Front Cover
 Realizing the potential of maize...


1989 / 90 CIMMYT world maize facts and trends : realizing the potential of maize in Sub-Saharan Africa
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00087042/00001
 Material Information
Title: 1989 / 90 CIMMYT world maize facts and trends : realizing the potential of maize in Sub-Saharan Africa
Physical Description: Book
Language: English
Creator: International Maize and Wheat Improvement Center
Publisher: International Maize and Wheat Improvement Center
Place of Publication: Mexico, D. F.
Publication Date: 1990
 Record Information
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UF00087042:00001

Table of Contents
    Front Cover
        Front Cover
    Realizing the potential of maize in Sub-Saharan Africa
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
Full Text
Oh &3
3?. "r




Part 1: Realizing the
Potential of Maize in Sub-Saharan Africa*

* Prepared with the collaboration of IITA.


In the wake of the Green Revolu-
tion of the 1960s and 1970s, when a
sharp surge in wheat and rice
production greatly reduced food
imports in many developing coun-
tries, an end to hunger appeared in
sight. More recently, however,
some of the greatest beneficiaries of
the technological innovations of the
1960s and 1970s-including China,
India, Pakistan, and Mexico-have
experienced slower rates of growth
in cereal production, and there is
evidence that historical growth
rates may be difficult to sustain
(CIMMYT 1989). The realization
has slowly been dawning that the
Malthusian race between popula-
tion growth and food production is
not yet won, at least not in the de-
veloping countries, and that the
world's ability to feed itself is more
precarious than had previously
been thought.

The problem of lagging food produc-
tion is most evident in sub-Saharan
Africa, a region largely unaffected
by the technological innovations
that so profoundly transformed the

1 Throughout this report, "Sub-Saharan
Africa" and "Africa" are used inter-
changeably to denote all countries in
Africa except those in North Africa
(Morocco, Algeria, Tunisia, Libya, and
Egypt) and the Republic of South Africa.
2 The primary data sources used in prepar-
ing this report include the AGROSTAT
data base compiled by the Food and
Agriculture Organization of the United
Nations (FAO), as well as the World
Development Report series published
annually by the World Bank. Some of the
data in these sources are questionable,
particularly the African data, because
official statistics are weak in a number of
countries. Empirical results appearing in
this report must therefore be interpreted
with some degree of caution.

world's wheat and rice economies.'
Unlike Asia, where rice and wheat
are the major staple foods, in Africa
diets are based primarily on coarse
grains (maize, millet, and sorghum)
or on roots and tubers (cassava,
yams, and sweet potatoes). Many
African staple food crops did not
benefit directly from the technologi-
cal advances associated with the
Green Revolution, either because
improved seed-fertilizer technolo-
gies were lacking for the African
crops, or, if they were available,
because government policies dis-
couraged their adoption by farmers.
As a result, food production in
Africa has not kept pace with rising
population. While population
growth averaged 2.9% from 1961-65
to 1984-88 for the region as a

whole, growth in cereal crop
production averaged only 1.9%
during the same period.2 Thus, per
capital cereal production in Africa
has actually fallen during the
past 25 years, a phenomenon
unparalleled elsewhere (Figure 1).

Most African nations have re-
sponded to widening food deficits by
importing increasing amounts of
cereals, primarily wheat and rice.
Consumption of imported cereals
has been encouraged by economic
policies that have made imports
cheaper than domestically produced
staples, and by demographic
changes (such as urbanization and
women's entry into the formal labor
market) which have increased the
demand for convenience foods. For

340 -m







60 i I I I I I I I I I I I I I I i i i" r
1961 1966 1971 1976 1981 1986
Source: Calculated from FAO data.

Figure 1. Cereal production per capital by developing country region.



sub-Saharan Africa as a whole, per
capital consumption of imported
cereals grew at an annual rate of
3.9% from 1961-65 to 1984-88.
Meanwhile, per capital consumption
of traditional coarse grains, roots,
and tubers declined (Figure 2).

These trends are viewed with
alarm by many policy makers,
because increasing dependence on
imported cereals not only uses
valuable foreign exchange, but also
is perceived as a threat to national
food security. A few wealthier
countries in Africa can afford
commercial imports of cereals, but
many others rely heavily on food
aid. Despite obvious short-term
benefits, food aid is undesirable in
the long run whenever it depresses
domestic food production incen-
tives. The danger of depending
heavily on food aid is heightened by
the fact that major donors such as
the USA and the European Com-
munity recently have pledged to
restrain the overproduction of
cereal crops, especially wheat,
which has made large amounts of
food aid possible.

Hardly anyone questions the need
to revitalize food production in
Africa as a first step toward stimu-
lating economic growth. But how to
increase food production is not
always obvious, particularly in view
of the large and widening gap be-
tween the kinds of food that can be
produced and the kinds of food that
consumers prefer. Demand for
wheat and rice is increasing
throughout much of the continent,
yet climatic and economic factors
limit the production of these two

crops. At the same time, consump-
tion of millet, sorghum, roots, and
tubers is declining, even though
these crops are suited to local
production conditions.

That leaves maize. Maize produc-
tion and consumption patterns vary
greatly throughout sub-Saharan
Africa. In large parts of eastern and
southern Africa, maize is the
principal staple food, produced and
consumed by most farming house-
holds. While relatively less impor-
tant in western and central Africa,
maize still provides a major source
of calories, especially in parts of
Nigeria, Ghana, Benin, and Cote
d'Ivoire. But whatever its present
importance, maize clearly has
enormous potential, because

improved technologies offer the pos-
sibility of greatly increasing yields
and thus production.

This report examines the status of
maize in sub-Saharan Africa. To
provide a perspective on the per-
formance of maize in Africa, the
report begins with a review of
global and regional trends in the
production of the world's major
cereal crops. Subsequent sections
present a descriptive overview of
the maize economy of Africa,
focusing on production systems and
technologies, utilization patterns,
institutions, and policies. A discus-
sion of critical production con-
straints is followed by an assess-
ment of future prospects for maize
in Africa and priorities for



1966 1971 1976 1981 1986

Source: Calculated from FAO data.

Figure 2. Consumption of millet and sorghum, roots and tubers, and
imported cereals in sub-Saharan Africa, 1961-88.

Trends in
Cereal Production

Cereal Production
in the Developing World
Of the major cereal grains grown in
developing countries, rice is the
most important in terms of quantity
produced, followed by wheat, maize,
and finally millet and sorghum
(Table 1).3 These rankings have not
changed significantly over the past
three decades, although the impor-
tance of millet and sorghum has
diminished as the area planted to
these crops has declined.

When historical production data for
each of the four major cereals are

just as substantial. Yield increases
in wheat and rice were concentrated
in the decade after the mid-1960s,
whereas those in maize began only
in the early 1970s and continued
well into the 1980s.

Maize Production
in the Developing World
Maize yields in the developing
world have experienced sustained
growth during the past 30 years,
but these gains have not been
distributed uniformly. At least three
factors explain the pronounced
geographical variability in maize
yield gains. First, much of the
world's maize is grown in marginal
environments characterized by un-
I;-l IA 4 --mi f .ll i l-, C il f ti -ilitx

a wide range of agroclimatic condi-
tions that improved germplasm or
cropping practices cannot always be
diffused rapidly or far. (Wheat and
rice, on the other hand, frequently
are grown in extensive, relatively
homogeneous agroclimatic zones-
often irrigated-where new tech-
nologies can be disseminated more
easily.) Third, hybrid maize seed, a
principal source of yield gains in in-
dustrialized countries, generally re-
quires specialized production and
distribution facilities which are
lacking in many developing coun-

In view of these considerations, it is
not surprising that maize produc-
tion trends have varied geographi-

decomposed into area planted and Second, maize is grown under s call. Progress has been noticeably
yield, a better perspective emerges
on the sources of production gains
in the developing world (Table 2).
From 1961-65 to 1984-88, average
From 1961-65 to 19 rage Table 1. Area, yield, and production of major cereal crops in develop-
annual production growth rates
were highest for wheat, followed by ing countries, 1984-88
maize, rice, and finally millet and Bh frion
sorghum. These production growth .- rea__
rates were largely attributable to _- -
yield gains, which were particularly Rice 140 3.20 447
strong for wheat (3.7%), maize Wheat 100 2.13 212
(2.6%), and rice (2.3%). These data Maize 80 2.20 176
suggest that the conventional Mie 0.92 69
wisdom that the Green Revolution Millet/sorghum 75 0.92 69
affected primarily wheat and rice is Calculated from FA data.
Source: Calculated from FAO data.
not quite correct. Over the past
three decades, although yield gains
in wheat more than equalled those
in the other cereals, yield gains in Table 2. Growth rates in area, yield, and production of major cereal
maize kept pace with those crops in developing countries, 1961-65 to 1984-88
achieved in rice. The only difference -.-, -... -...
was that the gains in maize began E-.. rate (%)
later and proceeded more slowly,-
which made them less conspicuous. 77 .. -.oduction
However, the gains in maize were
sustained over a longer period, so Wheat 1.2 3.7 4.9
the cumulative effects have been Maize 1.1 2.6 3.8
Rice 0.8 2.3 3.0
Millet/sorghum -0.3 1.5 1.2

Source: Calculated from FAO data.
3 By convention, data for millet and
sorghum are combined.



slower in countries lacking strong
research and extension systems, or
lacking well-developed networks for
distributing inputs.

The following summary of regional
trends in maize production in the
developing world shows which
areas have benefited most from
new technologies and which areas
remain relatively unaffected (Table
3, Figures 3 and 4).

Latin America-In 1984-88, 27
million hectares were planted to
maize in Latin America; more than
three-quarters of this area was in
Argentina, Brazil, and Mexico. Use
of improved germplasm is quite
high in Latin America compared to
other regions of the developing
world: in 1985/86, 50% of the maize
area was sown to hybrids, 10% to
improved open-pollinated varieties,
and 40% to local varieties
(CIMMYT 1987). Combined with a
modest expansion in maize area, in-
creased yields have helped fuel an
average annual growth rate in total
maize production of 3%.

South Asia, East Asia, and
Southeast Asia-In 1984-88, 35
million hectares were planted to
maize in this region. Area ex-
panded between 1961-65 and 1984-
88, resulting in a cumulative
increase of nearly 9 million hec-
tares. Use of improved germplasm
is high: in 1985/86, 42% of total
maize area was planted to hybrids,
14% to improved open-pollinated
varieties, and 44% to local varieties
(CIMMYT 1987). Yield increases
have been impressive, mostly
because of large yield gains ob-
tained in China through greater
use of nitrogenous fertilizer, im-
provements in irrigation infrastruc-
ture, and the introduction of

hybrids. The years from 1973-77 to West Asia and North Africa-
1984-88 saw regional maize produc- Maize area in this region is com-
tion accelerate sharply, rising at paratively small and since the early
about 4.2% per year. Average yields 1960s has remained around 2.4
currently stand at 2.8 t/ha, among million hectares, much of it irri-
the highest in the developing world. gated. Use of improved germplasm

Table 3. Growth in maize area, yield, and production in developing
countries by region, 1961-65 to 1984-88

.1 .M


South Asia, East Asia,
and Southeast Asia

Latin America

West Asia and North Africa

Sub-Saharan Africa

1.3 3.5 4.8

Source: Calculated from FAO data.

(million ha)
40 --

Latin South, East, West Asia Sub-Saharan
America and Southeast and Africa
Asia North Africa

Source: CIMMYT survey data.

Figure 3. Use of maize germplasm by developing country region,

is modest: 15% of the region's total
maize area is planted to hybrids,
32% to improved varieties, and 53%
to local varieties (CIMMYT 1987).
Yields in West Asia and North
Africa are high compared to those
in the rest of the developing world,
mainly because yields are high in
Egypt, which accounts for half of
the region's production. Since
1961-65, maize yields have grown
at an annual rate of about 2.4%.
Because area planted to the crop
has increased very little, the
growth rate of total maize produc-
tion only slightly exceeded growth
in yields during the past two

Sub-Saharan Africa-In 1984-88,
nearly 15 million hectares were
planted to maize in sub-Saharan
Africa. This represented about 19%
of developing world maize area but
accounted for only 10% of the
developing world maize crop. Maize

yields in Africa barely exceed
1.2 t/ha, by far the lowest in the
world, and since 1961-65 have
grown at an average rate of just
0.8% per year from a low initial
base. Largely because of the low
rate at which yields increased, pro-
duction grew at a modest rate
between 1961-65 and 1984-88.
Significantly, production growth
rates have declined considerably
during the most recent period for
which data are available, increas-
ing at a rate of just over 2% per
year from 1973-77 to 1984-88.

