To feed ourselves

Material Information

To feed ourselves a proceedings of the First Eastern, Central, and Southern Africa Regional Maize Workshop, Lusaka, Zambia, March 10-17, 1985
International Maize and Wheat Improvement Center
Eastern, Central, and Southern Africa Regional Maize Workshop, 1985
Place of Publication:
Mexico D.F. Mexico
International Maize and Wheat Improvement Center
Publication Date:
Physical Description:
viii, 307 p. : ill. ; 25 cm.


Subjects / Keywords:
Corn -- Congresses -- Africa ( lcsh )
Corn -- Research -- Congresses -- Africa ( lcsh )
Corn as food -- Congresses -- Africa ( lcsh )
Corn ( jstor )
Commercial production ( jstor )
Farmers ( jstor )
bibliography ( marcgt )
non-fiction ( marcgt )
conference publication ( marcgt )
Spatial Coverage:


Includes bibliographies.
Statement of Responsibility:
sponsored by the Government of Zambia and CIMMYT.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.
Resource Identifier:
17298253 ( OCLC )
87102973 ( LCCN )
9686127054 (pbk.) ( ISBN )


This item has the following downloads:






























































































































































































































































































































Full Text





To Feed Ourselves

A Proceedings of the First Eastern, Central
and Southern Africa Regional Maize Workshop
Lusaka, Zambia, March 10-17, 1985
Sponsored by: The Government of Zambia and CIMMYT


I would like to thank the following
groups, organizations and individuals
for contributing to the success of the
First Eastern, Central and Southern
Africa Regional Maize Workshop and to
the publication of the proceedings:

* The Government of the Republic of
Zambia in general, and the Ministry
of Agriculture and Water
Development in particular, for
graciously hosting the workshop and
meeting all local expenses;
* The steering committee, composed
of members of the Cereals Research
Team of the Ministry of Agriculture
and Water Development of Zambia
and chaired by R.K. Chungu,
Assistant Director of Agriculture-
Research, for working tirelessly in
co-organizing the workshop;
* The Zambia-based organizations
that provided assistance in funding
and transport, among others,
Zamseed, Power Equipment, the
EEC, SIDA, Barclays Bank and Shell
* The management of the Lusaka
Intercontinental Hotel for providing
excellent conference facilities, food
and accommodations;
* The International Development
Research Centre (IDRC), the East
African Regional Economic
Development Service of the United
States Agency for International
Development (USAID/REDSO/ESA),
the USAID program in Nairobi,
Kenya, the University of Florida/
USAID/Malawi Government
Agricultural Research Project, the
International Institute of Tropical
Agriculture (IITA) and the
International Maize and Wheat
Improvement Center (CIMMYT) for
their generous financial contribution
in meeting travel and per diem
expenses of the 34 delegates from
17 countries;

* My co-editor Elizabeth Cullar for
her outstanding contribution in the
preparation of conference handouts,
registration of participants, and
above all, for her patience and
perseverance in editing the entire
manuscript of the proceedings;
* Maria Kamau, secretary, CIMMYT,
Nairobi, for capably handling all
correspondence, ticketing and other
logistical support;
* Drs. R.P. Cantrell and R.L. Paliwal,
director and associate director of the
CIMMYT Maize Program, for their
encouragement and continued
assistance from the inception of the
workshop until the end;
* The CIMMYT Communications
Group, especially Nathan Russell for
his advice and encouragement,
Silvia Bistrain, Maricela A. de
Ramos, Patricia Martinez and
M6nica Hernmndez for typesetting
the proceedings, and Miguel
Mellado, Rafael De la Colina, Jos6
Manuel Fouilloux and Bertha
Regalado for design and layout;
* The participants, without whom
there would have been no workshop,
* My wife, Almaz, for her love and
understanding while this work was
going on.

Bantayehu Gelaw
Workshop Organizer

Table of Contents

vii Preface

1 Opening Ceremonies
Welcome to the First Eastern, Central and Southern Africa Regional Maize
Workshop, The Honourable G.K. Chinkulu, MP, Minister of Agriculture and
Water Development, Zambia

6 The Delegates' Response to the Honourable Minister, A.J. Moshi, National Maize
Research Programme. Tanzania

7 Initiation of the First Eastern, Central and Southern Africa Regional Maize
Workshop, B. Gelaw, Workshop Organizer, CIMMYT East African Maize
Program, Nairobi, Kenya

10 Country Reports
Maize Research Activities in Angola, F. Marcelino and M. Girao, Instituto de
Investigacao Agronomica, Chianga, Huambo, Angola

12 Farmer Rejection of Late-Maturing. High-Yielding Maize in Burundi, R.S. Zeigler
and M. Kayibigi, Programme Mais et Petit Pois, Institut des Sciences
Agronomiques du Burundi, Bujumbura, Burundi

20 Maize Research and Production in Ethiopia, A. Debelo, Institute of Agricultural
Research, Awassa, Ethiopia

26 Maize Research in Kenya: An Overview, J.A.W. Ochieng, National Agricultural
Research Station, Kitale, Kenya

32 The Maize Program in Kenya, E.W. Mwenda, Embu Agricultural Research
Station, Embu, Kenya

37 Maize Research in Lesotho, P.P. Ntlhabo, Agricultural Research, Thaba Tseka,
and M.T. Matl, Agricultural Research, Maseru, Lesotho

43 Maize Production and Research in Madagascar, L. Rondro-Harisoa and
R. Ramilison, Ministere de la Recherche Scientifique et Technologique pour le
Developpement, Antananarivo, Madagascar

50 Maize Research and Production in Malawi, L.D.M. Ngwira and E.M. Sibale,
Department of Agricultural Research, Chitedze Agricultural Research Station,
Lilongwe, Malawi

57 Maize Production, Constraints, Research and Development in Mauritius;
N. Govinden, Food Crop Agronomy Division, Mauritius Sugar Industry Research
Institute, and S.P. Mauree, Extension Services, Ministry of Agriculture,
Fisheries and Natural Resources, Reduit, Mauritiusi

67 Research on the Constraints to Maize Production in Mozambique, E. Nunes,
Institute Nacional de Investigacao, Posto Agronomico de Umbeluzi, and D.
Sousa, Posto Agronomico de Lichinga. Mozambique, and I. Sataric, Maize
Research Institute, Zemun Polje, Yugoslavia

80 The Reunion Island Maize Breeding Program, J.L. Marchand and E. Hainzelin,
Institute de Recherches Agronomiques Tropicales et des Cultures Vivrieres, St.
Denis, Reunion, Indian Ocean

86 Maize Research in the Economic Community of the Great Lakes Countries
(Burundi, Rwanda a Zaire), A. Mpabanzi and E. Ntawuyirusha, Institut de
Recherche Agronomique et Zootechnique de la CEPGL, Gitega, Burundi

98 Maize Improvement in Somalia, B. Abbanur, Agricultural Research Institute,
Agfoi, and M.F. Shirdon, Somali National University, Mogadishu, Somalia

100 Maize Research Activities in Swaziland, J.P. Shikhulu and E. Mavimbela,
Malkerns Research Station, Malkerns, Swaziland

112 Maize Research in Tanzania, A.J. Moshi, National Maize Research Programme,
TARO-Ilonga Research Institute, Kilosa, and W. Marandu, Uyole Agricultural
Centre, Mbeya, Tanzania

118 Maize Research and Seed Production in Uganda, E.R. Kaahwa and F. Kabeere,
Uganda Seed Project, and E. Rubaihayo, Kawanda Research Station, Uganda

130 Maize Research and Production in Zaire, N.N. Mulamba and M.Y. Asanzi, Zaire
National Maize Program, Lubumbashi, Zaire

138 The Zimbabwe Maize Breeding Program, R.C. Olver, Crop Breeding Institute,
Harare, Zimbabwe

143 Contributed Papers
I. Maize Research
Integration of Research Activities and Planning, W.E. Sprague, Maize
Consultant, Hull, Georgia, USA

151 CIMMYT's Maize Improvement Program, R.P. Cantrell, Director, Maize Program,
CIMMYT, Mexico

160 II. Breeding
Evaluation of Population Improvement in the Kenya Maize Breeding Methods
Study, L.L. Darrah, Agricultural Research Service, US Department of
Agriculture, University of Missouri, Columbia, Missouri, USA

177 Breeding for Drought Tolerance in Maize, O. Myers, Jr., Department of Plant
and Soil Science, Southern Illinois University, Carbondale, Illinois, USA, and
W. Mwale, Mount Makulu Research Station, Chilanga, Zambia

186 Development and Evaluation of Maize Hybrids in Zambia,
D. Ristanovic and P. Gibson, Mount Makulu Research Station, Chilanga, and
K.N. Rao, FAO, Lusaka, Zambia

197 Progress in Breeding for Resistance to the Maize Streak Virus Disease,
M. Bjarnason, CIMMYT/IITA, Ibadan, Nigeria

208 CIMMYT's Maize Improvement Role in East, Central and Southern Africa,
B. Gelaw, CIMMYT East African Maize Program, Nairobi, Kenya

229 II. Agronomy
On-Farm Research with a Systems Perspective: Its Role in Servicing Technical
Component Research in Maize, with Examples from Eastern and Southern
Africa, M. Collinson, CIMMYT Eastern and Southern African Economics
Program, Nairobi, Kenya

237 IV. Plant Protection
Maize Diseases in Africa and Their Role in the Varietal Improvement Process,
J.M. Fajemisin, International Institute of Tropical Agriculture, Ibadan, Nigeria

251 Maize Resistance to Stalk Borers [Chilo partellus (Swinhoe) (Lepidoptera:
Pyralidae)]: Some Aspects of Insect Responses to the Plant and Implications for
Breeders, J.K.O. Ampofo and K.N. Saxena, International Centre for Insect
Physiology and Ecology, Mbita, Kenya

259 The Maize Pathology Program in Zambia, K.N. Rao and L.D. Ristanovic, Mount
Makulu Research Station, Chilanga, Zambia

265 V. Seed Production
Kenya Seed Company: Growing for the Future, C. Ndegwa, N.K. arap Tum and
F. Ndambuki, Kenya Seed Company Limited, Kitale, Kenya

270 Zambia Seed Company: The Maize Seed Situation in Zambia, W.M. Chibasa,
Zamseed, Lusaka, Zambia

275 Field Visits
Visit to Golden Valley, W. Mwale (reporter), Mount Makulu Research Station,
Chilanga, Zambia

277 Visit to Mount Makulu Research Station, Chilanga, and the Maize Research
Institute Farm, Mazabuka, R.Watts (reporter), Mount Makulu Research Station,
Chilanga, Zambia

280 Visit to Small-Scale Farmers, Chipapa, Lusaka District, A.F.E. Palmer (reporter),
Maize Program, CIMMYT, Mexico

281 Closing Ceremonies
Conclusion of the First Eastern, Central and Southern Africa Regional Maize
Workshop, The Honourable D. Munkombwe, MP, Minister of State, Ministry of
Agriculture and Water Development, Zambia

283 The Delegates' Response to the Honourable Minister, A. Mpabanzi, Institut de
Recherche Agronomique et Zootechnique de la CEPGL (Burundi, Rwanda a
Zaire), Gitega, Burundi

284 Appendix I
Varieties, Composites and Hybrids Released by African National Programs

301 Appendix II
Participants, First Eastern, Central and Southern Africa Regional Maize
Workshop, Lusaka, Zambia, March-10-17, 1985


The last 20 years have seen little if any
increase in maize production in many
African countries, while population has
increased considerably, leading to a
decline in per capital production in
these countries. The result has been a
growing dependency on imports and
food aid, and adverse impacts on
foreign exchange holdings. This
situation has been aggravated by
drought. The worst famine in recent
African history took place in 1984, and
1985 was predicted to be still worse.

Many formidable problems lie in the
path of African farmers, barring the
way to more vigorous and efficient
maize production. Overcoming these
problems will require determined
action by many groups and a firm
resolve on their part to work together.
The scope for cooperation and its
potential benefits are particularly great
for Africa's agricultural researchers,
who stand to gain, among other things,
better access to ideas and techniques
from inside and outside the continent.

For several years, CIMMYT has been
helping construct a framework for
research cooperation through its two
regional maize programs in Africa. One
of the fruits of that work was the
Eastern, Central and Southern Africa
Regional Maize Workshop (held in
Lusaka, Zambia, March 10-17, 1985),
the first meeting of African maize
researchers since the termination of
the East African Community in 1977.
The chief aim of the workshop was to
create a better awareness among
researchers of their mutual problems
and of various approaches to solving

With this proceedings, our aim is to
further strengthen that awareness,
which is the foundation of regional
cooperation in maize research. The
proceedings consists of 17 country
reports and 13 contributed papers by
prominent maize scientists from both
developed and developing countries.
Some of the reports and papers are
followed by questions and answers or
comments that were made at the end
of the presentations and give further
information on the subject under
discussion. The contributed papers
address many critical issues (research
planning, breeding strategies, on-farm
research, seed production) that African
nations are confronting as they seek
more effective agricultural research
strategies. Many of the papers treat
some aspect of maize improvement,
with particular emphasis on genetic
resistance to insects, diseases and
drought. These resistances are vital to
the improvement of grain yield
stability, which in the African context
is at least as important as increased
yields, if not more so.

Maize scientists should find much
useful information in this fairly
detailed and comprehensive account of
the conditions, problems and activities
of their counterparts throughout the
region, as well as of maize research
being carried out by the international
agricultural centers. We hope that this
proceedings will not only make those
scientists better informed about maize
research in Africa, but that it will also
help them identify specific
opportunities for research cooperation.

Bantayehu Gelaw
Workshop Organizer

STropic of Cancer

Tropic of Capricorn


Participating Countries in the First Eastern,
Central and Southern Africa Regional Maize Workshop


Opening Ceremonies

Welcome to the First Eastern, Central
and Southern Africa Regional Maize Workshop

The Honourable G.K. Chinkulu,
Water Development, Zambia

The need to expand the production of
maize, as well as that of other cereals,
is recognized as one of the most
critical issues presently facing Zambia
and other countries in the region.
Therefore, I want to express my
sincere pleasure and appreciation at
being invited to open this maize
workshop, which I understand is being
attended by delegates from 17
neighboring countries in eastern,
central and southern Africa.

It is extremely gratifying to note that
many organizations, including the
International Maize and Wheat
Improvement Center, the International
Institute of Tropical Agriculture, the
International Development Research
Centre and the United States Agency
for International Development, have

MP, Minister of Agriculture and

provided funds to sponsor the
attendance here of many of the
delegates from outside Zambia.

Scientists of international renown are
also here, sponsored by their own
institutions. We are indeed fortunate to
have such experts join us. They will be
presenting stimulating papers, leading
discussions and providing the cohesion
required for this regional workshop.

I welcome you all and hope that you
will be well satisfied with the workshop
and by your visit and experiences in

This week's workshop is the first of its
kind to be organized in Zambia since
the Third East Africa Cereals Research
Conference in 1968, some 17 years
ago. That conference was jointly
sponsored by Zambia and Malawi, and
was part of a tri-annual gathering of
agricultural specialists which was held
in various countries. I understand that
some of the delegates here today were
also at that conference. This maize
workshop is part of an attempt to
revive such regular meetings.

As many of you know, the Consultative
Group for International Agricultural
Research (CGIAR) now has a number
of centers under its financial wing.
They conduct research in the major
food crops and livestock and have
mandates to work in close cooperation
with national research and
development programs. CIMMYT has a
global responsibility for maize, and
works in close cooperation with IITA,
which concentrates particularly on the
humid tropics of Africa.

The international centers were
established to help regions such as
ours. They are there to guide us in our
long- and short-term research projects
and to spread improved technologies
around the world. They cooperate in
providing guidance in the training of
personnel, in establishing procedures
for conducting sound research, in
organizing production and marketing
programs and in helping to transfer
knowledge regarding crop

I am glad that the international
institutions responsible for maize have
recognized the importance of our
region. Hopefully, production can be
increased to such an extent that we
will be able not only to feed ourselves
and supply our agroindustries, but also
to export to those who are less
fortunate and thus earn needed foreign

The objectives of this workshop are to
outline the "state of the art" for maize
in the region, to clarify the major
constraints to increased production
and to identify priorities. An effort has
been made to bring together as many
scientists and other lay personnel as
possible, so as to create a forum for a
fruitful exchange of ideas and regional
priorities and to discuss the possibility
of sharing resources and materials.
This workshop provides an excellent
opportunity to communicate, to think,
to evaluate and to improve upon these
interactions. Each country, however, is
ultimately responsible for its own
destiny, and must develop sound,
meaningful and long-term maize-
improvement strategies, with
assistance from the international

In Zambia, the party and its
government has made a clear
statement that this country should
become self-sufficient in the major
foodstuffs by 1990, "Operation Food

Production Programme." The program,
announced in 1980, is spearheaded by
small-scale subsistence farmers, lima
farmers (those cultivating one lima,
about 1/4 acre or 625 m2), commercial
farmers and state farms. For this
program to succeed, careful thought
and action need to be given to the
positive and careful use of the
country's resources; this means that
crop production will have to be tailored
to appropriate agroecological zones. In
Zambia, we are now well-aware that
we cannot continue to push production
of maize to all areas, especially those
where other crops can be grown with
comparative advantage. The country's
resources should be used to produce
food crops and other agricultural
products in areas where they are best

In terms of rainfall, its distribution and
potential evapotranspiration, a large
part of this region can be considered to
be in the semiarid tropics. Rainfall
tends to occur over only a few months
of the year, while evapotranspiration
exceeds rainfall for most of the year; it
is important that sensible and careful
use be made of the rain that falls.
There are some very large rivers in the
region, and it is extremely important
that agricultural practices not result in
our precious soil being transported
away from the farms and down the
rivers to be irretrievably lost in the sea.

The soil is our heritage, and it must
not be lost. The conservation of soil
and water and the use of irrigation is
of utmost importance, as is the
development of more drought-tolerant
crop varieties. We are all aware that
large areas of Africa have been
severely affected by drought conditions
over the last 'several years, and in
many countries this continues to be
the sad situation. I trust that this
aspect of maize research will be given
urgent consideration at this workshop.

Farmers live in a risky physical
environment, and many are in a
constant poverty cycle, especially those
who operate on a small scale. Many
factors contribute to this situation,
such as inadequate education and
training, limited resources, variable
marketing and pricing policies, a lack
of credit, land tenure problems and
limited or poor extension services. The
farmers can do very little to change
their physical environment, which also
has a part in keeping them in this
cycle of misery.

What the farmers can do, however, is
to make an all-out effort to protect
their environment and grow crops and
follow farming systems which are
better adapted to their own particular
conditions. They need guidance for
solving many day-to-day problems,
which crops to grow and where, which
varieties to use, how to lay out storm
drains and contour systems, where to
put access roads, how to manage
wetland areas, and how to manage the
wet dambos such as we have in

The overall objective that you have as
agricultural scientists, like all those
connected with the agricultural sector,
is to increase the well-being of your
fellow human beings. The problems of
rapid population growth and the
necessity of increasing national food
production are of concern worldwide.
There is a general belief that people in
Africa are eating less food now than
they were ten years ago, and that the
food is nutritionally poorer. This trend
must be reversed, and strategies for
increasing production must be adapted
to local conditions in each of our
countries. We should not and cannot
afford to continue to import foods
which can be grown locally; our scarce
resources must be used to import
materials and equipment to help us
produce the goods and finished

Every nation interested in maize
improvement has no doubt received
seed and other support from the
international centers and has benefited
from it. If your country has not already
done so, take the appropriate action so
that your national programs can
become more productive. It is most
important that our African scientists
have continuous access to the centers'
generous flow of germplasm and
technologies, and that no unnecessary
quarantine requirements or plant
breeders' rights are in the way.

Plant breeders and other scientists
must produce varieties which minimize
farmers' constraints, particularly in
marginal areas. You are the key to our
future health and prosperity. Your
shoulders must be strong, your skills
and judgment sound. Good scientific
procedure demands that you be
energetic, humble, both constructive
and critical, open-minded but not
credulous, accustomed to think before
you act and then to act upon your
conclusions. You must safeguard the
public interest in matters of health and
safety, and discharge your professional
responsibilities with integrity.

The staff working on maize in Zambia
have gone a long way in this respect,
and I am sure you will be pleased to
learn that the research branch of the
Department of Agriculture has made
excellent progress; they have released
eight new hybrids and two open-
pollinated varieties in the past year.
When multiplied and available, these
will reduce our 20 years of dependence
on the excellent and productive, but
late-maturing, hybrid SR52. The new
varieties will provide a choice for all
types of farmers in the various
ecological regions of the country. The
maize staff are to be congratulated on
this tremendous achievement.

The major thrust now must be to
ensure that technical solutions be
found to the production problems faced
by small-scale farmers; as I mentioned
before, the majority live in a harsh
environment with limited resources.
Many large-scale farmers also operate
within a difficult financial
environment, with the banks breathing
down their necks. Care must be taken
so that small-scale farmers are not
locked into their traditional practices,
and large-scale farmers are not
burdened with financial millstones.
This workshop should endeavor to
ensure that technology is relevant for
both of these groups. An effort is being
made in Zambia and elsewhere to
tackle this problem through on-farm
research. The adaptive research
planning team in Zambia functions on
a regional/provincial basis, and
supports the work of the
multidisciplinary commodity research
teams that operate nationally.

There are many examples in which a
change in crop management and/or
variety brings an array of new
problems to be solved. Each country
has to understand its own constraints
and utilize its own research capabilities
to provide appropriate solutions. The
international centers are ready to
provide assistance in germplasm,
materials, technologies and advanced
training. Sometimes scientists with the
necessary skills can be loaned by the
centers and placed at strategic
locations in a region. Perhaps this
workshop can recommend action and
decide on a possible location or
locations in this region. It is also
important that each country encourage
the creative ability of its scientists, so
as to ensure maximum productivity in
those areas where maize has
considerable potential. It is sometimes
difficult when only a few staff
members are available to develop well
thought-out strategies, but even a few,
supported by funds and equipment,
can make remarkable progress.

Training is imperative to success. All
of our countries need to train and have
available competent agriculturists. We
greatly appreciate assistance from
scientists from other countries, but
how much longer can we rely on
outside aid? The international centers
have excellent courses for scientific
and technical personnel in most
disciplines. It is the task of all
countries in the region to utilize these
opportunities, arranging sponsorship
for their workers, either from those
centers or elsewhere. This workshop
may reveal what can be done and
possibly suggest the number of
students that can be accommodated
annually. There should also be
opportunities for training within the
region. I understand that CIMMYT
recently held a successful training
program for maize technicians in
Malawi, and that plans are underway
to hold one in Zambia in 1986. This
workshop can further stimulate such

All of us are faced with a very
considerable challenge. Zambia offers
excellent conditions for increased
production through its soil, water
resources, rainfall and other climatic
factors. We look to the maize section of
the cereals research team in the
research branch of the Department of
Agriculture for guidance, commitment
and enthusiasm. Researchers must
continue to select improved, disease-
resistant, high-yielding varieties
suitable for all types of farmers. The
grain of these varieties must be of the
quality demanded by the consumer.
The research program must identify
packages of technologies for achieving
production targets in the shortest
possible time. Similarly, the extension
services and the various development
and aid programs must vigorously
disseminate the acquired knowledge
and help farmers during all stages of

We hope that everyone here will
contribute effectively to the objectives
of this workshop and that all of our
countries will benefit from it. Good
research and production strategies
must be worked out for each country.
Let us collect and collate the facts of
maize production. How much maize is
or could be grown? Where is it grown,
and what is the potential area within
each ecological zone? In which areas
would it be better to concentrate on
more drought-tolerant crops? What are
the major constraints of maize
production, and how can they be
eliminated? What are the market
prices for maize, and are they
sufficient to encourage farmers to stay
on the land rather than seeking
alternative employment in the towns?
What are consumer demands as to
amount and quality, and are the
marketing organizations and millers
able to fulfill those demands on a
regular basis? What are the training
needs of researchers, extension
workers, farmers and others in
agriculture? Let us produce first to
feed ourselves, and then hopefully to

In conclusion, I would like to repeat
my very sincere thanks to the sponsors
of this workshop, especially to
CIMMYT and our own Department of
Agriculture and its cereals team. I
understand that many Zambia-based
organizations have provided assistance
by way of funding and transport; I
thank them all most sincerely. Let me
also say how grateful we are to the
management of our Lusaka
Intercontinental Hotel for providing the
conference facilities and accommodations
for our international guests.

To all participants, may I wish you
well in your deliberations at this
workshop and success in your
challenging tasks when you get back
home. I will look forward to receiving a
copy of the workshop proceedings and
recommendations. And most of all, of
course, I look forward to seeing a
surplus of maize in all of our countries!

It is with great pleasure that I declare
this workshop open.

The Delegates' Response
to the Honourable Minister
A.J. Moshi, National Maize Research Programme, Tanzania

On behalf of my fellow participants, I
would like to extend our thanks to the
Government of Zambia for allowing us
to attend this maize workshop. We also
wish to thank the Honourable Minister
of Agriculture and Water Development
for the very sound advice he has just
given us. We hope that by the end of
the workshop we will know more about
the problems and successes of maize
production in the countries represented
here and will have identified ways of
solving those problems.

We extend our thanks to the
organizers of this workshop, especially
Dr. Bantayehu Gelaw of the CIMMYT
East African Maize Program, who has
worked so tirelessly in its organization.

We also thank the Zambian research
team for assisting him, as well as for
welcoming us so warmly to Zambia.
Our thanks also go to the several
institutions and international
organizations without whose
sponsorship many of us would not be

We hope that this workshop will not be
the last and look forward to another in
two years. Perhaps, in the interim, we
can put into practice the
recommendations of this workshop and
will have new achievements to report
in 1987.

Honourable Minister, again we wish to
say, "Thank you."

Initiation of the First Eastern, Central
and Southern Africa Regional Maize Workshop

B. Gelaw, Workshop Organizer, CIMMYT East African Maize
Program, Nairobi, Kenya

On behalf of the workshop steering
committee and the East African
Regional Maize Program of the
International Maize and Wheat
Improvement Center (CIMMYT), I
would like to welcome all of you to the
First Eastern, Central and Southern
Africa Regional Maize Workshop here
in Lusaka, Zambia. CIMMYT's East
African Regional Maize Program was
formally established in September
1982 and is headquartered in Nairobi,
Kenya; one maize breeder is assigned
to the program as coordinator. Since
that time, I, as the program's
representative, have been attempting
to organize such a workshop as this,
but due to a number of circumstances,
I have been unable to do so sooner. I
am pleased that the time has finally
come for this long-awaited meeting of
African maize research workers.

We feel that Zambia is a good place to
hold such a workshop; the Zambian
National Maize Research Program
successfully released eight hybrids and
two open-pollinated varieties in 1984,
and other hybrids and varieties are in
the pipeline for possible release. Their
hybrid maize work is backed up by a
strong population improvement
program. It is my belief that Zambia's
experience can provide an excellent
opportunity for maize scientists in the
region to see the results of an
integrated breeding approach.

