Agriculture in Urban Planning

This page intentionally left blank

Agriculture in Urban Planning

Generating Livelihoods and Food Security

Edited by Mark Redwood

First published in the UK and USA in 2009
by Earthscan and the International Development Research Centre (IDRC)

Copyright © International Development Research Centre 2009

All rights reserved

ISBN: 978-1-84407-668-0

IDRC publishes an e-book edition (ISBN: 978-1-55250-427-7)
For further information:
International Development Research Centre
PO Box 8500
Ottawa, ON K1G 3H9, Canada
info@idrc.ca/www.idrc.ca

Typeset by 4word Ltd, Bristol
Printed and bound in the UK by Cromwell Press, Trowbridge
Cover design by Yvonne Booth

For a full list of publications, please contact:

Earthscan
Dunstan House
14a St Cross St
London EC1N 8XA, UK
Tel: + 44 (0)20 7841 1930
Fax: + 44 (0)20 7242 1474
Email: earthinfo@earthscan.co.uk
Web: www.earthscan.co.uk
22883 Quicksilver Drive, Sterling, VA 20166-2012, USA

Earthscan publishes in association with the International Institute for Environment and Development

A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

Agriculture in urban planning: generating livelihoods and food security
edited by Mark Redwood

p. cm.

Includes bibliographical references.
ISBN 978-1-84407-668-0 (hardback)
1. Urban agriculture–Developing countries. 2. Sustainable development–Developing
countries. I. Redwood, Mark.
S494.5U72A42 2009
630.9173’2–dc22

2008023954

The paper used for this book is FSC-certified and totally
chlorine-free. FSC (the Forest Stewardship Council) is
an international network to promote responsible management
of the world’s forests.

Contents

List of Figures and Tables

FIGURES

TABLES

Foreword

In 1998, the AGROPOLIS programme was launched by the International Development Research Centre (IDRC) to encourage young scholars to undertake research on urban agriculture (UA). In seven years, the programme granted more than 60 awards to graduate students from places as diverse as Argentina, the Congo and Canada, which were worth approximately US$1.2 million.

AGROPOLIS awardees have helped build the new field of UA and launch it into the mainstream as a viable tool for fighting hunger and poverty in cities. Their research projects have helped to generate new thinking in the field for the betterment of the poor. AGROPOLIS awardees have gone on to teach and continue UA research in academia (University of Guelph, University of Rosario), and their work continues to influence and develop municipal and national policies on UA. In other instances, UA has been incorporated as part of the work of international organizations such as the World Food Program. Others continue to be advocates in the NGO sector and with community-based organizations.

In 2005, IDRC published AGROPOLIS: The Social, Political and Environmental Dimensions of Urban Agriculture, which summarized the work of nine AGROPOLIS awardees. Due both to demand plus the availability of many excellent unpublished reports, need for a second volume became evident. This book describes the research of more recent AGROPOLIS awardees on topics ranging from food security, livelihoods and community building.

AGROPOLIS has now evolved into ECOPOLIS – a new programme that will build on the former programme’s success. ECOPOLIS drives IDRC into a new direction: support for design. ECOPOLIS will continue to fund scholarly research, but it will also encourage engineers, architects and planners to create and build innovative solutions to poverty in cities. IDRC’s Cities Feeding People programme has evolved into a broader urban poverty and environment problematique. In addition to urban agriculture, ECOPOLIS funds research on water and sanitation, solid waste management, vulnerability and land tenure.

IDRC is proud to have supported AGROPOLIS and is convinced that funding graduate research is an exceptional investment that unleashes the knowledge, creativity and enthusiasm of young researchers. We are very proud of our AGROPOLIS awardees and hope that you will find their research as enlightening as we have.

Naser I. Faruqui

Program Leader, Urban Poverty and Environment
September 2008, Ottawa

Acknowledgements

The AGROPOLIS awards programme is the result of the efforts of many individuals and the support of many institutions. Originally conceived by the Cities Feeding People (CFP) programme of Canada’s International Development Research Centre (IDRC), AGROPOLIS continues within a new programme initiative created in 2005, Urban Poverty and Environment. As the team leader of CFP, Dr Luc Mougeot was the main voice that drove IDRC in the direction of urban agriculture (UA) in the mid-1990s. He deserves great credit for his efforts and vision.

The CFP programme team comprised Dr Ola Smith, Naser Faruqui, Dr Daniel Buckles, Dr Ana Boischio, Brenda-Lee Wilson, Kristina Taboulchanas and Karen Trebert. Wendy Storey and Liliane Castets-Poupart managed the daily functioning of AGROPOLIS and became a welcome source of support and advice for many of the awardees.

AGROPOLIS awards were adjudicated by a committee of international professionals who have dedicated significant portions of their careers to furthering the field of UA. Over the years, the following individuals provided their insights and overviews to the programme and ensured its rigour: Dr Beatriz Canabel Cristiani, Dr Chris Furedy, Dr Diana Lee-Smith, Professor Godfrey Mudimu, Dr Donald Cole, Dr Tony Binns, Dr Paule Moustier, Dr David Midmore, Dr Raphael Yuen, Dr Juan Izquierdo, Dr Barry Shapiro and Dr Daniel Buckles.

The awardees have received the support of a large number of academic supervisors who steered the daily research activities of every awardee. While too numerous to name, these individuals have acted as champions of UA within their institutions. In addition, they continue to work with students to improve their research skills and to encourage an ongoing interest in UA.

Dr Donald Cole, Dr Joe Nasr, Dr Nancy Keranja and Dr Guéladio Cissé are particularly thanked for their role in helping to provide editorial advice on the content of the chapters, while Eric and Katharine Fletcher aided in the style editing. I am grateful for their help and collegiality. IDRC’s Urban Poverty and Environment (UPE) team, in particular Naser Faruqui, has been supportive of our awards programmes from the outset and has been available to discuss and to offer sound advice. On an ongoing basis, Nicole Mayer assists our entire team and also helped a great deal with this book.

Finally, without the help of three colleagues in particular, this book would not have been possible. First, I must thank Alison Clegg and Anne-Marie Legault, whose organizational efforts kept many things moving along. They ensured that the ‘t’s’ were crossed, and the ‘i’s’ dotted. Bill Carman, who manages IDRC publications, has always a sober and well-informed perspective and, in fact, offered the initial suggestion that a second volume of AGROPOLIS papers would be a welcome addition to the literature on the topic.

Mark Redwood

Senior Program Officer, UPE
September 2008

List of Acronyms and Abbreviations

Introduction

Mark Redwood

Urban agriculture (UA) has long been dismissed as a fringe activity that has no place in cities; however, its potential is beginning to be realized. In fact, UA is about food self reliance: it involves creating work and is a reaction to food insecurity, particularly for the poor. Contrary to what many believe, UA is found in every city, where it is sometimes hidden, sometimes obvious. If one looks carefully, few spaces in a major city are unused. Valuable vacant land rarely sits idle and is often taken over – either formally, or informally – and made productive. Urban agriculture is a long-established livelihood activity that occurs at all scales, from the small family-held market garden to the large agri-business located on the fringe of a city. It supplies food to the city and income to those who farm. Above all, UA is making an important contribution to food security for those who do not have easy access. In essence, UA is the true realization of the statement that ‘necessity is the mother of invention’.

In the 21st century, food comes with baggage. Mechanized farming and the increased yields associated with fertilizer and pesticide usage have reduced employment. Accordingly, farmers are relocating to cities in search of work. As wealth spreads, appetites change, and food is travelling further and further from where it is produced as people demand specialty goods. While food choices increase for the wealthy few, others are exposed to nutrition and health risks because of their lack of secure food sources. Market changes associated with biofuels, high oil prices and inflation are raising the cost of basic goods, which leads people to seek alternative ways to secure their food.

Meanwhile, the historic separation of the uses of ‘urban’ land from what has traditionally been considered ‘rural’ has relegated UA to a position of being a minor economic sector at best or irrelevant at worst. In general, policy has followed suit. Many cities, for different reasons, have ignored the contribution of UA and settled on disingenuous prohibition of the activity. But this is changing for the better, since acceptance of UA is growing in many municipalities (Mougeot, 2006; Van Veenhuizen, 2006).

The fourth World Urban Forum in 2006 showcased the crucial importance of UA in cities of the 21st century. During the forum, statistics were presented which show that by 2006 more than 50 percent of the world’s population is living in urban areas. Moreover, projections indicate that by 2050 it is expected that two-thirds of humanity will live in urban areas. Thus, the Forum confronted delegates with the challenges of such a rapid and historic change in human geography. The Forum was also notable for bringing UA in from the fringe and introducing it during a major international event whose attendees included mayors, government ministers, international organizations, researchers and members of civil society. Urban agriculture was the main topic in a number of networking events, product launches and in the booths of at least 20 institutional partners and eight cities. Around 1000 delegates attended the networking events that took place specifically on UA. The acknowledgement of UA and its presence at such a major event is indicative of wide changes that are taking place with regard to the politics of how cities are viewed and how the value of land – and food production – is perceived.

ENTER AGROPOLIS

Between 1998 and 2005, the AGROPOLIS programme of the International Development Research Centre (IDRC) supported graduate research on UA. In 2005, AGROPOLIS: The Social, Political and Environmental Dimensions of Urban Agriculture (Mougeot, 2005) was published. This book contained ten chapters of primary research conducted by graduate students who did their research with support from IDRC. Much of the work presented in AGROPOLIS was unique because, for the first time, an audience was exposed to baseline work on UA in many parts of the world that had not previously been studied. Now, with more than 60 awardees and significantly more work to reveal and examine, IDRC has decided to publish this second volume.

The appeal of UA is that it is a subject that crosses many disciplines. In an age where research is challenged to deliver results based on an integrated approach to research, UA offers an uncommon breadth. This volume, for instance, contains research from engineers, agricultural scientists, sociologists, planners and geographers. It is IDRC’s view that participatory research, done in an environment that balances biophysical and social approaches, achieves aims that are beyond those of singularly focused research. This is indeed a challenge, and one that takes practice, but the effort is worthwhile. Also, in keeping with the spirit of AGROPOLIS, each researcher in this second volume had an association with a development group in order to increase the likelihood of the practical application of their work.

This introduction provides an overview of the chapters in this book and organizes them around key themes associated with UA. The first section represents a general background explaining the crises that the rise of cities presents – along with the opportunities. This is followed by a review of how UA represents a valid way to address and minimize the difficulties marginalized people face on a daily basis. Therefore, this introduction presents a synthesis, while the chapters that follow offer a more substantive look at each case.

THE RISE OF CITIES AND CITY FARMING

No longer isolated nodes of intense economic development and human settlements, many cities are now growing into large, continuous urban corridors containing millions of people. In cities, questions related to livelihoods and society are complex. In cities in developing countries, migration from rural areas is still the primary source of people. However, these days a second generation of children of migrants is increasing urban populations, making the competition for scarce resources even more fierce. Poverty in the fringe of cities and people being pushed into confined spaces inside cities has led to a situation where one in six people on the planet is living in a slum (UN-HABITAT, 2003), a fact that Davis (2006) uses to justify what he calls the ‘Planet of slums’.

The pace of urbanization is unprecedented. Economic migrants seek better opportunities and jobs in cities while many rural economies stagnate. The trend of urbanization (see Figure 1) is most notable in the rapidly expanding economies of China and India, as well as in areas of the world where there are significant population and environmental pressures without much economic growth. The Gulf of Guinea along the coast of West Africa, for instance, represents an extreme example of massive urbanization. In 1960, there were 17 cities with a population of more than 100,000; now there are more than 300 cities of that size (Davis, 2006).

In the 1950s and 1960s, the drive to modernize economies in the south shifted the onus of economic activity onto the city. An important social result of this trend was the attraction of people to places of investment and employment, which resulted in an out-migration from rural areas. Seen by liberal development planners as the way out of poverty, the transition of emphasis to urban development and the encouragement of industrial economic activity concentrated on cities. This became known as the ‘urban bias’.

In the 1970s and 1980s, a competing strategy became more popular, and development planning shifted to supporting rural livelihoods and decentralizing development efforts into smaller cities and towns. These towns were seen as important market centres for food and other goods produced in rural areas. Despite such efforts to revitalize rural areas, the pace of urbanization remains rapid. The sheer numbers of rural people moving to cities, and the growth of city populations due to high birth rates, continues to strain natural resources and the capacity of governments and states to cope.

Figure 1 Urban growth rates compared with the expansion of informal settlements in African cities

Settlements like Kieran (Nairobi), Pinkie (Dakar), Mainsheet Knars (Cairo) and Villa Maria Triumph (Lima) are testimony to the impact of rural to urban migration. As Figure 1 shows, the growth of unserviced areas, or ‘slums’, has far outpaced the overall growth of cities.

Many who move to urban areas do not find the jobs and opportunities they seek. Therefore, adopting UA is a common survival strategy used by the poor not only to deal with food insecurity and poverty, but also to organize with fellow citizens and improve the quality of life in their communities. Urban growth has also concentrated food demand in cities. Smit et al (1996) estimate that 15–20 percent of the global food output is grown in cities.

In times of geopolitical or economic crisis, UA has become one way to secure food. As it turns out, the macroeconomic climate has been a significant influence in upsetting food security throughout the world. Structural adjustment programmes (SAPs) in the 1980s led to widespread currency devaluations, price increases for basic goods and the removal of subsidies for food production. In their effort to stimulate economic growth, SAPs actually removed some of the critical lifelines of the poor, thus shifting the focus of some economic planners onto the informal economy – including UA – as the poor discovered ways to survive. More recent economic crises such as happened in Argentina in 1999–2002 dramatically altered the landscape of employment and encouraged a growth in UA accompanied by policy support, as evidenced by the Pirouette programme of the federal government and city of Rosario. Another example of thriving UA is the Gaza Strip: with 3600 people per square kilometre, it is one of the world’s most densely populated regions, with limited access to food imports.

Defining urban agriculture

Where there are people there is a market for food. In the eyes of planners, architects, politicians and developers, trained as many across the world are, in the arts of town planning by European colonial systems, farming in a city was considered a practice either to be discouraged or ignored. Meanwhile, UA thrived in confined spaces because people in search of a source of income who have access to land and water will practise it, regardless of restrictive policies.

Because of the history of agriculture in the European context, many people consider agriculture and the city as separate and distinct, but this has never reflected reality. Urban agriculture has always existed. Just as settlement patterns rely on good arable land in order to secure a nearby productive food source, as cities grow, their footprint paves over agricultural land. Therefore, a city’s expansion de facto means there is an ongoing need for more sources of food.

The classic and widely used definition of UA comes from Mougeot (2000):

Urban Agriculture is an industry located within, or on the fringe of a town, a city or a metropolis, which grows and raises, processes and distributes a diversity of food and non-food products, (re)using largely human and material resources, products and services found in and around that urban area, and in turn supplying human and materials resources, products and services largely to that urban area. (p10)

This definition links confined space production, related economic activity, location, destination markets (or home consumption) and the types of products produced in a dynamic interaction that can vary from one urban area to another. The breadth of this definition has not been challenged, and it influences the extent of research on the subject. Research, and increasingly policy, is now acknowledging that peri-urban and urban agricultural systems operate in a very different context than rural systems. Urban agriculture not only presents research conundrums associated with natural science (agronomy, pollution, water and soil quality), but also important questions of a social nature (land markets, rural to urban migration). Actually, UA research is needed so as to study policy and ‘technocratic’ responses in the form of planning, law and legislation.

Urban agriculture also nicely bridges the gap associated with the unfortunate tendency for development research and practice to gravitate towards two poles, ‘urban’ vs ‘rural’. This polarization has missed some of the important nuances that occur in what some are calling the ‘peri-urban interface’ (Allen, 2003). This notion is important to our understanding of UA since it broadens the notion to include those distinctly non-rural agricultural systems located on the outskirts of cities. In fact, the term ‘UA’ has now expanded to include peri-urban areas as part of its context. The reader will note that in this volume several chapters draw on both ‘UA’ and ‘peri-urban agriculture’. This is because some authors use the terms to differentiate forms of UA that take place in the downtown core rather than the periphery of cities.

Why urban agriculture?

Perhaps the most dramatic argument in favour of UA comes from existing data on the proportion of income that city dwellers spend on food. Table 1 illustrates the need to find more reasonable sources of food and provides a strong argument for UA as a household supplement that can counteract the worst effects of poverty. Its effectiveness is not limited to poverty reduction at the household level: it also creates economic spin-off industries and employment, plus it improves the urban biophysical environment (Moskow, 1999).

Urban agriculture also acts as a catalyst for political organization. A survey of producer groups in 2005–2007 identified that organizations based around UA play a significant role in social cohesion, offering technical training and providing a platform for political lobbying. Such groups have successfully lobbied for municipal policy change (Amsterdam), acted as the voice of farmers to lobby for recognition (Dakar, Villa Maria Triunfo) or provided technical assistance (Montreal) (Santanderau and Castro, 2007).

The administrative limits of a city also play a role in determining the extent of the practice. IDRC-supported work in Latin America demonstrated that within the city limits in that region’s cities, there are large areas of vacant land. For instance, in Quito, Ecuador, 35 percent of city land is vacant and often being used for agriculture (2001 data). In Rosario (2003 data), the amount is 80 percent (IDRC, 2004). Recent data from Abomey

Table 1 Percentage of income spent on food by low-income residents in selected cities

Source: Akinbamijo et al (2002)

and Bohicon, two cities located in Benin, West Africa, shows that agriculture is the main activity for 3–7 percent of people living in their downtown core. However, 6 km from the city limits, in the peri-urban area, the percentage grows to 50 (Floquet et al, 2005). In the five urban districts that make up downtown Hanoi, 17.7 percent of land is used for agriculture (Mubarik et al, 2005).

The status quo response of governments in their reaction to UA has tended to be to prohibit the practice. Often this is a policy that stems from simply regarding UA as a form of resistance to urban development priorities as determined by planners. Some cities have, by virtue of being exposed to UA and farmer groups, changed their perspective and put in place systems that are designed to support UA, or at least remove the most draconian restrictions on the activity. However, even when rules are in place, they are often not well understood or enforced. In this volume, Mutonodzo points out that in Harare 40 percent of the people practising UA were unfamiliar with any laws related to it. Moreover, one in five considered the existing legislation to be hostile towards the practice.

Nonetheless, progress is being made. The number of municipalities that have policies in favour of UA has increased dramatically in recent years. Accra, Beijing, Brasilia, Buluwayo (Zimbabwe), Governador Valdares (Brazil), Havana, Hyderabad, Kampala, Rosario (Argentina) and Nairobi are a short list of a growing number of cities that are being proactive on the topic. Another popular way of supporting urban farming has been food-policy councils. These represent an increasingly common way of bridging community groups with municipal politicians and bureaucrats. Amsterdam, Toronto, Vancouver, London, Detroit and Pittsburgh all have councils that encourage locally-based food systems.

The efforts of researchers, farmers and other institutions have been built upon by visionary political leaders. In Accra, for instance, UA policy has been developed and the deputy minister of food and agriculture has signed a statement of vision on UA, and the government is giving awards to UA farmers to help implement better practices associated with UA. In 2003, the Harare Declaration was signed. Represented by ministers of local government and agriculture, it commits their intention to develop UA in five African countries (Kenya, Malawi, Swaziland, Tanzania and Zimbabwe).

The Food and Agriculture Organization (FAO) now readily accepts UA has the potential to strengthen urban food security. A project of FAO conducted with IDRC worked with farmer organizations in Phnom Penh, Harare and Cairo amongst other cities to develop first-hand knowledge of how farmers organize around UA to lobby for their rights to land, water and safe food (FAO, 2007). These groups often exist as a reaction to the pressure of development and land markets that presses UA to the margins.

It has taken some time to develop appropriate policy responses, but where city governments are supportive the reaction has been very strong. The development of the ‘farm to fork’ approach by the International Water Management Institute (IWMI) is helping policy makers realize where interventions are most useful (IWMI, 2008). By analysing the chain of production through the lens of the farm-market-consumer continuum, policy makers can apply specific targeted interventions with maximum impact. For instance, interventions at the market level are an important policy entry point for municipalities. Consider, for example, the results from Soto in this volume that show an increase in microbiological risks from washing vegetables with dirty water. This helps us realize the importance of well-managed health campaigns that articulate best health management practices from the production of food through to the market and onto the consumer.

Urban agriculture, once a fringe topic, could now arguably be considered a field in and of itself due to the combined efforts of numerous organizations. What is most evident is that UA has become popular mostly because it makes enormous sense to many people. Some disagreement on the influence of UA still exists, but the logic that food will naturally be grown close to the market for that food is not debated. The contents of this book add fuel to the concept that UA, as a livelihood strategy, will continue to exist as long as the demand for food exists. Moreover, the importance of UA is such that it should not be left in a policy vacuum. Its positive and negative attributes need to be confronted by well-informed policies and programmes.

OVERVIEW OF THIS BOOK

The content of this book stems from primary research conducted by AGROPOLIS awardees in Africa and Latin America between 2002 and 2005. In Africa, research was conducted in Ghana, Senegal, Malawi and Nigeria. These chapters present work on urban food systems, urban planning, the use of participatory methodologies and wastewater use. Areas where the amount of research is relatively limited such as Zimbabwe and the Democratic Republic of Congo are included. A study in Kenya focused on household food security and nutrition in a poor area of Nairobi. In Latin America, two papers explore the use of UA and wastewater in Lima, Peru. A chapter written from work in Nicaragua (Managua) explores social ecologies and networks though a review of home gardening practices, while another chapter presents initial findings related to the health of farmers in peri-urban Rosario (Argentina).

In order to maximize the possible impact of research, the AGROPOLIS programme insisted that awardees associate their work with ongoing action, programmes and projects in the field, linking their research with practice. This type of relationship acknowledges the role UA can play in addressing poverty, while recognizing UA practices must minimize some negative environmental and health impacts. Urban agriculture is only now entering the mainstream and support for policy development is still required. By encouraging work on UA, these researchers are encouraging UA to be recognized as a viable commercial activity, thus reducing the possibility that its practitioners will remain marginalized.

Access to resources: Land and water

Not surprisingly, the environmental inputs required for UA – land and water – are the focus of a number of the papers presented here. Both resources are scarce in cities and a lack of tenure and legal status forces poor farmers to work on marginal lands with water of low quality. Indeed, wastewater has attracted a great deal of attention in recent years (Faruqui et al, 2004; Redwood, 2004; IWMI, 2008). Considered valuable by farmers for its nutrient properties, raw or poorly treated wastewater poses a significant health risk to those that use it; therefore, it has often been banned. In his chapter on wastewater use in Ghana, Amoah points out that banning its use is now accepted as a failed policy. Soto’s research also illustrates that farmers rely on observed phenomena when assessing risk. In his study of farmers in the Rímac River watershed, Soto notes 75 percent of the farmers surveyed rated domestic sewerage as a significant problem – something they see every day. Meanwhile, only 11 percent considered mining a major concern, despite the prevalence of that industry and related pollution in the watershed. In her analysis of wastewater farmers in Dakar, Senegal, Chaudhuri observes that farmers in the area she studied were relatively aware of the dangers posed by pesticides. However, they were not concerned with pathogenic risks posed by wastewater despite the fact that many farmers exhibited illnesses associated with water-borne diseases.

Meanwhile, urban development and resulting pressure on the land is changing the nature and function of UA. Access to land is limited by increasing urban sprawl and upward pressure on land markets. In Lima, Maldonado documents a 26 percent reduction in land available for agriculture in Carapongo, a suburb of Peru, between 2002 and 2006. During this same period, the amount of land used for housing increased by 75 percent, placing a strain on farmers, particularly those who do not own their own property. Using statistical data from Zaria, Nsangu shows that UA was most often present on sites where planned developments had not yet occurred. In fact, he concludes that most of the land in Zaria’s master plan (year 2000) is actually being used for agriculture regardless of the original intent. In this case, UA is happening on lands that are designated for development, where construction has not yet started. As happens with many cities, the ambitious plans of politicians, architects and planners remain unfulfilled, allowing UA to fill the void.

The intensity of UA as a result of growing in limited space places stress on soil fertility and leads to the reuse of liquid and solid wastes as a way to capture nutrient value (Mougeot, 2006). These wastes possess an economic value which is not captured by formal markets and, therefore, offer a potential source of revenue to those who can provide them. This ‘closing’ of the nutrient cycle also serves to reduce the amount of waste destined for landfills or along roadsides. Several writers (Amoah; Eriksen-Hamel and Danso) tackle this issue of nutrient recycling and waste use.

In Kumasi, Ghana, research documented farmers’ willingness to pay (WTP) for both compost and poultry waste as fertilizer (Eriksen-Hamel and Danso). These authors found that both compost and poultry manure are highly valued by farmers. The benefit of the use of compost is supplied by increased yields: for instance, both tomato and lettuce yields increase by 12 percent when compost is applied. In terms of poultry manure, its inherently high nutrient content makes the potential development of a natural fertilizer market very appealing. They also note that because chemical fertilizers are much more expensive, farmers were willing to pay a considerable amount for poultry waste. Also, according to the WTP survey, compost fetches a price of up to US$3/50kg amongst vegetable farmers. This proves that, in the right quantities, compost can be supplied for profit. In urban areas, farmers are more reliant on highly valued and plentiful poultry manure – as an example, there are 300 poultry farms in Kumasi.

The same chapter on Kumasi, Ghana, emphasizes that the importance of research on these economic questions cannot be understated. If economic benefits are made clear, farmers will opt for what they see as the most beneficial path. In fact, they will employ production methods even if there may be associated health risks. Therefore, it is important for a policy approach to have the twin objectives of enhancing economic benefits along the chain of UA production, while also being creative in regards to how to promote hygiene considering that health risks are often imperceptible to farmers.

Gender analysis, participation and innovation in methodological approaches

Development research has also evolved to a point where subjective work can be ordered in a systematic way to enhance more empirical, positivist methods (Thomas and Mohan, 2007). Some traditional methods of qualitative and quantitative enquiry have worked well in UA research, but increasingly, analysis associated with power, ethical and moral issues is being used to corroborate positivist analysis. Being such a broad topic, UA requires variety in the types of methods used in enquiry. AGROPOLIS awardees were encouraged to make ‘action research’ the cornerstone of their methodology and approach. The premise of this was that universities and research institutes working in partnership with non-governmental organizations (NGOs) and community-based organizations (CBOs) tend to support the link between the generation of new knowledge and the field implementation of that new knowledge. In fact, however, neither of the extremes are ideal as stand alone approaches. Research can benefit a great deal when scientists ask questions about political ecology, while social analysts attempts to include more scientific rigour.

In the first book presenting AGROPOLIS awardees’ work, Mougeot (2005) noted that awardees were working at different places in the research cycle. Some researchers were presenting baseline work on UA because their cities had only performed limited work, previously. In this book, several chapters are in locations where information on UA is basic at best, or non-existent – Zaria and Kinshasa are examples of this. Meanwhile, other chapters are building on an already considerable knowledge base, as is the case in Kumasi, Rosario and Harare where baseline research has already been done. As a result, the methodology in each chapter differs and we are exposed to a broad range of analysis. Standard quantitative methods are used in some of the work while in others, participatory methods are employed to explore a very specific aspect of UA. On one hand, having a systematic approach across all of the research would have yielded some strong comparative evidence; however, the richness of this breadth of approaches illuminates different learning within each context.

Ten years ago, hardly any data existed on UA, such that traditional forms of enquiry were employed to generate baseline data. Now, the field has moved into a new phase where more sophisticated methods are employed and research is able to answer the questions raised by earlier work. In this book, traditional methods of enquiry are interspersed with some attempts at new methodological approaches. This section of the introduction attempts to tease out some lessons from this variety.

Participatory learning and action

One clear trend in development research in the past two decades has been the rise of popularity of participatory methods in social science research. At its most basic, participatory research implies that researchers are engaging with ‘beneficiaries’ in a way that will lead to an empowerment of that group. Participatory research also implies that the group researched will have more ownership of the project and its results and is notable as a way that research can empower communities and not simply be a process of extraction.

Nevertheless, there are voices that critique the approach as overly reliant on information-gathering from sources that are already powerful and entrenched politically (Thomas and Mohan, 2007). This critique also extends the fact that in the participatory research process, the researcher has significant leeway to subjectively interpret results. However, this can be very misleading as local knowledge is not infallible or without its own political baggage. For this reason two things are necessary in participatory research. First, ensuring that the participatory research process is done in a systematic way, and second, to ensure that results using participatory techniques are triangulated with results acquired using other methods. To work systematically, researchers should be active in developing their background research, typologies, categories, interview questions and so on in advance, but should incorporate some flexibility in the process to adapt as required.

Critiques aside, one advantage of PLA is that it provides a context for the use of other analysis such as quantitative methods that are used in biophysical research. Also, PLA can help to ensure that the right questions are being asked. In terms of UA, the following are examples of particularly relevant questions: Who is growing what and why? Are the concerns of the local people the same as those of the researcher? What economic factors influence their decisions? Which institutions are involved and what role do they play?

Participatory education and action research

Most of the research conducted in the AGROPOLIS programme included in this book uses PLA or a variant of this method such as participatory education and action research (PEAR). Chaudhuri’s chapter is devoted to the presentation of PEAR as a method of uncovering health-related information. PEAR requires a high degree of involvement on the part of the researcher, in order to untangle power dynamics and lay the foundation for advocacy on the part of the farmers’ group with whom she worked. While this experiment can yield very interesting results, assessing the transformative nature of the researchers‘ involvement is more difficult.

Others made use of gender-disaggregated participatory workshops to explore the role of women in UA (Soto), to develop a historical timeline of the community, for mapping (Shillington) and to analyse feasible alternatives to UA for income diversification.

Notwithstanding the critique, small investments in participatory research can make a difference. Njogu’s work is an example. In her chapter, she illustrates that her research involved not only a baseline but also an intervention in a subset of households within her sample. Her activities – an attempt at raising dietary diversity – actually led to an increase in the number of crops consumed with whom she worked. For calories, protein, vitamin A and iron, there was a statistically significant increase in the number of children achieving the recommended daily allowance.

The livelihoods framework

Linking research with policy requires that research be multidisciplinary and involve as many stakeholders as possible. As a way of tackling the multiple stakeholders and disciplines, several authors used the livelihoods framework in developing their research protocol. The basis of the framework is analysis that is people-centred and looks at structural issues such as access to resources, human capacity, social and political assets (Buelcher, 2004). It is designed to capture the interplay between development, environment and livelihoods.

In this volume, some authors use the approach (Soto, Maldonado) as a way to capture information using a diverse set of methods and stakeholders. In Maldonado’s work, the livelihoods approach was important in teasing out detailed information about the motivation and situation of farmers in Carapongo, Lima. Maldonado is right to have identified self-perception and esteem as important elements of human capital that can impede (or facilitate) development. As part of his research, he asked farmers about their self perception. Perhaps not surprisingly, a lack of land was cited by one farmer as the main reason for having a ‘poor’ perception of oneself. On the other hand, access to basic services and other sources of income were reasons for an improved self perception. It will come as no surprise to those who have worked in field research that few participants described themselves as ‘poor’. Although the sample was small, the effort to explore self-esteem amongst farmers suggests that future research can consider this as an enabling factor when designing strategies in a participatory manner of how to deal with problems associated with UA.

One lesson that can be drawn from the work involving the livelihoods approach in this book is the effective support it can offer a researcher trying to categorize multiple sources of information. By disaggregating our understanding of poverty into the five different categories of assets, it becomes a flexible and detailed framework where both quantitative and qualitative information can fit. Using a livelihoods framework also allows a researcher to corroborate learning and needs analysis associated with different stakeholders.

Gender

It is well known that women and children are important actors in UA in terms of production and marketing as well as in associated composting and recycling activities. Women also face problems in accessing some services such as credit and extension and training services, and also in access to legal or customary land tenure – in fact, anything associated with asset ownership (Hovorka and Lee-Smith, 2006). In research, gender-disaggregated data is frequently provided, but often with little actual analysis.