Several factors contribute to low
maize yields in Africa. Much maize
in Africa is grown at low density in
mixed stands with one or more as-
sociated crops, including cassava,
sorghum, pumpkin, squash,
cowpea, groundnut, yam, and sweet
potato. Mixed cropping lowers
maize yields, but it helps farmers
increase the overall productivity of

Yield (t/ha)

the resources invested in agriculture
and reduces losses if any one crop
fails. While mixed cropping is the
primary reason for low maize yields,
other factors (discussed later in this
report) also come into play. For
example, land is still relatively
abundant in sub-Saharan Africa
compared to other regions of the
developing world; farmers take
advantage of easy access to land by
farming extensively rather than in-
tensively, using low levels of pur-
chased inputs, especially fertilizer.
Also, many African farmers continue
to plant unimproved local varieties.
In 1985/86, only 16% of the maize
area in Africa was planted with
hybrids, virtually all of it in eastern
and southern Africa, and an addi-
tional 15% of maize area was planted
with improved open-pollinated
materials (CIMMYT 1987).

Maize Production
in Regions of Africa
Maize production varies greatly
among the major subregions of
Africa (Table 4 and Figure 5).
Among sub-Saharan Africa's four
major regions (see Annex 2, p. 69, for
a list of countries in each region),
southern Africa is by far the great-
est producer and consumer of maize.
The region's extensive maize area
surpassed 6.4 million hectares in
1984-88 after growing at an average
annual rate of 1.4% since 1961-65.
Although yields are modest, total
production has grown at an average
annual rate of 2.8% since 1961-65
and currently stands at just under
8 million tons-96% of the region's
maize requirement.

0 8gP


1961 1966 1971 1976 1981 19
Source: Calculated from FAO data.

Figure 4. Evolution of maize yields by developing country region,

SIn 1984-88, farmers in eastern
Africa planted maize on about 4.6
Million hectares. Although still low
86 by global standards, eastern African

_Ij~l~ ~

maize yields are the highest in
Africa (reflecting cooler growing
conditions), averaging 1.6 t/ha. The
growth in yields from 1961-65 to
1984-88 was largely the result of
adoption of hybrids and increased
use of fertilizer. Expanding area
and rising yields have helped raise
total maize production in eastern
Africa at a rate of 2.7% per year
since 1961-65. Despite considerable
year-to-year variability in produc-
tion, the region is nearly self-
sufficient in maize.

From 1961-65, area planted to
maize in western and central
Africa grew slowly, reaching
almost 5 million hectares in
1983-87. Average yields now barely
exceed 1 t/ha, after increasing at an
average annual rate of 1% since
1961-65. Total production has
grown to just under 5 million tons,
a level of production that nearly
satisfies regional demand.

Area planted to maize in the Sahel
has expanded over the years to
reach its current level of approxi-
mately 0.5 million hectares. Much
of the growth in area occurred
when early maturing, input respon-
sive maize varieties moved into
zones traditionally occupied by
sorghum and millet. Maize produc-
tion remains modest, currently to-
talling just above 0.5 million tons.
Despite growth in production of
3.2% per year from 1961-65 to
1984-88, self-sufficiency in maize
actually declined because of rapid
population growth combined with
an increase in consumption per
capital. During the same period,
maize imports rose from a very low
base at an average of 8% each year,
equivalent to an average annual
rate of 5.4% per capital.

Although some of these statistics
present a less than optimistic view
of African maize production, they
do not tell the whole story. They
only hint at the diversity and com-
plexity of the maize economy in
sub-Saharan Africa, which is
shaped by countless biotic and
abiotic factors, ranging from

cultural preferences and historical
influences to agronomic conditions
and institutional constraints. The
next part of this report gives a
more detailed account of the varied
and often complex conditions under
which maize is produced, used,
marketed, and traded.

Table 4. Sources of growth in maize production in Africa, by region,
1961-65 to 1984-88

Regfie*.. A... .-.

Southern Africa 1.4



Eastern Africa

Western and central Africa



Source: Calculated from FAO data.

Production (million t)


1966 1971 1976 1981 1986

Source: Calculated from FAO data.

Figure 5. Evolution of African maize production by region, 1961-88.

The Maize
Economy of Africa

Production Zones

Several major maize production
zones can be distinguished in sub-
Saharan Africa. The chief charac-
teristics of each zone are described
in the following paragraphs.

Eastern and Southern
Africa-Maize is the primary
staple food of most people in east-
ern and southern Africa. Presently
maize has little competition from
other staples, except in a few agro-
ecological zones. CIMMYT has
identified eight distinct maize
production environments in sub-
Saharan Africa, based on agrocli-
matic factors and grain maturity
characteristics (see "Maize Produc-
tion Zones in Sub-Saharan Africa,"
p. 8). In eastern and southern
Africa, these eight maize environ-
ments can be grouped into four
basic agroecological zones: lowland
tropical (<900 meters above sea
level), wet subtropical midaltitude
(900-1,500 masl), dry subtropical
midaltitude (900-1,500 masl), and
highland (>1,500 masl).

Lowland tropical production
zones cover about 18% of the maize
area in eastern and southern
Africa, including the coastal areas
of Kenya, Mozambique, Somalia,
and Tanzania, as well as parts of
Malawi. Rainfall patterns vary;
some lowland tropical areas are
characterized by a distinct rainy
season, whereas in other areas
rainfall is distributed bimodally.
Soils range from sandy loams to
heavier clays. Maize may be mono-
cropped with grain legumes or
intercropped with sesame, cassava,
cowpea, pigeon peas, tomatoes, or
rice. Population growth and the
increasing scarcity of land have
practically eliminated the long
fallows that traditionally were part

of the shifting cultivation system,
and now continuous cropping with
limited rotation is common in most
lowland tropical zones.

Subtropical midaltitude zones
can be classified as wet (>1,000 mm
rainfall annually) and dry (<1,000
mm rainfall annually). Wet sub-
tropical zones cover 49% of the
area planted to maize in eastern
and southern Africa, including
parts of Angola, Burundi, Kenya,
Malawi, Mozambique, Rwanda,
Swaziland, Tanzania, Uganda,
Zambia, and Zimbabwe. Rainfall
varies considerably at different
altitudes and may be either uni-
modal or bimodal; in the latter case,
two maize crops can be grown. Soils
range from deep fertile soils along
river bottoms and in lake basins
(some of which may be prone to
waterlogging) to better drained and
more easily worked upland soils.
Maize may be monocropped, par-
ticularly by commercial producers,
but more commonly it is inter-
cropped with beans, cowpeas,
groundnuts, pumpkins, or pigeon
peas. Since livestock are a
significant part of the farming
system in this zone, animal manure
may be an important source of
nutrients for crops.

Dry subtropical zones constitute
approximately 16% of the area
planted to maize in eastern and
southern Africa and are located
chiefly in Ethiopia, Kenya, Tanza-
nia, Uganda, and Zimbabwe.
Rainfall is unreliable and inade-
quate. Soils include sandy, sandy
loam, alluvial, and volcanic types.
Maize is sometimes monocropped,
but more frequently it is associated
with beans, groundnuts, cassava,
cowpeas, or pigeon peas. Planting
dates are usually staggered to
reduce the risk of losing crops to
drought early in the growing
season. The unreliability of rainfall
discourages farmers in most dry
subtropical areas from using inor-
ganic fertilizer, so inadequate soil
fertility is a widespread problem.

Approximately 16% of the area
planted to maize in eastern and
southern Africa is located in
highland zones in Burundi,
Ethiopia, Kenya, Lesotho, Rwanda,
Tanzania, and Uganda. Highland
zones are characterized by ade-
quate to excessive rainfall, cool
temperatures, and long growing
seasons. Soils are generally deep
and well drained, with a high
content of organic matter. Maize is
monocropped or intercropped with
squash, beans, potatoes, peas, rape
seed, or even coffee. Depending on
the population density, land may be
continuously cropped or fallowed
after two to three years of cultiva-
tion. The practices that farmers use
to maintain soil fertility reflect
cropping patterns. For example,
inorganic fertilizers are widely used
in continuously cultivated areas.

Western and Central Africa-
Five major maize production zones
can be distinguished in western
and central Africa: the humid low-
land forest, the semideciduous
lowland forest, the derived and
southern Guinea savannas, the
northern Guinea savanna, and the
midaltitude zone (see "Maize
Production Zones in Sub-Saharan
Africa," p. 8). Cropping patterns are
highly diversified and vary from
zone to zone (Figure 6).

In the humid lowland forest,
maize is a minor crop generally
planted with the first rains at
fertile spots in the field. The main
crop is most often cassava, but in
some areas, such as southern
Cameroon, a groundnut/cassava
mixture is grown. Plantain bananas
are often interplanted as well,
producing a complex system of
short-cycle groundnutss, maize),
medium-cycle (cassava), and long-
cycle (plantain) crops. Monocropped
maize may be grown as an off-
season crop in low areas. A chief
reason for the relative unimpor-
tance of maize in the humid low-
land forest is that the crop is

(continued on p. 11)

Maize Production Zones in Sub-Saharan Africa

Maize is grown virtually throughout
sub-Saharan Africa in a range of
agroecological and economic envi-
ronments. It is the main staple in
18 countries and an important food
crop in another 13. Although
maize's ability to tolerate diverse
growing conditions enhances the
crop's importance, this same charac-
teristic complicates the task of
classifying production environments
and, eventually, of organizing and
conducting research. To better
determine breeding priorities for
the varied environments of sub-
Saharan Africa, work on maize

zoning has been done by CIMMYT
and IITA, among other organiza-

CIMMYT mega-environments--
CIMMYT scientists, in cooperation
with colleagues in national re-
search programs, have begun work
on a mapping system based on
so-called mega-environments to aid
in defining specific breeding objec-
tives. Mega-environments are
production zones, not necessarily
geographically contiguous,
delineated by ecological conditions
(temperature, rainfall, soils); crop

characteristics (maturity cycle,
grain color, grain texture); biotic
and abiotic constraints (pests and
diseases); and socioeconomic factors
(production systems, cropping
patterns, consumer preferences).
While work on defining mega,
environments is still preliminary
because reliable microlevel data are
scarce, eight mega-environmento
have tentatively been identified for
Africa. Information on these mega-
environments, including their ap-
proximate size and the incidence
within each of biotic and abiotic
stresses, appears in Table 1 and
Map 1.

Table 1. CIMMYT maize mega-environments for sub-Saharan Africa


Early and



88% white, 69% flint,
12% yellow 31% dent

Major insect

Stem borers

Intermediate 88% white, 58% flint, Stem borers,
12% yellow 42% dent armyworms,

Late and

Subtropical/ Early and
mid-altitude extra-early

92% white, 54% flint,
8% yellow 46% dent

Stem borers

100% white, 54% flint, Stem borers,
0% yellow 46% dent storage pests

Subtropical/ Intermediate 100% white, 0% flint, Stem borers,
mid-altitude 0% yellow 100% dent rooteutwori~
storage gqpsa

Subtropical/ Late and
mid-altitude extra-late

92% white, 48% ffihia St .* bocere.
8% yellow 52% d&ast W-mite.
storage peate- -

Highland/ Early and 83% white, 82% fint-: ItuaBtctworni.
transitional intermediate 16% yellow 18% detf tri.

Major diseases
of maize

Streak virus,
southern leaf blight,
stalk rot, ear rot

Streak virus,
southern leaf blight,
ear rot, stalk rot,
southern rust

Streak virus,
southern rust,
northern/southern leaf blight,
ear rot, stalk rot

Northern leafblight
streak virus, ear rot

Streak virus, ear rot,
northern leaf blght
-commo rult

Nowthern lea blight,
eat"** raw
streak virus, ear ro -
commonn a t U.

Northern Sbli.iJ
ammwsxa r

Highland/ Late and
transitional extra-late

95% white, 41% flin Stemi W
2% yellow. 554 dear I "-
__*^^te 'Aki-^^^^

Northern ie blight
commknrra --
ear rust, streak iua.
suadkre ""

Source CIMMYT Maize Program iprelminary data. .





Estimated area
(million ha)


SLowland tropics


SHighland tropics





Map 1. CIMMYT maize mega-environments for sub-Saharan Africa.

