Restoring The Soil

Restoring the Soil
A Guide for Using Green Manure/Cover Crops to Improve

the Food Security of Smallholder Farmers
Roland Bunch
Canadian
Foodgrains
Bank
A Christian Response
to Hunger
Praise for Restoring the Soil

Hundreds of thousands of small farmers in many countries are using, experimenting
with, and sharing green manure/cover crops, thereby reducing fertilizer use while
improving the sustainability of crop production systems. Restoring the soil by using
these technologies to feed food-insecure people will have a tremendous impact on
improving global food security. Roland Bunch brilliantly shows how green manure/
cover crops can best be used to achieve these objectives.
Elmer Lopez Rodriguez
Secretary of Agrarian Affairs
Government of Guatemala
Roland Bunchs book on cover crops and green manures could not have come at
a better time. Decades of experimentation and experience by farmers, technicians
and scientists are presented clearly and practically, making this work an essential
contribution not just for world food security but for the food sovereignty of the
2 billion smallholders who produce the bulk of the worlds food.
Dr. Eric Holt-Gimenez
Executive Director
Food First/Institute for Food and Development Policy
Roland Bunchs book distills for us decades of rich field experience in countries all
around the world, where millions of farmers have given practical meaning to the
concept of synergy. Farmers are doing this by growing a great number of crops in
intelligent and profitable association with a huge variety of green-manure/covercrop plants. These marvelous plants enrich the soil at the same time that they
enhance crop yields, conserve soil moisture, and save farmers labor. Thanks to
Bunchs keen observations, systematic analysis and critical thinking, this agroecologically-based strategy is now accessible to anyone who will read, digest and apply
what is in this book. It fills a huge gap in our published knowledge about how
to make the most of our land, labor and water for meeting our food needs in
sustainable ways.
Dr. Norman Uphoff
Professor of Government and International Agriculture
Cornell University
A timely and extremely valuable book from one of the worlds leading experts on
green manures/cover crops. Clearly written, it is both practical and analytical, and
richly illustrated with photographs from a great number of agroecosystems worldwide.
Highly recommended.
Dr. Jules Pretty
Author of Regenerating Agriculture
University of Essex
This is the book we have been waiting for. Restoring the Soil distils and condenses a
lifetime of learning and a wealth of experience. It is a wonderful source of knowledge
and advice about a vital and neglected area of huge potential, a treasure trove for small
farmers around the world and those who work with them. Only Roland Bunch with
his lifetime of learning and his unique worldwide experience could have written this.
Restoring the Soil should have a major impact, widening farmers choices and enhancing their sustainable production. It is practical, accessible and wide-ranging across an
astonishing variety of green manures/cover crops. Let me hope that it will be widely
available in several languages, and widely used.
Dr. Robert Chambers
Institute of Development Studies
University of Sussex
Restoring the Soil

A Guide for Using Green Manure/Cover Crops to Improve
the Food Security of Smallholder Farmers
Roland Bunch
Copyright Canadian Foodgrains Bank, 2012. All rights reserved.

No part of this book may be reproduced in any form without permission from Canadian Foodgrains Bank.
For more information, contact the Foodgrains Bank at [email protected]
Library and Archives Canada Cataloguing in Publication
Bunch, Roland, Restoring the soil: a guide for using green manure/cover crops to improve
the food security for smallholder farmers / Roland Bunch.
ISBN 978-0-9688546-4-8
1. Green manuringDeveloping countries. 2. Green manure cropsDeveloping countries.
3. Soil fertilityDeveloping countries. 4. Soil managementDeveloping countries.
5. Food securityDeveloping countries. I. Canadian Foodgrains Bank II. Title.
S661.B85 2012
631.874091724
C2012-905464-X
Founded in 1983, Canadian Foodgrains Bank is a partnership of 15 churches

and church-based agencies working together to end hunger.
Members: ADRA Canada (Adventist Development and Relief Agency) Canadian Baptist Ministries
Canadian Catholic Organization for Development & Peace Canadian Lutheran World Relief
Christian and Missionary Alliance in Canada World Renew (formerly Christian Reformed World Relief
Committee of Canada) Emergency Relief & Development Overseas (Pentecostal Assemblies of Canada)
Evangelical Missionary Church of Canada Mennonite Central Committee Canada
Nazarene Compassionate Ministries Canada Presbyterian World Service & Development
Primates World Relief & Development Fund (Anglican Church of Canada) The Salvation Army
The United Church of Canada World Relief Canada

Canadian
Foodgrains
Bank
to Hunger
Canadian Foodgrains Bank

Box 767, Winnipeg, MB Canada R3C 2L4
1.204.944.1993 www.foodgrainsbank.ca
Canadian International
Development Agency
Agence canadienne de
dveloppement international
Canadian Foodgrains Bank programs are undertaken with the financial support of CIDA.
Design: Roberta Fast
Cover Page Photo: Chris Woodring
Book Photos: Roland Bunch
Printed in Canada by CP Printing Solutions. Printed on recycled paper.
Table of Contents
Preface.................................................................................................................................i
Acknowledgements............................................................................................................. ii
How To Use This Book....................................................................................................... iv
Introduction........................................................................................................................1
1. The Definition of Green Manure/Cover Crops............................................................3
2. A Brief History of Green Manure/Cover Crops...........................................................5
3. Existing Green Manure/Cover Cropping Systems Around The World.....................8
Advantages of Green Manure/Cover Crop Systems.......................................................8
Disadvantages of Green Manure/Cover Crops.............................................................12
Green Manure/Cover Crops Systems are Widespread.................................................14
Characteristics of Known Green Manure/Cover Crop Systems....................................14.
Why Not Just Use Chemical Fertilizer?.........................................................................15
4. Improving Soils: The Basic Rules.............................................................................17
5. Choosing The Right Green Manure/Cover Crop System For A Specific Area......19
What About Farmer Participation?................................................................................19
How to Achieve the Adoption of Green Manure/Cover Crop Systems..........................21
Collecting the Necessary Information............................................................................24
6. Using The Decision Tree............................................................................................27
How to Use the DecisionTree........................................................................................27
The Green Manure/Cover Crop Systems Included in Decision Tree............................27
Characteristics of the Green Manure/Cover Crop Systems Recommended.................27
A Note About Scientific Names.....................................................................................29
7. Decision Tree...............................................................................................................30
Introduction....................................................................................................................30
Decision Tree Diagram..................................................................................................31
Decision Tree Guide......................................................................................................35
8. Green Manure/Cover Crop Systems..........................................................................57
Latin America Green Manure/Cover Crop Systems......................................................57
African Green Manure/Cover Crop Systems.................................................................68
Asian Green Manure/Cover Crop Systems...................................................................72
Annex 1: Glossary of Agricultural Terms....................................................................77
Annex 2: The Evidence..................................................................................................79
Annex 3: List of Recommended Green Manure/Cover Crop Species.......................87
Annex 4: Seed Sources for Green Manure/Cover Crops...........................................90
Annex 5: Additional Resources on Green Manure/Cover Crops...............................94
Restoring the Soil
Preface
As world population numbers tick ever higher, ensuring that food production keeps
pace is one of the biggest challenges facing humanity. Large-scale agriculture will
provide part of the solution, but smallholder farmers (farmers in developing countries
who often have limited land and capital, are poorly linked to markets and are vulnerable to risks) will also play a vital role in feeding the next generations.
Millions of smallholder farmers around the world, however, are facing a serious
soil fertility crisis, and many of these families also suffer from food insecurity. Soil
infertility and erosion losses in many regions of the world are standing in the way of
increased food production and improved livelihoods for many smallholder farmers.
Maintaining, and in many cases recovering soil fertility, has become a major challenge
facing agricultural professionals and farmers.
Green manure/cover crops are proving to be an effective, locally appropriate and lowexternal-input solution to this crisis. This strategy for improving livelihoods of some
of the worlds most food-insecure people needs to be shared with agriculture development workers and smallholder farmers around the world.
With this objective in mind, the Canadian Foodgrains Bank (CFGB) contracted
agroecologist Roland Bunch to write this book. Roland has years of international
field experience with green manure/cover crops and is widely recognized as one of the
worlds experts on this subject. He also has gained international respect for his passion
to help smallholder farmers through a people-centered agriculture revolution, summarized in his popular book Two Ears of Corn: A Guide to People-Centered Agriculture
Improvement.
This new book, Restoring the Soil, synthesizes Rolands extensive field-based research
gathered from thousands of smallholder farmers he has visited around the world who
incorporated green manure/cover crops into their farming systems. This book presents
the information in a user-friendly format intended to help agricultural development
workers and farmers decide what systems may be most appropriate for the geographical area they work in.
The CFGB network of members and partners is committed to building capacity
around the use of green manure/cover crops in agriculture development projects.
We are convinced that green manure/cover crops are among the best solutions to
sustainably increase production while improving soil fertility for smallholder farmers.
It is our sincere desire that this book will assist not only the CFGB network, but also
a wider network of organizations and smallholder farmers in designing sustainable
cropping systems.
Jim Cornelius, Executive Director
Canadian Foodgrains Bank
Restoring the Soil
Acknowledgements
This book could never have been written if it were not for a handful of people who
had the foresight to realize, 30 years ago, that sometime soon, the world demand for
energy would outstrip the supply, and energy prices would increase dramatically. Furthermore, they knew that a major increase in energy prices would inevitably increase
the price of chemical fertilizer, which would, in turn, mean that over a billion people
around the world who were dependent on chemical fertilizer would have no known,
economically feasible way of maintaining their soils fertility. These peoples predictions could not have been more accurate: energy prices have increased five-fold over
the last decade, real fertilizer prices have doubled in less than five years, world food
prices have soared and millions of smallholder farmers are watching their soils become
rather similar to infertile bricks.
But due to the foresight of this handful of people, we now have a large selection of
effective, very inexpensive alternatives to chemical fertilizer.Thousands of people have
worked on green manure/cover crops over the last 30 years, but the pioneers and leaders in this effort are fairly well-known:

Ana Primavesi has been the pioneer and theoretical leader of the entire Brazilian zero tillage and green manure/cover crop movement.Her classic book, The
Ecological Management of the Soil, still has no rival worthy of the name.
Claudio Monegat probably did more than anyone else to further the early
experimentation with green manure/cover crops in Brazil.He later wrote his own
classic, Green Manuring in Southern Brazil.
Valdemar Hercilio (Salgado) de Freitas and Aldemir Calegari have headed up

the state agricultural organizations for the Brazilian states of Santa Catarina and
Parana, where most of the early green manure/cover crop work was done in Brazil
and where over a million farmers now regularly use green manure/cover crops.
Rolf Derpsch worked with the movement in Brazil for many years, and then
spread it to much of Paraguay.
Steve Gliessman and Roberto Garcia did pioneering work with jackbeans on the
Yucatan Peninsula in Mexico.
In 1983, within a year or two of the time the Brazilians initiated their work on
green manure/cover crops, those of us in World Neighbors/Honduras began
experimenting with half a dozen green manure/cover crop species.
Milton Flores became the director in 1989 of the International Green Manure/
Cover Crop Clearinghouse (CIDICCO), which was founded by World Neighbors/Honduras. He ably ran CIDICCO for two decades, spreading information
about green manure/cover crops to more than 75 countries around the world.
In addition to these pioneers, I would like to thank the members of the Green
Manure/Cover Crop Taskforce organized by the Canadian Foodgrains Bank (CFGB)
ii
iii
Restoring the Soil
who recognized the need to have this book written. Dr. Tom Post and Dr. Wondimu
Kenea from World Renew (formerly Christian Reformed World Relief Committee)
first captured the vision of this book. They also came up with the crucial idea of
building the book around a decision tree, so that the often difficult and complicated
task of choosing among scores of green manure/cover crop systems could be simplified
for the common practitioner. Dawn Berkelaar from ECHO Inc. has done an incredible job of editing the book, simplifying technical language where it was needed, and
making sure, against all odds, that all the numbers in the text actually do correspond
to the numbers in our latest version of the decision tree. Dr. Tim Motis from ECHO
assisted in editing and adding a valuable seed resources section. Additional help in
editing was provided by Philip Bender, Carol Thiessen, Rachel Evans, Tiffany Hiebert,
Emily Cain, and John Longhurst. Lastly, this book would definitely not exist if Alden
Braul from CFGB had not initiated the idea and pushed this project forward to its
completion.
Roland Bunch
Restoring the Soil
How To Use This Book

This book is written primarily as a practical manual. It is meant to support
agricultural workers and smallholder farmers around the world in maintaining and/
or improving soil fertility and effective weed control through the use of green manure/
cover crops (gm/ccs). If you are exploring whether you wish to work with green
manure/cover crops, the definitions, history and practical field level advantages and
disadvantages of this technology will aid you in making that decision. The answer
to why one would not use chemical fertilizer and the basic considerations of soil
improvement are practical and non-ideological. If you are already convinced of the
importance of green manure/cover crops, these sections will refresh your commitment
and certainly provide extra motivation.
The methodological considerations of extension methodology that should be considered
are covered in the sections of What About Farmer Participation, How to Achieve the
Adoption of Green Manure/Cover Crops, and Collecting the Necessary Information. If the
decision is made to promote this technology, then the elements mentioned in
Collecting the Necessary Information should be carefully followed as a guide. Errors
in this stage can seriously affect the probability of achieving positive results in the
application of successful systems in the future.
The decision tree is a thoughtfully designed step-by-step process. It is presented to
help the reader decide which of the 91 systems presented in the book have a better
chance of being helpful in his/her particular work environment. This environment is
by and large defined by answering the questions mentioned in the Collecting the
Necessary Information section. The decision tree uses parameters such as altitude,
predominant crops and cropping systems, the slope of the land, rainfall patterns,
whether animals graze in a controlled fashion or not, etc., to guide the readers
decision process.
The decision tree will logically guide you, step by step, to green manure/cover crops
systems to be considered for situations you have in your work area. Each system is
assigned a number (from S1 to S91) which is used throughout the manual to help
you identify and to note the systems you might be interested in testing as you work
through the tree. Explanations of the 91 systems are given in the Green Manure/Cover
Crop Systems section.
Annex 2 (The Evidence) deals with a few issues related to the scientific basis or
justification of the systems. It provides the field level practitioners, who are primarily
interested in practical and real results, clear, practical evidence that these systems have
a scientifically defendable basis.
Phil Bender, Independent Agricultural Consultant
Santa Cruz, Bolivia
iv
Restoring the Soil
Introduction
During the last 30 years, green manure/cover crops have become an important
agricultural technology for the developing world. Books on tropical soil management
written during the 1950s, 1960s or 1970s seldom mention green manures except,
perhaps, to briefly suggest they were a failure. But then in the 1980s, green manures
began making a regular appearance in the literature.1 Since the 1990s and 2000s, most
books on sustainable or ecological agriculture for the tropics (and even some books on
conventional agriculture) have included significant sections on the subject.
Despite all this recent attention to the subject, far too many people who hear the term
green manure/cover crops (gm/ccs) still picture fields completely covered by mucuna
(Mucuna species) or jackbeans (Canavalia ensiformis). Agronomists frequently run
experiments with three or four easily available gm/cc species, and if these species dont
work, they conclude that gm/ccs are not appropriate for the area.
In fact, over a hundred green manure/cover crop (gm/cc) species are currently in
use around the world, in hundreds of different gm/cc cropping systems. Gm/ccs are
grown together with all the worlds basic subsistence crops, as well as with vegetables,
root crops and trees.
Faced with literally hundreds of different possibilities, even many well-informed
workers in agricultural development feel it is almost impossible for them to be able to
choose the best gm/cc system for the farmers in their area. In the past, a few documents were written to help development workers make the best decisions about this
jungle of possibilities, but often these documents were written by people with little
grassroots experience, or people who only had experience with gm/ccs in one region of
the world.
Unfortunately, some 20 different factors must be taken into account in order to select
the two or three gm/cc systems that have the greatest potential in a specific situation.
These factors include local food preferences, current market conditions, dominant
cropping systems, the major weeds in farmers fields, as well as local economic needs,
environmental conditions and land ownership patterns.
This book presents a decision tree designed to make this difficult task much simpler.
The decision tree guides the reader through a series of simple questions in order to
arrive at the gm/cc cropping systems that would have the highest probability of success in a particular situation.
In all but three or four cases, the gm/cc systems recommended are ones that over a
hundred smallholder farmers have used for at least five years with no outside subsidies
or encouragement. They are systems that have proven to be successful. I have observed
nearly all of these gm/cc systems and have talked with the farmers who use them. To
One of the first times they were mentioned positively would be in Roland Bunch, Two Ears of Corn, A Guide to PeopleCentered Agricultural Improvement (Oklahoma City: World Neighbors, 1982), p. 124.
the best of my knowledge, only two of the gm/cc systems recommended in this book
are no longer being used.
By and large, this book (except for the Annex) is written in such a way that anyone
with a basic understanding of the English language will be able to understand its
content. However, I have included a few technical agricultural terms that many readers of English may not understand. I could have avoided using these terms, but they
are terms that anyone who works with gm/ccs will find useful. In each case, I have
explained the words meaning the first time it is used in the text, and have included it
in the small glossary at the end of the book. A list of the different species mentioned
in the book is also included at the back of the book.
The Definition of Green Manure/Cover Crops
1. The Definition Of Green Manure/Cover Crops

We need to distinguish the term green manure/cover crops (gm/ccs) from what
agronomists have traditionally called green manures. The basic idea of traditional
green manures is that they are plants grown in a field all by themselves (they are
monocropped) and then plowed into the soil (as one does with manure) when they
are in the flowering stage, while they are still green.
But most gm/ccs are managed in an entirely different way. First, rather than always
being planted alone, gm/ccs are usually planted together with farmers traditional
crops and at about the same time (intercropped), or they are planted among the traditional crops just before these crops are harvested (relayed into the traditional crops).
Occasionally gm/cc species are even planted under tree crops. In these ways, farmers
can fertilize their soil without dedicating any extra land to growing the fertilizer.
Second, gm/ccs are almost always cut down after they have produced seeds, rather
than at flowering time. This is done because the farmers want to eat the seeds, sell
them, or feed them to their animals. And, finally, they typically save at least some of
the seeds to plant the next year.
Third, the gm/cc species are almost always cut down and left on top of the soil, rather
than being plowed or dug into the soil. This saves the expense of plowing or digging
them into the soil, and the organic matter on the surface protects the soil from the hot
tropical sun.
While we are still using the basic green manure idea of using plants to fertilize the
soil, most of how we manage the gm/cc species today has been changed by smallholder farmers as they developed better and better ways to use them.
Smallholder farmers also want the chosen species to do a lot more than just fertilize
their soils. One of the most important benefits smallholder farmers want is for the
gm/ccs to control weeds. In fact, most farmers in the tropics are considerably more
interested in controlling weeds than they are in improving their soil fertility. That is
why we have joined the concept of cover crops (whose job is to control weeds) with
that of green manures (whose job is to fertilize the soil).
Because the idea of gm/ccs has changed so completely in the last 35 years, it is
important that we define exactly what we mean by gm/ccs. The following definition,
used in this book, is the most commonly used definition today: a green manure/cover
crop is a species of plant, usually a legume, whether it is a tree, a bush, a vine, a
crawling plant or an algae, which is planted by farmers to maintain or improve their
soil fertility or control weeds, even when they have many other reasons for growing
these plants.
The two main objections to this definition by some agronomists is that systems
that use trees to fertilize the soil (what some people call agro-forestry systems) are
The Definition of Green Manure/Cover Crops
included in the definition, as are systems using legume crops that people eat (grain
legumes). These systems are purposely included here as gm/cc systems for a number
of reasons. First, there is no good reason for excluding legumes that happen to be trees
or that produce food. Trees and grain legumes can also fertilize the soil and control
weeds. Furthermore, they all work similarly in how they fertilize the soil. Lastly, there
is not always a clear line between which plants are trees, which are viny legumes,
and which are grain legumes.
As an example, lablab beans (Lablab purpureus or Dolichos lablab) are known in most
of the world as annual viny legumes, but in some countries, such as Haiti, they are
allowed to grow into trees with woody stems 30 centimetres in diameter. Furthermore, in many places, no one has any idea that they can be eaten, whereas in India
and Kenya they are a valued edible bean that is packaged and sold in supermarkets.
Thus, some people might call lablab beans an agro-forestry species, others a food
legume, and still others a viny legume. Whatever they are, by our definition they
will be included as a gm/cc.
A Brief History of Green Manure/Cover Crops
2. A Brief History Of Green Manure/Cover Crops

Green manures and cover crop systems have been used for at least 2,000 years. The
Romans wrote papers advising people on how best to use green manures. In the richer
countries of Europe and the United States, green manures were by far the most common method of maintaining soil fertility until the end of the Second World War.
Some books from the 1930s include a tremendous amount of information about
green manure systems in the tropicsinformation that today has been almost totally
forgotten.2
Since 1945, the ability of organic matter to effectively fertilize the soil has often
not been emphasized or even mentioned in crop production manuals or agriculture
textbooks (Photos 1, 2, 3 and 4 show examples of the impact of organic matter in
improving crop yields).
History shows beyond doubt that green manuring and gm/cc systems are capable of
maintaining soil fertility for thousands of years. After all, they have done precisely that
in virtually all the worlds different soils, environments and farming systems (Photos 5
and 6).
In the tropics, the main method of maintaining soil fertility historically was by
fallowing. This process involved leaving agricultural land to grow into forests or
grasslands for 10 to 15 years, until the soil had recuperated its natural fertility. Fallowing, like green manuring, maintained the fertility of the soils of whole continents for
thousands of years. Since gm/ccs fertilize the soil in virtually the same way as fallows
do, we can be quite confident (much more than we can be with chemical fertilizers)
that gm/ccs are capable of keeping the soil fertile over the long term. Other methods
are used in the tropics for maintaining soil fertility. Animal manure is, of course, an
extremely good material for maintaining crop fertility, but only rarely do smallholder
farmers have enough for more than about a tenth of their fields (Photo 7). Compost
is probably the best of all soil amendments, but if we calculate the amount of labor
it requires, we will almost always find that it is too expensive for basic grain crops
among smallholder farmers, except in the case of rice (Photo 8). On the other hand,
compost is usually far more useful than gm/ccs for farmers planting irrigated vegetables, since the cost of the compost is well-remunerated and the opportunity cost of
the land used in planting gm/ccs is very high (Photos 9 and 10).
See, for instance, Colombo, Charles, Use of Leguminous Plants in Tropical Countries as Green Manure, as
Cover and as Shade (Rome: Printing Office of the Chamber of the Deputies, 1936).
1.Honduras. Farmers constructing rock walls

to prevent more erosion and to make sure the
organic matter they apply stays in the field.
2.Honduras. This photo was taken on the same

field one year later, now with bean and maize
crops.The maize is only fair, but the beans are
growing very well.
3.Honduras. This field had been abandoned

because it would not produce a decent crop of
maize.
4.Honduras. One year later in the same field as
#3. The maize crop at the top produced close to
4 MT/ha.The bean crop at the bottom produced
better than the average bean crop in the United
States that year.The treatment consisted only of
soil conservation techniques (contour ditches with
Napiergrass hedgerows) and organic matter.
5.Honduras. The contrast between the upper half

of the field and the lower half consisted only of two
years of mucuna that were intercropped with maize
in the upper half of the field and then incorporated
into the soil.
6.Honduras. In the same field as #5, it can easily

be seen that the harvest in the upper half of the
field will be at least four times the harvest in the
control plot below.This increase is somewhat better
than average.Usually, the harvest after two years
of intercropping mucuna will only be about double
that of a control plot.
7. Honduras. Animal manure is a valuable

resource, and farmers should always be
encouraged to use it if the transportation costs
are not too high. Nevertheless, smallholder farmers
almost never have enough easily-collected manure
to maintain the fertility of their fields. All the manure
of approximately 8 to 10 well-fed animals would
be needed to maintain the fertility of one hectare
of cropland.
9. Honduras. The only difference between the

maize on the left and the row of foot-tall maize on
the right is the turning into the soil of a very large
amount of compost in the left-hand field. This
farmer experiment in the early 1980s shows the
potential of organic matter in improving soils. We
were not aware of this experiment by Conrado
Zavala of Guinope until the day this photo was
taken. The amount of compost used was far more
expensive in labor than the increased harvest, but
the potential of organic matter was nevertheless
demonstratedand has been proven true in country
after country. What this (and many later similar
experiments in other climates and soils) taught
us was that dramatic soil improvement was not a
problem of soil chemistry (except where the soil is
salty, is chemically polluted or has a manganese
deficiency), but rather of economics: How can you
get enough organic matter into the soil at a price
that smallholder farmers can afford? After seeing
this experiment, we began looking for cheaper ways
of getting organic matter into the soil, an effort that
quickly led us to gm/ccs as by far the cheapest
technology for recuperating soils.
8. Guatemala. The transport of composting

materials, management of the pile, transport to
the field and spreading across the field involve so
much labor that compost is also usually unprofitable if used on traditional basic grains (except rice)
and root crops.
10. Ecuador. We do not yet have gm/cc systems

for use with most low-lying vegetables. This photo
shows an experiment in which a vicia is being
intercropped with cabbages. Usually in irrigated
vegetable fields, animal manure, compost or
chemical fertilizer will be economically more advantageous than most gm/ccs because the opportunity
cost of growing a gm/cc on such valuable land is
very high.
11. Honduras. Gm/cc species should produce

large amounts of biomass quickly. This photo
shows the growth of mucuna on a small trellis only
45 days after planting
Existing Green Manure/Cover Cropping Systems Around The World
3. Existing Green Manure/Cover Cropping Systems