In this volume, several authors present data on gender and UA. Mkwambisi did some focused analysis with female-headed households. He suggests these homes are more efficient farms and notes that women are highly represented as UA farmers. In the study conducted on UA in Harare, the dependency ratio (defined as the proportion of adults to dependants, young and old) was found to be significantly higher in female-headed households than in male-headed households, corroborating earlier evidence from sub-Saharan Africa SSA (Maxwell et al, 2000). This suggests more challenges in achieving food security in households headed by women as they are expected to cope with more mouths to feed.

Evidence presented in this book also highlights improved efficiency in households practising UA headed by women. Female-headed households are more efficient at farming (Mkwambisi) and also over-represented as UA farmers. Maldonado’s work in Lima found that on average, women spend seven hours working in agricultural activities and another eight hours in household activities, whereas men spent nine hours working only in agricultural activities. In addition, Maldonado found that women are more present in the marketing of UA produce. Thus, the labour-intensive activity of urban farming is closely associated with the classic model of the ‘double day’. Mkwambisi demonstrates that female-headed households and poorer households benefit significantly more in terms of financial gain than wealthier households for whom UA is considered more of a ‘hobby’.

Gender analysis requires asking questions that relate to social and cultural norms and behaviours that are not always adequately captured by scientific enquiry. Shillington’s work presents the notion of ‘social ecologies’ through which men and women interact. In her mapping of these ecologies in patio and balcony gardens in Managua, she found that women identified and used a greater diversity of species than men, suggesting that they pay more attention to the patios. As she puts it, ‘patios tended to reflect women’s desires and needs’. This corroborates gender analysis of women being more active in private spaces in contrast to public ones.

Constraints facing researchers

In some parts of the South, conducting research is becoming increasingly difficult. More and more people are unwilling to participate since research too often is not seen to lead to a tangible change in one’s livelihood. The distance between scientists and researchers is often huge and difficult to surmount (Bentley, 1994). Stories of researchers making promises that are never fulfilled abound. There is a common expectation that study will lead to change, and while this is sometimes the case, it is not always true. Njogu describes challenges encountered in her follow-up survey where some of those who had participated in the baseline analysis refused to be interviewed in the follow-up, because they had not been selected for the pilot intervention.

The challenge of gathering data on incomes and financial information is well known and was mentioned by a number of the authors. It was also suspected in some cases that some respondents reported lower incomes than they actually received in hopes that the researcher might be able to facilitate access to government programmes and subsidies. In his chapter on Lima, Maldonado points out that during his baseline survey of income sources related to their sources of employment, accessing people’s true incomes was difficult as many are hesitant for reasons of privacy. This suggests two things. First, that the ethics of research need to be tightened when it comes to accessing information on peoples’ financial situation and other sensitive matters. Second, it reflects the need to use better collection methods – or at least corroborate – income data, such as reviewing tax rolls, or using income-expense journals that are monitored on an ongoing basis.

A similar problem was recognized by Eriksen-Hamel and Danso in their WTP survey. To increase the accuracy of their results, these authors avoided direct questions about WTP and instead modified their approach by introducing a bidding game. In a workshop asking participants their WTP for compost in different hypothetical scenarios, the game was introduced as a way of making their conceptual notions more concrete, accurate and as close to reality as possible.

Research related to health presents those working with few resources and over a short period of time with specific challenges. Accessing health information and identifying relationships between UA activities and health can be difficult. The chapter by Propersi about peri-urban agriculture and health alludes to problems associated with gathering health information. Soto found that in order to understand the different ways that health risks might be present, the researcher needs to assess multiple data sources and perform pathway analyses to understand how contaminants reach different foods. In Propersi’s case, some respondents were hesitant to discuss health problems associated with working on farms for privacy reasons plus for fear of their employment situation. In order to overcome this, Propersi corroborated her observations with visits and data from health clinics in the area of study. She also used photographic evidence to highlight certain types of practices. Cross-referencing survey data with photographic evidence can be a useful way to triangulate information.

The social and economic implications of urban agriculture

The purpose of UA is often to raise income through sale of produce or to reduce expenditure associated with the purchase of food (see Table 1). Surprisingly, despite this perhaps obvious conclusion, there have been few direct and strong economic analyses conducted related to the value of UA (Mougeot, 2005). In fact, none of the chapters in this volume actually present primary data on the financial value of UA which suggests that it is difficult information to gather. Another interpretation is that from the outset, the researchers assumed enough was already known about the income potential of UA – where assumptions such as this are common in work of this kind. In fact, the lack of strong economic data on UA is a major weakness that is impeding its entry into acceptance, particularly by economic planners.

The contribution of UA to food security varies. In the two chapters presenting quantitative evidence on this, both found that households employing UA do see an increase in food security. For instance, according to Mkwambisi’s work in Malawi, 17 percent of household food entitlements come from UA. Mutonodzo’s research found that in households that practise UA, 76 percent of daily food requirements were met. This compared to households not practising it, where 65 percent of daily food requirements were met according to FAO standards. This suggests that urban farmers are increasing their dietary diversity and nutritional intake through UA. The implication for poverty reduction is important in several respects. For one, if restriction on UA were to be upheld, one can safely assume that the population of urban poor using UA would be worse off. Second, the burden and cost of replacing UA would fall to the state/city taxing resources which are, in many cases, already stretched. These findings on the contribution of UA to food security illustrate the value it makes to poverty reduction.

Another finding that was expressed in several papers is that the poor are more reliant on UA than the rich. Kifuani’s work on the UA sector in Kinshasa illustrates the poverty of some UA farmers. In his study, he found that most (54 percent) of the farmers’ daily expenditures were less than US$0.50 per day. Mkwambisi’s study in Malawi noted that while UA contributed 9 percent of high-income households’ livelihoods, it was the main livelihood activity of 42.5 percent of low-income households. While UA tends to be concentrated among the poor in aggregate numbers, he points out that the larger economic contribution comes from wealthier farmers who operate farming businesses that are capital- and labour-intensive. This implies that policy makers have a choice to make in their policy support depending on their strategic objectives: If it is to reduce poverty, policy needs to be oriented towards poor farmers. However, if the objective is to increase aggregate food availability, it should concentrate on the businesses of wealthier farm operators.

Nutrition and environmental health: Benefits and burdens

Much has been exposed by recent work on environment and health matters and UA in recent years. Its importance is clearly illustrated by the relevance some international organizations are placing on it. For instance, two of Consultative Group on International Agricultural Research (CGIAR) systemwide initiatives (International Water Management Institute and Urban Harvest) focus on supporting research on environmental and health matters related to the reuse of waste. Moreover, WHO’s new guidelines for irrigated agriculture (2006) make explicit reference to the breadth of the problem in urban areas because cities generate the most waste. Such guidelines also provide crucial epidemiological evidence, as well as management strategies, on how to safely reuse various wastes in agriculture.

Several chapters are devoted to exploring the link between UA, environment and health. Soto notes that using water from the Rímac River in Lima for irrigation represents a major source of risk for farmers. The primary concern is one that most cities are faced with: much waste is disposed directly into rivers. Fully 97 percent of samples taken from the river were above acceptable limits for the presence of fecal coliforms (FC). Such micro-biological risks can be extrapolated to other rivers that are not only repositories for waste, but also sources of water for agricultural use. Heavy metals, especially cadmium and arsenic, also constitute a health problem in the Rímac River, where their presence as a result of mining was found in vegetables grown in the watershed.

Amoah shares his research from a large body that has been developed by IWMI-Ghana on the topic of wastewater use. The level of pathogenic and pesticide contamination of produce at the farm level due to inappropriate irrigation methods is high. In produce sampled in three Ghanaian cities, both pesticides and FC levels were above regulatory limits, indicating significant health concerns. Part of the problem is that irrespective of its source such as from shallow wells, rivers or from sewerage, irrigation water is usually well above allowable limits. The author stresses that health interventions must be adapted along the entire ‘farm to fork’ continuum. Hygienic practices in a market, for instance, play a significant role in the quality of produce purchased and consumed ‘downstream’. Importantly, however, Amoah shows that most contamination takes place at the farm and overshadows any hygienic danger at the sale point. The implicit suggestion is that effective interventions related to health are needed at different points along the food chain: at the farm level, transportation, market and home. The WHO guidelines, released in 2006, are sure to play a significant role in the policy development related to UA.

The policy response to these types of environmental problems can be a mix of basic hygiene procedures – probably the most economically efficient – combined with more comprehensive regulatory policies. In fact, in the case of the Rímac River, Soto found that historical levels of arsenic have been reduced following regulation of the use of the heavy metal by miners. But, as many already noted, the challenge for hygiene campaigns to surmount is that of a lack of any perception of risk. For example, both Soto and Chaudhuri point out that farmers are not very aware of risks that are not observable. This poses a challenge to health educators and policy makers, who must design educational programmes that educate farmers about the risks posed by using wastes in agriculture.

The 2006 WHO guidelines for wastewater use in irrigation are oriented around a multi-barrier approach that is designed to reduce risk. For example, these guidelines envision that at different stages of the production-consumption chain, barriers preventing contamination can be adopted. Such barriers can encompass treatment, irrigation methods, on-farm prevention of direct contact with polluted water and interventions at the market.

Conflicts and other constraints

While UA offers enormous potential, it is a mistake to believe that producing food in cities is free of problems. Although a number of major constraints exist, three are particularly important: access to land, access to water and the absence of comparative economic values generated by UA in the face of other activities such as land development.

Access to land

In cities, land is at a premium. The high value of urban land often makes UA at any scale particularly vulnerable, especially in terms of how the poor can access space to grow food. In the Mkwambisi chapter, most farmers in Malawi indicate that for them, access to land is the most pressing constraint on UA. Land-tenure security determines the extent of investment and self-help put into a property. Without secure tenure, farmers face eviction, uncertainty and are less willing to invest in long-term solutions related to their properties.

Access to water

Access to sources of water for irrigation causes similar problems. In the absence of safe water for irrigation, and due to the high nutrient contents present, wastewater is highly valued by farmers. As Chaudhuri discusses in her chapter, accessing raw wastewater through breaking mains and other means, with little consideration for health consequences, is a common practice in Dakar. Farmers also value this resource and have organized resistance to efforts by others to impose and enforce rules around access. In Dakar this led to the formation of the Union des Producteurs de la Vallée des Niayes (Producers’ Union of the Niayes Valley, UPROVAN), a farmers’ organization which actively disputes any attempt to infringe on their land by developers and others.

Absence of comparative economic values

The third major constraint is the failure to adequately measure and account for the relative contributions that UA make to alleviate poverty and contribute to a city’s economy. Setting aside social and cultural ‘returns’ such as those outlined by Shillington in her chapter – returns that as yet have not been well measured by economists – UA does not produce the same economic value on land that would otherwise be realized as a result of housing or commercial development. The lack of high economic output and, importantly, the lack of any assets that can be used for collateral restricts access to credit and financing mechanisms that urban farmers need so as to purchase inputs and build their enterprises.

In Maldonado’s study, 53 percent of UA practitioners in Carapongo (Lima) are working on land to which they do not have title or are otherwise farming illegally. Moreover, the author found that although 14 percent had received formal credit, few farmers use the formal banking system because they have no collateral to offer. As is the case in many unregulated or slum areas, land ownership functions within an informal system and thus credit is also acquired informally, for instance through banks, family, credit unions or NGOs. Family sources of credit seems to be the most common, probably due to the tenuous nature of UA and the perceived risks associated with offering loans to UA producers.

Other factors such as the predominance of the poor in UA, the illegal use of lands for UA, balancing health issues with social equity and the large number of women involved also contribute to the contentiousness of UA. In Kinshasa, Kifuani highlights other conflicts between various parts of the production chain in UA. Problems between producers are mostly associated with plot boundaries, access to water resources, the perception of a lack of fair competition and problems linked to agricultural equipment or inputs. Kifuani also found that 18 percent of market gardeners have conflicts with consumers usually over prices, non-payment and various other misunderstandings. Perceptions amongst consumers can also be negative to the production of UA. For instance, those consumers who are aware of produce grown with wastewater are less inclined to purchase them or will only purchase them at lower rates.

CONCLUSION

Despite the significant challenge of being perceived as a problem, UA is now part of the debate on how to improve our cities. Policy recognition of UA is on the rise, and as it increases, so too does the level of effort amongst scientists and researchers to document the practice. Once ignored by researchers, UA now is a recognized field of research and an area that has seen some rapid policy development in the past 15 years. This is because UA supports livelihoods and generates an economic value from land that would otherwise be idle or vacant. Its economic impact reaches far beyond farmers: it stimulates employment further along the chain of production for those who market, transport and sell produce, as well as for those who provide the tools and other means of production (Mougeot, 2000). In addition, UA is a catalyst for community organization and for the greening of cities. It also increases dietary diversity because it enables a local source of fresh produce; and it plays a role in the use of solid and liquid waste.

In the first volume of work by AGROPOLIS awardees, Luc Mougeot refers to UA as the ‘the oxymoron par excellence’ for the simple reason that strong perceptions of what constitutes ‘urban’ do not include ‘agriculture’. This, however, is less and less the case. The increasing acceptance of UA is a significant fact which illustrates changing attitudes towards the city as a whole. AGROPOLIS awardees have been the vanguard of research onto this topic. Many have gone on to complete their PhDs on UA, to teach and research the subject in universities, to practise it themselves, and to deliver new programmes and projects that support the activity. Although UA may always have its detractors, it is now absolutely clear that UA elicits sympathy and support. As research improves, so too will our understanding of the pitfalls and opportunities that UA can offer.

REFERENCES

Akinbamijo, O., Fall, S. and Smith, O. (eds) (2002) Advances in Crop-Livestock Integration in West Africa, IDRC Ottawa, Canada

Allen, A. (2003) ‘Environmental planning and management of the peri-urban interface: Perspectives on an emerging field’, Environment and Urbanization, vol 15, no 1, p135

Bentley, J. W. (1994) ‘Facts, fantasies, and failures of farmer participatory research’, Agriculture and Human Values, vol 11, no 2, pp140–150

Buechler, S. J. (2004) ‘A sustainable livelihoods approach for action research on wastewater use in agriculture’, in Scott, C., Faruqui, N. and Raschid-Sally, L. (eds) Wastewater Use in Irrigated Agriculture: Confronting the Livelihood and Environmental Realities, CABI Publishing, London

Davis, M. (2006) Planet of Slums, Verso Books, London

FAO (2007) ‘Urban farming against hunger’, www.fao.org/newsroom/en/news/2007/1000484/index.html (accessed 28 February 2008)

Faruqui, N., Niang, S. and Redwood, M. (2004) ‘Untreated wastewater use in market gardens: A case study of Dakar, Senegal’, in Scott, C., Faruqui, N. and Raschid-Sally, L. (eds) Wastewater Use in Irrigated Agriculture: Confronting the Livelihood and Environmental Realities, CABI Publishing, London

Floquet, A., Mongo, R. and Nansi, J. (2005) ‘Multiple functions of agriculture in Bohicon and Abomey, Benin’, Urban Agriculture Magazine, vol 15, pp9–10

Hovorka, A. J. and Lee-Smith, D. (2006) Gendering the Urban Agriculture Agenda, in Van Veenhuizen, R. (ed) Cities Farming for the Future: Urban Agriculture for Green and Productive Cities, International Institute for Rural Reconstruction and IDRC, Manila, Philippines

International Development Research Center (IDRC) (2004) Optimizing the Use of Vacant Land for Urban Agriculture Through Participatory Planning Processes, Final Technical Report, Project number 100983, IDRC, Ottawa

International Water Management Institute (IWMI) (2008) ‘Agriculture, water and cities’, www.lk.iwmi.org/rthemes/AgricultureWaterCities/index.asp (accessed 28 February 2008)

Maxwell, D. G., Levin, C., Armar-Klemesu, M., Ruel, M., Morris, S. and Ahiadeke, C. (2000) Urban Livelihoods and Food and Nutrition Security in Greater Accra, Ghana, IFPRI, Washington, DC

Moskow, A. (1999) ‘Havana’s self-provision gardens’, Environment and Urbanization, vol 11, no 2, p127

Mougeot, L. J. A. (2000) ‘Urban agriculture: definition, presence, potentials and risks’, in Bakker, N., Dubbeling, M., Gundel, S., Sabel-Koschella, U. and de Zeeuw, H. (eds) Growing Cities, Growing Food: Urban Agriculture on the Policy Agenda, German Foundation for International Development (DSE), Feldafing, Germany, pp1–42

Mougeot, L. J. A. (2005) AGROPOLIS: The Social, Political and Environmental Dimensions of Urban Agriculture, Earthscan, London

Mougeot, L. J. A. (2006) Growing Better Cities: Urban Agriculture for Sustainable Development, IDRC, Ottawa

Mubarik, A., De Bon, H. and Moustier, P. (2005) ‘Promoting the multifunctionality of urban and periurban agriculture in Hanoi’, Urban Agriculture Magazine, vol 15

Redwood, M. (2004) Wastewater Use in Urban Agriculture: Assessing Current Research and Options for National and Local Governments, Cities Feeding People Report no 37, IDRC, Ottawa

Santanderau, A. and Castro, C. (2007) ‘Social organisations of agricultural producers in Latin America and Europe: Lessons learned and challenges’, Urban Agriculture Magazine, vol 17, pp5–7

Smit, J., Ratta, A. and Nasr, J. (1996) Urban Agriculture: Food, Jobs and Sustainable Cities, United Nations Development Programme (UNDP)

Thomas, A. and Mohan, G. (2007) Research Skills for Policy and Development: How to Find Out Fast, Sage, London

UN-HABITAT (2003) The Challenge of Slums: Global Report on Human Settlements, Earthscan Publications, London, UK

Van Veenhuizen, R. (ed) (2006) Cities Farming for the Future: Urban Agriculture for Green and Productive Cities, International Institute for Rural Reconstruction and IDRC, Manila, Philippines

World Health Organization (WHO) (2000) Water Supply and Sanitation Assessment, Part II, World Health Organization, http://www.who.int/water_sanitation_health/Globassessment/Global1.htm (accessed 2 January 2008)

World Health Organization (WHO) (2006) Guidelines for the Safe Use of Wastewater, Excreta and Greywater (vols 1–4), WHO, Geneva, Switzerland

1
Household Food Security Among Urban Farmers in Nairobi, Kenya

Eunice Wambui Njogu1

INTRODUCTION

Since the 1960s, residents of Kamae had lived in slum conditions, but in the year 2001 they were allocated small landholdings by the local administration of the Government of Kenya. Prior to the start of this research project, observation indicated that these landholdings could potentially increase the diversification and intensification of food-production systems. A baseline survey conducted between June and July 2005 revealed none of the 300 respondents surveyed had received any form of organized agricultural training. Therefore, no organized farming was observed. Because the extension service of the Ministry of Agriculture and Livestock and Fisheries Development was farmer-driven, only more prosperous farmers sought crop and livestock extension services. As a result, the majority of poor farmers in Nairobi manage on their own, getting no assistance or advice, which is necessary for the enhancement of their food production skills particularly in the diversification of agricultural systems (Foeken and Mwangi, 2000).

The government has been restructuring and partially privatizing the public extension system to improve extension-service delivery to all farmers regardless of their socio-economic class (Government of Kenya, 2004). However, according to Ishani and Lamba (unpublished work, 2007), urban and peri-urban agriculture and livestock (UPAL) is not recognized in this policy. Fortunately, the National Livestock Extension Programme (NALEP) has been introduced in Nairobi province to support farmers (Z. Ishani and D. Lamba, unpublished work, 2007).

Without the inclusion of UPAL as a policy issue in agriculture and livestock policies, the social and economic benefits UPAL has on vulnerable groups cannot be fully realized (Z. Ishani and D. Lamba, unpublished work, 2007). When well developed, UPAL plays an important role in cities because it improves the nutrition status of household members, generates income, provides employment and conserves the environment.

‘Stunted’, ‘underweight’ and ‘wasted’ are indices used by the World Health Organization (WHO) based on the measurement of height, skin thickness (fat) and age to determine the nutritional status of the subject (WHO, 1995). In most cases, children are used as they represent an indicator of the general well-being of a community based on their intake of food. Children who are wasted are considered chronically malnourished. The results of the baseline survey showed that 62 percent of children below five years old were stunted, 53.7 percent were underweight and 31 percent were wasted, whether mildly, moderately or severely. Their caloric, vitamin A and iron intakes were below recommended dietary allowances (RDA) established by the WHO. The assessment of people’s nutrition knowledge on such important macronutrients (carbohydrates and proteins) and micronutrients (vitamin A and iron) showed that 49.7 percent had no idea about them. Research in Korogocho, Nairobi showed higher energy intake among urban farmers than non-farmers (Foeken and Mwangi, 2000). The exposure of this lack of agricultural and nutritional knowledge combined with the observed poor nutritional status necessitated the design and implementation of an intervention. The intervention was to introduce the concept of growing a diversity of crops and rearing small livestock to enhance households’ food security.

As a result of this diversification of crops and livestock-rearing, household income increased in several ways. First, if they were marketed, the surplus food crops and livestock as well as their products fetched extra income. Second, the increased availability of food crops relieved the households’ income from food purchases. Then, due to this increased income, households’ food purchases became more diverse so people could incorporate higher quality foods in their diet (Bonnard, 2001). Therefore, households’ food security was enhanced by increased availability and access to diversified diets.

Selected respondents were trained on how to utilize their small land-holdings to produce a diversity of crops, rear small livestock and also to teach nutrition. This was done in collaboration with the Ministry of Agriculture and Ministry of Livestock and Fisheries Development. The households were provided with the fencing material, seeds, seedlings and small livestock of their choice. All this was expected to lead to enhancement of household food security as the initial step towards poverty eradication. This paper covers the objectives, methodology and findings of the research project, in addition to offering a conclusion and recommendations based on the findings.

Objectives

The five specific objectives of the project were to:

1 Determine knowledge and attitudes of the farmers towards peri-urban agriculture (UPA) in the households.

2 Assess farming practices among the farmers in the households.

3 Determine each household’s food security status in this area.

4 Assess the nutrition status of children below five years of age in the households.

5 Establish the relationship between farming practices and the household’s food security.

Hypotheses

There were two hypotheses:

1 There was a positive relationship between household food security and the farming practices adopted by the peri-urban farmers of Kamae.

2 There would be a significant improvement in household food security after the intervention.

METHODOLOGY

Research design

This was an intervention study project which was semi-longitudinal in nature with an intervention period of one year. The research adopted the participatory approach to nutrition security (PANS) using the Triple-A cycle, as shown in Figure 1.1.

PANS facilitated community members to map out their problems, design their solutions and take action to implement these solutions. A baseline survey was carried out to assess the food production and food security situation in Kamae settlement. Kamae area is located north of Nairobi in Kahawa Ward,

Figure 1.1 The (PANS) Triple-A approach

Kasarani Division. It borders the Kiambu district, which is in the central province of Kenya. Kamae, like Nairobi City, has a bimodal rainfall pattern. The long rainy season is from April to August, while the short rainy season is from late October to early December. Kamae has a total of 2036 households and a population of 9847 persons according to the 1999 Population Census (Government of Kenya, 1999).

Data were analysed to identify the food production and food and nutrition security problems. Those that were given first priority during focus-group discussions were lack of agricultural and nutrition knowledge, as well as inadequate supplies of clean, safe water. The action plan included offering education in agriculture and nutrition by the officers from the Ministry of Agriculture and Livestock and the Ministry of Fisheries Development, as well as from the researcher. Project participants each contributed approximately US$58 for the provision of water. The researcher, with support from the International Development Research Centre (IDRC), provided fencing materials, seeds and seedlings, as well as small livestock. An evaluation was done through a survey to ascertain the impact of the project.

Sample size

Three hundred households were studied during the baseline survey for situation assessment.

To obtain a representative sample, Fisher’s formula was employed (Fisher et al, 1998). That is:

where nf is the desired sample when the population is less than 10,000; n is the sample2 when the total population is more than 10,000; and N is the estimated population of the households in Kamae area (2036). Therefore, the sample consisted of 323 households:

However, for the purpose of this study, this figure was rounded down to the nearest hundred, that is, to 300 households. Cluster sampling was employed during the baseline survey to obtain these. Every respondent was interviewed in those households within the selected cluster that contained a child below five years of age. For the sample related to the household interventions, 10 percent of the 300 households (30) were selected randomly.

Three focus group discussions took place during the baseline survey, consisting of ten participants each, where one group consisted of men, one of women and one of youth. There was one focus group discussion during the evaluation consisting of those who were in the intervention project. Observations were made for the 300 households during baseline survey, whereas the evaluation survey included the 30 homes where interventions took place and a further 150 households. Evaluation covered 180 households because approximately 120 respondents had either moved or, for others, children were older than five years of age. Still others refused to be interviewed because they were not selected for the interventions.

Interview sessions, observations and focus group discussions

The interviews were carried out in a face-to-face situation using Kiswahili language. They were conducted during a baseline survey between June and July 2005 and at the end of the project (March–April 2006) during the evaluation process. Monitoring was continuously carried out throughout the duration of the project and observations were recorded on the observation checklist. The characteristics observed were: farming systems, labour providers, crops produced, tools used for food production, general sanitation of the household compounds and cooking facilities. Focus group discussions were held during problem analysis in September 2005 and also during evaluation on 16 April 2006. The researcher and research assistant guided and tape-recorded these sessions.

Awareness campaigns, demonstrations and intervention

To sensitize the community members on the objectives and importance of this project, an awareness campaign was held on 25 July 2005 in collaboration with the staff of the Ministry of Agriculture and Ministry of Livestock and Fisheries Development. The 30 households chosen for intervention were those that had children with high levels of wasting, and these households were divided into two groups, which were trained in two separate sessions. The first session was on organic farming while the second was on nutrition education with a bias on macronutrients (carbohydrates and proteins) and micronutrients (vitamin A and iron).

The first group was trained by the researcher and an organic farming expert in collaboration with the staff of the Ministry of Agriculture and Ministry of Livestock and Fisheries Development on 29 October 2005, followed by a demonstration on 2 November. The second group, which was trained on 16 November, was facilitated by one of the participants who had been trained in the first session, who was assisted by the researcher. The second demonstration was done on the same plot on the same day. Small livestock were kept on the two demonstration plots. The other 28 participants in the intervention were guided on how to plant crops and rear small livestock on their plots.

This study focused on the actual intake of calories, proteins, iron and vitamin A. The intakes were assessed using a 24-hour dietary recall of children who were less than five years old. Adequacy was determined by assessing those children whose intakes were below and above the RDA.

Data analysis

The quantitative data were analysed using the Statistical Package for Social Sciences (SPSS) programme. The food production systems were determined by assessing the farming practices adopted by the households. Anthropometric data were analysed using the Nutrition Package (EpiInfo, 2000). Those children with −2 to < −1 standard deviations were classified as mildly malnourished, those with −3 to < −2 standard deviations were designated moderately malnourished, while those below −3 standard deviations were considered severely malnourished. Anthropometric measurements were compared with the US National Centre for Health Statistic/World Health Organization (NCHS/WHO) international reference standards.

Household food consumption was determined by assessing the intake of calories, protein, vitamin A and iron. The nutrient intake data were analysed using computer software (Nutrisurvey, 2004). Dietary diversity was assessed by totalling the number of food groups consumed in the households for the past 24 hours. These groups are starch (cereals and non-cereals), fresh and dry legumes and nuts, vegetables, fruits, eggs, meat, fish and milk. Households were classified as follows: those consuming 1–3 groups were considered low diversity, those consuming 4–6 groups were medium, while those consuming 7–9 were considered the high-diversity group. The level of agricultural and nutrition knowledge and the attitudes of the farmers was assessed by totalling the scores in each household.

Transcription was done using tape-recorders to determine the problems encountered in crop and small livestock production and to capture the farmers’ attitudes towards urban agriculture (UA). The Pearson product moment (r) established relationships between farming practices and household food security; it tested the hypotheses at 0.05 significant levels. The t-test was used to evaluate the intervention project by comparing intakes of calories, protein, vitamin A, iron and levels of nutrition knowledge both before and after the intervention at 0.05 significance levels. Cross-tabulation was used to determine the changes in occupations of respondents, crop diversity, nutrition knowledge and dietary diversity in the households after intervention. The results are presented in tables and graphs below.

Research findings

Knowledge and attitudes

The farmers’ knowledge, attitudes and skills depended upon the training they received. Knowledge in certain production techniques such as crop protection and rotation led to high levels of production (Mougeot, 2000). The project participants were offered agricultural and nutrition knowledge.

Agricultural knowledge

The results of the baseline survey revealed that none of the 300 respondents had received any kind of organized agricultural training. Because of their lack of knowledge, a training programme on organic farming was developed covering soil preparation, crop management, harvesting techniques and post-harvest handling. Increases in agricultural knowledge was observed through improved farming practices.

Nutrition knowledge

The results of the baseline survey showed that of the 300 respondents, 49.7 percent had no knowledge of macronutrients (carbohydrates and proteins) or micronutrients (vitamin A and iron). Approximately 30 percent knew what carbohydrates, proteins, vitamin A and iron were, and could give their sources. However, out of the 30 respondents selected for intervention, 15 (50 percent) had no idea about any of the macro- or micronutrients. After training, of these 30 respondents who were taught about carbohydrates, proteins, vitamin A and iron and their sources, nine (20 percent) still showed no knowledge of them. After re-examining respondents’ educational levels, it was discovered they had never gone to school and, therefore, they had a limited ability to recall the topic of the training session.

Attitudes towards urban farming

The respondents were given statements that represented attitudes towards UA and were asked whether they agreed or disagreed with them. In the baseline survey, 97 percent expressed positive attitudes. All 30 households both before and after intervention had positive attitudes towards urban farming and all thought it represented a good thing. They recommended their neighbours should consider adopting it as it enhanced household food supply. Also, they noted that the government should support urban farming. Therefore, the research showed that stakeholders in the area of UPAL should collaborate with the urban and peri-urban residents to push UPAL into the national policy agenda.

Farming Practices

Household food production systems

According to the results of the baseline survey, there was either no farming, or else little or unorganized farming being done. Focus-group discussions revealed that residents could not grow crops on their plots because livestock destroyed them as they could not afford fencing materials. The most vulnerable households selected for interventions were provided with fencing materials and were only expected to provide labour for fencing, land preparation, planting and managing the crops.

The 30 households chosen for intervention had neither crops nor small livestock. After intervention, identified food-production systems included subsistence farming systems and market-oriented crop production, where 46 percent of the intervention households practised the latter type of farming. All 30 households produced crops and reared small livestock – such livestock were chosen because they required low inputs in terms of space and feeding. Moreover, the Nairobi City Council by-laws allow rearing of small livestock unless someone complains of a nuisance. In fact, during the intervention process, the 30 households received cockerels and chickens or ducks, which the project team delivered. In addition, the National Livestock Extension Programme (NALEP) has also been introduced in Nairobi province (Z. Ishani and D. Lamba, unpublished work, 2007).

FOOD SECURITY

Dietary diversity

Out of the 300 households that were in the baseline survey, 57 percent were in the medium group that had consumed 4–6 groups of different foods in the past 24 hours (Table 1.1). Similarly, out of the 150 households in the evaluation survey, 54.7 percent were also in this medium group.

Among the 30 households in the intervention, 30 percent had low dietary diversity of 1–3 groups before intervention, but after intervention shifted to the medium diversity of 4–6 food groups. Those 70 percent with medium diversity before intervention increased to 86.7 percent after intervention. Also post-intervention, 10 percent of households moved from medium diversity to high diversity where 7–9 groups were consumed. These results are associated with increased crop and dietary diversity due to interventions. High levels of dietary diversity ensure adequate intakes of essential nutrients such as vitamins and minerals, as well as increased nutrient density (Hoddinott and Yohannes, 2002).

Nutrient intakes by children <5 years

Improvements were noted in both caloric and protein intakes of the children belonging to the two groups, as presented in Table 1.2. These improvements could be associated with increased amounts of both food and income from the sale of the surplus produce.