During the 1960s and 1970s, there was
a forum known as the East African
Cereals Research Conference, where
scientists from various countries in the
region got together every so often to
exchange ideas and research results.
This conference was terminated in
1977 with the discontinuation of the

East African Community, and since
then we have had only a few sporadic
conferences and workshops organized
by national programs and assisted by
international centers and donor

Production of maize, the primary food
crop in Africa, has remained stagnant
over the past 20 years in most African
countries, averaging about 1 ton per
hectare. In certain countries,
production per capital has actually
declined. To cover deficits in
consumption, several African countries
have become dependent on imports
and food aid, which has seriously
affected their balance of payments.
Africa is the only continent where
Malthus' prediction of food production
not being able to keep pace with
population growth seems to be a

The year 1985 was the target of the
famous 1980 Lagos Plan of Action,
which was adopted by African leaders
for eliminating hunger from their
continent. However, FAO's forecasts
show that 21 African countries will
face more severe food shortages in
1985 than they faced in 1984, one of
the worst famine years in the
continent's recent history. There is,
nevertheless, some hope that existing
improved maize hybrids and varieties,
as well as production practices, can
increase maize yields in Africa.

To make this hope a reality, research
activities must be stepped up to
generate new and more appropriate
technological components that can
increase yield dependability across a
wide range of environments. More
emphasis should be given to

developing varieties with greater
tolerance to drought, mineral stresses
and temperature extremes and with
improved resistance to economically
important diseases and pests. For this
reason, we have invited internationally
renowned maize scientists to share
their experiences with us by presenting
papers and contributing to our
discussion sessions.

In addition, each participating country
has been officially requested to present
a "country report," outlining current
maize research activities, materials
released, pressing problems and needs
and future plans of action. An
exchange of such information, as well
as of techniques and breeding
materials, should facilitate maize
production in the region. It is my firm
belief that this workshop will help
encourage cooperation in maize
research efforts by broadening our
awareness of each country's conditions
and activities and by strengthening our
relationships with one another. It
should also help familiarize the
participants with the current and
planned maize research activities of
the international centers in the region.
Scientists from this region have often
met outside the region or continent,
but not in one another's trial plots.

Twenty-two countries received formal
invitations for two senior national
maize scientists to attend this
workshop. Sudan and Botswana sent
their regrets. The Seychelles and
Comoros Islands have not responded,
and the expected Rwanda delegates
have not arrived. The other 17
countries accepted our invitation and
have sent their delegates. Special
invitations were also sent to a few
prominent maize scientists from
developing as well as developed
countries to join us here; all have
accepted our invitation, for which we
are indeed grateful. Some international
and bilateral donor agencies have also

sent representatives as observers. All
in all, over 100 participants are here at
the workshop, including observers
representing a number of organizations
in Zambia. Thank you all for your
interest in this endeavor.

This workshop would not have been
possible without the generous financial
and moral support of a number of
governmental, bilateral and
international organizations. First and
foremost, I would like to extend my
sincere thanks and appreciation to the
government of the Republic of Zambia,
in general, and to the officials of the
Ministry of Agriculture and Water
Development, in particular, for
graciously hosting this First Regional
Maize Workshop, and for organizing it
in cooperation with CIMMYT. Zambia
has also contributed immensely in
meeting all local expenses, including
hotel bookings, transportation,
secretarial assistance, receptions,
banquets, refreshments and many
other incidentals. Many senior officials
and scientists have spent a great deal
of time in organizing this workshop. A
number of organizations in Lusaka
have also provided assistance in
funding and transport, including
Zamseed, Power Equipment, the EEC,
SIDA, Barclays Bank and Shell

The International Development
Research Center (IDRC) fully funded
two candidates each from Ethiopia,
Uganda and Burundi; their regional
office in Nairobi was also very helpful
in encouraging the holding of this
workshop. The East African Regional
Economic Development Service of the
United States Agency for International
Development (USAID/REDSO/ESA)
covered the cost of plane tickets and
per diem expenses of many
participants. USAID's Kenya office
funded the two Kenyan delegates
nominated by their government, and
the University of Florida/USAID/Malawi

Government Agricultural Research
Project funded one of the participants
from Malawi. The International
Institute for Tropical Agriculture (IITA)
has pledged to fund one participant
each from Tanzania, Malawi and
Zimbabwe. CIMMYT's Eastern and
Southern African Regional Economics
Program is covering many of the
expenses of the workshop.

Last but not least, CIMMYT's Maize
Program was ultimately responsible for
meeting the expenses of all other
participants, as well as being
responsible for inviting the
participants, guest speakers and
observers. The arrangement of travel
plans within and between countries,
telephone calls and the sending of
letters, cables and telexes was
undertaken by CIMMYT's East African
Regional Maize Program office. To all
others who have contributed in any
way, and there are many, I am indeed

I hope that, before this workshop ends,
a recommendation will be made that
such a Regional Maize Workshop be
held every other year, rotating the host
countries so as to give each country an
opportunity to hold such a meeting.

In closing, I wish once again to express
our heartfelt thanks to the government
of Zambia for the wonderful hospitality
they have extended to us.


Country Reports

Maize Research Activities in Angola

F. Marcelino and M. Girao, Instituto de Investigacao Agronomica,*
Chianga, Huambo, Angola

Maize is the staple food of nearly one-
third of the 2.5 million people of
Angola and is also the main food crop
of the country. Annual maize
production in the 1970s, the last
period for which figures are available,
was about 700,000 tons, of which
about one-sixth was exported. The
three main maize-growing areas of the
country were the central highlands
(70%), the Huila highlands (15%) and
the Malange highlands (10%).

About 85% of the total production in
the 1970s was by small-scale farmers,
who had an average yield of about 500
kg/ha. The other 15% was produced by
larger farmers, with yields ranging
between 1 and 4 t/ha.

The principal grain types grown in
Angola at that time were white flint
(80%), white dent (10%) and yellow
flint (10%). The main cultivars grown
were Branco Redondo (a white flint,
open-pollinated variety), SAM2 (a
yellow flint synthetic) and SR52 (a
white dent, single-cross hybrid).

Today, because of problems due to the
war, no accurate statistical data is
available, but it is supposed that
present production does not exceed
200,000 tons per year.

Maize Research Activities

Maize research in Angola was begun in
the 1940s and is conducted by the
Agronomic Research Institute (IIA).
Presently Angola and Yugoslavia have
a bilateral contract for cooperation in
two projects, one in maize breeding
between IIA and the Maize Research

Institute (MRI) of Zemun Polje,
Yugoslavia, and the other in maize
seed production, with MRI, IIA and
DNOPA (the National Department for
Agricultural Production).

Maize breeding
The principal activities of the maize
breeding project are:

* Conservation of a small germplasm
bank, including about 800
heterozygous populations;
* Development of inbred lines (1400
lines are being studied, of which 600
are in their sixth or later generation
of self-pollination and 800 are
between the third and fifth
* Study of approximately 400 hybrids
of various types;
* Collaboration with CIMMYT in
international trials (with the best
results obtained in 7 EVTs, 2 ELVTs
and 1 QPMT);
* Selection for resistance to streak
virus, the principal maize disease in
Angola (phenotypic selection is
being carried out with late material
and MSR-EVT materials introduced
from IITA);
* Selection for tolerance to
Helminthosporium turcicum, which
causes leaf blight in the central
* Selection for tolerance to soil acidity
(most of the Angolan areas suitable
for maize growing have acid,
ferralitic soils); 2 apparently very
tolerant varieties, 8 tolerant varieties
and 18 varieties with medium
tolerance to aluminium toxicity have
been selected, and

* Agricultural Research Institute

*Selection for reduced plant height
for lodging resistance. The CIMMYT
materials have better plant height,
but are normally more susceptible to
Helminthosporium turcicum than
the Angolan materials.

Seed production
The controlled production of maize
seed presently involves only one single-
cross hybrid, ZPSC852b, a tall, late,
white dent variety similar to the SR52
grown at Malange, and one synthetic,
SAM3, a yellow flint with broad
adaptation, moderate tolerance to H.
turcicum and good yield potential
(sometimes more than 7 t/ha in small
trials at Huambo). In the 1983-84
season, 31 tons of ZPSC852b seed was
produced, and 2 tons of SAM3 basic
seed was ear selected.

Agronomic improvement
A small program of field trials has
been conducted on the use of nitrogen,
phosphorus, potassium, sulfur and
magnesium fertilizers and rock
phosphates. Also, tests on herbicides
have been conducted, in which
Primextra was found best for chemical
control of weeds, especially nut grass.

Other research achievements
Other accomplishments of the national
maize program have been:

* Inbred lines-12 inbred lines have
been selected as most promising,
including three of late and one of
medium-late maturity;
* Hybrids-18 experimental hybrids
have been selected, 7 white (4 single
crosses, 1 double cross, 1 top cross
and 1 triple cross), 6 yellow (4 single
crosses and 2 top crosses), 5 white-
yellow (mainly from one inbred line
developed from the hybrid Pioneer
44), and

* CIMMYT populations-the
populations showing best
performance in Angola among the
CIMMYT trials have been SIDS
7844, Poza Rica 8022, Across 7921,
El Paraiso 7929 and Across 8043.
The best one, Poza Rica 8022, has
yielded more than 8 t/ha.

Research Staff

The staff of the national maize
program includes two Angolan maize
breeders, two Yugoslavian maize
breeders and six Angolan technicians.
At times additional part-time staff are
employed; for example, more
Yugoslavian maize breeders and
Angolan technicians are needed at
pollination time. There is also close
collaboration with the staff of the Soils
and Climate and the Phytopathology
departments of IIA.

Constraints to Maize Research

The principal problem in the maize
breeding project is that experimental
trials are carried out only in the central
highlands. However, it is hoped that, in
the near future, they can be extended
to two other locations, Malange and
Huila (Matala).

The principal problems in the maize
seed production project are the
inadequacy of seed processing facilities
in the country and the lack of farms
that are well adapted for seed
production. The first problem will be
partially solved with the installation of
two seed processing plants, at Chianga
and Malange, by an FAO-UNDP project.
The establishment of state farms for
seed production at Huambo, Malange
and Huila will further increase the
amount of quality maize seed available
in the country.

Farmer Rejection of Late-Maturing,
High-Yielding Maize in Burundi

R.S. Zeigler and M. Kayibigi, Programme Mais et Petit Pois,
Institute des Sciences Agronomiques du Burundi,* Bujumbura,

Burundi is a landlocked country in
central Africa. The country is mostly
mountainous, except for the plain at
the tip of Lake Tanganyika. The
climate varies from tropical to
temperate according to elevation,
which ranges from low (800 to 1200
meters) to medium (1200 to 1800
meters) to high (1800 to 2600 meters).

Maize is the most important cereal
crop in Burundi and is grown by all
farmers in all localities. It is most
important at high altitudes, where it is
consumed as fresh green ears or made
into flour for the making of ugali. In
low altitudes, maize is consumed only
in the form of immature ears. In
medium-altitude areas, tuber crops
such as cassava and sweet potato tend
to be more important in the human
diet than maize.

Maize is cultivated in association with
legumes, particularly beans (Phaseolus
vulgaris L.) in the first season and peas
(Pisum sativum L.) in the second. It is
grown by small-scale farmers, and land
preparation, sowing, weeding and
harvesting are all carried out by hand.

The principal constraints to maize
production are:

* Diseases, especially maize streak
virus, but also leaf blight and rust;
Lack of early varieties with high
yield potential, which would permit
a second-season rotation crop;
Storage insect pests, particularly
Sitotroga cerealela and Sitophilus

Program for Maize and Peas, Burundi
Institute of Agricultural Sciences

* Infertile, acid soils, which are found
in many regions, and
* Poor infrastructure within the
country, which makes it difficult for
farmers to get the necessary inputs
and technical advice.

The maize program is one of the
research programs organized by the
Burundi Institute of Agricultural
Sciences (ISABU). Financial assistance
for this program is provided by the
International Development Research
Centre (IDRC) of Canada. Staffing
includes one expatriate adviser, two
graduate agriculturists, two
technicians, one administrative
assistant and one agricultural

The principal objectives of the maize
program are:

* Selection and improvement of
varieties that meet the needs of
farmers (with active collaboration
from the farmers in the testing of
* Assessment of new problems
identified at the farm level, and
* Establishment of research projects
for solving these problems.

Results of the program to date have
included the release of Kitale
Composite A (KCA) for the high-
altitude zone, Igarama-4 for the
medium-altitude zone and GPS5 for
the low-altitude zone.

Presently, special emphasis is being
placed on the development of a high-
altitude population that is resistant to

maize streak disease and early enough
in maturity to permit double cropping.
Resistance is being developed by
selecting within local materials under
artificial inoculation, as well as by
introducing genes for resistance by
backcrossing local materials with an
IITA source. Two low-altitude materials
derived from CIMMYT experimental
varieties (Population 43) have entered
verification trials on farms, following
their promising performance in variety
trials; these varieties are being
multiplied and will probably be
distributed to farmers in 1986. It is
hoped that the program can identify a
synthetic variety, using IITA inbreds,
to serve the medium-altitude zone.

The Burundi maize program
collaborates with international
organizations, including CIMMYT and
IITA; it is felt that the successes
registered by the program are a result
of that collaboration.

Maize Research

In 1980, the Burundi maize program
recommended for release a high-
yielding, long-season maize variety
selected from Kitale Composite A
(KCA), which is of Kenyan origin.
Maturing in approximately 220 days,
this tall ( > 2.5 m) variety is harvested
in late May; it is rather susceptible to
lodging. Although it yields 20 to 40%
more than previously released varieties
(8), KCA has met with considerable
farmer resistance because of its
lateness. The principal complaints are
that farmers have to wait six weeks
longer before they can begin their
harvest, and that this does not permit
a good second-season pea crop.

Informal surveys by the Burundi Maize
Program have revealed that one of the
most common cropping systems in the
highlands is a maize/bean-pea
intercrop/rotation. Maize is typically
planted with beans in late September
at the onset of the rains. The beans are
harvested dry in late December and

January. After a short dry season,
farmers plant peas in mid-March
among the ripening maize, which is
harvested a few weeks later. The peas
mature as the rains taper off in May
and June, and are harvested dry in
July. Thus, harvests are staggered over
the year. Almost all production is
consumed in the home, with any
surplus entering the local market.
Commodity prices are set in the local
market place with no government

This paper presents a critical analysis
of the variety KCA within the context
of this local cropping system. In 1982
and 1983, field trials were conducted
according to farmers' traditional
practices to quantify the impact of the
use of KCA on the system. The three
principal questions addressed were
whether the farmers gained by
planting high-yielding, late-maturing
maize, whether it was possible to
modify the cropping system to make
the variety more attractive, and
whether by closely following traditional
practices, researchers could modify
selection strategies to insure that new
varieties would be accepted by

Trial 1. Impact of maize maturity on
following-season pea yield-In
September 1982, 16 varieties and lines
of various maturities were planted in a
multilocational yield trial, generally
following CIMMYT's methods for their
EVTs, at three high-elevation sites,
Kisozi at 2090 m, Munanira at 2140 m
and Nyakararo at 2100 m. In early
March, 10 t/ha of manure was lightly
tilled into the soil of the plots, and on
March 15, peas of the local variety
Kyondo were planted at a density of
62,500 plants/ha. Because of the
differences in the maturity of the
various varieties and lines, the peas
were planted in some cases as much as
six to eight weeks before the maize
was harvested. In some plots, the
maize was harvested to determine if

tilling for peas had had an adverse
effect on maize grain yield. Total
economic yield was based on the
market prices of 25 Burundi francs/ha
for maize (120 FBu = USS 1) and 65
FBu/ha for peas. Although the crops
under consideration were not cash
crops, these freely set prices were
thought to fairly represent their
relative values.

Trial 2. The impact of maize maturity
and density on yield in a three-crop
system-A second trial was planted at
Munanira in September 1983 to
compare the performance of KCA with
that of Igarama-4, an improved local
maize which matured 35 days earlier
than KCA in the typical maize/bean-
pea system. Both varieties were
planted with uniform spacing at six
densities (55,000, 45,000, 35,000,
25,000, 15,000 and 5,000 plants/ha) to
determine which density led to
maximum yield for each crop in the
system. Plots measured 5 x 5 meters
and manure was applied at 30 t/ha; the
trial was arranged in four randomized
blocks. The local bean mixture was
interplanted at a density of 125,000
plants/ha. As in Trial 1, peas were
planted among the maize plants on
March 15, but with a density of
125,000 plants/ha, which more closely
approximated farmers' plant densities.

Maize was harvested at maturity, and
yields were corrected to a 14%
moisture level. Disease severity data
were noted for all crops during the
year. Total economic yield was
calculated with actual market prices at
harvest, beans at 65 FBu/kg, maize at
25 FBu/kg and peas at 90 FBu/kg. The
pea harvest was poor in 1984 because
of a lack of rain; this accounted for the
higher price. Total protein yield was
based on beans at 22%, maize at 9%
and peas at 22.4%. Simple land
equivalent ratio (LER) was calculated
for each maize density and variety over
the two seasons from the expected

yields determined from regressions of
yield data on planting density. The
standardizing factor for beans (1430
kg/ha) was the intercept of the
combined KCA and Igarama-4 bean
yields as a function of maize density;
that for peas (1003 kg/ha) was the
mean of the two intercepts for the pea
regressions. Maize standardizing
factors were their expected maximum
yields (8807 kg/ha for KCA and 6120
kg/ha for Igarama-4). Effective LER
(ELER) was calculated according to
Mead and Willey (5). All regressions or
correlations referred to in the text or in
the figures are significant at least at
P = 0.05.

Trial results
Trial i-There was no significant
correlation between maize maturity
and total economic yield, but there was
a highly significant correlation between
maize efficiency and total economic
yield; this differed among varieties.
High-yielding, late-maturing varieties
such as KCA didn't have total
economic yields significantly different
from those of earlier varieties. Maize
and pea yields as functions of maize
maturity followed a similar distribution
at the three sites (Figure 1). A
significant maize x site interaction was

Nyakararo 0
** Maize
S 40 00 oo 1000 -

u2 3000. O Peas 6n i
0> O
a W g
S2000 /C O 200

160 180 200 220
Days to maize harvest

Figure 1. Maize and pea yields as functions
of maize maturity at Nyakararo, Burundi
Note: Each point is the mean of four replicates

detected for maize yields and appeared
to be due to the lower yields (compared
to previous years) of the late-maturing
lines at Munanira than at the other
sites. This interaction precluded
across-site comparisons. Pea yields as a
function of maize maturity at all sites
closely followed exponential decay,
with high levels of significance. Peas
grown with late maize varieties were
observed first to etiolate and then die,
leaving sparse stands of spindly plants
after maize harvest. No difference in
maize yield was measured between
border rows with no interplanted peas
and maize interplanted with peas.

Trial 2-First-season bean yields as a
function of maize density for the two
varieties were identical, as indicated in
Figure 2. Grain yields (kg/ha) of the
two maize varieties were best
described as logarithmic functions of
density (Figure 3). Pea grain yields as
functions of maize density are
presented in Figure 4. There was no
significant correlation between maize
density and pea grain yield for
Igarama-4, with yield expressed as the
overall mean. In contrast, pea yield as
a function of KCA density followed
exponential decay.

0 Igarama-4


1200 -


C 800

The distribution of total economic yield
over the two seasons as a function of
maize density followed a binomial
distribution for both varieties, with no
statistically significant differences
between predicted yields at their
maximum levels (Figure 4). Differences
in disease levels were not related to
maize density.


- 8000

- 6000 .


2000 -


5 15 25 35 45 55
Maize density (000 plants/ha)
Figure 3. Maize yield as a function of maize
Note: Each point is the mean of four replicates


0 0 0 O


- 700 -




5 15 25 35 45 55
Maize density (000 plants/ha)


5 15 25 35 45 55
Maize density (000 plants/ha)

Figure 2. Bean yield as a function of maize
Note: Each point is the mean of four replicates

Figure 4. Pea yield as a function of maize
Note: Each point is the mean of four replicates

In 1984, the rains in the second pea
season were below normal and poorly
distributed, resulting in very dry soils
at pod filling and a heavy attack of
aphids and powdery mildew. Based on
the same pea variety in the same field
at the same density in the previous
year, predicted pea yield could be
calculated from the regression
equations. Using the "normal" price of
peas (65 FBu/kg), predicted total
economic yield could also be
calculated. The conservative figure of
75% of the 1983 pea yield, or 1712
kg/ha, was taken as normal for the
field in question. The predicted total
economic yield as a function of maize
density of both varieties followed
identical binomial distributions (Figure
5). Total predicted protein yield,
calculated in a similar manner and
plotted as a function of maize density,
also followed a binomial distribution
for the two varieties (Figure 6).

Simple LERs, as a function of maize
density over two seasons for the two
maize varieties, followed a similar





- *

5 15 25 35 45 55
Maize density (000 plants/ha)

Figure 5. Predicted total yield of three
crops over two seasons as a function of
maize density

pattern of decreasing LER with
increasing density after a peak at low
density. This decline was less marked
for Igarama-4, regardless of whether
maximum yields for KCA or Igarama-4
were used as standardizing factors for
maize yield (Figure 7). With the yield
of KCA as the standardizing factor,
Igarama-4 had a higher LER over the
range of densities found in farmers'
fields (35 to 55,000 plants/ha). This
trend, though less striking, is also
present when Igarama-4 yield is the
standardizing factor.

The ELER might be a more
appropriate comparison of the two
varieties within this system, but it is
difficult to calculate since there are no
direct data available on the desired
proportion of the crops. Based on data
from this trial, however, one can
estimate from the average on-farm
maize density (approximately 45,000
plants/ha) that the desired crop
proportions within local constraints
(with a maize having a maturity such
as that of Igarama-4) would be 0.49 for


S320 -



o 260 -
* 240




5 15 25 35 45 55
Maize density (000 plants/ha)

Figure 6. Predicted total protein yield of
three crops over two seasons as a function
of maize density

2.4 -

w 2.2

I = Igarama-4
1.8 -

5 15 25 35 45
Maize density (000 plants/ha)
Figure 7. LER over two seasons as a
function of maize density

Note: Numerator is variety in question,
denominator indicates maximum yield
used as maize standardizing factor

peas and 0.51 for maize. The ELER for
KCA (calculated at 15,000 plants/ha,
which gives approximately the same
grain yield as Igarama-4 at 45,000
plants/ha) was 1.46 and that for
Igarama-4, 1.90. Adding the bean
component for each appropriate
density gives a total ELER of 2.29 for
KCA and 2.52 for Igarama-4.


If maize were selected solely on the
basis of yield in Burundi, late-maturing
varieties could be recommended.
However, increases in maize yield
would be gained at the expense of the
second-season legume crop, thus
eliminating the yield advantage of
long-season maize. No advantage in
total economic yield and protein yield
for the late maize was shown in either
trial, and at local planting densities
LERs were less favorable for the late
variety. The ELERs indicate that there
is no density at which KCA is clearly
superior to Igarama-4.

Site x variety interaction is striking in
long-season maize. At Munanira, the
two maize varieties were damaged by
heavy wind, and at Kisozi they
1/1 suffered from a short but severe
drought that coincided with flowering.
S Thus late maize, although it has a
K/I higher yield potential, leads to greater
instability, perhaps simply by being
I/K exposed longer to the vagaries of
nature. Since risk avoidance is a
characteristic of subsistence
agriculture, farmers may prefer to
K/K divide potential productivity among
their principal food crops.

When comparing late- and medium-
maturity maize, a most significant
consideration becomes yield quality
and distribution. With KCA, most of
the protein in the system is
incomplete, being entirely from cereals;
with Igarama-4, a significantly greater
proportion of the protein is
nutritionally complete because it
comes from both cereals and legumes.
Igarama-4 is consumed with the beans
remaining from the earlier harvest, and
peas are consumed with other starch
staples which are grown on land too
poor for maize. If KCA were the
principal variety grown, maize would
only be available after most of the bean
crop had already been consumed, and
no legume would be available as a
complement to maize during the dry

The different maxima among the
curves of total economic yield and
protein yield and LER as functions of
maize density are noteworthy. LER
reaches its maximum at 20,000
plants/ha; total yield maxima are
reached at 40,000 plants/ha. This
commonly encountered discrepency (5)
agrees with Riley's observation that
intercrops with similar LERs may have
very different levels of attractiveness to
farmers (7).

Thus, selecting maize on the basis of
yield alone for the Burundi highland
cropping system could result in the
release of a variety that would disrupt
the system unacceptably from the
farmer's perspective. It is clear that,
within the context of the whole
system, farmer objections are quite
rational. Biological and economic yield
analyses show that a high-yielding,
long-season maize offers no particular
advantage, and an examination of
LERs and food availability and
distribution indicates that there may
even be a serious disadvantage to
pursuing a selection strategy of
maximizing maize yield with no regard
for the other components of the

This understanding of the place of
maize in the local cropping system has
permitted the Burundi maize program
to develop selection and evaluation
methods that ensure that maize
varieties released in the future are
compatible with farmer needs and
limitations (9). These methods include
the improvement of populations under
intercropping and low-input conditions,
as well as on-farm, farmer-managed
evaluations of promising material.
Particular attention is given to
maturity and problems that may
increase yield instability across years.

In discussions on crop improvement in
the developing world, the criticism is
often encountered that many of the
recommended varieties and
technologies are not relevant to the
needs of the farmers or are actually
incompatible with the constraints they
face (6). Farming-systems research was
developed, in part, to address this
problem and to guide agronomists to
more appropriate technologies (2).
However, even in well-integrated
projects, friction can develop among
researchers having different
perspectives (3). Farming-systems
teams frequently exist as separate
entities, and this may result in there

being little communication between
these teams and the commodity
programs; there may,in fact, be
competition for scarce resources (1).
The results presented here show the
inappropriateness of some
recommendations developed from a
strict commodity orientation, and they
suggest how a farming-systems
approach may be incorporated into a
national commodity research program
with relative ease.

National commodity research programs
should select and evaluate their
materials and technologies on the basis
of farmers' limitations and
requirements. Informal surveys
.conducted by the commodity
researchers themselves can suggest
how this may be done for a given
situation. It is essential that the
commodity program be given the
responsibility for on-farm work,
regardless of whether or not there is an
independent farming-systems team.
This allocation of responsibility does
not necessarily require major
institutional reorganization, but it does
require sensitive commodity
researchers who conduct a substantial
portion of their work in farmers' fields.


The authors wish to acknowledge the
assistance of the International
Development Research Centre of
Ottawa, Canada, which partly finances
the Burundi maize program. They are
also grateful to B. Come for his
excellent field assistance.


1. Eicher, C., and D. Baker. 1982.
Research on agricultural
development in sub-Saharan Africa:
A critical survey. Department of
Agricultural Economics, Michigan
State University, East Lansing,
Michigan, USA. International
Development Paper 1.

2. Gilbert, E., D. Norman and F.
Winch. 1980. Farming systems
research: A critical appraisal.
Department of Agricultural
Economics, Michigan State
University, East Lansing, Michigan,
USA. Rural Development Paper 6.

3. Horton, D. 1984. Social scientists in
agricultural research: Lessons from
the Mantaro Valley Project, Peru.
International Development Research
Centre, Ottawa, Ontario, Canada.

4. ISABU. 1984. Rapport annuel du
Programme Mais. In Rapport Annuel
de l'Institut des Sciences
Agronomiques du Burundi 1983.
Bujumbura, Burundi.

5. Mead, R., and R.W. Willey. 1980.
The concept of a "land equivalent
ratio" and advantages in yields from
intercropping. Experimental
Agriculture 16:217-228.

6. Poleman, T., and D. Freebairn.
1973. Food, Population and
Employment. Praeger. New York,
New York, USA.

7. Riley, J. 1984. A general form of the
"land equivalent ratio."
Experimental Agriculture 20:19-29.

8. Schalbroeck, J., and P. Bagona.
1983. Interaction varite~s x
environments dans les essais
varietaux de mais conduits au
Burundi en 1978 et 1979. Institute
des Sciences Agronomiques du
Burundi. Bujumbura, Burundi. Note
Technique 31.