Around The World
Advantages of Green Manure/Cover Cropping Systems
The proven advantages of gm/ccs are numerous:
1. Increased organic matter and soil nutrients (Photo 9). Gm/ccs are capable of
adding as much as 50 metric tons/hectare (MT/ha) or more of organic matter (green
weight) to the soil each year (Photo 11).3 This organic matter has various positive
effects on the soil, such as recycling nutrients back into the soil, pumping nutrients up
to the soil surface, and improving the soils water-holding capacity. It can also increase
the total amount of nutrients in the soil, improve its nutrient balance, increase the
number of macro and microorganisms (very small animals in the soil, many of which
also help a farmers crops grow better), make soil softer and easier to plow, improve the
acidity of soil (i.e.: buffer soil pH) and sequester carbon. Organic matter makes soil
nutrients, including those supplied by chemical fertilizers, more accessible to crops,
though this impact has been greatly underestimated. In the case of phosphorus, this
is particularly important: in acidic soils, phosphorus may become four to five times
more available to plants when surrounded by organic matter.
2. Nitrogen fixation. Legumes (plants that
produce their seeds inside pods) are able
to fix nitrogen (N) from the atmosphere
into a plant-usable form that accumulates
in plant tissues (Photo 12). Legumes can
thereby add large quantities of nitrogen
to farmers soils. Most of the widely used
legume gm/ccs are capable of producing
more than 50 kg N/ha, while a few gm/cc
species fix significantly more. Mucuna can
fix 140 kg N/ha, the jackbean up to 240
kg N/ha and tarwi (Lupinus mutabilis),
fava beans (Vicia faba) and Sesbania rostrata up to 400 kg N/ha.4
12. Honduras. When most of us think of nitrogenfixing nodules, we think of tiny little spheres about
one millimetre in diameter. This photo shows the
nodules of just one mucuna plant grown under
ideal conditions with no competition from other
plants.
Even allowing for considerable losses of

nitrogen to the air (volatilization), the 140
kg N/ha added by the gm/cc would cost at
least US$75/ha if purchased as a chemical
fertilizer. This increase in both nitrogen
Lathwell, Douglas J., Legume Green Manures, Principles for Management Based on Recent Research in TropSoils
Bulletin No. 90-01 (Raleigh, North Carolina: Soil Management Collaborative Research Support Program, North Carolina
State University, June 1990).
4
National Association of Sciences (NAS) Tropical Legumes, Resources for the Future (Washington, D. C.: NAS, 1979) and
Tisdale, Samuel L., et al., Soil Fertility and Fertilizers, Fifth edition (New York: MacMillan Publishing Co., 1993).
3
and organic matter from gm/ccs can

increase soil fertility tremendously. As a
result, programs from India to Brazil and
throughout Central America now speak
not just of soil conservation, but of soil
restoration and soil recuperation.
3. Zero transportation costs. Unlike most
other materials that improve the soil, these
additions of organic matter and nitrogen
have no associated transportation costs.
They are produced right in the field where
they will be used, and are already welldistributed across the field.
13. Honduras. Gm/ccs that can effectively control

weeds, such as the mucuna in this photo, will be
greatly appreciated by farmers.
4. Very low cost. Gm/ccs cost the farmer

no money whatsoever once the farmer has
purchased his or her first handful of seed.
5. Weed control. Gm/ccs can also be an

important factor in reducing the cost and
the labor required for controlling weeds
(Photo 13). Especially in Africa, this often means that womens total workloads can be
reduced by 20% or more.
In Africa, a particularly noxious weed is striga (Striga hermonthica). Striga can significantly reduce yields of maize, sorghum and millet, and is a major concern for farmers
in areas with low soil fertility. The solution, however, is quite simple. Striga dies out if
enough organic matter is added to the soil. Thus almost any gm/cc that produces a
decent amount of biomass will gradually rid the fields of this weed.5 In areas where
striga is a serious problem, this fact in itself will often convince farmers that they
should use gm/ccs.
Imperata grass (Imperata cylindrica) is another serious weed that grows best in poor
soils and therefore becomes less and less problematic as soils recuperate.
6. Reduction in the use of agrochemicals. If farmers are using chemical fertilizer,
gm/ccs can usually reduce the amount used by 60% to 80% without lowering yields.
Herbicide use is either reduced or eliminated, since many gm/cc species are able to
smother weeds. Some species of gm/cc can be used in place of other chemicals. For
example, mucuna and lablab beans kill nematodes, while sunnhemp (Crotalaria
ochroleuca) can be used to control pests that eat stored grain.
7. Soil cover. The soil cover provided by many gm/ccs can be very important for soil
conservation (Photo 13). In general, the value of soil cover (especially in the humid
Dyck, E., et al., Legumes in Every Farm, Strategies for Adoption of Soil-Improving Legumes by Smallholders in Kenya, 1996.
10
tropics) has been greatly underestimated. High rainfall and substantial slopes on
agricultural land create ideal conditions for erosion to occur. Rain drops hitting bare
ground cause soil particles to be dislodged. These are transported downhill in water
flow, and pick up more soil as momentum builds. But covered soil is less vulnerable
to water erosion. One study showed that farmers cultivating maize and mucuna on
hillsides with a 35% slope and more than 2,000 millimetres of annual rainfall in
northern Honduras are actually increasing the productivity of their soil year by year,
without using any other soil conservation practices. Their soils are totally protected
from erosion because they are covered by mucuna 10 months of the year.6
8. Improved soil moisture. The soil cover, or mulch, that is provided by a gm/cc also
greatly improves drought resistance. The gm/cc residues add organic matter to the soil,
which increases infiltration of water into the soil and increases the water-holding
capacity of the soil. In one experiment carried out during a drought in southern
Honduras, maize fertilized with chemical fertilizer died one month into the drought,
maize fertilized with animal manure died about two months later, and maize fertilized
with jackbean still managed to produce a rather small harvest.
9. Zero tillage. The experience of hundreds of thousands of farmers in Brazil,
Paraguay, Argentina and Honduras shows us that after two to four years of heavy
applications of organic matter from gm/ccs (over 50 MT/ha, green weight), farmers
can move to zero-till systems that retain very high levels of productivity. A tremendous
amount of time and effort is saved when farmers no longer need to plow their soil.
Using mucuna, zero tillage and no chemical fertilizer, smallholder farmers working on
hillsides in northern Honduras have maintained yields of maize of over 2.5 MT/ha for
over 30 years. They achieve maize yields of 4 MT/ha with very small additional applications of urea. Their cost of producing each sack of maize is about 30% less than that
of the richer farmers nearby who use tractors, chemical fertilizers and herbicides, and
have flat land.7
In Brazil, farmers who use gm/ccs combined with rotations and medium applications
of chemical fertilizer regularly harvest 7 to 8 MT/ha of maize without having tilled the
soil in over 10 years.8
10. Competitiveness with farmers using tractors. Gm/ccs are very effective in smothering weeds. They can also allow farmers to move to zero tillage. Since the primary
method used in conventional agriculture for achieving both weed control and soil
preparation is tillage, farmers who use tractors have traditionally had a huge advantage
over smallholder farmers who cannot afford tractors. Now, by using gm/ccs, smallholder farmers using animal traction or hoes and sometimes even those working on
Buckles, Daniel, et al., Cultivos de Cobertura en Agricultura de Laderas, Innovacin Campesina con Mucuna (Ottawa:
Centro Internacional para el Mejoramiento de Maz y Trigo and Centro Internacional de Investigaciones para el
Desarrollo, 1998).
7
Flores, Milton and Nicolas Estrada, Estudio de Caso, La Utilizacion de Frijol Abono (Mucura spp.) Como Alternativa
Viable para el Sostenimiento Productivo de Sistemas Agricolas del Litoral Atlantico. Paper presented to the Center for
Development Studies at the Free University of Amsterdam, 1992.
8
Bunch, Roland, El Trabajo de EPAGRI en el Estado de Santa Catarina, Brasil, Nuevas Posibilidades Importantes para
Agricultores de Escasos Recursos. No date, unpublished.
6
hillsides can compete with much wealthier

farmers. After all, eliminating these operations is more efficient than merely making
them cheaper by using a tractor. In an
age of falling trade barriers, this fact alone
could justify the use of gm/ccs among the
worlds smallholder farmers.
14. Peru. This wasteland in highland Peru is being

rehabilitated with tarwi. About 1/5th the amount of
manure is needed to rehabilitate the land compared
to only using manure. Basically, the manure just
helps the tarwi reach maximum yield.
15. Honduras. The dramatic increase in soil

organic matter and the addition of nitrogen make
gm/ccs ideal for recuperating wastelands. Only in
two cases (central Cambodia and Rwanda) have I
ever seen lowland soil so degraded that jackbeans
would not grow. In such cases, a light application
of manure will usually allow the jackbean to grow,
and the soil will be ready for cropping the following
year.
11. Provision of additional benefits. In

addition to improving the soil and controlling weeds, gm/ccs can provide many
other benefits:

high-protein human food;

commercial products to be sold,
thereby improving farmers incomes;
wasteland restorationof both
wastelands with extremely low fertility
and those invaded by very bad weeds,
such as imperata grass (Photos 14
and 15);
shade for other crops;
soil moisture conservationeither as
green or dead mulches;
high-quality fodder and bedding
for cattle;
prevention of plant diseasesfrom
sun scorch on citrus fruits to root rots
in vegetables;
reduction of the incidence of pests,
including nematodes in growing crops
and various pests in stored grain;
contour bunds;
firewood.
Of course, when we use gm/ccs as fodder

or fuel, we reduce the quantity of organic
matter and nutrients that are added to the
soil, thereby making them somewhat less
valuable in improving the soil.
12. The capture of additional nutrients. In places such as the Sahel in Africa, wind
erosion is a major problem. Low-lying bushes that grow year-round, such as
Piliostigma reticulatum, can capture large quantities of the soil that is blowing in
the wind, adding important nutrients to farmers fields.
11
12
13. The ecological advantages of trees. When we use tree gm/ccs as dispersed shade,
we also gain the ecological advantages provided by trees: soil and crops are protected
from the tropical sun, soil moisture is conserved (because evaporation and transpiration rates are reduced), and desertification is curbed.
When we compare the above advantages to those of composting, we find that in the
vast majority of cases, gm/ccs will be far more attractive to farmers than composting.
Gm/ccs provide weed control, food, and many other products and advantages, with
much lower transport and labor costs than compost (Photo 8). The exception would
be those cases where farmers are growing very high-value crops and/or own a very
limited amount of land (less than half a hectare, for instance), so that there is virtually
no space where the gm/ccs can grow.
Each of the above advantages should be analyzed and weighed when choosing which
gm/ccs to use and promote. Experience shows that farmers are rarely attracted
primarily by the gm/ccs ability to increase soil fertility. More commonly, farmers are
most motivated by the gm/ccs potential for human consumption (usually the highest
priority of the above advantages) or their ability to control weeds. Therefore, gm/
ccs should be promoted mostly by emphasizing these other advantages, not just the
advantage of increasing soil fertility. For instance, sometimes we refer to a gm/cc as a
food crop or a green herbicide.
Disadvantages of Green Manure/Cover Crops

Despite all these advantages, gm/ccs are sometimes difficult to introduce to farmers.
While working with the international development organization World Neighbors
during the 1980s, I tried for eight years to introduce the use of mucuna to farmers in
central Honduras, with very little success. This was because I didnt understand some
of the important disadvantages of gm/ccs. If we fail to overcome these disadvantages,
we will not be successful at introducing gm/ccs. In a few situations, these disadvantages will mean we should not try to introduce gm/ccs at all.
1. The opportunity cost of the land. Farmers normally will not plant something that
only fertilizes their soil if the land could instead be planted with either food crops or
cash crops. Unless the gm/ccs also produce food, the land used to grow gm/ccs must
have no other valuable use (i.e.: it must have no opportunity cost, meaning the
cost represented by not taking advantage of an opportunity to do something else
with the land). Traditional green manure systems from temperate countries require
farmers to monocrop green manures on their land before planting their regular crop,
thus taking otherwise useful land out of production for a few months in order to produce the green manure. These systems did not recognize the need to use land that had
no other profitable usewhich is probably why the introduction of green manures
into the developing world before the 1970s was rarely successful.
2. The slow results. Soil improvement is a long-term process that may not be immediately noticeable to the farmer. Usually, significant improvement in productivity does
not occur until after the first crop of gm/cc has been applied to the soil, which means
that concrete, visible results are not apparent until well into the second cropping cycle.
This slow appearance of resultsimproved soilsthat are often difficult for people to
believe, further complicates the adoption of gm/ccs. Once again, it is often preferable
to promote gm/ccs for some reason other than soil fertility. If farmers are not aware
of the value of organic matter in their soils, a heavy application of animal manure
on a small plot of land the first year will help make farmers aware of the value of the
organic matter the gm/ccs are producing in their fields (Photo 7). Simple demonstrations, such as showing how organic matter increases the water-holding capacity of soil,
can also be used.
3. Dry season problems. Often gm/ccs must produce their organic matter at the end
of the wet season, or must continue to grow during the dry season. Grazing animals,
wild animals, termites, agricultural burning, bush fires or several other problems may
destroy organic matter or growing plants before the farmer can use them the following
rainy season. In very hot climates and on soils with no shade, the nitrogen and much
of the organic matter will be burned off by the tropical sun. Thus, almost no benefit
from the gm/cc will be available for the next crop.
4. Difficult growing conditions. Smallholder farmers in the tropics commonly cope
with many challenging conditions, including extremely low or irregular rainfall,
extremes in soil pH, severe drainage problems, or a combination of these. Such conditions will reduce the growth of gm/ccs, thereby reducing or destroying their impact.
Through the years, we have learned how to overcome an increasing number of such
problems. The solution is often to use gm/cc species that are particularly resistant to
certain problems. For example, jackbean can withstand very poor soils and is often
used for recuperating wastelands. However, these solutions are often achieved by using
gm/cc species that produce less organic matter, dont fix as much nitrogen, dont have
additional benefits, or dont fit as well into the local farming system.
5. Timing (also called synchronization). The nutrients provided by the gm/ccs,
especially nitrogen, must be available to crops when they need them in order to raise
productivity. Gm/ccs will boost farmers productivity only if the gm/ccs nutrients are
available to the crops at the right time. In many gm/cc systems, the correct timing is
either impossible or very difficult to achieve. Therefore, the efficiency of the systems is
reduced.
Very often, this problem can be solved using natural foliar nutrient sprays at the
appropriate time of year to supplement soil nutrients. Solutions of cattle urine or
crushed mother of cacao (Gliricidia sepium) leaves are often used. Very small amounts
of chemical fertilizer can have the same effect. These remedies can be very useful in
supplying nutrients at exactly the time the crops are likely to run out of the nutrients
supplied by gm/ccs.
13
14
Green Manure/Cover Crops Systems are Widespread

Gm/cc systems are surprisingly common around the world. This fact is understandable if we take into account all the advantages listed above. During the 1990s, the
NGOs CIDICCO (International Center for Dissemination of Cover Crops) and
COSECHA (Consultants for a Sustainable, Ecological and People-Centered Agriculture) in Central America each compiled a list of known gm/cc systems in about 45
countries. The lists were admittedly incomplete. One list included 96 gm/cc systems
and the other 51 systems.9 Statistically, if we assume that the definitions of gm/cc
systems in the two lists are identical and that each list was a random sampling of the
worlds existing systems, a simple mathematical extrapolation indicates that over 500
gm/cc systems probably exist around the developing world.
Time and time again, in more than 30
developing countries, I have found that
gm/cc systems are thriving where local
agronomists were absolutely sure they did
not exist. For instance, Indonesian agronomists assured me that because Indonesians
do not eat beans, gm/cc systems would
not exist in their country. However, on
my first day out in the field in Indonesia,
we ate lunch at a small roadside restaurant
16. Indonesia. Our lunch at a roadside restaurant,
(Photo 16). The lunch included three
which included three gm/cc species, in a country
in which many agronomists had assured me no
legumes, all of which we later found
gm/ccs were being used.
farmers using as gm/ccs. None of these
leguminous foods were actually cooked as
beans, per se: the soybeans were served as
tofu, which was in the soup along with green cowpea pods, and the mungbeans had
been made into some delicious sweet cakes. In the following four days, I observed six
different gm/cc systems in just a very small area of southern Sumatra.
Characteristics of Known Green Manure/Cover Crop Systems

Of the known gm/cc systems that have been implemented by farmers around the
world for at least five years without any subsidies, over 70% were basically developed
by smallholder farmers themselves. This fact, perhaps more than anything else, shows
how appropriate these systems can be for smallholder farmers, and how interested
smallholder farmers are in finding and adopting alternatives to both fallowing and
chemical fertilizers. (Data on adoption and abandonment of these systems would be
a much better indicator of farmer interest, but such figures are not available in most
cases.) It is also surprising that over 100 species of gm/ccs are used, with new ones
being discovered all the time.
Bunch, Roland and Milton Flores, Unpublished lists.
17. Haiti. Using expensive chemical fertilizers

is not economically feasible for most smallholder
farmers growing basic subsistence crops (except in
the case of rice). Even when it is feasible, a normal
application of unsubsidized chemical fertilizer will
almost always cost the farmer over US$100 for a
half-hectare, a tremendous amount of money to
take out of a farmers pocket when similar yields
can be produced with inexpensive green manure/
cover crops.
In semi-arid areas where droughts are

common, farmers often adopt perennial
gm/ccs. This type of gm/cc has much
deeper roots than annuals, allowing them
to withstand the frequent droughts. For
example, in the African Sahel, farmers
tend to use mother of cacao (Gliricidia
sepium), Faidherbia albida and nyama
(Piliostigma reticulatum) when they can,
rather than gm/ccs that only last one year.
Even lablab beans, if used in such climates,
should usually be cut off at ground level
and allowed to sprout again, rather than
being planted from seeds again. This way
the plants can take advantage of the root
growth achieved during previous years.
Why Not Just Use Chemical

Fertilizer?
Quite simply, chemical fertilizer has

become unprofitable for most smallholder
farmers who produce basic subsistence
crops. The situation of each farmer is difThis farmer will have to travel more than 6 hours
ferent, but in most cases the costs related
from the roads end to apply this fertilizer to his
to the use of chemical fertilizers would be
fields. Most calculations as to the profitability of
greater than the benefit provided by the
chemical fertilizer do not include the risk factor,
high transportation costs and the cost of neutrafertilizer (Photo 17). A few major exceplizing the acidity caused by the fertilizer. When
tions to this rule exist; the most important
these costs are added, chemical fertilizer is
includes those farmers who grow rice. The
almost never a profitable technology for basic
increasing demand for rice in Asia has
grain or root-crop producers.
resulted in a relatively high world price of
rice, making it by far the most profitable
of the worlds basic grains for smallholder farmers to produce. Farmers growing highvalue vegetables and fruits can also make a profit using chemical fertilizer.
But for smallholder farmers in most of the world, especially those who produce maize,
sorghum, millet, beans and root crops, using chemical fertilizer is no longer profitable.
Many scientists still produce calculations that show fertilizers to be profitable. But if
all the costs are taken into account, such as taking a bus to a local town; purchasing
the fertilizer at a small-town retail price; returning on another bus (which will also
charge for the transport of the fertilizer); applying the fertilizer; purchasing inputs to
neutralize the acidification caused by the fertilizer (which can ruin an acidic soil if the
acidity of the fertilizer is not neutralized periodically); paying interest on the money
invested in the fertilizer (or the moneys opportunity cost); andmost important of
allshouldering the risk of crop failure, the chemical fertilizer is usually unprofitable.
15
16
This is particularly true in drought-prone areas, where the risk of crop failure is high.
As a result, farmers in many parts of the developing world (except where chemical
fertilizer is being subsidized) are already buying less and less chemical fertilizer or none
at all.
But lets suppose that chemical fertilizer still could be used profitably. Most smallholder farmers in the world now possess less than a hectare of land per family. Even a
light application of chemical fertilizer on that hectare of land will cost about US$200.
It is virtually impossible for a smallholder farmer to feed his or her family on what is
left from the harvest of a hectare of basic grains, even when fertilized, if he or she has
to sell enough of those grains to pay for the fertilizer. By contrast, gm/ccs, which can
improve the soils fertility just as well or better than chemical fertilizers, cost a fraction of what chemical fertilizers cost (or can be a free by-product of the production
of high-protein grain legumes). Therefore, farmers can feed their families with high
quality food on much smaller pieces of land. This is imperative if most of the worlds
smallholder farmers are to achieve anything approaching food security.
While it is true that highly efficient gm/cc systems do not exist for absolutely every
farming situation, this manual shows that systems do exist for the vast majority of
smallholder farmers around the world.
Improving Soils: The Basic Rules
4. Improving Soils: The Basic Rules

Books on agriculture seldom mention how much organic matter we need on each
hectare of soil to stop losing fertility or to actually improve the soils fertility. This
question has not been researched to any great extent; yet, it is of extreme importance
in recovering, maintaining or improving soil fertility.
Long-term experience with a series of different species of gm/ccs around the world
gives us a very approximate answer. Of the more than 100 gm/cc cropping systems
I have observed, those that can keep soil fertility at about the same level over many
years are those that supply somewhere around 20 to 25 MT/ha of leguminous organic
matter (green weight) to the soil. These would include maize/runner bean systems at
high altitudes and maize/lablab bean systems at lower elevations. Systems that supply
30 MT/ha or more of organic matter to the soil, such as the maize/mucuna (Mucuna
spp.) system of northern Honduras, northern Guatemala and eastern Mexico, will
gradually increase soil fertility over the long term, until a high level of productivity is
reached.
In many cases, several soil fertility practices are being used in the same field at once:
several different gm/ccs are grown, or gm/ccs are used along with manure. In these
cases, the weight of the various kinds of organic matter can be added together.
For many people, 20 to 25 MT/ha seems like a surprisingly large amount of organic
matter. Those who have worked for years with compost will immediately realize that
virtually no smallholder farmers are using enough compost to come anywhere near
this amount of organic matter. Although a kilogram of compost may be a little more
beneficial to the soil than a kilogram of leaves, most smallholder farmers observe
little or no difference between the impact of a kilogram of compost as compared to
a kilogram of gm/cc plant material. Thus, compost will almost never help a farmer
maintain his or her soil fertility without the added contribution of gm/ccs (or chemical fertilizer).
A second very important factor in maintaining soil fertility without the use of chemical fertilizers is the heat in the tropical lowlands. Many experiments with gm/ccs have
failed because the gm/ccs (such as mucuna) produced their organic matter primarily at
the end of the wet season. The organic matter then lays on the ground for five to seven
months of sweltering heat and often with high winds, until the next rainy season.
During this time, the intense heat basically burns up all the nitrogen and much of the
organic matter produced. By planting time at the beginning of the next wet season,
the gm/ccs impact on soil fertility is almost all lost. (Animals grazing the gm/cc by
roaming free during the dry season will, of course, cause the same loss of nitrogen and
organic matter.) As a rule, the calculations of total organic matter applied to a piece of
land in the lowland tropics should not include any organic matter that has lain on top
of the soil throughout the dry season, without any shade.
The loss of both nitrogen and organic matter is one of the most difficult problems to
overcome in using gm/ccs in the lowland tropics. Nevertheless, there are ways to do it.
17
18
Improving Soils: The Basic Rules
First of all, legumes such as the jackbean (Canavalia ensiformis), lablab bean, tephrosia
and most leguminous trees can grow through most or all of the dry season, thereby
providing fresh organic matter close to the beginning of the rainy season. If animals
are roaming wild, they will destroy the lablab beans. If they get extremely hungry,
they may also eat the jackbeans. In this case, the tephrosia or leguminous trees may
have to be used.
To overcome these challenges, a combination of two or three gm/ccs may provide the
best results. For example, using trees and another type of low growing gm/cc could
be a practical answer to this problem. By reducing the ambient temperature at least
10C, dispersed trees can cool the fields enough so that the gm/ccs organic matter
will not be burned off, and the soils fertility can be maintained. The temperature can
be lowered even more if the trees are not pruned until the months right before the
next rainy season, as is normally the case. Thus, dispersed shade can largely eliminate
the problem of dry season burn-off of the gm/ccs organic matter and nitrogen.
Choosing The Right GM/CC Cropping System For A Specific Area
5. Choosing the Right Green Manure/Cover Crop