There was also improvement in both vitamin A and iron intakes, as seen in Table 1.2. This was probably due to increased production and higher consumption of fruits. Consumption of vegetables and livestock products increases vitamin A and iron intakes. Research demonstrated that children in

Table 1.1 Dietary diversity based on RDA before and after intervention

Table 1.2 Nutrient intakes based on RDA before and after intervention

the urban households with self-provisioned food showed better health than do those children who do not have such access (Mougeot, 2006). Prior to this research, better health was discovered to be linked to improved nutrition brought about by increased dietary diversity and its accompanying increased nutrient intake.

Nutrition status of the 300 children

Of the 300 children observed during the baseline survey, 62 percent were stunted, 53.7 percent were underweight and 31 percent were wasted (the children were classified into one category at a time and overlapping of conditions were not established). The children with high levels of wasting were included in the intervention.

The nutrition status of children in the 30 households before and after intervention showed mixed results. However, there was a general improvement in their health as shown by reduced numbers both of children who were severely and moderately stunted, as well as those who were underweight (Figure 1.2).

Those children who were mildly wasted increased from 16.7 percent to 46.7 percent after intervention. Those who had normal weight decreased from 73.3 percent to 43.3 percent after the intervention.

In general, the above results do indicate changes. However, the results cannot necessarily be directly linked to the impact of the project because one

Figure 1.2 Nutrition status of children from households involved in the intervention

year is a short time for the project’s activities to influence the children’s nutrition status. Although adequate availability of food at the household level is necessary in order to achieve nutrition security, it is not sufficient. Other key contributors to good nutrition are adequate health, childcare, access to clean water and sanitation (FSAU, 2003).

RELATIONSHIPS BETWEEN VARIABLES

Between crop diversity and dietary diversity

A positive and significant relationship was found to exist between crop diversity and dietary diversity (r = 0:123 and P = 0:03), which shows that dietary diversity enhanced household nutrition and food security. Therefore, there is a need to promote production of a diversity of crops because it diversifies the households’ diets, thereby enhancing household food security.

Positive relationships existed between the levels of nutrition knowledge and intakes of calories, proteins, vitamin A and iron, as shown in Table 1.3. This may have been due to availability of vegetables.

Improvement in the household food security after intervention

There was a significant difference in iron intake before and after intervention t = 2:715 and P = 0:009 (Table 1.4). This could be attributed to increased consumption of animal proteins and dark green leafy vegetables.

Table 1.3 Nutrition knowledge and nutrient intakes

*Significant at 0.05 confidence level.

Table 1.4 t-test for differences after intervention

*Significant at 0.05 confidence level.

There were insignificant differences found in the intakes of calories, proteins, vitamin A and nutrition knowledge before and after intervention.

CHANGES IN AGRICULTURE AND LIVESTOCK EXTENSION SERVICES

The staff of Ministry of Agriculture and Ministry of Livestock and Fisheries Development became interested in the activities of the project and introduced the multi-storey gardening (use of sacks to plant crops) in other households. Moreover, strong cooperation between the residents of Kamae and the agricultural and livestock extension officers at the divisional level has continued. This is a positive move towards pushing UPAL into the agriculture and livestock national policy agenda.

Other benefits associated with the project

During the focus-group discussions, the problem of access to clean and safe water came out strongly. The participants came together and formed a community-based organization under the guidance of the project team. They each decided to contribute approximately US$58 to purchase water pipes and water meters, while 15 participants connected to piped water and the remaining 15 were due for connection at a later date. Many non-participants have also managed to access the same water through the efforts of World Vision’s Kahawa Area Development Programme. Non-participants developed an interest in interventions: they fenced off their plots and planted crops.

CONCLUSION AND RECOMMENDATIONS

The findings of the study are very significant because NGOs and other interested individuals in Kamae saw it as a worthy cause and assisted community members in other innovations, especially in the provision of clean and safe water. Also, gardens were introduced into this community. Moreover, adjacent communities asked to be similarly trained and as a result of this, World Vision (Kahawa Area Development Programme) sponsored a six-day training programme in Soweto and also in Ngomongo, in Kasarani, Nairobi. Through the Ministry of Agriculture and Ministry of Livestock and Fisheries Development, the government should implement more training to all willing urban dwellers.

Fifty percent of the households that participated in the interventions produced surplus produce they could sell. Thus, they became both farmers and business people. Thirty-six percent in this same group were involved in farming as well as being engaged as casual labourers in the neighbouring coffee plantations. Only 13.4 percent continued to be involved in farming without any other economic activity. Diversification of economic activities broadened the households’ financial base or capacity and, as a result, access to household food security was enhanced. However, there is a need to sensitize urban dwellers on the importance of diversifying their economic activities for their overall development, which would include farming.

Agricultural knowledge increased significantly, as evidenced in the improved farming practices and diversity of crops which became accessible to households. Organic farming was the preferred system of crop production because it does not require expensive inputs; hence, it is affordable to people of all socio-economic classes. Those households participating in the intervention phase experienced an improvement in their food security and this also indicated an increase in dietary diversity due to improved caloric, protein, vitamin A and iron intakes. Nutrition education was essential so participants could make informed choices in the selection of food for their household’s members and also so as to properly utilize available food.

Therefore, the research showed that when agricultural programmes are combined with well-designed nutrition education, significant changes in the participants’ consumption behaviour can occur. Also, when there is an increase in nutrition knowledge there is a related increase in intakes of important nutrients, particularly vitamins and minerals. Increase in consumption of micronutrient-rich food is notable when production systems are diversified to include intercropping and rearing of small livestock.

The nutrition status of children during the baseline survey showed very high levels of stunting, which meant that they had experienced long-term deprivation of sufficient food and probably disease. However, the results of the intervention showed mixed responses, proving that nutrition education needs to be included in such interventions. Nutrition knowledge should be given to community members because it is absolutely essential for making informed choices in the selection and preparation of food for the household members. It is important to educate mothers on how to provide adequate nutrition and healthcare so as to improve the nutrition and health of children (especially those in their early childhood) to prevent stunting.

ACKNOWLEDGEMENTS

I want to acknowledge everyone who dedicated their time, effort and input to ensure the success of this project. I thank my supervisors Dr Elizabeth Kuria and Dr Margaret Keraka of Kenyatta University for their committed guidance and support.

Sincere thanks, also, to project-team members Ruth Nyansera, Jane Kaberia and Paul Okuta from the Ministry of Agriculture and Ministry of Livestock and Fisheries Development for their technical support. I also acknowledge the support of Gideon Ndung’u, my research assistant, Elizabeth Wambui and Josephat Wahiu, my field assistants, as well as to Onesmus Muinde and Robert Ofwago for technical advice in the interpretation of data. Lastly and most importantly, I sincerely appreciate the financial support of the AGROPOLIS granting programme of the Canadian-based International Development Research Centre (IDRC), which enabled me to complete the project successfully.

NOTES

1 Eunice Wambui Njogu, Ph.D. (Foods, Nutrition and Dietetics), PO Box 62697, Nairobi 00200, Kenya; email: koieunnice@yahoo.ca

2 In Fisher’s formula, n was determined as follows: n = Z2pq divided by d2 where n = the desired sample size when the population is more than 10,000; Z = the standard normal deviation at the required confidence level; p = the proportion in the target population estimated to have characteristics being measured; q = 1 p; and d = the level of statistical significance set. The calculation was: n = (1:96)2 (0.50) (0.050) divided by (0:05)2 = 384:00.

REFERENCES

Bonnard, P. (2001) Improving the Nutrition impacts of Agriculture Interventions: Strategy and Policy Brief, Food and Nutrition Technical Assistance Project (FANTA) and Academy for Education Development (AED), Washington DC

EpiInfo (2000) Anthropometric analysis software. US National Centre for Health Statistic NCHS and World Health Organization (software accessed in 2000; http://www.cdc.gov/epiinfo/)

Fisher, A. A., Lang, J. E., Stoeckel, J. E. and Townsend, J. W. (1998) Handbook for Family Planning Operations Research Design, 2nd edition, Population Council Nairobi, Nairobi

Foeken, D. and Mwangi, A. M. (2000) ‘Increasing food security through urban farming in Nairobi’, in Bakker, N., Dubbeling, M., Gundell, S., Sabel-Koschella, U. and de Zeeuw, H. (eds) Growing Cities, Growing Food Urban Agriculture on the Policy Agenda: A Reader on Urban Agriculture, Deutsche Stiftung fur Internationale Entwicklung (DSE), Feldafing, Germany

Food Security Analysis Unit for Somalia (FSAU) (2003) Nutrition: A Guide to Data Collection, Analysis, Interpretation and Use, Food Security Analysis Unit for Somalia, Nairobi, Kenya

Government of Kenya (2004) Strategy for Revitalizing Agriculture (2004–2014), Ministry of Agriculture and the Ministry of Livestock and Fisheries Development, Nairobi Government Printer, Nairobi

Hoddinott, J. and Yohannes, Y. (2002) Dietary Diversity as a Household Food Security Indicator, Food and Nutrition Technical Assistance Project (FANTA) and Academy for Education Development (AED), Washington DC

Iannotti, L. and Gillespie, S. (2002) Successful Community Nutrition Programming Lessons from Kenya, Tanzania and Uganda, LINKAGES, Academy for Education Development (AED), Washington, DC

Mougeot, J. A. (2000) ‘Urban agriculture: Definition, presence, potentials and risks’, in Bakker, N., Dubbeling, M., Gundell, S., Sabel-Koschella, U. and de Zeeuw, H. (eds) Growing Cities, Growing Food Urban Agriculture on the Policy Agenda: A reader on Urban Agriculture, Deutsche Stiftung fur Internationale Entwicklung (DSE), Germany

Mougeot, J. A. (2006). Growing better Cities, Urban Agriculture for Sustainable Development, International Development Research Centre, Ottawa, ON, Canada

Nutrisurvey (2004) Nutrient analysis computer software, University of Indonesia, SEAMEOTROPMEN (software accessed in 2005 http://www.nutrisurvey.de)

World Health Organization (WHO) (1995) Uses and Interpretation of Anthropometry, Expert Committee on Nutrition and Physical Status, WHO, Geneva, Switzerland

2
Urban Compost: A Socio-economic and Agronomic Evaluation in Kumasi, Ghana

Nikita S. Eriksen-Hamel and George Danso1

INTRODUCTION

Uncollected and poorly managed solid and liquid wastes are a health and environmental hazard, especially to the urban poor who live near informal, and often illegal, waste dumps. The lack of facilities to collect, transport and treat municipal, agricultural and industrial wastes poses a major challenge to the rapidly expanding cities of West Africa. To address this, new and innovative methods of collection, transport, processing and storage need to be identified. Composting of municipal wastes and using the compost for agricultural purposes is a quickly growing and viable option to managing urban wastes in both the developed and developing world (Drechsel and Kunze, 2001). Composting provides the environmental benefit of diverting waste from landfill sites, and health benefits by reducing the survival and spread of pathogens in wastes. Furthermore, the end product is a valuable soil fertilizer. The use of urban composts in urban and peri-urban agriculture (UA and UPA) effectively closes the nutrient cycle in urban areas and reduces nutrient losses to the environment (Drechsel and Kunze, 2001).

Kumasi is the second largest city in Ghana, with a population of 1,017,000 and a growth rate of 3.1 percent (Ghana Statistical Service, 2002). It has a semi-humid tropical climate, with average annual rainfall of about 1500 mm. Its central location in the country and relatively well-developed infrastructure make it a major market and distribution centre for Ghana and other West African countries. Within the peri-urban area, which covers an area of approximately 40 km radius around the city centre (Blake and Kasanga, 1997), are many agri-food industries such as breweries, saw mills and commercial poultry farms, which produce large quantities of organic wastes. The daily domestic waste generation in Kumasi is about 610 tonnes. The two main markets generate an additional 250 tonnes per day, most of which is organic; thus, the total city-wide generation is about 860 tonnes per day (KMA-WMD, 2000). Farming and trading are the main income-generating activities in Kumasi. Commercial crop production is dominated by vegetable horticulture and staple crops such as plantain, maize and cassava. In partially waterlogged valley bottoms, sugarcane and taro are also grown.

Kumasi is an ideal city to promote composting as a waste management strategy due to the large amount of organic wastes that are generated, the high biodegradable organic fraction of solid waste ranging between 65 and 75 percent and the extensive agriculture within its peri-urban boundaries (Salifu, 2001). A pilot composting plant was established in the Buobai suburb of Kumasi in 2001 by the IWMI and Kumasi Metropolitan Assembly (KMA). Different composts were produced from market wastes, household solid wastes and dewatered human sewage from septic and public toilets. During the inception of this composting project, the stakeholders wished to know the farmers’ perceptions and demand for compost in Kumasi. At the onset of compost production, its agronomic potential and benefits to urban and peri-urban farmers had to be evaluated. This paper presents results of two composting studies.

OBJECTIVES

The objectives of this project were to conduct a social, economic and agronomic evaluation of compost made from urban organic solid wastes and fecal sludge, as well as an evaluation of its potential benefits in urban agricultural systems of Kumasi.

The social and economic evaluations were conducted through individual interviews and focus group discussions with relevant stakeholders. The objectives of these interviews were to: examine and document the farmers’ perceptions, knowledge and socio-cultural acceptability of the compost; estimate the farmers’ WTP for the compost compared with other common soil inputs; determine the socio-economic factors affecting farmers’ WTP; and determine the demand for compost by farmers. Agronomic benefits of the compost were evaluated in seed germination and phytopathology experiments in the screenhouse and field trials. The objectives of these experiments were to: determine the success of germination and growth of vegetables to increasing rates of compost; determine whether the compost suppresses root knot nematodes, Fusarium wilt and Sclerotium rolfsii in lettuce and tomato plants; compare the fertilizer value of the compost with that of poultry manure; and determine the residual nutrient effects of compost applications on lettuce yield and soil fertility.

METHODOLOGY

Data collection from field interviews and focus group discussions

Interviews and focus group discussions with stakeholders, farmers and potential customers used a partially pre-coded questionnaire to elicit WTP and determine attributes of farmers’ perception and farming practices. Additional data were collected from government institutions, IWMI urban and peri-urban agriculture (UPA) library, Ghana office of the Food and Agriculture Organisation of the United Nations (FAO), Ghana Organic Agriculture Network (GOAN) library and farmer associations. A total of 200 individual interviews were conducted with farmers from different urban farming systems in and around Kumasi. Few women were involved in agriculture in Kumasi; therefore, specific attempts were made to interview as many women as possible to elicit their opinions. The contingent valuation method (CVM) was used during interviews with farmers to estimate the price they would pay for compost based on their knowledge of similar products such as poultry manure or black soil. The CVM approach involved surveying individuals to determine whether and how much they are willing to pay for compost under different hypothetical scenarios (Whittington et al, 1990; Asenso-Okyere et al, 1997; Alberni and Cooper, 2000). Farmers were asked whether they were willing to pay for compost as a soil improver and how much they were willing to pay for a 50-kg bag. A bidding game was introduced in situations where farmers found it difficult to understand and provide their bids. This bidding game helped to determine the zero WTP, and the mean and maximum amount the farmers would pay for the compost (Field, 1994; Hanley et al, 1997; Akpalu, 2000; Nugent, 2001).

Further questions focused on socio-economic characteristics, income, experience with compost and perception of compost quality. Since the dependent variable was a dichotomy (yes – 1 or no – 0), it was deemed appropriate to use a probit model instead of traditional regression models. The probit model was used to explain the factors that could affect farmers’ WTP for compost (Kennedy, 1990) (Table 2.1). The probit model is defined as:

Pr(y = 1|x) = Φ(xb)

where Φ is the standard cumulative normal probability distribution and xb is the probit score and takes care of the explanatory variables (Pindyck and Rubinfeld, 1983). The parameters in the probit model were estimated by maximum likelihood methods (Pindyck and Rubinfeld, 1983; Hailu, 1990).

Compost production and analysis

The compost plant produced four types of compost with differing ratios of household waste, market waste and dewatered fecal sludge. The first compost

Table 2.1 Chemical characteristics of soil fertility amendments

EC, electrical conductivity; TKN, total Kjeldahl nitrogen.

produced and evaluated in field trials was made from a 3:1 ratio of market waste and dewatered fecal sludge. Organic market wastes were obtained from the main waste bins of the Kumasi Central Market. The sludge was obtained from public toilet septic tanks and was dewatered on a drying bed at the compost plant. Details of the dewatering facilities and compost production are described in the IWMI report (2003). The compost, poultry manure and test soils were analysed using established methods at the Soil Research Institute (SRI) in Kumasi and are presented in Table 2.1.

Germination experiment

The eight common vegetables used in the germination experiment were tomato (Lycopersicon esculentum), sweet pepper (Capsicum annuum), okra (Hibiscus sabdariffa/esculentus), carrot (Daucus carota), garden eggplant (Solanum melongena), lettuce (Lactuca sativa), cabbage (Brassica oleracea capitata alba) and spring onion (Allium cepa). Thirty seeds of each vegetable were planted into trays filled with one of five soil treatments. The soil treatments were a mixture of potting soil and fresh compost at 0, 12.5, 25, 50 and 100 percent. The seeds were loosely placed on the surface and lightly covered with soil. Okra seeds, being the largest seeds used, were soaked in water overnight (16 h) before planting. Seedling trays were placed in a screenhouse for three weeks, where temperatures ranged between 25 and 30°C, and watered daily to maintain sufficient moisture.

Germination was counted every day for 15 days and was considered successful when the seed radicle protruded from the surface of the soil. The average time to emergence (ATE) was calculated as:

After three weeks seedlings were harvested by uprooting and washing the root soil. The length of the primary taproot, shoot and fresh weights were also measured.

Nematode suppression experiment

A factorial experiment was designed with two vegetables (lettuce and tomato plants), six soil treatments and two nematode treatments (with and without nematodes present) and replicated four times. The six soil treatments were prepared with potting mixture and the following fresh-weight percentages of compost: 0, 6.25, 12.5, 25, 50 and 100 percent. For each vegetable, two three-week-old seedlings were transplanted into a half-litre pot containing a different soil treatment. Into half of the pots, a 1-ml aliquot of nematode egg suspension, prepared using established methods, was injected into the soil 1 cm from the seedlings. Pots were kept in a screenhouse for six weeks, with temperatures ranging between 25 and 30°C, and watered every two days to maintain sufficient moisture. At harvest, plants were uprooted and shoot, root lengths and fresh weights determined. The incidence and severity of nematode root-galling was rated using a five-point scale.

Field fertilizer experiment

Two field experiments were conducted with the farmers of the Gyenyasi Farmers Cooperative. This cooperative has about 16 farmers, who produce lettuce, cabbage and spring onions in a 2-ha plot. Fourteen raised lettuce beds belonging to three farmers were divided into 30 plots, with sizes ranging from 7.5 to 12 m2. The fertilizer treatments included the compost and poultry manure applied at rates of 1, 2 and 4 kg m−2 on each crop for three crops, compost applied at 2.5, 5 and 10 kg m−2 on the first crop, and a control with no compost applied.

The first experiment compared the compost and poultry manure applications on three crops. Since farmers evaluate the impact fertilizers based on how much they have to carry into the field, the two fertilizers were compared on a weight basis and not a nutrient basis. The second experiment compared lettuce yields over three harvests from a single application of compost to the first crop. Lettuce plants were watered daily with approximately 61 m−2 of irrigation water containing about 28 mg N l−1. The equivalent fertilization was about 50 kg N ha−1 per month. Hand weeding was done on the same day each week and no pesticides were applied on the first crop. An outbreak of septoria leaf spot (Septoria lactucae Pass.) in the second and third crops required the use of Dithane fungicide. Composite soil samples were taken for chemical analyses and bulk density determined from each plot at the beginning of the experiment and after each cropping season. The fresh yield of all lettuce plants within a 1-m2 grid was determined four weeks after transplanting.

Statistical analysis

Descriptive and inferential statistical tools were used to analyse interview data using SPSS software. The data from the CVM survey were analysed with a probit model. Correlations and means comparisons were conducted for the different explanatory variables to avoid multi-colinearity in the probit model. Due to difficulties in obtaining sufficient materials and labour, the germination experiment was not replicated and only one experimental unit per treatment was available. The unreplicated data were analysed by regression analysis. The effects of different compost treatments on nematode galling and yield were evaluated using ANOVA procedures. All the treatment comparisons were carried out at the P = 0:05 significance level.

RESEARCH FINDINGS

Social perceptions of compost made from urban organic wastes

Generally, the farmers had a good perception of the quality of compost regardless of how much experience they had using it. There was consensus that compost improved soil quality and increased crop yield. The major reasons given for this were: compost resembles ‘black soil’, which is known to farmers as a good soil input; high quantities of ‘night soil’ mixed with compost enriched the biofertilizer; they perceived long-term effects on the soil; and they associated compost use with benefits related to waste reduction and urban sanitation. About 42 percent of the farmers thought that handling compost did pose the same risks associated with the handling of chemical fertilizers.

The 20 percent of farmers who had no experience with compost perceived it to be ineffective, largely due to cultural beliefs or health concerns of the night-soil component of the compost. They believed that handling compost with night soil would lead to spreading of diseases such as HIV, typhoid and cholera. Furthermore, these farmers assumed that consumers would avoid crops on which compost had been used for fear of infections. Other reasons for being sceptical about the compost were the high labour requirements for compost production (57 percent); insufficient amount of raw materials to produce compost (28 percent); and the small market for organic products in the Kumasi region (15 percent). A similar survey in Kumasi found that farmers had the same perceptions and reasons for accepting compost made from urban wastes (Warburton and Sarfo-Mensah, 1998).

The majority of farmers indicated that compost would be moderately (51 percent) or very difficult (10 percent) to apply to soils compared with chemical fertilizers, which could easily be applied with irrigation water. Poultry manure was smelly, cloddy and inconvenient to apply when wet, whereas compost could be spread more easily using rakes. Many farmers perceived compost to be expensive (48 percent), but were willing to handle it (42 percent) and believed it could control the incidence of pest and diseases (42 percent). The cost of establishing a compost plant and the increased cost of transporting the material were the main economic concerns of farmers. Few were unwilling to handle the compost (6 percent) or believed it could not control the incidence of pests and diseases (2 percent). However, the majority of farmers were indifferent about handling compost (52 percent) or had no opinion about its ability to control pests and diseases (56 percent). Their claim was that, since they had not used it before, they could not judge its effectiveness at controlling pests and diseases.

Farmers’ willingness to pay for compost

There was a positive correlation between how farmers perceived compost and their WTP for it. About 70 percent of them expressed positive WTP, while 30 percent were unwilling to pay even though they perceived it as a good soil input. Farmers were unwilling to pay for it largely for economic reasons. They were generally satisfied with the high returns achieved from their current soil input and did not see a need to replace or try new ones. Many suggested that it should be the government’s responsibility to subsidize or supply inputs free to farmers.

Comparing WTP bids across different farming systems in Kumasi revealed that urban vegetable farmers expressed the lowest mean WTP (US$0.10), although their systems were profitable, with annual revenues ranging between US$600 and $1000 per hectare. Peri-urban vegetable farmers proposed the highest mean WTP (US$3; Table 2.2). The lower bid from urban vegetable farmers may have been influenced by the cheap and readily available poultry manure in urban farming systems.

A previous study in Kumasi found that more than 90 percent of the urban vegetable farmers preferred poultry manure because of the low price (US$0.01 per 10 kg) and availability (Drechsel and Kunze, 2001). On the other hand, the peri-urban vegetable farmers used expensive chemical fertilizers, which cost about US$2.80 per 10 kg. Although the nutrient content was greater in chemical fertilizers, the cost of nitrogen was still cheaper for the poultry manure than the chemical fertilizer.

The poultry industry in Kumasi may have the potential to provide large amounts of bio-fertilizer. There are about 300 registered commercial poultry farms in and around Kumasi (Kindness, 1999), and manure production by the largest farm alone is about 50 tonnes per week (Drechsel, 1996).

Decision variables for farmers’ WTP

Farmers’ experience of the use of compost had significant influence on the WTP for compost, as did gender and household dependency. However, their

Table 2.2 Farmers’ willingness-to-pay for 50 kg of compost

Table 2.3 Probit model of explanatory variables of farmers’ WTP for compost in Kumasi

1Coefficient is significant at 5 percent (*) and 1 percent (**).

perceptions as related to compost, household income, level of education and age were all significant variables that influenced WTP (Table 2.3).

As expected, a positive perception of compost implied a higher probability of WTP for the compost. Similarly, farmers with higher income were willing to pay more than poorer farmers. Farmers who could read and write have a higher probability of paying for compost than those who could not read or write. This was most likely because more educated farmers were better informed about innovations and better understood the advantages and disadvantages of compost. Older farmers had a lower WTP than younger farmers, possibly due to older farmers not valuing the need to invest in the land they were using, whereas younger farmers were more motivated to invest in their land to realize long-term benefits. The present soil input the farmer was using had a significant impact on his or her WTP for compost. This was primarily because farmers used inexpensive soil fertility improvement inputs which were familiar to them. The farmers who were not interested in applying compost to their crops had less security on the plots they farmed and long-term investment in soil quality was not a priority to them.

Demand for compost in Kumasi

Demand for compost in Kumasi was estimated through WTP bids by farmers for both the subsidized and unsubsidized compost plants. Compost produced from Kumasi was valued at US$5 per 50 kg bag and its demand was 940 tonnes per year. Peri-urban farmers were willing to pay US$4 per 50 kg bag, which created a difference of US$2 with the prevailing market prices. This meant that, by sustaining demand for compost at this higher price, the government would have to subsidize the prices by US$2, which would lead to an explosion in demand estimated at 11,000 tonnes per year.

Although the compost plant is centrally located with respect to the urban farming areas of Kumasi, the distances between it and the farms are still

Figure 2.1 WTP for compost at different distances from the compost station

large, plus the road access to these farms is often poor. Even with subsidized compost production, high transportation costs make it difficult for those living beyond a 35-km radius of the compost plant to afford to buy compost (Figure 2.1). Since peri-urban farms are located beyond a 40-km radius of the city (Adam et al, 1999), it appears unlikely that composting of urban waste in Kumasi would be a realistic option for recycling nutrients from urban areas to peri-urban and rural areas.

Germination experiment

The germination percentage was consistent for all vegetables and ranged from 70 to 100 percent except for okra and garden eggplant, whose seeds were infested by red ants. The germination of most vegetables was unaffected by the compost application, with the exception of carrot and onion. Germination of carrot (P = 0:050, R2 = 0:55) and onion (P = 0:050, R2 = 0:61) decreased as the rates of compost application increased. Seedling survival of most vegetables was similarly unaffected by compost rates, with the exception of carrot (P = 0:048, R2 = 0:78) and onion (P = 0:060, R2 = 0.74). The average time of germination increased with increasing compost rates for tomato (P = 0:031, R2 = 0:83), carrot (P = 0:013, R2 = 0:86), lettuce (P = 0:012, R2 = 0:91) and okra (P = 0:001, R2 = 0:98). For these vegetables, the average time for germination was 1–2 days longer at high compost rates (>50 percent) than low rates (<6.25 percent). Differences in shoot weight were not apparent after 21 days except for okra, which had the greatest shoot biomass at 6.25 percent compost.

With the exception of carrot and onion plants, seed germination and seedling growth of most vegetables were unaffected by the phytotoxic compounds in the compost. Compost applications to carrot and onion nursery beds should be limited to low amounts or entirely avoided. The delay in germination of some plants is most likely due to low levels of salinity in the compost, which interfered with the imbibition phase of the seeds.

Nematode suppression experiment

Root galls were observed for the majority of plants inoculated with root knot nematodes. Tomato and lettuce plants without compost and those infected with nematodes were either dead or highly infected at the time of harvest. Root galling in tomato (P = 0:001, R2 = 0:70) and lettuce plants (P = 0:002, R2 = 0:97) decreased with increasing compost rates. A significant reduction in root galling from the 0 percent compost treatment was observed in tomato plants at 6.25 percent compost and in lettuce plants at 50 percent compost. Phytotoxic effects of compost significantly reduced plant weight when compost rates exceeded 50 percent in tomato and 25 percent in lettuce. The optimum compost rate for greatest yield was between 6.5 percent and 12 percent for tomatoes and 12 percent for lettuce. This research supports other findings which have shown that composted biowaste reduces the incidence of many diseases in a wide range of vegetables (Tuitert et al, 1998; Blok et al, 2002).

Field fertilizer experiments

Three farmers volunteered a limited number of lettuce beds to be used in the experiment. The resulting experimental design had each farmer managing a single replicate of each treatment on their lettuce beds. This design introduced the farmers as a new factor along with fertilizer type and application rate. To discount the effects of the farmers, we attempted to control the management of the lettuce beds. The date of planting and harvesting, quantities and method of fertilizing, watering and hand-weeding were the same for all of them. Nonetheless, differences between them were confirmed through statistical analysis (P = 0:01) and these were caused by differences in management history, location of lettuce beds and pest incidence among the three farmers’ plots. Contrast analysis showed that poultry manure treatments had greater yield (P = 0:004) than compost treatments. The compost and poultry manure treatments with the same application rates

Table 2.4 Yield of lettuce following application of the compost and poultry manure

1Values in each column for each crop and the mean of crops followed by similar letters are not significantly different than Tukey’s HSD test (P < 0:05). Letters indicate treatment groupings that are statistically different from one another. Grouping rank decreases with alphabetical order (e.g. group ‘a’ is greater than group ‘b’).

were paired on the same bed to allow for direct visual comparisons. Although the only significant differences between paired treatments were at the lowest application rate (1 kg m−2) in the third crop (Table 2.4), all farmers agreed that lettuce plants were larger and greener in poultry manure treatments compared with compost treatments. Furthermore, the farmers’ perceptions agreed with the statistical analysis, which showed that 4 kg m−2 compost had greater lettuce yield (P < 0:05) compared with the control but only in the first crop. Marketable yields of 1.5 kg m−2 were obtained from all poultry manure treatments and from the highest compost application (4 kg m−2) (Table 2.4). Lettuce yields increased linearly with increasing compost application in the first crop and marketable yields (about 1.5 kg m−2) were achieved for the 50 and 100 kg N m2 treatments. In the second crop, marketable lettuce yields were attained only with addition of the 100 kg N m2 treatment.

Lettuce yields in treatments that only had a compost application on the first crop were not different from the control in each crop nor for the mean of all crops. Yields obtained from the control treatment remained the same for the crops in the three cropping seasons. Since no soil nutrients were added, it would have been expected that the continued extraction of nutrients in harvested lettuce would deplete the nutrients in the control plot to a level where no crop would have grown. This is because nitrogen mineralized from the organic matter in the sandy soil was unlikely to provide sufficient nutrients to explain these yields. However, irrigation water used in this study supplied about 50 kg N ha−1 per month and acted as fertilizer to the control treatment.

CONCLUSIONS AND RECOMMENDATIONS

Farmers were aware of the potential benefits compost could give to improving soil quality, increasing yields and controlling the incidence of pests and diseases. However, there were many concerns regarding the potential costs of producing compost, the difficulty in using the bulky compost and the health risks associated with handling compost made from night soil.

WTP for compost was lowest for urban vegetable farmers (US$0.10 per 50 kg), who rely on readily accessible and cheap poultry manure, whereas peri-urban vegetable farmers were ready to pay US$3 per 50 kg.

Demand for compost in Kumasi was estimated at about 11,000 tonnes per year based on the assumption that subsidies were provided. But without subsidies actual demand was only 940 tonnes per year.

Adding compost increased growth and did not delay germination of a wide range of vegetables when applied at concentrations between 6.25 and 12.5 percent, with the exception of carrot and onion. Nematode damage in lettuce and tomato plants was also suppressed, and poultry manure was the organic fertilizer of choice for the urban farmers in Kumasi.

In conclusion, we can infer from this study that large-scale compost production from urban organic waste and night soil in Kumasi is not an economically viable project without some kind of subsidy. Sorting and transportation costs for compost at this scale outweigh the economic benefits associated with the use of compost. There is a need to explore also the viability of community-based composting initiatives at the farm level, where sorting and transportation costs are minimal.

