⊕ ⊕ 

Barringtonia edulis navele ⊕ ⊕ ⊕ ⊕ 

Canarium indicum nangaie ⊕ ⊕ ⊕ ⊕ ⊕ 

Carica papaya paw paw ⊕ ⊕ ⊕ ⊕ 

Citrus grandis pomelos ⊕ ⊕ ⊕ 

Citrus limon lemon ⊕ ⊕ ⊕ ⊕ 

Citrus reticulata mandarine ⊕ ⊕ ⊕ 

Dioscorea spp. soft yam ⊕ ⊕ ⊕ 

Discorea nummularia strong yam ⊕ ⊕ 

Erythrina variegata narara ⊕ ⊕ 

Helicona indica leaf lap lap ⊕ ⊕ 

Hibiscus tiliaceus bourrao ⊕ ⊕ ⊕ ⊕ ⊕ ⊕ 

Inocarpus fagiferus namambé ⊕ ⊕ ⊕ 

Macaranga spp. navenue ⊕ ⊕ ⊕ ⊕ 

Mangifera indica mango ⊕ ⊕ ⊕ 

Manihot esculenta manioc ⊕ ⊕ 

Metroxylon warburghii natangora ⊕ ⊕ ⊕ 

Musa spp. banana ⊕ ⊕ ⊕ ⊕ 

Pometia pinnata nandao ⊕ ⊕ ⊕ ⊕ ⊕


Psidium guajava guava ⊕ ⊕ ⊕ 

Saccharum officinarum sugarcane ⊕ ⊕ 

Spondias dulis naos ⊕ ⊕ ⊕ 

Vanilla 

planifolia/tahitensis 

Xanthosoma 

sagittifolium 

vanilla ⊕ 

taro Fiji ⊕ ⊕ ⊕ 

121

CHAPTER 8 

DIVERSITY AND DYNAMICS 

IN HOMEGARDENS OF SOUTHERN 

ETHIOPIA 

TESFAYE ABEBE 1 , K.F. WIERSUM 2 *, F. BONGERS 3 

AND F. STERCK 3 

1 Debub University, Awassa College of Agriculture, Ethiopia. 2 Forest and Nature 

Conservation Policy group, and 3 Forest Ecology and Management group, 

Wageningen University, The Netherlands; *E-mail: 

Keywords: Adaptability, Species composition, Socioeconomic change, Sustainability. 

Abstract. Most homegarden studies have focused on Asia, where homegardens constitute a 

component of a spatially separated farming system consisting of cultivated fields with staple 

food and/or commercial crops away from homes complemented by the homegardens with 

supplementary crops such as fruits and vegetables surrounding residential houses. In the 

highlands of East and Central Africa, another type of homegarden is found in the form of an 

integrated farming system within itself and without additional cultivated fields. In these 

‘integral’ homegardens, not only supplementary crops such as fruits and vegetables, but also 

staple food crops and cash crops are grown. The enset (Enset ventricosum) and coffee (Coffea 

arabica) homegarden system in southern Ethiopia is a typical example of such integral 

homegardens. An assessment of 144 of these homegardens was made to gain insights into 

their structure and vegetation composition and the relation between composition and 

geographic and socioeconomic factors. Four specific garden types are identified, which vary 

in commercial crop composition and diversity. These variations are related to farm size and 

access to roads and markets, and illustrate the dynamic character of homegardens. Overall, the 

diversity of the integral homegarden system seems to be somewhat lower than that of the 

‘complementary’ homegarden systems in Asia, probably due to the inclusion of light 

demanding staple food crops and a relatively large number of commercial crops. The dynamic 

pathways of the integral homegarden systems because of commercialization appear similar to 

reported trends in the ‘complementary’ homegarden systems in Asia. Although the 

composition of the homegardens is influenced by socioeconomic dynamics, overall the 

Ethiopian homegardens can be characterized as being ecologically and socioeconomically 

sustainable. This can be attributed not only to species diversity but also to the presence of two 

keystone species—coffee and enset. 

123 




124 TESFAYE T ABEBE ET AL. 


Homegardens have commonly been characterized as biodiverse and sustainable land 

use systems (Soemarwoto, 1987; Torquebiau, 1992; Kumar and Nair, 2004). 

Recently, it has been acknowledged that this does not mean that the structure and 

composition of homegardens should be assumed as being stable (Kumar and Nair, 

2004). From an ecological point of view, the production processes are not 

necessarily negatively affected by changes in vegetation structure and composition, 

if the nutrient cycling processes, hydrological conditions, and synergetic relations 

are not compromised. From a social point of view, the concept of sustainability 

incorporates the notion of adaptation to social change (Peyre et al., 2006). Similar to 

any land use system, homegardens are faced with constant pressure of change 

brought about by demographic, economic, technological, and social dynamics, and 

they are constantly adapted to changing livelihoods. Several studies in Asia indicate 

that with commercialization, often a gradual change from subsistence to commercial 

crops occurs in homegardens, while the crop diversity decreases (Michon and Mary, 

1994; Kumar and Nair, 2004; Peyre et al., 2006; Abdoellah et al., 2006). 

Most homegardens studies are focused on gardens that constitute a component of 

a spatially separated farming system consisting of cultivated fields away from 

homes complemented by the homegardens surrounding residential houses. In such 

multi-locational farming systems, homegarden production is mostly supplementary 

to the staple food production and mainly focuses on vegetables, fruits, and 

condiments (Wiersum, 2006; Soemarwoto, 1987; Hoogerbrugge and Fresco, 1993). 

The notion of tropical homegardens as components of integrated farming systems, 

which also include cultivated fields for staple food production, prevails in much of 

the homegarden literature. These ‘complementary’ homegardens typically consist of 

small (0.01 to 1 ha) plots around houses with a more or less randomly organized 

cropping pattern. A part of the garden may be devoted to ornamentals or tree crops. 

As these homegardens complement other components of the overall farming system, 

crop diversity and homegarden dynamics are influenced by the nature and 

characteristics of the other components of the overall farming system (Stoler, 1978; 

Karyono, 1990). However, in the highlands of East and Central Africa, a somewhat 

different type of homegardens exists in the form of an integrated farming system 

within itself without additional cultivated fields (Tesfaye Abebe, 2005). These 

‘integral’ homegardens consist of medium-scale (0.4 to ~3 ha) multipurpose farm 

fields around homes that form the principal means of livelihood for the households. 

In these gardens, not only supplementary crops such as vegetables, fruits, 

condiments, and/or medicinal crops, but also staple food crops and cash crops are 

cultivated. The motivating factor for this multiplicity of species is that farmers have 

no or very little additional land devoted to specialized types of production, for 

instance cereals. Consequently, these homegardens function as a total rather than a 

partial farming system. Most of the homegardens in the highlands of East Africa 

belong to this category (Fernandes et al., 1984; Okigbo, 1990; Oduol and Aluma, 

1990; Rugalema et al., 1994; Tesfaye Abebe, 2005). They have been much less 

intensively studied than the ‘complementary’ homegardens of Asia. An interesting 

question is whether the diversity and dynamics as observed in the ‘complementary’

HOMEGARDENS OF SOUTHERN ETHIOPIA 

125 

homegardens are also present in these ‘integral’ homegardens. This question is 

examined in this chapter by analyzing the structure and composition of the 

homegardens of southern Ethiopia as an example. 

2. HOMEGARDENS IN SOUTHERN ETHIOPIA 

The traditional agroforestry homegardens of southern Ethiopia are located at 

altitudes of 1500 to 2300 m above sea level where moisture and temperature 

conditions are favorable for agriculture. These gardens are popularly known as 

‘enset-coffee homegardens’ after the two major perennial crops dominating this 

system (Fig. 1; Tesfaye Abebe, 2005). Enset [Enset ventricosum (Welw.) 

Cheesman], sometimes called false-banana, is a multipurpose crop that provides 

subsistence food for about 10 million people in Ethiopia (Bezuneh and Feleke, 1966; 

Desalegn Rahmato, 1995; Almaz Negash, 2001). Because of the possibility to 

harvest this perennial crop in times of famine, it has been termed as a ‘tree against 

Figure 1. Coffee (Coffea arabica)-enset (Enset ventricosum) m homegarden in southern 

Ethiopia. This ‘integral’ homegarden is not leveled as usual in Asian homegardens (Photo: 

Tesfaye Abebe). 

hunger’ (Brandt et al., 1997). Coffee (Coffea arabica) is also a native crop, which is 

not only grown for household use, but also as a cash crop. Other components of this 

agroforestry system include roots and tubers, fruits, vegetables, cereals, spices, and 

other crops such as the stimulant chat (Chata edulis). Moreover, livestock is kept in 

the gardens and different tree species are grown to serve productive as well as

126 

TESFAYE T ABEBE ET AL. 

ecological functions. Structurally, the gardens resemble the coffee-banana 

agroforestry systems of Uganda (Oduol and Aluma, 1990) and northern Tanzania 

(Fernandes et al., 1984; Rugalema et al., 1994; Soini, 2005) with enset taking the 

position of banana. 

The enset-coffee homegardens have for centuries supported very dense 

populations in the mid-altitude highlands of southern Ethiopia (Kippie Kanshie, 

2002). Although some studies have been made on the system (Westphal, 1975; 

Okigbo, 1990; Tessema Chekun, 1997; Zemede Asfaw and Zerihun Woldu, 1997), 

still only limited information is available about the (variations in) diversity and 

composition as well as the dynamics of the system. For instance, the gardens have 

been mostly described as being predominantly subsistence-based, although the 

presence of coffee and chat is indicative of the fact – and the authors’ experiences 

support this – that these gardens are also used for commercial production. 

3. RESEARCH ON STRUCTURE AND COMPOSITION OF SOUTHERN 

ETHIOPIAN HOMEGARDENS 

In order to assess the structure and composition of the enset-coffee homegardens as 

well as the main factors influencing them, a study was conducted in the Sidama 

Figure 2. Map of Sidama administrative zone (southern Ethiopia) showing the study areas 

with location of selected woredas (or districts) and names of selected Peasant Associations 

within each woreda.

127 

administrative region of southern Ethiopia during 1999 to 2002 (Tesfaye Abebe, 

2005). This region is one of the most densely populated areas of the country with a 

population density of 320 persons km –2 

m . The most important agroecological zone in 

the area is locally known as Gammoje (Sidama) or Woyna-Dega (Amharic). This 

zone is situated between 1500 and 2300 m above sea level, and characterized by a 

moist to subhumid warm subtropical climate with average annual rainfall of 1000 to 

1800 mm, and a mean temperature of 15 to 20 o C. The dominant soils are Eutric 

Nitosols (corresponding to Alfisols in the USDA soil taxonomy). Within this zone, 

detailed data on homegarden composition were collected from 144 homegardens 

located in 12 different Peasant Associations (PA = smallest Ethiopian administrative 

unit) distributed over four Woredas or districts (Fig. 2). The administrative units 

were selected purposefully in order to systematically cover the range of geographic 

conditions in the study area; and, within each administrative unit, the homegardens 

were selected randomly. For each homegarden, data were collected on the size and 

layout and all species (except spontaneously grown weeds) were inventoried and 

enumerated. Through farmer’s interviews, data were also collected on physical 

and socioeconomic characteristics of the farms, such as altitude, distance to markets 

and roads (physical data collection in the field), and on household characteristics 

such as family size, labor force, age and educational status. The interviews also 

served to collect data on the production of various crops and their market prices. 

Tesfaye Abebe (2005) gives further details on that. 

4. STRUCTURE AND COMPOSITION OF HOMEGARDENS IN SIDAMA 

REGION 

4.1. Structure 


The size of the 144 selected homegardens varied from 0.18 to 7.46 ha with a mean 

size of 0.90 ha. These homegarden holdings included residential areas and 

specialized grazing areas (with mean share of 14% of the holding size), cultivated 

lands (mean 82%), and sometimes some specialized woodlots (average 4%; Fig. 3). 

A major variable influencing plot size was the wealth status of the households with 

average values of 0.55, 1.46 and 2.75 ha for poor, middle income, and rich farmers 

(according to the local Peasant Association classification) respectively. Coffee and 

enset dominate in over 50% of the homegarden area, while the other crops occupy 

much smaller areas. Cash crops such as chat, sweet potato (Ipomoea batatas) and 

pineapple (Ananas ( comosus) 

often are grown in special zones. 

Within the homegardens, the vegetation structure was not uniform; often zones 

distinguished by specific crop combinations were found. For instance, zones dominated 

by coffee mixed with fruit and other trees, enset, and miscellaneous auxiliary crops; 

zones dominated by enset mixed with vegetables and miscellaneous trees; zones 

with maize (Zea mays) mixed with other food crops; zones with cash crops such as 

chat; and residential and grazing zones. The diversity of homegardens can, therefore,

128 

not only be assessed based on species composition, but also of the area share of 

main crop components. 

Figure 3. Mean area share of major homegarden components in Sidama administrative zone, 

southern Ethiopia. 

4.2. Species composition 


Overall, 198 species of cultivated crops (78) and trees (120) were recorded from the 

144 homegardens in four woredas. Within each woreda, the total number of plant 

species present in homegardens varied from 84 to 159 (Table 1) demonstrating 

significant intra-regional variations. The mean number of plant species per 

homegarden was 37, with values ranging from 15 to 78. Appendix 1 gives an 

overview of the recorded crop species. In addition to species diversity, a high level 

of genetic diversity was found with respect to the two major crops, enset and coffee, 

being represented by 42 and 24 cultivars respectively. Homegardens also included 

seven livestock species: cattle, goats, sheep, donkeys, horses, mules, and poultry 

(mainly chicken). 

Homegarden composition can also be characterized by the diversity of functional 

crop types. Besides miscellaneous tree species, 10 functional groups of plants were 

recognized: fruit crops (24%), root and tuber crops (16%), vegetables (15%), 

stimulant crops (10%), cereals (9%), pulses (6%), spices and condiments (5%), oil 

crops (3%), medicinal crops (3%), and miscellaneous crops (9%). Each functional 

crop type was represented by 3 to 15 species. Cereals and root/tuber crops provide 

carbohydrate-rich staple foods; fruits, vegetables, pulses, spices/condiments, and 

medicinal crops mostly yield supplementary food and household products; the 

stimulants and oil crops mostly serve as cash crops. Miscellaneous tree species 

provide fuel- and construction wood and serve as shade trees. This combination of

different functions of crops, coupled with the presence of livestock in the 

homegardens, illustrates its character of forming an integrated farming system. 

129 

Table 1. Number of crop, tree, and livestock species in the homegardens of four woredas in 

southern Ethiopia. 

Woreda (n) Crop Tree Total useful Livestock 

species species plant species species 

Dara (36) 56 72 128 5 

Aleta Wondo (48) 64 95 159 6 

Dale (36) 57 94 151 5 

Awassa Zurya (24) 33 51 84 6 

Overall combined 78 120 198 7 

n = number of homegardens sampled. 

4.3. Homegarden types 


In addition to species composition, the extent of area under major crops varied 

significantly among different geographic zones. Four homegarden types could be 

Figure 4. Detrended correspondence analysis (DCA) scatter plots of tree composition of the 

farms. The four homegarden types are indicated with different symbols and indicated by a 

different letter, and are spatially separated. A, Type 1 Enset-coffee-maize type; B, Type 2 

Enset-chat-maize-coffee type; C, Type 3 Enset-coffee-sweet potato type; D, Type 4 Ensetcoffee-maize-chat-pineapple 

type.

130 


identified based on the extent of area under major crops in different PAs (Tesfaye 

Abebe, 2005). This comparative deductive assessment of crop data for the four types 

is compared with the results of a Detrended Correspondence Analysis of tree species 

present in the homegardens. DCA reduces the multidimensional space of a speciesabundance 

matrix into a two-dimensional one. We used DCA as implemented in 

CANOCO 1 . The first axis represents the main variation in species composition, the 

second axis the main variation once the first axis variation is removed. The 

homegarden types are indicated in the DCA graph (Fig. 4). The spatial separation of 

the homegarden types in the graph indicates that the two methods (deductive 

assessment of crop data and DCA of tree species data) resulted in a similar 

categorization into four homegarden types having the following characteristics (see 

Tables 2 and 3): 

1. The enset-coffee-maize type. In a large part of the research area (almost 60% of 

all inventoried homegardens), the homegardens belong to this original type in 

which coffee and enset dominate on about 75% of the farmland. In addition, 

maize is grown on about 10% of the land. As reflected by their high wood 

volume, trees form an important component in the system (Appendix I). Species 

diversity is relatively high with a mean of 41 cultivated crop and tree species. 

These homegardens are predominantly subsistence-oriented with enset and 

maize serving as main staple food crops and coffee serving as a cash crop. The 

overall financial value (based on production amounts and market values) of the 

combined annual yields amount to Birr 5084 ha –1 , which is relatively low (1 

Birr ~ 0.1 US$). 

2. The enset-coffee-maize-sweet potato type is present in 8% of all sampled 

homegardens. It is even more subsistence-oriented than the first homegarden 

type. The share of the staple crop enset is relatively lower than in type 1. 

Instead, farmers produce mainly maize and sweet potato as staple foods. The 

proportion of land devoted to coffee as a cash crop is much lower than in type 1. 

Some farmers are cultivating eucalyptus (Eucalyptus spp.) as an alternative cash 

crop. This homegarden type has the highest species richness (43) of crops and 

trees. The overall financial value of the annual yields was lowest with Birr 4362 

ha –1 . 

3. The enset-chat-maize-coffee type is found in 16% of all sampled homegardens. 

It is much more cash-oriented than the types 1 and 2. Staple food production 

dominates in 56% of the land with maize occupying more land than enset 

production. The importance of coffee as a cash crop is low, and chat has taken 

over this role. The diversity of plant species is relatively low, however, the 

number of livestock per farm is the highest (3.4) of all types. This can be 

attributed to the higher farm income of farmers (overall annual financial value 

of all crops = Birr 6802 ha –1 ) which enables them to buy cows and feed them to 

produce milk for home consumption and the market. 

4. The enset-coffee-maize-chat-pineapple type was represented in 8% of all 

sampled homegardens. This homegarden type accommodates a relatively 

balanced proportion of the different major crops. This garden type has the 

lowest area share of grazing and housing lands, and the highest proportion of 

croplands, where enset production dominates. The staple food crops enset,

131 

maize, and sweet potato occupy 41% of the land area against 46% for the cash 

crops. In addition to coffee and chat, pineapple is an important cash crop. 

Species diversity of this type is relatively low, but higher than in type 3. The 

overall financial value of all crop annual yields is high with Birr 6809/ha. 

Table 2. Area share of main crops in different homegarden types of four woredas in southern 

Ethiopia. 

Homegarden type (n) 

Area coverage of different crops 1 (%) 

enset coffee maize chat sweet 

potato 

pineapple 

others 

Enset-coffee-maize (84) 29.1 46.5 10.5 0.6 1.2 0.3 12.2 

Enset-coffee-maize-sweet 

potato (12) 

17.2 27.2 33.0 0.8 10.6 0 11.2 

Enset-chat-maize-coffee 

(24) 

24.8 13.7 31.6 19.8 1.4 0 8.7 

Enset-coffee-maizepineapple 

and chat (24) 

23.5 31.1 12.2 6.5 5.3 8.5 13.1 

Mean 26.4 36.6 16.4 4.5 2.6 1.6 11.9 

n=number of homegardens sampled. 1 Percentage area coverage of different crops was 

calculated considering the crop areas only. That t is, residential and grazing areas and separate 

woodlots were not included in the calculation. Overall area share including these are shown in 

Fig. 3. 

Homegarden types (n) 


Table 3. Composition of different homegarden types of four woredas in southern Ethiopia. 

Number of crop 

and tree species 

Number of 

livestock 

species 

Number of 

livestock 

(TLU ha –1 ) 

Mean SD Mean SD Mean SD 

Enset-coffee-maize (84) 41 a 

12.3 2.3 a 0.9 2.1 b 1.9 

Enset-coffee-maize-sweet 

potato (12) 

43 a 

12.2 2.0 a 0.4 1.9 b 1.0 

Enset-chat-maize-coffee (24) 25 b 5.6 2.1 a 0.7 3.4 a 3.6 

Enset-coffee-maize-chatpineapple 

(24) 

30 b 7.9 2.2 a 0.6 1.7 b 1.6 

Mean 37 12.0 2.2 0.8 2.2 2.2 

F test ( (p) 

< 0.001 ns < 0.05 

n = number of homegardens sampled; SD = standard deviation; TLU = tropical livestock unit; 

ns = not significant. Homegarden types with different letters differ significantly (F test and 

Duncan’s Multiple Range test, p < 0.05); p = probability level of significance.

132 


4.4. Factors influencing presence of different homegarden types 

The different types of homegardens were not evenly distributed over woredas 

(Tesfaye Abebe, 2005): type 2 was found in only one PA of Dale woreda, type 3 

was found in Awasa Zurya woreda only, type 4 was found in one PA in Dara and in 

one PA in Aleta Wondo. Type 1 was most extensively found, in three woredas. This 

indicates that the presence of different homegarden types cannot be explained by 

variation in physical conditions only; but socioeconomic conditions might account 

for a significant extent of such variations. For instance, the homegardens of type 1 

are located far from major roads, while homegardens of type 4 have good access to 

roads, which facilitates the sale of homegarden products. Homegarden type 3 is 

located in areas with a very high population density, which necessitated an increase 

in staple food production (e.g., maize). The impact of several ecological and 

socioeconomic factors on homegarden composition was further tested by means of 

multiple step-wise regressions between crop and tree diversity and possible 

explanative factors. The factors that were included in this analysis were: altitude and 

slope of the farm, farm size, farm labor force, involvement in off-farm activities and 

distance to major roads and markets (Tesfaye Abebe, 2005). Among these, the 

following two factors emerged as the most important determinants of homegarden 

diversity (Table 4). 

Farm size: Although a decrease in farm size was not significantly correlated with 

overall crop diversity, it negatively affected the relative proportion of a homegarden 

covered by cash crops, indicating how smallholders give priority to produce food 

crops rather than cash crops. Also, species richness of trees and livestock decreased 

with decreasing farm size. Small landholders grew the same number of crop species 

as the large holders; but with increasing land size, farmers increased the number of 

tree species. The density of dominant native timber and multipurpose species such 

as Podocarpus falcatus, Cordia africana and Milletia ferruginea also decreased with 

decreasing farm size, while that of fast-growing eucalyptus increased because of the 

need for wood for home consumption as well as for income generation. 

Access to major roads: Although access to highways did not significantly 

correlate with overall crop diversity, it affected significantly the area share of the 

major crops. The share of annual crops, mainly maize, increased at the expense of 

enset. Also the importance of the new cash crops chat and pineapple increased, 

while the traditional cash crop coffee declined. Proximity to major roads also 

negatively affected the richness in tree species. The share of native and multipurpose 

trees declined with increased road access, but the share of eucalyptus increased. This 

reflects the ability of eucalypts to grow fast and produce wood for consumption as 

well as income generation. 

These two factors are, however, not static, but depend on socioeconomic 

development. They can logically be related to the processes of population growth 

and commercialization. It can, therefore, be concluded that under the current local 

conditions, these two developments have a major impact on the dynamics in 

homegarden structure and composition. In respect to the first factor, it should be 

remembered that agroforestry homegardens of southern Ethiopia already carry a 

very dense population of 300 to 600 persons per km 2 . Its high growth rate (2.2%) is

133 

likely to increase the fragmentation of farmlands. The resulting increasingly smaller 

farms may lead to a reduction of the perennial crop and tree components as well as 

livestock. Regarding commercialization, it appears that the access to road networks 

often results in a gradually greater emphasis on commercial crops and crop 

specialization in homegardens. Consequently, the share of the perennial crops and 

native tree species tend to decline with proximity to highways and, hence, access to 

markets. 

Table 4. Multiple linear regression of species richness and number of livestock on physical 

and socioeconomic environments of homegardens in southern Ethiopia. 

Factors 


Species richness 

Crops Trees Livestock 

No. of 

livestock 

(TLU) 

Physical environment 

Altitude (1520 – 2040 m above sea level) ns ns ns 0.19* 

Slope (0 – 45%) ns 0.14* ns ns 

Socioeconomic environment 

Distance to markets (0.04 – 6.0 km) 0.17* ns ns ns 

Distance to highway (0.02 – 26 km) ns 0.35*** 0.17* ns 

Farm size (0.18 – 7.46 ha) ns 0.42*** 0.28*** 0.48*** 

Farm labor force (2 – 12) 0.18* ns ns ns 

Population density (2 – 35 inhabitants/ha of 

farmland) 

– 

0.20* –0.17* ns ns 

Involvement in off-farm work (yes/no) ns 0.14* ns ns 

Parenthetical values under “factors” denote the range for each parameter. R 2 values species 

richness of crops, trees and livestock were 0.15, 0.53 and 0.11 respectively and that of TLU 

was 0.48; * = p < 0.05; ** = p < 0.01; *** = p

134 

here as well. In the first place, the differences in lowland/highland location may 

impact species diversity with humid lowland homegardens having more biodiversity 

than highland homegarden systems (Wiersum, 2006). In the second place, one 

should take care in comparing the diversity figures, because of differences in the 

types of plant species considered. Some reports considered all plant species 

including ornamentals and sometimes weeds. For instance, Mendez et al. (2001) 

reported a total of 324 plant species in the homegardens of Nicaragua, out of which 

180 (56%) were ornamentals. Likewise, 219 plant species occurred in one village in 

West Java and 60 (27%) were ornamentals. As mentioned earlier, in the present 

study only deliberately grown crops and trees were recorded. Due to such 

differences in inventory, broad comparisons of species diversity in homegardens in 

agroecologically and socioeconomically different regions have several drawbacks, 

and should be considered as providing only indicative information. 

Ecological 

zone 

Humid 

lowlands 

Table 5. Species richness of selected homegardens in the tropics. 

Location Total no. of 

plant species 

West Java, 

Indonesia 

219 (60 

ornamentals) 

Average no. 

of plants per 

homegarden 

Sources 

56 Soemarwoto (1987); 

Soemarwoto and 

Conway (1991) 

39 De Clerck and 

Castillo (2000) 

70 (22–106) Mendez et al. (2001) 

Quintanana Roo, 150 useful 

Mexico 

plants 

Nicaragua 324 (180 

ornamentals) 

Santa Rosa , 168 35 (18–74) Padoch and De Jong 

Peruvian Amazon 

(1991) 

Humid Kandy, Sri Lanka 125 (93 usable) 46 (37–65) Perera and Rajapakse, 

lowlands 

1991 

to mid Kerala, India 127 woody 22 Kumar et al. (1994) 

altitudes 

species 

Highlands Chagga, Tanzania 111 (58 woody na Fernandes et al. 

and 53 herbs) 

(1984) 

Bukoba, Tanzania 57 na Rugalema et al. 

(1994) 

Wolayita and 60 14.4 Zemede-Asfaw and 

Gurage, southern 

Zerihun-Woldu 

Ethiopia 

(1997) 

Sidama, southern 198 crop and 37 (15–78) Tesfaye Abebe (2005) 

Ethiopia tree species 

na = not available. 

TESFAYE T ABEBE ET AL.


135 

Regarding the pattern in changing homegarden composition in the Sidama region 

in relation to decreasing plot size and commercialization respectively, similar trends 

have also been observed in ‘complementary’ homegarden systems (Kumar and Nair, 

2004; Peyre et al., 2006; Wiersum, 2006). For the ‘integral’ Chagga homegardens in 

Tanzania, changing livelihoods due to dynamics in socioeconomic conditions as 

well as market prices for garden products are reported to affect the composition of 

the gardens (Soini, 2005). We hypothesize that these trends are stronger in the 

‘integral’ homegarden system compared to the ‘complementary’ type because the 

former essentially incorporates cash crops, while this is not necessarily the case in 

the latter. 

6. IMPLICATIONS OF HOMEGARDEN COMPOSITION AND DYNAMICS 

FOR SUSTAINABILITY 

The description of the homegardens in southern Ethiopia illustrates that their 

structural characteristics are similar to the general features of tropical homegardens. 

The multispecies composition is often considered as a basic feature contributing to 

sustainability (Kumar and Nair, 2004). However, the presence of different types of 

homegardens illustrates that the homegarden composition is not always similar, but 

that it varies in response to socioeconomic differences and changes. Thus, when 

considering the sustainability of homegardens, a differentiation between ecological 

sustainability and socioeconomic sustainability seems warranted (Peyre et al., 2006). 

6.1. Homegarden composition and ecological sustainability 

Many studies have discussed the relation between species diversity in homegardens 

and their ecological sustainability (Soemarwoto, 1987; Torquebiau, 1992; Kumar 

and Nair, 2004). In the case of the Sidama homegardens, the presence of animal 

species is also a noteworthy phenomenon. Although such presence has been noted in 

several studies (e.g., Soemarwoto, 1987; Okafor and Fernandes, 1987), the animal 

component of homegardens is often neglected. However, our data demonstrate that 

livestock form an important component of the system (see also the Mesoamerican 

gardens described by Montagnini, 2006). In addition to their economic contribution 

by fulfilling various functions such as providing food in the form of milk and meat, 

traction and transport, they also play an important ecological role providing manure 

for the improvement of soil fertility and crop productivity. The animals contribute 

towards the maintenance of a closed nutrient cycling system with minimum dependence 

on external inputs such as fertilizers. The plant species diversity contributes toward 

the maintenance of animals. For instance, in the dry season, when fodder grass is in 

short supply, the animals are fed with immature thinned-out plants and leaves of 

enset, banana and other plants, as well as crop residues. 

Thus, within the Sidama homegardens not only the diversity in crop species and 

related multilayered vegetation system, but also the inclusion of animals in the 

system contributes toward their ecological sustainability. Moreover, the gardens also 

demonstrate that the stability of the system should not exclusively be related to its

136 


diversity, but can also be attributed, at least in part, to the specific characteristics of 

the two main components: enset and coffee. As an evergreen perennial crop, enset 

gives a permanent shade to understorey crops, including coffee. Soil management is 

facilitated by the use of enset residues as a mulching material. Coffee is an ideal 

complementary crop to enset. Not only is it architecturally and ecologically 

compatible with enset, but the harvest of both enset and coffee involve only selected 

plant parts and do not involve major export of soil nutrients. Thus, enset and coffee 

can be considered as keystone species contributing to ecological sustainability of the 

system. In ecological studies, the role of keystone species in maintaining ecosystem 

stability has received some attention (Mills et al., 1993; Khanina, 1998), but the 

notion of keystone species has still received little attention in agroforestry research. 

6.2. Homegarden composition and socioeconomic sustainability 

The maintenance of high species diversity in the Sidama homegardens also contributes 

to socioeconomic stability. As in other homegarden systems, the diversity of crop, tree, 

and livestock species with different uses and production cycles enables year-round 

production of different products, reduces risk of production failure, allows spreading of 

labor-use and flexibility, and enables efficient cycling of locally available resources, thus 

reducing dependence on external inputs (Kumar and Nair, 2004). In addition, the 

Sidama homegardens also incorporate several specific features in respect to 

socioeconomic sustainability. They not only have high species diversity, but also a 

high diversity in functional crop types; notable is the presence of both staple food 

crops and cash crops in addition to the more usual supplementary homegarden 

crops. The basic food crops (enset and maize), which are rich in carbohydrates, are 

supplemented by pulses, vegetables, fruits, and animal products that provide 

proteins, fats, and vitamins, and by trees that provide resources for construction and 

household energy. Also cash crops are incorporated in the homegarden, not only 

coffee, but also chat and pineapple. The proportion of subsistence and cash crops is 

often adjusted to meet the household requirements. Moreover, the spreading of risk 

from crop failures is not only facilitated by the crop diversity, but also by the 

inclusion of enset. The flexibility in harvesting enset for staple food production has 

been indicated as one of the main reasons why the Southern Highlands are relatively 

free of hunger (Desalegn Rahmato, 1995; Brandt et al., 1997). 

Thus, similar to the ecological sustainability, the socioeconomic sustainability 

cannot only be explained by species diversity, but also by the specific features of the 

two key species enset and coffee. Enset is both a food crop and a provider of 

different products such as fiber and fodder. It is therefore ideally suited to lowexternal 

input agricultural production systems, while its high productivity and 

multiple functions provide sustenance for a very dense population which is often 

two to three times higher than that in the cereal-based systems found in other parts 

of Ethiopia. Moreover, due to its perennial production cycle, enset can serve well as 

an excellent drought-relief crop. Coffee serves as a main cash crop supplementing 

the mainly subsistence-oriented enset production. The combined production allows 

for a good safety net in times of crop or market failures. Also, processing and 

marketing of coffee creates employment for many people. Consequently, not only


from an ecological point of view, but also from a socioeconomic point of view, 

coffee and enset can be considered as keystone species. 

6.3. Impact off system dynamics on sustainability 

137 

Even if the Sidama homegardens can be characterized as being sustainable, this does 

not mean that they do not change. The system is affected by decreasing farm size 

resulting from population growth and increased commercialization. The shift from 

the traditional enset-coffee systems towards inclusion of other food and cash crops 

has diversified the diet and increased household incomes. But the expansion of 

open-field food crops, such as maize and sweet potato, and of monocultural cash 

crops, such as chat and pineapple, are not only causing a gradual loss of species 

diversity and tree biomass, but also a gradual decrease in the dominance of the two 

key species enset and coffee. This results in a gradual reduction of the ecological 

benefits of these integrated and complex systems, e.g., by decreasing soil cover and 

thus increasing erosion hazards, as well as a reduction of the keystone enset species 

serving as ‘a tree against hunger’ in favor of quickly producing cash crops. Although 

the hazards of such changes in vegetation structure and composition could 

potentially be offset by more intensive management practices including use of 

external inputs, in case that no proper adaptive management activities are undertaken 

this may threaten the long-term sustainability of the homegardens. In view of the 

call for stimulating new development of forest-analogous land use systems 

combining production and biodiversity conservation (Wiersum, 2004), attempts 

should be made to integrate new crops into the existing multistory systems without 

affecting its biodiverse nature and without losing essential keystone species. 


The enset-coffee homegarden system can be considered as an integral homegarden 

system as it forms a spatially delineated farming system in contrast to the more 

commonly studied homegarden systems that are spatially complementary to 

cultivated fields and have a supplementary role to the overall family farming system. 

Nonetheless, both types of systems have several common features. The diversity of 

crops that are predominantly perennial in nature, with high diversity of trees and the 

presence of livestock allow a multitude of ecological interactions among the 

homegarden components and allow ecological sustainability. Moreover, the species 

richness combined with presence of different functional crop groups permit a 

balanced year-round production of both subsistence and cash crops. 

The enset-coffee homegarden systems also have some characteristics that are 

different from those of the more common spatially integrated (or supplementary) 

homegarden systems. Due to the absence of additional fields for staple food production 

or cash crop cultivation, the enset-coffee systems form a sort of “complete” farming 

system, producing a much higher proportion of basic food and cash crops than in the 

‘complementary’ homegarden systems. The system is characterized by the presence 

of two crops, enset and coffee, which are keystone species, due to their important

138 


economic and ecological roles. The large number of varieties of both species reflects 

their great importance in the system. The combination of these two, mutually 

compatible, native perennial crops and their dominance in the systems are essential 

features of these homegardens. 

In a similar manner as reported on complementary homegarden systems, the 

recent developments in land use systems resulted from increasing commercialization 

and continuing population growth affects the enset-coffee homegarden system. The 

growing population requiring basic foods has resulted in a gradual replacement of 

enset by annual staple food crops. The advent of commercialization has resulted in 

the development of new lucrative cash crops such as chat and pineapple requiring 

monoculture-cropping practices. These patterns have led to the decline in the areas 

of enset, coffee, and other trees. The decline in the share of these perennial 

components and their replacement particularly with annual crops could reduce some 

of the multiple benefits derived from these integrated and complex traditional 

systems, but the impact of such changes on long-term sustainability of the system is 

speculative, at best. 

Within homegarden studies, little attention has been given to differences in 

homegarden structure and function in relation to their position either in the overall 

farming system or to the role of keystone species in respect to the sustainability of 

these systems. As demonstrated by the features of the Sidama homegarden systems, 

these aspects deserve further research attention. 

ENDNOTE 

1. Ter Braak C.J.F. and Smilauer P. 1998. CANOCO reference manual and user’s 

guide to CANOCO for Windows: Software for canonical community ordination 

(version 4). Centre for Biometry, CPRO-DLO, Wageningen, 351p. 

REFERENCES 



functional changes. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A 

time-tested example of sustainable agroforestry, pp 233 – 250. Springer Science, 

Dordrecht. 

Almaz Negash. 2001. Diversity and conservation of enset (Enset ventricosum (Welw.) 

Cheesman) and its relation to household food and livelihood security in southwestern 

Ethiopia. PhD dissertation, Wageningen University, Wageningen, 247p. 

Bezuneh T. and Feleke A. 1966. The production and utilization of the genus Ensete in 

Ethiopia. Econ Bot 20: 65 – 70. 

Brandt S.A., Spring A., Hiebsch C., McCabe J.T., Endale Tabogie, Mulugeta Diro, Gizachew 

Wolde Michael, Gebre Yntiso, Masyoshi Shigeta and Shiferaw Tesfaye. 1997. The tree 

against hunger. American Association for the Advancement of Science, Washington, DC, 

56p. 

De Clerck F.A.J. and Negreros-Castillo P. 2000. Plant species of traditional Mayan 

homegardens of Mexico as analogs for Multistrata agoforests. Agroforest Syst 48: 

303 – 317.


139 

Desalegn Rahmato 1995. Resilience and vulnerability: Enset agriculture in southern Ethiopia. 

J Ethiopian Stud 28: 23 – 51. 

Fernandes E.C.M., Oktingati A. and Maghembe J. 1984. The Chagga homegardens: a 

multistoryed agroforestry cropping system on Mt. Kilimanjaro, Northern Tanzania. 


Hoogerbrugge I.D. and Fresco L.O. 1993. Homegarden systems: Agricultural characteristics and 

challenges. International Institute for Environment and Development, Gatekeeper series no. 

39, London, 23p. 


M. (eds), Tropical home gardens, pp 138 – 146. United Nations University Press, Tokyo. 

Khanina L. 1998. Determining keystone species. Conservation Ecology 2(2):R2 [online]: 

URL:http:// www.consecol.org/Journal/vol2/iss2/resp2 (last accessed: December 21, 

2005). 

Kippie Kanshie T. 2002. Five thousand years of sustainability? A case study on Gedeo land 

use. PhD dissertation, Wageningen University, Wageningen, 295p. 


135 – 152. 

Kumar B.M, George S.J. and Chinnamani S. 1994. Diversity, structure and standing stock of 

wood in the homegardens of Kerala in peninsular India. Agroforest Syst 25: 243 – 262. 

Mendez V.E., Kok L. and Somarriba E. 2001. Interdisciplinary analysis of homegardens in 

Nicaragua: micro-zonation, plant use, and socioeconomic importance. Agroforest Syst 51: 

85 – 96 


strategies of rural households in the area of Bogor, Indonesia. Agroforest Syst 25: 

31 – 58. 

Mills L.S., Soule M.E. and Doak D.F. 1993. The keystone-species concept in ecology and 

conservation. BioScience 43: 219 – 224. 

Montagnini F. 2006. Homegardens of Mesoamerica: biodiversity, food security, and nutrient 

management. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time- 

tested example of sustainable agroforestry, pp 61 – 84. Springer Science, Dordrecht. 

Oduol P.A. and Aluma J.R.W. 1990. The banana (Musa spp.)-Coffea robusta: traditional 

agroforestry system of Uganda. Agroforest Syst 11: 213 – 226. 

Okafor J.C. and Fernandes E.C.M. 1987. Compound farms of southeastern Nigeria: A 

predominant agroforestry homegarden system with crops and small livestock. Agroforest 

Syst 5: 153 – 168. 

Okigbo B.N. 1990. Homegardens in tropical Africa. In: Landauer K. and Brazil M. (eds), 

Tropical home gardens, pp 21 – 40. United Nations University Press, Tokyo. 

Padoch C. and Jong W. de 1991. The house gardens of Santa Rosa: Diversity and variability 

in an Amazonian agricultural system. Econ Bot 45: 166 –175. 

Perera A.H. and Rajapakse R.M.N. 1991. A baseline study of Kandian forest gardens of Sri 

Lanka: structure, composition and utilization. For Ecol Manag 45: 269 – 280. 

Peyre A., Guidal A., Wiersum K.F. and Bongers F. 2006. Homegarden dynamics in Kerala, 

India. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 


Rugalema G.H., Okting’ati A. and Johnson F.H. 1994. The homegarden agroforestry systems 

of Bukoba district, Northwestern Tanzania. 1. Farming systems analysis. Agroforest Syst 

26: 53 – 64. 

Soemarwoto O. 1987. Homegardens: A traditional agroforestry system with a promising 


pp 157 – 170. International Council for Research in Agroforestry, Nairobi.

140 


Soemarwoto O. and Conway G.R. 1991. The Javanese homegarden. J Farming Syst Res Extn 

2: 95-118. 

Soini E. 2005. Changing livelihoods on the slopes of Mt. Kilimanjaro, Tanzania: challenges 


Stoler A. 1978. Garden use and household economy in rural Java. Bull Indonesian Econ Stud 

14: 85 – 101. 

Tesemma Chekun 1997. The culture of coffee in Ethiopia. Agroforest Today 9: 19 – 21. 

Tesfaye Abebe 2005. Diversity in homegarden agroforestry systems of southern Ethiopia. 

Wageningen University, the Netherlands, Tropical Resource Management Paper No. 59, 

143p. 


Environ 41: 189 – 207. 

Westphal E. 1975. Agricultural Systems in Ethiopia. Centre for Agricultural Publication and 

Documentation, Wageningen, 278p. 


continuum: characteristics and future potential. Agroforest Syst 61: 123 – 134. 




Zemede Asfaw and Zerihun Woldu 1997. Crop associations of homegardens in Welayita and 

Gurage in southern Ethiopia. Sinet: Ethiopian J Sci 20: 73 – 90. 

APPENDIX I 

List of crop species in Sidama homegardens listed under their main functional crop 

group, sorted by frequency (% of homegardens in which the species is found, out of 

a total of 144). 

Scientific name Family English common Freque 

Roots and tubers 

name ncy (%) 

Enset ventricosum (Welw.) Cheesman Musaceae enset, false 100 

Dioscorea alata L. Dioscoreaceae yam 59 

Colocasia esculenta (L.) Schoot Araceae taro 51 

Ipomoea batatas (L.) Lam. Convolvulaceae sweet potato 42 

Manihot esculenta Cranz. Euphorbiaceae cassava 8 

Solanum tuberosum L. Solanaceae potato 6 

Beta vulgaris L. Chenopodiaceae beet root 5 

Daucus carota L. Apiaceae carrot 2 

Dioscorea bulbifera L. 

Vegetables 

Dioscoreaceae aerial yam 1 

Brassica integrifolia (West.) O.E. Brassicaceae kale 99 

Cucurbita pepo L. Cucurbitaceae pumpkin 83 

Capsicum frutescens L. Solanaceae hot pepper 43 

Brassica oleraceae L. Brassicaceae Ethiopian kale 33 

Lycopersicon esculenta L. Solanaceae tomato 16


141 

Capsicum annuum L. Solanaceae chilly 12 

Solanum villosum L. Solanaceae African 9 

Allium cepa L. Alliaceae shallot 3 

Brassica oleracea L. var. capitata Brassicaceae cabbage 3 

Lactuca sativa L. Asteraceae head lettuce 2 

Allium porrum L. Alliaceae leek 2 

Allium sativum L. 

Pulses 

Alliaceae garlic 1 

Phaseolus vulgaris L. Fabaceae common bean 99 

Phaseolus lunatus L. Fabaceae lima bean 30 

Vicia faba L. Fabaceae faba bean 3 

Pisum sativum L. Fabaceae pea 2 

Cajanus cajan (L.). Mill. 

Cereals 

Fabaceae pigeon pea 2 

Zea mays L. Poaceae maize 100 

Sorghum bicolor (L.) Moench Poaceae sorghum 31 

Eragrostis tef (Zucc.) Trotter Poaceae tef 6 

Hordeum vulgare L. Poaceae barley 2 

Triticum sativum L. 

Fruits 

Poaceae wheat 2 

Persea americana Mill. Lauraceae avocado 88 

Musa paradisiaca L. Musaceae banana 83 

Psidium guajava L. Myrtaceae guava 43 

Citrus sinensis (L.) Osbeck Rutaceae sweet orange 38 

Casimora edulis La Llave & Lex. Rutaceae white sapota 29 

Ananas comosus (L.) Merr. Bromeliaceae m 

pine apple 24 

Prunus persica (L.) Batsch Rosaceae peach 15 

Carica papaya L. Caricaceae papaya 15 

Passiflora edulis Sims. Passifloraceae passion fruit 13 

Annona reticulata L. Annonaceae bullock’s heart 11 

Mangifera indica L. Anacardiaceae mango a 

8 

Cyphomandra betacea (Cav.) Sendt. Solanaceae tree tomato 8 

Fragaria vesca L. Rosaceae strawberry 5 

Citrus aurantifolia (Christm.) Swingle Rutaceae lime 4 

Punica granatum L. 

Stimulants 

Punicaceae pomegranate 1 

Coffea arabica L. Rubiaceae coffee 100 

Chata edulis (Vahl.) Forssk. ex Endl. Celastraceae khat 57 

Nicotiana tabacum L. 

Spices and condiments 

Solanaceae tobacco 8 

Capsicum frutescens L. Solanaceae hot pepper 43 

Ruta chalepensis L. Rutaceae rue 17 

Appendix 1 (contd.)

142 


Capsicum annuum L. Solanaceae chilly 12 

Afromomum korarima (Braun) Jansen Zingiberaceae false cardamom 6 

Zingiber officinale L. Zingiberaceae ginger 3 

Rosmarinus offwinalis L. Lamiaceae rosemary 3 

Ocimum basilicum L. Lamiaceae sweet basil 3 

Lippia adonensis Hochst. ex Walp. Verbenaceae 3 

Piper nigrum L. Piperaceae black pepper 1 

Nigella sativa L. 

Oil crops 

Ranunculaceae black cumin 1 

Ricinus communis L. Euphorbiaceae castor 43 

Brassica carnata A. Br. Brassicaceae Ethiopian 19 

mustard 

Arachis hypogaea L. Fabaceae ground nut 9 

Carthamus tinctorius L. Asteraceae safflower 3 

Linum usitatissimum L. 

Medicinal plants 

Linaceae linseed 1 

Ocimum gratissimum L. Lamiaceae 13 

Foeniculum vulgare Mill. Apiaceae fennel 2 

Otostegia integrifolia Benth. Lamiaceae 1 

Artemisia absinthium L. 

Fragrance plants 

Asteraceae absinthe 1 

Ocimum gratissimum L. Lamiaceae 13 

Lippia adoensis Hochst. ex Walp Verbenaceae 12 

Cymbopogon citratus (DC.) Stapf. 

Other crops 

Poaceae lemon grass 1 

Rhamnus prinoides L’herit Rhamnaceae rhamnus 70 

Saccharum officinarum L. Poaceae sugarcane 54 

Lagenaria siceraria (Mol.) Stardl. Cucurbitaceae bottle gourd 5 

Agave sisalana Perr. Agavaceae sisal 4 

Gossypium herbaceum L. Malvaceae cotton 2 

Sorghum dochna (Forsk.) Snowden Poaceae sweet stalk 

sorghum 

1 

Pennisetum purpureum Schumach 1 

Poaceae elephant grass 12 

Chloris gayana Kunth 1 

Poaceae Rhodes grass 2 

Desmodium unicinatum (Jacq.) DC 1 

1 

introduced forage crop. 

Leguminoseae desmodium 1

CHAPTER 9 

HOMEGARDEN PLANT DIVERSITY IN 

RELATION TO REMOTENESS FROM 

URBAN CENTERS: A CASE STUDY FROM 

THE PERUVIAN AMAZON REGION 

A. WEZEL 1 AND J. OHL 2 

1 Institute of Landscape and Plant Ecology (320), University of Hohenheim, 70593 

Stuttgart, Germany; E-mail: . 2 School of Biological 

Sciences, University of East Anglia, Norwich NR4 7TJ, United Kingdom 

Note: Adapted from: Wezel A. and Ohl J. 2005. Does remoteness from urban centres 

influence plant diversity in homegardens and swidden fields: a case study from the 

Matsiguenka in the Amazonian rainforest of Peru. Agroforestry Systems 65: 241 – 251. 

Keywords: Floodplains, Indigenous people, Manu National Park, Shifting cultivation, Slash 

and burn, Swidden agriculture 

Abstract. Swidden cultivation is the traditional agricultural system in most parts of the Amazonian 

rainforest, and in many situations swiddens lead to the establishment of homegardens. In a remote 

area of the Manu National Park, Peru, such a system was investigated in two indigenous 

Matsiguenka communities for diversity of cultivated plants on swidden fields and in homegardens. 

The cultivated plants were identified from two to four plots per field in 46 fields in a total of 126 

survey plots and 19 homegardens. Altogether 71 species were found in the homegardens and 25 in 

the swidden fields. Cassava (Manihot ( esculenta) 

was dominant in the cultivated fields, whereas 

fruit trees such as peach palm (Bactris gasipaes), guava (Psidium guajava), and Inga edulis; and 

cotton (Gossypium barbadense) and a medicinal plant (Cyperus s sp.) predominated more than 75% 

of the homegardens. Species diversity increased steadily with age (length of cultivation) of the 

swidden fields. Diversity of species cultivated in the homegardens was low compared to other 

studies reported from the Amazon. This seemed to be due to remoteness from urban areas, relative 

isolation and consequently little interaction of the farmers with outside communities, and easy 

availability of plant products from nearby forests. Although these findings appear to contradict the 

premise that subsistence farming in such remote areas encourages farmers to produce a broad 

variety of species and, therefore, remoteness from urban centers increases species richness on 

farms; the extent to which the situation is impacted by easy availability of plant products from 

nearby forests, however, was not investigated in this study. In contrast to the homegardens, 

swidden fields in this study did not show any difference in species richness compared to other 

reported studies. 

143 




144 

A. WEZEL W AND J. OHL 


The indigenous agricultural system in most parts of the Amazonian rainforest is 

based on swidden cultivation (also known as slash-and-burn or shifting cultivation; 

Dufour, 1990). Besides cropping on swidden fields and cultivation of plants in 

homegardens, other activities of resource use such as hunting, fishing, and forest 

extraction are employed by rural peoples in the Amazonian floodplains and on the 

terra firme (elevated river terraces or hills). 

In a basic swidden cultivation cycle, the first step is to select a new field site in 

primary or secondary forest areas (Thrupp et al., 1997). The forests are then cleared 

and burnt, and different crops planted. Cropping is normally abandoned after 2 to 3 

years because of declining yields, partly due to increasing weed pressure. Finally, 

secondary forests develop on the fallowed fields and might be cleared again after 

several years. Analyses of different swidden systems in the Peruvian and Colombian 

Amazon including lists of plant species cultivated on swidden fields were reported 

by various authors previously (e.g., Johnson, 1983; Hiraoka, 1986; 1989; Eden and 

Andrade, 1987; Salick, 1989; Dufour, 1990; Coomes and Burt, 1997). In most such 

systems, cassava (Manihot esculenta) is the main crop during the first three years. In 

some cases, this is followed by another 2 to 3 years of plantain and banana (Musa 

spp.) production. Finally, the fields are fallowed (often called swidden fallows) 

when forest regrowth takes place, although fruit trees and other tree species still 

occupy the sites (e.g., Hiraoka, 1986; 1989). 

In systems with permanent or semi-permanent settlements, different plants are 

cultivated around the dwelling units. These homegardens are generally characterized 

by different vegetation strata (trees, shrubs and herbs) composed of annual and 

perennial agricultural crops and small livestock within the house compounds 

(Fernandez and Nair, 1986). Normally, the whole tree-crop-animal unit is 

intensively managed by family labor. Homegarden systems around the world have 

been comprehensively analyzed and summarized by Kumar and Nair (2004). In 

Peru, homegardens in the villages on the Amazonian floodplains have been 

investigated by Padoch and de Jong (1991) and Lamont et al. (1999), and those in 

the upland locations (upper Amazon) by Salick (1989). Works 1 described the 

homegardens of a small but steadily growing town in the Alto Mayo region of upper 

Amazon. 

Ohl (2004) analyzed the traditional economic system as well as the influences of 

new economic activities of two indigenous communities in a remote area in the 

Peruvian Amazon. Among other things, that study focused on the extent of changes 

currently taking place and its implications for sustainable resource use in the Manu 

National Park. The research of the broader project included an investigation on 

different aspects such as the socioeconomic situation of the households, land use 

practices, hunting and fishing activities as well as health care issues. This chapter 

focuses on the plant aspects of this broad study: plants cultivated on fields and in 

homegardens for meeting the basic food needs of the indigenous population, and the 

diversity as well as similarities among such plants in the two communities. The 

results were used to examine if remoteness from the urban centers increases plant 

diversity in homegardens and swidden fields, as often reported in the literature.

DIVERSITY IN RELATION TO REMOTENESS FROM URBAN U CENTERS C 

2. STUDY AREA 

145 

The study area is located in the lowland part of the Manu National Park in southeastern 

Peru (Fig. 1). The National Park comprises an area of about 1.72 million ha. 

The area studied within the National Park receives 2000 – 2500 mm annual precipitation 

2 . The average annual temperature is around 23°C. Two seasons are distinguished: 

a dry period from May to September and a rainy period from October to 

April. The vegetation is characterized by different types of tropical lowland 

rainforests on both periodically inundated alluvial plains (varzea) and on more 

elevated river terraces or hills (terra firme). Predominant soil types are Fluvisols and 

Gleysols (according to FAO classification) on the alluvial plains and Cambisols, 

Luvisols, and Acrisols on river terraces and on the hilly terrain 3 . 

This region is also the most sparsely populated area 4 of Peru with 0.1 

persons/km². Two Matsiguenka communities in the Fitzcarrald district of the Manu 

province were selected for the study following several visits to them during 2000 – 

2002; the community Tayakome (123 people) is located at 11°43.8’ S/71°38.8’ W 

(368 m altitude) and Yomibato (183 people) at 11°48’ S/71°54.4’ W (419 m 

altitude). Both are remote locations that can be reached only by boat on the river 

Manu, taking 1 to 4 days from Boca Manu, the nearest important settlement. Besides 

Tayakome and Yomibato, a third, smaller settlement of colonists from the Andes 

exists in the Manu National Park. 

2.1. Matsiguenka villages 

The two selected villages are relatively new communities founded in 1968 (Tayakome) 

and 1978 (Yomibato). A few decades ago, the Matsiguenka were mainly 

semi-nomadic hunter-gatherers, who used various sites along their treks to cultivate 

food crops (Johnson, 1989). After establishing permanent settlements, most people 

started to practice slash-and-burn, hunting, and fishing around their villages. 

Nevertheless, there are still some people, mainly in Yomibato, who wander around, 

moving between two or three distant huts spending a week or two in each, for 

successful hunting. In general, every household consists an average of 5.9 persons in 

Tayakome and 6.1 in Yomibato (range: 4 to 26 members per household in both 

villages: a household is defined as a group of people eating regularly together in one 

house), who actively cultivate 2.0 fields (a ‘field’ in this context is an area of land 

cleared in the primary or secondary forest for cropping) in Tayakome (0.73 ha) and 

1.8 fields in Yomibato (0.92 ha). The cropping period for a single field varies 

normally between 2 to 3 years during which cassava is mainly cultivated. In the 

third and fourth years, the swiddeners mostly harvest species such as plantains 

/bananas or papaya (Carica papaya), which were planted in the first year. After that, 

they return every now and then to the swidden fallows to collect mainly fruits from 

previously planted trees. 

The Matsiguenka often cultivate different crops in homegardens around their 

houses. Some chickens and ducks are also kept. Eggs are sometimes used, but their 

meat is only consumed in case of food shortage. Hunting trips are made only by

Figure 1. Location of the Manu National Park in Peru and the two villages studied. 

146 

A. WEZEL AND 

W J. OHL

. 

147 

men, mainly during the rainy season. Fishing is the main activity in the dry season 

when rivers are low. In addition, joint fishing is done in small rivers using fish 

poison. 

3. METHODS 

Cultivated plants were counted from 2 to 4 plots per field in a total of 46 fields in 

both communities; each plot was 5 x 10 m in size. There were a total of 126 survey 

plots in 1 to 2 year-old fields, implying that if the distribution of cultivated plants in 

the field appeared to be homogeneous, two plots were randomly selected and four if 

the distribution was non-homogeneous. Cultivated plants were counted in 19 

homegardens as well, and were organized into functional groups according to their 

preferred uses. Ornamental plants and timber species were not included in this study. 

Normally the homegardens are located within a zone of 5 to 25 m around the 

houses. The boundaries of most homegardens were quite evident because of the 

regular weeding that takes place within this zone. In few cases, however, it was 

difficult to distinguish between homegardens and the adjacent fields. Three age 

classes were recognized for the homegardens: young = up to 2 years since establishment; 

medium-old = 3 to 10 years; and old = more than 10 years. 

Plants which could not be directly identified were collected in a field herbarium. 

Local names provided by the owners were referenced to corresponding scientific 

names following Brack Egg (1999), and in few cases with the help of Shepard and 

Chicchón (2001) and Baer (1984). The collected plant samples were then verified 

with the help of botanists from the Universities of Cusco and Lima. Scientific plant 

names follow Brack Egg (1999). 

The similarity of species composition of fields and homegardens between the two 

communities was calculated using the Sørensen coefficient of similarity (Müller- 

Dombois and Ellenberg, 1974), according to the formula (2A/B+C) x 100 (where A = 

number of species common to two villages; B = total number of species in village 1 

and C = total number of species in village 2). For the comparison of species 

composition between fields and homegardens within each community also the same 

formula was used (where A = number of species common in fields and homegardens; 

B = total number of species in fields and C = total number of species in homegardens). 

4.1. Swidden fields 


4. RESULTS 

The dominant crop cultivated by the Matsiguenka on all fields is cassava (for details 

see Wezel and Ohl, 2005). It is consumed daily and used frequently to brew cassava 

beer. In Tayakome, 19 different cassava varieties were cultivated, and as many as 56 

in Yomibato. Other important plants included plantains/bananas and maize (Zea 

mays). Plants such as sugarcane (Saccharum officinarum), sweet potato (Ipomoea 

batatas), Lonchocarpus sp. (used as fish poison), papaya and cush-cush yam 

(Dioscorea trifida) were also cultivated, but less frequently; pineapple (Ananas ( 

comosus) and guava (Psidium guajava) were grown only occasionally.

148 

Furthermore, some species such as plantains/bananas, sugarcane, Lonchocarpus 

sp. and cotton (Gossypium barbadense) were found much more frequently in the 

second year of cultivation than the first. By contrast, maize is only planted during 

the first year. Maize, sugarcane, sweet potato, and papaya are cultivated generally 

more often in Tayakome than in Yomibato. In total, 25 species of crops were found 

in the study villages, with 21 in each village; besides, 18 and 24 species were found 

in one- and two-year-old swidden fields respectively. Species composition of the 

fields was comparable between the study villages with 81% similarity (Sørensen’s 

coefficient). 

4.2. Homegardens 


In Matsiguenka homegardens, fruit trees such as peach palm (Bactris gasipaes), 

guava, and Inga edulis; cotton and a medicinal plant (Cyperus sp.) predominated 

more than 75% of the observations (Table 1). Other fruit-producing trees and shrubs 

such as cashew (Anacardium ( occidentale), 

Pouteria caimito, mango (Mangifera 

indica), papaya, plantains/bananas, lemon (Citrus limon) and orange (C. sinensis) 

were also frequent with an occurrence of over 50%. Tubers such as cassava and 

Xanthosoma poeppigii, as well as pineapple and sugarcane were also noted on 50% 

of the gardens. One homegarden had many medicinal plants (but only some could be 

identified). Guava, Inga edulis, and Pouteria caimito were the most frequent tree 

species in the homegardens. 

Papaya, cassava, and pineapple were most frequent in young homegardens and 

their frequency decreased with age (Table 1). By contrast, Inga edulis, Pouteria 

caimito, orange, Genipa americana, mandarin (Citrus reticulata) and Crescentia 

cujete were found more often in the older homegardens. Out of 71 species found in 

all homegardens studied, 25 species, not considering medicinal plants, were 

cultivated in young homegardens, 27 in medium-aged and 50 species in old 

homegardens. For medicinal plants, the numbers were 26, 29 and 66 species, 

respectively. Eighteen species were found only in old homegardens. The almost 

exclusive occurrence of medicinal plants in the category of ‘old homegardens’ might 

be related to the fact that they were mainly found in one homegarden whose owner 

had a broad knowledge of the medicinal uses of such plants. The average number of 

species per garden increased with age: 14 in young, 16 in medium, and 20 in old 

homegardens. 

On an average, 18 species of plants were found in the homegardens of the two 

villages (range: 7 to 31). In both communities, the homegardeners mostly cultivated 

the same species, although a few disparate species with low occurrences (e.g., 

coconut palms, Cocos nucifera, with less than 15% occurrence) were noted. This is 

also reflected in the Sørensen coefficients of similarity, which showed that 75% of 

the species were similar in both villages when medicinal plants were excluded, and 

65% when they were included. 

Similarity of species between swidden fields and homegardens was 46% for both 

Yomibato (46% without medicinal plants) and Tayakome (54% without medicinal 

plants). Plants cultivated widely in swidden fields and homegardens were cassava, 

plantain/banana and, to a lesser extent, sugarcane.

Functional groups/species Occurrence (%) in different 

age classes of homegardens 1 

Matsiguenka name Local name English name Family 

Fruit trees 2 

Table 1. Cultivated plants in homegardens of the Matsiguenka in the Amazonian rainforest of Peru. 

Young Medium Old 

(n = 3) (n = 4) (n = 12) 

Bactris gasipaes 67 100 92 kuri pijuayo, chonta peach palm Arecaceae 

Psidium guajava 100 75 92 komashki guava guava Mirtaceae 

Inga edulis 67 75 92 intsipa guallaba, pakay Mimosaceae 

Anacardium occidentale 67 75 67 kasho marañon cashew Anacaridaceae 

Carica papaya 100 50 58 tinti papaya papaya Caricaceae 

Pouteria caimito 0 25 83 segorikashi caimito Sapotaceae 

Mangifera indica 0 75 67 manko mango mango Anacaridaceae 

Musa paradisiaca 3 

33 75 50 parianti platano plantain/banana Musaceae 

Citrus limon 33 75 58 irimoki limon lemon Rutaceae 

Citrus sinensis 0 50 67 naranka naranja orange Rutaceae 

Table 1 (cont.) 


149





(n = 3) (n = 4) (n = 12) 

Persea americana 33 0 58 tsivi, inchatoki palta avocado Lauraceae 

Genipa americana 0 25 58 ana huito Rubiaceae 

Citrus reticulate 0 25 58 tasharina mandarina mandarin Rutaceae 

Annona chirimola 67 0 50 tsirimoito chirimoya cherimoya Anonaceae 

Mauritia flexuosa 33 50 25 koshiki, achoariki aguaje Arecaceae 

Crescentia cujete 0 25 50 pajo, pamoko calabaza tree gourd Bignoniaceae 

Citrus limetta 0 0 50 lima 4 lima sweet lime Rutaceae 

Artocarpus altilis 0 50 25 pan de arbol 4 pan de arbol breadfruit Moraceae 

Attalea phalerata 0 25 25 tsigaro chapaja Arecaceae 

Inga sp. 0 0 25 intsipa orompiano pacay “colombiano” Mimosaceae 

Citrus grandis 0 0 17 toronja 4 pomela, toronja pummelo Rutaceae 

Cocos nucifera 0 0 17 koko coco coconut palm Palmae 

Oenocarpus bataua 0 0 17 sega(ki) ungurahui Arecaceae 

Solanum sessiliflorum 0 0 17 kokona cokona Solanaceae 

Bactris sp. 0 0 8 manataroki chontilla Arecaceae 

150 A. WEZEL AND 

W J. OHL

Tubers 

Manihot esculenta 100 75 42 sekatsi yuca cassava Euphorbiaceae 

Xanthosoma poeppigii 33 25 67 tsanaro uncucha cocoyam Araceae 

Calathea allouia 33 0 25 shonaki dale-dale Guinea arrowroot Marantaceae 

Dioscorea trifida 33 0 17 magona sacha papa cush-cush yam Dioscoreaceae 

Pachyrhizus ahipa 0 0 8 poi ashipa yam bean Fabaceae 

Ischnosiphon killipii 0 0 8 shirina sachaoca Marantaceae 

Ipomea batatas 5 

0 0 8 koriti camote sweet potato Convolvulaceae 

Vegetables and pulses 

Capsicum pubescens 33 0 17 tsitikana aji Solanaceae 

Lycopersicon cf. 

tomate 4 

tomate tomato Solanaceae 

peruvianum 33 0 0 

Cyclanthera pedata 33 0 0 iritsima poreatsiri caihua wild cucumber Cucurbitaceae 

Cucurbita moschata 0 50 0 kemi zapallo seminole pumpkin Cucurbitaceae 

Solanum mommosum 0 25 8 ivoniaro nuña huaca Solanaceae 

Cajanus cajan 0 0 8 ivinkoki poroto de palo pigeonpea Fabaceae 

Capsicum annuum 0 0 8 masekagana aji chili Solanaceae 

Lablab niger 0 0 8 tsitstita 

poroto hyacinth bean Fabaceae 



151





(n = 3) (n = 4) (n = 12) 

Others 

Gossypium barbadense 100 75 75 ampei algodon cotton Malvaceae 

Ananas comosus 100 75 50 tsirianti piña pineapple Bromeliaceae 

Saccharum officinarum 67 75 42 impongo caña azugar sugarcane Poaceae 

Bixa orellana 33 100 33 potsoti achiote annato Bixaceae 

Banisteriopsis sp. 33 75 25 kamarampi ayahuasca Malpighiaceae 

Lonchocarpus sp. 33 25 25 kogi, shimaaro, barbasco Fabaceae 

komo 

Brugmansia sp. 33 0 25 saaro, jayapa, toé Solanaceae 

kepigari 

Gynerium sagittatum 0 25 8 chakopi caña de flecha arrow cane Poaceae 

Nicotiana tabacum 33 0 0 seri tabaco tobacco Solanaceae 

Cedrela odorata 0 0 8 santaviri, santari cedro Meliaceae 

Curcuma longa 0 0 8 porikano palillo curcuma Zingiberaceae 

152 A. WEZEL AND 

W J. OHL

Hymenaea courbaril 0 0 8 koveni azucar huayo Caesalpinaceae 

Miconia sp. 0 0 8 savotaroki Melastomataceae 

Crescentia sp. 0 25 0 oeshinta, pamoko Bignoniaceae 

Medicinal plants 6 

Cyperus sp. 100 100 67 ivenkiki piri piri Cyperaceae 

Jatropha gossypiifolia 0 0 25 piñon piñon negro Euphorbiaceae 

Eryngium foetidum 0 0 8 sacha culantro sacha culantro Apiaceae 

Plukenetia volubilis 0 0 8 mani sachamani Euphorbiaceae 

Justicia pectoralis 0 0 8 viriorioshi Acanthaceae 

Cordia nodosa 0 0 8 matiagiroki Boraginaceae 

Martinella obovata 0 0 8 pocharo Bignoniaceae 

Eleutherine bulbosa 0 0 8 kapirokotapini yahuar piri-piri Iridaceae 

1 Young: 0 to 2 years, Medium: 3 to 10 years, Old: >10 years; 2 Inclusive of Musa paradisiaca and Carica papaya; 3 No distinction was between 

plantains and bananas. Anyhow, plantains were reported to be mainly cultivated; 4 The Matsiguenka use the common name as they have no own name 

for this species; 5 In few cases it might be a second Ipomoea species; 6 Additionally, nine different plants said to possess medicinal value were noted in 

the homegardens; but they could not be identified other than by their local names and are, therefore, not reported. 


153

154 


5. DISCUSSION 

5.1. Remoteness of homegardens and richness of cultivated species 

One question that is discussed in homegarden studies is the relation between species 

richness and distance to urban markets (Fernandez and Nair, 1986; Padoch and 

de Jong, 1991; Lamont et al., 1999). Often it is mentioned that urban-market 

pressure results in decreased total species diversity in the homegardens, whereas 

subsistence farmers in remote areas are compelled to produce diverse products and, 

therefore, species diversity increases in remote areas (but see: Lamont et al., 1999). 

In the present study, differences in total species numbers between homegardens 

of the two study villages were small, ranging from 49 in Yomibato to 58 in 

Tayakome. These numbers, however, were lower than the species richness reported 

by Lamont et al. (1999) from north-east Peru, where they documented 104, 111, and 

125 different species in three villages located 3 to 10 hours away by boat ride from 

the nearest urban center. Padoch and de Jong (1991), however, recorded as many as 

168 species in the homegardens of another de-tribalised and market-influenced 

village in north-east Peru. It needs to be noted that Lamont et al. (1999) included 

species for construction in their analysis, whereas Padoch and de Jong (1991) 

included species for construction as well as ornamental plants; albeit t their numbers 

were relatively low. The remoteness of the villages and ethnical differences seem to 

be important in determining total species richness. For example, in the village with 

the highest species number, residents included former members and descendants of 

at least four tribal groups as well as a few families who trace their ancestry to 

Europe (Padoch and de Jong, 1991). Peoples of mixed European and Amazonian 

ancestry live in the village with 125 homegarden species (Lamont et al., 1999). The 

other two villages are considered native communities although peoples of mixed 

ancestry have migrated to one of these villages over the years. Similar results are 

mentioned by Works 1 with more than 120 different species in the homegardens of 

Moyobamba. This town is a steadily growing urban center in the upper Amazon area 

where many newcomers settled in recent decades. In contrast, the Mastiguenka 

communities are native, without mixture of different tribes and located most 

remotely from the urban centers. The Matsiguenka homegardens can thus be 

characterized as relatively “pristine” with fewer cultivated plants. This seems to be 

similar to the situation of the Andoke and Witoto Indians in the Colombian Amazon, 

who cultivate only 33 species in their homegardens (Eden and Andrade, 1987). The 

contact of the Andoke and Witoto Indians with the outside world is relatively 

limited and local production is largely subsistence-oriented. Although the 

Matsiguenka exchange cultivated plants with other Matsiguenka communities, they 

are presently not able to sell any plant products from homegardens in the urban 

market because of remoteness and transportation problems. This could be a 

disincentive for planting many commercial species. Although Matsiguenka 

communities rely on subsistence production, they do not cultivate a broad variety of 

different species. Furthermore, these communities are of relatively recent origin,


155 

having been founded only in 1968 (Tayakome) and 1978 (Yomibato). The 

Matsiguenka still collect many products from the forest including medicinal plants. 

This seems to be the reason why only few medicinal plants were found in most 

homegardens. About 55% of the medicinal plants that were noted in this study were 

found in one single homegarden whose owner is a traditional healer and he planted 

medicinal plants from the forest as well as from other places in his homegarden. 

In the present study, 25 species were found in the one-and two-year-old swidden 

fields, 21 in each village. Johnson (1983) reported 26 species in his random samples. 

For young swiddens in the Colombian Amazon, Eden and Andrade (1987) recorded 

a total of 38 cultivated species, with an average of 12 per field, and Dufour (1990) 

reported nine different crops per field – but that could be because only four plots 

were studied. Contrary to the situation in the homegardens, however, differences 

in species numbers in the swidden fields were relatively small and factors such as 

remoteness did not seem to have an influence. This might be due to the fact that 

in swidden fields the most common crops and fruit trees are cultivated, whereas in 

homegardens, factors such as remoteness and cultural difference play a much more 

important role in species selection. 

5.2. Frequently cultivated species in the homegardens and swidden fields 

Many plants found in the Matsiguenka homegardens with high frequency are also 

typical plants of homegardens throughout the tropics in the world, e.g., plantains/ 

bananas, guava, mango, avocado (Persea americana), papaya, Citrus s spp., breadfruit 

(Artocarpus ( altilis), 

cassava, and sugarcane (Jensen, 1993; De Clerck and Negreros- 

Castillo, 2000; Méndez et al., 2001; Wezel and Bender, 2003). By contrast, coconut 

palms, which are planted very frequently in the homegardens worldwide, are rarely 

found in Matsiguenka homegardens, except for a few young trees in Tayakome. 

Instead, the peach palm is cultivated in 89% of the gardens analyzed. This species is 

a domesticated natural hybrid of different native Amazonian palms (Brack Egg, 

1999). Another frequently planted tree is Inga edulis, also a domesticated species in 

the tropical America. Both species have been reported from the homegardens of 

Latin America (Peru: Padoch and de Jong, 1991; Lamont et al., 1999; Colombia: 

Eden and Andrade, 1987; Brazil: Smith, 1996; Costa Rica: Zaldivar et al., 2002), 

although homegardens elsewhere seldom contain these species (Brazil, 1990). Other 

native plants which are most frequently cultivated in Mastiguenka homegardens and 

cultivated worldwide at present include cashew and guava, the latter having its 

origin in Peru itself (Brack Egg, 1999). 

Plant species found in the swidden fields of Matsiguenka are also reported from 

other areas of the Peruvian Amazon. For example, Johnson (1983) reported that 

cassava, maize, cocoyam (Xanthosoma sp.), pineapple, cotton, sugarcane, papaya 

and yam (Dioscorea sp.) are frequently planted crops of young Matsiguenka fields. 

In general, the most frequently planted crop that dominates the swiddens is cassava 

(Eden and Andrade, 1987; Dufour, 1990).

156 

5.3. Changes in species richness 


The longer an area of land is used by the Matsiguenka, the more different will be the 

plant species cultivated. For instance, species diversity increased steadily from 18 

and 24 species on one-and two-year-old fields to 26, 29, and 66 species in young, 

medium-old, and old homegardens, respectively. Some species such as maize, 

Calathea allouia, Dioclea virgata, Citrullus lanatus, and tobacco (Nicotiana 

tabacum) are only cultivated in swidden fields, and not in the homegardens. In 

contrast, many tree species, e.g., Citrus spp., Pouteria caimito, mango and avocado, 

are solely observed in the homegardens. The tree species present in fields are 

Lonchocarpus sp., peach palm, guava, Inga edulis, and cashew. These species 

except Lonchocarpus sp. are the most common ones in homegardens too and they all 

are native to Amazonia. 