System for a Specific Area
What about Farmer Participation?
Many people working in agricultural development strongly believe that farmers should
participate in their own agricultural development. I would go one step further than
that. I believe that farmers should become the protagonists, the architects or, as Paulo
Freire wrote, the authors of their own development. In fact, if the smallholder farmers
are not in charge of the process, at least after the first few years, I would question
whether what is happening is really development.
The achievement of broad-based authentic farmer participation is not easy. Serious
questions need to be answered about participation. Are we talking about participation of the powerful, or of everyone? And if a few of the powerful can manipulate a
group, does participation exist, even if everyone is giving his or her opinion? Is true
participation occurring if people are giving us the answers they think we want to hear,
or the answers other development people have told them are the right answers? Is it
participation if the men do all the talking, or women also talk but are afraid to
contradict what the men have said? Can we assume that the farmers know all the
technological possibilities in a given situation? If not, of how much value is their
participation?
For 40 years, development organizations have been told they are not using farmer
participation as much as they should, yet seldom have the above questions been asked.
When they are asked seriously, they can have a tremendous impact on what we call
participation.
These questions about participation should be in the back of peoples minds as they
work with the decision tree in this book. Most questions in the decision tree should
be discussed with the farmers, and most of the answers should be provided by the
farmers, in an atmosphere of complete freedom, trust and confidence.
Furthermore, the gm/cc systems recommended by the decision tree are ones that have
been tried and, at least for some farmers, have been successful over a period of several
years. Still, they should not be considered the only possibilities that might be used in
each situation. It is quite possible, that over time, smallholder farmers will find better
gm/cc species or better ways of using them. I am sure that dozens of good gm/cc species are yet to be discovered, and perhaps hundreds of as-yet untried systems will also
be discovered. The world is in serious need of more, and better, gm/cc systems. Farmers and programs should always be looking for more gm/cc species and more ways to
use them, just as farmers are always looking for more crops to grow or better ways of
growing them. This decision tree should not be allowed to reduce peoples creativity.
We must allow creativity to blossom.
19
20
But programs should also not underestimate the difficulties involved in finding
successful gm/cc systems. Hundreds of programs have tried, and many have failed.
Some people will inevitably respond, But why should we worry about finding
solutions to the smallholder farmers problems? They already know everything they
need to know. Dont they have prodigious amounts of indigenous technical knowledge
(ITK)? Yes they definitely do. But if they know the best solutions to their problems,
why havent they solved their problems already? Are we to believe that, knowing the
most appropriate solutions, they have failed to implement them? I dont believe this,
in part because it would require us to believe the farmers are not very smartand that
is definitely not the case.
Farmers have spent hundreds of years learning about technologies and cropping
systems they needed at different times. But during all those centuries, they were
able to use fallowing, so there was little need to learn much about gm/ccs. Only in
approximately the last 20 years has population growth restrained their ability to
maintain their soil fertility through fallowing. Therefore, it has only been in the last
20 years that farmers have had a crucial need to learn about gm/ccs. Twenty years
is not enough time for smallholder farmers to learn all elements of such a complex
subject through their systems of informal experimentation.
Anyone who feels the smallholder farmers already know everything they need to
know should look through the list of 91 gm/cc systems in this book. I think they will
find that no smallholder farmers (except perhaps in Brazil) know about more than
a handful of these systems, and yet many of these systems might be of use to them.
There will be, in most cases, information here that the smallholder farmer themselves
will admit they dont know, and wish they had known before now.
So what should be our role as people who wish to promote true farmer-protagonist,
farmer-led agricultural development? And how can a guide such as this support that
role? Certainly we should be well-informed, we should listen carefully to the farmers,
we should value their experience and we should be very slow to ever discount their
knowledge and priorities. We also need to learn how technologies that smallholder
farmers themselves have selected can best be adapted to local needs and most effectively communicated and/or promoted.
Then we need to train farmers in the relevant cropping systems mentioned below,
while respecting the smallholder farmers priorities, their knowledge and their abilities. The training process needs to become more and more a dialogue in which both
farmers and outsiders provide information and learn from each other. Eventually, by
experimenting to obtain more knowledge on their own, by learning to train other
farmers, and by having access to this decision tree in a language they can understand,
the farmer leaders should take over the development process themselves. To learn
more about these issues of agricultural extension processes, I would suggest my book,
Two Ears of Corn: A Guide to People-Centered Agricultural Improvement.10
Bunch, op. cit.
10
How to Achieve the Adoption of Green Manure/Cover Crop Systems

A great deal of discussion centers on the levels of adoption or abandonment of gm/
cc systems around the world. In general, evidence indicates that many traditional
gm/cc systems are gradually being abandoned as Green Revolution technology (for
example, using chemical pesticides and genetically modified organisms) has spread, as
traditional foods have become unfashionable, as chemical fertilizer has become more
widely available, and as professional extension workers have criticized farmers for
doing things that arent recommended in university textbooks. This gradual process of
abandonment has apparently affected such widespread systems as the maize/cowpea
and maize/rice bean systems that once existed from Mexico through Nicaragua, and
the scarlet runner bean/maize system that once stretched from the northern U.S. to
Chile.
At the same time, some other gm/cc systems have spread widely and quickly right up
to the present time. One of them is the maize/mucuna system that has spread spontaneously through Mexico, Guatemala, Belize and Honduras over the last 60 years. So
the question persists: will gm/cc systems become sustainable, or will they, after much
blood, sweat and tears, become just another footnote of history?
Quite frankly, a soil fertility crisis is approaching. It will affect the whole world, but
the lowland, drought-prone areas of Africa will be the hardest hit. Population growth
has drastically diminished the amount of land per family. Approximately 80% of small
holder farmers in Africa have less than two hectares of land. Therefore, most people
today have to farm every inch of what little land they have in order to survive; there is
nothing left for fallowing. Thus, for the first time in history, most of Africas lowland
farmers can no longer fallow their soils. In addition to this unprecedented problem,
farmers animals have decreased in numbers (because grazing lands have also been
swallowed up by cultivated farms), chemical fertilizers have increased dramatically in
price because we have used up all the worlds cheap fuels, and global warming is also
taking its toll. Because of this perfect storm of factors, harvests in much of lowland,
drought-prone Africa are dropping by 10 to 15% a year. Within the next five years, we
could witness widespread hunger across much of Africa that will be among the worst
in recent history, with the livelihoods and productive capacity of tens of millions of
smallholder farmers severely damaged.
Most of the smallholder farmers of lowland Africa know something is terribly wrong.
In hundreds of villages that I recently visited, the people are intensely worried. Their
food supply is diminishing at a rapid rate, and they know of no way to solve the
problem.
For years, a large majority of African farmers really werent very interested in gm/ccs.
But now, I can honestly say that I have never seen farmers as interested in gm/ccs as
are the farmers today in most of the Sahel and lowland southern Africa. I think agencies working in the Sahel and the rest of lowland Africa will find that the farmers have
much more interest in improving their soils than ever before.
21
22
The same problemsexpensive fertilizer, smaller plots, little or no opportunity to

fallow and decreasing soil fertilityare affecting smallholder farmers elsewhere, even
though the crisis situation seems to be approaching more slowly. Thus many farmers
in Latin America and Asia are also becoming increasingly interested in the benefits of
gm/ccs.
It is highly unlikely that gm/ccs will ever become just a footnote in history. Nevertheless, if we are to make introduced gm/cc systems as sustainable and widely popular
as possible, we need to discover what makes certain specific gm/cc systems so much
more popular than others. What is the difference?
Most of the programs and organizations that have been successful in introducing
sustainable gm/cc systems have used the following guidelines:11
1. The land occupied by non-food-producing gm/ccs must have no opportunity cost.
This means that if the gm/cc does not produce food, it must occupy land that cannot
be used for producing either a cash crop or a subsistence crop during the time the gm/
cc is using it. This sounds like a very difficult, if not impossible, rule to follow, but we
are finding more and more places and times that gm/ccs can be used:
a) If the gm/cc does produce a valued food, it can be grown in any way that fits into
the system like any other equally valuable food crop.
b) If the gm/cc does not produce food, it can be grown intercropped together with
another food or market crop. For instance, jackbeans are intercropped with maize
or cassava, or perennial peanut (Arachis pintoi) is intercropped with coffee. Intercropping is currently the most popular way of growing gm/ccs.
c) The gm/cc can be grown on wasteland or during the first year or two of a fallow as an improved fallow. For instance, farmers in Vietnam broadcast tephrosia
(Tephrosia candida) seeds into their first year fallow. That way, instead of fallowing
the land for five years, as they normally do, the soil will be fertile enough to farm
again in just one or two years. In this case, since the soil is very poor when the
fallow starts and the gm/cc is planted, farmers must use gm/cc species that can grow
well in very poor soils. These include gm/cc species such as jackbeans, tephrosia, or
certain very hardy trees.
d) The gm/cc can be grown mostly during the dry season. This can be done in
three ways. First, the gm/cc can be relayed (intercropped with another crop that has
already been growing for some time) into traditional crops. Examples of this are
the maize/cowpea and maize/lablab bean systems in Thailand, which are S70 and
S71 in the decision tree. The second way is to plant the gm/cc after the harvest of
normal crops, such as in S78, the system in Vietnam in which rice bean is planted
after the rice. The third way is to intercrop the gm/cc with the normal crop and
11
Bunch, Roland, El Uso de Abonos Verdes por Agricultores Campesinos, Lo Que Hemos Aprendido Hasta la Fecha.
Reporte Tcnico No. 3, Segunda Edicin, Tegucigalpa, Honduras: Centro Internacional de Investigacin y Divulgacin de
Cultivos de Cobertura (CIDICCO), 1995.
then leave it growing through the dry season after the traditional crop is harvested.
An example of this method is S7, the maize/sweet clover system in Mexico.
e) The gm/cc can be grown under fruit or forest trees or almost any perennial
crops. In this case, we choose particularly shade-resistant gm/ccs, like jackbeans or
Centrosema pubescens.
f ) Occasionally we can use other ways of growing gm/ccs without affecting the
main crop. These include growing gm/ccs during periods of frost (lupines, such as
tarwi, often do well), in soils that are too acidic for our crops (mucuna or buckwheat), or during very short periods of time (as in the case of Sesbania rostrata).
2. Gm/ccs must not cost money. This rule implies that farmers must be able to produce their own gm/cc seed year after year, and that the gm/ccs must have no disease
or insect problems that are serious enough to significantly diminish their growth.
Insect attacks that do not affect the growth of the plants are fine; they merely help
process the organic matter. If an insect or disease problem does significantly reduce a
gm/ccs growth, we must usually discard that species and start using another one. The
Brazilians, for instance, quit using lablab beans some 20 years ago because it was too
heavily attacked by insects. This rule also means we cannot use inoculants (commercial products that increase the fixation of nitrogen).
3. Gm/ccs must not increase labor costs very much. This rule means that, except
where animal traction or tractors are being used, gm/ccs will have to be left on the soil
surface. It also means that the intercropping of gm/ccs is particularly advantageous,
because the weed control the gm/cc provides when it is intercropped can often save
more labor than the labor needed for planting and cutting down the gm/cc.
The labor problem is also the main reason why farmers appreciate zero till systems.
Farmers will often decide to plant gm/ccs primarily because of the hope that they will
never again have to plow or hoe their fields. Since farmers must maintain high levels
of organic matter in the soil in order to be able to use a zero till system indefinitely,
this desire to eliminate the plow or the hoe is a motivating factor for the long-term use
of gm/ccs.
4. Gm/ccs must fit into the existing farming systems. Gm/ccs will be seen as much
less important than food or cash crops, at least for the first few years. Thus farmers
and extension workers have to adjust the gm/cc to fit into the already-existing cropping system, rather than adjust the farming system to fit some way of growing a gm/
cc.
5. The gm/cc should provide at least one important benefit in addition to improving
the soil. In a worldwide study of gm/cc systems that were taught to farmers through
programs, almost all of the gm/cc systems that persisted after the development organization had left the area were those systems that produced important benefits above
and beyond soil improvement. Thus, whenever possible, we should promote gm/cc
23
24
species that can be eaten, fed to animals, or provide some other benefit for which a
strong felt need exists among the farmers.
In general, good gm/cc species should establish themselves easily and grow rapidly
under local conditions; be able to cover weeds quickly; and be able to fix plenty of
nitrogen. They also should be resistant to insects, diseases, grazing animals, bush fires,
droughts, or any other problem they may have to face within the cropping system. As
well, they should produce enough seed for future plantings. If the gm/ccs are to be
used for intercropping, they should withstand shade and fit in with the growing cycle
of the main crops.
Note that virtually all of the gm/cc systems recommended in this book fall within the
above rules that favour easy adoption by farmers.
Gm/ccs have also become very useful for large-scale farmers who cultivate as much as
100,000 ha of land in Brazil. However, among smallholder farmers, gm/ccs tend to be
most useful for those who have between 0.25 ha to 10 ha of land. Farmers who have
more than 10 ha can still use fallows to maintain the fertility of their soils. It is fairly
difficult, though definitely possible, for gm/ccs to compete with fallowing, because
gm/ccs require more labor. For farmers who have less than 0.25 ha, the use of the land
is usually so intense that there are virtually no times or places when the farmers arent
using the land. Thus, there is much less opportunity to use gm/ccs. In these cases, it is
often better for farmers to make compost or buy organic or chemical fertilizer.
Farmers who dedicate their land to year-round paddy rice production also may not
have any good way to use gm/ccs.
Collecting the Necessary Information

The first step in choosing one or more gm/cc system(s) for a particular area is to get
to know the area well. Some information is absolutely essential before you can choose
the best systems for a given area. This includes:
1. Do the farmers have any preferences as to how they want to solve the soil fertility
problem, and if so, what are they? Have they tried some techniques already?
What were the results? Why?
2. At what elevation above sea level is the area? Zero to 800 metres (m) above sea
level, 800 to 1,500 m, 1,500 to 3,000 m or above 3,000 m?
3. What leguminous grains or leaves do people already know and eat?
4. For what leguminous plants is there a good market?
5. When asked what their major agricultural problems are, do farmers (men and
women) see soil fertility as one of the top two or three? Take care not to bias this
question by former comments, by how the question is asked, or even by such
things as the organizational name or logo on the vehicles in which one arrives in
the village.
6. Are yields of peoples subsistence crops generally increasing or decreasing each
year? By how much?
7. Have people already attempted to use natural means of improving soil fertility?
What were these methods? Did any of them involve plants that fertilize the soil
(that is, gm/ccs)? What do the farmers feel about using plants to fertilize the soil?
Have they ever seen an example where this was done successfully? If so, what
species did they see being used? Do people use animal manure? How much per
hectare? Do they use compost, and if so, how much per hectare?
8. What are the dominant agricultural crops? Are other plants intercropped with
these dominant crops? If so, what percentage of the land dedicated to these crops
is intercropped? Do farmers use a crop rotation? What is the rotation, including
the management and the seasons for each crop?
9. What percentage of farmers in the area still fallow land? For how many years at a
time? Do farmers plant or harvest anything together with the natural vegetation
on their fallowed land?
10. Do grazing animals exist in the area? Are they set free during the agricultural
off-season(s)? Are there any limits on the areas where they roam?
11. What is the average rainfall? During which months does it rain the most?
12. What is the dominant land tenure system in the area? Do farmers own the land
they farm, or can they be sure that they will be able to continue farming the land
they have improved? What is the average size of the plot(s) belonging to each
household? How many households have larger holdings, and how much land do
they have?
13. Do women possess land? Do they have any use rights to land owned by others?
Are these use rights long-term? How do womens priorities for the agricultural
system differ from those of the men? (e.g. Do women want to plant different
crops than men?)
14. What agricultural activities are carried out by women? What rights do they have
related to crop or animal selection and use?
15. Have other agricultural programs worked in the area previously? What practices
did they recommend, and what were the sustainable results of those efforts? What
does the farm family feel about these efforts? Are other organizations working in
the area presently? If so, do they work with gm/ccs, and what are their priorities
among the various possible gm/cc systems?
25
26
These questions should be open-ended, to the extent possible. Farmers should be

allowed to digress, bring up issues they want to mention, ask questions and express
their opinions. As development agency personnel come closer to selecting a particular
gm/cc technology to try, they need to consult farmers regarding how they feel about it.
In addition to these interviews, make careful observations. Often it is important to
verify information by observation, since interviews alone can lead to wrong or incomplete information. Observe things like average yields, the prevalence of intercropping,
and mens and womens roles in agriculture. Other factors that can and should be
observed include the following:
1. Is nitrogen a major limiting factor in the soil? (If the leaves of maize or sorghum
all turn yellow, nitrogen is almost surely the limiting factor.)
2. Do different fields vary widely in soil fertility? Why?
3.
Are there large areas of fallowed land? Wasteland? Forest? How old are the trees
in most of the fallowed land?
4. Is the agriculture quite intensive? For instance, are the crops well weeded? Is every
scrap of land in use? Are technologies appropriate to intensive systems being used,
such as terracing, contour ditches or barriers, diguettes, zai holes, etc.?
5. What kinds of erosion are problematic? Is the land mostly hilly or steep?
6. Is there a great variation in how the farmers of the area are using their land?
7. Is most of the organic matter in the fields consumed by animals? Is it burned?
8. Are farmers applying large amounts of animal manure, compost and/or sweepings
from their compounds to their soil? Are they using chemical fertilizers? To which
crops are these fertilizers applied? In what amounts?
Using The Decision Tree
6. Using The Decision Tree

How to Use the Decision Tree
The following decision tree should not be seen as a straitjacket. It should be used to
get new ideas, but not to restrict what you can do. The options given may be expanded on. Other species can be tried. That said, the gm/cc systems listed in this decision
tree are definitely worth a try. Nearly all of them are proven systems that hundreds of
farmers have used for years, usually in the absence of any outside program. In many
cases, thousands of farmers have used them for centuries.
Green Manure/Cover Crop Systems Included in the Decision Tree

The gm/cc systems included in this decision tree are almost all systems that either I or
Gabino Lopez, a Guatemalan colleague, have observed during our more than 25 years
of observing gm/cc systems around the world. Thus we know the systems, have a feel
for the situations in which they have prospered, and have been able to interview farmers using these systems. The list includes systems that are used in about 25 developing countries. Obviously, there must be many systems that we have not observed in
these countries, and additional systems most likely exist in approximately another 60
developing countries.
This list includes only gm/cc systems that use legumes. It does not include nonleguminous gm/ccs such as the oats, forage turnip (Raphanus sativa) and sunflowers
often used in South America, the fonio used in West Africa or the three-year cassava
used in Vietnam. It also does not include the vast majority of gm/cc systems based on
tree species in agro-forestry systems, which are also extremely common around the
developing world.12 In most cases, the number of farmers mentioned is only a rough
estimate, since it has often been 10 to 20 years since Gabino or I visited these systems.
The gm/cc systems are distinguished from each other by the species of gm/cc that
is used, by the cash or subsistence crop, or by the date of planting or management
of each system. When essentially the same gm/cc system is used in more than one
country or continent, I will describe the system only once and list the other places it is
practiced. In many cases, especially when gm/ccs are used in rotations or to recuperate wastelands, they are associated with any of a number of traditional crops. In these
cases, each gm/cc species is counted as one system.
Characteristics of the Green Manure/Cover Crop Systems Recommended

As mentioned, the gm/cc systems included in this decision tree are those that have
proven themselves most successful among smallholder farmers, according to two
criteria. The first is that their use has spread fairly widely. A particularly dramatic case
would be that of S1, the maize/mucuna system in Mexico and Central America,
12
Cairns, Malcolm, ed., Voices from the Forest, Integrating Indigenous Knowledge into Sustainable Upland Farming (Washington, D.C.: Resources for the Future, 2007).
27
28
which has spread almost continuously from north of Veracruz in Mexico, through
Guatemala and parts of Belize, to half way across the northern coast of Honduras, in
less than 60 years.
The second criterion is that adoption of the system has lasted a long time, as exemplified by the case of S13, the maize/runner bean (Phaseolus coccineus) system. This
system has lasted from long before Columbus arrived in the Americas to the present
day. At one time, this system was utilized in both temperate and highland tropical
regions from New York State in the U.S. (where it is known as the Seneca bean),
through Mexico (ayocote), Guatemala (piloy) and Honduras (chinapopo) to Colombia
(frijol cacha), Ecuador, Peru and Bolivia.
I have not included a few long-term, widely used systems that fall into one of two
categories. First are some of those systems that use thorny gm/cc species. Anyone who
has seen Dogon farmers in Mali trying to bury troublesome spines in their field, or
seen a childs foot that was badly infected because it had been punctured by a thorn
from a local bush, will understand why I have not included these.
The second group of gm/cc species not included here is that of plants, such as the
kudzus (Pueraria spp.) or perennial soybeans (Glycine wightii) that have become, in
one environment or another, serious invasive species. Invasive species are species
introduced to a non-native environment, where they act like weeds or otherwise
negatively affect the local ecosystem. These plants might sometimes serve as very good
gm/ccs, but the danger that they will spread and negatively affect tens of thousands of
smallholder farmers is just not worth the risk.
In both of these cases (the thorny and the invasive species), we usually have other
species that can serve just as well, or nearly so. The comparative advantage that the
former species might provide is not worth the risk of their becoming a problem.
Other systems included in the list of gm/cc systems are not recommended by the
decision tree. These systems can add to your knowledge of the potential variety of gm/
cc systems, but I consider them to be a distant second- or third-best system for any
particular situation. They are mentioned so you can read about and consider them,
and even visit them if they happen to exist near where you work, but they are notto
my way of thinkingthe best possible solution to any particular situation.
In places where soils are so poor that most of the gm/cc species listed below will not
grow well, we have to start by using a gm/cc species that is more resistant to poor
soils, such as the jackbean (Canavalia ensiformis) or tephrosia (Tephrosia spp.). After
one to three years, these species will have improved the soil enough so that less hardy
species can take the place of these hardier pioneer species. The original gm/cc species
will not have many secondary advantages (almost all of the hardiest species cannot be
eaten because they have serious anti-nutritional factors), but within a few years, farmers will be able to switch to species that do have additional advantages.
Another faster, but more expensive, way of improving very poor soils is to use a
moderate amount of animal manure the first year or two, in order to allow the ideal
gm/cc to develop adequately. After one or two years, the gm/cc will have improved
the soil to the point that there is no more need to use manure.
Elevations listed for the gm/cc species in the decision tree should be taken as a general
guide, not as hard and fast rules. The best elevations will vary somewhat according to
the variety of the gm/cc species being used, the quality of the soil (with better soils,
they will grow well in a wider range of altitudes), and the distance from the equator
(the further from the equator, the lower the elevation at which they will grow best).
Still, we should remember that gm/ccs are only valuable if they grow rather vigorously,
and altitude above sea level has a tremendous impact on the growth of most gm/cc
species. A gm/cc may grow outside of the range of elevations mentioned, but grow so
slowly that it would be better to use a different gm/cc species that grows more vigorously under the given conditions. Other local management considerations can also
affect what species is grown at a given altitude. The lablab bean, for instance, is only
recommended in this decision tree between the altitudes of 0 to 1,500 m. Nevertheless, near Jesus de Otoro, Honduras, hundreds of farmers working at 1,800 m in
elevation are growing lablab beans as a gm/cc despite their less-than-vigorous growth
at that altitude because the farmers are very interested in the high quality dry-season
fodder produced by the plant.
Some agronomists looking at this decision tree will be bothered by the lack of attention to different soil types. Our experience around the world has been that soil type
rarely has a tremendous impact on the growth of gm/cc species, except when soils
are waterlogged or have extremely high or low pHs. When pHs are extremely low,
jackbeans (or even Desmodium ovalifolium) may have to be used until the increase in
organic matter buffers the pH up to a more hospitable level of acidity.
Tree species are apparently more susceptible to pH levels. For example, mother of
cacao (Gliricidia sepium) grows better in acid soils and leucaena (Leucaena leucocephala
and L. diversifolia) grows better in neutral to alkaline soils.
A Note about Scientific Names

The idea of establishing scientific names for each species of plant in the world was
to have only one name for each species. Yet over time, scientists have changed the
names of many gm/cc species, in some cases over and over again, which to some
extent defeats the purpose of having scientific names. Lablab beans have, at different
times, been called Dolichos lablab, Lablab purpureus, Dolichos purpureus and Lablab
niger. Mucuna has been called Stizolobium pruriens, Stizolobium aterrium, Stizolobium
niveum, Stizolobium deeringiana, Mucuna deeringiana, Mucuna pruriens and Mucuna
utilis. Some of these names are used only for one color of mucuna seed (that is, one
particular species of mucuna), but many of them have been used to refer to all the
different types of mucuna in general.
In this book, we use the scientific name that from our experience is most commonly
used around the world today.
29
30
Decision Tree
7. Decision Tree
Introduction
The decision tree presented in this section is relatively simple to use. Start with #1 on
the diagram and answer the questions in each box to help identify the most appropriate gm/cc system for a given location.
Note that there are two types of numbering system used in the decision tree. The
first numbering system simply numbers each box in the decision tree. Additional
information on each of these boxes is provided in the section immediately following
the decision tree named Decision Tree Guide.
The second numbering system in the decision tree identifies each of the recommended
gm/cc systems one arrives at after working through the decision tree. Each number
in this second numbering system begins with an S. Immediately following the
Decision Tree Guide is the section with detailed explanations about the 91
recommended gm/cc systems, titled Green Manure/Cover Cropping Systems.
The most important characteristics of each of the more important gm/cc species
in each system will be included when that species is first mentioned in the decision
tree. A list of where each species description appears is provided in Annex 3: List of
Recommended Green Manure/Cover Crop Species.
Decision Tree
START: Discuss with farmers
No
2. Will you
continue
anyway?
Yes
3. END
of work on
gm/ccs
No
1. Are
farmers interested
in gm/ccs?
Yes
4. Is there a
successful system
nearby?
No
Yes
5. Do field trials
Successful
Not
Successful
6. Organize
educational
field trips
7. Did
farmers decide
to try this
system?
No
Yes
8. Disseminate
this gm/cc system
9. Do
farmers
feel this system is
enough?
Yes
No
10. Is
the land
at <10% slope or
will gm/ccs maintain
cover?
Yes
Go to next page
No
11.
Incorporate
hedgerows
31
32
Decision Tree
Higher than 3,000 m
12.
At what
elevation do you
work?
Below 1,500 m
1,500 - 3,000 m
13.
What is
the dominant
crop?
Maize
Potatoes
Fruit trees
14. Use
S6, S7,
S12 or S13
15. Use
S32, S33,
S34 or S35
16. Use
S40
17. What is
the dominant
crop?
Maize
Potatoes
18. Use
S6 or S13
19. Use
S32, S33,
S34 or S35
Fruit trees Various crops
21. Use
S40 or
S63
20. Use
S39
Go to
#5
22. Are you

working below 800m
& in the tropics?
No
Yes
23. Use trees as
dispersed shade PLUS
some other gm/ccs
24.
Do people
prefer exotic
species?
Yes
No
25. Use
S57
Go to #5
26. Use
S16, & S58
Go to next page
Decision Tree
27.
Do nearly all
farmers use fallow
of 2 years?
No
Yes
28.
Is rainfall
<1,000 m a
year?
No
30.
Grazing animals
are let loose?
Yes
Yes
29. Use S60,

maybe with
jackbean
31. Use S60
No 32. Use S1, S60

or S77, maybe
with jackbean
Go to #5
33.
Are the fields
relatively free of nutgrass & imperata
grass?
Yes
No
35.
Are any grain
legumes widely
consumed?
Yes
36.
Which
ones?
No
34. Use S59

or S89
Lablab beans
Cowpeas, rice,
bean & mungbean
37. Use S23,

S24 or S71
38. Use S9, S22,
S68, S69, S70,
S73, S74, S78,
S79, S80 & S85
Pigeon peas
39. Use S30 or

S72. See #26 &
#32.
Peanuts & bambara

ground nuts
40. Use S43 &

see #36d
Go to #5
Go to next page
33
34
Decision Tree
41.
What are the
dominant
crops?
Maize, sorghum or millet
Other
crops
Wasteland
42.
Are some of
the fields not
intercropped?
Yes
No
43.
Would you
prefer a rotation or
hedgerows?
Dry-season
rotation
Hedgerows
44. Use S14