9. Zeigler, R., and K. Manasse. 1984.
Toward a maize program responsive
to Burundi farmers. In Crop
Improvement in Eastern and
Southern Africa: Research
Objectives and On-Farm Testing,
R. Kirkby, ed. International
Development Research Centre,
Ottawa, Ontario, Canada. Pp. 30-35.


Mr. Olver: We have no intercropping in
Zimbabwe. Since you find that it is
important, and since plant population
is reduced in intercropping systems,
how do you select for prolificacy?

Mr. Kayibigi: We select by routine
extraction under standarized
intercropping systems. One of our
released varieties has tow ears.

Mr. Olver: Is prolificacy a genetic

Mr. Kayibigi: Prolificacy is influenced
by the environment, and given the
right conditions, that variety will give
two ears.

Maize Research and Production in Ethiopia
A. Debelo, Institute of Agricultural Research, Awassa, Ethiopia


Ethiopia is a large country (1,222,000
km2) with a range of climatic
conditions suitable for the growing of
both temperate and tropical crops. A
great variety are grown, ranging from
teff, millet and cotton in the tropics to
such temperate crops as wheat, barley
and potatoes. The main native food
crops are sorghum, maize, teff, millet,
wheat, barley and enset. Coffee, cotton,
castor beans, peanuts and pulses are
the major cash crops.

The amount of cultivated land in the
country is small in proportion to the
total area; only about 10 million
hectares are farmed (Table 1). Of this
amount, about 50% are planted to the
cereals teff, barley, wheat, maize,
sorghum and millet (Table 2).

In Ethiopia, maize is grown throughout
the country, with the bulk of
production concentrated in the
southern, southwestern 'and western

Table 1. Land use, Ethiopia

o/o of
Type 000 km2 total

Cultivated land 104.3 8.5
Fallow land 20.9 1.7
Orchards and stimulants 7.2 0.6
Meadows 0.1 0.1
Pastures (rough and dry) 656.7 53.7
Swamps 51.8 4.2
Forestland, open 88.1 7.2
woodland and bush
Barren land and
built-up areas 172.0 14.1
Lakes and rivers 120.9 9.9

Total 1222.0 100.0

Source: National Atlas of Ethiopia, 1981

Table 2. National estimates of areas planted to the major cereal crops, Ethiopia, 1974
to 1982

Year Teff Barley Sorghum Maize Wheat Millet Total

1974 1247.8 864.6 755.7 802.5 785.1 205.6 4661.3
1975 1470.7 648.1 782.2 786.1 556.9 378.4 4622.4
1976 1365.7 807.9 751.3 723.0 867.3 199.0 4414.2
1977 1333.5 897.8 767.5 901.4 512.3 232.3 4644.2
1978 1423.2 940.6 731.1 964.1 531.6 238.3 4828.9
1979 1513.3 909.8 1026.3 870.9 786.7 215.9 5022.8
1980 1362.0 830.9 979.1 735.5 536.3 232.9 4711.8
1981 1331.5 810.4 844.3 652.3 684.9 226.5 4629.3
1982 1399.8 908.0 905.7 819.7 715.0 225.2 5029.2

Rank #1 #2 #3 #4 #5 #6

Source: Central Statistics Office of Ethiopia, 1985

regions. Among the major cereals,
maize ranks first in total production
and in yield per hectare; it ranks
fourth in total area (Tables 2 and 3).

Almost all of the maize produced in
Ethiopia is used for human
consumption in the form of injera (a
large, thin pancake-type bread), kitta
(bread), porridge, or boiled or roasted
as a vegetable, particularly in the milk
to early dough stage. The rest is used
for making tella (local beer) and araki
(a local alcoholic beverage); a very
small amount is used for animal feed.
The stalk is used for construction, fuel
and animal feed.

History of Maize in Ethiopia

The precise date and route by which
maize was introduced into Ethiopia is
not known. However, it is generally
believed that it was brought to East
Africa, and hence into Ethiopia, during
the late sixteenth or early seventeenth
century (5). Since its introduction, the
crop has gained much popularity and
has become adapted to the various
ecological conditions of the country.

There are many different varieties, the
most important ones being dent and
flint types (1).

In reporting on their visit to Ethiopia
in 1967, Harrison, Eberhart and
Hazelden mention that the main
source for Ethiopian maize varieties
may be Tuxpefio with some Caribbean
and US Corn Belt dents. However,
further investigation will be needed
after local collections are made from
the major maize-growing areas and
pockets in the country, and those
varieties are identified.

Even though Ethiopian maize is well
adapted to the environmental
conditions in the country, the average
national yield from 1974 to 1982 was
only about 1.5 t/ha, far below the
world average (Table 3). This low yield
may be attributed to one or more of
the following factors:

*The majority of the farmers still use
varieties that have not been
improved for yield potential,
agronomic traits (such as ear
placement, plant height and lodging)

Table 3. National yield estimates for the major cereal crops, Ethiopia, 1974 to 1982

Year Maize Sorghum Barley Wheat Millet Teff Average

1974 10.5 8.3 7.2 8.9 7.6 6.8 8.2
1975 17.4 11.2 8.3 9.6 10.0 6.8 10.2
1976 13.1 10.1 11.1 10.7 8.7 7.3 9.9
1977 10.3 9.2 7.7 8.4 8.9 7.7 8.6
1978 10.2 9.3 7.4 8.4 8.0 7.6 8.5
1979 17.5 16.0 11.6 11.0 9.9 9.4 12.7
1980 12.9 14.4 12.9 11.5 8.8 9.6 11.9
1981 18.4 14.3 11.5 10.3 8.7 8.1 11.7
1982 19.6 15.0 12.9 12.8 10.7 9.8 13.4

Rank #1 #2 #3 #4 #5 #6

a/q = quintals (100,000 kg)
Source: Central Statistics Office of Ethiopia, 1985

and resistance to the major insect
pest (Buseola fusca) and diseases
(Helminthosporium turcicum and
Puccinia sorghi);
* Farmers use unimproved cultural
practices in land preparation,
planting date, seed rate, weed and
insect control and fertilizer
application, and
* Farmers lack proper storage

Maize Research

Agricultural research was initiated in
the Ethiopian institutions of higher
education in the late 1950s and early
1960s to boost agricultural production
and feed the country's large
population. However, a well-organized
research program with research
stations located in the varying
agroclimatic areas of the country was
initiated only in the late 1960s. The
national crop research program is
organized and conducted by the
Institute of Agricultural Research
(IAR); research is also carried out by
the College of Agriculture.

The IAR coordinates all crop research
activities in the country through its
crop teams, which were organized in
1980. A team comprises various
disciplines, including breeding,
agronomy, soil science, entomology,
pathology and weed science. Due to
the shortage of trained manpower, staff
are involved in more than one crop

maize program objectives
The national maize research program
has the following long-term objectives:

SDevelop high-yielding maize
varieties with desirable agronomic
characters, to be made available to
state farms and farmer associations;

* Create gene pools from which open-
pollinated varieties, lines and
hybrids can be developed at
different stages of the breeding
program, and
* Improve the nutritional quality of
maize through the selection and
release of highly nutritious maize

Short-term objectives
The short-term; objectives of maize
research are:

* Develop open-pollinated varieties
adapted to the different agroclimatic
areas of the country, making them
available to small farmers;
* Develop hybrid varieties for state
farms which have trained personnel
and production know-how;
* Determine proper cultural practices
for the different agroclimatic areas
of the country;
* Strengthen the research and
extension linkage so that improved
varieties and management practices
can be disseminated to farmers;
* Strengthen cooperation with
international and national research
centers of various countries, and
Screen varieties for Ethiopia's
moisture-stress areas.

The varieties which have already been
developed or will be developed can be
classified into three groups, according
to their maturity. The early group (90
to 100 days to maturity) are adapted to
those areas with a short rainy season
or erratic rainfall, and the intermediate
group (120 to 130 days to maturity), to
areas with intermediate rainfall; the
late group (160 to 170 days to
maturity) are suitable for areas with a
long rainy season.These varieties may
be either open-pollinated materials or

Current Research Activities

The breeding program is working to
meet the germplasm requirements of
moisture-stress, intermediate- and
high-rainfall areas. For intermediate-
and high-rainfall areas, open-pollinated
varieties, mainly composites, with
experimental yield potentials of 80 to
100 quintals per hectare (1 quintal =
100,000 kg), have been developed and
released. The weak point of these
varieties is their lodging susceptibility,
which makes them unsuitable for
mechanization and reduces final grain
yield. The good qualities of these
varieties are their high yield potential
and resistance to H. turcicum and P.
sorghi. A full-sib family selection
breeding approach is being followed to
reduce plant height among lodging-
resistant varieties with high yield

In the late 1970s, a breeding program
for moisture-stress areas was initiated
to develop early maturing, open-
pollinated varieties which could escape
stress through earliness. The breeding
program was begun with different
populations, of which four were found
to have good general combining ability.

Then, an Si selection procedure was
followed to improve agronomic traits,
such as days to flowering, plant height
and final yield potential. The result
from two cycles of the Si selection
program is shown in Table 4. From
this breeding program, an open-
pollinated variety was developed and
has been presented to the National
Variety Release Committee (NVRC) for
final approval.

An improved variety per se is not
sufficient for producing high yields, but
must be accompanied by proper
cultural practices. Agronomic research
is a necessary component for
developing packages of
recommendations for improving
production. The agronomic studies
emphasized to date are in the areas of
planting date, optimum population
density for varieties from the different
maturity groups, intercropping, crop
sequence, tillage practices, the effect of
row planting and crop water
requirements. Since Ethiopia is a large
country with many different
agroclimatic areas, the studies have
been carried out in different zones with
varying local conditions.

Table 4. Improvements in three characters after two cycles of S1 selection in four maize
populations, Ethopia

Yield Days to flower Plant height
Total o/o gain/ Total o/o red./ Total o/o red./
Population gain cycle reduction cycle reduction cycle

47B/52 8.6** 4.30 12.5** 6.25 9.8* 4.90
Indian synthetics 7.7** 3.85 13.0** 6.50 9.3* 4.60
Katumani (MI) 14.2* 7.10 12.5"* 6.50 16.8** 8.40
Neghelle (MI) 12.5** 6.25 10.8* 5.40 15.8** 7.90

x 10.75** 5.38 12.2** 6.16 12.9"* 6.45

*, Significant at P = 0.05 and 0.01, respectively

Soil fertility
In this research activity, various
aspects of fertilizer use are being
tested, such as timing and method and
rate of application (particularly of
nitrogen and phosphorus for different
soil types). Studies are also being
carried out on the use of green manure

Insect control
Entomology-Surveys are being made
to identify the different species of
insect pests found in the country and
their importance in maize production.
This step is necessary to arrive at
control measures (chemical, cultural or
biological) for an integrated pest-
management system. In the survey, B.
fusca was found to be the major insect
pest, causing heavy damage from the
seedling stage to maturity. It occurs at
altitudes of 1235 to 2600 meters.

Pathology-Very limited studies on
maize diseases have been conducted
over the last few years. They include a
disease survey, a loss assessment
study and a study of control measures,
such as varietal screening and
chemical control of some of the major
diseases. According to the routine
survey, there are fifteen diseases
affecting maize in Ethiopia, among
which rust (P. sorghi Schw.) and leaf
blight (H. turcicum Pass.) are the
major ones. Head smut (Sphacelotheca
relliana Kuktn. Clint.) has also been

found on certain state farms in the
southern part of the country. Other
diseases are minor, or their intensity is
not yet known.

Weed control
Weed problems in maize were found to
be varied and complex. In general,
annual and perennial grasses are less
of a problem than are the broad-leaf
species, which cause the greatest loss
and are less easily controlled. Apart
from these, Striga asiatica and
S. hermonthica were found to be the
major parasitic weeds in specific areas
of the country.' Some cultural and
chemical means of weed control have
been identified for the nonparasitic

Seed Production,
Marketing and Distribution

The national seed program was
implemented in July 1978 with the
establishment of the Ethiopian Seed
Corporation (ESC). However, large-
scale seed production and distribution
has been carried out only since 1980.
At present, maize seed production is
concentrated in the wet western
lowlands and in the southern part of
the Rift Valley. These regions are
important for both seed and grain
production. The production of maize
seed during the last four years (1980 to
1984) has been substantial (Table 5).

Table 5. Crop seed distributed by the Ethiopian Seed Corporation (ESC), 1980 to 1984

Year Maize Wheat Barley Sorghum Teff Haricot- Rape Soybean Sunflower Total
bean seed

1980 19,996 194,792 1,656 250 4,147 612 660 214,113
1981 25,746 224,413 1,596 1,757 1,834 146 832 40 267,364
1982 16,967 256,815 23,430 3,046 1,490 2,797 500 217 305,262
1983 26,155 186,088 8,936 3,256 1,047 860 36 47 226,425
1984 13,190 122,473 16,476 1,081 2,581 1,832 257 12 157,901
1985 118,8311/ 229,630 12,878 11,752 56,000 2,490 432 1,050 160 433,218
a Of this, 74,8312 is commercial seed, not certified but field approved

Seed prices before and after processing
are fixed by the government's central
planning council. Based on these
prices, the ESC delivers seed to the
state farms and sells it at their
processing stations; they do not have
extended marketing and distribution
facilities. Farmer associations obtain
their seed through the Agricultural
Marketing Corporation (AMC); the
AMC, along with other organizations
and individual farmers, gets that seed
at the ESC processing stations.

Research Challenges

Since the introduction of maize into
Ethiopia, various varieties have been
grown by farmers under the different
agroclimatic conditions of the country.
Cultivation techniques used by 95% of
the maize farmers include hand hoeing
and plowing with oxen. Row planting
is still not used, despite the efforts of
the IAR. The need for improved
cultural practices as well as improved
varieties is obvious.

In the past, late-maturing varieties of
maize were developed and distributed
to a few farmers in the major maize-
growing areas of the country. The
importance of these varieties is now
declining as a result of the changing
weather pattern over the last three or
four years; rains have begun late and
have stopped before crops have
reached physiological maturity. Hence,
the development of medium-maturing
varieties (120 to 130 days) is now

On the state farms, which account for
about 5% of total maize output,
production is semi-mechanized. The
need is great for uniform, high-
yielding, lodging-resistant varieties for
this sector, which has the necessary
manpower and sufficient production
know-how. The farmers in this sector
grow hybrid maize as well as open-
pollinated varieties.

Research Constraints

As mentioned earlier, since 1980 the
maize research program has been
organized into teams combining
different disciplines. Due to a shortage
of highly trained researchers,
individuals are involved in more than
one crop team, which leads to
inefficiency. In addition, most team
members still lack experience and/or
high-level training. Hence, although
there is a pressing need for upgrading
the present staff in terms of training,
this is not being done because of the
economic situation in the country.

Research activities are also affected by
a lack of facilities, such as laboratory
equipment, cold storage facilities,
irrigation at some research stations
and transport vehicles. For the present,
the removal of these constraints is
probably beyond the economic
capacity of the country.


1. Acland, J.D. 1971. East African
Crops. Longmans, London, England.

2. Ethiopian Mapping Agency. 1981.
National Atlas of Ethiopia. Addis
Ababa, Ethiopia.

3. Ethiopian Statistical Abstract, CSO.
1985. Addis Ababa, Ethiopia.

4. Gugsa, E. 1985. Seed Production in
Ethiopia. Ethiopian Seed
Corporation, Addis Ababa, Ethiopia.

5. Huffnagel, M.P. 1961. Agriculture in
Ethiopia. FAO, Rome, Italy.

Mr. Watts: Does Ethiopia receive any
foreign aid that is specifically for maize

Mr. Debelo: No, we do not. However,
the World Bank is aiding crop
production in general in our country.

Maize Research in Kenya: An Overview
J.A.W. Ochieng, National Agricultural Research Station,
Kitale, Kenya

Maize is the staple cereal diet of over
80% of Kenya's population of some 19
million. More than 90% of the maize is
currently produced by small-scale
farmers, often on farms as small as
0.25 ha or even less in some heavily
populated parts of the country. The
Kenyan Government aims at self-
sufficiency in the production of food,
including maize, through a 4% per
annum increase in crop production

Only about 33% of the land area in
Kenya is arable, a limitation imposed
mainly by rainfall regime. The arable
regions are divided into the following
agroecological zones:

* High-potential (HP) zone-unimodal
rainfall pattern: 1000 to 2200 mm,
1600 to 2300 meters altitude
* Medium-potential (MP) zone-bimodal
rainfall pattern: 700 to 1800 mm,
1000 to 1700 meters altitude
* Low-potential (LP) zone-scanty,
short-duration rains
* Coastal strip (CS)-hot, humid belt,
some saline soils

Maize varieties for Kenya have to be
tailored to fit these climatic patterns,
i.e., late-maturity varieties, designed to
take full advantage of the whole
season, for HP areas, medium-maturity
varieties, grown in two seasons a year,
for MP areas, early maturity varieties,
for drought escape, for LP areas, and
special varieties capable of
withstanding the soil conditions
prevailing in the CS.

Maize Research Achievements

Maize breeding in Kenya began in
1955, and since then has gone through
many phases:

* Assembling local land races of maize
of Tuxpeflo origin from farmers'
fields (1950s);
* Building synthetic populations to
form basic breeding stocks (Kitale
Synthetics II and III);
* Introduction of exotic germplasm,
notably Ecuador 573 and Costa Rica
76 from Central America, and a
search for maize with heterosis in
crosses with local strains (1959);
* Inbreeding and hybridizing local
germplasm and the release of the
first series of hybrids (early 1960s);
* Mounting a maize breeding
methodology study (MBMS) to
identify selection methods
appropriate for each objective, i.e.,
intrapopulation selection for
improved open-pollinated (OP)
varieties, interpopulation selection
methods, e.g., reciprocal recurrent
selection, for hybrids (HYB) (1965 to
* Incorporation of a comprehensive
breeding program into the MBMS for
developing the products of the
breeding program, i.e., OP for small-
scale farmers, HYB for large-scale
farmers (1967);
* Initiation of a qualitative genes
program, using maize endosperm
mutants, e.g., opaque-2 and floury-2
to improve the amino acid profile in
local maizes and brachytic-2 to
reduce plant height (1969 to 1978,
discontinued in 1980);
* Initiation of systematic maize
variety testing through National
Performance Trials (NPT) before
their release to farmers (1979), and

*Initiation of a breeding component
aimed at reducing field losses (pre
and post-harvest), i.e., phyto-
pathology and entomology (in
collaboration with the International
Centre of Insect Physiology and
Ecology (ICIPE) and selection for
maize stalk strength (planned to
start 1985-86).

The achievements of the Kenya maize
improvement program are summarized
in Table 1. Hybrid maize has become
so popular in Kenya that most farmers
in the HP and MP ecozones will not
accept anything else. To date,
however, Katumani Composite B and
Coast Composite are the only
commercial maize varieties available
for the LP and CS zones, respectively;
replacements for these are in the

In the late-maturity maize breeding
program based at Kitale, experimental
maize varieties yielding far better than

Hybrid 625 (the latest commercial
variety) have been identified. The new
varieties, in the final stages of the
National Performance Trials, yield 4 to
24% higher than H625, and some of
them have greater stability (lower
regression coefficients) over

Progress from population improvement
in Kitale Synthetic II (KSII), Ecuador
573 (Ec573) and the variety cross KSII
x Ec573, over eight cycles of reciprocal
recurrent selection (RRS), is
summarized in Figure 1. No significant
genetic improvement was detected in
either KSII (-14.3%, b = -0.46) or in
Ec573 (1.9%, b = 0.53). However,
significant genetic advance was
attained in the variety cross KSII x
Ec573 over six cycles of selection
(28.5%, b = 3.10), although a plateau
effect was discernible after the eighth
cycle. A 28.5% yield gain over eight
cycles is equivalent to a gain of 3.6%
per cycle or 1.8% per annum, still far

Table 1. Maize varieties released by the Kenya national breeding programs

Year of Yieldd-
release/ Yied (baqs/hal (o/o of Altitude Days to Potential Special
Variety Type- introd. Farm' Potential./ of KSM) (m) maturity ecozones problems Observations
KSM OP -- -- 100 -- -- Never grown
Ec573 OP 1959 Over 2200 -- -- Never grown
KS II OP 1961 -- 107 1700-2200 .- -- Not grown
H611 VC 1964 -- 142 1800-2400 105 High Too tall Not grown
H621 DC 1964 -- 132 1000-1700 100 High ? Not grown
H631 TWC 1964 -- 140 1000-1700 100 High ? Not grown
H622 DC 1965 54 62 135 1000-1700 100 High Streak
H632 TWC 1965 54 55 140 1000-1700 100 High Streak
H612 TC 1966 63 75 155 1500-2100 90 High
KCB 1967 25 -- 500-1600 65 Marginal Streak
H511 VC 1967 40 52 -- 1000-1700 60-70 Medium Headsmut, streak
H512 VC 1970 45 62 -- 1000-1700 65-80 Medium Headsmut, streak
H611C VC 1971 63 75 155 1800-2400 105 High Too tall
H613 TC 1972 68 75 166 1500-2100 100 High
CMC OP 1974 35 -- 0-1000 80 Coastal P. sorghi rust
H614 TC 1976 68 77 166 1500-2100 100 High
H625 DC 1981 76 87 176 1550-2100 95 High
a KCB = Katumani Composite B, CMC = Coast Maize Composite, OP = open-pollinated, VC = variety-cross hybrid,
DC = double-cross hybrid, TWC = three-way cross hybrid, TC = top-cross hybrid
SSource: Report on Research Programmes: Achievements, Constraints and Training, Director NARS, Kitale,
Kenya, 1982 (adjusted down 100/o)
cSource: National performance trials (late-maturity maize), transformed means from combined analysis over
nine environments
d Source: Crop Improvement in East Africa, C.L.A. Leakey, ed., 1970

below the target of 4% per annum. A
previous preliminary evaluation from
estimates of genetic variance
components had indicated a genetic
advance of 7.3% per cycle by the RRS
method in KSII x Ec573.

A program of selection for prolificacy
in Kitale Composite B (KCB) and Kitale
Composite E (KCE) by the full-sib
method was initiated at Kitale in the
late 1960s. Four cycles of selection
revealed inconsistent changes in yield
over cycles, but percent prolificacy was
increased by 20 to 30%. The
composites, now under RRS for
increased prolificacy, have a fairly high
percent of prolificacy, approximately
15 to 25%. Table 2 summarizes the
average heterosis for yield in the
varietal cross KCB x KCE over two
cycles of full-sib selection for



/ C

/0 l-e R11 x R12 cycle crosses:
70 / Y=67.05 + 3.71 x 0.15X2
KSII (R11) cycles:
SY Y= 53.75-0.46X
rT- -- Ec 573(R12) cycles:
-6 60 Y=44.82+82+0.53X

S a
a a
a b
a Y
b a b b
b mb

40 I
0 1 2 3 4 5 6 8
Cycle of population improvement

Figure 1. Regression of population and
hybrid yields on cycles of improvement,
Cycle Evaluation Trials, Kenya, 1983

Maize Research Constraints

The main constraints in the Kenya
maize improvement and production
programs fall within three principal

* Technical
Plateau effect in the basic breeding
stocks KSII and Ec573 due to
erosion of genetic variability
Inconsistent progress from selection
in KCB and KCE due to poor
heterosis between the two
Poor harvest index ( < 30%) in most
late-maturity hybrids
Pre-harvest losses to maize diseases
(maize streak, headsmut, common
smut), pests (stalk borers) and
lodging, especially stalk lodging
before grainfill
Post-harvest losses to insects, such
as weevils, grain moths and the
greater grain borer

Resistance to hybrid maize adoption
by some farmers on the claimed
basis of low palatability
Low test weight of the kernels
Poor tolerance to witchweed (Striga
spp.) in western Kenya

Limited arable land for expanding
maize production Erratic rainfall
patterns in some traditional maize-
growing areas, presumably caused
by the encroaching desert

The Use of Quality Seed

These has been a distinct upward
trend in the use of improved seed in
Kenya between 1963 and 1981. This is
reflected in Figure 2, which shows the
number of hectares planted to hybrids.
Large-scale farmers predominated in
hybrid use until 1968, when the
number of small-scale farmers growing
improved varieties began to increase

Table 2. Heterotic patterns for yield from cycles of full-sib selection in Kitale
Composite B (KCB), Kitale Composite E (KCE) and the variety cross

Cycle of
selection No. of
and year of obser- Yield (q/ha) O/o average
evaluation nations KCB KCE KCB x KCE heterosis

Cycle 0
1968 11 48.9 52.4 56.1 110.9
1969 7 55.2 59.9 69.9 121.5
7 68.2 75.9 88.2 122.4
7 37.9 38.4 45.5 119.3
1972 4 42.6 46.9 46.1 103.0
1973 4 31.8 26.6 33.5 114.7
1974 2 67.0 78.4 93.4 128.5
1975 5 69.9 46.8 57.1 100.0
1976 6 59.0 55.2 52.9 92.6
Meanra/ 52.87 53.05 59.44 112.24

Cycle I
1972 4 46.3 47.6 47.5 101.2
1973 4 32.3 34.1 33.8 101.8
Meana-/ 39.30 40.85 40.65 101.43

Cycle 2
1973A 4 33.6 36.4 35.7 102.0
1973B 4 78.4 86.9 90.2 109.1
Meana- 56.00 61.65 62.95 107.01

a/ Means weighted according to number of observations

Hectares (0000)


40 | Large-scale farms (> 5 ha)

| Small-scale farms (<5 ha)


20 ---


63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81

Figure 2. Hectares of hybrid maize, Kenya, 1963 to 1981
Source: Sales Department, Kenya Seed Company

Production of certified seed in Kenya is
the responsibility of the Kenya Seed
Company (KSC), which obtains
parental materials (inbred lines) from
national research programs. The KSC
then puts the lines through a screening
program to check on synchronization
of flowering dates for the male-
designated and female-designated lines.
The company is responsible for
maintenance of the lines as well as of
the composites (Katumani Composite B
and CCM), once they are released from
breeders' stock. Cross-checking for
trueness to type is performed annually
by the National Seed Quality Control
Service (NSQCS) by growing out the
lines under maintenance by the KSC
against breeders' stock in post-control
plots. The NSQCS also routinely grows
all commercial maize varieties (both
old and new) in large plots to check for
distinctness, uniformity and stability
(DUS) every year; they also inspect
seed-production fields through all of
the required stages for the purpose of

Currently, the major constraint to seed
production seems to be a lack of
adequate isolation, since farms are
becoming smaller and smaller in the
face of increasing population pressure.
The former large-scale farms are being
subdivided into small parcels under
settlement schemes.

On-Farm Research

The difficulty in the dissemination to
farmers of new information based on
research findings has tended to be a
constraint in Kenya. This has been
due, in large part, to the researchers'
inappropriate approach to technology
transfer and to deeply entrenched
beliefs and practices, especially among
small-scale farmers.

Lately, new approaches have been
sought to address this problem and to
bridge the gap 'between the practices of
farmers and those of researchers. The

Training and Visits (T and V)
Extension Project, funded by the World
Bank since 1982,1 is one such
approach. The method involves
creating and maintaining close links
between agricultural research scientists
(ARS), especially crop agronomists, and
the agricultural extension subject
matter specialists (SMS) through
monthly workshops. There such
matters are discussed as land
preparation (technique and timing),
fertilizer types and rates, intercropping
maize with various crops, planting
density and weed management.