IITA agroecological zones-
Within its mandate area, IITA dis-
tinguishes three broad agroecologi-
cal zones: the forest, the moist (or
Guinea) savanna, and the forest/
savanna transition zone (IITA
1988). This classification is based
primarily on differences in vegeta-
tion-strikingly evident as one
moves northward from the coast-
which in turn reflect differences in
soil conditions, climate, and human
activity (Map 2).

The three broad agroecological
zones are subdivided into five
maize production zones (Table 2):

Humid forest: Located in the
southern part of the forest zone,
the humid forest production
zone experiences more than
seven humid months (i.e.,
months in which precipitation
exceeds potential evapotranspi-
ration). Rainfall is distributed
unimodally or bimodally. Soils
are acidic.

Maize production zones

Humid forest

I I Semideciduous forest and forest-savanna

S Derived/southern and northern Guinea

S Midaltitude

Source: Adapted from IITA (1988).

Map 2. Maize production zones, western and central Africa

Table 2. IITA Maize-growing zones for western coastal and central Africa


Humid forest
(Southern forest)

(Northern forest)

Derived and
southern Guinea

Northern Guinea

(Savanna vegetation)







Rainfall Soil
distribution types



Climax Crops planted
vegetation with maize

Ultisol, Evergreen
oxisol forest


UnimodalM Alfaol,
bimodal ultisa






Wosannd yam
savanna yafl

AlftsP~ Shub --S~n

Other major crops

Oil palm,

oil palm



mille .-
~ararn~ es

ltisdL mhi mL cocoi
ulti"L. waidb ".i~


Source: IITA. .... .... ........... ... ... ...
.tii-eea I<. -";;--:,ay -.'>K; .-

& V

* Semideciduous forest:
Located in the northern part of
the forest zone, the semidecidu-
ous forest production sone expe-
riences six to seven humid
months, with rainfall distrib-
uted bimodally in most areas.
Two growing seasons are distin-
guished. Soils are generally not

* Derived and southern
Guinea savannas:
Encompassing the forest/
savanna transition zone and
the southern part of the Guinea
savanna, this production zone
experiences five to six humid
months. Vegetation in much of
the transition zone is derived
savanna, because more forest
has been cleared and fallow
periods are becoming shorter.
Rainfall may be unimodal or

* Northern Guinea savanna:
The northern Guinea savanna
experiences four to five humid
months. Insolation during the
brief maize-growing season is
high, and the long dry season
limits the incidence of pests and

* Midaltitude zone: This "zone"
actually consists of isolated
areas ranging from 1,500 to
1,800 masl. High insolation
and cool temperatures make
midaltitude areas possible high-
potential locations for maize

harvested during the rainy season,
which leads to storage problems
and reduces the attractiveness of
maize compared to other food crops.
Partly because of the difficulty of
storing dried grain, most maize
produced in the humid lowland for-
est is consumed green.

In the semideciduous lowland
forest, which covers a significant
part of the Cocoa Belt of West
Africa, maize is second in impor-
tance after cassava, with which it is
often associated. The two crops are
planted mainly during the first
rainy season; after maize is har-
vested, cassava may occupy the
land for up to two additional years.
Groundnuts are grown with low-
density maize in some areas,
including southeastern C6te
d'Ivoire and central Ghana. Second-
season maize is often grown in the
semideciduous forest zone, but its
success depends on the incidence of
certain pests and diseases (e.g.,

stem borers and maize streak
virus), as well as on the amount
and distribution of rainfall during
the second rainy season.

The derived and southern
Guinea savannas are the most
important maize-growing areas in
western and central Africa. Crop-
ping patterns vary widely, but
typically maize is planted first in
the rotation. After one and a half to
two months, sorghum is often relay
planted into the maize. This prac-
tice takes advantage of the full
rainy season, as photoperiod-
sensitive sorghum completes its
cycle on residual moisture after the
rains have stopped. In many areas
of the derived and southern Guinea
savannas, farmers interplant
groundnuts with low-density maize.
Sorghum is also relayed into this
mixture. In wetter areas, mounds
for yams are prepared in the
sorghum crop following the harvest
of maize or groundnuts; the yams

-c .- ---~ -. -- ---~


Derived and
Guinea savanna



(same or other

(same or other combinations)

M (same or other combinations)

Note: 4as Yam emerges with first rains after maize and sorghum are harvested.

Figure 6. Typical mixed cropping patterns involving maize in major
agroecological zones, western and central Africa.

are planted after the rains stop.
The early rains are often erratic,
and where the rainfall regime is
unimodal, staggered maize planting
is common.

Maize in the northern Guinea
savanna is grown in mixed stands
along with sorghum, groundnuts,
cowpeas, cotton, and yams, al-
though monocropped maize is
becoming more prevalent. Tradi-
tionally, maize was a minor crop
grown near family compounds,
where it benefited from regular ap-
plication of household refuse and
organic manure. With the advent of
chemical fertilizers, maize has
acquired importance as a field crop,
particularly in areas where soils
are good and fertilizer supply is
assured. The minor role played
until recently by maize in the
northern Guinea savanna is diffi-
cult to explain, since temperature
and rainfall in this area are favor-
able for maize (Kassam 1976).
However, this situation seems to be
changing, as maize has moved
rapidly into the zone in recent
years (see "Expansion of Maize into
the Northern Guinea Savanna,"

Maize is a major crop in the mid-
altitude zone of central Africa
(e.g., parts of Cameroon), where it
is generally grown with grain leg-
umes groundnutss, beans, cowpeas)
or tubers cocoyamm). Staggered
maize planting is common, so ear-
lier plantings mature during the
rainy season and later plantings
mature in the dry season. Late-
planted maize frequently suffers
from maize streak virus and de-
creased soil fertility caused by
nutrient leaching. Soils in this zone
are often acidic, and nutrient defi-
ciencies (particularly phosphorus)
are common.

Production Technologies
Because of the diversity of agrocli-
matic conditions, production
systems, and producer groups, any

summary of maize production tech-
nologies in Africa risks oversimpli-
fication. Without attempting to be
exhaustive, the following discussion
provides a general idea of how
maize is produced across sub-
Saharan Africa.

Eastern and Southern Africa-
Three basic groups of producers can
be distinguished in this region:
1) small-scale hand-hoe cultivators,
2) medium-scale cultivators who
use draft animals, and 3) large-
scale commercial farmers whose op-
erations are heavily mechanized.

Small-scale hand-hoe
cultivators. Approximately 45% of
the total area planted to maize in
eastern and southern Africa is
cultivated by small-scale farmers
(also known as smallholders) who
rely primarily on family labor to
grow maize on 1-3 ha of land held
under traditional tenure arrange-
ments. Cultivation with hand-hoes
is often, though not always, associ-
ated with heavier soils. Land
preparation generally begins before
the wet season to take full advan-
tage of the rains but is rarely
completed on time, as the dry soil is
difficult to work by hand. Maize is
usually the first major crop
planted, except where high-value
cash crops such as cotton or tobacco
take precedence. Most maize is
sown shortly after the start of the
rains, although farmers may make
three or more plantings to accom-
modate limited draft power re-
sources, to reduce the risk of
drought losses, and to ensure an
extended food supply. Planting
method varies. In some areas maize
seed is planted in randomly ar-
ranged hills, although more com-
monly it is planted in rows. Seeding
rates depend on soil fertility levels,
plant spacing, and expected germi-
nation rates. Maize may be mono-
cropped or intercropped with other
food crops, especially beans, pump-
kins, cowpeas, pigeon peas, ground-
nuts, sweet potatoes, and cassava.

Many small-scale farmers choose to
plant unimproved local materials
because the grain quality of im-
proved materials is unacceptable,
or because improved materials offer
little yield advantage under the
limited level of inputs and manage-
ment that smallholders can pro-
vide. On the other hand, in areas
where fertilizer and other inputs
are readily available, interest in
early maturing varieties has been
strong, since these materials give
farmers greater flexibility to
stagger maize planting. Kenya,
Zambia, and Zimbabwe have been
particularly successful in delivering
improved maize varieties to a
large percentage of small-scale
farmers, including some hand-hoe

Fertilizer use varies greatly among
hand-hoe cultivators, depending on
soil fertility levels and on the
availability of organic or inorganic
fertilizers. Many farmers rely on
crop rotation strategies to maintain
soil fertility, such as intercropping
or relay cropping maize with
legumes. Animal manure can be a
significant source of nutrients in
areas where livestock are part of
the farming system, although most
hand-hoe cultivators have no
access to manure. In areas where
inorganic fertilizers are available,
modest amounts of fertilizer may be
used. However, fertilizer is fre-
quently applied late and/or in low
doses, so its effect is limited.

Weeds are a serious problem
seldom controlled by hand hoeing.
One or two weedings are normal,
although the first weeding is fre-
quently late, and subsequent
weedings are sometimes omitted.
As the season progresses, farmers
must often compromise between
planting more land and weeding
maize that has already emerged.
The use of herbicides is rare.

(continued on p. 14)

Expansion of Maize into the Northern Guinea Savanna

The high amount of radiation and
low night temperatures character-
istic of the northern Guinea sa-
vanna make this zone the most fa-
vorable ecology for maize in west-
ern and central Africa, provided
adverse soil conditions do not limit
production. Yet until the mid-
1970s, maize in the northern
Guinea savanna was a minor crop
valued primarily because, if har-
vested green, it was the earliest
food available after the "hungry pe-
riod." Families grew just enough
maize to sustain themselves until
the main cereal crops, sorghum
and millet, were harvested.

Today maize production appears to
be increasing substantially in this
zone. The effect is most striking in
Nigeria, where a recent study
shows that, in northern Kaduna
and southern Katsina States, maize
is the most important food crop in
over 50% of 15 randomly selected
villages and the top cash crop in
60%. The increased importance of
maize is also evident, though to a
lesser extent, in Bauchi and south-
eastern Sokoto States. Almost all
the maize grown in these areas
(with the exception of Sokoto)
appears to be improved varieties.

In the past, maize grain was
consumed mainly in southern
Nigeria. Although maize could be
produced in the north, demand
there was minimal, and poor
transportation made it uneconom-
ical to "export" maize from the
north to consumption centers in
the south. Beginning in the mid-
1970s, however, several factors
helped raise maize production in
the north. Oil revenues were used
to improve roads between the
urban south and rural areas of the
north, providing northern farmers
with better access to southern

markets. At the same time, subsi-
dized fertilizer and improved early
maturing maize varieties well
adapted to the ecology were made
available through World Bank-
assisted agricultural development
programs. Higher yields rather
than higher prices made maize
more profitable than competing
crops; the domestic price of maize
relative to competing crops such as
millet, sorghum, and groundnuts
either remained constant or de-
clined, although it remained high
relative to the world price of maize
(converted at official exchange

In addition to replacing other crops
on land already under cultivation,
maize also began to be cultivated on
land not previously used for crop
production. This expansion in area
was made possible by the adoption
of animal traction and by increased
use of fertilizer, which permitted
the elimination of fallow periods in
many areas. Also, fertilizer subsi-
dies favored maize because maize is
more responsive to fertilizer than
the sorghum and millet that it re-
placed. Eventually the greater
availability of maize in the north
led to its adoption as a staple food
in that area, further reinforcing its
attractiveness to small-scale

The key question now is whether or
not expanded maize production is
sustainable. Three issues in par-
ticular will have to be resolved
before maize establishes itself as a
crop with long-term prospects in
the northern Guinea savanna.

The first issue relates to the future
cost and availability of inorganic
fertilizer. Maize production in the
northern Guinea savanna relies
heavily on added nutrients, espe-
cially nitrogen. In most countries of

western and central Africa, particu-
larly Nigeria, fertilizer has been
highly subsidized; when available
at official prices, it has generally
been cheap, although the quantities
available through official distribu-
tion channels have often varied. A
number of countries in the region
are now committed to removing
subsidies and privatizing fertilizer
distribution, which could lead to
improved availability, but at
substantially higher prices. The
likely impact of these policy
changes on future maize production
requires further investigation.

A second issue is the future profita-
bility of maize relative to export
crops such as cotton and ground-
nuts. In the past, overvalued
exchange rates reduced the profita-
bility of export crops. Now many
countries in the region have drasti-
cally devalued their currencies,
which would be expected to lead to
a resurgence in the production of
export crops, assuming domestic
producer prices reflect inter-
national prices. Whether increased
production of export crops would
occur at the expense of maize
remains unclear.