S20, S50, S51,
S55, S70,
S71 or S75
38. Use S1, S2, S3,

S8, S9, S18, S22,
S24, S30, S52, S53,
S58, S64, S69, S70,
S71, S75 or S85
45. Use
S21, S25,
S58 or S82
Rice
60. Use
S77
46.
Is it upland
or paddy rice?
Paddy
Upland
48. Use S74

S76, S78, S79,
S80 or S81
47. Use
S72 or S73
Go to #5
50.
Are they root
crops, vegetables
or perennials?
Vegetables
& root crops
Perennials
51.
Are they
irrigated or
rainfed?
Irrigated 55. Go to #2
56.
What are the
main perennials?
or do some
research.
Rainfed
52.
Do you
want to use an
intercrop or a
rotation?
Rotation 54. Use S48,

S74 or S75
Intercrop
53. Use S10, S11,
S42, or S43
Go to #5
Coffee
Oil palm,
coconut,
cacao
Fruit trees
57. Use
S16, S28,
or S29
58. Use
S26
or S27
59. Use
S26,S37,
S38, or S67
Decision Tree Guide
Decision Tree Guide

Each of the points below refers to a numbered box in the decision tree. Each point
provides additional information on the decision tree boxes and will assist in selecting
options as one moves through the decision tree.
1. Farmers priorities. If farmers do not include soil fertility or the need for fertilizers as one of their two or three most important problems, and dont seem to
show much interest in working on soil fertility, go to #2. If they do identify soil
infertility as a major problem and show a strong interest in trying to solve the
problem, go to #4.
2. The importance of gm/ccs. Seriously consider the possibility of not working in
soil fertility. If the decision is not to work in soil fertility, go to #3. If you still
believe strongly that soil fertility is a major problem in the area and success in
tackling it is still definitely possible, go to #4.
3. The end. This is the end of the programs work in gm/ccs. Maybe something will
change (for instance, peoples soils become poorer, so they will have more motivation to improve them) that will make gm/ccs more attractive or easier to manage
in the future, but for now, gm/ccs will not be part of the programs agenda.
4. A successful system nearby? Is there a successful gm/cc system near the project
location or in an area of roughly similar farming systems and ecological environment? Or do the farmers know of such a system, and are they interested in it?
Are more smallholder farmers adopting this gm/cc system, or are they gradually
abandoning it? Are they adopting it in the absence of any artificial subsidies or
any program that is promoting it? If the answer is yes to most of these questions,
go to #5. If there is no such system, go to #10.
If there is such a system, but you cannot answer yes to the questions, interview
the farmers that are using it, being careful not to just take their answers at face
value. Analyze the systems economics and its pros and cons, both according to
the farmers and according to your own analysis. If the system does not seem to
provide more benefits than costs, go to #10. If it does, go to #5.
5. Field trials. Organize at least three or four trials of the technologies in question.
Preferably, trials can be done on the land of the best farmer leaders, or of those
farmers who will do the best job of managing the experiments and whose farms,
if possible, are accessible. If there is not a strong relationship with the farmers, or
the farmers may not properly manage the experiments, conduct a trial on a small
plot of rented or borrowed land.
Keep very close records of the costs and benefits as compared to traditional systems, being especially careful to include all labor costs. Far too often those of us
35
36
Decision Tree Guide
who work in agricultural development recommend certain technologies without

knowing at all whether their costs are actually covered by their benefits. If the
experiments fail, go to #10. If they are successful, go to #6.
The germination of all seeds that are bought should be tested by placing 50 to
100 seeds between about 30 pages of a moistened newspaper. The newspapers
should be kept moist throughout the test period. One to two weeks later, the
germination rate of the seeds will be obvious. Note that with some of the more
drought-resistant seeds, such as mucuna, some of the seeds will not germinate
until up to a months time.
6. Educational field trips. Organize educational field trips so that influential farmers of both sexes can see the results of the gm/cc experiments. If possible, it is
important that farmers take note of two things in particular. One is that gm/ccs
may well also be food crops. The second is that after the gm/cc crop residues have
been applied to the soil, traditional crops grow healthier and are more productive. Two educational field trips will probably be needed to observe both of these
features of the system. Go to #7.
7. Farmer interest. If the farmers still do not seem interested, find out why. If their
objections are justified, go to #10. If they seem excited about the new possibility,
or want to learn more about what was done in the experiments, go to #8.
8. Extension work. Start promoting the system in the area, by getting farmers to
test the system on a small scale on a plot anywhere from 10 m x 10 m to 25 m
x 25 m. To do this, use a participatory extension methodology, whereby farmers
first experiment with the new technology on a small scale (known as participatory technology development) and then teach each other the technology once
they have successfully tried it out (known as farmer-to-farmer extension). See my
book, Two Ears of Corn: A Guide to People-Centered Agricultural Improvement,
where I refer to villager extensionists.
It is very important to get as much feedback as possible on these experiments
done by farmers. Have frequent conversations with these farmers. If problems
occur, try to trouble shoot very quickly. In addition to listening closely, watch
closely whether farmers who have tried the technology use it on larger and larger
plots of land, whether those not in the program adopt it (both of these are good
indications), or whether those who tried it abandon it (a bad indication). Go to
#9.
9. Do the farmers want to learn about more gm/ccs? If the farmers understand the
potential of gm/cc systems yet feel that, with this first system working well they
already are using enough gm/cc systems, go to #2. If they would like to learn
about more gm/cc systems, go to #10.
10. Hillsides. If the land most farmers in the area are using is at a 10% slope or less,
Decision Tree Guide
or the gm/cc systems they would most likely apply will keep the soil covered during almost all of the annual rainy season or seasons, go to #12. If the soil is at less
than a 35% slope and will be well-covered during the rainy season, go to #12. If
neither of these conditions applies, go to #11.
11. Contour hedgerows. To prevent or reduce water erosion, many farmers around
the world are using contour hedgerows. This is a subject that could fill a book. I
will mention just a few points that are very important and controversial among
some practioners. First of all, ground cover is far more important than physical
barriers in preventing water erosion, whether the barriers are contour hedgerows,
rock walls, ditches, or whatever. If the gm/cc covers the ground well and the rains
are not too plentiful, farmers may not need hedgerows at all, even when the land
is at a 40% slope. The ranges mentioned in #10 are applicable to most situations.
Second, trees are not as effective as grass at holding the terrace face that may
gradually build up (to as much as 75 cm), so probably the best over-all species
for hedgerows is Napier grass (Pennisetum purpureum), which is sometimes called
elephant grass. This species has a double function, as it holds the eroded soil
very well and provides a large amount of fodder for animals. If possible, use the
varieties of napier grass that do not have small hairs on the leaves, as they are
much more palatable for animals and do not spread into farmers fields nearly as
quickly. If grazing animals roam free during the dry season, the farmer who owns
the hedgerows will receive no benefit from the napier grass as a good source of
fodder. A better choice in that case would be vetiver grass (Chrysopogon zizanioides). If the slope is less than 20% (so the terrace face will never get too high),
the fields are at less than 1,000 m in elevation and the animals are under control,
sugarcane is usually much more popularbut it will fall over if the terrace face
grows more than 25 centimetres (cm) in height.
If farmers really want to plant non-grass crops in their hedgerows, such as pineapples or fruit trees, a good option is to plant the grass and then space the other
crops along the grass barrier. This would also apply to trees planted to create
dispersed shade. If farmers want to plant other grasses, such as lemon grass or a
shorter-stature grass among vegetables, they can use these instead of napier grass,
at least on parts of their land.
Hedgerows should not be planted at less than 12 m apart, because farmers will
usually reject any shorter spacing. The rows can be as far as 20 m apart on less
inclined slopes. If the hedgerows are too far apart to catch all the soil, do not
move them closer together than 12 m; instead, find a way to cover the soil better
with a gm/cc. Hedgerows should be laid out by the farmers using A-frame levels.
Hedgerows can be planted after the gm/ccs or at the same time, except where
erosion is excessive. In the latter case, the gm/ccs will not provide any increase
in crop yields until the hedgerows are developed enough to prevent the gm/ccs
organic matter from being washed down the hill. Go to #12.
37
38
Decision Tree Guide
12.

Altitude above sea level. At what range of altitudes do you work?

a) If mostly below 1,500 m above sea level, go to #22.
b) If mostly between 1,500 m and 3,000 m above sea level, go to #17.
c) If mostly above 3,000 m above sea level, go to #13.
13. Above 3,000 m. What is the main crop in the area? If it is maize, go to #14. If it is
potatoes, go to #15. If it is fruit trees, go to #16.
14. Maize. Fava beans (also called broad
beans) and runner beans (sometimes
called scarlet runner beans because many
varieties have bright orange-red flowers)
can usually be intercropped with maize at
these altitudes (Photos 18 and 19). Both
produce widely consumed and tasty edible
beans. Fava beans can be eaten both green
and dry. The runner bean generally produces more organic matter, and increases
soil fertility more, than do the fava beans.
18. Mexico. In the Americas, fava beans are only

produced at high altitudes, but are produced at low
elevations in the Sudan and other parts of the Arab
world. In either case, they make a very valuable
and tasty gm/cc for intercropping with maize,
sorghum or millet. In fact, since the plants are erect
and therefore will not pull down the sorghum or
millet, they might be one of the best intercrops for
these grains. However, I know of no one who has
ever tried such a system.
Farmers from central Honduras have

intercropped runner beans with their
maize for 20 years without using chemical
fertilizer and without seeing any decrease
in their soils productivity. Runner beans
often dont grow well the first year in a
given field (presumably because of a lack
of appropriate rhizobium), but should
grow very well after the first year. They
also have a tuber that sprouts each year
for up to ten years after the first planting,
thereby eliminating the cost of reseeding
them, but farmers say it is better to replant
them every fifth year.
See information on the systems using both
fava beans and runner beans in S6, S12
and S13. If you choose one of these, go
to #5.
19. Honduras. Scarlet runner beans have tremendous potential for use in highland maize systems.
Sometimes tephrosia will also have to be grown
in the same field, to keep the weight of the runner
beans biomass from pulling down the maize stalks.
Another possibility is sweet clover

(Melilotus albus), used in areas where
people also raise cattle. Sweet clover is a
very hardy perennial that will fertilize the
soil efficiently, and also provide fodder for
cattle during the entire dry season. It is so
Decision Tree Guide
hardy that the biggest worry for years with respect to this species was that it was
difficult to eliminate. One method to remove this species from a field is by pruning it down to the soil surface at the beginning of the dry season. If you select
this system (described in S7), proceed to #5.
15. Potatoes. Tarwi (Lupinus mutabilis) can be used at these altitudes (Photo 14).
Tarwi is used in a number of ways to fertilize white potatoes in the Andean countries. Tarwi produces a traditional, still widely-consumed edible bean, although it
must be washed and cooked before consumed. Tarwi is an erect plant that reaches
about 1.5 m in height. It is also one of the best legumes in the world in terms of
fixing nitrogen and produces up to 400 kg N/ha. Choose between systems S32,
S33, S34 and S35, and then go to #5.
16. Fruit trees. See information on tarwi in #19. See S40 and then go to #5.
17. Areas main crop? What is the main crop in the area? If maize, go to #18. If
white potatoes, go to #19. If fruit trees, go to #20. If it is various other annual
crops, go to #21.
18. Gm/ccs for medium-altitude maize. Both fava beans (Vicia faba) (Photo 18) and
runner beans (Phaseolus coccineus) are commonly intercropped with maize. (See
descriptions of these in #14.) If the farmers prefer fava beans, use S6 and go to
#5. If they would prefer runner beans, use S13 and go to #5.
19. Potatoes. Tarwi (see #15) can be planted with potatoes in a variety of ways. If
farmers prefer to use tarwi in a rotation with potatoes, choose S32. If they prefer
to grow it along the borders of their potato fields, use S33. If they choose to
intercrop it among their potatoes, use S34. In each case, return to #5.
If farmers would prefer to grow fava beans (described in #14) with their potatoes,
see S35, and go to #5.
20. Fruit trees. Select S40, and then go to #5.
21. Other crops. Consider S40 and S63. If you decide in favor of either, go to #5.
22. Below 800 m? If the region is between 800 m and 1,500 m in elevation, or in a
temperate zone, go to #27. If within the tropics and below 800 m in elevation,
consider using dispersed shade along with the low-stature gm/ccs. Go to #23.
23. Dispersed shade. Dispersed shade refers to a light tree cover (about 15% to 20%
shade) that is maintained over a field in the lowland tropics. In these areas, the
mid-day heat is so intense that all unshaded crops stop growing for two or three
hours in the middle of the day. A light shade will create a favorable microenvironment that can increase crop yields by about 40%. Scientific experiments
have shown that 15% shade will also increase the growing period of crops by at
39
40
Decision Tree Guide
least a week or two. This happens because with the reduced temperature under
the trees, both the rate of evaporation and the rate of transpiration are reduced.
Therefore, the soil dries out slower allowing crops to grow longer. Soil nitrogen
also burns off more slowly. These differences often mean that farmers can, for
instance, grow maize instead of sorghum.
The benefit to crops from shade trees provides a win-win situation. In addition
to improving crop yields, the trees can produce fruit, timber, firewood, natural
pesticides, and/or soil fertility. All these advantages come with the simple cost of
planting the trees, protecting them for one or two years from animals (including
termites), and pruning them once a year (or twice a year if there are two rainy
seasons). Even the pruning offers a benefit to families, because the firewood produced in this way usually requires less labor than cutting and carrying firewood
from distant forests.
Most dispersed shade systems consist of trees planted 8 to 12 m apart in each
direction, creating a population of about 64 to 150 trees/ha. If only bushes are
used (for example, to supply shade to low-growing vegetables), the spacing might
be only 1.5 to 2.5 m between the bushes. Gm/ccs used in this way include pigeon
peas (Cajanus cajan) and tephrosia (either Tephrosia vogelii or T. candida).
Dispersed tree systems will also provide significant protection against the effects
of global warming. As the weather gets hotter, the trees can be pruned a little less,
thereby creating the same favorable temperatures for the crops below them. Go to
#24.
24. Native trees or exotic species? Several rapid growing non-native tree species
can improve crop yields, simplify management and provide multiple benefits for
farmers. These are mother of cacao (Gliricidia sepium), leucaena (Leucaena leucocephala or L. diversifolia) and pigeon pea (Cajanus cajan). Mother of cacao grows
three to four times faster under Sahelian conditions than does Faidherbia albida,
one of the best native Sahelian species for use as dispersed trees. The mother of
cacao has edible flowers, no thorns and bark that is a natural pesticide. If exotic
tree species are desired, go to #26.
However, native species are usually more resistant to local pests and diseases, ecologically more desirable, and better known by the local farmers. Sometimes fields
already contain trees of these species, or still harbor root systems that grow every
year but are often cut off at ground level by farmers. Advantages of local trees
include the fact that they do not need to be propagated, and their already welldeveloped root systems will provide good growth and drought resistance from the
beginning. The system of allowing tree roots to grow into trees in farmers fields
is now widely called Farmer Managed Natural Regeneration, or FMNR. It has
been successfully used to dramatically increase the number of trees in major parts
of Niger and Mali.
Decision Tree Guide
Native trees and exotic trees can both be used to provide dispersed shade. A combination will provide a wider variety of trees and increase biodiversity, and will
therefore be a more sustainable option. If the farmers select native tree species, go
to #25.
25. Dispersed shade with native species. The species you decide to use will probably depend on what is locally available. If a lot of trees already exist in the fields,
you can pick and choose among those species, selecting leguminous trees without
spines, if possible. Though it does have the problem of thorns, Faidherbia albida
(previously Acacia albida) is easily the most desirable in very drought prone
areas like the northern Sahel because it drops its leaves during the rainy season
and therefore does not need to be pruned. (Other trees will need some pruning immediately before the rainy season to reduce crop shading.) Furthermore,
because it is dormant during the rainy season, it does not compete with nearby
crops for water. Therefore, despite its thorns, F. albida is preferred by farmers in
northern Sahel. Nevertheless, as one moves from the drier areas to higher rainfall
regions, the advantages of the F. albida will diminish in importance, and the
problems of its thorns and relatively slow growth will make it less desirable than
some other species.
Trees like the shea butter trees that produce an income are also a desirable option.
Those wanting to use native species in dispersed shade systems in very dry areas
(such as the northern Sahel) should consider using the Farmer Managed Natural
Regeneration (FMNR) system. In this system, instead of establishing nurseries
and planting trees (a process that in these extremely difficult environments fails
more often than it succeeds), farmers are encouraged to quit cutting or burning,
and to protect the trees that grow naturally each year from stumps or underground roots in their fields. In many of these environments, well over 50 such
stumps or roots per ha already exist. Trees grown this way almost always survive,
and often grow much faster than trees grown from seeds because the new trees do
not have to grow a new root system. On the other hand, farmers have to resign
themselves to having trees of various species growing in their fields in a random
spacing. In Mali, many Dogon farmers prune these trees in order to maximize
crop production under the trees, and some plant additional trees by feeding
viable seeds of desired species to their cattle. FMNR has been used to successfully populate hundreds of thousands of hectares of cropland with trees in Mali,
Burkina Faso and southern Niger.
26. Dispersed shade with exotic species. First, remember that having many different species in your field (that is, having high biodiversity) is always good. If you
can combine some native trees with the exotic trees, the system will be less risky
and perhaps more sustainablebecause if one tree species dies out, others can
still be used. It should be noted that most native trees grow much more slowly
than leucaena or mother of cacao.
41
42
Decision Tree Guide
Leucaena trees have been written about extensively. They are the miracle trees
for neutral and alkaline soils. The leaves are extremely good for soil improvement (the impact on the soil of various tree species depends on a lot of poorly
understood factors, only one of which is the amount of nitrogen in the leaves).
The leaves can be eaten by animals, but leucaena leaves should make up no
more than half of the animals diet. The branches are good for firewood and the
immature seeds are edible. The trees sprout vigorously after being trimmed or cut
off, so that a forest of leucaena, like one of eucalyptus, never really disappears. L.
leucocephala generally works better below 800 m of elevation and L. diversifolia
grows better above that altitude.
For acid soils, mother of cacao replaces leucaena as the miracle tree. It also
produces vast amounts of leaves that fertilize the soil. The leaves are edible by
animals, though cattle will generally not consume it unless they are hungry.
The branches are good for firewood, the bark can be used to kill rats and mice,
the leaves are used to make insecticides and foliar fertilizers, and the flowers
are edible for humans. The tree also grows vigorously after having been cut off,
which makes it very easy to control the amount of shade under it.
Bushes, such as pigeon peas and tephrosia, can be planted every 1.5 to 2.5 m to
provide dispersed shade for low-growing crops such as vegetables or beans.
Pigeon peas (Cajanus cajan) can be used for shade only when associated with
crops like vegetables that grow low to the ground. The pigeon pea plant lives for
about four years, usually producing its best harvests of edible peas in the second
and third years. Sometimes it grows rather poorly the first year in a given field,
probably because of a lack of rhizobia. Pigeon pea is usually 2 to 4 m tall with its
height strongly influenced by the soil fertility level. If over-shading is a concern,
the pigeon pea can be pruned each year to a minimum height of about 60 cm.
The pea is edible, tasty and highly nutritious. In fact, pigeon pea is the fourth
most widely eaten grain in the entire world (after rice, wheat and maize). It is
eaten in most of India, West Africa and the Caribbean. The grain can also serve
as an excellent feed for animals. It should be cooked for fowl, but not for other
animals.13
It is preferable to use pigeon peas where livestock grazing is controlled during
the dry season. Tephrosia is best used in uncontrolled grazing systems since only
goats will eat this species if it is planted quite near the homestead. See #29 for a
more complete description of tephrosia.
Dispersed shade trees can often be combined with hedgerows planted with
another gm/cc species. The hedgerows can be planted between the shade trees in
every tree row or every other tree row. In the Sahel, for instance, nyama or
tephrosia is grown in hedgerows under dispersed shade provided by mother of
Kowal, Torsten Mark, Manual sobre el Manejo y Aprovechamiento del Frijol Gandul, 1994.
13
Decision Tree Guide
cacao or leucaena. This arrangement is potentially beneficial because dispersed

trees are planted so far apart that they normally do not provide enough organic
matter to maintain the soils fertility.
If interested in a system designed for the northern to central Sahel, go to S58. If
interested in finding a temporary dispersed shade for coffee, go to S16. In either
case, try to organize an educational field trip to somewhere where such a system
already exists (go to #6). If a dispersed shade tree system doesnt exist in the
region, it is impractical to do field trials of this technology because these species
take so long to mature. Thus it is best to plant some dispersed shade examples
with interested farmersnot as experiments, but as eventual demonstrations,
with the idea that the systems will be available as demonstrations for many years.
Since dispersed shade trees also need low-growing gm/ccs to improve soil organic
matter, go to #27.
20. Haiti. Jackbean seeds can often be found in

the forests of Latin America or Africa. In Africa,
village women will often know where to find them.
The jackbean pods can be eaten when they are
long and still thin, but the seeds should not be
eaten, as they contain Concanavalin A, which
blocks calcium absorption by the body. In areas
with a major lack of firewood, the pod can work
quite well as an alternative to firewood.
21. Mexico. In this dry area of Mexico, most of

the maize in this field died because of a lack of
rain, while the jackbean shows no sign of moisture
stress. Jackbean is the most drought-resistant of all
the annual gm/cc species we know.
27. Fallows still exist? If a large majority of the farmers in the area still use a
two-year or longer fallow, go to #28. If
not, go to #33. In the case of improved
fallows (#28 to #32), farmers who cannot
fallow their land will not be able to use the
technology. Avoid promoting improved
fallowing systems where a large part of
the farming population has limited access
to land, as this may further reduce land
accessibility and exacerbate inequalities.
28. Rainfall levels for improved fallows.
The simplest way of introducing a gm/cc
in a region is by using it in an improved
fallow. By focusing on what grows on
the fallowed land, rather than change
the farming system, soil fertility will be
improved at a significantly faster rate than
what farmers would achieve in five to ten
years with a natural fallow. These systems
of improved fallows are normally very
popular, if kept as simple as possible. If
the rainfall is less than 1,000 mm per year,
go to #29. If it is more than 1,000 mm
per year, go to #30.
29. Improved fallows in lowland, droughtprone areas. Probably the best approach
here is to test the following three varieties/
species to determine and compare which
43
44
Decision Tree Guide
grow well and effectively control weeds

given the local conditions of altitude and
weed pressure. The three varieties/species
are bushy-type jackbean, climbing jackbean or swordbean and tephrosia (either
Tephrosia vogelii or T. candida). A week or
two before the beginning of the first rainy
season in the fallow period, broadcast the
seed of both varieties of jackbean and the
tephrosia. The jackbean seed (Photo 20)
should be sown at a rate of about three
seeds per square metre and the tephrosia
about one to two seeds per square metre.
The jackbean (Canavalia ensiformis - white
seed) and its close relative, the swordbean
(Canavalia gladiatus - red seed), are by far
the most drought resistant gm/cc species
22. Honduras. Tephrosia is most widely used as
a temporary shade for coffee, but has tremendous
used, even when just a week or two old
potential as a smallholder gm/cc crop where
(Photo 21). They are also the most resisuncontrolled grazing is practiced. Being poisonous,
tant to poor soils, and grow well in moderanimals will not bother it.
ate shade. The jackbean can fix up to 240
kg N/ha each year, so where nitrogen is
the limiting factor, soil fertility will increase rapidly in one year. Furthermore, like
the tephrosia, jackbean can grow through much or all of the dry season, greatly
reducing the nitrogen loss problem.
Both species of tephrosia are woody-stemmed bushes that will reach three or
four metres in height under good growing conditions. The tephrosias are very
drought-resistant, but usually dont grow well below 300 to 400 m of elevation.
Because they are poisonous, animals will not eat them except under conditions
of serious drought (Photo 22).14 Tephrosia plants will die after about four years,
but the soil is generally fertile before then. African farmers who use either of the
tephrosias as an improved fallow usually cut it down and kill it after only one or
two years, because the soil is ready for cropping by that time. Weed control is
usually very good if it is sown thickly, as mentioned above. Tephrosia can be used
to make an insecticide. If tephrosia is the best fit, see S60, then go to #5.
30. Are animals let loose during the off season? Are the domestic animals (cattle,
sheep or goats) let loose when crops are not in the fields? If so, go to #31. If not,
or if there are no grazing animals in the villages, go to #32.
31. Improved fallows where the animals are loose. Here the best species are the
two tephrosia species and the jackbean. As with the tephrosias, the jackbean may
be attacked by animals if they are very hungry, but this rarely occurs in areas with
14
See, for instance, Cherry, Stefan, Tephrosia fallow management in Cameroon, in ECHO Development Notes, Issue #65,
1999.
Decision Tree Guide
more than 1,000 mm per year. These systems would be the same as those in #29.
Go to #29.
32. Improved fallows where the animals are kept away from the fields. In these
areas, improved fallows can use any one of several species. Try mucuna (Mucuna
spp.) (Photos 15 and 23) and pigeon pea (Cajanus cajan), along with the jackbean
and tephrosia.
23. Honduras. This mucuna was planted in another

part of the same field as shown in photo 15. The
mucuna has only been growing for 45 days, in spite
of the poor condition of the soil. For extremely poor
soil, jackbeans will grow better than mucuna.
24. Honduras. The mucuna we use as a gm/

cc must not be confused with the itchy type of
mucuna, which has a cover of small hairs on the
pod (the variety to the right). This itchy kind should
never be used in farmers fields. Some farmers
traditionally use the itchy kind (e.g., in central
Ghana). These farmers should be told that the nonitchy kind is just as strong, meaning it fertilizes
their fields just as well.
This may be the ideal situation for