Typical plants reported by the Matsiguenka to be harvested on the old abandoned 

fields are peach palm, avocado, Lonchocarpus sp., sugarcane and plantains/bananas. 

On such abandoned sites in other parts of Amazon, often described as the agroforest 

stage, preference for harvested species is, however, different, except for peach palm 

and plantains/bananas. Hiraoka (1986; 1989) described that peach palms and 

plantains/bananas as well as Inga edulis, star apple (Chrysophyllum caimito), Brazil 

nut (Bertholletia excelsa), and Poraqueiba sericea are still used in the floodplains of 

Peruvian Amazon. In the Colombian Amazon, Inga edulis, Theobroma bicolor, 

breadfruit, Poraqueiba sericea, Pourouma cercropiifolia, and the West Indian locust 

(Hymenaea courbaril) are cultivated (Dufour, 1990). 

5.4. The swidden cultivation system in Manu 

Before the Matsiguenka settled in communities, they used to move their fields and 

huts around in the rainforest area of the Manu. The swidden process typically 

involved clearing a patch of rainforest for cultivation (2 to 4 years), building the huts 

in the center of the field, and cultivation of mainly cassava in the first two years. 

They also used to plant some trees in a very simple form of homegardens around the 

huts. Once the field is abandoned, they move to a new field, and build a new hut. 

However, every now and then, they return to the old fields to harvest plantains /bananas 

or papaya. Some of the homegardens investigated in this study also originated in this 

manner. Presently, however, most homegardens are created anew – around scattered 

huts of the villages. Although this swidden cultivation system is very similar to the 

traditional one, the difference is that the Matsiguenka now cultivate fields within a 

certain range of the village – without having to move through the forests. They also 

re-establish new fields on fallows within a short period of time. As calculated for 

Yomibato, 29% of the field areas cultivated in 2000 or 2001 are located within 

formerly cultivated fallow areas (Ohl, 2004). About 28% of these fields were 

established on 2 to 10 year fallows, 38% on 10 to 14 year fallows and 34% on 14 to 

21 year fallows. However, most fields (71% field area) were created by clearing 

primary forests or very old fallows of at least 26 years. On satellite images these 

differences could not be clearly seen; but it seems that mainly primary forests have 

been cleared for establishing the new fields.



157 

The diversity of species cultivated by the Matsiguenka communities in homegardens 

is relatively low as compared to results of other studies reported from the rainforests 

of Amazon. Relative isolation from other communities and remoteness from urban 

areas seem to be the most important reasons for this low diversity. These findings 

are somewhat contrary to the often perceived notion that remoteness from urban 

centers increases species richness because subsistence production is based on a 

broad variety of species. Furthermore, these communities are still able to extract 

several plant products from the surrounding forests, and the impact of this factor on 

the observed low species diversity in homegardens was not investigated in this 

study. 

ACKNOWLEDGMENTS 

We thank E. Mannigel, H. Ohl and G. Shepard very much for their corrections and 

comments on the manuscript. The fieldwork of this study was financed by the 

Tropical Ecology Support Program of the GTZ, Germany. 

ENDNOTES 

1. Works M.A. 1990. Dooryard gardens in Moyobamba, Peru. Focus (Summer 

1990): 12 – 17. 

2. CTAR (Consejo Transitorio de Administracíon Regional, Madre de Dios), 

Parque Nacional del Manu and IMA (Instituto de Manejo de Agua y Medio Ambiente) 

1998. Propuesta de ordenamientio territorial de la provincia del Manu - 

Resumen ejectivo oficina de proramcíon y planeamiente ambiental – Unidad de 

estudios. Cusco, Peru, 231p. 

3. CTAR (Consejo Transitorio de Administracíon Regional, Madre de Dios) and 

IIAP (Instituto de Investigacíon de la Amazonia Peruana) 2000. Zonificacíon 

Ecologica Economica de la Region Madre de Dios, Medio Fisico II, Puerto 

Maldorado, Peru, 156p. 

4. Instituto Nacional de Estadistica e Informatica 2003. Censos – informacíon 

distrital. www.inei.gob.pe. (last accessed: December 2005). 

REFERENCES 

Baer G. 1984. Die Religion der Matsiguenka - Ost-Peru. Wepf & Co. AG Verlag, Basel, 

516p. 

Brack Egg A. 1999. Diccionario enciclopédico de plantas útiles del Perú. CBC, Cuzco, 550p. 

Brazil M.A. 1990. A list of herbaceous and woody plants grown in home gardens world-wide. 

In: Landauer K. and Brazil M. (eds), Tropical home gardens, pp 214 – 230. United 

Nations University Press, Tokyo. 

Coomes O.T. and Burt G.J. 1997. Indigenous market-oriented agroforestry: dissecting local 

diversity in western Amazonia. Agroforest Syst 37: 27 – 44. 

De Clerck F.A.J. and Negreros-Castillo P. 2000. Plant species of traditional Mayan homegardens of 

Mexico as analogs for multistrata agroforests. Agroforest Syst 48: 303 – 317. 

Dufour D.L.1990. Use of tropical rainforest by native Amazonians. BioScience 40: 652 – 659.

158 


Eden M.J. and Andrade A. 1987. Ecological aspects of swidden cultivation among the 

Andoke and Witoto Indians of the Colombian Amazon. Hum Ecol 15: 339 – 359. 

Fernandes E.C.M. and Nair P.K.R. 1986. An evaluation of the structure and functions of 

tropical homegardens. Agric Syst 21: 279 – 310. 

Hiraoka M. 1986. Zonation of Mestizo riverine farming systems in northeast Peru. Natl Geogr 

Res 2: 354 – 371. 

Hiraoka M. 1989. Agricultural systems of the floodplains of the Peruvian Amazon. In: 

Browder J.G. (ed.), Fragile lands of Latin America: Strategies for sustainable 

development, pp 75 – 101. Westview Press, Boulder, CO. 

Jensen M. 1993. Soil conditions, vegetation structure and biomass of a Javanese homegarden. 


Johnson A. 1983. Machiguenka gardens. In: Hames R.B. and Vickers W.T. (eds), Adaptive 

responses of native Amazonians, pp 29 – 63. Academic Press, New York. 

Johnson A. 1989. How the Machiguenga manage resources: Conservation or exploitation of 

nature. In: Posey D.A. and Balee W. (eds), Resource management in Amazonia: 

Indigenous and folk strategies, pp 213 – 223. New York Botanical Society, New York. 


135 – 152. 

Lamont S.R., Eshbaugh W.H. and Greenberg A.M. 1999. Species composition, diversity, and 


Méndez V.E., Lok R. and Somarriba E. 2001. Interdisciplinary analysis of homegardens in 

Nicaragua: micro-zonation, plant use and socio-economic importance. Agroforest Syst 51: 

85 – 96. 

Müller-Dombois D. and Ellenberg H. 1974. Aims and methods of vegetation ecology. John 

Wiley & Sons, New York, 547p. 

Ohl J. 2004. Die Ökonomie der Matsiguenka im Nationalpark Manu, Peru – Tourismus als 

Chance für eine nachhaltige Entwicklung? PhD thesis. University of Greifswald, 

Germany (on CD-ROM). 


an Amazonian agricultural system. Econ Bot 45: 166 – 175. 

Salick J. 1989. Ecological basis of Amuesha agriculture, Peruvian upper Amazon. Adv Econ 

Bot 7: 189 – 212. 

Shepard G.H.J. and Chicchón A. 2001. Resource use and ecology of the Matsiguenka of the 

eastern slopes of the Cordillera de Vilcabamba, Peru. In: Alonso L.E., Alonso A., 

Schulenberg T.S., and Dallmeier F. (eds), Rapid assessment program, pp 163 – 174. 

Smithsonian Institution/Monitoring and Assessment of Biodiversity Program, Washington 

D.C. 

Smith N.J.H. 1996. Home gardens as a springboard for agroforestry development in 

Amazonia. Int Tree Crops J 9: 11 – 30. 

Thrupp L.A., Hecht S. and Browder J. 1997. The diversity and dynamics of shifting 

cultivation: myths, realities, and policy implications. World Resources Institute, 

Washington D.C. 49p. 



Wezel A. and Ohl J. 2005. Does remoteness from urban centres influence plant diversity in 

homegardens and swidden fields: a case study from the Matsiguenka in the Amazonian 

rainforest of Peru. Agroforest Syst 65: 241 – 251. 

Zaldivar M.E., Rocha O.J., Castro E. and Barrantes R. 2002. Species diversity of edible plants 

grown in homegardens of Chibchan Amerindian from Costa Rica. Hum Ecol 30: 

301 – 316.

CHAPTER 10 

GENDER AND SOCIAL DYNAMICS IN 

SWIDDEN AND HOMEGARDENS IN LATIN 

AMERICA 

P.L. HOWARD 

Department of Social Sciences, Wageningen University, Hollandseweg 1, 6706 KN 

Wageningen, the Netherlands; E-mail: 

Keywords: Gender relations, Social structures, Women’s status. 

Abstract. Structure, composition, and functions of homegardens are said to be closely related 

to the social structure of households, but this issue is not often researched. An analysis of the 

literature on swidden and homegardens in Latin America shows that such interrelationships 

become transparent when examining the gender division of labor, gendered access to garden 

resources including land, trees, and other plants, and gendered control over subsistence and 

cash crops and income derived from them. Social status related to gardening, gendered 

knowledge distribution and transmission, and social dynamics leading to change in gardening 

and gardens are also important parameters in this matrix. A review of 39 Latin American case 

studies dealing with swidden or homegardens revealed that women are by far the prominent 

garden managers across its sub-regions. Aside from the multiple material benefits provided by 

gardens, other drivers that tend to ensure that women will strive to maintain them include 

their emotional and spiritual values and the positive social status that productive and beautiful 

gardens confer. Homegardening is a ‘respectable’ way for women to contribute to subsistence 

production and manifest specialized knowledge and skills without competing with men. 

However, commercialization may be undermining both women’s control and the benefits they 

derive from homegardening as well as the complex structure and function of homegardens. 


Past research on homegardens shows that the composition, structure, and functions of 

gardens are interrelated with their economic, social, and cultural functions (see for 

example Wiersum, 2006). However, the social dimensions of homegardens have only 

rarely been researched in-depth. Social factors influencing swidden and homegardens 

159 


Sustainable Agroforestry, 159–182 . 


160 

P.L. HOWARD 

have not been discussed in any depth in the agroforestry literature in Latin America. 

It is particularly by examining gender relations within swidden and homegardens 

that the complex interrelationships between social structures and gardens as land use 

systems become transparent. Examining gender relationships is also of great 

importance since, across most of Latin America, swidden and homegardening are 

largely women’s domains, and homegardens may help to mitigate the inequalities 

between the sexes that are evident across the region. 

This chapter is based on a review of the literature on swidden and homegardens 

in Latin America (that which is published in English as well as the little available to 

the author in Spanish) that reported sex-disaggregated information. Eight cases were 

found that focus on homegardens within Mayan production systems in 

Mesoamerica, whereas 12 cases refer to non-Mayan indigenous or mestizo (mixed 

Indian-Spanish descent) populations in the same region. In South America, 14 cases 

were found that deal with Amazonian Amerindian populations and swidden gardens, 

whereas only five cases were found that focus on homegardens among non- 

Amerindian South American populations, four of which are also in the Amazon 

basin. While swidden gardens and homegardens are distinct land use systems, they 

are both agroforestry systems that are rich in species diversity, possessing 

“sophisticated spatial structures and dynamics” and manifesting sustained yields 

(Michon, 1983). Further, while there has been very little study of homegardens 

among Amazonian Amerindians (for reasons for this see Heckler, 2001), there is a 

rich literature on swidden gardening. 

It must be recognized that a thorough comparative effort would require a 

substantially richer bibliographical underpinning. Further, the information provided 

in the 39 case studies that are reviewed here is very uneven and hence often difficult 

to compare. Thus, in analyzing this literature, the emphasis is on setting out certain 

similarities and identifying some of the potential explanatory factors in order to 

illustrate the nature and complexity of homegardens as social systems and of gender 

relations in swidden and homegardening, and to begin to relate these to the structure, 

composition, and functions of these gardens as agroecological systems. Finally, it is 

acknowledged that the 39 case studies analyzed herein do not cover the full 

spectrum of gardens across the sub-regions, or its ethnic, racial, and indigenous 

groups, and therefore the results can only be generalized within the limits of the 

study. 

2. THE GENDER DIVISION OF LABOR IN GARDENING IN LATIN 

AMERICA 

The gender division of labor not only provides many insights into how households 

organize homegarden production; it also highlights how contributions and 

responsibilities of individuals differ according to their positions within the 

household, which is very important for understanding the incentives, opportunities, 

and constraints that they confront when managing homegardens and how such 

individual factors influence homegarden structure, composition and functions. Many 

studies across the world seek to analyze the household division of labor in 

homegardens by sex and age, principally with the aim of understanding how

GENDER G AND SOCIAL DYNAMICS IN LATIN L AMERICA 

161 

production is organized. Some studies conceptualize and measure the division of 

labor in terms of tasks, where the breakdown may be gross or fine, e.g., land 

preparation, planting material procurement, varietal selection, planting, weeding, 

water management and irrigation, soil management, pest management, and 

harvesting. Others present the division of labor in relation to specific types of crops 

(e.g., medicinals, vegetables, spices, trees), to specific species (e.g., coffee [Coffea 

spp.], manioc [cassava, Manihot esculenta]), or to specific varieties (e.g., red maize 

[Zea mays] used for rituals, versus white maize used for daily consumption). Other 

relevant factors include the amount of time required and the timing and intensity of 

work, and the relation between homegarden work and individuals’ other labor 

obligations and physical mobility. Yet another way it is approached is in terms of 

the division of decision-making responsibilities (e.g., for location and design of the 

garden, selection and arrangement of species, cultural practices, destination of 

output). Divisions of labor are also sometimes discussed in relation to physical 

spaces such as zones within gardens or gardens in different locations that are 

considered to ‘belong’ to particular persons. Irrespective of how it is measured, the 

division of labor is based on cultural associations between sex, age, and kinship 

relations (e.g., senior male, wife, daughter-in-law) and obligations that people with 

such “social identities” (Boster, 1985a) have to provide particular resources for the 

household or for themselves, as well as the differential access to resources (land, 

labor, capital, markets, livestock, knowledge, and skills) that is related to these 

obligations. These cultural associations are rooted in cosmologies (‘world views’; 

understandings of the universe and human beings’ place in it) and related concepts 

of what is appropriate behavior for people of different social identities, which at 

least for the past generation have been undergoing rapid change nearly across the 

region. 

In general, homegardening studies in Latin American still do not mention the 

gender division of labor, and those that do often present only one of the possible 

indicators without specifying why that particular indicator was chosen. Who 

provides information about the gender division of labor is also an issue. Lerch 

(1999) pointed out that men and women gave very different answers when asked 

who was mainly responsible for the homegarden: men said that both were 

responsible whereas women said that they themselves were responsible. In fact, 

women were most observed working in homegardens. Dufour (1981) also reported 

that Tukanoan men in her study site in the Colombian Amazon insisted on 

representing their households to outsiders and socially it was not acknowledged that 

women are knowledgeable about plants; however, she found that the plant 

knowledge of men and women of the same social status did not differ significantly. 

Such problems affect much more than only data on the gender division of labor – 

getting the informants ‘right’ is necessary to avoid all sorts of research bias, but 

particularly bias about women’s work and knowledge around plants (Howard, 

2003). 

Table 1 presents the reported sex of the ‘main gardener’ (exclusively women, 

mainly women, or both men and women together) across the 39 cases, disaggregated 

by sub-region. Reporting the sex of the ‘main gardener’ does not mean that the other 

sex is not involved – even where it is indicated that “only women” are responsible,

162 P.L. HOWARD 

men may occasionally “help out” and, where “mainly women” garden, men often 

take on certain tasks or manage certain crops. In those case studies where men were 

reported to be the “only” or “main” gardeners, this was in relation to a minority of 

the households studied. 

2.1. Mayan Mesoamerica 

Geographically, Mesoamerica includes the seven countries of Central America as 

well as Mexico. Culturally it “joins present day middle and south Mexico, Belize, 

Guatemala and parts of Honduras and El Salvador,” much of which was historically 

dominated by the Mayan civilization (see Montagnini, 2006). As Montagnini 

indicates, homegardens are a complex and much-studied feature of the traditional 

land use system among the Mayan people, which have evolved in conjunction with a 

particular system of shifting cultivation and bush fallow (milpa) agricultural 

production that is organized around the ‘milpa triad’: maize, beans (Phaseolus 

coccineus, P. polyanthus, P. vulgaris) and squash (Curcurbita moschata, 

C. argyrosperma, C. pepo). In all but one (Murray, 2001) of the Mayan homegarden 

studies reviewed, it was found that women were the exclusive or main homegarden 

managers, although children and other household members might provide labor. For 

Mayan populations, the milpa, which provides the bulk of subsistence staples and 

cash crops, is “symbolically the male domain and is the source of male prestige” 

(Stavrakis, 1979; Greenberg, 1996; Murray, 2001; Lope Alzina, 2006). Women 

often provide ‘additional’ or seasonal labor for milpa production, but it is considered 

improper for women to be seen in the milpa without the company of males, a 

proscription which is enforced by social sanctions, gossip, and even the threat of 

witchcraft (e.g., Murray, 2001). On the other hand, homegardens (solares, huertos, 

patxokon na) are perceived as female domains or spaces where a great diversity of 

vegetables, condiments, ornamentals, medicinals, and other utilitarian or ritualistic 

plants are maintained along with most useful trees, and where women are primary 

decision-makers (Benjamin, 2000; Patterson, 2000). When men are involved, this is 

either related to particular tasks such as land clearing, tree pruning and thinning, 

construction of structures and fences and chopping undesirable growth (Benjamin, 

2000; Patterson, 2000), or to specific species or crops. Men also use homegardens as 

experimental stations and in situ gene banks for crop diversity - for example, in a 

case study site in the Yucatan, Mexico, men use homegardens to test new maize 

varieties and preserve traditional varieties that they do not wish to plant in their 

fields (Lope Alzina, 2006). Trees and tree crops may be of particular concern to 

Mayan men especially when these have commercial value (Patterson, 2000), but 

they are usually principally women’s responsibility (Gillespie et al., 2004) because 

they fall within the physical space of the homegarden. 

In one case, in Belize among the Kekchi Maya, the interrelationship between the 

gender division of labor, production spaces, crops, and conceptions about what is 

appropriate behavior for women was shown to have an effect on the species 

diversity and size of different types of gardens. Women manage homegardens and 

‘milpa gardens’. The latter are established after milpa fields are left fallow and are 

generally much larger than homegardens (Patterson, 2000). Women maintain

Sex of the 

“main 

gardener” 

Women only i 4 50 

i 

Table 1. Sex of the “Main Gardener” in the 39 homegardens across Latin America. 

Number and percentage of cases, countries, and literature citations forr different sub-region/ethnic groups 

Mayan Mesoamerica Non-Mayan Mesoamerica Amazonian Amerindian Non-Amerindian S.A 

# % Country (literature # % Country (literature # % Country (literature # % Country 

reference) 

reference) 

reference) 

Belize (35), 

Guatemala (18), 

Mexico (12, 14) 

Mainly women 3 37.5 Belize (29), 

Mexico (5, 22) 

3 25 Belize (28); Mexico (6, 13) 5 35.7 Brazil (30); 

Colombia (17, 38); 

Peru (1); 

Venezuela (16) 

3 25 Honduras (9), Mexico (3, 

24) 

6 42.9 Colombia (10), 

Ecuador (8, 36), 

Peru (7), 

Venezuela (15, 37) 

2 3 

0 

3 7 

0 

(literature 

reference) 

Brazil (34); 

Ecuador (11) 

Brazil (23, 

39), Peru (20) 


GENDER G AND SOCIAL DYNAMICS IN LATIN L AMERICA 163

Men and 

women 

1 12.5 Mexico (26) 6 50 Costa Rica (27); Guatemala 

(31); Honduras (21); 

Mexico (19); Nicaragua 

(25); Panama (33) 

3 21.5 Brazil (4), Peru (2, 

32). 

Total 8 12 14 5 

i Men may ‘help’ or not, but their labor is otherwise not systematically involved. 

Literature citations and countries of the case studies. 

1. Aikman, 1999 (Peru); 2. Alexiades, 1999 (Peru); 3. Angel Perez et al., 2004 (Mexico); 4. Baleé, 1994 (Brazil); 5. Benjamin, 2000 (Mexico); 6. 

Blanckeart et al., 2004 (Mexico); 7. Boster, 1985a (Peru); 8. Descola, 1994 (Ecuador); 9. Doxon, 1988 (Honduras); 10. Dufour, 1981 (Colombia); 11. 

Finerman and Sackett, 2003 (Ecuador); 12. Gillespie, et al., 2004 (Mexico); 13. Govers, 1997 (Mexico); 14. Greenberg, 1996 (Mexico); 15. Heckler, 

2004 (Venezuela); 16. Hoffmann, 1993 (Venezuela); 17. Irvine, 1987 (Colombia); 18. Keys, 1999 (Guatemala); 19. Lazos Chavero and Alvarez Buylla, 

1988 (Mexico); 20. Lerch, 1999 (Peru); 21. Lok (endnote 1; Honduras); 22. Lope Alzina, 2006 (Mexico); 23. Madaleno, 2000 (Brazil); 24. Martin, 1996 

(Mexico); 25. Mendez et al., 2001 (Nicaragua); 26. Murray, 2001 (Mexico); 27. Ochoa et al. (endnote 2; Costa Rica); 28. Palacio, 1980 (Belize); 29. 

Patterson, 2000 (Belize); 30. Posey, 1984 (Brazil); 31. Ruonavaara, 1996 (Guatemala); 32. Salick, 1997 (Peru); 33. Samaniego and Lok (endnote 3; 

Panama); 34. Sereni Murrieta and Winklerprins, 2003 (Brazil); 35. Stavrakis, 1979 (Belize); 36. Uzendoski, 2004 (Ecuador); 37.Veth and Reinders 

(endnote 4; Venezuela); 38. Wilson, 1997 (Colombia); 39. Winklerprins, 2002 (Brazil). 

0 0 

164 P.L. HOWARD

milpa gardens because of constraints in terms of soil quality, animal predation, and 

lack of space in homegardens. It was found that women who cultivate a large 

number of edible crops and herbs in their homegardens are those whose husbands 

are frequently away for considerable periods, during which time women rarely if 

ever travel to milpa gardens. Women whose husbands do not leave the village 

cultivate larger milpa gardens and maintain fewer species in their homegardens. In 

another case, in a Mayan community in the Yucatan (Lope Alzina, 2006) where 

communal land has become available within city limits, people use it as a second, 

non-traditional gardening space. Men and women share labor and decision-making 

to a much greater extent and women are allowed to work in these gardens 

unaccompanied by men, even though these gardens are organized as a kind of 

“miniature milpa,” containing the traditional milpa staple crop triad. Thus, gendered 

norms appear to be more flexible when people work outside of the traditional 

production system, and such flexibility also affects the structure, composition, and 

functions of homegardens. 

2.2. Non-Mayan Mesoamerica 

GENDER G AND SOCIAL DYNAMICS IN LATIN L AMERICA 165 

In the division of labor among non-Mayan Mesoamerican populations, men are 

typically responsible for field crop and cattle production and women for 

homegardening and small livestock (usually pigs and chickens). The staple crops 

produced across much of Mesoamerica are similar to those produced in the Mayan 

milpa: especially maize and beans are prominent. The exception is presented by the 

one case study on the Garifuna (Palacio, 1980) which is a Black Carib population in 

which women are the main crop producers. Eleven of the studies presented 

information on the “main gardeners” – in six of these cases, women are reported to 

be exclusive or principal gardeners, whereas in the other five cases homegardens are 

managed by both sexes. The main differences with the Mayan division of labor in 

terms of men’s participation appears to be that men are involved in homegardening 

principally in relation to crops with high commercial value, especially tree crops 

[e.g., citrus (Citrus spp.) and coffee], and there are more cases where men use 

homegardens to test exotic crops that they wish to introduce into agricultural 

production (Angel Peréz and Mendoza, 2004; endnotes 2 and 3). Homegardens may 

also be considered in general as women’s spaces in non-Mayan Mesoamerica and 

the restrictions on women’s work in milpas appear to be strict (Govers, 1997; 

Roquas, 2002). 

2.3. Amazonian Amerindians and swidden gardening 

Nearly all Amazonian Amerindian societies have traditionally depended for their 

livelihoods on a combination of hunting, fishing, gathering, and gardening activities, 

where men hunt and fish and women are responsible for gardening, although these 

relations are changing mainly due to commercialization (Knauft, 1997; Heckler, 

2004). Amazonian Amerindians often have highly complex land use systems that 

combine multiple types of swidden gardens (including fallow field gardening) and


homegardening to provision themselves with starchy staples, particularly manioc 

(both bitter and sweet varieties), which are complemented especially by plantain and 

banana (Musa spp.), yam and sweet potato (Dioscorea spp.), taro (Colocasia 

esculenta), vegetables, fruits and medicinals. The complexity of their agroforestry 

systems and social organization has made them the subject of much in-depth 

research. 

In all but three of the 14 case studies on Amazonian swidden gardens that were 

reviewed, women were the exclusive or principal gardeners. In one case, among the 

Ese Eja in Peru (Alexiades, 1999), there is also an age division of labor that gives a 

more prominent role to older men in gardening than in other Amazonian Amerindian 

cases, mainly due to the fact that certain cultivars are associated with malevolent 

spirits that may harm fetuses and infants, which effectively prohibits women of 

childbearing age from cultivating or consuming them. In another case, among the 

Ka’apor in the Eastern Amazon of Brazil (Baleé, 1994), men invest a slightly greater 

amount of time in swidden gardening than women, and neither sex invests much 

time in homegardening. However, men’s involvement in swidden gardening over 

much of Amerindian Amazonia is often restricted mainly to clearing undergrowth 

and felling trees for new gardens, whereas all other tasks are left to women (Posey, 

1984; Hoffman, 1993; Descola, 1994). In some cases it is reported that men assist in 

garden maintenance (Hoffman, 1993; Uzendoski, 2004), particularly among 

Guyanese groups where the gender division of labor is less rigid – for example, 

among the Piaroa of Venezuela, men are reported to help in weeding, harvesting and 

carrying crops from swidden gardens (Heckler, 2004). 

The gender division of labor is not only reflected in tasks associated with 

gardening – it is also often strongly related to crops as well as to physical spaces, 

associations which are embedded within cosmology and concepts of masculinity and 

femininity that are in turn related to prestige and to complementarities and conflicts 

between the sexes. Manioc is by far the most important crop across Amazonian 

Amerindian cultures, and it is strongly culturally associated with women – in only 

two cases (Baleé, 1994; Salick, 1997) was it found that men had a substantial role in 

manioc cultivation and in one of these (Salick, 1997) it was reported that this 

probably represented a deviation from the traditional division of labor due to labor 

shortages. Manioc and manioc beer figure importantly not only in the diet, but as 

well in ritual and exchange. The highly complex cosmology associated with women, 

manioc, and gardening is discussed in relation to the Achuar (Descola, 1994) and the 

Warua 4 . 

Apart from tubers, other crop-sex associations are also quite evident. Among the 

Ka‘apor (Baleé, 1994), both men and women plant manioc, but women are 

exclusively responsible for planting cotton (Gossypium spp.), Indian shot (Canna 

indica), job’s tears (Coix lacryma jobi) and pipiriwa (Cyperus corymbosus), which 

are used only by women for textiles or for body ornamentation. Only men plant 

maize. Among the Piaroa, it is also men who plant maize, and they exclusively plant 

tobacco (Nicotiana tabacum) (Heckler, 2004). In fact, Amazonian Amerindian men 

are often strongly linked to particular species and have exclusive power to manage 

these species – among the Achuar, only men may plant botanical fish poisons since 

“if these were to be handled by women, they would lose their effectiveness”


(Descola, 1994). Achuar men are also predominantly associated with hallucinogens, 

tobacco, maize, and bananas, which are only planted outside the main swidden 

garden around the edges of the house yard. This reveals yet another aspect of the 

gender division of labor. Women’s swidden gardens are out-of-bounds to men since 

they are “the only absolutely female space in the Achuar social topography, the only 

place where women truly exercise a material and symbolic hegemony” (Descola, 

1994). Uzendoski (2004) also noted the very strong association between physical 

production spaces and gender among the Napo Runa of the Ecuadorian Amazon, 

who see the forest (sacha) as masculine while gardens (chagra) are seen as mainly 

feminine. 

2.4. Non-Amerindian South America 

Few case studies were found that focus on homegardening in non-Amerindian South 

American societies and that discuss the gender division of labor. Four out of the five 

studies reviewed deal with the Amazonian region but not with Amerindian 

populations, whereas only one was found that relates to the Andes (Finerman and 

Sackett, 2003). 

All five cases reported that women are the exclusive or principal gardeners in the 

majority of the households that were investigated. However, in three of these cases, 

a number of households were found where men were main gardeners, although those 

households were in the minority. In the case of three villages in rural Amazonian 

Peru (Lerch, 1999), in those households where men were the main gardeners, it was 

clear that women were also involved in the work. In the two urban cases in Para 

State, Brazil, 70% of the urban growers in Belém were women, whereas in 

Santarém, 67% of the homegardens were maintained by women (Madaleno, 2000; 

Winklerprins, 2002). On Ituqi Island, also in Para State, among a Caboclo 

population (mixed Brazilian Amerindian and European or African ancestry), 

homegardens are said to be the “unquestionable domain of women” (Sereni Murrieta 

and Winklerprins, 2003). In the Ecuadorian Andes, Finerman and Sackett (2003) 

also found that women are unquestionably the heads of gardens. These gardens are 

“medicine cabinets” (one contained 194 species of which 132 were medicinals and 

on average nearly 70% of species in homegardens were used for medicine), where 

lay medicine is clearly defined as a female domain. 

The case studies reviewed above demonstrate that the gender division of labor 

can be viewed in multiple ways – as a division of tasks or responsibilities, of crops, 

or of resources or physical domains, and typically as a combination of these. 

Regardless of how it is viewed, it is related to culturally established norms of 

behavior that often have their roots in cosmology and that clearly differ according to 

ethnicity and tribal affiliation. What is considered appropriate behavior for women 

differs strongly between Amerindian and non-Amerindian populations insofar as 

women have primary responsibility for staple crop production in many Amerindian 

societies. Yet they are excluded from such production over much of Latin America, 

where instead they are responsible for a myriad of so-called ‘minor’ crops, 

particularly those with cultural, culinary, and medicinal values that are typically 

produced in homegardens. In both cases, the responsibility for swidden and


homegardens, and ipso facto for the skills, resources, knowledge, and biological 

diversity that are entailed, falls mainly to women. 

3. GENDER AND ACCESS TO GARDEN RESOURCES 

From the foregoing discussion, it appears that the gender division of labor is closely 

related to men and women’s differential access to homegarden resources, especially 

land, trees, and other plants. Terms of access and rights to these resources are more 

complex and significant than most case studies suggest and have significant 

consequences for garden structure, composition and functions, for the investments 

made and benefits derived from gardening, and for the distribution of such benefits 

between households and among household members. 

As is the case with property rights of all types, rights to swidden and 

homegarden resources are also differentiated by sex. Variations in homegarden 

resource access according to sex that are found in the case studies are summarized in 

Table 2. This table includes only a subset of the total number of case studies 

reviewed since the majority did not provide sufficient sex-disaggregated information 

about resource access. 

Table 2. Patterns of gendered access to garden resources in the 39 case studies across Latin 

America. 

Agricultural 

land 

owner- 

ship ii 

Agricultural 

land 

usufruct 

Garden 

land 

ownership 

ii 

Garden 

land 

usufruct 

Garden 

tree 

ownership 

iii 

i 

Garden 

plant 

ownership 

iii 

Countries (literature 

references: see 

Table 1 for author 

details) 

MF M MF F F F Ecuador (11) 

M M M MF MF MF Belize (29) 

Mexico (5, 12, 22) 

M M M MF M F Costa Rica (27) 

Mexico (3, 19) 

Panama (21, 33) 

- - M MF MF MF Ecuador (8) 

Note: signifies physically separate; M = male; F = female. 

i Includes only cases where gendered resource access and control were discussed. 

ii Where there is no private ownership, this means customary control over land allocation. 

iii May be explicit, or may be inferred from the case studies. 

Although homegardens in Latin America are nearly by definition small 

(generally considerably less than 1 ha and at times only a dozen square meters), land 

is obviously still a crucial production factor, and homegarden land can be even more 

productive on a per hectare basis than agricultural land 3 . Irrespective of the sex of 

the main gardener, it is reported that poorer households have greater difficulty 

obtaining access to land for homegardening and greater difficulties meeting


household needs from gardens when they do have access. For these people, lack of 

tenure security represents the greatest threat since many occupy land illegally, 

especially in urban or urbanizing areas (Madaleno, 2000; Finerman and Sackett, 

2003). Lower land access is reported to prevent cultivation of species that require 

substantial amounts of space (Doxon, 1988). As well, tree planting and security of 

land tenure are often interrelated since tree planting often creates rights to land or, 

conversely, only landowners may plant trees (Bruce and Fortmann, 1988). Thus, tree 

and land tenure may affect composition and structure through the number and type 

of trees planted in homegardens and through garden size. 

It is normally presumed that whoever owns or formally controls homegarden 

land will control homegarden production, but this is certainly not so in the case 

studies reviewed here. Deere and Leon (2001) show that male land ownership and 

control predominate over most of Mesoamerica and non-Amerindian South 

America, which is confirmed by the homegarden studies that report such 

information: in only one case was it said that men and women jointly own land 

(Finerman and Sackett, 2003). Generally, it is men who have the ultimate right to 

dispose of land, although decision-making may be joint as Patterson (2000) showed 

for Mayan homegardens in Belize. In all homegardening cases, women appear to 

obtain informal usufruct rights to homegarden land from their husbands. Among 

populations where there is no clear concept of land ownership, such as among many 

swidden horticulturalists, ‘spaces’ are frequently gendered and gardens may be 

considered strongly or weakly as ‘women’s spaces’. Generally, men formally 

control swidden garden land and allocate it to women (Descola, 1994; Goldman, 

1963; Posey, 1984). However, neither of these phenomenon is pervasive – for 

example, “The Cubeo always speaks of a particular manioc plot as belonging to a 

woman, the only instance of individual possession of land in Cubeo society” 

(Thompson, 1977). 

Only two studies were found that elucidate why or how it is that women gain 

which type of usufruct rights to homegarden land that men control (Descola, 1994; 

Lope Alzina, 2006), and these discuss how such rights are influenced by 

negotiations between men and women. Among the polygamous Achuar of 

Amazonian Ecuador (Descola, 1994), each co-wife must cultivate her own garden 

plot. Men divide plots that they have cleared and assign them to each co-wife by 

planting rows of banana. The size of the garden is negotiated: while both men and 

women wish to have large gardens, women consider their access to labor for 

weeding and negotiate with their husbands considering this constraint. It is also 

notable that once each co-wife’s patch “has been materially marked out under male 

authority, the garden finally becomes the closed area of a purely female praxis.” 

Still, it is clear that women have strong obligations to produce manioc beer for the 

men who provide them with land. 

Access to homegarden land must be very important for those Latin American 

women who lack access to other land to cultivate, particularly when women are able 

to use this land in a way that is highly productive and to make most decisions 

regarding its use and management. The fact that women obtain land through their 

husbands implies that divorce or separation may deprive them of access altogether,


and men can also ultimately decide to use such land in other ways, which is 

discussed further below. 

Land is not the only resource that is crucial to homegardening. Agrobiodiversity 

is also a major resource and rights to trees and other plants cannot simply be 

assumed to pertain to those who control the land. What appears to be most common 

in the cases reviewed is that men may grow specific homegarden crops of their own, 

but most homegarden species belong to women. Within this, a relatively consistent 

pattern is discernable in relation to trees, since trees are often related to male 

ownership (Bruce and Fortmann, 1988) and tree tenure also differs from “plant 

tenure” (Howard and Nabanoga, 2006). In the Mayan cases, tree ownership does not 

appear to be clearly related to either men or women and women generally plant and 

manage trees in homegardens (Benjamin, 2000; Gillespie et al., 2004). Nevertheless, 

in the Mayan case that reported that trees provide cash crops, men also participate in 

their management and use (Patterson, 2000). In many non-Mayan Mesoamerican 

communities, however, it appears that men make decisions about trees and control 

income from them, particularly when they have high commercial value (Lazos 

Chavero and Alvarez Buylla, 1988; endnotes 2 and 3). 

One case provides insights into the relationship between land tenure, tree tenure, 

and cosmology. Samaniego and Lok 3 report that, among the Ngöbe in Panama, 

women attribute greatest importance to homegardening while men value tree 

(especially coffee) production most highly. Land is generally communally owned, 

but the person that plants a tree becomes the owner of the land upon which the tree 

is planted. The head of a household will bury the placenta pertaining to every 

newborn and plant a tree on that site, and only the head knows the tree species and 

the site. The well-being of the tree and of the person whose birth the tree marked are 

directly interrelated, so household heads should always be consulted regarding the 

management of trees in homegardens. 

Aside from trees, women may have exclusive rights to plants growing in 

homegardens and their husbands or other household members may have no right to 

harvest or otherwise destroy these plants without their permission. Finerman and 

Sackett (2003) reported that, in their study village in the Ecuadorian Andes, men and 

women jointly own land and animals. Men make most decisions concerning 

farmland and cattle, while households and homegardens are women’s domains, both 

in terms of management and of rights to the plants growing therein. Anyone wishing 

to have access to a plant in a woman’s garden must ask her permission. Dufour 

(1981) observed that, among the Tukanoan Indians of the Colombian Northwest 

Amazon, women often plant manioc in a section of another woman’s garden. This 

highlights the fact that there are myriad social relations and subtle social norms 

about property in land and in plants that are related to factors other than 

“ownership”: other homegarden research shows that even a person who “owns” or 

manages trees or plants in a homegarden might not have exclusive rights to them 

(Howard and Nabanoga, 2006). 

Lok 1 provided an example that demonstrates how rights to plants and trees may 

be circumscribed depending upon the species and upon who controls the zone in 

which they are planted, as well as how gendered rights to homegarden resources 

affect their structure and composition. She researched homegardens in a Mestizo


(mixed Spanish-Indian) community in north central Honduras and found 253 useful 

plant species in a sample of 10 gardens, with an average 60 species per garden 

distributed in nine management zones that could be discriminated by examining 

vertical strata, geophysical characteristics and “gender access or responsibility.” 

Women have responsibility for particular zones, such as the residential zone where 

ornamentals, vegetables, and medicinal plants are produced, as well as tree and plant 

nurseries. Men are in charge of the coffee zone, which provides much shade for 

vegetables and medicinal plants that women plant within it, but men simply 

‘tolerate’ these plants and eliminate them without their wives’ permission if they see 

fit in order to plant more coffee. Lok 1 concluded that the study of management zones 

“makes it possible to relate agroecological variables to social and economic ones, 

which is of great importance in homegarden analysis.” It is clear that one factor in 

this analysis is gendered rights to homegarden trees, plants and zones. 

4. GENDER AND COMMODITY PRODUCTION IN HOMEGARDENING 

Both the gender division of labor in homegardening and gendered rights to 

homegarden resources appear to be related to the control over income generated 

through cash crops. In the author’s experience with rural homegardens in Honduras, 

Nicaragua and El Salvador from 1982 – 1990, women use such income to pay for 

school fees, pharmaceuticals and medical services, and as “pocket money” for 

making daily purchases of food and other goods to meet household needs. 

Homegarden produce is available in small quantities year-round, so it is unsurprising 

that women are responsible both for such small daily purchases and for the 

production that provides the income for these purchases. The cases reviewed suggest 

that, when women market homegarden produce, they do so nearly exclusively in 

local markets, and the amount of income generated is generally quite small in 

relation to total household income. However, the amount generated can certainly be 

more substantial, as several homegarden studies across the globe attest. Finerman 

and Sackett (2003) reported that, in the Andes of Ecuador, women sell sufficient 

surplus from their homegardens to contribute to household income and improve 

their own status. 

Table 3 presents an overview of the gender division of labor in relation to the 

production of subsistence crops, of crops that are marketed on a small-scale, and of 

high-value crops or crops marketed on a larger scale for those case studies that 

provided information (15 of the 39 reviewed). It shows that women are more likely 

to manage crops destined principally for subsistence (in 80% of the cases) or for sale 

in small quantities in local markets (in 88%). As cash cropping occurs on a larger 

scale or high value crops are produced, men’s involvement and control are much 

more evident (86%). 

The associations between women and subsistence production are quite strong, as 

are the associations between women and medicinals, spices, condiments, and 

ornamentals. A typical example is found in Angel Peréz and Mendoza (2004) in 

relation to a traditional Totonac community in Veracruz, Mexico. They reported that 

women manage culturally important plants (for subsistence, ritualistic, and 

medicinal purposes) and are responsible for backyard gardens and orchards, whereas


men use homegardens to test and adapt exotic plants that they later introduce into 

commercial field crop production. Patterson (2000) found a similar pattern among 

the Kekchi Maya in Belize where homegarden cash crops have recently increased in 

number and are mainly introduced by male heads. 

Table 3. Responsibility for subsistence and cash crop production in homegardens by sex i in 

15 case studies across Latin America. 

Subsistence Small 

scale 

marketing 

M = male; F = female. 

Major or 

high value 

cash crops 

Countries (literature references: see 

Table 1 for author details) 

F - M Belize (29), Honduras (21), Mexico (3 ii , 

26), Panama (33) 

MF M Mexico (Alvarez Buylla et al., 1989) iii 

F F - 

Guatemala (18), Ecuador (11), Mexico 

(6, 13, 14, 22), Venezuela (16) 

MF - M Costa Rica 27) 

MF - MF Nicaragua (25) 

12 80.0% 7 87.5% 0 0.0% Total cases and percent of women in 

them 

3 20.0% 0 0.0% 1 14.3% 

Total cases and percent of both men and 

women 

0 0.0% 1 12.5% 6 85.7% Total cases and percent of men 

15 8 7 Total cases 

i 

Includes only cases where some garden produce is reported to be sold. 

ii 

Men experiment with cash crops destined eventually for agricultural fields. 

iii 

Implicit. Men manage fruit trees, and only citrus is sold in small quantities. 

Given the strong influence of women’s decision-making in homegardening 

across most of the cases in Latin America, it is interesting to examine whether 

commodity production plays a role in the gender division of labor in homegardens 

that are managed by both men and women. Table 4 presents the cross-tabulation of 

the 10 cases where data on the sex of the main gardener and the production of major 

market crops were both reported. In five of the six cases where both men and 

women share responsibility for homegardening, men produce major or high value 

cash crops in homegardens and, in one case, both men and women produce them. In 

only one case was it reported that, while men and women are main gardeners, no 

major or high value cash crops are produced; in another two cases, women are the 

main homegardeners but men manage high value crops. It is important to note that 

all of these cases refer to the Mesoamerican context: the South American cases 

presented insufficient data. 

A few studies have discussed what occurs in terms of shifting responsibilities 

for, and benefits from, homegardening when it begins to generate substantial 

amounts of cash income or cash crops are introduced, even in contexts where 

homegardening is culturally strongly associated with women. Murray (2001)

eported that, among the highland Maya of Chiapas, Mexico, people are quite 

dependent on the cash economy. Men emigrate and secure paid jobs, and much cash 

crop production occurs in homegardens where chemical inputs are also used. 

Commercial flower production is one of the activities that men have integrated into 

traditional homegardens. The strong integration into the market economy has 

undermined women’s economic and decision-making power in these households and 

gardens. As homegarden production becomes more lucrative or more marketoriented, 

women’s roles in them as managers, sellers, and earners of cash income 

appear to shift. Other studies report that commercialization may have negative 

effects with respect to agrobiodiversity and household food security. For example, 

Baleé (1994) reported that, in the eastern Amazon of Brazil, agricultural 

extensionists encouraged the production of rice as a cash crop. The result was that 

the space for traditional crops such as foods, spices, and other utilitarian plants was 

reduced to the point that these crops are no longer found in swiddens in these 

villages, which would obviously have a substantial impact on women. 

Table 4. Cross-tabulation of sex of the main gardener and responsibility for major or high 

i 

value cash crops in 10 case studies across Latin America . 

Sex of main 

gardener 


Subsistence Major or high 

value cash 

crops 

Countries (literature references: 

see Table 1 for author details) 

Men and women Men and 

women 

- Mexico (19) 

Women Women Men Belize (29), Mexico (3) ii 

Men and women Men Costa Rica (27), Honduras (21), 

Mexico (26), Panama (33) 

Men and women Men and 

women 

Nicaragua (25) 

i 

Ruonavaara (1996 – Guatemala) reported that both women and men manage homegardens 

and also reported small-scale marketing, but did not report who was responsible. 

ii 

Men only experiment with cash crops destined eventually for agricultural fields. 

Commercialization may leave women’s gardening responsibilities intact but may 

create other shifts that affect the composition and structure of homegardens and 

therefore agrobiodiversity and dietary composition. In Ecuador, dependency on 

global markets caused an economic crisis when the nation’s economy collapsed in 

the late 1990s. Residents of the village that Finerman and Sackett (2003) studied in 

the Andes have lost property and been forced to emigrate, so that “increasing 

number of homegardens lie abandoned by absentee landowners, or are plowed under 

to make way for cash crops that have done little, thus far, to ease the financial 

burdens of the owners.” 

Women are not necessarily marginalized when homegardens increase in 

economic importance; rather, both women and homegardens may provide a buffer 

against the worst effects of economic or environmental crisis affecting men’s 

agricultural production. Greenberg (1996) reported that, due to decreasing viability


of men’s agricultural production in rural areas of the Yucatan Peninsula, families 

have migrated to the tourist resort of Quintana Roo in search of wage labor. Men no 

longer engage in agricultural production, but women maintain traditional homegardens 

and agrobiodiversity in this urban setting, and homegardens generate cash 

for these families in many ways. Still, there may be other negative implications of 

shifting gender roles: this change in gender domains may partly account for social 

problems and men’s excessive drinking. 

Yet other trade-offs for women and their households must be considered, since 

production for subsistence and for cash income generation are certainly not the only 

measures of the value that homegardens provide. As many authors point out, 

homegardens are often sources of non-monetary exchange values through giftgiving 

and reciprocal exchange. These are very important especially to women 

gardeners as sources both of material goods and of status and social autonomy. 

5. SOCIAL STATUS, GENDER, AND GARDENS 

Much of the research that has been done on homegardens has emphasized the 

economic and ecological functions and benefits of homegardening and has stressed 

these as principal reasons for their creation and maintenance, without examining in 

any depth other ways in which homegardens provide social or material advantages 

for their owners. Even so, several of the articles reviewed acknowledge the social 

status that is associated with homegardening, especially with having a particularly 

large, beautiful or genetically diverse garden. As Sereni Murrieta and Winklerprins 

(2003) noted, a homegarden “says much about its keeper.” 

The same also appears to be true of swidden gardening. Descola’s (1994) work 

highlighted how researchers often mistakenly assume that the diversity that swidden 

gardeners create or maintain is due to ecological or economic motivations rather 

than to status-seeking behavior. At the same time, he showed how gardening may 

increase women’s status in the eyes of men. For the Achuar, it is a “point of honor” 

for women to cultivate large swidden gardens. The garden diversity evident, 

particularly in tubers, cannot be attributed to nutritional or culinary needs since 

“men - whose attitude openly encourages their wives’ agronomic capacities - 

recognize by taste alone only a very low proportion of the varieties of manioc, yams, 

or sweet potatoes.” Nor can it be attributed to the need to reduce species-specific 

diseases since only one serious manioc disease is recognized, and only a few plants 

are usually affected. Rather, “a woman who successfully grows a rich pallet of 

plants thereby demonstrates her competence as a gardener and fully assumes the 

main social role ascribed to women by proving her agronomic virtuosity” (Descola, 

1994). 

Finerman and Sackett (2003) found in the Ecuadorian Andes that people observe 

each other’s homegardens and deduce information about the owners’ wealth status, 

occupation and market orientation, as well as health status. The abundance and 

diversity of a garden is an important source of status for women who develop 

reputations as skilled gardeners whom people continually approach for planting 

materials, for advice and to exchange produce. Women boast about their 

homegardens and about the independence these afford. Yet the implications of


homegardening for women’s status are not only related to their production 

capacities; they are also clearly linked to the roles women are expected to perform as 

family caretakers and as representatives of their households. Homegardens reveal: 

the extent of the owner’s commitment to family well-being . . The presence of a garden 

rich in . . [medicinal plants] epitomizes her exertions on behalf of kin, and her 

proficiency as primary health provider; a spacious and productive garden filled with 

medicinal plants suggests that the family, too, is prosperous and fit . . Gardens 

themselves [are] a manifestation of the community’s most deeply held values: 

autonomy, status, religious piety, and personal investment in family. . A garden 

demonstrates a woman’s freedom from dependence on products from neighbors and 

commercial vendors; her fiscal standing evidenced by her ability to expend valuable 

land on a garden; her faith displayed by a sacrifice of resources to adorn the church; and 

her industriousness and devotion to family exhibited by her investment in plant 

cultivation (Finerman and Sackett, 2003). 

It is clear that the status provided through gardening is not confined to gardens’ 

visible characteristics or the skills of their owners. Many studies show that garden 

produce that is not consumed is much more commonly given as gifts or exchanged 

with others rather than sold in markets, and most homegarden studies also report that 

the vast majority of garden planting materials that are not self-provisioned are 

acquired through gifts and exchange (Blanckaert et al., 2004), predominantly 

between women (Boster, 1985b; Alvarez Buylla et al., 1989; Hoffman, 1993; 

Greenberg, 1996; Lerch, 1999; Patterson, 2000; Ruonavaara, 1996; Finerman and 

Sackett, 2003; Sereni Murrieta and Winklerprins, 2003). Such exchanges are not 

only important in terms of the garden products or planting materials that gardeners 

access – they are just as important as a means to create and maintain social 

networks. Gift giving and exchange of planting materials often help Mayan women 

maintain kinship and neighborly ties with people in distant places (Greenberg, 1996) 

and provide additional opportunities to accumulate knowledge (Patterson, 2000). 

Likewise, Finerman and Sackett (2003) referred to women’s plant “borrowing” in 

the Andes as an important basis for household exchanges, which are most common 

among female relatives and close friends. Lerch (1999) researched homegarden 

plant diversity and exchange in the Amazon where networks for exchange of 

indigenous planting material have been strong historically. In the villages she 

studied, reciprocal exchange among neighbors (who might also be kin) was the most 

important source of plant material acquisition, and households with high plant 

diversity exchanged plants at a higher rate. 

Among Amazonia Amerindians, male prestige is often related to ceremonial 

exchange of food products such as manioc beer (Descola, 1994; Heckler, 2004; 

Thompson, 1977). Women may also gain prestige as producers of the crops that men 

exchange as occurs among the Cubeo (Goldman, 1963), the Achuar (Descola, 1994) 

and the Piaroa (Heckler, 2004). Among Piaroa groups, women manioc cultivators 

can assert themselves as agronomic experts, which is evident in the great diversity of 

manioc cultivars they produce. They create alliances by exchanging this diversity, as 

well as by processing manioc in “processing parties” which are events of 

“communality and congeniality” in which women gain prestige as hard workers and 

food providers (Heckler, 2004).


To the degree that women’s status is positively affected by their homegardening 

activities, their status may erode as homegardening itself declines. Stavrakis (1979) 

noted in the villages that she studied in Belize how kitchen gardens lost prestige as 

people began to reject local fruits and vegetables in favor of imported varieties, and 

gardens became obsolete. “As the garden loses its social value, so naturally do 

women’s gardening activities.” Aikman (1999) and Hoffman (1993) argue that 

women’s traditional knowledge and management of local crop diversity that they 

maintain in home or swidden gardens may become valueless, and their high social 

status turn to social stigma, as such knowledge and production become increasingly 

associated with poverty and backwardness. 

6. KNOWLEDGE AND GENDER IN HOMEGARDENING 

The status derived from gardening is in part due to the knowledge and skills that are 

necessary to create and maintain them. Depending on the degree to which gardening 

knowledge is specialized, it will be unevenly distributed and this distribution will 

reflect factors such as age, sex roles, and differential ‘opportunities to learn’ (Boster, 

1985a). That gardening knowledge is specialized is widely reported in the literature 

reviewed. To the degree that the species diversity in homegardening is greater than 

in agricultural fields, this implies greater breadth of ethnobotanical and agronomic 

knowledge than what is common in agricultural production. Further, because so 

many species and varieties are intercropped in homegardens, knowledge of plant 

associations is also likely to be greater. Such associations are also very likely to be 

related to microclimates that are created within homegardens and that do not exist 

elsewhere, which “enables: (i) the growing of varieties with different climatic 

requirements . . . (ii) the elaboration of a management calendar independent of the 

climatic functions, and (iii) the experimentation with new varieties” (Alvarez Buylla 

et al., 1989; endnote 1). 

When Benjamin (2000) examined Mayan cultural homegarden practices in 

depth, she found that women homegardeners’ knowledge is based on “principles” 

that maximize micro-environmental conditions for successful plant propagation, 

which are passed on across generations. Similarly, Gillespie et al. (2004) found that 

Mayan women’s management of Ramón trees (Brosimum alicastrum), a dry-season 

forage source found in all homegardens in their study area in the Yucatan, is based 

upon an intimate knowledge of environmental factors that are taken into account 

when propagating the species, where their management techniques were found to 

increase growth very substantially. 

Gardening knowledge is not confined to agroecology and agronomy. Garden 

planning for subsistence purposes must combine an understanding of vegetative 

cycles, of perishability and processing and storage characteristics, and of timing and 

quantity of demand, including the needs for ingredients for specific dishes and/or 

medicines and substitutability of those ingredients (Lope Alzina, 2006), and of 

the need to meet nutritional and medicinal requirements of households whose 

composition also changes over time. Finerman and Sackett (2003) showed that 

the composition of homegardens in their study area closely reflects the stage in the 

life cycle, where the medicinals produced reflect in part the specific needs of young


families or elderly household members. Several other researchers confirm that 

homegardens provide the basis for acquiring much environmental, agronomic, 

cultural and other knowledge related to plants and plant uses (Alvarez Buylla et al., 

1989; Angel Peréz and Mendoza, 2004; Greenberg, 1996). 

Indigenous knowledge associated with gardening is also related to plants as 

cultural capital, where individual plants take on social meaning. For example, Sereni 

Murrieta and Winklerprins (2003) found that women were able to relate the history 

of many individual plants, their origins, utility, and their status as a gift, a symbol of 

someone’s affection or a commemoration of an event (see also Finerman and 

Sackett, 2003). Much ritualistic knowledge may also be entailed in gardening as 

Descola’s (1994) work on the Achuar amply testifies. 

It can therefore be presumed that knowledge entailed in managing complex 

gardens takes a considerable part of a lifetime to accrue, involves considerable 

hands-on experience and trial and error (experimentation), and entails continual 

exchange of information. It is clear that, across the region, women are more often 

homegardening specialists; as principle knowledge holders, it can be hypothesized 

that it is also women who are primarily responsible for the transmission of 

homegardening knowledge. In other words, homegardening knowledge and 

knowledge transmission are largely gender-related. 

There is ample testimony to gendered gardening knowledge in the literature 

reviewed in this chapter. In one case where women are nearly exclusively 

responsible for homegardening, it was reported that “men generally disavow any 

knowledge of homegardening, deferring to their wives for even basic information 

about gardens and their products” (Finerman and Sackett, 2003). In another such 

case, Descola (1994) came to an even more dramatic conclusion: Achuar men “are 

. . . totally incapable of replacing their wives if the need arises, and moreover have no 

desire to do so. When a man no longer has any woman (mother, wife, sister, or 

daughter) to cultivate his garden and prepare his food, he has no choice but to kill 

himself.” 

However, it is more common that both men and women have homegardening 

knowledge and that the division of such knowledge reflects the nature of their 

involvement. Such a conclusion is born out by an unusual study 2 that researched 

gendered species knowledge in 23 households in the Nicoya Peninsula of Costa 

Rica, where both women and their husbands participate in homegardening. The 13 

homegarden species most frequently used were selected: four exclusively for 

medicinal use, five for medicinal and food use, and four exclusively for food. The 

results showed that women’s knowledge of medicinal plants was always higher than 

men’s. Regarding food plants, only for Musa spp. (plantains and bananas) did the 

knowledge between men and women differ significantly, where men had greater 

knowledge than women. The authors related these findings to the gender division of 

labor where women were responsible for health care and food preparation and men 

for cash crop production, and six out of nine of the food items studied had 

commercial values. 

It is also important to stress that gardening knowledge, like ethnobotanical or 

ethnobiological knowledge in general, varies not only between men and women, but 

as well according to factors such as kinship, age, social class, ethnicity,


specialization, and personal propensity (Howard, 2003). As Greenberg (1996) 

reported among the Maya of Quintana Roo, Mexico, “There are individual 

differences in the intensity of peoples’ interests in plants and their cultivation.” 

When examining knowledge transmission networks and processes of knowledge 

erosion, the influence of kinship and age also comes to the fore. Knowledge 

transmission is a dynamic and continual process since household circumstances and 

ecological and economic conditions change continually, and homegardens must be 

adapted to such changes. In this, women and their social networks play a 

predominant role. Several authors show that homegardening knowledge is transmitted 

largely among women and then principally among closely related kin (Boster, 

1985a; Descola, 1994; Greenberg, 1996; Keys, 1999; Patterson, 2000). Children’s 

labor in homegardening is so common that it is not surprising that much general 

knowledge is transmitted to them as they work under the supervision of their 

mothers. Keys (1999), whose research specifically focused on homegarden 

knowledge transmission among the Kaqchikel Maya of Guatemala, showed how 

homegardens act as veritable classrooms for both girls and boys where women teach 

children how to use farm tools and to cultivate and manage crops. What boys learn 

is not only applicable to the homegarden. Keys observed that boys have already 

learned the basic concepts of cultivation before they accompany their fathers to the 

milpa. Not only cultivation techniques, but as well knowledge about the use of 

plants for food, medicine and handicrafts, are transmitted from mother to child 

through homegardens. Patterson (2000), working within Mayan communities in 

Belize, showed that it was often not only mothers, but also other female relatives, 

who formed key knowledge transmission networks. All gardeners interviewed stated 

that they acquired environmental and homegarden management knowledge from 

older female family members, whereas 94% said they also acquired environmental 

knowledge from “other” female family members including younger sisters and more 

distant relations. The process of knowledge transmission begins at age five or six 

when girls accompany their female relatives to gardens where they learn to identify, 

water, and harvest or collect plants and to tend small animals. Hoffman (1993) 

found that not only was gardening knowledge transmitted between mothers and 

daughters: plant material as well as knowledge, material, skills and practices were 

often part of a “package” of cultural and physical capital that flows among women 

and between women and their offspring. Boster (1985b) also reported this for 

women Aguaruna manioc cultivators in northern Peru. 

Some homegarden researchers remark that homegarden knowledge is eroding or 

is likely to erode in the near future. They provide several reasons for this, some of 

which are gender-specific. One is cultural erosion: as young people assimilate into a 

dominant culture through education and migration, they learn less about plants and 

homegardening (Angel Peréz and Mendoza, 2004). Benjamin (2000) cites 

emigration among Mayan youth as the main risk. Keys (1999) pointed out that 

particularly young Guatemalan women are affected by off-farm employment in 

textile factories, which leaves them no time for homegardening, and Hoffman 

(1993) stressed not only off-farm employment, but also migration and participation 

in formal educational systems that denigrate women’s traditional gardening 

practices, which leads to loss of knowledge.



Many women in Latin America contribute to subsistence and to meeting the cash needs 

of their families but usually do so in ways that are not permitted to be predominant or 

very visible. Swidden gardening and other work that Amazonian Amerindian women 

perform accord them social status and prestige. Although men mainly clear and allocate 

swidden garden land to women, women have strong if not exclusive claims to most 

swidden garden resources. Many Mayan and mestizo women are far less likely than men 

to own property, and are generally not permitted to engage in agricultural production or 

to generate substantial amounts of income. Still, as ‘acceptable’ social and environmental 

spaces where domesticity is centered and esteemed, homegardens offer these women 

sources of authority, autonomy, status, social networks and visible ‘public’ spaces of 

recognition without challenging male dominance. Homegardens are clearly essential to 

women: they fit in well with their domestic duties, labor patterns, productive decisionmaking 

spheres, aesthetic sensibilities, and cultural roles. Through gardening, women 

develop great knowledge and proficiency in relation to the plant world and to the 

environment, which permits them to shape and manage these to meet the needs of their 

households. In addition to utilitarian or monetary values, homegarden species have 

deeper, spiritual emotional, and symbolic meaning for women whose spaces and 

relations are circumscribed by historically and culturally-specific phenomena that 

relegate them to subordinate positions; they are also assertions (and continual 

reassertions) of women’s importance, contributions and the continuity of traditions and 

identity that they bring to their societies, families and communities. In this, homegardens 

serve men as much as they serve women. They permit women to contribute to family 

subsistence, status, and identity in ways that are ‘respectable’. 

Women can more readily enter markets where they do not compete with men or 

when they do not earn so much income that they challenge men’s economic 

predominance. Beyond this, they also appear to be able to negotiate change with 

husbands and other family members based upon their authority as garden managers. 

The terms of such negotiations may be restrictive, but they may also afford women 

the ability to meet their own particularistic needs, and may contribute positively to their 

status and increase their ability to “have a greater say” in the management of their 

households and communities. On the other hand, this does not negate the fact that 

women’s command over homegarden resources is tenuous and likely to shift as 

commercialization increases, and they may also lose access altogether in the case of 

separation or divorce. Homegardening may also be seen as a source, or a continuing 

reminder, of women’s subordinate status, and change processes may leave them 

bereft of control over or access to homegarden resources. 

Threats to homegardening are many – as are the driving forces to maintain them 

that are mentioned above. Commoditization, the decreasing status of local 

agrobiodiversity in human consumption and health due to acculturation, urbanization 

which draws youth away from primary production, and formal education that denigrates 

‘traditional’, ‘peasant’ or ‘indigenous’ ways of life, can all be major threats since they 

subvert many of the dynamics that have maintained the value of homegardens – 

diversity, independence and autonomy, cultural identity, local adaptability, home- and 

needs-centeredness, and multi-value production. Many of these threats at the same time


may offer women greater formal equality and autonomy. But it is likely that women will 

attempt to negotiate the trade-offs between such potential gains and losses – as ample 

literature on homegardens among urban migrant populations attests, women are very 

likely to continue to exert every effort to create, maintain and manage the most socially 

and agroecologically complex systems known to the region. 

ENDNOTES 

1. Lok R. 2001. A better understanding of traditional homegardens through the use 

of locally defined management zones. Indigenous Knowledge and Development 

Monitor 9: 14 – 18. 

2. Ochoa L., Fassaert C., Somarriba E. and Schlonvoight A. 1998. Conocimiento 

de mujeres y hombres sobre las especies de uso medicinal y alimenticio en 

huertos caseros de Nicoya, Costa Rica. Agroforestería en las Américas 5: 

17 – 18, 7 – 11. 

3. Samaniego G. and Lok R. 1998. Valor de la percepción y del conocimiento 

local de indígenas Ngöbe, en Chiriqui, Panamá. Agroforestería en las Américas 

5: 17 – 18, 12 – 16. 

4. Veth B. and Reinders M. 1995. Planten zijn als mensen: genezen met 

medicinale planten bij de Warao en Wayana in de Guyana's. Indigo 3(5): 4 – 7. 

REFERENCES 

Aikman S. 1999. Schooling and development: Eroding Amazon women’s knowledge and 

diversity. In: Heward C. and Bunwaree P. (eds), Gender, education and development: 

Beyond access to empowerment, pp. 99 – 153. Zed Press, London. 

Alexiades M.N. 1999. Ethnobotany of the Ese Eja: Plants, health and change in an 

Amazonian Society (Plant Medicinals). PhD dissertation, City University of New York. 

University Microfilms International, Ann Arbor, Michigan. 

Alvarez Buylla M., Lazos Chavero M.A. and García Barrios J.R. 1989. Homegardens of a 

humid tropical region in South East Mexico: An example of an agroforestry cropping 

system in a recently established community. Agroforest Syst 8: 133 – 156. 

Angel Peréz, A. del and Mendoza B., M.A. 2004. Totonac homegardens and natural resources 

in Veracruz, Mexico. Agric Hum Val 21: 329 – 346. 

Baleé W.L. 1994. Footprints of the forest: Ka’apor ethnobotany–The historical ecology of plant 

utilization by an Amazonian people. Columbia University Press, New York pp. 50 – 60; 158. 

Benjamin T.J. 2000. Maya cultural practices in Yucatecan homegardens: An ecophysiological 

perspective. PhD dissertation, Purdue University. University Microfilms International, 

Ann Arbor, Michigan. 

Blanckaert I., Swennen R.L., Paredes Flores M., Rosas Lopez R. and Lira Saade R. 2004. 


Coxcatlán, Valley of Tehuacán-Cuicatlán, Mexico. J. Arid Env 57: 39 – 62. 

Boster J.S. 1985a. Requiem for the omniscient informant: there’s life in the old girl yet. In: 

Dougherty J. (ed.), Directions in cognitive anthropology, pp. 177 – 198. University of 

Illinois Press, Champaign, Illinois. 

Boster J.S. 1985b. Selection for perceptual distinctiveness: Evidence from Aguaruna cultivars 

of Manihot esculenta. Econ Bot 39: 310 –325.


Bruce J. and Fortmann L. 1988. Why land tenure and tree tenure matter: Some fuel for 

thought. In: Fortmann L. and Bruce J.W. (eds), Whose trees? Proprietary dimensions of 

forestry, pp 1 – 9. Westview Press, Boulder and London. 

Deere C.D. and Leon M. 2001. Empowering women: Land and property rights in Latin 

America. University of Pittsburgh Press, Pittsburgh pp. 264 – 291. 

Descola P. 1994. In the society of nature: A native ecology in Amazonia, pp 93 – 212. 

Cambridge University Press, Cambridge. 

Doxon L.E. 1988. Diversity, distribution and use of edible and ornamental plants in 

homegardens in Honduras. PhD Dissertation, Kansas State University. University 

Microfilms International, Ann Arbor, Michigan. 

Dufour D.L. 1981. Household variation in energy flow in a population of tropical forest 

horticulturalists. PhD dissertation. State University of New York at Binghamton. 


Finerman R. and Sackett R. 2003. Using homegardens to decipher health and healing in the 

Andes. Med Anth Quart 17: 459 – 482. 

Gillespie A.R., Bocanegra Ferguson D.M. and Jiménez Osornio J.J. 2004. The propagation of 

Ramón (Brosimum alicastrum Sw.; Moraceae) in Mayan homegardens of the Yucatan 

peninsula of Mexico. New Forests 27: 25 – 38. 

Goldman I. 1963. The Cubeo: Indians of the North-west Amazon. University of Illinois Press, 

Champaign, Illinois. 

Govers C. 1997. Agriculture, cosmology and the gender division of labour in a Totonac 

highland village (Mexico). In: de Bruijn M. and van den Hombergh H. (eds), Gender and 

land use: Diversity in environmental practices, pp. 27 – 47. Thela Publishers, Amsterdam. 

Greenberg L.S. 1996. You are What you eat: ethnicity and change in Yucatec immigrant 

house lots, Quintana Roo, Mexico. PhD dissertation, University of Wisconsin-Madison. 


Heckler S.L. 2001. The ethnobotany of the Piaroa: Analysis of an Amazonian people in 

transition. PhD dissertation, Cornell University. University Microfilms International, Ann 

Arbor, Michigan. 

Heckler S.L. 2004. Tedium and creativity: The valorization of manioc cultivation and Piaroa 

women. J Roy Anth Inst 10: 241 – 259. 

Hoffman S.D. 1993. Subsistence in transition: Indigenous Agriculture in Amazonas, 

Venezuela. PhD Dissertation, University of f California Berkeley. University Microfilms 

International, Ann Arbor, Michigan. 

Howard P. 2003. Women and plants: An exploration. In: Howard P. (ed.), Women and plants: 

Gender relations in biodiversity management and conservation, pp. 1 – 47. Zed Press and 

Palgrave Macmillan, London and New York. 

Howard P. and Nabanoga G. 2006. Are there customary rights to plants? An inquiry among 

the Baganda (Uganda), with special attention to gender. World Devel (in press). 

Irvine D. 1987. Resource Management by the Runa Indians of the Ecuadorian Amazon. PhD 

dissertation, Stanford University. University Microfilms International, Ann Arbor, Michigan. 

Keys E. 1999. Kaqchikel gardens: Women, children, and multiple roles of gardens among the 

Maya of Highland Guatemala. Yearb Conf Lat Am Geogr 25: 89 – 10. 

Knauft B.M. 1997. Gender identity, political economy and modernity in Melanesia and 

Amazonia. J Roy Anth Inst 3: 233 – 259. 

Lazos Chavero E. and Alvarez Buylla M. 1988. Ethnobotany in a tropical-humid region: The 

homegardens of Balzapote, Veracruz, Mexico. J Ethnobio 8: 45 – 79. 

Lerch N.C. 1999. Homegardens, cultivated plant diversity, and exchange of planting material 

in the Pacaya-Samiria National reserve area, Northeastern Peruvian Amazon. Masters 

thesis, McGill University. University Microfilms International, Ann Arbor, Michigan.


Lope Alzina D. 2006. Gendered production spaces and crop varietal selection: a case study in 

Yucatan, Mexico. Sing. J. Trop. Geog (in press). 

Madaleno I. 2000. Urban agriculture in Belém, Brazil. Cities 17: 73 – 77. 

Martin G.J. 1996. Comparative ethnobotany of the Chinantec and Mixe of the Sierra Norte, 

Oaxaca, Mexico. PhD dissertation, University of California, Berkeley. University 


Mendez V.E., Lok R. and Somarriba E. 2001. Interdisciplinary analysis of homegardens in Nicaragua: 

Micro-zonation, plant use and socioeconomic importance. Agroforest Syst 51: 85 – 96. 

Michon G. 1983. Village forest gardens in West Java. In: Huxley P.A. (ed.), Plant research and 

agroforestry, pp. 13 – 24. International Centre for Research in Agroforestry, Nairobi, Kenya. 

Montagnini F. 2006. Status of homegardens in Mesoamerica In: Kumar B.M. and Nair P.K.R. 

(eds), Tropical homegardens: A time-tested example of sustainable agroforestry, pp. 61– 

84. Springer Science, Dordrecht.. 

Murray S.J. 2001. Plants in the ‘Patxokon Na’: Tzotzil Maya homegardens in the highlands of 

Southeastern Mexico. PhD dissertation, Wayne State University. University Microfilms 


Palacio J. 1980. Food and social relations in a Garifuna village (Belize). PhD dissertation, 

University of California at Berkeley. University Microfilms International, Ann Arbor, 

Michigan. 

Patterson M.L. 2000. Agroforestry in Belize: Maya homegardens in San Lucas. Masters 

thesis, University of Alberta. University Microfilms International, Ann Arbor, Michigan. 

Posey D.A. 1984. A preliminary report on diversified management of tropical forest by the 

Kayapó Indians of the Brazilian Amazon. In: Prance G.T. and Kallunki J.A. (eds), 

Ethnobotany in the neotropics: Proceedings. Ethnobotany in the Neotropics Symposium, 

Society for Economic Botany (13-14 June 1983), Oxford, Ohio. Adv Econ Bot 1: 112 – 

126. New York Botanical Garden, New York. 

Roquas E. 2002. Stacked law. Land, property and conflict in Honduras. PhD dissertation, 

Wageningen University. Rozenberg Publishers. Amsterdam. 

Ruonavaara D.L. 1996. Traditional household gardens of the Petén, Guatemala. Masters thesis, 

Michigan State University. University Microfilms International, Ann Arbor, Michigan. 

Salick J. 1997. Subsistence and the single woman among the Amuesha of the Upper Amazon, 

Peru. In: Sachs C. (ed.), Women working the environment, pp. 139 – 153. Taylor and 

Francis, Washington DC. 

Sereni Murrieta R.S. and Winklerprins A. 2003. Flowers of water: Homegardens and gender roles 

in a riverine Caboclo community in the lower Amazon, Brazil. Cult Agri 25: 35 – 47. 

Stavrakis O. 1979. The effects of agricultural change upon social relations in a village in 

Northern Belize, Central America. PhD dissertation, University of Minnesota. University 


Thompson S. 1977. Women, horticulture, and society in tropical America. Am Anthrop 79: 

908 – 910. 

Uzendoski M.A. 2004. Manioc beer and meat: Value, reproduction and cosmic substance 

among the Napo Runa of the Ecuadorian Amazon. J Roy Anth Inst 10: 883 – 902. 



agroforestry, pp. 13 – 24. Springer Science, Dordrecht. 

Wilson W.M. 1997. Why bitter cassava (Manihot esculenta Crantz)? Productivity and 

perception of cassava in a Tukanoan Indian settlement in the Northwest Amazon 

(Colombia). PhD dissertation, University of Colorado at Boulder. University Microfilms 


Winklerprins A. 2002. House-lot gardens in Santarém, Pará, Brazil: linking rural with urban. 

Urb Ecosyst 6: 43 – 65.

SECTION 3 

SOME NEW THRUST AREAS

CHAPTER 11 

CARBON SEQUESTRATION POTENTIAL 

OF TROPICAL HOMEGARDENS 

B.M. KUMAR 

Department of Silviculture and Agroforestry, College of Forestry, Kerala 

Agricultural University, Thrissur 680656, Kerala, India; E-mail: 

 

Keywords: Carbon stocks, Multistrata systems, Net primary productivity, Soil organic 

matter. 

Abstract. This chapter examines the premise that tropical homegardens have a special role in 

carbon (C) sequestration because of their ability for carbon storage in the standing biomass, 

soil, and the wood products. In doing so, it analyzes the potential for C storage in 

homegardens and the role of homegardens in reducing CO 2 concentration in the atmosphere. 

Lack of reliable inventories/estimates and uncertainties in the estimation of C sequestration 

potential of homegardens present formidable difficulties in the analysis. Nevertheless, 

available information indicates that homegardening has a higher potential to sequester C 

compared to monospecific production systems, and the costs are lower than emission 

reduction or sequestration by other means. Indeed, the C sequestration potential of 

homegardens that mimic the structure and diversity of mature evergreen forest formations is 

comparable to that of such forest stands. Although experimental evidence suggests that 

species diversity does not necessarily mean high C sequestration, complementary or 

compensatory gains in resource acquisition, possibility of biological N 2 fixation and the 

relatively low herbivory pressure, may explain this high C sequestration ability of 

homegardens. Extension of homegardens into more lands and adaptive management of the 

existing gardens offer scope for enhanced C sequestration and economic gains. 