The information coming out of the
workshops is written in language the
farmers can understand by the SMS
and is relayed to the farmers by
technical assistants. Projects such as
adaptive trials are then jointly
conducted by SMS and ARS in the
fields of contact farmers. These
adaptive trials serve as demonstrations
for those follow-up farmers who live
near the contact farmers.

A similar project, involving not only
crop agronomists but also agricultural
socioeconomists, has been launched in
some parts of Kenya under the
auspices of CIMMYT. The new project
does not differ fundamentally from
T and V, except that it is divided into
several stages:

* On-station research, which Involves
precision experimentation requiring
a high degree of error control, high-
risk research with new chemicals,
experiments requiring back-up
laboratories, etc.;
Exploratory research, which
includes agroeconomic farmer
surveys and agronomic
experimentation simulating farmer
practices. The surveys enable
researchers to identify
recommendation domains, become
more familiar with farmer practices,
and define areas in which further
research is needed;

* Levels experimentation, in which
economic levels of agricultural
inputs versus sufficient levels
required to elicit crop response are
investigated for the formulation of
technological alternatives for maize
Verification experiments, in which
comparisons between farmer
practices and research
recommendations are made to guide
future research, and
Experimental production plots,
which are used for demonstrating
factors of production on a large scale
(at least one hectare) in maize plots
managed jointly by the extension
staff and a farmer.

If, after these various steps have been
followed, the farmer still does not
adopt the technological packages
recommended by the researchers, it
can be concluded that nonadoption is
due to socioeconomic constraints and
not to an information gap. This project
is still in its infancy (began late 1984),
and time will tell whether it will be
effective in Kenya for bridging the gap
between research findings and farmer


1. Annual Maize Tour Report 1979.
Western and Nyanza provinces,

2. Annual Maize Tour Report 1984.
Kericho, Nandi and Trans-Nzoia
districts, Kenya.

3. Darrah, L.L., S.A. Eberhart and
L.H. Penny. 1972. A maize
breeding methods study in Kenya.
Crop Science 12:605-608 (reprint).

4. Darrah, L.L., S.A. Eberhart and
L.H. Penny. 1978. Six years of
maize selection in Kitale Synthetic
II, Ecuador 573, and Kitale
Composite A, using methods of the
comprehensive breeding system.
Euphytica 27:191-204.

5. Eberhart, S.A., M.N. Harrison and
F. Ogada. 1967. A comprehensive
breeding system. Der Zuchter

6. Harrison, M.N. 1970. Maize
improvement in East Africa. In
Crop Improvement in East Africa
Commonwealth, C.L.A. Leakey, ed.
Agricultural Bulletin, Farham
Royal, Bucks, UK.

7. Harrison, M.N., S.A. Eberhart,
P.T.S. Mirie, A.Y. Allan, W.H.
Verburgt and F. Ogada. 1968. How
hybrid seed is revolutionizing
maize growing in Kenya.
International Development Review.
Washington, D.C., USA.

8. Kitale maize: The limits of success.
1980. AID Project Impact
Evaluation Report 2.

9. National Agricultural Research
Station. Annual Reports 1966 to
1983. Maize research programs.
Kitale, Kenya.

10. Onyango, R.M.A. 1984. Report on
maize production training course
at CIMMYT, Mexico.

11. Republic of Kenya. 1981. National
Food Policy. Sessional Paper No. 4.
Nairobi, Kenya.

12. Tranferring technology for small-
scale farming. 1981. In ASA
Special Publication 41, Noble R.
Usherwood, ed.

The Maize Program in Kenya
E.W. Mwenda, Embu Agricultural Research Station, Embu, Kenya

Kenya lies astride the equator on the
high East Africa plateau. It averages
about 1500 meters in elevation,
although altitude over the country as a
whole ranges from sea level to about
3500 meters. Because of the country's
geographic position, daily variations in
temperature are very small, usually
less than 50C; however, temperatures
can vary significantly as a result of
seasonal changes and altitude.

Kenya has a land area of about 60
million hectares. Maize is the most
important crop and is high on the list
of marketed agricultural products. It is
grown on about 1.5 million hectares, a
large portion of the country's limited
arable land.

Maize Breeding

Kenya's maize breeding program,
which was begun in 1955 at Kitale
Agricultural Research Station, has
released more than ten improved
hybrids and varieties for commercial
production. The early ones yielded 30
to 80% more than local varieties,
depending on the ecological area. For
the high-potential areas, the newer
hybrids are at least 20% better than
those early ones, with a potential of 12
t/ha. This exceeds current maize yields
by nearly 700%.

Figure 1 shows the location and
rainfall of the five agricultural research
stations in Kenya. Currently, there are
five major maize breeding programs,
the National Agricultural Research
Station at Kitale for late-maturity
maize, the Agricultural Research
Station at Embu for intermediate-
maturity maize, Nyandarua for high-
altitude maize and Msabaha for coastal
maize; the Dryland Farming Research
Station for early maturity maize is at

Late-maturity maize
The late-maturity maize breeding
program was begun in 1957 at the
National Agricultural Research Station
at Kitale, a substation of the Njoro
Plant Breeding Station. Kenyan inbred
lines were developed from Kitale
Station maize, which was basically
Kenya flat white. The first generation
of these inbred lines was tested by a
form of progeny testing, and a
minimum of ten of the best performers
were merged to form the new
synthetic, Kitale II. In 1961, Kitale
Synthetic II was released to farmers
west of the Rift Valley, where it
outyielded the ordinary Kitale maize
by 10 to 20%.

Single-cross hybrids were also tried
and proved promising, to the extent
that a group of inbred lines, as an
average of all their crosses, yielded
30% better than Kitale Synthetic II.
Consequently, the double-cross hybrid
622 and the three-way cross H632
were released for commercial

SMore than! 750 mm
SB 500 750 mm
E[ Less than 500 mm

Figure 1. Mean annual rainfall in Kenya

production in 1965. It was also found
that certain crosses of Kenya maize
with Central American open-pollinated
varieties on both sides had very good
crossing value. Crosses of Kitale
Synthetic II to Costa Rica 76 and
Ecuador 573 yielded 40% above the
best parent, Kitale Synthetic II. This
led to the release of H611, the variety-
cross hybrid of Kitale Synthetic II x
Ecuador 573, in 1964. Others that
have been released to date are the
hybrids 612, 613, 614 and 625.

The station at Kitale serves the areas
west of the Rift Valley, where there is
precipitation of 750 to 1778 mm
during the seven-month rainy season.
The type of maize bred for these areas
takes six to eight months to mature.
Additional hybrids are being tested in
the National Late-Maturity Maize
Performance Trials.

Intermediate-maturity maize
Breeding of intermediate-maturity
maize for the central part of the
country is carried out at the Embu
Agricultural Research Station. Maize in
this area, where 350 to 750 mm of rain

falls in two distinct seasons (April to
September and October to February or
March), takes five to six months to
mature. Table 1 is a summary of the
climatic data for Embu for a 6-year

The Embu program was started in
1965. Two hybrids, 511 and 512, are
commercially available, and many
others are under study. Rapid progress
has been made because of the two
seasons per year which allow for
breeding two generations; with
irrigation, a third generation is
possible. Also, the work has gone faster
because of the experience already
gained at Kitale.

Two composite populations, Embu 1
(El) and Embu 2 (E2), form the basic
breeding stock at Embu. They are
currently being improved through
reciprocal recurrent selection. Embu
Composite 3 was formed in 1980 and
has been improved by mass selection;
five local varieties were merged in its
formation. More hybrids are being
tested in the National Medium-Early
and Coast Maize Performance Trials.

Table 1. Climatic data for Embu Agricultural Research Station, Kenya, 1977 to 1983

Total Mean Mean rel. hum. Mean daily open pan
Days of rainfull temp. (0/0) sunshine evaporation
Year rain (mm) (OC) 9:00 a.m. 3:00 p.m. (hrs) (mm)

1977 119 1508 18.9 58.5 -- 6.7
1978 139 1581 17.9 80.3 62.9 6.2 1402
1979 112 1320 18.3 83.7 60.1 6.2 1538
1980 95 1071 19.3 79.2 52.1 6.5 1803
1981 100 1226 17.4 76.1 52.3 5.7 1539
1982 114 1464 19.4 81.6 57.4 7.0 1744
1983 76 1081 19.6 80.7 55.3 6.7 1628

Mean 108 1322 18.7 77.1 56.7 6.4 1609

Note: The station has a bimodal rainfall pattern, the long rains occurring March to June and
accounting for about 600/o of annual rainfall. The short rains occur October to
December and account for about 300/o. The period January to mid-March is hot and
dry. The annual P/E is about 680/o.

Early maturity maize
This program is based at Katumani
(Machakos). The two breeding streams
which have formed the basis of the
program are Katumani Synthetic VII
and Katumani Synthetic VIII. The
Katumani maize currently on the
market is Katumani Composite B,
which was derived from the original
cross of Katumani Synthetic V and
Katumani Synthetic VI and was
released for commercial production in
1967. A second population, which is
much earlier than Katumani, has been
developed from the lines derived out of
Taboran and top crossed by the
progeny (ex French via Malagasy); it is
at present being evaluated under the
name of Makueni.

Coastal maize
This program is located at Msabaha
Agricultural Research Station at Kilifi.
Coastal Composite Maize has been
developed there and is already
commercially grown in large areas of
the coast. This open-pollinated variety
compares favorably with the
intermediate-maturity hybrids 511 and
512. It has been decided that a second
breeding population will be formed of
Jamaican lines and materials from

High-altitude maize
The Nyandarua Agricultural Research
Station at 01-Joro-Orok is breeding
maize for altitudes above 2000 meters;
at those elevations, the local maize
planted by farmers takes 12 months or
more to mature. Ten percent of the
total area of Kenya is located at this
elevation, and farmers have
continuously tried to grow maize
without success.

The program of maize improvement for
high altitudes was started later than
that for other areas. The first maize
population formed is now referred to as
High Altitude Composite (HAC.) After
its formation, the population was
slowly improved until the 1970s, when
intensive ear-to-row selection was

begun. Later, half-sib selection was
used to extract lines, since a tester
with good combining ability was
already available in Ecuador 573. The
lines so far extracted are intended to
produce hybrids for these areas. The
main problems of maize grown at this
altitude are maturity and the danger of

Maize breeding system
The comprehensive maize breeding
program developed at Kitale has been
the basis of all of the breeding
programs in Kenya. The essential
features are:

* Evaluation of local and exotic
materials to assess their merits for a
long-term breeding program;
* Formation of two or more
composites of the selected material,
so that each population has wide
genetic variablility and the potential
for crossing well with the other
* Use of recurrent selection in the
populations so that their crosses are
improved with each cycle, and
* Release of a commercial variety as a
cross of two populations, as single,
double or three-way crosses from
the elite lines or as a synthetic
variety derived from the advanced
generations of the population

Seed Production and Distribution

The importance of a program capable
of supplying good quality seed to
farmers cannot be over-emphasized;
the success of maize production in
Kenya has been largely due to well-
organized seed production and
marketing. The relationship between
the various agencies involved is
described in the following seed-
production sequence:

Stage 1. Breeders' seed
Inbred development and maintenance
Open-pollinated variety development
National Performance Trials
Variety release and naming

Stage 2. National Seed Quality Control
Quality control and testing for yield
Inspection of seed growers' fields

Stage 3. Kenya Seed Company
Seed multiplication
Seed drying and conditioning
Seed processing, sizing, treating and
Seed storage, labeling and distribution

Stage 4. Seed marketing and
Kenya Farmer Associations
Farmer cooperatives
Small stockists

The seed for hybrids and open-
pollinated varieties that are presently
available in Kenya is shown in Table 2.

Table 2. Hybrid maize seed available in Kenya, 1985

Yield Altitude Length of
(/o of range rainy season
Hybrid H 613) (m) (months) Observations




134 1500-2100

Open-pollinated variety
Katumani 54
Coast 64



6-8 To obtain good results
5-7 from hybrid maize, a high
5-7 level of inputs is essential
5-7 Land preparation: Early
4-7 plowing for a good
3-4 seedbed ready for planting
3-4 when the rains start
Population: Between 40,000
and 53,000 plants/ha,
depending on rainfall reliability
and soil fertility; two plants
per hill spaced 75 x 50 cm
or one plant per hill spaced 75
x 25 cm
Fertilizer: Phosphate essential;
farmers should seek the
advice of their agricultural
officers regarding top-dressing
3-4 Heat-tolerant tropical hybrid
produced under licence with
the Pioneer Hybred Seed
5-7 Similar to H614 but
higher yielding

2-3 Short-season crop

3-4 Good heat tolerance; tolerant
to leaf rust

Maize Agronomy

The maize agronomy research program
has identified the factors which limit
maize yields in Kenya. In order of
importance, these are:

* Land preparation and planting date
(estimated as the main contributors
to yield);
* Weed control and planting density;
* Use of suitable hybrids or varieties;
* Use of fertilizers in appropriate
quantities, and
* Pest control and harvesting date.

In the Kenya Ministry of Agriculture
food policy document (8), projections
were made that, if the country were to
become self-sufficient in maize,
production growth rates of 12.7%
between 1980 and 1983 and of 6.8%
between 1980 and 1989 would be
necessary. These estimates included
maize for livestock feed and industrial
uses. The Economic Planning Division
of the MOA forecast the need for a
maize production growth rate of 85%
between the years 1983 and 2000 if
self-sufficiency were to be maintained.

Dissemination of
Research Information

To extension
There are many ways in which
extension can help in the
dissemination of research findings. It is
important that staff members be
included at seminars and field days at
the research stations, and that the
agricultural information centers keep
them informed so that they can pass
information on to farmers. In order for
extension to perform this task
effectively, there must be direct
contact between extension subject
matter specialists and the research
stations; this can be accomplished
through training and visits to projects.
The annual maize tours are also a
valuable part of the program.

To the farmer
The group approach has been found to
be a good method for reaching farmers,
through field days at research stations
and through field demonstrations.
Training sessions can also be held at
the farmers' training centers. Each
research station: should organize at
least one field day per year for farmers,
although the farmers should also be
encouraged to visit their research
stations whenever they have problems
that they wish to discuss.


1. Annual Reports for 1978-79, Kitale,
Embu, Katumani and Coast
Msabaha Research Stations, Kenya.

2. Darrah, L.L., S.A. Eberhart and L.H.
Penny. 1972. A breeding methods
study in Kenya. Crop Science
12:605-608 (reprint).

3. Eberhart, S.A., M.N. Harrison and F.
Ogada. 1967. A comprehensive
breeding system. Der Zuchter

4. Eberhart, S.A., L.H. Penny and M.N.
Harrison. 1973. Genotype by
environment interactions of maize in
Eastern Africa. East Africa
Agriculture and Forestry Journal

5. Harrison, M.N., S.A. Eberhart, P.T.S.
Mirie, A.Y. Allan, W.H. Verburgt and
F. Ogada. 1968. How hybrid maize
is revolutionizing maize growing in
Kenya. International Development
Review. Washington, D.C., USA.

6. Mwenda, E.W., and K. Njoroge.
1982. Maize production in Kenya.
Paper presented at an IITA seminar,
Ibadan, Nigeria.

7. Njogu, N. 1979. Fifteen years of
maize agronomy research: What
dividends? Paper presented at a
National Agricultural Research
Service seminar, Nairobi, Kenya.

8. Republic of Kenya. 1981. National
Food Policy. Sessional Paper No. 4.
Nairobi, Kenya.

Maize Research in Lesotho
P.P. Ntlhabo, Agricultural Research, Thaba Tseka, and M.T. Matli,
Agricultural Research, Maseru, Lesotho

The nation of Lesotho is entirely
surrounded by the Republic of South
Africa. The country may be divided
into three main agroecological regions,
the lowlands, the foothills and the
Maloti mountain range, where land is
being farmed at elevations as high as
3048 meters. Each region has its own
distinct climatic characteristics and
agroeconomic potentials and problems.

For the low elevations in Lesotho, most
of the maize varieties (usually hybrids)
are imported from South Africa; the
hybrids are found to outyield open-
pollinated varieties. In the
mountainous regions, hybrids usually
do not yield well because of the shorter
growing season and cooler conditions.
There, open-pollinated composite
varieties better meet the farmers'

Maize Research

The maize research program of
Agricultural Research Lesotho is
carrying out a number of experiments
to increase maize production in the
country. Maize variety trials are
usually planted in September with
varieties from CIMMYT (highland
Mexico) and South Africa, as well as
local varieties. No fertilizer is used;
Thiodan drench is usually applied at
planting to control cutworms and other
insects. The maize is harvested in May
or June, with counts made of the
number of plants per plot and the
number of plants harvested. Earworm
and ear rot incidence is recorded. The
ears are shelled manually, and grain
weight and moisture content are
determined. The weights are adjusted
to 15% moisture content and the yield

In the trials, the standard pattern of
75 cm between rows and 50 cm
between hills is used; three seeds are
planted per hill. Plot size is 3 x 3
meters, planted in a randomized block
arrangement with four replications of
each variety. The plants are thinned to
two per hill after emergence. After
harvest, the results are analyzed

Considering the yield of maize varieties
in both station and on-farm
experimental plots, Mexican highland
maize has proved more successful than
other varieties over the past six years.
The best two varieties, highland early
white flint and highland early white
dent, have yielded 62% better than the
two best South African hybrids, SA4
and SA11, and they have yielded
108% better than local varieties
(Figure 1).

Insect control
Cutworm control-In insect control
experiments, highland early white dent
is sown in plots measuring
3 x 3 meters with four replications.
Chemicals are applied either at seeding
time or on emergence of the young
plants. Thiodan is applied at the rate of
16.5 liters per hectare, using one
tablespoon of the 35% emulsifiable
concentrate in 12 liters of water,
applied to the soil with a watering can.
In addition to controlling cutworms,
Thiodan controls insects around the
germinating seeds.

Endosulfan 0.175% bait, which is
specifically for cutworms, is applied at
10 kg/ha on the soil surface, above the
seed or around the emerged plants.
Curaterr (10% carbofuran) is applied at
the same rate as Endosulfan, but is
placed in the ground with the seed; it
is a systemic insecticide and is mainly
effective against stalk borer. The plots
are harvested and average grain yields
determined from the two central rows
(yields are adjusted to 15% moisture
content). Yield results have shown that
it is more effective to apply the
chemicals at seeding than at
emergence. Thiodan gives the best
control, followed by Endosulfan and
Curaterr (Figure 2).

Stalk borer control-To test for stalk
borer control, highland early white
dent is planted at various locations in


3- -

plots measuring 3 x 3 meters. Curaterr
is placed in the ground with the seed
in one set of treatments as
recommended for stalk borer control.
In others, Curaterr, Thiodan and
Endosulfan are poured into the funnels
of the plants just before flowering. The
plots are harvested, the grain weight
and moisture content measured and
the weights converted to 15% moisture
content. The results of the trials are
shown in Figure 3.

Two applications of Curaterr, at
planting and preflowering, have been
found to be the most effective
treatment. One application of Curaterr,
either at planting or at preflowering,
gave only a slight increase in yield
over the control, as did treatment at
preflowering with Thiodan. When
Endosulfan was applied at


'C 0

.. E >

CD 0 CD C a)
"- I 0 I

r I J I I I 1 0
-- -- -v vo S .0
.5 .-S .r '. < "0 rC

Figure 1. Average yields of Mexican, South African and Basotho maize varieties, Thaba-Tseka,
Lesotho, 1979 to 1984

preflowering, yield was slightly lower
than the control. The systemic effect of
Curaterr persists only until the first
wave of stalk borers appears, when the
plants are still young. It seems that a
second application may be necessary
to control later attacks by the insect. It

Yield (t/ha)

3g -
a g8 5- -

Yid a) o o
2--a. -" 0.
C 0

0 .9 m
C 3 e- C 0
I-- LUJ h- U" 1

Figure 2. Yield of highland early white dent,
according to chemical treatment for cutworm
control, Lesotho

Yield (t/ha)
.' ---

2-- 1
2. S- --------


0 C
S- -- C- C-

S o o -
borer control, Lesotho-
s o -0 0

(..) UI- u LU

Figure 3. Yield of highland maize,
according to chemical treatment for stalk
borer control, Lesotho

had been supposed that both Thiodan
and Endosulfan would be effective
when applied in the funnels of the
plants, but only Thiodan was found to
be beneficial.

Maize agronomy
Maize yield tests have been conducted
under varying cultural regimes.
Highland early white dent was planted
in October in the standard plots of
3 x 3 meters with four replications. For
planting, the ground was either plowed
or left untilled. Seeds were planted in
rows or broadcast, and plots were
unweeded, weeded once or weeded
twice. The middle two rows of each
plot were harvested in June, and grain
weights converted to 15% moisture
content. Figure 4 shows the results of
the trial.

No-till planting, even with two
weedings, was not shown to be
significantly better than regular
cultivation with only one weeding. No-
till with a single weeding yielded

Tons per hectare

2- -


e 4. Yield of maize grown with
ig cultural practices, Lesotho




:5 5 | :
0 0

-- o 6
"0 -1 "

0 0 6
-- S --* .' --
! jo o o
2 i .
Z L Z a

significantly less than regular
cultivation with one weeding. This was
in contrast to the previous season,
when no-till with one weeding had
yielded significantly more than regular
cultivation with one weeding, probably
due to the better rainfall distribution in
the current season, which resulted in
heavier weed growth. In the previous
season, a drought occurred in the
spring (October and November), and
better spring growth was evident in the
no-till plots. As in other seasons, maize
yield was negligible with no weeding.

Planting date experiments-To test
maize yield against planting date,
highland early white dent. SA11 and
CIMMYT-German early maize were
sown at two-week intervals between
September 16 and December 8. Plots
measured 3 x 4 meters, with two rows
of each variety per plot and outside
barrier rows of HEWD; the plots were

replicated four times. Emergence was
very irregular, especially in the first
replications at the edge of the field.
The graph of yield and planting date
(Figure 5) shows that, as in past years,
yields diminished with each
succeeding planting in the beginning.
There was also the usual dip in the
middle of the season due to cutworm

For very late maize plantings, yields
increased with later plantings, starting
on November 24 and continuing until
December 8. This increase was less
marked for SA1l, but the yield of
HEWD went from 0.6 t/ha, when
planted on November 10, to 2.39 t/ha
when planted on November 24 and to
2.92 t/ha when planted December 8.
The CIMMYT-German maize showed a
similar pattern, yielding 3.37 t/ha for
the December 8 planting. The early
planted maize had showed lower

Figure 5. Maize yield as a result of planting date, Lesotho

yields, partly because of poor
emergence; this was probably caused
by drought at flowering in December
and January. The interesting
conclusion that can be drawn from this
experiment is that, in some seasons, it
is possible to get higher yields by
planting late, if a fast-maturing maize
is used and there is not an early killing

Genetic improvement
Half-sib recombination blocks-The
half-sib recombination block method of
genetic improvement is used at
CIMMYT to increase yields of maize
lines and to change them as desired.
This method consists of growing seed
from selected ears and plants in double
rows, with a single row in between
which is seeded with a mixture of all of
the selected ears. The tassels are
removed from the double rows, and
these serve as females. They are
pollinated by the plants in the single
rows, the males. The best plants in the
female rows are then used the next
year along with additional selections.

Highland maize has a few
disadvantages which are presently
being addressed. The plants are prone
to lodging in the strong winds of the
mountainous areas. Selection is being
carried out for shorter plants with
lower ears to correct this problem, and
at the same time, for large ear size,

good plant type and frost tolerance.
Plants in the seed multiplication fields
were selected the previous season and
either self- or cross-pollinated with
other selected plants, the ears being
saved for row planting in the half-sib
blocks. This procedure has been used
for all four highland varieties.

Hybridization-A second approach for
adapting the highland maize to local
conditions is to cross it with well-
adapted Basotho varieties. The latter
have low yields, but some are able to
withstand high winds and cutworm
attack better than the highland maize.
Basotho varieties tend to send out
several tillers which mature almost as
fast as the main shoot, i.e., there is
less apical dominance and the main
shoot is shorter. This gives lower
plants with several ears each. The goal
is to produce a plant with this form,
but with ears like those of highland
maize and the resistance to cutworms
that is found in the Lehalesebere (flint)
maize. Table 1 shows the crosses that
have been made to date.

The CIMMYT-German line is a fast-
maturing maize which was developed
by selecting for earliness in Mexico in
the winter and in Germany in the
summer. It can mature at Thaba Tseka
even when planted as late as the
beginning of December. The Basotho
yellow floury maize (Rantsimane

Table 1. Maize crosses made by Agricultural Research Lesotho

Female Male

Highland early white flint Basotho white flint
Basotho white flint (Lehalesebere) Highland early white flint
Highland early yellow flint Basotho yellow floury
Basotho yellow floury (Rantsimane Highland early yellow
khutseanyane) flint
CIMMYT white German Basotho yellow floury
Basotho yellow floury CIMMYT white German
CIMMYT white German Basotho white flint
CIMMYT yellow German Basotho yellow floury
CIMMYT yellow German Highland early white flint

khutseanyane), with which it is being
crossed, produces small ears, a
characteristic that has discouraged
farmers from growing it. However, last
season it yielded well in trials because
of the maturation of the ears on the
tillers. The cross-pollinated ears in the
trials were individually harvested and
labeled for planting the following

Seed Multiplication
and Distribution

From the multiplication of highland
maize varieties, 100 5-kg lots were
distributed to 50 leading farmers with
the understanding that they would
return the same amount of seed after
harvest. Each lot had information on
recommendations for planting.

Farmers had been told of the
advantages of planting highland maize
at courses at the farmers' training
centers, and their response has been
enthusiastic. Highland maize seed is
also sold by the Crops Research Seed
Multiplication Section.

Extension Activities

The extension service holds maize
demonstrations in the fields of leading
farmers. Field days are also held in
various locations to provide
opportunities for both farmers and
officials to see how seed multiplication
is carried out. The farmers and officials
can also visit on-farm experiments to
see the difference between local
varieties and varieties being developed
through the maize research program.

Maize Production and Research in Madagascar
L. Rondro-Harisoa and R. Ramilison, Minist6re de la Recherche
Scientifique et Technologique pour le Developpement,*
Antananarivo, Madagascar

Maize is the staple food in the southern
part of the island of Madagascar. In
other parts of the country, it is used as
a food complement and as feed for

Maize production is scattered over
nearly all parts of the island (Figure 1),
although most is produced on the high
plateaus (Table 1). Table 2 shows the
characteristics of the principal maize-
growing areas of Madagascar.

The maize crop is grown currently on
some 136,000 hectares, an increase
from 92,000 ha in 1962 (Table 3). Both
area and production have increased
over nearly all of the island districts
since 1980 (Table 4).

o North'

Mahajanga P ort-Berg0
el k. !
.; .; ..... Ambatondrazaka
,,,,r, :'.'.: Antananarivo Toamasina
T....,... ...v. .,....3,.ille
T .-,cL. l

Morondava..,. \ J -. ...

olryT of Capricorn
Toliary s

Figure 1. Maize production areas, Madagascar

* Department of Scientific and
Technological Development Research

There are two government farms, FEO
and FESA, in the midwestern part of
the island, as well as one
agroindustrial firm (SAGRIM), which
produce maize for feed and for export.
They use improved varieties yielding
up to 4.5 t/ha. Local market needs
cannot be met by small-scale growers
using local varieties and achieving
average yields of only 1 t/ha. Since
1981, maize production has been
increasing because of an increase in
price of almost 100%, from 53 to 100
FMG/kg (600 FMG = US$ 1).