mucuna, a gm/cc species that has, in the
past, been tried in many situations that
were not ideal for it. If there is enough
rain mucuna will grow most of the year,
or will reseed naturally in January (July
in the southern hemisphere). The use of
mucuna usually results in a very good,
high-nitrogen level improved fallow that
is also very efficient at eliminating most
weeds. The farmer obtains a virtually
weed-free, fertile soil in 12 months (Photo
13). The mucuna seed can be broadcast
either when there is no vegetation because
the field has just been plowed or the
vegetation cut, or when there is naturally
very little vegetation, such as just before
the rains start (although the field must
have been dug or plowed sometime within
the last year or so). In either case, the
mucuna seed should be planted when
under normal climatic conditions there
will be significant rainfall for at least the
following three or four months. The use of
mucuna can restore the soil in six months
to a year. (See, for instance, S1 and S77.
The mucuna need not be a perennial
mucuna, like that used in Vietnam.)
Mucuna is the best-known of all the gm/
cc species (Photo 24). It is also probably
the most widely used gm/cc species in the
world, since it is extensively grown in the
Americas, Asia and Africa. This may seem
strange, given that it produces no easily
edible bean or widely useful by-product.
The beans can be eaten, and are high in
protein, but only after some major
45
46
Decision Tree Guide
25. Honduras. Mucuna can also be consumed,

once it has been treated to rid the bean of its
L-Dopa (which is a medicine used for treating
Parkinsons disease). Gm/ccs like cowpeas, rice
beans, mungbeans and lablab beans for low
altitudes, and fava beans, scarlet runner beans and
tarwi for higher altitudes, can be consumed without
the need to take any such precautions.
processing to get rid of the L-Dopa they

contain (Photo 25). On the other hand,
for people with Parkinsons disease,
L-Dopa is the drug of choice, so a cup or
two of coffee each morning, made with
well-toasted mucuna seeds, can provide
some relief for those Parkinsons patients
too poor to buy pills. If mixed half and
half with normal coffee, the slightly
different taste of the mucuna coffee is not
noticeable.
Very likely, mucunas popularity is due to

its ability to control weeds. It is by far the
best gm/cc species for this purpose. The
fact that farmers have given it so much use
and prominence is a major indication of
farmers priorities: reducing labor demands is very high on their list and manual
weed control demands a lot of labor.
Most mucunas are superior weed killers because they crawl and climb aggressively. (Photo 13only the perennial mucuna is a bushy type, but it has not
been popular with farmers because it often fails to produce seed and it does not
control weeds as well.) Generally, we prefer not to use mucuna as an intercrop for
this very reasonit is too difficult to keep it from smothering the crop(s) with
which it is associated. At first, program personnel and farmers who see mucuna
growing over everything in sight might fear it will become a pest, crawling over
and suffocating everything. (I have seen it grow up to the top of a 20 m tall pine
tree.) But it is very easy to get rid of. If mucuna is cut below the first branches
before it has produced viable seed, it will disappear from a field.
Mucunas aggressiveness and ability to place its next leaf precisely where it will
prevent weeds from getting any sunlight at all are probably what make it such
a superior green herbicide. Some people also attribute mucunas weed control
attributes partly to allelopathic substances in its leaves that may deter the growth
of weeds. Furthermore, the 140 kg of N that is fixed per hectare and high levels
of organic matter production (in some cases as high as 75 MT of organic matter
(green weight) per hectare each year) make it a wonderful fertilizer (Photo 12).
It also decomposes quickly, like most legumes containing a high percentage of
nitrogen. When the leaves are incorporated into a moist, lowland tropical soil,
they will disappear within two to three weeks.
Rapid decomposition is usually not desired, because a lot of the nitrogen will be
washed out of the soil (and into the groundwater) as a result. That is one reason
to leave the plant material on top of the soil, or bury it together with the rest of
the organic matter from the fallow. If the mucuna is planted with crops, we leave
Decision Tree Guide
it on top of the soil (together with the residues of crops such as maize or millet),
so the decomposition process will take place more slowly. Leaving the mucuna
vegetation on the surface is far preferable in nearly all instances especially since
this practice also saves labor. A fair amount of nitrogen will be lost into the air
(it will volatilize), but this is normally not a major worry. Smallholder farmers
rarely need more than 100 kg N/ha each year.
In the case of any of the maize/mucuna systems, if farmers also raise pigs, they
can experiment with using the mucuna to feed them. Mix one part of maize flour
with one part of ground mucuna seeds, then lightly cook the dough to make a
very cheap (and therefore profitable) feed for pigs that is just as nutritious as commercial feeds. See S8.
In rare cases, mucuna is not effective. For example, in Kalimantan Island in
Indonesia, the local weeds will outgrow the mucuna. But this is rare. Another
situation is in areas where there are lots of wild animals, such as deer (Kalimantan) or wild grazing animals (near the Sierra de Las Minas in Guatemala) which
can eat enough of the mucuna to kill it.
ICRAF, the World Agroforestry Center, recommends planting trees as improved
fallows. This requires a good deal more labor than just broadcasting mucuna or
jackbeans, and it may be several years before you see the desired impact on fertility or weed control. Nevertheless, if the areas natural fallows are quite long and
firewood can be sold at a good price, this would be an option worth considering.
Gliricidia and leucaena would be among the best species.
Pigeon pea was mentioned as another appropriate species for improved fallows
where animals are kept away from the fields. A description of pigeon peas can be
found in #26.
Once you have tried and chosen any one or several of these fallow systems, go to
#5.
33. Noxious weeds? Is your area relatively free of nut-grass (Cyperus rotundus) and
imperata grass (Imperata cylindrica)? If so, go to #35. If not, go to #34.
34. There are serious weed problems. Use S59 or S89, and then go to #5.
35. Do farmers grow grain legumes? Are any grain legumes widely consumed? If so,
go to #36. If not, go to #41.
36. Local grain legumes. What grain legumes are locally known and consumed, or
have a good price in local markets, with major demand?
a) Lablab beans. Go to #37.
b) Cowpeas (Vigna unguiculata), rice beans (V. umbellata) or mungbeans
(V. radiata). Go to #38.
47
48
Decision Tree Guide
c) Pigeon peas. Go to #39.

d) Peanuts (also called groundnuts) or bambara groundnuts (Voandazeia
subterranean). Go to #40.
26. Honduras. Lablab seeds can be black, red,

brown, cream-colored, and speckled, but the hylum
is always a bright white. The black seeds may be
better tasting, at least according to some cultural
preferences.
37. Lablab beans. Lablab beans are

widely consumed in northern Peru (as
green peas), Kenya (as a dry bean, called
black beans), India (as a pulse, called
horsepeas), Bangladesh (called hyacinth
beans), Cambodia and Vietnam (as dry
beans) and the Philippines (as green pods,
for which they have a tender variety),
just to name the places I have personally encountered them being eaten. They
contain no anti-nutritional factors, and
therefore require no special processing
before being eaten (Photo 26).
Lablab beans are especially valuable in

areas of moderate to poor rainfall. Once they have grown for five or six months,
they are extremely drought-resistant and are capable of growing and producing
grain for four to six additional months with almost no rain at all. This means
they can provide fresh beans up to six months after the harvest of all the other
field beans has endeda very important consideration for the nutritional wellbeing of smallholder farmers. However, lablab beans are not nearly as droughtresistant if they have rain for only three months, so they normally will not do
well in areas with two rainy seasons that last only three months each.
Lablab beans can also be cut off at ground level after their first year of production. They will then sprout again. Particularly in drought-prone climates, this
practice can help the lablab bean get a fast start the next year in order to acquire
drought tolerance earlier in its growth period.
The main problem with lablab beans is that they do not grow well in poor soils
that have not been inoculated with rhizobia. In relatively poor soils or in a field
in which they have not been planted before, it is almost always necessary to fertilize them with either chemical or organic fertilizers for the first year or two. After
those first two years, they will grow very well with no additional fertilizer.
Lablab beans are capable of producing over 60 MT/ha of organic matter (green
weight), which means that if grazing animals are not consuming a large part of
the organic matter, the lablab beans will maintain or gradually increase the soil
fertility of maize fields for many years.
Lablab beans produce better when they can climb, so they are frequently planted
as an intercrop in maize fields, where they can climb up the maize stalks. They
Decision Tree Guide
grow somewhat slowly during their first few months, so they never become a
problem for the maize. When intercropped with maize, they should be planted
about three seeds per square metre. If weed control is particularly important, four
seeds could be tried, though the early slow growth of lablab beans makes them
less able to control weeds than, for example, mucuna. Lablab beans should not be
intercropped with sorghum or millet. Since the stems of these crops are not very
strong, the lablab beans will cause the sorghum and millet to fall over.
Animals very much like lablab beans. Cattle will prefer them to virtually any
other fodder. Since the plant as a whole contains about 23% protein, it is a very
valuable animal feed. Where cattle are allowed to graze freely during the dry
season, small plots of lablab beans will be grazed down to the ground, unless they
are protected from the animals. This greatly reduces the value of the lablab bean,
since it does not produce much seed before the dry season starts when most of
the organic matter needed for soil fertility has not yet been produced. Lablab
beans are therefore largely useless when planted where there are free-ranging
cattle, sheep or goats.
27. Honduras. These lablab plants, intercropped

with maize, will stay green and produce seeds all
through a 6-month dry season, thereby providing
highly nutritional, vitamin-rich fodder for cattle right
through the time when animals have the least
fodder to eat.
On the other hand, for farmers who can

ensure that the lablab beans are consumed
by their own animals, lablab beans can be
a wonderful feed. Cattlemen in Honduras grow lablab beans with their maize
during the wet season. After the maize is
harvested, they let their cattle feed on the
lablab plants and maize stalks during the
entire 6-month dry season. This way, their
animals continue to gain weight even during the long dry season (Photo 27).
Probably the best way to manage freeranging cattle is to have farmers grow the
lablab beans for several years, either in
fenced areas or areas where cattle and other animals have no access. Once farmers
realize the value of the lablab bean for human consumption and also for dry season grazing and fertilizing the soil, they will often make village-level decisions to
limit the grazing areas of the cattle or to control grazing completely. This process
is facilitated when the cattle-owners have lablab beans in their own fields, and
therefore will have plenty of dry season fodder.
Use S23, S24, or S71, and then go to #5.
38. Cowpeas, rice beans and mungbeans. Cowpeas, rice beans and mungbeans all
belong in the genus Vigna. They grow in similar ways, except that certain varieties
of cowpeas and rice beans crawl or climb, while others are bushy. The bushy-type
Vignas can be intercropped with maize, sorghum or millet, but the climbing vari-
49
50
Decision Tree Guide
eties should not be intercropped with sorghum or millet, because the sorghum
and millet stalks will not support the weight of the intercropped beans.
Vigna species fix much more nitrogen than do the common beans (Phaseolus
vulgaris). They can fix approximately 80 kg N/ha as opposed to about 30 kg N/
ha for Phaseolus beans, so just switching from Phaseolus beans to a Vigna species
as an intercrop can make a major difference in the maintenance of soil fertility
(assuming all the nitrogen is not being burned off during the dry season).
In general, among the Vignas, cowpeas and rice beans are more drought-resistant,
while mungbeans require more soil moisture. For the driest situations, especially
at the end of the wet season when soil moisture can be expected to gradually
decrease, short-cycle (60- to 70-day) cowpeas are probably the best option.
In places like the Sahel, the introduction of short-cycle cowpeas can improve
soil fertility, because the crops with which they are intercropped usually last at
least a month longer. This fact means that the cowpea residuesat least that
part of them that is not carried off to sell or store for foragewill die and be
incorporated into the soil by termites before the millet or sorghum is harvested.
In this way, the organic matter is protected from the heat and from other farmers
grazing animals, which are not allowed to roam free until the millet or sorghum
is harvested.
If you are working with cowpeas, select S22, S68, S70, S73, S74 or S80, and then
go to #5. If you are working with rice beans, select S9, S69 or S78 and then go to
#5. If working with mungbeans, select S79 or S85, and then go to #5. Hedgerows can also be used with these systems, and are advisable especially under 800
m in elevation. See #45.
39. Pigeon peas. Pigeon peas can be used as a gm/cc in many different ways.
a) Pigeon peas can be a very good temporary, dispersed shade (up to four years)
for vegetables or other low-stature crops. They allow flexibility in complicated
and changing farming systems because they can be taken out or replanted at any
time. They also provide a high-protein food and fertilize the soil very well. See
#26.
b) Pigeon peas can be used as an improved fallow. See #26.
c) If you want to try a system in which pigeon peas are intercropped with maize,
go to S30 and then go to #5. For upland rice, go to S72 and then #5. Pigeon peas
can also be grown together with many other crops, often benefiting them because
of the shade while producing additional food and organic matter for the soil.
40. Peanuts and bambara groundnuts. Peanuts and bambara groundnuts are often
planted together with maize, sorghum, millet and even cassava. They are usually
planted at the same time as the basic grains with which they are being intercropped. Both species can also be intercropped with cassava (see S43, and then
#5, if you and the farmers wish).
I did not describe these systems below. They do not fertilize the soil well because
Decision Tree Guide
the nitrogen and organic matter are burned off or eaten by animals, so farmers
rarely use them for the purpose of fertilizing the soil. This means they are not
gm/cc systems by our definition. Nevertheless, these crops can be used to fertilize
the soil somewhat, especially where dispersed shade is used, dry seasons are short
or animals are absent.
If either of these species is used, dispersed shade and perhaps other gm/ccs should
also be used in order to achieve a noticeable improvement in soil fertility. If you
are working below 800 m in elevation, go to #24. Additional ways of fertilizing
the soil should also be used. Go to #41.
41. Major crop. What is the major subsistence crop in the local farming system?
a) If the main subsistence crops are maize, sorghum or millet, go to #42.
b) If the main subsistence crop is rice, go to #46.
c) If the main subsistence crops are vegetables, a root crop or a perennial,
go to #50.
d) If a lot of the areas land has no crops because it has been turned into
wasteland, go to #60.
42. Maize, sorghum and/or millet are the main crops. Are many of the maize,
sorghum or millet fields free of intercropped species? If not, go to #43. If so,
go to #49.
43. All fields are intercropped. First of all, it is important to make sure this is actually
the case. Observation of the fields will be more reliable than asking people. In
many areas of the world, people feel they should intercrop all their fields, and
therefore claim they do. If a significant portion of fields are not intercropped,
see #49 also. Where all the fields are, in fact, intercropped, the best approach is
usually to plant the gm/ccs in a hedgerow or (if the rains permit) use them in a
dry-season rotation. Both these systems are usually a little less popular than most
other gm/cc systems, but they have been successful and sustainable in many areas.
If farmers prefer a dry season rotation, go to #44. If they prefer hedgerows, go to
#45.
44. Dry season rotations. Grain legumes can often be relayed into maize, sorghum or
millet crops (if the intercrops are harvested before the maize), or planted after the
main crops at the beginning of the dry season (Photo 28). The limiting factors
here will be the length of the wet season, how much moisture is retained in the
soil and whether or not at least some rain falls during the dry season. Dry season rotations are usually used in areas where the total annual rainfall is well over
1,000 mm, although S20, the maize/wild sunflower (Tithonia diversifolia) system,
is used in a droughty area that receives an average of much less than 1,000 mm.
For dry season rotations that are used with maize and its intercrops, you could
choose S14, S20, S69, S70, S71 or S75, and then go to #5.
S50 and S51 can also be used in the same sort of situation, even though they are
51
52
Decision Tree Guide
not dry-season rotations. These systems

violate the gm/cc rule of not spending
money or effort on transportation, but
thousands of farmers use this system in
Burkina Faso. These systems do usually
require a lot less labor than collecting and
taking animal manure out to the fields
from the homestead.
28. Thailand. Lablab beans, cowpeas and rice
beans are each grown during the dry season in
maize fields in northern Thailand. The practice was
originally developed by the farmers mostly to keep
down the weeds and maintain soil moisture, but
these gm/ccs now are also being consumed and
sold in the local market.
45. Hedgerows. These are not usually

very popular among farmers because they
occupy land that farmers crops could
be using and because they often require
pruning during the wet season, when the
demand for labor is highest. This is why
alley cropping was almost universally
rejected by farmers.
Several things can be done to make hedgerows less objectionable. First, farmers
should be consulted about the distance between hedgerows. If they choose (as
they likely will) a distance such as 25 m between hedgerows, we should agree even
though the impact on soil fertility will be minimal due to this large distance. In
time, when they see the impact of the hedgerows on soil fertility, they may want
to add a hedgerow between each of the two they already have, making the distance 12 m instead of 25 m. Perhaps a better practice would be to add a second
row of bushes to the already existing headgerows.
Second, if dispersed trees are planted in the hedgerows, the hedgerows will be less
objectionable, because they will be occupying the area of maximum shade, where
crop growth may be somewhat poor anyway if the pruning is not done heavily
enough. Furthermore, the hedgerows will be seen as part of a total system with
benefits of moisture retention and crop shade in addition to soil fertility.
Third, if on hillsides the hedgerows can double as contour hedgerows to prevent
erosion, they will be more acceptable. This is the case on several islands in Indonesia, where the alley cropping hedgerows have met with considerable acceptance.
Fourth, some hedgerow species, such as nyama (Piliostigma reticulatum), can
be pruned to ground level, which gives the field crops a head start. Nyama is
a native bush in much of the Sahel, from Senegal to Kenya. It is a perennial
whose leaves do an extremely good job of fertilizing the soil. It is very resistant to
animals (although some protection may be required the first year) and maintains
its leaves until June, when the rains come. A field that is no longer fertile can be
plowed and nyama seedlings planted, either from a nursery or using seedlings
that volunteer near where the bushes already exist. The seeding rate will depend
on the number of seedlings available, but they could be planted one seedling
Decision Tree Guide
each metre in hedgerows 8 to 12 m apart. Soil improvement using nyama will

be a slower process than with jackbeans, because the nyama grows much more
slowly. However, this technology could be used in areas where it is so dry that the
animals are eating the jackbean. There is also a tremendous advantage in having a
permanent gm/cc, rather than one that has to be planted every few years, like the
jackbean. When nyama is planted in maize, sorghum or millet fields, it does not
need to be pruned at all during the wet season, which is a major advantage over
alley cropping systems.
For examples of other hedgerow technologies, consider S21, S25, S58 or S82,
and then go to #5. Recently, mother of cacao hedgerows have been gaining
popularity in Haiti, Indonesia and the Philippines. However, they had not been
used long enough when I saw them to justify being included in the list of successful gm/ccs below.
46. Gm/ccs for rice. The search for gm/cc systems that work well with rice, whether
upland rice or paddy rice, has been a frustrating endeavor. Rice is such a shortstature crop that it is difficult to find anything that can be intercropped with it.
Besides, very few legumes can survive for long in standing water. Luckily, there
is not too much pressure to find gm/ccs for rice. For one thing, the System of
Rice Intensification (SRI) is raising rice yields dramatically around the world
without gm/ccs, and many rice farmers are busy learning about and refining their
use of SRI. [For information about SRI, see http://sri.ciifad.cornell.edu/index.
html]. Secondly, rice has such a good international price that farmers can easily
use chemical fertilizer, gather organic matter from other fields or apply compost
without losing money. Both these factors reduce the urgency of finding gm/ccs
for rice. Even so, there are a limited number of options for using gm/ccs with
rice.
If the dominant system is upland rice, go to #47. If it is paddy rice, go to #48.
47. Upland rice. In northern Laos, pigeon peas and a crawling variety of cowpea (also
found in certain areas of Africa, including parts of the Sahel) are intercropped in
some upland rice fields. Choose S72 or S73, then go to #5.
48. Paddy rice. If development workers and the farmers are interested in using a
traditional green manure system, see S76 and then #5. If a system with a rotated
or relayed gm/cc sounds more interesting, see S74, S78, S79 and S80. Afterwards, go to #5. If people are interested in using azolla, an algae that grows in
paddy water and fixes nitrogen, go to S81.
49. Many of the fields are not intercropped. By and large, maize fields that have no
other crops growing in them can be intercropped with any one of a large number
of leguminous gm/ccs. In high-rainfall areas, mucuna can be intercropped with
maize. Plant it at least 30 to 40 days after the maize, and allow the mucuna to
grow until the next crop of maize is planted. The mucuna may need to be pruned
53
54
Decision Tree Guide
at least once, depending on the height and days-to-harvest of the maize. Be careful to plant the mucuna late enough that it will not cover the maize (see S1 and
S2). If the soil is too infertile for mucuna to grow well, jackbeans can be grown
instead. Jackbeans will fix even more nitrogen (up to 240 kg per hectare), but
will usually not provide the same degree of weed control. The jackbean can be
grown alone or associated with maize, sorghum, millet or cassava. When planted
together with other crops, the bushy type of jackbean should be used rather than
the climbing type. (For more information on jackbean, see #29. For information
on mucuna, see #32.)
If farmers are interested in one of the edible legumes and if conditions allow
them to grow well, lablab beans, cowpeas, rice beans or mungbeans are good
options.
Short-cycle legumes, like 60-day cowpeas, are often best for this use. For areas
with less than 1,000 mm of rain, select S3, S9, S22, S53, S58 or S64 and then
go to #5. For wetter areas, select S1, S8, S18, S30, S53 or S85 and then go to #5.
In wetter areas where grazing animals are not a problem, relay systems in which
the legume grows well into the dry season can also be used. Select S24, S69, S70,
S71 or S75, and then go to #5. When shorter-cycle legumes are planted together
with maize, a relay crop might be used in the same field after the first legume is
harvested but the maize is still growing.
For millet and sorghum, any of these systems may be used, as long as the legume
does not climb. In the drier areas of the Sahel, Piliostigma reticulatum, which is a
native bush, can also be used with millet or sorghum (decide whether to try S52,
and then go to #5).
50. Gm/ccs for vegetables, root crops or perennials. If people need gm/ccs in
vegetables or root crops, go to #51. If the people need gm/ccs in perennials,
go to #56.
51. Gm/ccs for vegetables or root crops. If these crops are irrigated, go to #55. If
they are rain fed, go to #52.
52. Rainfed vegetables or root crops. Would the farmers prefer to intercrop the gm/
ccs with their vegetables or use them in a rotation? The latter process is far more
widely used and easier to manage. If they prefer intercropping, go to #53. If they
prefer a rotation, go to #54.
53. Rainfed vegetables or root crops with gm/cc intercrops. If the vegetables are of
short stature, like carrots or radishes, we dont have any proven systems except for
dispersed shade from trees or pigeon peas. Research on this issue is needed and
could prove to be very important. If the vegetables are of medium stature, such
as tomatoes, chilis or eggplants, use S10 and S11, then go to #5. If the vegetables
are on bushes that are fairly large (like cassava, for example), use S42 or S43, and
then go to #5.
Decision Tree Guide
54. Vegetables or root crops with rain-fed rotations. Choose S48, S74 or S75, and
then go to #5. In this case, depending on the nature of the crops in the rotation,
many other possibilities of gm/cc species could also be tried.
55. Irrigated vegetables. Our present level of knowledge doesnt provide us with any
effective options for gm/ccs to use with irrigated vegetables. In fact, the possibilities for gm/ccs in these circumstances will always be rather limited, for a number
of reasons. In irrigated vegetable systems, the crops are usually of short stature,
the system is in constant change and is highly variable, the land is extremely
valuable (making the opportunity cost of growing gm/ccs high), and the value of
the crops is also usually relatively high. Every one of these factors works against
finding advantageous gm/ccs for this situation. In most cases, growing irrigated
vegetables is quite profitable, so the use of chemical fertilizers, compost, animal
manure or some other purchased fertilizer will not only be feasible, but will
prove even more advantageous than gm/ccs. Therefore, the best approach here is
either to go to #2, or try to do some original research into gm/cc use under these
circumstances.
56. Perennials. Where perennials are being grown as a main crop, it is normally more
efficient (though not always possible) to use gm/cc crops that are also perennials.
Using perennial gm/ccs avoids the labor of replanting every year. More importantly, it normally allows a much better control of weeds across the entire field or
orchard once the perennial gm/cc has become established. Thus, the advantages
of a perennial gm/cc include improved soil fertility, reduced weeding and reduced
maintenance costs after the second year.
What is the main perennial being grown? If it is coffee, go to #57. If it is coconuts, oil palm or cacao, go to #58. If it is fruit trees, go to #59.
57. Gm/ccs for coffee. Most of the systems recommended currently for growing
coffee use shade. Leguminous tree gm/ccs are recommended for this role. But the
trees used for permanent shade in coffee fields tend to grow rather slowly, leaving
the young coffee trees unprotected for the first four or five years of their growth.
In Guatemala, using Tephrosia vogelii as a temporary shade for coffee has become
very popular among both smallholder farmers and plantation owners. Check
out S16; if you choose to use it, go to #5. See also #29 for more information on
tephrosia.
Many smallholder farmers are interested in intensifying their coffee fields and
guarding against those years when the coffee price crashes. Some have found that
using diversified fruit trees as shade can improve both their incomes and their
diets. In Guatemala, such farmers will prune their fruit trees fairly heavily in
years when the coffee price is high in order to reduce the shade to 50%, which is
ideal for maximum coffee production. When the coffee price is low, they let their
fruit trees grow without any pruning. This increases their fruit production while
reducing their unprofitable coffee production.
55
56
Decision Tree Guide
However, if fruit trees are used for shade, something must be done to provide
soil fertility. Farmers who want to improve their coffee fields fertility can use
perennial peanuts (Arachis pintoi) as a low-lying gm/cc, if they can get the
cuttings. Perennial peanuts are a little difficult to get established, but once they
are established, they provide a cover less than 20 cm thick that will completely
control weeds and fix nitrogen for as long as it is in the ground. Choose S28 and
go to #5.
If farmers cannot obtain perennial peanut cuttings, they will usually settle for
jackbeans as an intercrop with their coffee. See S29, then go to #5. Also see the
description of jackbeans in #29.
58. Coconuts, oil palms or cacao. Select S26 (or S29 if perennial peanut cuttings are
not available) and go to #5.
59. Gm/ccs for fruit trees. I have seen jackbeans, mucuna, perennial peanuts and
Centrosema pubescens used as gm/ccs for fruit trees. I believe the perennial peanut
is easily the best of these systems. See S26 then go to #5. Centrosema works well
as a perennial, but forms a mat of about 30-40 cm in depth; this means fruit that
drops is often lost. As well, the Centrosema might become a cover for snakes. If
extension workers and the farmers favor this system, go to S67 and then to #5.
Jackbeans can be used, but have to be replanted every year or two. Go to S38,
and then #5. Mucuna would have to be replanted every year in most climates,
and would also have to be trimmed regularly around the trees. I consider mucuna
as advantageous only where farmers have a serious problem with sun scorch (as
they do in Paraguay, where farmers are very happy with mucuna). In this case,
use S37 and then go to #5.
60. Much of the land is wasteland. See S77, which can be used with the normal
annual mucuna, or the perennial species. In most situations, however, it would
be better to plow the soil and broadcast jackbean seeds on it, at a rate of approximately 4 seeds/m2. If the soil is so infertile that even jackbeans will not grow on it
(an extremely rare phenomenon), you might use a little animal manure to get the
jackbean started.
Green Manure/Cover Crop Systems
8. Green Manure/Cover Crop Systems