The Intergovernmental Panel on Climate Change (IPCC) Special Report on Land 

Use, Land Use Change and Forestry (LULUCF) suggests that the average annual 

accounted carbon stock changes in the first commitment period (2008–2012), 

resulting from afforestation and reforestation, would be between 197 and 584 Tg C 

year –1 (Watson et al., 2000). Agroforestry, including the homegarden, plays a 

cardinal role in this respect—net changes in global C stocks are estimated to be 26 

185 




186 B.M. KUMAR 

Tg C year –1 for better agroforest management and 390 Tg (million tons) C year –1 

r for 

agroforestry-related land use changes in 2010 (Watson et al., 2000). Information 

relating to this substantial, but under-exploited potential of agroforestry as a carbon 

sequestration strategy has been recently reviewed by Albrecht and Kandji (2003) 

and Montagnini and Nair (2004). This chapter is a follow up to these reports, with 

the objective of examining the role of tropical homegardens as a mechanism for 

carbon sequestration, on which presently little or no concrete information exists. 

Furthermore, an attempt is made to examine carbon restitution (above- and 

belowground) as a function of species richness in homegardens, using biomass 

productivity as a ‘proxy’ of carbon sequestration by comparing several woody 

perennial-based polycultures. 

2. EXTENT OF HOMEGARDENING 

Although homegardens are an age-old practice in many parts of the tropics and even 

other parts of the world (Kumar and Nair, 2004; Nair and Kumar, 2006), only 

limited data are available on their extent and distribution. The available information 

suggests that Indonesian homegardens or pekarangan cover about 5.13 million ha of 

land, of which 1.74 million ha are in Java 1 . Homesteads cover about 0.54 million ha 

in Bangladesh 2 and 1.05 million ha in Sri Lanka 3 (which constitutes about 60% of 

the land holdings

CARBON C SEQUESTRATION 

S POTENTIAL OF HOMEGARDENS 

187 

than 2000 g C m –2 

m yr –1 , is imperative; Kaye et al., 2000), carbon conservation 

(easing of anthropogenic pressure on existing stocks of C in forests through 

conservation and management efforts), and carbon substitution (substitution of 

energy demand materials by renewable natural resources, fuelwood production, 

increased conversion of biomass into durable wood products for use in place of 

energy-intensive materials; Kürsten, 2000). 

While most agroforestry systems are important in respect to one or the other 

mechanisms mentioned above (Ruark et al., 2003), the homegardens perhaps are 

unique in that all three mechanisms are relevant; i.e., they sequester C in biomass 

and soil, reduce fossil-fuel burning by promoting woodfuel production, help in the 

conservation of C stocks in existing forests by alleviating the pressure on natural 

forests and ensure greater synergy with the Convention on Biological Diversity 

(CBD). Moreover, there is no complete removal of biomass from the homegardens 

(Gajaseni and Gajaseni, 1999), signifying the permanence of the system. While lack 

of stability or permanence of the C sequestered being a major concern in LULUCF 

C sequestration projects (UNFCCC, 2002), the homegarden system is remarkably 

resilient. Additionally, C storage can last for decades if boles, stems or branches are 

processed in any form of long-lasting products (Roy, 1999) and the homegarden 

system has reasonable prospects in that respect too (Kumar et al., 1994). 

Available reports and case studies on biomass production/carbon sequestration 

potential of tropical homegardens are summarized in Table 1, and are compared 

with other tropical land use systems. One particular problem, however, is the 

profound age-related variations in the C stocks of different land use activities. 

Although Tomich et al. (2002) suggested that time-averaged C stocks (e.g., half the 

system’s C stock at its maximum age or rotation length) will be appropriate to 

compare C stocks of different land use systems on a scale that adjusts C stocks of 

the systems to their ages, adequate information on this aspect is not available in 

many case studies. Notwithstanding such intrinsic variability and assuming that 

homegardens are “steady-state systems” (Kumar and Nair, 2004), Table 1 is an 

attempt to compare different systems. As expected, the multi-layered woody 

perennial dominated systems have higher C sequestration potentials than other 

comparable systems. For example, the Javanese and Sumatran homegardens 

accumulated C in the range of 55.8 to 162.7 Mg ha –1 [19 Sumatran homegardens of 

12 to 17 years age (Roshetko et al., 2002) and a Javanese garden of undefined age 

(Jensen, 1993)], which is considerably greater than monocultures of annual crops, 

most woodlots and simple agroforests (with one dominant species such as oil palm, 

cacao or coffee). Likewise, the data show that a shift from single-crop production 

systems to multistrata systems increased the C sequestration potential. For example, 

conversion of all “sun-coffee” to “shade coffee” systems in Sumatra increased 

average landscape level C stocks by an estimated 10 Mg C ha −1 during a 20-year 

period (van Noordwijk et al., 2002). Monospecific woodlots also accumulate 

substantial C in their biomass – which is, however, dependent on the species, site, 

and management (Table 1). 

In certain cases, the aboveground C in homegardens is on par with the C stocks 

reported for similar-aged secondary forests (e.g., Jensen, 1993); but lower than that

Land use practice Duration 

[years] 

Table 1. Carbon uptake rates and carbon stocks of prominent land use systems in the tropics 1 . 

C uptake [Mg 

C ha –1 yr –1 ] 

C stocks [Mg 

C ha 1 ] 1 

Remarks and source 

Primary and logged forest Unknown 0 192 to 276 Sum of above-and below-ground C; summary of 

116 sites within different land uses before and 

after slash and burn from Brazil, Cameroon, 

Indonesia and Peru (time averaged system C 

stocks; Sanchez, 2000) 

Cropping after slash and burn 2 –76 to –112 39 to 52 As above 

Crops/bush fallow 4 2 to 4 32 to 36 As above 

Tall secondary forest fallows 23 5 to 9 95 to 142 As above 

Complex agroforests 25 to 40 2 to 4 65 to 118 As above 

Simple agroforests 15 5 to 9 65 to 92 As above 

Pastures, Imperata grasslands 4 to 12 –0.2 to –0.6 27 to 31 As above 

Indonesian homegardens, Sumatra 13.4 - 107.2 ± 37.2 

[range: 55.8 to 

162.7] 

Sum of aboveground, litter, herbs, soil and roots 

(35.3, 2.0, 0.3, 60.8 and 8.8 Mg C ha –1 

respectively; Roshetko et al., 2002) 

Javanese homegarden, Legokole - - 63 Total biomass (including 4.4 Mg ha –1 of ground 

litter; Jensen 1993) were converted to carbon 

stocks by multiplying with 0.5 

– 

188 

B.M. KUMAR

Forest remnants, Lampung, Indonesia - - 262 Time averaged total C stock above –0.3m in the 

soil (van Noordwijk et al., 2002) 

Shade-coffee, Lampung, Indonesia - - 82 As above 

Sun-coffee (monoculture), Lampung, - - 52 As above 

Indonesia 

Monospecific stands of Casuarina 

equisetifolia, Eucalyptus robusta and 

Luecaena leucocephala, Puerto Rico 

Woodlots off Acacia auriculiformis, 

Ailanthus triphysa, Artocarpus 

heterophyllus, Artocarpus hirsutus, 

Casuarina equisetifolia, Emblica 

officinalis, Leucaena leucocephala, 

Paraserianthes falcataria and 

Pterocarpus marsupium, Kerala, 

India 

Amazonian forests: Terra firme, Tall 

Caatinga and Tall Bana forests 

4 - 128.3, 115.7 

and 116.9 

respectively 

Sum of aboveground, litter, herbs, soil and root 

C; unplanted control had 83.2 Mg C ha –1 

(Parrotta, 1999) 

8.8 - 26.3 to 178.4 Sum of bole, branch, foliage, roots (Kumar 

et al., 1998) and detrital C (Jamaludheen and 

Kumar, 1999): converted to carbon stocks by 

multiplying with 0.5 

Unknown - 152, 178 and 

155 respectively 

Above + belowground C (Cuevas and Medina 

1986) 


CARBON 

C SEQUESTRATION 


189

Land use practice Duration 

[years] 

C uptake [Mg 

C ha –1 y –1 ] 

C stocks [Mg 

C ha –1 ] 

Remarks and source 

Natural forests, Jambi, Indonesia 120 - 500 Aboveground C: (cited from Roshetko et al., 

2002) 

Central American lowland forests Unknown 146 Average for six forest types s (above + 

belowground: 114 + 32 Mg ha –1 ); Sanford and 

Cuevas (1996) 

Mature agroforests, Sumatra 30 - 101 Aboveground C: Roshetko et al. (2002) 

Secondary forests, Sumatra 30 - 86 As above 

Young agroforests, Sumatra 9 - 14 As above 

Cassava, Sumatra 0.3 - 0.5 As above 

Cordia alliodora + cacao, Turrialba, 

Costa Rica 

initial 

5 

10 

98 (soil) 156.6 

213.8 

Perennial C stock [i.e., soil + tree + cacao (cacao 

branches and stems), tree stems, estimated 85% 

of roots (coarse root proportion), and Cordia 

branches)]; Beer et al. (1990) cited in 

Montagnini and Nair (2004) 

190 B.M. KUMAR

Erythrina poeppigiana + cacao, 

Turrialba, Costa Rica 

initial 

5 

10 

220.8 

Mature forests, Mekoe, Cameroon Unknown 270 Biomass (sum of tree, understorey, litter and 

roots; Duguma et al., 2001) converted to carbon 

stocks by multiplying with 0.5 

Cacao agroforests, Mekoe, Cameroon 26 152 As above 

115 

159.8 

As above; except for Cordia branches 

Tectona grandis plantation, Panama 20 120.2 Sum of above (104.5 Mg ha –1 ) and belowground 

C (15.7 Mg ha –1 ); Kraenzel et al. (2003) 

Food crop fields, Mekoe, Cameroon - 43 Biomass (sum of tree 2 , understorey, and roots; 

Duguma et al., 2001) converted to carbon stocks 

by multiplying with 0.5 

1 Also, see Schroeder (1994) who estimated the average carbon storage by agroforestry practices as 9, 21, 50, and 63 Mg C ha −1 for semiarid, subhumid, 

humid, and temperate regions and Albrecht and Kandji (2003) who reported a value between 12 and 228 Mg C ha −1 with a median value of 95 Mg ha −1 . 

2 When land was cleared, indigenous fruit, medicinal and timber trees (e.g., Ricinodendron heudelotii, Cola nitida, Voacanga africana, Triplochiton 

sclerozylon etc.) were deliberately retained. 

CARBON 



191

192 

accumulated by the mature forests in the region (114 to 500 Mg aboveground C ha –1 

Table 1). Indeed, the homegardens resemble young secondary forests in structure 

and biomass accumulation and may be considered as a human-made forest kept in a 

permanent early successional state with considerable productive potential. Consistent 

with this, in a study on Kerala homegardens, Kumar et al. (1994) showed that the 

average standing stock of commercial timber ranged from 6.6 to 50.8 m 3 

ha –1 . Overall, the data presented shows that the homegardens that mimic the 

structure and diversity of mature evergreen forest formations (Fig. 1) rank very close 

to mature forests in their biomass C storage potential (Table 2). This observation is 

based, however, on a few datasets, and should be followed up with more rigorous 

studies. 

Figure 1. Diversity, multistrata canopy structure, and various functional groups of food, fuel, 

fruit and nut yielding plants in a Kerala homegarden [coconut palms (Cocos nucifera), areca 

or betel nut palms (Areca catechu), jackfruit tree (Artocarpus heterophyllus), black pepper 

vines (Piper nigrum), plantains (Musa a spp.) and the like]. 

3.1. Uncertainties in estimating homegarden C stocks 

Since net ecosystem productivity generally reflects the overall gain or loss of 

terrestrial C pools (Nair and Nair, 2003), larger C sinks are probable when croplands 

(input-intensive production systems) are converted into homegardens (sensu 

. 

B.M. KUMAR



193 

Houghton and Goodale, 2004) than simple agroforests/plantations. Information on 

the actual rate of change in homegarden coverage and the spatial and temporal 

heterogeneity in C stocks are, however, not available. Lack of such data at the 

landscape-level particularly hampers our understanding of the potential of 

homegarden systems to sequester C and its eventual use in C sink projects, which is 

a situation that is common to most agroforestry systems (Montagnini and Nair, 

2004). 

Yet another challenge is the difficulty in estimating tree biomass itself. Despite 

the fact that most trees accumulate C in their wood, precise estimates on the C 

sequestration potential of several tropical trees are not available (Roshetko et al., 

2002). Aboveground biomass is usually estimated with general regression equations 

developed for trees in the natural forests. However, the size of individual tree 

canopies in a forest and in an open agroforestry setting could be variable, as the trees 

in some agroforestry systems have more space and access to light. In addition, the 

crown and root architecture and tree management practices are different; the 

resultant variations in structure could probably result in erroneous estimates. 

A more important technical issue is the definition of a standard set of methods 

and procedures for the inventory and monitoring of C stocks in current and potential 

land use and management approaches (FAO, 2004). Differing interpretations of 

source and sink category or other definitions, use of simplified representations with 

“averaged” values and uncertainties in the basic processes leading to emissions 

and/or removals further complicate the matter (de Jong, 2001). In addition, to 

estimate the effects of harvest on homegarden C stocks, accurate information on 

three items is required: pre-harvest biomass, the fraction of this biomass harvested 

or damaged, and the fraction of the harvested biomass removed; much of these are 

not available, making estimation of the C sequestration potential of homegardens at 

the landscape-level a difficult issue. 

4. PLANT DIVERSITY IN HOMEGARDENS AND C SEQUESTRATION 

High biodiversity is an intrinsic property of the homegardens (Kumar and Nair, 

2004), which presumably favors greater NPP (Vandermeer, 1989) and higher C 

sequestration potential than monospecific production systems. This could be because 

diverse assemblages (Fig. 2) have a greater likelihood of containing species with 

strong responses to resources compared to species-poor assemblages (Tilman et al., 

1997). The inference that diversity leads to greater NPP and thus stability of 

ecosystems, however, is the subject of an ongoing debate in ecology (McCann, 

2000). That is, although homegardens and other multistrata systems are assumed to 

promote NPP and improve the soil and biomass C sequestration (Table 2), often 

doubts are expressed concerning the productive capacities of species mixtures 

(FAO, 1992; Wedin and Tilman, 1993). In particular, asymmetric competition 

(resource acquisition at differential rates; Wedin and Tilman, 1993) and thereby 

resource pre-emption by the dominant component of a competing mixture may 

retard their productive potential.

Attributes Types of land use system 

C stocks 1 

Aboveground 

Soil 

Fossil fuel inputs/subsidies (C costs) 2 

Table 2. Summary of the relative attributes of a land use continuum in the tropics. o 

Intensive monoculture Polyculture 

Annual crops Perennial Simple Homegardens 

crops/ 

and complex 

plantations 

multistrata 

systems 

agroforests 9 

Secondary 

forests 

Mature 

forests 

low 

low-high medium-high medium-high very high 

low 

low-high low-high medium-high medium-high very high 

high medium-high low- low zero zero 

medium 

Ecosystem services 3 low low-medium low-high medium-high high very high 

Diversity 3 

low low medium high very high very high 

Herbivory pressure 4 

high high medium low low very low 

Loss rate of soil C (decomposition) 5 high-very high medium medium low-medium low-medium low 

Nutrient outputs 6 (leaching/other high medium medium low low very low 

losses) 

Soil biota 7 low low medium high high very high 

low-high 8 

194 B.M. KUMAR

1 Low (121 Mg C ha –1 ); upper limits represent the midpoints 

of the range of values reported by Sanchez (2000) for pastures, simple agroforests and complex agroforests rounded to the nearest multiple of 10 (see 

Table 1). 

2 Annual/perennial crops are usually fertilized, irrigated, and managed with heavy doses of plant protection chemicals; agroforestry in 

general is less input intensive and for homegardens, in particular, little or no chemical inputs are used, while the natural systems are selfnourished. 

3 Agricultural intensification (e.g., large-scale use of agricultural chemicals) reduces diversity and abundance of biota; for example, the 

bees, which render pollination services (Kremen et al., 2002). 

4 The natural enemy complex of crop pests/pathogens is generally low in intensive monospecific production systems than in polycul 

n tures; 

consequently, the herbivory pressure is much lower in the natural and woody perennial-based mixtures (Keenan et al., 1995; Ball et al., 

1995; Jactel et al., 2005). 

5 Soil organic matter (SOM), a keystone component of the ecosystem (sensu Swift et al., 2004), is related to the quantity and variability of 

plant litter inputs. Higher floristic diversity generally ensures greater litter heterogeneity (Hättenschwiler et al., 2005) and the “speciesrich” 

systems generally have a greater chance of maintaining soil organic matter relations than the “species-poor” ones (Russell et al., 

2004). 

6 Loss of perennial vegetation leads to erosion, reduced soil quality, and low productivity (Singh et al., 1992; Vinod et al., 2003). 

7 Greater organic matter fluxes in woody perennial-based systems favor soil biota (Vohland and Schroth, 1999; Kumar and Nair, 2004). 

8 Wherever ranges are mentioned, it denotes variations because of stand age, species, management and/stage of succession. 

9 with one dominant species such as oil palm, cacao, coffee and the like. 

CARBON 



195

196 

The implicit assumption in studies reporting the positive “mixture effect,” 

however, is that one or more of the components improve the environment 

(facilitative production principle; Vandermeer, 1989) and/or share site resources 

harmoniously. The contribution of biologically fixed N2 to the associated non-N2 

fixing component is particularly relevant in this respect. Legumes in general are 

thought to be soil improvers—and may promote the growth and productivity of 

components in such systems (Kaye et al., 2000); yet there is no agreement on the 

role of woody legumes in promoting growth and NPP of associated woody nonlegume 

components. Lack of consistent impacts of the legume components in 

experimental mixtures (Parrotta, 1999; Gathumbi et al., 2004), can be explained 

based on species, site attributes—especially soil N content and soil management, 

ensuring the availability of appropriate rhizobial strains and maintenance of 

conditions suitable for their multiplication. 

In certain cases, productivity has been linked to site quality; for example, higher 

productivity for mixtures on nutrient-poor sites (Montagnini et al., 1995). 

Furthermore, there are considerable variations in the C sequestration potential of 

individual gardens and species, implying both within- and between- garden 

variations (Table 1). Yet, no comparative accounts on homegarden productivity as a 

function of its floristic attributes could be found. Issues such as what contributes to 

the superior performance of multistrata systems and homegardens also have not 

been adequately addressed. Such an analysis, however, is relevant to the CBD to 

which land use change, agriculture and forestry activities recognized by the Kyoto 

Protocol are closely linked. Aside from the ecological benefits of biological 

diversity conservation and improved site fertility, species mixtures offer greater 

resistance to insect infestation or disease outbreak (Table 2). A recent review, based 

on a meta-analysis of more than 50 field experiments, which contrasted pure stand 

vs. mixed stand of the same tree species, demonstrated a significant increase in 

insect pest damage in single-species stands (Jactel et al., 2005). 

It is probable that the relative superiority is dependent on species/circumstances, 

and is not amenable to sweeping generalizations; i.e., the effect may be positive, 

negative, or neutral. Ideally, in a mixture, the components should exploit different 

vertical layers—both above- and belowground—which signifies greater resource 

utilization efficiency. This idea, however, pre-supposes that species with divergent 

growth characteristics, be mixed for optimizing resource capture (Kumar et al., 

2001; Gathumbi et al., 2002). An interesting aspect of belowground resource use, 

however, is that the proximity of species/individuals often favors competitive 

downward displacement of tree roots (Kumar and Divakara, 2001). That is, in 

certain cases, species may develop vertically stratified root systems, and this spatial 

segregation of the roots of associated plants may abate possible inter specific 

competition in species-mixtures (Divakara et al., 2001). By extension, in homegardens, 

depending upon the nature of associated tree components, a greater 

potential to capture the lower leaching nutrients and accomplishing on-site nutrient 

conservation is probable (safety-net mechanism). Therefore, if planned with 

consideration for each species’ growth characteristics, mixed stands and homegardens 

could, theoretically, be more productive than single species stands and 

would probably sequester more C. 

. 

B.M. KUMAR



5. PRODUCTIVITY UNDER RISING ATMOSPHERIC CO2 LEVELS 

197 

Although it is now clear that high CO2 emission levels (Houghton, 1995) will have 

several adverse fallouts, indications are that the elevated CO2 may increase plant 

photosynthesis and NPP to some extent (Mingkui and Woodard, 1998). Given that 

the capacity of the photosynthetic machinery of C3 plants remains unsaturated at 

current concentrations of close to 370 ppm of CO2 (Körner, 2003), this seems 

reasonable too (but see Luo et al., 2004). Some experimental evidences also suggest 

that plant diversity and composition influence the enhancement of biomass and C 

acquisition in ecosystems subjected to elevated atmospheric CO2 concentrations. For 

instance, Reich et al. (2001) reported that biomass accumulation was greater in 

species-rich than in species-poor experimental populations under conditions of CO2 

and N fertilization. By extension, homegardens, which are inherently species-rich, 

may trap progressively greater quantities of atmospheric CO2 under rising levels of 

this gas. In view of the limited nature and range of the experimental studies reported 

(mostly from temperate regions and none on tropical homegardens), however, it is 

difficult to draw firm generalizations on the effects of enriched CO2 levels on C 

sequestration, especially in the tropics. 

6. SOIL CARBON SEQUESTRATION 

More than half of the C assimilated by woody perennials is eventually transported 

belowground via root growth and organic matter turnover processes (e.g., fine root 

dynamics, rhizodeposition, and litter dynamics), making soil organic carbon (SOC) 

a significant pool of terrestrial C (~2500 Pg C globally; Lal, 2004). In view of the 

great diversity and abundance of woody perennial components, it is perhaps 

reasonable to assume that the magnitude of such processes will be greater in 

homegardens compared to other systems (Gajaseni and Gajaseni, 1999; Kumar and 

Nair, 2004). Judicious management of plant residues as it is often practiced in 

homegardens also can contribute to increases in soil organic matter content 

(Montagnini, 2006). There is, however, great variation among homegardens in this 

respect. For instance, Roshetko et al. (2002) found that SOC of Indonesian 

homegardens ranged between 10.4 to 103.7 Mg C ha –1 . 

The C stored within the soil may increase and under certain conditions biomass 

production also increases, augmenting C inputs (root biomass, litter and prunings) 

into the soil (sensu Lal et al., 1998). Consistent with this, Russell (2002) noted that 

total SOC may increase directly with basal area of the trees included in the system. 

Both inputs and decomposition rates are, however, strongly affected by a host of 

factors (Lal et al., 1998) including climate change (Schimel et al., 2000). Warmer 

temperatures generally accelerate litter decomposition. However, in view of the 

possible stimulatory effects of rising atmospheric CO2 levels on photosynthetic 

production and the associated greater litterfall production rates, the effects are 

seemingly more complex (Kumar et al., 2005). 

Soil organisms such as microflora (bacteria, fungi, actinomycetes and algae), 

mesofauna (mites, collembola, micro-arthropods and enchytraeid worms), microfauna 

(protozoa, nematodes and mites) and macrofauna (earthworms, spiders, slaters,

198 

B.M. KUMAR 

centipedes, larvae, molluscs, etc.) fulfill a wide range of ecosystem services that 

underpin C sequestration and eventually the sustainability of the homegarden system 

(Table 2). However, as on many other aspects of belowground diversity, few data 

are available on the composition of soil biota or its determinants in the 

homegardens. This is partly because soil research in multistrata agroforestry systems 

poses methodological difficulties. Owing to variations in soil microenvironment, 

profound intra-garden variations in soil biotic activity are also probable. 

7. CARBON SEQUESTRATION PROGRAMS AND LIVELIHOOD 

SECURITY OF RURAL PEOPLE 

The Kyoto Protocol, the main instrument of the United Nations Framework 

Convention on Climate Change (UNFCCC), has set up the Clean Development 

Mechanism (CDM) concept as a cost-effective process to reduce rural poverty by 

extending payments to low-income farmers who provide carbon storage through 

land use systems 5 . Projects under the CDMs usually have the dual mandate of 

reducing greenhouse gas emissions and contributing to sustainable development. 

Implicit in this are, trade-offs between carbon sequestration, local social development, 

economic well-being and access to resources, and other aspects of 

environmental changes. Moreover, C storage through agroforestry is less costly 

(range $1–69/Mg C, median $13/Mg C) than through other CO2 mitigating options 

such as pure tree-based systems, carbon dioxide capture and storage or emission 

reduction (Albrecht and Kandji, 2003). It allows investors in developed countries to 

receive carbon credits in exchange for greenhouse gas emission reductions, whilst 

the developing countries where such investments are made, receive investments. 

There are many examples of how payment for environmental services to farmers can 

be made, while implementing mitigation projects (Brown et al., 2004; Montagnini 

and Nair, 2004). Carbon finance projects, thus, could transcend the existing barriers 

in resource mobilization for sustainable development of the developing countries. 

Although a number of such projects have been initiated as pilot activities around 

the globe, in alliance with non-governmental or development agencies, none of these 

as of date, are on tropical homegardens, implying that the potential of homegardens 

as a strategy for carbon sequestration has not yet been fully recognized, let alone 

exploited. Yet, the homegarden system offers considerable scope to improve biomass 

accumulation, and overcome “excess problems” (i.e., ameliorating “soil sickness” 

through mechanisms such as phytoremediation). Three pathways could be explored 

to promote externalities in agroforestry, in general, and homegarden systems, in 

particular. These are: 

• “Bringing more land under homegardens”: More land should be brought under 

agroforestry, resulting in more C sequestered in the landscape. There are already 

plenty of degraded lands available in most developing countries. For example, an 

estimated 1900 million ha of land is affected by soil degradation worldwide; of 

these, the largest area (around 747 million ha) is in the Asian region (van Lynden 

and Oldeman, 1997); India alone has an estimated 130 million ha of degraded 

lands 6 . The bottom line is that degraded sites could be molded into reasonably



199 

productive systems by appropriate policy and/or management interventions. For 

example, the indigenous Mayan groups have survived the extreme conditions by 

developing the multistrata homegardens over the karst topography, formed by 

limestone bedrock, and limited amounts of precipitation (Benjamin et al., 2001). 

• Intensification: More C can be sequestered per unit of land by improving 

efficiency of production through the choice of optimal species combinations and/or 

appropriate stand management practices, on which little scientific information 

exists, however. Moreover, restoring soil C triggers soil quality improvements 

(Lal, 2004). Multistrata stands and polycultures such as homegardens not only 

increase C sinks in soil and vegetation but also improve agricultural productivity 

and livelihood security, and are thought to be one step closer in the 

transformation of barren landscapes to “perpetually natural looking forests”— 

clearly a “win-win” situation (FAO, 2004). 

• Conservation: Ensuring long-term stability and sustainability—if such 

polycultures at least partially alleviate the anthropogenic pressure on natural 

forests— improves biodiversity conservation and reduce fossil fuel consumption. 

Many such traditional land use systems are, however, experiencing severe strains 

(Kumar and Nair, 2004), especially in the backdrop of technological changes; 

and to preserve them, appropriate land use policies/managerial interventions are 

needed. 


Under the Kyoto Protocol, one clear strategy for mitigating the increase in 

atmospheric CO2 is to expand the size of the terrestrial C sink, using trees on 

agricultural lands as “biological scrubbers.” The magnitude of such C sequestration 

may, however, be dependent on the nature and extent of agroforestry system 

involved, and its structure and function, which in turn, are dependent on species 

composition and system management. Apparently, the homegardens have a special 

role in such abatement processes. Overall, they occupy the penultimate position in a 

tropical land use continuum ranging from annual crops to mature forests (Table 2). 

In particular, aspects such as higher biomass production potential and the return of a 

greater proportion of plant materials to the soil to increase its C stock compared to 

other agroforestry systems have been adequately demonstrated. In addition, they 

ensure “carbon permanence,” which the “carbon contracts” require, farmers to 

adopt; maintain sustainability and exploit the synergies between CBD and the Kyoto 

Protocol. 

One of the major constraints in employing homegardens to provide 

environmental benefits, however, is the lack of quantitative data on such potential 

advantages. Nevertheless, in view of the substantial coverage of homegardens in 

some geographical regions (e.g., south and southeast Asia) and especially if 

effective policies to promote such land use systems are implemented especially for 

degraded lands, they could become large carbon sinks; and the mitigation costs are 

probably lower than what is required for emission source controls. Indeed, the 

traditional knowledge has shown that the homegarden system is ideally suited for

200 

B.M. KUMAR 

regions characterized by highly weathered soils with relatively lower nutrient 

endowments as in the lateritic soils of Kerala and the karst deposits of Yucatán 

Peninsula. On a final note, science and natural resource policy should recognize the 

work of local people who still maintain agroecosystems with high agrobiodiversity 

as part of their culture, lifestyle, or practice. 

ENDNOTES 

1. Badan Pusat Statistik (BPS) 2000. Statistical yearbook of Indonesia 2000, 

Jakarta, 590p. 

2. Bangladesh Bureau of Statistics (BBS) 2001. Statistical yearbook 2000. 

Ministry of Planning, Government of People’s Republic of Bangladesh, Dhaka, 

452p. 

3. Department of Census and Statistics (DCS) 2003. Census of agriculture – Sri 

Lanka 2002. Agricultural holdings, extent under major crops and livestock 

statistics by district and DS/AGA division: based on operator’s residence— 

small holding sector. Preliminary Release– No. 2, P.O. Box 563, Colombo, 62p 

(www.statistics.gov.lk). Last accessed: November 2005. 

4. Kerala State Land Use Board (KSLUB) 1995. Land resources of Kerala State. 

Thiruvananthapuram, Kerala, 209p. 

5. Smith J. and Scherr S.J. 2002. Forest Carbon and local livelihoods: assessment 

of opportunities and policy recommendations. CIFOR Occasional Paper 37, 

Centre for International Forestry Research, Jakarta, 45p. 

6. Indian Council of Forestry Research and Education (ICFRE) 2000. Forestry 

statistics—2000. Directorate of Statistics, Indian Council of Forestry Research 

and Education, Dehra Dun, pp 55. 

REFERENCES 


Agric Ecosyst Environ 99: 15 – 27. 

Ball J.B., Wormald T.J. and Russo L. 1995. Experience with mixed and single species 

plantations. Commonwealth For Rev 74: 301 – 305. 

Beer J., Bonnemann A., Chavez W., Fassbender H.W., Imbach A.C. and Martel I. 1990. 

Modelling agroforestry systems of cacao (Theobroma cacao) with laurel (Cordia 

alliodora) or poro (Erythrina poeppigiana) in Costa Rica. V. Productivity indices, organic 

material models and sustainability over ten years. Agroforest Syst 12: 229 – 249. 

Benjamin T.J., Montanez P.I. Jimenex J.J.M. and Gillespie A.R. 2001. Carbon, water and 

nutrient flux in Maya homegardens in the Yucatan peninsula of Mexico. Agroforest Syst 

53: 103 – 111. 

Brown K., Adger W. N., Boyd E., Corbera-Elizalde E. and Shackley S. 2004. How do CDM 

projects contribute to sustainable development? Tyndall Centre Technical Report No. 16. 

University of East Anglia, Norwich, UK, 54p. http://www.tyndall.ac.uk/publications/ 

tech_reports/tech_reports.shtml (last accessed: December 2005). 

Christanty L. 1990. Homegardens in tropical Asia with special reference to Indonesia. In 

Landauer K. and Brazil M. (eds), Tropical home gardens, pp. 9 – 20. United Nations 

University Press, Tokyo.



201 

Cuevas E. and Medina E. 1986. Nutrient dynamics within Amazonian forests. 1. Nutrient flux 

in fine litterfall and efficiency of nutrient utilization. Oecologia 68: 466 – 472. 

de Jong B.H.J. 2001. Uncertainties in estimating the potential for carbon mitigation of forest 

management. For Ecol Manag 154: 85 – 104. 

Divakara B.N., Kumar B.M., Balachandran P.V. and Kamalamm N.V. 2001. Bamboo hedgerow 

systems in Kerala, India: root distribution and competition with trees for phosphorus. 


Duguma B., Gockowski J. and Bakala J. 2001. Smallholder cacao (Theobroma cacao Linn.) 

cultivation in agroforestry systems of West and Central Africa: challenges and 

opportunities. Agroforest Syst 51: 177 – 188. 

FAO. 2004. Assessing carbon stocks and modelling win–win scenarios of carbon 

sequestration through land use changes. Food and Agriculture Organization of the United 

Nations, Rome, 156p. 

FAO. 1992. Mixed and pure forest plantations in the tropics and sub-tropics. FAO Forestry 

Paper 103, Food and Agriculture Organization of the United Nations, Rome, 152p. 



Gathumbi S.M., Cadisch G. and Giller K.E. 2004. Improved fallows: effects of species 

interaction on growth and productivity in monoculture and mixed stands. For Ecol Manag 

187: 267 – 280. 

Gathumbi S.M., Ndufa J.K., Giller K.E. and Cadisch G. 2002. Do species mixtures increase 

above- and belowground resource capturing in woody and herbaceous tropical legumes? 

Agron J 94: 518 – 526. 

Hättenschwiler S., Tiunov A.V. and Scheu S. 2005. Biodiversity and litter decomposition in 

terrestrial ecosystems. Ann Rev Ecol Evol Syst 36: 191 – 218. 

Houghton J.T. 1995. Determining emissions of carbon from land: a global strategy. In: Murai 

S. (ed.), Toward global planning of sustainable use of the earth, pp 59 – 76. Elsevier, 

Amsterdam. 

Houghton R.A. and Goodale C.L. 2004. Effects of land use change on the carbon balance of 

terrestrial ecosystems. Ecosystems and land use change. Geophys Monogr Ser 153: 

85 – 98. 

Jactel H., Brockerhoff E. and Duelli P. 2005. A test of the biodiversity–stability theory: metaanalysis 

of tree species diversity effects on insect pest infestations, and re-examination of 

responsible factors. In: Scherer-Lorenzen M., Körner Ch., and Schulze E.-D. (eds), Forest 

diversity and function: Temperate and Boreal Systems. Ecological Studies, Vol. 176. pp 

235 – 262. Springer-Verlag, Berlin, Heidelberg. 

Jamaludheen V. and Kumar B.M. 1999. Litter of nine multipurpose trees in Kerala, Indiavariations 

in the amount, quality, decay rates and release of nutrients. For Ecol Manag 

115: 1 – 11. 



Kaye J.P., Resh C.S., Kaye M.W. and Chimner R.A. 2000. Nutrient and carbon dynamics in a 

replacement series of Eucalyptus and Albizia trees. Ecology 81: 3267 – 3273. 

Keenan R., Lamb D. and Sexton G. 1995. Experience with mixed species rainforest 

plantations in North Queensland. Commonwealth For Rev 74: 315 – 321. 

Körner C. 2003. Carbon limitation in trees. J Ecol 91: 4 – 17. 

Kraenzel M., Castillo A., Moore, T. and Potvin C. 2003. Carbon storage of harvest –age teak 

(Tectona grandis) plantations, Panama. For Ecol Manag 173: 213 – 225. 

Kremen C., Williams N.M. and Thorp R.W. 2002. Crop pollination from native bees at risk 

from agricultural intensification. Proc Natl Acad Sci 99: 16812 – 16816.

202 

Kumar B.M. and Divakara B.N. 2001. Proximity, clump size and root distribution pattern in 

bamboo: A case study of Bambusa arundinacea (Retz.) Willd., Poaceae, in the Ultisols of 

Kerala, India. J Bamboo Rattan 1: 43 – 58. 


135 – 152. 

Kumar B.M, George S.J. and Chinnamani S. 1994. Diversity, structure, and standing stock of 

wood in the homegardens of Kerala in peninsular India. Agroforest Syst 25: 243 – 262. 

Kumar B.M., George S.J., Jamaludheen V. and Suresh T.K. 1998. Comparison of biomass 

production, tree allometry and nutrient use efficiency of multipurpose trees grown in 

wood lot and silvopastoral experiments in Kerala, India. For Ecol Manag 112: 145 – 163. 

Kumar B.M., Haibara K. and Toda H. 2005. Does plant litter become more recalcitrant under 

elevated atmospheric CO2 levels? Global Environ Res 9: 83-91. 

Kumar B.M., Thomas J. and Fisher R.F. 2001. Ailanthus triphysa at different density and 

fertilizer levels in Kerala, India: tree growth, light transmittance, and understorey ginger 

yield. Agroforest Syst 52: 133 – 144. 

Kürsten E. 2000. Fuelwood production in agroforestry systems for sustainable land use and 

CO2 mitigation. Ecol Eng 16: 69 – 72. 

Lal R. 2004. Soil carbon sequestration impacts on global change and food security. Science 

304: 1623 – 1627. 

Lal R., Kimble J.M., Follett R.F. and Stewart B.A. 1998. Soil Processes and the Carbon 

Cycle. CRC Press LLC, MA, 609p. 

Luo Y., Su Bo, Currie W.S., Dukes J.S., Finzi A., Hartwig U., Hungate B., McMurtrie R.E., 

Oren R., Parton W.J., Pataki D.E., Shaw, M.R., Zak D.R. and Field C.B. 2004. 

Progressive nitrogen limitation of ecosystem responses to rising atmospheric carbon 

dioxide. BioScience 54: 731 – 739. 

McCann K.S. 2000. The diversity-stability debate. Nature 405: 228 – 233. 

Mingkui C. and Woodard F.I. 1998. Dynamic responses of terrestrial ecosystem carbon 

cycling to global climatic change. Nature 393: 249 – 252. 

Montagnini F. 2006. Status of homegardens in Mesoamerica. In: Kumar B.M. and Nair 



B.M. KUMAR 

Montagnini F. and Nair P.K.R. 2004 Carbon sequestration: an underexploited environmental 

benefit of agroforestry systems. Agroforest Syst 61: 281 – 295. 

Montagnini F., Gonzalez E., Porras C. and Rheingans R. 1995. Mixed and pure forest 

plantations in the humid neotropics: a comparison of early growth, pest damage, and 

establishment costs. Commonwealth For Rev 74: 306 – 314. 

Nair P.K.R. and Kumar B.M. 2006. Introduction. In: Kumar B.M. and Nair P.K.R. (eds), 

Tropical homegardens: A time-tested example of sustainable agroforestry, pp 1 – 10. 

Springer Science, Dordrecht. 

Nair P.K.R. and Nair V.D. 2003. Carbon storage in North American agroforestry systems. In: 

Kimble J., Heath L.S., Birdsey R.A., and Lal R. (eds), The potential of U.S. forest soils to 

sequester carbon and mitigate the greenhouse effect, pp 333 – 346. CRC Press, Boca 

Raton, FL. 

Parrotta J.A. 1999. Productivity, nutrient cycling and succession in single- and mixed-species 

stands of Casuarina equisetifolia, Eucalyptus robusta and Leucaena leucocephala in 

Puerto Rico. For Ecol Manag 124: 45 – 77. 

Reich P.B., Knops J., Tilman D., Craine J., Ellsworth D. Tjoelker M., Lee T., Wedin D., 

Naem S., Bahauddin D., Hendrey G., Jose S., Wrage K., Goth J. and Bengston W. 2001. 

Plant diversity enhances ecosystem responses to elevated CO2 and nitrogen deposition. 

Nature 410: 809 – 812.



203 

Roshetko M., Delaney M., Hairiah K. and Purnomosidhi P. 2002. Carbon stocks in 

Indonesian homegarden systems: Can smallholder systems be targeted for increased 

carbon storage? Am J Alt Agr 17: 125 – 137. 

Roy C. 1999. Options techniques et socio-économiques des émissions de CO 2 

et d’augmentation des stocks de carbone. CR Acad Agric. France 85: 311 – 320. 

Ruark G.A., Schoeneberger M.M. and Nair P.K.R. 2003. Agroforestry–Helping to Achieve 

Sustainable Forest Management. UNFF (United Nations Forum for Forests) Intersessional 

Experts Meeting on the Role of Planted Forests in Sustainable Forest Management, New 

Zealand (24 – 30 March 2003). www.maf.govt.nz/unff-planted-forestry-meeting (last 

accessed: July 2005). 

Russell A.E. 2002. Relationships between crop-species diversity and soil characteristics in 

southwest Indian agroecosystems Agric Ecosyst Environ 92: 235 – 249. 

Russell A.E., Cambardella C.A., Ewel J.J. and Parkin T.B. 2004. Species, rotation, and life 

form diversity effects on soil carbon in experimental tropical ecosystems. Ecol Appl 14: 

47 – 60. 

Sanchez P.A. 2000. Linking climate change research with food security and poverty reduction 

in the tropics. Agric Ecosyst Environ 82: 371 – 383. 

Sanford R.L. and Cuevas E. 1996. Root growth and rhizosphere interactions in tropical 

forests. In: Mulkey S., Chazdon R.L., and Smith A.P. (eds), Tropical forest plant 

ecophysiology, pp 268 – 300. Chapman and Hall, New York. 

Schimel D., Melillo J., Tian H., McGuire A.D., Kicklighter D., Kittel T., Rosenbloom N., 

Running S., Thornton P., Ojima D., Parton W., Kelly R., Sykes M., Neilson R. and Rizzo 

B. 2000. Contribution of increasing CO 2 and climate to carbon storage by ecosystems in 

the United States. Science 287: 2004 – 2006. 

Schroeder P. 1994. Carbon storage benefits of agroforestry systems. Agroforest Syst 27: 

89 – 97. 

Singh G., Babu R., Narain P., Bhushan L.S. and Abrol I.P. 1992. Soil erosion rates in India. 

J Soil Water Conserv 47 (1): 97 – 99. 

Swift M.J., Izac A.-M.N. and van Noordwijk M. 2004. Biodiversity and ecosystem services in 

agricultural landscapes—are we asking the right questions? Agric Ecosyst Environ 104: 

113 – 134. 

Tilman D., Lehman C.L. and Thomson K.T. 1997. Plant diversity and ecosystem 

productivity: theoretical considerations. Proc Natl Acad Sci USA 94: 1857 – 1861. 

Tomich T.P., de Foresta H., Dennis R., Ketterings Q., Murdiyarso D., Palm C., Stolle F., and 

van Noordwijk M. 2002. Carbon offsets for conservation and development in Indonesia? 

Am Alt Agr 17: 125 – 137. 

UNFCCC 2002. Activities implemented jointly under the pilot phase. Sixth synthesis report. 

Subsidiary Body for Scientific and Technological Advice. FCCC/SBSTA/2002/8. United 

Nations Framework Convention on Climate Change, Bonn 5p (http://unfccc.int/resource/ 

docs/2002/sbsta/08. pdf; last accessed: December 2005). 

van Lynden G.W.J. and Oldeman L.R. 1997. The Assessment of the Status of Human- 

Induced Soil Degradation in South and Southeast Asia. International Soil Reference and 

Information Centre, Wageningen, 35p. 

van Noordwijk M., Rahayu S., Hairiah K., Wulan Y.C., Farida A. and Verbist B. 2002. 

Carbon stock assessment for a forest-to- coffee conversion landscape in Sumber-Jaya 

(Lampung, Indonesia): from allometric equations to land use change analysis. Science in 

China Series C-Life Sciences 45: 75 – 86 Suppl. S Oct 2002. Science in China Press, 

Beijing. 

Vandermeer J. 1989. The Ecology of Intercropping. Cambridge University Press, Cambridge, 

249p.

204 

B.M. KUMAR 

Vinod V.R., Syed Anwarulla M. and Vishwanth D.P. 2003. Run off and soil loss under 

different land use systems in the Western Ghats of Karnataka. Indian J Soil Conserv 31: 

131 – 138. 

Vohland K. and Schroth G. 1999. Distribution patterns of the litter macrofauna in agroforestry 

and monoculture plantations in central Amazonia as affected by plant species and 

management. Applied Soil Ecol 13: 57 – 68. 

Watson R.T., Noble I.R., Bolin B., Ravindranath N.H., Verardo D.J. and Dokken D.J. (eds). 

2000. Intergovernmental Panel on Climate Change (IPCC). Land use, land use change, 

and forestry. A special report of the Intergovernmental Panel on Climate Change. 

Cambridge University Press, 377p. 

Wedin D. and Tilman D. 1993. Competition among grasses along a nitrogen gradient: initial 

conditions and mechanisms of competition. Ecol Monogr 63: 199 – 229.

CHAPTER 12 

MEDICINAL PLANTS IN TROPICAL 

HOMEGARDENS 

M.R. RAO 1 AND B.R. RAJESWARA RAO 2 

1 Former ICRAF Scientist; Current address: Plot No. 11, ICRISAT Colony (Phase- 

I), Brig. Syed Road, Manovikasnagar (P.O.), Secunderabad 500 009, India; E-mail: 

. 2 Central Institute of Medicinal and Aromatic Plants 

(CIMAP) Resource Centre, Boduppal, Uppal P.O., Hyderabad 500 039, India 

Keywords: Aromatic plants, Bioprospecting, Indigenous knowledge, Phytochemicals, 

Traditional medicine, Value addition. 

Abstract. Nearly 80% of the people living in developing countries depend on medicinal 

plants (MPs) for primary healthcare, and homegardens are an important source of production 

of these plants. Homegardens can fulfill the dual role of production and in situ conservation of 

MPs to overcome their dwindling supplies and threat of extinction from natural sources. MPs 

in homegardens are either deliberately cultivated or they come up spontaneously. They are an 

important constituent of homegardens, next only to food crops and fruit trees; yet their 

economic value is not fully recognized, let alone exploited. Homegardens offer an 

economically and socially viable option for large-scale production of phytochemicals from 

important MPs under organic cultivation. Promoting organic production of selected 

commercially valuable species of MPs through homegardening can, thus, augment the 

farmers’ income, enhance rural employment opportunities, and help reduce migration of rural 

youth to urban centers in search of jobs. Research is needed to improve the existing 

germplasm, introduce suitable commercial MPs in different agroecosystems, and develop 

cultivation and processing techniques to increase yield and improve product quality, and 

exploit indigenous knowledge and market opportunities. 


Humans depended on certain plants for healthcare since time immemorial. Centuries 

of experimentation on the use of plants or products derived from them has led to the 

development of indigenous systems of medicine that are still respected and used in 

many societies. Plants have been a source of medicines for humans and livestock 

and pesticides to protect crops from certain pests and diseases. In India, over 200 

types of vegetable drugs were in use during the Vedic period (3700 – 2000 BC). 

Charak Samhita (600 BC) mentioned 1270 medicinal plants (MPs), while Sushruta 

205 




206 M.R. RRAO AND B.R. RRAJESWARA RAO R 

Samhita (450 BC) and Vagbhatta’s Astangahridaya (342 BC) mention about 1100 

and 1150 MPs, respectively (Chadha and Gupta, 1995). America, Arabia, China, 

Egypt, Greece, Mexico, and many other countries in Europe and Asia too recorded 

the use of MPs (Principe, 1991). Furthermore, about 1800 species of MPs are 

reported to be used in the traditional Indian medical system of Ayurveda, 750 

species in Unani or Tib, 500 species in Siddha, 400 species in the Tibetan medicine 

and 5000 species in the Chinese medicine. Traditional medical systems in Japan, 

Korea (Kampo system), Indonesia (Jamu system), South Africa (Julu system), 

Bhutan (Gso-ba-rig-pa), Sri Lanka (Deshiya Chikitsa), and Malaysia (Malay herbal 

medicine) also recorded a number of MPs and their uses (Principe, 1991). 

An estimated 14 to 28% of the 422 000 plants occurring on earth had been used 

by human cultures for medicinal purposes at one time or another (Farnsworth and 

Soejarto, 1991). Approximately 80% of the people in developing countries rely even 

today mainly on traditional medicines for humans (FAO, 1996) as well as domestic 

animals, a major portion of which are extracts of medicinal plants or their active 

principles. More than 6500 species of such medicinal plants have been identified in 

Asia, 1900 species in tropical America and 1300 species in north-west Amazon 

(Farnsworth and Soejarto, 1991). Global trade in plant-based drugs was estimated at 

US$ 100 billion, of which traditional medicines using medicinal plants accounted 

for 60 billion (WHO, 2004). In addition, trade 1 in herbal teas, drug adjuncts, dietary 

foods etc. (sold over the counter) was estimated at US$ 5 billion in 1997. India has 

approximately 150 000 practitioners of traditional systems of medicine, 10 000 

licensed pharmacies manufacturing plant-based drugs. The trade in medicinal herbs 

in India was estimated at US$ 1 billion (EXIM Bank, 2003) and the country exports 

medicinal herbs worth US$ 287 million annually 2 . 

Most of the medicinal plants (70 to 90%) have traditionally been collected from 

forests and natural habitats. Indiscriminate extraction over years not only reduced 

their supplies but also endangered some of these valuable species. The growing 

demand for plant-derived drugs both in modern and traditional systems of medicine 3 

further exacerbated the problem in many natural habitats. This has led to the 

extinction of about 75 species between 1600 and 1900 and a similar number in a 

short span between 1900 and 1970 (Principe, 1991; Rao, 1999). It is feared that if 

this trend continues, about 60 000 species will become extinct in the next century 

(Principe, 1991). Considering the economic importance of medicinal plants, there is 

an urgent need to systematically cultivate them to exploit their full potential and to 

save them from extinction. MPs can be cultivated like any other crop(s) in different 

systems including agroforestry – in forest plantations, homegardens, as intercrops 

between trees, and as components of multistrata systems (Rao et al., 2004). This 

chapter reviews the status of medicinal plants in tropical homegardens and examines 

the scope for improving their relative contribution to the economy of rural families. 

2. MEDICINAL PLANTS IN HOMEGARDENS 

Homegardens being one of the earliest forms of agroforestry practiced in the tropics 

(Kumar and Nair, 2004), it is only logical to be expected that MPs have been an 

essential component of these production systems (Tables 1 and 2). Indeed, the

MEDICINAL M PLANTS IN HOMEGARDENS 

207 

homegardens make a substantial contribution to the supply of MPs, which may be 

traded or consumed locally by the family or community (Albuquerique and 

Andrade, 2002). There is, however, no reliable data on the extent of homegardens in 

different countries (see Nair and Kumar, 2006), yields of medicinal plants, or 

products extracted and sold at national and international levels. Majority of MPs in 

homegardens are herbs/vines/climbers and they together with vegetables and spices 

generally constitute the lower layer (0 – 1 m), unless they are vines and climbers. 

Additionally, a number of homegarden shrub and tree species also have medicinal 

value and they constitute the second (1 – 3 m) and upper (>10 m) layers respectively 

(Wezel and Bender, 2003). Some species that grow spontaneously in homegardens 

may possess medicinal value which may or may not be recognized and used. For 

example, in Chiriqui, Panama, the Ngöbe community utilizes the land fallowed for 

soil fertility replenishment as a source of MPs (Samaniego and Lok, 1998). Nearly 

half of the 41 weed species found in the homegardens of Central Sulawesi, 

Indonesia, possess medicinal value (Kehlenbeck and Maass, 2005). In India, 

seasonal weeds such as Phyllanthus amarus, Boerhaavia diffusa, Achyranthus 

aspera, Tribulus terrestris, Sida cordifolia, and Aerva lanata that occur both in 

cultivated fields (including homegardens) and wild are collected for medicinal 

purposes (Rao et al., 1999). 

2.1. Relative importance of MPs in homegardens 

While some components in the homegardens have exclusive medicinal value, others 

are multipurpose species combining medicinal value with food, ornamental, fiber, 

and spice values. For example, in the Kandyan homegardens of Sri Lanka, 30% of 

the total 125 species found were exclusively mentioned for medicinal uses and 12% 

combined medicinal with other uses. Among the medicinal species, trees constituted 

7%, shrubs 5%, herbs 15%, and creepers 3% of the total species (Perera and 

Rajapakse, 1991). Homegardens in Bukoba district in northwestern Tanzania 

contained species that were said to be used exclusively for medicine (Baphiopsis 

spp., Cyperus dives, Leonotis nepetifolia, Vernonia amygdalina, and Solanum 

incanum), those that combined medicine and fuelwood (Senecio multicorymbosa 

tree for medicines to cattle), medicine, fruit, and fuelwood (Psidium guajava and 

Citrus limon), and propping poles and medicine (Ricinus communis; Rugalema 

et al., 1994). The Chagga homegardens on Mt. Kilimanjaro in Tanzania were 

dominated by woody components; nearly 50% of the 111 species found in the region 

were trees, of which 30% were mentioned as medicines for humans and livestock 

(O’Kting’ati et al., 1984). Of the 77 useful plants (shrubs, vines, and forbs) found 

across 80 traditional Mayan homegardens in Quitana Roo, Mexico, nine were 

reported to have exclusive medicinal value and 26 species combined medicine, food, 

spice, and ornamental values (De Clerck and Negreros-Castillo, 2000). About 70% 

of 301 species in the forest and homegardens in the Yucatan, Mexico were classified 

for medicinal purpose; however, only 16 species were exclusively used for medicine 

and the rest had multiple uses (Rico-Gray et al., 1991).

Table 1. Spices, condiments, and aromatic plants possessing medicinal value grown in tropical homegardens. 

Species Family Part(s) used Uses (for treatment of diseases/other 

applications mentioned) 

Trees 

Cinnamomum Lauraceae bark diarrhea, gastric debility, flatulence, nausea, 

zeylanicum (cinnamom) 

vomiting, herbal tea 

Citrus aurantifolia Rutaceae fruit source of vitamin C, cataract, bleeding gum, 

(lime) 

herbal tea, smallpox 

Syzygium aromaticum Myrtacae flower buds carminative, antispasmodic, galacto purifier, 

(clove) 

antibacterial, appetizer, rubifacient 

Myristica fragrans Myristicaceae seeds, aril dyspepsia, diarrhea, hepatopathy, impotency, 

(nutmeg) 

insomnia, cardiac disorders 

Murraya koenigii Rutaceae leaves carminative, skin diseases, anorexia, 

(curry leaf) 

dyspepsia, flatulence, hair tonic, stomach 

ache 

Tamarindus indica Caesalpiniaceae root, leaves, fruit jaundice, scabies, smallpox, alcoholic 

(tamarind) 

pulp, seed intoxication, carminative, refrigerant 

Shrubs/herbs/grasses 

Allium sativum (garlic) Liliaceae bulb antiperiodic, antibacterial, diuretic, skin 

diseases 

Where grown? 

Sri Lanka, Indonesia, 

Madagascar, Brazil, 

Seychelles 

Many countries in 

tropics 

Southeast Asia , Sri 

Lanka, Tanzania, Brazil 

Southeast Asia, Sri 

Lanka, West Indies 

India 

South Asia, East, and 

West Africa 

throughout tropics 

208 M.R. RRAO AND B.R. RR RAJESWARA RAO R

Allium cepa (onion) Liliaceae bulb pulmonary phthisis, whooping cough, colic 

dyspepsia, reduces cholesterol 

Capsicum annuum Solanaceae fruits gout, arthritis, dyspepsia, hoarseness, 

flatulence 

Cymbopogon flexuosus, Poaceae leaves source of vitamin A, leprosy, epilepsy, 

C. citratus (lemongrass) 

mosquito repellent, herbal tea 

Coriandrum sativum 

(coriander) 

Cuminum cyminum 

(cumin) 

Curcuma longa (C. 

domestica) 

(turmeric) 

Elettaria cardamomum 

(small cardamom) 

Kaempferia galanga 

(candramula) 

South and Southeast Asia, 

Spain, Brazil, Egypt 

throughout tropics 

Mexico, Brazil, China, 

Haiti, South and 

Southeast Asia, Africa 

South and Southeast 

Asia 

Apiaceae fruit, leaf colic, laxative, blood purifier, indigestion, 

sour throat 

Apiaceae seed dyspepsia, flatulence, diarrhea, skin diseases East Asia, India 

Zingiberaceae rhizome antiseptic, skin allergies, viral hepatitis, antibacterial, 

wounds, anti-inflammatory, soar 

throat 

Zingiberaceae fruit nausea, indigestion, abdominal pains, 

Zingiberaceae rhizomes, rootstock. 

leaves 

bronchitis, respiratory infections 

digestive, vulnerary, anthelminthic, 

dyspepsia, leprosy, skin diseases, 

rheumatism, asthma, bronchitis, malaria, 

urolithiasis 

South, Southeast, and 

East Asia 

South and Southeast 

Asia 

India 



209

Species Family Part(s) used Uses(for treatment of diseases/other Where grown? 

applications mentioned) 

Mentha arvensis (mint) Labiatae leaves cough syrups, flavoring agent, expectorant, Mexico 

pain reliever 

Mentha piperita (mint) Labiatae leaves flavoring agent Mexico 

Piper betle (betel vine) Piperaceae leaves antiseptic, aphrodisiac, expectorant, South and Southeast 

bronchitis, rheumatism, stimulant, 

Asia 

carminative, wounds 

Piper nigrum 

Piperaceae dried berries indigestion, chronic rheumatism, asthma, Asia, Africa, Brazil 

(black pepper) 

cough, throat complaints 

Trigonella foenum- Fabaceae seeds anti-diabetic, flatulence, carminative, India, Middle East, 

graecum (fenugreek) 

emollient, galactagogue 

Egypt, Morocco 

Zingiber officinale Zingiberaceae rhizome asthma, skin diseases, de-worming, nausea, South and Southeast 

(ginger) 

carminative, common colds 

Asia, China, Nigeria 

Source: Padoch and De Jong (1991), Perera and Rajapakse (1991), Rugalema et al. (1994), Lamont et al. (1999), Rao et al. . (1999), De e Clerck 

and Negreros-Castillo (2000), Millat-e-Mustafa et al. (2002), and Wezel and Bender (2003). 



211 

Many of the economic species grown in homegardens possess complementary 

medicinal values. Such species may or may not be exploited commercially for their 

medicinal properties but are used locally within the family and community. For 

example, people in southeastern Nigeria uses a number of species that they grow in 

their compound farms—for purposes other than healthcare, for medicinal purposes 

(Okafor and Fernandes, 1987). Such species include Cajanus cajan (leaves for 

treating measles), Carica papaya (leaves for treating malaria), Cola lipidota/ 

C. nitida/C. pachycarpa (stimulant), Kigelia africana (bark for treating sores), 

Jatropha curcas (leaves for ringworm treatment), Neubouldia laevis (stem and roots 

medicinal), and Invingia gabonensis var. gabonensis (leaves and bark medicinal). 

Similarly, many plants are collected for medicinal uses from multistoried 

agroforestry systems in west Sumatra (Indonesia), although none was grown in the 

system consciously for that purpose (Michon et al., 1986). Majority of spices, a 

number of vegetables and ornamentals grown in homegardens also have medicinal 

uses (Table 1). The homegardens in Java and Sumatra were reported to contain 26 

medicinal species and a similar number of spices (Kubota et al., 1992). 

Agelet et al. (2000) made a detailed analysis of medicinal plants found in 155 

homegardens in the mountain zones of Catalonia (north-eastern Iberian Peninsula, 

Spain). The gardens contained nine distinct categories of species: plants exclusively 

cultivated for medicinal purpose (23) mostly close to the house, the medicinal wild 

plants favored by homegarden structure and care (105), and seven kinds of 

horticultural plants with complementary medicinal values (117). There was, 

however, loss of about 56 taxa or 23% of the total over the years. 

Despite the presence of many medicinal species in homegardens, only a few 

species stand out as economically important in any given region. The most 

frequently found species in 31 homegardens in three villages in Cuba, were 

Jatropha gossypiifolia, Senna occidentalis, Xanthoxylum pistacifolium, Pluchea 

odorata, and Rhoeo spathacea (Wezel and Bender, 2003). Common among species 

expressly cultivated for medicinal purpose in Catalan homegardens were Tanacetum 

parthenium – a plant used for intestinal antiseptic – and Liliun candidum for 

vulnerary use (Agelet et al., 2000). In the state of Kerala (India), Kaempferia 

galanga – which has been traditionally collected from forests, is now being 

commercially cultivated in the homegardens (Kumar et al., 2005) and as intercrop in 

orchard crops (Maheswarappa et al., 1998). Tribals living in the Eastern Ghats of 

Andhra Pradesh (India) have been growing Piper longum and Curcuma angustifolia 

extensively for medicinal purposes along with turmeric (Curcuma longa) using 

Jatropha curcas as a bio-fence in homegardens (K.P. Sastry, CIMAP Resource 

Centre, Hyderabad, pers. comm., July 2005). In the ‘Dai homegardens’ of 

Xishuangbanna province in China, the prominent medicinal species found were 

Acanthopanax trifoliatus, Toona sinensis, Sapindus rarak, Tamarindus indica, 

Bryophyllum pinnatum, Euphorbia antiquorum, and Prunus persica (Saint-Pierre, 

1991). Ammomum villosum, which requires about 70% shade, is planted under forest 

cover after clearing the undergrowth and it yields 30 to 150 kg rhizomes ha –1 year –1 

depending on water resource availability. Homegardens even in an isolated Soqotra 

island in the Republic of Yemen despite containing on average 3.9 to 8.4 species per 

garden included medicinal plants such as Aloe perryi, Jatropha unicostata, and

212 

Commiphora ornifolia (Ceccolini, 2002). This should indicate the importance given 

to MPs by rural people in the tropics. 

Table 2. Relative importance of medicinal species in relation to total species in tropical 

homegardens. 

Region/location Homegardens 

examined 

(no.) 

M.R. RRAO AND B.R. RRAJESWARA RRAO 

Total and 

medicinal a 

species 

across 

gardens 

Total and 

medicinal a 

species per 

garden 

Reference 

Santa Rosa, Peruvian 21 168 (46) 18 to 74 Padoch and de Jong 

Amazon 

(9.7) (1991) 

Bukoba, North- 72 57 (10) N/A Rugalema et al. 

western Tanzania 

(1994) 

Amazon, 

Northeastern Peru 

51 161 (56) N/A (9.5) Lamont et al. (1999) 

Catalonia, Iberian 145 N/A (250) N/A (30 to Agelet et al. (2000) 

Peninsula, Spain 

60) 

Congo (Zaire) N/A 273 (74) N/A Mpoyi et al. (1994) 

Masatepe, Nicaragua 1 98 (10) N/A (10) Viquez et al. (1994) 

Floodplain Jamuna 17 125 (48) N/A Yoshino and Ando 

tributary, 

Bangladesh 

(1999) 

Dhamrai, Bangladesh 243 N/A (71) N/A Millat-e-Mustafa 

et al. (2001) 

Deltaic, dry land, 200 120 (31) N/A Millat-e-Mustafa 

hilly, and plain 

regions, 

Bangladesh 

et al. (2002) 

Eastern Cuba 31 101 (39) 18 to 24 (4) Wezel and Bender 

(2003) 

Tixcacaltuyub and N/A 301 (152) N/A Rico-Gray et al. 

Tixpeual, Mexico 

(1991) 

Kerala, India 252 127 (25) 3 to 25 Kumar et al. (1994) 

Kandy, Sri Lanka 50 125 (52) 37 to 65 Perera and 

Rajapakse (1991) 

Central Sulavesi, 

30 149 28 to 37 Kehlenbeck and 

Indonesia 

(2.8) Maass (2005) 

a Values in parentheses refer to medicinal species; N/A = information not available. 

Immigrants from Southeast Asia to USA continued the tradition of growing 

many species in homegardens wherever they settled – for family use as well as for 

sale in the Asian markets. A survey of 59 gardens of Laotian Hmong settlers in the 

central Sacramento Valley, California, USA, revealed 59 taxa of which 38 had food

213 

value, 36 had medicinal value and a few others had uses like fiber and ornamental. 

Nineteen taxa had exclusive medicinal value, 15 combined food and medicine, and 

one or two combined medicinal, with ornamental or fiber uses. Many species that 

are categorized as being used for both food and medicine were primarily used for 

food seasoning or as additives (Corlett et al., 2003). 

2.2. Diversity of MPs in homegardens 

The species diversity including medicinal species in homegardens primarily depends 

on climate, altitude, socioeconomic and cultural factors, and nearness to markets. 

The diversity and density of plants generally increase with rainfall and elevation. In 

Venezuela, high diversity was positively correlated with age and remoteness of the 

garden, its use for subsistence, age of the farmer, and extent of participation of 

family labor in the activities of the garden (Mulas et al., 2004). In Bangladesh, 

species number decreased with increase in homegarden size and from deltaic region 

to dry region (Millat-e-Mustafa et al., 2002). Homegardens in West Java, Indonesia, 

contained the greatest diversity with an average number of 56 species per garden, 

the number of species being more in the wet season than in the dry season 

(Soemarwoto, 1987). In contrast, species composition of Cuban gardens differed 

across sites, especially in terms of medicinal plants, with gardens in the semiarid 

climate showing greater range than those in the humid region (Wezel and Bender, 

2003). Medicinal plants were recognized as the second most important group next 

only to cash value species in Sri Lanka (Perera and Rajapakse, 1991) and 

Bangladesh (Millat-e-Mustafa et al., 2002), food crops in Peruvian Amazon (Padoch 

and de Jong, 1991) and fruits in Cuba (Wezel and Bender, 2003) and Peruvian 

Amazon (Lamont et al., 1999). Homegardens close to cities were noted to capitalize 

on their relatively easy access to market in exploiting medicinal/other plants 

(Padoch and de Jong, 1991; Drescher et al., 2006). 

Aromatic species are less common compared to medicinal species in 

homegardens. Vetiver (Vetiveria zizanioides) cultivation was, however, observed in 

the homegardens of Kerala, India (Nair and Sreedharan, 1986) and the Chagga 

gardens on Mt. Kilmanjaro in Tanzania (Fernandes et al., 1984). Likewise, 

lemongrass (Cymbopogon citratus) was found in the homegardens of Thailand 

(Boonkird et al., 1984), Kerala (Nair and Sreedharan, 1986), and Nicaragua 

(Mendez et al., 2001), and citronella (Cymbopogon nardus) in the Kandyan 

homegardens of Sri Lanka (Perera and Rajapakse, 1991). Homegardens in Ethiopia 

also contained aromatic plants (Zemede and Ayele, 1995). 

2.3. Uses of MPs grown in homegardens 


The MPs grown in homegardens are used to treat a variety of ailments ranging from 

common colds, fevers, headache, snake bites, and digestive problems to infectious 

and complicated diseases (Tables 1, 3, and 4). Thus, we find species yielding 

curatives, preventives, placebos, palliatives, nutrition supplements, and energizers. 

Some of the species provide medicaments to treat livestock diseases, fish baits, and

214 

M.R. RRAO AND B.R. RRAJESWARA RAO R 

piscicides. Medicinal and aromatic species found in the homegardens are also 

used as biopesticides. For example, leaves of sacred basil (Ocimum sanctum/ 

O. enuiflorum) are traditionally used as a toxicant against insect pests in grain legume 

storage. Clove (Syzygium aromaticum) powder was found to cause adult mortality of 

bruchids (Callosobruchus maculatus; Rajapakse et al., 2002). Essential oils of 

citronella, Eucalyptus citriodora, and lemongrass are widely used as mosquito 

repellants. Parts of MPs used for medicinal purpose could be whole plants, young 

shoots, flowers, young leaves, stem, seed, bark, pods, rhizomes, bulbs, fruits, roots, 

and inflorescence depending on the species (see Tables 1, 3, and 4). 

3. GENDER ISSUES AND MEDICINAL PLANTS 

In many traditional societies, women are actively involved in the cultivation of food 

crops, while men are more concerned with the cash crops. This is true generally for 

Africa, the Ngöbe community of Panama (Samaniego and Lok, 1998), and the 

natives of Soqotra Island, Yemen (Ceccolini, 2002). Commercialization of certain 

products in the homegardens, however, reduced the diversity of species and income 

to women in a number of communities in Latin America (Howard, 2006). The 

proverbial reference to household treatment for common ailments, which generally 

are based on MPs as ‘grandmother’s remedies’, perhaps indicates the understanding 

of women on these aspects. Women also may have as much role as men, if not more, 

in the cultivation of traditional medicinal plants, use, and sale of herbal products in 

village markets because of proximity. In Nicoya, Costa Rica, it was noted that 

although men and women had equal knowledge of the parts used, women had 

greater knowledge of medicinal species, the forms of preparation, and application 

than men (Ochea et al., 1999; Howard, 2006). In Tanzania, men harvest fuel and 

fodder trees, while women harvest fodder grasses and herbs (Fernandes et al., 1984). 

Understanding the role of women in homegardens in general and possible impact of 

introduction of high value medicinal plants in homegardens on gender equity and 

well-being of women within the family and society is important; yet, in-depth 

studies are lacking on these aspects. 

4. SHADE TOLERANCE OF MEDICINAL PLANTS 

Several MPs, especially those grown in homegardens, require or can tolerate 

overstorey shade. Ginger (Zingiber officinale) can withstand light interception by 

the overstorey up to 48% without experiencing appreciable yield reduction (Kumar 

et al., 2001). Yield and quality of galangal or kacholam (Kaempferia galanga) – a 

medicinal and aromatic oil-yielding herbs were, however, not affected by light 

interception levels by the upperstorey canopy up to 82% of the open (Kumar et al., 

2005). In fact, rhizome yield of galangal as an intercrop in coconut garden was 6.1 

Mg ha –1 compared with 4.8 Mg ha –1 in the open in Kerala, India. Essential oil and 

oleoresin contents were also greater in the rhizomes of the intercropped kacholam 

(Maheswarappa et al., 1998). Likewise, Plumbago rosea, K. galanga, and 

Asparagus racemosus performed better as intercrops in 20 year-old coconut

Table 3. Multipurpose trees with medicinal uses grown in, or suitable for, r homegardens. 

Latin name Family Parts used Medicinal uses (treatment of the diseases 

Where grown 

mentioned) and other applications 

at present? 

Albizia lebbeck k (siris tree) Mimosaceae flowers, seeds, 

bark 

asthma, thoracic pain, skin diseases, leprosy, 

sprains, wounds, ulcers, neuralgia, night blindness, 

diarrhea 

asthma, bronchitis, leprosy, ulcers, fevers, tumors, 

India, Africa 

Alstonia scholaris/A. // boonei Apocynaceae leaves, bark, 

India, Africa 

(devil tree) 

milky exudates cardiopathy, helminthiasis, debility, elephantiasis 

Azadirachta indica (neem tree) Meliaceae leaves, sticks, bronchitis, diabetes, ulcers, haemorrhoides, skin India, Africa 

flowers, seeds, diseases, tumors, syphilis, antiseptic, dandruff, 

oil, bark contraception, dental care, insecticide 

Bombax buonopozense (bombax) Bombacaceae leaves antipyretic Africa 

Cinnamomum camphora/ C. Lauraceae leaves fever, eruptions, measles, delirium, whooping India, China, 

parthenoxylon (camphor tree) 

cough, melancholia, chronic bronchitis, uterine Sri Lanka 

pains, myalgia 

Cedrela odorata (cedro) Meliaceae bark snake bites, fever Peru, Brazil, 

East Africa 

Commiphora mukul 

Burseraceae stem, leaves, rheumatic disorders, hyperchhloroesterolaemia India 

(Indian bedellium tree) 

gum, resin 

Croton lechleri 

Euphorbiaceae latex swellings, gastric ulcers, contraception Peru 

(sangre de drago) 



215

Latin name Family Parts used 

Medicinal uses (treatment of the diseases 


Where grown 

at present? 

Emblica officinalis 

Euphorbiaceae fruits aging and general debility, acid-peptic diseases, hair India 

(Indian gooseberry) 

loss, dyspepsia, laxative, cooling, diuretic, ulcers 

Erythrina spp. (Indian coral tree) Fabaceae bark, leaves sedative, vulnerary, lactogogue, collyrium, sterility India, Africa 

in women, diabetes, dysentery, eye infections, 

insomnia, worms, joint pains, whooping cough 

Eucalyptus citriodora (lemon- Myrtaceae leaves essential oil, perfumery, mosquito repellant India, China 

scented gum) 

Eucalyptus spp. Myrtaceae leaves essential oil, colds India, China 

Euterpe precatoria 

Arecaceae roots diabetes, vaginal infections Peru 

(chonta, pana) 

Garcinia cola, 

Clusiaceae branch sticks, anthelmenthic, cardiotonic, astringent, demulcent, India, Africa 

G. afzelii, G. efunctata 

seeds, fruits emollient, antiobesity, dental care 

Ginkgo biloba 

Ginkgoaceae leaves, fruits old-age problems, memory enhancer, general tonic, China, Japan, 

(ginkgo) 

adaptogenic 

other east 

Asian 

countries 

Gliricidia sepium (Mexican lilac) Fabaceae leaves insecticide India, Africa 

Hagenia abyssinica (cusso) Rosaceae flowers anthelminthic, purgative Africa 

Jatropha curcas (furging tree) Euphorbiaceae leaves, stem, laxative, lactogogue, leprosy, rheumatism, eczema, India, Africa, 

seeds, oil blisters, inflammations, ear-ache 

China 

216 M.R. RAO R AND B.R. RR RAJESWARA RAO R

Leucaena leucocephala (lead tree) Mimosoideae 

bark, root emmengogue, ecbolic, depilatory, contraceptive India, Africa, 

America 

Madhuca longifolia 

Sapotaceae bark, heartwood, inflammations, sprains, pruritus, epilepsy, India 

(butternut tree) 

flowers, seeds strangury, verminasis, haemotysis, hepatopathy, 

dipsia, bronchitis, dermatopathy, cephalalgia, 

rheumatism, skin diseases 

Maytenus macrocarpa 

Celastraceae bark arthritis, diarrhea, stomach disorders, anemia Peru 

(chuchuhuasha) 

Melia volkensii, 

Meliaceae different parts insecticide, anthelmintic, antiseptic, astringent, India, Africa 

M. azadirach 

emetic, febrifuge, anti-rheumatic 

Okoubaka aubrevillei (oku) Octoknamaceae bark vomiting, influenza, infections, diarrhea, gastritis Africa 

Pausinystalia johimbe (yohimbe) Rubiaceae bark male impotency, aphrodisiac, hypotensive, cardio Central Africa 

tonic 

Prunus africana Rosaceae bark prostatitis Central Africa 

Saraca asoca (ashoka) Caesalpiniaceae bark, leaves, luecorrhoea, anthelmintic styptic, dyspepsia, ulcers, India 

flowers, seeds visceromegaly, pimples, cervical adenitis, vesicle 

calculi, haemorrhagic dysentery, diabetes 

Terminalia arjuna Combretaceae bark, leaves, wounds, ear-ache, heart diseases, fractures, India, Africa 

fruits 

contusions, febrifuge, dysentery, diuretic, tonic, 

deobstruent, hypertension 



217

Latin name 

Terminalia bellirica 

(beleric myrobalan) 

Family Parts used Medicinal uses (treatment of the diseases 

Where grown 


at present? 