Maize Research on Madagascar

Maize research was started by the
IRAM (Institut de Recherches
Agronomiques a Madagascar) in
1961-62. In 1974, FOFIFA or
CENRADERU (Centre National de
Recherche Appliquee au
D6veloppement Rural) took over the
research on varietal improvement and

Table 1. The location of maize production in
Madagascar, 1980

Location (O/o)

High Plateaus (Antananarivo
and Fianarantsoa) 62.1
South (Toliary and Morondava) 15.7
Alaotra 4.0
North (Antsiranana) 3.5
West (Mahajanga and northern
Morondava) 3.2

Table 2. Characteristics of the principal maize-growing areas of Madagascar

Temperature farmersa-
Principal Altitude ( C) Rainfall (principal groups
Region towns (m) Climate Max. Min. (mm) Soil types underlined) Observations

High Tananarive 1000- High 22.3 12.3 1250 Very desaturated 1-2-3 Dense population

Plateaus Antsirabe

Middle west Mandoto

Lake Ambatondrazaka

1600 tropical

700 High warm
900 tropical

700- High warm
900 tropical

Montage d'Ambre 250-500
Ambanja 30-500

Northwest Port-Berge
and West Maevatanana

23.3 10.3
23.7 13.2


ferralitic soils

1550 Fairly desaturated
ferralitic soils

26.6 14.7 1200 Fairly desaturated
ferralitic soils and
fluviatile alluvium

Semi- to very 29.0 18.0
humid tropical 31.0 20.4

1300 Fairly desaturated
2150 ferralitic soils on

Fluviatile alluvium
and ferruginous
tropical soils

Semihumid 30.0 19.8
warm tropical 29.7 21.7
32A 18.1

Very humid 31.0 20.4
warm tropical 29.4 20.1

10-500 Semiarid warm 29.7 17.9
100-400 tropical 28.7 17.1

2-3 Most favorable

2-3 Rice growing

2-3 Favorable

2-3 Irrigation

750 Ferruginous red
1000 tropical sand and
900 ferralitic soils

Ferralitic soils
and alluvium

350-700 Ferralitic and
600 ferruginous soils
and tropical soils
and red sand

1-2 High presence
of viral diseases

1 High risk area
for agriculture

a/ Target farmers: 1 = traditional farmers, 2 = farmers in the process of modernization, 3 = commercial farmers






20- 40






Maize breeding
Improvement of local populations-
Two stable and improved populations
were obtained by recurrent selection
for general combining ability in the
southern part of the island. They are
synthetics from Fianarantsoa and from
Tulear. These two populations were
intercrossed, forming the variety
Plata 264, which is now proposed for
release. The variety Tsakomalady has
shown resistance to virus diseases on
the east coast.

Formation of lines-More than 150
lines have been created from
introduced varieties.

Table 3. Area and production of maize,
Madagascar, 1962 to 1984

Area Production
Year (ha) (tons)

1962 92,000 85,000
1972 95,000 106,000
1980 128,000 127,200
1981 128,000 127,600
1982 116,500 112,800
1983 131,100 150,500
1984 136,300 160,500

Source: Agricultural Statistics Service,
Ministry of Agriculture

Hybrid formation-Several hybrids
have been obtained, three of which are
being recommended by the Extension
Service. They have had yields of 10 to
12 t/ha in station trials.

Polyhybrids-Polyhybrids have been
created in Madagascar by crossing
hybrids; they are used as synthetics or
composites. The best ones are the
white-grain polyhybrids 266 and 377,
which were obtained from H632, SR11,
SR13 and three lines from Natal. Three
polyhybrids with yellow grain are 374,
384 and 387.

Intervarietal hybrids-Three hybrids
recommended by the Extension
Service are 321, 375 and 383. The
latter is one of the best maize varieties
obtained to date, with a mean yield of
8 t/ha in station trials. This variety is
best known as a polyhybrid for use in
advanced generations, at which stage it
has given a mean yield of 6 t/ha in

Composites-Three composites have
been created, but only one is still in
use, Composite High Plateau, with 60

Table 4. Maize production in each faritany (district), Madagascar, 1980, 1983 and 1984

Faritany Area (ha) Production (tons)
1980 1983 1984 1980 1983 1984

Antsiranana 3,800 3,700 3,900 3,000 3,500 4,700
Mahajanga 13,000 13,900 15,200 10,400 15,700 15,200
Toamasina 10,400 13,500 12,100 10,700 13,500 12,500
Antananarivo 65,100 60,200 63,400 70,300 76,000 83,700
Fianarantsoa 15,000 15,400 16,300 14,700 16,200 18,000
Toliary 20,600 24,400 25,400 18,100 25,600 26,400

Total 127,900 131,100 136,300 127,200 150,500 160,500

Source: Ministry of Agriculture

Fertilizer studies
In fertilizer trials it was shown that
yield is lower on land being cultivated
for the first time; this was especially
true on the high plateaus. There,
production was almost nil, even when
large amounts of inorganic fertilizer
was applied. It was found that this
phenomenon could be overcome
through the use of a mixture of
chemical and organic fertilizers
containing manure.

Nutrient deficiency studies-The first
step in the fertilizer studies was the
qualitative determination of
deficiencies in maize plants grown in
pots. The results of these studies are
shown in Table 5.

Chemical fertilizer studies-The results
of the study of elements showed that:

* Nitrogen, phosphorus and potassium
are essential on the high plateaus (K
and sometimes P are less needed in
other regions);
* It is beneficial to apply N and K in
split applications;
* Dolomite is needed on acid soils and
on soils deficient in calcium and
* On virgin lands, potassium can be
lost through leaching, and
* Phosphorus should be applied
annually at the rate of 45 kg of
P205/ha, rather than in a large, one-
time dose to bring the phosphorus
level up to par.

Studies of mixtures of organic and
chemical fertilizers-These studies
showed that:

* When the crop residue is plowed
under, less fertilizer is needed
(especially potassium);
* The use of compost has a negative
effect, and
* Manure, composed of animal matter
and harvest residues, is often poorly
decomposed. The amount of manure
available is often limited to 5 t/ha
per year.

It is recommended that organic matter
be combined with chemical fertilizers.

The Present Status of
Maize Research in Madagascar

Varietal improvement
The varietal improvement program has
resulted in a national collection which
is utilized for:

* Maintenance of 112 pure lines and
176 varieties;
* Increase of certain varieties for
subsequent use in trials or in
forming promising varieties;
* Formation of promising varieties,
* Introduction of nine parent lines
(which were lost because of
maintenance problems), and the
introduction of two new varieties,
IRAT83 and IRAT200.

The recommended varieties and others
released to extension are evaluated in
two ways:

* Tests of varieties in order to identify
promising ones for a given location,
* Comparative variety trials for
studying their performance in other

From the 1983-84 variety trials, two
varieties were found to be best, 383 for
yield in high elevations and 374 for

The main areas studied since 1983-84
have been fertilizer use and seed

Fertilizer use-The main objective has
been the identification of economical
fertilization schemes adapted to
selected sites and to different levels of
production. Areas under study include
the maximum use of resources, such
as manure, straw, harvest residues and
dolomite, and rotating cereals with

Table 5. Fertilizer recommendations resulting from nutrient deficiency studies, Madagascar, 1960 to 1980

Recommended fertilization
Station Principal Observed fertilizer Annual fertilizer Observations
Region (altitude) soil type deficiencies (kg/ha) It/ha)

P205 K20 Dolo- N P205 K20 Dolo- Manure
mite mite (2)

High Ampangabe Ferralitic soils P,Ca,K,
Plateaus (1300 m) on acid rock S,Mg
(very poor)

High Ambohi- Ferralitic soils K,Mg,Ca,
Plateaus mandroso on basalt P (in
(1600 m) (very poor) fields)


400 300 2000 90-135 45-60 30-60 500 P and K main
60- 90 45 0-30 250-500 10-20 limiting factors,
also dolomite

600 350 2000 90-135 30-45 30 250-500 a/

Ferralitic soils P (medium) 250
on basalt
(fairly poor)

After 4 a/ N main limiting
years factor

Middle Kianjasoa Ferralitic soils P (medium) 300
west (1000 m) on glacial K (low)

North Anketrakabe Ferralitic soils P (medium) 250
(300 m) on basalt
(medium poor)

N main limiting


a/ Alluvial soils/ None

a/ Vertisolsa

South Ihosy

South- Ankazoabo
west (TulIar)

a/ No information available

soils (hydro-
morphic soils
on red sand)

0 0 0 90-135 0-45

a/ a/ a/ 90-120 45-90

P (medium) 100

P (low)

45- 90 50-60

5-15 N main limiting

a Liming not
necessary, slow
evolution of
organic matter

The areas in which fertilizers have
been tested are the high plateau, with
an elevation between 1000 and 1600
meters, and the midwest, with an
elevation of 700 to 1000 meters, a
relatively cool and humid tropical
climate (1200 to 1500 mm of rainfall
per year) and ferralitic soil.

Some of the recommendations from
the fertilizer studies are:

* The use of nitrogen fertilizers, even
on maize planted after legumes;
The use of a nitrogen supplement
with NPK fertilizers;
The use of a balanced mixture of
organic and chemical fertilizers, and
Further study of the use of NPK
with dolomite and manure and of
NK with manure.

Studies have also been made on soils,
and their responses to the essential
elements (N, P and K), to dolomite (Ca,
Mg) and to manure. These have been
carried out on the northern part of the
island, which has a low-elevation
climate (250 to 300 m) and moderately
poor volcanic soil.

Rotations-Studies have been made to
determine promising rotation systems
which include the two main food
cereals, maize and rice, and the
legumes, groundnuts, beans and
soybeans. The studies have been

conducted on the volcanic soils in the
north and on the ferralitic soils of the
Lake Alaotra area.

Cultural practices-Studies have also
been conducted in various parts of the
island on cultural practices, such as
time of planting x depth of seeding,
time of planting x variety and methods
of land preparation x Ca and Mg

Seed production
The seed production studies are a
continuation of the work on varietal
improvement and formation. FOFIFA
produces prebasic and basic seed for
the agronomy complex of Lake Alaotra
(CALA). Since 1976, the seed of three
polyhybrids (383, 377 and 266) has
been released to growers. It is
multiplied and distributed by CALA.
Table 6 shows the amount of seed
distributed in 1983-84.

Maize Research Staff

The genetic and varietal improvement
program has two Malagasy staff
members, a maize breeder and an
agronomist; there is one expatriate
breeder in the program. There are also
personnel in the areas of entomology,
plant pathology and soil science.

Table 6. Seed production and distribution, Madagascar, 1983-84

Variety (kg)
Receiving agency 383 377 266 374 Total Observations

MPARA MPAEF 690 8 8 706
Private societies 400 400 SAGRIM (Morondava)
Farmers 120 20 10 150
Testing program
(FOFIFA) 106 30 60 15 211 For testing
DRZV (FOFIFA) 1000 200 1200 For food

Total 2316 58 278 15 2667

The Relationship between
Research and Extension

Until 1982, varietal release was not
efficient in Madagscar because of a
lack of coordination between FOFIFA
and the Ministry of Agriculture
(MPARA). In 1982, a liaison service
(SALIAR) was created at MPARA to act
as a bridge between research and
extension. The service of plant
material (SMV) at MPARA has
established a national seed policy for
maize and rice. FOFIFA's seed
production program will be determined
by this service to satisfy the national


Important results have been obtained
since maize research was begun in
1961, in the areas of both varietal
improvement and agronomic practices.
After a four-year interruption, research
was begun again in 1983. Since then,
it has been oriented, in the short and
medium term, to the maintenance of
the main seed collection at Lake
Alaotra, the reconstitution of
degenerated polyhybrid lines intended
for extension, the reintroduction of
parental lines that had been lost, as
well as of new lines and varieties, and
to research into economical fertilizer

Maize Research and Production in Malawi
L.D.M. Ngwira and E.M. Sibale, Department of Agricultural
Research, Chitedze Agricultural Research Station, Lilongwe,

Maize is the staple food and major
source of carbohydrates for over 80%
of the people of Malawi. The rising
demand for food in recent years has
turned maize into not only an essential
staple food crop, but also into a cash
crop; it can be sold to the Agricultural
Development and Marketing
Corporation (ADMARC), which later
sells it to meet urban food
requirements and other needs. Maize is
now being grown for food even in those
areas where cassava, sorghum or
millet used to be the principal foods.

Maize Production

Although most of Malawi's maize is
produced on the central plateau, it is
widely grown throughout the country,
primarily by smallholders.
Approximately 970,000 hectares were
grown in 1980-81, of which about 80%
was grown in pure stands. This reflects
a dramatic shift from mixed cropping,
as more than 90% of the 1,070,000
hectares planted to maize in 1968-69
was in stands mixed with pulses. This
decline in hectarage was accompanied
by a modest production increase (13%)
between the two surveys. About 90%
of the production is of local flint
varieties which are preferred for home
consumption; most of the composite
and hybrid production is sold. The
production of maize in rotation with
tobacco on estate farms has increased
in recent years.

Maize yields in Malawi vary widely,
from less than 1000 to over 4000 kg
per hectare, depending on such factors
as location, variety and fertilizer use.
Conditions are suitable for maize
production in the drier areas of the
Shire Valley and the lake shore; there
the production potential is high.

Maize is grown at altitudes ranging
from a few meters above sea level to
1700 meters or more. The main maize-
growing areas are between 600 and
1300 meters above sea level, although
some maize is also grown in the
marginal areas above or below 'this
range. The marginal areas offer the
biggest challenge to research for the
breeding of varieties suited to their

There are three marginal areas for
growing maize. Parts of the Shire
Valley in the southern tip of the
country has a semi-desert climate with
a very short rainy season (two to three
months); rainfall is erratic and
unreliable. The very hot lake shore
area also has a short rainy season of
only three to four months. In the hills
in the northern and central parts of the
country, cool temperatures and
overcast conditions are unfavorable for
maize production.

Maize is grown in Malawi by two types
of farmers. Estate farmers have large
land holdings and can afford to invest
large amounts of capital in their crops.
They grow maize purely as a cash crop
and, therefore, prefer growing high-
yielding dent hybrids. The maize from
the estates is sold either to ADMARC at
a premium price or directly to the
Grain and Milling Company in Limbe,
Lilongwe or Mzuzu at a price higher
than that of ADMARC (to cover
transportation costs). The Grain and
Milling Company processes most of
this maize into flour for human
consumption, with a small amount
being processed into animal and
poultry feed. The company also buys
maize from ADMARC to meet the ever-
increasing demand for maize flour
from the urban population. The estate

sector plays an important role in
feeding the ever-growing urban
population, as well as in the National
Food Reserve Program.

Smallholders farm small parcels of
land (8 ha or less). These farmers grow
maize for food, with only surpluses
being sold. They normally have limited
capital and cannot afford to grow
hybrids, which have high demands for
fertilizer, without assistance loans for
fertilizer and seed. Most of them grow
local unimproved flint maize.

The average yield of unfertilized local
maize in Malawi is less than one-third
that of fertilized hybrid maize.
However, the local, unimproved flint
maize has some grain characteristics
that farmers like. The grain is resistant
to weevil attack in storage; therefore, it
stores well, even without pesticide
treatment. There is also less grain
breakage when the seed coat is
removed by pounding in the traditional
mortar and pestle. As a result, less
grain is lost along with the seed coat.
There is a need for the development of
improved semi-flint, open-pollinated
varieties so that farmers have available
higher-yielding varieties which demand
less fertilizer, as compared to the
hybrids, and are closer to the
unimproved local maize in grain
characteristics. Table 1 is an estimate
of the types of maize grown in Malawi
in the 1982-83 season.

National Policy

Malawi national policy is to increase
maize production in order to maintain
self-sufficiency in the rural areas and
to provide enough food for the growing
urban population. The government is
also trying to accumulate sufficient
grain reserves to meet the country's
needs in times of adverse weather. The
goal is to increase yields per hectare,
through improved seed and cultural
practices, the use of both manure and
chemical fertilizers, and effective
disease and pest control measures.

Maize Research

Maize research in Malawi is the sole
responsibility of the Department of
Agricultural Research (DAR) of the
Ministry of Agriculture. Maize seed
production and distribution is carried
out by the National Seed Company of
Malawi (NSCM), the Seed Technology
Unit of the DAR, and ADMARC. The
Seed Technology Unit was set up to be
responsible for seed certification and
quality, and the National Seed
Company of Malawi, seed production
an'd processing; ADMARC has the
responsibility for seed distribution. The
maize program is designed to develop
high-yielding varieties and cultural
practices for both the high-potential
and marginal areas of the country and
for both estate farmers and

Table 1. Estimated maize production, Malawi, 1982-83

Total o/o of
Area O/o of production total Yield
Type of maize (ha) total (tons) production (kg/ha)

Local unimproved 1,067,525 90.2 1,017,114 77.7 952
Composite 26,954 2.3 46,097 3.5 1709
Hybrid 89,004 7.5 245,573 18.8 2759
Total 1,183,493 100.0 1,308,784 100.0 1106

Maize research is coordinated from the
DAR Chitedze Research Station,
situated 16 kilometers west of
Lilongwe. The research is
multidisciplinary, with the program
divided into the areas of breeding,
agronomy, pathology and entomology.
A unit has been formed recently to
carry out, among other things, on-farm

The maize breeding program
A systematic maize-breeding program
was set up in the early 1950s, and the
production of inbred lines from a wide
range of base materials was begun at
that time. Until 1971, emphasis was on
the production of synthetic varieties
(SV) and double-cross hybrids (coded
LH for local hybrid). The lines were
recombined in the early 1960s to form
synthetic varieties or crossed into
double-cross hybrids. This approach
was successful, and a number of
synthetic varieties (SV17, SV28 and
SV37) and a hybrid variety (LH11)
were released to farmers after testing
in the mid-1960s. These varieties were
grown for a long period.

In 1967, a breeding program for the
formation of composite varieties was
initiated. Random pollination of some
20 varieties, which included local
synthetics and hybrids and a few
introduced materials, was carried out
for three generations. In 1971,
recurrent selection was started with
the new population thus formed; it was
named Chitedze Composite A (CCA).
At that time, emphasis was shifted
from synthetics to composites, and the
hybrid program was suspended.
Chitedze Composite B (CCB) was
formed almost exclusively from exotic
materials and was very broad based.
The selection criteria used in the
program was grain yield, grain
characteristics and ear and plant
height. The S2 testing method of
selection was chosen, since it appeared
to make the best use of the resources
available. Table 2 shows the schedule
that was followed.

However, problems were experienced
with this method. It was difficult to get
enough S2 ears because the S1 lines
did not grow well because of reduced
vigor and consequent poor seed set. In

Table 2. Maize S2 selection method used in Malawi, 1971 to 1973

Cycle Site Operation

First wet season Chitedze From 1 ha, 1500 plants were selfed for
(1971-72) Sls, 300 ears selected
First dry season Makhanga-/ 300 Sls planted ear-to-row and selfed
(1972) to S2
Second wet season Chitedze Yield trials for best 200 S2 entries,
(1972-73) Bvumbwe two replications at each site, best 10
Mbawa to 20 entries selected based on yield
and desirable agronomic characteristics
Second dry season Makhangaa/ Selected entries recombined by the
(1973) Kitale "Irish method" of planting best ears
from selected families ear-to-row, the
resulting best ears mixed in equal
proportions to represent one improved
cycle of the population

a/ Grown under irrigation

1974, it was decided to change the
selection method to modified S1
testing. Therefore, the top-crossing of
the S1 lines to the original population
(CO) in the first dry season was
substituted for the S2 production. The
remnant seed of the best 10 to 20
families of the S1 lines were then
recombined in the same manner as in
the S2 method. Two other populations,
Ukiriguru Composites A and B (UCA
and UCB), were introduced from
Tanzania and also underwent this
method of selection.

In the mid-1970s, the two composite
varieties, UCA and CCA, were released
for high-potential and low-potential
areas, respectively. Farmers expressed
dissatisfaction with them because they
were too tall. A recurrent selection
program was therefore started to
reduce the ear height of these
composites. Some progress has been
achieved in ear height reduction in
both varieties (Table 3).

The hybrid program was revived in
1977 in order to satisfy the country's
demand for hybrid seed, which had
risen due to an increase in the number
of commerical maize growers. Until
that time, all hybrid seed had been

Table 3. Mean grain yields of two maize
varieties over three seasons of testing, Malawi

Grain yield Plant height Ear height
Cycle (kg/ha) (cm) (cm)

C0 5785 301 184
C1 6253 282 165
C2 6646 293 172
CO 6678 285 185
C1 6851 293 165
C2 6525 277 158
SE I 666 8 64
CV (o/o) 5.6

imported from neighboring countries.
By producing its own hybrid seed,
Malawi could free much-needed foreign
exchange for other development
projects. The hybrid program was to
be followed along with the composite
program (for open-pollinated varieties)
which catered to smallholders who
grew maize with fewer inputs, either
for home consumption or for the

The hybrid program-In 1977, the
development of inbred lines from local
and exotic populations was initiated.
Inbred lines were also acquired from
cooperating institutions outside the
country. The ear-to-row inbreeding
method was used. Selections were
made both within and between
families, with the populations involved
being CCA, UCA, Ecuador 573,
Cortazar and TL73B. At S2 the lines
were evaluated for general combining
ability (GCA), and selfing was
continued only in the lines which
showed good GCA. At S6 the lines
were evaluated for specific combining
ability (SCA). Three new high-yielding
hybrids have been released from this
program, for both the high-potential
and the marginal areas. Three lines of
these hybrids are now with NSCM for
bulking and seed production and
distribution to farmers; they are
CXH66, CXH74 and CXH43.

The composite program-In the
composite program, there is
continuous population improvement
for both high-potential and marginal
areas. For high-potential areas, the
emphasis is on high yield, late-to-
medium maturity, low ear placement,
short plant height and disease and pest
resistance/tolerance. The populations
undergoing improvement in this
program are UCA and CCC. Work is
also underway to synthesize new
populations using local and exotic
materials. Like the hybrid program,
the collection and evaluation of new
introductions is an ongoing process.

The present maize agronomy program
is an outgrowth of the ODM project.
However, the size and scope of the
program has changed, because maize
cultivation in Malawi has extended to
areas that in the past were under
cassava, sorghum or millet as food
crops. New problems have also come
about because of increased maize
production. In some areas, maize
monoculture has become a practice,
since maize is now being used as both
a food and a cash crop, and in a few
cases, because of population pressure
on the land. Nutrients which were in
abundant supply in the soil are now
becoming deficient, because of the use
of high-yielding varieties which are
more demanding of soil nutrients. This
is particularly true of the micro-
nutrients, especially boron and sulfur,
and in some areas, potassium. Weeds,
especially witchweed (Striga asiatica),
are also becoming a problem because
of monoculture or insufficient rotation.
The land is not allowed to rest long
enough between crops to reduce the
incidence of witchweed.

The maize agronomy program is
charged with developing improved
cultural practices for the new high-
yielding varieties coming from the
national program or those introduced
from outside the country. Research
work in maize agronomy presently
includes plant density and spacing, soil
fertility and crop nutrition, the
intercropping of maize with legumes,
and weed control. Maize physiology
studies are also investigating the
efficiency of the various maize varieties
in partitioning dry matter into grain.

The maize pathology program
With the intensification of the
production of maize, its disease status
has changed, with the occurrence of
more diseases in epidemic proportions
every year. Therefore, it has became
necessary to engage a pathologist to
initially screen existing materials for
resistance or tolerance to the most

common diseases. These materials can
later be incorporated into the disease-
resistance breeding program in case
any of the diseases reach economic

The most serious diseases at present
are maize streak virus, maize leaf
blight (Helminthosporium turcicum
and Trichometa-sphaeria turcicum),
rust (Puccinia sorghi) and leaf
anthracnose. Since the economic
importance of these diseases has not
been studied previously in Malawi, the
preliminary program consists of
investigations into the economic
importance of maize leaf blight and
rust and into varietal reactions to the
diseases and disease development over

The maize entomology program
The entomology section of the maize
research program is responsible for
monitoring the incidence of
economically important pests in
Malawi and investigating control
measures. The maize entomologist, like
other maize scientists, is also involved
in advisory work with farmers.

The most serious pests are stalk borer,
termites and armyworm. While
termites are difficult to control, stalk
borers can be controlled by 2%
Dipterex granules; armyworm can be
controlled by Sevin 85% wettable
powder or Dipterex 95% soluble

Maize Research Staffing

The maize breeding program is staffed
by three maize breeders. They are
supported by staff at the technical
officer and technical assistant grades
and a labor force which fluctuates,
depending on the season. Two of the
maize breeders are assigned to the
breeding project for the high-potential
areas; the third breeder is responsible
for breeding for the marginal areas.

The maize agronomy program is
manned by a senior maize agronomist
and a professional officer, two technical
officers, a senior technical assistant
and a technical assistant. The labor
force again fluctuates, according to
season. The maize pathology program
is composed of one professional officer,
a technical officer and a technical
assistant. Maize entomology is manned
by a senior entomologist, a technical
officer and a technical assistant. All
maize research scientists work as a
team under the coordination of the
Maize Community Team Leader.


Maize production in Malawi has
dramatically increased in recent years
in all parts of the country. Malawi is
now not only self-sufficient in maize,
but has become an exporter as a result
of increased production in recent years.
Table 4, which shows ADMARC

purchases of maize from smallholder
farmers, is an indication of this
increase in production. The increase in
farmer sales may be attributed to
increased yields per hectare, to
farmers' adoption of improved cultural
practices and to the use of improved
seed and fertilizer. These factors are
partly the result of the Malawi maize
improvement program and partly of
good government policy and intensified
extension efforts.

Table 4. Amount of maize sold to ADMARC by
smallholder farmers, Malawi, 1981 to 1983

Amount of grain sold (000 tons)
Northern Central Southern
Year region region region Total

1980-81 20,723 65,559 4,923 91,205
1981-82 36,387 96,186 4,018 136,591
1982-83 46,306 152,993 45,617 244,916

Maize Production, Constraints,
Research and Development in Mauritius
N. Govinden, Food Crop Agronomy Division, Mauritius Sugar
Industry Research Institute, and S.P. Mauree, Extension Services,
Ministry of Agriculture, Fisheries and Natural Resources, Reduit,

Mauritius and its island district,
Rodrigues, form part of the Mascarene
Archipelago in the southwest Indian
Ocean. Mauritius is situated at 200S
latitude and 570E longitude, about 880
km east of the Madagascar. It is of
volcanic origin, and the land rises from
a coastal plain to a central plateau,
with elevations ranging between 73
and 275 meters. It has a maritime
climate, tropical in summer and
subtropical in winter (8); temperature
is mild the year around. The mean
maximum temperature in the warmest
areas varies from 25.90C in August to
31.20C in February, and the mean
minimum in the coolest areas, 14.90 to
20.50C. Annual rainfall is less than
1000 mm on the coast and more than
5000 mm on the central plateau, but
the amount varies from year to year.
Most of the rain falls between
December and April, the cyclone
season. Tropical cyclones cause
considerable damage to crops.

Mauritius occupies an area of about
1840 km2, of which 57% is cultivated.
Sugarcane occupies about 90% of the
cultivated area, and tea, about 6%.
The agricultural economy is, therefore,
dominated by the production of sugar.