Latin American Green
Manure/Cover Crop Systems
Mexico
S1. Maize/mucuna-1. The primary maize/
mucuna (Mucuna spp., also called velvetbean) system extends along the east coast
29. Honduras. The maize/mucuna-1 system covof
Mexico, from about halfway between
ers thousands of hectares in Mexico, Guatemala
the U.S. border and the city of Veracruz
and Honduras. It is still spreading in some areas of
each country, but is losing ground in others. The
to Tabasco State, just west of the Yucatan
factors deciding spread or contraction include the
Peninsula (Photo 29). The mucuna dies
abandonment of maize as a crop, the spontaneous
in December, maize is planted in January,
spread to new farmers who just didnt know about
and the mucuna reseeds itself naturally
it, and the competition for maize land from cattle
ranchers and oil palm producers.
in January and grows until the following
December. A major advantage with this
system is that very little weeding, if any, is necessary. The system results in slightly
improved harvests over a period of at least 40 years, with maize grown every year and
with no use of chemical fertilizers (Photo 30). The system apparently either spread
originally from Guatemala, or was independently developed by Mexican farmers when
the seed was introduced into Mexico from the southeastern U.S. during the early
to middle 20th century. An estimated 10,000 to 15,000 farmers use this system in
Mexico alone, and it is still expanding in Tabasco State and north of Veracruz.
The same system has spread into Guatemala along the Polochic River Valley of the
Department of Alta Verapaz. During the last 30 years, it has spread into many areas of
the Department of El Peten. Nevertheless, its use along the Polochic River has been
reduced, largely because much of the land there has been bought up to grow sugarcane
for making biofuels. The system apparently was initiated about 60 years ago when
the mucuna was first brought to Guatemala by the United Fruit Company to feed the
companys mules. Over 5,000 farmers likely use this system in Guatemala.
The maize/mucuna-1 system has also spread from the Guatemalan border north into
patches of southern Belize and east to the mountains of north-central Honduras,
where it was once used by an estimated 10,000 Honduran farmers. In this area,
however, the system is losing ground rapidly because maize farmers are being replaced
by large cattle ranchers who now prefer to use the hillsides to graze their cattle because
Hurricane Mitch killed many of the animals they previously grazed on the coastal
flatlands.
It should be mentioned that rainfall basically has to be year-round for this maizemucuna-1 system to work.
57
58
30. Honduras. This photo is not of a handful of

compost, though it may seem like it is. This is
the soil of a typical field in the maize/mucuna-1
system, after planting maize every year for 40
years. Without the mucuna in this poor tropical soil,
a farmer can grow maize for three to four years,
and then will have to fallow it for ten to twelve
years. By using mucuna (which can be seen in
the background of the photo), farmers can plant
maize every single year with no loss in productivity.
Instead, production gradually increases, up to a
point of diminishing returns at which maize yields
are about 35% less than those produced on the
lower-elevation flatlands using hybrids, irrigation,
plows, etc.. However, even though the productivity
is lower, the total cost per ton produced is about
30% lower than in the conventional system,
because the expenses are so much lower. Thus,
the maize/mucuna system provides a much larger
profit while operating on less valuable land and
increasing the productivity of that land from year
to year.
S2. Maize/mucuna-2. This system is used

west and south of the maize/mucuna-1
system, where there is less rainfall and a
dry season occurs from December through
April. It is used on irrigated land, with the
mucuna intercropped among the maize 30
to 40 days after the maize is planted. The
earlier planting increases the danger of the
mucuna choking out the maize, but the
later dates reduce the organic matter that
the mucuna will eventually produce. Often
one pruning is needed to keep the mucuna
from covering the maize, although this
will depend on the height of the maize,
how many months the maize requires to
mature, etc.. Different planting patterns
are used; one is to plant three seeds per hill
of maize, but only in every other row. The
planting seasons depend on the supply
of irrigation water. This system seems to
only be used by a few hundred farmers in
Mexico, and a few others in Costa Rica.
This system is probably an adaptation of
maize/mucuna-1 developed by farmers
themselves for use in lower-rainfall areas.
S3. Maize/jackbean. Jackbean (Canavalia

ensiformis) is intercropped with maize in
a number of areas of northern Yucatan
and the state of Oaxaca. The jackbean is
planted at the same time as the maize, with
two or three seeds per square metre, or
sometimes four if the weeds are particularly problematic. Sword beans (Canavalia
gladiatus) can be used in place of the
jackbean. In Brazil, both jackbean and C.
paraguayensis are intercropped with maize. Thousands of farmers in both Mexico and
Brazil use this traditional system. For this system, use only the bushy-type jackbean,
not the climbing type.
S4. Maize/lima bean. Lima beans (Phaseolus lunatus), locally called ibes, are intercropped with maize on the Yucatan Peninsula. The beans produced are, of course,
eaten. The lima beans are planted either together with the maize or up to a month
afterwards. Tens of thousands of farmers are using this traditional system. A similar
system is used in western Guatemala (where the lima bean is called piligua) and
southern Honduras (where it is called chilipuca).
S5. Maize/vetch. Vetch (Vicia toluca) is widely intercropped with maize in the State of
Michoacan, both to fertilize the soil and to feed animals during the dry season. Probably well over 10,000 farmers use this traditional system. The vetch is planted at the
same time as the maize. After the maize harvest, most of the vetch is cut and carried to
the homestead to be stored and fed to animals during the dry season.
S6. Maize/fava beans. Fava beans are intercropped with maize by hundreds of
thousands of farmers on the central plateau of Mexico, both for soil fertility maintenance and food (Photo 18). Usually two or three seeds are dropped in each hill
of maize. The same system is used in the higher areas of the Guatemalan highlands
(thousands of farmers) and in Lao Chai Province of Vietnam (by only a few hundred
farmers, to my knowledge). This traditional system maintains maize yields for at least
twenty to thirty years.
S7. Maize/sweet clover. In this system, which I believe exists in only two or three
villages in northern Oaxaca State near the town of Tlaxiaco, the sweet clover (Melilotus
albus) is planted together with the maize. The sweet clover is a perennial and grows to
about two m in height after the maize is harvested. After the harvest, the maize stalks
and sweet clover are grazed to ground level, if necessary, during the dry season. When
the rains start again, what is left of the sweet clover is cut down to ground level and
the maize is planted again. In this way, the sweet clover maintains the soils fertility
for decades and feeds the animals during the dry season. This system should only be
used where farmers value raising cattle, because it is fairly permanent. The sweet clover
can only be killed by cutting or grazing it to ground level at the beginning of the dry
season. Probably around 50 farmers use this system, which was introduced in the
mid-20th century.
S8. Maize/mucuna-3. The difference between this system and S1 or S2 is that farmers are using the mucuna seeds to feed pigs, which is a very lucrative business.15 As
a result, the farmers are not concerned if the mucuna reduces their maize harvests a
little. In the southern Yucatan Peninsula town of Xpujil, farmers intercrop mucuna
with maize, and then feed the mucuna seeds (cooked and mixed half and half with
maize meal) to their pigs. Introduced in the 1990s, the system never included more
than a hundred farmers, to my knowledge.
S9. Maize/rice bean-1. Rice bean (Vigna umbellata) is intercropped with maize, largely
to be eaten. Both crops are planted at the same time, with rice bean seeds planted
at about three seeds per square metre. This practice was once (and perhaps is still)
practiced by thousands of farmers on the Isthmus of Tehuantepec. It was also common at one time on Guatemalas south coast and in much of El Salvador (where it was
used by hundreds of thousands of farmers). A similar maize/rice bean system is also
practiced in Vietnam.
People like the taste of the rice bean, but Mesoamerican women complain that the
tiny bean is hard to thresh and clean. Anyone promoting this system in Latin America
15
CIDICCO, et al., Experiencias Sobre Cultivos de Cobertura y Abonos Verdes (Tegucigalpa, Honduras: CIDICCO, 1997),
pp. 112-15.
59
60
should therefore also teach the community to winnow the grain as Asians do their rice
and other small grains.
S10. Tomatoes/jackbeans. Jackbeans are intercropped in the rows between tomato
plants. They are usually planted at the same time as the tomatoes. When they begin to
shade the tomatoes, they are pruned back to a height of about 30 to 40 cm, which
forces them to sprout more branches and grow laterally rather than vertically. At
most one hundred farmers use this practice that was developed in Yucatan during the
1980s.
S11. Chili peppers/jackbeans. Jackbeans are managed in chili pepper fields in Yucatan
the same way as in the tomato fields in S10.
Guatemala
S12. Maize/runner bean-1. In the Department of San Marcos, scarlet runner beans
are intercropped very sparsely with maize. The runner beans produce a good deal of
organic matter. Since they are climbers, they can cause the maize to fall over toward
the end of the growing season. Therefore, farmers tend to plant it quite sparsely, using
as few as one or two seeds in each 3 m x 3 m sector. The runner bean is planted at the
same time as the maize. At least 3,000 to 5,000 farmers use this traditional system.
This problem of the production of a massive amount of organic matter by the runner bean could be a tremendous advantage in terms of soil fertility and productivity. I
suspect that if the runner bean and maize could be supported by a third, fast-growing,
woody-stemmed plant (perhaps tephrosia?), the system could become one of the most
profitable and sustainable highland maize-growing systems anywhere. This possibility
needs to be researched.
S13. Maize/runner bean-2. In the central highlands a variety of runner bean is intercropped much more densely with maize (up to one seed for every 2 square metres)
(Photo 19). Maize yields have been maintained for up to 20 years in many places
with this system. Before the Europeans arrived, similar systems were apparently used
in virtually all the temperate or highland areas from New York State in the United
States, south to Mexico, Guatemala, Honduras, Colombia, Ecuador, Peru, Bolivia and
northern Chile. Hundreds of thousands of farmers, stretching from Mexico to Bolivia,
use this system today.
S14. Maize/mucuna-4. Mucuna is rotated with maize in the Cerro San Gil region,
Department of Izabal (Photo 31). The maize is planted when the rains start in May
or June, and mucuna is relay-cropped into the maize 30 to 40 days before the maize is
harvested (Photo 32). Maize following mucuna, without chemical fertilizer, has produced up to 7.4 MT/ha on Brazilian experimental stations.16 This is evidence of the
potential yields that can be achieved by rotating or relaying mucuna with maize. This
system is also used in Nicaragua and Brazil. In Paraguay some farmers wait to plant
Lathwell, op. cit.
16
33. Guatemala. For centuries, choreque was intercropped with maize. This photo, taken at the end of
the dry season, shows that only the choreque and
a few trees and bushes have remained green.
31. Honduras. Mucuna in a maize field after the

maize harvest. Most farmers in Honduras who
intercrop a gm/cc with their maize no longer use
mucuna because there is too much danger that it
will smother the maize. Farmers prefer to use the
less aggressive jackbean, or something that is
more useful, such as lablab beans.
32. Honduras. The maize/mucuna-4 system as

used in Honduras
34. Guatemala. We failed to spread choreque to

other areas of Central America because it has a
very narrow ecological niche. The climate must
be cold, but the plant will not withstand frosts.
Furthermore, the runner bean and sweet clover
can do what the choreque does, with a number of
important additional advantages.
35. Honduras. In this field of second-season maize, the

maize is tall enough to have needed a second weeding.
Nevertheless, because the owner of this field is using the
maize/mucuna-6 system, the field was not weeded. This
is one very clear example, among many, of the fact that
the oft-stated belief that eco-agricultural practices always
require additional work is not correct.
61
62
the mucuna until 40 to 60 days before the maize harvest. In this way, the mucuna will
keep growing through much of the dry season, rather than dying in December, as it
usually does. While thousands of farmers in Paraguay and Brazil use some version of
this system, only a few hundred farmers in Guatemala and about 2,000 farmers in
Nicaragua do.
S15. Maize/choreque. Many farmers in the west of the Department of Chimaltenango
traditionally used to intercrop choreque (Lathyrus nigrivalvis) with their maize (Photo
33). They then fed their animals on the choreque throughout the six-month dry season. The system has died out largely because the area where it was practiced has gone
heavily into vegetable production, and few people have cattle any more (Photo 34).
Forty years ago, slightly over 2,000 farmers used this traditional system.17 This system
and one other in this section are the only ones close to dying out.
S16. Coffee/tephrosia. Tephrosia vogelii has become popular in the last 35 years among
large-scale coffee producers on Guatemalas south coast as a temporary shade for coffee
during plantation establishment. Farmers plant the tephrosia at the same time as the
coffee, and it shades the coffee for three to four years before the tephrosia dies out. By
that time, the slower-growing permanent shade trees are providing adequate shade.
Adopters probably number in the thousands.
Honduras
S17. Maize/mucuna-5. Some farmers have tried to intensify the maize/mucuna-1
system by growing two crops of maize a year. This complicates management of the
mucuna, but pays off in increased over-all yields per hectare. The maize is planted
in both January and May. The mucuna has mostly died out in December, but drops
seeds that will sprout in January/February. However, in April the farmers must cut
down all the mucuna or eliminate it with an herbicide. As far as we know, only a few
hundred farmers use this recently developed system.
S18. Maize/mucuna-6. Around the town of Omoa on Honduras northwest coast,
farmers have developed another mucuna system. They plant maize in May/June intercropped with mucuna. After the harvest, they cut down the mucuna to form a dense
mulch and injection plant the second crop of maize through the mulch in September,
without planting mucuna. The mucuna mulch allows them to grow the second maize
crop without ever having to weed it. Probably not more than 200 or 300 farmers use
this now traditional system (Photo 35).
S19. Maize/mucuna-7. In this surprising system near La Entrada in the Department of
Santa Barbara, the mucuna is grown just as in S1, and then is burned just before the
next crop of maize is planted. Perhaps a few thousand farmers use this system.
CIDICCO, op. cit., pp. 57-61.
17
36. Honduras. Wild sunflower is a phosphorus

accumulator, so it can be especially important as
a green manure in areas where phosphorus is the
limiting factor in crop production.
S20. Maize/wild sunflower-1. Farmers

around Sabana Grande in central Honduras sometimes broadcast wild sunflower
(Tithonia diversifolia) seeds into their fields
after the maize harvest in November/
December (Photo 36). It grows throughout the dry season, and provides a good
source of fertility for the following years
crops. Probably a hundred farmers, at
most, use this recently developed system.
S21. Maize/wild sunflower-2. Farmers in
western Honduras sometimes plant the
wild sunflower as a contour hedgerow,
cutting and spreading the branches across
the field before planting each year in order
to fertilize the field. Again, this is another
system recently developed by innovative
farmers. It is probably only used by a
hundred farmers or so.
S22. Maize/cowpea-1. Cowpea (Vigna

unguiculata) varieties, locally known as
37. Honduras. Cowpeas were traditionally
alacin and pochote, are intercropped
intercropped with maize in much of El Salvador and
with maize in much of southern Honduras
the Pacific coasts of Guatemala, Honduras and
(Photo 37). They are usually planted at
Nicaragua. Now, because of the buying up of land
the same time and within the maize row
by large-scale farmers and the active promotion of
monocropping by government extension services,
to facilitate weeding. Probably somewhere
the practice is used by perhaps only 10% of the
between 5,000 and 10,000 farmers use
farmers it once served.
this traditional system. Several thousand
farmers in El Salvador and an equal
number in northern Nicaragua continue to use the same system, as do thousands
of farmers on the south coast of Guatemala, where cowpeas are called frijol rienda
or frijol de tierra. In Panama, tens of thousands of farmers use virtually the same
traditional system.
S23. Maize/lablab bean-1. Farmers in the village of Pacayas, near Guinope, intercrop
lablab beans (Dolichos lablab) with their maize, planting both of them at the same
time in May/June when the rains start, at a rate of two to four seeds per square metre
(Photo 27). In this case, farmers are motivated in large part because the organic matter from the lablab controls the nematodes that previously did tremendous damage
to the irrigated garlic produced during the dry season (December to April). Probably
about 50 farmers use this system, which was recently developed by farmers. A similar
traditional system of lablab intercropped with maize is used in northern Peru (notably
between Cajamarca and San Ignacio), where the lablab beans are eaten as green peas.
In this case, tens of thousands of farmers use the system.
63
64
S24. Maize/lablab bean-2. During the last 20 years, large-scale lowland cattle ranchers
in southern Honduras have begun intercropping lablab beans with their maize, so that
the cattle will have a plentiful, green, very palatable and high-protein fodder throughout the 6-month dry season. The number of ranchers using this system is well over
100, and is increasing.
S25. Maize/tephrosia. Tephrosia is sometimes planted in hedgerows and then cut and
scattered across the fields to fertilize crops right before farmers plant their maize. I
have no idea how many people do this, but probably not many.
S26. Oil palm/perennial peanut. The perennial peanut or forage peanut (Arachis pintoi)
is planted under new oil palm trees (when they are about 50 cm tall) in large plantations, mostly to prevent weeds from ever growing in the fields, but also to supply
nitrogen to the oil palms. It takes about a year to eighteen months to establish the
perennial peanut, but once it has covered the ground, virtually no weeding ever needs
to be done again. Several thousand large-scale farmers use this system in Honduras
and Costa Rica. Occasionally the perennial peanut is used the same way to produce
palm hearts (pejibaye) in Costa Rica.18
S27. Oil palm/desmodium. Desmodium (Desmodium ovalifolium) is used as a cover
crop in oil palm plantations by perhaps a few hundred farmers.19 Desmodium is also
being used this way in Belize in situations where the soil pH is below 5.0.
Costa Rica
S28. Coffee/perennial peanut. The perennial peanut is used in coffee fields, much as it
is used in Honduras for oil palms. Somewhere between five hundred and several thousand farmers use this system that I believe was developed during the 1970s to 1980s.20
S29. Coffee/jackbean. Jackbeans are also used among coffee plants, both to fix nitrogen and control weeds. I believe the number of farmers presently doing this is well
into the hundreds.21
Panama
S30. Maize/pigeon pea. Pigeon pea (Cajanus cajan) is widely consumed in Panama,
and is also widely intercropped with maize as a gm/cc species and food crop. Approximately one seed of pigeon pea should be used for every two square metres, planted at
the same time as the maize. Tens of thousands of farmers use this traditional system,
both here and in Brazil. Farmers near the town of Machakos in Kenya also plant
pigeon peas in their maize, but the number of pigeon pea plants per unit of land
varies widely, often in the same field. The number of farmers doing this runs into the
thousands.
Ibid., pp. 87-90, 94-98.
Flores, Milton, La Utilizacion de Leguminosas de Cobertura en Plantaciones Perennes. n.d.
20
21
18
19
S31. All crops/mucuna. Mucuna is used to restore land that has been taken over
by pajablanca, a tall grass of the genus Saccharum, closely related to sugarcane.
Pajablanca was imported into Panama to stabilize the steep banks along the canal. It
has since become a very noxious weed. Mucuna is being used to control it, in ways
similar to that used to control imperata grass (see S59). Probably a few hundred farmers are using this system, which farmers first developed about 15 years ago.
Peru
S32. Potatoes/tarwi-1. Tarwi is described in #19. In the Andes, it is traditionally
grown in rotation with potatoes, both to fertilize the potatoes and to provide a highprotein bean. The tarwi is planted during the rainy season one year, potatoes the next
year, and often a small grain the third year, before returning to tarwi. Potato yields
often triple under this system. Used in Peru, Ecuador and Bolivia, this traditional
technology has hundreds of thousands of practitioners.22
S33. Potatoes/tarwi-2. Tarwi is often also grown around the edges of potato fields, for
the same two purposes. Tens of thousands of farmers also use this traditional practice
in both Peru and Ecuador.
S34. Potatoes/tarwi-3. In Peru some farmers intercrop the tarwi with their potatoes.
The latest information I have is that a few hundred farmers continue to use this
practice.
S35. Potatoes/fava beans. Fava beans are planted in a rotation with potatoes, much
as the tarwi is in S32. Again, this is a traditional system used by tens to hundreds of
thousands of farmers.
S36. Potatoes/peas. Common peas (Pisum sativa) are used just like the tarwi in S32.
This system is used by hundreds of thousands of farmers.
Paraguay
S37. Citrus trees/mucuna. A handful of farmers use mucuna as a gm/cc for citrus
(Photo 38). Though this requires more work than using a perennial, they have found
the mucuna very useful because they let it climb over the tree to cover it right before
harvest, thereby preventing sun scorch on the fruit. Mucuna is also used as a cover
crop for citrus trees near Veracruz, Mexico. In this case no problem of sun scorch
exists, so a perennial cover crop would probably be preferable.
S38. Citrus/jackbean. Jackbean is a more conventional gm/cc for citrus. It requires less
work because it can grow and maintain the cover for several years. When a bushytype is used, it does not climb up into the trees. Perhaps several hundred farmers in
Paraguay use this system, and even fewer in Honduras.
22
Beingolea Ochoa, Victor Julio, Inventario de Sistemas de AVCC de Pequeos Productores en los Tropicos, 1997.
Unpublished.
65
66
S39. Citrus/lupine. In Paraguay and Brazil,

some lupine species (i.e.: Lupinus albus) are
used as a gm/cc for citrus (Photo 39). The
number of farmers who used this practice
previously numbered in the thousands.
Anthracnose has now decimated the lupine
crop in Brazil, in part because of not rotating gm/ccs (personal communication with
Rolf Derpsch).
S40. Various other tree crops/lupine.
Lupines were also used with other tree
crops to improve soil fertility.
38. Paraguay. Gm/ccs are often used among fruit
trees (in this case, a citrus grove), more because
of the weed control they achieve than because of
the improvement in soil fertility.
39. Brazil. Lupines are grown in Brazil as a

perennial cover crop for fruit trees. They are used
primarily in areas where frosts can be a problem,
because they are highly resistant to frost.
40. Paraguay. Because of the jackbeans growing

between the rows of this cassava field, the field has
required only one weeding, rather than the three to
four weedings a crop of cassava normally requires.
S41. Maize/mucuna-8. Mucuna is used

to prepare land for tobacco. One crop of
mucuna, often intercropped with maize,
is used to rehabilitate worn-out land to
permit farmers to then grow tobacco,
which requires fairly fertile land. Sometimes farmers can also shift to zero tillage
after only one year of mucuna.23 Thousands of farmers use this technology.
S42. Cassava/jackbean. In a traditional
system, jackbean is intercropped with
cassava (Photo 40). Both are planted at
the same time. The jackbean provides
nitrogen to the cassava crop, but farmers
appreciate even more not having to weed
their cassava again after the first weeding.
The jackbean is planted between the rows
of cassava. Tens of thousands of farmers
use this traditional practice. Several small
groups of farmers in Honduras have taken
up the same practice.
S43. Cassava/peanuts. Peanuts (Arachis hypogaea) are also quite commonly
intercropped with cassava to help with soil
fertility and to provide a second cash crop.
Probably tens of thousands of farmers use
this traditional technique.
23
Ministerio de Agricultura y Ganaderia de Paraguay/GTZ, Extracto de los Sistemas de Produccion del Proyecto de Desarrollo Rural de San Pedro Norte, n.d.
Brazil
EPAGRI, the agricultural development arm of the state government of Santa Catarina
in Brazil, has introduced well over a hundred gm/cc systems, using over 70 gm/cc species, in the state. Many of these systems, and some others, have also been introduced
into huge areas of Parana, Rio Grande do Sul and other Brazilian states, to a sum total
of well over a million farmers. I have personally observed the handful of systems listed
below, but they are only the tip of a very large iceberg.
41. Brazil. Brazilian farmers use gm/ccs, among

other things, to be able to increase the organic
matter in their soils as rapidly as possible, so they
can begin using zero tillage within a year or two.
Therefore, gm/ccs like this crotalaria, that was
intercropped with maize, are very popular, because
they commonly produce a total biomass of well
over 80 t/ha (green weight).
S44 to S46. Maize/crotalaria. At least three

species of the genus Crotalaria are used as
intercrops with maize in Santa Catarina
State (Photo 41). The crotalarias are generally planted at the same time as the maize.
In Brazil, all of these systems including
those using other intercrops such as forage
turnips, jackbeans and oats (the latter not
included here because they are not leguminous) are designed not so much to provide
nutrients to the soil as to produce large
quantities of organic matter. Farmers use
the systems so they can increase their soil
organic matter content enough to move to
zero tillage as quickly as possible. Over 25
million ha of land in Brazil and Argentina
are now in zero tillage.24
S47. Fruit/mucuna. Mucuna is used as a cover crop under various species of fruit trees
in Acre State. Perennial peanut would probably be a better choice of cover crop, to
reduce the labor of cutting the mucuna out of the trees and of reseeding the mucuna
each year.
S48. Onions/mucuna. Mucuna is used in a rotation for onion production in Santa
Catarina. Hundreds of farmers use this system introduced by EPAGRI.
S49. Various crops/common peas. For a number of different crops, peas are grown as
a winter rotation crop in Santa Catarina, both as a gm/cc and as a cash crop.
Hundreds, if not thousands, of farmers use this traditional system.
24
Brown, Lester R., World on the Edge, How to Prevent Environmental and Economic Collapse (New York: Earth Policy
Institute, 2011), pp. 143-44.
67
68
African Green Manure/Cover Crop Systems