Combretaceae bark, fruits, gum bronchitis, sore throat, biliousness, inflammations, 

strangury, asthma, astringent, dropsy, diarrhea, 

leprosy, gum purgative 

Combretaceae fruits asthma, soar throat, vomiting, hiccoughs, eye 

diseases, heart, and bladder diseases 

India, Africa 

Terminalia chebula 

India, Africa 

(cherubulic myrobalan) 

Raphia hookeri (raphia palm) Arecaceae exudates cosmetics, wine Africa 

Uncaria tomentosa 

Rubiaceae bark infections, cancer, gastritis, birth control, allergies Peru, Brazil 

(uña de gato) 

Zanthoxylum rhoifolium Rutaceae leaves, fruits chest infection, dental care, analgesic, antibacterial Brazil, 

Columbia 

Source: Saint-Pierre (1991), Chadha and Gupta (1995), and Rao et al. (1999; 2004). 

218 

M.R. RRAO AND B.R. RR RAJESWARA RAO R

Table 4. Shade-tolerant medicinal species of commercial potential that can be grown in homegardens. 

Latin name (common name) Family Habit Part(s) used Uses (treatment of the diseases mentioned) and 

other applications 

Ranunculaceae herb tuberous root astringent tonic, anti-diarrhea, dyspepsia, 

Where grown at 

present? 

Aconitum heterophyllum 

East Asia, 

(monks hood) 

cough, alexitexic, anti-periodic, anthelmenthic, Western 

hemorrhoids, general debility 

Himalayas 

Adhatoda zeylanica, 

Acanthaceae shrub leaves, roots, asthma, menorrhagia, psoriasis, chronic India 

A. beddomei (Malabar nut tree) 

stem bark bronchitis, cough, body inflammation 

Aloe vera, A. barbadensis (aloe) Liliaceae herb leaves health drink, burns, cuts, skin diseases, leprosy, Many countries 

piles, liver ailments, dysentery 

Alpinia galanga, A. calcarata Zingiberaceae herb rhizomes asthma, bronchitis, hiccoughs, dyspepsia, Malaysia, 

(galangal) 

diabetes, obesity, rheumatoid arthritis, Indonesia 

stimulant, tonic 

Ammomum villosum Apiaceae herb fruit stomachic, carminative, expectorant, tonic, China 

antiemetic, antispasmodic 

Andrographis paniculata Acanthaceae herb shoots antipyretic, antiperiodic, anti-inflammatory, India 

(king of bitters) 

ulcers, bronchitis, skin diseases, intestinal 

worms, jaundice, leprosy, hemorrhoids 

Asparagus racemosus, Liliaceae climber tuberous lactogogue, urinary and gynecological India 

A. adscendens (asparagus) 

roots disorders, diseases of nervous system, 

hyperacidity, gastritis, cardiac debility, 

hypertension, oligospermea 



219



Boerhaavia diffusa (hog weed) Nyctaginaceae creeping 

herb 

Cassia senna, C. acutifolia 

(senna) 

Catharanthus roseus 

(periwinkle) 

Centella asiatica (Indian 

pennywort/goticola) 

Chlorophytum borivilianum, 

C. tuberosum, 

C. arundinaceum (safed musli) 

whole plant aphrodisiac, diuretic, cardiac disorders, 

stimulant, diaphoretic, anti-inflammatory, 

jaundice, anemia, general debility, myalgia, 

scabies, oedema 


present? 

many tropical 

countries 

Caesalpiniaceae shrub leaves, pods constipation, skin diseases India, Sudan 

Apocynaceae small 

shrub 

leaves, roots cancer therapy (leaves), hypertension (roots) India, China, 

Apiaceae creeper whole plant memory enhancer, anxiety, neurosis, general India 

debility, wound healing, leprosy, eczema, 

sporiasis 

Liliaceae herb tubers aphrodisiac, nervine tonic, India 

Costus speciosus Zingiberaceae shrub rhizomes contraception, aphrodisiac astringent, digestive, India 

skin diseases, fevers 

Curculigo orchioides 

Amaryllidaceae herb roots erectile impotency, mm spermatorrhoea, general India 

(black musli) 

weakness, burning and fatigue piles, 

menorrhagia, jaundice 

Curcuma angustifolia Zingiberaceae herb tubers anti-diarrheal, anti-dysenteric, coolant, health India 

(arrow root) 

drink 

Central and South 

America 


Cymbopogon martinii var. Poaceae shrubby flowering perfumery, flavoring, joint pains, galactogogue, South and 

motia 

grass shoots febriluge, aromatherapy 

Southeast Asia 

Decalepis hamiltonii Asclepiadaceae twining tuberous health drink, tonic, promotes digestion, cures India 

straggler roots fever 

Dioscorea deltoides/ 

Dioscoreaceae tuberous tubers steroidal drugs, contraception, anthelmintic, India 

D. floribunda (medicinal yam) 

twines 

leprosy 

Gloriosa superba (glory lily) Liliaceae climbing tuber, seeds gout, polyploidy, rheumatism, abortifacient, India 

herb 

chronic ulcers, piles, diarrhea, antiperiodic, 

anthelmintic, snake bites, scorpion stings, 

gonorrhea 

Glycyrrhiza glabra (liquorice) Fabaceae shrub roots cough, general tonic, acid peptic disease anti- India, China, 

inflammatory, sweetener 

Eurasia 

Gymnema sylvestre (pepricola Asclepiadaceae climbing leaves antidiabetic, cardiac stimulant, eye diseases, India 

of the wood) 

shrub 

diuretic 

Hippophae rhamnoides Elaeagnaceae shrub berries skin care, analgesic, antioxidant, antibacterial, USA, Canada, 

(seabuckthorn) 

anti-inflammatory, nutraceutical 

Europe, India 

Holostemma adakodien Asclepiadaceae climber roots opthalmopathy, fever, arthritis, cough, burning India 

(swallow wort/ring coronet) 

sensation, stomachalgia 

Mucuna pruriens (velvet bean) Fabaceae climbing seeds Parkinson’s disease, anthelmintic, laxative, many countries 

shrub 

tonic for male virility, elephantiasis 



221




present? 

Ocimum sanctum 

Lamiaceae herb flowering cold, cough, bronchospasm, general debility, India, West 

(sacred basil) 

shoots stress disorders, skin infections, wounds, Indies 

indigestion, nausea, essential oil in flavoring, 

perfumery 

Phyllanthus amarus Euphorbiaceae herb whole plant hepatoprotective, oedema, anorexia many countries 

Piper longum (long pepper) Piperaceae climbing fruit, stem, bronchial asthma, throat infections, flatulence, India 

shrub roots dyspepsia, respiratory diseases, analgesic, 

carminative, sedative, insomnia, epilepsy, 

abortifacient 

Plumbago zeylanica, 

Plumbaginaceae herb roots acro-narcotic poison, abortifacient, rheumatic India 

P. rosea 

and paralytic affections, ulcers, leprosy, 

(white/red flowered lead wort) 

enlarged spleen, rubefacient, piles, skin 

diseases, leucoderma, syphilis, influenza 

Pueraria tuberosa 

Fabaceae large tubers arthritis, agalactia, cardiac debility, pharyngitis, India 

(Indian kudzu) 

climber 

leprosy, tuberculosis, spermatorrhoea 

Rauvolfia serpentina, Apocynaceae shrub roots leaves hypertension, insanity, insomnia, psychological India, Malaysia, 

R. tetraphylla, 

disorders, epilepsy eczema, skin diseases Indonesia 

R. vomitoria (serpentine root) 

opacities of cornea, psoriasis, snake and insect 

bites, toxic goiter, angina pectoris 


Rosemarinus officinalis Lamiaceae herb flowering perfumery, aromatherapy, digestive, nervine Mediterranean 

shoot tonic, stimulates kidneys 

region 

Salvia officinalis 

Lamiaceae herb leaves, oil perfumery, digestive, nervine tonic, antiseptic, Mediterranean 

(sage) 

deodorant, diaphoretic 

region 

Stevia rebaudiana (stevia) Asteraceae herb leaves sweetener Brazil, Japan, 

Paraguay, 

China, 

Indonesia, 

Thailand 

Tinospora cordifolia (tinospora) Menispermaceae woody stem seminal weakness, urinary affections, tonic, India 

climber 

fever, jaundice, syphilis, rheumatism, general 

debility, leprosy 

Source: Saint-Pierre (1991), Rao et al. (1999), Chadha and Gupta (1995), and Kehlenbeck and Maass (2005). 


223

224 


plantations spaced at 7.5 x 7.5 m, and gave 69 to 97% higher net returns compared 

to sole crops. The performance of Adhatoda beddomei and Holostemma adakodien, 

however, was unaffected by the cropping systems (Kurien et al., 2003), implying 

that they could perform well under disparate cropping situations. Patchouli 

(Pogostemon patchouli), an important aromatic crop, is grown as an intercrop in the 

coconut gardens of India. Its biomass yield and quality of oil were better under 

shade than when grown in the open (E.V.S. Prakasa Rao, CIMAP Resource Centre, 

Bangalore, India, pers. comm., July 2005). Black musli or golden eye grass 

(Curculigo orchioides) planted at 10 x 10 cm spacing under 25% shade performed 

better than the crop in the open in terms of vegetative growth, rhizome yield, harvest 

index, and nutrient uptake 4 . 

Most of the medicinal plants harvested from forests are shade tolerant or prefer 

some degree of shade, so that they can be cultivated in the homegardens as well, 

provided they are adapted to the prevailing climatic and soil conditions. A number 

of medicinal and aromatic crops that are traditionally grown outside forests can also 

withstand some shade (Jha and Gupta, 1991; Nair et al., 1991) and such species too 

can be integrated into homegardens. Tables 3 and 4 list a number of species that can 

be promoted in the homegardens. Species requiring mild shade may be grown in the 

early years of newly established homegardens or in patches under partial shade, 

whereas those that withstand intense shade can be grown in ‘mature’ homegardens. 

5. PROMOTING MEDICINAL CROPS IN HOMEGARDENS 

With the future of homegardens themselves being uncertain (Kumar and Nair, 2004; 

Wiersum, 2006), its role in providing a steady supply of medicinal plants and other 

products is unclear. Consistent with this, some reports indicate a reduction in the 

supply of MPs from homegardens. For example, an analysis of the species 

composition of homegardens in West Java, Indonesia in 1980 and 1999 revealed that 

fruit trees and ornamentals constituted a high proportion of plant species in both the 

years. There was, however, a decrease in the number of useful species from 126 to 

100 during the 1999 enumeration. The utilization of useful plants, except for fruit 

trees and plants for miscellaneous uses largely changed in the past 20 years 

especially in the case of vegetable, industrial, and ornamental plants (Kubota et al., 

2002). In Catalonia, MPs declined because of the loss of original significance of 

certain species and death of people with particular knowledge on the cultural 

requirements of some plants (Agelet et al., 2000). 

In spite of the above uncertainties, homegardens offer an opportunity to produce 

some high value medicinal crops and help smallholders earn additional incomes. For 

instance, in the Ba Vi National Park in northern Vietnam, the Dao people have taken 

up cultivation in the homegardens some of the 44 commercially important medicinal 

species identified in the area including Alstonia scholaris, Cinnamomum 

zeylanicum, Tradescantia zebrine, Piper retrofractum, and Travesia palmatet t (On 

et al., 2001). Ammomum villosum in China (Saint-Pierre, 1991) and Piper longum 

and Kaempferia galanga in India (Kumar et al., 2005) are similarly grown for 

commercial purposes. In the Peruvian Amazon, younger generations were as keen as 

the older ones to add species potentially useful as medicine, food, cosmetics, and


225 

other items to their collections as well as gathering knowledge on such plants 

(Padoch and de Jong, 1991). 

In the humid tropics, the active slash-and-burn agriculture (120 million ha), 

secondary forest fallow (203 million ha), logged forests (136 million ha), secondary 

forest fallows (203 million ha), Imperata-infested grasslands in Southeast Asia (40 

million ha), and degraded pastures in the Amazon (10 million ha) present vast 

degraded and abandoned areas, some of which can be put under permanent crop 

production systems (Sanchez et al., 1994). Homegardens and multistrata systems are 

regarded as some of the best bet alternatives to slash-and burn system both for the 

newly cleared lands as well as to bring degraded lands into permanent production. In 

the uplands of northern Vietnam, the need for improved homegardens using 

medicinal crops, rattan, quality timber, and livestock was recognized to replace 

shifting cultivation and to prevent opium m production (Tai et al., 1995). Homegardens 

have been taken up by smallholders in the re-settlement projects in Southeast Asia 

(e.g., Indonesia) and Amazon (e.g., Brazil). The native people and migrants in the 

course of developing their homegardens have used a wealth of plant materials 

including recently developed germplasm. A survey of 33 homegardens in the 

uplands and 18 in the floodplains of Brazilian Amazon revealed that a total of 77 

and 80 commercially valuable perennial species respectively are present (Smith, 

1996). These species included, in addition to those providing food, beverages, 

juices, nuts, oils, thatch, and wood, those that provided folk remedies such as juca 

(Caesalpinia ferrea), piao roxo (Jatropha gossypiifolia), yellow mombim or 

taperebá (Spondias mombim), fish bait (e.g., Colossoma macropomum, C. bidens, 

and Brycon sp), and piscicide (e.g., Ichthyothere cunabi). The species diversity was 

greater if medicinal, ornamental, and vegetable species meant mostly for family use 

were also considered. The number of such species in gardens ranged from 4 to 27. 

Homegardens established recently as alternatives to slash-and-burn agriculture in 

cleared forests or degraded lands, however, did not contain as many medicinal 

species as the traditional gardens. Similarly, recently established homegardens in 

southern Andaman, India did not contain medicinal plants (Pandey et al., 2002). 

Official recognition of traditional medicine will promote growing of medicinal 

plants, which in turn would help farmers earn better price to their products and 

citizens to get healthcare at reduced costs. Homegardens and health resorts could 

also promote ecotourism or ‘health tourism’, as is happening in the Kerala state of 

India (www.ktdc.com and www.keralatourism.org; last accessed: December 2005). 

The social benefits include revival of local traditions and protection of traditional 

knowledge. It is possible to patent indigenous knowledge about medicinal plants and 

preparations of products so that the society associated with the development of such 

knowledge derive the economic benefits thereof. Patenting of the stress relieving 

properties of Trichopus zeylanicus, a medicinal plant used by the Kani tribals of 

Agasthyar hills in Kerala is worth mentioning in this context (TBGRI, 2003). 

Indeed, a share of the royalty paid by the firm, which commercialized the 

technology, has been passed on to the tribal community that possessed this 

knowledge as part of their traditions. Value-addition and product development at 

local level wherever possible would also increase the earnings of farmers as well as 

create rural employment to skilled people and reduce migration to cities.

226 


6. MARKETING OF MEDICINAL PLANTS FROM HOMEGARDENS 

Local markets may not be adequate in most cases to absorb all the commercially 

valuable MPs and offer an equitable price to the producer; prices offered at these 

markets are often only a small fraction of those at the national and international 

markets. Lack of organized market channels, poor infrastructure, and involvement of 

middlemen in the supply chain from farm to factory deprive the farmers of 

remunerative prices to their produce. Strategies that will promote marketing of MPs 

and offer competitive prices to farmers are needed; these include establishment of 

farmers’ cooperatives, contract farming with ‘buyback’ arrangements by the industry, 

declaration of minimum support price to promising MPs, and subsidies to exporters 

of MPs as in other sectors. Examples of such proactive policies include development 

of a marketing network for Piper longum in Andhra Pradesh (India) and the 

intervention by government agencies in the case of Ammomum villosum in China, 

which encouraged large-scale cultivation of these MPs in homegardens. The Girijan 

Cooperatives in many Indian states also help the tribals living at forest margins to 

market non-wood forest products. Likewise, the Mayan farmers in the Yucatan 

region of Mexico have organized a cooperative project for the sale of aloe ( (Aloe 

barbadensis) and orange juice produced from forest gardens (Neugebauer and 

Mukul, 2000). Dabur India Ltd., a pharmaceutical company that makes herbal 

medicines, relies on contract farming for the supply of Indian gooseberry (Emblica 

officinalis), Rauvolfia, and Piper longum. Maintaining quality of the produce all 

through the supply chain is, however, very important to earn a premium price for 

which the farmers, transporters, and processors should be trained properly. 

7. OUTLOOK AND RESEARCH NEEDS 

Clearly, not all medicinal and aromatic plants found in the homegardens are used by 

people, and the relative importance of these plants to local societies also varies 

greatly from place to place. As a first step, therefore, priority species need to be 

identified based on their medicinal importance, ailments for which they are used, 

commercial value, cost effectiveness of alternate medicines, and the potential for 

synthesizing alternative compounds. Research efforts could then concentrate on a 

few priority species in terms of improving germplasm and developing agronomic 

techniques, particularly effective propagation techniques, and field establishment in 

homegardens and forest gardens. Sustainable harvesting methods have to be 

developed, especially for species harvested from the wild. 

Basic research is needed on the response of important medicinal species that are, 

and can be grown, in the homegardens to variations in quantity and quality of light; 

and to determine the effects of varying light regimes and organic and inorganic 

sources of nutrients on yield and quality. Such information helps to develop 

appropriate canopy management practices for multistrata systems to facilitate the 

growth of understorey crops. The use of MPs is based on indigenous knowledge and 

customs passed down from generations; the principal chemical compounds in many 

of these plants and their curative properties and mode of action have not yet been


227 

elucidated properly. Such studies will give authenticity to the use of traditional 

medicines and help protect genuine herbalists from unscrupulous practitioners. 

Globalization of agricultural trade under the World Trade Organization (WTO) 

regime brought with it several challenges and opportunities in the medicinal and 

aromatic plants sector too. The challenges include price competition, maintenance of 

quality, and scientific validation of claims for traditional medicines. The 

opportunities include global positioning of natural products obtained from medicinal 

plants, which have large demand. Bioprospecting for molecules of pharmaceutical or 

flavor/fragrance value from these plants and patenting of these molecules is going to 

be a future source of conflict between developed and developing countries. While 

the developed countries have the technology and fiscal resources, the developing 

countries in the tropics, where most of these MPs are grown, lack such resources. As 

a first step, therefore, tropical countries should make efforts to develop databases on 

MPs, indigenous medicinal practices, and herbal preparations in use. These will not 

only prevent loss of indigenous knowledge but also help promote the use of MPs. 

Documentation further helps native communities to protect their intellectual 

property rights on their genetic resources and indigenous knowledge systems and 

safeguard from biopiracy (Jose, 2004). 

7.1. Processing of homegarden produced MPs 

Medicinal and aromatic plants in homegardens can be produced at a lower cost 

compared to input intensive sole crops, as they benefit from common field 

operations and minimal use of chemical inputs. Organically produced MPs may also 

attract premium prices in the international markets. Processing and packaging of 

MPs at local level instead of selling the raw materials will further increase the value 

of the products and benefit the growers. Some typical value-addition practices are: 

(1) drying and powdering of relevant plant parts, (2) distillation of aromatic plants, 

(3) isolation of menthol crystals from mentha oil (Mentha spp.) following chilling 

and centrifuging, (4) pulverizing and encapsulation (e.g., peeled and dried tubers of 

Chlorophytum borivilianum in India), (5) preparation of herbal extracts, and (6) 

preparation of simple products such as incense sticks, perfumed candles, soaps, and 

herbal drugs. Powdering medicinal plant parts is the simplest activity, which can be 

taken up at the farm-level; e.g., tribals cultivating Curcuma angustifolia in Andhra 

Pradesh state, India, prepare a white powder from the tubers of this plant. Other 

processes may need establishment of facilities at village- or community-level as 

cottage industries. Nevertheless, it may increase profits to the farmers and generate 

employment to the local people. For instance, in Karnataka state of India, incense 

sticks are made mostly by women and children using plant-derived raw materials, 

adding value to these products and enhancing household incomes. Farmers have to 

be encouraged and trained, if necessary, to take such value-addition processes either 

individually or collectively at the farm- or village-level to realize better prices for 

their products. Good packaging, branding, organic labeling, and quality certification 

by authorized agencies for finished products will also increase the value of herbal 

medicines and its consumer acceptability.

228 

M.R. RAO R AND B.R. RRAJESWARA RAO R 


Homegardens will continue to be an important land use system for the small-scale 

farmers in humid and subhumid tropics. They can be turned into future ‘biofactories’ for 

the production of commercially important phytochemicals. Furthermore, organically 

grown MPs can be an important income and employment generating village 

enterprise in many rural localities. Promotion of ecotourism to herbal/homegardens 

and health resorts catering to aromatherapy or herbal therapy will have its spin off in 

terms of additional income and rural employment. Training farmers in improved 

cultivation and processing practices, contract farming, and establishment of 

institutions that provide market information and ensure quality standards will go a 

long way in promoting MPs in the homegardens. 


The authors sincerely thank Dr. J.S.K. Prasad, Central Research Institute for 

Dryland Agriculture (CRIDA), Santhoshnagar, Hyderabad 500 030, India, and Mr. 

Solomon G. Haile, School of Forest Resources and Conservation, University of 

Florida, Gainesville, Florida, USA for providing some useful literature in the course 

of preparing this paper. 

ENDNOTES 

1. World Bank stresses importance of coming phytomedicines. Newsletter of the 

Asian Network on Medicinal and Aromatic Plants. 23: 5 – 6 (1997). 

2. DGCIS 2004. Monthly statistics of foreign trade of India. Annual report for 

2003 – 2004 (Vol. 1). Exports including re-exports. Directorate General of 

Commercial Intelligence and Statistics, Ministry of Commerce, Kolkata. 

3. Emerging trends in productivity of medicinal and aromatic plants. Newsletter of 

the Asian Network on Medicinal and Aromatic Plants. 18: 7 (1996). 

4. Joy P.P., Savithri K.E., Mathew S. and Thomas J. 2005. Optimum shade and 

spacing for black musli (Curculigo orchioides Gaertn.). In: Book of Abstracts 

‘National seminar on achievements and opportunities in post-harvest 

management and value addition in roots and tuber crops’, 19 – 20 July 2005, 

Central Tuber Crops Research Institute, Thiruvananthapuram, Kerala, p114. 

REFERENCES 

Agelet A., Bonet M.A. and Valles J. 2000. Homegardens and their role as a main source of 

medicinal plants in mountain regions of Catalonia (Iberian peninsula). Econ Bot 54: 

295 – 309. 

Albuquerique de U.P. and Andrade L. de H.C. 2002. Conhecimento botánico tradicional e 

conservação em uma área de caatinga no estado de Pernambuco, nordeste do Brasil. Acta 

Botanico do Brasil 16: 273 – 285. 

Boonkird S.A., Fernandes E.C.M. and Nair P.K.R. 1984. Forest villages: an agroforestry 

approach to rehabilitating forest land degraded by shifting cultivation in Thailand. 

Agroforest Syst 2: 87 – 102.


229 

Ceccolini L. 2002. The homegardens of Soqotra island, Yemen: an example of agroforestry 

approach to multiple land use in an isolated location. Agroforest Syst 56: 107 – 115. 

Chadha K.L. and Gupta R. 1995. Medicinal and aromatic plants. Advances in horticulture 

(Vol 11). Malhotra Publishing House, New Delhi, 932p. 

Corlett J.L., Dean E.A. and Grivetti L.E. 2003. Hmong gardens: Botanical diversity in an 

urban setting. Econ Bot 57: 365 – 379. 

De Clerck F.A.J. and Negreros-Castillo P. 2000. Plant species of traditional Mayan 

homegardens of Mexico as analogs for multistrata agroforests. Agroforest Syst 48: 

303 – 317. 


gardens for sustainable livelihoods: Management strategies and institutional environments 

In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested example of 

sustainable agroforestry, pp 317 – 338. Springer Science, Dordrecht. 

EXIM Bank 2003. Export potential of Indian medicinal plants and products. Publication No. 

OP 98. Export and Import Bank of India (EXIM Bank), Mumbai, India (see also 

www.eximbankindia.com/publications; last accessed: October 9, 2005). 

Farnsworth N.R. and Soejarto D.D. 1991. Global importance of medicinal plants. In: Akerele 

O., Heywood V., and Synge H. (eds), The conservation of medicinal plants, pp. 25 – 51. 

Cambridge University Press, Cambridge. 

Fernandes E.C.M., O’kting’ati A. and Maghembe J.M. 1984.The Chagga homegardens: a 

multistoreyed cropping system on Mount Kilimanjaro (Northern Tanzania). Agroforest 

Syst 2: 73 – 86. 

Food and Agricultural Organization (FAO) 1996. Forests, food and health. 

www.fao.org/forestry/site /28813/em m (last accessed: October 10, 2005). 


America In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 


Jha K.K. and Gupta C. 1991. Intercropping of medicinal plants with poplar and their 

phenology. Indian For 117: 535 – 544. 

Jose R. 2004. US firm hijacks Kerala patent. www.huk.org/articles/0204/78.html (last 

accessed: October 10, 2005). 


Central Sulavesi, Indonesia. Agroforest Syst 63: 53 – 62. 

Kubota N., Hadikusumah H.Y., Abdoellah O.S. and Sugiyama N. 2002. Changes in the 

performance of the homegardens in West Java for twenty years. 2. Changes in the 

utilization of cultivated plants in the homegardens. Jpn J Trop Agr 46: 152 – 161. 

Kubota N., Shimamura K. and Ogo T. 1992. Useful plant species observed in homegardens, 

fields and local markets in Java and Sumatra islands. 2. Spice, medicinal, industrial and 

miscellaneous plants. Jpn J Trop Agr 36: 298 – 308. 

Kumar B.M. and Nair P.K.R. 2004. The enigma of tropical homegardens. Agroforest Syst 

61/62: 135 – 152. 


wood in the homegardens of Kerala in Peninsular India. Agroforest Syst 25: 243 – 262. 

Kumar B.M., Kumar S.S., and Fisher R.F. 2005. Galangal growth and productivity related 

light transmission in single-strata, multistrata and no-over-canopy systems. J New Seeds 

7: 111 – 126. 

Kumar B.M., Thomas, J. and. Fisher, R.F. 2001. Ailanthus triphysa at different density and 

fertilizer levels in Kerala, India: tree growth, light transmission and understorey ginger 

yield. Agroforest Syst 52: 133 – 144. 

Kurien A., Augustin A. and Nybe E.V. 2003. Economic analysis of resource-based cropping 

in selected medicinal species. In: Mathur A.K., Dwivedi S., Patra D.D., Bagchi G.D.,

230 

M.R. RAO R AND B.R. RRAJESWARA RAO R 

Sangwan N.S., Sharma A., and Kanuja S.P.S. (eds), Proceedings of the first national 

interactive meet on medicinal and aromatic plants, pp. 47 – 49. Central Institute of 

Medicinal and Aromatic Plants, Luknow. 

Lamont S.R., Eshbaugh W.H. and Greenberg A.M. 1999. Species composition, diversity and 


Maheswarappa H.P., Hegde M.R, and Nanjappa M.V. 1998. Kacholum (Kaempferia galanga) 

– a potential medicinal-cum-aromatic crop for coconut gardens. Indian Coconut J 

(Cochin) 29 (5): 4 – 5. 

Mendez V.E., Lok R. and Somarriba E. 2001. Interdisciplinary analysis of homegardens in 


85 – 96. 

Michon G., Mary F. and Bompard J. 1986. Multistoreyed agroforestry garden system in west 

Sumatra, Indonesia. Agroforest Syst 4: 315 – 338. 

Millat-e-Mustafa M., Khodeja Begum, Mohammed-Al-Amin and Shafiul Alam Md. 2001. 

Medicinal plant resources of the traditional homegardens in Bangladesh. J Trop Med 

Plants 2: 99 – 106. 

Millat-e-Mustafa M., Teklehaimanot Z. and Haruni A.K.O. 2002. Traditional uses of 

perennial homestead garden plants in Bangladesh. Forests Trees Livelihoods 12: 

235 – 256. 

Mpoyi K., Lukebakio N., Kapendo K. and Paulus J. 1994. Inventaire de la flore domestique 

des parcelles d’habitation. Cas de Kinshasa (Zaire). Revue de Mé et Pharmacopée 

Africaine 8(1): 55–66. 

Mulas M.G., Quiroz C., Perez S. D.M., Rodriguez D., Perez T., Marques A. and Pacheco W. 

2004. Conservacion in situ de diversas especies vegetales en ‘conucos’ (home gardens) in 

the states of Carabobo y Trujillo de Venezuela. Plant Gen Resour Newsl 137: 1 – 8. 






Nair G.S., Sudhadevi P.K. and Kurian A. 1991. Introduction of medicinal and aromatic plants 

as intercrops in coconut plantations. a In: Raychaudhuri S.P. (ed.), Recent advances in 

medicinal, aromatic and spice crops, pp 163 – 165. Today and Tomorrow’s Printers & 

Publishers, New Delhi. 

Neugebauer B. and Mukul Ek A. 2000. Trees for people – a Mayan strategy towards organic 

agriculture. In: Alfoldi T.T., Lockeretz W., and Niggli U. (eds), The world grows organic: 

Proceedings 13th International IFOAM scientific conference (28-31 August 2000), 

IFOAM, Basel, Switzerland, 428p. 

Ochea L., Fassaert C., Somarriba E. and Schlonvoight A. 1999. Medicinal and food plants in 

Nicoya, Costa Rica: there are differences in what men know and women know. 

Agroforest Today 11: 1–2, 11 – 12. 

Okafor J.C. and Fernandes E.C.M. 1987. The compound farms of southeastern Nigeria: a 

predominant agroforestry homegarden system with crops and small livestock. Agroforest 

Syst 5: 153 – 168. 

O’Kting’ati A., Maghembe J.A., Fernandes E.C.M. and Weaver G.H. 1984. Plant species in 

the Kilimajaro agroforestry system. Agroforest Syst 2: 177 – 186. 

On T.V., Quyen D., Bich L.D., Jones B., Wunder J., and Russel-Smith J. 2001. A survey of 

medicinal plants in Ba Vi National Park Vietnam: methodology and implications for 

conservation and sustainable use. Biol Conserv 97: 295 – 304. 


an Amazonian agricultural system. Econ Bot 45: 166 – 175.


231 

Pandey C.B., Kanak Lata, Venkatesh A., Medhi R.P. and Lata K. 2002. Homegardens: its 

structure and economic viability in South Andaman. Indian J Agroforest 4: 17 – 23. 

Perera A.H. and Rajapakse R.M.N. 1991. A baseline study of Kandian forest gardens of Sri 

Lanka: structure, composition and utilization. For Ecol Manag 45: 269 – 280. 

Principe P.P. 1991. Valuing the biodiversity of medicinal plants. In: Akeele O., Heywood V., 

and Synge H. (eds), The conservation of medicinal plants, pp 79 – 124. Cambridge 

University Press, Cambridge. 

Rao B.R.R. 1999. Medicinal plants for dry areas. In: Singh R.P. and Osman M. (eds), 

Sustainable alternate land use systems for drylands, pp 139 – 156. Oriental Enterprises, 

Dehra Dun. 

Rao M.R., Palada M.C. and Becker B.N. 2004. Medicinal and aromatic plants in agroforestry 

systems. Agroforest Syst 61/62: 107 – 122. 

Rao P.S., Venkaiah K. and Padmaja R. 1999. Field guide on medicinal plants. Forest 

Department, Government of Andhra Pradesh, Hyderabad, India, 208p. 

Rajapakse R., Rajapakse H.L. de Z. and Ratnasekera D. 2002. Effect of botanicals on 

oviposition, hatchability and mortality of Callosobruchus maculatus L. (Coleoptera: 

Bruchidae). Entomon 27: 93 – 98. 

Rico-Gray V., Chemas A. and Mandujano S. 1991. Use of tropical deciduous forest species 

by the Yucatecan Maya. Agroforest Syst 14: 149 – 161. 

Rugalema G.H., Okting’ati A. and Johnsen F.H. 1994. The homegarden agroforestry system 

of Bukoba district, Northwestern Tanzania. 1. Farming system analysis. Agroforest Syst 

26: 53 – 64. 

Saint-Pierre C. 1991. Evolution of agroforestry in the Xishuangbanna region of tropical 

China. Agroforest Syst 13:159 – 176. 

Samaniego G. and Lok R. 1998. Valor de la percepcion y del conocimiento local de indigenas 

Ngöbe, en Chiriqui, Panama. Agroforesteria en las Americas 5 (17/18): 12 – 16. 

Sanchez P.A., Woolmer P.L. and Palm C.A. 1994 Agroforestry approaches for rehabilitating 

degraded lands after tropical deforestation. In: Rehabilitation of degraded forest lands in 

the tropics: Technical approach. JIRCAS International Symposium Series No. 1, pp 

108 – 119. Japan International Research Centre for Agriculture Research, Tsukuba, 

Ibaraki. 

Smith N.J.H. 1996. Homegardens as a springboard for agroforestry development in 

Amazonia. Int Tree Crops J 9: 11 – 30. 




Tai N.D., Nhan H.D., Yen N.T. and Cameron D.M. 1995. Socio-economic considerations in 

the planning of agroforestry systems for the acid uplands of northern Vietnam. In: Date 

R.A., Grundon N.J., Rayment G.E., and Probert M.E. (eds), Plant–soil interactions at low 

pH: Principles and management. Proceedings of the Third International Symposium 

(September 12–16, 1993), pp 697 – 702. Brisbane, Queensland. 

Tropical Botanic Garden and Research Institute (TBGRI) 2003. TBGRI, Palode, 

Thiruvananthapuram, Kerala, www.tbgri.org/tbgri/patent/htm (last accessed: October 10, 

2005). 

Viquez E., Prado A., Onoro P., Solano R. and Solano A.R. 1994. Characterization of the 

tropical mixed garden ‘La Asuncion’, Masatepe, Nicaragua. Agroforesteria en las 

Americas 1(2): 5 – 9. 



World Health Organization (WHO) 2004. www.WHO.int/entity/mediacentre/news/notes/ 

2004/np3/en (last accessed: October 9, 2005)

232 





Yoshino K. and Ando K. 1999. Utilization of plant resources in homestead (bari-bhiti) in 

floodplain in Bangladesh. Japanese J Trop Agric 43: 306 – 318. 

Zemede A. and Ayele N. 1995.Homegardens in Ethiopia: Characteristics and plant diversity. 

Sinet-An Ethiopian J Sci 18(2): 235 – 266.

CHAPTER 13 

COMMERCIALIZATION OF 

HOMEGARDENS IN AN INDONESIAN 

VILLAGE: VEGETATION COMPOSITION 

AND FUNCTIONAL CHANGES 

O.S. ABDOELLAH 1* , H.Y. HADIKUSUMAH 2 , K. TAKEUCHI 3 , 

S. OKUBO 3 , AND PARIKESIT 2 

1 Institute of Ecology and Department of Anthropology and 2 Department of Biology, 

Padjadjaran University, Bandung, Indonesia. 3 Department of Ecosystem Studies, 

Graduate School of Agricultural and Life Sciences, University of Tokyo, Japan; * 

E-mail: 

Note: Adapted from: Abdoellah O.S., Hadikusumah H.Y., Takeuchi K., Okubo S. and 

Parikesit. 2006. Commercialization of homegardens in an Indonesian village: Vegetation 

composition and functional changes. Agroforestry Systems (in press). 

Keywords: Agricultural transformation, Cash crops, Owner preferences, Plant diversity. 

Abstract. With rapid development of Indonesia’s agricultural sector in response to market 

pressures, homegardens and other traditional forms of agriculture are increasingly being 

transformed into income-generating enterprises through the introduction of cash crops. We 

examined the impact of this commercialization on the structure and function of homegardens in the 

upland area of the Citarum watershed, West Java, Indonesia, and analyzed the ecological, social, 

and economic implications of these changes. Results of a vegetation survey and a survey of 94 

respondents indicated plant diversity in commercialized (intensively managed) homegardens 

decreased owing to the introduction of commercial crops. The change from subsistence to 

commercial farming was accompanied by decreased plant diversity, higher risks, higher external 

input use, increased instability, and reduced social equitability. The needs and preferences of the 

owners and market pressures were the main factors that triggered the development of intensive 

agriculture and increased the commercialization of homegardens. Commercialization adversely 

impacted the socio-cultural value that homegardens have traditionally provided to the society. 

Likewise, the long-term impacts and sustainability of commercial homegardens are also 

uncertain. 

233 




234 O.S. ABDOELLAH ET AL . 


Traditional homegardens have received special attention in Indonesia since the 

1970s, when the Institute of Ecology of Padjadjaran University discussed the role of 

these homegardens in rural development. Soemarwoto (1987) defined homegardens 

as a land surrounding houses in which the structure resembles that of a forest, 

combining the natural aspects of a forest with solutions to the socioeconomic and 

cultural needs of the people. Homegardens are centuries-old components of the rural 

ecosystems and are usually cultivated with a mixture of annual and perennial plants 

that can be harvested on a daily or seasonal basis. 

The structure of homegardens, however, varies from place to place according to 

the local physical environment, ecological characteristics, and socioeconomic and 

cultural factors (Christanty et al., 1986; Abdoellah, 1990; Karyono, 1990; Ceccolini, 

2002; Kumar and Nair, 2004). The high diversity of plant species in these 

homegardens and the mixture of annuals and perennials at different heights result in 

a complex horizontal and vertical structure. The multi-layered plant canopies prove 

to be beneficial in terms of the utilization of sunlight and in terms of water and soil 

conservation (Wiersum, 1982; Brownrigg, 1985; Torquebiau, 1992). 

Homegardens have several functions: economic, social and cultural, esthetic, and 

ecological (Abdoellah, 1990; Soemarwoto and Conway, 1991; Wezel and Bender, 

2003). In addition, the multiple uses of homegarden products contribute significantly 

to meeting the various needs (such as nutrition and income) of the households 

(Abdoellah and Marten, 1986; Christanty et t al., 1986; Abdoellah, 1990; Karyono, 

1990; Michon and Mary, 1994; Ceccolini, 2002; Blanckaert et al., 2004). Income 

derived from homegardens in West Java, for example, ranged from 6.6% to 55.7% 

of the family’s total income (Soemarwoto, 1987). The diversity of plants in 

traditional homegardens is beneficial from the nutritional point of view as well. 

Many of the plants are important sources of non-food necessities such as fuelwood 

and building materials too (see Shanavas and Kumar, 2003). 

Apart from their economic and ecological functions, rural homegardens also play 

important social roles (Abdoellah, 1990; Soemarwoto and Conway, 1991). For many 

rural people, the homegarden is an important place for socializing with family and 

neighbors. Many homegarden products also have social functions, since it is 

common for neighbors to let each other obtain such products freely. Many species 

are believed to have “magical” values or to serve as weather indicators. The 

homegarden is also an important status symbol; those who do not have their own 

homegardens and who must build their homes in another’s homegarden are 

considered poor. In the light of these multiple functions, many authors have 

concluded that homegardens are sustainable production systems (Karyono, 1990; 

Soemarwoto and Conway, 1991; Ceccolini, 2002; Nair, 2001; Wezel and Bender, 

2003; Blanckaert ett al., 2004; Kumar and Nair, 2004). 

However, during the rapid development of f Indonesia’s agricultural sector in 

response to market pressures, commercialization and the adoption of new 

technologies have been forcing major changes upon the agroecosystems, and 

homegardens are no exception to that general rule (Abdoellah et al., 2001; Kumar 

and Nair, 2004). Some villagers are already transforming their homegardens to meet

235 

the need for more cash as consumer goods become increasingly available. The 

introduction of commercial crops into this system to generate income is a potential 

source of structural and functional changes. Coincidentally, some homegardens have 

become dominated by few plant species; or have even become similar to 

monocultures. Examples include the gardens comprising of cash crops such as 

vegetables that are in high demand in urban markets. 

We examined the impact of this commercialization on the structure and function 

of homegardens in Sukapura village in the upland area of Citarum watershed, West 

Java, Indonesia. We addressed the following key questions: Does commercialization 

of homegardens affect their structure? Does it affect their social and economic 

functions? The answers to these questions will increase our understanding of the 

homegardens in relation to the multidimensional socioeconomic, ecological, and 

cultural dynamics of the people in this region. 

2.1. Study site 

2. METHODS 

The study was conducted in the Upper Citarum watershed, West Java, Indonesia 

during 2000 and 2001. With a total catchment area of approximately 6000 km 2 , the 

watershed covers seven districts, and its main river, the Citarum, runs approximately 

350 km northward from Mount Wayang to the Java Sea. This watershed, particularly 

in the upper part, has been experiencing rapid agricultural development since the 

1970s (after the Green Revolution), which caused major changes in its agricultural 

landscape, with a strong trend towards homogenization. 

Sukapura village, the study site, is located about 30 km southeast of Bandung 

Municipality and 20 km to the Majalaya sub-district, a center of the textile industry, 

in the Upper Citarum river basin. Most families in Sukapura depend on agriculture 

with little influence of urbanization. A major share of the total land area of the 

village (163.5 ha out of the administrative area of 187 ha) is devoted to agriculture 

consisting of cash crop gardens and mixed gardens. Being located at about 1250 m 

above sea level, the climate of the village is slightly cooler than in the lower part of 

the watershed and thus more suitable for cultivation of leafy vegetable such as 

Chinese cabbage (Brassica sinensis) and green onion ( (Allium fistulosum). 

The soil, 

which is an Andosol, volcanic in origin, is well drained and quite fertile. The village 

also has easy access, with an asphalt road, to nearby urban centers (Majalaya and 

Bandung), which allows the villagers to easily market their agricultural products. 

Because of these, cash crop cultivation in homegardens is nowadays quite common 

in this village. Thus, we considered Sukapura village as one of “typical” places that 

have the potential for homegarden commercialization. 

2.2. Sampling design 

COMMERCIALIZATION OF HOMEGARDENS IN INDONESIA I 

We interviewed respondents by using a standard questionnaire. In addition, we 

conducted a vegetation survey to characterize the composition and structure of the 

homegardens.

236 

Sample selection: We determined the required number of homegardens by the 

following formula of Lynch et al. (1974): 

n = 

NZ 2 p(1– p) 

Nd 2 +Z 2 p p(1– p) 

Where 

n = number of samples, 

N = number of households in the study village, 

Z = the value of the normal variable (1.96) for a confidence level of 0.95, 

p = the highest possible proportion (0.5), and 

d = the sampling error (0.1). 

Using the above formula, we randomly selected 94 households out of 3433 for 

interviews and for our vegetation survey. This total number of households was based 

on data obtained from the village office. In fact, landlessness in the study village 

was very high (>50%). Based on preliminary interviews with the 94 landowners, 

we defined the homegardens as commercial (if more than half of the products from 

the homegarden were sold for cash) or non-commercial (if more than half of the 

products were consumed by the family). Fifty-nine homegardens were thus found to 

be non-commercial and 35 commercial homegardens. We also collected data on 

household profiles, including main occupation, income from the homegardens, 

resources used as inputs, and the presence or absence of livestock, fences, and 

buruan (places in front of the house used for socializing and as playgrounds) by 

conducting interviews and through direct observation. Regarding income, the data 

obtained on annual and currency bases (Indonesian Rupiah: IDR) were converted 

into the value equivalent to rice weight at the rate of 1250 IDR per kg rice, which 

was the average of selling rate for the variety being cultivated in the village during 

2001. 

Vegetation survey: For the vegetation survey, we recorded the following data: 

species name, number of individuals of each species per plot/farm, number of 

structural layers based on plant height, and the plant category based on the main use. 

Vegetables were categorized into cash crops (for sale) and subsistence foods (for 

own consumption). Land utilization in the homegardens for nursery for plants and 

for growing cash crops was also recorded. 

To describe the dominance of a given species, we calculated the summed 

dominance ratio (SDR; Numata, 1966) for each species. This index was based on the 

density and frequency of the species. We also calculated plant diversity using the 

Shannon–Wiener diversity index and Pielou’s evenness index (Magurran, 1988). 

2.3. Data analysis 

O.S. ABDOELLAH ET AL . 

The vegetation and interview data for each homegarden were summarized, and the 

differences between commercial and non-commercial homegardens were compared 

using non-parametric tests in the SPSS for Windows software, Version 10.0 (SPSS


237 

Inc., Chicago, Illinois, USA). The Mann–Whitney U-test was used for the number of 

species and individuals, diversity and evenness indexes, area of the homegarden, 

ownership of other agricultural land, and income from the homegarden. For a 

bivariate comparison between the types of homegardens and the presence or absence 

of fences, livestock, and buruan, the use or non-use of external inputs, and either 

off-farm activities or on-farm activities as the main occupation, the Chi-square test 

was used. 

3.1. Structure of homegardens 

3. RESULTS 

The size of the homegardens averaged 341.73 m 2 , but commercial homegardens 

were larger than the non-commercial gardens (Table 1). The correlation between the 

number of species and the size of the homegarden was, however, poor for both types 

(Fig. 1a). The number of individual plants tended to increase with increasing size of 

the commercial homegardens, but not for the non-commercial ones (Fig. 1b). 

Table 1. Plant diversity parameters in commercial and non-commercial homegardens in 

Sukapura village, West Java, Indonesia. 

Structural attributes Commercial Non-commercial 

Area (m 

homegardens (n = 35) homegardens (n = 59) 

2 ) 

Average 461.54 270.66 

Range 

Number of species 

120 – 2000 85 – 1400 

Total 145 181 

Average 15.71 15.37 

Range 

Number of all plants 

4 – 49 4 – 41 

Total 42952 3893 

Average 1227.20 65.98 

Range 

Shannon–Wiener diversity index 

95 – 8388 6 – 159 

Average 1.11 2.03 

Range 

Pielou’s evenness index 

0.16 – 2.00 0.96 – 3.12 

Average 0.42 0.78 

Range 0.07 – 0.86 0.39 – 0.95 

The numbers of species and individual plants in the commercial and noncommercial 

homegardens are shown in Table 1. The total number of species found 

in both types of homegarden was 127 (out of 199 species; data not presented). There

238 


was no significant difference in the number of species per homegarden between the 

commercial and non-commercial gardens (Fig. 1a and Table 1); however, there was 

significant difference in the number of individuals between the two categories 

(U- U test, p

239 

between the number of individuals and area particularly in the commercial ones. A 

comparison of the area density (individuals/ha) between the two types of 

homegardens, showed that the average number in commercial ones was significantly 

higher than that in the non-commercial ones (27 154 and 3486 individuals/ha 

respectively; U-test, p < 0.01). The average Shannon–Wiener diversity and evenness 

indexes in commercial homegardens were significantly lower than that in the noncommercial 

homegardens (U-test, p < 0.01). The five most-dominant species in the 

commercial homegardens were vegetables (Table 2). Green onion ( (Allium 

fistulosum) had by far the highest SDR among these species. In the non-commercial 

homegardens Duranta erecta was dominant because of its use as a living hedge 

surrounding the homegardens. The numbers of species and their SDR values suggest 

that although the number of species did not differ significantly between homegarden 

types, the species distribution differed. 

Table 2. The five most-dominant species (based on the summed dominance ratio) in the 

commercial and non-commercial homegardens in Sukapura village, West Java, Indonesia. 

Type of 

homegarden 

and species 

rank order 


Dominant species Relative 

density 

Relative 

frequency 

Summed 

dominance 

ratio 

Commercial 

1 Allium fistulosum 39.27 3.27 21.27 

2 Daucus carota 12.15 2.00 7.07 

3 Ipomoea batatas 10.48 2.00 6.24 

4 Brassica sinensis 9.87 0.55 5.21 

5 Raphanus sativus 8.38 0.36 4.37 

Non-commercial 

1 Duranta erecta 15.75 3.64 9.69 

2 Manihot esculenta 4.98 1.87 3.43 

3 Psidium guajava 3.39 3.31 3.35 

4 Alternanthera philoxeroides 4.44 1.43 2.94 

5 Musa paradisiaca 3.06 2.32 2.69 

Based on the main use of each species, we defined eight plant categories in all 

sampled homegardens: vegetable, ornamental, food, fruit, spice, medicinal, building 

material, and “other” (Table 3). In commercial homegardens, vegetables were 

dominant, whereas ornamental plants were dominant in the non-commercial 

gardens. Table 4 presents the proportion of individuals in each plant category as a 

function of the size of the homegarden and it shows that the proportion of vegetables 

was highest for all sizes of commercial homegardens. These figures suggest that 

villagers who used homegardens for commercial purposes did so regardless of the 

size of the gardens. In addition, 65% to 93% of the total area of the commercial

240 

homegardens was planted with vegetable crops (data not shown). In contrast, in the 

non-commercial homegardens, inedible ornamental plants were dominant (Table 3). 

The relative proportion of the number of individuals in each plant category did not 

seem to be related to the size of the homegarden, but ornamental plants occupied the 

highest percentage for all size categories (Table 4). Even though the average number 

of individuals of ornamental plants was not different between the commercial and 

non-commercial homegardens, there were pronounced variations in this respect 

concerning vegetables and food crops (Table 4). 

Table 3. Dominance ratios of the main categories of plants in commercial and noncommercial 

homegardens in Sukapura village, West Java, Indonesia. 

Plant category Summed dominance ratio 

Commercial Non-commercial 

Vegetable 44.30 9.61 

Ornamental 23.63 56.51 

Food 14.53 7.85 

Fruit 11.30 16.76 

Spice 1.80 3.06 

Medicinal 1.35 2.46 

Building 1.25 1.52 

Other 1.84 2.23 

Total 100.0 100.0 

In terms of growth form, 88.6% of the individual plants in commercial 

homegardens occupied the first (ground) strata of the vegetation structure and were 

shorter than 1 m tall (Fig. 2). Of this, 90.1% comprised commercial crops such as 

Allium fistulosum, D. carota, Ipomoea batatas, Brassica sinensis, and Raphanus 

sativus. Figure 2a also indicates that the non-commercial homegardens kept the 

multistrata structure better than the commercial gardens. 

3.2. Functions of homegardens 


In general, homegarden functions depended on their species composition. In the 

commercial homegardens, the choice of species is determined largely by market 

demands. The number of respondents conducting off-farm activities as the main 

occupation was about 22% in the non-commercial category and about 11% for the 

commercial-homegarden owners; albeit t the differences were not significant (Table 

5; Chi-square test, p = 0.20). Moreover, based on the area of other agricultural lands 

owned by farmers, there was no difference between the types of the homegardens 

(U-test, U p > 0.05; n = 31 for the commercial homegardens and n = 46 for the noncommercial 

class).


241 

Figure 2. Vertical structural differences between the commercial and non-commercial 

homegardens in Sukapura Village, West Java, Indonesia. (a) Relative proportion of the 

number of species in each story of the vegetation structure to the total number of species. (b) 

Proportion of the total number of individuals in each story. 

Table 5 also shows the income derived from homegardens. The annual income 

from commercial homegardens was significantly higher than that from noncommercial 

homegardens (14 553 versus 2467 kg rice equivalent per ha). It is 

interesting to note that income per unit area in each sample was almost similar 

among the commercial gardens, but it varied among the non-commercial 

homegardens. The actual income in each sample (kg rice per year) had a significant 

and positive correlation with the area of the gardens only for the commercial 

category (Table 6). It was also significantly correlated with the number of 

individuals of all species, as well as the numbers of vegetable species, timber 

species (producing building materials), and food plants in commercial homegardens 

(Table 6). Although there was significant correlation between the income (kg rice 

per year) and the number of plants in non-commercial homegardens, it was only 

correlated with the number of fruit- and food plants (Table 6). These results imply 

that the main sources of income in the commercial homegardens were commercial

Table 4. Proportion of the individual plants in each of the eight plant categories as a function of the size of the homegardens in Sukapura village, 

West Java, Indonesia. 

Type of 

homegardens and 

area classes 

n 

Proportion of the number of individuals in each plant category (%) 

Vegetable Ornamental Food Fruit Spices Medicinal Building 

material 

Commercial 

< 100 m 2 0 – – – – – – – – 

101–200 m 2 7 45.9 (295) 23.2 (149) 25.6 (164) 1.5 (10) 0.6 (5) 0.2 (1) 0.0 (1) 3.0 (19) 

201–300 m 2 11 69.9 (410) 7.2 (43) 18.3 (108) 3.1 (19) 0.4 (3) 0.7 (5) 0.2 (1) 0.2 (1) 

301–400 m 2 5 81.8 (1061) 4.1 (54) 11.2 (145) 2.0 (26) 0.0 (1) 0.1 (1) 0.4 (6) 0.5 (7) 

401–500 m 2 5 74.2 (313) 15.5 (66) 6.8 (29) 2.9 (13) 0.0 (1) 0.2 (1) 0.1 (1) 0.2 (1) 

> 501 m 2 7 73.8 (2470) 2.3 (79) 22.6 (758) 0.8 (28) 0.0 (1) 0.0 (1) 0.4 (15) 0.0 (1) 

Non-commercial 

< 100 m 2 7 11.5 (6) 66.9 (35) 1.1 (1) 16.1 (9) 2.2 (2) 0.3 (1) 0.3 (1) 1.6 (1) 

101–200 m 2 24 5.3 (4) 60.3 (36) 12.4 (8) 15.5 (10) 0.8 (1) 1.2 (1) 2.1 (2) 2.4 (2) 

201–300 m 2 8 13.6 (10) 62.2 (44) 10.1 (8) 6.7 (5) 1.6 (2) 4.2 (3) 1.1 (1) 0.5 (1) 

301–400 m 2 7 5.9 (5) 57.5 (44) 14.3 (11) 16.4 (13) 2.5 (2) 1.0 (1) 0.4 (1) 2.1 (2) 

401–500 m 2 7 16.4 (13) 46.6 (35) 19.3 (15) 7.4 (6) 4.2 (4) 4.0 (3) 0.6 (1) 1.5 (2) 

> 501 m 2 6 7.6 (7) 66.6 (55) 1.0 (1) 12.1 (10) 10.5 (9) 0.6 (1) 1.6 (2) 0.0 (0) 

n = number of homegardens; parenthetic values are the average number of individuals per homegarden. 

Other 

242 O.S. ABDOELLAH ET AL .

243 

Table 5. Differences in socioeconomic data for the owners of commercial and noncommercial 

homegardens in Sukapura village, West Java, Indonesia. 

Socioeconomic attributes 

Type of homegarden 

Commercial 

(n = 35) 

Ownership a of agricultural lands (m 2 ) 

Paddy field 23.57 

(0 – 700) 

Crop field 881.20 

(0 – 1000) 

Mixed garden 116.00 

(0 – 2100) 

Total agricultural lands (area) 1020.77 

(0 – 11600) 

Income from homegarden b 

(kg rice equivalent ha –1 yr –1 ) (13757 – 14960) 

Off-farm activities as the main 

occupation c 

Use of external inputs in the 

homegarden c 

Existence of fences around the 

homegarden c 

Raising livestock in the 

homegarden c 

Presence of buruan in the 

homegarden c 


Noncommercial 

(n = 59) 

90.68 

(0 – 2400) 

653.05 

(0 – 6000) 

28.98 

(0 – 840) 

772.71 

(0 – 6000) 

Statistical 

significance d 

NS 

NS 

NS 

NS 

14565 2467 

(144 – 12318) 

** 

4 13 NS 

33 16 ** 

29 13 ** 

7 42 ** 

10 51 ** 

a 2 

Mean followed by range in parentheses (m ). 

b –1 

Mean followed by range in parentheses (kg rice equivalent ha a yr –1 

r ). 

c 

Number of respondents doing off-farm activities as the main occupation, using external 

inputs such as chemical fertilizers and pesticides in their homegardens, having fences around 

their homegardens, raising livestock in their homegardens and having a buruan in front of 

their homegardens. 

d 

For comparing commercial and non-commercial homegardens, we used the Mann–Whitney 

U-test for the three types of ownership of agricultural lands and for income from the 

homegarden and Chi-square test—for external inputs, existence of fences and livestock, and 

presence of a buruan; ** p < 0.01; NS = no significant difference or correlation. 

crops such as vegetables and timber species while in the non-commercial 

homegardens it was the fruit-producing species. This relatively higher income from 

the commercial homegardens reflected the change in function of homegardens from 

subsistence to commercial purposes. Based on our observation, the owners of

244 


commercial homegardens managed their homegardens much more intensively, for 

example, by routinely watering the homegarden plants and using external inputs 

such as chemical fertilizers and pesticides (Table 5). Thus, commercialization of the 

homegardens increased the demand for external inputs. Almost all respondents with 

commercial homegardens (94.3%) used these inputs to enhance crop yields and to 

protect crops from pests. There was a significant correlation between the use of 

external inputs and the type of homegarden: most commercial gardens used those 

inputs, while very few non-commercial gardens did. 

Table 6. Pearson’s correlation coefficient between the annual income from the 

homegardens and some homegarden characteristics in Sukapura village, West Java, 

Indonesia. 

Correlation coefficient to the annual 

income (kg rice equivalent yr –1 Homegarden characteristics 

) 

Commercial Non-commercial 

homegardens homegardens 

(n = 35) 

(n = 59) 

Size of homegarden 

Number of individuals 

1.00** –0.09 

Total 0.77** 0.31* 

Vegetable 0.72** 0.18 

Ornament 0.11 0.04 

Food 0.53** 0.26* 

Fruit 0.14 0.29* 

Spice –0.21 0.24 

Medicinal –0.06 0.15 

Building material 0.60** 0.05 

Other –0.15 –0.17 

To protect the commercial homegardens, 82.9% owners established fences, 

although the fences did not completely enclose the homegardens. In contrast, 78% of 

the owners of non-commercial homegardens did not establish fences (Table 5). In 

addition, 80% of the owners of commercial homegardens did not raise animals such 

as chickens, goats, and sheep in their homegardens, partly because of lack of space 

to raise the animals and build livestock pens, and partly because of the desire to 

protect their cash crops from grazing animals. Conversely, 71.2% of the owners of 

non-commercial homegardens raised livestock, and this difference was significant. 

Furthermore, 71.4% of commercial homegardens lacked a buruan where children 

could play in front of the house. The main reason for the decision not to create a 

buruan was, again, the lack of space and the worry that the children might damage 

the crops. In contrast, 86.4% of the non-commercial homegardens had a buruan, and 

this difference was significant.


4. DISCUSSION 

245 

Although the average number of species present did not differ significantly between 

the types of homegardens and many species were planted in both types, floristic 

composition of commercial homegarden was characterized by an increasing number 

of individuals of cash crops (vegetables) and a significantly decreasing diversity 

index. Implicit in this is that owners of both types of homegardens desired variety in 

products for both self-consumption and for sale, but that the latter goal probably 

outweighed the former for the owners of commercial homegardens. 

The total number of species found in all sampled homegardens in Sukapura did 

not differ from the results of the previous studies conducted in the lower part of the 

Citarum watershed. However, the dominant species in the present study (and 

especially those in the commercial homegardens) differed strongly from those of the 

previous studies. For example, in a study conducted by Chistanty et al. (1986), the 

homegardens greatly resembled the non-commercial homegardens of the present 

study, which were dominated by ornamental plants and had only few cash crops. 

This difference may have been strongly influenced by the specific needs and 

preferences of the landowners as well as by the different climatic and edaphic 

factors prevailing in the lower parts of the watershed. The fertile and well-drained 

soils and cooler climate of the upper watershed (present study) could have 

encouraged the local farmers to intensify the land use, including homegardening. 

The low correlation between the number of species and size of the homegardens 

in the present study suggests that homegarden size is probably not the main factor 

that governs species diversity. Instead, the structure and composition of the 

homegardens depended most likely on the role of various species required to fulfill 

the owner’s cultural, nutritional, social, and economic needs. For example, the fact 

that buruan were far more common in the non-commercial homegardens suggests 

that these landowners gave a high priority to the social and cultural roles 

traditionally supported by homegardens. 

Unlike in rural areas located at lower altitudes, the structure of the commercial 

homegardens in the present study was characterized by a more complex lower 

canopy, which, for example, was different from that described by Karyono (1990). 

In our study, some homegardens were dominated by only a few plant species that 

occupied the lower layers of the canopy structure, and some had even become 

monocultures, with the dominant species comprising cash crops such as vegetables 

that were usually found in the lowest layer (less than 1 m tall; 88.6% of the total). 

Interestingly, plants were more evenly distributed throughout the vertical structure 

of the non-commercial gardens. This indicates the presence of a multistrata canopy 

structure in most of these homegardens, as has been suggested by many others too 

(Karyono, 1990; Soemarwoto and Conway, 1991; Michon and Mary, 1994; 

Blanckaert et al., 2004). However, there were both inter-site and intra-site variations 

that complicate this inference. The specific needs and preferences of the owners 

were clearly important factors that influence the structure and the number of strata 

that were preserved or created in the homegardens (De Clerck and Negreros- 

Castillo, 2000).

246 


Many authors have pointed out that the structural pattern of the vegetation cover 

is influenced by specific physical circumstances, ecological characteristics, 

economics, and social and cultural factors (Christanty et al., 1986; Abdoellah, 1990; 

Karyono, 1990; Soemarwoto and Conway, 1991; Wezel and Bender, 2003). Although 

the landowners in our study were living under similar biophysical conditions, the 

structural pattern of the vegetation cover differed between the commercial and noncommercial 

homegardens (Fig. 2). Given the high degree of variation among gardens 

(Tables 1 and 4), there was clearly no single “typical” homegarden. Although tree 

species taller than 10 m were found in both types of homegardens, they were clearly 

more common in the non-commercial category (Fig. 2). These tree species were 

grown by the owners of non-commercial homegardens without distinct spatial 

arrangements; these owners grew big trees in any part of the yard and on any side of 

the house. It is likely that the owners tried to make the better use of available space 

in their homegardens, besides, tree planting is an old custom aimed to fulfill 

subsistence needs and, to some extent, to provide a restful micro-environment 

around the house. Besides, as the non-commercial homegardens often function as 

spots for social activities, the presence of tree canopies providing shade may 

facilitate such activities. In contrast, the owners of commercial homegardens mostly 

planted such trees in the backyard areas to mark the border of their gardens. It was 

very rare to find a tall tree in front of or at the side of a house in a commercial 

homegarden where cash crops were planted. According to the owners, this was 

because growing a big tree would inhibit their ability to grow commercial vegetable 

crops due to excessive shading. 

This difference suggests that the structural pattern of the vegetation in the 

homegardens was strongly influenced by the specific needs and preferences of the 

owners. We assumed that the owner whose main occupation was farming and who 

did not have much agricultural lands might commercialize his/her homegarden and 

vice-versa. However, there was no significant correlation between the commercialization 

of an owner’s homegarden and ownership of other agricultural lands or 

the main occupation status (Table 5). Thus, the structure of the homegarden 

depended on the owner’s management objectives as has been reported by several 

previous workers too (De Clerck and Negreros-Castillo, 2000; Mendez et al., 2001; 

Kumar and Nair, 2004). 

Many authors, such as Abdoellah (1990) and Soemarwoto and Conway (1991), 

have also reported that the increased intensity of cultivation of homegardens and the 

domination of these homegardens by particular species has reduced the overall 

number of plant species. However, our study showed that the number of species did 

not change significantly because of commercialization, but that the diversity index 

did indeed decrease, most likely because of the greatly increased number of 

individuals of certain species (Table 1). One consequence of the rising demand for 

better vegetable crops is that the species evenness has decreased substantially. The 

increased reliance on a limited number of species is likely to increase the risk of pest 

and disease outbreaks, as has already occurred in the sweet orange (Citrus nobilis) 

and clove (Syzygium aromaticum) orchards of Java. Soemarwoto and Conway 

(1991) reported that these crops, which were being extensively introduced in the 

homegardens, were already being severely damaged by Phyllosticta spp. and by


247 

citrus vein phloem degeneration disease, respectively. Similar problems have also 

been reported by Ceccolini (2002) in the homegardens on Soqotra Island. This 

suggests that commercialization of homegardens may eventually create ecological 

instability, leading to an increased incidence of pests and diseases. 

Furthermore, commercialization of homegardens by focusing on cash crops has 

resulted in only short-term improvements in farmers’ incomes. It is, however, not 

certain whether the high initial levels of productivity can be sustained. Cash crops 

also require high-energy inputs in the form of fertilizers and pesticides (Abdoellah, 

1990; Abdoellah et al., 2001). Our study confirmed that the use of these external 

inputs is significantly higher in the commercial gardens (Table 5), and that such 

increased use of chemical fertilizers and pesticides are inevitable for the commercial 

gardeners. Thus, although the gross income is higher, at least in the short term, net 

income will increasingly suffer and the long-term stability of this income is also 

uncertain, particularly for the products for which market demand fluctuates greatly. 

An additional consequence is the need for credit from banks and other sources of 

capital. Inadequate credit facilities in the public sector, however, have driven the 

villagers to unscrupulous middlemen and moneylenders, potentially leading to future 

changes in land ownership, and making the continued existence of somewhat 

autonomous homegardens doubtful. This seemed to reflect what has been stated by 

Michon and Mary (1994) that apart from high population density, major factors that 

threatened the existence of traditional homegardens in West Java were increased 

scarcity of agricultural lands, conflicts between commercial agriculture and 

traditional food production system, and development of a market economy. 

Soemarwoto and Conway (1991) also stated that the income generated from the 

sale of homegarden products tended to be used for ceremonies and other forms of 

consumption. There is also a danger that the dietary role of homegardens in 

providing protein, vitamins, and minerals may be neglected or even lost (Abdoellah 

and Marten, 1986; Soemarwoto and Conway, 1991; Wezel and Bender, 2003; 

Blanckaert et al., 2004), because traditional vegetables with low commercial value 

but high nutritional value may be the first to disappear from the commercial 

homegardens. Furthermore, commercialization of these homegardens has led to a 

decline in animal husbandry, thereby eliminating another source of nutrition that 

might compensate for the loss of these vegetables. These factors, taken together, 

undoubtedly decrease the ecological and economic sustainability of the commercial 

homegarden production system. 

Commercialization of homegardens has eliminated or reduced some of their 

multiple functions also. Traditionally, many products such as fruits, vegetables, and 

other useful plants were shared within the local communities (Abdoellah, 1990; 

Soemarwoto and Conway, 1991), thereby adding a unique social role to these 

homegardens (Kumar and Nair, 2004). Commercialization, however, has impeded 

this practice and has thus reduced the equitability of farming. Soemarwoto and 

Conway (1991) already pointed out that traditionally the Sundanese who live in this 

area have abided by the prospect of living harmoniously (rukun) with both relatives 

and other members of the community. Soemarwoto and Conway (1991) reported 

that an important way of expressing rukun was by offering useful homegarden 

products to relatives or neighbors daily, and particularly to the poor or unfortunate

248 


who needed this gift to survive, thereby maintaining, and strengthening social 

networks. Unfortunately, based on our interviews with several respondents, commercialization 

has decreased this sharing, even with relatives, and this has undermined the 

community’s social linkages, particularly concerning the poor. 

Commercialization of homegardens has forced more owners to establish fences 

around their homegardens (Table 5). Although these fences do not completely 

enclose the homegarden, they prevent people from entering or passing through 

freely, and force these people to request the owner’s permission to enter. This 

represents an important negative change from the traditional free access, as there 

was originally no concept of trespassing (Soemarwoto and Conway, 1991). This 

access has been retained in many non-commercial homegardens, which still mostly 

lack fences (Table 5). Most owners of non-commercial homegardens feel that 

establishing a fence around a homegarden is socially inappropriate, and that the 

owners of completely fenced homegardens are “conceited” (Soemarwoto and 

Conway, 1991). Furthermore, commercialization of homegardens has significantly 

decreased the number of buruans (Table 5). Children can no longer play in front of a 

house that lacks a buruan, thereby removing an important location for socializing 

with family and neighbors. Even more seriously, the buruan has traditionally been a 

place for children to learn cultural and social values from their elders (Soemarwoto 

and Conway, 1991). As a result of decreasing the sharing of products from 

homegardens and disrupting the social networks that are encouraged by free passage 

through homegardens and the existence of buruan, the commercialization of 

homegardens has done serious damage to the social fabric of these communities. 


Our results suggest that the homegardens of Sukapura village, in the Upper Citarum 

watershed of Indonesia, have changed dramatically over the past two decades. The 

ecological characteristics and social roles of these homegardens have been adversely 

affected, and the traditional system of sustainable agriculture that has kept people 

safe and well fed for centuries may no longer be sustainable without external inputs. 

Although income from commercialized homegardens has increased, these gardens 

have decreased plant diversity and evenness, heightened the ecological and financial 

risks to the owners, increased the requirements for external inputs such as fertilizers 

and pesticides, lowered community equitability, and increased overall instability. 

To revitalize the traditional functions of homegardens, we must convince the 

owners that the complex vegetation structure of these homegardens is more 

advantageous in the long-term, than the simpler and less stable structures of the 

commercial homegardens. In order not to go to the dangerous extent of full 

commercialization, heavily relying on external inputs as occurring in the research 

site, efforts should be made to improve the economic functions of the homegardens 

by manipulating their species composition. To succeed in this endeavor, a detailed 

analysis of the plant associations in traditional homegardens is required. This will 

provide a better knowledge of the ecological and economic compatibility of various 

plant species. The perceptions of landowners related to the preferred plant species 

must also change to reflect these findings, leading to new planting patterns based on


249 

improved selection of species, based on both their ecological roles and economic 

potential. Improved homegarden designs should also consider integrating crop-based 

activities with animal husbandry, both of which are crucial “social capital” for 

sustaining traditional homegardens and permitting future development. Post-harvest 

technology related to the products of these homegardens should also be investigated so 

as to reveal opportunities to add value to the products and create jobs that generate 

income without undermining the sustainability of the homegardens. Finally, 

supportive regional land use planning and management policies must be developed 

to encourage landowners to maintain the structure and function of traditional 

homegardens. 


This research was supported by the Institute of Ecology and the Research Institute of 

Padjadjaran University, and the Core University Program in Applied Biosciences 

funded by the Japanese Society for the Promotion of Science and the Directorate 

General of Higher Education, Indonesian Ministry of Education and Culture (Grantin-Aid 

for Scientific Research Category B #16405037, JSPS). The authors thank 

their students: Luppy Handinata, Deyna Handiyana, Dendi Muhamad, and Fazar R. 

Zulkarnaen, for assistance in fieldwork. This article was written while the first 

author was a visiting professor at the Laboratory of Landscape Ecology and 

Planning, Department of Ecosystem Studies, Graduate School of Agricultural and 

Life Sciences, University of Tokyo, Japan. 

REFERENCES 

Abdoellah O.S. 1990. Homegardens in Java and their future development. In: Landauer K. 

and Brazil M. (eds), Tropical home gardens, pp 69 – 79. United Nations University Press, 

Tokyo. 

Abdoellah O.S. and Marten G.G. 1986. The complementary roles of homegardens, uplands 

fields, and rice fields for meeting nutritional needs in West Java. In: Marten G.G. (ed.), 

Traditional agriculture in Southeast Asia: A human ecology perspective, pp 293 – 325. 

West view Press, Boulder and London. 

Abdoellah O.S., Takeuchi K., Parikesit, Gunawan B. and Hadikusumah H.Y. 2001. Structure 

and function of homegarden: revisited. In: Proceedings of First Seminar of JSPS-DGHE 

Core University Program in Applied Biosciences “Toward harmonisation between 

development and environmental conservation in biological production” (21–23 February 

2001), pp 167–185. University of Tokyo, Tokyo. 

Blanckaert I., Swennen R.L., Paredes Flores M., Rosas López R. and Lira Saade R. 2004. 


Coxcatlan, Valley of Tehuacan-Cuicatlan, Mexico. J Arid Environ 57: 179 – 202. 

Brownrigg L. 1985. Home gardening in international development: what literature shows? 

The League for International Food Education, Washington, DC, 330p. 

Ceccolini L. 2002. The homegardens of Soqotra islands, Yemen: an example of agroforestry 


Christanty L., Abdoellah O.S., Marten G. and Iskandar J. 1986. Traditional agroforestry in 

West Java: the pekarangan (homegarden) and kebun-talun (perennial/annual rotation)

250 


cropping systems. In: Marten G.G. (ed.), Traditional agriculture in Southeast Asia: A 

human ecology perspective, pp 132 – 156. Westview Press, Boulder and London. 

De Clerck, F.A.J. and Negreros-Castillo P. 2000. Plant species of traditional Mayan 

homegardens of Mexico as analogs for multistrata agroforests. Agroforest Syst 48: 

303 – 317. 

Karyono 1990. Home gardens in Java: their structure and function. In: Landauer K. and Brazil 



135 – 152. 

Lynch F., Hollnsteiner M.R. and Corvar L.C. 1974. Data gathering by social survey. Trial 

edition. Social Science Council Inc., the Philippines, Quezon City, 227p. 

Magurran E.A. 1988. Ecological diversity and its measurement. Princeton Univ. Press, New 

Jersey, 179p. 

Mendez V.E., Lok R. and Sommarriba E. 2001. Interdisciplinary analysis of homegardens in 


85 – 96. 



31 – 58. 



Numata M. 1966. Ecological judgment of grassland condition and trend. II: Judgment by 

floristic composition (in Japanese with English abstract). J Jpn Soc Grassland Sci 12: 

29 – 36. 

Shanavas A. and Kumar B.M. 2003. Fuelwood characteristics of tree species in homegardens 

of Kerala, India. Agroforest Syst 58: 11 – 24. 




Soemarwoto O. and Conway G.R. 1991. The Javanese homegarden. J Farming Syst Res Extn 

2: 95 – 118. 

Torquebiau E. 1992. Are tropical homegardens sustainable? Agric Ecosyst Environ 41: 

189 – 207. 



Wiersum K.F. 1982. Tree gardening and taungya in Java: examples of agroforestry 

techniques in the humid tropics. Agroforest Syst 1: 53 – 70.

CHAPTER 14 

TRANSPIRATION CHARACTERISTICS 

OF SOME HOMEGARDEN TREE SPECIES 

IN CENTRAL SRI LANKA 

W.A.J.M. DE COSTA*, K.S.P. AMARATUNGA, 

AND R.S. UDUMULLAGE 


Peredeniya 20400, Sri Lanka; *E-mail: 

Keywords: Artocarpus heterophyllus, Cedrela toona, Radiation interception, Soil water 

deficit, Swietenia macrophylla, Vapor pressure deficit. 