In the past decade, Mauritius' food
imports have increased alarmingly. In
1982, the value of imported food
represented about 25% of total
imports, with the balance of trade
suffering a heavy deficit. About two-
thirds of the foreign currency earnings
from the main export, sugar, was
absorbed by this cost of imported food
(4). Hence, the government declared a
policy "to achieve the greatest
autonomy in the control and
production of our food supplies" (2).

In 1984, 54% of the country's food
imports could not have been produced
locally, 39% could have been produced
and would not have required arable
land, and 7% could have been
produced but would have required
arable land. In this last category, maize
and vegetables were the most
important items (10). Therefore, much
emphasis is being placed on increasing
maize production.

Maize is not a new crop in Mauritius;
its cultivation has a long history dating
back to the first decades of
colonization. On several occasions in
the past, particularly when sugar
prices were low or in times of crisis,
such as wars, interest was also shown
in agricultural diversification through
the production of maize and other food

Presently, efforts are again being made
to increase maize production, with a
view to attaining self-sufficiency by
1990. The annual per capital
consumption of maize in Mauritius is
16 kg, most of which is used as
livestock feed. In 1984, only a quarter
of this maize was produced on the
island; the rest was imported.

About half of the maize produced in
Mauritius is grown in sugarcane
interrows; the other half is grown in
rotation with cane. Most is produced
by the sugar estates, but recently other
producer groups have started to show
an interest in the crop.

The main constraints to maize
production are land scarcity, the
occurrence of tropical cyclones and of
drought, insufficient shelling and
drying facilities and relatively low

economic profitability. These
constraints are discussed in this paper
in relation to research objectives and
achievements, as well as the role of
extension and the planning and
organization of production. Emphasis
is on the attempts being made to
remove the constraints through
research and development. The
uniqueness of some of the Mauritian
approaches are underlined.

Maize Production and Utilization

In the eighteenth century, the early
French colonists grew maize in
Mauritius in order to avert the threat of
famine. When a guaranteed market for
sugar was established in 1825,
sugarcane became the dominant crop,
and by the end of the nineteenth
century, maize was no longer grown; it
continued to be of some importance in
Rodrigues, where cane was not grown.
During the first and second world
wars, when rice supplies were
disrupted, schemes were launched for
the production of maize, but these
were not very successful. In 1944, the
production of maize was only 5500
tons, whereas demand for rice was
about 55,000 tons. After the Second
World War, maize production declined,
reaching its lowest level in 1962 to
1964. There was some increase in
production in the 1970s, but no further
progress was achieved until 1984,
when production again increased to
4000 tons (Figure 1). On the basis of
orders for seed, it is estimated that
production in 1985 will reach more
than 7000 tons.

There has been a large increase in the
country's demand for maize. The
average amounts utilized annually for
the periods 1969 to 1973, 1974 to
1978, and 1979 to 1983 were 4280,
5630 and 14,180 tons, respectively.
This rapid increase was associated
with an increase in the demand for
livestock products as a result of rising
incomes. At present, about 99% of the
maize in Mauritius is used as feed for

livestock, especially poultry. Mauritians
do not eat much maize, and surveys
have revealed that if imported rice and
flour, the main staples, were not
available, they would prefer manioc,
sweet potatoes and potatoes to maize
(3). By contrast, maize figures largely
in the Rodriguan diet, although rice is
the preferred staple (9).

Agricultural Production Systems

There are presently three main maize
production systems in Mauritius,
extensive pure-tand cultivation,
intensive pure-stand cultivation and
intensive intercropping with sugarcane.

Extensive pure-stand
cultivation (System 1)
System 1A-Extensive pure-stand
cultivation is practiced in Mauritius by
small-scale farmers, many of whom are
squatters on Crown Lands. Maize


In CD ?D (0 r% re Wc 0 co0 Coo
6 'co 0 Mn co 6i i 1 MM

Figure 1. Maize production, Mauritius, 1955
to 1985

cultivation is not their main
occupation. Small plots on mountain
slopes are cleared with a machete and
hoe at the beginning of the rains, and
three or four seeds are sown in holes
dug at a spacing of approximately
1 x 1 m; this usually results in a stand
of two plants per hill. Grain yields are
about 2 t/ha, and the total production
of these farmers in 1984 was estimated
at about 100 tons. In the extensive
pure-stand cultivation of maize, there
are no cash inputs in the form of
fertilizers, herbicides or irrigation.
Invariably, the local maize variety is
utilized; due to its rusticity it is well
adapted to this traditional method of

System 1B-Extensive cultivation is
also practiced in much the same way
on Rodrigues, although in contrast to
Mauritius, small-scale maize farmers
there do not always plant in pure
stands; their maize is often

intercropped with manioc and sweet
potato. Local Rodrigues varieties are
used, and yields reach about 2 t/ha.
Total production was estimated at
2000 tons in 1983, with most of the
maize being grown for home use.
Maize is the most important crop in

Intensive pure-stand
cultivation (System 2)
Intensive pure-stand cultivation of
maize has been carried out on
Mauritius mainly by the sugar estates;
only in 1984 did it begin to be used by
other farmers. In this system, which is
very important since it presently
accounts for about half of the maize
produced, maize is grown in rotation
with sugarcane. The maize is grown on
the land lying fallow between the
sugarcane harvest and the next cane
planting. Inputs such as fertilizers,
herbicides, insecticides and, often,
supplementary irrigation are provided.

Maize planted in sugarcane interrows, Mauritius

Highly responsive varieties, presently
hybrids, are planted at 62,500 plants
per hectare, and average yields are
about 4.3 t/ha.

Intensive intercropping
with sugarcane (System 3)
Intensive intercropping with sugarcane
(growing maize between the rows of
cane), has also been done on the sugar
estates; only in 1984 did other groups
adopt the practice. This system
accounts for about half of the present
production. Although the system exists
in a few other sugar-producing
countries, such as India, Taiwan and
the Philippines, nowhere is it as
important as it is in Mauritius. The
most common pattern of intercropping
sugarcane with maize is to grow one
row of maize in alternate interrows of
the sugarcane crop and of the first and
second ratoon crops. In this system,
the maize population density is one-
third that of pure-stand maize. The
maize is fertilized, and it benefits from
the irrigation given to the young cane;
sometimes it also receives
supplementary irrigation. Grain yields
are about 1.4 t/ha, the equivalent of a
yield of about 4.2 t/ha of pure-stand

Production Structure

Maize production increased remarkably
in 1984, and is expected to increase
still further in coming years. Recently,
there has been a gradual change in the
structure of maize production in
Mauritius (Table 1). The amount
produced by small planters by
traditional methods (System 1A) has
increased slightly in response to an
increase in price. However, it is
anticipated that in the future this
production will decrease as pressure is
exerted on squatters on Crown Lands
to stop cropping erosion-susceptible
mountain slopes.

A campaign was begun in 1984 to get
producer groups other than sugar
estates to grow maize on sugarcane
lands, and this has started to bear
fruit. In that year for the first time,
small planters ventured into maize
production in sugarcane interrows and
in rotation with sugarcane. Sugar
estates, which own about 55% of the
cane lands, produced 93% of the total
maize crop in 1983. The proportion
decreased to 85% in 1984 and is
expected to be about 86% in 1985. The
proportion may decrease further in the

Table 1. Maize farming systems, Mauritius, 1983 to 1985
Estimated Production
1983 1984 1985
Farming system Type of farmer Tons Percent Tons Percent Tons Percent

Extensive pure stand Small-scale 50 3 100 3 200 3

Intensive pure stand Sugar estates 600 40 1600 40 2500 35
Large-scale -- -- 50 1
Small-scale -- 100 3 150 2

Intensive intercropping Sugar estates 800 53 1800 45 3600 51
Large-scale -- 50 1 100 1
Small-scale 50 3 350 9 500 7

Total 1500 99 4000 101 7100 100

future if other producer groups
increase their production. The
proportion of maize produced in
sugarcane interrows is expected to
increase from about 55% in 1983 and
1984 to about 58% in 1985. This can
be attributed to an increase in first-
season (March) plantings when the
only free land is found in sugarcane

Constraints to Maize Production

Numerous constraints limit the
production of maize in Mauritius and
account for the failure of past attempts
at increasing production. The main
ones are described here in relation to
the research objectives and
achievements of the maize program of
the Mauritius Sugar Industry Research
Institute (MSIRI), the organization
responsible for maize research.

Land scarcity
Since most of the arable land in
Mauritius is presently cropped, there is
very little scope for increasing the
cultivated area. The only lands that
could eventually be developed are
located in the dry zones, and not only
is it too expensive to provide them
with irrigation, but water is not always
available. Also, the present official
policy is to maintain sugar production
at current levels because about 80% of
the sugar is exported at a guaranteed,
negotiated price; this policy implies
that diversification should not come
about at the expense of sugarcane. The
solution, therefore, is to intensify
maize production on sugarcane lands.
The challenge of Mauritian agriculture
is to find a way to produce more food
crops without an increase in arable
land and while maintaining the level of
sugar production. This can only be
done by rotating crops in the
sugarcane lands (System 2) and by
making maximum use of sugarcane
interrows (System 3).

Research on intercropping sugarcane
with food crops started at the end of
the nineteenth century, but intensive
studies have been recent. Intercropping
patterns have been considered, and in
the case of potato, widely adopted;
85% of potato production in Mauritius
is in sugarcane interrows. This does
not reduce sugar production. With
maize, the recommended pattern of
growing one row of maize in alternate
interrows of plant and first and second
ratoon sugarcane accounts for about
50% of maize production. Research is
being pursued to further intensify the
system by growing such crops as
potato, groundnut and beans in the
plant cane interrow which presently is
not planted. This should make the
practice of intercropping maize with
plant cane still more attractive.
Another approach being studied is that
of increasing maize density in
sugarcane interrows. Presently, the
density utilized is 20,800 plants per
hectare, one-third of the density
recommended for pure-stand
cultivation. The average maize grain
yield of about 1.4 t/ha is also one-third
of the yield of that of pure stands,
somewhat low compared to other crops
which compete with maize for the
sugarcane interrows. The density of
potato intercropped with plant cane is
50% that of sole-cropped potato, and
yield, about 60%.

A way to increase intercropped maize
density, and hence yields, has now
been proposed. This consists of
planting two rows of maize in large
sugarcane interrows created by pairing
cane rows. This would increase maize
yield by more than 60% over that of
the presently recommended pattern.
Moreover, it would allow the
intercropping of older (third and
fourth) sugarcane ratoons. The system
of pairing cane rows has not yet found
favor with sugarcane producers
because of a number of problems
which are presently being studied.

The success of intercropping maize
with sugarcane depends on the use of
early maturing, short-statured, high-
yielding maize cultivars. Such cultivars
were not available in the past, and this
may explain why past attempts to
encourage the intercropping of maize
with sugarcane met with little success.
Hybrids imported from Europe were
not suitable to Mauritian conditions
because of their susceptibility to
diseases, and quarantine regulations
made it impossible to import seed
except from a few European countries,
Zimbabwe and the Republic of South
Africa. Therefore, Mauritius started its
own breeding program in 1970. Since
then, much research at MSIRI has
been devoted to the selection of
cultivars for intercropping, with the
first objective of the breeding program
being the development of cultivars for
use in sugarcane interrows. The first
two hybrids were developed in 1980,
and they are now the only cultivars

The second approach to intensified
cropping is to make maximum use of
sugarcane rotation lands. This land lies
fallow for four or five months between
the harvest of the last sugarcane
ratoon and the replanting of the field.
As sugarcane is replanted only after an
average of ten cycles (one plant crop
and nine ratoon crops), one-tenth of
the area under sugarcane is replanted
every year.

For various reasons, the area available
for growing pure-stand maize is less
than one-tenth of the sugarcane area.
First, not all sugarcane lands are
suitable for maize. Second, in many
instances and particularly in
nonirrigated zones, the fallow period is
too short for a maize crop. Finally,
other crops, such as tobacco and
vegetables, compete with maize for
sugarcane rotation lands. In order to
boost maize production on rotation
lands, a project has been launched to
identify the reasons why lands suitable

for maize are not used for cropping
and to remove the constraints. In most
cases, this can be done by shifting the
harvest date of the last cane ratoon
and the date of replanting.

Other alternative cropping systems
have been proposed but have not yet
been studied. For instance, reducing
the cane cycle from ten to eight years
has been suggested for increasing the
amount of available rotation lands. It
might also be possible to increase the
fallow period from four to eight months
to permit the growing of two
successive short-cycle maize crops.
The feasibility and the economics of
these suggestions will be studied.

The second most important constraint
limiting maize production in Mauritius
is the occurrence of tropical cyclones.
Sugarcane has become the dominant
crop in Mauritius because, of the many
crops that have been planted over the
years, it is least vulnerable to cyclones.
The history of maize in Mauritius lists
innumerable occasions when cyclones
have seriously reduced production,
causing hardship for the people of the
island. Maize can become an important
crop only if ways are found to avoid
the destructive effects of cyclones.
These occur in the wet summer
months, December to March, the
period which otherwise is the most
suitable for maize cultivation. It is
recommended, therefore, that maize be
planted in August or September,
before, or in March or April, after the
cyclone-prone period. The use of short-
cycle cultivars is also important.

In Mauritius, cyclones bring gusts of
wind of 120 kph; no maize cultivar can
resist such winds. Also, however, in
the cyclone season, and even in mid-
winter, winds of 35 to 60 kph are
common, and there are cultivars that
can withstand these winds. It is
imperative that lodging resistance be

incorporated into all cultivars grown
on Mauritius, and this is one of the
principal objectives of the maize
breeding program.

While the effects of cyclones can be
avoided by growing maize before and
after the wettest months, this increases
the risk of crop failure due to drought.
In Rodrigues, most of the annual
precipitation is associated with
cyclones, and farmers there have to
risk growing maize in the cyclone
season. In some parts of Mauritius,
where supplementary irrigation is
available, maize can be grown in the
dry season. In rainfed areas, drought is
a serious factor limiting yield.

There are two main approaches for
overcoming drought in Mauritius, the
use of short-cycle cultivars for drought
avoidance, and irrigation. Irrigation is
provided when and where it is
available, although the rise in the cost
of energy has been responsible for the
abandonment of a number of irrigation
schemes. Except in the drier regions, it
usually does not pay to irrigate
sugarcane unless gravity-fed systems
are employed; for maize, on the other
hand, supplemental irrigation is
usually worthwhile. The intercropping
of sugarcane with maize will make
irrigation schemes more economically
viable, since irrigation applied to maize
benefits the cane and vice versa.

Insect pests
At present, insect pests are not limiting
factors in maize production in
Mauritius. In the past, insect pests of
stored grains were important because
maize was stored for long periods, but
currently the production is regularly
absorbed by the feed mills. A few
insects, such as the webworm
(Angustalius malacellus), the greasy
cutworm (Agrotis ipsilon) and the
defoliating caterpillar (Spodoptera
littoralis), attack maize seedlings, but
they are controlled reasonably well by

the use of insecticides (5). Earworms
(Heliothis armigera and Cryptophlebia
leucotretra) do not appear to cause
economic damage.

Leaf diseases are important yield-
limiting factors. In Mauritius, the leaf
blights Helminthosporium maydis and
H. turcicum prevail in the maize-
growing areas (5). The latter is more
serious and may cause reductions in
yield by as much as 50% for
susceptible cultivars. Rust, Puccinia
polysora, is prevalent during the warm
season, and can also seriously reduce
yields. In addition to leaf blights and
rust, maize streak virus (MSV) is also
important in Rodrigues. The approach
to the control of leaf diseases is the
breeding of resistance into the cultivars
grown on Mauritius (6). The cultivars
presently grown in pure stands are
tolerant, but those grown in the
sugarcane interrows are moderately
susceptible to leaf blights. A number of
resistant inbred lines have recently
been introduced from the USA and the
Republic of South Africa, and they will
be crossed with the best local inbred
lines. MSV-resistant populations have
been introduced from IITA and will be
used in the development of a
composite for Rodrigues. So far, no
material with good resistance to rust
has been found; the CIMMYT
populations and gene pools that have
been grown so far appear to be
susceptible to rust.

Weeds create a serious problem in
maize fields in Mauritius. At the
beginning of the nineteenth century,
the root parasitic weed Striga hirsuta
was a major pest, but it has now
almost disappeared. The nutgrasses,
Cyperus rotundus and C. esculentus,
are particularly noxious weeds in
maize and sugarcane fields. A number
of herbicides are recommended against
nutgrass (5), but they are costly.

A major difficulty with maize
production on Mauritius is the shelling
and drying of the crop at the producer
level. Because of the need to keep the
crop cycle as short as possible, it is
necessary to harvest maize when the
grain moisture content is still at 25 to
30%. The ears then have to be shelled
and the grain dried to 12% moisture
content. The first maize-drying plant
was erected in 1917, and three modern
regional maize-processing plants for
the shelling, drying and temporary
storage of maize are now operational.
One is privately owned and has been
in existence since 1980; two other
government-owned plants were built in
1984. Some sugar estates also use
tractor-driven sellers and bagasse-
fired dryers. These facilities are
adequate to process the present
production, although more will be
required as production increases.

The need to dry maize artificially
increases the cost of production by
about 10%. For this reason,
alternative, low-cost drying methods
are being examined. At the small-farm
level, crib drying appears to be a
possibility. Solar dryers also offer some
potential, and they are being developed
at the University of Mauritius. In
Rodrigues, maize is hand-shelled and
then dried in the sun.

Economic Considerations

In 1983, it was calculated that the cost
of the production of maize grain at
12% moisture content was about
2885 Rs approximately USS 185) per
ton for non-mechanized and rainfed
crops (Table 2). The guaranteed price
for local maize was 3750 Rs per ton or
130% of the cost of production; this
was not considered attractive by the
farmer. In 1984, the price for local
maize was increased to 4050 Rs per
ton, and the effect on production was
remarkable. This situation is
reminiscent of what happened in the
1946 to 1950 period, when maize

subsidies resulted in a three-fold
increase in production. When the
subsidies were removed, production
dropped back to low levels (7). The
present price is not considered a
subsidy since the price for imported
maize is also about 4000 Rs per ton. It
is, instead, an incentive price.

An analysis of the breakdown of the
cost of production reveals that material
inputs, especially fertilizers, labor, and
shelling and drying, are the main
components of the cost of production.
Material inputs cannot be reduced
without reducing yields. The work
being done on natural drying has
already been mentioned, and there are
also other areas where a reduction in
production costs is possible.

In spite of the amount of
unemployment on Mauritius, it is felt
that some labor-intensive operations
should be mechanized. It has been
shown that the cost of labor could be
cut in half through the use of
appropriate implements for planting
and harvesting. The mechanical
planters presently utilized by some
planters are not very efficient, and new

Table 2. The cost of production of rainfed
maize, Mauritius, 1983

Cost Percent
component Rs/haa/ Rs/ton/ of total

Material inputs
Seed 1000 285 9.9
Fertilizer 3150 900 31.2
Biocides 1150 330 11.4
Labor 2550 730 25.2
Transport 350 100 3.5
Shelling and
drying 1750 500 17.3
Interest 150 40 1.5

Total 10100 2885 100.0

US $ 1 = 15.60 Rs
b/Based on an average yield of 3.5 t/ha

models, including pneumatic planters,
are being tested. Corn pickers have
been in use in pure stands since 1977.
In 1978, a prototype one-row corn
picker for use in sugarcane interrows
was designed and built and found to
work reasonably well (1). On sugar
estates it is not used, because
presently the permanent labor force
can be shifted from cane to maize.

Finally, the cost of production per ton
can be reduced by increasing yield. For
instance, irrigation could improve the
efficiency of the other resources, such
as fertilizers, and increase yields.


The Mauritius Sugar Industry Research
Institute, which conducts research on
maize, is also responsible for extension
for the sugar estates and large
planters, the owners of 40 hectares or
more, through its Extension and
Liaison Division. Lectures on maize are
regularly given at the MSIRI; research
recommendations are also made in its
annual reports and advisory bulletins,
and updates appear in mimeographed
recommendation sheets. Moreover,
maize specialists visit the maize
plantations regularly, and trials are
conducted on land belonging to sugar
estates and large planters.

Extension for the 33,000 small cane
planters is the responsibility of the
Extension Services of the Ministry of
Agriculture. There are two or three
extension officers in each of the ten
districts of Mauritius, and they collect
information from MSIRI and pass it on
to the farmers whom they visit
regularly. They also broadcast
recommendations on radio and
television and publish a news bulletin.

The Organization and
Planning of Production

The removal of technical constraints
alone, through research and efficient
extension, are not necessarily
translated into increased production.
The Mauritian experience with the
potato has demonstrated the need to
devote more attention to development.
Even when information is quickly
made available to producers, new
technologies are not adopted for
several years, and some not at all.

On Mauritius, maize is a controlled
product. The Agricultural Marketing
Board is responsible for calculating
costs of production, fixing prices and
quotas and allocating subsidies, as well
as for conditioning, storing and
marketing maize. This is done through
a Maize Production Committee, which
is made up of representatives of
producers, research and extension
services, feed mills, the Chamber of
Agriculture and the Consumer

The Chamber of Agriculture monitors
the use of sugarcane lands for the
production of food crops; lands not
used by sugar estates are leased to
small planters. The seed requirements
of the different producer groups are
also channeled through the Chamber.
The activities of seed importers and
local seed producers are coordinated so
that there will be no shortfall in
production because of a lack of seed.

At a higher level, the High-Powered
Committee for Agricultural
Diversification has a maize
subcommittee which makes
recommendations to the government
on all matters pertaining to the
development of maize production; it
also monitors the activities of all of the
other organizations. The subcommittee
reviews past performances in the light
of government policy and production
targets. The target set by the

government in 1983 was to satisfy the
country's annual needs by 1987 (4).
This has now been revised, with the
present goal that of attaining self-
sufficiency by 1990.


After several decades, Mauritius now
has the possibility of increasing its
maize production substantially. This is
the result of concerted effort on the
part of all those concerned with maize.
Producers are motivated by the
attractive price. Research workers are
determined to face the challenge of
developing methods to increase
production without reducing sugarcane
yields. The government has
demonstrated that it has the political
will to support local production.
Therefore, the goal of achieving self-
sufficiency by 1990 should be


1. Anonymous. 1979. Mechanization
(of maize). Report of the Mauritius
Sugar Industry Research Institute
(1978) 25:61-62.

2. Boolell, S. 1980. Opening address.
In Proceedings of the National
Agricultural Producers'
Conference, A. Peerally, ed.
University of Mauritius, December
10-15, 1979. Pp. 11-17.

3. Chan Ki Chan, D. 1982. Changer
ses habitudes alimentaires les
aspects sociaux, culturels et
psychologiques. Paper presented at
the Workshop on Food Habits,
University of Mauritius, November
25-December 4, 1982.

4. Government of Mauritius. 1983.
White Paper on Agricultural
Diversification. Government
Printer, Port Louis, Mauritius.

5. Govinden, N., J.C. Autrey, H. Dove
and G. McIntyre. 1984. Growing
maize. Advisory Bulletin 4.
Mauritius Sugar Industry Research

6. Govinden, N., and J. Dintinger.
1984. The Mauritius maize
breeding programme. In
Proceedings of the Sixth South
African Maize Breeding
Symposium. Pietermaritzburg,
South Africa, March 20-21, 1984.
(In press.)

7. North-Coombes, A., and A.
d'Emmerez de Charmoy. 1960.
The Production of Foodcrops in
Mauritius. Revue Agricole et
Sucriere de l'ile Maurice

8. Padya, B.M. 1984. The Climate of
Mauritius. Meteorological Office,

9. Vas, M. 1982. Certains facteurs qui
influencent les habitudes
alimentaires a Rodrigues. Paper
presented at the Workshop on
Food Habits, University of
Mauritius, November 25-December
4, 1982.

10. Wiehe, J.B., N. Govinden and P.
Rouillard. 1984. Achievements and
prospects in crop diversification on
sugarcane lands in Mauritius. In
Proceedings of the Eighth Congress
of Societies of Technical
Agriculture. Mauritius, October
15-19, 1984. (In press.)

Research on the Constraints to
Maize Production in Mozambique
E. Nunes, Instituto Nacional de Investigacao, Posto Agronomico de
Umbeluzi,* and D. Sousa, Posto Agronomico de Lichinga,
Mozambique, and I. Sataric, Maize Research Institute, Zemun
Polje, Yugoslavia

Maize is the preferred staple food of
most Mozambicans, and it is grown
throughout the country, both single
cropped and intercropped. The total
area under maize is about 500,000
hectares annually. Yields are generally

South of the Save River, the principal
factors limiting maize yields are lack of
rainfall (or for irrigated maize, poor
irrigation techniques) and diseases,
mainly downy mildew and maize
streak virus; pests, such as stalk
borers, are also constraints. In the
higher rainfall areas of central and
northern Mozambique, the principal
agronomic factors limiting maize
production are low soil fertility,
periodic droughts in the lowland areas,
and diseases, such as fusarium and
diplodia ear rots and
Helminthosporium spp. Weed control
is a major problem under both peasant
and commercial maize-growing
conditions throughout the country,
with the worst problems being the
weeds Striga lutea for the peasant
sector and Rottboellia exaltata and
perennial sedges for the mechanized

The most serious limitations to maize
production, however, are economic.
For the peasant sector, the most
important is that, in the countryside,
there is a lack of consumer goods that
would serve as incentives to
production. For the mechanized sector,
problems include a serious shortage of
trained manpower, insufficient

* National Research Institute, Agricultural
Research Station (at Umbeluzi and at

management expertise, organizational
difficulties and frequent shortages of
vital inputs, such as fuel.

The national maize research program
was started in 1977. Initially, the main
research emphasis was on the selection
of introduced maize germplasm for
Mozambican conditions. As a result,
three varieties (based on CIMMYT
materials) are ready for release.

Selection for resistance to drought,
downy mildew and maize streak virus
continues in the southern part of the
country; in the north there is a small
program of hybrid seed production. It
has become clear, however, that the
use of improved varieties is only a
minor factor influencing maize
production. Since 1982, the maize
program has adopted a more balanced
farming systems approach. This paper
summarizes results from planting date,
plant density, fertilizer use, insecticide
and herbicide trials. Land preparation
studies have also been initiated.

Mainly post-independence work is
described in this paper. Because of the
transfer of power from the colonial to
the independent government of
Mozambique, there was an interruption
both in agricultural production and in
research. There was complete
discontinuity in terms of research staff,
and much unpublished data was lost.
It has taken time to rebuild research
capability, and work at times is still
rudimentary. While not attempting to
present experimental details, an

overview is provided here of maize
production and its constraints, as well
as of progress made in maize research
during the ten years since

Areas of Mozambique

According to the agricultural census of
the 1960s (3), the country could be
divided into three categories in terms
of maize production, the area where
maize was the major staple food crop,
the area where it was of equal
importance with sorghum and the area
where it was of secondary importance.
At that time, maize was the major
staple crop in the highland areas of
Niassa, Tete and Manica provinces, the
lower Zambezi Valley and most of the
southern part of the country, Maputo,
Gaza and Inhambane provinces. In the
central area, most of lowland Manica
and inland Sofala, maize had about
equal importance with sorghum, and
in the rest of the country either
sorghum, cassava or rice was
dominant (Figure 1). There is no
evidence to suggest that this pattern
has changed in the last 20 years.