ACKNOWLEDGEMENTS

This research project benefited from the support and time of many individuals and organizations. The authors would like to thank the generous financial assistance of the International Development Research Centre (IDRC; AGROPOLIS programme), the IWMI, Wageningen University and Kwame Nkrumah University of Science and Technology (KNUST). The compost plant was funded by the IDRC, the French Foreign Ministry, the Swiss Federal Institute for Water and Sanitation in Developing Countries (SANDEC) and the KMA. We would like to acknowledge the support, help and guidance of Dr S. Fialor, Dr W. Blok, Prof. C. Quansah, Dr R. Awuah, Mr K. Osei, Dr P. Drechsel, Mrs K. Dapaah and all the technicians, labourers and farmers who helped us throughout this project. A special mention to the head of the composting project, Dr C. Olefunke, for her dedication to this project.

NOTE

1 Eriksen-Hamel, 34 Maple St. Apt. 31, Ste-Anne-de-Bellevue, Québec, Canada, H9X 2E6, tel: +1-514-457-7352/fax: +1-514-398-7990; email: nikita.eriksenhamel@gmail.com

REFERENCES

Adam, M. G., Sarfo-Mensah, P. and Warburton, H. (1999) ‘Urban blight and farmers’ plight: Peri-urban agriculture around Kumasi, Ghana’, The Land 2(2), pp131–148

Akpalu, W. (2000) ‘Willingness to pay for fish conservation and to reduce health risks associated with pollution of Fosu Lagoon, application of contingent valuation methods’, Proceedings of the Beijer Research Seminar, Morocco, March 2000

Alberni, A. and Cooper, J. (2000) ‘Application of the contingent valuation methods in developing countries’, FAO Economic and Social Development Paper No 146, F, Rome, Italy

Asenso-Okyere, W. K., Osei-Akoto, I., Anum, A. and Appiah, E. N. (1997) ‘Willingness to pay for health insurance in developing economy. A pilot study of the informal sector of Ghana using contingent valuation’, Health Policy, 42, pp223–237

Blake, B. and Kasanga, K. (1997) Kumasi Natural Resource Management Research Project, Inception Report, Natural Resource Institute, University of Greenwich, UK

Blok, W. J., Coenen, T. C. M., Piji, A. S. and Termorshuizen, A. J. (2002) ‘The Netherlands – suppressing disease in potting mixes with composted biowastes’, Biocycle, 43, pp58–68

Drechsel, P. (1996) ‘Applied research for peri-urban areas’, IBSRAM Newsletter, 42, pp5–7

Drechsel, P. and Kunze, D. (eds) (2001) Waste Composting for Urban and Peri-urban Agriculture – Closing the Rural–Urban Nutrient Cycle in Sub-Saharan Africa, CABI Publishing, Wallingford, UK

Field, C. B. (1994) Environmental Economics: An Introduction, McGraw-Hill College, Maidenhead, UK

Ghana Statistical Service (2002) Population Census of Ghana, Ghana Statistical Service, Accra, Ghana

Hailu, Z. (1990) The Adoption of Modern Farm Practices in African Agriculture. Empirical Evidence about the Impacts of Household Characteristics and Input Supply Systems in the Northern Region of Ghana, Nyankpla Agricultural Research Report 7, CSI/GTZ, pp114–167

Hanley, N., Shogren, J. and White, B. (1997) Environmental Economics in Theory and Practice, Oxford University Press, Oxford, UK

Henn, P. (2000) ‘User benefits of urban agriculture in Havana, Cuba: An application of the contingent valuation methods’, MSc Thesis, McGill University, Canada

IWMI (2003) Co-composting of Faecal Sludge and Solid Waste for Urban and Peri-urban Agriculture in Kumasi, Ghana, Final Technical Report, International Water Management Institute (IWMI), Accra, Ghana

Kennedy, P. (1990) A Guide to Econometrics, 2nd edition, Basil Blackwell, Oxford, UK

Kindness, H. (1999) Supply and demand for soil ameliorants in peri-urban Kumasi, KNRMP R6799, NRI-DFID-KNUST Report, Kumasi, Ghana

KMA-WMD (Kumasi Metropolitan Assembly – Waste Management Department) (2000) Yearly Report 2000, Kumasi Metropolitan Assembly, Kumasi, Ghana

Nugent, R. A. (2001) ‘Using economic analysis to measure the sustainability of urban and peri-urban agriculture: A comparison of cost-benefit and contingent valuation analyses’, Proceedings of the CIP-SUIPA and ETC-RUAF workshop on Appropriate Methodology in Urban Agriculture Research, Planning, Implementation and Evaluation, Nairobi, Kenya

Pindyck, R. S. and Rubinfeld, D. L. (eds) (1983) Econometric Models and Econometric Forecasts, McGraw Hill, Tokyo, Japan

Salifu, L. (2001) ‘An integrated waste management strategy for Kumasi’, in Drechsel, P. and Dagmar, K. (eds) Waste Composting for Urban and Peri-urban Agriculture: Closing the Rural–Urban Nutrient Cycle in Sub-Saharan Africa, CABI Publishing, Wallingford, UK, pp112–115

Tuitert, G., Szczech, M. and Bollen, G. J. (1998) ‘Suppression of Rhizoctonia solani in potting mixtures amended with compost made from organic household waste’, Phytopathology, 88, pp764–773

Warburton, H. and Sarfo-Mensah, P. (1998) The Use of Composted Urban Waste in Integrated Pest Management Systems to Control Pests and Pathogens in Peri-Urban Agriculture, Technical Project Report C1045, DFID, RNRRS, and NRSP Peri-Urban Interface Program, Kumasi, Ghana

Whittington, D., Briscoe, J., Mu, X. and Barron, W. (1990) ‘Estimating the willingness to pay for water services in developing countries: A case study of the use of contingent surveys in Southern Haiti’, Economic Development and Cultural Change, 38, pp293–311

3
Urban Agriculture as a Livelihood Strategy in Lima, Peru

Luis Maldonado Villavicencio1

INTRODUCTION

This study was a component of the baseline of the project entitled ‘Agricultores en la Ciudad’ (Farmers in the City), carried out by the Urban Harvest Program hosted by the International Potato Center (CIP). This research used the neighbourhood of Carapongo, in the city of Lima, capital of Peru, as a case study. Lima has a population of more than seven million people and, like many Latin American cities, has a high concentration of urban poverty along with rampant and chaotic urban sprawl. At present, approximately 3.3 million poor people live in the city. Poverty in Peru has become a problem with a distinctly urban face. The explosive growth of Lima can be traced to 1940 when people migrated from rural areas to the capital in search of better living and working conditions. Growth occurred largely through the illegal occupation of both privately and publicly owned uncultivated lands.

Carapongo was selected as a case study because it represents a peripheral area of Lima where urban agriculture (UA) is commonplace. Here, one of the principal threats to urban producers is rampant urban sprawl. In addition, urban households involved in agriculture contend with a lack of recognition and understanding from policy makers. City authorities often still perceive agriculture as a rural activity conducted far away from the city and thus there is a marked lack of interest in the sector. In this context, an understanding of the strengths and limitations of UA becomes a crucial starting point for developing adequate political strategies.

This paper analyses UA as a household strategy which is used to confront urban poverty, and it examines institutional and political factors that block and/or support the strategy. Specifically, the paper identifies which assets households use in order to overcome urban poverty and how they engage with institutions and municipal authorities. Gender analysis between men and women in the use of household assets was also done.

The paper begins by explaining the methods used as well as the livelihoods approach. An analysis of access to assets by urban farmers is presented, followed by an overview of the dimensions of gender related to home-based agriculture. The paper then identifies the influence of UA in Carapongo and presents an analysis of the stakeholders and some of the conflicts between them. Finally, the study concludes with the main research results.

METHODOLOGY

Definition of the study area

Carapongo is located on the Rímac River approximately 15 km from the centre of metropolitan Lima (Figure 3.1). Here, many vegetables are produced by smallholder farms by a predominantly migrant population who come from rural areas. This locality was selected for several reasons: agriculture represents a way of life for more that half of its population (INEI, 2002); there is a strong interconnection and integration with the economic and ecological system of the urban centres; there is ample future potential due to the abundance of natural resources such as water and land; and producers can benefit from contacts with nearby organizations.

Methodology

In this study, several techniques and instruments of quantitative, qualitative and participatory research were used. Additionally, a review of literature, including a statistical overview, was carried out in order to prepare the theoretical framework.

In order to better understand the actors’ various perceptions of UA, semi-structured, in-depth interviews were carried out. In total, 15 farmers and five local authority representatives were interviewed. Also, participatory workshops explored various subjects such as historical events that have affected the community in recent years. An example of this was the identification and prioritization of the principal historical problems affecting UA along with discussions and suggestions of alternative solutions – the matrix of alternatives.

Two discussion groups were conducted to collect information about gender roles, one comprised of men and the other of women. Other tools such as seasonal calendars, transect walks and problem analysis using drawing were also carried out for gender analysis.

For the survey, 125 households were randomly selected using a list of farmers. In each household, the person identified as the main individual responsible for the farm was interviewed. Interviews contained 110 questions and lasted approximately 45 minutes. All interviews were conducted by the author and a team of students during two months.

Figure 3.1 Location of the field research

The survey sought to identify the ways households identify assets. In addition to these activities, video images were recorded as a means of direct observation so as to validate and/or obtain basic information on the economic condition and level of services available in the area.

CONCEPTUAL FRAMEWORK

The livelihoods system framework

Chambers and Conway (1992) state:

A livelihood comprises the capabilities, assets (including both material and social resources) and activities required for a means of living. A livelihood is sustainable when it can cope with and recover from stresses and shocks, maintain or enhance its capabilities and assets, while not undermining the natural resource base.

The livelihoods framework was initially designed to improve the understanding of rural households, but it is now seen as a generic framework for use in urban as well as rural areas (Singh and Gilman, 1999; Martín et al, 2000; Sanderson, 2000). The framework does not attempt to provide an exact representation of reality, but it does provide a way of thinking about livelihoods designed to stimulate debate and reflection. The livelihoods framework views poor households as being dependent upon a diversity of strategies in order to face urban poverty (Figure 3.2). These strategies are based on a set of household ‘assets’: natural capital (land and water); financial capital; physical capital (houses, equipment, animals, seeds); human capital (in terms of both labour power and capacity, or skill); and social capital (networks of trust between different social groups). The deployment of assets also depends on external influences such as dealing with regulations, policies,

Figure 3.2 The sustainable livelihoods systems framework

urban authorities and local marketing practices. The inability to adequately use and employ the various assets at their disposal can leave households vulnerable to economic, environmental, health and political stresses and shocks.

Various studies illustrate how UA is used as a strategy by the urban poor to generate income and provide fresh produce to urban consumers. The practice of UA is considered an important tool of the urban poor in order to contend with poverty, generate income and deal with insecurities such as procuring food (Rakodi, 1991; Maxwell, 1994; O’Reilly and Gordon, 1995; Antweiler, 2000; Arm-Klemesu, 2000; Nugent, 2000). Other work has analysed the crucial role played by women in UA (Rakodi, 1988; Mbiba, 1993; Maxwell, 1994; Mianda, 1996; Mudimu, 1996; UNDP, 1996; Hovorka, 1998; Mougeot, 2000).

This paper analyses UA as a livelihood strategy, using the following definition:

An industry located within (intra-urban) or on the fringe (peri-urban) of a town, a city or a metropolis, which grows and raises, processes and distributes a diversity of food and non-food products (re-)using largely human and material resources, products and services found in and around that urban area, and in turn supplying human and material resources, products and services largely to that urban area. (Mougeot, 2000)

RESEARCH FINDINGS

Urban agriculture: The Carapongo case

In Peru, the importance of UA increased with the implementation of land reform during the 1970s. Prior to land reform, much of the agricultural land around Lima was occupied by large estates – haciendas – which primarily grew crops on an industrial scale. Reforms led to the establishment of production cooperatives for former hacienda workers, followed by a gradual redistribution of this land into small plots of less than one hectare. Initially, former hacienda workers and their families planted these plots to enhance food security by using crops such as maize and sweet potato. More recently, commercial vegetable production and livestock raising are increasingly common.

The neighbourhood of Carapongo rests at approximately 200 m above sea level in the lower zone of the Rímac River watershed. The community encompasses an area of 400 ha and has more than three times as much cultivated land than an average residential area. The Rímac River is the source of irrigation for Carapongo’s agricultural plots. The water is conveyed through a system of irrigation channels which permit farmers to produce three or four harvests annually. The most commonly grown vegetables are beets, lettuce, cabbages, turnips, celery and radishes (Figure 3.3). Agriculture

Figure 3.3 Urban agriculture in Carapongo

constitutes an important strategy for obtaining income; however, it has developed in a context of unregulated urban growth, limited access to markets and restrictive policies.

Urban agriculture: a sustainable livelihood strategy

In order to analyse UA as a livelihood strategy, this section is divided into two parts. First, the types of assets of urban households involved in agriculture are discussed. Second, the different gender roles of men and women within the household are examined in order to present a more nuanced understanding of UA in Carapongo.

Assets of households practising urban agriculture

As noted above, urban producers use a variety of assets, which they combine in order to deal with risks and vulnerabilities. These assets are divided into five categories: natural, physical, human, financial and social.

Natural capital

Access to land – The process of urbanization is having a serious impact on Carapongo as agricultural fields make way for urban development. Of the 289 ha used for agriculture in 2002, 213 ha remained by 2006: a 26 percent reduction. On the other hand, for the same period, the area devoted to housing increased from 113 to 197 ha, representing a 75 percent increase (Figure 3.4).

Figure 3.4 Carapongo land use change: 2002–2006

The study found that 37 percent of UA producers in Carapongo occupy land that they own, which has an average area of 0.72 ha, whereas 19 percent occupy rented land (average area 0.61 ha). Seven percent of farmers occupy land located on the border of the river (average area 0.87 ha). This last category of farmers are farming illegally and are known as pocesionarios. Seventeen percent of producers occupy land with a combination of ownership and rental, with an average area of 0.96 ha per holding.

Access to water – The principal water source for irrigation is the Rímac River, which runs through the city of Lima. The river nurtures an ecosystem in the urban and peri-urban areas that generates both benefits and risks for people. One of these risks is water pollution, which greatly affects agricultural producers of Carapongo, because the edges of irrigation channels have become deposits of solid and liquid wastes from nearby residents. Where quality requirements for food are high, such pollution prevents urban producers from accessing new markets. To access irrigation water, farmers pay a fee of around US$62.42 (206 nuevo sol/ha per year, where US$1 = 3.3 nuevo sol) to the water user association (Junta de Usuarios del Rio Rímac – JUR) that manages the resource.

Physical capital

Animal breeding – Animal husbandry represents physical capital for households practising UA. The purchase and raising of animals is a source of savings, as animals can be sold to obtain access to financial capital. In the area studied, 49 percent of the farmers raise sheep, 42 percent raise guinea pigs and 21 percent raise pigs. Forty-two percent of farmers raise chickens and 25 percent reported that they raise ducks. Only sheep and pigs are sold for income (Figure 3.5).

Access to roads and transportation – Carapongo has easy access to transportation routes, which makes it possible for produce to reach the principal urban markets rapidly. Farmers use various strategies for transportation of their harvest. One way is to rent pick-up trucks or trucks. In such scenarios, two or three producers join together to share the transportation expenses. Larger producers use their own private transportation. Motorcycles are also commonly used by producers who sell their farm produce to an intermediary, who is then in charge of transporting the goods to market by truck (Figure 3.6).

Means of production – Urban producers mainly use the horse-drawn plough to work the land. Heavy machinery such as tractors is rarely used

Figure 3.5 Livestock farmer in Carapongo

Figure 3.6 Different ways of transporting food from the field to urban market

(Figure 3.7). On average, the majority of farmers have two backpack sprayers to fumigate the crops. The irrigation system is primarily comprised of bed and furrow irrigation and very few farmers utilize sprinkle irrigation. The urban production of vegetables is characterized by an intensive use of labour, mainly carried out by family members. It is common to find whole families

Figure 3.7 Using traditional tools for soil management

participating in agricultural tasks, from seeding to harvest (Figure 3.8). Contracting local labour is also common. Men are hired for certain jobs (washing produce, loading sacks and transporting), while women are employed for seeding, weeding and harvesting. The daily wage is approximately US$5 for men and US$4 for women.

Access to housing – More than half of the urban farmers interviewed stated that they owned their homes. In many cases, however, ownership does not have legal backing: 61 percent of the people interviewed in this study do not have title to their property. According to de Soto (2000), 53 percent of homes in the cities are ‘extralegal’ dwellings, that is, they are traded in an informal market without formal legal status. Further, de Soto points out that the lack of legality of these assets makes it difficult for households to generate financial capital because people are unable to access credit.

Human capital

Local knowledge and training – Urban farmers’ knowledge is transferred in various ways: from generation to generation, from parents to children, and by the knowledge acquired over time through decisions made along the way thanks to the adoption of new and innovative ideas. In this zone, 75 percent of farmers have come from rural areas (ENCUESTAS, 2004). They arrive with knowledge of how to cultivate land and raise animals. Others who do not

Figure 3.8 Families participating in agricultural tasks

have this traditional knowledge have had to learn it while living in the urban zone. Many families grow corn, potatoes and beans, and raise guinea pigs – a traditional activity in rural areas. Urban farmers also practise weed control, companion planting and use animal dung for fertilizer.

The formal educational level of farmers is low, with only 18 percent of those surveyed having received any training. Nonetheless, on many occasions, farmers expressed a lack of interest in participating in training workshops offered to them. This is symbolic of an ongoing lack of trust directed towards any external agents who attempt to intervene.

Health status – Urban farmers in Carapongo have adequate access to health care because of good proximity to health centres and hospitals. Still, producers are exposed to health risks and in particular to gastrointestinal diseases, mainly because of polluted water. Some studies (TECNIDES, 2002) indicate that the water in this community is highly contaminated by fecal coliforms, parasites and other pathogens that surpass acceptable limits established by health legislations, both for drinking and irrigation water.

Perception of poverty – Low self-esteem, lack of self-confidence and the perception of poverty impede the ability to find solutions to one’s problems and, by extension, the ability to develop one’s ‘human capital’. This study attempted to identify some issues facing farmers in Carapongo. To this end, one participatory workshop involved a discussion concerning whether farmers considered themselves to be poor or not. Results from the six participants are presented in Table 3.1 and an interesting trend was discovered: most farmers do not consider themselves to be poor. Factors influencing this personal

Table 3.1 Perception of poverty

assessment include their access to land, crops and animals, all of which give food security to the family and strengthen self-esteem. On the other hand, feelings of being vulnerable and poor were expressed in the case where the participant did not have access to land.

Financial capital

Financial assets are the most significant limiting asset of poor people. Financial assets are also one of the most important as they can be used to leverage other types of capital. To speak about income and expenditures with the urban farmers was quite a delicate subject because they are not open about this issue; however, in the study we made an approximation of these financial components. This study identified that 51 percent of the population in Carapongo only receive income derived from agriculture. For the analysis of the income variable, an approximation of the availability of cash to meet monthly expenditures is shown. In this study, 28 percent declared that their families have less than US$150 available for their monthly expenditures; 41 percent had more than US$150 but less than US$300; and 31 percent had equal to or more than US$300 a month. Twenty-eight percent of families with an average family size of five persons who live on US$150 a month – less than US$1 per person/day – indicates extreme poverty (Table 3.2).

Another important financial indicator is the proportion of total expenditures that are made on food. For 64 percent of urban producer families in Carapongo, food expenditures comprise more than 50 percent of their total expenditures. For the remaining 40 percent of the families, food expenditures represent more than 70 percent of their total household expenditures.

Table 3.2 Income range and per capita income

Loans and credits – Access for formal credit from banks is limited, and only 14 percent of the producers reported having received bank loans. The remaining 86 percent have not received any loan. Thirty-five percent of all of the farmers surveyed have received credit (credit being different from a loan). For these farmers that have received credit, the main sources are stores that sell agricultural products and fertilizer, and other small enterprises. The principal reason for asking for financial assistance was to cover the input costs of growing vegetables. Lack of access to credit restricts the growth of their businesses and illustrates an important problem associated with smallholder farming in urban areas.

Savings – There does not seem to be a very broad range of options for saving money for the population living in Carapongo. Interviews revealed there is no custom of formal saving (in banks, for instance). Some people indicated that they save by keeping money safe at home or by purchasing animals that will be sold later.

Social capital

Carapongo is made up of a population who are originally from various regions and who grow different crops. For Durston (1999), common socio-cultural elements such as outlooks and values, memory of historical events, religion and myths, identity, rules of relationship, principles of reciprocity, as well as the social phenomena of neighbourhood and friendship, comprise the necessary precursors for the formation of social capital. He identified four levels of analysis: the individual; closed small groups, where there is shared experience and a high degree of trust; the wider community where there are common interests and objectives; and external relationships and interactions.

After noting Durston’s constructs, four categories of social capital in Carapongo were identified: individuals; family or closed groups; community groups; and ‘exogenous’ groups, such as local government:

1 Individual actors, especially community leaders, have access to networks and contacts with other groups that permit them a greater control of information and resources.

2 Family and/or closed groups have kinship ties and linkages that influence their actions. This is especially true among those coming from rural areas who share land and help new arrivals to establish their housing and employment. Furthermore, in many cases it was discovered that activities in the field are performed by the members of the family.

3 Community groups are an important source of social capital and include organizations such as farmers’ associations and groups that are initiated with the purpose of defending property rights both of land and housing. This research found that in some cases, the leadership of these groups was not completely trusted. The leadership of the Water Users Association was cited as more trustworthy than leadership of other groups.

4 Exogenous groups such as, for instance, local government are characterized by very weak ties and a high degree of mistrust within the community. An example of conflict includes the issue of how lands are zoned in Carapongo. Farmer and community groups are advocating for the right to conduct agriculture, while local authorities do not permit such a classification.

Other indicators of the social capital are indexes of confidence and network formation.

Indexes of confidence – The strength of social ties can be quantified using a ‘confidence index’ – essentially, a way of quantifying different degrees of confidence between different actors. With regard to the index of confidence required to form an organization, the highest degree of trust is for neighbours (36 percent), then family members (22 percent). However, when enquiries were made about confidence in different institutions, the local government was rated poorly, with a 60 percent level of mistrust, while a higher amount of confidence was shown in the Irrigation Board (with 35 percent average confidence and 15 percent high confidence).

Networks and other forms of organization – Urban farmers participate in organizations that have been formed primarily because of the presence of common interests in specific areas (such as water and land). To date in Carapongo no networks have been formed involving inter-sectoral or comprehensive approaches. Working with others in networks involves being part of a process where information is exchanged, new knowledge generated, experiences strengthened, resources exchanged, practices integrated and replicable models built for other projects. Lack of networking is a limitation since it impedes the ability to take advantage of opportunities generated by the market. Lack of networking also makes negotiating with other agents of the production and consumption chain more difficult.

Gender division of labour within households

We examined the gender division of labour to look at the roles of men and women in UA. Among other gender issues, this section examines gender division in access to and control over assets (with respect to control over the benefits of production) and productive resources. The survey of 125 farmers included questions related to gendered labour responsibilities, reproductive responsibilities and control over means of production.

Division of labour in agricultural work – One of the more interesting findings relating to the division of labour between men and women in UA is the frequency with which tasks are shared. Primary responsibility for agricultural tasks is summarized in Table 3.3. In only two types of tasks – land preparation and pest control – do men clearly possess sole responsibility. Women are generally responsible for these tasks only when they are single, widowed, or under other special circumstances. Where the task of land preparation is shared, women help to prepare vegetable beds that cannot be done by ploughing alone. Many other tasks are more frequently shared than defined specifically as male or female roles. Although men play a stronger role in vegetable production, this does not carry through to post-harvest and marketing activities, where women clearly play a bigger part. This is because

Table 3.3 Primary responsibility for agricultural tasks of men and women in Carapongo (n = 125)

women are considered better and tougher negotiators. Women are also actively involved in caring for livestock, including feeding, health care and marketing. For some types of livestock such as poultry, women are principally responsible in almost two-thirds of all cases (Figure 3.9).

Domestic responsibilities – Women are totally responsible for fulfilling the role of family caretaker and are responsible for the raising of children and the maintenance of the household. Moreover, women must deal with the ‘double day’. An activity profile developed with women in Carapongo illustrated that women have to combine a large number of activities during the day when they are at home, before and after going to the field. The work at home in the evening is similar to that done in the morning, but women consider it to be heavier because at the time they are very tired. The survey found that, on average, women spend seven hours working in agriculture activities and another eight hours in household activities, whereas men spend nine hours working only in agricultural activities. Women also undertake household management of waste. For instance, in the course of the discussion groups, many women identified the need to learn how to better manage and recycle organic wastes to produce nutrient-rich fertilizer, including the management of wastewater from household sewage and how it could be safely applied for use in vegetables production.

Control over the benefits of production – In group work, participants specified the degree of access and control they believe they have over different productive resources and the benefits of UA (Table 3.4). The discussion group itself determined the definitions of ‘control’ and ‘access’.

Figure 3.9 Women play a central role in urban agriculture in Carapongo

There were many parallels between men and women in reporting on their tenure security (Table 3.5).

None of the types of tenancy mentioned above are gender-specific. Nevertheless, men are more likely to assume responsibility for crop production in Carapongo. Of the 125 households surveyed, men were

Table 3.4 Access to and control over resources in Carapongo

*•••, complete access/control;••, partial access/control;•, limited or no access/control.

Table 3.5 Land tenure classified by gender and reported by the household head

* Posecionarios are those farmers farming land illegally (i.e. on land that has a different designated use).
Source: Survey data.

identified as being primarily responsible for the farm in 70 percent of cases, and women in the remaining 30 percent. Of the men responsible for the farm, 38 percent lived on their own land, but less than half had a formal title. Among the 30 percent of households where women were mainly responsible for farming, the pattern was the same: although 36 percent owned their own land, only 49 percent held formal title to it. The principal problems identified by farmers include contaminated water and the resulting contamination of food, particularly vegetables.

Exactly who has access to and control over inputs for crop and livestock production was related directly to the purpose of production: either for market or for home consumption. Both men and women invest significant inputs for commercial production (i.e. cash crops and animals for sale), whereas it is mostly women who perform subsistence production (small-scale plantings of root and tuber crops, beans, green maize and herbs, along with small animals such as poultry and guinea pigs). The purchase and use of pesticides is mostly done by men, partly because of the physical exertion involved, but also because of the risk of contaminating children and food if women handle pesticides. In some cases, hired labour is used to replace family labour in the use of pesticides and other inputs. Male farmers have better access to agricultural credit because of more frequent interactions with suppliers. In almost 90 percent of cases, men are responsible for crop protection: both purchasing and applying chemicals.

Most community organization activities among the farming population in Carapongo are male dominated. Men predominate in existing associations of water users. Women make up only 10 percent of membership in the committees, but play an important role at the household level with domestic water management in Carapongo.

In terms of gender differences and human capital, both men and women lack access to training and information about crops and livestock husbandry practices, though women are at a much greater disadvantage. Only 22 percent of the farming population had received agricultural training, and of these, 86 percent were men and 14 percent were women. Women were particularly interested in learning more about basic methods for treating animal health problems, whereas men were more interested in information about sources of credit and government training programmes.

Decision making – Women predominate in the marketing of farm products, while both men and women share in the decision to sell and in the control of the proceeds of sales. Where a woman has control over land, she most often has the right to decide on the sale of produce. She also maintains control over the proceeds and, as in one case recounted by a participant, a woman might choose to use the proceeds from the sale of inherited land to pay rent on other land. On the other hand, men are more commonly in control of property and they generally have more authority in decisions about selling land.

Precisely how decisions are made within the family depends on how farming tasks are divided. One of the most important dimensions in the organization of farming systems is the division of labour based on gender. Both the position of individuals within the household and the division of labour affect an individual’s knowledge of the crop–livestock system. This, in turn, determines who makes the decisions. Commercial farmers make more decisions alone as well as fewer in consultation with family members; however, farmers producing for home consumption or who operate non-landowning households with animals make fewer decisions alone and more in consultation with family members. The extent of consultation also depends upon how farmers perceive the importance of the decision. A summary of decision-making is shown in Table 3.6.

Outcomes of urban agriculture in Carapongo

Even though farmers who were interviewed during the study generally declared that UA is not profitable, the majority obtain some income from the sale of crops or animals which makes it possible for them to make a subsistence living. Those farmers who have secure access to the land (either owned or rented) generally have more opportunities to earn a higher income and attain a better quality of life. At the same time, some farmers generate employment for other people, mainly those who live in the vicinity of Carapongo or along the river bank (marginal zone). Many farmers also reported that, thanks to the income generated by UA, they have managed to get more land and build their house with more robust construction materials. Furthermore, they have managed to obtain education for their children, some of whom reach a technical and/or university education.

Most agricultural produce – lettuce, turnip and beets – is sold by farmers and not eaten in their homes. However, in the orchards space is reserved for seeding other products they usually eat. Furthermore, in small-animal husbandry meat and eggs represent other food sources.

Table 3.6 Decision-making matrix in Carapongo households

Source: Results from workshop discussions.

There is also evidence that the practice of UA strengthens family ties which, in some cases, make access to housing or land possible. Also, many farmers, particularly those of rural origin, are supportive of others who share their rural roots. This has enhanced a sense of community amongst farmers in Carapongo. Moreover, farmers feel that despite various inherent problems associated with an agricultural lifestyle, they fulfil an important role in feeding the city. They tend not to feel poor, because they have access to food, work and education for their children.

It was clear that urban farmers recognize that their lifestyle is risky because of exposure to water pollution, solid waste and agricultural chemicals. Another problem they emphasized relates to their exposure to market forces that are unpredictable. Farmers identified a number of reasons: price instability, the lack of information networks, the lack of farmer organization and fear that the quality of their products may be substandard. Finally, many farmers are increasingly concerned that agriculture will no longer be a profitable activity due to the low price of crops as compared to the high cost of inputs.

The assets and strategies developed by urban agricultural households in Carapongo are influenced by many complex institutional and political relations, both of which positively and negatively affect farming families.

In metropolitan Lima, the main wholesale market does not possess an adequate infrastructure and is currently saturated with an excess of agricultural products. This has emerged mainly due to a lack of planning and ineffective regulations, both of which contribute to instability in crop pricing. This instability has negatively affected farmers’ household economies because they do not have other markets where they can sell their products. As noted above, the local government does not recognize UA as an important activity when developing city planning. Therefore, despite the existence of a local ‘municipality law’ that promotes green space within the city, local decision-makers give priority to urban development in favour of UA. This situation generates conflict between the urban farmers and municipal authorities.

Perceptions and prospects of the main stakeholders

Farmers perceive UA as an activity that concerns neither the local authorities nor the central government. Most feel betrayed by failed promises pledged by politicians and believe that the prospects for improvements are uncertain. As a result, most farmers report being pessimistic about their future. They fear the time when they no longer have land to work because of increasing urbanization. Despite this unfortunate situation, older farmers cannot imagine themselves performing any other type of economic activity because agriculture is all they know and at their age they believe they will be unable to find another job. Hence, they expect to continue working in agriculture, possibly in another location. On the other hand, younger farmers are not as concerned and they are willing both to take risks and to undertake another business if necessary. The following quotes represent farmers’ various feelings:

In a couple of years all this land will disappear, because all this is being populated. Everything was land but now it’s all houses. Samuel Yupanqui, Carapongo farmer.

In five years this surely is going to be urbanized because people will no longer want to plant vegetables. Edgard Palacios, Carapongo farmer.

The land is no longer agricultural, this is already housing. I already have noticed that, because the council has forced us to accept that this is now an area of urbanization. Pasión Reynoso, Carapongo farmer.

Five years from now, what can I do – work on the farm until I die? As I say, when you are already old, no one can give you work. Edmundo Vicuña, Carapongo farmer.