Abstract. Deep-rooted trees that dominate the multilayered homegardens (MHG) in Central Sri 

Lanka might adversely impact the catchment water yield because of their high transpiration rates. 

Our objectives were to quantify the water use of three representative tree species in an MHG and to 

identify the major determinants of transpiration. The species were Artocarpus heterophyllus 

(diameter at breast height, DBH = 40.5 cm), Cedrela toona (DBH = 9 cm) and Swietenia 

macrophylla (DBH = 3 cm) representing the upper, middle and lower canopy layers respectively of 

an MHG in the high-rainfall zone of central Sri Lanka. Transpiration was measured as trunk sap 

flow rate using thermal dissipation probes in Artocarpus s and Cedrela and sap flow gauges in 

Swietenia. Measurements during a 72 h period, when soil moisture was not limiting, showed that 

sap flow of Artocarpus s was significantly greater than those of Cedrela and Swietenia. Daily 

transpiration of Cedrela and Swietenia ranged from 19% to 27% of that in Artocarpus and it 

increased linearly with incident solar radiation and saturation vapor pressure deficit (VPD). 

Measurements during a 54-day rainless period also showed that Artocarpus s and Cedrela had high 

transpiration rates despite reduced water availability in the top 1 m soil layer, indicating water 

extraction by roots from deeper horizons. Transpiration rate increased with increasing irradiance up 

to 13 MJ m –2 2 –1 

m d and with increasing VPD up to 0.8 kPa. Decreases in transpiration at irradiances 

and VPDs greater than the above values indicated that stomata had begun to exert significant 

control over water loss from the foliage canopy. 


Multi-layered homegardens (MHG) are a common agroforestry system, which 

covers a considerable part of the Central Province of Sri Lanka. It is the 

251 




252 

W.A.J.M. DE COSTA ET AL . 

predominant land use system in the Kandy district of Central Sri Lanka, where this 

traditional system of perennial cropping has been practiced for several centuries, and 

hence known as the ‘Kandyan forest gardens’ (Jacob and Alles, 1987). The MHGs 

consist of a mixture of trees of varying heights, canopy spreads, life cycle durations, 

and uses. The high diversity of plants in a Kandyan MHG offers economic stability 

to farmers because of the wide range of economically important products obtained 

from them such as food, beverage and fruits, spices and timber 1 . 

The predominantly perennial vegetation in MHGs could absorb a significant 

amount of water from the soil through their deep root systems. This water is 

ultimately transpired from the foliage after it is translocated through the xylem 

system. Therefore, water use by MHGs forms an important consideration in the 

catchment water balance of the Central Province of Sri Lanka, which is the most 

important rainfall catchment on which much of the Sri Lankan agriculture and 

power generation depend on. Catchment water balance of this region thus exerts a 

critical influence on the national economy. Strong evidence available from different 

parts of the world suggests that high rates of transpiration by forests and other taller, 

perennial vegetation could lead to decreased water yields in rivers and reservoirs 

(Calder, 1996). 

Following an analysis of 94 forest catchment experiments in different parts of 

the world across a range of climates, Bosch and Hewlett (1982) concluded that a 

10% increase in forest cover would decrease annual water yield by 40 mm for Pinus 

and Eucalytpus and by 25 mm for deciduous hardwood forests. Although such 

reports generally predict reduced streamflow and ground water availability in areas 

planted with Pinus and Eucalytpus, it could be argued that such reductions may not 

occur in the humid tropics because of their high rainfall. This question has been 

frequently raised in Sri Lanka and based on limited and incomplete data 2,3 , it was 

concluded that water use of forests and taller perennial vegetation in the Central 

Province was greater than that of grasslands and annual cropping systems. 

Quantification of water use of MHGs and determination of its impact on catchment 

water balance are difficult because of the complex and diverse vegetation structure 

of MHGs. The total water use of an MHG would be the sum of transpiration rates of 

different tree species whose canopies are arranged into different vertical layers 

(Jacob and Alles, 1987). As a first step in quantifying the total water use of MHGs, 

transpiration rates of a limited number of trees representing different vertical layers 

in the canopy structure was measured. Our specific objective was to relate the 

measured transpiration rates to the microenvironmental factors in the relevant 

canopy layers. Developing mechanistic relationships between transpiration and its 

driving environmental factors could pave the way to estimate water use of MHGs 

through an adequately comprehensive modeling approach. 

Transpiration rates depend on soil water availability (Monteith, 1986; Jarvis and 

McNaughton, 1986; Wallace, 1996); and in the absence of soil water deficits it is 

primarily driven by microenvironmental factors such as incident solar radiation and 

vapour pressure deficit. With increasing soil water deficits, stomata may exert a 

significant control over transpiration by partial closure. Therefore, a secondary 

objective of our study was to see whether stomata could exert an appreciable control 

over water use of the MHGs during rainless periods of gradually increasing soil

253 

water deficits. Considering the widely predicted reductions in rainfall and water 

availability (McCarthy et al., 2001), this information would be relevant to the 

discussion on sustainability of MHGs. 

2.1. Experimental site 

TRANSPIRATION 

T CHARACTERISTICS C OF HOMEGARDEN TREES T 


The study was conducted from June 2001 to February 2002 in a typical multilayered 

homegarden in Kandy (7 o 17’N, 80 o 40’E), in the central highlands of Sri Lanka. The 

site is in the mid-elevational (500 m above sea level), high rainfall (2000 mm yr –1 ) 

zone, which is known as the Mid-country Wet Zone, WM2, according to the 

classification of agro-ecological zones of Sri Lanka by Panabokke (1996), with a 

mean daily temperature in the range of 25 to 27 o C and relative humidity between 70 

to 90%. 

This agro-ecological zone contains gently to steeply undulating terrain consisting 

of hills and valleys. Multilayered homegardens are located in the hills while the 

valleys are cultivated with lowland rice (Oryza sativa). Soil in the hills is deep and 

well-drained, dark reddish brown with a sandy clay loam texture; it is classified as 

‘Reddish Brown Latosolic Soils’ according to the local classification (Panabokke, 

1996), Dystric Cambisols by FAO/UNESCO and Typic Troporthents by USDA 

(Senarath and Dassanayake, 1999). Bulk density ranges from 1.50 x 10 3 to 1.60 x 

10 3 kg m –3 

m and pH from 5.0 to 5.5 throughout the soil profile. Available water, i.e. 

soil water content between field capacity (–0.01 MPa of soil water potential) and 

permanent wilting point (–1.5 MPa) is 111.4 mm m –1 

m . 

2.2. Composition of the homegarden 

The area of the MHG selected for this study, including the house within, was 0.15 

ha. The house was surrounded by vegetation consisting mainly of trees varying in 

heights, canopy spread and depth, trunk diameter and age. The woody perennials at 

the study site were enumerated by measuring their height and diameter at breast 

height (DBH). The probable age of trees was ascertained from the garden owner. 

The MHG had 56 trees belonging to 14 species (Table 1), most of which were of 

economic value providing food (Artocarpus ( heterophyllus and Cocos nucifera), 

beverages (Coffea arabica), fruits (Mangifera indica and Persea americana), spices 

(Syzygium aromaticum and Myristica fragrans), timber (Swietenia macrophylla and 

Artocarpus heterophyllus) and variety of other products and services. 

The trees were categorized into three groups based on their canopy position: top 

layer (10 m or taller), middle layer (5 to 10 m), and bottom layer (less than 5 m). 

Three representative trees occupying these three vertical layers were selected for 

measuring transpiration rates. The species were Artocarpus heterophyllus, Cedrela 

toona, and Swietenia macrophylla representing the top, middle, and bottom layers, 

respectively. Main attributes of the three species are given in Table 1. The trees were 

located within a 15 m radial zone to avoid using excessively long wires conducting

254 

voltage signals from sapflow sensors and to have the data-logger in a central 

location. The trees were selected based on their characters at the time of 

transpiration measurement. For example, a sapling of Swietenia was selected to 

represent the bottom canopy layer because of its lower height. However, this does 

not imply that Swietenia always would occupy the lower canopy layer of MHGs. 

Among the several tree species occupying the top layer, Artocarpus heterophyllus, 

which is the most abundant species in Kandyan MHGs 4 , was selected. 

Table 1. List of tree species in a multilayered homegarden in central Sri Lanka and their 

structural attributes. 

Sl. 

No 

Tree species Height 

(m) 


Canopy 

depth 

(m) 

Canopy 

spread 

(m) 

DBH 

(cm) 

1 Persea americana 7.5 6 3.5 16 10 

2 Artocarpus heterophyllus 15.5 14 14 55 40 

3 Gliricidia sepium 13 11 3 14 27 

4 Persea americana 11 6 5 18.5 10 

5 Mangifera indica 12.5 8.5 5.5 17 10 

6 Cedrela toona 8 4 1.5 8.5 8 

7 Persea americana 13 10 4.5 25.5 11 

8 Swietenia macrophylla 15 6 5 37 15 

9 Swietenia macrophylla 15.5 9.5 6.5 46 15 

10 Cocos nucifera 15 6 9.5 30 10 

11 Swietenia macrophylla 1.9 1.3 0.9 1 3 

12 Cocos nucifera 15 6 9 24.5 10 

13 Cocos nucifera 14 5.75 9 28.5 10 

14 Cocos nucifera 16 7 9 32 10 

15 Swietenia macrophylla 4 1.5 0.6 3.5 3 

16 Alstonia macrophylla 4 1.5 2 3 5 

17 Alstonia macrophylla 9.5 7.5 4 9.5 2 

18 Persea americana 9 7 4 15.5 5 

19 Artocarpus heterophyllus 15 13 10 40.5 25 

20 Persea americana 8.5 4 2 17 8 

21 Cedrela toona 8 1.9 2.8 9 8 

22 Cedrela toona 8.75 4.5 2.3 9.5 8 

23 Alstonia macrophylla 8 3 4.2 7.5 5 

24 Michelia champaca 13 8.5 5 18 10 

25 Swietenia macrop o hylla 6.5 2.5 2 8 8 

26 Swietenia macrophylla 6.5 2.5 3.5 10.5 8 

27 Swietenia macrophylla 7 2.5 2.5 10 8 

28 Cedrela toona 4.3 0.25 1.5 3 8 

Probable 

age (yr)

TRANSPIRATION 


29 Cedrela toona 4.3 0.25 1 3.2 8 

30 Cedrela toona 4 0.25 0.5 3 8 

31 Myristica fragrans 4.3 1.7 1.4 3 8 

32 Michelia champaca 13.5 7.5 6 30 8 

33 Alstonia macrophylla 8 4.5 5 11 15 

34 Swietenia macrophylla 7 4.3 2 7 8 

35 Swietenia macrophylla 7 4 1.8 6.5 8 

36 Cedrela toona 5 0.75 2 4 8 

37 Swietenia macrophylla 7.5 5 3 11 8 

38 Swietenia macrophylla 7.5 5 1.5 7.5 8 

39 Litsea iteadaphna 7.5 6 5.5 22.2 15 

40 Michelia champaca 14 6 4.7 16.5 10 

41 Cedrela toona 10 * * 7 8 

42 Caryota urens 13 6.5 6 33.5 8 

43 Cocos nucifera 14.5 5 9 30 10 

44 Cedrela toona 13.5 0.5 2.8 3 8 

45 Syzygium aromaticum 13 9.5 6 21.5 25 

46 Coffea arabica 7.5 6.5 3.5 5.25 5 

47 Cocos nucifera 14 6.5 9.5 27.5 10 

48 Persea americana 8 5 5.5 15 6 

49 Gliricidia sepium 7 6.5 3 6.5 15 

50 Duria zibethinus 22 7 10.2 40 28 

51 Gliricidia sepium 8 5 2 12 14 

52 Cocos nucifera 13.5 5 4.5 22 28 

53 Alstonia macrophylla 17 7 1.5 15 8 

54 Psidium guajava 1.9 1.5 1.5 2 10 

55 Swietenia macrophylla 1.9 1 1.2 1 3 

56 Cassia roxburghii 23 1.2 7 45 40 

Note: The trees selected for transpiration measurement are indicated in bold; DBH = diameter 

at breast height. 

2.3. Measurement of transpiration rates of different trees 

255 

Among a variety of techniques available for measuring transpiration rates in trees, 

thermal methods (Swanson, 1994) involve measurement of the rate of heat flow in 

the xylem transpiration stream. Under steady-state conditions, the rate of xylem sap 

flow would be equal to the rate of transpiration of a tree (Van den Honert, 1948; 

Cowan, 1965). An added advantage of thermal methods is their ability to measure 

transpiration rates of individual trees, which is essential in a species mixture such as 

the MHG. In the present study, we used two thermal methods, thermal dissipation 

probes and stem sap flow gauges (Dynamax Inc, USA/Delta-T, UK), the former for 

Artocarpus (DBH = 40.5 cm) and Cedrela (DBH = 9 cm), and the latter for

256 


Swietenia. Both instruments measure the rate of sap flow in the xylem, but the 

thermal dissipation probe (TDP) relates the rate of heat dissipation to sap flow rate 

through an empirical relationship, while the sap flow gauge directly calculates the 

rate of sap flow as the residual of an energy balance equation. 

The TDP consisted of two ‘needles’, each 30 mm long and 1.2 mm in outer 

diameter, spaced 40 mm apart; these were inserted horizontally to the trunk at a 

height of 1.2 m above the ground. Both needles contained thermocouples to measure 

the average sap temperature. In addition, the upper needle contained a heating 

element, which was supplied with a constant input of energy (0.15 Js –1 ) by a 12 V 

car battery through a voltage regulator (AVRD Dual Regulator, Delta-T Devices). 

As the lower un-heated needle measured the reference sap temperature, temperature 

difference (dT) T between the two needles was determined by rate of heat dissipation 

due to sap flow. As such, dT T was inversely related to sap velocity. It was measured 

and recorded as a voltage signal every 30 seconds, averaged every 5 minutes and 

stored in a DL2e data logger (Delta-T Devices). Equation 1 (Granier, 1985; 1987) 

was used to compute a dimensionless ‘flow index’ K as follows, 

K = (dTm – dT)/dT (1) 

Where dTm is the maximum recorded value of dT T at times of zero sap flow. In the 

present experiment, the maximum dTm values were recorded between midnight and 

0400 h. Mean dTm during this four-hour period each day was taken as dTm for the 

–1 

respective day. K was related to the average sap velocity, V, V in cm s , by an 

empirical relationship of Granier (1985; 1987) as, 

V = 0.0119 K 1.231 

K (2) 

This relationship, which did not differ significantly among tree species, was used in 

the present study. V was converted to sap flow rate, F , as, 

V Fs 

Fs F = AsV (3) 

Where As is sap wood cross-sectional area, which was measured by taking several 

core samples using an increment borer on Artocarpus and Cedrela trees that were 

not being used for sap flow measurements. As recommended by the manufacturer 

(Dynamax), three thermal dissipation probes were used at different locations around 

the trunk of Artocarpus while two probes were used on Cedrela. Signals from 

different probes were averaged before sap flow calculations. 

The Dynagage (SGB25-ws) consisted of a heating plate, 110 mm high and 28 

mm in diameter, which was wrapped around the stem of sapling Swietenia at a 

height of 1.2 m above the ground. A constant, regulated power input (Pin) of 0.5 Js –1 

was supplied to the heating plate. A set of precise electronic sensors attached to the 

gauge measured the radial heat transfer from the gauge to the ambient air (Qr) and

TRANSPIRATION 


257 

axial heat fluxes through the trunk (Qv). The variable amount of heat carried by the 

sap flow (Qf ) was calculated from an energy balance equation (Sakuratani, 1981; 

Baker and Van Bavel, 1987) as, 

Qf = Pin – Qr – Qv (4) 

Qf f can be converted to the sap flow rate ( F) by dividing it with specific heat capacity 

of water (Cp) and sap temperature increase (dT ) as, 

F = (Pin – Qr – Qu – Qd)/(Cp dT) (5) 

Where Qu and Qd d are the upward and downward components of the axial heat 

transfer (Qv). Voltage signals from sensors were recorded every 30 seconds, 

averaged every 5 minutes and stored in the data logger. 

2.4. Measurement of environmental variables 

Solar radiation incident on the canopies was measured by tube solarimeters installed 

at appropriate heights above and within the MHG. Relative humidity and air 

temperature at the top of the three selected trees were measured by solid-state 

sensors (TDK Inc., Hiroshima, Japan) installed at the respective heights. Output 

signals from all these sensors were recorded every 30 seconds, averaged every 5 

minutes and stored in the data logger. Relative humidity (h) and air temperature (Ta T ) 

data were used to compute the saturation vapor pressure deficit (VPD) of air as, 

VPD = es(Ta) – e (6) 

Where es(Ta) is the saturation vapor pressure at air temperature Ta T and e the actual 

vapor pressure. es(Ta) was computed by an empirical equation developed by Tetens 

(1930) and adopted by Murray (1967) as, 

[17.27 [(T – 273)/(T – 36)] 

es(Ta) = 0.611 e 

The actual vapor pressure (e) was calculated from measured relative humidity and 

es(Ta) as, 

(7) 

e = (h es(Ta))/100 (8) 

Soil moisture content was measured gravimetrically at weekly intervals by taking 

samples at 20 cm depth intervals down to a maximum depth of 1 m. Soil moisture 

was measured at a point approximately equidistant from the three trees on which 

transpiration was being measured.

258 

2.5. Computation of canopy conductance 

Canopy conductance (gc) is the overall stomatal conductance of the entire foliage 

canopy. It was computed by inverting a simplified version of the Penman-Monteith 

equation adopted for tall vegetation by Granier et al. (1996) as, 

gc = (F )/( ( Cp D) (9) 

–2 –1 

Where F is mean sap flow rate (kg m s ), the latent heat of evaporation of water 

(J kg –1 ), the psychrometer constant (kPa K –1 ), the density of air (kg m –3 

m ), Cp the 

specific heat capacity of air (J kg –1 K –1 ) and D the vapor pressure deficit of air (kPa). 

The measured daily total sap flow values obtained in kg tree –1 d –1 were divided by 

–2 

tree leaf area and day length to convert them to F in kg m s –1 . Therefore, gc 

obtained from equation 9 were daily mean canopy conductance values. 

2.6. Measurement of canopy area 

Approximate leaf area of all three trees selected for sap flow measurements was 

computed by measuring area of a sample of 50 leaves representing a range of sizes 

and counting the number of branches, leaf cohorts, and leaf number per cohort as 

follows: 

Approximate canopy area per tree = No. of branches per tree x mean no. of leaf 

cohorts per branch x mean no. of leaves per cohort x mean area per leaf (10) 

Projected leaf area index (LAI) I was computed as the ratio between approximate 

canopy area per tree and projected horizontal ground area covered by the tree 

canopy, assuming a circular canopy having a diameter equal to the measured canopy 

spread (Table 1). 



Data on transpiration and environmental variables recorded at 5-minute intervals 

were plotted against time of the day to obtain their diurnal variation patterns. 

Regression analysis was used to obtain relationships between transpiration and 

environmental variables. 

3. RESULTS 

3.1. Short-term variations in tree transpiration rates and its determinants in the 

absence of soil water deficits 

The data presented in Fig. 1a show the diurnal variation of sap flow rate over a 72-h 

period (from 23 to 25 June 2001) when the soil was fully saturated. All three species 

showed a similar pattern of sap flow, with the maximum rate being achieved from 

1300 to 1500 h. Minimum sap flow rates were observed from midnight to 0400 h.

TRANSPIRATION 


259 

The very small rates of sap flow that were observed during the night were probably 

to replace the water that was lost during the daytime from water storage tissues (i.e., 

capacitors) surrounding the xylem vessels. Artocarpus, which occupied the upper 

canopy layer of the MHG, showed the highest sap flow rate. There was, however, no 

significant difference in sap flow rate between Cedrela (middle canopy) and 

Swietenia (sapling in the lower canopy). Furthermore, the highest sap velocity 

Sap Flow (g/300 s) 

Sap Velocity (mm/min) 

140 

120 

100 

8 

7 

6 

5 

4 

3 

2 

1 

80 

60 

40 

20 

(a) 

0 

0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 

(b) 

Time (hours) 

0 

0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 

Time (hours) 

Figure 1. Diurnal variation of sap flow (a) and sap velocity (b) of three tree species 

representing different vertical layers in a multi-layered homegarden in central Sri Lanka 

during a 72 h measurement period of adequate soil water availability. Art. – Artocarpus 

heterophyllus; Ced. – Cedrela toona; Mah. – Swietenia macrophylla. 

Art. 

Ced. 

Mah. 

Art. 

Ced. 

Mah.

260 

was observed in Swietenia (Fig. 1b) while the lowest was noted in Artocarpus, with 

Cedrela having intermediate values. The respective sapwood cross-sectional areas 

for Artocarpus, Cedrela and Swietenia were 605, 51 and 9.6 cm 2 . Hence, an inverse 

relationship between sap velocity and sapwood cross-sectional area can be deduced. 

This is because sap has to be translocated at a higher velocity to provide enough 

water to compensate for the transpirational losses in individuals having a lower 

sapwood cross-sectional area. 

The upper canopy Artocarpus trees also received the highest irradiance (Fig. 2) 

during most of the day. However, irradiance on the upper canopy showed substantial 

short-term fluctuations because of partially cloudy weather. Middle canopy Cedrela 

Irradiance (W m -2 ) 

800 

700 

600 

500 

400 

300 

200 

100 

0 


0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 

Time (hours) 

Art. 

Ced. 

Mah. 

Figure 2. Diurnal variation of incident radiation on canopies of three selected tree species in 

a multi-layered homegarden in central Sri Lanka during a selected 72 h measurement period. 

Artocarpus heterophyllus (Art.), Cedrela toona a (Ced.), and Swietenia macrophylla (Mah.) 

represented the upper, middle, and lower layers of the homegarden. 

and lower canopy Swietenia received radiation, which was approximately similar to 

each other and less fluctuating compared to the upper canopy. Even the lower 

canopy Swietenia was exposed to direct sunlight during part of the day (from the 

direction of the house), which explains the similarity in irradiance level of the 

middle and lower canopies. When the data for the different canopy layers were 

pooled, there was, however, a clear linear relationship between transpiration and 

incident radiation on a daily basis (Fig. 3). Daily total transpiration of middle and 

lower canopy trees were 15 to 27% of that of upper canopy trees (Table 2). 

Vapor pressure deficit above the MHG (i.e., near the upper canopy) was greater 

than that within it (data not shown), especially during daytime. However, the diurnal

TRANSPIRATION 


261 

patterns of both were similar with maximum values around 1200 to 1400 h. 

Transpiration rates of trees representing all canopy layers increased with increasing 

VPD (Fig. 4). However, above a threshold VPD of about 0.8 kPa the increase of 

transpiration slowed down, because of decreasing stomatal conductance at higher 

VPD (data shown elsewhere). 

Sap Flow (kg d −1 tree −1 ) 

14 

12 

10 

8 

6 

4 

2 

0 

y = 1.2105 x + 0.8196 

R 2 = 0.9941 

0 2 4 6 8 10 

Irradiance (MJ m −2 d −1 ) 

Figure 3. Relationship between daily total sap flow of the three selected tree species and the 

daily total solar radiation incident on their respective foliage canopies during a 72 h 

measurement period. 

Table 2. Daily totals of transpiration and incident radiation during a three-day period of 

adequate soil water availability (june 2001) in a multi-layered homegarden in central Sri 

Lanka. 

Transpiration (kg d –1 tree –1 ) Incident radiation (MJ m –2 d 

2 –1 Species 

) 

23 Jun ‘01 24 Jun ‘01 25 Jun ‘01 23 Jun ‘01 24 Jun ‘01 25 Jun ‘01 

Artocarpus 

heterophyllus 

9.46 11.80 10.34 6.73 9.44 7.77 

Cedrela toona 2.06 2.64 2.00 1.04 1.37 1.16 

(0.22) (0.22) (0.19) (0.16) (0.15) (0.15) 

Swietenia 

2.04 3.10 2.77 1.38 1.71 1.49 

macrophylla (0.22) (0.26) (0.27) (0.20) (0.18) (0.19) 

Note: Fraction of the respective upper canopy value is given in parenthesis. 

3.2. Medium-term variation of transpiration during a period of increasing soil water 

deficits 

Daily transpiration rates of the upper canopy Artocarpus and middle canopy Cedrela 

were monitored during a two-month period of very little rainfall (25 Dec. 2001 to 20

262 

Feb. 2002). Despite the gradual decrease of soil water content in the top 1 m of the 

soil, both trees continued to have high rates of transpiration (Fig. 5). This probably 

indicated that root systems of both tree species absorbed water from deeper layers of 

the soil profile during periods when the top soil was dry. Total transpiration of 

Artocarpus during the 54-day experimental period was 3881 kg tree –1 , with the daily 

transpiration ranging from 17.84 to 95.87 kg day –1 tree –1 . The corresponding values 

for Cedrela were 463 kg tree –1 and 0.64 to 21.60 kg day –1 tree –1 . 

Sap Flow Rate (g / 300 s) 

140 

120 

100 

80 

60 

40 

20 

0 

(a) 

0 0.3 0.6 0.9 1.2 1.5 

Vapour Pressure Deficit above MHG (kPa) 

Sap Flow Rate (g / 300 s) 

30 

25 

20 

15 

10 

5 

0 

(b) 

0 0.2 0.4 0.6 0.8 1 

Vapour Pressure Deficit within MHG (kPa) 

Sap Flow Rate (g / 300s) 

35 

30 

25 

20 

15 

10 

5 

0 

(c) 

0 0.3 0.6 0.9 1.2 

Vapour Pressure Deficit within MHG 

Figure 4. Variation of sap flow rate with vapour pressure deficit in Artocarpous 

heterophyllus (a), Cedrela toona a (b) and Swietenia macrophylla (c) in a central Sri Lankan 

homegarden during a selected 72 h measurement period. 

Daily Sap Flow (kg/tree/d) 

120 

100 

80 

60 

40 

20 

0 

25-Dec-01 

01-Jan-02 


08-Jan-02 

15-Jan-02 

22-Jan-02 

29-Jan-02 

Calendar Date 

05-Feb-02 

y = -0.5641x + 21071 

R 2 = 0.9701 

12-Feb-02 

19-Feb-02 

Artocarpus Cedrella Soil Water Linear (Soil Water) 

Figure 5. Medium-term variation of daily sap flow of two selected upper- (Artocarpus 

heterophyllus) and middle canopy (Cedrela toona) trees in multi-layered homegarden in 

central Sri Lanka during a selected 54-day experimental period. Variation of soil water 

content in the top 1 m of the soil profile is also shown along with a fitted linear regression. 

70 

60 

50 

40 

30 

20 

10 

0 

Soil Water Content (cm/m)

TRANSPIRATION 


263 

The clear increase of daily transpiration in Cedrela was due to the significant 

increase of its projected LAI from 1.4 to 3.0 during the experimental period. In 

contrast, the projected LAI of Artocarpus was around 5.7 throughout. Medium-term 

fluctuations in sap flow of Artocarpus were closely related to fluctuations of daily 

total irradiance (Fig. 6a) and daily mean VPD (Fig. 6b). Transpiration of Artocarpus 

increased with increasing irradiance up to about 13 MJ m -2 d -1 and with increasing 

VPD up to about 0.8 kPa. The subsequent leveling-off and reduction was probably 

due to partial stomatal closure induced by higher VPD. This was confirmed by 

observed reduction of canopy conductance (gc) with increasing vapor pressure 

deficits both on a daily basis (Fig. 7) and on a diurnal basis (data not shown). 

Sap Flow (kg tree −1 d −1 ) 

100 

80 

60 

40 

20 

(a) 

y = -0.513x 2 + 17.433x - 67.407 

R 2 = 0.6746 

0 

0 5 10 15 20 

Incident Radiation (MJ m −2 d −1 ) 

Figure 6. Relationships between daily sap flow of upper canopy Artocarpus heterophyllus 

with daily total irradiance (a) and daily mean vapour pressure deficit (b) in a multi-layered 

homegarden in central Sri Lanka during a selected 54-day experimental period. 

Canopy Conductance (cm s −1 ) 

5 

4 

3 

2 

1 

0 

Figure 7. Relationship between canopy conductance during the day time in upper canopy 

Artocarpus heterophyllus and vapour pressure deficit in a multi-layered homegarden in 

central Sri Lanka during a selected 54-day experimental period. The canopy conductance was 

estimated by inverting a simplified version of the Penman-Monteith equation. 

Sap Flow (kg/tree/d) 

120 

100 

80 

60 

40 

20 

(b) 

y = -47.386x 2 + 124.71x + 2.063 

R 2 = 0.511 

0 

0 0.5 1 1.52 2.5 

y = -1.0738x + 2.9371 

R 2 = 0.438 

Daily Mean VPD (kPa) 

0 0.5 1 1.5 2 2.5 

Vapour Pressure Deficit (kPa)

264 


4. DISCUSSION 

4.1. Water use of multi-layered homegardens and its determinants 

Results of the present study clearly show that the taller trees occupying the upper 

canopy, which experience higher levels of incident radiation and vapour pressure 

deficits, dominate the total water use of MHGs. However, the contribution of trees 

occupying the middle and lower strata cannot be ignored. For instance, at the 

individual tree level, the combined transpiration of these two strata accounted for 30 

to 33% of the total water use of the three strata (based on data in Table 2). 

A linear increase in sap flow with incident solar radiation (Fig. 3) and vapour 

pressure deficit (Fig. 4) implies that these are the main drivers of transpiration in the 

MHG trees, irrespective of the level of water availability in the top 1 m of the soil. 

Furthermore, the linear relationship between transpiration and irradiance on a daily 

basis can be used to predict the daily transpiration rates of trees during periods of 

adequate soil water availability. The dependence of tree transpiration on VPD and 

irradiance is consistent with several other studies on a range of tree species (Granier 

and Loustau, 1994; Granier et al., 1996). 

Decreasing sensitivity of transpiration rate to increasing VPD and irradiance, as 

shown by decreasing slopes of the relevant relationships above 0.8 kPa (Fig. 4a and 

Fig. 6b) and above 13 MJ m –2 d –1 (Fig. 6a), indicates some degree of stomatal 

control of transpiration. These observations suggest that canopy conductance 

decreases with increased VPD above 0.8 kPa, which was confirmed by the observed 

reduction in canopy conductance with increasing VPD (Fig. 7). This is in agreement 

with the findings of several other studies (Roberts et al., 1990; Granier et al., 1996; 

Hogg and Hurdle, 1997; Meinzer et al., 1997) on several tree species in different 

forest types. However, the high levels of daily transpiration observed during the 

prolonged rainless period (Fig. 5) show that the level of stomatal control observed 

was not strong enough to reduce transpiration substantially. It is probable that both 

Artocarpus and Cedrela had root systems that were deep enough to extract water 

from soil depths below 1 m. 

4.2. Implications on sustainability of multi-layered homegardens in central 

Sri Lanka 

The MHGs in the Central Province of Sri Lanka are generally found on deep soils 

with high potential for water storage. Presence of deep-rooted trees capable of 

absorbing water from the lower soil layers is, however, a matter of concern under 

certain circumstances—especially during the rainless periods. Although the Sri 

Lankan MHGs generally predominate the humid tropical climatic zone having welldistributed 

rainfall (~2000 mm yr –1 ), the predicted drop in total rainfall and its 

increasingly non-uniform distribution in a future climate change scenario (McCarthy 

et al., 2001), is becoming a matter of concern. Perhaps this may be an overcautious 

scenario considering that the MHGs have sustained themselves for several centuries 

in areas with shallow soils and limited ground water resources. For example, the

TRANSPIRATION 


265 

Maya homegardens of the Yucatan Peninsula of Mexico (Benjamin et al., 2001) 

have provided sustainable livelihoods under rather harsh environmental conditions 

with limited water resources and soil nutrients. This could be yet another aspect of 

the ‘mysteries’ or the ‘enigma’ of tropical homegardens that defy the conventional 

scientific wisdom developed based on single-species systems (Nair, 2001; Kumar 

and Nair, 2004). 

5. CONCLUSIONS AND DIRECTIONS FOR FURTHER RESEARCH 

It is acknowledged that this study is based on a set of data, which is limited in 

several aspects. First, the transpiration measurements are not replicated and are 

based on only three trees out of 56 present in the 0.15 ha extent of the MHG. This 

was because of the practical difficulties involved in installing TDPs or sap flow 

gauges on an adequate number of trees and saplings and recording their output 

signals. Moreover, the highly uneven nature of tree distribution made replications 

difficult. For example, the two trees of Artocarpus heterophyllus were situated in the 

opposite parts of the MHG. We acknowledge that adequately replicated measurements 

of several tree species have to be done before making firm conclusions on the 

dynamics of transpiration in a highly complex vegetation system such as the MHGs. 

Subject to the above limitations, the study suggests that water use of multilayered 

homegardens of Central Sri Lanka is dominated by the upper canopy trees, 

with appreciable contributions from middle and lower canopy trees. Transpiration 

rates of MHGs are driven by incident radiation and vapour pressure deficit during 

periods of both adequate soil water availability and significant soil water deficits. 

Upper and middle canopy trees of MHGs maintain high rates of transpiration even 

during prolonged rainless periods by absorbing water from deeper soil layers. These 

findings should prompt concern on the impacts of high transpiration rates of MHGs 

on catchment water yield in a predicted future climate of reduced rainfall. We have 

not investigated the extent and depth of ground water availability. Despite the 

enormous practical difficulties involved, further in-depth studies are needed to 

quantify this impact at the catchment scale and to understand the ability of MHGs to 

sustain the livelihoods and ecosystem stability. 


We thank the National Research Council of Sri Lanka for financial support 

(Research grant no. NRC-99-23) and Ms. M.R.H.L. Karunasinghe and Mr. P. 

Surenthran for help in the fieldwork. A special word of thanks should go to Mr. 

Tennakoon, the land owner of the homegarden in which this research was done for 

all support and cooperation.

266 


ENDNOTES 

1. McConnell D.G. and Dharmapala K.A.B. 1973. The economic structure of 

Kandyan Forest Farms. The Management Report No. 7, UNDP/FAO, 

Agricultural Diversification Project, Peradeniya, Sri Lanka. 

2. Finlayson W. 1998. Effects of deforestation and of tree planting on the 

hydrology of the Upper Mahaweli catchment: A review of the published 

evidence. Environment and Forest Conservation Division, Mahaweli Authority 

of Sri Lanka, Polgolla. 

3. Gunawardena E.R.N. 1998. Overview of the hydrology project. In: Gunasena 

H.P.M. (ed.), Proceedings of the final workshop: University of Peradeniya - 

Oxford Forestry Institute Link Project, July 1998, pp. 20 – 29. UP-OFI Link 

Project, Peradeniya. 

4. Hitinayake H.M.G.S.B., De Costa W.A.J.M. and Jayaweera K.G.D. 1996. Food 

trees in multi-layered homegardens in different agro-ecological regions of 

Kandy district. In: Gunasena H.P.M. (ed.), Multipurpose trees for food security. 

Proceedings of the seventh regional workshop on multipurpose tree species, 

Kandy, Sri Lanka (24 – 26 October 1996), pp. 252 – 264. UP-OFI Link Project, 

Peradeniya. 

REFERENCES 

Baker J.M. and Van Bavel C.H.M. 1987. Measurement of mass flow of water in stems of 

herbaceous plants. Plant Cell Environ 10: 777 – 782. 

Benjamin T.J., Montanez P.I., Jimenez J.J.M. and Gillespie A.R. 2001. Carbon, water and 

nutrient flux in Maya homegardens in the Yucatan peninsula of Mexico. Agroforest Syst 

53: 103 – 111. 

Bosch J.M. and Hewlett J.D. 1982. A review of catchment experiments to determine the 

effects of vegetation changes on water yield and evapotranspiration. J Hydrol 55: 3 – 23. 

Calder I.R .1996. Water use by forests at the plot and catchment scale. Commonwealth For 

Rev 75: 19 – 30. 

Cowan I.R. 1965. Transport of water in the soil-plant-atmosphere system. J Appl Ecol 2: 

221 – 239. 

Granier A. 1985. Une nouvelle methode pour la mesure du flux de seve brute dans le tronc 

des arbres. Ann For Sci 42: 81 – 88. 

Granier A. 1987. Evaluation of transpiration in a Douglas-fir stand by means of sap flow 

measurements. Tree Physiol 3: 309 – 320. 

Granier A. and Loustau D. 1994. Measuring and modeling the transpiration of a maritime 

pine canopy from sap-flow data. Agric For Meteorol 71: 61 – 81. 

Granier A., Huc R. and Barigah S.T. 1996. Transpiration of natural rain forest and its 

dependence on climatic factors. Agric For Meteorol 78: 19 – 29. 

Hogg E.H. and Hurdle P.A. 1997. Sap flow in trembling aspen: implications for stomatal 

responses to vapour pressure deficit. Tree Physiol 17: 501 – 509. 

Jacob V.J. and Alles W.S. 1987. Kandyan gardens of Sri Lanka. Agroforest Syst 5: 123 – 137. 

Jarvis P.G. and McNaughton K.G. 1986. Stomatal control of transpiration: scaling up from 

leaf to region. Adv Ecol Res 15: 1 – 49. 


61/62: 135 – 152.

TRANSPIRATION 


267 

McCarthy J.J., Canziani O.F., Leary N.A., Dokken D.J. and White K.S. 2001. Climate change 

2001: Impacts, adaptation and vulnerability. Contribution of Working Group II to the 

Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). 

Cambridge University Press, Cambridge, 1032p. 

Meinzer F.C., Andrade J.L., Goldstein G., Holbrook N.M., Cavelier J. and Jackson P. 1997. 

Control of transpiration from the upper canopy of a tropical forest: the role of stomatal, 

boundary layer and hydraulic architecture components. Plant Cell Environ 20: 

1242 – 1252. 

Monteith J.L. 1986. How do crops manipulate water supply and demand? Phil Trans Royal 

Soc London, Ser A 316: 245 – 259. 

Murray F.W. 1967. On the computation of saturation vapour pressure. J Appl Meteorol 6: 203 

– 204. 



Panabokke C.R. 1996. Soils and Agro-ecological Environments of Sri Lanka. Natural 

Resources, Energy and Science Authority of Sri Lanka, Colombo, 220p. 

Roberts J., Cabral O.M.R. and Aguiar L. de F. 1990. Stomatal and boundary layer 

conductances in Amazonian Terra firme rain forest. J Appl Ecol 27: 336 – 353. 

Sakuratani T. 1981. A heat balance method for measuring water flux in the stem of intact 

plants. J Agric Meteorol 37: 9 – 17. 

Senarath A. and Dassanayake A.R. 1999. Soils of the mid-country wet zone. In: Mapa R.B., 

Somasiri S., and Nagarajah S. (eds), Soils of the wet zone of Sri Lanka: Morphology, 

characterization and classification, pp. 92 – 121. Soil Science Society of Sri Lanka, 

Peradeniya. 

Swanson R.H. 1994. Significant historical developments in thermal methods for measuring 

sap flow in trees. Agric For Meteorol 72: 113 – 132. 

Tetens O. 1930. Uber einige meteorologische Begriffe. Zeitschrift Geophysic 6: 297 – 309. 

Van den Honert T.H. 1948. Water transport in plants as a catenary process. Discuss Faraday 

Soc 3: 146 – 153. 

Wallace J.S. 1996. The water balance of mixed tree-crop systems. In: Ong C.K. and Huxley 

P.A. (eds), Tree-crop interactions: A physiological approach, pp. 189 – 233. CAB 

International, Wallingford.

CHAPTER 15 

ECOLOGY VERSUS ECONOMICS 

IN TROPICAL MULTISTRATA 

AGROFORESTS 

E. TORQUEBIAU* AND E. PENOT 

CIRAD TERA, TA 60/15 – 34398, Montpellier CX5, France; *E-mail: 

 

Keywords: Environmental services, Externalities, Modeling, Risk buffering, Rubber. 

Abstract. Homegardens and other multistrata agroforests are often described as ecologically 

sound, economically viable, and socially equitable land use activities. As in a majority of 

sustainable management situations, there are no widely accepted norms for a “perfect” 

combination of these attributes; what is often envisaged is a compromise among them. We 

argue that the development of ecological features of homegardens can be fostered by an 

“innovative” economic analysis. Performance of homegardens cannot be fully assessed by 

using conventional economic criteria and approaches such as yield, cost-benefit analysis, and 

net present value. Alternatively, if micro- and meso-level economic analyses (farming 

systems and upper level systems) are applied, the internalization of externalities such as 

agrobiodiversity management, carbon sink value, improved nutrient cycling or integrated pest 

management may turn homegardens into highly profitable ventures. Economic analysis 

methods should integrate risk buffering, outputs of mixtures of plants with different cycles, 

and allow to take a into account farming strategies with long-term objectives as well as the 

patrimonial (asset inheritance) components. Additionally, the merits of homegardens in terms 

of subsistence food for families, flexibility in production, reduced external-input 

requirements, enhanced aesthetic-, landscape-, and societal values, should also be 

incorporated into such an analysis. 


In the realm of agroforestry, homegardens and other multistrata, multispecies 

associations occupy an odd place. They are the most elaborate manmade, tree-cropanimal 

associations, and as such the only agroforestry system which can claim a 

resemblance to natural forests; hence their alternative name “agroforests.” Although 

269 




270 E. TTORQUEBIAU AND E. PENOT 

these systems have been studied in several countries (Indonesia, Brazil, India, and 

Sri Lanka: see chapters in this volume), the fact remains that they are seldom 

advocated as a land use option in agricultural or forestry development paradigms. 

Before pursuing further, a clarification on the use of the term homegardens 

versus other multistrata agroforestry systems is relevant. Both are multistrata 

combinations of trees and crops (sometimes with a livestock component). 

Homegardens are located next to human dwellings, managed for the production of 

subsistence items, and sometimes includes a cash objective. They are practiced on 

small parcels of land and are usually intensively managed. Not all multistrata 

agroforestry systems may, however, qualify as homegardens. Examples are the 

village-forests (village-forest-gardens), which are multistrata agroforestry systems 

developed on larger areas (at least a few ha per family) and managed mainly for 

cash income generation through the production of resins, jungle rubber, wood, fruits, 

etc. These are often considered as ‘intermediates’ between natural forests and treecrop 

plantations (Wiersum, 2004). Neither can all agroforestry systems be called 

“agroforests.” For example, the term does not cover agroforestry systems such as 

scattered trees on croplands, windbreaks, or woody hedgerows. As the term 

suggests, agroforests resemble forests and mimic their ecology (Michon and de 

Foresta, 1999; Wiersum, 2004). This resemblance is important in the context of the 

present chapter, as ecological and economic analyses applied to multistrata systems 

partly draw on their forest equivalents. Agroforest is thus the term used to represent 

both homegardens and other multistrata agroforestry systems and will hereafter be 

used in this chapter (instead of homegardens). 

1.1. Attributes and spread of agroforests 

Because of their resemblance to forests both physiognomically and ecologically, 

agroforests have a “good reputation.” Most statements recognizing the quality of 

agroforests, including in recent papers, refer to their ecological attributes, in 

particular biodiversity conservation and the long-term benefits of soil fertility 

maintenance and water conservation (Gajaseni and Gajaseni, 1999; Kaya et al., 

2002; Penot, 2001), even under harsh environments (e.g., the Soqotra Island of 

Yemen; Ceccolini, 2002). In some studies, the socioeconomic variables are taken 

into account (e.g., Mendez et al., 2001; Penot, 2003; Wezel and Bender, 2003) for 

analyzing the system’s functions but most do not describe the socioeconomic 

attributes with the same rigor as that of the ecological variables. Some studies 

dealing with bio-economic modeling of agroforests are also restricted to the 

cropping system-level (e.g., Purnamasari et al., 2002). Issues such as labor needs 

and returns, investments and return-on-investments in the mid- and long-term, 

product benefits, and income generation might be described, but they are seldom 

presented as arguments for adoption, or even taken into account in the innovation 

process behind the adoption of agroforests. In other words, the overall advantages as 

well as positive externalities of agroforests are widely recognized but not properly 

valued. Direct benefits of agroforests at farm-level are generally underrated and 

more so at the community and landscape levels.

ECOLOGY VERSUS ECONOMICS IN MULTISTRATA AGROFORESTS 

271 

The only two economic variables which seem to provide convincing arguments 

are: (1) diversification linked with the spreading of risk, income and labor, and (2) 

income generation as a whole (e.g., Torquebiau, 1992; Mendez et t al., 2001; Penot, 

2003; Wezel and Bender, 2003; Wiersum, 2004). The large number of products of 

agroforests and their uses may explain the difficulty to go beyond mere description 

and quantify these in economic terms. Similarly, the links between diversification, 

risk buffering capacity and long-term economic and ecological sustainability have 

not been sufficiently taken into account so far. The role of risk and uncertainty has 

been studied in agroforestry adoption (Mercer, 2004) but not as an innovation 

process in itself. 

Yet, tree homegardens cover significant land areas. For example, they occupy 

20% of arable land of Java, Indonesia (Jensen, 1993). It has been shown that the 

economic functions of these “pekarangans 

“ ” (see Wiersum, 2006) contribute to 

social equilibrium (Mary and Dury, 1997). A study of their patrimonial value 

demonstrated that durians (Durio zibethinus, a popular fruit tree) in these Javanese 

homegardens have a significant economic importance, both as a source of income 

and as an insurance mechanism in informal financial systems (Dury et al., 1997). 

There are more than 5 million homegardens in Kerala State, India (Kumar and Nair, 

2004) – another homegarden ‘hotspot’. Three million hectares of jungle rubber 

(Hevea brasiliensis-based agroforest) provide more than 50% of the total rubber 

production of Indonesia and there are another 2 million ha of various agroforests in 

Indonesia (Penot, 2001). Multistrata agroforests are also known in Brazil, 

Cameroon, Ghana, Nigeria, Tanzania, Sri Lanka, and other countries (Kumar and 

Nair, 2004). Agroforestry homegardens can also be observed in many tropical 

countries, both on agricultural frontiers and in stabilized agricultural landscapes. 

Although a worldwide estimation of the contribution of these cropping systems to 

agricultural production has not been made, it is now accepted that their contribution 

is far from negligible, be it in terms of traded products, fuelwood, subsistence crops, 

nutritional value, medicinal plants, timber, etc. If farmers worldwide have developed 

such systems, it is certainly not because they mimic the forests or foster biodiversity 

conservation—there must be something else as discussed hereunder. 

1.2. Need for a specific economic analysis 

We argue that there is an economic rationale explaining the importance of 

agroforests worldwide, but that this rationale is relatively complex to identify and 

measure. First, there is a well-known complementarity between direct sales of 

agroforestry products (timber, fruits, legumes, resins, nuts, rattan, medicinal 

products, etc.) and self-consumption by the garden owner, which leads to significant 

savings in the households’ day-to-day expenses. Secondly, it has been shown that 

long-term patrimonial strategies are of utmost importance to farmers growing 

agroforests (Mary and Dury, 1997); yet, conventional economic analyses based on 

discounting rates hardly serve for such perennial, multi-component and multi-cycle 

systems, where future discounted values of tree products are difficult to anticipate 

and as such seldom taken into account by farmers in their planting choices 

(Torquebiau et al., 2002), unless the harvested products are easily marketable and

272 

they generate a net margin which covers replanting costs (e.g., clonal rubber). 

Finally, farmers also plant and tend agroforests for their social functions (land 

tenure, social status and living environment). So, while scientists have repeatedly 

said that agroforests are environmentally sound, that alone is probably not a major 

motivation for farmers. 

The objective of this chapter is to try and show that the reason behind the 

“enigma of tropical homegardens” (Kumar and Nair, 2004) lies in elements of 

positive externalities, which are not accounted for in standard economic analyses, 

yet matter to the farmers and perhaps to other stakeholders (e.g., timber for 

sawmills). If agroforestry scientists want to convince farmers and policy makers that 

agroforests are more than just relics of the past and are worthy to be considered as 

land use options, then appropriate economic analyses of agroforests need to be done 

covering the ecological services (e.g., watershed protection, nutrient-cycling, carbon 

sink, bio-habitat functions, and biodiversity maintenance) as well as social, cultural 

and aesthetic values. 

Following Coase (1960)’s analysis of social cost, we make a difference between 

“giving a value to a service” (potentially, but not automatically tradable) and 

“paying for a service” (which leads to the “who is going to pay” question). Taking 

into account (assigning a value to a service) or internalizing positive externalities 

(paying for a service) relate to resources or services that cannot be included in 

private accounting because they are public goods (e.g., landscape beauty, pollinating 

insects) or because they are preserved for future generations (e.g., biodiversity, soil 

resources). We argue here that such “global goods,” considered as services to the 

community, need be taken into account not only by international negotiations such 

as discussions on climate change or biodiversity, but also in agricultural policies and 

incentives, and, as a consequence in the farmers’ day-to-day decision-making 

processes. 

One of the services that are likely to be taken into account in the future is the 

carbon sink function of the Clean Development Mechanism (CDM), as scheduled in 

the Kyoto Protocol. Rubber being the only tree crop (beside timber trees) eligible for 

CDMs, rubber-based and timber-based agroforests will theoretically be eligible. In 

such a case, their carbon sink service can be valued and considered in the trade or 

exchange of pollution rights (O.J. Cacho, pers. comm., 2002). 

2. AGROFORESTS AS CROPPING SYSTEMS PROVIDING MISCELLANEOUS 

GOODS AND SERVICES 

2.1. Multiple roles 

E. TTORQUEBIAU AND E. PENOT 

Farmers worldwide, but especially those in the developing countries, do not focus 

only on agricultural production. They are concerned first and foremost about their 

family priorities and are seldom sensitive to global issues such as biodiversity 

conservation or carbon sequestration; they nevertheless contribute to a series of 

goods and services that are not always marketed or even recognized. This multifunctional 

role of agriculture is now recognized and promoted in some regions (e.g.,


273 

Europe) in contrast to merely “production-oriented” agriculture. This has also led to 

the reduction in direct subsidies to production but subsidizing the environmental 

functions of farms. 

Agroforests can fulfill this multi-functional role better than other farming 

systems because they have more positive externalities than other monocultures or 

simpler agroforestry systems. So they deserve a specific economic analysis taking 

into account both goods and environmental services as well as short- and long-term 

issues. Agroforests, homegardens in particular, combine perennial-based production 

with long-term strategies (e.g., resin, nuts, fruits, and timber production) and 

shorter-duration food crops (e.g., legumes, cassava – Manihot esculenta, and banana 

– Musa sp.) with a short-term perspective. Farming systems models can include 

components on externalities or services to analyze this multifunctional feature. It 

might, however, be easier to handle the benefits of some services such as 

biodiversity conservation at regional- or macro-level. While priority has so far been 

on plant biodiversity, some studies have shown the role of agroforests as wildlife 

buffer zones (Nyhus and Tilson, 2004). 

Another important role is the generation of a “forest rent” as defined by Ruf 

(1987), i.e., the reduction of costs and risks of perennial plantation establishment – 

thanks to the forest’s positive externalities such as soil quality, weed and pest 

control. This concept has been extended to agroforests by Penot (2001), who showed 

that agroforests did maintain (sometimes improve) the forest rent while conventional 

monoculture plantation crops such as cacao (Theobroma cacao), coffee (Coffea 

spp.), and oil palm (Elaeis guineensis) generally consumed (part of) it. 

Agroforests have some constraints too, however. Since crop mixtures are the 

rule, some crops are favored while others are not; and agroforests may provide small 

quantities of a given crop that are not always saleable, except locally. Crops may 

also change with time; e.g., rice, maize or cassava may be initially intercropped with 

young trees but will not yield optimally under an increasing intensity of shade, 

which necessitates their replacement with shade tolerant crops (e.g., beans, some 

banana varieties). Similarly, rattan vines intercropped in rubber agroforests will not 

be harvested during peak rubber production but rather at the end of rubber trees’ 

lifespan because rattan harvesting tends to damage tree canopies. 

High reliance on manual labor and limited markets for specific products are 

other significant features in this respect. Delayed production (from large-sized trees) 

delays return on investment. Most farmers use non-improved plants and the quality 

can be variable, a potential problem for export of fruits, although there can also be a 

niche market for “organically grown” local varieties. However, some agroforests 

(e.g., rubber agroforestry systems) also rely on fertilizers and improved planting 

materials (e.g., rubber clones and grafted fruit trees). 

Overall, agroforests are specific cropping systems, which display a range of 

specifications making them more difficult to analyze than the monocropping or even 

multiple cropping systems with annual crop associations. It can be argued that it is 

this lack of analysis that has hampered the efforts of agronomists and extension 

agents to promote agroforests and hindered research to reach beyond the descriptive 

studies and into the stage of analytical research.

274 

3. SUSTAINABILITY OF AGROFORESTS 

Sustainability of agroforests can be explained based on different factors and criteria. 

Ecological sustainability stems principally from biodiversity conservation, natural 

resource management (soil and water), the control of pollution (little or no use of 

agrochemicals) and phytoremediation. Against today’s global change challenge, 

agroforests represent an important carbon sequestration potential (Kumar, 2006). 

Economic sustainability is based on the consideration that agroforests are able to 

provide in the long-run a stable and diversified source of income and are viewed as 

patrimonial assets (i.e., contributing to the long-term wealth and inheritance of the 

family; Mary and Dury, 1997). A large proportion of the local, traditional farming 

knowledge is related to agroforests. The risk buffering capacity of agroforests 

contributes to both ecological and economic sustainability. Social sustainability 

might be achieved through land tenure security linked to tree growing and 

preservation of community values. Institutional sustainability might be seen through 

the fact that agroforests can be individually or commonly managed. Table 1 

summarizes some arguments that link agroforests with sustainability. 

Table 1. A summary of sustainability attributes of agroforests. 

Ecological Economic Social and institutional 

- reduced soil erosion 

- high soil organic matter 

content 

- buffered soil moisture and 

temperature 

- closed nutrient cycling 

- improved soil physicochemical 

properties 

- efficient use of light and 

water 

- high wild plant and animal 

biodiversity 

- use of endogenous resources 

- contribution to on-farm 

production of wood and fuel 

wood 

- high soil biotic activity 

- better scope for evolution 

and diversification of 

economic plants 

- differentiated vertical and 

horizontal management 

zones and related ecological 

niches 

- potential for organically 

grown products 

E. TORQUEBIAU T AND E. PENOT 

- significant use of 

endogenous resources 

- high safety factor against 

marketing and 

seasonality hazards 

- reduced cash needs 

- high and diversified biophysical 

outputs (plant 

and animal food, 

medicines, fibers, etc.) 

- socio-economic outputs 

diversified and 

distributed over time 

- balance between 

subsistence and cash 

income 

- building up of capital 

- boosting rural industries 

and employment 

- adjustment to varied 

contexts 

- yield stability 

- management flexibility 

(intensive vs. extensive) 

- economic resilience 

(value as “land reserve”) 

- reduced and flexible 

labor needs 

- contribution to 

nutritional security 

- contribution to 

community 

socialization 

- preservation of 

traditional knowledge 

- biodiversity linked to 

traditions and practices 

- key role of women 

- equitable distribution 

of products 

- land reserve function 

(for alternative 

landuses) 

- maintenance of access 

rights to common 

goods (e.g., fruits) 

- flexibility of ownership 

(private vs. communal) 

Source: Adapted from Torquebiau (1992), Penot (2003), and Kumar and Nair (2004).


275 

Kumar and Nair (2004) rightly point out that homegardens (i.e., not all 

agroforests) may be on the verge of extinction due to new trends in agrarian 

structure, high market-orientation, demographic pressure, land fragmentation, and 

cultural dynamics. In the face of such constraints, the ecological foundations of 

homegardens may not be sufficient to warrant their survival. However, Javanese 

homegardens keep their place and role with an average population density of more 

than 800 persons km –2 

m , and a strong market-orientation of agriculture (Wiersum, 

2006). Presence of some high value crops (e.g., durian) may probably explain this. 

Interestingly, Java is not the only place where a positive correlation is observed 

between number of trees per unit area and human population density. Other 

examples include Kenya (Tiffen et al., 1994), Kerala (India) and Sri Lanka. 

Often multistrata agroforests are also under the influence of changing economic 

factors. For instance, jungle rubber and damar (Shorea javanica) gardens of 

Indonesia are facing international price fluctuations (e.g., rubber price moving from 

2 US $ kg –1 in 1996, to 0.6 in 2001, and then back to 1.2 in 2004). Furthermore, 

diversification of local farming may be at the expense of traditional agroforests, e.g., 

massive investments in industrial crops such as oil palm. The recent push toward 

globalization impacts the traditional farming practices in a myriad of ways among 

which access to market and marketing procedures rank high. In Asia, for instance, 

most export products have long been linked with international prices (rubber, oil 

palm, coffee and cocoa). The commodity boards established in Africa in the 1970s 

to protect farmers from price volatility have failed to deliver the expected results and 

their relevance is now being questioned. Thus, globalization has a stronger impact 

on African farmers than their Asian counterparts, who used to adapt better to the 

international markets and price cycles. We suggest that agroforests play a role in this 

adaptability; yet new policies of decentralization and local governance, new rules for 

access to credit, projects or information may impact it. It is, however, speculative as 

to whether agroforests will be able to react to such changes more efficiently than 

conventional monocropping. 

4. CHALLENGING THE REAL ECONOMIC IMPACT OF AGROFORESTS 

The sustainability advantages of agroforests come from a trade-off between 

ecological and socioeconomic attributes. Conventional economic approaches may be 

inadequate for integrating these attributes in a comprehensive manner, because (1) 

farmers manage agroforests for a variety of objectives, (2) the ecological benefits 

are not internalized in existing analyses, and (3) some ecological attributes have no 

current market value. 

Furthermore, if neoclassical economics are used to assess the performance of 

agroforests, the criteria of yield, cost-benefit analysis and net present value may end 

up giving agroforests poor ratings compared to conventional monocropping 

activities, because the analysis will exclude a series of agroforests’ outputs, which 

are not traded in the market or insufficiently taken into account in farm economics; 

Indonesia’s jungle rubber is a case in point. While it has been a major opportunity 

for poor farmers at the agricultural frontier for years, it is now becoming obsolete 

compared to clonal rubber monoculture, in terms of yields and labor productivity

276 

(Penot, 2001). However, it is difficult to measure or assign economic values to 

intangible services and positive externalities. For instance, carbon sink values of tree 

crops and forests are currently available but no one can choose among various prices 

suggested by experts as long as carbon markets are not functional. Risk-buffering 

potential of agroforests, as in situations of climatic variations and commodity price 

volatility, also deserves to be measured. The key question behind this is: how to 

make a measurement of the agricultural sustainability of agroforests? Perhaps farmsystem 

models used in farming system research could be a useful tool for such 

comparative assessments. 

4.1. Farming system level approach 


A pragmatic approach could be first to analyze at the household-level the cost saved 

by using products provided by agroforests for items that would otherwise need to 

have been purchased (e.g., building and fencing materials, food, medicines and raw 

materials for handicraft). Next, the accounting for environmental benefits might be 

performed at the household-level by compiling data over at least a year. Farming 

system modeling (e.g., with a software like “Olympe”) 1 is useful to process data on 

production, value, cost of production and labor, in order to be able to compare 

returns to labor and gross margin per cropping systems at the farm-level. Olympe 

performs whole-farm analysis in terms of resources, land, labor and other 

opportunities. It is a simulation tool for farm management advice which includes a 

“hazard” module that takes into account uncertainties, externalities (both positive 

and negative), as well as scenario definition according to risks. It can also be run at 

the regional level and with farmers’ groups. An analysis can be made in terms of 

income source, return to labor or investment, and linkage between strategic choices 

and production factor allocation, in order to assess the relative importance and real 

impact of cropping systems within the farming system. The combination of farm 

modeling with economic quantification, a historical perspective and the “contextualization” 

of farmers’ decisions according to political, socioeconomic, non-market 

(ecological)- and market factors provide the explanatory factors of a given farming 

system. Typically, the software allows re-interpreting the role of agroforests, as 

cropping systems within a farming or regional system. 

Under this approach, farmers’ strategies on labor, capital and land use are 

analyzed holistically (i.e., at the level of all enterprises of a farm, and not only at the 

level of one or the other cropping system). This is crucial to detect the place of 

agroforests in the overall farming strategy, because agroforests seldom produce the 

main staple food (Michon and de Foresta, 1999), and are invariably one cropping 

system among others on a farm. This approach, developed for the rubber farmers of 

Indonesia 2 , allows analyzing the diversification of opportunities for farmers facing 

an economic crisis and a political change that, in turn, can trigger significant 

changes in the social framework.


4.2. A social-ecological perspective 

277 

While a farming system approach can pave the way for a better understanding of 

agroforests’ roles, there is also a need for a renewed approach to agroforest analysis 

which can deal with higher levels of complexity and translate their “socialecological 

3 ” performance into economic performance. An apparently non-rational 

behavior that has been observed in Indonesia is the maintenance of old rubber 

agroforests along with economically very profitable oil palm plantations. One 

hypothesis was that agroforests would gradually leave the way for oil palm 

plantations. Social value (land control), possibilities of agroforest improvement 

(clonal rubber), and diversification strategies eventually may lead to a new 

development of improved rubber agroforests, which remain within the financial 

possibilities of local farmers with no access to credit, or even insufficient capital 

building capability. Meanwhile, whatever the important gains in return to labor and 

net margin provided by oil palm, agroforests have never disappeared – a proof of the 

value of such systems in a social perspective. Agroforests as “reserve land factor” or 

“long-term land control factor” might not have a direct value but do have an indirect 

value as a capital reproduction factor or as a potential expanding factor. 

Patrimonial analysis based on the evolution of capital building and asset 

transmission could be used for agroforests considered as reserves of land which can 

be traded, and since large-sized trees constitute a strategy for the build-up of capital 

for further investment. Long-term multi-cycle analyses may provide a framework to 

understand the farmers’ behavior and strategies. Economic analyses of mixtures of 

plants with different cycles can also be done through farming system modeling. 

Smoothening of long-term and patrimonial strategies (Mary and Dury, 1997; 

Torquebiau et al., 2002) may help taking into account the time factor and historical 

perspective (e.g., capital accumulation and building capacity). A multi-criteria 

analysis at both farm and community level is far more powerful than simple 

conventional cost-benefit analysis at cropping system level. Again, linking crucial 

social aspects (and their consequences in terms of use of production factors) with the 

economic analysis may provide a reliable framework that can take into account all 

cultural and non-merchantable aspects. Unfortunately, since methods for valuation 

of non-tangible social and cultural benefits of agroforestry are practically nonexistent 

(Kumar and Nair, 2004), it is difficult to substantiate the above (Penot and 

Deheuvels, 2006); rather, it is a plea for future research on these issues. 

4.3. Subsistence versus cash income generation 

The merits of agroforests in terms of subsistence for families, flexibility in crop 

production or reduced external input requirements also need to be taken into 

account. The comparison between farms with and without agroforests may show the 

savings and impact on household’s income. However, not all agroforests are food 

crop-based. Some agroforests are totally cash-oriented, e.g., rubber (jungle rubber), 

resin (damar agroforest), spices (e.g., cinnamon: Cinnamomum zeylanicum), fruits 

(durian) and timber-based agroforests.

278 


The flexibility in crop and tree production in agroforests relates to the different 

phases with mature and immature periods of trees or crops. Therefore, it is essential 

to take into account the life cycle of plants to implement an economic analysis in the 

long run. Specific discounting rates may be necessary as cycles may extend up to 40 

or 50 years. Different scenarios are necessary, as this may introduce bias in valuing 

products according to the discounting rates chosen. For instance, in tree crop-based 

agroforests, rubber or resin is produced for more than 30 years when annual and 

biennial crops are generally produced only in the first 3 to 6 years. Timber can be 

harvested only at the end of the agroforest’s life-span. Therefore, if detailed data are 

available to obtain a reliable assessment of real income (including self- 

consumption), system comparison will be more valuable than absolute data (Penot, 

2001). 

4.4. Landscape amenity and social conviviality 

The role of agroforests in providing services such as landscape beauty and aesthetics 

or social interaction or social status improvement has also to be incorporated in the 

assessment. It seems clear that in many situations, agroforests, and in particular, the 

non-private agroforests managed by local communities, and as such considered as 

public goods with limited and shared access (for fruits, timber etc.), have a social 

importance. The “Tembawang” of the Dayak people in Kalimantan (Indonesia) is a 

typical example. Besides being a reserve of forest products through “extractivism,” 

when original forests will have disappeared, such agroforests generally include 

important social components such as graveyards or may play a role of protection 

through the maintenance of a “green belt” around the village. Even if there is no 

economic value to this service, its social value will be a compelling reason for the 

maintenance of such agroforests and generally prevent its destruction. 

5. THE MICRO-ECONOMIC APPROACH 

Obviously, many specific features of agroforests might not be purely valued as 

goods. Social values, long-term strategic value of land, and risk buffering are 

examples; yet they provide powerful incentives to advocate agroforest development. 

With farming system modeling and a prospective approach, it is possible to assess 

the effects on risks. A prospective analysis with scenarios can lead to identification 

of economic thresholds and boundaries 1,2 and enables the definition of an economic 

feasibility domain (or expected economic outputs), i.e., the range within which the 

system is economically viable. 

If agroforests’ benefits can be analyzed through market values of their products 

and services, then neo-classical environmental economics can be used and 

externalities can be included (or re-internalized) into the process of income 

generation. Growth or pollution costs and delay may be taken into account as 

negative externalities or constraints to further development. Environmental services 

(for example, carbon sequestration potential: Albrecht and Kandji, 2003; 

Montagnini and Nair, 2004; Kumar, 2006) can be valued according to a “system of


279 

values” recognized locally as relevant at a higher, community or provincial level. 

The real problem is, therefore, to see whether farmers can potentially or do really 

take benefit of externalities and positive advantages of agroforestry. The payment of 

environmental services as promoted by the RUPES project (South-Sumatra and 

Lampung provinces, Indonesia) 4 provides some evidence in this respect. Other 

examples include the potential of agroforestry to reach the millennium development 

goals (Garrity, 2004) and the application of the Kyoto mechanisms to rubber trees 

(Hamel and Eschbach, 2001). Research on rubber agroforestry in Indonesia 

(Lawrence, 1996) provides an important data-set on these issues. 

In the context of most developing countries, huge income gaps due to strong 

social stratification, information asymmetry, high transaction costs and institutional 

failures have strong implications on local economies. Microeconomics allows 

accounting for environmental assets, complexity, and uncertainty, and involves 

stakeholder participation. When dealing with agroforests, benefits that relate to 

public goods or goods that cannot be given a market price because they are for 

future generations (e.g., biodiversity, landscape amenity, carbon sink and cultural 

and aesthetic values) need to be assessed through a new perspective. A multifunctional 

approach, similar to that developed by the Common Agricultural Policy 

for European farmers (Dévé, 2004), can provide ideas to take these externalities into 

account. New mechanisms such as the CDMs could be explored, in particular for 

global issues such as biodiversity conservation. 

Agroforest attributes should also be considered in national accounting. Policy 

makers should acknowledge the fact that if resource depletion is taken into account 

through an environmental economics approach; agroforests will rank very high 

among land use options because they generate an “agroforest rent” which is much 

higher than the rent from conventional agriculture or other forms of resource 

exploitation (e.g., logging, mining the soil through excessive harvests). Farmers 

contributing to this resource rent could hence be given direct or, better, indirect 

incentives (e.g., tax exemption) to stimulate land use options, which contribute to 

such public goods for current and/or future generations. 

To reach a status where agroforests could be recommended among other land use 

options, they need a reference framework, which takes into account these alternative 

economic analyses. Unfortunately, such analyses are lacking at present. In the 

meantime, multistrata agroforestry systems will continue being rejected or 

marginalized by conventional literature as not fitting into the mainstream economics 

and hence in development objectives. Be it for commercially oriented agroforests or 

subsistence oriented homegardens, a long-term perspective must be part of farmers’ 

strategy. However, there is obviously a biased debate between short-term 

(economics) vs. long-term (ecology) issues. In both cases, farmers have developed 

long-term farming practices through a long haul innovation process that eventually 

takes into account economics through the risk buffering capacity of agroforests. In 

most cases, social organization is deeply linked with technical constraints in 

production, food reliance, income securing and, eventually, land control. There is a 

strong coherence between technical pathways and social systems (Penot, 2003). 

Customary laws take into account this important point and are generally able to 

adapt to changes. There is an economic strategy behind maintaining agroforestry

280 


practices that have proved to be able to secure production and maintain control on 

land. In other words, long-term economics are totally associated with ecology and 

sustainability. An appropriate economic analysis should actually take care of the 

long-term aspects. One main challenge for the immediate future, however, is to take 

further steps towards the internalization of externalities, providing a value to 

services through a multifunctional approach and giving value-added objectives to 

ecological criteria. 


If an economic perspective with emphasis at local and regional level is applied to 

integrate positive externalities such as agrobiodiversity management, improved 

nutrient cycling, integrated pest management, ecological sustainability and services, 

decision-makers may be convinced that homegardens and agroforests are highly 

profitable ventures. If an “agroforest rent” approach is adopted, policy makers and 

development officers will see a long-term profitable investment in agroforests. 

Hopefully, this will lead to agroforests being given better consideration than at 

present in research and development programs worldwide. Furthermore, if 

agroforests are still a success-story with many farmers, it is obviously not because of 

biodiversity conservation. Other values such as social values, security, diversity, 

land control and reserve (including land and tree tenure) are probably important. 

There is also a need for a mechanism for the societal or community payment of 

those external and social benefits. A micro-economic analysis at the farming system 

level including all sources of income, cost-benefit per activity and return to labor, 

can explain such long-term strategies, provided they take into account the dynamics 

(“time effect”) of perennial crops in homegardens and other agroforests. 

Economic analysis methods using farming system modeling which integrate the 

outputs of mixtures of plants with different cycles and allow for the smoothening of 

long-term and patrimonial strategies are required to explain with accuracy what the 

farmers do and why they do so. Agroforests, despite their positive externalities and 

advantages are not a “panacea” but seem to be an ideal compromise between 

sustainability and risk spreading. 


We acknowledge the useful comments by Stefano Farolfi, Anne Marie Izac, Patrice 

Levang, and three anonymous referees on an earlier version of the manuscript. 

ENDNOTES 

1. Penot E., Le Bars M., Deheuvels O., Le Grusse Ph. and Attonaty J.M. 2004. 

Farming systems modelling in tropical agriculture using the software 

“Olympe.” ECOMOD Workshop, June 2004, Paris. 

2. Penot E. and Hébraud C. 2003. Modélisation et analyse prospective des 

exploitations hévéicoles en Indonésie: Utilisation du logiciel Olympe pour la


281 

définition de scénarios d'évolution en fonction de choix techniques et des aléas. 

Modélisation des exploitation agricoles: les multiples usages du logiciel 

Olympe. CIRAD Workshop, September 2003, Montpellier. 

3. The term “social-ecological” implies an interactive system of equally important 

social and ecological parts, while the conventional “socio-ecological” has the 

simple connotation of an ecological system with some social aspects (Sayer and 

Campbell, 2004). 

4. Van Noordwijk M., Chandler F.J. and Tomich T.P. 2004. An introduction to the 

conceptual basis of RUPES: rewarding upland poor for the environmental 

services they provide. ICRAF Southeast Asia, Bogor, 46p. 

REFERENCES 


Agric Ecosyst Environ 99: 15 – 27. 

Ceccolini L. 2002. The homegardens of Soqotra island, Yemen: an example of agroforestry 


Coase R.H. 1960. The problem of social cost. J Law Econ 3: 1 – 44. 

Dévé F. 2004. Major findings and conclusions on the role of agriculture. RAO Project of 

FAO, Phase 1. FAO, Rome, 18p. 

Dury L., Vilcosqui L. and Mary F. 1997. Durian trees in Javanese homegardens: their 

importance in informal financial systems. Agroforest Syst 33: 215 – 230. 



Garrity D.P. 2004. Agroforestry and the achievement of the millennium development goals. 


Hamel O. and Eschbach J.M. 2001. Impact potentiel du MDP dans l’avenir des cultures 

pérennes: état des négociations internationales et analyse prospective à travers l'exemple 

de la filière de production du caoutchouc naturel. Oléagineux Corps Gras Lipides 8: 

599 – 610. 

Jensen M. 1993. Productivity and nutrient cycling of a Javanese homegarden. Agroforest Syst 

24: 187 – 201. 

Kaya M., Kammesheidt L. and Weidelt H.J. 2002. The forest garden system of Saparua 

island, Central Maluku, Indonesia, and its role in maintaining tree species diversity. 






135 – 152. 

Lawrence D.C. 1996. Trade-offs between rubber production and maintenance of diversity: the 

structure of rubber gardens in West Kalimantan, Indonesia. Agroforest Syst 34: 83 – 100. 

Mary F. and Dury S. 1997. Les fonctions économiques méconnues des jardins villageois à 

Java-Ouest. Fruits 49: 141 – 150. 


Nicaragua: mico-zonation, plant use and socioeconomic importance. Agroforest Syst 51: 

85 – 96. 

Mercer D.E. 2004. Adoption of agroforestry innovations in the tropics: a review. Agroforest 

Syst 61: 311 – 328.

282 


Michon G. and de Foresta H. 1999. Agro-forests: incorporating a forest vision in agroforestry. 

In: Buck L.E., Lassoie J.P., and Fernandes E.C.M. (eds.) Agroforestry in sustainable 

agricultural systems, pp 381 – 406. CRC Press, Boca Raton, FL. 

Montagnini F. and Nair P.K.R. 2004. Carbon sequestration: an underexploited environmental 

benefit of agroforestry systems. Agroforest Syst 61: 281 – 295. 

Nyhus P. and Tilson R. 2004 Agroforestry, elephants and tigers: balancing conservation 

theory and practice in human dominated landscapes of Southeast Asia. Agric Ecosyst 

Environ 104: 87 – 97. 

Penot E. 2001. Stratégies paysannes et évolution des savoirs: l’hévéaculture agro-forestière 

indonésienne. PhD Thesis, University of Montpellier, Faculty of Economics, France, 

360p. 

Penot E. 2003. Cohérence entre systèmes techniques et systèmes sociaux et territoires. 