The highland areas are those with
altitudes of 500 to 1300 meters, mean
annual rainfall between 1000 and 1300
mm and PET (potential evapo-
transpiration) (Penman) of less than
1500 mm per annum. There is a single
rainy season, from November to April,
giving a growing period of about 180
days. A recent analysis of the
agroclimatic suitability of Mozambique
for maize classifies these areas as
suitable or very suitable for rainfed
maize (5) (Figure 2). It is interesting to
note that the predominance of maize
as a staple does not always follow the
indicated agroclimatic suitability for
growing rainfed maize.

The whole of the northeastern part of
the country is classified as very
suitable for maize production, but in
fact, sorghum and cassava
predominate. This must be due then to
other factors, and the most probable
ones are soil fertility limitations and
the competition for land and labor
from cotton, the main cash crop.

Southern Mozambique is classified as
marginally suitable or unsuitable for
rainfed maize production. This region
is characterized by irregular rainfall in
terms of both total annual amount and
distribution. Prolonged drought during
the growing period is the norm, and
crop failure is common as a result of
either drought or flooding.

The coastal strip has more reliable
rainfall, with a mean annual total of
800 to 1000 mm and PET below 1350
mm. Moving from the coast, the
rainfall drops to less than 400 mm per
annum in the interior of Gaza
Province, while PET rises to over 2000
mm. In Gaza, maize is not a rainfed
crop; it is irrigated or grown in
depressions with residual moisture and
hence is susceptible to flooding.

In the peasant secor of the highland
areas, maize is grown on hand-formed
contour ridges in which plant residues
are incorporated. It is intercropped,
mainly with beans, cowpeas and other
legumes and sometimes potatoes. In
each of these areas, the system is well
developed and adapted to the heavy
soils, the abrupt beginning of the rains,
the high risk of erosion and the
shortage of chemical and mechanical
inputs. Planting dates in the highlands
are fairly well defined, from mid-
November to mid-December, as soon as
sufficient rain has fallen. Yields of the
intercropped maize are between 0.8
and 1.5 t/ha.




Yield potential (t/ha)

$ Very suitable (1.4 -1.6)

Suitable (1.1 1.41

: MModerately suitable (0.7 -

__ Marginal (0.4 0.7)

Not suitable (0 0.4)
nhambane Areas where maize is the
predominant crop



Figure 1. Climatic suitability for rainfed maize production at a low level of inputs,



- Maize + sorghum





um + cassava

um + millet


Figure 2. Dominant crops produced under peasant conditions, Mozambique
Source: Mario de Carvalho

In the south, oxen are used to cultivate
the heavier irrigable soils, although
hand cultivation is also common. In
these soils, maize is grown as a row
intercrop, usually with cowpeas, and
pumpkins are undersown at an
irregular density. In the sandy soils of
the coastal strip, cultivation is mainly
by hand. Maize is planted at a low
density (less than 10,000 plants per
hectare) and with irregular spacing in
fields of cassava, groundnuts and
cowpeas. Although the maize-planting
season in the south is considered to be
from the beginning of September to
mid-October, maize is planted
throughout the year following any
sizable rainfall.

Mechanized farming with high
agrochemical input is largely confined
to the state farms. These were set up
soon after independence in 1975 to
secure continuing agricultural
production despite the exodus of the
Portuguese farmers; they were formed
by putting several smaller units under
one management in order to make full
use of the limited expertise available.
They are very large, usually over 1000
ha. On each farm the crop mix is
dominated by one crop, either maize,
cotton, sugarcane or rice, and on the
cotton farms a considerable amount of
maize is also planted. The combination
of crop farming and the raising of
livestock on one farm is rare. The
location of these state farms is shown
in Figure 3.

Cooperatives and private farms also
exist on a more limited scale, with a
lower level of inputs than the state
farms and with mechanization usually
limited to land preparation. The
distribution of these farms loosely
follows that of the state farms.

Maize Research

Maize research in Mozambique is
carried out by the National
Agricultural Research Institute (INIA)
and the Rural Development

Department (DDR) of the Ministry of
Agriculture. In 1977 a program,
mainly of variety selection, was started
at INIA with FAO help; the FAO expert
was joined by a Mozambican
agronomist in 1981. In 1982, the FAO
expert left the country, and since 1983
research has been carried out with the
help of a five-person Yugoslavian team
subcontracted by FAO.

Expatriate scientists have also worked
on pest and disease problems and
fertilizer requirements since 1977 and
on weed control since 1981. These
were initially separate, uncoordinated
programs, but in 1982 all research on
maize was coordinated under the
National Maize Program. In 1983,

A Agricultural crops

* Livestock
0 Forestry

Figure 3. Location of state farms and their
production, Mozambique

attempts were made to make research
more meaningful in the solving of
immediate problems limiting maize

In the early years, conditions of
staffing and equipment on the
experiment stations were such that
experiments tended to be sown and
harvested by visiting scientists (based
in Maputo) and looked after by poorly
qualified station staff. This led to a
high percentage of failures, high CVs
and low yields. In 1983, priority was
given to three main experimental sites,
Lichinga for the highlands, Namapa for
the northern lowlands and Chokwe for
the south; experienced research
scientists were assigned to each,
resulting in a dramatic increase in the
quality of experimental results.
However, the present war is making
this quality difficult to maintain.

The Rural Development Department
has its own centers, in six of which
research on maize has been carried out
for the last couple of years. These
centers work with the peasant sector
and cooperative farms, while INIA
research has been aimed at fully
mechanized agriculture.

Factors Limiting Maize Production

In the peasant sector
Although it is estimated that 50% of
the marketed maize in Mozambique is
produced by peasant farmers, this
sector receives very little help from the
state. There is practically no extension
service and little distribution of seed
and agrochemicals; even tools are
unavailable some years. Manufactured
goods in general are in short supply in
the countryside, including basics such
as salt, oil, sugar, cloth and kerosene,
as well as such items as bicycles and
spare parts, radios, lamps and sewing
machines. It is probable that the lack
of consumer goods is a major factor
limiting the amount of maize

Apart from this lack of assistance, the
major constraints to peasant maize
production are drought, flooding, pests
and diseases in the south and declining
soil fertility in the north. The soil
fertility problem has been aggravated
in recent years by the movement of
population from the countryside into
villages. This increases the pressure on
the land and reduces the utilization of
traditional bush fallows.

Trials carried out by the DDR centers
have shown that the varieties used by
the peasants are as productive, within
the constraints of their farming
system, as are the improved varieties
presently available (1); the grain also
has the added advantages of
palatability and resistance to storage
pests. The improved varieties tend to
have a greater potential under high-
fertility conditions, although this is not
always the case. Some local maize
populations have given experimental
yields of up to 5.6 t/ha with the
application of fertilizer.

In about 100 nonreplicated trials on
peasant farmers' fields in two areas of
the lowlands in the north, row planting
of maize in maize-legume intercropping
led to an average increase in yield of
about 24%. Neither the yield of the
companion crop nor the densities of
the non-row planted maize intercrop
was measured, but it is possible that
row planting leads to an increase in
plant density.

In the mechanized sector
Soon after independence, most farmers
with experience in mechanized
agriculture left the country, and
because of the very low level of
education and work experience that
had been allowed the black population
in colonial times, this sector still
suffers from a shortage of skilled
agricultural workers. This fact is
especially serious because of the need
to deal with the problems which arise
from external factors, such as badly
organized transportation and

distribution services, the shortage of
crucial inputs and spare parts,
flooding, drought and, at present, the
war. However, in the last few years,
there has been an improvement in
yield on some of these farms as
farmers have gained more experience.

The lack of experienced managers and
skilled workers at all levels has been
the cause of many of the problems
which have led to average yields of less
than 2 t/ha on many mechanized
farms. One of the most serious of these
problems is that work is consistently
late, from the preparation of the land
to the harvesting of the crops. This
stems partly from a lack of
understanding of the crucial
importance of timeliness, but even
more from organizational difficulties,
which are further aggravated by
deficient machine maintenance. As a
result, a high percentage of the farm
tractors are not operational at any one
time, and tractor life is short. The size
of the farms also implies that
considerable skill is necessary to
productively manage the large work

Limitations of a more agronomic
character are also influenced by the
lack of skilled personnel. Often
machines are badly adjusted, irrigation
water is not properly controlled, and
agrochemicals are inappropriately
applied. These problems, although
identified, cannot be solved by
agricultural research unless the
production system can be simplified.

Research Challenges

Land preparation and sowing
One of the major areas in which
research needs to provide viable and
economic recommendations is in that
of land preparation for the mechanized
section. Even when high-quality hybrid
seed is used, plant density is often low,
serious erosion is evident in maize
fields, and there is little weed control
by mechanical means. In the highland

areas, the start of the rainy season is
often abrupt, allowing only about a
month for land preparation and sowing
after the rains have begun.
Alternatives have been proposed to the
present system of plowing and
harrowing the hard dry soil; some have
been tried, but have not yet been
properly tested. These alternatives
include various forms of minimum or
reduced tillage and a rescheduling of
present operations; for instance,
plowing could be done immediately
after harvest, when there is still some
moisture in the soil.

Irrigation and drainage
In the south, the alluvial plains of
almost all of the rivers are used for
growing maize on both mechanized
and peasant farms. In some areas, the
maize is not irrigated and adequate
ground water control is the major
problem. Drainage problems, at both
superficial and deep levels, occur in
these generally heavy soils, including
that of increased soil salinity. Pumped
drainage is sometimes practiced.
Irrigation is by means of furrows,
although sometimes small basins or
sprinkler irrigation are used.

Major drainage problems and salinity
occur in the country's most extensive
irrigation scheme at Chokwe. This area
is presently being used more and more
for growing maize instead of the
traditional rice. On both peasant and
cooperative farms, small amounts of
maize are grown in the limited, well-
drained seepage zones called
machongos. Peat may have developed
in some places, but generally minor
open drains are adequate for maize.

The medium to large irrigation
schemes date from before
independence. More recently, small
self-help schemes are being
constructed for the peasant sector. On
the state farms and the cooperatives,
limited expertise in water management

hampers the optimization of
production from irrigated agriculture.
There is major need for training at all

Soil fertility
In both the mechanized and the
nonmechanized farming sectors, the
principal nutrient limiting maize yield
is nitrogen; it is deficient in all soils on
which maize is grown. The response of
maize to nitrogen is greatest in the
high-rainfall areas. There, experimental
results indicate that doses of nitrogen
up to 150 kg/ha are economic in
mechanized farming (6). The
corresponding figure for the northern
lowlands is 110 kg/ha. In the south,
the economic optimum has not yet
been determined, and in this area
response may be suppressed by a lack
of water and by pests and diseases.

Phosphorus deficiency is common in
the soils of Mozambique, but
experimental results show that
economic response to phosphorus
fertilizer can only be expected when
other factors permit a maize yield of at
least 2.5 t/ha. Test results have shown
that yield is limited when there are
fewer than 16 ppm of available
phosphorus in the soil (North Carolina
or Mehlich extract) (6).

Sulfur deficiency has been detected in
widely dispersed areas of the country,
but responses to sulfur have not been
quantified. Nitrogen, phosphorus and
potassium are considered major
nutrients by the fertilizer importing
and distributing bodies, and attempts
are being made to ensure that
fertilizers destined for areas with sulfur
deficiencies contain sulfur in the
future. The results of sulfur deficiency
in the peasant sector, where almost no
fertilizer is used, has not been studied.

In general, soil potassium levels are
satisfactory for maize growth, although
one important area of low-potassium
soils exists in southern Cabo Delgado
and northern Nampula provinces. No

results from systematic studies of
micronutrient deficiencies are
available. Zinc deficiency has been
detected from leaf analyses of maize
grown on sandy soils in the northeast
part of the country, and it is expected
to become more evident as yields
increase. Molybdenum deficiency is
also to be expected but has not yet
been detected, due possibly to the
widespread use of molybdenum-treated
seed on the mechanized farms.

Weed control
The principal weed species found in
maize in Mozambique are similar to
those of neighboring countries, with
the addition of Mucuna pruriens in the
north, Parthenium hysterophonus in
the south and Brachiaria deflexa and
Urochloa mossambicensis throughout
the country asia whole (2).

In peasant farming, the principal
problem weeds in maize seem to be
Striga lutea and some perennial weeds,
such as Panicum maximum. Little is
known about traditional methods of
control, although the use of maize
companion crops, cowpeas,
groundnuts, squash and sweet
potatoes, obviously helps to smother
weeds; virtually all cultivation is
carried out with heavy hoes. The
presence of Striga lutea and some
other species,, such as Eragrastis
arenicola and Rhychyletrum repens, is
taken by the peasants as an indication
of low soil fertility and as a signal to
abandon a site for two to ten years.

In mechanized farming, herbicides are
used on most of the maize area. The
principal one utilized is Atrazine mixed
with Alachlor or Metolachlor. In large
areas of the north, the annual grass
weed Rottboellia exaltata, which is
resistant to these herbicides, has
become dominant, particularly on
farms where maize is grown
continuously without rotation.
Pendimethalin (mixed with Atrazin) is
the standard herbicide on these farms,
but it is very expensive and requires

good soil moisture. Organizational
difficulties prevent the implementation
of alternative measures, such as
rotation with broad-leaf crops
combined with the use of Trifluralin.

Other problem weeds include the
perennial sedges Cyperus esculentus,
C. rotundus and Scirpus maritimus in
the Limpopo Valley. Incorporated
herbicides such as EPTC are rarely
used against these weeds, due to a lack
of suitable equipment and to
organizational problems. Mucuna
prurlens (buffalo bean) poses a problem
in some of the central and northern
areas. As well as competing with the
maize plant, it produces pods with
stinging hairs which can make
harvesting impossible. A control
measure found to be successful on one
large farm was the application of 2,4-D
at the knee-high stage of the maize.

The weed control sector of INIA has
carried out weed surveys and some
herbicide trials. These have mainly
been devoted to demonstrating well-
tried and economical chemicals as a
countermeasure to the hard sell of new
expensive herbicides by the
agrochemical companies.

Pest control
Stalk borers (Chilo partellus, Sesamia
calamistis and Busseola fusca) are the
most important maize pests in
Mozambique. As a result of continuous
planting in the southern part of the
country, stalk borer incidence is
severe. In the north infections of only
about 10% of the plants are common,
causing losses as low as 1%; in the
south, infestation may reach 100% of
the plants with considerable yield loss.
Two generations of the borers can
develop on the same plant, the first in
the stem and the second in either the
stem or ear, also causing ear rot in the
latter case. Experiments show that two
insecticide applications in the first five
weeks of the growing season can
reduce stalk borer incidence to an
acceptable level. A study on the

biological control of stalk borers has
shown promising results, but is still in
the early experimental stage.

Termites (Microtermes spp.) cause
lodging at harvest time, which may
result in severe losses if harvest is
delayed. Soil treatment with Aldrin or
Dieldrin is a common practice on state
farms in the northern part of the
country. In some years, birds are also
important pests. Occasionally field rats
and black maize beetles (Heteronychus
spp.) are problems.

Disease control
Four viruses have been identified in
maize in Mozambique, of which maize
streak virus is the most important. The
others are maize mosaic virus, maize
stripe virus and sugarcane mosaic
virus. Three vectors have been
identified for maize streak virus, all
Jassids of the genus Cicadulina; they
are C. mbila, C. parazeane and C.
triangular. This disease seriously
affects the maize crop in the south. It
is very common throughout the year,
and no crop there completely escapes
damage; with late planting, losses are

Recommended control measures are
early planting and good weed control
in the vicinity of the crop. Promising
results have been obtained in
experiments on the control of the
vector by using either Carbofuran at
sowing or Pyrethroids in two
applications during the first four weeks
after sowing. The only real solution,
however, is varietal resistance, and
experiments have been begun this

The other group of diseases of
importance in the southern lowlands is
that of the downy mildews. The major
pathogen involved is Peronosclerospora
sorghi. Recently there has also been a
considerable increase in the incidence
of crazy top, which has been attributed
to Sclerophthora macrospora.
Experimental downy mildew-resistant

varieties and populations have revealed
a marked susceptibility to maize streak
virus, which severely limits their direct
usefulness. Recommended control
measures include the manual
elimination of diseased plants and seed
treatment with Metalaxyl.

In the north, the principal problems
are the helminthosporium (Drechslera)
blight complex and diverse ear rots,
with associated serious seed infections.
Ear rots caused by Fusarium spp. and
Diplodia maydis are the most
important and generally give rise to
primary ear rot infections in the
highlands. All three helminthosporium
leaf blights and spots are in evidence,
H. turcicum, H. maydis and H.
carbonum. They can occur
simultaneously, although the first is
the most frequent and damaging. In
the lowlands, natural climatic
conditions do not seem to favor the
spread of these diseases, although
inoculation experiments have shown
great potential susceptibility.

Rusts of maize, either singly or
collectively, are not very important in
the general disease pattern, although
they can be observed regularly in the

Planting date and density
Even though the importance of correct
planting date is recognized on the
mechanized farms, external factors,
such as the late arrival of seed and
agrochemicals, or internal
organizational problems still lead to
late planting. In the north, the
optimum planting time for the
mechanized sector ends the last week
of December; loss of yield in later-
planted crops is assumed to stem from
a decreased availability of water and
nitrogen. Experimental results from
Lichinga and Lioma show yield losses
of 1 to 2% for each day that planting is
delayed between about December 7
and January 7 (4). In the south, the
optimum planting time for irrigated

maize ends on November 15, because
of the occurrence of disease attack
after that date. No consistent
experimental results are available, but
after the end of October, disease attack
increases rapidly.

Plant densities that farmers anticipate
are about 50,000 plants per hectare in
the highlands and 42,000 in the
lowlands, both in the north and the
south. However, these densities are
rarely achieved. Problems of erosion,
land preparation and machine
availability have already been
mentioned. Where nationally produced
seed is used, seed quality also plays a
part, as do certain pests (field rats,
termites, etc.). Experimental trials on
planting date were held at eight sites,
but only for one year each, so that
conclusions cannot yet be drawn.
However, in both highland and lowland
areas, yields seem to increase when
planting densities are increased to
40,000 to 50,000 plants per hectare.

Maize Seed Production

The three national institutions involved
in maize seed production and quality
control in Mozambique are the
National Agricultural Research
Institute, the National Seed Company
(ENS) and the National Seed Service
(SNS). In addition, two state companies
and the Provincial Agriculture
Directorates (DPA) are involved in the
distribution of seed.

Responsibility for seed production and
control has been defined as set out in
the following diagram:

by SNC

'Breeder seed
Prebasic seed
Basic seed

Certified seed
- first generation


ENS seed farms

Certified Seed ENS seed farms and
- second generation contract growers

Breeder and
pre-basic seed production
At the time of independence, the major
varieties in mechanized production
were Silver Mine, Silver King, Kalahari,
White Cango, Hickory King and Yellow
Sahara, as well as the hybrid SR52. All
of this seed was imported. Seed
renewal took place on the farms at
intervals of three or four years in the
case of open-pollinated varieties. Silver
Mine and Kalahari continue to be in
production today. Silver Mine is highly
productive but very susceptible to
lodging, and Kalahari is very
susceptible to pests and diseases.

Since 1977, materials of various
origins, both varietal and hybrid, have
been tested, and certain varieties from
CIMMYT's collection have shown
themselves consistently superior to the
pre-independence varieties. For five of
these, Obregon 7643, Cotaxtla 7921
and Ferke 7822 for high altitudes, and
San Andres 7823 and Mexico 8049 for
the northern lowlands, prebasic seed
has been produced. For the south, no
adequate varieties have yet been
identified; all of the above are highly
susceptible to maize streak virus.
Materials with horizontal resistance to
maize streak have been obtained from
IITA in Nigeria, and they are being
tested this season. The trials are not
yet harvested, but good resistance to
attack is evident in some of the
varieties (Figure 4).

Eight drought-resistant varieties from
CIMMYT are also being evaluated this
season. However, six have been
eliminated by virus attack and/or
drought, and even the remaining two
(Ilonga 8043 and Ikenne 8243) are
moderately susceptible to virus attack.
Other breeding programs have been
started on a modest scale to improve
the varieties selected for the north, as
well as Silver Mine and Kalahari.
Germplasm is being collected from the
peasant sector for evaluation and use
in these programs. Conditions for long-
term seed storage are presently poor;

the program's cold storage facility was
damaged in the floods of 1984.
However, the most important
materials, including the drought and
virus-resistant materials, are being

In experiments, well-adapted hybrids
have consistently given higher yields
than open-pollinated varieties in the
highland areas. However, the
production of any of these hybrids
implies the annual importation of large
quantities of seed, bought with scarce
foreign exchange. At present, this is
not justified by the yields obtained on
the vast majority of farms.

80- 8
1. Mayo Galke 82TZESR-W
2. EV8431-SR (BC3)
3. Gusau 82TZESRW 10
70- 4. kenne 82TZESR-W
5. IBK
6. EV8430-SR (BC)
7. EV8435-SR (BC
60- 8. Gusau 81 Pool 16
S 9. R200
10. Kalahari

1 2 3 4 5 6 7 8
Weeks after planting
Figure 4. Results of trials of IITA maize
varieties for resistance to maize streak virus,

Since the arrival of the Yugoslavian
team, local hybrid seed production has
begun. Various tropical hybrids from
the Yugoslavian Maize Institute are
being tested in Lichinga, with SR52 as
the control. Last season, 180 hybrids
were tested and ZP752b significantly
out-yielded SR52; the yield of 27
others did not differ significantly from
the control. Breeder and prebasic seed
of the parent lines of ZPSC852b was
produced last year in pilot production,
together with breeder seed of other
inbred lines.

Basic seed production
Mozambique suffers from a lack of
adequate basic seed of the varieties in
production, because no organized
maintenance of these varieties took
place for a number of years. INIA is
now responsible for basic seed, and
last year some was produced for both
Obregon 7643, which is almost ready
for release, and Kalahari, which
remains in production. Also, small
quantities of seed of the parental lines
of the experimental hybrid ZPSC852b
have been produced.

Certified seed production
ENS has three large farms for certified
seed production, Namialo (for the
northern lowlands), Chimoio (for the
highlands) and Lionde (for the
southern lowlands) in Nampula,
Manica and Gaza provinces,
respectively. This seed is for open-
pollinated varieties. All of the problems
of mechanized agriculture are also
found in seed production, with the
added problem of the lack of drying
facilities, which jeopardizes seed

Quality control and seed pathology
The SNS operates a central laboratory
in Maputo and two small laboratories
in Namialo and Lionde. An additional
laboratory is being constructed in
Chimoio. The national staff at SNS is
able to conduct purity and germination
tests, but as yet has little experience in
field inspection.

No seed population of any significant
magnitude has proved to be free from
seed-borne infectation, and it is
common to encounter from two to
seven different pathogens in a single
seed sample. As a result, an average of
three seeds are required to generate
one plant. Damping-off losses are
estimated to account for approximately
30% failure in plant establishment.

Seed is generally treated with
insecticide (Damfin) and fungicide
(Captan). However, experiments on the
treatment of seed with a wide range of
fungicides and bactericides have
indicated that the contribution of
bacteria to poor establishment is as
important as that of fungal pathogens.
The nitrofuran antibacterial
formulation Furasol at 100 ppm (by
seed weight) was found to be the most
effective single seed treatment.
Bacterial infection is caused by
members of the genera Erwinia and
Pseudomonas and often occurs in the
presence of Fusarium spp.
E. carotovora is widespread.

Seed distribution
A state company (BOROR) distributes
seed to state, cooperative and private
farms, whereas a second state
company (Agricom) and the provincial
agricultural directorates are involved in
the distribution of seed to the peasant
sector. The present unsettled
conditions in the country pose major
difficulties in distribution, as do the
lack of transportation and storage
facilities and the shortage of trained

Maize Research Staff and Training

The shortage of trained personnel in
Mozambique is grave. There are fewer
than 50 Mozambican agronomists; of
those, four work in INIA with two
assigned to the maize program. Two
more are located at the university and
one of the largest state farms; they also
collaborate with the maize program.

The maize program staff also includes
five agricultural technicians, one at the
diploma and four at the certificate
level. None of the personnel have much
experience in maize research, and all
need further training.

At present, the certificate and diploma-
level staff attend two- to three-week
annual courses at INIA on various
aspects of research and agronomy. For
the graduate staff, courses of four to
six months are planned at
international institutes, currently in
Yugoslavia and at CIMMYT. It is
intended that this staff specialize in
breeding, agronomy or basic seed

The national staff of the DDR centers
have only certificate-level training, and
the work is headed by expatriate


The war is at present disrupting both
maize production and research.
However, the placement of qualified
and experienced staff in the
experiment stations is already showing
results, not only in improved
experimental work and basic seed
production, but also in the
identification of factors limiting maize
production. Efforts to improve maize
production must concentrate on the
following aspects:

* Further staff training, especially at
the international institutes;
* Production of good-quality seed of
the best open-pollinated varieties
presently available;
* Selection and breeding for resistance
to maize streak virus and downy
mildew, and
* Organization of available data for a
clearer understanding of the major
agronomic limitations, and the
selection of appropriate solutions.

As soon as the security situation
returns to normal, more emphasis
must be placed on field diagnosis and
economic evaluation to provide a
firmer basis for the design of the
research program.


1. Bruno, A., and J.C. Duhart. 1984.
O milho no sector campones.
Experimentacdo aplicada.

2. Compton, J. Manual das Ervas
Daninhas de Mocambique. (In press.)

3. De Carvalho, M. 1969. A agriculture
traditional de Mocambique. Missao
de Inqu6rito Agricola de
Mocambique, Lourenco Marques,

4. Hoekstra, S. 1983. Maize research
on a state farm in northern
Mozambique. (Unpublished.)

5. Kassam, A.H., et al. 1982.
Assessment of Land Resources for
Rainfed Crop Production in
Mozambique. Field Document 32.

6. Woodhouse, P.J., and C.J. Rendle.
1983. Program de Ensaios de
Adubac&o. Relat6rio Trienal
1979-1982. Comunicacoes, S6rie
Terra e Agua 23. Institute Nacional
de Investigacao Agronomica,
Maputo, Mocambique.

The Reunion Island Maize Breeding Program
J.L. Marchand and E. Hainzelin, Institut de Recherches
Agronomiques Tropicales et des Cultures Vivrieres, St. Denis,
Reunion, Indian Ocean

Maize has a long history in the islands
of the Indian Ocean. It was introduced
very early and is an important source
of food. Today, its importance varies
from island to island. It is important in
Reunion, and even more so in Comores
and Rodrigues. It is of secondary
importance in Madagascar, Mauritius
and the Seychelles.

Reunion is a small island, located
between Madagascar (700 km) and
Mauritius (200 km). Sixty thousand
tons of maize are utilized on the island
annually, of which 40,000 to 45,000
tons are imported. Most of the maize is
used as feed for livestock; however,
that consumed by the islanders
themselves, although the amount is
not known exactly, is far from
negligible. Probably in certain regions,
such as the western highlands and
Cirques, maize is the principal staple

There is widespread cultivation of
maize on the island; at least one-third

of the farmers grow maize although
the total area under cultivation is
small. This is estimated at less than
6% of total arable land (Table 1).
However, for several years, there has
been an upward trend in maize
cultivation, presumably in response to
the scarcity of maize on the local
market. As cultivation patterns are
extremely varied (there is both single
cropping and intercropping of various
degrees of complexity), and the harvest
does not pass through the type of
market system that would make
counts readily available, figures
presented here are estimates.