Burkina Faso
S50 and S51. Millet and sorghum/nyama-1. Nyama (Piliostigma reticulatum) is a
perennial bush native to the Sahel (see #45 for information on nyama). In this system,
nyama leaves are carried from the bush to fertilize millet and sorghum in northeastern
Burkina Faso. We know that at least hundreds of farmers use this traditional system.
S52. Millet/nyama-2. Nyama is allowed to volunteer in farmers fields. Just before the
rains come, it is pruned down at ground level and the leaves are spread across the field
to fertilize it. This traditional practice is used by thousands of farmers in both Burkina
Faso and Mali.
Ghana
S53. Maize/mucuna-9. This system is a simple rainy season intercrop of mucuna in
maize. The mucuna is intercropped late enough in the maizes growing season so the
maize is not overcome by the mucuna. In one area, the people were actually using
the variety of mucuna that provokes severe itching, meaning they couldnt go near
their fields for two months. They switched to the non-itchy mucuna immediately
after learning about it. Thousands of farmers use this traditional system. In addition,
thousandsif not tens of thousandsuse the same system in Benin and Oaxaca State
in Mexico.
S54. Maize/nyama. At the end of the fallow period, when the fields are to be burned,
the nyama bushes are cut and covered with dirt, so that less nitrogen will be burned
off. Actually, this system was used traditionally for all local crops, but is disappearing
as fallowing itself is disappearing in the Sahel. Quite likely almost no one is using this
traditional system today.
S55. Maize/calopogonium. Calopogonium (Calopogonium mucunoides) is used in
rotation with maize in parts of Ghana and neighboring countries. There are probably
thousands of adopters. This system started being used after calopogonium was introduced into the country. To the dismay of many farmers, calopogonium has become a
noxious weed. This system is not recommended for use anywhere else.
S56. Maize/leucaena. In another interesting system, Leucaena leucocephala is intercropped (three seeds per square metre) in maize and allowed to grow the whole rainy
season. After the maize harvest, the field is burned lightly to kill the leucaena without
burning away all the nitrogen. This is done in Benin (where the practice originated
and where it is apparently widely practiced) and southern Ghana, where perhaps a few
hundred farmers use it.
Mali
S57. Millet/Faidherbia albida. Farmers in Mali, Burkina Faso and Niger purposely
leave existing Faidherbia albida (previously Acacia albida) trees in their fields. These
trees, different from all the other Sahelian trees, drop their leaves during the rainy season, therefore supplying a very beneficial 15 to 25% shade to the crops under them.
This shade allows crops to keep on growing all day, instead of closing down during
the excessive mid-day heat, thereby increasing productivity by up to 40%. The lower
soil surface temperature achieved by the shade also dramatically reduces evaporation
and transpiration rates. As a result, soil moisture losses are reduced and the annual
growing season is extended by a week or two. Farmers who practice this traditional
system probably number in the hundreds of thousands.
S58. Millet/nyama-3. This is the only system in the decision tree that has never been
practiced by any farmers exactly as described. But I believe it would be an ideal system
for the African Sahel. I venture this bit of hubris for two reasons. First, the system
I am suggesting is basically a combination of two already-existing systems (S52 and
S57). A fast-growing exotic tree and organized on a grid to make animal traction
plowing easier will reduce the labor requirement of the traditional systems.
Second, the Sahel suffers from a lack of good gm/cc systems. This has happened in
part because of a widespread but mistaken belief that they will not work under semiarid conditions. It also has happened in part because the environment in the Sahel
is definitely somewhat hostile to all living things, especially gm/cc species, with their
high nitrogen, and therefore high protein content.
Yet even in the northern Sahel, two gm/cc systems are presently working very well.
Farmer managed natural regeneration (FMNR) has populated some 5 million hectares
of Niger with trees, as well as another half million hectares in northern Mali. Since
a large number of these trees are leguminous, the soil in this system has gained both
an increase in organic matter and an increase in nitrogen. Millet yields have increased
significantly. Another system, developed by the Dogon around the town of Koro,
involves leaving naturally occurring trees in the fields and protecting certain other
desired species of trees (such as Faidherbia albida) that sprout naturally in their fields.
These trees are all pruned in the shape of a funnel, which prevents the area under the
tree from getting too much shade, yet shades all the field at some time during the day
as the sun moves across the sky. Under this dispersed shade, the Dogon farmers practice a rotation that includes such gm/ccs as peanuts, Bambara groundnuts and fonio.
Just as with FMNR, yields have increased as the organic matter content of the soil has
also increased. The last time I visited this area, the Dogon farmers using this system
were collecting the largest harvest they had had in years, while everyone around them
for many miles was complaining that the drought had destroyed their crops.
For the southern Sahel (from about 13oN latitude towards the south), the gm/cc
system that I would like to see tried more widely consists of dispersed shade provided
by mother of cacao (Gliricidia sepium) and, if farmers wish, Faidherbia albida trees,
69
70
spaced anywhere from 8 m to 12 m square. Parallel rows of nyama would run under
every row or every other row of trees (i.e. every 12 to 24 m across the field). In drier
northern areas, where the nyama grows rather slowly, tephrosia (Tephrosia vogelii or T.
candida) could be planted along with the nyama to provide plenty of organic matter
until the nyama is large enough to do so.
With sufficient rainfall, the mother of cacao trees will produce well within five to
seven years, and the nyama would produce sometime before that. Within seven years,
this system would produce enough gm/cc organic matter to maintain soil fertility. By
including the already traditional cowpeas, peanuts and Bambara groundnuts (whose
nitrogen would be better-maintained because of the dispersed shade), I am sure the
system would be capable of increasing soil fertility in the Sahel, without using any
chemical fertilizer.
Benin
S59. All crops/mucuna. Mucuna is used in
Benin to control imperata grass (Imperata
cylindrica, generally called speargrass in
West Africa). Imperata grass is one of
the worlds most noxious weeds (Photo
42). With mucuna, imperata-dominated
wastelands are returned to cultivation.
42. Benin. Two plots on a research station show
The
specific techniques vary somewhat,
the relative growth of speargrass with and without
depending
on the level of infestation of
mucuna. (This is the only photo of green manures
used in this book that was not taken on the field of
imperata grass. Usually the imperata grass
a farmer who had already adopted the use of green
must be burned. Mucuna is then planted.
manures.)
The imperata grass may need to be cut
once more to let the mucuna develop
sufficiently to smother it. The imperata grass will die after being shaded four or five
months. Often the last 5% of the imperata grass must be eliminated by hand. The
last I knew, 14,000 farmers were using this system, which was largely introduced by
IITA (International Institute of Tropical Agriculture) in Nigeria.25 Sometimes tithonia
(Tithonia diversifolia) is used in a similar manner to control imperata grass. Leucaena
and mother of cacao can also be used, but these take three or four years to shade out
the weed.
Cameroon
S60. Tephrosia fallow. Tephrosia vogelii is used as an improved fallow near the town
of Bamenda. A year or two after the tephrosia is planted, it is cut down and cropping
begins again. The fertilizing effect of this one-year improved fallow equals the impact
of between two and four years under a natural fallow (Photo 43). One farmer started
doing this in the late 1990s; eight years later the practice had spread spontaneously
25
Okon, Paul B. and Uche C. Amalu, Rehabilitation of Degraded Tropical Farms Using Weed to Fight Weed, n.d.
Unpublished.
to over one thousand neighboring farms,

without the support of any outside institution.26 A very similar system is used in
parts of Vietnam by thousands of farmers.
Ethiopia
S61. Teff/grasspea. Grasspea (Lathyrus
sativa) is planted in a rotation with teff
43. Cameroon. An improved fallow using tephrosia
and other crops. Traditional grasspea
near Bamenda, Cameroon.
varieties should not be eaten, although
ICARDA (the International Center for
Agricultural Research in the Dry Areas) has now developed low-toxin varieties that are
safe for human consumption. Tens of thousands of farmers practice this traditional
system.
Rwanda
S62. Other crops/buckwheat. Buckwheat (Fagopyrum esculentum) is rotated with a
number of other crops throughout much of Rwanda. Tens of thousands of farmers use
this traditional system.
Uganda
S63. Various crops/common peas. In southwestern Uganda, common peas are grown
with various highland crops to maintain fertility, either in a rotation or as contour
hedgerows on steep hillsides. Thousands of farmers use this traditional system.
Tanzania
S64. Maize/sunnhemp. Sunnhemp (Crotalaria ochroleuca) is intercropped with maize,
both to fertilize the soil and to provide a very good pesticide to control insects during
grain storage. The sunnhemp is broadcast at a rate of about 30 kg/ha of seed, mixing
one part sunnhemp seed with two parts sand to get a fairly even distribution of the
seed. The sunnhemp is planted at the same time as the maize. Tens of thousands of
farmers use this introduced system.27
26
Anagho, Richard, Green Manure for Improved Fallows in the Northwest Province of Cameroon. Paper delivered at a
Follow-Up Seminar Workshop Eco-Farming in Africa, on November 10, 2000.
27
Kullaya, I. K., et al., Towards Improving Soil Productivity by Sunnhemp (Crotolaria ochroleuca) in the Highlands of
Kilimanjaro in Northern Tanzania, in Norwegian Journal of Agricultural Sciences, No. 21, pp. 99-106.
71
72
Asian Green Manure/Cover Crop Systems

Bangladesh
S65. Various crops/lablab bean. Land-hungry Bengalis often plant a few seeds of
lablab beans (known there as hyacinth beans) right next to their houses, so the bean
will grow up and cover the roof of the house. The seeds are eaten as a pulse, and the
greenery is spread across their fields as fertilizer. Probably hundreds of thousands of
Bengalis use this traditional system.
Thailand
S66. Citrus/mimosa. Spineless Mimosa species are grown near the town of Chiang
Mai as a cover crop under citrus trees. Probably only 50 to 100 people use this
introduced system.
S67. Citrus/centrosema. Centrosema pubescens is grown under citrus trees near Chiang
Mai as a perennial gm/cc, as well as for animal fodder. Perhaps a hundred farmers use
this system.
S68. Citrus/cowpeas. Cowpeas are also grown as a cover crop under citrus. Very few
farmers are as yet using this introduced system.
S69. Maize/rice bean-2. Rice bean is relayed into maize fields a month or two before
the maize is harvested. Then the rice bean continues to grow throughout much of the
dry season. The rice bean is grown partly to fertilize the soil, partly to produce the
rice beans as a cash crop, and partly to keep the soil covered during the dry season for
better weed suppression and protection of the soil. The system also reduces the loss of
soil nitrogen and, in many cases, allows farmers to use zero tillage (reducing costs once
again).28 Two or three hundred farmers use this system that was recently developed by
farmers.
S70. Maize/cowpea-2. The cowpea is relayed into maize, as is done with the rice bean
in S69. Most of the same farmers who use S69 also use this one, just alternating
between one legume and another in different years.
S71. Maize/lablab bean-3. The lablab bean is relayed into maize, just as it is done with
rice bean in S69 (Photo 28). Two or three hundred farmers use this system.
Laos
S72. Upland rice/pigeon pea. Pigeon pea (Cajanus cajan) is planted at the same time
as upland (non-irrigated) rice, at a rate of about one seed every 1.5 to 2 m in each
direction. Thousands of farmers use this traditional practice in northern Laos.
Prinz, Klaus and Somchai Ongprasert, Relay Cropping as an Improved Fallow in Northern Thailand, Unpublished.
28
S73. Upland rice/cowpea. A variety of creeping cowpea found only rarely in Asia (I
have seen it only in northern Laos), and also in some places in the Sahel, is planted
quite sparsely among the rice. Some farmers mix cowpea seeds with rice seed at a rate
of only about one part of cowpea seed to 200 parts of rice seed before broadcasting
the mixture. Thousands of farmers use this practice.
Cambodia
S74. Various crops/cowpea. In southern Cambodia, cowpea is relayed into or planted
after rice or vegetables, to grow during the dry season, much like in S70. Perhaps
several thousand people use this traditional system.
S75. Various crops/jackbean. The jackbean is used in a manner similar to that of the
cowpea in S69. Probably hundreds of farmers use this traditional system.
S76. Rice/sesbania. In a traditional green manure system, Sesbania rostrata is grown at
the beginning of the rainy season in rice paddies to fertilize them for the subsequent
rice crop. The sesbania is incorporated into the soil after it has grown just a month
or so. Then rice is planted, as is done in traditional green manure systems from
temperate climates. This practice is used sporadically in many parts of Southeast Asia
and Sri Lanka. Tens of thousands of farmers use it.
Vietnam
S77. Many crops/mucuna. In northern Vietnam, a perennial mucuna species is sometimes used to recuperate wastelands. It is planted at the beginning of the rainy season
on a piece of wasteland, and then allowed to grow until the land is judged fertile and
largely weed-free. Thousands of farmers use this traditional system.
S78. Rice/rice bean. Rice bean is frequently grown immediately after the paddy rice
harvest, during seasons when the rainfall is insufficient for rice. In fact, farmers in
Vietnam say rice bean has that name because it is grown after rice. However, in most
of the world, farmers think it is called rice bean because of the shape and very
small size of the grain. Hundreds of thousands of farmers (perhaps millions) use this
traditional system in northern Vietnam.
S79. Rice/mungbean. Mungbean (Vigna radiata, known in Vietnam as green bean
and in India as green gram) is planted after rice, just as in S78. Many thousands
of farmers use this traditional system. A similar system was also used traditionally in
much of Indonesia.
S80. Rice/cowpea. Cowpeas are used in the same way as the rice bean in S78 (Photo
44). Tens of thousands of farmers use this traditional system in northern Vietnam. A
similar system was traditional in Indonesia, but largely died out when the national
government began subsidizing chemical fertilizers. Once the subsidies were ended, the
gm/cc system was revived in some areas, such as in southern Sumatra.
73
74
S81. Rice/azolla. Azolla is an algae that

grows in rice paddy water and fixes up to
30 kg of nitrogen per hectare. It can be
spread from paddy to paddy to fertilize
the rice virtually for free.
44. Indonesia. Cowpeas and other edible gm/

cc species often grow well on paddy fields during
seasons when there is not enough water in the
rivers to irrigate paddy rice.
S82. Various crops/tephrosia. Tephrosia

vogelii is planted as a contour barrier in
mountainous areas to recuperate wastelands. This is done in various provinces of
northern Vietnam, especially Son La and
Thai Nguyen Provinces. Hundreds or
thousands of farmers use this introduced
system.
S83. Various crops/yam bean. Yam bean (Pachyrhizus erosus) is used in southern
Vietnam, both as a gm/cc in rotation with other crops, and as a home-made insecticide. This traditional system is practiced widely by an unknown number of farmers.
S84. Various crops/soybeans. Throughout Vietnam, soybeans (Glycine max) are
rotated with other crops to improve soil fertility and to produce a valued food.
Hundreds of thousands (if not millions) of farmers use this traditional system.
S85. Maize/mungbeans. Mungbeans are intercropped in maize fields in northern
Vietnam. Both crops are planted at the same time. Thousands of farmers use this
traditional system.
S86. Several crops/jackbean. The jackbean is occasionally grown on wastelands in
Yen Bai Province to recuperate the areas for the cultivation of many other crops.
Jackbean seeds are broadcast on the land a week or so before the beginning of the
rainy season, at about two or three seeds/square metre. Where weeds are particularly
aggressive, you may need to cut the weeds back once, in order to allow the jackbean to
grow vigorously. In such cases, planting the seeds (placing them about two cm under
the soil surface) might also solve the problem, because planted seeds will germinate
about two weeks sooner than if they are broadcast. The number of farmers who use
this traditional technology is probably well into the thousands.
S87. Several crops/Indigophera spp. Indigophera trees are used like the jackbean
in S86, except the tree is usually used where farmers expect to spend several years
recuperating the wasteland. Again, the number of farmers who use this traditional
technology is probably in the thousands.
S88. Cassava/rice bean. Bushy-type varieties of rice bean are intercropped with cassava. When the rice bean starts to compete with the cassava, the rice bean plants are
pruned down to between 1 and 1.5 m in height, forcing the plant to grow laterally
rather than vertically. Hundreds, if not thousands, of farmers use this traditional system.
S89. Nutgrass control with mucuna. Nut-grass (Cyperus rotundus), one of the worlds
most noxious weeds, is almost impossible to eliminate from a heavily infested field.
However, farmers in parts of Vietnam and Honduras know that nutgrass can be
eliminated by planting mucuna densely in an infested field and allowing the mucuna
to shade it for six months.
The Philippines
S90. Several crops/Stylosanthes. Stylosanthes spp. are used in rotation with several
local crops on Leyte Island. Hundreds of farmers were using this introduced system
when I was there.
S91. Vegetables/lablab beans. Many Filipinos plant lablab beans to climb up the
fences of their home gardens. They do this primarily because they have a soft-podded
lablab variety; they eat the pod like a snowpea. They also plant lablab to fertilize the
soil in their gardens. Tens of thousands of households on several of the islands use this
traditional system.
75
Annex 1: Glossary of Agricultural Terms

Agrochemicals: Chemicals that are used in agriculture. These include chemical fertilizers and chemical-based pesticides, such as insecticides, fungicides and herbicides
Annuals: Plants that die within one year or less of germinating
Biodiversity: Refers to an environment containing a wide variety of different species
living together
Dispersed shade: A condition of reduced sunlight in a field, produced by evenly
spaced trees that produce fairly light shadows
Exotic plant: A plant that is not native to a given region
Fallow: The practice of temporarily not planting a field for several years, usually for
the purpose of allowing natural vegetation to grow back and return the land to its
natural state of fertility
Global warming: The gradual over-all increase of the worlds temperature that is being
caused at least in part by peoples use of fossil fuels, by the destruction of the worlds
forests and other natural environments and by the decreasing amount of organic matter in agricultural soils
Grain legumes: Leguminous plants (plants with seeds that grow inside an elongated
pod) whose seeds are eaten by humans
Green manure/cover crop, or gm/cc: A species of plant, usually a legume, whether
it is a tree, a bush, a vine, a crawling plant or an algae, which is planted by farmers
to maintain or improve their soil fertility or to control weeds. Farmers may also have
many other reasons for growing these plants
Hedgerow: A line of plants grown across a field to protect it from erosion or to provide some other agricultural or environmental benefit
Inoculant: Microorganisms (extremely small microscopic plants or animals) used to
increase the production of nitrogen on the roots or stems of certain legumes
Intercrop: A crop grown simultaneously amongst another crop or crops, planted
within a month or two of the latter
Invasive species: Plants that have gradually propagated themselves in areas where
they were not planted or did not previously exist, and are damaging other human
production activities or the environment
77
78
Legume: A plant with seeds that grow inside an elongated pod. These plants are
particularly important in agriculture since, through a natural process, they take
nitrogen out of the air and make it available to other plants
Monocrop: A crop that is grown alone in a field at a given time
Nitrogen: A crop nutrient that is probably the greatest limiting factor to soil productivity around the world. It is becoming a more and more expensive part of chemical
fertilizer because of the increasing price of energy around the world
Opportunity cost: The money or time you lose (that is, the cost) by choosing one
option that eliminates the possibility (the opportunity) of taking advantage of
another option
Organic matter: Anything that, in its previous form, was part of a living organism,
such as parts of dead plants, bodies of dead animals, urine and manure
Perennial: A plant whose natural lifetime is two years or more
pH: A measure of acidity or alkalinity. A soils pH will affect almost all plants growth,
especially if the soil is either extremely acidic or extremely alkaline
Relay crop: A crop that is planted in a field where another crop is growing within a
month or two of the initial crops being harvested
Rhizobium: A microorganism that often grows on the roots of legumes and fixes
nitrogen. It takes nitrogen out of the air and puts it in the soil in a form that plants
can access
Rotation: A cropping system in which one crop is followed by another, or several
others, in a systematic way. The crop sequence usually being designed to maintain soil
fertility and reduce insect pests and diseases
Synchronization: The timing of the application of nutrients to the soil so they will
provide the amount and kind of nutrients a crop will need at any given stage of its
growth
Annex 2: The Evidence

Introduction
Some readers will feel that many statements in this book have been made with insufficient back-up. For a book of this length, the number of footnotes is meagre, and
virtually none of them refer to a peer-reviewed article.
The fact of the matter is that most of the English-speaking scientific community has
been very slow to do research on green manure/cover crops. The scientists who have
done a fair amount of research on these crops are mostly Latin American, especially
Brazilian. Their work on the subject is magnificent, but, of course, they write the
results of their research in Portuguese and Spanish, sometimes in journals that few
English speakers would have access to, even if he or she is able to read Spanish or
Portuguese.
For those who do read Spanish and/or Portuguese, I would heartily recommend the
following three books, more or less in descending order of priority:
Monegat, Claudino, Plantas de Cobertura del Suelo: Caractersticas y Manejo en
Pequeas Propiedades (Tegucigalpa, Honduras: CIDICCO, 1997). In Spanish and
Portuguese. The original in Portuguese was printed in Chapeco, Santa Catarina,
Brazil, and copyrighted in 1991. The Spanish edition is probably most easily obtained
from CIDICCO in Honduras.
Calegari, Ademir, et al., Adubacao Verde no Sul do Brasil (Rio de Janeiro, Assessoria e
Servicos a Projetosem Agricultura Alternativa, 1993). In Portuguese only. AS-PTA in
Rio de Janeiro would probably be the best source.
Binder, Ulrike, Manual de Leguminosas de Nicaragua, Volume I (Esteli, Nicaragua:
PASOLAC and the Escuela de Agricultura y Ganaderia de Esteli, 1997). In Spanish
only.
Research notwithstanding, the vast majority of what we know today about gm/ccs
has resulted from experimentation by smallholder farmers. Some of these experiments
were done through a process of action research, or participatory technology development (PTD). Others were merely informal, trial-and-error experiments that smallholder farmers have done for millennia in their fields. These practical processes are
admittedly not as carefully quantitative as scientific experiments, and often there are
problems of experimental design. However, they have been done under smallholder
farmers conditions. They reflect smallholder farmers priorities and they have been
conducted within the economic, social and climatic limitations under which smallholder farmers must work. Each kind of experimentation, whether it be scientific,
participatory or totally informal, has its advantages and disadvantages.
79
80
For example, smallholder farmers never do 15 replications of exactly the same experiment on one small piece of land. But if a technology is to the liking of the farmers,
we may have hundreds of replications of very similar experiments, spread across
dozens of soil types, topographies and different (though fairly similar) management
styles. Mathematically speaking, these trials will give us a much better approximation
than will a scientific experiment of what will happen when other smallholder farmers
replicate that technology on their own farms. This better predictability from farmers
experiments occurs because a scientific experiment will normally include dozens of
replications on fairly uniform micro-plots, carried out on one soil type (which usually
has a bizarre history of cultivation under past experiments), on flat land (hillside
experimental stations are virtually nonexistent) and under only one standardized,
and often very expensive, form of management. Each of these conditions makes the
scientific experiment less representative of the farmers conditions than do the farmers
experiments.
Despite the informal nature of their experimentation, the varied trials smallholder
farmers do can be analyzed mathematically to find out to what precise level of confidence we can generalize their results across a wider population of farmers.
I have not done any such analysis. I have visited the fields of thousands of farmers and
seen the results. This book is a summary of what I have learned. The vast majority of
this books content is based on those observations of individual fields and on conversations I have had with the owners, plus conversations with dozens of the owners
neighbors and collaborators. It also includes observations of tens of thousands of other
fields from moving vehicles.
For agricultural extension purposes, most agricultural research in the world is done
with the purpose of trying to predict what technologies farmers will find beneficial,
and will therefore adopt. When it comes to gm/ccs, we do not need to predict what
systems farmers will adopt in the future. We already know what technologies they
have adopted in the past. The only predictions needed are whether or not other
farmers working in similar conditions will act in the same way. This prediction is best
tested by showing other farmers the results of the technology, and then watching what
they do or dont do with it.
If we believe that farmers are, by and large, economically rational (given their own
priorities), we dont need to be guessing about whether the gm/cc technologies in this
book will be beneficial to smallholder farmers under the conditions in which they are
being used. The very fact of widespread farmer use, given the assumption of farmer
rationality, leads us to conclude that these technologies are, in fact, benefiting the
farmers.
Still, some readers may have questions about issues that are more in the realm of the
theory of soil fertility rather than farmer adoption of technology. I will refer to a few
of those questions and issues here.
The Amount of Organic Matter Required to Maintain or Improve

Soil Fertility
First, I have purposely not given a recommendation for the exact quantity of
organic matter that will maintain or improve soil fertility. The exact amount of
organic matter needed will depend on a series of factors that vary from one village
to another: the characteristics of the leaves, compost or manure being used; the density of nutrients in that organic matter; the amount of moisture it contains (when
we are using green weight as a measurement); the quality of the soil, the specific factors in the soil that limit productivity; the amount of erosion; the climate; etc. Any
exact figure of an optimum level of organic matter required would vary depending
on the particular situation, taking each of these factors into account.
Where did I get the approximate figure of 20 to 25 MT/ha, green weight, of
leguminous leaves (that is, above and beyond the normal crop residues)? It comes
basically from having made an analysis of the impact of dozens of different gm/cc
systems that I have observed around the world. I will present here a much-simplified version of that analysis:
Name and number of system:
Approximate amount
of organic matter produced:
Does it usually maintain

system or improve
soil fertility?*
Systems that tend to increase soil fertility:

Maize/mucuna-1 (S1)
70 t/ha
Improves
Maize/sweet clover (S7)
40+ t/ha
Roughly maintains, but

also feeds grazing animals
Maize/runner bean-2 (S13)
60 t/ha
Improves
Maize/lablab bean-1 (S23)
60 t/ha
Improves
Systems that more or less maintain soil fertility:

Maize/fava beans (S6)
25 t/ha (depending on the

density of the legume)
Maintains
Millet/native dispersed shade/

short-cycle cowpeas
20 t/ha
Roughly maintains
Tephrosia fallow (S60)
60+ t/ha, once every

3 - 4 years
Maintains
Systems that fail to maintain soil fertility (not including those where burn-off is a factor):
Maize/wild sunflower-2 (S21)
15 t/ha (depends on
spacing of hedgerows)
Fertility decreases
Maize/tephrosia (S25)
10 t/ha
Fertility decreases
Potatoes/tarwi (S33)
5-10 t/ha (depends on the

size of the field)
Fertility decreases
Various crops/common peas (S63)
5-15 t/ha
Fertility decreases
*Note: Those systems that improve the soil do so only up until a certain point of diminishing returns, at which
they, too, only maintain a specific level of soil fertility.
81
82
Though the data for some of these systems is very approximate, I think that these and
other gm/cc systems provide fairly strong evidence supporting the 20 to 25 t/ha figure
given. Furthermore, I have never seen a case where a gm/cc flagrantly violated this
rule, regardless of the species of gm/cc or management of it.
The source of our knowledge on the impact of dispersed shade?