Evolution des systèmes de production hévéicoles et gestion de la ressource foncière : le 

cas de la province de Ouest-Kalimantan, Indonésie. In: Dugué P. and Jouve P. (eds), 

Organisation spatiale et gestion des ressources et territoires ruraux, pp 60 – 68. UMR 

SAGERT (CIRAD – CNEARC - ENGREF), Montpellier. 

Penot E. and Deheuvels O. (eds) 2006. Du système de culture à la petite région: Modélisation 

du fonctionnement de l’exploitation agricole, simulation et aide à la décision avec le 

logiciel Olympe. L’Harmattan, Paris (in press). 

Purnamasari R., Cacho O. and Simmons P. 2002. Management strategies for Indonesian 

rubber production under yield and price uncertainty: a bio-economic analysis. Agroforest 

Syst 54: 121 – 135. 

Ruf F. 1987. Eléments pour une théorie sur l’agriculture des régions tropicales humides: de la 

forêt, rente différentielle au cacaoyer, capital travail. Agron Trop 42: 218 – 232. 

Sayer J. and Campbell B. 2004. The science of sustainable development: Local livelihoods 

and the global environment. Cambridge University Press, Cambridge, 268p. 

Tiffen M., Mortimore M. and Gichuki F. 1994. More people less erosion: Environmental 

recovery in Kenya. John Wiley, London, 234p. 


Environ 41: 189 – 207. 

Torquebiau E., Mary F. and Sibelet N. 2002. Les associations agroforestières et leurs 

multiples enjeux. Bois et Forêts des Tropiques 271: 23 – 36. 




continuum: Characteristics and future potential. Agroforest Syst 61: 123 – 134. 




CHAPTER 16 

FINANCIAL ANALYSIS OF HOMEGARDENS: 

A CASE STUDY FROM KERALA STATE, 

INDIA 

S. MOHAN*, J.R.R. ALAVALAPATI, AND P.K.R. NAIR 

School of Forest Resources and Conservation, University of Florida, Gainesville, 

FL 32611, USA; *Current Address: CREST-RESSACA, Texas A&M University, 

Kingsville, MSC 213, 700 University Blvd, Kingsville, TX 78363, USA; E-mail: 

 

Keywords: Adaptive management, Economic utility, Non-monetary benefits, Resilience, 

Sensitivity analysis. 

Abstract. Homegardens are touted as economically and biologically sustainable systems, but 

studies to quantify the economics of these gardens are limited. This study used inventories, 

survey information and market data to estimate the productivity of 75 homegardens in 

Thrissur district of Kerala state, India, and applied benefit-cost analysis to ascertain the 

current financial values of these systems. All homegardens were found to be economically 

profitable and also to be of better economic utility to the farmer than selling or leasing the 

land. Sensitivity analyses indicated that these systems were easily resilient to 10% shifts in 

the prices of hired labor and in the prices of the three most economically important crops: 

coconut (Cocos nucifera), arecanut (Areca ( catechu), 

and banana (Musa spp.). Profit value of 

the gardens tended to increase with holding size and with increasing years of cultivation. 

Labor hours (both household and hired) and gender of the decision-maker were not suitable 

predictors of profit. Intensity of profit generation was highest in the smaller gardens, thus 

perhaps indicating both adaptive management to land constraints, and the presence of other 

intangible benefits that might affect land management strategies. 


Homegardens are well developed agroforestry systems consisting of distinct 

assemblages of plants with or without livestock, intensively managed within the 

residential compound. Economic theories and methodologies relating to agroforestry 

283 




284 S.MOHAN ET AL . 

systems are well documented (Alavalapati et al., 2004); however, rigorous field 

studies that apply these concepts to homegardens are rare (Nair, 2001). One of the 

major constraints to implementing some of these concepts stems back to an 

observation made by Scherr (1992) regarding the lack of guidelines for data 

collection and analysis. Preliminary economic analyses in Central America and the 

Caribbean have indicated that many agroforestry systems are profitable even at real 

discount rates of 20% or higher (Current et t al., 1995), yet economic studies relating 

to homegardens are limited. The economic worth of homegardens is especially 

difficult to quantify due to three reasons: these systems have high, yet variable levels 

of biodiversity, making data collection time-intensive and error-prone; these systems 

provide some benefits that are designed to be of particular use to certain farmers 

only; and finally, these are established systems, some of which have existed for 

many hundreds of years, and the benefits realized in the past may not be accurately 

quantified because of the inadequate availability of data. Different methods of 

quantification of intangible benefits, which are outside the scope of this chapter, are 

now increasingly studied and might potentially, be used to address these non-market 

values. 

Homegardens, although primarily used for subsistence purposes of the 

household, are increasingly being used to generate cash income (Christanty, 1990; 

Torquebiau, 1992; Mendez et al., 2001). They are also used to generate non-market 

benefits such as aesthetics, ornamentation, improved food quality, and nutritional 

security to the farmers (Karyono, 1990; Jose and Shanmugaratnam, 1993; Drescher, 

1996). With the overall aim of using a combination of different methodologies to 

assess the current tangible financial status of existing homegardens and providing a 

set of guidelines for data collection and analysis, a study was carried out in Kerala, 

India (Mohan, 2004). This chapter forms a part of that investigation and deals with 

the cost-benefit analyses for one year, and sensitivity studies to ascertain the 

economic resilience of Kerala homegardens to market fluctuations. The net values of 

these gardens were also compared with other available economic alternatives. 

2. METHODOLOGY 

2.1. Study location, sampling and economic evaluation 

The study was conducted in Thrissur District of central Kerala (between 10 o and 

10 o 47’ N latitude, and 75 o 55’ and 76 o 54’ E longitude). Kerala is one of the 

southernmost states of India, with a coastline of approximately 600 km and a 

tropical monsoonal climate. Thrissur experiences an annual precipitation of 

approximately 2500 mm. Homegardens are the predominant form of agriculture in 

the district, along with rice (Oryza sativa) farming and commercial plantations of 

coconut (Cocos nucifera), arecanut or betel nut (Areca ( catechu) 

and bananas (Musa 

spp.). A wide variety of plants are grown in the homegardens including commercial 

crops such as coconut and arecanut, starchy foods such as cassava (Manihot 

esculenta), and a large number of vegetables and fruits.

FINANCIAL F ANALYSIS OF HOMEGARDENS 

285 

Seventy-five homegardens of Thrissur district were randomly selected, and 

systematically (based on location) inventoried during October 2002 – February 

2003. These homegardens were located in both rural and semi-urban areas. A 

comprehensive survey was administered and the productivity of all homegardens 

estimated. The values of the products were determined according to existing market 

prices (shadow prices for medicinal plants), and key decision-makers in the selected 

homegardens were interviewed. 

The study was based on the premise that an analysis encompassing the steps 

summarized below would provide an adequate understanding of the economic value 

of these gardens in steady state (i.e., no natural calamities or extenuating 

circumstances that distinguished the year of study from other years). 

• Accounting the costs and benefits for the farmer over a one-year period. 

• Assessing the economic resilience of homegardens to market shifts in labor or 

crop price patterns by conducting sensitivity analyses. 

• Comparing homegardens with other economic alternatives to evaluate the option 

that would provide optimal economic utility to the farmer. 

Cost and benefit sources were determined based on the farmers’ records, as well 

as inventory of the gardens. Plant productivity was based on both yield estimates 1 

and farmer records. Market values were determined based on existing prices. Costs 

and benefits were assessed at the actual existing prices that the participating farmers 

encountered in markets. Many of the costs had already been incurred, such as onetime 

costs for building wells and for the initial preparation of land, but they were 

added to the total cost involved in maintaining the garden if incurred during the 

lifetime of the farmer who owned and farmed the property during the time of the 

study. The benefits realized from these costs are usually continuous and stretch over 

several years. Therefore, the yearly worth of these benefits was also added to the 

annual profits generated from these gardens. 

2.2. Opportunity costs of land and household labor 

The land tenure and ownership system in Kerala makes land a very valuable 

commodity in an increasingly land-deprived social system. Furthermore, the land 

occupied by the homegarden almost always houses the residential building, and 

these homes are usually inherited by the next generation. Therefore, it is unlikely 

that a homegarden will be sold on its own, without the residential building. 

However, in order to avoid inflating the financial worth of these systems while 

adhering to the observed social and cultural norms of the land, the opportunity costs 

of land were assigned values equivalent to the rate at which farmers were able to 

lease out all or part of the land. This rent rate was calculated to be an average of 

Rupees 12 350 (~ $262) per ha per year (one US $ = ~ Rupees 47, October 2003). 

Opportunity cost of household labor (OCHL) was calculated as a function of 

time as OCHL= ƒ (t*labor rate), where t is time spent in the garden. If the daily rate 

for a hired male laborer in a particular area was Rupees 70, and the owner/farmer 

put in an average of four hours work in the garden per day, the household labor costs 

were calculated to be Rupees 35(30) = 1050 (~ $22) per month.

286 

2.3. Components of the annual financial cycle of Kerala homegardens 

Based on farmer surveys and farm inventories, Table 1 presents the inputs and 

outputs that are the main components of the annual finances of a typical Kerala 

homegarden in steady state. Inputs were determined as any monetary contribution 

to the annual economic cycle of the garden and were generally found to comprise of 

human labor, seeds, organic and chemical fertilizers, hired labor, one-time costs 

such as barn maintenance and equipment (if incurred during the year of study), and 

the associated transportation costs. Some of the associated maintenance costs 

included transportation of products to markets, de-husking of coconuts, and the 

harvesting of coconuts, areca nut, and other market products. Except for 

transportation, these tasks were usually performed by hired labor. The farmers 

sometimes employed a system called karar r (contract), in which the commercial 

produce is leased either in part or full to a buyer, who would undertake all associated 

tasks, such as harvesting, transporting and selling, after paying a fixed sum of 

money to the owner. Such local barter systems might exist in other geographic 

locations around the world, and any financial analysis should take into account these 

individual practices and the social and cultural factors that influence these decisions. 

The tangible benefits derived from the garden also included products for market 

sale, milk and other livestock products, and goods used for household consumption 

such as food, firewood and medicinal plants. 

Inputs Outputs 

Fertilizers 

Seeds and seedlings 

Animal feed 

One-time expenses 

Maintenance operations 

Land cost 

Household labor 

S.MOHAN ET AL . 

Table 1. Components of the annual finances of a typical homegarden in steady state in 

Thrissur district, Kerala, India. 

Household products 

Market products 

Animal products (milk, meat, dairy) 

Long-term benefits (timber) 

Medicinal plants 

Note: Intangible benefits (e.g., shade, aesthetics, and ornamentation) have not been quantified 

in this study. 

All economically important species were inventoried and the production over the 

period of one year was estimated based on farmer reports. The economic inventory 

included medicinal plants that might or might not have been used by the farmer 

during the course of one-year, but were present in the garden because the gardeners 

considered them essential. The values of these medicinal species were included in 

those instances where the farmer had occasion to utilize a medicinal plant, by using 

a shadow pricing mechanism of estimating the cost involved in obtaining a similar 

benefit elsewhere.

287 

Economic theory argues that the highest social utility is attained when producers 

adopt practices generating the highest rates of return to all available resources, 

including all costs and benefits (Scherr, 1992). Economic planners also prefer 

investment in those activities yielding the highest rates of return to total resources or 

total labor used. However, the adoption decision for farmers is more complicated, 

especially in the case of homegardens where they reside within the confines of the 

agricultural property. These decisions may be influenced by a desire to maximize 

utility of family labor, returns to land, or even nutritional security. Two alternatives 

to homegarden cultivation have been considered in this study in order to understand 

the extent to which the farmer-needs and desires affect the pure cash flow into the 

homegarden system: Option I, entails selling the entire property and the house 

(assuming that selling the property without the house might prove to be improbable 

in the case of Kerala state) and Option II, in which the homegarden land is leased to 

another farmer, while the owner resides in the same house. Both options would 

allow the decision-making farmer to seek employment (work as agricultural laborer) 

elsewhere, assuming there is a steady demand for labor; yet they would have to pay 

to attain all benefits from the homegarden. Option 1 would also require that the 

farmer seek out an alternate residence. 

2.4. Analysis 


The collected data were analyzed using the basic economic methods of benefits and 

costs comparison, i.e., net financial worth (NFW) = Br r – Cr, 

where, B = benefits, C = 

costs, and r = year of study. For this, the homegarden products were categorized as 

having one of the three levels of economic utility; primary utility: those that are 

essential to the household, e.g., cassava, coconuts, and banana; secondary utility: 

those that are not absolutely essential but without which the household might suffer 

from nutritional deficiencies or other losses, e.g., gourd vegetables, amaranth 

(Amaranthus 

( spp.), and medicinal plants; and tertiary utility: those that are grown 

primarily for personal pleasure, e.g., ornamental plants and flowers, e.g., roses (Rosa 

spp.). Some plants are grown for both decorative and medicinal purposes, e.g., the 

shoe flower plant (Hibiscus spp.). The value of primary utility plants was quantified, 

and the value of the secondary category including medicinals was estimated using 

shadow pricing; the tertiary category provides mainly intangible benefits. All tree 

and shrub species found in the homegardens are listed in Appendix I. 

The sensitivity analyses were conducted by adding a 10% increment to the price 

of hired labor, and reduction of 10% in market prices of coconut, arecanut and 

banana, which are the main market crops in Kerala. Data were analyzed using the 

statistical software, Statistica. Various statistical procedures utilized in the analysis 

included analysis of variance (ANOVA) to compare characteristics of different size 

categories of homegardens, t-tests for comparison of means assuming unequal 

variances, and multivariate regression analyses to determine the predictors of 

homegarden profitability.

288 

S. MOHAN ET AL . 

3. RESULTS 

The 75 gardens included in this study had a mean landholding size (excluding the 

residential area) of 0.34 ha (± 0.03; median = 0.26 ha). The smallest garden was 0.01 

ha in extent, and the largest, 1.0 ha. Although homegardens greater than 1 ha was 

initially included in the data collection as part of the random sampling scheme, they 

were subsequently excluded from the analysis because they were deemed to be very 

large farms that showed more characteristics of sole cropping than that of traditional 

homegardening. The gardens included in the study were also subdivided into four 

groups (small: 0.26 ha; medium: 0.26 to 0.52 ha; large: 0.52 to 0.78 ha and 

commercial: 0.78 to 1.0 ha). Following this, there were 24 small, 14 medium-sized, 

10 large, and 27 commercial gardens. 

3.1. Economic values of homegardens and annual economic profit 

The existing financial worth of all the surveyed gardens, estimated based on the 

quantitative values of costs and benefits experienced in the year of study, is 

presented in Table 2. All 75 homegardens generated a positive economic value for 

the year 2001 – ‘02. Intensity of cultivation as indicated by the generation of profit 

per unit area (mean profit/m 2 of homegarden) calculated for the four holding-size 

categories was highest for the small gardens. While the commercial gardens yielded 

an average profit of Rupees 40.61/m 2 , the small gardens yielded more than double 

the average profit at more than Rupees 84/m 2 . Implicit in this is that the intensity of 

production is much greater in the smaller gardens, despite net production being 

higher in the larger gardens. 

Table 2. Mean financial value of homegardens for 2002 – ‘03 based on the benefits and costs 

of 75 gardens surveyed in Thrissur district, Kerala, India. 

Size of homegarden Mean 

financial 

value 

(Rupees) 1 

Mean financial 

value including 

opportunity 

costs of land 

and labor 

(Rupees) 

Intensity of profit 

generation 2 

Mean 

profit/m 2 

(Rupees/ 

year) 

Standard 

error 

Small (≤0.26 ha, n = 24) 62261 46284 84.28 a 10.72 

Medium (≤0.52 ha, n = 14) 157524 132759 68.80 b 

9.61 

Large (≤0.78 ha, n = 10) 256639 225116 76.64 a 11.48 

Commercial (≤1.0 ha, n = 27) 275967 214899 40.6 c 4.15 

1 Financial worth measured in Rupees (1.00 $US ~ Rs. 47, October 2003). 

2 Intensity refers to the mean profit generated per m 2 of cultivated area in the homegarden. 

Superscripts (a, b, and c) following a value indicate significant changes in means at = 0.05 

in t-tests assuming unequal variances.

3.2. Economic importance of homegarden species 

289 

The most important contributors to the economic profit generated by homegardens 

were coconut, arecanut and banana (both cooking and dessert varieties), but the 

distribution of profit varied across garden sizes (Fig. 1). The other economically 

important categories in the homegarden were dairy, cashew (Anacardium 

( 

occidentale), spice trees such as nutmeg (Myristica fragrans), and vanilla (Vanilla 

planifolia) (data not presented). Household needs consumed a significant percentage 

of the products (more than 50%) in the smaller gardens, while the larger and 

commercial gardeners invested most in the commercial production of coconut and 

arecanut (Fig. 1). 

Figure 1. Contribution of three crop categories and household use to total profit generated 

by different size classes of homegardens in Thrissur district, Kerala, India. The holding sizes 

are: small: 0.26 ha; medium: 0.26 to 0.52 ha; large: 0.52 to 0.78 ha, and commercial: 0.78 

to 1.0 ha. 

3.3. Sensitivity analyses 


Sensitivity analyses are important when evaluating economic benefits, in order to 

ascertain the extent to which agricultural systems are susceptible to shifts in the 

prices of labor and market products. A majority of the households surveyed (96%) 

reported that the prices of hired labor to be the most restrictive aspect of managing 

these systems, and coconut, arecanut and banana are the most economically 

important crops. A comparison of the data in Table 3 indicates the changes in net 

value of the gardens when the labor prices are increased by 10%, and the market 

prices of coconut, arecanut, and banana are reduced by 10%. Some of the gardens 

that cultivated rubber trees (Hevea brasiliensis) as a component were also very 

dependent on it; but rubber was mainly found in the larger gardens, mostly as a sole 

crop. Hence, it was excluded from the sensitivity analysis. 

The results indicate very low changes in annual profit value across all classes of 

homegardens, ranging from 0.24% to 2.46%. The only statistically significant

290 


difference across means was the effect of raised arecanut prices in the commercial 

gardens, which ranged from 2.46% for commercial gardens to 0.81% for the small 

gardens. 

Table 3. Sensitivity analysis to ascertain the economic resilience of homegardens to price 

fluctuations in labor and price of three economically significant crops. 

Sensitive categories 

Percent response in financial worth 

(based upon a 10% change in price) 

Small Medium Large Commercial 

P (hired labor) 0.28 1.12 0.24 0.31 

P (coconut) 1.0 2.0 2.8 1.0 

P (arecanut) 0.81 1.65 2.21 2.46 * 

P (banana) 0.42 0.35 0.74 0.92 

P = existing market price; *significant at = 0.05 in comparisons involving small (n = 24), 

medium (n = 14), large (n = 10), and commercial (n = 27) using t-test assuming unequal 

variances. 

3.4. What factors affect the financial value of homegardens? 

The multivariate regression model developed to predict the effects of various factors 

on the financial values of the surveyed homegardens is as follows and its statistical 

parameters are given in Table 4. 

Financial worth of homegarden = 4.61 + 0.007(x1) + 0.003(x2) 

Where x1 = land area in m 2 and x2 = number of years in cultivation. 

Table 4. Coefficients, standard error and probability level of significance of the predictors of 

homegarden’s economic worth in Thrissur district in Kerala, India. 

Parameter B Standard 

error of B 

p values 

Intercept 4.61 0.073 0.000 

Land holding size (m 2 ) 0.007 0.056 < 0.005 

Age of garden (years) 0.003 0.001 0.017 

Adj. R 2 = 0.447; standard error = 0.319. 

The model indicates that the financial value of Kerala homegardens increases 

with increasing land holding size and with an increase in the number of years of 

cultivation, although both are only modest predictors of profit. The number of hours 

of household or hired labor and gender of the decision-maker in the household were, 

however, not significant predictors (p 1.00) of net profitability. Biophysical

aspects such as soil quality and availability of water might contribute to the financial 

value of these gardens, but such effects need to be investigated further. 

3.5. Economic alternatives to homegardens 


291 

Two possible alternatives were considered when comparing the economic rationale 

behind homegarden cultivation to other forms of investment. The first assigned 

alternative for a farmer was to sell the land, with the house and all associated crops 

and benefits, invest the capital in a bank at 6% compound interest rate (average 

prevailing rate at the time of study) and to live in a comparable neighborhood with a 

similar quality of life. The second option was to lease the land and all associated 

benefits to other farmers. Both alternatives and their profit values for all size classes 

of homegardens at the end of the investment year are considered in Table 5. The 

non-monetary benefits, however, were not quantified. 

Table 5. Comparison of homegarden finances to two alternate forms of economic 

investment 

India. 

1 for small, medium, large and commercial holdings in Thrissur district, Kerala, 

Variables 

Finances from gardens and two 

alternate land use options (Rupees) 

Garden 2 

Lease 3 Bank 4 

k 

a. Mean ‘small’ homegarden (n = 24) 

Land 0 1086 22012 

Labor 0 7250 7250 

Living expense 0 (20000) (20000) 

Rent 0 0 (15000) 

Transportation 0 (500) (500) 

Incidentals 0 (800) (800) 

Homegarden costs (7548) 0 0 

Benefits 65519 0 0 

Net income 57971 (12964) (7038) 

b. Mean ‘medium’ homegarden (n = 14) 

Land 0 2552 61329 

Labor 0 14914 14914 


Rent 0 0 (15000) 





Net income 162513 (5834) 37943 

Table 5 (contd.)

292 


Variables 

Finances from gardens and two 

alternate land use options (Rupees) 

Garden 2 

Lease 3 Bank 4 

k 

c. Mean ‘large’ homegarden (n = 10) 

Land 0 4240 101760 

Labor 0 11880 11880 


Rent (15000) 





Net income 224851 (7180) 75340 

d. Mean ‘commercial’ homegarden (n = 27) 

Land 0 8250 201370 

Labor 0 17862 17862 


Rent (15000) 





Net income 258222 812 178932 

1 2 

Estimated for a homegarden of mean size in each size category; maintained as 

homegardens; 3 lease option and the lease values were based on existing rent rate of Rupees 12 

350 per ha. 4 capital invested in a bank at 6% compound interest rate; parenthetical values are 

costs. 

Living costs were estimated based on a two-month survey of expenses incurred 

by four urban and rural households with no attached homegardens. All household 

expenses, not including meat, staple foods such as rice, potato, salt, and other goods 

not normally realized from the garden, were estimated to be an average minimum of 

Rupees 20 000 per year per household. A comparison of the data in Table 5 also 

indicates the financial effectiveness of maintaining homegardens as opposed to 

leasing or selling the land. Selling the garden becomes a reasonable yet not 

comparable alternative with increases in land area. Small farmers would be best 

served if they retained their homegardens. Leasing was not an economically viable 

option especially for the small, medium or large gardeners. 

4. DISCUSSION AND CONCLUSIONS 

All homegardens surveyed in this study generated profits at steady state, thus 

justifying the need to consider them by the policy makers as on par with other 

mainstream agricultural production systems. The positive financial value, regardless 

of the number of years in cultivation, implies the renewable nature of these gardens 

year after year. The profit generated per unit area was highest for the small gardens


293 

(Table 2) and was lowest in the commercial gardens, perhaps implying that the small 

farmers are particularly adept at adaptive management techniques. Holding size 

being a constraint, farmers intensify cultivation on available land in order to attain 

desired goals and objectives. Commercial farmers, however, may devote some part 

of their holding for intangible benefits such as aesthetics and ornamentation. Future 

studies could assess whether this difference in profit generation equals the 

opportunity cost incurred by those commercial farmers who do not intensify 

production. Coconut, arecanut and bananas were the three most economically 

important crops (Fig. 1). It was noted, however, that although market needs were 

extremely important in determining garden use, small gardeners used more than half 

their annual produce for household uses, e.g., vegetables, fruits and firewood. 

Allocation of garden space also was need-based; i.e., if the farmers possessed liquid 

cash at their disposal with which to buy subsistence products, they increased the 

acreage under commercial crops such as areca and spice trees. On an average, more 

than 75% of the household needs of an average family were met by their 

homegardens irrespective of the garden size. The sensitivity analyses (Table 3) 

reaffirmed the hypothesis that these systems are economically stable, not dependent 

on any one crop or factor, and that the farmers followed an age-old adaptive 

approach to farming. Harvests were staggered so as to retain food crops such as 

cassava, for times of the year when staple food crops such as rice were not readily 

available. No one crop formed a focal point in the garden. For example, the areca 

crop had been sustaining high returns during the 1990s, but suffered a crash in 

market prices during the past few years (2001 and 2002); many farmers would have 

sustained heavy losses had their gardens consisted of sole stands of areca palms 

alone. With the existing complexity and diversity of these gardens (Appendix I), 

however, the lagging arecanut prices did not substantially affect the overall profit 

from the gardens. After considering two potential alternatives to homegardening 

(Table 5), it was estimated that retaining the land under homegarden cultivation was 

more profitable than leasing or selling the land even without factoring in intangible 

benefits such as aesthetics, nutritional security, and improved quality of food. 

Plantation farming was not considered as an alternative because many of the gardens 

surveyed were deemed too small to be fit for plantation agricultural systems. 

The household labor associated with homegardening was an important 

component of the alternatives because it was assumed that if the land were no longer 

available to farmers, they would earn money by working as laborers in the nearby 

farms. This is another debatable point, however, because many of the farmers 

reported that they were not equipped to perform any skilled work, nor did they 

desire to perform farm labor outside their properties. Furthermore, many of the 

farmers were older, and cherished the relative freedom they enjoyed from working 

in their own fields, and in their ability to set their own timings. 

It needs to be acknowledged that the methodology used for the study had some 

constraints. Homegardens are so diverse in species richness and composition 

(Appendix I) that data collection becomes arduous and error-prone. Data analysis 

becomes further complicated because many homegarden species are retained to 

fulfill certain specific needs and functions, and these needs vary from farmer to 

farmer and from region to region. Intangible benefits of homegardens, such as

294 

S. MOHAN ET AL . 

aesthetics and ornamentation, nutritional security, food quality, and empowerment 

of women also need to be considered in order to obtain a more accurate assessment 

of the economic values as articulated also by Torquebiau and Penot (2006). 

Furthermore, some of the data presented here, especially the monetary values, are 

time-sensitive. Although these constraints set some limits to applicability of the 

findings to other regions, we believe that the methodology can be adapted in any 

geographic area to estimate the economic value of these multipurpose production 

systems. 

ENDNOTE 

1. The data were gathered during the first author’s field study, which involved 

interaction with farmers and discussion with various officials of the Kerala 

Agricultural University and local field extension personnel of the government 

agricultural and other departments. The authors thank the Kerala Agricultural 

University, Thrissur, India for extending support to this project. 

REFERENCES 

Alavalapati J.R.R., Shrestha R.K., Stainback G.A. and Matta J.R. 2004. Agroforestry 

development: An environmental economic perspective. Agroforest Syst 61/62: 299 – 310. 

Christanty L. 1990. Homegardens in tropical Asia with special reference to Indonesia. In: 

Landauer K. and Brazil M. (eds), Tropical home gardens, pp 9 – 20. United Nations 


Current D., Lutz E. and Scherr S. (eds). 1995. Costs, benefits, and farmer adoption of 

agroforestry. The World Bank environment paper 14: World Bank, Washington, DC. 

Drescher A.W. 1996. Management strategies in African homegardens and the need for new 

extension approaches. In: Heidhues F. and Fadani A. (eds), Food security and innovations 

– Successes and lessons learned. pp 231 – 245. Peter Lang, Frankfurt. 

Jose D. and Shanmugaratnam N. 1993. Traditional homegardens of Kerala: a sustainable 

human ecosystem. Agroforest Syst 24: 203 – 213. 





85 – 96. 

Mohan S. 2004. An Assessment of the ecological and socioeconomic benefits provided by the 

homegardens: A case study from Kerala, India. PhD Dissertation, University of Florida, 

Gainesville, FL. 



Scherr S.J. 1992. Financial and economic analyses of agroforestry systems: An overview of 

the case studies. In: Sullivan G.M., Huke S.M., and Fox J.M. (eds), Financial and 

economic analyses of agroforestry systems. Proceedings of a workshop held in Honolulu, 

Hawaii, USA. July 1991, pp 3 – 12. Nitrogen Fixing Tree Association, Paia, Hawaii. 

Torquebiau E. 1992. Are tropical agroforestry homegardens sustainable? Agric Ecosys 

Environ 41: 189 – 207. 

Torquebiau E. and Penot E. 2006. Ecology versus economics in tropical multistrata 

agroforests. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 

example of sustainable agroforestry, pp 269 – 282. Springer Science, Dordrecht.


APPENDIX I 

Woody perennials encountered in the sampled homegardens of Thrissur district, 

Kerala, India 1 . 

295 

Scientific name Local/common name Family Uses 2 

Tree and shrub species of primary economic utility to farmers 

Ailanthus triphysa matti Simaroubaceae b 

Anacardium occidentale cashew Anacardiaceae d,b,a 

Areca catechu arecanut Palmaceae a, f 

Artocarpus heterophyllus jackfruit Moraceae d, b 

Artocarpus hirsutus aini Moraceae b,c 

Bridelia retusa kaini Euphorbiaceae b 

Bombax ceiba poola Bombacaceae b,f 

Borassus flabellifer palmyra palm Palmaceae f,e 

Calophyllum inophyllum punna Clusiaceae b,c 

Caryota urens fish-tail palm Palmaceae f 

Cocos nucifera coconut Palmaceae a, c, f 

Coffea arabica coffee Rubiaceae d 

Corypha umbraculifera talipot palm Palmaceae f 

Dalbergia latifolia rosewood Fabaceae b 

Delonix regia poomaram Caesalpiniaceae b 

Garcinia cambogia kodampuli Clusiaceae d, c 

Grewia tiliifolia chadachi Tiliaceae b 

Hevea brasiliensis rubber Euphorbiaceae c, f 

Mangifera indica mango Anacardiaceae d, b 

Manihot esculenta cassava Euphorbiaceae e 

Michelia champaca kaatu chembakam Magnoliaceae b,c,g 

Morus alba mulberry Moraceae c,f 

Myristica fragrans nutmeg/mace Myristicaceae a 

Palaquium ellipticum pali Sapotaceae b,g 

Piper longum thippili Piperaceae g,b 

Pterocarpus marsupium venga Fabaceae b 

Santalum album sandalwood Santalaceae f 

Saraca indica asoka tree Caesalpiniaceae b,c 

Swietenia macrophylla mahogany Meliaceae b 

Syzygium aromaticum clove Myrtaceae a 

Tamarindus indica tamarind Caesalpiniaceae d 

Tectona grandis teak Verbenaceae b 

Terminalia tormentosa maruthy Combretaceae b 

Xylia xylocarpa irumullu Mimosoideae b,c 

Appendix 1 (contd.)

296 


Scientific name Local/common name Family Uses 2 

Trees and shrubs species of secondary economic utility to farmers, used 

mainly in the household 

Annona squamosa custard apple Annonaceae d,b,c 

Artocarpus altilis breadfruit Moraceae d 

Averrhoa bilimbi irimbampuli Oxalidaceae d,c 

Azadirachta indica neem Meliaceae g 

Cananga odorata ylang ylang Annonaceae f,g 

Carica papaya papaya Caricaceae d 

Casuarina equisetifolia kattaadi Casuarinaceae f,h 

Cinnamomum camphora camphor Lauraceae f,g 

Cinnamomum zeylanicum cinnamon Lauraceae e,c 

Citrus limon cherunarakam Rutaceae d,c 

Emblica officinalis Indian gooseberry Euphorbiaceae d,g 

Flacourtia inermis louvi-louvi Flacourtiaceae b,c,d 

Manilkara zapota sapota (sapodilla) Sapotaceae d,b,c 

Murraya koenigii curry leaf tree Rutaceae e,c 

Pimenta dioica allspice Myrtaceae e,g 

Pouteria campechiana eggfruit Sapotaceae d 

Psidium guajava guava Myrtaceae b,c,d 

Punica granatum pomegranate Punicaceae d 

Syzygium jambolana 

rose apple Myrtaceae d,b 

Terminalia catappa Indian almond Combretaceae e,c 

Theobroma cacao cacao Sterculiaceae f,a 

1 

In addition, 17 herbaceous species were identified under two categories each (having primary 

or secondary economic utility to the farmers); for details see Mohan (2004). 

2 

Uses: a = nuts, b = timber, c = fuelwood, d = fruits, e = leaves, bark and other parts of plant 

used as food, f = leaves bark and other parts of plant used for other purposes, g = ornamental 

or medicinal purpose, h = shade; Local names appearing in italics are vernacular names 

(Malayalam).

SECTION 4 

FUTURE OF HOMEGARDENS

CHAPTER 17 

THE ROLE OF HOMEGARDENS 

IN AGROFORESTRY DEVELOPMENT: 

LESSONS FROM TOMÉ-AÇU, A JAPANESE- 

BRAZILIAN SETTLEMENT IN THE 

AMAZON 

M. YAMADA* AND H.M.L. OSAQUI 


of Agriculture, Tokyo University of Agriculture and Technology, 3-5-8 Saiwaich, 

Fuch-shi, TT ky 183-8509, Japan; *E-mail: 

Keywords: Adaptive research, Farm cooperatives, Farmer innovation, Fruit trees, 

Institutional support. 

Abstract. Agroforestry systems developed by the Japanese immigrants and their descendants 

in the Eastern Amazon region have been the focus of attention as a model for sustainable rural 

development in the humid tropics. This paper looks at the role of homegardens in agroforestry 

development at the Tomé-Açu Nikkei settlement in Pará, Brazil during the past seven decades. 

Potential crop species – native as well as exotic – were gathered and nurtured by the farm 

families in these homegardens of size 1 to 3 ha. Although the Tomé-Açu Multipurpose 

Agricultural Cooperative (CAMTA) had experimental nurseries and the Japanese public 

agencies established local agricultural research stations for supporting emigrant farmers in the 

Amazon, the homegardens functioned as individual validation fields where the farmers 

‘experimented’ with new crops. Homegardens were also used for improvement and 

propagation of nursery stock making them on-farm laboratories for adaptive research and 

extension. The immigrants with the traditional tokun (master farmer) education of East Asia 

analyzed the local environment and ‘experimented’ with various plant associations and 

management techniques, which led to the evolution of the exceptionally successful and 

popular multistrata agroforestry systems in the Eastern Amazon region. 


Since the early 1980s, various authors of Amazonian studies have discussed 

agroforestry systems developed by the Japanese immigrants and their descendents 

299 




300 M. YYAMADA AND H.M.L. OSAQUI 

(the Nikkei farmers) as an economically viable and ecologically sustainable rural 

development option for the region (Jordan, 1986; Gradwohl and Greenberg, 1988; 

Uhl and Subler, 1988; Uhl et al., 1989; 1990; Anderson, 1990; Barrow, 1990; Subler 

and Uhl, 1990; Serrão and Homma, 1993; Fearnside, 1995; Serrão, 1995; Homma, 

1998). While there are more than thirty rural Japanese-Brazilian (Nikkei) 

communities in the Amazon, most authors have focused on the Tomé-Açu 

settlement, which was the center of commercial rice (Oryza sativa) and vegetable 

production in the Amazon, and the location where first commercial black pepper 

(Piper nigrum) production was started in the Americas. The Tomé-Açu settlement 

was founded in 1929 ca. 120 km south of Belém, the capital of Pará (2°31’ S and 

48°22’ W). At the end of 2002, there were 214 Nikkei farms in Tomé-Açu, covering 

77 500 ha, with 7200 ha of agroforestry fields 1 . By the end of 1996, the Nikkei 

farmers in Tomé-Açu planted 6500 ha of agroforestry fields, with three perennial 

vine species, four shade trees, 33 fruit trees, 68 multipurpose tall trees, and 

numerous vegetable-, herb-, grain-, tuber- and green manure plants, forming a 

spatial mosaic of different ages and species combinations (Yamada, 1999). The 

main farm fields (6100 ha) excluding the homegardens around the housing/barnyard 

areas were occupied by ~70 crop species (90% arboreal) and some leguminous 

shade trees (“eritrina” = Erythrina spp. and “palheteira” = Clitoria racemosa). Sixty 

percent of the area involved polycultures, with approximately 300 different crop 

combinations, while the remainder was temporary monocultures based on sequential 

intercropping or “successional” agroforestry (Subler, 1993; Tanaka, 1997). For 

example, rubber (Hevea brasiliensis) and cupuaçu (Theobroma grandiflorum) fields 

were planted with annual crops (grains and vegetables) and perennial vines, but 

seedlings of shade tolerant arboreal and herbal species were subsequently 

introduced. 

2. HOMEGARDENS IN TOMÉ-AÇU AS LABORATORIES FOR SPECIES 

INTRODUCTION, SCREENING AND BREEDING 


The first author lived in Tomé-Açu from January 1995 to January 1997 conducting 

field work on farm histories and crop inventories. He visited all 214 Nikkei farms 

and interviewed farm owners on land use (intact and explored primary forests, 

secondary forest, fallow, pasture, and area under agroforestry) and crop species 

(year planted, number, area, and nature of cropping, i.e., with or without intercrops). 

For the inventory of homegardens, the authors first acquired reference information 

from the Tomé-Açu Multipurpose Agricultural Cooperative (CAMTA), which listed 

about 30 candidates that typified the history and dynamics of plant introduction, 

screening and dissemination in the Nikkei settlement. Among them, a dozen gardens 

were randomly chosen. With assistance from the owners of these gardens, details on 

each species (year, route of introduction, from where it came, etc.) were gathered. 

Additional information was gathered by revisiting the site occasionally from May 

2002 to March 2005, and reviewing published and unpublished documents.

HOMEGARDENS IN AGROFORESTRY DEVELOPMENT: TOMÉ-AÇU T 

2.2. Extent and diversity of components 

301 

Homegardens are maintained in areas of 1 to 3 ha around the house of almost every 

Nikkei farmer. The gardens contain a mixture of fruit trees, vegetables, medicinal 

plants, ornamental plants, and tall trees for shade, timber, nuts, fruits, and resin. In 

many farms, the housing areas were distinctively noticeable from a distance by the 

presence of 35 to 40 m tall trees surrounding the house. Homegardens often had a 

plant nursery for the main fields and sometimes a henhouse, a pigsty, a tortoise pen, 

or a small pond for pisciculture. The nursery was often close to the farmhouse to 

facilitate irrigation and close monitoring. The locations of animal sheds and fishponds 

depended on the species, source of water, and security against possible attacks 

by predators and thieves (guard dogs were also used for protection especially at 

night). Each homegarden had a unique appearance due to its history, species 

diversity, and physical arrangements. The owner’s family, labor families, and 

domestic animals living on-farm consumed the homegarden produce. The native 

and/or traditional homegarden fruits such as açai (Euterpe oleracea), mango 

(Mangifera indica), jackfruit ( (Artocarpus heterophyllus), 

and guava (Psidium 

guajava), were often harvested freely for on-farm consumption even without 

the permission of the owner. In farms where the young successors took over the 

management of main fields, the retired but still active parents took care of the 

homegardens, which not only increased the product shipment off-farm but also 

provided engagement and recreational avenues to the senior citizens. Homegarden 

surpluses were also given to the neighbors, friends and relatives, and sold in 

unprocessed or home-processed form at the local markets. Each farm also had 

modest facilities for cleaning and packaging vegetables, extracting and freezing fruit 

juice, making jam and bonbons from fruit pulp, and baking cookies with fruits and 

nuts. The homegarden facilities were also used for processing the off-season main 

field produce, especially when there was not enough to ship to the CAMTA’s juice 

factory. Marketing the homegarden produce gave income to the garden’s caretakers, 

mostly elders, housewives and children, and provided them with an opportunity for 

socialization at the marketplaces. 

2.3. Source of homegarden components 

Most species found in the homegardens had been acquired by the male farmers, 

casually or purposefully, during their travels. In addition, the CAMTA used to send 

exploratory missions for collection of seeds and vegetative materials, when the 

members became interested in certain species/cultivars available at farther locations, 

such as the Caribbean and tropical Asia. Such materials were initially brought to the 

cooperative’s experimental nursery for multiplication and eventual distribution 

among the associated farmers. The local research stations established by the 

Japanese public agencies for emigrant support, such as the Amazon Tropical 

Agriculture Experiment Institute (INATAM; eventually incorporated by the 

Brazilian Agricultural Research Corporation for Eastern Amazon – EMBRAPA 

Amazônia Oriental) also introduced potential crop species for the agricultural 

cooperatives and interested individuals. In addition, some housewives made

Table 1. Inventory of tree species and other useful plants found in two Tomé-Açu Japanese-Brazilian homegardens, with information on their origin, route 

of introduction, and date. 

a. Sakaguchi Farm (Lot # Açaizal 185; given from father-in-law in 1959; homegarden 2 ha) 

Species (Latin name) English name Local name Year and route of introduction Origin 

Albizia spp. ? volunteer 

Annona muricata L. soursop graviola 1991 friend Tomé-Açu 

Araucaria angustifolia (Bertol.) Kuntze. Paraná pine pinheiro do 1988 market Belém 

Paraná 

Artocarpus incisa L. breadfruit fruta-pão 1992 friend Tomé-Açu 

Artocarpus odoratissima Blanco. marang marang 1987 friend Mindanao, Philippines 

Averrhoa bilimbi L. bilimbi bilimbi ? ? 

Averrhoa carambola L. starfruit carambola 1974 CAMTA survey Kuala Lumpur, 

Malaysia 

Azadirachta indica A. Juss. neem nim 1984 friend Thailand 

Bactris gasipaes H.B.K. peach palm pupunha 1970 IPEAN Belém 

Bertholletia excelsa H.B.K. brazilnut castanha do 1932 JPCB Tomé-Açu 

Pará 

Bischofia javanica Blume. Javanese bishopwood 1995 visitor Ogasawara, Japan 

Camellia sinensis (L.) Kuntze. (5 vars.) tea chá 1981 excursion Registro, São Paulo 

Carapa guianensis Aubl. andiroba andiroba 1960 friend (from JPCB) Tomé-Açu 

Carica papaya L. papaya mamão 1979 CAMTA survey Kao-hsiung, Taiwan 

Casuarina equisetifolia L. Australian pine casuarina 1985 friend Philippines 

302 M. YAMADA AND 

Y H.M.L. OSAQUI

Cedrela odorata L. cedro cedro 1984 farm laborer Bujarú, Pará 

Ceiba pentandra L. kapok sumaúma ? volunteer 

Cereus jamacaru D.C. mandacaru mandacaru 1976 INATAM excursion Pernambuco 

Cinnamomum zeylanicum Blume. cinnamon canela 1978 Takasago Farm Daini Tomé-Açu 

Citrus latifolia Tan. Tahiti lime limão Tahiti ? ? 

Citrus aurantium L. bitter orange laranja amarga 1977 INATAM Daini Tomé-Açu 

Citrus deliciosa Ten. (mexerica) common mandarin mexerica ? ? 

Citrus deliciosa Ten. (2 vars.) common mandarin mandarina 1980 Dienberger nursery Limeira, São Paulo 

comum 

Citrus junos Sieb. ex Tan. yuzu ? relative Japan 

Citrus noblis Makino mandarin mandarina 1977 friend Acará, Pará 

Citrus noblis Makino mandarin mandarina 1980 Dienberger nursery Limeira, São Paulo 

Citrus sinensis (L.) Osbeck. orange laranja 1980 Dienberger nursery Limeira, São Paulo 

Citrus unshiu Marc. Satsuma mandarin Satsuma 1986 friend Tomé-Açu (from 

unshiu 

Japan) 

Cocos nucifera L. (praia) coconut coqueiro 1962 relative Tomé-Açu 

Cocos nucifera L. (anão) coconut coqueiro 1970 friend Tomé-Açu 

Coumarouna odorata (Aubl.) Willd. tonka bean cumarú 1965 friend Tomé-Açu (from 

Belém) 

Crescentia cujete L. calabash tree cuieira 1970 excursion Igarapé-Açu, Pará 

Cryptomeria japonica D.Don. Japanese cedar cedro japonês 1980 Japanese consul Belém (from Yaku, 

Japan) 



303


Cryptomeria japonica D.Don. (obi) Japanese cedar cedro japonês 1981 friend Tomé-Açu (from 

Japan) 

Cycas revoluta Thunb. sago palm cica, palmeira- ? friend Tomé-Açu 

sagu 

Dalbergia spruceana Benth. Amazon rosewood jacarandá do 1974 JICA Monte Alegre, Pará 

Pará 

Delonix regia (Bojer ex Hook.) Raf. flamboyant tree flamboyant m 1966 friend 

Belém (from São 

Paulo) 

Durio zibethinus L. durian durião 1979 CAMTA survey Johor Bharu, Malaysia 

Endopleura uxi (Huber) Cuatrec. uxi uxi 1982 friend Daini Tomé-Açu 

Euphoria longan (Lour.) Steud. longan longan 1979 CAMTA survey P’ing-tung, Taiwan 

Ficus elastica Roxb. ex Hornem. Indian rubber tree figueira-da- 1978 friend Atibaia, São Paulo 

borracha 

Garcinia dulcis (Roxb.) Kurz. rata mangostão 1960 friend Tomé-Açu (from São 

falso 

Paulo) 

Garcinia mangostana L. mangosteen mangostão 1961 IPEAN Belém 

Gliricidia sepium (Jacq.) Steud. mother of cacao gliricidia 1985 friend Philippines 

Gmelina arborea Roxb. gmelina gmelina 1970 friend Belém (from Jari, Pará) 

Hevea brasiliensis (Willd.) Muell.-Arg. rubber tree seringueira 1959 Goodyear nursery Igarapé-Açu, Pará 

Illicium spp. 1973 CAMTA excursion Uruçuca, Bahia 

Inga edulis Mart. (long variety) inga inga 1984 excursion Abaetetuba, Pará 

Lansium domesticum Corr. langsat langsat 1979 CAMTA survey Kuala Lumpur, 

Malaysia 


Y H.M.L. OSAQUI

Leucaena leucocephala (Lam.) de Wit. leucaena leucaena 1976 INATAM excursion Juazeiro, Bahia 

Licaria puchury-major r (Mart.) Kosterm. pichury bean puxuri 1982 relative (from JPCB) Tomé-Açu 

Litchi chinensis Sonn. lychee lichia 1979 CAMTA survey P’ing-tung, Taiwan 

Macadamia ternifolia F. Muell. macadamia nut macadâmia 1978 Dienberger nursery Limeira, São Paulo 

Moringa oleifera Lam. horseradish tree moringa 1976 friend Tomé-Açu 

Murraya paniculata (L.) Jack. orange jasmine murta 1981 JICA survey Dominican Republic 

Myrciaria dubia (H.B.K.) Mc Vaugh. camu camu camu mm camu 1987 friend 

Belém 

Myrciaria cauliflora Berg. jaboticaba jaboticaba 1973 CAMTA excursion Uruçuca, Bahia 

Myristica fragans Houtt. nutmeg noz-moscada 1976 INATAM excursion Uruçuca, Bahia 

Nephelium lappaceum L. rambutan rambutan 1979 CAMTA survey Johor Bharu, Malaysia 

Ocotea cymbarum H.B.K. Brazilian sassafras louro- 1988 Takasago Farm Daini Tomé-Açu 

inhamuí 

Pachira aquatica Aubl. Guiana chestnut munguba 1974 JICA Belém 

Paulownia spp. 1993 friend Belém (from Taiwan) 

Persea americana Mill. avocado abacate 1977 friend Tomé-Açu (from São 

Paulo) 

Pimenta dioica (L.) Merr. allspice pimenta da 1976 INATAM excursion Uruçuca, Bahia 

Jamaica 

Platonia insignis Mart. bacurí bacurí ? volunteer 

Pouteria caimito Radlk. abiu abiu 1976 friend Belém 

Pouteria campechiana Baehni. eggfruit canistel 1976 friend Tomé-Açu 

Prosopis juliflora (Sw.) DC. mesquite algaroba 1982 CAMTA excursion Ceará 



305


Quassia amara L. amargo pau amargo 1976 friend Tomé-Açu 

Rollinia deliciosa Saff. biriba biriba 1995 volunteer 

Spondias tuberosa Arruda. umbú umbú 1976 INATAM excursion Juazeiro, Bahia 

Tamarindus indica L. tamarind tamarindo 1977 friend Pakistan 

Theobroma subincanum Mart. cupuí cupuí 1990 volunteer 

Veronica condensata Baker boldo verde boldo verde 1988 friend Daini Tomé-Açu 

Virola surinamensis (Rol.) Warb. virola ucuúba 1995 volunteer 

(herbs; a portion) 

Capsicum spp. 1988 friend Chili 

Curcuma zedoaria (Christm.) Roscoe. zedoary zedoária 1980 friend Campinas, São Paulo 

Elettaria cardamomum (L.) Maton. cardamom cardamom 1971 CAMTA Tomé-Açu (fr. 

Guatemala) 

Manihot aypi Spruce sweet cassava macaxeira ? ? Rio de Janeiro 

Pogostemon patchouli Pell. patchouli patchouli 1960s Takasago Farm Daini Tomé-Açu 

Spilanthes acmella Murr. jambu, toothache jambu 1959 relative Tomé-Açu 

plant 

Vanilla planifolia Andr. vanilla baunilha 1960s IPEAN Belém 

Vanilla planifolia Andr. vanilla baunilha 1969 JICA Mexico 

Vanilla planifolia Andr. vanilla baunilha 1970s Takasago Co. Madagascar 

Zingiber officinale Roscoe. ginger gengibre 1971 CAMTA Tomé-Açu (from 

Trinidad) 

b. Maki Farm (Lot # Cuxiu 2-241; purchased in 1969; homegarden 1 ha) 

Achras sapota L. sapodilla sapoti ? CAMTA nursery Tomé-Açu 


Y H.M.L. OSAQUI

Aleurites moluccana (L.) Willd. candlenut 

1975 San Stepano Farm Daini T-Açu (from 

noz da India 

Hawaii) 

Aniba canelilla (Kunth.) Mez. casca preciosa casca 1980 farm laborer 

preciosa 

Aniba rosaeodora Ducke. rose wood 

1976 INATAM Daini T-Açu (fr. 

pau-rosa 

Manaus) 

Annona muricata L. soursop graviola 1977 friend Daini Tomé-Açu 

Annona squamosa L. sugar apple fruta-do-conde 1990 market Belém 

Artocarpus heterophyllus Lam. jackfruit jaca 1960s former lot owner 

Averrhoa bilimbi L. bilimbi bilimbi 1977 relative Daini Tomé-Açu 

Averrhoa carambola L. starfruit carambola 1960s former lot owner 

Averrhoa carambola L. starfruit carambola 1990 friend Tomé-Açu 

Azadirachta indica A. Juss. neem nim 1987 COPAMASA Tomé-Açu 

Caryocar villosum (Aubl.) Pers. piquiá piquiá 1981 forest Daini Tomé-Açu 

Cedrela odorata L. cedro cedro 1976 forest Daini Tomé-Açu 

Citrus aurantifolia (Christm.) Swingle. lime limão 1960s former lot owner 

Galego 

Citrus latifolia Tan. Tahiti lime limão Tahiti 1982 friend Daini Tomé-Açu 

Citrus deliciosa Ten. (murcote) common mandarin murcote 1970s INATAM Daini Tomé-Açu 

Citrus deliciosa Ten. (tangerin) common mandarin tangerina 1980s relative Daini Tomé-Açu 

Citrus grandis (L.) Osbeck pomelo toranja 1975 friend Daini Tomé-Açu 

Citrus sinensis (L.) Osbeck. orange laranja ? volunteer 

Citrus spp. (limãozinho) 1960s former lot owner 



307


Cordia goeldiana Huber. freijó freijó 1974 forest Daini Tomé-Açu 

Dalbergia nigra (Vell.Conc.) Benth. Brazilian rosewood jacarandá da 1976 JAMIC Bahia 

Bahia 

Dendrocalamus giganteus Wall. ex giant bamboo bambu- 1976 friend Tomé-Açu 

Munro. 

gigante 

Diospyros kaki Thunb. Japanese persimmon caqui 1976 friend Tomé-Açu 

Endopleura uxi (Huber) Cuatrec. uxi uxi 1981 forest Daini Tomé-Açu 

Eugenia cumini (L.) Druce. Java plum, jambolan jamborão 1992 elementary school Daini Tomé-Açu 

Eugenia stipitata McVaugh. araçá-boi araçá-boi 1989 friend Daini T-Açu (fr. 

Manaus) 

Fortunella japonica (Thunb.) Swingle. round kumquat kumquat 1970 friend Tomé-Açu 

Garcinia mangostana L. mangostin mangostão 1990 relative Daini Tomé-Açu 

Macadamia ternifolia F. Muell. macadamia nut macadâmia 1985 CAMTA nursery Tomé-Açu 

Malpighia glabra L. acerola, Barbados 

1980s friend (fr. INATAM) Daini Tomé-Açu 

acerola 

cherry 

Mammea americana L. mammey apple abrico do 1980s friend Tomé-Açu 

Pará 

Mangifera indica L. (haden) mango manga 1978 friend Tomé-Açu (from São 

Paulo) 

Mangifera indica L. (haden) mango manga 1978 friend Tomé-Açu (from 

Mimosa caesalpiniaefolia Benth. sabiá sabiá 1990 CAMTA survey Ceará 

Manaus) 


Y H.M.L. OSAQUI

Morus bombycis Koidz. mulberry amora 1978 ? Tomé-Açu 

Myrciaria cauliflora Berg. jaboticaba jaboticaba 1976 friend Tomé-Açu 

Pachira aquatica Aubl. Guiana chestnut munguba 1970s friend Tomé-Açu 

Persea americana Mill. avocado abacate 1974 INATAM Daini Tomé-Açu 

Pilocarpus microphyllus Stapf. jaborandi jaborandi 1980s friend (fr. INATAM) Daini Tomé-Açu 

Platymiscium ulei Harms. macacauba macacauba 1976 Museu Goeldi Belém 

Prosopis juliflora (Sw.) DC. mesquite algaroba 1982 CAMTA survey Ceará 

Prunus mume Sieb. et Zucc. Japanese apricot umezeiro 1990 friend Daini Tomé-Açu 

Psidium guajava L. guava goiabeira ? volunteer 

Rauwolfia serpentina Benth. Indian snakeroot 1980s friend (fr. INATAM) Daini Tomé-Açu 

Rheedia macrophylla Planch. et Triana. biribá biribá 1988 CAMTA nursery Tomé-Açu 

Schizolobium amazonicum Hub. ex Ducke. paricá paricá 1980 friend Tomé-Açu 

Simarouba amara Aubl. marupá marupá 1977 IBDF Santa Isabel do Pará 

Spondias tuberosa Arruda. umbú umbú 1976 INATAM excursion Juazeiro, Bahia 

Stenocalyx pitanga O. Berg. pitanga pitanga 1980 friend Tomé-Açu 

Swietenia macrophylla King. mahogany mogno 1976 INATAM Daini Tomé-Açu 

Theobroma speciosum Willd. ex Spreng. cacaui cacaui 1980s volunteer 

Veronica condensata Baker boldo verde boldo verde 1992 CAMTA nursery Tomé-Açu 

herbs; a portion 

Cephaelis ipecacuanha (Brot.) A. Rich. ipeca ipecacuanha 1992 friend Daini Tomé-Açu 

Curcuma zedoaria (Christm.) Roscoe. zedoary zedoária 1990 friend Tomé-Açu 



309


Egletodendron pariri pariri pariri ? farm laborer 

Hibiscus sabdariffa L. red sorrel hibiscus ? relative Daini Tomé-Açu 

Kaempferia spp. ? CAMTA nursery Tomé-Açu 

Luffa operculata (L.) Cogn. luffa luffa ? ? 

Spilanthes acmella Murr. jambu, toothache 

? volunteer 

plant 

jambu 

CAMTA = Tomé-Açu Multipurpose Agricultural Cooperative; COPAMASA = Pará Cassava Corporation; IBDF = Brazilian Forest Defense Institute 

(today’s IBAMA); INATAM = Amazon Tropical Agriculture Experiment Institute (today’s EMBRAPA Eastern Amazon Research Station in Tomé- 

Açu); IPEAN = Northern Agriculture and Stockbreeding Research Institute (today’s EMBRAPA Eastern Amazon); JAMIC = Japan Migration and 

Colonization Corporation (today’s JICA); JICA = Japan International Cooperation Agency; JPCB = Japanese Plantation Company mm of Brazil. 


Y H.M.L. OSAQUI


311 

collections of local medicinal plants from the yards of neighboring Brazilians. 

Friends, relatives, neighbors, contract workers, and visitors also brought in plant 

species often as gifts, or in exchange. Table 1 lists the species present in two sample 

homegardens in Tomé-Açu, along with the year of introduction and from where they 

were obtained. 

2.4. Innovative approach of farmer-explorers 

The Tomé-Açu homegardens became well known in the region since the late 1970s, 

thanks mainly to the efforts of two leading farmers: Noboru Sakaguchi (1933 –) and 

Takur Maki (1947 –). Some details on these two farms are furnished hereunder. 

• Sakaguchi farm: Noboru Sakaguchi is a forest science graduate from the Tokyo 

University of Agriculture, who traveled extensively on CAMTA missions in 

search of alternatives to black pepper that had been seriously threatened by 

diseases. From such expeditions, he brought back several species to the 

CAMTA nursery and to his own homegarden. Moreover, after studying the 

species composition and structure of rural Brazilian homegardens and writing 

accounts on traditional farming systems in the Amazon, Sakaguchi reported to 

the CAMTA administrative board that native Theobroma species planted with 

native multipurpose tall trees for shading (such as rubber and andiroba = 

Carapa guianensis) would be most appropriate for sustainable production in the 

region. CAMTA thus introduced the Bahian hybrid cacao (Theobroma cacao) 

to Tomé-Açu in 1971. 

• Maki farm: Takur Maki was one of the pioneer farmers to plant freijó (Cordia 

goeldiana) and macacauba (Platymiscium ulei), two highly appreciated native 

timber trees, for shading cupuaçu and cacao. Maki loved to wander around the 

forests and collect seedlings of useful trees that he learned about from the rural 

Brazilians. Although he did not have frequent chances of travel as Sakaguchi 

did, he looked for interesting species in the homegardens of friends and 

relatives within the settlement. The species procured from distant sources by 

Sakaguchi and others thus spread among the farmers of Tomé-Açu. Maki also 

provided seeds and seedlings from his homegarden to other interested farmers. 

With support from a Japanese public agency for emigrant support, he even 

shipped freijó seeds to other Nikkei settlements in Amazon. 

Although the homegarden caretakers (elders, housewives, and children) 

evaluated the local performance of new plants, final decision regarding large-scale 

planting in the main fields was taken primarily by the male heads of households 

after considering the available market information. Moreover, those pioneering 

family heads had often received tokun (master farmer) education in Japan that 

emphasized diligent practices based on careful observation of nature and taking 

pride in the vocation of producing food. The following case studies of key 

agroforestry species introductions in Tomé-Açu further illustrate the innovative 

approach of these farmer-explorers. 

• Sait-Oshikiri farm [specialty crops: cacao, rubber, brazilnut (Bertholletia 

excelsa) and black pepper]: According to Aiko Oshikiri (1920 – 2000), who

312 

M. YAMADA Y AND H.M.L. OSAQUI 

wrote about the period in the 1930s when Tomé-Açu was called the “green hell 

of poverty and fatal endemics,” she and her mother took care of the homegarden 

plants collected by her father, Enji Sait (1891 – 1958). While acting as the 

president of the vegetable producers’ cooperative (the predecessor of CAMTA) 

for the daily survival of the impoverished Japanese immigrants in the interior 

settlements, Mr. Sait searched for seedlings of ‘permanent crop’ species 

including brazilnut, cacao, rubber, urucu (Bixa orellana), guaraná (Paullinia 

cupana), and black pepper (‘Singapura’ or Kuching variety). He later became 

known as the founder of black pepper culture in the Amazon and the Americas 

(Oshikiri, 1985). It was his son-in-law Tanio Oshikiri (1911 – 1987), Aiko 

Oshikiri’s husband and CAMTA president, who promoted rubber and brazilnut 

among the Tomé-Açu farmers during the mid-1960s as substitute crops for 

black pepper. Today more than forty 70-year-old brazilnut trees remain in the 

Sait-Oshikiri farm, with the largest ones attaining 200 cm diameter at breast 

height and a height of about 35 m. 

• Shimomaebara farm [specialty crop: passionfruit (Passiflora edulis)]: In the 

early 1970s, black pepper fields in Tomé-Açu were severely affected by fungal 

blight (Fusarium solani f. sp. piperis) causing great economic hardship to the 

Nikkei community, which had made this crop their principal source of income. 

Mitsuji Shimomaebara (1914 – 1994), an honors graduate from Matsuda Farmer 

School in Matsubase, Kumamoto, Japan, where he received the tokun 

education, however, developed a simple system of passionfruit culture through 

which the economic hardships of the Nikkei community could be partially 

mitigated. In this method, passionfruit, a common local homegarden vine grown 

on trellises, vigorously climbed on the abandoned black pepper stakes, taking 

advantage of the residual soil fertility and spreading horizontally on a single 

wire extended over the stakes. The fruit bearing vines hung from the wire like a 

curtain and produced excellent results. With growing demand from juice 

factories, passionfruit became a key crop in the black pepper plantations, which 

began succumbing to Fusarium five to six years after planting. In 1974, Mr. 

Shimomaebara was awarded the Marshal Rondon medal for interior 

development by the Brazilian government, as his method of growing passionfruit 

became popular nation-wide. In terms of importance to agroforestry, both 

black pepper and passionfruit provided temporary shade, wind protection, and 

residual soil fertility to the young trees planted between the rows of perennial 

vines. Consequently, native fruit trees previously screened in homegardens such 

as cacao and cupuaçu, and tall trees including rubber, brazilnut, freijó, and 

andiroba established very well in this system (Yamada, 1999). 

• Kusano and Yokokura farms (specialty crops: Theobroma spp. under shade 

trees): After the introduction of Bahian-hybrid cacao in 1971, it became the 

most popular tree crop in the region in the 1970s and early 1980s. However, it 

was susceptible to witch’s broom disease caused by the fungus Crinipellis 

perniciosa (Stahel) Singer. By the mid-1980s, Hisaharu Kusano (1927 – 2003), 

another honors graduate from Matsuda Farmer School, and his son Tsuneo 

Kusano (1948 –) developed disease resistant cacao cultivars. They along with


313 

families toiled for more than a decade in their homegardens and the adjoining 

cacao orchards, conducting individual selection, grafting, and cross-pollination 

of the tiny cacao flowers 2 . The cacao scions screened were disseminated to 

interested farmers along with information on grafting techniques. Their methods 

were also applied to cupuaçu, which had been established as a major field crop 

by Nobuyoshi Yokokura (1914 – 1997), a farming haiku poet who had learned 

tokun discipline in his youth at the Kitami Colonization Training Center, 

Hokkaido, Japan. In the early 1970s, Yokokura anticipated the potential fruit 

pulp market for this homegarden species, which has a growth habit similar to 

that of cacao. He planted the first cupuaçu field at Tomé-Açu with tree shading, 

and distributed seeds to his young followers. Due to misplaced worries of 

cupuaçu transmitting the witch’s broom disease to cacao, local agricultural 

extension authorities warned Yokokura to cut down his cupuaçu trees or lose 

institutional financing. However, he never gave up his orchard. In the 1990s, 

when cacao and the vine crops faced low prices, Tomé-Açu farmers were 

sustained by their cupuaçu pulp sales. The native and shade-tolerant 

Theobroma-based systems thus expanded to 3400 ha or 56% of the main field 

agroforests in Tomé-Açu and opened up new opportunities for planting cacao 

with various useful tall tree species screened in the homegardens (Yamada, 

1999). Again, farmers were initially warned by the extension agents to plant 

only leguminous shade trees (eritrina and palheteira) with cacao, but they 

pursued their own ideas and created productive multistrata/multispecies farms 

within three decades, which now serve as officially recommended models for 

family farms in the region. 

The Nikkei farmers of Tomé-Açu are perpetual innovators. In addition to the 

significant cases of local and regional agroforestry development history listed above, 

we identified during the farm visits various on-going studies involving promising 

crop species, such as acerola (Malpighia glabra), açai, araça-pera (Psidium 

acutangulum), avocado (Persea americana), lime and oranges (Citrus spp.), uxi 

(Endopleura uchi), bacuri (Platonia insignis), spice trees, and other tall tree species 

for timber and non-timber purposes. While each farm became specialized in certain 

species or cropping systems, successful results were shared quickly within the 

community and beyond, partially because of the easy access for curious visitors to 

the homegarden area near the farmhouse. However, it was essentially the 

multipurpose agricultural cooperative (CAMTA) that prompted the development and 

dissemination processes. CAMTA’s Technical Assistance and Extension Division 

(ATEA) had experienced agronomists, who regularly visited these farms. Besides, 

the cooperative received public supports from Japan, such as visiting experts of 

various specialties, training and excursion programs for farmers, introduction of new 

species and varieties, and financial support for the cooperative projects including 

construction of the experimental juice factory. Considering that institutional support 

is crucial in developing complex agroforestry systems (Follis and Nair, 1994), in 

Tomé-Açu the long-lasting and comprehensive collaboration between CAMTA, the 

cooperative representing immigrant farmers, and the Japanese public agencies for 

emigrant support led to the success of agroforestry development. In this scenario, the 

Nikkei homegardens functioned as an informal ‘institutions’ run by the networked

314 

farmers with tokun orientation, complementing the roles of the cooperative and 

public institutions and making their initiatives more effective. Tomé-Açu farmers 

thus realized the intensive and economically viable production systems that 

converted much less forested area to farmlands compared to other prevailing types 

of land development models in the Eastern Amazon, and generated rural 

employment (Yamada, 1999; Nair, 2001; Yamada and Gholz, 2002). 

2.5. Outreach and technology transfer 


Since the mid-1990s CAMTA board members became active in transferring agroforestry 

techniques to non-Nikkei family farms in the neighborhood. Michinori 

Konagano (1958 –), who was in charge of the cooperative’s extension division, told 

the authors that raising production on numerous small family farms would make the 

rural societies peace-loving and the society at large would also be free from criminal 

activities. During the weekends, Konagano would, therefore, visit his neighbors, 

distribute seedlings, and teach agricultural techniques. Konagano was later 

appointed the secretary of agriculture of the municipality of Tomé-Açu. 

Since the early 2000s, the Tomé-Açu agroforestry model gained wider attention 

as a viable alternative to mass forest destruction in the Amazon and orders to 

CAMTA for its products increased from the US, Europe, and Japan. In 2004, the 

Japan International Cooperation Agency (JICA) launched a project in Tomé-Açu, in 

collaboration with the local municipal office, CAMTA, EMBRAPA Amazônia 

Oriental, and POEMA (Poverty and the Environment in Amazonia – a local NGO), 

to establish an agroforestry training center for the young owners of the small family 

farms. In 2005, SAMBAZON, a US-based customer of CAMTA facilitated organic 

certification of açai products, which in turn led to doubling the capacity of the 

cooperative’s fruit juice factory to 2400 Mg month –1 . It encouraged CAMTA to 

disseminate agroforestry among small family farmers of the region, teach them how 

to organize marketing cooperatives, and buy products from these cooperatives for 

processing at the CAMTA juice factory. 


The individual, collective, and public efforts at Tomé-Açu over the past 75 years 

have led to the development of successful multistrata agroforestry systems that have 

attracted worldwide attention. The homegarden as the locus of individual experimentation 

with a variety of crops and their mixtures has been at the center of this 

historical process. Through this developmental process, immigrant farmers overcame 

difficulties in the unfamiliar climate, and established commercial crops such as 

vegetables, black pepper, passionfruit, fruit trees, and other products. This case 

implies that stimulating and supporting farmer initiatives in agroforestry 

homegardens is an effective approach to the development of sustainable rural 

development projects – perhaps more effective than the ‘conventional’ strategy of 

providing farmers with supposedly proven agroforestry modules for their main farm 

fields.


ENDNOTES 

315 

1. Nagasaki Y. 2003. Tomé-Açu ni okeru Nikkei Nka no Sakumotsu Uetsuke 

Jky ni kansuru Tsuiseki Chsa. ACTA, Tomé-Açu, Brazil,19p. 

2. The elder Kusano recounted his tokun philosophy to the authors that farm crops 

grow by listening to the owner’s footsteps (i.e., the owner needs frequent visits 

and careful observation of his field) and that a farm is established only after the 

pioneer’s wooden house has been returned to the soil (i.e., it takes long-term 

efforts to make good soil for sustainable production). Thus, even after the crash 

in international market of cacao, or the sudden drop in cacao bean prices in the 

international markets in the early 1980s, the Kusanos continued their on-farm 

research. However, the traditional tokun farmers sometimes overemphasized 

diligence over rationality and preferred clean culture rather than green mulch or 

grass cover methods in the tropical climate. 

REFERENCES 

Anderson A.B. 1990. Deforestation in Amazonia: Dynamics, causes, and alternatives. In: 

Anderson A.B. (ed.), Alternatives to deforestation: Steps toward sustainable use of the 

Amazon rain forest, pp 3 – 23. Columbia University Press, New York. 

Barrow C. 1990. Environmentally appropriate, sustainable small-farm strategies for 

Amazonia. In: Goodman D. and Hall A. (eds), The Future of Amazonia: Destruction or 

sustainable development? pp 360 – 382. Belhaven Press, London. 

Fearnside P.M. 1995. Agroforestry in Brazil’s Amazonian development policy: The role and 

limits of a potential use for degraded lands. In: Clüsener-Godt M. and Sachs I. (eds), 

Brazilian perspectives on sustainable development of the Amazon region, pp 125 – 148. 

UNESCO Man and the Biosphere Series 15, Parthenon Publishing Group, Carnforth. 

Follis M. and Nair P.K.R. 1994. Policy and institutional support for agroforestry: An analysis 

of two Ecuadorian case studies. Agroforest Syst 27: 223 – 240. 

Gradwohl J. and Greenberg R. 1988. Sustainable agriculture. In: Saving the tropical forests, 

pp 102 – 137. Earthscan Publications, London. 

Homma A.K.O. (ed.). 1998. Amazônia – Meio Ambiente e Desenvolvimento Agrícola. 

EMBRAPA, Brasília, 386p. 

Jordan C.F. (ed.). 1986. Permanent plots for agriculture and forestry. In: Amazonian 

rainforests: Ecosystem disturbance and recovery, pp 58–75. Springer-Verlag, New York. 

Nair P.K.R. 2001. Do tropical homegardens elude science or, or is it the other way around? 


Oshikiri A. 1985. Nyshoku tji no omoide. In: Hensh 

Iinkai (ed.), Tomé-Açu Kaitaku 

Gojusshnenshi, pp 74 – 75. Midori no Daichi, Tomé-Açu Cultural Association, Tomé- 

Açu. 

Serrão E.A.S. 1995. Possibilities for sustainable agriculture development in the Brazilian 

Amazon: An EMBRAPA proposal. In: Clüsener-Godt M. and Sachs I. (eds), Brazilian 

perspectives on sustainable development of the Amazon region, pp 259 – 285. UNESCO 

Man and the Biosphere Series 15, Parthenon Publishing Group, Carnforth. 

Serrão E.A.S. and Homma A.K.O. 1993. Sustainable agriculture in the humid tropics: Brazil. 

In: Committee on Sustainable Agriculture and the Environment in the Humid Tropics, 

Board on Agriculture and Board on Science and Technology for International 

Development of the National Research Council (ed.), Sustainable agriculture and the

316 


environment in the humid tropics, pp 265 – 351. National Academy Press, Washington 

D.C. 

Subler S. 1993. Mechanisms of nutrient retention and recycling in a chronosequence of 

Amazonian agroforestry systems: Comparisons with natural forest ecosystems. PhD 

Dissertation. The Pennsylvania State University, State College, 203p. 

Subler S. and Uhl C. 1990. Japanese agroforestry in Amazonia: A case study in Tomé-Açu, 

Brazil. In: Anderson A.B. (ed.), Alternatives to deforestation: Steps toward sustainable 

use of the Amazon rain forest, pp 152 – 166. Columbia University Press, New York. 

Tanaka N. 1997. Nettai ngy ni okeru takakuka no tenkai to agroforestry ni kansuru ichi 

ksatsu: Amazon chiiki Tomé-Açu mura wo jirei to shite. Hokkaid Daigaku Nkei 

Rons 53: 151 – 163. 

Uhl C. and Subler S. 1988. Asian farmers: Stewards of Amazonia. Garden 12: 16 – 20, 31. 

Uhl C., Nepstad D., Buschbacher R., Clark K., Kauffman B. and Subler S. 1989. Disturbance 

and regeneration in Amazonia: Lessons for sustainable land use. The Ecologist 19: 

235 – 240. 

Uhl C., Nepstad D., Buschbacher R., Clark K., Kauffman B. and Subler S. 1990. Studies of 

ecosystem response to natural and anthropogenic disturbances provide guidelines for 

designing sustainable land use systems in Amazonia. In: Anderson A.B. (ed.), 

Alternatives to deforestation: Steps toward sustainable use of the Amazon rain forest, pp 

24 – 42. Columbia University Press, New York. 

Yamada M. 1999. Japanese immigrant agroforestry in the Brazilian Amazon: A case study of 

sustainable rural development in the tropics. PhD Dissertation, University of Florida, 

Gainesville, 821p. 

Yamada M. and Gholz H.L. 2002. An evaluation of agroforestry systems as a rural 

development option for the Brazilian Amazon. Agroforest Syst 55: 81–87.

CHAPTER 18 

URBAN HOMEGARDENS AND ALLOTMENT 

GARDENS FOR SUSTAINABLE LIVELI- 

HOODS: MANAGEMENT STRATEGIES 

AND INSTITUTIONAL ENVIRONMENTS 

A.W. DRESCHER 1 , R.J. HOLMER 2 , AND D.L. IAQUINTA 3 

1 Albert-Ludwigs-Universität, Freiburg, Germany; E-mail: 

. 2 Xavier University College of Agriculture, 

Cagayan de Oro, The Philippines. 3 Nebraska Wesleyan University, Lincoln, 

Nebraska, USA 

Keywords: Food security, Households, Species diversity, Sustainable development, Urban 

agriculture, Urban planning. 