A third issue concerns the impact of
continuous maize cultivation on
savanna soils. Farmers once relied
on a combination of fallowing,
manuring, and crop rotation to
maintain soil organic matter and
preserve soil fertility. As continu-
ous cropping of maize for cash
receives greater emphasis, tradi-
tional soil conservation practices
are being replaced by increased use
of fertilizer. It will be important to
investigate whether this strategy is
sustainable over the long run.

Most smallholders harvest maize
when the plants are fully dried.
Either the cobs are picked or the
entire plant is cut and stocked for
later stripping. Much of the harvest
is stored on the farm. Most often
cobs are kept in cribs, either
outdoors or indoors over a fireplace
where the smoke helps control
insects. In some areas, raised clay
or brick outdoor granaries or
underground storage pits are used.
In other areas, maize may be
shelled and stored indoors in sacks,
earthen jars, metal bins, or other
containers. Given the long dry
season in much of eastern and
southern Africa, these traditional
storage methods perform well,
providing good aeration and offer-
ing some protection from insects
and rodents. The use of insecticides
to control storage pests, while rare,
is increasing.

Medium-scale cultivators.
Medium-scale cultivators who use
draft animals (usually oxen) to per-
form agricultural operations farm
approximately 50% of the total area
planted to maize in eastern and
southern Africa. Most of these
farmers rely on family labor and
grow maize on 1-10 ha of land held
under traditional arrangements.
Animal traction is often, though not
always, associated with lighter
soils. Land preparation generally
begins with the onset of the rains,
although in some areas farmers
plow their land after the previous
harvest, before the soil hardens.
Most of these farmers use mold-
board plows drawn behind oxen.
Generally only a single plowing is
done, although occasionally it is
supplemented by a harrowing
before planting.

Planting method varies depending
on the area to be planted, soil
moisture, and availability of labor.
In many dry areas, seed is broad-
cast directly onto the soil and then
plowed in, a method especially
suitable for planting a large area

rapidly. Dibbling seed behind the
plow is another method farmers use
to plant quickly while soil moisture
conditions are favorable. Seed is
dibbled in every other furrow and
covered by a return pass of the
plow. Hoe planting behind the plow
is favored in some places as a
means of ensuring uniform stands,
although this method requires con-
siderable labor. Finally, in parts of
southern Africa, drilling with an
ox-drawn planter has become
increasingly popular in recent

Medium-scale cultivators plant a
range of maize materials. As
improved varieties become avail-
able, farmers have begun to de-
mand germplasm with specific
characteristics, especially drought
avoidance or drought tolerance,
higher yield potential, and respon-
siveness to fertilizer. As in the case
of hand-hoe cultivators, interest in
early maturing varieties has been
strong because they provide greater
flexibility in management. Farmers
in some areas also value rapidly
maturing varieties because they are
ready for consumption earlier in
the season.

Both organic and inorganic fertiliz-
ers are used to maintain soil
fertility. Manure, when available,
tends to be of variable quality; since
it is bulky and expensive to trans-
port and store, it is applied to only
a small percentage of fields in most
years. The use of inorganic fertiliz-
ers has become more common since
many governments improved
fertilizer delivery to small- and
medium-scale farmers. Inorganic
fertilizer, often nitrogen alone, is
usually applied basally; less fre-
quently, it may also be applied as
an early postemergence dressing
and/or as a top dressing during

Weeds are controlled either by
hand hoeing or, less commonly,
with ox-drawn cultivators. Row

planting facilitates mechanical cul-
tivation, which is generally done
several weeks after emergence
while the maize plants are still
small. Frequently a late ridging is
also done to control weeds and
reduce lodging. Many farmers who
rent oxen to prepare land do not
have access to animals later in the
season and rely entirely on manual
labor for weeding. Chemical herbi-
cides are not commonly used by
medium-scale farmers, mainly
because herbicides and application
equipment are unavailable and
farmers do not know how to use
them, or because the chemicals
damage the intercrop. Harvesting
and storage practices resemble
those used by small-scale hand-hoe

Large-scale commercial
farmers. Large-scale commercial
producers (known as estate farmers
in some countries) farm approxi-
mately 5% of the total area planted
to maize in eastern and southern
Africa. Although the definition of
"large-scale farmer" varies from
country to country, these farmers
usually plant at least 50 ha of
maize and often as much as 100 ha
or more. Large-scale commercial
farmers typically live on their land,
which they hold under registered
titles. Many also rent land from
neighbors who do not farm. Much
of the land cultivated by these
farmers is located in the high po-
tential zones of Kenya, Malawi,
Zambia, and Zimbabwe. Commer-
cial farmers generally produce
maize as a cash crop, although in
some instances maize is grown to
feed workers (e.g., on the tobacco
estates of Malawi).

Land is prepared with tractors. An
early plowing before the onset of
the rains is followed by one, two,
and in some cases even three
harrowings. Recently this pattern
has begun to change. The high cost
of operating machinery, and the
difficulty of obtaining spare parts in

countries lacking the foreign
exchange to import them, have led
large-scale commercial farmers to
experiment with reduced tillage
and zero-tillage technologies that
do not require such intensive use of

Commercial farmers plant maize in
rows, using either hand-operated
seed drills or tractor-drawn
mechanical planters. Early plant-
ing is associated with higher yields
when rainfall is normal, although
maize planted early is subject to
greater risk in drought years. If a
dry spell occurs just after planting,
some farmers can irrigate to help
establish the crop. Large-scale
farmers tend to grow hybrids, since
these materials are well suited to
favorable production environments
and respond well to high manage-
ment levels. Hybrid seed is pro-
duced by private companies and in
some cases by public sector organi-
zations. Certified seed treated with
fungicide and sometimes pesticide
is usually sold through producers'

Almost all large-scale commercial
farmers apply inorganic fertilizers
to maize. Application rates vary
depending on soil conditions,
averaging around 150 kg/ha
nitrogen (N), 60 kg/ha phosphorus
(P20,), and 30 kg/ha potassium
(IK20) throughout the region
(Low and Waddington 1989;
Anandajayasekeram and Ransom
1989). In most cases, all of the
phosphorus and potassium and
about one-third of the nitrogen are
applied basally, with the rest of the
nitrogen top dressed or side dressed
four to six weeks after the crop

Weeds are controlled by mechanical
cultivation and/or with herbicides.
Herbicides are applied with tractor-
mounted sprayers or by air. Where
herbicides are unusually expensive,

many farmers reduce costs by
combining band spraying on the
crop row with tillage between rows.

Large-scale commercial farmers
harvest maize with combine har-
vesters or by hand. Combine
harvesters are faster and techni-
cally more efficient but expensive to
operate, especially in countries
where foreign exchange shortages
have reduced the availability and
raised the cost of imported machin-
ery. When combine harvesters are
unavailable or prohibitively expen-
sive, laborers are hired to harvest
maize by hand. The ears are
picked, deposited directly into
tractor-drawn wagons, and trans-
ported to storage facilities on the
farm. The ears are shelled later
using small mechanical sellers,
and the grain is bagged for sale.

Western and Central Africa-
The associations between specific
maize production technologies and
broad groups of producers are not
as distinct in western and central
Africa as they are in eastern and
southern Africa. Nor does the scale
of maize production vary as much,
partly because fewer settlers
arrived to foster large-scale com-
mercial farming. However, maize
production technologies in western
and central Africa are quite di-
verse, shaped by widely varying
agroclimatic and socioeconomic fac-
tors. The discussion that follows
therefore focuses not on producer
groups but on the principal factors
that determine which technologies
farmers use to grow maize.

Maize production in western and
central Africa is for the most part
still based on shifting cultivation
systems and slash-and-burn meth-
ods. In many areas, three or four
years of cropping alternate with
three or four years of bush fallow,
although in some places fallows
and/or cropping periods are much
longer, depending on population

density and soil conditions. Maize is
typically intercropped with other
food crops, with the predominant
combinations varying by production

In much of the humid forest zone,
increasing population and shorter
fallow periods have compelled
farmers to adopt a combination of
strategies to prevent soil degrada-
tion. The two most important
strategies are preserving trees in
cropped land and planting a range
of crops that provide good ground
cover early in the rainy season. If
fertilizer is used, it is applied at
very low levels.

In the derived and southern Guinea
savannas, maize can be grown
without chemical fertilizer on good
land, especially if the land is
located close to household com-
pounds where it can be fertilized
with organic refuse. In many areas,
farmers allow nomadic cattle
herders to keep their animals
overnight in a field to improve soil
fertility. Whenever farmers have
access to inorganic fertilizer, they
will apply it to maize rather than to
other cereals. Despite these fertility
management practices, nutrient
deficiencies are common, particu-
larly deficiencies of nitrogen and
phosphorus. Low response to major
nutrients is often exacerbated by
sulfur and zinc deficiencies.

In the northern Guinea savanna,
bush fallowing is still widely prac-
ticed. However, the fallow cycle
seems to be decreasing in some
areas and disappearing altogether
in others as fertilizer use increases.
The major constraint for maize
production in this zone is soil infer-
tility, partially brought about by
annual burnings that deplete soil
organic matter. Maize therefore
depends heavily on chemical
fertilizer, which is sometimes sup-
plemented by manure and by
rotations of maize with legumes. In

the absence of chemical fertilizer,
the yield potential of maize remains
low, around 1 t/ha.

Mechanization increases as one
moves northward and inland. In
the coastal forest zones, mechaniza-
tion is practically nonexistent, and
virtually all farming operations are
performed manually using cut-
lasses and hoes. Some degree of
mechanization has occurred in the
savanna zones, where fields are
less obstructed by trees and thus
more accessible to machinery. In
addition, labor constraints appear
to be more severe in the savanna
because of the much more clearly
defined growing seasons, which
lead to sharp peaks in the demand
for labor (Carr 1989). In the more
heavily forested areas of the south-
ern Guinea savanna, where try-
panosomiasis (sleeping sickness) is
a major problem for cattle, mecha-
nization is still not widespread, but
in many parts of the northern
Guinea savanna, where the effects
of the disease are less severe,
animal traction has become well
established. Plowing as well as
weeding are often done with ox-
drawn implements, and animal
carts are used for transport. Trac-
tor adoption has occurred in a few
areas, often with the help of direct
or indirect subsidies, but the use of
tractors has yet to prove economi-
cally viable throughout much of the

The choice of technology to prepare
land has important implications for
soil fertility, especially on the re-
gion's shallower soils. Unless it is
carefully managed, heavy machin-
ery can lead to serious soil degrada-
tion and can cause dramatic yield
declines after only three to four
years of continuous cultivation. Ox
plowing degrades the soil less than
tractor plowing, but in some areas
the use of oxen is seriously con-
strained by trypanosomiasis, as
well as by the absence of a tradition
of keeping cattle.

Well-developed land markets are
practically nonexistent in western
and central Africa. Credit markets
are generally underdeveloped,
although small loans are often
available from informal savings
associations and thrift societies.
These features are consistent with
the present relative abundance of
land. However, land for agriculture
is becoming increasingly scarce,
and this scarcity may lead to
further development of markets for
land and credit.

The prevalence of well-developed
labor markets is a bit puzzling, be-
cause labor markets tend to be

poorly developed where land is
relatively abundant (Binswanger
and McIntire 1987). The labor
markets of western and central
Africa are fed by two distinct types
of migration: seasonal migration
(which occurs when peak labor
demands in neighboring production
regions do not coincide), and per-
manent migration (which occurs
when the difference in wage rates
between one area and another
exceeds the costs of moving). Both
types of migration are expected to
increase as population pressure
leads to greater intensification of
production in zones of high

Eastern Africa

Figure 7. Maize germplasm color by region, sub-Saharan Africa, 1983-87.

Maize Types
Grown in Africa
African farmers plant many differ-
ent types of maize. White maize
predominates throughout the
continent, except for the Sahel,
where the mix is slightly more
balanced between white and yellow
materials (Figure 7). Within each
color type, the variability in other
physical grain characteristics is

In Africa as elsewhere, use of
improved maize germplasm is
difficult to estimate precisely,
because improved and unimproved
materials cannot always be distin-
guished easily. Since maize is an
open-pollinated crop, maize plants
in one field often cross with plants
in nearby fields if both crops flower
at the same time. When improved
varieties are introduced in an area
where unimproved varieties are
grown, mixtures often result, until
farmers-and breeders-cannot
always tell which varieties are
improved and which are not.
Alternative methods of determining
pedigrees, such as tracing the
sources of seed, may provide more
valid estimates than simple visual
inspection. A further complication
in assessing the adoption of im-
proved varieties is that the defini-
tion of "improved" material varies.
Improved materials are sometimes
defined as certified seed purchased
during the previous two to three
years; in other cases, seed contain-
ing mostly improved germplasm is
classified as improved. Statistics on
adoption of improved materials
should be interpreted with these
qualifications in mind.