Among smallholder farmers, the use of dispersed shade is traditional and surprisingly
widespread. Professionals, however, have only recently begun paying any attention to
the technology.
Probably the best study ever made of the traditional use of dispersed shade was done
by Malcolm Cairns, who worked in Southeast Asia with the International Center for
Research in Agroforestry. He found dozens of such systems spread from northeastern India to Indonesia. I have seen traditional systems in a dozen countries of Latin
America (most notably in Mexico, El Salvador, Guatemala, Honduras, Nicaragua
and Brazil). I am told that the most notable case in West Africa, the millet/Faidherbia
albida system (S57), is used from Senegal to Niger, although I have only personally
seen it in three of those countries.
Scientific research on dispersed shade was conducted for a number of years by Dr. Ian
Cherret and Ing. Luis Martinez in Honduras for the United Nations Food and Agriculture Organization (FAO). Much of the theoretical and quantitative information I
presented in earlier chapters came from conversations held with Ian and Luis and their
team during the years I lived in Honduras.
The effect of organic matter on phosphorus availability

Virtually all farmers, including those in the United States, know that phosphorus
availability is greatly enhanced if the phosphorus is not actually in contact with the
soil. This is why chemical fertilizers are commonly banded into the rows of crops
rather than broadcast or spread more evenly across the soil.
In tropical soils where extremes of pH are much more common, the tying up of phosphorus in contact with the soil is even more important. Most literature agrees that
very acidic soils (with a pH of less than about 5.0) will have virtually no phosphorus
that is available to the plants, even though reasonable levels of phosphorus are present
in the soil.
What is happening in developing world soils that have more moderate pHs? First, lets
admit that we dont really know for sure. We do know that the amount of total phosphorus in a soil that is actually available to crops can range from 0.5% to 2.0%, and
that much of this variation depends on the level of organic matter in the soil. Though
an increase in availability from 0.5% to 1% of the total soil phosphorus doesnt sound
like much, it represents a 100% increase in the phosphorus available to crops. My
observations of soil phosphorus levels in the developing world lead me to believe that
this level of increase is quite common when we use gm/ccs.

The one case I have followed most closely is the maize/mucuna system in northern
Honduras (S1). It is a very interesting case, and quite well documented (see, for
instance, Bernard Trombleys Cornell University doctoral dissertation). In this case,
thousands of farmers working on poor tropical soils with a parent material low in
phosphorus have been planting maize and mucuna on the same fields every year for
40 to 45 years. Most of these farmers have never applied chemical phosphorus to the
soil. Yet, after 40 or more years of harvesting maize at rates of 1.5 to 2.5 t/ha, they
still are having no problem with phosphorus availability in the soils. In fact, Bernard
found no yield response to adding phosphorus to these soils.
In most of my travels across more than 45 developing countries, I have observed
gm/cc systems in use. Yields have doubled or tripled in case after case. Yet I have
never observed a serious phosphorus deficiency, even where farmers have applied no
phosphorus for decades. The northern Honduras phenomenon is not an exception. It
is well within the normal range of cases.
So what is happening? How can farmers take significant levels of phosphorus every
year out of a soil that was never particularly rich in it anyway, and do so without running out of phosphorus?
One answer, of course, is that the incredible amounts of organic matter added to the
soil have increased the available phosphorus by 100%, or maybe even several hundred
percent. That would explain a good part of the mystery, but not all of it.
I think the problem may well lie with one inherent assumption that undergirds
everything that has been written above: the assumption that these farmers fields are
closed systems. Most agronomists dont even realize they are constantly making this
assumption. Up until now in our discussion, we have been assuming that the only
phosphorus that could possibly exist in each farmers field is the phosphorus that was
there 40 years ago, when they began using mucuna.
But this assumption is wrong. It is, in fact, very wrong. I once visited a scientific
research station in the Brazilian Amazon near the city of Belen. Farmers there were
cutting down virgin forest, largely because of the weed pressure on all land that had
been farmed for more than a few years. Scientists found that birds and bats deposited
an average of 10,000 viable weed seeds on each hectare of new farm land per year.
How much phosphorus would be deposited in the guano along with those seeds? We
dont know, but the amount would be significant, especially given that a large majority
of bird species are carnivorous and eat insects rather than seeds. In addition to those
depositions, phosphorus would be spread from forest to farm and from farm to farm
by microorganisms, insects, earthworms, wild animals, domestic animals, etc.
This transfer of nutrients does not only happen near the edges of virgin forests. Across
the Sahel, far larger transfers of phosphorus occur over larger areas because of the
83
84
seasonal winds called the harmattan. Thousands of tons of topsoil are carried across
the southern perimeter of the Sahara desert. To understand how much soil is moved,
you only need to look at the windward side of many homes on the windward side of
Sahelian villages. The build-up of topsoil against those walls is frequently over a metre
high, and occasionally reaches the eaves of the house. In much of the Sahel, every lowlying bush that has leaves year-round (including the nyama) has around 40 cm of soil
built up around its base. This represents a major transfer of nutrients from one field to
another, and a major influx of phosphorus to fields in which farmers are using nyama
or other bushes to fertilize their soil.
With these dynamics as background, lets return to the original question of the
effect of organic matter on the availability of phosphorus. Our present methods of
soil analysis do not give us a very good answer. Some scientists who have looked at
phosphorus levels in organic systems have admitted they cannot explain where all the
phosphorus is coming from (Cheryl Palm is one of them).
CGIAR (Consultative Group on International Agricultural Research) system scientists
have been telling us for over a decade that Africa is heading for a major crisis of soil
phosphorus deficiency. They have not been clear about exactly when the problem
would hit, but certainly sometime within the 25 years following the start of their
predictions. They normally give two major reasons for decreasing levels of available
phosphorus. One is the gradually decreasing levels of available phosphorus they have
found in the soil analyses theyve done. The second is a calculation based on crop
yields, from which they have calculated the annual loss to the soil of various nutrients.
But there is reason to believe that somehow more phosphorus is available to plants
than scientific studies would indicate. The CGIAR calculation based on crop yields
may be based on a major misunderstanding of the dynamics of smallholder agriculture
in Africa. In the United States, calculating nutrient losses on the basis of crop yields
can be quite accurate because American farmers consume virtually nothing of what
they harvest, and even what they do consume goes into flush toilets and far away from
their farms. However, the vast majority of African smallholder farmers produce only
enough food for 6 to 10 months of consumption. That is, most African villages are
net importers of food, and therefore net importers of phosphorus. Their food waste
all goes either onto the soil surface or into hand-dug latrines that go nowhere near as
deep as some of the roots of nearby bushes and trees. So quite likely, the total amount
of phosphorus recycled into peoples fields is even greater than the amount that is
taken out at harvest time.
That leaves soil analyses of available phosphorus as the main credible piece of evidence
for the predicted African soil phosphorus crisis. Remember, scientists point to gradually decreasing levels of available phosphorus in soil analyses. But do decreased levels
of available phosphorus correlate well with decreased levels of the total phosphorus in
the soil?
Soil analyses, by and large, do not give us a very accurate assessment of the phosphorus
content of a soil. First, they do not measure the phosphorus that is tied up chemically
and therefore is unavailable to plants. Secondly, they do not measure the phosphorus
that exists in organic forms, whether that be in decomposing plant material or in soil
fauna and flora. Furthermore, these inaccuracies may be quite large, leaving out a
large majority of the total phosphorus present in the soil.
During a study I did in Africa in late 2009 for the Christian Reformed World Relief
Committee, I traveled through significant areas of Malawi, Zambia, Kenya, Uganda,
Niger and Mali, interviewing farmers wherever I visited. When doing such studies,
I observe all the agricultural practices and conditions that I can as I travel along the
highways and dirt roads in order to triangulate that information with what I see in
smallholder farmers fields and hear from the farmers. I saw tremendous evidence
that yields were diminishing, primarily because of nutrient deficiency. But nitrogen
was clearly most deficient in at least 95% (probably closer to 98%) of the tens of
thousands of fields that I observed. Nitrogen deficiency is due to a lack of organic
matter in the soil. All over Africa, farmers are mining the nitrogen out of their soils.
Wherever one goes, one can observe the yellowish, stunted maize and millet fields that
indicate a nitrogen deficiency. On the other hand, the purple leaves and stems associated with phosphorus deficiency were visible on less than 5%, and probably closer to
2%, of the fields.
Something is obviously wrong with the CGIARs prediction of a major phosphorusbased crisis within the next decade or so. But why such a striking lack of phosphorus
deficiency symptoms? Deficiency symptoms in most grasses like maize and millet
tend to indicate only the single element that is most limiting. That is, if a nitrogen
deficiency is more limiting to the plants growth than a phosphorus deficiency, you
will see only the symptoms of nitrogen deficiency on the leaves or stems of the plants.
So my observation of fields around Africa did not indicate that there was no phosphorus deficiency. Rather, it indicated that in a widespread, consistent pattern, the soils
phosphorus deficiency (if present) was masked by a worse nitrogen deficiency.
This consistent masking of whatever phosphorus deficiency there wasacross a whole
continentleads me to believe there must be some relationship between the available
amounts of the two nutrients, causing the phosphorus deficiency to be consistently
less limiting than the nitrogen one.
If the original amount of phosphorus in the soil and subsoil minus the amount taken
out by African farmers over the years were the main factor in determining the amount
of available phosphorus, there would be no relationship between that quantity of
phosphorus and the quantity of nitrogen in the soil. There would, therefore, tend to
be many fields in Africa with the purplish signs of phosphorus deficiency. However,
if the amount of organic matter in the soil were the main determinant of how much
phosphorus was available to plants, then a very clear relationship would exist. In the
absence of chemical fertilizer, the amount of nitrogen in a soil correlates quite closely
with the amount of organic matter, which in turn would correlate quite closely with
the amount of available phosphorus in the soil.
85
86
To check out this theory, I asked about a dozen owners of the few fields with significant phosphorus deficiency symptoms whether they had used any chemical fertilizer
in the last year. In every single case (except one where worms had obviously done
major damage to the maizes roots), the farmers reported applying urea, and only urea,
to those fields. Application of urea would mean that nitrogen was plentiful. Therefore,
in the absence of a complete formula of fertilizer that also contained phosphorus,
phosphorus would become the limiting factor for the crops.
Thus, I propose that there is major evidence that a lack of organic matter is the main
reason African soils are deficient in available phosphorus, and that most evidence of
rapidly decreasing phosphorus availability is due not to a lack of over-all phosphorus,
but to a lack of organic matter to make those reserves of phosphorus available to farmers crops. As we increase the amount of organic matter in the soils, as we can do with
gm/ccs, the amount of available phosphorus is also going to increase substantially, as
has happened in northern Honduras.
The logical conclusion would seem to be that the organic matter supplied by gm/ccs
significantly raises the availability of phosphorus in the soil, making additional applications of chemical phosphorus unnecessary, at least in the short to medium term.
Of course, nothing can come from nothing. Where will all that phosphorus, that is
slowly becoming available, come from over the long run? If my thesis about the net
inflow of phosphorus into most African smallholders fields is correct, and maintains
itself, the whole theory that total phosphorus is diminishing is mistaken, and Africas
smallholder farmers may never have to worry about the so-called phosphorus deficit.
If my thesis is wrong, or if it is correct and African smallholder farmers increase their
productivity to the point of being net exporters of phosphorus, African farmers will
need to start buying replacement phosphorus for their fieldsbut probably not until
some 20 to 40 years from now, given the supplies of phosphorus from the harmattan,
birds, bats, termites, etc.. By that time, hopefully, they will have advanced enough,
and their harvests, or world food prices, will be high enough, that they will be able to
afford it.
Annex 3: List of Recommended Green Manure/Cover Crop Species
Annex 3: List of Recommended Green Manure/Cover

Crop Species
Note: A number immediately following the common name of the species
indicates the box of the decision tree in which there is a general description of
this species.
Common name
Scientific name
Gm/cc
Systems
Countries where systems

used
Azolla
Azolla pinnata
S81
Vietnam
Bambara groundnut
Voandazeia
subterranean
S58
Mali
Buckwheat
Fagopyrum esculentum
S62
Rwanda
Calopogonium
Calopogonium
mucunoides
S55
Ghana
Black bean (see lablab bean)

Broadbean (see fava bean)
Centrosema
Centrosema pubescens
S67
Thailand
Choreque
Lathyrus nigrivalvis
S15
Guatemala
Common pea
Pisum sativa
S36
S49
S63
Peru
Brazil
Uganda
Cowpea (#38)
Vigna unguiculata
S22
Honduras, Guatemala,
El Salvador, Nicaragua, Panama
Mali
Thailand
Thailand
Laos
Cambodia
Vietnam, Indonesia
S58
S68
S70
S73
S74
S80
Crotalaria
Crotalaria spp.
S44 to 46
Brazil
Desmodium
Desmodium ovalifolium
S27
Honduras, Belize
Faidherbia (#25)
Faidherbia albida,
formerly Acacia albida
S57
Mali, Burkina Faso, Niger
S58
Fava bean
Vicia faba
S6
Mexico, Guatemala, Vietnam
S35
Peru
S61
Ethiopia
Forage peanut (see perennial

peanut)
Grasspea
Green bean (see mungbean)
Green gram (see mungbean)
Lathyrus sativa
87
88
Groundnut (see peanut)

Horsepea
(see lablab bean)
Hyacinth bean
(see lablab bean)
Indigophera
Indigophera spp.
S87
Vietnam
Jackbean (#29)
Canavalia ensiformis
S3
S10
S11
S29
S38
S42
S75
S86
Mexico, Brazil
Mexico
Mexico
Costa Rica
Paraguay, Honduras
Paraguay, Honduras
Cambodia
Vietnam, Nicaragua
Lablab bean (#37)
Dolichos lablab
S23
S24
S65
S71
S91
Honduras, Peru
Honduras
Bangladesh
Thailand
Philippines
Leucaena (#26)
Leucaena leucocephala
and L. diversifolia
S56
Ghana
Lima bean
Phaseolus lunatus
S4
Mexico, Guatemala, Honduras
Lupines
Lupinus albus, L. luteus

and L. angustifolius
S39 and
S40
Paraguay and Brazil
Mimosa
(see spineless mimosa)
Mother of cacao (#26)
Gliricidia sepium
S58
Mali
Mucuna (#32)
Mucuna spp.
S1
S17
S18
S19
S31
S37
S41
S47
S48
S53
S59
S77
S89
Mexico, Guatemala, Belize,

Honduras
Mexico, Costa Rica
Mexico
Guatemala, Nicaragua, Brazil,
Paraguay
Honduras
Honduras
Honduras
Panama
Paraguay, Mexico
Paraguay
Brazil
Brazil
Ghana, Benin, Mexico
Benin
Vietnam
Vietnam
S2
S8
S14
Mungbean
Vigna radiata
S79
S85
Vietnam, Indonesia
Vietnam
Nyama (#45)
Piliostigma reticulatum
S49
S50
S51
S52
S54
S58
Burkina Faso
Burkina Faso
Burkina Faso
Burkina Faso, Mali
Ghana
Mali
Pea (see common pea)

Peanut
Arachis hypogea
S43
S58
Paraguay
Mali
Perennial peanut
Arachis pintoi
S26
S28
Honduras, Costa Rica

Costa Rica
Pigeon pea (#26)
Cajanus cajan
S30
S72
Panama, Brazil, Kenya

Laos
Vigna umbellata
S9
Mexico, Guatemala, El Salvador,

Vietnam
Thailand
Vietnam
Vietnam
Quickstick (see mother of

cacao)
Rice bean
S69
S78
S88
Runner bean (#14)
Phaseolus coccineus
S12
S13
Guatemala
Guatemala, United States,
Mexico, Honduras, Colombia,
Ecuador, Peru, Bolivia, Chile
Scarlet runner bean

(see runner bean)
Sesbania
Sesbania rostrata
S76
Cambodia, Sri Lanka, etc.
Soybean
Glycine max
S84
Vietnam
Spineless mimosa
Mimosa spp.
S66
Thailand
Stylo
Stylosanthes spp.
S90
Philippines
Swordbean
Canavalia gladiatus
S3
Mexico
Sunnhemp
Crotalaria ochroleuca
S64
Tanzania
Sweet clover (#14)
Melilotus albus
S7
Mexico
Tarwi (#15)
Lupinus mutabilis
S32
S33
S34
Peru, Ecuador, Bolivia

Peru, Ecuador
Peru, Bolivia
Tephrosia (#29)
Tephrosia vogelii and

T. candida
S16
S25
S58
S60
S82
Guatemala
Honduras
Mali
Cameroon
Vietnam
Vetch
Vicia toluca
S5
Mexico
Wild sunflower
Tithonia diversifolia
S20
S21
Honduras
Honduras
Yam bean
Pachyrhizus erosus
S83
Vietnam
Tithonia
(see wild sunflower)
Velvetbean
(see mucuna)
89
90
Annex 4: Seed Sources for Green Manure/Cover Crops
Annex 4: Seed Sources for Green Manure/Cover

Crops
In many cases, seeds can be found in-country; look for seeds in village markets,
NGOs with agriculture-related projects, and research institutions (e.g., agricultural
colleges and universities). If seeds of a particular crop or variety cannot be found in
the country or region where you are working, it is possible to order seeds from other
countries.
When transferring seeds across borders, you will likely need the following documents:
1) a seed import permit from the country that the seeds will be shipped to; 2) a
phytosanitary certificate from the country of origin (often the shipper/seed supplier
can supply this) stating that the seeds are free of pests and diseases. Check with the
Ministry of Agriculture in the country you work in to find out details regarding documents needed and how to obtain them. Typically, one would begin by obtaining the
import permit, as this document could contain specific seed treatments required for
entry of particular crops. One would then send the import permit to the seed supplier
in the country of origin. Based on requirements outlined in the import permit, the
seed supplier can carry out any specific seed treatments, list those in the phytosanitary certificate, and include both documents (originals may be required) in the box
of seeds to be shipped. Request that shipments be sent via a reputable carrier such
as DHL. Be aware that extra costs can be incurred if it becomes necessary to hire a
separate courier to clear seeds through customs once they arrive in your country.
Below are a number of seed suppliers that carry seeds of gm/ccs:
Abundant Life Seed Foundation
PO Box 279
Cottage Grove, OR 97424-0010 USA
www.abundantlifeseeds.com
Tel: (541) 767-9606
Fax: (866) 514-7333
Source of clover and hairy vetch
B & T World Seeds
Paguignan
34210 Aigues-Vives
France
www.b-and-t-world-seeds.com
Tel: 00 33 (0)4 68 91 29 63
Fax: 00 33 (0)4 68 91 30 39
Numerous species and varieties in stock; allow up to 8 weeks for seeds not in stock.
Baker Creek Heirloom Seeds

2278 Baker Creek Road
Mansfield, MO 65704 USA
www.rareseeds.com
Tel: (417) 924-8917
Fax: (417) 924-8887
Email: [email protected]
Source for a number of cowpea varieties
Banana Tree
715 Northampton St.
Easton, PA 18042 USA
www.banana-tree.com
Tel: (610) 253-9589
Fax: (610) 253-4864
Source for jackbean
CIAT
International Center for Tropical Agriculture
Km 17, Recta Cali-Palmira
Apartado Areo 6713
Cali, Colombia
Tel: +57 2 4450000
Fax: +57 2 4450073
E-mail: [email protected]
www.ciat.cgiar.org
URL for regional office contact info:
www.ciat.cgiar.org/AboutUs/Paginas/contact_us.aspx
CIDICCO
International Center for Information on Cover Crops
Apdo. Postal 4443,
Tegucigalpa, Honduras
Tel: +504 239 5851 +504 232 385
www.cidicco.hn
A CGIAR (Consultative Group on International Research) research institution with
an extensive collection bean and forage lines. URL for information on how to request
seeds: http://isa.ciat.cgiar.org/urg/inforequestmaterial.do
ECHO Seed Bank
17391 Durrance Road
North Fort Myers, FL 33917
www.ECHOcommunity.org
91
92
Trial packets of a number of gm/cc crops are available to ECHOs network of international
development workers. See the website for information on how to join ECHOs network and
request seeds.
Inland & Foreign Trading Co., LTD
Block 1090, #04-04/05
Lower Delta Road
Tiong Bahru Industrial Estate
Singapore 169201
www.iftco.com.sg
Tel: (65) 2722 711
Fax: (65) 2716 118
Email : [email protected]
See legumes under Legume Cover Crop and Pasture Seed categories. Source of bulk
quantities (50 kg).
Shivalik Seeds Corporation
05, Panditwari, P.O. Prem Nagar
Dehradun, 248007, Uttaranchal
India
www.shivalikseeds.com
TeleFax: +91 135 773348
Source for a number of annual gm/ccs
Setropa B.V.
Troelstralaan 4
Postbox 203
1400 AE Bussum
Holland
www.setropa.nl
Tel: +31 (0)35 5258754
Fax: +31 (0)35 5265424
Source of a number of gm/ccs
Wolf and Wolf Seeds
Rua Paulo Padovan, 81
Ribeiro Preto-SP
Brasil
Email: online form
www.wolfseeds.com
Source of gm/ccs in bulk amounts
Heritage Seeds Pty. Ltd

7-9 McDonalds Lane
PO Box 4020
Mulgrave, Victoria 3170
Australia
www.heritageseeds.com.au
Tel: 03 9501 7000
Fax: 03 9561 9333
Source for several tropical and temperate legumes
If you are aware of other sources of seeds, please communicate that to the moderator
of the Green Manure/Cover Crops discussion group on www.ECHOcommunity.org.
We hope to be able to add sources to the above list over time.
93
94
Annex 5: Additional Resources on Green Manure/Cover Crops
Annex 5: Additional Resources on Green Manure/

Cover Crops
Books
Centro International de Divulgacion sobre los Cultivos de Cobertura (CIDICCO),
et al., Experiencias sobre Abonos Verdes y Cultivos de Cobertura (Tegucigalpa, Honduras:
CIDICCO, 1998).
Eilitt, M., J. Mureithi, and R. Derpsch.Green Manure/Cover Crop Systems of
Smallholder Farmers:Experiences from Tropical and Subtropical Regions (Dordrecht, The
Netherlands: Kluwer Academic Publishers, 2004).
Florentin, M.A., M. Penalva, A. Calegari and R. Derpsch. Green Manure/Cover Crops
and Crop Rotation in Conservation Agriculture on Small Farms (Rome: Food and
Agriculture Organization 2011).
International Development Research Centre (Editors: D. Buckles, A. Eteka, O.
Osiname, M. Galiba, and N. Galiano). Cover Crops in West Africa/Plantes de couverture
en Afrique de lOuest (Sasakawa Global, 1999).
Kay, D.E. Food Legumes (Tropical Products Institute, 1979).
Monegat, C. Plantas de Cobertura del Suelo: Caractersticas y Manejo en Pequeas
Propiedades (Tegucigalpa, Honduras: CIDICCO, 1997).
National Research Council. Tropical Legumes: Resources for the Future (National
Academy Press, Washington, D.C., 1979)
Rinaudo, T. Farmer Managed Natural Regeneration: Exceptional Impact of a Novel
Approach to Reforestation in Sub-Saharan Africa (Technical Note #65, Educational
Concerns for Hunger Organization 2012)
Website: https://echocommunity.site-ym.com/?page=tech_notes
Virtual Community
ECHO Community:
https://echocommunity.site-ym.com/group/green_manure_crops
Roland Bunch has worked in agricultural development for

more than 42 years in more than 50 nations of Latin America,
Africa and Asia. He has done consultancies with the Ford
Foundation, Cornell University, CARE and the top nongovernmental organizations from Canada, Great Britain, the
Netherlands, Germany and Switzerland, as well as the governments of Guatemala, Honduras, Swaziland and Vietnam. In
1982, he published the book, Two Ears of Corn, A Guide to
People-Centered Agricultural Improvement, which has since
been published in ten languages and is an all-time bestseller
in the field of agricultural development. Two Ears of Corn pioneered the ideas of
development programs organizing smallholder farmers to teach each other and
organize experiments, both of which have now become major movements around
the world under the names of farmer-to-farmer extension and participatory
technology development.
Starting in 1983, Roland began investigating the use of plants that are particularly
good at fertilizing the soil, which are now called green manure/cover crops.
Together with an independent group of agronomists in southern Brazil, he has
spearheaded the effort that has successfully put this technology on the agenda of
development organizations around the world.
Roland has been nominated for the Global 500 Award, the End the Hunger Prize
of the President of the United States, and the World Food Prize.
Canadian
Foodgrains
Bank
to Hunger
Canadian Foodgrains Bank is a partnership of 15 churches and

church-based agencies working together to end global hunger.
foodgrainsbank.ca 1.800.665.0377
$12.95

Restoring The Soil

Uploaded by

Copyright:

Available Formats

Restoring The Soil

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Restoring The Soil

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What experience does the author have?

What experience does the author have?

Restoring the Soil

A Guide for Using Green Manure/Cover Crops to Improve

Praise for Restoring the Soil