Interviews with responsible authorities suggest a dramatic difference in perspective. Their perception of UA is that it is a marginal sector. They give no importance to the conservation of agricultural areas in the district. Instead, their main priority is the physical development of urban infrastructure. From their point of view, the concept of environmental care refers to issues related to the reforestation of the Rímac River and the maintenance of the city’s parks and gardens. In general, they do not fully understand the concept of UA and its importance within the city. The authorities believe that in the medium term agricultural activity will disappear in this area and everything will be urbanized.

There are several key elements that enhance the disagreement between these two main parties. For farmers, there is a high level of mistrust directed at the municipal management as well as a rejection of their authority. Three main conflicts exist:

1 A subject that causes particular conflict is public health, specifically because the farmers do not agree to having to pay municipal taxes for services that are not provided. For example, although local authorities collect taxes for garbage collection in agricultural areas, the municipal garbage truck does not always come.

2 Another point of disagreement with the municipality is sanitation. The farmers say that the municipality supports urbanization without considering that the area is mainly agricultural. Because they are not yet urbanized, these agricultural zones do not qualify for the necessary sanitation infrastructure, so there are no sewerage systems in the agricultural lands. As a result, settlers use the irrigation channels as sewers and garbage dumps. This behaviour contaminates the water use for irrigation purposes, and can damage the farmers’ health.

3 Conflict also exists relating to classification of land use in Carapongo. The Central Municipality of Lima has designated the area as being an urban residential zone, a classification which naturally conflicts with the perspective of the farmers.

CONCLUSION

This research project is one of the first initiatives in the country linking UA with the livelihoods framework. The analysis presented in this chapter provides a holistic and integrated view of the processes by which people achieve (or fail to achieve) sustainable livelihoods. This study has identified the principal problems and opportunities for UA which is being used in a larger initiative, Agricultores en la Ciudad (Farmers in the City) of the Urban Harvest Program of CIP. The initiative is designed to formulate strategies for intervention in the UA sector in Carapongo.

Furthermore, this research contributed to the development of policy guidelines developed by Carapongo’s local government. These guidelines are manifested in a municipal document, Regulation of Organization and Functions, which the management regulates. The document outlines the functions for the new Department of Urban Agriculture, a department that opened within the municipality with management support from the Urban Harvest Program of CIP.

ACKNOWLEDGEMENTS

I would like to gratefully acknowledge the financial support from AGROPOLIS. I thank my thesis adviser Professor Joel Jurado of Pontificia Universidad Catolica de Lima of Peru and my supervisor Dr Gordon Prain, coordinator of the Urban Harvest Program. Many thanks to Lina Salazar for her comments on the final report. And finally, a special thanks to all of the urban producers and local authorities who collaborated during the fieldwork.

NOTE

1Luis Maldonado Villavicencio, MSc. Social Management, Pontifical Catholic University of Peru. Address: Calle 51 no 280 Urb. Corpac San Isidro, Lima, Peru; tel: +51 12260309; email: l.maldonado@cgiar.org

REFERENCES

Antweiler, C. (2000) ‘Urban knowledge for a citizen science, experience with data collection in eastern Indonesia’, presented at the Association of Social Anthropologists (ASA) Conference 2–5 April 2000, Participating in Development, London, UK

Armar-Klemesu, M. (2000) ‘Urban agriculture and food security, nutrition, and health’, in Growing Cities, Growing Food: Urban Agriculture on the Policy Agenda, German Foundation for International Development (DSE), Feldafing, Germany

Chambers, R. and Conway, R. (1992) ‘Sustainable rural livelihoods: Practical concepts for the 21st century’, Institute of Development Studies (IDS), Brighton, UK, Discussion Paper No 296

Durston, J. (1999) ‘Construyendo capital social comunitario’, Revista de la CEPAL 69 (LC/G.2067-P), Santiago, Chile, December

ENCUESTAS (2004) Encuestas realizadas en Carapongo. Linea de Base del Proyecto Agricultores en la Ciudad. Cosecha Urbana-Centro Internacional de la Papa (CIP), Lima, Peru

Hovorka, A. J. (1998) ‘Metodología para el análisis de género para la investigación sobre agricultura urbana’, Cities Feeding People Series 26

Instituto Nacional de Estadística e Informatica (INEI) (2002) Encuesta Nacional de Hogares (ENAHO) IV, Trimestre 2001, Lima, Peru

Martin, A., Oudwater, N. and Meadows, K. (2000) ‘Urban agriculture and the livelihoods of the poor in Southern Africa: Case Studies from Cape Town and Pretoria, South Africa and Harare, Zimbabwe’, Paper presented at the International Symposium ‘Urban agriculture and horticulture: the linkage with urban planning’, 7–9 July 2000, Berlin, Germany. Sponsored by TRIALOG, Vienna, Austria; Humboldt-University, Berlin, Germany; HABITAT, Havana, Cuba

Maxwell, D. (1994) ‘Internal struggles over resources, external struggles for survival: urban women and subsistence household production’, paper presented at the Association for African Studies, 3–6 November 1994, Toronto, Canada

Mbiba, B. (1993) ‘Urban agriculture, the poor and planners, a Harare case study’, 10th Inter-schools Conference on Development, 29–30 March 1993, Philips House, The Bartlett Graduate School, University College London, London, UK, pp129–135

Mianda, G. (1996) ‘Women and garden produce of Kinshasa: The difficult quest for autonomy’, in Ghorayshi, P. and Belanger, C. (eds) Women, Work, and Gender Relations in Developing Countries, Greenwood Press, Westport CT, pp91–101

Mougeot, L. (2000) ‘Urban agriculture: definition, presence, potentials and risks’, in Bakker, N., Dubelling, M., Gründel, S., Sabel-Koschella, U. and de Zeeuw, H. (eds) Growing Cities, Growing Food. Urban Agriculture on the Policy Agenda, German Foundation for International Development (DSE), Feldafing, Germany

Mudimu, G. D. (1996) ‘Urban agriculture activities and women’s strategies in sustaining family livelihoods in Harare, Zimbabwe’, Singapore Journal of Tropical Geography, 17(2), pp179–194

Nugent, R. (2000) ‘The impact of urban agriculture on the household and local economies’, in Bakker, N., Dubelling, M., Gründel, S., Sabel-Koschella, U. and de Zeeuw, H. (eds) Growing Cities, Growing Food: Urban Agriculture on the Policy Agenda, DSE, Feldafing, Germany

O’Reilly, C. and Gordon, A. (1995) ‘Survival strategies of poor women in urban Africa. The case of Zambia’. NRI Socio-economic Series 10, Natural Resources Institute, Chatham, UK

Rakodi, C. (1988) ‘Urban Agriculture: Research Questions and Zambian Evidence’, The Journal of Modern African Studies, vol 26, no 3, pp495–515

Rakodi, C. (1991) ‘Women’s work or household strategies?’, Environment and Development, vol 3, no 2, pp39–45

Sanderson, D. (2000) Cities, Disasters and Livelihoods, CARE, United Kingdom. Also available at www.livelihoods.org/cgi-bin/dbtcgi.exe (accessed 20 November 2007)

Singh, N. and Gilman, J. (1999) ‘Making livelihoods more sustainable’, International Social Science Journal, vol 51, no 4, pp539–545

Soto de, H. (2000) El Misterio del Capital, Empresa Editora El Comercio S.A., Lima, Peru

TECNIDES (2002) ‘Uso de tecnologías no convencionales para el abastecimiento de agua potable y letrinización en zonas rurales y/o urbano marginales’, Programa APGEPSENREM, Convenio USAID-CONAM, Lima, Peru

UNDP (1996) ‘Urban agriculture: food, jobs and sustainable cities’, publication series for HABITAT II, vol 1, UNDP, New York

4
The Social and Economic Implications of Urban Agriculture on Food Security in Harare, Zimbabwe

Charity Mutonodzo1

INTRODUCTION

For decades, poverty, food insecurity and malnutrition in Africa were viewed as largely rural problems (Maxwell et al, 2000); however, the population of many African countries, including Zimbabwe, is becoming more urban. The urban population in Africa grew from 27 to 38 percent between 1980 and 2000, and is expected to reach nearly 50 percent by 2020 (WRI, 1999). Unfortunately, the urbanization of the continent does not mean economic opportunity and prosperity for the majority of Africans. On the contrary, global poverty is becoming more African, more urban and more feminine. Fifty percent of the world’s poor and 40 percent of Africa’s poor live in urban areas (Rabinovitch, 1999).

Earlier research in Zaire has shown that 60–80 percent of the total household budget of the poor is spent on food (Tabatabai, 1993). This finding makes it likely that urban poverty will be manifested at least in part as a problem related to food security. As food security continues to worsen in urban areas, households probably will turn to urban agriculture (UA) as a means of coping. Nevertheless, research on this topic is still relatively new despite an increasing but still limited recognition of UA (Mougeot, 2006). Part of the reason may be that many local governments and municipal decision-makers continue to view agriculture as essentially ‘non-urban’, which blinds them to UA’s contributions to food security and socio-economic improvements. Hence, there is an urgent need to apply food-security monitoring approaches developed in rural contexts (FEWSNET, 2001) to the urban environment.

OBJECTIVES AND HYPOTHESES

Definition of terms

These are the main terms used in the paper requiring definition:

• Urban agriculture (UA): Agricultural activities (farming) in an urban setting. As defined by Mbiba (1995), it is ‘the production of crops and/or livestock on land which is administratively and legally zoned for urban uses’. Mougeot (1995) defines it as ‘the production, processing, marketing and distribution of crops and animals and products from these in an urban environment using resources available in that urban area for the benefit largely of the residents of that area’.

• Food security: The United States Agency for International Development (USAID, 1996) defines food security as a situation where all people at all times have both physical and economic access to sufficient food to meet their dietary needs for a productive and healthy life. However, for the purpose of this limited study, the researcher opted to use a narrower measurement of household food security. The indicator of food security was based on energy intake, and people who met 70 percent of their energy requirements from cereals were considered food secure. Energy adequacy in the current study is taken as sufficiency of kilocalories (kcal) consumed in relation to requirements.

• Vulnerability: Moser (1996) defines vulnerability as the ‘insecurity of the well-being of individuals, communities or households in the face of a changing environment’. Vulnerability can include the risk of, or susceptibility to, food insecurity and can result from either chronic or transitory conditions.

Objectives

The research had five objectives, which were designed to lead to an understanding of the socio-economic implications of UA in food and livelihoods security:

1 To characterize sampled urban households in terms of their vulnerability, poverty levels and gender of household’s head among those that engage in UA compared with those that do not. This objective remained unchanged and was fulfilled during research.

2 To assess food security and livelihoods in terms of UA participation and other demographic, socio-economic and geographical characteristics.

3 To measure production performance of urban farmers compared with smallholder farmers throughout Zimbabwe.

4 To identify key factors accounting for variance in crop production performance and crop choices.

5 To explore policies and institutional innovations that are likely to safeguard and enhance the contribution of UA to the food security of the vulnerable urban population.

Unfortunately, a number of the objectives had to be modified during the research process due to a number of external influences. For instance, objective 3 had to be dropped due to changes that were brought about by Operation Restore Order, a project started on 19 May 2005 by the Government of Zimbabwe to clean up its cities. The operation resulted in the destruction of homes, business premises and vending sites, including most of those who were practising UA who were interviewed in April and early May 2005. Those affected moved to stay with friends and family or found temporary shelter in churches. Because many of the displaced persons in this sample were scattered in different locations and were highly mobile, it became impossible to measure their production performance.

Meanwhile, the fourth objective was also modified to mirror the changes brought about by Operation Restore Order because it was impossible to assess crop performance and farmer choices. This objective, therefore, became more generic, being restated as: To identify key factors accounting for variation in food security and livelihood status in an urban setting.

METHODS

A livelihoods-based vulnerability analysis framework, based on the assessment used during the Zimbabwe Vulnerability Assessment (ZIMVAC, 2003), was applied to guide data collection and analysis. This framework seeks to clarify the mechanisms by which people obtain access to food and other essential resources and services within communities, by offering a framework that allows for interpretation and monitoring of data such as market prices. It is a framework that provides a geographic and socio-economic basis on which to understand the strategies that people use to meet their basic needs (FEWSNet, 2005). Data sources included a secondary data review, household surveys, focus-group discussions and key informant interviews. The widely accepted minimum daily energy requirement of 2100 kcal per day for an average adult (FAO, 1993; WFP, 2000; UNHCR and WFP, 1997) was used as a benchmark for the study. However, since other nutrient requirements are such that 10–12 percent of energy in the diet should be in the form of protein and at least 17 percent of the energy in the diet should be in the form of fats, by difference 70 percent of the daily energy requirements of 2100 kcal come from cereals. Food security is achieved when at least 70 percent of the energy requirement is met.

Four focus groups for soliciting the community’s perspectives on the division of labour, income and food sources, coping strategies and challenges faced were assembled. The groups consisted of 8–10 men and women. Each group included a mix of people who were and were not involved in UA. A gender-specialist facilitator guided the discussion to ensure it stayed on topic, while the researcher and a note-taker recorded the discussion. A number of participatory approaches were used: seasonal calendars for understanding seasonal aspects, timelines to identify events affecting well-being, proportional piling to determine relative magnitudes and Venn diagrams to depict the various groups’ relationships with other institutions. The research was done with the participation of a number of teams within the departments of Agricultural Research and Extension and the Ministry of Agriculture (both the National Early Warning Unit and the Harare Urban Extension teams). Both work in the areas of UA, food security and livelihoods research.

A review of the available literature on UA in general and in Harare in particular was carried out, using both published and unpublished documents. Six weeks of observation and unstructured interviews with city authorities and households involved in UA were a starting point for the research.

Study sites

Harare Metropolitan Province is divided into four districts: Harare Rural, Chitungwiza, Epworth and Harare Urban. Harare Urban had the largest proportion (76 percent) of the population in the 2002 national census. The present study took place in the Harare Urban district, which is made up of 48 wards spread over 27 suburbs. In this district, on-plot agriculture exists and was provided for in low-density suburbs such as Borrowdale. Meanwhile, off-plot agriculture is found across the city in suburbs of varying densities, with varying degrees of intensity and tenure regimes. Peri-urban agriculture (UPA) exists in areas surrounding the city and some of the lands are smallholdings zoned for agricultural purposes.

A total of 16 suburbs where most of UA was assumed to be taking place were selected for the study. Generally speaking, high-income residents tend to inhabit low-density areas, medium-income residents live in suburbs of all density types and low-income residents mostly inhabit high-density suburbs. The distribution of the 372 households among the selected suburbs is shown in Table 4.1.

A household survey which included a consent form was used to collect information from sampled households. It collected information on demographics, assets, support to UA, UA-related activities and formal and informal employment. It also collected information on coping strategies related to income, expenditures and consumption.

Survey sampling and sample size

The study used a two-stage sampling strategy, where the first stage was the ward or suburb. A list of these was obtained from the Central Statistics Office, representing wards used for the 2002 population census. These were deliberately chosen on the basis of residents’ participation in UA. The second stage for sampling was households, chosen randomly by selecting people who were in their fields during the time of the interview.

The minimum sample size required was calculated using the formula (Equations 4.1, 4.2) for large samples given by Poate and Daplyn (1993):

Table 4.1 Distribution of sampled households by suburb name and type

where n is the minimum sample size required; z is 1.96, the value of z at the 95 percent confidence interval; c is the variation within the population, which has been assumed to be 60 percent; and x is the expected level of accuracy, which has been estimated at 6.1 percent.

Therefore:

Conducting the study

The questionnaire was pre-tested in three different areas of the city, modified and delivered by eight surveyors selected and trained by the researcher. The training sessions served to explain why the survey was being conducted, its relevance to urban and national development and the rationale for sampling. Surveyor training and pre-testing of questionnaires took place during a two-week period of November 2005. After training, enumerators received notes explaining the meaning, relevant concepts and definitions for each question, as well as the techniques for data collection. Enumerators helped translate the questionnaire into Shona, the language spoken in the study area. The survey was carried out over a period of two weeks in December 2006. Questionnaires were field-checked by the academic supervisor of this work and subsequently by the researcher.

Data management

Data entry took four weeks using the Statistical Package for Social Sciences (SPSS, version 10) for data entry. To minimize errors, a double-entry system was used and all discrepancies were corrected by referring back to the questionnaire. Data cleaning on the original data files was also conducted in SPSS (version 10). Tabulation and further analysis were done in SPSS versions 10 and 11. A total of 368 questionnaires remained from the original 372 after data cleaning.

Analytical techniques

Livelihood and food security analyses were conducted by looking at food and income sources, as well as amounts, expenditure patterns, coping strategies and required daily allowances compared with energy adequacy.

Statistical analyses

Statistical analysis used in this research (see below) involved descriptive statistics such as mean, median, proportions and odds ratios. These were used to compare those households practising UA with those that were not. Independent t-tests were applied similarly. Associations between categorical variables were investigated using chi-square tests. A P value of 0.05 or less was considered significant. A linear regression model was fitted to determine factors affecting food security. A binary logistic regression model was fitted for participation in UA, which has a binary outcome (yes or no). Odds were calculated for the categorical variables such as sex, educational level of household head and suburb type.

Food security analysis

Annual household food requirements

The widely accepted minimum required intake of 2100 kcal per day for an average adult was used as the benchmark during this research (WFP/UNHCR, 1997; WHO, 2000; WFP, 2000). This is the minimum requirement for adults against which people’s access to food was compared. However, it is acknowledged that differences in requirements between households with the same total size, but different demographic composition, are masked by using this method. Based on the information from these studies, we assumed that cereals supplied 70 percent of the required food energy, or 1470 kcal/person per day, for an average household member. This is equivalent to 148 kg of cereals per year. These energy requirements were disaggregated by age and gender, as is indicated in Table 4.2.

The total number of household members was calculated in the age and gender categories listed above. The number of people in each category was then multiplied by the quantity in kilograms of cereal required by that age and gender group in order for them to meet 70 percent of their minimum cereal needs.

Table 4.2 Household energy requirements by age and gender

Total cereals accessed in 2005–2006

Total cereals accessed by households in the 2005/06 marketing year was calculated by summing all sources of cereals:

• Personal production consumed: The amount of maize accessed – the most common cereal – was obtained by subtracting the amount sold, exchanged or given away from the total amount harvested from urban areas, plus the rural production consumed in urban areas. Tubers were converted to an equivalent amount of maize, then added to other cereals to give the total kilograms of cereal consumed from people’s own production. The conversion factor used was 1140 kcal/3630 kcal = 0.31; that is, kilograms of tubers were multiplied by 0.31 to get the maize equivalent. This was divided by 12 to get the monthly personal production contribution.

• Purchased food: Information was provided by households related to the quantity of food purchased. This was converted into kilocalories for the study month.

• Direct sources of food: Cereals – maize, sorghum and millet, wheat, rice and maize meal – obtained from relatives, churches or other donors and other friends other than derived from personal production were calculated for eight months and monthly quantities obtained.

Food security status

Food security status was calculated using the contribution of all cereal foods available to the family’s household calorie intake during the month of the survey. The quantity for each household per month as well as for each food item was calculated for the survey month. Then, using calorie requirements for different age groups, the ideal monthly energy intake for each household was calculated.

Limitations of the study

The study does not claim to be exhaustive and flawless, because there are some assumptions and drawbacks that one needs to be aware of when using the results (as shown below in Research Findings). Interpretation of results should be limited by the fact that these data are cross-sectional because of the single round of data collection. Also, consumption data do not capture seasonality or whether households experienced chronic or transitory food insecurity.

The household survey only recorded quantities of cereals and tubers accessed. It did not capture other foods consumed, such as legumes, meat, oils and fats or greens. It was assumed that 70 percent of required energy is derived from cereals (1470 kcal). This presented two problems: first, how to determine the cut-off point; and second, whether it made sense, given variation across households as well as individuals’ dietary requirements.

Another limitation is related to the survey sample. The sample was chosen purposively and is representative of those suburbs where UA is undertaken. As a result, the descriptive statistics reported in this study are not necessarily representative of all the households in Harare, which include neighbourhoods where UA is limited. Another important point is that a precursor of inclusion was that those people studied had to be occupying a household. This meant the homeless, as well as street children, were left out regardless of whether or not they were practising UA.

RESEARCH FINDINGS

Objective 1: Characterization of sample

Table 4.3 shows basic demographic characteristics of the sampled households (N = 368) by gender of head of household and participation in UA. A total of 69 households, or 18.8 percent of the total sample, were headed by females.

Table 4.3 Descriptive characteristics of surveyed Harare households by the gender of the household head and by participation in UA

Table 4.4 Factors associated with practising UA (logistic regression model)

In this study, households headed by males were, on average, significantly larger than households headed by females.

Though female-headed households were smaller in size than male-headed households, they had significantly higher dependency ratios. Maxwell et al (2000) noted a similar finding in Kampala, Uganda, in the early 1990s. Households headed by females contained a much larger proportion of relatives and grandchildren than those headed by males. Female heads tend to be older (mean age 50.9 years) than their male counterparts (mean age 45.9 years).

Results for the logistical regression on UA are shown in Table 4.4. A correct prediction value of 89.9 percent was obtained. This meant that the variables used were very good at statistically predicting the observed outcome of practising UA. Suburb type (P = 0:004), educational level of head (P = 0:04), household size (P = 0:05) and house ownership (P = 0:003) were significantly associated with participation in UA (see also Table 4.4).

Suburb type

Households in the high-density suburbs were more likely to practise UA than their counterparts from low-density suburbs, a finding consistent with ZIMVAC (2003), Mwakiwa (2004) and Mudimu et al (2005). From the odds ratio given by exp β, those from high-density suburbs were 21 times more likely to practise UA than their low-density counterparts.

House ownership

Owner-households were more likely to practise UA than those who were staying in rented accommodation, which corroborates what Mudimu et al (2005) found. The odds ratio for this variable is 0.22, meaning that owner-households were approximately 22 percent more likely to participate in UA compared with those in rented accommodations.

Household size

Larger households were more likely to participate in UA than small households. The Zimbabwe Vulnerability Assessment Committee makes the same observation (ZIMVAC, 2003).

Educational level of household head

Households where the head had up to primary level education were approximately eight times more likely to practise UA than those whose heads had educational levels beyond primary level.

Objective 2: Assessment of food security

Overall, based on the benchmark of having 70 percent of the recommended 2100 kcal/aeu per day (where ‘aeu’ is adult equivalence unit) for Zimbabwe, 92 households, or 25 percent of the sample, were considered to be food secure. The percentage of household energy requirements for food-insecure households increased with expenditure, from 66.9 percent for the lowest expenditure quartile to 80.9 percent for the highest. A rather surprising result is that the mean adequacy for female-headed households was higher than for their male counterparts. This was the case both for households that were food secure (107.3 percent of caloric requirements met compared with 106.9 percent) and for those that were food insecure (79.1 percent compared with 73.8 percent).

A summary of the relationships between food security and the practice of UA is given in Table 4.5. Among those that practised UA, about twice as many households (26.3 percent) were food secure compared with 13.5 percent among those not practising UA. Conversely, households defined as food secure and practising UA met 76 percent of their requirements as compared to food-insecure households not practising UA, who met 65 percent of their energy requirements.

Households whose household head contributed to UA on a full-time basis tended to meet more of their requirements (79 percent of caloric requirements vs 66 percent for households whose heads do not contribute to labour). Households represented with owners met more (79 percent) of their requirements, compared with 70 percent where housing is provided by employers (signifying less security of land tenure). Families coped with food stress by regularly reducing the number of meals eaten per day, as well as by rationing quantities of food eaten per meal.

Urban agriculture contributes significantly to urban food security in the period during which grain produced by UA lasts in a household. A total of

Table 4.5 Comparing UA and food security

32.9 percent of those households practising UA had enough grain to last one to three months, while 16.9 percent of the households had enough grain for four to six months. Furthermore, close to 9 percent of the households had enough grain to last seven to nine months and 14.8 percent of the households had enough grain for 10–12 months. Roughly 27 percent of the households had no grain at all, either because they did not produce maize that season or because the small amount they produced was consumed in less than a month.

Across all households surveyed, food is the largest item in the household budget (49 percent), followed by transportation (35 percent). Figure 4.1 shows that more than 70 percent of total expenditure of households in the lowest expenditure quartile went to food, compared with less than 20 percent in the highest quartile. The poor, who pay a disproportionate part of their income on food, end up being vulnerable to any unanticipated price changes or problems. This finding is consistent with the Engelian relationship between income and the percentage allocated to food (Colman and Young, 1996;

Figure 4.1 Average share of total budget allocated to food, by expenditure quartile

Deaton, 1997). As income increases, the percentage of the household budget allocated to food falls.

Harare urban diets were shown to be more varied than rural diets (ZIMVAC, 2003), with tubers and cereals dominating (62.2 percent), followed by meat and fish (13.7 percent), then by fats, oils and vegetables (15.7 percent). Expenditure-reduction strategies included cutting spending on consumption and purchases mainly by curtailing money spent on transportation and health care to make ends meet.

Objective 4: Factors associated with food security and energy adequacy

The results of the energy adequacy regression using household calories per adult equivalence unit are given in Table 4.6. Adjusted R2 was 0.487, meaning 48.7 percent of the variation in energy adequacy is explained by the given independent variables. The Durbin-Watson statistic was 1.653. This suggests that there is a slight positive correlation between adjacent residuals, since the value is close to 2, signifying that the residuals are uncorrelated. This was also confirmed prior to moving on to regression analysis, where residuals were independent of each other.

Some variables showed a significant link with household energy adequacy. Practising UA was found to be related to energy adequacy. This is consistent with a number of earlier studies on UA (Mbiba, 1995; Mutangadura and Makaudze, 1999; Mwakiwa, 2004; Mudimu et al, 2005). Correlations also

Table 4.6 Factors associated with adult diet and energy adequacy (linear regression model)

Adjusted R2 = 0:487; Durbin-Watson = 1.653; *Significant at the 0.05 level.

were found between energy adequacy and household size, the head’s contribution to UA labour, plot cultivation, per capita expenditure (which is being used as a proxy for income) and house ownership. Age and gender of the head of household and the type of suburb were not significantly associated with household energy adequacy.

As expected, households with more members were more likely to have lower energy adequacy values than those with fewer household members. This finding is consistent with ZIMVAC findings in both their urban and their rural assessments (ZIMVAC, 2003). Households whose heads contributed to UA labour were more likely to meet more of their energy requirements and be more food secure, according to the bivariate analyses already mentioned.

Urban agriculture taking place in spaces officially designated for UA – in contrast to in open spaces and small yards – was associated with a significantly greater energy adequacy. Therefore, households practising UA on plots are more likely to be food secure than those who are using open spaces and yards within residential areas.

Home owners had significantly higher energy adequacy compared with lodgers, leaseholders and those in accommodations provided by their company or employer. This might be because households that own their homes take much of the land on-plot if they are subletting from other renters and have greater claim to open spaces by virtue of long periods of residence in the suburbs (Mwakiwa, 2004).

Participation of household members in informal employment (vs the formal sector) was associated with greater energy adequacy. This probably reflects the fact that incomes from formal employment are much lower than returns from informal activities given the current status of the Zimbabwean economy.

Objective 5: Policies that enhance urban agriculture

Figure 4.2 is a representation of focus-group data and secondary data from the Municipal Development Partnership (MDP, 2003), the central government and Harare City Council. The views of both UA practitioners and policy makers from the central government and the municipal administration were included.

Figure 4.2 Mapping of views of policy makers and practitioners on UA

Their views of UA are categorized by the author as either positive or negative. Interestingly, some of the results were not as anticipated.

The most predominant issues identified were included in the matrix. Those in Quadrant 1 are considered by all parties to be the positive contributions of UA. Quadrant 2 represents issues that the policy makers consider to be good, while practitioners view them as unconstructive and negatively affecting UA. Issues that are seen as harmful and negative by policy makers while they are viewed as helpful by the practitioners are given in Quadrant 3. Quadrant 4 represents issues identified by all parties as the problems of UA that negatively affect society, the environment or the planning process.

Quadrants 1 and 3 of Figure 4.2 represent issues that both groups agree upon, which can represent the basis for dialogue that seeks to ensure common agreement on UA. Quadrants 2 and 4 are boxes of conflict, divergent views and issues that need to be resolved. It is in these boxes where there is tension between the practitioners and the metropolitan and central government policy makers. If issues in these areas are resolved, there could be a more peaceful coexistence between the two parties. Policy should aim to lessen the rift and stand-off in these two quadrants by increasing dialogue.

From the above analysis, it can be concluded that most of the existing policies and legislation address the negatives listed in Quadrant 3, but not as many issues from Quadrants 1, 2 and 4 are being addressed. For example, for optimum management of land where investment is needed for such things as conservation works, people need land titles and security of tenure. While full tenure is unlikely to be feasible in the short term, one possible solution would be to provide secure yet informal tenure – such as by issuing temporary permits.

The Grain Marketing Board (GMB), which collects surplus grain from farmers for redistribution, was able to purchase more than 60 percent of the annual national maize production in the early 1980s (Figure 4.3). This marks an improvement, because the GMB only managed to procure an average of 34 percent per year during the 1970s. However, since 2000, the GMB was only able to collect 18 percent of the total national production of maize per annum because of lack of surplus.

This poses a challenge regarding how to redistribute grain from surplus areas to urban centres. The situation necessitates the development of an integrated strategy of urban and rural development. Food security in Zimbabwe is achievable through a combination of urban production and improvements in the infrastructure for food distribution from rural areas.

Respectively, 62 and 42 percent of respondents cited shortages of inputs as well as high costs as problems. This suggests the need for support for production inputs. For example, the city government can play an important role in providing enough market stalls from which to sell their crops. The imposition of generic and restrictive policies on UA has not succeeded. Harare City Council needs to replace these with policies that actively regulate and guide UA. If such policies were implemented, producers would be able to fully

Source: GMB and CSO.

Figure 4.3 Trends in maize intake via the Grain Marketing Board

realize their potential and possible side-effects would be avoided (MDP, 2003). One option for Harare would be to create a platform on urban food and environmental policy issues that includes UA. In Bulawayo, the city has undergone a project called the Multi-stakeholder Process for Action Planning and Policy Design. Its purpose is to bring multiple stakeholders around the same table to discuss issues of common interest, challenges and solutions. A similar process for Harare City would improve the quality of decision making. The city can act as a convenor and facilitate the process of the development of UA. Such a municipal platform could lead to the development of a comprehensive, city-based food security plan as it has in Buluwayo. It could also stimulate the integration of UA into urban zoning and development plans, to advise on the revision of city by-laws related to UA, and to coordinate the planning of action programmes, as is the case in Kampala, Uganda.

CONCLUSIONS

As the population of Zimbabwe becomes increasingly urban, food insecurity becomes more of an urban issue. This study showed that UA can contribute significantly to food security. Therefore, policy guidelines – along with guidelines for sustainable use of open spaces for UA – should be both very clear and readily available. This is particularly relevant, given that 80 percent of UA takes place in open spaces and on vacant public land. Also, it is important to support UA in the form of production inputs, agricultural extension, viable prices and places for marketing.

The regulatory environment should not constrain the activities of the urban self-employed, given that close to 50 percent of the income of the lowest quartile income group comes from the informal sector. Historically, the municipality has tended to view the informal sector as a problem to be solved rather than as a viable resource for dealing with unemployment. Lately, the central government has also developed an intolerance for this sector, resulting in harassment of informal entrepreneurs. Female-headed households tended to have lower incomes because they had fewer assets, especially human assets such as education, labour and marketable skills. Therefore, investment in education for girls is likely to reduce poverty in the long term.

The municipality should invest in publicizing UA legislation and invest in dialogue with urban farmers. This is crucial, given that more than 40 percent of the people practising UA were unfamiliar with any such related by-laws, and 22 percent considered existing legislation to be hostile to UA. Given the potential environmental and health hazards that can emanate from UA, municipal administration should actively regulate, manage and guide UA. Only by doing this will they be able to fully realize its potentials and prevent negative side-effects.