Most African smallholders continue
to plant unimproved traditional
varieties, while most large-scale
commercial farmers grow improved

4 Smallholders in a few countries, including
Kenya, Nigeria, Senegal, Zambia, and
Zimbabwe, do plant improved maize.

maize.4 Little empirical research
has been done in Africa on eco-
nomic factors underlying farmers'
decisions for and against adopting
improved maize varieties. Differ-
ences in the rate and degree of
adoption can be explained partly in
terms of demand (for example,
farmers sometimes cannot afford
improved seed), but inadequate
supplies of improved germplasm
usually play a much larger role in
slowing the rate and degree of
adoption. Of particular importance
is the availability of appropriate
materials, which depends on local
crop improvement programs and/or
local seed production capacity.

Maize breeding in eastern and
southern Africa for a long time
focused primarily on the needs of
commercial farmers, who them-
selves contributed to the germ-
plasm improvement process by
selecting materials in their own
fields. Today, most national agricul-

Cumulative % of farmers adopting

1964 1966


tural research programs in the
region produce improved
germplasm with varietal character-
istics desired by both large- and
small-scale farmers. The fact that
both groups are often interested in
the same characteristics is illus-
trated by the experience of Kenya,
where the area under improved
hybrid maize increased from 120 ha
in 1963 to over 1,000,000 ha in
1988. In high potential zones,
small-scale producers as well as es-
tate farmers demonstrated great
enthusiasm for the commercial
hybrids developed by the national
program; smallholders' adoption
lagged only in less favorable pro-
duction environments where hy-
brids did not perform so well
(Figure 8). Hybrids have enjoyed
similar success in Zimbabwe and
parts of Zambia, demonstrating
what can happen when improved
germplasm is made available to
small-scale farmers (see "Hybrid
Maize in Sub-Saharan Africa:
Problems and Prospects," p. 21).




Source: Gerhart (1975).

Figure 8. Adoption of hybrid maize in four zones in Kenya.

Elsewhere in eastern and southern
Africa, improved maize materials
have not been adopted so readily,
often because they have not met
specific requirements of farmers.
For example in Malawi, even
though farmers expressed a clear
preference for flint maize, breeders
neglected flint types because they
had little access to improved flint
maize to use as source material in
their breeding work.5 For years
many of the improved maize varie-
ties released by the Malawian
breeding program were dent types.
Most farmers grew them only in
limited quantities for sale, prefer-
ring to plant unimproved flint
maize for home consumption. The
breeding program eventually recog-
nized this problem and now pro-
duces improved flint materials as
well as dents.

In western and central Africa,
adoption of improved maize
germplasm also has been uneven.
Use of improved materials, chiefly
open-pollinated varieties, is exten-
sive in some of the newer maize-
growing areas. In Nigeria's north-
ern Guinea savanna (parts of
Sokoto State excepted), almost all
the maize planted appears to
include improved germplasm,
mainly TZB and TZPB, varieties
that are particularly well adapted
to the ecology. Similarly, survey
data from the Brong-Ahafo region
of Ghana indicate widespread
adoption of improved maize materi-
als (Figure 9).

5 The difference between dent and flint
maize is discussed on p. 24.
6 Where government-fixed prices exist in
western and central Africa, they are
generally irrelevant, since market prices
in most years differ sharply from official

In the more humid savanna and
forest zones, improved materials
have not been as widely adopted.
More than 90% of farmers in Benin
continue to plant traditional varie-
ties, in part because improved
materials are not always available,
but also because traditional varie-
ties suffer less damage in storage
and are better suited for local
dishes (Yallou et al. 1989). But even
in the more humid zones, low adop-
tion rates are far from universal. In
southern Nigeria, the spread of
seed with predominantly improved
germplasm appears high. Inspec-
tion of sample cobs from farmers'
fields in the semideciduous lowland
forest showed that 74% contained
improved germplasm (yet only 19%
of farmers classified their materials
as improved). This high adoption
rate may not reflect a conscious
decision on the part of farmers;
seed storage problems in this area
force farmers to purchase new seed
every year, and most of it is

produced in the north where
improved materials predominate
(Smith et al., forthcoming).

Hybrid maize has been introduced
in a few countries in western and
central Africa, but it is unlikely
that the area planted to hybrids
exceeds 2% in any country. Inter-
estingly, about half the hybrid seed
sold in Nigeria is sold in 5-kg
packages, implying that smallhold-
ers too are adopting hybrids.

Marketing and Price Policy
Government policies pertaining to
maize marketing and pricing vary
across Africa. Maize markets in
eastern and southern Africa are
characterized by extensive state
participation, whereas most west-
ern and central African countries
leave marketing and pricing of
maize in the hands of the private
sector.6 The sections that follow
discuss the sources and implica-
tions of this regional variation.

Cumulative % of farmers adopting

1980 1981 1982 1983 1984 1985 1986

Source: Tripp et al. (1987).

Figure 9. Adoption of improved open-pollinated maize varieties in
Ghana, Brong-Ahafo region, 1980-86.

Eastern and Southern Africa- Production (000 t)
The economic and political impor- 4.0
tance attached to maize in eastern
and southern Africa is reflected in
extensive government involvement
in marketing and pricing. Follow- 3.0
ing independence, governments in
nearly all of the major maize-
producing countries fixed prices
from farm gate to final consumer,
required that farmers sell maize
only to authorized state grain
marketing agencies, monopolized 1.0
maize imports and exports, and
strictly regulated the private grain
trade. These actions were under-
taken for a combination of reasons: 0 -
to insure an adequate supply of
maize to meet national consump- 2.0
tion requirements, to reduce
variability in supplies and prices, to
ensure -remunerative prices for 1.6
producers and affordable prices for
consumers, and to minimize unnec-
essary marketing costs attributable 1.2
to inefficient private sector inter-
Although the empirical record is far
from complete, it is evident that
these ambitious-and sometimes 0.4
contradictory-policy goals have
not always been achieved. Produc-
tion instability continues to plague 0
many maize-producing countries of 3.0
eastern and southern Africa
(Figure 10), because most maize
farmers must depend on unreliable 2.5
and highly variable rainfall. With-
out improved production technolo-
gies to ensure that maize yields
remain fairly stable even when
rainfall varies from year to year, 1.5
marketing and price policy alone
have proved an ineffective means of
breaking the link between rainfall 1.0
and production. Consequently, most
countries in eastern and southern 0.5
Africa experience alternating
periods of over- and undersupply.
1961 1964 1967 '19'70 1973 1976 19 19 1982 1985 1988
Source: Calculated from FAO data.

Figure 10. Variability of maize production in selected countries of eastern
and southern Africa, 1961-88.

Several governments have at-
tempted to insulate maize markets
from the effects of unstable produc-
tion by introducing policies de-
signed to reduce price fluctuations.
Usually these policies set official
producer and consumer prices,
which for ease of administration
remain in place for the entire
season (pan-seasonal) and prevail
over the whole country (pan-
territorial). Given the difficulty of
eliminating unofficial marketing
activities, official prices-if
successfully defended-generally
only establish limits for the
movement of actual market prices
(a floor in the case of the producer
price, and a ceiling in the case of
the consumer price). The ability of
the marketing authority to defend
the limits established by official
prices eventually depends on the
resources at its disposal, such as
credit to finance purchases from
producers, transportation and
storage facilities to engage in
arbitrage, and stocks to sell to

Another common strategy that
many countries in eastern and
southern Africa use to control
instability in maize markets is
storage policy. Holding sufficient
stocks to cover periodic shortages,
although costly, allows the govern-
ment to act as a stabilizing influ-
ence in the domestic market; by
buying grain in times of surplus
and selling grain in times of scar-
city, the state theoretically can
dampen price and supply fluctua-
tions. The key policy issue, of
course, is deciding how much grain
to store. The experience of India
and other Asian countries suggests
that the marketing authority must
stockpile enough grain to exert
pressure on domestic prices in
times of real crisis, not just during
periods of "normal" production
variability. While it is difficult to
judge how much grain will be

needed during times of crisis, some
analysts have suggested that a
number of countries in eastern and
southern Africa are maintaining
excessively large-and hence un-
necessarily costly-reserves (see,
for example, Pinckney and Valdes
1988; Buccola and Sukume 1988).

Maize storage policies in eastern
and southern Africa have at times
succeeded in reducing market
instability, but only rarely have
price movements been eliminated
completely. Despite government ef-
forts, considerable variability in
maize prices is the norm in many
countries for both producers and
consumers. Market prices fre-
quently diverge from official levels,
often showing a marked seasonal
pattern (Figure 11).

Although the effectiveness of
marketing and price policy for
maize in eastern and southern
Africa is subject to differing inter-
pretations, one fact is beyond

dispute: government intervention
in maize markets has been expen-
sive. Direct costs-in the form of
subsidized prices and marketing
services-and indirect costs-in
the form of bureaucratic ineffi-
ciency and sometimes corruption-
have placed a considerable strain
on government budgets (for ex-
ample, see Muir and Takavarasha
1989). Widespread dissatisfaction
with the performance of public
marketing authorities and the need
to cut government expenditures
have spurred some governments to
encourage a more active role for the
private sector in maize marketing.
During the 1980s, Ethiopia,
Malawi, Tanzania, and Zambia ini-
tiated policy reforms designed to
free up maize prices and to reduce
direct state participation in market-
ing. So far, these measures are
having a positive effect, increasing
supply while reducing the strain
on government treasuries (Amani
et al. 1989; Sipula et al. 1989;
Christiansen and Stackhouse 1989).

(continued on p. 22)

Price (Tambala/kg)





0 Fl4 1 -1 1 1 ; I1
1984 1985 1986 1987 1988
Source: Kingsbury (1989).

Figure 11. Seasonal movements in official maize retail prices and open
market maize retail prices, Malawi, 1984-88.

Hybrid Maize in Sub-Saharan Africa: Problems and Prospects

Virtually no-one involved in maize
research and production would
disagree with geneticist Paul
Mangelsdorfs (1985) remark that
hybrid maize represents the most
far-reaching development in
applied biology in the latter half of
the 20th century. But as true as
Mangelsdorfs assertion may be in
general, it carries less weight in
many developing countries, where
the widespread adoption of hybrid
maize is frequently aspired to but
seldom realized.

Hybrid maize has performed well
in developing countries that have
temperate production zones, such
as Argentina, Chile, and China.
These countries contain cool, moist
environments with high potential
for maize production, and local
maize breeders can draw directly
on the stock of superior hybrids
available in the USA and Europe.

Successful hybrids have been de-
veloped for the more favored envi-
ronments of sub-Saharan Africa,
such as the higher altitude areas of
eastern and southern Africa. Yield
gains of 25% or more over local
materials attributable to hybrid
germplasm alone have been re-
corded in numerous on-farm trials.
Because the better hybrids deliver
reasonably high yields with com-
paratively low risk, they have been
widely adopted in a number of

In hopes of extending these suc-
cesses, breeders are currently at-
tempting to develop hybrids
adapted to the economically and
ecologically less favored environ-
ments of sub-Saharan Africa.
Although relatively little research
has been done on hybrid maize in
many developing countries,
hybrids show less yield advantage

when grown under lowland tropical
conditions, as in most countries in
4p-Saharan Africa. Furthermore,
.he superiority of hybrid
germplasm diminishes when it is
grown under low levels of inputs
and management: in some mar-
ginal environments under subsis-
tence farming conditions, the yield
difference between hybrids and
open-pollinated varieties becomes
narrow or nonexistent (Low and
Waddington 1989).

In sub-Saharan Africa, as else-
where in the developing world,
generating superior germplasm is
just half the battle. The other half
is delivering the materials to
resource-poor farmers. Institu-
tional and political barriers fre-
quently limit the production of high
quality seed, particularly of
hybrids. Poor seed quality often
results from inadequate production
and marketing facilities, as well as
from a lack of adequately trained
seed production specialists. These
are the predictable consequences of
severe budget constraints, espe-
cially in the public sector, which
tends to be heavily involved in seed
production. Poor seed quality can
mask the true genetic potential of
improved materials and reduce the
likelihood that farmers will
continue to use them.