Abstract. Diversity of food and income resources is one of the main buffers against 

vulnerability of the urban poor. Based on the authors’ field experience in the Philippines, 

Latin America, and southern Africa, and involvement with various other project evaluations, 

this chapter discusses the major differences between individual homegardens and allotment 

gardens and their respective roles in urban livelihood support programs. Major differences 

between these two systems of gardening are in their respective decision-making processes and 

impacts—in terms of both quantitative and qualitative outcomes. Current land use planning, 

multistory housing, and land use competition from different sectors limit both open space and 

space for gardening in the urban centers, necessitating lobbying and public advocacy to 

support such garden systems. While homegardens need public advocacy and extension 

services, allotment gardens additionally require significant political intervention to secure 

land, organize access, and support development. Implicit in this is the need for identifying the 

institutional barriers as well as gathering support for gardening projects in urban and 

periurban environments, prior to promoting the urban-gardening programs. 


Urban agriculture is the general term used to refer to a wide variety of food 

production practices in and around cities. Together with periurban agriculture, it 

317 




318 A.W. DRESCHER ET AL . 

represents a continually growing activity and consequently an emerging research 

area. Urban gardening is perhaps the most significant component of urban 

agriculture from the perspective of individual practitioners. It includes three types of 

practices: homegardens, allotment gardens and community gardens. While there is 

no universal agreement on the precise meaning of these terms, we adopt the 

following definitions. Homegardens are maintained – typically, but not always, near 

the homes – by individuals or households who have some access to land (either 

customary or legal), which they have arranged for themselves. Allotment gardens 

are separate parcels of land allocated to individuals or households for personal use. 

While contiguous, each household works on the parcels independently and the land 

is made available through either government action or private enterprises. The 

individual households are organized into self-governing associations. Community 

gardens are maintained by a group of individuals or households who produce 

agricultural goods collectively on a piece of land primarily for self-consumption. 

The extent and significance of urban gardening have been discussed elsewhere 

(Mougeot, 2005). Suffice to say that, it is widespread and that all three forms of 

gardening are growing throughout the world, particularly in response to income 

deprivation and the crises involving economic recession, natural disasters, and civil 

disorder (Jacobi et al., 2000). This is especially significant for the large segments of 

urban poor that continue to grow. 

In an extensive treatment of urbanization, urban and periurban agriculture, and 

urban poverty, Drescher and Iaquinta (2003) addressed an array of issues relevant to 

this chapter. In particular, they identified the characteristics of urban and rural poor, 

the very different socioeconomic conditions of urban poor in the developed and 

developing economies (see also UNCHS, 2001), and the gender-related aspects of 

urban gardens in different social and cultural systems. Another important 

consideration is that people practice urban gardening – whether home or allotment – 

for more varied reasons in the developed countries than in developing countries, 

principally because of the size of the economically stable population. Food 

production and income generation are important in both places, but the objectives of 

middle-income urban gardeners in developed countries (who grow flowers, create 

leisure environments, build kids’ playground, promote outdoor meeting places, etc.) 

are often different from those of their low-income counterparts. 

This chapter discusses the importance of urban gardens, highlights the 

constraints faced by urban gardeners, and addresses possible resolution of such 

constraints. Central to this discussion is the extension of the homegarden model to 

allotment gardens and the elaboration of the institutional contexts within which 

household livelihood strategies operate. 

2. DIVERSITY AND THE INVENTIVE SPIRIT OF URBAN GARDENING 

A common problem for urban gardens is the increased demographic pressure on 

available land, which we call spatial densification. This is caused or exacerbated by 

planning regulations intended to avoid urban sprawl and the desire of homeowners 

and users to create more living space. Because spatial densification means adding 

more people to the same area, it involves increased housing construction, which

URBAN U HOMEGARDENS AND ALLOTMENT GARDENS G 

319 

competes directly with the land available for gardening. This is a major problem for 

the poor, urban squatters and for the residents in spontaneous, periurban land 

occupations. Despite this apparent space constraint, homegardens are common in 

many urban environments. 

A probable solution to the constraints imposed by the overall urban situation is 

the development of alternative production systems adapted to lack of space, water, 

and other inputs. Soil-less cultures such as hydroponics, substrate cultures, and 

container gardens are just a few examples. Rooftop gardening is increasing in many 

densely populated cities. Even poultry farming and pig rearing take place on the 

rooftops or within houses, and are sometimes promoted by the local nongovernmental 

organizations (NGO) or international organizations such as the Food and 

Agriculture Organization (FAO, 1998; Drescher and Iaquinta, 2003). Since 1985, a 

cooperative of 100 poor women in Bogotá, Colombia, have used rooftops to grow 

hydroponic vegetables for city supermarkets. Unmarketable crops are either fed to 

livestock or used for home consumption. 

In Lima, Peru during the past two decades, the Ministry of Agriculture, FAO and 

United Nations International Children’s Emergency Fund (UNICEF) have promoted 

household and community kitchen gardens to avert widespread hunger. The Center 

for Education and Technology in Santiago, Chile, promotes 20 m 2 gardens, where 

plants are raised in containers stacked up in pyramids and walls are used for trailing 

vines (FAO, 1998). In Sri Lanka, “edible air-scapes” are promoted by the island 

nation’s department of agriculture as a strategy for rebuilding in the aftermath of the 

tsunami. Walls, bottles, bags, and fences are used for raising plants as part of this 

(Ranasinghe, 2005). The Cuban example of “organoponics” (Cruz and Medina, 

2003; Pinderhughes, 2004) is a well-established system incorporating both efficient 

water conservation and the use of compost and manure for fertilization in the urban 

context. 

In other parts of Latin America also, a particular form of organized homegardens 

exists (the so-called microgranjas), which involves the production of vegetables, 

fruits, other products, and small animals (e.g., chickens, pigs, guinea pigs, or rabbits) 

(Arias, 2000). The distinguishing characteristic of the microgranjas is that with 

governmental support, the homegardeners have been organized into groups. It thus, 

provides a platform for exchange of information and knowledge, despite the spatial 

dispersion of cultivated plots. Thus, while there are several forms of urban 

gardening, they can broadly be classified as homegardens, allotment gardens, and 

community gardens. 

3. URBAN HOMEGARDENS 

An urban homegarden, a multispecies production system on the area of land around 

the house to meet different physical, social, and economic needs and functions, is 

traditionally an important land use activity for individual households. Although its 

functions are similar throughout the world, focusing principally on subsistence or 

income generation, their structure and size vary considerably. For instance, in Papua 

New Guinea, Vasey (1985) reported that house plots generally range between 300 to 

400 m 2 but are often too small to meet the household demands. Consequently, many

320 

A.W. DRESCHER ET AL . 

households establish second gardens away from the house. Christanty (1990) 

reported that the size of homegardens in Bangladesh ranged between 30 and 700 m 2 , 

with an average of 200 m 2 . Hoogerbrugge and Fresco (1993) also found wide 

variations in the size of homegardens even within a given country. For example, 

they reported size estimates ranging from 10 to 120 m 2 and 5000 to 20 000 m 2 in 

two separate studies in Zambia and ranges of 172 to 500 m 2 and 200 to 1700 m 2 in 

two Javanese studies. Drescher (1998) also observed large variations in this respect 

(17 to 865 m 2 in Lusaka, Zambia), but indicated that the majority of urban 

homegardens were less than 300 m 2 and that on average, women’s gardens were 

more than double the size that of men’s. Prosterman and Mitchell (2002) suggested 

that a great majority of homegarden plots in Java were less than 200 m 2 . Christanty 

(1990) earlier showed that on the less densely populated Indonesian islands, 

homegarden plots averaged 2500 m 2 , sometimes reaching a size of three hectares. In 

Lima, Peru, Hetterschijt (2004) also found that the size of urban organic 

homegardens varied between 25 and 900 m 2 , with an average of 110 m 2 (n = 109). 

In several studies, however, the failure to designate the study areas as rural, urban, 

or periurban, and the lack of universally accepted definitions for these classifications 

complicate the matter. For example, in thickly populated regions such as Java 

(Indonesia) and Kerala (India), the distinction between rural and urban settings is 

rather blurred. Iaquinta and Drescher (2000) have shown that this is even more 

pronounced in periurban areas. 

Nonetheless, four points emanate from the discussion above. First, the size of 

homegardens varies considerably across cultures and even within them. Second, the 

size of the majority of homegardens tends toward the low end of the range 

(positively skewed distribution pattern). Third, households sometimes make 

managerial decisions to locate some or all homegardens geographically distant from 

the house due to space or other constraints. Fourth, households make numerous 

management decisions, which, along with environmental constraints, determine both 

the physical structure and the outputs of the homegarden. 

Urban homegardens integrate a variety of physical, social, and economic 

functions. Typical homegardens include (1) physical areas for living, storage, and 

waste disposal, (2) social areas for meetings, children’s playgrounds, and display, 

and (3) economic areas for raising animals and for growing food, medicinal plants, 

and fruit trees. Overall, the homegarden is a place for people to live but also a place 

to produce a variety of foods and products for home consumption and income 

generation (Landon-Lane, 2004). 

Homegardens also play an important role in the conservation of indigenous 

crops, thus enhancing biodiversity in rural, periurban and urban environments. 

Drescher (1998) and Boncodin et al. (2000) found a variety of indigenous vegetables 

in the homegardens and in the local markets of Lusaka, Zambia. In particular, 

Amaranthus sp. grows semi-cultivated in the gardens. Other examples include 

Bidens pilosa, Brassica sp., Corchorus sp., Solanum macrocarpum, Hibiscus sp., 

Cleome sp., Ipomoea batatas (sweet potato), and Cucumeropsis edulis (squash). 

Indigenous tree species providing multiple products such as firewood, food, 

fruits, and medicines also abound in the homegardens, yet they are often overlooked 

when talking about urban homegardens. For example, the leaves of the horseradish


321 

tree (Moringa oleifera) grown in the homegardens are the most frequently consumed 

vegetable among the households of Cagayan de Oro City, the Philippines (Agbayani 

et al., 2001). 

Urban homegardens compensate to a certain extent for the gardener’s restricted 

access to natural resources. While gathering wild vegetables and roots is still 

prevalent in periurban and rural areas, such options are limited in the urban context. 

For example, only 39% of households included in a study in Lusaka gathered wild 

fruits and vegetables, compared to 76% in periurban and 86% in the rural areas. 

Thus, in the urban context, homegarden produce provides an economical and 

nutritious substitute for wild vegetables and roots (Drescher, 1998). As seen in Fig. 

1, trees [e.g., peach (Prunus persica), papaya (Carica papaya), mango (Mangifera 

indica), and Morus alba] are an integral part of the urban homegardens in Lusaka. 

Figure 1. Map of an urban homegarden in Lusaka (Zambia) (Source: Drescher, 1998). 

While homegardening provides subsistence and supplementary household food 

supply, Boncodin et al. (2000) showed that it concurrently makes a significant 

contribution to the amount of nutrients and variety in the household food intake. In a 

study of the rural homegardens in the Philippines, they identified 33 different food 

crops, including green, leafy, and yellow vegetables; starchy roots and tubers; and 

legumes, beans, nuts, and spices. Homegardens thus provide year-round food 

supplements to households not only in terms of quantity but also in terms of food 

diversity and variation, and play an important role in providing Vitamin A and 

Vitamin C as well as supplying one-third or more of calcium and iron needs (Kumar 

and Nair 2004). This is consistent with the findings of a study on urban 

homegardens in the Philippines 1 .

322 

3.1. Community and allotment gardens 


Community gardens are defined as gardens where people share the basic resources 

of land, water, and sunlight (MacNair, 2002). Allotment gardens, a special type of 

community garden, were first developed in Germany. Introduced as Schrebergärten 

in the mid-1800s, they flourished over a century and a half (Kasch, 2001). Allotment 

gardens are characterized by a concentration in one place of several small land 

parcels (usually 200 to 400 m 2 each). Individual families are organized into an 

association, which assigns the land parcels. In allotment gardens, the parcels are 

cultivated individually, as compared to community gardens where the entire area is 

tended collectively by a group of people (Holmer et al., 2003). Community gardens 

are often organized around a particular institution such as school, workplace, faith 

organization, hospital, etc. They may also be organized around social characteristics 

such as ethnicity, age, or religious orientation. 

In the Philippines, the production practices for vegetables in urban allotment 

gardens are similar to those in the rural areas; however, they differ particularly in the 

choice of cultivars and in the reduced application of agrochemicals due to the 

proximity to populated areas (Guanzon and Holmer, 2003). Although allotment 

gardeners are not excessively environment-oriented, nor are there many government 

restrictions on the use of agrochemicals, they are usually market-oriented. That is, 

about 70% of the produce is marketed directly within the garden itself—mostly to 

close neighbors; the consumers are generally well aware of the production practices 

and do not accept produce that has been heavily sprayed with chemicals. This 

situation differs greatly from the general system where vegetables are anonymously 

produced in far away locations and the customers mostly make assumptions 

regarding the production practices. This contrast is particularly true in the 

developing countries where government food safety controls are lax and quality 

labeling is either non-existent or unreliable. 

A preliminary study 2 in Cagayan de Oro, the Philippines indicated that the local 

people perceive the multiple benefits of allotment gardens. While 25% of the 

vegetables produced were consumed by the family or shared among friends, 75% 

were sold to neighbors or walk-in clients who come directly to the garden and who 

appreciate the freshness of the produce, the convenience of proximity, and the 

relatively lower price than the public markets. The gardening activities, a secondary 

occupation for all association members, thus augmented their incomes by about 

20%, while vegetable consumption also increased by about 75%. This is especially 

notable since the average vegetable consumption in Cagayan de Oro is only 36 kg 

per capita per year, about one-half of the minimum recommended intake suggested 

by FAO (Agbayani et al., 2001). In addition to these direct effects, the gardeners 

appreciate the strengthening of the community values brought about by allotment 

gardening. 

The gardens are also essential for the successful implementation of the city’s 

integrated solid waste management program. Segregated biodegradable wastes from 

neighboring households are delivered to the allotment gardens where they are 

converted into compost. The amount of residual waste delivered to the landfill site 

from these areas could theoretically be reduced by more than one-third, if all


323 

biodegradable wastes are channeled to a system of composting by allotment gardens 

in the city 3 . The city government of Cagayan de Oro is presently mainstreaming this 

concept into its overall city planning and development, using participatory GISbased 

approaches to identify suitable areas for further expansion of allotment 

gardens (Emmanuel Abejuela, pers. comm., August 2005). The advantages of such 

an approach are manifold. For example, mineral fertilizer application can be 

drastically reduced by using enriched compost, m thus reducing the danger of ground 

water pollution. The difference between urban and rural gardens is that the former 

uses biodegradable household wastes from many nearby households organized by 

the local government, and not only the bio-waste generated within the garden. Thus, 

urban gardens have a comparative advantage in their access to organic inputs 

generated by urban households, which is important in view of projected increase in 

demand for organically grown food. 

3.2. Urban homegardens, allotment gardens, and community gardens: a comparison 

The most important feature of allotment gardens is that they are institutionally 

administered and organized, and they serve as a community facility and a place of 

social interaction. In the German context, each gardener in an allotment garden 

needs to be a member of the respective Kleingartenverein (allotment garden 

association). In developing countries, however, gardeners are often not members of 

any associations, but are part of the community. The gardens are not necessarily 

near the homes, but rather located where sufficient space is available, and sometimes 

where favorable soil and water conditions exist. Obviously, transportation issues 

arise as the distance between homes and gardens increases. In other cases, the 

gardens are located in areas unsuitable for buildings or they are established as buffer 

zones along rail corridors and highways. They may even be located in protected 

areas necessary to balance the urban microclimate, as is the case in some German 

cities. Here allotment gardens are used on green belts to facilitate cold drainage, thus 

reducing urban heat island effects and the demand for air conditioning 

(Landeshauptstadt, 1998; Innenministerium, 2004). 

In Lusaka, Zambia, community gardens can be found on the edges of densely 

populated compounds that do not allow people to grow food within the housing area. 

A similar situation existed formerly in Harare, the capital of Zimbabwe. In Port 

Elizabeth, South Africa, some allotment gardens are situated on common areas and 

on the grounds outside churches, schools and hospitals (Jarlöv, 2000). Unlike 

Germany where allotment gardens are located on public lands owned by the city or 

railway, all allotment gardens of Cagayan de Oro, the Philippines, were established 

on private lands, due to the lack of publicly owned open spaces (Holmer et al., 

2003). In Cagayan de Oro, the chairpersons of the barangay (city district) simply 

asked the local private landowners if poor residents of the barangay could use their 

vacant land for food production. To preclude residential occupation, however, the 

gardeners are only permitted to construct a small shed for tools and other garden 

implements, and not allowed to establish residential structures. Conditions for land 

use are then formalized into a memorandum of agreement jointly signed by all

324 


stakeholders to legitimize access to the land for horticultural purposes (Vélez- 

Guerra, 2004). 

Urban homegardens have several characteristics that are similar to those of nonurban 

homegardens: (1) their location adjacent to homes (2) close association with 

joint family activities and (3) wide diversity of crop and livestock species used to 

meet family needs (Landon-Lane, 2004). Thus, homegardening is not simply a 

spatial integration but is characterized by a social and socioeconomic integration of 

the families involved. Importantly, overemphasis on the first characteristic 

oversimplifies the realities faced by residents in places such as the seasonal tropics 

of Africa, where homegardens are not necessarily near the homes; rather, they are 

located near water sources due to long dry seasons (Fig. 2). This could be a major 

reason why these production systems have often been overlooked in the past 

(Drescher, 1998). 

Figure 2. Homegardens and adjacent community gardens in Southern Zimbabwe. Some 

gardens are near the houses (upper left) some others are distant to the houses near the water 

source. 

The choice homegardeners make in locating their plots represents a clear 

environmentally informed management decision, and not a decision to engage in 

something fundamentally different from homegardening, or one made by the local 

authorities. Such gardens are household-based small-production entities and it is 

best that our definition be concerned less with the location than the nature and aim 

of the activity. 

The geographic separation of homegardens and dwellings is not unique to 

Africa. During the first half of the twentieth century in urban United States, 

immigrant groups such as Southern Italians practiced substantial levels of


325 

homegardening for both cultural and economic reasons similar to the poor around 

the world today. Evidence on the extent of such practices in the US is found in 

historical collections of folklore and ethnographic narratives. A particularly good 

example is from Pennsylvania 4 , where the Southern Italians practiced urban 

gardening to produce subsistence items (direct consumables) and preserved food 

stock, highlighting the centrality of food to Italian culture, family, and gender roles. 

More recent examples such as MacNair (2002) demonstrate that the same groups 

such as Italians continue to practice urban gardening often under conditions of land 

scarcity. In this example from Montreal, it resulted in policy interventions to 

regularize the practice under municipal authority. Often the gardens were at some 

distance away from the urban dwellings due to the nature of the land market. 

Multiple tenancy dwellings, however, made no provisions for land use by occupants. 

Available land was scattered and had to be purchased or leased. This meant that 

many households maintained multiple plots scattered around the neighborhood or 

further afield. As in the case of the African gardens in the semiarid tropics discussed 

above, we view these gardens as clear examples of homegardens because of their 

centrality to family activities and the absence of municipal facilitation. The 

difference in this case is that the management decisions on the part of the 

practitioners were conditioned primarily by land availability rather than the labor 

costs associated with transporting water. Nonetheless, the majority of homegardens 

are located adjacent to the houses. 

4. ALLOTMENT GARDENS: A SPECIAL CASE OF HOMEGARDENS 

OR AN ‘INDEPENDENT’ SYSTEM? 

4.1. Secured access to space: an important question 

Stakeholders themselves differ in important ways, mostly in terms of economics and 

land tenure. Generally, homegardeners are not the “poorest” residents on the 

socioeconomic scale since they already have access to land. Their challenge is 

primarily political, mobilizing a fragmented group of individual households with 

shared but unrecognized common interests. 

The establishment of allotment gardens also requires space within the city 

boundaries. This is important only partially to minimize transportation issues. 

Women with children need to be near the house because they are generally involved 

in multiple household tasks such as cooking, firewood collection, cleaning, and 

childcare. However, the reservation and allocation of land for allotment gardens is a 

major problem in most urban settings. City authorities rather tend to create public 

parks or golf courses, which they consider more in line with the urban disposition. 

Generally, gardening does not fit into the conceptualization of urbanization or the 

philosophy of urban planners, and this makes it difficult to convince urban 

authorities that agriculture in the city is not inherently a problem, but a solution to 

various other urban problems (Drescher, 2001). 

Securedd access to land is especially relevant when considering the role of trees in 

urban gardens of all kinds. To justify capital and labor inputs, the urban gardener

326 


must realize a return on investment. For some crops such as sweet potatoes grown 

along roadsides, this occurs in a single growing cycle. However, where soil building, 

tree planting, pond construction, or landscaping, for example, are involved, the time 

span of guaranteed d access increases, and a number of years is required to recover 

labor and capital inputs. For allotment gardeners this means sufficiently long sitespecific 

‘lease’ arrangements for the allotment as a whole combined with binding 

association ‘rules of access’ for individual participants. For homegardeners this is an 

issue of legally formalized land tenure rights, meaning either a deed/title system or 

the legal enfranchisement of usufruct rights for individuals in communal ownership 

contexts or in long-term spontaneous land occupations. 

4.2. Lobbying or political mobilization is particularly relevant to allotment gardens 

Overcoming limited vision, economic constraints, and political resistance requires 

effective lobbying. Yet, allotment gardeners typically represent some of the most 

politically alienated and economically impoverished residents, requiring external 

expertise to facilitate participatory lobbying and support for allotment gardens. 

Often nongovernmental organizations (NGOs), educational institutions, or 

sympathetic municipal agencies play this role. However, effective lobbying also 

requires solutions tailored to local conditions. Urban and periurban environments 

represent a “lumpy continuum” of human settlement with important and varied 

institutional capacities (Iaquinta and Drescher, 2000; 2001; 2005). The respective 

roles of home and allotment gardens differ dramatically across this continuum, as do 

the lobbying strategies necessary, even within a single municipality. 

4.3. Who makes the key management decisions that directly affect outputs? 

Observations made in Zambia and Zimbabwe show that community gardens are 

often located in unsuitable locations, distant from water sources and/or with bad soil 

conditions. Further, as compared to homegardens, trees are generally lacking in the 

observed community gardens (Drescher, 1998). These facts suggest that some of the 

limitations experienced in community gardens are due more to poor administrative 

planning, constrained extension support and gaps in data collection related to the 

importance of trees than to problems inherent to community gardens per se. 

Bilateral and reciprocal transfer of knowledge between gardeners, extension 

officers, and local officials is required to properly understand the smallholders’ land 

use system and management strategies (Drescher, 1996). 

Development projects tend to over-regulate the maintenance and management of 

allotment gardens. Typically, “the authorities” select and provide the seeds and 

fertilizers and this minimizes growing traditional vegetables, mitigates the role of 

the garden as a site for household experimentation, and marginalizes the gardener as 

innovator. Further, attitudes of municipal authorities also reflect the general 

underestimation of the multifunctional role of trees in cities, including food security.


5. THE HOMEGARDEN MODEL 

5.1. Relationship to household livelihood strategy 

327 

The homegarden model (Fig. 3) is based on the assumption that homegardening is a 

process that forms part of the household livelihood strategy. Household livelihood 

security is defined as “adequate and sustainable access to income and resources to 

meet basic needs, including adequate access to food, potable water, health facilities, 

educational opportunities, housing, time for community participation and social 

integration” (Frankenberger and McCaston, 1998). For the present discussion, our 

emphasis is on the access to food, directly, or by means of access to income and 

resources. 

The model in Fig. 3 was originally designed for homegardens, but has been 

modified to incorporate allotment gardens. The major differences between the two 

systems of gardening are in the decision-making process and their differing impacts, 

in terms of both quantitative and qualitative results. Allocation of assets in the model 

follows the treatment of Swift (1989). Collective assets (e.g., tools, stores and 

buildings) are of particular importance for allotment gardens. 

The model implies that the household decision to get involved in gardening 

depends on factors such as the existence of a supportive general environment, access 

to land and water resources, and the availability of specific inputs (seeds, 

knowledge, work and time). The model also identifies household vulnerability 

factors that either stimulate or inhibit household involvement in subsistence and 

market-oriented food production. The risk of livelihood failure determines the level 

of vulnerability of a household to income, food, health, and nutritional insecurity. 

Therefore, livelihoods are secure when households have secured ownership of—or 

access to—resources and income earning activities, including reserves and assets, to 

offset risks, ease shocks, and meet contingencies (Chambers, 1989). 

The model in Fig. 3 implicitly depicts homegardening as embedded in the 

livelihood system, interacting with the socioeconomic and environmental conditions 

of the larger system. It helps to identify factors that promote or inhibit household 

gardening and assists in the development of scenarios for different contexts 

regarding climate, space, politics, institutional framework, culture, and economics. 

Homegardens support important farm-development activities; some farm inputs 

come from homegarden activities such as plant propagation, raising and housing 

draught animals, and making and repairing tools. New crops and farming techniques 

are often first tried out in the homegarden, which is also an area for drying, 

processing, and storing farm products (Landon-Lane, 2004; Yamada and Osaqui, 

2006). In small compound homegardens in Zambia, sweet potato seedlings are 

planted and later transferred into the fields during the rainy season. Sweet potato 

leaves are also used as a vegetable both in dry and rainy season. Nearly 40% of the 

compound residents of Lusaka are engaged in staple crop production during the 

rainy season, while about 25% are engaged in homegardening and community 

gardening (Drescher, 1998).

328 


Figure 3. The Homegarden Model (adapted from Drescher, 1998). 

5.2. Why so important in the urban context? 

Urban environments differ considerably from rural environments. Urban poor 

dwellers are more likely to report food insecurity and heavy dependency on urban 

markets (Zalilah and Khor, 2004). At the same time, there are clear signs that poverty 

and malnutrition in cities are increasing, especially in slum areas and high-density


329 

compounds (FAO, 2001; Iaquinta and Drescher, 2002). Low purchasing power and 

declining access to food are major problems in many rapidly growing cities in the 

developing world (Drescher and Iaquinta, 2003). 

Both home- and allotment gardening can partly compensate for the deficiencies 

of urban poor households and alleviate food insecurity. Thus, secured access to 

resources such as land, water, seeds, and tools is key to increasing food security in 

cities. Further, food production is only one dimension, albeit t an important one, of 

the benefits to be derived from urban and periurban production. With proper 

assistance and management techniques, these environments are positively impacted 

through provision of shade, microclimate modification, waste recycling, soil 

stabilization, and soil building. 

5.3. Applying the model to allotment gardens 

If we apply the homegarden model (Fig. 3) to allotment gardens, it is evident, that 

the two systems are similar regarding access to resources, assets, activities, and 

outcomes. However, the arena of decision-making becomes more complex. 

Essentially, we need to “nest” the individual family as a decision-maker within the 

allotment association as the decision-broker. The decision module appears as a 

shaded pentagon in the center of the model and represents the decision-making 

process (e.g., what to grow, when to grow, where to grow, with whom and how to 

interact and cooperate, the balance between short and long-term investments, etc.). 

For homegardens, it is simply a single module (pentagon). For allotment gardens, 

however, it appears as a series of household pentagons nested within a larger 

“association” module. In practice, the entire decision-making process becomes more 

transparent and probably more standardized in the case of allotment gardens because 

individual family decisions are now directly guided and influenced by the 

association in a more public forum. Assets and activities can be better shared in 

allotment gardens and the transcendent outcomes are eventually more visible, 

mainly regarding community empowerment, social peace and status, and economic 

power. In this sense, we do not consider individual parcels in allotment gardens as 

independent farming systems in the way that homegardens are. However, an 

allotment garden association is an institution and it may conflict with preexisting 

institutions—particularly but not exclusively—those of a traditional form. That is 

why it is so important to understand the relationship between the type of urban or 

periurban environment and the respective institutions (Iaquinta and Drescher, 2000; 

2001; 2005). 

5.4. Contrasting roles of homegardens vs. allotment gardens within the general household 

gardening model 

Allotment gardens, properly institutionalized and integrated into urban planning have 

their biggest influence on the level of the (non-) supporting structures and can 

regularize better entitlement and access to resources. Thus, allotment gardeners have

330 


an enhanced voice through the association, while homegardeners typically remain 

isolated, or “just “ gardeners!” 

Because homegardening is seen primarily as a private activity, there is little 

public support for these gardens. Homegardening is only done when the specific 

circumstances permit it; for the most part space availability is the major determinant. 

Public support is more likely to occur in relation to allotment gardens, because of 

their greater visibility. 

With respect to the environmental impact and output of both activities, the 

differences depend primarily on the management strategies employed. For both 

systems, space is restricted. Allotment gardens allow gardening for those who do not 

have access to land near their residences. 

Thus, in terms of access, both homegardens and allotment gardens have a clear 

political dimension. On the one hand, The World Food Summit 2002 (FAO, 2002) 

reaffirmed the right of everyone to have access to safe, nutritious, and culturally 

relevant food. The ability to grow food is one important dimension of this access 

relevant to both types of gardens. On the other hand, community empowerment, 

which is especially relevant to allotment gardens, is inherently a process of 

advocacy and political negotiation among municipal authorities, local residents, and 

various interest groups. 

6. HOW DO MANAGEMENT STRATEGIES FOR HOMEGARDENS 

AND ALLOTMENT GARDENS DIFFERENTIALLY IMPACT OUTPUTS? 

6.1. The example of species diversity: observations from Africa 

Species diversity or “garden biodiversity” provides an excellent example of the way 

management strategies and their outcomes differ between the two systems. Species 

diversity is determined by two factors: the number of species per garden and the 

abundance of each species within the community in a given area. Which species are 

planted and how much of each species gets planted represent fundamental 

management decisions. 

Important inferences can be drawn from a study of rural homegardens and rural 

community gardens in southern Zimbabwe (Drescher et al., 1999), where, clear 

differences were found in the diversity of plant species in community and individual 

gardens. Individual gardens (average 8.6; range 5 to 12 species) showed a higher 

species diversity than community gardens. Community gardens averaged only four 

species (Drescher et al., 1999). In a related study, garden species-diversity was 

shown to be positively correlated with the prevalence of biological antagonists of 

crop pests (Drescher, 1998). Together, these results point to the benefits of 

promoting individual gardens and the need for extension strategies adapted 

specifically for community gardens. 

Sweet potato, grown for both its nutritious leaves and starchy tubers, further 

illustrates the different approaches to management in the two types of gardens. 

Sweet potatoes were cultivated in all individual gardens but in only one community 

garden (Drescher et al., 1999). In homegardens, sweet potatoes serve as early patch

331 

leafy greens and are transplanted to fields later during the rainy season. This is a 

household decision that contributes to a more balanced diet, provides good early 

ground cover, and releases land for subsequent cultivation. Authorities decided, 

however, that the sweet potato would not be recommended for planting in 

community gardens, hence distributed no sweet potato seeds to gardeners. These 

two management systems could be brought into greater alignment with appropriate 

research and support. One opportunity for this is to transfer greater management 

authority to allotment associations, which typically function more interactively with 

households than do formal institutions and authorities. 

6.2. Other outputs 


In private homegardens, output in terms of yield can suffer due to the lack of labor 

and time. This can be more easily compensated for in allotment gardens, first by 

more cooperation between families and second by economies of scale relative to 

water supply and other capital investments and crop management techniques. 

6.3. Marketing of surplus easier in allotment gardens? 

Allotment gardens can produce marketable surplus produce too. Greater 

concentration of output creates economies of scale wherein both direct marketing 

and production are facilitated. In Cape Town, South Africa, for instance, the 

Siyazama community allotment garden produced both for the market as well as for 

home consumption of 15 dependent families. Public support for allotment gardens 

might help urban poor to get better access to markets 5,6 . 

The lack of public support for homegardens, however, reduces the number of 

such gardens in cities. Current land use planning, multistory housing and land use 

competition from different sectors often limit the open space and space for 

gardening. Nonetheless, in most cities, open space, unused sites, and idle land are 

still widely available. Fig. 4 shows the use of open spaces for gardening near Manila 

International Airport. Elsewhere in the Philippines, tax policy has been used to 

stimulate changes in land use. For example, in Cagayan de Oro, the municipal 

authorities taxed unused open lands motivating property owners to make the sites 

available to poor city dwellers for crop production. 

In homegardens, important information flows through informal channels whereas 

the allotment gardens have enhanced information access through associations and 

extension support. This includes information on pest abatement strategies, 

management, and technologies. Effective systems combine the delivery and 

scientific advantages of extension with the firsthand user knowledge of gardeners in 

a reciprocal fashion. 

Social interaction is higher in allotment gardens because of the joint activities of 

different families in close spatial proximity. Social interaction is even higher if there 

is a corporate identity like membership in an allotment garden association. Exchange 

of information, joint activities, and family-based participatory learning is more

332 


likely to happen in allotment gardens while homegardening is in most cases a purely 

family-based activity. 

Figure 4. Gardens near Manila International Airport with densely populated compound 

at the top. 

In contrast to homegardens, allotment gardens enable users to learn democratic 

rules more efficiently because they have to resolve many problems within the 

association. Usually, problems arise regarding the use of land, equitable distribution 

of land and water, and joint community work. Where democratic rules and civil 

society associations differ significantly from customary practices, participants need 

guidance and support in acquiring the necessary skills. New cultural forms and 

institutions may be strongly resisted since they can upset existing political and social 

arrangements. In such situations participatory process planning can be combined 

with knowledge of the institutional context to direct the structure and functioning of 

the association.


7. INSTITUTIONAL CONTEXT 

333 

In Table 1, we present the institutional context surrounding homegardens and 

allotment gardens. While it does not survey the complete range of possible relevant 

institutions, it gives a good idea of the complex ways that homegardens and 

allotment gardens are differentially affected. A more complete accounting of the 

context would classify institutions along at least two important dimensions: formal 

versus informal and traditional versus modern. Importantly, these two dimensions 

are neither collinear nor orthogonal. The utility of such a classification goes beyond 

the simple question of support or lack of support for one or the other type of 

gardening. The exercise points to the linkages between urban gardening and other 

social problems in the community, fostering the possibility for constituency building 

through integrated problem-solving. Nonetheless, even without such synergistic 

system gains, elaboration of the institutional context provides valuable insights into 

the gardening model presented and into the processes by which urban gardening – 

whether homegardening or allotment gardening – is facilitated or hindered. 

8. RECOMMENDATIONS 

• Allotment gardens need to be institutionalized. Ideally, they should be part of the 

concept of urbanization wherein land is specifically set aside for such activities 

in the planning process. 

• Lobbying or public advocacy is required to support both garden systems. 

However, the nature of such lobbying efforts differs significantly between the 

two systems. Homegardens need public advocacy and extension services while 

allotment gardens additionally require political intervention to secure land, 

organize access, and support development. 

• Allotment gardens should be developed as a package of services, including for 

example extension outreach, community and infrastructure building, delivery of 

health care, etc. Allotment gardens will be more protected and access to them 

will be better coordinated in such a configuration. For example, allotment 

gardens are well adapted to periurban environments when authorities are willing 

to regularize spontaneous occupations. Small- and medium-sized towns provide 

ideal conditions for such early intervention and land preservation when 

combined with proper waste management and extension services. 

• Housing design and planning for backyards should facilitate homegardens. 

• Provision of adequate water is a problem in many cities, and public water use 

often restricted. Solutions include urban rainwater harvesting and the use of grey 

water (i.e., non-septic household wastewater), but their effectiveness depends on 

developing cost-effective locally adapted designs/technologies. Other solutions 

such as modern irrigation techniques can be implemented rather cheaply in 

conjunction with the preceding but depend more on extension information 

services to be effective. 

• The urban-periurban continuum is not uniform. Participatory process planning 

should be framed within the components of the periurban-urban typology and 

their corresponding institutions.

Institutional 

Type of urban gardens 

(non-) 

Homegardens Allotment gardens 

supporting 

structures 1 Characteristics Impact Characteristics Impact 

Allotment 

garden 

associations 

Not applicable Not applicable Sometimes existent Supporting 

City councils Mostly neglecting Strongly inhibiting Some recognition Sometimes 

supporting 

Town Planning Generally not aware of the importance 

of homegardens 

Educational 

system 

Health 

Authorities 

Table 1. Institutional contexts of urban homegardens and allotment gardens. 

Strongly inhibiting Generally not aware of the importance 

of urban food production 

Strongly 

inhibiting 

Ignorant in many countries Little support Ignorant in many countries Little support 

Generally not involved and not aware Inhibiting but 

probably no effect 

Generally not involved Inhibiting 

334 

A.W. DRESCHER ET AL .

Water 

authorities 

Prohibiting the use of public water for 

gardening without promoting 

alternatives (e.g. wells, rain water 

harvesting, grey water use) 

Inhibiting Prohibiting use of public water for 

gardening without promoting 

alternatives (e.g. wells, rain water 

harvesting, grey water use) 

Inhibiting 

Credits Generally not available Inhibiting Generally not available Inhibiting 

Extension 

Services 

Generally not available, rarely in the 

context of development projects 

Markets Mostly bartering with neighbors; 

production too small 

Markets and 

market 

information 

Direct marketing, bartering, household 

networks 

No market information system in place 

Inhibiting Generally not available, sometimes in 

the context of development projects 

Supporting social 

interaction 

Sometimes 

supporting 

Sometimes 

inhibiting 

Access to markets difficult, direct 

marketing possible but not supported 

Direct marketing, bartering, 

household networks 

No market information system in 

place 

Inhibiting 

Inhibiting 

Sometimes 

supporting 

Sometimes 

inhibiting 

1 By non-supporting structures we mean those institutions and bodies, which do not have policies on mechanisms to directly support gardens, 

and may even impede their development. Supporting structures in contrast clearly support the development of gardens. 

URBAN U HOMEGARDENS AND ALLOTMENT GARDENS GG 

335

336 


• Extension strategies should be elaborated in close cooperation among all 

stakeholders, especially smallholders and extension officers. 

• Strong advocacy for the multifunctional role of trees in both gardens and the 

broader urban context is required. Integration of this concept into extension 

programs aimed at both homegardens and allotment gardens is needed. 

ENDNOTES 

1. Velez C. 1997. Homegardening as a strategy for food and nutrition security: a 

case study of selected households in Lantapan, Bukidnon. MSc paper in 

Applied Nutrition, submitted to the Faculty of the Graduate School, University 

of the Philippines, Los Baños. 

2. Urbina C.D., Miso A.U., Holmer R.J. 2005. Socio-economic impact of the 

allotment garden project in Cagayan de Oro. Paper presented at the 6th 

PUDSEA Network Conference. July 11-15, 2005, Cagayan de Oro, Philippines. 

3. Segne J.B., Salcedo J.M. and Guiral H. 2004. Implementation of an integrated 

solid waste management system in two Sitios of Cagayan de Oro. Proceedings 

of the 16th NOMCARRD Regional symposium on research and development 

highlights, August 5-6, 2004, Central Mindanao University, Musuan, Bukidnon, 

Philippines (in print). 

4. Saverino J.L. 1995. ‘Domani Ci Zappa`’: Italian immigration and ethnicity in 

Pennsylvania. In: Pennsylvania folk life. 45 (Autumn): pp. 2-22. 

5. Abalimi, pers. comm., 2000; see also in Motion Magazine 2002, interviews 

related to the United Nations World Summit on Sustainable Development in 

Johannesburg, South Africa, August 26 - September 4, 2002. 

6. Food and Agriculture Organization (FAO) 2001. Proceedings of the expert 

consultation on urban horticulture in Southern Africa. Stellenbosch, South Africa. 

REFERENCES 

Agbayani A.L.P., Holmer R.J., Potutan G.E. and Schnitzler W.H. 2001. Quality and quantity 

requirements for vegetables by private households, vendors and institutional users in a 

Philippine urban setting. Urban Agr Mag 5: 56 – 57, Leusden. 

Arias G. 2000. Análisis de las políticas publicas en la agricultura urbana caso Texcoco, México. 

Dirección de Desarrollo Rural, Ayuntamiento de Texcoco, México, 20p. Programa de Gestão 

Urbana para a América Latina e o Caribe – PGU-ALC (CNUAH-HABITAT/PNUD) 

http://www.ipes.org/au /estudioc/texcoco.pdf (last accessed: January 2006). 

Boncodin R., Prain G. and Campilan D. 2000. Dynamics in tropical homegardens. Urban Agr 

Mag 1: 19 – 20, Leusden. 

Chambers R. 1989. Editorial: introduction: vulnerability, coping, and policy. IDS Bull 20(2): 1 – 7. 

Christanty L. 1990. Home gardens in tropical Asia with special reference to Indonesia. In: 

Landauer K. and Brazil M. (eds), 1990. Tropical home gardens, pp. 9 – 20. United 

Nations University Press, Tokyo. 

Cruz M.C. and Medina R.S. (eds). 2003. Agriculture in the city. IDRC, Ottawa, 244p. 

Drescher A.W. 1996. Management strategies in African homegardens and the need for new 

extension approaches. In: Heidhues F. and Fadani A. (eds), Food security and 

innovations: Successes and lessons learned, pp 231 – 246. Peter Lang, Frankfurt.


337 

Drescher A.W. 1998. Hausgärten in Afrikanischen Räumen – Bewirtschaftung nachhaltiger 

Produktionssysteme und Strategien der Ernährungssicherung in Sambia und Simbabwe. 

Sozioökonomische Prozesse in Asien und Afrika, 4. Centaurus, Pfaffenweiler, 290p. 

Drescher A.W. 2001. The integration of urban agriculture into urban planning – An analysis of the 

current status and constraints. In: Annotated bibliography on urban agriculture, pp 343 – 357. 

ETC Urban Agriculture Programme and Swedish International development Agency (SIDA), 

Leusden. http://www.ruaf.org/bibliography/annotated/014.pdf (last accessed: January 2006). 

Drescher A.W. and Iaquinta D.L. 2003. Urbanization – linking development across the changing 

landscape. Economic and social department publication (ESAC), FAO, Rome, 121p. http:// 

www.fao.org/fcit/docs/sofa_01.pdf (02.pdf – 03.pdf) (last accessed: January 2006). 

Drescher A.W., Hagmann J. and Chuma E. 1999. Home gardens - a neglected potential for 

food security and sustainable land management in the communal lands of Zimbabwe. In: 

Der Tropenlandwirt, 2/1999, pp 163–180. Kassel-Witzenhausen. 

Food and Agriculture Organization (FAO) 1998. Food security and community nutrition, No. 

22, Food and Agriculture Organization of the UN, Rome, 72p. 

Food and Agriculture Organization (FAO) 2001. FAO warns of increasing malnutrition 

among urban poor. FAO Press release 01/36. 1p. http://www.fao.org/WAICENT/OIS/ 

PRESS_NE/PRESSENG /2001/pren0136.htm (last accessed: January 2006). 

Food and Agriculture Organization (FAO) 2002. Declaration of the World Food Summit: five 

years later. In: Report of the World Food Summit: five years later. FAO, Rome, 

http://www.fao.org/DOCREP/MEETING/005/Y7106E/Y7106E09.htm (last accessed: January 

2006). 

Frankenberger T.R. and McCaston M.K. 1998. The household livelihood security concept. In: 

FAO 1998. Food Nutr Agr 22: 30-35 http://www.fao.org/docrep/X0051T/X0051T00.htm 

(last accessed: January 2006). 

Guanzon Y.B. and Holmer R.J. 2003. Basic cultural management practices for vegetable 

production in urban areas of the Philippines. Urban Agr Mag 10: 14 – 15, Leusden. 

Hetterschijt T. 2004. Our daily realities: Urban organic homegardens in Lima, Peru. Urban 

Agr Mag (Gender and Urban Agriculture, May 2005) 12: 10-11, Leusden. 

Holmer R.J., Clavejo M.T., Dongus S. and Drescher A.W. 2003. Allotment gardens for 

Philippine cities. Urban Agr Mag 11: 29 – 31, Leusden. 

Hoogerbrugge I. and Fresco L.O. 1993. Homegarden systems: Agricultural characteristics and 

challenges. Gatekeeper, No. 39, International Institute for Environment and Development, 

London, 23p. 

Iaquinta D.L. and Drescher A.W. 2000. Defining periurban: towards guidelines for 

understanding rural-urban linkages and their connection to institutional contexts, pp 8 – 

27. Land Reform 2000/2, FAO, Rome. 

Iaquinta D.L. and Drescher A.W. 2001. More than the spatial fringe: An application of the 

periurban typology to planning and management of natural resources in the periurban, 

DPU International Conference, “Rural-urban encounters: Managing the environment of 

the periurban interface”, College of London, 9 – 10 November 2001. http://www.ucl. 

ac.uk/dpu/pui/events/EPM_conf_ abstracts.htm# DefiningthePeri-Urban (last accessed: 

January 2006). 

Iaquinta D.L. and Drescher A.W. 2002. Food security in cities – A new challenge to 

development. In: Brebbia C.A., Matrin-Duque J.F., and Wadhwa L.C. (eds), The 

sustainable city II – Urban regeneration and sustainability. Advances in Architecture, pp 

983 – 994. Wessex Institute of Technology, WIT Press, Wessex. 

Iaquinta D.L. and Drescher A.W. 2005. Defining periurban: A framework for transportation 

planning in India, periurban workshop, Leeds (April 11 – 12, 2005). https://www.periurban. 

org/pub/hom / me/events/0504_leeds/documents/PRESENTATION- DefiningPeriurban-Iaquinta. 

pdf (last accessed: January 2006).

338 


Innenministerium B-W. 2004. Climate booklet for urban development references for zoning 

and planning. Amt für Umweltschutz, http://www.staedtebauliche-klimafibel.de/ 

Climate_Booklet/index-2.htm (last accessed: January 2006). 

Jacobi P., Drescher A.W. and Amend J. 2000. Urban agriculture: Justification and planning 

guidelines. GTZ, Eschborn, 70p, http://www.cityfarmer.org/uajustification.html (last 

accessed: January 2006). 

Jarlöv L. 2000. Urban Agriculture in South Africa. In: Hoffmann H. and Mathey K. (eds), 

Urban agriculture and horticulture: the linkage with urban planning. International 

Symposium (July 2000), Berlin (on cd-rom). http://www.ruaf.org/conference/ 

info_market/econf_papers/14jarlov.doc (last accessed: January 2006). 

Kasch G. 2001. Deutsches Kleingärtnermuseum in Leipzig: Deutschlands Kleingärtner vom 19. 

zum 21. Jahrhundert. Band 4, Sächsische Landesstelle für Museumswesen, Chemnitz, 128p. 


61/62: 135 – 152. 

Landeshauptstadt D. 1998. Umweltbericht. Stadtklima von Dresden. Umweltamt, 39p. http:// 

www. dresden.de/pdf /infoblaetter/umweltbericht_text.pdf (last t accessed: Janua ary 

2006). 

Landon-Lane C. 2004. Livelihoods grow in gardens: Diversifying rural incomes through 

homegardens. FAO Diversification booklet 2. Agricultural Support Systems Division, FAO, 

Rome. 58p. 

MacNair E. 2002. The garden city handbook: How to create and protect community gardens 

in Greater Victoria. Polis Project on Ecological Governance. University of Victoria, 

Victoria BC, 34p. http:// www.polisproject.org/polis2/PDFs/the%20garden%20city%20 

handbook.pdf (last t accessed: January 2006). 

Mougeot L.J.A. (ed.). 2005. AGROPOLIS: The social, political, and environmental 

dimensions of urban agriculture. CRDI, Earthscan. 308p. 

Pinderhughes R. 2004. Alternative urban futures: Planning for sustainable development in 

cities throughout the world. Rowman & Littlefield Pub Inc. Lanham, MD, 271p. 

Prosterman R. and Mitchell R. 2002. Concept for land reform on Java. Paper presented at the 

seminar: Rethinking land reform in Indonesia, Jakarta (8 May 2002). On file with the 

Rural Development Institute, Seattle. http://www.rdiland.org/PDF/RDI_LandReform 

OnJava.pdf (last accessed: February 2006). 

Ranasinghe T.T. 2005. Family business gardens: Agricultural options in remodelling and 

modernising tsunami devastated townships in Sri Lanka. Cityfarmer, 9p. 

www.cityfarmer.org/subsouthasia.html (last accessed: 27 January 2006). 

Swift J. 1989. Why are rural people vulnerable to famine? In: Chambers R. (ed.), IDS Bull 

20(2): 8 – 15. 

United Nations Centre for Human Settlements (UNCHS) 2001. The state of the world’s cities 

report 2001.United Nations Centre for Human Settlements (UNCHS), Nairobi, 126p. 

Vasey D.E. 1985. Household gardens and their niche in Port Moresby, Papua New Guinea. 

Food Nutr. Bull 7(3): 37 – 52. 

Vélez-Guerra A. 2004. Multiple means of access to land for urban agriculture: A case study 

of farmers’ groups in Bamako, Mali. Cities feeding people report series, December 2004, 

International Development Research Centre (IDRC), Ottawa, 88p. 

Yamada M. and Osaqui H.M.L. 2006. The role of homegarden for agroforestry development: 

lessons from a Japanese-Brazilian settlement in the Amazon. In: Kumar B.M. and Nair 



Zalilah M.S. and Khor G.L. 2004. Indicators and nutritional outcomes of household food 

insecurity among a sample of rural Malaysian women. Pakistan J Nutr 3: 50 – 55.

CHAPTER 19 

ARE TROPICAL HOMEGARDENS 

SUSTAINABLE? SOME EVIDENCE FROM 

CENTRAL SULAWESI, INDONESIA 

K. KEHLENBECK AND B.L. MAASS* 

Institute for Crop and Animal Production in the Tropics, Georg-August-University 

Göttingen, Grisebachstr. 6, D-37077 Göttingen, Germany; *E-mail: 

 

Keywords: Biodiversity, In situ conservation, Soil fertility, Sustainability indicators, 

Vegetation dynamics. 

Abstract. Homegardens are regarded as sustainable agricultural production systems, although 

support for this statement by quantitative data has been rare. Out of the suggested 

indicators/descriptors for assessing sustainability, plant diversity has been frequently studied. 

However, species diversity is not static: it varies with time and according to ecological and 

socioeconomic factors and/or characteristics of the gardens and gardeners. In order to evaluate 

sustainability of the homegarden system, we assessed soil fertility parameters and changes in 

diversity of useful plants over time during 2001 – 2004 in 30 homegardens from three villages 

adjacent to the Lore Lindu National Park in Central Sulawesi, Indonesia. Soil carbon (C) and 

nitrogen (N) contents decreased over time. In large gardens with different production zones, soil 

of vegetable zones contained less C and N than that of cacao (Theobroma cacao) zones. 

Richness of useful plant species was high and increased over time, from 149 species in 2001 to 

168 in 2003. Species composition of homegardens from one village, mainly inhabited by 

migrants, contrasted strongly with those from the other two, inhabited by native farmers. 

Diversity of useful plants was lower in the migrant village, where soil fertility was low, too. 

Plant diversity appeared to be influenced to varying extent by a combination of factors such as 

garden size/age, soil fertility, ethnicity and age of gardener, and market access. The surveyed 

homegardens did not seem to be managed appropriately to ensure sustainability in terms of soil 

fertility although they had a high diversity of useful plants. 


Tropical homegardens are generally regarded as sustainable production systems 

(Christanty, 1990; Landauer and Brazil, 1990; Soemarwoto and Conway, 1991; 

339 




340 

K. KEHLENBECK K AND B.L. MAASS M 

Torquebiau, 1992; Abdoellah et al., 2001; Kumar and Nair, 2004). However, 

quantitative support for this statement is mostly lacking, particularly because of the 

difficulties in measuring sustainability (Kumar and Nair, 2004). Therefore, 

researchers rely on indirect evidences using certain sustainability descriptors and/or 

indicators (Torquebiau, 1992; Huxley, 1999). 

Among the available indicators, perhaps the criterion most used in homegarden 

research is biodiversity, particularly plant species diversity. The wide spectrum of 

useful plants creates a multilayered vegetation structure in homegardens, which is 

responsible for many benefits and advantages of the system. This diversity results in 

favorable microclimate, reduced risk of pests and diseases, efficient use of 

resources, year-round availability of products, and soil fertility maintenance. Thus, 

plant diversity is considered as contributing substantially to the sustainability of the 

system (Soemarwoto and Conway, 1991; Torquebiau, 1992). 

Because of their diversity, homegardens are also regarded as an ideal production 

system for in situ conservation of plant genetic resources (Watson and Eyzaguirre, 

2002), crucial for long-term sustainability. However, crop diversity is influenced by 

different factors such as size and age of homegardens or age of gardeners 

(Abdoellah et al., 2001; Gutiérrez et al., 2004). Besides, environmental and 

socioeconomic characteristics are known to influence homegarden diversity 

(Michon and Mary, 1994; Wezel and Bender, 2003; Gutiérrez et al., 2004). 

Nevertheless, the suitability of biodiversity as a sustainability indicator needs to be 

critically examined because there is no threshold value for an ideal number of 

species in a sustainable system. In addition, diversity seems to be highly variable 

over time, and the homegarden research so far has neglected to quantify such 

changes. 

Another sustainability indicator generally accepted is soil fertility (Torquebiau, 

1992; Huxley, 1999; Kumar and Nair, 2004). In homegardens, soil fertility is said to 

be maintained due to the closed nutrient cycling and low nutrient-export through 

harvested products (Gajaseni and Gajaseni, 1999; Kumar and Nair, 2004). Dense 

layers of litter and undergrowth are supposed to prevent or at least reduce soil 

erosion in homegardens (Karyono, 1990; Soemarwoto and Conway, 1991). 

Investigation of soil fertility parameters is common in homegarden research (Jensen, 

1993; Gajaseni and Gajaseni, 1999), whereas soil erosion has rarely been assessed 

(Torquebiau, 1992). Usually, statements on sustainable soil fertility management in 

homegardens are supported only by a single ‘snapshot’ of the status quo without any 

further consideration on soil fertility variation over space and time. The role of 

different management practices leading to this variation in the long-term is not 

sufficiently investigated. 

In association with the multidisciplinary German-Indonesian collaborative 

research program STORMA (Stability of Rainforest Margins in Indonesia, SFB 

552), this study aimed at assessing the sustainability of selected homegardens on the 

island of Sulawesi with the help of selected sustainability indicators. A first 

assessment from a comprehensive dataset is presented here, focusing on aspects of: 

• Stability/dynamics in diversity of useful plants over time 

• Changes in soil fertility over time

341 

• Specific influences of selected factors on diversity of useful plants 

The ecological indicators ‘diversity of useful plants’ and ‘soil fertility’ were 

chosen because data from a previous study of the same homegardens were available 

(Kehlenbeck and Maass, 2004) and both indicators are essential for assured 

productivity of the system. Besides ecological indicators, social and economic 

indicators (e.g., labor requirement, cash input and biophysical output) as suggested 

by Torquebiau (1992) and Kumar and Nair (2004) have been assessed under the 

overall project, but those results will be presented elsewhere. 

2.1. Study area 

ARE TROPICAL T HOMEGARDENS SUSTAINABLE? 


The study was conducted from March to November 2001 and from June 2003 to 

June 2004 in the Napu Valley (1°23’ to 37’S, 120°18’ to 20’E), located on the 

eastern margins of the Lore Lindu National Park in Central Sulawesi (Indonesia), 

about 100 km south of the city of Palu. Elevation is around 1100 m above sea level; 

annual precipitation is about 2000 mm with a mean temperature of 21°C. Natural 

vegetation is classified as lower montane rainforest (Whitten et al., 1987); soils are 

mostly Cambisols (FAO; USDA: Tropepts, Inceptisols) and Fluvisols (Fluvents, 

Entisols). 

The initially low human population density has been increasing in the region, 

especially since the 1980s, due to migration. Most inhabitants are farmers, and offfarm 

employment opportunities are scarce. Agricultural production is mainly based 

on paddy rice (Oryza sativa) production for subsistence, agroforestry with cacao 

(Theobroma cacao) and coffee (Coffea spp.) as cash crops, and rain-fed annual 

crops (Kehlenbeck and Maass, 2004). Large areas of the Napu Valley are under 

fallow or degraded grasslands. 

Table 1. Characteristics of three villages studied in the Napu Valley, Central Sulawesi, 

Indonesia. 

Parameters Wuasa Rompo Siliwanga 

Year of foundation 1892 1915 1992 

Inhabitants (no.) 2600 (2003) 400 (2004) 600 (2004) 

Ethnicity mixed >75% indigenous >75% migrants 

Distance to paved road 0 km 5 km 0 km 

Market access good poor medium 

Source: Zeller et al. (2001) and Kehlenbeck (unpublished data). 

For this research, three villages, which differed in their market access and origin 

of inhabitants, were chosen (Table 1). Wuasa is the administrative center of the 

Napu Valley with a junior and senior high school, a small hospital, many shops and 

offices as well as a market place. Rompo is a small village surrounded by forest,

342 

accessible by a dirt road. Siliwanga was founded only recently for settling migrant 

families, mostly from Bali, in the context of the transmigration program of the 

Indonesian government (Mayor of Siliwanga, pers. comm., 2001). For convenience, 

the three villages were labeled as ‘market village’ (Wuasa), ‘forest village’ 

(Rompo), and ‘migrant village’ (Siliwanga). 


Ten households with homegardens were randomly selected from each village. 

Information about local knowledge and management of the same homegardens was 

gathered in 2001 and 2003/2004, except for one garden in the migrant village that 

was abandoned in 2002. Gardeners were individually interviewed using an 

unstructured questionnaire with questions on age and functions of the homegarden, 

inputs and outputs, and the use of homegarden products, among others. Data 

concerning household characteristics, such as age, formal education, ethnic group, or 

occupation of the household members were also gathered through interviews, partly 

within larger surveys of the STORMA project. 

Homegarden size was measured, excluding the area occupied by the house. 

Complete inventories were carried out in 2001 (July – October) and 2003 (July – 

August) to assess number of species and abundance of crops and ornamentals. In 

this study, the term ‘crops’ is applied to all useful plant species, including planted 

and spontaneously occurring except the ornamentals. Presence of weeds, defined as 

undesired plants from the gardener’s view, was documented but not quantified. 

Plants were recorded with local and/or scientific names. Crop species were classified 

into different use categories (Kehlenbeck and Maass, 2004). 

In 2001, 20 soil samples per garden were randomly collected from 0 – 15 cm 

depth and mixed, except for four large gardens, where soil was sampled separately 

according to production zone (vegetables, coffee/cacao, or fruit trees). In 2003/2004, 

five soil samples per garden were randomly collected at 0 – 15 cm depth and mixed, 

2 

distinct production zones, five samples per zone were collected and mixed. Due to 

these different sampling strategies, soil fertility change over time was analyzed only 

in a subgroup of homegardens with comparable soil sampling in both years, i.e., 

gardens with one mixed vegetation zone only (n = 10) as well as large gardens 

(n = 4) already sampled by zones in 2001. Total C and N were quantified by C/N- 

Autoanalyser and pH with an electrode (soil: water ratio, 1:2.5). Bulk density was 

determined in 2003 only by assessing the dry weight of soil samples with known 

field volume. 



if the garden was small (< 350 m 

) or planted uniformly. In large gardens with 

Species density (no. of spp./100 m 2 ), Shannon index (H’), and Pielou evenness index 

(E = H’/Hmax H ) were calculated for every garden (Magurran, 1988). To compare 

floristic similarity between the three villages, Sørensen’s coefficient was computed 

(Magurran, 1988). Data were analyzed using the statistical package SPSS 11.0.

343 

Differences between means were determined by Mann-Whitney U-Test U or Kruskal 

Wallis H-Test. Changes over time as well as spatial differences of soil fertility 

parameters between production zones within one garden were analyzed as ‘paired 

samples’ using the Wilcoxon-test. Influence of relevant factors on crop diversity was 

determined by correlation analysis (Spearman). 

3. RESULTS 

In 2003/2004, size of the homegardens ranged from 240 to 2400 m 2 , and they had 

been established 4 to 41 years ago. Homegardens in the migrant village were 

significantly younger than those in the market village, and were managed by 

younger families (Table 2). In all three villages, homegarden size, farm size, and 

homegarden proportion in relation to the overall farm size were highly variable. 

Compared to 2001 (Kehlenbeck and Maass, 2004), farm size increased significantly 

only in the migrant village due to purchase or clearing of additional land. Therefore, 

the proportion of the homegarden in relation to overall farm size as well as its 

importance for staple food production recently decreased in the migrant village. 

Table 2. Characteristics of households and homegardens surveyed in three villages of the 

Napu Valley, Central Sulawesi, Indonesia, 2003. 

Parameters Market village 

(Wuasa) 

Forest village 

(Rompo) 

Migrant village 

(Siliwanga) 

Median Range Median Range Median Range 

Age of household 

head (years) 

55 a 

34 – 69 

ab 

50 25 – 89 

b 

35 30 – 50 

Gardener’s age 

(years) 

48 a 

32 – 67 

a 

40 20 – 60 

a 

34 28 – 50 

Household members 8 

(no.) 

a 

3 – 14 

a 

5 1– 11 

a 

4 3 – 6 

Farm size (ha) 2.6 a 

0.9 – 11.1 

a 

5.9 1.7 – 11.5 

a 

3.1 1.5 – 5.5 

Homegarden size 

(m 2 ) 

720 a 

236 – 1134 

a 

610 

a 

287 – 1450 820 471 – 2383 

Homegarden size/ 

farm size (%) 

2.2 a 

0.5 – 10.0 

a 

1.2 0.5 – 4.3 

a 

2.3 1.2 – 11.9 

Age of homegarden 

(years) 

28 a 

14 – 37 

ab 

16 4 – 41 

b 

10 6 – 11 

Medians in a row followed by different superscripts are significantly different at p≤0.05. 

3.1. Crop diversity and its changes 


Crop species richness was high and increased markedly over time both per village 

and per garden (Fig. 1). In the three villages, a combined total of 149 and 168 crop 

species were identified in 2001 and 2003 respectively. Distribution of crops into 

different use categories was comparable in different sampling years (Kehlenbeck

344 


and Maass, 2004). Out of the 168 crop species grown in homegardens, about 35 

were wild species (mainly used as fuelwood/timber or medicine) and about 44 were 

classified as underutilized species (mainly used as vegetable). In addition to the 168 

crop species, 99 ornamental and 62 weed species were found in the homegardens 

surveyed in 2003. 

Figure 1. Changes of total and mean crop species richness in homegardens per village, 

studied in the Napu Valley, Central Sulawesi, Indonesia, 2001 (n = 30) and 2003 (n = 29). 

Figure 2. Changes of mean species density in homegardens of three villages studied in the 

Napu Valley, Central Sulawesi, Indonesia, 2001 and 2003. 

Mean species density increased significantly over time, particularly in the market 

village (Fig. 2). However, in the migrant village, species density continued to be 

significantly lower than that in the forest village in 2001 or the market village in 

2003. Changes in Shannon diversity and Pielou evenness indices were not so clear 

apart from the migrant village, where both indices were significantly higher in 2003 

than in 2001 (Fig. 3). In the market village, Shannon and Pielou indices showed a 

slight tendency to decrease because in 2003 some gardeners started to grow spring

345 

onion (Allium ( fistulosum) 

for sale in relatively large plots, which dramatically 

reduced the indices. For example, in one homegarden an area of about 190 m 2 out of 

865 m 2 was planted with a mixture of vegetables and spices in 2001, but only spring 

onion during 2003. This resulted in a decrease of Shannon index from 2.1 to 1.2 and 

Pielou index from 0.59 to 0.31. However, the total number of crop species increased 

in this particular garden from 35 to 47 during the same period. 

Figure 3. Changes of mean Shannon diversity and mean Pielou evenness indices in 

homegardens of three villages studied in the Napu Valley, Central Sulawesi, Indonesia, 2001 

and 2003. 

Crop species composition was clearly different among the three villages in both 

years. Sørensen’s coefficients showed a higher similarity between the market and 

the forest villages (0.71) than between these two and the migrant village (market vs. 

migrant village: 0.63; forest vs. migrant village: 0.58). Compared to 2001 

(Kehlenbeck and Maass, 2004), Sørensen’s coefficient decreased slightly in all 

cases. The species common to all three villages remained rather stable over time, 

while obvious changes occurred in those crop species unique for one village and, 

hence, not found in the other two. Particularly in the market village, 22 unique crops 

were recorded in 2003 instead of 15 in 2001. 

3.2. Soil properties 


In large gardens with different production zones, soil fertility was obviously 

different among these zones. Across all 12 gardens where distinct vegetable and 

cacao zones existed, soil of the vegetable zone contained significantly less N and C 

than soil of the adjacent cacao zone (Table 3). Soil pH did not differ among 

vegetable and cacao production zones. Bulk density was significantly higher in 

vegetable zones than in adjacent cacao zones. 

Because of these large differences in soil fertility between production zones 

within one single homegarden, it did not appear meaningful to compare mean values 

of the homegardens investigated in the three villages. Instead, soil fertility of cacao

346 

zones only was compared among the villages as this particular zone existed in most 

of the homegardens (n = 16), apart from the very small ones. In five homegardens, 

the cacao zone was even the only obvious production zone. 

Table 3. Properties of homegarden topsoil (0–15 cm) from different production zones in three 

villages of the Napu Valley, Central Sulawesi, Indonesia, 2003/2004. 

Soil attributes Vegetable zone Cacao zone 

Mean Range Mean Range 

Ctotal (%) 1.64 b 

0.93 – 2.96 2.31 a 

Ntotal (%) 0.13 

1.40 – 3.42 

b 

0.06 – 0.21 0.18 a 

pH (H2O) 5.87 

0.10 – 0.27 

a 

4.65 – 6.88 5.62 a 

Bulk density (g/cm 

5.24 – 5.83 

3 ) 1.17 a 

0.90 – 1.48 1.03 b 

0.77 – 1.17 

Means in a row followed by different superscripts are significantly different at p ≤ 0.05. 

Table 4. Properties of topsoil (0–15 cm) from cacao production zones of 21 homegardens in 

three villages of the Napu Valley, Central Sulawesi, Indonesia, 2003/2004. 

Soil attributes Market village 

(Wuasa; n = 8) 


Forest village 

(Rompo; n = 7) 

Migrant village 

(Siliwanga; n = 6) 

Mean Range Mean Range Mean Range 

C total (%) 2.02 a 1.43 – 2.95 2.31 a 1.40 – 3.21 2.83 a N total (%) 0.17 

2.32 – 3.42 

a 0.12 – 0.21 0.19 a 0.10 – 0.27 0.19 a pH (H2O) 5.65 

0.15 – 0.24 

a 5.24 – 5.84 5.48 a 5.16 – 5.80 5.63 a Bulk density 

(g/cm 

5.21 – 5.85 

3 ) 

1.10 ab 0.92 – 1.24 0.98 b 0.77 – 1.16 1.16 a 1.08 – 1.24 


Table 5. Changes of soil fertility parameters of topsoil (0–15 cm) from 14 homegardens in 

three villages of the Napu Valley, Central Sulawesi, Indonesia, 2001 and 2003. 

Soil attributes 2001 2003 

Mean Range Mean Range 

C total (%) 2.35 a 1.20 – 3.58 2.12 b 0.92 – 3.21 

N total (%) 0.19 a 0.11 – 0.29 0.16 b 0.07 – 0.27 

pH (H2O) 5.72 a 4.70 – 6.50 5.75 a 4.88 – 6.75 


Soil C and N contents of cacao production zones were highly variable in all 

villages, although these values were slightly lower in the market village (Table 4). 

Soil pH was relatively similar in all villages. Only soil bulk density was significantly 

higher in the migrant village and lower in the forest village. When comparing soil

fertility over time, C and N contents decreased significantly from 2001 to 2003, 

whereas soil pH did not change (Table 5). 

3.3. Influence of selected factors on crop diversity 

347 

To detect factors that possibly influence crop diversity, correlations between crop 

diversity parameters and several variables describing characteristics of homegardens 

(e.g., age, size, soil fertility parameters), the gardener (e.g., age, education), or 

socioeconomics (e.g., wealth status of household, size of paddy rice fields, market 

access) were analyzed (Table 6). Socioeconomic characteristics of the gardeners or 

households did not play an important role in determining crop diversity. 

Table 6. Spearman correlation coefficients between crop diversity parameters and different 

characteristics of homegardens, gardeners, and households in three villages of the Napu 

Valley, Central Sulawesi, Indonesia, 2001 and 2003. 

Parameters Species 

richness 


Species density Shannon 

index 

Pielou index 

2001 2003 2001 2003 2001 2003 2001 2003 

Garden age 0.45* 0.41* ns ns ns ns ns ns 

Garden size 0.45* 0.52** –0.83*** –0.81*** ns ns ns ns 

Soil pH value ns –0.40* 0.43* 0.40* ns ns ns ns 

Soil N content ns ns –0.50** –0.38* ns ns ns ns 

Soil C content ns ns –0.58** ns –0.43* ns –0.42* ns 

Gardener’s age 0.47** 0.49** ns ns 0.53** ns 0.41* ns 

Gardener’s 

education 

ns ns ns ns ns ns ns ns 

HH members 

(no.) 


Wealth status 

of HH 


Size of HH’s 

rice fields 


Garden 

size/farm size 

ns ns –0.68*** –0.43* ns ns –0.47* ns 

Market access ns ns ns ns ns ns ns ns 

*: p ≤ 0.05; **: p ≤ 0.01; ***: p ≤ 0.001; ns = not significant; HH = household. 

Crop diversity was mainly influenced by the gardener’s age and by variables 

describing homegarden characteristics such as size, age, or soil fertility parameters. 

In large and old homegardens, higher crop species richness could be expected than 

in small and young homegardens. Furthermore, the older the gardener, the higher 

was the species richness, diversity, and evenness. However, the influence of all 

variables was rather weak, particularly of soil parameters. Within the tested

348 

socioeconomic variables, only the ratio of homegarden-size to farm-size showed a 

weak but significant negative influence on evenness index. Ethnicity of the gardener 

probably was linked with crop diversity because mean species richness and density 

were significantly higher in gardens of local families than that of the migrants. No 

differences in crop diversity were observed by grouping gardeners into male and 

female subgroups. However, direct influence of these two nominal variables on crop 

diversity could not be assessed by the correlation analysis. 

4.1. Changes of crop diversity 


4. DISCUSSION 

In 2001 and 2003 total crop species richness as well as the mean per garden were 

rather high, but comparable to the data reported from other regions in Indonesia or 

even from the tropics as a whole (Kumar and Nair, 2004). Crop diversity in the 

homegardens surveyed was not only maintained, but even increased over time. 

Seasonal effects could not be made responsible for this because in both years species 

inventories were carried out in the same season. Partly, the increase in diversity can 

be explained by interventions of development projects. For instance, in all villages, 

seedlings of mandarin trees (Citrus reticulata) were provided to most gardeners in 

2002/2003. Another project promoted the cultivation of medicinal plants in 

homegardens at the same time. As a result, the Mayor of the market village pushed 

gardeners to grow these recommended plants. This led to an increase of medicinal 

species from a total of 16 in 2001 to 21 in 2003 and a mean per garden from 3.4 to 

5.4, respectively. However, in the other two villages the impact of these 

development projects on homegarden diversity seemed to be rather low. 

Additionally, research activities in 2001 (Kehlenbeck and Maass, 2004) have 

possibly stimulated interest of the gardeners in crop diversity. As a result, gardeners 

might have revived the networks of seed and plant exchange within their 

neighborhoods, and were more open for experimental cultivation of new crops. 

At the same time, gardeners stopped to grow some crop species (a mean of six 

species per garden). According to the gardeners, many of these species died during 

an unusual dry period in 2002. Another reason for decrease of diversity in 

homegardens could be that production became more market-oriented, as described 

by Soemarwoto and Conway (1991). However, in this study, the market-oriented 

production of spring onions in the market village resulted only in a slight decrease of 

Shannon and evenness indices but not richness or density of crop species. Besides, 

in 2004 it was observed that the gardeners already stopped growing spring onions 

for sale due to a decline in prices and problems with diseases. 

It can, therefore, be concluded that crop diversity in homegardens is very 

dynamic and every species inventory reflects only the diversity at the very instant of 

assessment. Thus, the temporal dynamics observed in this study might not reflect 

long-term trends. Nevertheless, the suitability of homegardens for in situ 

conservation of plant genetic resources needs to be critically revised based on these 

results. For this purpose, specific target groups of crops or even key species instead

349 

of the overall diversity should be emphasized m (e.g., Watson and Eyzaguirre, 2002). 

Furthermore, it is crucial to make gardeners active stakeholders of such conservation 

efforts by sharing both responsibility and benefits. Finally, crop diversity should not 

be used as the only sustainability indicator of this system because of its changes with 

time. 

4.2. Soil fertility 


According to Landon (1991), soil of vegetable and cacao zones surveyed had low to 

very low mean C and N contents, whereas mean pH values were classified as 

medium. Therefore, the current situation with limited N available in the soil most 

likely restricts the level of production, particularly for N-demanding vegetables. 

Considering the significant decrease in soil C and N contents over time, crop 

production may become more constrained in the near future, particularly in the 

market and forest villages, where C and N contents were already very low in many 

garden soils. Insufficient soil fertility management by the gardeners caused this 

alarming situation. For example, only about 30% of the gardeners used farmyard 

manure as a fertilizer, although it was available to all of them. Many gardeners 

removed weeds including their roots for burning or depositing in garbage pits 

instead of using them for compost preparation. Use of compost or mulch was 

virtually unknown, and industrial N fertilizer was available to only 15% of the 

gardeners, an overall situation that has not changed since 2001 (Kehlenbeck and 

Maass, 2004). Deterioration was accelerated also by the habit of gardeners to 

remove the litter layer by daily sweeping and regular burning. Typical reasons given 

by the gardeners for this practice were keeping away snakes and insects from the 

house. Sweeping and total weeding was carried out in all front gardens, in most 

vegetable and ornamental zones and in some cacao or fruit tree zones, which led to 

severe soil erosion (Fig. 4). 

In general, soil fertility is said to be maintained in homegardens in the long-term 

(Gajaseni and Gajaseni, 1999; Kumar and Nair, 2004). Only few reports 

(Soemarwoto, 1987; Soemarwoto and Conway, 1991; Hvoslef, 1994; Benjamin 

et al., 2001) stated problems of soil deterioration and erosion due to insufficient 

management practices similar to those identified here. Soil management in the 

present study, however, needs also to be seen in the context of changing traditional 

land use in the Napu Valley. The dominant shifting cultivation was replaced by 

permanent agriculture only about 10 to 30 years ago (Burkhard, 2002). Therefore, 

indigenous as well as newly arrived migrant farmers may not be familiar with 

appropriate sustainable land management practices. Negative environmental 

consequences have similarly been documented for other cases of resettlement, e.g., 

in Ethiopia (Wood, 1993) and Tanzania (Charnley, 1997). 