ACKNOWLEDGEMENTS

I want to thank Canada’s International Development Research Centre for fully funding my research project through its AGROPOLIS award programme. I am grateful to Agricultural Research and Extension Services (AREX) and the City of Harare for providing an environment conducive for my research. I also want to thank my supervisors, Dr R. Mano and Mrs C. Muchopa, and the many Harare residents who answered my questions.

NOTE

1 Charity Mutonodzo; MSc Agricultural Economics; 10 Hillside Gardens, 15–17 Ferreira Avenue, Hillside, Harare, Zimbabwe; tel: +263 4 747800; email: mutonhodzac@yahoo. co.uk

REFERENCES

Colman, D. and Young, T. (1996) Principles of Agricultural Economics: Markets and Prices in Less Developed Countries, Department of Agricultural Economics, University of Manchester, UK

Deaton, A. (1997) The Analysis of Household Surveys: A Micro-econometric Approach to Development Policy, World Bank, Washington, DC

FEWSNet (Famine Early Warning Systems Network) (2002) Emergency Food Security and Vulnerability Assessment Report, Washington, DC

FEWSNet (2005) Approach to Livelihoods-Based Food Security Analysis Methodology, FEWSNET, World Bank, Washington, USA

Food and Agriculture Organization (FAO) (1993) Food and Nutrition in The Management of Group Feeding Programmes. FAO, Rome, Italy

Maxwell, D., Levin, C., Amar-Klemesu, M., Morris, S., Ruel, M. and Ahiadekeke, C. (2000) ‘Urban livelihood and food and nutrition security in greater Ghana’, Research Report No 112, International Food Policy Research Institute, Washington, DC, USA

Mbiba, B. (1995) Urban Agriculture in Zimbabwe: Implications for Urban Management and Poverty, Avebury, Hants, UK

Moser, C. (1996) ‘Confronting crisis: A comparative study of household responses to poverty and vulnerability’, in Four Poor Urban Communities, World Bank, Environmentally Sustainable Development Studies and Monographs Series No 8, Washington, DC, USA

Mougeot, L. J. A. (1995) Urban Food Production: Evolution, Official Support and Significance (with Special Reference to Africa), Urban Agriculture Notes City Farmer: Canada’s office of Urban Agriculture; IDRC, Ottawa, ON, Canada

Mudimu, G., Matinhure, N., Mushayavanu, D., Chingarande, S., Toriro, P., Muchopa, C. (2005) Research Project on Improving Access to Land for Urban Agriculture by the Urban Poor in Harare, IDRC, University of Zimbabwe, Harare, Zimbabwe

Municipal Development Partnership (MDP) (2003) ‘The Harare declaration on urban and peri-urban agriculture in Eastern and Southern Africa’, www.mdpafrica.org.zw/Publications/Harare%20Declaration.pdf (accessed 15 September 2008)

Mutangadura, G. and Makaudze, E. (1999) ‘Urban vulnerability to chronic poverty and income shocks and effectiveness of current social protection mechanisms: The case of Zimbabwe’, Harare Consultancy Report (draft) submitted to the Ministry of Public Service, Labour and Social Welfare, and the World Bank, Harare, Zimbabwe

Mwakiwa, E. (2004) ‘Evaluation of the socio-economic determinants and benefits of urban agriculture: the case of Kadoma, Zimbabwe’, MSc Thesis, Agricultural Economics, University of Zimbabwe, Harare, Zimbabwe

Poate, C. D. and Daplyn, P. F. (1993) Data for Agrarian Development, Cambridge University Press, Cambridge, UK

Rabinovitch, J. (1999) ‘Practical approaches to urban poverty reduction’, paper presented by Urban Development Team at International Forum on Urban Poverty, Governance and Participation, Florence, UNDP, November 1999

Tabatabai, H. (1993) ‘Poverty and food consumption in urban Zaire’, Cornell Food and Nutrition Policy Programme Working Paper 46, Cornell Food and Nutrition Programme, Ithaca, NY, USA

United States Agency for International Development (USAID) (1996) ‘Performance indicators for food security’, Final output of the December 1995 USAID Workshop on Food Security Performance Measurements, USAID, CDIE, Washington, DC, USA

World Food Programme (WFP) (2000) Food and Nutrition Handbook, WFP, Rome, Italy

WFP (2005) Emergency Food Security Assessment Handbook: Methodological Guidance for Better Assessments, WFP, Rome, Italy

WFP/UN High Commission for Refugees (UNHCR) (1997) Guidelines for Estimating Food and Nutritional Requirements, WFP, Geneva, Switzerland

World Resource Institute (WRI) (1999) 1998–1999 World Resources: A Guide to the Global Environment, WRI, Washington, DC, USA

Zimbabwe Vulnerability Assessment Committee (ZIMVAC) (2003) Zimbabwe Urban Areas, Food Security and Vulnerability Assessment, Harare, Zimbabwe

5
Urban Agriculture and Food Security in Lilongwe and Blantyre, Malawi

David Dalison Mkwambisi1

INTRODUCTION

Despite continued economic growth around the world, food insecurity remains a pressing problem in many parts of Africa (Garrett and Ruel, 1999; Maxwell, 1999; Mougeot, 2005a; UN-HABITAT, 2006). For example, the World Food Programme (FAO, 2001) and Food and Agriculture Organisation (FAO, 2003) both estimate that approximately 800 million people are unable to obtain an adequate and secure supply of food year round. The FAO (2002) further estimates that about 33 percent of people in sub-Saharan Africa are undernourished. The UN-HABITAT (2006) reports that the percentage of urban residents in sub-Saharan Africa is expected to rise from 30 to 47 percent of the total population. This will bring about new and very severe challenges for urban policy, especially when trying to ensure household food security (Huddad et al, 1998).

UA is one strategy where recent research suggests that food insecurity could be tackled (Pothukuchi and Kaufman, 1999; Mougeot, 2001, 2005a, b). UA has been shown to be an important source of food in developing countries and a critical food ‘insurance policy’ for poor urban households (Mougeot, 2000; Nugent, 2000). UA also affects household nutrition as it provides a source of fresh, locally grown crops that meets the micronutrient requirements in poor households’ diets (FAO, 2001; Maxwell, 2001). Other benefits have been documented, such as increases in household incomes due to the sale of urban agricultural produce (Sanyal, 1985; Smit, 1996; Sabates et al, 2001; Henn, 2002; IFPRI, 2002).

Nowhere are these issues more pressing than in Malawi, where persistent poverty, HIV/AIDS and rapid urbanization have brought huge numbers of poor, unemployed and hungry people into the cities. Currently, over 55 percent of the population is living on less than US$1 per day (Government of Malawi, 2005a, b; USAID, 2005). Despite this, there is a real gap in Malawian policy, as UA is not seriously considered by the Malawian government as a viable livelihood option.

For example, despite UA being mentioned within the ‘Town and Country Planning Act’ (Government of Malawi, 1998a), there are no practical regulations to guide and support urban food production (Kwapata et al, 2001). The latest policy instruments put forward by the government still favour rural farmers but do not support or acknowledge urban food producers. As a result, the Malawi Growth Strategy (Government of Malawi, 2004), Malawi Poverty Reduction Strategy Paper (Government of Malawi, 2002) and previous models to reduce poverty in Malawi all fail to recognize the potential of UA.

The understanding of UA as a concept in Malawi is so limited that the mention of it is often based on speculation without any real awareness of who urban farmers are, what crops they produce, or the importance it plays in the local economy (Mkwambisi, 2005). Therefore, the purpose of this study was to evaluate the contribution UA makes to food security in Malawi. The information presented is designed to contribute to policy recommendations that are empirically grounded.

LITERATURE, CONCEPTS AND OBJECTIVES

The 1996 World Food Summit in Rome defined food security as a state when all people, at all times, have both the physical and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life (FAO, 1996). The FAO (2002) explores this definition in more detail, pointing out that food security has two components: it is a phenomenon relating to individuals whereby the nutritional status of the individual household member is the focus; and food security should highlight the risk that this nutritional status may be undermined.

Garrett (2000) and the World Bank (1986) have developed similar themes by pointing out that urban food insecurity and malnutrition may be different from rural food insecurity because most urban dwellers depend on incomes to purchase their food. This relatively recent discussion hearkens back to Sen’s (1991) approach that considered food security as a function of a person’s – or household’s – bundle of ‘food entitlements’. According to this argument, people have different ways of meeting their household’s food entitlements using the totality of rights and opportunities that they have (Sen, 1991).

Broadly speaking, Sen identified four types of entitlement: direct or production-based entitlement, which occurs when a person consumes or sells the food they produce; labour-based entitlement, which is obtained through working for a wage and purchasing food from the market; trade-based entitlements obtained through sale or barter of assets; and transfer-based entitlement, where entitlement is transferred through charity or food aid. Sen’s work is important because it focuses on the ability of people to command food through legal means available to society (Sen, 1991). The entitlements framework puts the economic ability of individuals or families to deal with their own problems at the centre of the food security debate (Fraser et al, 2006).

The entitlement approach is useful for policy makers as it does not presuppose how people should be living, or the types of food they should be eating. Instead, it attempts to provide a framework that measures how much freedom an individual or a family has. Sen (2000) indicates that this freedom involves both the processes that allow freedom of actions and decisions, and the actual opportunities that people have, given their personal and social circumstances. Seen in this light, Sen’s entitlement framework can help explore the complexity of UA. Food grown in the city can be a source of a range of different entitlements. It may provide a direct entitlement for those urban farmers who consume the food they produce. It can also provide two different types of indirect entitlement: first by providing marketable produce that a poor family could sell; and second by providing a source of paid employment for workers on larger-scale urban farms. Finally, if UA is used by charities, community- and faith-based organizations (CBOs and FBOs), it could also be used as the basis for transfer entitlements.

RESEARCH METHODOLOGY

Study location

Research was carried out on urban farms within the cities of Blantyre and Lilongwe in Malawi. They were useful for the purposes of this research because they are considerably different in terms of economic and demographic structure. Blantyre had a population of 711,233 in 2005, being the largest commercial and industrial centre in the country (Figure 5.1). It covers 228 km2of hilly ground, with a reasonably temperate climate, where 71 percent of the city’s residents live in unplanned settlements characterized by poor living conditions (Government of Malawi, 1998b).

Lilongwe lies on the Lilongwe-Kasungu Plain in the central fertile region of the country at an altitude of 1100 m above sea level. The capital of Malawi since 1975, Lilongwe is an administrative and commercial centre. Vast amounts of arable land and a low population density have resulted in an active UA and peri-urban agriculture (UPA) sector. According to the Government of Malawi (2005), Lilongwe has a population of 669,114 residents, where 40 percent live below the poverty line. Within these two cities there is a range of low-, medium- and high-income neighbourhoods whose details are given in Table 5.1.

Data collection

Primary data was collected based on sustainable livelihood analysis through a structured questionnaire (Scoones, 1998; Ellis, 2000; Solesbury, 2003) that was administered to 330 heads of households who had access to agricultural land

Source: Government of Malawi, 2005a.

Figure 5.1 Modified map of Malawi showing Blantyre and Lilongwe

within the two cities. Specifically, 165 households were studied in Blantyre (112 male-headed and 53 female-headed) and 165 in Lilongwe (131 male-headed and 34 female-headed). The sample was stratified into high-income (n = 70: 68 males, two females) and low-income households (n = 120: 72 males, 48 females), as well as middle-income (n = 140: 127 males, 13 females) and also between male- (n = 243) and female-headed (n = 87) households. Due to several similarities in medium- and high-income households, results

Table 5.1 Household characteristics in urban Malawi

Source: Government of Malawi (1998b).

have been presented for two groups only (low and high). Income status in urban Malawi was based on the government definition of poverty (Government of Malawi, 1998b).

Low- and high-income neighbourhoods were identified using consumption and expenditure data from the 1997–1998 Malawi Integrated Household Survey (Government of Malawi, 1998b), where a set of daily basic food and non-food requirements of individuals were identified. Respondents were selected using the snowball sampling method amongst low- and high-income households. Within each neighbourhood, key informants were identified through community workshops. Each respondent was then asked to identify two other urban farmers. This process was continued until 165 interviews were completed in each city.

The questionnaire used the ‘five practical methods’ outlined by Neefjes (2000). This method proposes that data be collected on basic household demographic characteristics such as marital and socio-economic status; income and employment; food frequency, allocation and food-related coping strategies; and urban agronomical practices. Logistic regression analysis and t-tests for independent samples were used to determine which groups (income, gender, education or location) benefited the most in terms of food security from UA.

Determining urban agriculture’s contribution to food security

To assess UA’s contribution to direct entitlements, respondents were asked to calculate the amount of food consumed that came from their own urban agricultural plots. Specifically, this research analysed the amount of maize each household produced and converted non-maize crops into ‘cereal equivalents’. For example, fresh cassava and sweet potato yields (in kilograms per hectare) were considered to be worth 25 percent of the equivalent fresh weight in cereal (FAO/WFP/GoM, 2005).

Results were then compared with the Government of Malawi’s recommended annual consumption levels of 181 kg of cereal per capita.2Finally, to facilitate comparison with other entitlement strategies, data on food consumption were converted into a monetary value by using the 2005 market prices for maize in May of 2005 (21 Malawian kwacha/kg).

RESULTS

Results of the study revealed that the majority of households in Malawi were not dependent on UA as the main household food entitlement.

The findings from our study contrasted with other research conducted elsewhere in Africa which suggested that urban farming was an important source of food for urban populations (Maxwell, 1995). Our study showed that 70 percent of all households gave formal employment as the most important component in their livelihoods’ strategies, informal employment came second, while UA was ranked third.

Table 5.2 Main livelihoods’ sources as identified by household heads in urban Malawi

Despite this low rating of UA, considerable variation between groups was observed. For example, while UA contributed 9 percent of high-income households’ livelihoods, it provided 42.5 percent of low-income households. Thus, low- and high-income households are statistically different (P < 0.05), as shown in Table 5.2.

Results presented in Figure 5.2 show that high-income households were more productive in terms of the amount of food they produce than lower-income

Figure 5.2 Average cereal equivalents (expressed in kilograms per member of each household per year) from UA

Figure 5.3 The relationship between education and UA to total household’s food bundle

groups. This is due to the fact that higher-income groups had more access to larger plots and inputs such as fertilizer.

Female-headed households sold more of their crops as compared to other groups. Similar findings in Kampala were reported by Maxwell (1995). Logistic regression showed that there was statistical difference (P < 0.05) between income groups and gender but not between locations. Households headed by individuals with higher education recorded higher cereal yields than those who had not attended formal school (Figure 5.3).

Based on level of education, independent sample tests revealed a statistically significant difference between household heads with pre-school, no education and higher education (P < 0.05) on total yield and cereals consumed. However, there was no statistically significant difference between primary and secondary levels, secondary and higher levels, or illiterate and pre-school education levels. Also, there was no significant difference between education levels on cereals sold. The list of constraints farmers experienced are presented in Figure 5.4, which shows that female (49.4 percent), low-income households (62.5 percent) and farmers from Blantyre (53.9 percent) mentioned agricultural land as the major constraint.

The average plot size available to male-headed households was 0.24 ha, while females had 0.08 ha, and the high-income households had an average of 0.27 ha as compared to 0.06 ha for low-income households. Lilongwe residents had on average 0.22 ha, while those from Blantyre had 0.17 ha per household.

Figure 5.4 Main agricultural constraints identified by urban farmers (n = 330)

One interesting result was that, at small-farm sizes, both low- and high-income urban farmers obtain similar yields per hectare. However, for larger farms, high-income households are able to capitalize on economies of scale and become more efficient (Figure 5.5). This suggests that, even if land is

Figure 5.5 Average cereal yield (kilograms per hectare) between low-income (n = 120) and high-income (n = 210) households in urban Malawi

Figure 5.6 Average cereal yield (kilograms per hectare) between female-headed (n = 87) and male-headed (n = 243) households in urban Malawi

made available to the land resource poor farmers, the yield will remain low due to other production constraints such as inputs and access to information and technology. Overall, however, yields on UA plots remained low: the government suggests that farmers should obtain 6000–8000 kg/ha (Government of Malawi, 1999), and this reveals that even at its most productive, UA under-performs.

Despite the fact that women on average were poorer and less educated than their male counterparts, when they accessed large plots they were more efficient farmers because they had more time to attend to the crops (Figure 5.6). Thus, there were significant differences (P <0.05) on the cereal yields between male- and female-headed households.

DISCUSSION

Income status and urban food production

The study revealed that UA in Malawian cities will not provide the solution to food insecurity because it provides more food for middle- or upper-class families than for the poor. These results have been reported in other studies (Machethe et al, 1997; Maxwell, 1999). However, UA is playing a crucial role in generating extra income to some of the marginalized groups, especially the low-income, female-headed households and those with no formal education.

Gender and urban agriculture

According to Hovorka (2005), women have limited access to agricultural land and only 4.5 percent participated in UA. When provided with adequate land, however, women were more efficient producers than men, which agrees with Scanlan (2004), who commented that women’s role in agriculture covers all the production stages, which include acquisition, processing and preparation. UA in Malawi is dominated by men and high-income people who are able to invest and undertake UA as a ‘luxury’ livelihood strategy, and are able to access information, private and expensive agricultural consultants/experts and agro-input.

Urban agriculture and education

Education is known to be a major determinant of living standards, and information on education and literacy status is essential for planning and evaluation of existing policies; and statistics for Malawi showed that the majority of poor farmers were illiterate (World Bank, 1986; Government of Malawi, 2005a). Correlation between education and food security reveals that another challenge for the policy makers to surmount before the country can achieve sustainable UA is literacy: the majority of people in Malawi remain illiterate.

Production constraints

Different social groups cited different constraints including access to land, poor product presentation, unstructured markets, lack of government support and shortage of extension services (see also Maxwell, 1995). UA was one of the livelihood strategies for the many low-income households, but street vending and small-medium enterprises are preferred to farming. This observation confirms Ellis’s (2000) comment that diversification into different sources of income has been used as a mechanism by poor households, especially in the informal market, and should be taken into consideration in urban development issues (Mougeot, 2005b). An increased consumption of street food has negative consequences on the nutrition and health status of the people due to a shift in dietary patterns (Ruel et al, 1999; quoted by Kennedy, 2003).

CONCLUSION AND POLICY RECOMMENDATIONS

Urban residents in Malawi engaged in agricultural activities mainly for consumption irrespective of their socio-economic status. However, many households did not consider urban farming as a commercial entity due to lack of support from the government and other institutions. The study has shown that it represents a safety net to marginal female-headed and low-income households. If the development strategy remains as it is in rural Malawi (to provide improved living conditions to the poor and disadvantaged societies), then government policies should encourage collective groups that would practise farming as a business.

ACKNOWLEDGEMENTS

The author wishes to express sincere gratitude to the International Development Research Centre (IDRC) through AGROPOLIS for funding the research work. Thanks also to the Country Director and Staff of Africare, along with research assistants and technical staff at Bunda College of Agriculture, for their contributions and considerable knowledge.

NOTES

1 David Dalison Mkwambisi; Ph.D. Environment and Development, University of Leeds, The School of Earth and Environment, Sustainability Research Institute, Leeds LS2 9JT, UK; tel: +44 113 343 6466; fax: +44 113 343 6716; email: Mkwambisi@env.leeds.ac.uk

2 This is made up of 150 kg maize and 13 kg of rice, sorghum, millet and wheat, and an equivalent of 17-kg-worth of grain from other sources such as sweet potato – crops such as sweet potato are considered to be worth 25 percent of the equivalent amount of grain (FAO/WFP/GoM, 2005).

REFERENCES

Ellis, F. (2000) Rural Livelihoods and Diversity in Developing Countries, Oxford University Press, Oxford, UK

FAO (1996) ‘Rome declaration on world food security and world food summit plan of action’, World Food Summit, 13–17 November 1996, Rome, Italy

FAO (2001) Urban Agriculture and Peri-urban Agriculture: A Briefing Guide for Successful Implementation of Urban Agriculture and Peri-urban Agriculture in and Countries of Transition, FAO, Rome, Italy

FAO (2002) ‘Food security and trade: An overview’, in Trade and Food security: Conceptualising the Linkages Expert Consultation, Rome, 11–12 July 2002, FAO, Rome, Italy, www.fao.org/DOCREP/005/Y4671E/y4671e05.htm (accessed 24 September 2006)

FAO (2003) The state of food insecurity in the world. Monitoring progress towards the World Food Summit and Millennium Development Goals, FAO, Rome, Italy

FAO/WFP/GoM (2005) ‘Crop and food supply assessment mission to Malawi’, Special report by FAO/WFP mission to Malawi, July 2004, Ministry of Agriculture, Lilongwe and FAO, Rome, Italy, www.fao.org/docrep/008/J5509e/J5509e00.htm (accessed 1 September 2006)

Fraser, E. D. G., Hubacek, K. and Korytarova, K. (2006) ‘Food and the environment: (Neo-) Malthusianism arguments and population growth’, in O’Hara, P. (ed) International Encyclopaedia of Public Policy, Governance in a Global Age, Cambridge University Press, London

Garrett, J. L. (2000) ‘Achieving urban food and nutrition security in the developing world: A 2020 vision for food, agriculture and the environment’, Focus 3, Brief 1 of 10; www.ifpri.org/2020/focus/focus03/focus03.pdf (accessed 24 September 2006)

Garrett, J. L. and Ruel, M. T. (1999) ‘Are determinants of rural and urban food security and nutrition status different? Some insights from Mozambique’, World Development, vol 27, no 11, pp1955–1975

Government of Malawi (1998a) Town and Planning Act, Ministry of Local Government, Lilongwe, Malawi

Government of Malawi (1998b) Integrated Household Survey I: Household Socio-Economic Characteristics, National Statistics Office, Zomba, Malawi

Government of Malawi (1999) Guide to Agricultural Production, Agricultural Extension Services Department, Ministry of Agriculture, Lilongwe, Malawi

Government of Malawi (2002) Malawi Poverty Reduction Strategy Paper, Government of Malawi/IMF/World Bank, Lilongwe, Malawi

Government of Malawi (2004) Malawi Growth Strategy, vol III, Ministry of Economic Planning and Development, Lilongwe, Malawi

Government of Malawi (2005a) Integrated Household Survey II (2004–2005), vol 1, Household Socio-Economic Characteristics, National Statistics Office, Zomba, Malawi

Government of Malawi (2005b) Malawi Food Security and Nutrition Policy, Government of Malawi, Ministry of Agriculture, Lilongwe, Malawi

Henn, P. (2002) ‘User benefits of urban agriculture in Havana, Cuba: An application of the contingent valuation method’, City Farmer, Vancouver, Canada, www.cityfarmer.org/havanabenefit.html (accessed 23 September 2006)

Hovorka, A. (2005) ‘Gender, commercial urban agriculture and urban food supply in Greater Gaborone, Botswana’, in Mougeot, L. J. A (ed) AGROPOLIS: The Social, Political and Environmental Dimensions of Urban Agriculture, International Development Research Council, Ontario, Canada, pp137–152

Huddad, L., Ruel, M. and Garrett, J. (1998) Growing Urban Poverty and Undernutrition and some Urban Facts of Life: Implications for Research and Policy. IFPRI, Washington DC, USA

IFPRI (2002) ‘The Accra urban food and nutrition study’, IFFPRI Issue Brief No 9, Addendum; IFPRI, Washington DC, USA

Kennedy, G. (2003) ‘Food security in the context of urban Sub-Saharan Africa’, http://foodafrica.nri.org/urbanisation/urbspapers/GinaKennedyFoodsecurity.pdf (accessed 23 September 2006)

Kwapata, M., Nkanaunena, G. A., Maliro, M. F. A., Kamwendo, P. M., Dawa, G. and Bokosi, J. (2001) The Current Status and Contribution of Urban Agriculture and Identification of Information, Communication and Training Needs in Malawi, Bunda College, Lilongwe, Malawi

Machethe, C. L., Reardon, T. and Mead, D. C. (1997) ‘Promoting farm/non-farm linkages for employment of the poor in South Africa: A research agenda focused on small-scale farms and agro industry’, Development Southern Africa, vol 14, no 3, pp377–394

Maxwell, D. G. (1995) ‘Alternative food security strategy: A household analysis of urban agriculture in Kampala’, World Development, vol 23, no 10, pp1669–1681

Maxwell, D. G. (1999) ‘The political economy of urban food security in Sub-Saharan Africa’, World Development, vol 27, no 11, pp1939–1953

Maxwell, D. G. (2001) The importance of urban agriculture to food and nutrition. Annotated biography, ETC-RUAF, CTA publishers, Leusden, Netherlands

Mkwambisi, D. D. (2005) ‘Impact of urban agriculture on child labour, health and education in Lilongwe city, Malawi’, in Zeleza Mtafu Manda (ed) Malawi Physical Planning, Alma Publishers, Mzuzu, Malawi, pp 94–111

Mougeot, L. J. A. (2000) ‘Achieving urban food and nutrition security in developing countries: The hidden significance of urban agriculture’, IFPRI, Brief paper number 6, 2000 www.ifpri.org/2020/focus/focus03/focus03.pdf (accessed 23 September 2006)

Mougeot, L. J. A. (2001) ‘Urban agriculture: Definition, presence, potential and risks’, in Barker, N., Dubbelling, M., Gindel, S., Sabel-Koschella, U. and de Zeeuw, H. (eds) Growing Cities, Growing Food – Urban Agriculture on the Policy Agenda, DSE, Eurasburg, Germany, pp1–42

Mougeot, L. J. A. (2005a) Introduction, in Mougeot, L. J. A. (ed) AGROPOLIS: The Social, Political and Environmental Dimensions of Urban Agriculture, International Development Research Council, Ontario, Canada, pp1–29

Mougeot, L. J. A. (2005b) ‘Neglected issues on form and substance of research on urban agriculture’, in Mougeot, L. J. A. (ed) AGROPOLIS: The Social, Political and Environmental Dimensions of Urban Agriculture. International Development Research Council, Ontario, Canada, pp267–279

Neefjes, K. (2000) Environments and Livelihoods: Strategies for Sustainability, Oxfam, London

Nugent, R. (2000) ‘The impact of urban agriculture on the household and local economies’, in Barker, N., Dubbelling, M., Gindel, S., Sabel-Koschella, U. and de Zeeuw, H. (eds) Growing Cities, Growing Food-Urban Agriculture on the Policy Agenda, DSE, Eurasburg, Germany, pp67–97

Pothukuchi, K. and Kaufman, J. L. (1999) ‘Placing the food system on the urban agenda: The role of municipal institutions’, Agriculture and Human Values, vol 16, no 2, pp213–224

Sabates, R., Gould, B. W. and Villarreal, H. J. (2001) ‘Household composition and food expenditures: a cross-country comparison’, Food Policy, vol 26, pp571–586

Sanyal, B. (1985) ‘Urban agriculture: Who cultivates and why?’, Food and Nutrition Bulletin, vol 7, no 3, pp15–25

Scanlan, S. J. (2004) ‘Women, food security and development in less-industrialized societies: Contributions and challenges for the new century’, World Development, vol 32, no 11, pp1807–1829

Scoones, I. (1998) ‘Sustainable rural livelihoods – A framework for analysis’, University of Sussex, Institute for Development Studies, Brighton, UK

Sen, A. K. (1991) Poverty and Famines: An Essay on Entitlements and Deprivation, Oxford University Press, Oxford, UK

Sen, A. K. (2000) Development as Freedom, 1st edition, Anchor Books, New York, NY, USA Smit, J. (1996) Urban agriculture – Food Jobs and Sustainable Cities, United Nations Development Programme, New York, USA

Solesbury, W. (2003) ‘Sustainable livelihoods: A case study of the evolution of DFID Policy’, DFID Working Paper No 217; www.odi.org.uk/publications/working_papers/wp217.pdf (accessed 1 October 2006)

UN-HABITAT (2006) The State of the World’s Cities, 2001, United Nations Center for Human Settlements, Nairobi, Kenya

USAID (2005) Facts about Malawi, USAID, Lilongwe, Malawi

World Bank (1986) Poverty and Hunger: Issues and Options for Food Security in Developing Countries, World Bank Policy Study, World Bank, Washington DC, USA

This page intentionally left blank

6
An Analysis of the Quality of Wastewater Used to Irrigate Vegetables in Accra, Kumasi and Tamale, Ghana

Philip Amoah1

INTRODUCTION

Cities in developing countries, including Ghana, are experiencing unparalleled population growth. Rapidly increasing water supply and sanitation coverage generates large volumes of wastewater, which is often released untreated into the environment (streams, drains, etc.). In Ghana, most urban centres have no means of treating wastewater and the sewerage network serves a small percentage (4.5 percent) of the total population (GSS, 2002). The need for year-round production of vegetables in or near urban areas makes irrigation necessary; hence, farmers in search of water for irrigation often rely on wastewater. However, despite its importance for countless livelihoods, the use of urban wastewater for irrigation is not without health risks. Wastewater contains the full spectrum of pathogens found in the urban population, many of which can survive for several weeks when discharged onto fields.

Effective wastewater treatment can reduce pathogen levels, but in most developing countries it is not an option for the municipal authorities due to the high costs involved (Keraita et al, 2002). Most new sewerage treatment plants in Ghana are also operating below their design capacity. As wastewater treatment does not appear to be a realistic option, banning the use of polluted water by urban farmers has also been tried in Accra and other cities. This has failed because such bans threaten the livelihoods of many individuals, which run contrary to poverty-alleviation strategies. In these circumstances, urban farmers express significant concerns because their livelihoods are at permanent risk. Any solution to reduce health risks without forcing them to change their (market-driven) cropping patterns or access to water would be appreciated. In addition, Ghana’s Tourism Board has started a campaign directed at consumers to promote ‘safer vegetables for healthier cities’. This was prompted because tourists were suffering from outbreaks of gastrointestinal disorders after consuming vegetables in urban areas.

Another potential health risk derives from the use of pesticides, although this is beneficial in decreasing crop loss both before and after harvest (Clarke et al, 1997). Despite the recognition of urban agriculture (UA) as a source of urban food security, concerns are growing among city authorities on the indiscriminate use of pesticides. Insufficient data exist, however, on the actual gravity of the problem. These would provide guidance on appropriate interventions or policy formulation.

Hence research into other risk-reduction options is required for regions or countries where wastewater treatment is not a realistic option. Therefore, this study was designed to explore the contamination and decontamination of wastewater-irrigated crops. The structure of this chapter first outlines the concepts, objectives and hypotheses, then describes the research methods followed by results, a discussion of impacts and finally some recommendations.

KEY CONCEPTS

Wastewater

Wastewater is the liquid portion of waste. It may be defined as a combination of liquid or water-carried wastes that are removed from residences and institutions, as well as commercial and industrial establishments. In addition, a combination of groundwater, surface water and storm water may be present (Metcalf and Eddy, 1995). In this section, it is assumed that urban wastewater may be a combination of some or all of the following:

• Municipal wastewater consists of domestic effluent made up of black water (excreta, urine and associated sludge) and grey water (kitchen and bathroom wastewater). It may also include water from commercial establishments and institutions, including hospitals.

• Industrial effluent is water polluted by industrial processes and containing high levels of heavy metals or other chemical or organic constituents. Industrial effluent does not normally contain high levels of microbiological pollution unless it emanates from slaughterhouses or food-processing plants.

• Storm water is run-off precipitation that finds its way across surfaces into receiving waters. Urban storm run-off is collected and transported in storm or combined sewers. The composition of storm water reflects the composition of precipitation and the surfaces with which it is in contact (Environment Canada, 2006).

Other terms

• Marginal-quality water: water whose quality might pose a threat to sustainable agriculture and/or human health, but which can be used safely for irrigation provided certain precautions are taken (Abbot and Hasnip, 1997). Such water is polluted as a consequence of mixing with wastewater or agricultural drainage (Cornish et al, 1999).