Enterprises that produce hybrid
maize seed, whether public or
private, have a commercial interest
in maintaining high standards of
seed quality. However, policy
barriers often restrict the produc-
tion and distribution of
high-quality seed. Many govern-
ments control maize seed prices,
with the laudable objective.of
making seed readily accessible to a
greater number of farmers.
Unfortunately price controls may

reduce incentives for seed produc-
ers to provide a high-quality
product. In establishing retail seed
prices, for example, the govern-
ments of some countries have
restricted marketing margins
earned by seed enterprises to such
an extent that they have not been
able to meet processing and distri-
bution costs.

To help accelerate the progress of
national programs in developing
hybrids, CIMMYT and IITA pro-
vide information on the combining
ability of different materials,
improve parents for combining
ability, and develop methodologies
for producing conventional and
nonconventional hybrids. It is
important to recognize that the
eventual impact of hybrid breeding
efforts will depend very much on
improvement in local seed produc-
tion and distribution facilities,
which in turn will depend on the
general economic climate facing
seed producers and farmers. On
that front, there are encouraging
signs that local seed production
capacity is improving in many
countries, sometimes with support
from multinational seed companies.
The experiences of Kenya, Zim-
babwe, and Nigeria suggest that
the availability of hybrid materials
can contribute to the development
of a local seed industry and at the
same time have an important
catalytic effect upon suppliers of
fertilizer and other purchased
inputs. Significantly, companies
that produce and distribute hybrid
seed can also handle seed of im-
proved open-pollinated varieties
and synthetics, which in many
countries will continue to be the
most commonly used type of im-
proved germplasm well into the
next century.

Western and Central Africa-In
contrast to the highly regulated
maize markets of eastern and
southern Africa, maize markets
elsewhere in Africa are rarely
subject to government control. Few
governments in western and
central Africa attempt to partici-
pate directly in maize marketing
activities, and official maize prices,
if even announced, are rarely en-
forced. The difference in govern-
ment intervention reflects the fact
that maize, which is generally not a
primary food in western and
central Africa, has relatively less
political and economic importance
in that region than in eastern and
southern Africa.

The interesting question, of course,
is whether differing levels of gov-
ernment participation have any
appreciable effect on market per-
formance. If proponents of an active
role for government are correct, it
should be possible to discern
problems in the largely unregulated
maize markets of western and
central Africa. However, market
performance is not always easy to
evaluate. Numerous studies have
sought to determine if maize mar-
kets in western and central Africa
are economically efficient. In
efficient markets, price differences
from one location to another ap-
proximately equal transportation
costs, and price differences from
one season to another approxi-
mately equal storage costs. Price
spreads can also be expected to
include normal profits earned by
intermediaries on their investment
capital, as well as reasonable
compensation for risk. If transpor-
tation and storage costs, profits,
and/or risk premiums are excessive,
a market may be inefficient.

There is considerable evidence that
marketing margins for maize in
western and central Africa are
generally compatible with levels

that would prevail in competitive
markets. The differences between
producer prices and retail prices
are relatively large because of the
high real costs of marketing maize,
not because intermediaries wield
excessive power that allows them to
operate inefficiently or to earn
inflated profits. Williams and
Oludimu (1986) found that the
average 32% marketing margin for
shelled maize in Ondo State,
Nigeria, was reasonable given the
high cost of capital and transport.
Margins also were not found to be
excessive in northern Nigeria (Hays
and McCoy 1977; Delgado 1985)
and in Ghana (Southworth et al.

Another determinant of the effi-
ciency of maize markets is spatial
integration, which depends on such
factors as the regional stability of
production on the one hand, and
transportation infrastructure,
access to information, and availabil-
ity of credit on the other hand. If
regional maize production varies
considerably from year to year,
flows of marketed grain are likely
to change directions unpredictably,
and spatial integration is likely to
be poor. However, if production is
reasonably stable from one region
to another, spatial integration can
be expected to improve over time as
infrastructure improves and as
production increases (allowing
intermediaries to capture econo-
mies of scale in transportation and
storage). However, it is often
difficult to establish these relation-
ships empirically. Jones (1984)
reviewed several studies and
concluded that spatial market
integration in western Africa
frequently is poor because deficien-
cies in transportation, information,
and cash availability may seriously
impede efficiency.

Extreme price variability from one
season to another is sometimes
regarded as a sign of poor market

performance. Judged by this
criterion, maize markets in western
and central Africa appear to per-
form poorly, since seasonal price
differences are particularly large
for maize (see Figure 12 for an
example from Senegal). Southworth
et al. (1979) observed 100% vari-
ations in Ghana, and a recent IITA
survey done in northern Nigeria
found maize prices to be 66%
higher during the hungry period
(the time just before new crops are
harvested, when reserves from
previous crops are low or ex-
hausted) than after harvest. How-
ever, seasonal price fluctuations of
these magnitudes do not necessar-
ily indicate poor market perform-
ance, since average storage losses
of maize are believed to be as high
as 30-40% in many areas.

On the whole, it is difficult to say
whether the largely unregulated
maize markets in western and
central Africa perform better or
worse than the highly regulated
markets in eastern and southern
Africa. Clearly, the economic and
political costs of market instability
are higher in eastern and southern
Africa because maize is so impor-
tant there. Whereas most consum-
ers in western and central Africa
can substitute other staples (millet,
sorghum, rice, roots and tubers) for
maize when supplies are short,
consumers throughout large parts
of eastern and southern Africa do
not have easy access to substitutes.
For this reason, government efforts
to stabilize maize supplies and
prices in eastern and southern
Africa, while costly, are seen as
politically necessary.

International Trade
International trade in maize in sub-
Saharan Africa occurs mostly
between neighboring countries. In
normal years, only limited quanti-
ties of maize are imported from

Price (FCFA/kg)
100 -- .




60 -

1986 1987

Source: Ouedraogo and Ndoye (1988).

Figure 12. Seasonal movements in official maize producer prices and open
market maize producer prices, Senegal, 1985-87.

Europe or the Americas, partly
because of the type and quality of
grain available in global markets:
most of the maize traded interna-
tionally is yellow, but much of the
maize produced and consumed in
Africa is white. Worldwide, over 65
million tons of yellow maize are
traded annually, compared to less
than 3 million tons of white maize,
which frequently is unavailable in
global markets. Furthermore, most
of the maize traded internationally
is used for livestock feed, so it is not
handled to preserve the high
milling quality required in maize
destined for human consumption.
These factors induce most African
countries to import maize from
local sources or to import wheat or

For individual regions in Africa, the
significance of international trade
in maize depends on the local
importance of the crop. Trade is
much more significant in eastern
and southern Africa than in west-

ern and central Africa, where the
trade in cereal grains involves
mainly rice and wheat.7

Eastern and Southern Africa-
At the start of this century, trade in
agricultural commodities was
extremely limited in eastern and
southern Africa. However, the
commercial farming sector's need
for guaranteed market outlets
eventually led to the development
of export markets. Although non-
food crops such as cotton, oilseeds,
and tobacco received the most at-
tention, efforts were also made to
promote the marketing of food
grains. The government of South-
ern Rhodesia (now Zimbabwe)
began encouraging maize exports
soon after the turn of the century;
by 1914, nearly two-thirds of the
maize crop was exported, mostly to
Europe for use as livestock feed
(Muir-Leresche 1985). The main
period of international maize
exports from eastern and southern
Africa ended in the 1930s, when

world maize prices fell sharply. The
decline in international export op-
portunities coincided with colonial
authorities' attempts to insulate the
region from the uncertainties of
global markets.

Eastern and southern African
countries continue to sell relatively
little maize in world markets, since
low global prices, combined with
high transport costs, have made
exporting unattractive. Trade in
maize is seen as a means of dispos-
ing of occasional surpluses, rather
than as a strategy for ensuring
reliable national supplies of the
main staple food. However, despite
the lack of participation in world
markets, some trade continues
within the region.

Western and Central Africa-
Maize trade is limited in western
and central Africa because domestic
production in most countries does
not fall far short of human needs.
However, trade in other cereals is
substantial, particularly imports of
wheat and rice, whose consumption
has long been encouraged by a
combination of price policy
distortions (e.g., overvalued
exchange rates, retail price con-
trols, food subsidies) and structural
phenomena (e.g., urbanization,
changing employment patterns)
(Delgado and Miller 1985). Seeking
to increase food self-sufficiency, a
number of countries in western and
central Africa recently introduced
policy reforms designed to promote
production of local cereal crops
while discouraging imports. In
Nigeria between 1978 and 1982,
import duties on cereals were
raised 50-100c, and quantitative

7 In interpreting official trade statistics, it
is important to note that maize trade
between neighboring countries is often
not recorded, particularly in western and
central Africa. Thus. official maize trade
data probably understate actual trade

restrictions were placed on cereal
imports (maize and rice imports
were banned completely in 1985;
wheat imports were banned two
years later). Ghana, Senegal, and
Zaire also tightened restrictions on
imported cereals, although less
drastically, and relaxed price
controls on locally produced staples.
While the effects of these actions on
maize trade were minimal, the
growth of wheat and rice imports
decreased, and in some cases maize
production increased substantially.

Nearly all maize grown in sub-
Saharan Africa is used for human
food, with the exception of a small
amount fed to livestock (less than
10%). Some maize is consumed
green as a snack food, either
roasted or boiled. More often, dried
maize grain is processed into
porridge, soup, fermented paste, or
a kind of couscous. In all cases,
quality is important, especially
texture, color, taste, ease of proc-
essing, storage quality, and cooking

Grain texture can be hard (flint) or
soft (dent). The denting feature of
maize grain comes from the propor-
tion of hard (or vitreous) en-
dosperm in the kernel to soft (or
floury) endosperm (Figure 13). In
flint materials like popcorn, virtu-
ally all of the endosperm is vitre-
ous, and the kernel retains its
rounded shape during drying. In
dent materials, a core of floury
endosperm is embedded in a shell of
flinty endosperm; the floury core
shrinks during drying, causing the
surface of the kernel to collapse
inward and giving the grain its
characteristic dented appearance. If
maize is used as whole grain, flint
and dent types differ little in
processing efficiency. However, if
maize is consumed without the

germ, as is common in many parts
of Africa, the keeping quality of
maize flour improves but process-
ing losses may be greater, espe-
cially for dent maize.

In many parts of eastern and
southern Africa where maize is the
primary staple, rural households
show a strong preference for flint
maize, which is made into meal.
Refined meal is usually produced at
home, although partially processed
meal may be taken to a village mill
for final grinding. In some areas,
maize intended for household
consumption is taken directly to the
village mill to be ground into whole,
unrefined meal. However, prefer-
ences may be changing. Consumer
demand for more refined types of
meal in both rural and urban areas
has been steadily increasing,
probably because more refined meal
stores better and cooks faster (FAO

In western and central Africa,
preferences for different grain tex-
tures vary depending on how maize
is consumed. In areas where grain
is wet milled (i.e., milled after being
soaked in water for several days),
grain texture is less important, and
consumers generally prefer flint
maize because it stores better. But
in areas where grain is milled dry,
consumers prefer dent and floury
maize types because they are easier
to process by traditional milling

Feed use of maize is still modest in
Africa compared to the rest of the
developing world (Figure 14). At
present, only small amounts of
maize grain are fed to animals,
mostly poultry. However, two
developments could change this
situation. First, when economic
growth resumes in sub-Saharan
Africa, rising consumer incomes are

Figure 13. Structure of dent and flint maize types.

likely to push up demand for meat
and dairy products, increasing the
derived demand for livestock feed.
Second, if incentives for dairy
production improve, intensification
of the dairy industry will likely lead
to increased demand for maize as
feed. While the timing of these two
developments remains uncertain,
most analysts agree that sooner or
later demand for feed maize will in-
crease dramatically in Africa.

In addition to being used for human
food and animal feed, maize is also
used in processed foods like break-
fast cereals and in beer. Industrial
uses of maize are expected to grow
as more affluent urban consumers
shift to foods that are easier and
quicker to prepare. Future efforts
in maize improvement will there-
fore need to consider the quality re-
quirements of industrial users such
as brewers and food manufacturers.

To a certain extent, increased
utilization of maize in sub-Saharan
Africa will depend on whether
myriad production constraints can
be overcome. The next section of
this report deals specifically with
that issue.

Critical Constraints to
Maize Production

Throughout sub-Saharan Africa,
farmers' efforts to increase and
stabilize maize production are
frustrated by numerous con-
straints, ranging from low soil
fertility and unavailability of
improved germplasm, to unremu-
nerative prices and uncertain
access to markets. However, al-
though the production constraints
faced by African farmers are
serious, in many instances strate-
gies have been or can be devised to
overcome them. Success in some

Use (million t)

Latin South, East. West Asia Sub-Saharan
America and Southeast and Africa
Asia North Africa
Source: Calculated from FAO data.