Spread of household waste materials in homegardens might cause a new problem 

affecting long-term soil fertility and, consequently, system productivity. This has 

never been mentioned in the homegarden literature. Due to lack of opportunities for 

waste disposal, many gardeners in the research area spread all garbage on the soil of 

the backyard, including non-biodegradable items such as glass and plastic bottles, 

tins, plastic bags, and old batteries. Mixed with organic wastes from the kitchen, this

350 


garbage formed the ‘litter’ layer in many backyard gardens. This practice will 

probably cause soil contamination; the spread of biodegradable waste on soil 

will, however, contribute to better nutrient cycling and reduced soil erosion. 

Figure 4. Example of soil erosion in the front yard of a homegarden in the forest village 

Rompo, Napu Valley, Central Sulawesi, Indonesia, 2004. The broken line indicates soil 

surface during planting of the ornamentals along the fence; the dotted line shows the present 

surface. 

To achieve sustainable soil fertility management in the study region, the existing 

extension service should not exclusively focus on paddy rice production but also on 

agroforestry systems (including homegardens) with their great significance for cash 

income generation (Maertens et al., 2002). Advantages of using compost, mulch, 

and farmyard manure should be explained. Growing N2-fixing cover crops ought to 

be promoted, not only in the homegardens, but also in other cropping systems. 

Besides, villagers should be enlightened about disadvantages and risks of soil and 

water contamination in order to preserve the resources on which they rely.

4.3. Factors influencing crop diversity 


351 

Within the major factors influencing crop diversity, garden size is one of the 

frequently analyzed. Among others, Abdoellah et al. (2001) and Gutiérrez et al. 

(2004) reported a positive relationship between garden size and crop species 

richness. Results from the present study (Table 6) showed a slightly positive, but 

non-linear relationship. In very large gardens, crop species richness tended to reach 

a plateau. On the other hand, the larger the garden the lower was crop species 

density because of more uniform planting patterns in very large gardens. A positive 

influence of garden age on species richness was also stated by Gutiérrez et al. 

(2004). In this study, however, garden age had a highly significant positive 

correlation with gardener’s age because, generally, young families establish a new 

homegarden, starting with a rather small set of crop species. 

Besides age and size of homegarden, soil fertility is another factor describing 

garden features, but its influence on crop diversity has not yet been studied in detail. 

Hodel et al. (1999) assumed an influence of soil factors on diversity without 

quantifying this. In forest gardens, Kaya et al. (2002) reported lower species 

diversity on marginal soils compared to soils that are more fertile. Many crop 

species, particularly vegetables and spices, do not give adequate yield under 

unfavorable soil conditions. Therefore, gardeners stop cultivating these species 

while switching to a reduced set of crops that can cope with low soil fertility. In the 

migrant village Siliwanga with its rather poor soil conditions, for example, acidtolerant 

species such as tea (Camellia sinensis), cassava (Manihot esculenta) and 

cashew ( (Anacardium occidentale) 

were found in many homegardens, whereas 

vegetable cultivation was rare (Kehlenbeck and Maass, 2004). However, influence 

of soil fertility parameters on crop diversity in this study must be seen in the context 

of the significant correlations between garden size and soil pH (negative) as well as 

between garden size and soil C and N contents (positive) that probably biased the 

results of analysis (see Table 6). 

Gardener’s age can influence crop diversity positively, possibly because, over 

the years, gardeners try to cultivate new crops while they continue to plant well-tried 

species (Gutiérrez et al., 2004). Besides, older gardeners often have more time for 

homegardening and are supported by their grown-up children. A higher timeallocation 

to homegardening leads to a higher diversity of useful plants (Hodel et al., 

1999; Gutiérrez et al., 2004). In the present study, however, the positive relationship 

between gardener’s age and plant diversity was rather weak. 

Within gardener’s characteristics, ethnicity also may be a factor explaining 

variation in crop diversity (Soemarwoto and Conway, 1991; Hodel et al., 1999). 

Contrary to the findings of Soemarwoto and Conway (1991), crop diversity in the 

present study was lower in homegardens of migrants as compared to locals. 

Admittedly, migrant gardeners brought various useful species from their home 

regions. Due to the unfavorable soil and climate conditions of the lands assigned to 

them, a large part of these plants did not establish. Another reason for the low crop 

diversity in the migrant village might be the socioeconomic status of the gardeners. 

After arrival, young migrant families focused strongly on paddy rice production, 

with a resulting shortage of labor for homegarden management. Furthermore, field

352 


crop failures and poor access to suitable agricultural land might have led to 

cultivation of additional staple crops in the migrants’ homegardens. The result of 

correlation analysis that a high portion of homegarden size to overall farm size was 

related with a low Pielou evenness index (Table 6) support this statement. A 

reduction in diversity of homegardens is known to be caused by a high proportion of 

staple food crops (Soemarwoto and Conway, 1991) as well as by labor shortage 

(Hodel et al., 1999; Gutiérrez et al., 2004). 

Among socioeconomic factors, the negative influence of market proximity and 

commercially oriented production on crop diversity has frequently been recorded 

(Christanty, 1990; Soemarwoto and Conway, 1991; Michon and Mary, 1994; 

Abdoellah et al., 2001; Gutiérrez et al., 2004). In the study area, however, this effect 

was only slightly recognized (Kehlenbeck and Maass, 2004). Nevertheless, there 

seemed to be a high risk of decreasing crop diversity with an associated loss of plant 

genetic resources, if production of cash crops such as spring onions were to be 

successful. In summary, our results suggest that diversity is not only influenced by 

clearly identifiable single factors but rather by a complex interaction among several 

factors studied and probably others. This interaction is not yet understood, and 

additional intrinsic characteristics of gardeners, such as individual preferences and 

practices might play an overriding role. Obviously, further research is needed for a 

better understanding of these interrelationships and the processes leading to them. 

This would help assessing the sustainability of the system as well as its suitability 

for in situ conservation of plant genetic resources. 


Although the homegardens surveyed maintained high crop diversity over time, their 

management at present in the study region was not conducive to sustainability in 

terms of soil fertility management. The set of the two common sustainability 

indicators chosen was found to be adequate only for a temporary assessment of 

homegardens. Nevertheless, an estimation of soil erosion as an additional indicator 

of sustainability should be considered, particularly where soil fertility monitoring is 

not practicable over time. Homegardens can play an important role in in situ 

conservation of plant genetic resources as long as gardeners participate in the whole 

process. 


Many thanks to all the people in Wuasa, Rompo and Siliwanga, especially to the 30 

respondent families, for their cooperation and ready help during the field survey. 

The support provided by members and staff of STORMA (SFB 552) in Bogor, Palu 

and Göttingen is also gratefully acknowledged, particularly the subproject 

‘Economic Analysis of Land Use Systems of Rural Households’ (A4) for making 

available their survey data set and subproject ‘Central Laboratory Unit’ (Z3) for 

carrying out soil analyses. The German Academic Exchange Service (DAAD) and 

German Science Foundation (DFG) provided financial support.


REFERENCES 

353 

Abdoellah O.S., Takeuchi K., Parikesit, Gunawan B. and Hadikusumah H.Y. 2001. Structure 

and function of homegarden: a revisited. Proc. Seminar ‘Toward harmonisation between 

development and environmental conservation in biological production’ (21-23 February 

2001), pp 167 – 185. University of Tokyo, Tokyo. 

Benjamin T.J., Montañez P.I., Jiménez J.J.M. and Gillespie A.R. 2001. Carbon, water and 

nutrient flux in Maya homegardens in the Yucatán peninsula of México. Agroforest Syst 

53: 103 – 111. 

Burkhard G. 2002. Natural resource management in Central Sulawesi: Past experience and 

future prospects. STORMA Discussion Paper Series No. 8 (online). Universities of 

Goettingen and Kassel, Germany, Institut Pertanian Bogor and Universitas Tadulako, 

Indonesia. http://www.storma.de /DPS/pdf/SDP8.pdf (last accessed: July 29, 2004). 

Charnley S. 1997. Environmentally displaced peoples and the cascade effect: Lessons from 

Tanzania. Hum Ecol 25: 593 – 618. 

Christanty L. 1990. Home gardens in tropical Asia, with special reference to Indonesia. In: 

Landauer K. and Brazil M. (eds), Tropical home gardens, pp 9 – 20. The United Nations 

University, Tokyo. 



Gutiérrez M., Quiróz C., Pérez D., Rodríguez D., Pérez T., Marques A. and Pacheco W. 2004. 

Conservación in situ de diversas especies vegetales en ‘conucos’ (home gardens) en los 

estados Carabobo y Trujillo de Venezuela. Plant Gen Resour Newsl 137: 1 – 8. 

Hodel U., Gessler M., Cai H.H., Thoan V.V., Ha N.V., Thu N.X. and Ba T. 1999. In situ 

Conservation of plant genetic resources in home gardens of Southern Vietnam. 

International Plant Genetic Resources Institute, Rome, 106p. 

Hvoslef H. 1994. Homegardens of Javanese transmigrants in Seberida subdistrict: 

Description, agroecological constraints and evaluation of potential solutions to declining 

productivity. In: Sandbukt O. and Wiriadinata H. (eds), Rain forest and resource 

management. Proceedings NORINDRA seminar, pp 127 – 136. Indonesian Institute of 

Sciences (LIPI), Jakarta. 

Huxley P.A. 1999. Tropical agroforestry. Blackwell Science, Oxford, 371p. 



Karyono 1990. Home gardens in Java. Their structure and function. In: Landauer K. and 

Brazil M. (eds), Tropical home gardens, pp 138 – 146. The United Nations University, 

Tokyo. 

Kaya M., Kammesheidt L. and Weidelt H.-J. 2002. The forestt garden system of Saparua 

island, Central Maluku, Indonesia, and its role in maintaining tree species diversity. 



Central Sulawesi, Indonesia. Agroforest Syst 63: 53 – 62. 


135 – 152. 

Landauer K. and Brazil M. (eds). 1990. Tropical home gardens. The United Nations 

University, Tokyo, 257p.

354 


Landon J.R. (ed.). 1991. Booker tropical soil manual: A handbook for soil survey and 

agricultural land evaluation in the tropics and subtropics. Longman Scientific & 

Technical, Essex, 474p. 

Maertens M., Zeller M. and Birner R. 2002. Explaining agricultural land use in villages 

surrounding the Lore Lindu National Park in Central Sulawesi, Indonesia. STORMA 

Discussion Paper Series No. 4 (online). Universities of Goettingen/Kassel, Germany; 

Institut Pertanian Bogor/Universitas Tadulako, Indonesia. http://www.storma.de/DPS/ 

index. htm (last accessed: July 27, 2004). 

Magurran A.E. 1988. Ecological diversity and its measurement. Croom Helm, London, 179p. 



31 – 58. 

Soemarwoto O. 1987. Homegardens: A traditional agroforestry system with a promising 

future. In: Steppler H.A. and Nair P.K.R. (eds), Agroforestry: A decade of development. 


Soemarwoto O. and Conway G.R. 1991. The Javanese homegarden. J Farm Syst Res-Ext 2: 

95 – 118. 

Torquebiau E. 1992. Are tropical agroforestry home gardens sustainable? Agric Ecosyst 

Environ 41: 189 – 207. 

Watson J.W. and Eyzaguirre P.B. (eds). 2002. Home gardens and in situ conservation of plant 

genetic resources in farming systems. Proceedings of the second international home 

gardens workshop (17 – 19 July 2001), Witzenhausen, Germany. International Plant 

Genetic Resources Institute, Rome, 184p. 



Whitten A.J., Mustafa M. and Henderson G.S. 1987. The Ecology of Sulawesi. Gadjah Mada 

University Press, Yogyakarta, 777p. 

Wood A.P. 1993. Natural resource conflicts in south-west Ethiopia: State, communities, and 

the role of the National Resource Strategy in the search for sustainable development. Nord 

J Afr Stud 2: 83 – 99. 

Zeller M., Schwarze S. and van Rheenen T. 2001. Statistical sampling frame and methods 

used for the selection of villages and households in the scope of the research program on 

stability of rainforest margins in Indonesia (STORMA). STORMA Discussion Paper 

Series No. 1 (online). Universities of Goettingen and Kassel, Germany, Institut Pertanian 

Bogor and Universitas Tadulako, Indonesia. http://www.storma.de/DPS/pdf/SDP1.pdf 

(last accessed: January 20, 2003).

CHAPTER 20 

WHITHER HOMEGARDENS? 

P.K.R. NAIR 

School of Forest Resources and Conservation, University of Florida, Gainesville, 

Florida 32611, USA; E-mail: 

Keywords: Commercialization, Species diversity, Sustainability, Urban homegardens. 

Abstract. Although homegardens provide sustenance to millions of households in the tropics, 

their underlying scientific foundations have not been fully explored, and therefore they are not 

a part of development agendas. While their integrated and complex nature are a challenge to 

scientific investigations that are often compartmentalized, these very same attributes form the 

bases of the ecological, economic, and social sustainability of homegardens. In the wake of 

recent trend towards commercialization and consequent conversion of homegardens to 

produce market-oriented crops, concerns have been raised about the future of traditional 

homegardens. Lack of rigorous scientific evidence makes it difficult to make predictions. 

Nevertheless, experiences about the role and value of homegardens from around the world 

suggest that homegardens are not on the path to extinction. They will continue to be an 

essential part of the way of life, but their nature and functions will change in tune with the 

rapid changes happening all over. The concept of homegardens will increasingly be adopted 

in urban and periurban areas, not only in the tropics, but also in industrialized societies, 

reflecting the society’s increasing appreciation of traditional values and ecosystem functions. 


“… that whoever could make two ears of corn, or two blades of grass, to grow on a spot 

of ground where one grew before, would deserve better of mankind and do more 

essential service to his country …” 

Jonathan Swift 

The above quote that I included at the beginning of my first book nearly three 

decades ago (Nair, 1979) is as apt now as it was then. The subject matter of that 

book “Intensive Multiple Cropping with Coconuts in India,” written before the 

355 




356 

P.K.R. NAIR N 

advent – or just at the beginning – of “modern” agroforestry, is not very different 

from the subject matter of this book, i.e., homegardens: multiple cropping with 

coconuts (Cocos nucifera) and other tree crops, now commonly referred to as 

multistrata agroforestry, is a distinguishing feature of (most) tropical homegardens. 

What Jonathan Swift envisioned in making two ears of corn, or two blades of grass, 

to grow on a spot of ground where one grew before is exactly what homegardeners 

have been practicing, especially in the warmer biomes, for centuries, i.e., growing an 

array of herbaceous species, shrubs, vines, and trees, all in intimate association on 

the same piece of land around their homes. Yet, these magnificent farming practices 

and intriguing plant associations are seldom recognized as worthy of consideration 

in development paradigms and ecological studies, nor are their practitioners treated 

as “… better of mankind doing more essential service to their countries …” 

In spite of this apparent neglect of homegardens and homegardeners, the reasons 

for which have been discussed in several previous writings (Nair, 2001; Kumar and 

Nair, 2004), the appeal, relevance, and lessons to be learned from this time-tested 

practice are so overwhelming and fascinating that time and again it attracts the 

attention of some researchers. For example, publications on homegardens can be 

found in almost all volumes of Agroforestry Systems. While some of them are at best 

scientific descriptions of a set pattern (characteristics of systems at specific 

locations), some deal with examining homegardens in the context of current trends 

and issues in land use systems, such as environmental integrity, carbon 

sequestration, biodiversity conservation, economic valuation of intangible benefits, 

and social equity, to name a few. Only very few of these are scientific analyses, 

however. Nevertheless, all such publications – old and new – on homegardens have 

had only “good things” to say about the practice: irrespective of its focus – be it C 

sequestration, biodiversity, soil fertility, or whatever – the study will have the 

inevitable conclusion that homegardens are “great” on that score. 

Other than these occasional researcher-motivated efforts – and, of course, the 

incessant individual efforts of the homegardeners – there has been no organized 

institutional initiative to promote homegardens either locally anywhere or 

internationally. That is hard to understand: if homegardens have all these desirable 

characteristics, why have they not earned a rightful place as a development vehicle? 

If homegardens are the “epitome of sustainability” (Torquebiau, 1992), how is it that 

they “defy” scientific explanation, or is it that homegardens are just a “backyard” 

activity with little prospects as a development tool and therefore not worthy of any 

serious scientific investigation? No answer has yet been found to the question that 

was posed five years ago: “Do homegardens defy science or is it the other way 

around?” (Nair, 2001). In the meanwhile, commercialization seems to make its way 

to homegardens that have traditionally been known as anything but commercial. 

Two chapters in this book report the recent tendency for growing crops in 

homegardens mainly for commercial use, in Java, Indonesia (Abdoellah et al., 2006) 

and Kerala, India (Peyre et al., 2006), the two best-known bastions of traditional 

homegardens. Is this an indication of the heralding of a new genre of homegardens 

and possibly the demise of the traditional ones? Is such an “evolution” of 

homegardens good or bad? In other words, what does the future hold for 

homegardens?

WHITHER W HOMEGARDENS? 

357 

In order to address the above key question, we need to discuss why homegardens 

(especially their species diversity) have traditionally been important to the 

households and what the relevance is of the much-acclaimed sustainability attributes 

of homegardens to the current context and future prospects. 

2. SPECIES DIVERSITY IN HOMEGARDENS AND HOUSEHOLD FOOD 

SECURITY 

The most distinguishing and possibly important characteristic of all homegardens is 

their species diversity: the intimate admixture of plants of all types – herbs, shrubs, 

vines, trees, other perennials, and so on – on the same small parcel of land (Fig. 1). 

From the homegardener’s point of view, the primary objective of growing all these 

plants together is to produce food, often as a supplementary source. In order to 

appreciate the role of these plants grown in apparent disarray, we have to first of all 

recognize the fact that ‘he’, the traditional homegarden practitioner, is a ‘she’: 

Figure 1. A “typical” rural homegarden in Kerala, India, showing a large number of 

economic species in intimate association around the home (Photo: B. Mohan Kumar). 

women have primary responsibilities, or are as involved as men, for homegarden 

maintenance. This is common wherever homegardening is practiced. Considering

358 

P.K.R. NAIR N 

that it is primarily the woman’s responsibility in many societies to feed the families, 

it is perhaps a combination of both inspiration and desperation that prompt them to 

grow food around their homesteads: inspiration from experience and innovative 

instinct, and desperation from the lack of other avenues for finding food for the 

family. Species diversity in these systems may be a consequence of the interplay of 

these forces of inspiration and desperation. Mixing annual food crops with 

frequently harvestable tree crops that provide food and sometimes cash income to 

the family represents a confluence of human ingenuity with ecological ambience, 

such that the opportunity offered by year-round growing seasons and the 

amenability of the various species to grow in mixed stands makes it a “win – win” 

situation. Tracing the historical development of homegardens, Wiersum (2006) 

observes that in the most widely studied homegarden systems in South- and 

Southeast Asia, homegardens are used to produce products with high nutritional 

va1ue (proteins, vitamins, minera1s), medicina1 plants and spices, firewood, and 

sometimes a1so forage crops and construction wood, and homegardening is always 

combined with field-crop cultivation often in the form of wetland rice (Oryza sativa) 

in South- and Southeast Asia. These regions with good farming conditions and high 

population densities contributed to optimal development of the complementary 

system of staple food cultivation in open fields and supplementary diversified 

homegarden production for the family’s self-sufficiency and trade. 

Whatever be the reason for species diversity, and irrespective of whether it will 

continue to be a conspicuous feature of future homegardens in the wake of the push 

to commercialization, researchers seem to be quite obsessed (perhaps more than the 

practitioners) with species diversity of homegardens. Cataloging of species lists is 

such a common feature of most homegarden literature to the extent that many 

authors believe that a paper on any aspect of homegarden is incomplete without a 

species list (Nair and Kumar, 2006). An interesting point that comes out of such 

species lists is that, irrespective of the geographical focus of the study, the species 

that dominate such lists are the same from similar ecological regions. This is evident 

from the species listed in four chapters of this book, summarized in Table 1, from 

homegardens in Kerala, India (Mohan et al., 2006); Peruvian Amazon (Wezel and 

Ohl, 2006); and two locations in the Pacific islands (Lamanda et al., 2006; and 

Thaman et al., 2006). The situation may not be different if the study is extended to 

all the 135 case studies included in Fig.1 of Nair and Kumar (2006), with the 

exception that in some locations, the locally important species that are not common 

outside their limited geographical areas of distribution will be common in 

homegardens as well. Examples of this category include the peach palm (Bactris 

gasipaes) and various other palm species in Central and South America, fruit trees 

such as durian (Durio zibethynus) in Southeast Asia and breadfruit (Artocarpus ( 

altilis) in the Pacific islands, and various fruit trees in West Africa (Cola spp., 

Dacroydes edulis, Pterocarpus spp., Treculia africana: Okafor and Fernandes, 

1987). Similarly, in the tropical highlands, the dominant species in homegardens 

will be different from those in tropical lowlands (e.g., Fernandes et al., 1984; and 

Soini, 2005; for the Chagga homegardens of Tanzania and Tesfaye Abebe et al., 

2006, for the homegardens of Ethiopian highlands).

359 

The bottom line is, dominant food crops, both herbaceous and woody, that are 

locally adapted have been the dominant species of homegardens in different 

ecological regions. The easy access to these crops in the backyard and the 

opportunity offered by many of them for staggered harvesting as needed (e.g., tuber 

crops, vegetables, plantain) make them quite attractive to the women who take it on 

themselves as their obligation and responsibility to find food for the family. 

Nutritional security (rather than food security) of the homegarden is another 

important benefit of homegardens. It is well known that several of the tree fruits in 

the gardens (Table 1) are nutritionally richer than the common, carbohydrate-rich 

grain crops, and are indeed the main sources of vitamins and minerals to the family 

(Niñez, 1984; Okafor and Fernandes, 1987; Kumar and Nair, 2004; Nair, 2006). The 

cash-income opportunity offered by saleable products (especially tree products) 

from the homegardens make it an attractive proposition for men too. Social and 

cultural value of the species in the homegardens is yet another important factor to be 

considered (discussed later). Species diversity of homegardens is thus quite an 

appealing feature to the homegardeners for a variety of reasons, and has been a 

major driving force in the maintenance of the gardens over centuries. 

Table 1. Commonly reported plants in homegardens of humid tropical lowlands. 

Category Species in homegardens 

Root and 

tuber crops 

Other food 

crops 

Fruit and nut 

yielding 

perennials 

Spices, Social 

beverages, 

and 

stimulants 


Colocasia esculenta (taro), Dioscorea alata (greater yam), 

Dioscorea esculenta (sweet yam), Ipomoea batatas (sweet potato), 

Manihot esculenta (cassava), Xanthosoma spp. (tannia or 

cocoyam) 

Ananas comosus (pineapple), Arachis hypogaea (peanuts), Cajanus 

cajan (pigeon pea), Passiflora edulis (passion fruit), Phaseolus, 

Psophocarpus and Vigna spp. (beans and other legumes), 

Saccharum officinarum (sugarcane), Zea mays (corn = maize), and 

various vegetables 

Anacardium occidentale (cahew nut), Annona spp. (soursop and 

sweetsop), Averrhoa carambola (carambola), Artocarpus 

heterophyllus (jackfruit), A. altilis (breadfruit), Carica papaya 

(papaya), Citrus spp. (lemon, lime, orange, tangerin), Cocos 

nucifera (coconut), Ficus spp. (edible figs), Mangifera indica 

(mango), Musa spp. (bananas and plantains), Persea americana 

(avocado), Psidium guajava (guava), Spondias dulcis (vi apple, 

hogplum), Syzygium malaccense (Malay apple), Tamarindus indica 

(tamarind) 

Areca catechu (betel nut), Cinnamomum zeylanicum (cinnamon), 

Curcuma longa (turmeric), Cymbopogon citratus (lemon grass), 

Piper betle (betel vine), Piper methysticum (kava), Zingiber 

officinale (ginger).

360 

P.K.R. NAIR N 

3. SUSTAINABILITY AND HOMEGARDENS 

Sustainability is perhaps the most widely discussed, yet least well-defined, term 

across disciplines in contemporary agricultural and land use literature. Even before 

publication of the much-acclaimed and so-called Brundlandt Commission report 

(WCED, 1987), sustainability has been a cornerstone of many traditional land use 

systems and it used to figure prominently in the early debates on agroforestry (Bene 

et al., 1977). Without going into any discussion on this much-discussed issue, 

suffice it to say that sustainability is about meeting today’s needs without 

compromising the ability of future generations to satisfy their needs; it is not a new 

concept, simply the retrieval of ancient wisdom dictating that “you don’t eat your 

seed corn”; and it strives to achieve a balance between ecological preservation, 

economic vitality, and social justice. 

Much of the discussion on ecological sustainability of homegardens is linked to 

their species diversity. While dealing with species of various forms, life cycle, and 

nature of products, the number or frequency of occurrence of a species in the 

homegarden is not a sufficient indicator of the importance or dominance of the 

species. Ecological parameters and indices that are commonly used to express 

population complexity and diversity such as Sorenson’s index of similarity, 

Shannon-Weiner and Margalef Indices of species diversity, and Importance Value 

Index, have lately been reported in homegarden studies (Kumar and Nair, 2004 – for 

literature until then; Mohan et al., 2006; Abdoellah et al., 2006; Kehlenbeck and 

Maass, 2006). Some authors have also used statistical procedures such as cluster 

analysis and correspondence analysis to group descriptive characteristics of 

homegardens, and to find out factors that may play a significant role in explaining 

patterns of floristic composition of the complex system; one such study is reported 

by Tesfaye Abebe et al. (2006) in this volume. 

The rationale is to use these indices as a basis for comparing homegardens with 

nearby natural vegetation – usually forests – on the assumption that in terms of 

species abundance and diversity, homegardens are in between natural systems and 

managed systems. Homegardens are perhaps the most diverse agroforestry practice, 

and among all agroforestry practices, they are at one end of the spectrum, twospecies 

(a tree and a crop) associations such as alleycropping being at the other end 

(Nair, 1993; Rao et al., 1988). Species abundance and diversity of homegardens 

should not, however, be equated with ecological succession that is characteristic of 

natural systems and the benefits of which are exploited in some traditional low-input 

agricultural systems such as shifting cultivation. The fact that natural systems are 

more diverse than agricultural systems has been known for long, one of the most 

widely cited articles on the subject being that of Odum (1969). In the very few 

examples of low-input agriculture that take advantage of the process of succession, 

the species are all carefully selected, but are not random successional species that 

seed-in naturally. In homegardens too, the species are selected carefully, and are 

therefore similar to such systems. Homegardens start off from one particular stage of 

the natural successional process, but keep natural succession from carrying the 

community to a so-called “climax” community. On the other hand, agroforestry 

practices such as alleycropping that are at the “other end” of the species-diversity


361 

spectrum have little similarity with the natural systems and do not fit into the realm 

of successional processes. Thus, in terms of complexity and species diversity, 

homegardens represent a unique set of ecological sustainability characteristics of 

natural systems as well as production benefits of agricultural systems. Another 

aspect of ecological sustainability in homegardens is the benefit of nutrient cycling 

experienced in multistrata systems, which is again a consequence of the species 

diversity (Nair et al., 1999). 

It needs to be pointed out in this context that the premise that diversity provides 

stability to ecosystems, which is the basis of the concept of ecological sustainability 

of homegardens, is being debated by ecologists: the so-called “diversity – stability 

debate” (e.g., McCann, 2000). Although the consensus of this debate as of now is 

that diversity can be expected, on average, to give rise to ecosystem stability, 

diversity is not the driver of this relationship; rather, ecosystem stability depends on 

the ability of communities to contain species, or functional groups, that are capable 

of differential responses. At present, in ecological studies, the role of keystone 

species is receiving increasing attention; this concept has hardly been used in 

homegarden studies yet, but seems to offer scope for further studying the diversity – 

stability issue in homegardens (see Tesfaye Abebe, 2006). If simplified communities 

are more vulnerable to invasion by other communities/species, then the trend 

towards commercialization of homegardens (discussed later) should result in higher 

frequency of invader species as well as pests and diseases in homegardens. The 

profit-oriented commercial homegarden enterprises will then resort to keeping such 

invading species under check through use of chemicals, which will inevitably 

disrupt the harmonious biodiversity and species associations (including microorganisms 

and species other than plants) that have been so characteristic of 

traditional homegardens. 

Economic and social sustainability attributes of homegardens are even less well 

studied than ecological-sustainability attributes. A common problem seen mentioned 

in most attempts to study economic benefits of homegardens is, again, lack of 

widely accepted procedures to measure economic benefits of intangible benefits and 

services. Alavalapati and Mercer (2004) described some procedures for economic 

valuation of agroforestry systems. Most attempts at economic valuation have two 

common features: first, they acknowledge the importance and need for “proper” 

evaluation of the intangible benefits of homegardens, such as aesthetics and 

ornamentation, nutritional security, food quality, and empowerment of women; then 

they highlight the difficulties involved in collecting realistic data and therefore 

caution about the error-prone nature of such analyses. The two chapters on 

economic analysis presented in this volume are no exception to this general trend: 

Torquebiau and Penot (2006) articulate the importance of including valuation of 

such benefits in homegarden evaluation, but stop short of suggesting any new 

procedures; and, Mohan et al. (2006), following a study applying conventional and 

some “non-conventional” economic procedures in some Kerala homegardens, 

confirm that the results are along expected lines and caution that their study 

procedure will need considerable “fine-tuning” to adapt to local conditions before it 

is applied elsewhere. Thus, economic sustainability of homegardens remains another

362 

attribute, the importance of which can only be felt qualitatively and intuitively, but 

is difficult to quantify. 

The same can be said about social sustainability. All social studies on 

homegardens exclaim the social attributes of homegardens, ranging from their role 

in ensuring gender equality and nutritional security to societal harmony and cultural 

heritage. Several chapters in this book touch upon these issues. Howard (2006) 

presents a well researched account of the major role of women in homegardens in 

Latin America: the presence of a garden rich in a variety of plants epitomizes the 

woman’s exertions on behalf of kin and her proficiency as primary provider of food, 

health, and overall well-being of the family, and demonstrates her freedom from 

dependence on products from neighbors and commercial vendors. Abdoellah et al. 

(2006) describe how the tendency towards conversion of homegardens to produce 

commercially valuable crops for market in Indonesia has disrupted the community’s 

equality, sharing, and harmonious living (rukun) that used to be built around 

traditional homegardens, and decreased the number of common grounds (buruan) in 

front of homes that serve as playground for children, and as a place for socializing 

with neighbors and for children to learn cultural and social values from their elders. 

The strength of these threads that are woven together in the fabric of social 

sustainability of homegardens cannot be expressed in quantitative terms. 

4. HOMEGARDENS AND SOME CURRENT LAND USE ISSUES 

4.1. Biodiversity 

P.K.R. NAIR N 

Biodiversity (short form for biological diversity) is often used as a synonym for 

species diversity. The importance of maintaining biodiversity in sustaining food 

production and protecting human and ecosystem health is now universally 

recognized, and land use systems that promote biodiversity are considered to be 

quite desirable from that perspective. A classification based on the production 

systems and species diversity ranked homegardens top with its highest biological 

diversity among all manmade agroecosystems (Swift and Anderson, 1993). Species 

richness and extent of biodiversity in homegardens depend, however, on ecological 

and socioeconomic factors and household preferences. Gajaseni and Gajaseni (1999) 

have reported, for example, the existence of non-commercial indigenous varieties of 

durian (Durio sp.) and rare varieties of mango (Mangifera indica) in homegardens 

of Thailand. Large numbers of cultivars of banana (Musa paradisiaca), coconut, and 

breadfruit have been reported in the homegardens of Micronesia (Falanruw, 1990; 

Thaman et al., 2006). Indeed, as already mentioned, most publications on 

homegardens from around the world (see Fig. 1: Nair and Kumar, 2006) report the 

large numbers of species present. The role of homegardens as repositories of plant 

biodiversity is thus indisputable. In a recent study from seven New- and Old-World 

tropical forest dynamic plots, Wills and 33 collaborators from 21 institutions around 

the world reported that an erosion of an ecological community's species diversity 

(that tends to happen as a result of stochastic extinction, competitive exclusion, and 

unstable host-enemy dynamics) can be prevented over the short-term through

363 

preferential introduction of rare species (Wills et al., 2006). They found that when 

species were rare in a local area, they had a higher survival rate than when they were 

common, resulting in enrichment for rare species and increasing diversity with age 

and size class in these complex ecosystems. Thus, it can be surmised that the 

preferential introduction of rare species such as medicinal plants (Rao and 

Rajeswara Rao, 2006) and fruit trees that homegardeners have been practicing for 

centuries around the world contributes to species biodiversity even if economic and 

social gains are the primary motivations for such introductions. 

4.2 Genetic-diversity conservation and species domestication 

In addition to the wide array of plants grown in homegardens for a variety of 

reasons, homegardens have high potential for in situ conservation of genetic 

resources (Watson and Eyzaguirre, 2002; McNeely, 2004; Schroth et al., 2004). An 

important issue, the significance of which is seldom recognized in the extant 

species-listing-dominated literature on homegardens, is the continuous interaction of 

homegardeners with these large groups of plants and the resultant contribution to 

species domestication. Simons and Leakey (2004) describe the deliberate selection 

and management of trees (domestication) by humans that has been going on for 

millennia in agroforestry systems. For example, Leakey et al. (2004) present 

evidence that subsistence farmers have domesticated locally popular indigenous 

fruits (Dacroydes edulis and Irvingia gabonensis) in Cameroon and Nigeria. It is 

reasonable to assume that much of this in situ domestication has taken place in 

homegardens. It is also likely that similar patterns of domestication have happened 

for other plant species in homegardens around the world, especially in those with 

long history as in South- and Southeast Asia (Wiersum, 2004). 

4.3. Carbon sequestration 


Most discussions on carbon sequestration potential of homegardens – and, indeed 

agroforestry systems in general – are based more on hypothetical considerations 

than empirical results. The argument is that these systems have high carbon storage 

(sequestration) potential in their multiple plant species, especially in woody 

perennial species, and soil; they help in conservation of C stocks in existing forests 

by alleviating the pressure on natural forests (Schroth et al., 2004); and, to some 

extent, in C substitution by reducing fossil-fuel burning through promotion of wood 

fuel production. Most reports indicate that the addition of a large proportion of the 

relatively high quantity of plant materials produced in a system will increase C stock 

in soils (Lal, 2004); therefore it is reasonable to surmise that homegardens will help 

substantially in C sequestration. All reports on C sequestration potential of 

homegardens (e.g., Montagnini and Nair, 2004; Kumar, 2006), however, are related 

to aboveground biomass. In the case of soils, C stored in surface soils has received 

some mention. But C exists in soils in labile (mobile) or recalcitrant (stable) form; 

the latter is more important for C sequestration; and, no study has been reported on 

this “real” form of C sequestration within soil profiles in homegardens. Most C

364 

sequestration reports also have disclaimers and caveats that lack of reliable 

inventories/estimates and uncertainties in the methods of estimation present serious 

difficulties. Thus, as in the case of other intangible and difficult-to-measure benefits 

and services, C sequestration benefit of homegardens remains one of the “potential 

benefits” that has not been even quantified, let alone exploited. 

5. NEW DIMENSIONS OF HOMEGARDENS 

5.1. Commercialization of homegardens 

Consequent to liberalizations in many formerly tightly controlled economies, agricultural 

enterprises, just as other production enterprises, are becoming increasingly subject to 

market pressures. A direct consequence of this is development and adoption of new 

strategies to promote commercialization of even traditional operations such as 

homegardens. Abdoellah et al. (2006) describe a case study of such a transformation 

in a West Java village in Indonesia, where some villagers, attracted by economic 

possibilities, have transformed their homegardens in such a way that they have 

become dominated by few plant species or are approaching even monocultures; the 

dominant species are cash crops such as vegetables that are in high demand in 

nearby urban markets. Similar examples are also prevalent in the Pacific islands as 

described by Thaman et al. (2006), where promotion of a wide range of export cash 

crops in rural areas has led to the clearing of diverse agroforests. Increasing trend 

towards commercialization has also been reported from Kerala homegardens 

(Kumar and Nair, 2004). 

This so-called commercialization is, however, not new to homegardens. It has 

been in existence to varying degrees in most well-known homegardens (of South 

and Southeast Asia). Perennial species that produce commercial products such as 

spices, fruits and nuts, medicinal plants, and even timber have been a component in 

many of these systems. As Kumar and Nair (2004) have pointed out, although 

interest in homegardens has been primarily focused on producing subsistence items, 

its role in generating additional cash income has been quite substantial in many 

places. Considerable variations from place to place have also been reported in the 

proportion of homegarden products that are used for household consumption as 

opposed to sale, and the contribution of the net income derived from sale of products 

to the total household income. Conversion of homegardens to intensive production 

units of market-oriented systems as described by Abdoellah et al. (2006) is not a 

totally new phenomenon; similar trends have occurred in several rapidly urbanizing 

and periurban centers. A case in point is the conversion of the traditional shamba 

gardens of Kenya’s highlands to produce vegetables for sale in Nairobi, the capital 

city, and for export to Europe (author’s personal experience). 

5.2. Urban homegardens 

P.K.R. NAIR N 

Another relatively new trend related to commercialization of homegardens is the 

extension of the homegarden practice from its conventional rural settings to urban


365 

environments. Two chapters in this book (Drescher et al., 2006; and Thaman et al., 

2006) describe such developments; while the former includes examples from several 

places around the world representing both developing and developed countries, the 

latter deals primarily with such developments in the Hawaiian Islands, USA. These 

urban homegardens are often the “modern” cousins of their traditional relatives in 

the sense that while they maintain the species diversity that is characteristic of the 

traditional homegardens, their aesthetic and recreational value is as important as – if 

not more than – their nutritional role. As Fig. 2, a photograph of an urban 

homegarden in Kona, Hawaii, USA, shows, the gardens with manicured lawns and 

hedges, well tended fruit trees, and attractive ornamentals surrounding a “modern” 

home look more like tourist resorts, in sharp contrast to the “natural” look of the 

subsistence-oriented homegardens and the type of “traditional” homes they surround 

(Fig. 1). 

Figure 2. An urban homegarden with fruit trees such as avocado (Persea americana), litchi 

(Litchi chinensis), mango (Mangifera indica), papapya (Carica papaya, a and various 

ornamentals, in Kona, Hawaii, USA (Photo: Craig Elevitch). 

This trend towards urban homegardening may be seen in the context of other 

similar activities such as urban forestry and organic agriculture that have gained 

considerable prominence in urban and periurban areas during the recent past. These 

activities constitute a substantial portion of the green space and are considered to be 

the lungs of the cities. For example, the role of urban vegetation in mitigating 

atmospheric greenhouse gas concentrations and improving air quality in Santiago, 

Chile (a city of more than 4 million inhabitants), was illustrated in a recent study

366 

P.K.R. NAIR N 

(F. Escobedo, personal communication; January 2006). Gaston et al. (2005) reported 

that the ‘domestic gardens’ with mean area of only 151 m 2 per garden covered 

approximately 33 km 2 or 23% of the predominantly urban area of the city of 

Sheffield, U.K., and provided tremendous opportunities for maintenance of 

biodiversity and provision of ecosystem services in urban areas. Furthermore, there 

is a revival of appreciation of recreational and social values of ornamental and other 

types of homestead gardening in the industrialized world such as the United States 

(Westmacott, 1992) and Europe (Vogl and Vogl-Lukasser, 2003). An increasing 

number of gardeners are now finding pleasure in growing plants for various uses and 

deriving satisfaction from agrarian life-style, self-reliance, and private ownership – a 

clear expression of the appreciation of the aesthetic, cultural, and landscape values 

of such integrated systems, and perhaps the bygone days. 

6. FUTURE OF HOMEGARDENS 

Prompted by the lack of appreciation of the value of homegardens in development 

paradigms and the trends towards commercialization of homegardens and urban 

homegardens, the question has been posed “are homegardens becoming extinct?” 

(Kumar and Nair, 2004). Wiersum (2006) argues that this illustrates that “the notion 

of socioeconomic sustainability of homegardens should be interpreted as referring 

not only to their ability to contribute towards the livelihood needs of traditional rural 

dwellers, but also to their ability to adjust to the process of rural change.” 

Obviously, no one can accurately predict the future of an activity such as 

homegardening that is deeply rooted in ecological, socioeconomic, and cultural 

milieu of the land and its people. Some of the well-known predictions such as the 

200-year-old Malthusian theory are even better known today for their failures to 

hold up in a changing world. As the old adage goes, change is the only constant 

thing. Homegardens are no exception; they will certainly be affected by the changes 

happening in the local ecology, economics, and culture. The rate and extent of the 

impact of such changes will depend on a myriad of factors. Economic and cultural 

forces often pull the society and people’s attitudes in opposite directions. If some 

farmers in periurban centers are attracted by the forces of economics to convert their 

homegardens or sections of them to growing crops that can fetch money in the 

market, there will be an equally strong (if not stronger) section of farmers who are 

not attracted by the lure of money to abandon their age-old traditions. When, rather 

than if, some genetically modified crops find their way to homegardens, that may 

not necessarily mean a proliferation of transgenic homegardens – at least in the near 

future. In fact, homegardens are “testing grounds” of many innovations of the 

gardeners, and today’s gardens of long standing are a result of such continuous 

innovation and improvement. The migration of the youth to urban and even overseas 

centers in search of jobs and cash income, a common feature in many homegardendominated 

societies, naturally raises concerns about the future of homegardens, 

particularly the scope for bringing any technological innovations to the practice of 

homegardening. What is seldom recognized, however, is the reverse migration of 

older workforce who, after long stays in industrialized urban centers get 

disenchanted and seek to return to their roots in increasing numbers and take up


367 

hobby farming and homegardening for the pure pleasure of doing something they 

have grown up with and to which they possess a cultural bondage; this reverse 

migration seldom gets the media attention of out-migration of youth. 

What conclusion can, then, be drawn on the future of homegardens? Will they 

survive or will they become extinct? It is anybody’s guess. I, for one, have 

relentlessly argued for quantitative and measurable evidence in support of a 

conclusion. But I don’t have much evidence of that nature to draw upon in this case. 

So, I would rather make no prediction. Nevertheless, my intuition is that 

homegardens will not become extinct. Because of the difficulties in quantitative 

valuation of the sustainability attributes of homegardens, it is unlikely that 

homegardens will become a part of the development bandwagon; therefore it is 

unlikely that there will be any “big push” towards research on homegardens. But 

that will not lead to the demise of homegardens. I have only my personal 

experiences of interactions with homegardeners around the world to support this 

intuitive prediction: the innovative spirit of the Japanese settler farmer in Tomé-Açu 

(Brazil), the sentimental attachment to ancestral land and way of life of the 

homegardeners in Kerala (India), the tenacity of the farmers who maintain 

economically attractive Kandyan homegardens (Sri Lanka), the community’s 

commitment to traditional life style of the homegardeners in Nakhon Sawan 

(Thailand), the intuitive skills of the industrious and tradition-bound homegardeners 

of Java even after they were transmigrated under government pressure to unfamiliar 

and distant lands in Kalimantan (both in Indonesia), the friendliness and confidence 

of the ecotourism-oriented homegardeners of the Blue Mountain region (Jamaica), 

the hope and aspirations built around homegardens of the hapless rural folks in 

Koutiala (Mali) and Cap Haitien (Haiti), the satisfaction of the gardeners in being 

able to produce a variety of food and other essential needs in their homegardens in 

mountainous landlocked terrains in Mount Hagen (Papua New Guinea) and waterlocked 

Gizo (Solomon Islands), the pride and self-confidence effused by the female 

gardeners in the shambas of the Kikuyuland (Kenya) and the chagga in Arusha 

region (Tanzania), the ingenuity of the farmers who have successfully introduced 

rearing in captivity through stall-feeding of the African grasscutter (Thryonomys 

swinderianus, a herbivorous rodent that is harvested for delicious and pricy bush 

meat) in Kumasi (Ghana), … – the list can be long – all point to continuation of the 

homegardens, of sorts, in perpetuity. So, my submission is, homegardens will 

undergo changes; but they will not become extinct; they will continue to exist with 

their mysterious, enigmatic charm to provide sustenance, satisfaction, income, and 

aesthetic appeal to many, and fascination to scientists who care to look at them. 


I thank Drs. Michael Bannister, Vimala Nair, and Freerk Wiersum for comments 

and suggestions on the manuscript.

368 

P.K.R. NAIR N 

REFERENCES 



functional changes. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A 

time-tested example of sustainable agroforestry, pp 233 – 250. Springer Science, 

Dordrecht. 

Alavalapati J.R.R. and Mercer D.E. (eds). 2004. Valuing agroforestry systems: Methods and 

applications (Advances in Agroforestry # 2). Kluwer, Dordrecht, 314p. 

Bene J.G., Beall H.W. and Côte A. 1977. Trees, food and people. International Development 

Research Centre, Ottawa, Canada. 52p. 


gardens for sustainable livelihoods: Management strategies and institutional 

environments. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A timetested 


Falanruw M.V.C. 1990. The food production system of the Yap Island. In: Landauer K. and 

Brazil B. (eds), Tropical home gardens, pp 94–104. United Nations University Press, 

Tokyo. 

Fernandes E.C.M., O’Kting’ati A. and Maghembe J. 1984. Chagga homegardens: a 

multistory agroforestry cropping system on Mt. Kilimanjaro, northern Tanzania. 




Gaston K.J., Warren P.H., Thompson K and Smith R.M. 2005. Urban domestic gardens (IV): 

the extent of the resource and its associated features. Biodivers Conserv 14: 3327 – 3349. 


America. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 


Kehlenbeck K. and Maass B.L. 2006. Are tropical homegardens sustainable? Some evidence 

from Central Sulawesi, Indonesia. In: Kumar B.M. and Nair P.K.R. (eds), Tropical 







135 – 152. 

Lal R. 2004. Soil carbon sequestration impacts on global change and food security. Science 

304: 1623-1627. 

Lamanda N., Malézieux E. and Martin P. 2006. Structure and dynamics of coconut-based 

agroforestry systems in Melanesia: a case study from Vanuatu archipelago. In: Kumar 



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. and Ukafor V. 2004. Evidence 

that subsistence farmers have domesticated indigenous fruits (Dacryodes edulis and 

Irvingia gabonensis) in Cameroon and Nigeria. Agroforest Syst 60: 101 – 111. 

McCann K.S. 2000. The diversity – stability debate. Nature 405: 228 – 233. 

McNeely J.A. 2004. Nature vs. nurture: Managing relationships between forests, agroforestry 

and wild biodiversity. Agroforest Syst 61: 155 – 165.


369 

Mohan S., Alavalapati J.R.R. and Nair P.K.R. 2006. Financial analysis of homegardens: A 

case study from Kerala State, India. In: Kumar B.M. and Nair P.K.R. (eds), Tropical 



Mohan S., Nair P.K.R. and Long A.J. 2006. An assessment of homegarden diversity: A case 

study from Kerala State, India. J Sust Agric (in press). 

Montagnini F. and Nair P.K.R. 2004. Carbon sequestration: An under-exploited 

environmental benefit of agroforestry systems. Agroforest Syst 61: 281 – 298. 

Nair P.K.R. 1979. Intensive multiple cropping with coconuts in India: Principles, 

programmes and prospects. Verlag Paul Parey, Berlin, 149p. 

Nair P.K.R. 1993. An Introduction to Agroforestry. Kluwer, Dordrecht, 499p. 



Nair P.K.R. 2006. Agroforestry for sustainability of lower-input land use systems. J Sust. 

Crop Production (in press). 




Nair P.K.R., Buresh R.J., Mugendi D.N. and Latt C.R. 1999. Nutrient cycling in tropical 

agroforestry systems: Myths and science. In: Buck L.E., Lassoie J.P., and Fernandes 

E.C.M. (eds), Agroforestry in sustainable agricultural system, pp 1 – 31. CRC Press, Boca 

Raton, FL. 

Niñez V.K. 1984. Household gardens: theoretical considerations on an old survival strategy. 

Potatoes in Food Systems Research Series Report No. 1, International Potato Center, 

Lima, Peru, 41p. 

Odum E.P. 1969. The strategy of ecosystem development. Science 164: 262 – 270. 

Okafor J.C. and Fernandes E.C.M. 1987. The compound farms of Nigeria: A predominant 

agroforestry homegarden system with crops and small livestock. Agroforest Syst 5: 153 – 

168. 

Peyre A., Guidal A., Wiersum K.F. and Bongers F. 2006. Homegarden dynamics in Kerala, 

India. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 


Rao M.R. and Rajeswara Rao B.R. 2006. Medicinal plants in tropical homegardens. In: 

Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested example of 

sustainable agroforestry, pp 205 – 232. Springer Science, Dordrecht. 

Rao M.R., Nair P.K.R. and Ong C.K. 1988. Biophysical interactions in tropical agroforestry 

systems. Agroforest Syst 38: 3 – 50. 

Schroth G., da Fonseca A.B., Harvey C.A., Gascon C., Vasconcelos H.L. and Izac N. (eds). 

2004. Agroforestry and biodiversity conservation in tropical landscapes. Island Press, 

Washington, DC. 523p. 

Simons A.J. and Leakey R.R.B. 2004. Tree domestication in tropical agroforestry. Agroforest 

Syst 61: 167 – 181. 

Soini E. 2005. Changing livelihoods on the slopes of Mt. Kilimanjaro, Tanzania: Challenges 


Swift M.J. and Anderson J.M. 1993. Biodiversity and ecosystem function in agricultural 

systems. In: Schulze E.D. and Mooney H.A. (eds), Biodiversity and ecosystem function, 

pp 15 – 42. Springer, Berlin. 

Tesfaye Abebe, Wiersum K.F., Bongers F. and Sterck F. 2006. Diversity and dynamics in 

homegardens of southern Ethiopia. In: Kumar B.M. and Nair P.K.R. (eds), Tropical 


Science, Dordrecht.

370 

P.K.R. NAIR N 

Thaman R.R., Elevitch C.R. and Kennedy J. 2006. Urban and homegarden agroforestry in the 

Pacific islands: current status and future prospects. In: Kumar B.M. and Nair P.K.R. (eds), 




Environ 41: 189 – 207. 

Torquebiau E. and Penot E. 2006. Ecology versus economics in tropical multistrata 

agroforests. In: Kumar B.M. and Nair P.K.R. (eds), Tropical homegardens: A time-tested 


Vogl C.R. and Vogl-Lukasser B. 2003. Tradition, dynamics and sustainability of plant species 

composition and management in homegardens on organic and non-organic small scale 

farms in Alpine Eastern Tyrol, Austria. Biol Agric Hortic 21: 349 – 366. 

Watson J.W. and Eyzaguirre P.B. (eds). 2002. Home gardens and in situ conservation of plant 

genetic resources in farming systems. Proceedings of the Second International Home 

Gardens Workshop (17–19 July 2001), Witzenhausen, Germany. International Plant 

Genetic Resources Institute (IPGRI), Rome, Italy, 184p. 

World Commission on Environment and Development (WCED) 1987. Our common future. 

Oxford University Press, Oxford, 370p. 

Wills C, Harms K.E., Condit R., King D., Thompson J., He J., Muller-Landau H.C., Ashton 

P., Losos E., Comita L., Hubbell S., LaFrankie J., Bunyavejchewin S., Dattaraja H.S., 

Davies S., Esufali S., Foster R., Gunatilleke N., Gunatilleke S., Hall P., Itoh A., John R., 

Kiratiprayoon S., de Lao S.L., Massa M., Nath C., Noor M.N.S., Kassim A.R., Sukumar 

R., Suresh H.S., Sun I-F., Tan S., Yamakura T. and Zimmerman, J. 2006. Nonrandom 

processes maintain diversity in tropical forests. Science 311: 527 – 531. 

Westmacott R.N. 1992. African-American gardens and yards in the rural south. University of 

Tennessee Press, Knoxville, TN, 198p. 

Wezel A. and Ohl J. 2006. Homegarden plant diversity in relation to remoteness from urban 

centers: a case study from the Peruvian Amazon region. In: Kumar B.M. and Nair P.K.R. 

(eds), Tropical homegardens: A time-tested example of sustainable agroforestry, pp 

143 – 158. Springer Science, Dordrecht. 

Wiersum K.F. 2004. Forest gardens as an ‘intermediate’ land use system in the nature – 

culture continuum: characteristics and future potential. Agroforest Syst 61: 123 – 134. 




Adaptability 66, 67, 87, 123, 179, 275 

Adaptive management 137, 185, 283, 

293 

Adaptive research 299 

Aesthetic garden18 

African-American yards and gardens 6, 

7 

Agricultural landscape 65, 203, 235, 

271 

Agrobiodiversity (see also biodiversity) 

46, 55, 103, 117, 170, 173, 174, 

179, 200, 269, 280 

Agrodeforestation 25, 26, 34, 36, 39, 

41, 118 

Allotment gardens 9, 229, 317–319, 

321–323, 325–327, 329–337, 

368 

Amazon 10, 23, 43–60, 62, 64, 73, 83, 

84, 134, 139, 143, 144, 149, 

154–158, 160, 161, 163, 

165–167, 169, 170, 173, 175, 

179, 180–182, 189, 201, 204, 

206, 212, 213, 224, 225, 230, 

231, 267, 299–301, 304, 

309–312, 314–316, 338, 358, 

370 

Amazonia 23, 43–60, 62, 83, 144, 

156–158, 160, 163, 165–167, 

169, 175, 179–181, 189, 201, 

204, 230, 231, 267, 299, 301, 

314–316 

Amerindians 75, 84, 160, 165, 175 

Andes 145, 167, 170, 171, 173–175, 

181 

Animal husbandry 34, 247, 249 

Aromatherapy 221, 223, 228 

Aromatic plants 205, 208, 213, 226, 

226–231 

Benefit-cost analysis 283 

Biodiversity (see also 

Agrobiodiversity, Crop diversity 

and Species diversity) 8, 27, 39, 

41, 61–64, 69, 73, 74, 77, 80, 82, 

SUBJECT INDEX 

83, 87, 88, 101, 134, 137, 139, 

158, 181, 193, 199, 201, 203, 

231, 270–274, 279, 280, 284, 

320, 330, 339, 340, 356, 

361–363, 366, 368, 369 

Biodiversity conservation (see also 

Convention on Biological 

Diversity) 8, 77, 137, 199, 

270–274, 279, 280, 356, 369 

Biofactory 228 

Biopesticides 214 

Biopiracy 227 

Budget garden 20 

Caboclo 43, 44, 46–48, 50–52, 54, 55, 

58, 167, 182 

Canopy area 258 

Canopy conductance 258, 263, 264 

Canopy layers 251, 252, 260, 261 

Carbon sequestration 9, 61, 80–83, 

117, 119, 185–187, 189, 191, 

193, 195, 197, 198, 200–203, 

272, 274, 278, 281, 282, 356, 

363, 368, 369 

Carbon stocks 185, 188, 189, 191, 201, 

203 

Cash crops 9, 22, 38, 39, 90, 97, 100, 

102, 123, 124, 127, 128, 131, 

132, 134–138, 159, 162, 170–173, 

214, 233, 235, 236, 244–247, 

341, 352, 364 

Catalonia 3, 4, 6, 8, 211, 212, 224, 

228 

Cattle grazing 106, 111, 112, 117 

Chagga 2, 3, 9, 134, 135, 138, 139, 

140, 207, 213, 229, 358, 367–369 

Children 54, 57, 117, 162, 178, 181, 

227, 244, 248, 301, 311, 319, 

320, 325, 351, 362 

Classification, homegarden 23, 89, 90, 

92, 99, 108, 109, 112, 113, 127, 

229, 253, 320, 333, 353, 362 

Clean Development Mechanisms 

198–200, 272, 279

372 SUBJECT INDEX 

Cluster analysis 69, 87, 89, 90, 92, 360 

Commercial crops 18, 19, 47, 48, 50, 

55, 98, 123, 124, 133, 233, 235, 

240, 284, 293, 314 

Commercialization 7, 8, 15, 21, 22, 37, 

40, 60, 88, 101, 123, 124, 132, 

133, 135, 137, 138, 159, 165, 

173, 179, 214, 233–235, 237, 

239, 241, 243–249, 355, 356, 

358, 361, 364, 366, 368 

Commoditization 179 

Commodity production 94, 171, 172, 

186 

Community gardens 318, 319, 322, 

322–324, 326, 330, 331, 338 

Composition, homegarden 8, 9, 13, 15, 

17–22, 35, 40, 51, 53, 58, 61, 66, 

68, 69, 74, 81, 87, 88, 90, 

100–102, 123–132, 135–139, 

147, 148, 158–160, 165, 168–170, 

173, 176, 180, 186, 194, 197–199, 

201, 213, 224, 230, 231, 233, 

235, 240, 245, 248–250, 253, 

293, 311, 339, 345, 360, 368, 

370 

Contract farming 226, 228 

Convention on Biological Diversity 

187 

Correspondence analysis 69, 129, 130, 

360 

Cosmology 54, 161, 166, 167, 170, 

181 

Crop diversity (see also Species 

diversity and Biodiversity) 18, 

19, 23, 45, 124, 132, 136, 162, 

176, 229, 340, 343, 347–349, 

351–353 

Cropping system dynamics 105 

Diversity index (see also Shannon 

index, Shannon-Weaver index, 

and Margalef index) 94, 236–238, 

245, 246, 345 

Division of labor 159–162, 165–168, 

171, 172, 177 

Domestic animals (see also Livestock) 

1, 2, 20, 32, 48, 61, 76, 77, 81, 

206, 301 

Domestication 43–46, 54, 56, 57, 59, 

60, 73, 80, 82, 363, 369 

Dynamics, homegarden 9, 13, 15, 20–23, 

40, 82, 87–89, 91, 93, 95, 97, 99, 

101–103, 105–110, 112, 115, 

118, 123, 124, 126, 132, 135, 

137, 139, 158–161, 163, 165, 

167, 169, 171, 173, 175, 177, 

179–181, 197, 201, 229, 235, 

265, 275, 280, 300, 315, 336, 

339, 340, 348, 362, 368, 369, 370 

Ecological instability 247 

Ecological succession 360 

Ecological sustainability 78, 87, 88, 

101, 102, 135–137, 271, 274, 

280, 360, 361 

Economic alternatives 284, 285, 291 

Economic resilience 274, 284, 285, 

290 

Economic utility 283, 285, 287, 295, 

296 

Economic valuation (see also Financial 

worth) 356, 361 

Ecosystem services 194, 198, 272, 203, 

366 

Ecotourism 225, 228, 367 

Environmental economics 278, 279 

Environmental services (see also 

Ecosystem services) 62, 81, 198, 

269, 273, 278, 279, 281 

Ethnobiodiversity 39 

Ethnobotany 84, 180–182 

Ethnohistory 23, 43, 44, 46, 83 

Extension 10, 17, 21, 40, 43, 44, 56, 

57, 106, 108, 110, 111, 118, 119, 

173, 185, 273, 294, 299, 313, 

314, 317, 318, 326, 330, 331, 

333, 335, 336, 350, 364 

External inputs 87, 88, 91, 95, 98, 100, 

101, 135–137, 237, 243, 244, 

247, 248

Externalities 198, 269, 270, 272, 273, 

276, 278–280 

Extractivism 278 

Fallow 17, 26, 33, 46, 54, 55, 58, 65, 

84, 105, 106, 109, 111, 112, 115, 

117, 144, 145, 156, 162, 165, 

188, 201, 207, 225, 300, 341 

Farm size 123, 132, 133, 137, 343, 

347, 348, 353 

Farmer innovation 299 

Farmer-explorers 311 

Farmers’ practices 106, 112 

Farming systems 2, 9, 13, 15, 17, 19, 

21, 23, 103, 106, 124, 139, 158, 

230, 269, 273, 280, 311, 329, 

354, 370 

Financial worth (see also Economic 

valuation) 285, 287, 288, 290 

Floodplains 51, 143, 144, 156, 158, 

225 

Floristic composition 69, 81, 180, 245, 

249, 250, 360 

Food security 3, 25, 26, 39–41, 55, 61, 

62, 74, 75, 77, 81, 117, 119, 139, 

173, 202, 203, 266, 294, 317, 

326, 329, 336, 337, 357, 359, 368 

Forest fallow 188, 225, 300 

Forest garden 2, 17, 23, 24, 62, 139, 

140, 182, 226, 231, 252, 270, 

281, 282, 351, 353, 370 

Forest rent 273, 279, 280 

Fragmentation of holdings 133, 186, 

275 

Fruit trees 2, 3, 6, 7, 17, 26, 31, 35, 

43–47, 49, 51, 53–56, 61, 62, 65, 

66, 68, 72, 74, 75, 80, 94, 95, 98, 

110–113, 115, 143, 144, 148, 

149, 155, 172, 205, 224, 273, 

299, 300, 301, 312, 314, 320, 

342, 358, 363, 365 

Functional characteristics 91, 94, 99 

Functional differentiation 89 

Functions 6, 16, 18, 21, 23, 52, 69, 91, 

98, 117, 126, 129, 135, 136, 

158–160, 165, 168, 174, 176, 

SUBJECT INDEX 373 

234, 235, 240, 247, 248, 270–273, 

293, 319, 320, 342, 355 

Gender relations 159, 160, 181 

Genesis, homegardens 2 

Global distribution, homegardens 2–5 

Globalization 26, 227, 275, 

Handicraft plants 27, 30 

Health care 144, 177, 333 

Herbal medicine 206, 226, 227 

Hillside gardening 38 

Historic development, homegardening 

15, 17 

Historical accounts, homegardening 

45 

Homegarden management 53, 101, 

178, 351 

Homegarden model 318, 327–329 

Homegarden, types 2, 15–18, 20, 89, 

91, 92–100, 129–132, 237, 239, 

240, 245, 246, 330, 336 

Homegardens, extent 3, 4, 38, 186, 

207, 301, 325 

Importance value index 360 

In situ conservation 69, 103, 205, 339, 

340, 348, 352–354, 363, 370 

Incident radiation 61, 68, 80, 81, 251, 

252, 257, 260, 261, 263–265 

Indigenous knowledge 8, 41, 43, 48, 

62, 177, 180, 205, 225–227 

Indigenous people 46, 80, 143 

Institutional support 299, 313, 315 

Intangible benefits (see also nonmarket 

benefits) 8, 283, 284, 

286, 287, 293, 356, 361 

Integrated farming system 9, 16, 123, 

124, 129 

Intellectual property rights 227 

Intensification 2, 19, 20, 25, 101, 106, 

118, 195, 199, 201 

Intercropping 108, 118, 203, 229, 

300 

Internalization of externalities 269, 

280


Japanese-Brazilian 10, 60, 84, 300, 

302, 338 

Javanese 2, 15–18, 21, 35, 30, 41, 140, 

158, 187, 188, 201, 250, 271, 

275, 281, 302, 320, 353, 354 

Kandyan gardens 22, 266 

Keystone species 123, 136–139, 195, 

361 

Kitchen garden 18, 38, 40, 176, 319 

Knowledge transmission 177, 178 

Knowledge erosion 178 

Kyoto protocol 196, 198, 199, 272 

Land tenure 51, 53, 169, 170, 181, 272, 

274, 285, 325, 326 

Leaf area index 258 

Livelihood security 138, 198, 199, 327, 

337 

Livelihood strategies 15, 16, 24, 318 

Livestock (see also Domestic animals) 

4, 17, 26, 38, 53, 55, 62, 65, 76, 

125, 128–133, 135–137, 139, 

144, 161, 165, 200, 205, 207, 

213, 225, 230, 236, 237, 243, 244, 

270, 283, 286, 319, 324, 369 

Living edible pens 34 

Malnutrition 15, 41, 61, 66, 74, 77, 

328, 337 

Management intensity 17, 91, 95, 100 

Management practices 26, 56, 78–81, 

87, 90, 91, 95, 96, 100, 101, 137, 

180, 193, 199, 226, 249, 337, 

340, 349 

Management strategies 9, 61, 229, 282, 

283, 294, 317, 326, 330, 336, 368 

Margalef index 360 

Market access 19, 339, 341, 347 

Market garden 18, 20 

Market pressure 154, 233, 234, 364 

Market value 130, 275, 278, 284, 285 

Marketing 16, 19, 40, 56, 108, 118, 

136, 172, 173, 226, 274, 275, 

301, 314, 331, 335 

Market-oriented systems 364 

Matsiguenka 143, 145, 147–158 

Maya 61, 62, 64, 65, 68, 70, 72–74, 76, 

78, 80–84, 138, 157, 160, 162, 

163, 165, 169, 170, 172, 173, 

175, 176, 178–182, 199, 200, 

207, 226, 229–231, 250, 265, 

266, 353 

Medicinal plants 8, 16, 27, 28, 35, 36, 

46, 49, 68, 69, 91, 98, 142, 148, 

153, 155, 171, 175, 177, 205–207, 

211, 213–215, 224–231, 271, 

285–287, 301, 311, 320, 348, 

363, 364, 369 

Melanesian agriculture 105, 118 

Mesoamerica 2, 6, 44, 61–78, 80–83, 

135, 139, 160, 162, 163, 165, 

169, 170, 172, 182, 186, 202 

Mestizo 48, 158, 160, 170, 179 

Micro-economic approach 278 

Microgranjas 319 

Mixed gardens 17, 18, 235 

Modeling 252, 266, 269, 270, 276–278, 

280 

Modernization 22, 66, 87, 101 

Monocultures 187, 235, 245, 273, 300, 

364 

Multispecies systems 87 

Multistrata systems 1, 2, 68, 82, 88, 

105, 111, 116, 117, 138, 157, 

185, 187, 192–194, 196, 199, 

206, 225, 226, 229, 240, 245, 

250, 269–271, 273, 275, 277, 

279, 281, 294, 299, 313, 314, 

356, 361, 370 

Net primary productivity 185 

Nitrogen fixation 78, 82, 185, 196, 

294, 350 

Non-food plants 27, 29, 30 

Non-market benefits (see also 

Intangible benefits) 284 

Non-monetary benefits (see also 

Intangible benefits) 283, 291 

Nutrient cycling 55, 62, 67, 78, 80, 81, 

124, 135, 202, 269, 272, 274, 

280, 281, 340, 350, 361, 369

Nutrient outputs 194 

Nutritional security 7, 14, 117, 274, 

284, 287, 293, 294, 359, 361, 

362 

Nutritional value 16, 61, 117, 247, 

271 

Nutrition-related non-communicable 

diseases 26, 39 

Olympe 276, 280–282 

Opportunity costs 285, 288 

Organoponics 319 

Ornamentals 2, 18–20, 31, 36, 49, 71, 

90, 98, 99, 101, 124, 133, 134, 

162, 171, 211, 224, 342, 350, 

365 

Patrimonial analysis 277 

Patrimonial strategies 271, 277, 280 

Pekarangan 2, 3, 16, 17, 22, 186, 249, 

271 

Penman-Monteith equation 258, 263 

Periurban typology 337 

Physical environment 133, 234 

Phytochemicals 205, 228 

Phytoremediation 198, 274 

Pielou evenness index 342, 344, 345, 

352 

Plant diversity (see also Crop 

diversity, Biodiversity, and 

Diversity index) 17, 30, 62, 70, 

72, 143, 144, 158, 175, 181, 

193, 197, 202, 203, 232, 233, 

236, 237, 248, 339, 340, 351, 

370 

Plant genetic resources 55, 103, 340, 

348, 352–354, 370 

Policies 101, 199, 226, 249, 272, 275, 

335 

Population density 14, 27, 65, 66, 68, 

108, 127, 132, 133, 247, 275, 

341 

Poverty 25, 36, 39, 61, 65, 66, 74, 75, 

77, 80, 176, 198, 203, 312, 314, 

318, 328 

Profit 227, 283, 285, 287–293, 361 


Radiation interception 251 

Rainforest 59, 64, 83, 143–145, 149, 

156–158, 201, 315, 340, 341, 

354 

Regression 132, 133, 193, 258, 262, 

287, 290 

Residue management 61, 78, 81 

Resource utilization efficiency 196 

Ribereño 43, 44, 46, 47, 50–52, 54, 55, 

58 

Rights 32, 51, 168–171, 181, 227, 272, 

274, 326 

Rights to plants 170, 181 

Risk buffering 269, 271, 274, 276, 278, 

279 

Roadside gardening 69, 76, 326 

Rooftop gardening 319 

Rural development 22, 56, 61, 77, 90, 

102, 234, 299, 300, 314, 316, 

338 

Rural transformations 13, 15, 20, 21 

Sap flow gauges 251, 255, 265 

Sap flow rate 251, 256–259, 262 

Sapwood cross-sectional area 260 

Sensitivity analysis 283, 289, 290 

Shannon index 91, 93, 94, 342, 344, 

345, 347, 348 

Shannon-Weaver index /Shannon- 

Wiener index 105, 109, 110, 

111, 114, 116, 236–239, 360, 

Shifting cultivation 1, 2, 14, 16, 17, 20, 

21, 106, 143, 144, 158, 162, 225, 

228, 349, 360 

Slash and burn 75, 143–145, 188, 225 

Social and economic functions 235, 

320 

Social capital 249 

Social dynamics 82, 88, 124, 159, 229, 

368 

Social networks 175, 178, 179, 248 

Social status 159, 161, 174, 176, 179, 

272, 278 

Social sustainability 13, 22, 88, 101, 

102, 274, 355, 361, 362 

Social-ecological perspective 277, 281


Socioeconomic change 13, 14, 87–89, 

101, 123 

Socioeconomic sustainability 14, 15, 

87, 89, 135, 136, 366 

Soil biota 194, 195, 198 

Soil carbon sequestration 82, 197, 202, 

368 

Soil erosion 78, 203, 274, 340, 349, 

350, 352 

Soil fertility 19, 37, 56, 117, 118, 135, 

207, 270, 312, 339–343, 345–347, 

349–352, 356 

Soil organic matter 81, 185, 195, 197, 

274 

Sørensen’s Index 147, 148, 342, 345 

Spatial densification 318 

Species abundance 130, 174, 195, 197, 

330, 342, 360 

Species composition (see also Species 

diversity) 9, 17, 53, 58, 61, 66, 

68, 87, 100, 123, 128–130, 135, 

147, 148, 158, 186, 199, 213, 

224, 230, 240, 248, 311, 339, 

349, 370 

Species density 87, 90, 92, 342, 344, 

347, 351 

Species diversity (see also Species 

composition) 7, 57, 61, 69, 72, 

73, 84, 98, 101, 109, 111–113, 

116, 123, 128, 130, 131, 133–137, 

143, 154, 156–158, 160, 162, 

176, 185, 201, 203, 213, 225, 

231, 245, 250, 281, 282, 301, 

317, 330, 339, 340, 351, 353–355, 

357–362, 365 

Species introduction 49–51, 53, 56, 91, 

98–100, 214, 230, 233, 235, 300, 

302, 304, 306, 308, 310–313, 363 

Species richness 91, 109, 130, 132, 

133, 134, 137, 143, 154, 156, 

157, 186, 293, 343, 344, 347, 

348, 351, 362 

Staple crops 32, 38, 80, 165, 352 

Streuobst 6, 9 

Structural complexity 8, 61, 67 

Structure of homegardens (see also 

Vegetation structure) 13, 18, 22, 

66, 69, 89, 116, 173, 234, 237 

Structure, horizontal 61, 66–68, 69, 80, 

109, 111–113, 234, 

Structure, vertical 61, 66–68, 80, 89, 

105, 109, 111, 113, 116, 171, 

196,, 234, 241, 245, 252, 253, 

259, 274 

Subsistence crops 17, 57, 171, 271 

Subsistence garden 6, 18, 20 

Sundanese 16–18, 247 

Survival garden 20 

Sustainability indicators 339, 340, 352 

Sustainable production 72, 234, 311, 

315, 339 

Swidden agriculture 51, 143 

Swidden fallow/gardens 46, 58, 65, 84, 

144, 145, 160, 165–167, 174, 

176 

Talun-kebun 2, 16, 17, 22, 249 

Temporal dynamics 108, 110, 112, 

115, 348 

Thermal dissipation probes 251, 255, 

256 

Tokun 299, 311–315 

Totonac 70, 76, 79, 82, 171, 180, 181 

Traditional agroforestry 9, 13, 22, 23, 

25, 74, 83, 102, 103, 125, 139, 

231, 249, 250, 354 

Traditional homegardens 22, 52, 78–80, 

82, 87, 88, 100, 101, 103, 173, 

174, 180, 230, 234, 247–249, 

294, 355, 356, 361, 362, 365 

Traditional medicine 205, 206, 225, 

227 

Traditional vegetables 73, 247, 326 

Transmigrants 18, 353 

Transmigration 20, 342 

Transpiration 251–267 

Tree management 45, 69, 72, 73, 80, 

87, 89–91, 95–102, 137, 161, 

170, 176, 193, 195, 226, 228, 

274, 330, 331

Tree root systems 190, 193, 196, 197, 

201–203, 251, 252, 262, 264, 

366 

Tree tenure 170, 181, 280 

Urban homegardens 7, 8, 19, 229, 317, 

319–321, 323–325, 327, 329, 

331–338, 355, 364–366, 368 

Urban poor 317, 318, 328, 329, 331, 

337 

Urbanization 18, 25, 26, 179, 186, 235, 

318, 325, 333, 337 

Value addition 205, 225, 227, 228 

Vegetation dynamics 339 


Vegetation structure (see also 

Structure of homegardens) 17, 

90, 91, 98, 99, 105, 106, 108–110, 

112, 116, 124, 127, 137, 158, 

201, 240, 241, 248, 252, 340, 

353 

Whole-farm analysis 276 

Women 27, 74, 77, 159–182, 214, 216, 

227, 230, 274, 294, 319, 320, 

325, 338, 357, 359, 361, 362 

World Trade Organization 227 

Zonation 83, 103, 139, 158, 182, 230, 

250, 281, 294

1. 

Advances In Agroforestry 

P.K.R. Nair, M.R. Rao and L.E. Buck (eds.): New Vistas in Agroforestry: A Comp- 

st 

endium for the 1 World Congress of Agroforestry, 2004. ISBN 1-4020-2501-7 

2. Janaki R.R. Alavalapati and D. Evan Mercer (eds.): Valuing Agroforestry Systems: 

Methods and Applications, 2004. 

ISBN 1-4020-2412-6 

3. 

B.M. Kumar and P.K.R. Nair (eds.): Tropical Homegardens: A Time-Tested Example 

of Sustainable Agroforestry, 2006. 

ISBN 1-4020-4947-1 

springer.com

Tropical Homegardens - library.uniteddiversity.coop

Create successful ePaper yourself

Delete template?

Save as template?