• Indirect use of wastewater: This is the unplanned application to land of wastewater from a receiving water body. Municipal and industrial wastewater is discharged without treatment or monitoring into the watercourses draining an urban area. Irrigation water is drawn from rivers and streams or other natural water bodies that receive wastewater flows. There is often no control over the use of water for irrigation or domestic consumption downstream of the urban centre. Consequently, many farmers indirectly use marginal-quality water of unknown composition that they draw from many points downstream of the urban centre.

Research objectives

This chapter has four primary objectives:

1 To assess the water quality (biological and chemical) of irrigation water sources used for vegetable cultivation.

2 To trace the (microbiological and helminth) contamination pathways of vegetables in urban and peri-urban sites to identify where interventions should take place along the production–consumption continuum.

3 To isolate and identify fecal coliform (FC) bacteria found on irrigated vegetables from urban and peri-urban sites.

4 To determine the level of pathogen and pesticide contamination of vegetables produced on urban agricultural sites.

Hypotheses

Three basic hypotheses underlie this research:

1 Fecal coliform (FC) and helminth egg population levels in wastewater from different urban sources exceed common standards recommended for irrigation.

2 Microbiological (FC and helminth levels) contamination of wastewater-irrigated vegetables is increased through handling and distribution within the production marketing chain.

3 Potential health risks to consumers are not reduced to acceptable levels after the normal household treatment of vegetables.

METHODS

Phases of research and cities studied

The study was divided into three phases. The first phase (market sampling) was conducted in the three Ghanaian cities of Accra, Kumasi and Tamale (Figure 6.1). The second phase (water to field to market sampling) took place

Figure 6.1 Agro-ecological map of Ghana showing the three urban study areas of Tamale, Kumasi and Accra, with details for Kumasi and Accra

in Accra and Kumasi. The third phase (field trials) was only conducted in Accra.

Accra is the capital city of Ghana, with a population of approximately 1.7 million (GSS, 2002). It is located at the Gulf of Guinea in the coastal savannah belt. Two sites were chosen. Dzorwulu is one of Accra’s suburbs, with a major vegetable production site within the metropolis. It has a total land area of roughly 12 ha, which is cultivated by more than 300 farmers. Marine Drive is a smaller vegetable production site in Accra near Independence Square. Its area is close to 4 ha and has more than 100 vegetable growers.

Kumasi is the capital town of the Ashanti Region and the second largest city in Ghana, with a population which is a little more than one million (GSS, 2002). It is located in the forest belt of Ghana. Two sites were chosen here, also. The Gyenyase site is the largest urban vegetable growing site in Kumasi, with a total land area of approximately 6 ha which supports approximately 60 vegetable farmers. D Line, a suburb close to the Kwame Nkrumah University of Science and Technology (KNUST), is a popular vegetable farming spot in Kumasi, with close to 40 farmers and a total area of approximately 3 ha.

Tamale is the administrative and regional capital of Ghana’s northern region, located in the savannah zone with a population of approximately 300,000 (GSS, 2002). In contrast to Accra and Kumasi, Tamale Municipality has few bodies of water and only a few seasonal streams.

Phase I: Sampling of vegetables at the markets

The purpose of this phase was to determine pathogens and pesticide contamination of vegetables produced at UA sites. From October to December 2002, a total of 180 vegetable samples (lettuce, cabbage and spring onion) were collected from nine major markets and 12 specialized, individual vegetable and fruit sellers (i.e. sellers with permanent stalls outside of designated markets) in Accra, Kumasi and Tamale (see Figure 6.1). At each market, samples were collected under normal purchase conditions from three randomly selected sellers. A minimum of three composite samples – each containing two whole lettuce heads, three bunches of spring onions (each containing two bulbs) and three cabbages – were collected from the upper, middle and lower shelves of each seller, put in sterile polythene bags and transported on ice to the laboratory, where they were analysed immediately or stored at 4°C until analysis could occur within 24 hours.

These and all other samples collected (see below) were analysed for coliform and helminth egg populations using the most probable number (MPN) method (APHA/AWWA/WEF, 2001) and the floatation and sedimentation method following a modified US Environmental Protection Agency (US–EPA) method (Schwartzbrod, 1998), respectively. Gas Chromatography (Flame Ionization Detector; Hewlett-Packard 5890 series II) was used for pesticide residues on lettuce following the method adopted by Ntow (2001). Sample peaks were identified by their retention times compared with those of the corresponding pesticide standard obtained from the International Atomic Energy Agency. The ability of the laboratory to identify these substances has been verified by cross-tests of river sediments in Ghana.

Phase II: Contamination pathway study

This study was undertaken to determine the microbiological contamination levels at various entry points along the production-marketing chain. In Accra and Kumasi, two major irrigated vegetable production sites were selected based upon the source of irrigation water and the type of vegetables grown, with emphasis on exotic vegetables such as lettuce that were probably going to be consumed raw. Both sites in Accra used water from drains and streams, while shallow wells and streams provided the sources in Kumasi. Farmers in at least one of the two sites in each city used irrigated, piped water as their source over a period of at least three years. All sites had a similar history of land use. For instance, all were under vegetable cultivation for periods of not less than five years, and all farmers used poultry manure as a source of fertilizer.

Irrigation water

This study monitored the microbiological quality of irrigation water from different urban sources. One composite sample per week was collected from each source for 52 weeks from May 2003 to April 2004. In all, six were involved – stream, shallow well and piped irrigation water sources in Kumasi and drain, stream and piped water in Accra. Sampling at all sites was carried out between eight and ten in the morning in keeping with farmer’s irrigation practices (APHA/AWWA/WEF, 2001). At each site, 200-ml glass bottles were used to take water from three different points in the wells or in 20-m intervals along the drain or stream. Piped water was collected directly from the water hose used by the farmers for irrigation. Samples from a particular site were later joined into one composite sample per source and transported to the laboratory on ice. A total of 312 composite water samples were analysed for total and fecal coliform populations. Sampling for helminth egg quantification in irrigation water was done twice every month for five months from November 2003 to March 2004 at all the selected sites. Two-litre samples were taken after deliberately disturbing the bed of the irrigation water source to stimulate agitation that might occur when farmers are filling their watering cans. This was intended to bring out the eggs, as they usually settle under their own weight (Cornish et al, 1999).

Lettuce sampling in farms, wholesale and retail markets

Over a period of 12 months, from May 2003 to April 2004, a total of 1296 lettuce samples were collected at different entry points2 from farm to the final retail outlet. The original sets of lettuce were either irrigated with stream, drain, well or piped water (microbiological quality of these water sources were monitored as described above). Twice a month, a minimum of three composite samples (each containing two whole lettuces) from each of the selected farm sites were randomly collected using sterile disposable gloves just before harvesting for sale at the market. These were put into separate sterile polythene bags and labelled as farm samples. The seller was followed to the wholesale market where another sample from the same original stock was collected, before being finally sold to a retailer. At the final retail point, three composite samples were again sampled after vegetables were displayed on the shelves for at least two to three hours, which is a typical turn-over period at the retail point. Producers and sellers were paid for their produce. Sampled vegetables were transported on ice to the laboratory where they were analysed for TC, FC and helminth counts. To eliminate potential biases during analysis, staff working in the laboratories were blind to the source of the samples.

Phase III: Field trials

In order to further understand the importance of the different possible sources of contamination of irrigated lettuce (e.g. wastewater, poultry manure, soil), a field trial was set up at two farm sites in Accra. Lettuces were grown on raised beds of the common size of approximately 3 m2. For each trial two adjacent plots, A and B, were used, with each plot subdivided into four blocks. Each block contained four beds, making a total of 16 beds per plot. Using a randomized complete block design, each of the four beds in a block was randomly allocated to one of four treatments (three representative poultry manure samples, PM1, PM2 and PM3, with average FC populations of 4.3 × 107, 2.4 × 105 and 3.3 × 103 100 g–1, respectively, and one inorganic fertilizer NPK (F)), which were applied to the beds two weeks after transplanting following farmers’ usual practices. Each of the two plots (A and B) at a site was irrigated with either clean piped water or wastewater from a drain. The crops on a bed were irrigated once daily with approximately 30 l of water every morning excluding rainy days. All work was done by the farmers as part of their standard routine.

Soil samples from the top 10 cm were collected from six randomly selected points on each bed, just before transplanting from the nursery, and combined into one composite sample per bed for the initial count of FC and helminth egg populations in duplicates. Samples were also taken from nearby plots with no previous history of vegetable production to serve as controls. Six lettuce heads were aseptically collected from each bed four weeks after transplanting, after which they were randomly combined into three composite samples containing two lettuce heads. The samples were kept in a refrigerator at 4°C and analysed within 24 hours for FC and helminth eggs using standard methods.

Data analysis

The data were analysed using SPSS for Windows 10. Using MPN, FC populations were normalized by log transformation for the analysis of variance (ANOVA). t-tests, used to compare data sets, were also employed where necessary. Inferences for significant differences were set at P < 0.05.

RESULTS AND DISCUSSION

Microbiological quality of market vegetables from urban markets

Bacterial contamination of market vegetables

Most of the vegetable samples showed high FC contamination levels (see Table 6.1). The highest level of FC contamination was recorded in lettuce (geometric mean count of 1.1 × 107 per g wet weight), probably due to the

Table 6.1 Ranges of total and fecal coliform population on selected vegetables

Note: MPN, most probable number.

larger surface area exposed. Cabbage and spring onion showed geometric mean counts of 3.3 × 106 g–1 and 1.1 × 106 g–1 wet weight, respectively. No sample had less than 4000 FC per gram wet weight.

The mean FC levels of all the three crops exceed the International Commission on Microbiological Specifications for Food (ICSMF, 1974) recommended level of 103 fecal coliforms per gram fresh weight. Several factors may account for the high levels recorded in most of the analysed vegetables. Among these is the use of polluted irrigation water and fresh poultry manure, both of which are applied on top of the crops. Significantly high FC contamination levels (between 4.8 × 103 and 2.8 106 100 ml–1) which usually exceed common standards have been recorded in irrigation water (Cornish et al, 1999; Drechsel et al, 2000; Mensah et al, 2001; Keraita et al, 2002). High FC populations (between 3.6 × 104 and 1.1 × 107) were also reported in poultry manure in the same study areas.

Another potential source of contamination is market-related handling, especially where provision for better sanitary standards (e.g. clean water for crop washing and refreshing) is lacking. A relatively high total and FC population recorded on some vegetables was also reported by Johnson (2002) and Armar-Klemesu et al (1998) analysing street food and market crops in Accra, respectively.

Mean helminth egg population on vegetables

About 30 percent of vegetables had no helminth eggs. Lettuce, cabbage and spring onion carried mean helminth egg populations of 1.1 g–1, 0.4 g–1and 2.7 g–1 wet weight, respectively. No significant difference was observed in the mean helminth egg populations recorded in lettuce and cabbage; however, the difference between spring onion and both lettuce and cabbage was significant (P < 0.05). If distributions skewed, the mean separations were compared using one-way ANOVA. The eggs identified included Ascaris lumbricoides, Ancylostoma duodenale, Schistosoma heamatobium and Trichuris trichiura, with Ascaris lumbricoides eggs being the predominant contaminant (60 percent of lettuce, 55 percent of cabbage and 65 percent of spring onions showed Ascaris lumbricoides eggs).

Biologically, the highest health risk from pathogens is infections due to helminth contamination. Because helminths persist for longer periods than pathogens in the environment, host immunity is usually low to non-existent and the infective dose is small (Gaspard et al, 1997). Such microbial and parasitic contamination probably contributes to the high number of food-borne and water-related diseases in Accra such as diarrhoea (sometimes caused by typhoid or cholera), as well as intestinal worm infections. However, these also have to be seen in the context of generally sub-optimal sanitary conditions in parts of the metropolis (Arde-Acquah, 2002).

Pesticide residues on lettuce leaves

Table 6.2 shows pesticide detection prevalence and residues recorded on lettuce leaves, with maximum residue limits (MRL) as comparators. Only

Table 6.2 Pesticide residue detection and concentrations on lettuce (n = 60)

14 percent of samples had no detectable pesticide residues. More than 60 percent of the lettuce samples had two or more pesticide residues, with 78 percent of samples showing chlorpyrifos, an organophosphate of moderately acute hazard (WHO, 2005). Chlorpyrifos was the only pesticide with higher levels in one city, Kumasi. In most cases, pesticide residue levels observed exceeded the MRL.

The results of tests for pesticide residue indicate that several pesticides (particularly chlorpyrifos) are widely used by vegetable producers in Ghana, in keeping with other studies (Okorley and Kwarteng, 2002). As also described by Danso et al (2002), farmers mix cocktails of various pesticides to increase their potency. Vegetables are often eaten raw so it is not surprising to read about evidence of chlorpyrifos contamination such as can be found in waakye, a popular Ghanaian dish (Johnson, 2002). Lindane and endosulphan are restricted for the control of capsids on cacao trees, stem borers in maize and for pests on coffee plants, while DDT is banned in Ghana. However, the data show clearly that these potent agrochemicals are used irrespective of whether approved for vegetable production or not. It has been reported that in several African countries the legislation on importation and regulation of pesticides is sketchy, nonexistent or imbedded in bodies of legislation indirectly related to pesticides (Tallaki, 2005). Because of the lack of proper regulations, organochloride pesticides banned in industrialized countries for their retention in the environment or their high toxicity are still commonly used.

This widespread pesticide contamination, often exceeding the MRL, indicates potential health risks to consumers. Washing vegetables before consumption is highly recommended, but the majority of pesticides cannot just be washed away and may still pose health risks (Volpi, 2005). A rough calculation helps to elucidate this potential: The acceptable daily intake (ADI)3 of chlorpyrifos, for example, is 0.01 mg kg–1 body weight (WHO, 1997). To exceed the ADI, a child weighing 30 kg would have to consume at least 0.3 mg of chlorpyrifos per day. With a residue level of 1.6 mg

Figure 6.2 Fecal coliform contamination levels of irrigation water used during the study period (piped water was excluded because no fecal coliforms were detected during the study period)

chlorpyrifos kg–1 lettuce, the child would have to eat close to 200 g lettuce per day. The amount of lettuce (usually served with other staples, e.g. rice) is usually less than 30 g daily. However, if a child was malnourished they might be more susceptible; moreover, fetal (via maternal) or chronic neurodevelopmental effects might occur, since they are not always included in MRL analysis.

Contamination pathway study

FC levels of irrigation water used

With the exception of piped water, FC contamination levels of irrigation water from different sources used in the study significantly exceeded a geometric mean of 1 × 103 100 ml–1 (WHO recommended level for unrestricted irrigation4). For details, see Figure 6.2. Significantly higher FC counts were recorded in water samples from streams than from shallow wells in Kumasi.

The maximum FC contamination densities for shallow well and stream samples in Kumasi were 4 × 106 100 ml1 and 4 × 108 100 ml–1, respectively, while drain and stream samples in Accra showed densities of 9 × 106 100 ml–1 and 2 107 100 ml–1, respectively. This corresponds well with data reported by several researchers (Cornish et al, 1999; Mensah et al, 2001; Keraita et al, 2002) in the same study area.

Helminth egg contamination levels of irrigation water used

No eggs were recorded in piped water during the study period. Arithmetic mean5 helminth egg contamination levels in irrigation water from drains,

Table 6.3 Mean numbers of helminth eggs in irrigation water from different sources

streams and wells in Accra and Kumasi also exceeded the WHO recommended level of one or less eggs per litre for unrestricted irrigation (Table 6.3). The main species of helminth eggs isolated in water and on lettuce included those of Ascaris lumbricoides (most predominant species in all samples), Hymenolepis diminuta, Fasciola hepatica, T. trichura and Strogyloides larvae. The results from Kumasi showed a higher helminth population in stream water than in shallow wells, probably due to run-off.

The results of the microbiological quality of irrigation water confirm earlier reports (Cornish et al, 1999; Mensah et al, 2001; Keraita et al, 2002) that low-quality water is being used for urban vegetable production in most Ghanaian cities. Shallow wells or dugouts might be expected to meet the WHO recommended standard due to the natural filtering of aquifer materials and long underground retention times. However, those used in this study were often not protected against surface inflow.

The significant differences recorded in FC levels between the two sources in Kumasi may suggest that shallow well water may pose relatively less risk to farmers and consumers, although the coliform levels still exceed 1000 counts per 100 ml. Similar results have been reported from Kenya (Hide et al, 2001). However, the situation can also change. For example, Cornish et al (1999) recorded in Kumasi temporarily higher FC population in shallow wells than in nearby streams. This may be due to the fact that probably the wells used were shallower and got more easily contaminated through surface run-off on the field (Drechsel et al, 2000).

Microbiological quality of lettuce at different entry points along the production consumption pathway

Irrespective of the irrigation water source, mean FC levels on produce exceeded recommended standards. Figure 6.3 shows FC populations on lettuce samples collected in Kumasi at the farm gate, wholesale market and retail outlets over a 12-month period and for three irrigation water sources. There were no significant differences in average lettuce contamination levels at different entry points (farm, wholesale market and retail outlet). Similar contamination levels were recorded for the three irrigation water sources in Accra (Amoah et al, 2007). High levels of FC counts (usually above the

Figure 6.3 Fecal coliform counts at different entry points on the production–consumption pathway for irrigated lettuce using water from (A) wells, (B) streams and (C) piped water in Kumasi

common acceptable standard of 1 × 103 100 g–1 wet weight) were recorded on all irrigated lettuce including those irrigated with piped water.

Apart from lettuce from Accra that was stream-water irrigated, higher levels of FC contamination were recorded in the rainy season than in the dry season. The difference was significant (P < 0.05) only in the cases of well- and stream-water irrigated vegetables from Kumasi (Amoah et al, 2005). The results also showed that in 80–90 percent of the weeks’ produce sampled in Accra and Kumasi, there was no significant difference in the FC counts of samples analysed from the farm gate to the markets and final retail points.

Helminth eggs

Helminth eggs, including those of Ascaris lumbricoides, Hymenolepis diminuta, T. trichuris, F. hepatica and Strongyloides larvae, were counted on lettuce samples at the different entry points. The helminth egg population ranged from one to six eggs 100 g–1 wet weight. In the majority of cases, significantly higher levels (P < 0.05) were recorded in lettuce that was irrigated with polluted water than those from piped water irrigated sources. However, mean helminth egg population on lettuce from the same original stock and irrigation water source did not show any significant difference from field to market (see Table 6.4).

The study revealed that the contamination of lettuce with pathogenic micro-organisms does not significantly increase through post-harvest handling and marketing, in contrast to earlier work (Armar-Klemesu et al, 1998). These results were unexpected because of the alarmingly unhygienic conditions on market sites, which are related to washing habits, display and handling of food and the availability of sanitation infrastructure. Only 31 percent of the

Table 6.4 Helminth egg contamination at different entry points along the production–consumption pathway

* Mean numbers represent the mean of all the different types of eggs, as well as Strongyloides larvae (n = 15 for each irrigation water source). Eggs were rounded to the nearest whole number.

** The same letters indicate no significant difference in column numbers between water sources for each city (P > 0.05). For differences between farm, wholesale market and retail, see Figure 6.3.

markets in Accra have a drainage system and only 26 percent have toilet facilities, while 34 percent are connected to pipe-borne water (Nyanteng, 1998). The results suggest that the initial on-farm contamination is so high that it overshadows any possible post-harvest contamination.

Significantly higher levels of FCs in well- and stream-water irrigated lettuce were recorded in Kumasi during the rainy season than in the dry season. A similar but non-significant trend was observed in Accra. In fact, it might be expected that in the rainy season, when contamination of the irrigation water is more diluted and farmers irrigate less, FC population on lettuce at the farm gate would be lower than in the dry season. Another observation was that crop contamination also takes place under irrigation with piped water. One reason could be that poultry manure does not dry sufficiently; another reason is that splashes from the soil during heavy rains may contain pathogens from already contaminated soil (Bastos, 1992).

Phase IV: Field trial

The results of the field trial (Amoah et al, 2005) confirmed that, even at the farm level, wastewater is only one of several sources of crop contamination. The soil and poultry manure were identified as other sources of microbiological contamination, although wastewater is the most significant one. The need to reduce the potential health risks resulting from FC and helminth contamination of urban and peri-urban vegetables thus requires a more holistic approach rather than concentrating solely on wastewater.

Health risk reduction

The study indicates it is crucial to reduce health risks associated with the consumption of contaminated lettuce at the farm level through good agricultural practices. This, however, is not as easy as is often suggested. Farmers might not be able to make changes to irrigation methods, timing and crops (Drechsel et al, 2002). Therefore, other options are under investigation under the Challenge Program on Water for Food (CPWF), a major international effort being conducted by the Consultative Group on International Agricultural Research (CGIAR). Although first trials by IWMI show that the on-farm contamination levels can be reduced through minor changes in practices, it is unlikely that contamination can be minimized below the threshold for safe consumption, as data from the use of piped water show. Thus, it will always be necessary to wash the crops, in addition to improving farming practices.

The last stage in the production–consumption chain, where food for consumption or fast-food for street sales is prepared, appears to be a good entry point for improving hygiene. This is because awareness for food safety is generally high at this point. Also, more than 90 percent of the food vendors and consumers wash salad. However, individual methods vary widely and seldom meet the required standards (Amoah, in press). Moreover, consumers reduce health risks by trusting only food vendors with neat appearance and visually clean food (Olsen, 2006), which is a first step but not sufficient to avoid contaminated food completely (Mensah et al, 2002).

Impacts or outcomes associated with the work

The study has provided relevant information and advice to city authorities such as the Ministry of Food and Agriculture (MoFA) and the Metropolitan Assembly on wastewater use in urban and peri-urban vegetable production in Ghana. Authorities are aware of the need to have a more holistic approach in addressing issues related to the use of wastewater in peri-urban agriculture (UPA) rather than application of ‘hard’ solutions such as banning the use of wastewater, which are often impractical and difficult to enforce. In view of this, the authorities at the MoFA and the Metropolitan Assembly have requested assistance in the formulation of more appropriate policies on UA in the future.

Researchers have seen the need to do more research on non-treatment options for health risk reduction (e.g. CPWF and IWMI on such concepts as farm risk-reduction techniques, vegetable washing techniques). The results have contributed to the formulation of the new WHO guidelines on wastewater use in agriculture which include several non-treatment options for health risk reduction (WHO, 2006).

Leading farmers and sellers involved in the activities of the study have become aware of the risks involved in the use of wastewater and are taking

Figure 6.4 Farmer irrigating with lowered watering can

Figure 6.5 Seller displaying vegetables in the market

the necessary precautions. For example, these farmers have reduced contact with wastewater during fetching it by not stepping into the water. To further decrease contamination from the soil, watering cans are lowered during irrigation (see Figure 6.4) to reduce contamination from splash. Also, only mature compost is applied. The sellers have stopped displaying their produce on the floor, now choosing raised shelves and tables (see Figure 6.5). These practices are expected to be introduced to other farmers and sellers at the same sites as well as throughout the cities.

CONCLUSIONS AND RECOMMENDATIONS

The research demonstrates that typical microbiological and pesticide contamination levels of vegetables in Ghanaian markets pose a threat to human health. It has also shown that the majority of microbial and parasitological contamination of lettuce produced from urban sources in Accra and Kumasi comes from wastewater, application of manure and residual soil contamination. Results also show that the post-harvest sector is probably a relatively minor contributor to lettuce contamination.

Although there is enough general epidemiological evidence in support of disease transmission through the consumption of contaminated vegetables that are irrigated with wastewater (Shuval et al, 1986), it is difficult for farmers and municipalities to comply with the WHO wastewater irrigation guidelines (WHO, 2006) for various reasons (Drechsel et al, 2002). Simply banning the use of wastewater for unrestricted irrigation would deprive many farmers and sellers of their livelihood and drastically reduce the amount of many perishable vegetables in Ghana’s cities (Cofie and Drechsel, 2007). It also would not solve the problem of post-harvest contamination. Therefore, more efforts are needed to test possible options for risk reduction at farm, market, street restaurant and household level (Drechsel et al, 2002). For example, market sellers could improve their vegetable washing to better meet the demands of their customers (Drechsel et al, 2000). Further requirements are that the sanitary facilities in markets provide permanent access to clean running water and that support through education and awareness campaigns takes place. Risk reduction should also include sensitization of consumers (especially parents) on potential health implications of unhygienic handling and consumption of contaminated vegetables.

In conclusion, adoption of the ‘multiple barrier approach’ (WHO, 2006), where health risk intervention methods (such as good irrigation practices and vegetable washing before food preparation at the household) are used, appears to be the most appropriate overall risk-reduction strategy.

ACKNOWLEDGEMENTS

This study was made possible by financial support from IDRC’s International Graduate Research Awards in Urban Agriculture (AGROPOLIS), the Challenge Program on Water and Food (CPWF) projects CP38 and CP51, and a capacity building grant of the International Water Management Institute (IWMI). We are grateful to Mr Osei Tutu from the Biological Science Department of the Kwame Nkrumah University of Science and Technology, KNUST, Ghana, and Richmond Kofi Yawson, a laboratory assistant at IWMI lab, for their support in data collection and laboratory analysis.

NOTES

1 Philip Amoah, Ph.D. Biological Sciences, IWMI West Africa Office, PMB CT 112 Cantonments, Accra, Ghana; tel: +233-21-7847452 or +233-20-8154651; email: p.amoah@cgiar.org

2 Products pass through three main sampling stages from the field where they are harvested to the retail outlet where consumers buy them: the farm, where samples of crops are collected; the wholesale market, where samples are taken from crops wholesalers will purchase; and the retailer, where samples were taken two to three hours after vegetables had been displayed and where some had been refreshed.

3 The acceptable daily intake (ADI) is a measure of the quantity of a particular chemical in food which, it is believed, can be consumed on a daily basis over a lifetime without harm.

4 World Health Organization’s (WHO, 2006) recommended level for uses that include crops likely to be eaten raw.

5 Arithmetic mean, which included Strongyloides larvae.

REFERENCES

Abbott, C. and Hasnip, N. (1997) The Safe Use of Marginal Quality Water in Agriculture – A Guide for the Water Resource Planner, HR Wallingford and DFID, UK, p59

American Public Health Association, American Water Works Association, Water Environment Federation (APHA, AWWA, WEF) (2001) Standard Methods for the Examination of Water and Wastewater, 22nd edition, Washington DC, USA

Amoah, P. (in press) ‘Report on a study to test and improve the efficacy of common home-washing methods for vegetables’

Amoah, P., Drechsel, P. and Abaidoo, R. C. (2005) ‘Irrigated urban vegetable production in Ghana: sources of pathogen contamination and health risk-reduction’, Irrigation and Drainage, vol 54, pp49–61

Amoah, P., Drechsel, P., Abaidoo, R. C. and Henseler, M. (2007) ‘Irrigated urban vegetable production in Ghana: Microbiological contamination in farms and markets and consumer risk group’, Journal of Water and Health, vol 5, no 3, pp455–466

Amoah, P., Drechsel, P., Abaidoo, R. C. and Klutse, A. (in press) ‘Effectiveness of common and improved sanitary washing methods in selected cities of West Africa for the reduction of coliform bacteria and helminth eggs on vegetables’, Tropical Medicine and International Health,

Arde-Acquah, A. A. (2002) ‘Sanitation and health-related problems of street-vended foods in Accra’, in SADAOC Street Food in Ghana: Types, Environment, Patronage, Laws and Regulations, Proceedings of a roundtable conference on 6 September 2001, ISSER, University of Ghana/WILCO, Accra, pp57–60

Armar-Klemesu, M., Akpedonu, P., Egbi, G. and Maxwell, D. (1998) ‘Food contamination in urban agriculture: vegetable production using wastewater’, in Armar-Klemesu, M. and Maxwell, D. (eds), Urban Agriculture in the Greater Accra Metropolitan Area, Final Report to IDRC (project 003149), University of Ghana, Noguchi Memorial Institute, Accra, Ghana

Bastos, R. K. (1992) ‘Bacteriological aspects of drip and furrow irrigation with treated wastewater’, Thesis report submitted in accordance with the requirements for Ph.D, Department of Civil Engineering, University of Leeds, UK, December 1992

Clarke, E. E., Levy, L. S., Spurgeon, A. and Calvert, I. A. (1997) ‘The problem associated with pesticide use by irrigation workers in Ghana’, Occupational Medicine, vol 47, pp301–308

Cofie, O. O. and Drechsel, P. (2007) ‘Water for food in the cities: The growing paradigm of irrigated (peri)-urban agriculture and its struggle in sub-Saharan Africa’, African Water Journal, vol 1, no 2, pp23–32

Cornish, G. A., Mensah, E. and Ghesquire, P. (1999) ‘An assessment of surface water quality for irrigation and its implication for human health in the peri-urban zone of Kumasi, Ghana’, Report OD/TN 95, September 1999, HR Wallingford, UK

Danso, G., Drechsel, P. and Fialor, S. C. (2002) ‘Perception of organic agriculture by urban vegetable farmers and consumers in Ghana’, Urban Agriculture Magazine, vol 6, pp23–24

Drechsel, P., Abaidoo, R. C., Amoah, P. and Cofie, O. O. (2000) ‘Increasing use of poultry manure in and around Kumasi, Ghana: Is farmers’ race consumers’ fate?’, Urban Agricultural Magazine 2, pp25–27

Drechsel, P., Blumenthal, U. J. and Keraita, B. (2002) ‘Balancing health and livelihoods: Adjusting wastewater irrigation guidelines for resource-poor countries’, Urban Agriculture Magazine, vol 8, pp7–9

Environment Canada (2006) ‘Municipal wastewater sources and characteristics’, www.atl.ec.gc.ca/epb/issues/wstewtr.html (accessed June 2006)

Gaspard, P., Ambolet, Y. and Schwartzbrod, J. (1997) ‘Valorisation des boues de stations d’épuration en vue de l’amélioration des sols destines a l’agriculture: Contamination parasitaire et modélisation en vue de la gestion du risqué sanitaire’, Bulletin de L’ Académie National de Médicine, vol 181, no 1, pp43–138

Ghana Statistical Services (GSS) (2002) 2000 Population and Housing Census; Summary Report of Final Results, Accra, Ghana

Hide, J. M., Hide, C. F. and Kimani, J. (2001) ‘Informal irrigation in the peri-urban zone of Nairobi-Kenya. An assessment of surface water quality for irrigation and its implication for human health in the peri-urban zone of Kumasi, Ghana’, Report OD/TN 105, March 2001, HR Wallingford, DFID, UK

IMCSF (1974) Micro-organisms in foods. Sampling for microbiological analysis: Principles and specific applications. The International Commission on Microbiological Specifications for Food, University of Toronto Press, Toronto, Canada

Johnson, P. N. T. (2002) ‘Overview of a recent DFID/NRI/FRI project on street-vended food in Accra’, in SADAOC (2002) Street Food in Ghana: Types, Environment, Patronage, Laws and Regulations, Proceedings of a roundtable conference on 6 October 2001, ISSER, University of Ghana/WILCO, Accra, pp41–51

Keraita, B., Drechsel, P., Huibers, F. and Raschid-Sally, L. (2002) ‘Wastewater use in informal irrigation in urban and peri-urban areas of Kumasi, Ghana’, Urban Agriculture Magazine, vol 8, pp11

Mensah, E., Amoah, P., Drechsel, P. and Abaidoo, R. C. (2001) ‘Environmental concerns of urban and peri-urban agriculture: Case studies from Accra and Kumasi’, in Drechsel, P. and Kunze, D. (eds), Waste Composting for Urban and Peri-urban Agriculture: Closing the Rural-Urban Nutrient Cycle in Sub-Saharan Africa, IWMI, FAO, CABI, pp55–68

Mensah, P. D., Yeboah-Manu, K., Owusu-Darko, K. and Ablordey, A. (2002) ‘Street foods in Accra, Ghana: How safe are they?’ WHO Bulletin, vol 80, pp546–554

Ntow, W. J. (2001) ‘Organochlorine pesticides in water, sediment, crops, and human fluids in a farming community in Ghana’, Archives of Environmental Contamination and Toxicology, vol 40, pp557–563

Nyanteng, V. K. (1998) ‘Draft summary report on food markets and marketing in the Accra metropolis’, in Report on the National Seminar on Food Supply and Distribution Systems, AMA/FAO, Accra, 1998

Okorley, E. L. and Kwarteng, J. A. (2002) ‘Current status of the use of pesticides in urban and peri-urban vegetable production in the central region of Ghana’, Paper presented at RESEAU Ghaneen/SADOAC Workshop on Sustainable Food Security in West Africa, 30–31 January 2002, Milkin Hotel, Accra

Pesticide Safety Directorate (2005) Excel Spreadsheet of current MRLs. Department for Environment, Food, and Rural Affairs, London, UK. Available at: www.pesticides.gov.uk/uploadedfiles/web-Assets/PSD/MRL-Spreadsheet.xls (accessed August 2005)

Schwartzbrod, J. (1998) Methods of Analysis of Helminth Eggs and Cysts in Wastewater, Sludge, Soils and Crops, Université Henry Poincare, Nancy, France

Shuval, H. I., Fattal, B. and Yekutiel, P. (1986) ‘State of the art review: an epidemiological approach to the health effects of wastewater reuse’, Water Science and Technology, vol 18, pp147–162

Tallaki, K. (2005) ‘The Pest-control System in the Market Gardens of Lomé, Togo’, in Mougeot, L. J. A., Agropolis: The Social, Political and Environmental Dimensions of Urban Agriculture, Earthscan Publications, London, UK

Tchobanoglous, G., Franklin, B. and Stensel, D. (1991) Wastewater Engineering: Treatment, Disposal and Reuse, Metcalf and Eddy Inc., New York

Volpi, A. M. (2005) ‘Pesticides, can we avoid them?’, www.annamariavolpi.com/pollutants.html (accessed June 2006)

WHO (2005) The WHO Recommended Classification of Pesticides by Hazard and Guidelines to Classification: 2004, WHO, Geneva, Switzerland

WHO (2006) Guidelines for the Safe Use of Wastewater, Excreta and Grey Water: Wastewater Use in Agriculture, vol 2, WHO, Geneva, Switzerland

7
Water Contamination and its Impact on Vegetable Production in the Rímac River, Peru

Henry Juarez1

INTRODUCTION

The rapid growth of population in Lima is leading to an expansion of unplanned informal settlements which lack many basic urban services such as waste management, clean water and drainage. Combined with poor farming practices, these settlements create the perfect conditions for large amounts of run-off. This run-off contains waste products, for instance from overflowing septic systems, that affect surface waters, the environment and ultimately the health of food producers and consumers. The latter face risks from waterborne diseases, while the farmers are also affected by skin ailments and intestinal problems. Due to the scarcity of clean water and lack of adequate treatment of domestic sewage, the use of contaminated water is a common practice which sustains farmers’ livelihoods in urban and peri-urban areas. Lima is merely one of many international cities located at the mouth of a large river, many of which face the same environmental problems that directly impact on water quality. Typically, such waters are polluted by excessive quantities of nutrients, plus they are contaminated with pathogens and toxic chemical substances that affect both the ecosystem and the public’s health (Lee-Smith and Prain, 2006).