Figure 14. Food, feed, and other uses of maize by developing country
region, 1986-88.

areas, such as parts of eastern and
southern Africa where improved
maize has been adopted by small-
holders, has proven that progress
against even a few constraints can
go a long way toward realizing the
potential of maize for raising food

Soil Fertility Problems
Inadequate soil fertility ranks
among the most serious constraints
on maize production in sub-
Saharan Africa. Management
practices have depleted soil fertility
in many maize-growing areas, espe-
cially the reduction in fallowing
brought about by increased popula-
tion pressure. After years of con-
tinuous cropping, most African soils
are deficient in macronutrients.
Nitrogen is usually the most limit-
ing nutrient, followed in impor-
tance by phosphorus. Sulfur and
zinc deficiencies frequently also
depress maize yields, particularly
in the savanna.

Providing sufficient nitrogen and
other essential nutrients to maize
can be difficult. Without long fallow
periods, slash-and-burn land
preparation cannot support high
and sustained maize yields unless
nutrients are added regularly. In
view of limited availability of
organic nutrient sources (animal
manure and green manure), many
farmers rely on chemical fertilizer
as the principal method of main-
taining soil fertility. Although it is
difficult to estimate how much fer-
tilizer is applied specifically to
maize, because data on fertilizer
use are usually not disaggregated
by crop, total fertilizer use per hec-
tare of arable land for sub-Saharan
Africa remains extremely low by
global standards, having increased
from 3.3 kg of mineral nutrients in
1970 to 8.6 kg in 1986. However,

these figures conceal tremendous
variability between regions, be-
tween crops, and between groups of
producers. Maize usually receives
more fertilizer than other food
crops, some of which show a low
response to added nutrients (millet,
sorghum, cassava), but less than
high-value cash crops (cotton,

Probably the single biggest obstacle
to fertilizer use in Africa is cost.
Nutrient-to-maize grain price ratios
(which indicate the number of
kilograms of maize grain needed to
pay for one kilogram of nitrogen)
reveal that fertilizer is relatively
more expensive in Africa than in
either Asia or Latin America,
particularly in landlocked countries
with no ready access to an ocean
port (Table 5). Relatively high
prices for fertilizer can be attrib-
uted to higher real marketing costs
(which are incurred when distribu-
tors must move fertilizer over long
distances and transportation is
poor), as well as to weak and dis-
persed demand (which prevents
distributors from capturing econo-
mies of scale). The importance of
marketing costs in Africa becomes
evident in comparing the cost of
nitrogenous fertilizers: the unit cost
of nitrogen contained in urea (46%
N) is consistently lower than that of
nitrogen contained in ammonium
sulfate (20.5% N), reflecting the
lower cost of transporting the more
concentrated formulation.

Relatively high prices do not
preclude the use of fertilizer on
maize if the yield response is high
enough to make fertilizer use prof-
itable. Trial data from sites
throughout Africa suggest that
modest doses of fertilizer-espe-
cially nitrogen-on maize often
generate significant yield increases
(Figure 15). A study carried out by
International Fertilizer Develop-
ment Center (IFDC) (1985) con-

Table 5. Nitrogen fertilizer price in relation to maize grain price,
selected countries, 1988/89









(Farime .U

pm.TT- M ilpggto-
(U^ ... p. 3

Source: CIMMYT survey.

Maize grain yield (kg)

0 50 100
Nitrogen (kg)

Source: Maize Commodity Research Team, Ministry of Agriculture, Malawi.

Figure 15. Maize yield response to nitrogen fertilizer on estate and
smallholder farms, Malawi.

eluded that, at current low levels of
nitrogen application, crop response
to nitrogen is frequently around 15
kg grain/kg N in forest zones and
33 kg grain/kg N in savanna zones
(the difference is partly explained
by the higher natural level of
nitrogen in forest soils). Data from
FAO trials show a crop response of
around 10-20 kg grain/kg N at rates
of 20 kg N/ha, falling to 6-14 kg
grain/kg N at rates of 40 kg N/ha
(McIntire 1986).

Whether or not these crop re-
sponses are sufficient to justify the
increased cost of purchasing and
applying fertilizer depends on a
number of factors, including the
price of fertilizer, the price of maize
grain, and the cost of additional la-
bor required. Partial budget analy-
ses done in several countries indi-
cate that fertilizer use on maize is

often profitable at prevailing
market prices. On the other hand,
researchers, extension agents, and
especially policy makers should not
be lulled into believing that it is al-
ways profitable to use fertilizer on
maize. On-farm trial data from a
number of sites suggest that, when
other factors that limit yields are
present (e.g., lack of water, inade-
quate weed control, or deficiencies
of complementary nutrients) maize
may respond only modestly to in-
creased applications of fertilizer. If
such constraints are present,
increased fertilizer use on maize
may be unprofitable, and farmers
would be making an economically
rational decision in choosing not to
apply additional fertilizer.

Subsidies often encourage ineffi-
ciently high levels of fertilizer use.
In a series of on-farm trials at 222
sites in different agroclimatic zones

Table 6. Economic profitability of fertilizer use on maize in Ghana
under two levels of fertilizer prices

-...-- .. -:~~aie n practice from:

throughout Ghana, two levels of
fertilizer application were com-
pared to the practice of using no
fertilizer (Table 6). The marginal
rate of return to the resources
invested in fertilizer is positive for
both fertilizer treatments, if 1987
prices for inputs and maize grain
are used in the calculations. How-
ever, 1987 market prices in Ghana
included a large fertilizer subsidy,
which reduced the nutrient-to-
maize grain price ratio to approxi-
mately 2:1. If the subsidy had not
been present and fertilizer had
been sold to farmers at a price
reflecting its true import cost, the
nutrient-to-maize grain price ratio
would have risen to approximately
4:1. Using this higher ratio, the
marginal rate of return to the
resources invested in fertilizer
would have been positive for the
lower fertilizer level (Treatment 2),
but negative for the higher one
(Treatment 3). These data support
the view that fertilizer use on
maize in many parts of Africa can
be justified economically only at
relatively modest levels (Carr

rtwilizer price level

T Treatment la
- tr to
AWreatment 2b


Nitrogen-to-maize grain price ratio = 2:1

Marginal costs (cedis/ha)
Marginal benefits (cedis/ha)
Marginal rate of return to
additional investment ( h)



Nitrogen-to-maize grain price ratio = 4:1

Marginal costs (cedis/ha)
Marginal benefits (cedis/ha)
Marginal rate of return to
additional investment (i%)



lent 2 In addition to high cost, a second
obstacle to increased fertilizer use
ent 3C in Africa is limited availability.
Many farmers cannot obtain
fertilizer when they need it, or in
the formulations they desire. Many
222 factors contribute to fertilizer
778 supply problems. Planning and
administering a national fertilizer
63 program requires skills that are not
always available in the government
agencies that oversee input supply,
and the private sector also may ex-
740 perience problems in distributing
260 fertilizer (Shepherd 1989). Further-
more, fertilizer imports have to be
-62 financed with foreign exchange,
which is often in short supply.
These problems are compounded by
lack of facilities for importing,

Source: Ghana Grains Development Project.
a Treatment 1 = no fertilizer; yield = 1.6 t/ha maize.
b Treatment 2 = 50 kg N/ha, 25 kg PO/ha; yield = 2.7 t/ha maize.
c Treatment 3 = 100 kg N/ha, 50 kg P,05/ha; yield = 3.2 t/ha maize.

storing, transporting, and distribut-
ing a bulky input whose chemical
composition can be affected by ex-
posure to the elements. Conse-
quently, even when fertilizer is
available to farmers on time, its
potency often has been lowered by
improper handling.

Of course, applying chemical
fertilizer is only one strategy for
maintaining soil fertility. Other
technologies being investigated
include crop rotation, crop residue
management, use of live mulches,
planting of cover crops, and agro-
forestry techniques such as alley
cropping, in which woody legumi-
nous species are grown in hedge-
rows with food crops in between.
Research on experiment stations
has demonstrated the effectiveness
of these technologies in maintain-
ing soil fertility, but further on-
farm testing is required to deter-
mine whether or not they are
economically viable from the
farmer's point of view.

In some areas, large-scale commer-
cial farmers have begun to experi-
ment with zero-tillage techniques
(which use herbicides to replace
cultivation) to manage soil fertility.
Although early results are promis-
ing, it is important to keep in mind
that herbicides are complex to
manage; thus, zero-tillage can
succeed only where the supply of
inputs is dependable, site-specific
adaptive research is possible, and
support from extension is available
(Carr 1989).

Generally speaking, the adoption of
improved soil management tech-
nologies has been limited in Africa.
The abundance of arable land has
meant that pressure to adopt inten-
sifying technologies-including soil
management practices and chemi-
cal fertilizers-has not been as
great in Africa as in other parts of
the world. However, the situation is

changing rapidly. As population
has grown and land use has inten-
sified, soil fertility levels have
declined drastically in many maize-
growing areas, to the point where
inadequate soil fertility now poses
the single most important con-
straint to maize production.

The importance of pressing ahead
with research on soil management
technologies that offer an alterna-
tive to continuous use of chemical
fertilizer is highlighted by recent
evidence showing that fertilizer
alone may be insufficient in the
long run. Results of long-term IITA
fertilizer trials in alfisols (a major
soil type for maize production)
indicate that continuous application
of nitrogenous fertilizer can lead to
declines in soil pH, organic matter,
and nutrient status. Adding large
quantities of organic material can
alleviate some of these problems
(Kang and Balasubramanian 1990).
Clearly, an integrated approach to
soil fertility management that
combines biological nutrient
sources with inorganic fertilizer
will be needed to maintain sustain-
able production.

Limited Use of
Improved Germplasm
Improved maize germplasm has not
yet been developed for all African
agroecologies, but improved materi-
als are available for most of the
major lowland and subtropical pro-
duction environments. The superi-
ority of these materials has been
confirmed under a variety of
production conditions. In experi-
ment station trials involving high
levels of management, improved
materials yield substantially better
than local checks (CIMMYT 1989).
Under farmers' conditions, im-
proved materials usually retain
some yield advantage, even in
marginal zones. For example,
Rohrbach (1988) reports that in
Zimbabwe the hybrid R200 yields

30% more than local varieties
without fertilizer, and that the
yield advantage is even greater
under drought stress. Similarly, a
decade of on-farm experimentation
by the Ghana Grains Development
Project affirms that certain im-
proved maize varieties regularly
yield better under varied agrocli-
matic conditions and management

Given the superior performance of
many improved materials, it is
reasonable to assume that African
farmers would grow them if they
could. As noted earlier, seed of
improved varieties simply may not
be available to farmers. Most maize
seed in Africa is produced and
distributed by the public sector;
private sector involvement is rare,
because private seed companies
make most of their money on
hybrids, and use of hybrids is still
minimal in most areas. A high
degree of specialization and
considerable expertise are required
to produce maize seed (CIMMYT
1987). Constrained by insufficient
resources and inadequately trained
personnel, public sector seed com-
panies in Africa have not always
demonstrated the ability to produce
and distribute adequate supplies of
high quality maize seed.

This idea-that low use of im-
proved germplasm often can be
attributed to poor availability of
seed-is reinforced by the experi-
ences of Kenya and Zimbabwe. In
each of these countries, strong
demand from the commercial
farming sector for improved maize
materials led to the emergence of
an efficient private seed industry.
After building a solid base of sales
among commercial farmers, this in-
dustry extended its distribution
network into rural areas, making
improved materials readily avail-
able to smallholders lacking the
means to travel long distances to

procure seed. Widespread adoption
of improved germplasm followed.
Significantly, the key to success in
Kenya and Zimbabwe lay not only
in the development of appropriate
germplasm, but equally impor-
tantly, in the emergence of an
effective seed production and
distribution system capable of
delivering its product to small-

In certain cases the reason for low
use of improved germplasm is more
fundamental than poor seed pro-
duction and distribution: suitable
germplasm has not been fully
developed, let alone released to
farmers. The yield advantage
offered by improved materials re-
mains slight in some areas.
Dahniya et al. (1986) found that, in
on-farm trials conducted on a range
of sites in Sierra Leone, the local
check performed better than two
improved materials when no
fertilizer was applied. In such
cases, farmers are making a ra-
tional decision in electing not to
adopt improved materials, and
additional breeding work is

Approximately 40% of the maize
area in sub-Saharan Africa experi-
ences occasional drought (defined
as causing average yield losses of
10-25/ ), whereas 25c experiences
frequent drought (causing average
yield losses of 25-50%). In eastern
Africa, southern Africa, and the
Sahelian zone of western Africa,
almost all ecologies in which maize
is produced are characterized by
unpredictable dry periods of one to
three weeks or more.