The remarkable adaptability of maize
is evident from the various
microclimates of Reunion where it is
grown; these environments range from
those of the coast (sea level) to 800 to
1000 meters elevation. Maize is also
grown under all types of cultivation
patterns, pure-stand intensive
cultivation (sometimes under
irrigation), intensive cultivation

Table 1. Maize area and production, Reunion Island, 1983

Undeveloped Developed o/o of total
Cultivation area area a/ cultivated Production
Scheme (ha) (ha) area (tons)

maize 2780 4450 4.33 10,690
maize 580 580 0.90 920

Total 3360 5030 5.23 11,610

a/ Developed area represents total area planted to maize; if there are two
harvests per year, these figures will double
Source: Hainzelin (4); Ministry of Agriculture (13)

* Institute for Research in Tropical
Agriculture and Food Crops

intercropped with sugarcane (with
manure fertilization and good cultural
practices, in some cases completely
mechanized), semi-intensive cultivation
intercropped with legumes, flowers,
etc., and very intensive garden-style
cultivation. For all of these cultivation
schemes, farmers use local varieties
which have evolved from old
introductions which have undergone
decades of natural selection and have
become remarkably well adapted to
local conditions. A large number of
different maize ecotypes are in use,
due to the varied environments found
on such a small island.

Research on Maize Viruses

In tropical Africa, certain viruses, such
as maize streak virus (MSV) and maize
stripe virus (MStpV), often cause severe
damage, sometimes destroying entire
crops, i.e., MSV epidemics in East
Africa and MStpV epidemics in Sao
Tome. In Reunion Island, climate
peculiarities have caused several
viruses to merge into a particularly
aggressive "cocktail;" MSV, MStpV,
maize mosaic virus (MMV) and
sugarcane mosaic virus (SCMV) are all
found on the island. The insect vectors
of MSV (Cicadulina mbila) and MMV
and MStpV (Peregrinus maydis) thrive
under island conditions.

All of the island's cultivated maize
materials have undergone extreme
natural selection in order to survive
not just one virus, but this
combination of viruses which varies
from place to place. Therefore, the
collection of local varieties (at least
those grown on the coast where the
viruses thrive) presents remarkable
levels of tolerance to the different
viruses. This becomes evident in the
comparison of local and introduced
varieties. Several hundred
introductions from all parts of the
world have been tested, but none have
been found to have tolerance
comparable to that of the Reunion
varieties. Research programs in other

tropical countries are stiving to create
varieties tolerant to the virus diseases,
but in the Reunion environment their
tolerance is not found to be sufficient.
This may be because the Reunion
Island virus strains are particularly
aggressive, or because a tolerance to
one virus does not necessarily imply a
tolerance to another one. For this
reason, Reunion ecotypes are choice
genetic materials for programs
selecting for virus tolerance, and many
leading programs have used them.
Exhaustive trials in the USA have
recently confirmed the exceptional
characteristics of the island varieties

In 1979, the Institute for Research in
Tropical Agriculture and the
Department of Subsistence Crops
(IRAT) in Reunion began research on
virus tolerance to take advantage of
the island varieties and the
exceptionally favorable environment
for virus work. The program has now
grown considerably with additional
financing from the European Economic

Maize Breeding

IRAT's program for the improvement
of maize varieties has two objectives,
the meeting of the specific needs of
Reunion farmers and the search for
varieties or hybrids suitable for the
tropics in general; in both cases there
is emphasis on tolerance to streak and
stripe viruses. Happily, these two
objectives are compatible, at least as
regards yellow maize, which is the
type grown in Reunion. The island's
conditions have been found to be
particularly favorable for the selection
of maize in its two environments. On
the coast, with its low-altitude, hot,
tropical climate, streak viruses and the
two classic low-altitude diseases,
Puccinia polysora and
Helminthosporium maydis are found,
as well as "warmonger" birds (Pioceus
cucullatus), which attack the maize
ear. In the mountains (above 800

meters altitude), with cooler, high-
altitude tropical climates, viruses are
rarely a problem, but
helminthosporium (H. turcicum) is
severe. Puccinia sorghi is also present,
and insect attacks and soil diseases
can do serious damage. Stalk borers
(Sesamia spp.) often cause considerable
damage. At even higher altitudes
(above 1000 meters), soils become
highly acidic, with pH levels as low as
4. In all of these regions, winds are
often strong and rainfall can be either
excessive or lacking altogether.

Socially, the farmers in Reunion
include the three principal groups,
traditional farmers, those in the
process of modernization, and
commercial farmers using intensive
cultivation systems. This justifies
research into improved local varieties
as well as into hybrids and other
improved varieties. There is a need for
maize of early, medium and late
maturity. The classic situation is,
therefore, found in the breeding
program, that of working with the
three maturity groups of maize for the
three levels of farmers, both for the
coastal areas and for the highlands.

The program has two objectives, the
development of two groups of varieties
and hybrids, those suitable for hot,
tropical climates and those for high-
altitude environments, and the
selection of streak-resistant varieties.
Thus, the work of the program consists
of the collection, study and
preservation of plant materials, the
development of varieties for low-
altitude sugarcane-growing regions, the
development of varieties for high-
altitude regions and the study and
utilization of virus-resistant varieties.

The collection and
preservation of material,
The collection and, above all, the
preservation of plant materials are
subject to various problems. The
sensitive coastal varieties are difficult
to renew, because of virus presence.

Also, labor costs limit the volume of
renewals. Storage has been a problem,
but it should be solved with the
construction of a cold storage building
in St. Denis. Because of these
problems, the current collection
numbers only 300, of which 85 are
local varieties and some 60 are
introduced varieties (mostly from the
Indian Ocean area and East Africa); the
rest are either local or introduced lines.

The development of
varieties for low-altitude areas
Figure 1, a varietal grid, summarizes
the origins of the plant materials that
are presently being used by the maize
breeding program for the three groups
of farmers in the lowlands.

Traditional farmers (Target 1)-For
traditional farmers, local varieties are
used. Ecotypes collected on the island
have been tested for several years and,
from them, twelve varieties will be
chosen. Early maturing maize ecotypes
were collected on Rodrigues and
testing began in 1984; their resistance
to viruses makes them promising. A
simultaneous selection of early
maturing varieties is in progress in

Farmers in the process of
modernization (Target 2)-For these
farmers, improved varieties are best
adapted to their needs. CIMMYT
varieties perform well for them in the
absence of viruses. Both the yellow and
white CIMMYT maize are excellent for
carrying out resistance transfers (over
the long term), as well as for planting
in winter, when viruses are less of a
problem. IITA's streak-resistant
selections, which were tested in 1984,
are also good in a number of respects,
but are sensitive to MMV and MStpV.

Commercial farmers (Target 3)-For
these farmers whose production
system is intensive and mechanized,
hybrids are best. Introductions from
France, South Africa and Madagascar
have proved disappointing due to virus

sensitivity. However, tests are being
continued with materials originating in
the tropics. Work on local x temperate
zone hybrids has led to the recommen-
dation of IRAT 143 = (Revolution x
INRA 508). However, it will have to be
replaced as it is very sensitive to

Although the 1137 line from South
Africa seems to show a certain
tolerance to viruses, it cannot replace
INRA 508 because it is much later
maturing. It can, however, provide
promising hybrids in crosses with
Revolution; the hybrid IRAT 279 =
(Revolution I137TN) is proposed for
development. It can also be used in
crosses with other lines originating in
Revolution. Other sources of genes
being tested are Rodrigues lines
crossed with Revolution, and
Rodrigues and Reunion varieties
crossed with Revolution.

All of this work is being done with
yellow maize. White-grain maize is not

being used at this time due to its
susceptibility to virus attack.

The development of
varieties for high-altitude areas
Work on high-altitude maize is less
advanced than that for coastal areas
for a number of reasons. Historically,
before 1975 most trials were conducted
on the coast; later, those planted at
higher altitudes often failed as they
were planted in farmers' fields and
were not adequately controlled. Also,
the maize teams are located on the
coast, and limited land is available for
their use in the highlands. Although
the demand for highland maize
varieties calls for increased research,
this will only be possible when
sufficient facilities become available.

Traditional farmers (Target 1)-Of the
85 local varieties in the collection, 20
are being used for selection; all are of
yellow maize. Traditional white
varieties will probably be introduced
from various African countries in the

Figure 1. Origins of maize materials used for low-altitude tropical environments, Reunion

Target Early maturity Medium maturity Late maturity
Group Yellow White Yellow White Yellow White

Traditional Reunion local ? Reunion local ? Reunion local ?
farmers (1) varieties varieties varieties
Rodriguez local Revolution
varieties improved varieties
Revolution Reunion composites
varieties Revolution streak-
and lodging-
resistant varieties

the process and varieties lines and lines and varieties and IITA IITA lines and IITA
of moderni- varieties lines and and varieties lines and
nation (2) varieties varieties

Commercial Locally ? Locally ? Locally ?
farmers with developed developed developed
intensive, varieties varieties varieties
systems (3)

Farmers in the process of
modernization (Target 2)-The answer
to the need for greatly improved
varieties by these farmers may be
provided by CIMMYT, in both yellow
and white maize, in spite of some
sensitivity to H. turcicum. Selection
made in 1984 points to the
recommendation of Tocumen (1) 7931
for this group.

A breakthrough in the understanding
of resistance to H. turcicum has come
about as a result of the work of
phytopathologist J.C. Girard in St.
Pierre, and from introductions from
South Africa being multiplied in Mon
Caprice. Lines carrying different
resistance genes are presently being
tested, using artificial inoculations.

Commercial farmers (Target 3)-
Hybrids developed specifically for the
coast, as well as French and South
African hybrids, are being tested
simultaneously. Although in the
beginning stages, tests have already
shown that the early-maturing French
hybrids are poorly adapted to the
region. Nevertheless, it is felt at
present that it may be better to
continue to utilize introductions, rather
than to develop original hybrids. This
will lead to quicker results in the

The study and utilization
of virus-resistant sources
Preservation of local genetic material-
With the increase in the exchange of
materials and the opening up of the
islets, the diversity of the island's
ecotypes was beginning to be
threatened. In 1979, the entire island
was searched and 85 local varieties
collected (5). Their behavior under
severe virus pressure allowed for the
selection of the most promising
material for crossing for composite
virus research. It has been noted that
the island of Rodrigues has promising
local maize. Therefore, a search of
Rodrigues Island was made in 1980 in
close collaboration with MSIRI

(Mauritius Sugar Industry Research
Institute) (3). Selected ecotypes
demonstrated tolerance to viruses, and
several varieties were included in
composite virus research.

Exploitation of tolerance to virus
diseases-Existing tolerance to virus of
island maize offers the challenge of
complex biological phenomena that
researchers have long sought to
understand. They have also searched
for an answer to how this tolerance is
genetically determined; no doubt it
involves numerous genes in multiple
allelic combinations. The varieties
must be resistant to a combination of
viruses, which varies from one area to
another and even from one field to
another. This study thus involves a
complexity of the highest order. The
Reunion breeders have chosen the
simplest and most effective route of
classic repetition to concentrate
tolerance. Simultaneously, there will
be an attempt to transfer this tolerance
to various materials, such as breeding
stock, composites used in other
selection projects and varieties popular
in West Africa, where virus disease is
beginning to take on importance.

Selection principles are simple,
although their execution and
effectiveness require a great deal of
practice and the existence of certain
tools, such as techniques for rapid and
simple virus detection, a knowledge of
biology and the insect vectors, and the
mass rearing of a great number of
vectors for artificial infestation.

Detection of viruses-Several analyses
have been carried out in France, with
the help of the phytopathologists of
MSIRI, to catalog the maize viruses
present on Reunion. The four viruses
cited earlier (MSV, MMV, MStpV and
SCMV) were detected either by
microscopic observation, by serum
testing or by transmission testing.
During the next two years, within the
framework of the CCE project, various
techniques will be used to continue

virus detection. It will be necessary to
perfect rapid serological techniques
(the ELISA test in particular) to allow
for the identification of a great number
of on-site samples. These techniques,
which are indispensable for repeated
selection tests, should be operational
by early 1986.

Mass rearing of vectors-For efficient
selection for virus resistance, vector
stress must be the result of artificial
infestation that is severe, homogeneous
and reproducible. This assumes the
availability of a great number of
vectors capable of transmitting the
disease; it also assumes the capability
for mass rearing the Cicadulina mbila
and Peregrinus maydis vectors and the
ability to optimize their performance in
acquiring and transmitting the virus
during the selection tests. A beginning
study of insect populations and their
vector mechanisms has now been
completed. In the next two years, work
will continue on improving mass
rearing techniques.

. Maximum collaboration among
breeders will determine the success of
the last phase, multilocational testing
of selections for tolerance, especially
those varieties into which tolerance
has been transferred. Their working
together can lead to the refinement of
methods and the success of the


1. Damsteegt, V.D. 1983. Maize
streak virus: Host range and
vulnerability of maize germplasm.
Plant Disease Reporter

2. Hainzelin, E. 1979. Situation
6conomique du Maiz a la Reunion.
Trial record 79/21.

3. Hainzelin, E. 1980. Rapport de
mission A l'ile Rodrigues:
Collectage mais. IRAT-REUNION
Document 145.

4. INSEE-REUNION. 1983. Panorama
de l'tconomie de la Reunion 1983.

5. IRAT-REUNION. 1980. Rapport
Annuel 1979. Reunion.

6. IRAT-REUNION. 1981. Rapport
Annuel 1980. Reunion.

7. IRAT-REUNION. 1982. Rapport
Annuel 1981. Reunion.

8. IRAT-REUNION. 1983. Rapport
Annuel 1982. Reunion.

9. Ministere de l'Agriculture SCEES.
Bulletin de statistiques agricoles.

10. Ministere de l'Agriculture. 1977.
Recensement g6enral de
l'agriculture 1974. SCECS Study
154. Reunion.

11. Ministere de l'Agriculture. 1982.
Importations-exportations des
principaux products agricoles de
1970 A 1982. Reunion.

12. Ministare de l'Agriculture. 1982.
Recensement g6enral de
l'agriculture 1980-1981: lers
r6sultats. Reunion.

13. Ministere de l'Agriculture. 1984.
Annuaire de statistiques agricoles
1983. Direction D1partementale de
la Reunion. Reunion.

14. Service des Douanes de la
Reunion. 1984. Personal

In Zaire, maize is first among the
cereals and is one of the basic
subsistence crops, along with rice,
cassava and bananas. Maize is
cultivated in all regions, although the
major production zones are Kasai,
Shaba and Bandundu. In 1975,
675,100 hectares were planted to
maize (16% of the subsistence crop
area and 67% of the area planted to
cereals), producing 495,400 tons (2.6%
of subsistence production and 65% of
cereal production). Thus, in terms of
area, maize is second only to cassava;
in volume it is third, after cassava and
bananas. The Agricultural Renewal
Plan estimated 1980 maize production
at 562,340 tons; in 1977, 509,600 tons
had been produced.

Maize is not used only as food in Zaire,
but also by the national breweries for
beer. Its recent introduction into urban
and semi-urban centers has increased
demand considerably, resulting in a
shortage of approximately 200,000
tons annually. One of the objectives of
the Agricultural Renewal Plan 1982 to
1984 (an extension of the Mobutu Plan)
was to eliminate maize shortages by
1984 by raising production from the
1982 figure of 687,785 tons to 810,630

Maize Research
in the Great Lake Countries

In Burundi and Rwanda, research is
essentially aimed towards the
development of varieties adapted to the
various ecological zones of the two
countries. In Rwanda, particular
research emphasis is on developing
varieties adapted to high altitudes and
for meeting the needs of the glucose
and starch mill at Ruhengeri
(Mukamira). In Zaire, maize research is
conducted for the development of an
integrated program to improve plant
breeding techniques (cultivation trials)
and increase maize production, both
quantitatively and qualitatively.

Maize Production Constraints

The factors limiting maize production
in Burundi are:

* Lack of high-yielding, early-
maturing varieties adapted to high-
altitude regions (the variety
currently distributed in these
regions is late maturing, thus
preventing the timely planting of the
following crop);
* Poor soil fertility, and
* The diseases maize streak virus and
downy mildew, which cause great
economic losses.

The research program is seeking
solutions to these problems.

The factors limiting maize production
in Rwanda are:

* Diseases, the most important of
which are maize streak virus and
downy mildew;
* Poor soil, and
* Climatic conditions, which are
unfavorable for the cultivation of
maize. These include drought in the
eastern regions and the high altitude
of the northern regions, which
prolong the maize growth cycle,
making it susceptible to
cryptogamic diseases.

The maize program is searching for
solutions to all of these problems for
maize cultivation.

The factors limiting maize production
in Zaire exist at five levels:

* Non-utilization of appropriate seed
and technologies;
* Severe diseases, such as maize
streak virus and downy mildew;
* Lack of adequate roads;

* Lack of a high-quality training
program for maize program staff,
* Lack of sufficient vehicles and/or
gasoline for program staff use.

The solution to these problems must
come from high government levels and
through the research of the National
Maize Program.

Maize Research in Burundi

Varietal selection
The station at Kisozi, created in 1929,
conducts research on maize and other
high altitude crops. Breeders work with
both the local maize variety Igarama
and introduced varieties. Selection
criteria are productivity, grain quality,
adaptability to various ecological
conditions and early maturity. The
selected varieties made available to
farmers have been Kisozi 41 (in 1941)
and Bambu and GPS5 (after 1964). In
1977-78, with Belgian cooperation,
multilocational, semi-annual trials were
set up at altitudes of 830 and 2250
meters. The objective of the program
was the identification of varieties
adapted to each of the country's
ecological areas. These trials were
concluded in 1980, and the results
analyzed in 1980-81. Table 1 shows
the recommendations made as a result
of the trials.

Table 1. Maize varieties recommended for
four ecological zones as a result of varietal
selection trials, Burundi, 1981

(m) Variety

800 to 1250 GPS5
1250 to 1500 GPS4 x SR52
1500 to 2000 Igarama-4
Above 2000 Kitale and Isega

The Selected Seed Service has
observed that the GPS4 x SR52
varieties are very sensitive to stalk
borers. Moreover, all of the varieties
are sensitive to maize streak virus,
especially under conditions of low

A new varietal improvement program
was begun in Burundi in 1979,
following a cooperative agreement
signed with the International
Development Research Center.
Objectives of that program included:

* Identification of varieties superior in
yield and disease resistance to those
currently recommended;
Crossing of local adapted maize with
promising introduced varieties at
Imbo and Kisozi (this will be a long-
term program);
Selection of varieties resistant to
maize streak virus, utilizing IITA
varieties (this disease is found
throughout the country, especially
on the Imbo plains in the second
season and in the marshes of the
intermediate and high-altitude
areas), and
Improvement of currently
recommended varieties for yield and
resistance to lodging and the major
diseases (leaf blight, rust and maize
streak virus). This will be a
permanent, ongoing program.

The program also conducts CIMMYT
international trials (EVTs and ELVTs),
and in 1982 it also began participating
in the regional trials of the Burundi
National Maize Program. The following
changes in recommendations were
made as a result of the 1983
multiregional trials:

For low and midaltitudes, the high-
yielding varieties with acceptable
characters were Across 7643 (a
variety with relatively wide
adaptation) and Cotaxtla 7929; these
varieties would continue to be field
tested. The GPS4 x SR52 variety
was very susceptible to weevil.

GPS5 was proven to be the best for
the plains areas, due to its
satisfactory resistance to maize
streak virus.

For high altitudes, the introduced
varieties performed poorly, with low
yields, disease susceptibility and/or
late maturity. Kitale continued to be
the best choice among late varieties.
Igarama-4 had equal or only slightly
higher yield than local varieties;
however, it would be recommended
until a better variety became
available. Isega, a local variety
which matured a month earlier than
Kitale, had good potential and was
very resistant to cold; it was chosen
for selective breeding in 1984. For
the improvement of the
recommended varieties, 250 families
were selected for each variety, and
these were used for basic seed in

Varietal improvement
In 1981, the Maize Program began
crossing varieties as part of its
improvement objectives. GPS5 was
crossed with 25 varieties of the 1981
collection, and hybrid testing was
conducted in 1981B (results are not
available). The IITA hybrids TZSR-W
and TZSR-Y, which possess genetic
resistance to maize streak virus, were
crossed with GPS5, and the resulting
hybrids were tested in 1982B (results
not available).

Agronomic testing
Planting density and spacing trials-At
Kisozi in 1981A, the variety Kitale was
tested in trials comparing five planting
densities, from 30,000 to 70,000 plants
per hectare (with two seeds per hole);
each density had three different
spacings between the rows, 50, 70 and
100 cm. The highest yield was
obtained with a planting density of
approximately 50,000 plants per
hectare, but the difference was not
significantly different from that of
30,000 plants per hectare.

Intercropping trials-In the research
program on beans, the Burundi
Institute for Agricultural Sciences
(ISABU) conducts trials on
intercropping maize and green vine
beans. Since this program is new
(1981A), the trials have not yet
resulted in recommendations to

Fertilizer-use trials-A fertilizer-use
trial, also utilizing the Kitale variety,
was conducted at Kisozi in 1982 in
order to define adequate dosages of
nitrogen and phosphorus and their
interaction. The results of this trial are
not yet available.

Maize Research in Rwanda

Varietal selection
Maize research here began in 1930 at
the Rubona station of the National
Institute for the Study of Agronomy in
the Congo (INEAC). Research goals for
the program were:

* Introduce varieties that would be
more productive and earlier
maturing than the local varieties
(from 1930 until 1982, over 650
varieties were introduced from more
than 40 countries; in 1982, the
collection totaled 483 varieties), and
Utilize genetic improvement to raise
maize yield and shorten the growing

In 1978, the Rwanda Institute for
Agricultural Sciences (ISAR) replaced
INEAC and continued to pursue these
goals. The best varieties developed in
the program were:

* Golden Corn (1951), which was
rejected by consumers and millers
because of its hard grain, and its
distribution suspended.
Nevertheless, it is still
recommended, due to its high
productivity and medium maturity.

* Bambu (until 1966), which is
predisposed to lodging due to its
height. However, its yield and
resistance to diseases make it
suitable for the central regions and
areas with sufficient rainfall. In high
altitudes, its growing cycle is long
(eight to nine months), and under
prolonged dry conditions the stalk is
weak and very susceptible to corn
borers. This variety is in high
demand at the Selected Seed

* Katumani, which is early maturing
and well adapted to the eastern
parts of the country. The first
adaptability tests conducted around
Lake Kivu revealed its positive
potential in that region.

* Nyirakagoli, which emerged from
massive selection of local materials
in 1975. It is adapted to high
altitudes, but can also be grown in
other environments in the country.
Its early maturity, large grains and
sweet flavor make it more desirable
than either Bambu or Golden Corn.
The Selected Seed Service notes that
its growing cycle is quite long in
high altitudes.

Table 2 is a summary of some of the
features that characterize each of the
recommended varieties.

Varietal improvement
Pedigree selection for grain color and
early maturity was begun at Rubona in
1955, and a white maize having the
same yield as the yellow Plata was
perfected. This program was later
dropped, but was reinitiated in 1980.

The Nyirakagoli, Bambu and Golden
Corn varieties were crossed with
materials with brachytic-2 genes for
short plant height and opaque-2 genes
for high lysine content. These same
varieties were also freely crossed with
different CIMMYT varieties in attempts
to raise their yield levels. Before the
br-2 genes were incorporated into
Bambu, this variety was treated with
Cycoccel Extra and with X-rays to
inhibit the elongation of the internodes
and thus reduce its height; this
treatment was very expensive and did
not prove to be effective. In 1981, the
br-2 and opaque-2 genes were
combined and introduced into Bambu.
Work on all of the resulting materials
continues, and apart from reduced
height and increased lysine content,
early maturity and resistance to
disease and insects are also being

Table 2. Characteristics of recommended maize varieties, ISAR, Rwanda

Variety and Year of Growth Plant Yield of
country intro- Area of cycle height dry grain
of origin duction adaptation (days) (m) (kg/ha)

Golden Corn 1930 Midaltitude, 155-225 2.75 4000
(Zaire) eastern area
Bambu 1959 Midaltitude, 155-225 3.15 5000
(Zaire) eastern area
Katumani 1972 Low elevation, 99-110 2.25 2500
(Kenya) eastern area
Nyirakagoli 1975 Mid- and high 130-160 2.75 3800
(Rwanda) altitude

Note: ISAR has participated in comparative trials in collaboration with East Africa (RMSX,
Kms, WMsX) and with CIMMYT international trials (EVTs, ELVTs)
Source: F. lyamuremeye, 1983, Synthdse des Acquis Scientifiques et Techniques de la
Recherche Agricole au Rwanda.

FAO is developing a program for the
creation of photo-insensitive maize
adapted to both tropical and temperate
regions. Within the framework of this
program, the seeds of the base
varieties were planted at Rubona and
were crossed with Bambu in 1980. The
best crosses have been sent to Rome
for further study.

Agronomic testing
Planting density and spacing-Planting
density and spacing trials at the
station have shown that, for Bambu,
optimal yield is obtained with a density
of 40,000 to 60,000 plants per hectare,
with respective spacings of 80 x 60 cm
with two seeds per hole and 80 x 20
cm with one seed per hole. Highest
yields of Katumani were achieved at
125,000 plants per hectare under
station conditions at Rubona.

Intercropping trials-Experiments in
intercropping maize with other crops
have shown that maize and beans,
maize and soybeans, and maize and
peanuts are the most satisfactory. The
combination of maize and vine beans
is good as the maize plants serve as
stakes for the beans, although the
maize decreases the bean yield.

Fertilizer-use trials-In several trials
conducted at Rubona station, yields of
up to 6 t/ha have been achieved with
the application of 120-60-60 units of
nitrogen, phosphorus and potassium
per hectare. The Golden Corn variety
responded better to a complete
fertilizer (100-100-100).

Irrigation trials-In Karama in 1971A,
irrigation trials were conducted to
determine the optimum water regimes
for maize. The average daily
evapotranspiration (ETP) was
determined to be 4.5 mm or a total of
651 mm of water for the growth cycle
of 138 days for Bambu. The ETP
coefficient versus the evaporation of a
free sheet of water determined
irrigation frequency and quantity;
natural precipitation and evaporation

were also taken into account. The
water provided by sprinkler irrigation
averaged 170 mm. No trials were
conducted in 1979-80 as the sprinkler
system was out of order, but irrigation
research was re-initiated in 1981.

Maize research in Zaire

Selection and varietal improvement
Maize research began in Zaire at the
Gandajika station in the 1940s and
was later extended to the stations at
Kiyaka, Mulungu, Nioka, Bambesa and
Yangambi. Each station focused on a
different set of activities, and there was
very little contact between stations.
Therefore, the selected varieties were
highly specific to their regions of
origin, leading to difficulties in
adaptation in other environments.
Approximately 12 years passed
between the initiation of the selection
process and the launching of a new
variety. Until 1972, the varieties of
maize recommended in southern Zaire
were GPS4, GPS5, double hybrid and
Hickory King.

The National Maize Program has
conducted research since 1971 at
Kisanga in the Shaba region and other
stations throughout the country. The
PNM focuses on maize research,
production, and the training of national
staff. Since the beginning, the goal of
the program has been the development
of high-yielding varieties that fulfill
farmer preferences (exclusively white
maize), are adapted to the various
agroecological regions of the country
and are resistant to the principal
diseases (maize streak virus and
downy mildew). Over the years, the
National Maize Program has developed
and distributed six improved varieties,
Shaba Safi, Salongo, Salongo 2,
PNM 1, Kasai 1 and Shaba 1. However,
since September of 1978, only
Salongo 2, Kasai 1 and Shaba 1 have
been recommended by the program.
Table 3 lists some characteristics of
the National Maize Program varieties,
as well as their cultivation zones.