The Rímac River basin is one of the most important in Peru. It provides enough drinking water for approximately 60 percent of Lima’s population of 7.7 million people. The river supplies a large population with a wide range of socio-economic activities including mining, an industry established long ago in the upper and middle part of the basin. In addition, hydroelectric generation occurs along the river, as well as agricultural irrigation.

Water contamination of the Rímac River historically has been related to the discharge of mining waste in the upper and middle part of the basin (Castro, 1993; Infante and Sosa, 1994; MEM-DGAA, 1997; Bedregal et al, 2002). In addition, the disorderly growth of human settlements around areas of crop production – where 35 percent of the vegetables marketed in Lima are produced – aggravates this situation, contaminating the horticultural products with enteric pathogens (Castro and Sáenz, 1990; Moscoso and León, 1994; Moscoso, 1998; Acosta et al, 2001; Manrique et al, 2002a, b).

To date, it has not been clear whether vegetable production sites irrigated with contaminated water have also been affected by heavy metals (HM) and enteric pathogens to the same magnitude. Information about plant uptake of HM, patterns of water quality in the river basin and causes and sources of contaminants are lacking, as well as farmers’ awareness of the risks inherently involved in their work. Furthermore, there are significant public health risks associated with food consumption where urban agriculture (UA) and peri-urban agriculture (UPA) use contaminated water. Given the lack of information on related health risks, this study was initiated to assess and provide guidance to stakeholders at different levels, from producers to consumers.

Research objectives

This research involved an analysis of HM and fecal contamination in the Rímac River basin to determine the environmental risks and the impact on soil, water and vegetables in the eastern part of Lima. The objectives of the present study were to first study historical data (spatial and temporal) on the quality of water in the basin. The specific objectives were to document and analyse the sources of pollution in the river basin both currently and in the past; to map the spatial distribution of HM and fecal contamination in the entire Rímac River basin using secondary data; and to determine whether municipal and national environmental regulations have influenced a reduction in the pollution.

The second major objective was to evaluate existing environmental risks affecting agricultural land, water and vegetables produced. Specifically, the study aimed to understand farmers’ perceptions concerning water quality used in vegetable irrigations; to characterize the actual levels of HM and fecal contamination in water located both in the main weir and irrigation canals; to determine the environmental risks due to the absorption of HM in the soil and the risks to human health due to concentrations of HM and fecal contamination in vegetables; and to propose recommendations to improve the quality of agriculture products.

METHODS

Assessment of historical water quality of the Rímac River

Sources of information

HM such as arsenic (As), cadmium (Cd), chromium (Cr) and lead (Pb) were selected because they are known to be contaminants that can be absorbed by

Figure 7.1 Sampling location of water quality in the Rímac River basin conducted by SEDAPAL (white dots) and DIGESA (black dots)

plants, enter the food chain and steadily accumulate in the organs of human beings, resulting in chronic poisoning (Zhou et al, 2000). The World Health Organization (WHO) guidelines on wastewater identify fecal coliform (FC) bacteria as an indicator of fecal pollution in wastewater used in agriculture and aquiculture (Mara and Cairncross, 1989).

The study included monthly data collection from the General Environmental Health Directorate of the Ministry of Health (DIGESA) on Cd, Cr, As, Pb and FC from 24 sampling stations between 2000 and 2004. Also, we used annual data from the Drinking Water and Sewerage Service of Lima (SEDAPAL) on Cd, As and Pb from 28 sampling stations between 1996 and 2004. It was difficult to obtain monthly data from SEDAPAL due to political issues surrounding As contamination in water. Figure 7.1 shows the sampling stations from both sources used in this research.

Mapping spatial–temporal water quality of the Rímac River

Each sampling station was geo-referenced through a global positioning system (GPS). Spatial and temporal surfaces of water quality were created using spatial analysis from ArcGIS 9.0.2 Annual average numbers for each sample station were used for interpolation. A buffer of 2000 m around the Rímac River was defined in order to delimit the interpolated values, and the inverse distance weighted (IDW) tool was used to interpolate the parameters for water quality.

Table 7.1 Water quality index based on Class III stipulated by the Peruvian law for vegetable irrigation and animal consumption

aAbove permissible limits for vegetable irrigation. MPN, most probable number.

To assess the contamination status of water, the General Water Law of Peru was consulted (MINAG, 1969, 1976). A water quality index was designed to provide a simple and concise method for expressing the water quality of the river. The index was classified in four levels, taking into account the stipulations for water used for vegetable irrigation and animal consumption (known as Class III waters) as determined by the Peruvian General Water Law. Low and medium indexes consider values below the maximum permissible limits set for Class III, while a high index considers values two to ten times higher than the maximum permissible set for Class III. A very high index considers values that are extremely high which are set for Class III, but which are unsuitable for vegetable irrigation (Table 7.1).

Water quality maps were prepared of each of the water quality parameters sampled. While they are based on average annual values, and hence do not give us detailed information about monthly variation, they offer a general idea of what is influencing their concentrations across the Rímac River.

Perception of risk of pollution or exposure to contaminants

A detailed survey of 125 of the 276 livestock- and vegetable-producing households had been previously carried out in the Lurigancho Chosica district.3 Survey interviewers asked questions on characteristics of members of the household, the context of the livestock activity, agronomic management, livestock management, post-harvest and market activities, water use, excreta and solid waste disposal, complementary activities, expenditures and family income, division of labour, and institutional and organizational networks that are assisted by the activities of UA (Lozano, 2004). We used this survey to understand what farmers think about their exposure to pollution or contaminants.

Assessment of water, soil and crop quality in eastern Lima

Lurigancho Chosica is one of the most important farming areas, supplying as much as 35 percent of the vegetable market in Lima. It is located in the lower

Figure 7.2 Urban and peri-urban agriculture in and around Lima (agricultural plots shown in black, urban areas in grey) and three research locations: Carapongo, Nievería and Huachipa

watershed of the Rímac River in the eastern part of the city. Three of its four vegetable production sites were selected in this study to assess the impact of irrigation water on soils and in vegetables. The three sites selected for study were Carapongo, Nievería and Huachipa (Figure 7.2). Site selection was chosen because of the extent of the farms (871 of 1115 ha of all vegetable production areas are located in these three sites). In addition, land use change is occurring very rapidly in the fourth farming area (Ñaña) because of the conversion of agricultural land into permanently built-up areas which are covered with infrastructure (Juarez et al, 2007).

Water quality of irrigation canals

Twenty-five water samples were collected in both October 2004 and March 2005 in Carapongo for HM and FC analysis. Additionally, 20 water samples were collected in Huachipa and Nievería in September 2005 for FC. Sampling sites included main and lateral irrigation canals, water gates and proximity to populated places. Criteria included samples taken during both the dry and the wet seasons.

Water samples for HM were preserved with nitric acid for subsequent laboratory analysis. Water samples for FC were collected in sterilized glassware containers. Each sample was geo-referenced by a GPS location. To assess the contamination status of water, the General Water Law of Peru was consulted (MINAG, 1969, 1976) and the indexes displayed in Table 7.1 were used to classify all maps.

Quality of vegetables

Vegetables represent the main crop around Carapongo, Nievería and Huachipa. In total, 57 vegetable samples were collected and sampled.

Carapongo was divided in five agricultural zones to allocate samples proportionally (Campo Sol, Guadalupe, Huancayo, Nuevo Horizonte and Tulipanes). Five vegetable samples within each zone were collected in December 2004 (FC analysis) and April 2005 (FC and HM analysis). Crops included were those noted in previous surveys conducted in the area (Lozano, 2004): huacatay (Tagetes minuta L., used as seasoning or as a condiment in Peruvian foods), lettuce (Lactuca sativa L.), radish (Raphanus sativus L.), turnip (Brassica rapa L. var. rapa) and the common beet (Beta vulgaris L. var. crassa). An additional seven samples were collected in Nievería and Huachipa in March 2006 for FC analysis. Selected crops were turnip, common beet, huacatay, celery (Apium graveolens L.) and grass (Lolium perennial L.).

Five to six sub-samples were taken for each vegetable sample during harvest season. For the assessment of HM, vegetables were classified in accordance with their edible parts: roots (turnip, radish and common beet) and foliage (lettuce, huacatay and celery). In addition, vegetables were collected before and after washing for FC assessment.

The freshly harvested vegetables were brought to the laboratory and washed, initially with running water to remove the soil particles, followed by three washings with distilled water. Samples were cut into small pieces before being oven-dried at 105°C to constant weight. The samples were then pulverized with a mortar and subjected to wet digestion in the conical flask with HNO3 and HClO4 (2:1) to extract total HM (AOAC, 1984). Then 10 ml of HCl was added to dissolve inorganic salts and oxides. Care was taken to prevent contamination in all steps of the process. HM in all samples were determined with atomic absorption spectroscopy (AAS) for Cd and Cr, AAS + graphite furnace for Pb, and AAS + hydride generation for As.

Guidelines for HM in vegetables were taken from several sources including the 1995 Chinese National Standards Analytical Methods (Zhou et al, 2000) and Codex Alimentarius (2006). The guidelines for HM were set at 0.50 mg As/kg, 0.20 mg Cd/kg, 0.50 mg Cr/kg and 0.3 mg Pb/kg in fresh matter.

FC bacteria count was analysed by the five-tube most probable number method (MPN) (APHA, 1992). Limits for FC were taken from those recommended by the WHO (1989).

Heavy metals in soils

Soil samples were obtained from the same place as vegetable samples. All soils were sampled manually with a soil sampler (Model J, Spectrum Technologies).4Samples were extracted from topsoil (approximately 20 cm depth) and air-dried for four days. Tiny roots and other residues were removed before the soil was ground and sieved in a 2-mm mesh. Fractions less than 2 mm were analysed for total as well as dissolved HM.

Total HM content was determined using an acid mixture. One gram of soil for each sample, in duplicates, was transferred into a 200-ml digestion flask. Ten millilitres of a mixture of concentrated HClO4 and HNO3 (ratio of 2:1) was added before covering the digestion flask with a watch glass. The mixture was heated progressively and boiled under reflux for two hours after which the digestion flask was cooled (Baker and Amacher, 1982; AOAC, 1984). Tests for available HM in soils were carried out with 1N ammonium acetate shaking for two hours (Bradi, 1984; Yanai et al, 1998). Heavy metals in all samples were determined with AAS for Cd and Cr, AAS + graphite furnace for Pb and AAS + hydride generation for As.

Guidelines for HM concentrations in soils were taken from the Taiwan standards for assessment of soils contaminated with HM (Wang et al, 1994; Chen et al, 1996). These included A values (the upper limit of background concentration), B values (the acceptable level) and C values (the intervention level, at which pollution control is needed). These guidelines take into account

Table 7.2 The Taiwanese standards for assessment of soils contaminated with heavy metals

A value, reference top value of the background range; B value, further monitoring level; C value, pollution control level.

Source: Chen, 1992; Chen et al, 1996.

total HM in soils and available HM for plants (extracted by 0.1 M HCl) (Wang et al, 1994). The C values were used as guidelines for assessment of HM in soils.

Statistical analysis of the data

Data were subjected to an analysis of variance and means were compared by Fisher’s protected least significance difference (LSD) test. Statistical analysis was carried out with assistance of CIPSTAT7 (CIP Statistical Analyser) developed at the International Potato Center. Descriptive statistics were used to explain the basic features of the data in a study to provide simple summaries about the samples.

RESEARCH FINDINGS AND DISCUSSIONS

Historical contamination of Rímac River

The incorporation of spatial–temporal data on water quality in a geographical information system (GIS) proved to be a useful tool in assessing pollution trends for different metals in different parts of the watershed and the associated potential risks. As the maps track changes over time, they are useful in identifying trends and offer a general idea of what is influencing their concentrations throughout the Rímac River. Looking at these summary maps is faster and easier than doing extensive detailed analysis of the raw water-quality data. Also, they help to focus the analysis directly on areas of concern, locate areas of high risk due to the presence of HM and FC and assesses the discrepancies between different data sets.

Figure 7.3 shows the annual averages over time of As, Cd, Cr, Pb and FC reported by SEDAPAL (1997–2004) and DIGESA (2000–2004), while Figure 7.4 shows the spatial distribution of As, Pb and FC along the Rímac river corresponding to the year of greatest contamination (2000 or 2001) and the last reported (2004).

Cd and Cr did not affect any part of the basin and were always below the maximum permissible level for vegetable irrigation. However, vegetable growing areas in the lower part of the basin were affected with contaminated water for at least two years with As (2000 and 2002) and for the entire evaluated period with Pb (1997–2004).

As expected, the presence of Pb and As in the Rímac River is related to discharges of mining wastes in the upper and middle part of the basin (Infante and Sosa, 1994; MEM-DGAA, 1997). The data also indicates that As levels have reduced in the last two years (2003–2004) (Figure 7.5A). This is probably due to an obligatory implementation of the Environment Adequacy Program (Programa de Adecuación Ambiental, PAMA), as applied to ongoing mining operations and the Environment Impact Assessments which are applied to new mine operations (MINEM, 1993). However, there is no evidence that water quality has been improved in terms of Pb (Figure 7.5B). High levels of

Figure 7.3 Annual means for As, Cd, Cr, Pb and FC in the Rímac River, where the vertical lines represent the standard error of the mean

Figure 7.4 Spatial patterns of water contamination with As, Pb and FC in the Rímac River for the year of greatest contamination (2000 or 2001) and the most recent year (2004)

Pb still exist in the Rímac River basin which could pose an important risk for agricultural activity in the Rímac Valley. Figure 7.6 shows the degree of mining activity established long ago in the basin.

A different situation regarding microbiological contamination was observed. FC were found to be higher than 105/100 ml in almost the entire river except for a small section of the highest part of the basin (see Figures 7.3 and 7.4). These findings were also reported in other studies (Castro and Sáenz, 1990; Moscoso and León, 1994; Moscoso, 1998; Acosta et al, 2001; Manrique et al, 2002a, b). These high levels of FC present in the river are due to the

Figure 7.5 As and Pb in the upper part of the basin from 1997 to 2004, where vertical lines represent the standard error of the mean

inadequate treatment of domestic sewage from human settlements along the Rímac River. Improper disposal thus contaminates the surface water. As an indicator bacteria, the presence of such a level of FC contamination suggests that the water poses a risk for agricultural activity in the Rímac Valley.

Perception of risk to pollution or exposure to contaminants

Among survey participants, 74 percent of farmers perceived that water used for irrigation is contaminated and 73 percent thought domestic sewage and

Figure 7.6 Location of mines and populated places in the Rímac River basin

solid waste from residential communities comprised the principal sources. Eleven percent of farmers thought that waste from mining activities affected water quality and were aware that HM can affect soils, crops and human health. The appraisal of how farmers perceive the risk to pollution or exposure of contaminants can be very complex (Grasmück and Scholz, 2003). These results show that farmers’ perception of water quality is attributed mainly to factors that are easily observed such as domestic sewage (excreta) and solid waste from urban communities (such as bottles and papers). If the exposure is not visible (e.g. the presence of parasites and pathogens or HM in irrigation water or the accumulation of HM in soil or plants), the perception of the risk is less apparent.

Assessment of water, soil and crop quality in eastern Lima

Water quality of irrigation canals

None of the water samples collected from the canals exceeded the maximum permitted limits for As, Cd, Cr and Pb according to Codex Alimentarius (2006). Thus the water is suitable for use in irrigating vegetables from the standpoint of HM.

Contamination of irrigation water with pathogens and parasites is a more serious problem for vegetable production. The Rímac River is a main source of contamination due to the sewage and excreta effluents that are emptied directly into irrigation canals that supplement farms. More than 97 percent of water samples taken from irrigation canals were found to contain five million FC, which is above the maximum permitted limits. The quality of irrigation water worsened downstream as the river passed through more densely populated settlements in the Rímac Valley.

Heavy metal contents in soils

From the analysis of the soils it was observed that more than 40 percent of the samples contained high levels of As and Cd, as well as Pb that exceeded the safe limits according to Chen (1992), Chen et al (1996) and Wang et al (1994). The total concentration (Figure 7.7) indicates the location of potential contamination pathways, such as soil ingestion by children, inhalation of dust, soil adhesion on edible leaves and other sources associated with handling the soil (Nabulo, 2002). Extractable concentrations of HM in soils are considered to be indicators of availability to plant roots (Kimberly and William, 1999). Observations in this study (Figure 7.8) showed that uptake by plants was only a fraction of the total HM concentrations (Sauve et al, 2000). Because there are no guidelines for safe levels of available As in soils, it was impossible to compare amounts.

Some of the available Cd measured in soil solution may have come from the weathering of parent rock, as it was also observed that 68 percent of the variability of available Cd came from total Cd in soils (P < 0:001). The levels of trace elements present in soils result from the interactions between the geology of the parent rock and the soil-forming factors and human activities (Kabata-Pendias and Adriano, 1995). However, available As found in the soil may have originated from pesticides, fertilizers or sewage sludge (Alloway and Ayres, 1993).

The results also suggest different amounts of HM in soils across evaluation sites. Localities near the main gate of Nuevo Horizonte accumulated more Cd (P < 0:001) and Pb (P = 0:019) than other sites that were far from the gate.

Quality of vegetables

In spite of the mechanisms involved in elemental uptake by roots (non-metabolic or metabolic), plants are known to respond to the amounts of available inorganic nutrients in soil solution (Li and Shumas, 1996; Baban, 1999; Madrid et al, 2002). Analysis of some of the vegetables in the irrigated fields showed that both Cd and As were accumulated more in foliage than in roots. But only the indigenous aromatic plant huacatay exceeded the maximum permitted levels (guideline: 0.50 mg As/kg and 0.20 mg Cd/kg) (Figure 7.9). The plant is used in small quantities for seasoning or as a condiment in various Peruvian foods and so may not pose a health risk to consumers.

Seventeen percent of lettuce and 31 percent of radish samples had very high levels of FC contamination, presumably due to exposure to the highly contaminated canal water (WHO, 1989). Contamination loads in the two

Figure 7.7 Total heavy metals in soils. Vertical lines represent the standard error of the mean

Figure 7.8 Available heavy metals in soils. Vertical lines represent the standard error of the mean

Figure 7.9 Total heavy metals in vegetables. Vertical lines represent the standard error of the mean

Figure 7.10 Levels of FC in vegetables in selected areas: Huachipa, Nievería and Carapongo

Figure 7.11 Effect of washing vegetables in the irrigation canals

vegetables varied depending on the location of sampling (Figure 7.10). Washing vegetables with irrigation water from the canal increased the level of contamination, and 57 percent of clean vegetables were found to be contaminated after the washing process (Figure 7.11).

CONCLUSIONS AND RECOMMENDATIONS

This research exemplifies the use of multiple data sources and multiple pathway analyses to capture different ways through which food gets contaminated. The integration of spatial–temporal data on water quality in a GIS model proved to be a useful tool to evaluate pollution trends for different contaminants flowing through a watershed and downstream into a coastal city, and the associated risks of contamination of vegetable-growing areas. The project demonstrated the usefulness of participatory mapping (GIS) as an effective approach to dissemination of environmental pollution information and risk reduction strategies. These methods also will help the development of appropriate information dissemination packages for the communities who depend for their livelihood upon these ecosystems.

Vegetables grown in areas contaminated by water from the Rímac River had high levels of As and Pb, although a reduction in As has been observed over the last two years due to improvements in the Ministry of Energy and Mines’ regulatory systems. Regulations now exist which involve obligatory programmes of environmental adequacy as well as environmental impact assessments (MINEM, 1993). Despite the high levels of Pb found in the river water, the content found in soils and crops does not pose a significant health risk. However, significant levels of As and Cd were found in soil, and these could be harmful. Investigation of the Cd and As levels in vegetables showed that there was more accumulation in foliage than in roots. The indigenous aromatic plant huacatay contains Cd and As concentrations that are above those permitted by Peruvian law. Contamination of irrigation water with pathogens was a more serious problem for vegetable production. FC were found to reach levels higher than permitted by Peruvian law: more than 97 percent of water samples from irrigation canals contained five million MPN/100 ml, which is beyond the maximum permitted limits for FC bacteria. Analysis of lettuce and radish samples showed 17 percent and 31 percent of the samples having bacterial loads that are above permitted limits. The practice of washing vegetables using water from the irrigation canals further contaminated these vegetables: 57 percent of clean vegetables were contaminated during the rinsing process.

To complement the chemical and biological data collection, the study sought to understand how farmers perceive environmental risks. Not surprisingly, farmers tended to depend on observed phenomenon rather than hearsay references to mining risks. Almost three-quarters associated domestic sewage and other urban solid wastes in the irrigation channels, with only 11 percent mentioning mining. Provision of correct information about the sources of pollution in the environment would go a long way to enhancing awareness of such risks to the community.

The high levels of microbiological contamination raises the need for measures to be applied which would make it possible to improve the quality of water used for irrigation and washing in the study area. While the ideal scenario would be to prevent untreated domestic sewage from entering the river through sewerage, this is not realistic at the moment. Consequently, other mechanisms that permit in situ quality improvement should be considered, such as the use of washtubs of clean water for vegetables or the implementation of reservoirs so as to treat irrigation water.

Based on this research, efforts were made to find an appropriate technology for a simple, low-cost way to improve water quality. Several small reservoirs have been built to test their feasibility for reducing potential contaminants in irrigation water. The cleaning process is quite simple and greatly assisted by solar radiation and increased temperature. First, water enters the reservoir, where it is retained for about 10–14 days. While there, the water is exposed to physical processes that act on the pathogens and parasites. For instance, the action of keeping the suspended bacteria isolated from their biological cycle for more than ten days reduces their concentration and viability. At the same time, parasites sink to the bottom of the reservoir and eventually die, leaving the water clean for vegetable irrigation.

Furthermore, the use of small reservoirs has been shown to have additional benefits. Introducing productive fish to the reservoir compensates for the loss of agricultural land and provides families with additional nutritious food to consume or sell. In addition, this aspect makes the concept of reservoirs appealing for other farmers in the area who might wish to construct one on their land.

ACKNOWLEDGEMENTS

My sincere thanks go to the International Potato Center (CIP), the Urban Harvest Initiative and in particular to the International Development Research Centre (IDRC), whose AGROPOLIS grant programme funded the field research. I would specially like to thank my research supervisor Gordon Prain from Urban Harvest and Julio Moscoso from the Centro Panamericano de Ingeniería Sanitaria y Ciencias del Ambiente (CEPIS) for their time and effort in helping me develop a research product. They provided excellent scientific guidance pertaining to their specialties. I also want to thank my thesis advisor, Victor Aguilar from Universidad Nacional Agraria la Molina, for his advice. Finally, I want to thank Reinhard Simon for providing the freedom and opportunity to pursue this research as part of my activities at CIP.

NOTES

1 Henry Juarez, MSc, Environmental Sciences, PO Box 1558, Lima 12, Peru, tel: + 51-1-3175311 ext 2156; email: h.juarez@cgiar.org

2 Copyright 2002 ESRI. 380 New York Street, Redlands, CA 92373-8100, USA.

3 Urban Harvest baseline survey (Huachipa, Ñaña, Nievería y Carapongo)

4 www.specmeters.com/Soil_Moisture/Soil_Samplers.html

5 http://riuweb.cip.cgiar.org/cipstat/

REFERENCES

Acosta, L., Alvaro, H., Jiménez, H., Manrique, R. and Torralba, R. (2001) Proyecto Regional Sistemas Integrados de Tratamiento y Uso de Aguas Residuales en América Latina: Realidad y Potencial, Estudio General del Caso, Villa el Salvador, Lima, Peru

Alloway, B. J. and Ayres, D. C. (1993) Chemical Principles of Environmental Pollution, Oxford University Press, Oxford, UK

APHA (1992) Standard Methods for the Examination of Water and Wastewater, 18th edition, American Public Health Association, Washington, DC, USA

Association of Official Chemists (AOAC) (1984) Official Methods of Analysis, 14th edition, Sidney Williams (ed), AOAC, Arlington, VA, USA

Baban, S. M. (1999) ‘Use of remote sensing and geographical information systems in developing lake management strategies’, Hydrobiologia, vol 395–396, pp211–226

Baker, D. E. and Amacher, M. C. (1982) ‘Nickel, copper, zinc and cadmium’, in Page, A. L., Miller, R. H. and Keeney, D. R. (eds) Methods of Soil Analysis, Chemical and Microbiological Properties, SSSA Publishers, NY, USA, pp323–336

Bedregal, P., Montoya, E., Torres, B., Olivera, P., Mendoza, P., Ubillus, M., Ramos, R., Creed-Kanashiro, H., Penny, M., Aleman, M., Gil, A., Junco, J. and Ganoza, L. (2002) ‘Determination of trace elements and heavy metals in agricultural products cultivated at the River Rimac in the city of Lima’, Co-ordinated research project on use of nuclear and related analytical techniques in studying human health impacts of toxic elements consumed through foodstuffs contaminated by industrial activities, Report on the First Research Coordination Meeting, Vienna, Austria, 18–22 March 2002, International Atomic Energy Agency

Bradi, M. A. (1984) ‘A study of comparative behaviour of selected heavy metals in polluted and unpolluted estuarine and coastal sediments’, Ph.D. Thesis, University of Lancaster, Lancaster, UK

Castro, M. L. (1993) Evaluación de Riesgos para la Salud por el uso de las Aguas Residuales en Agricultura: Aspectos Toxicologicos, Centro Panamericano de Ingeniería Sanitaria y Ciencias del Ambiente (CEPIS)

Castro, M. L. and Sáenz, R. (1990) Evaluación de Riesgos para la Salud por el uso de Aguas Residuales en Agricultura; Aspectos Microbiológicos, Centro Panamericano de Ingeniería Sanitaria y Ciencias del Ambiente (CEPIS)

Chen, Z. S. (1992) ‘Metal contamination of flooded soils, rice plants, and surface waters in Asia’, in Adriano D. C. (ed), Biogeochemistry of Trace Metals, Lewis Publishers, FL, USA, pp85–107

Chen, Z. S., Lee, D. Y., Lin, C. F., Lo, S. L. and Wang, Y. P. (1996) Contamination of Rural And Urban Soils in Taiwan, Boston, London

Codex Alimentarius (2006) ‘Codex General Standard for Contaminants and Toxins in Foods’, CODEX STAN 1993–1995, Rev.2–2006, www.codexalimentarius.net/download/standards/17/CXS_193e.pdf (accessed December 2007)

Grasmück, D. and Scholz, R. W. (2003) ‘Risk perception of heavy metal soil contamination by high-exposed and low-exposed inhabitants’, ETH–UNS Working Paper No 42, Zurich, Switzerland

Infante, L. and Sosa, S. (1994) ‘Cuantificación espectrofotométrica de arsénico en aguas de consumo humano en la vertiente del Río Rímac’, tesis presentada a la Universidad Nacional Mayor de San Marcos, Facultad de Farmacia y Bioquímica para el Grado de Químico Farmacéutico

Juarez, H., Prain, G. and Castro, M. (2007) ‘The use of spatial analysis in urban and peri-agriculture research and development in Lima’, Working Paper, International Potato Centre, Lima, Peru

Kabata-Pendias, A. and Adriano, D. C. (1995) ‘Trace metals’, in Rechcigl, J. E. E. (ed) Soil Amendments and Environmental Quality, Lewis Publishers, Boca Raton, FL, USA, pp139–168

Kimberly, M. F. and William, H. (1999) ‘Trace metals in Montreal urban soils and the leaves of Teraxacum officinale’, Canadian Journal of Soil Science, vol 79, pp385–387

Lee-Smith, D. and Prain, G. (2006) ‘Urban Agriculture and Health’, in Hawkes, C. and Ruel, M. T. (eds) Understanding the Links Between Agriculture and Health. International Food Policy Research Institute (IFPRI), Washington, DC, USA

Li, S. and Shumas, L. M. (1996) ‘Heavy metals movement in metal-contaminated soil profiles’, Soil Science, vol 161, pp656–66

Lozano, M. (2004) Documentacion de las bases de datos de la encuesta de linea base del Cono Este: Nieveria, Carapongo, Huachipa, Ñaña, Cosecha Urbana – Centro Internacional de la Papa (CU-CIP)

Madrid, L., Diaz-Barrientos, E. and Madrid, F. (2002) Distribution of heavy metal content of urban soils in parks of Seville, Chemosphere vol 49, pp1301–1308

Manrique, R., Jiménez, H., Álvaro, H., Torralba, R. and Acosta, L. (2002a) ‘Estudio de Viabilidad: Sistema de Aprovechamiento de las Aguas Residuales en el Fundo San Agustín, Callao – Perú’, Proyecto Regional Sistemas Integrados de Tratamiento y Uso de Aguas Residuales en América Latina: Realidad y Potencial, Convenio IDRC–OPS/HEP/CEPIS 2000–2002, vol 86

Manrique, R., Jiménez, H., Álvaro, H., Torralba, R. and Acosta, L. (2002b) ‘Estudio complementario del caso fundo de San Agust