Maize farmers in Africa have
developed many strategies to cope
with drought, including selecting
drought resistant or drought toler-

ant germplasm, using water har-
vesting techniques to take maxi-
mum advantage of available mois-
ture, staggering maize planting
dates, and diversifying cropping
systems to reduce the risk of crop
failure. Strategies used by subsis-
tence farmers often are designed to
enhance yield stability (to ensure at
least enough production to meet
minimum household needs), even if
that means average yields will be

Current work on drought problems
in maize includes both germplasm
improvement and crop manage-
ment research. IITA, CIMMYT,
and several national breeding
programs are developing maize
populations with improved drought
resistance. IITA and a number of
national research systems in the
Sahelian countries collaborate on
drought work under an initiative
funded by the Semi-Arid Food
Grain Research and Development
(SAFGRAD) project. Scientists in
the SAFGRAD project have fol-
lowed two breeding strategies: they
have sought to improve drought
tolerance to mitigate the effects of
the dry spells that occur during the
growing season, and they have also
tried to reduce the time to maturity
so that maize escapes periods of un-
reliable rainfall at the beginning
and end of the rainy season.

At CIMMYT, four broadly adapted
elite populations have been chosen
for improvement as drought toler-
ant materials; the objective is to
provide national programs with late
white, late yellow, early white, and
early yellow maize populations to
use as sources of drought resis-
tance. In addition, cultivars that do
well despite drought conditions
have been combined to form the
Drought Tolerant Population
(DTP). Recurrent selection within
these populations is based on mor-
phological and physiological traits

that have been linked to drought
tolerance, including delayed leaf
senescence and reduced anthesis-
silking interval, in addition to high
grain yields under moisture stress
(Edmeades et al. 1988).

Crop management research for
drought environments has focused
on soil management techniques
such as mulching and ridging,
which are designed to increase and
preserve soil moisture levels, as
well as on techniques designed to
achieve optimal plant population
densities and spatial arrangements
for maize grown in dryland condi-
tions. Evidence from experiment
station trials as well as farmers'
fields indicates that tied ridging is
particularly effective for preserving
moisture and raising grain yields,
except in sandy soils (Figure 16).
However, the economic feasibility of
tied ridging must be evaluated case
by case, because the yield increase
does not always justify the signifi-
cant labor input required to con-
struct tied ridges.

Most weed control in Africa is
performed manually or using
animal-drawn cultivators. The fre-
quency and timing of weeding vary,
depending on the severity of weeds
and the availability of labor. One or
two weedings are common in most
areas, although they are often
delayed because of seasonal labor
constraints. Weeding maize during
a critical period 10-30 days after
crop emergence greatly enhances
grain production; uncontrolled
weed growth during this period can
reduce maize yields by 40-60%
(Akobundu 1987). Weed competi-
tion is particularly problematic in
arid and semiarid zones, since
moisture lost to weeds translates
directly into yield losses in the
maize crop.

Although agronomic data from
experiment station trials demon-
strate a clear link between timely
weeding and enhanced maize
yields, economic data from on-farm
trials do not always confirm the
profitability of "improved" weeding
practices. On-farm weed control
experiments are notoriously diffi-
cult to manage, since weed growth
varies across sites and between sea-
sons, and defining a standard
baseline "farmer practice" is not
easy. Economic analysis of trial
data is complicated by the difficulty
of estimating the opportunity cost
of labor (often family labor), which
may change throughout the season.
Many African farmers in fact adjust
their weeding practices according
to the level of weed infestation and
labor availability, which suggests
they appreciate the necessity of
weeding and allocate labor to
weeding when it pays to do so.

Maize grain yield (t/ha)

No e



0 _
Management Low High
Land Hand hoe

Where weeding competes for labor
with other income-generating ac-
tivities, herbicides have often been
investigated as an alternative form
of weed control. As with many
other improved technologies,
returns to herbicide use are highly
sensitive to site-specific factors.
Work done in Zambia suggests that
adoption of chemical weed control
in maize is more likely to be profit-
able for large-scale farmers with
access to tractors than for small-
holders (Vernon and Parker 1983).

Striga, also known as witchweed, is
an indigenous parasitic weed that
attacks the traditional crops of the
African savanna, including maize,
sorghum, pearl millet, groundnuts,
and cowpeas. As maize has become
more widely cultivated, striga has
become an increasingly important
maize parasite. Screening methods
to assess striga resistance are being
developed and several inbred lines

Low High Low High
Ox plow Tractor

and hybrids tolerant to one or more
species of striga have been identi-
fied, but much additional research
on the ecology and control of this
pest is needed.

Stem borers, termites, rootworms
and cutworms, and storage insects
are a major cause of low maize
yields and grain losses after har-
vest. The trend in sub-Saharan
Africa toward greater cropping in-
tensity, the increasing use of
minimum or conservation tillage
practices, and the growing practice
of incorporating crop residues all
contribute to rising insect popula-
tions. Although many insect pests
can be controlled effectively with
chemicals, pesticides are often
difficult to obtain, particularly for
smallholders who lack access to
production credit. On the other
hand, chemical control is increas-
ingly viewed as undesirable by
many policy makers, since improp-
erly used pesticides threaten
human safety and damage the
environment. Concern for small-
scale farmers is particularly great,
given their generally poor knowl-
edge about pesticide safety. In view
of the implications for human and
environmental health, it is likely
that the use of toxic chemicals for
pest control will eventually be
reduced (Mihm and Renfro 1987),
and breeding for insect tolerance
and resistance is now seen as an
integral part of controlling insect

In Africa, where maize grain is
frequently stored before it is
properly dried and/or without
insecticide treatment, insect dam-
age to stored grain can cause losses
of 50% or more. Problems with
storage pests typically develop as a
result of high temperatures and

Source: Rodriguez (1989).

Figure 16. Effect of different soil management practices on maize grain
yield, Kamboise, northern Nigeria, 1981.

intermediate humidity, which
encourage insect populations to
grow. Scientists are developing
methods to reduce losses before and
after harvest, either through
strategies to reduce the buildup of
borers, weevils, and other insect
pests that attack mature grain in
the field or in storage, or by seeking
to improve resistance to these
insects. Work on improving husk
cover (to present a physical barrier
to insects) and grain hardness (to
inhibit boring) is also underway,
and specialists are developing
better methods for handling,
treating, and storing grain at the
farm level after harvest.

Maize in sub-Saharan Africa is
attacked by numerous fungal, bac-
terial, and viral diseases. Prelimi-
nary CIMMYT data show wide-
spread incidence of ear rots, leaf
blight, maize streak virus, and
stalk rot. Ear rot is probably the
most serious disease, because it
reduces the yield and the nutri-
tional value of infected grain and
may cause the formation of myco-
toxins, a health threat to humans
and animals. Leaf blight and maize
streak virus also are extremely
widespread. More localized diseases
include downy mildew, sugarcane
mosaic virus, maize chlorotic mottle
virus, maize dwarf mosaic virus,
curvularia leaf spot, and brown

A variety of chemical and cultural
practices can help control maize
diseases. However, using resistant
germplasm is the most cost-effec-
tive method for disease control, as
well as the least harmful to the
environment. Breeding for disease
resistance thus remains a primary
objective of national and interna-
tional maize research programs.

Developing materials resistant to
maize streak virus receives high
priority. Epidemics of maize streak
occur periodically in many coun-
tries and in all maize-growing
ecologies of Africa. The virus is
most serious in crops that are
planted late, sown during the
minor rainy season in bimodal rain-
fall zones, or grown during the cool
season on residual moisture in
swampy areas. The international
research centers have played a
leading role in combatting maize
streak. During the 1970s, scientists
at IITA pioneered methods for mass
rearing the leafhopper that is a
vector for the disease; with the help
of these methods, IITA, CIMMYT,
and national program breeders
developed and improved maize
populations and varieties with
streak resistance. Additional streak
research is done at the CIMMYT
Maize Research Station outside
Harare, Zimbabwe.

Labor Shortages
Shortages of agricultural labor
occur throughout many parts of
Africa. These shortages tend to be
particularly severe if the returns to
farm work are low compared to
returns to outside employment, for
example, in areas where mining
offers steady and relatively remu-
nerative work.

While agricultural labor shortages
are basically caused by the limited
supply of farm workers, two types
of competition on the demand side
exacerbate the problem. In zones
where growing seasons are re-
stricted by rainfall distribution,
labor-intensive cropping operations
(such as land preparation, planting,
weeding, and harvesting) fre-
quently overlap (Figure 17). When
this happens, farmers must choose,
for example, between planting a
larger area and weeding the area
that has already been planted.

Labor (h/ha)
120 -


SNo e Jn Fb MIr Apr
Oct Nov Dec Jan Feb Mar Apr

May Juy Jul Aug

Source: Malawi Ministry of Agriculture (1977).

Figure 17. Household labor use on maize cropping operations, Chisasa,
Malawi, 1972/73.


In addition, competition between
crops commonly occurs when the
demand for labor is at its peak
(Figure 18). This competition can
have severe consequences for food
crops. Faced with allocating scarce
labor either to food crops or to high-
value cash crops such as ground-
nuts and cotton, farmers often
make the economically rational
decision in favoring the cash crops.
Many operations for food crops
must be postponed or omitted
entirely, reducing yields.

Large-scale maize producers can
reduce the peak demand for labor
by investing in machinery (e.g.,
tractors to enable earlier and faster
planting, and cultivation equip-
ment for weeding). However, the
high cost of tractors and their
limited utility on unimproved plots
generally make them an unattrac-
tive investment for small-scale
farmers, unless machinery rental
services are available. Where graz-
ing capacity is sufficient to support
cattle or donkeys (and where
trypanosomiasis is not present),
animals can substitute for human
labor, but many operations must
still be performed manually.

Faced with labor constraints, many
African maize producers adopt
strategies to spread out the demand
for labor-for example, intercrop-
ping, planting sequentially, and
distributing crops between plots
with different soils and moisture re-
gimes. Diversification strategies
typically reduce maize yields but
frequently increase total agricul-
tural output per unit land area and
raise overall returns to labor. Also,
such strategies tend to reduce the
risk of losses if a single component
in the cropping system fails.

When evaluating potential produc-
tion technologies for maize, re-
searchers have not always consid-
ered the relationship between the
opportunity cost of labor and labor
productivity. Productivity increases
are generally reported in terms of
yield gain per unit land area,
rather than in terms of yield gain
per unit of labor. However, the
latter measure may be more rele-
vant if labor is the limiting factor.
New technologies that increase the
productivity of land may also in-
crease the productivity of labor and
other scarce resources. But if the
opportunity cost of agricultural
labor is high (for example, because
of attractive off-farm employment
opportunities), returns to labor
invested in yield-increasing tech-
nologies for maize may be small or

nonexistent (Table 7). Farmers may
then decide not to adopt the new
technologies, to adopt them only
partially, or to adopt them while
modifying other practices to
accommodate the increased labor

Work done in Zambia by Ndiaye
and Sofranko (1988) illustrates the
importance of labor constraints in
farmers' decisions to adopt maize
technology. The study focused on a
sample of farmers who shifted from
traditional open-pollinated varieties
to hybrids. The authors discovered
that farmers' inability to mobilize
the additional labor required for
managing hybrids resulted in
partial adoption and compromises
in the performance of recommended
practices, such as weeding and
earthing up.

Labor (person days/mo)
240 -

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul
Source: MLARR survey data.

Figure 18. Seasonal labor requirements for maize and groundnuts,
Makoholi region, Zimbabwe (1985-87 averages).


Of course, not all new technologies
increase the demand for labor;
some, such as use of chemicals to
control weeds, have just the oppo-
site effect. While labor saving tech-
nologies theoretically enable
farmers to expand their farming op-
erations, the result is not always an
increase in agricultural production.
Work in Zimbabwe has shown that
when off-farm employment is
particularly remunerative, maize
farmers may use the labor saving
effect of a new technology to reduce
the cost of satisfying household food
requirements, rather than to
increase their total output. Thus,
technical change in agriculture
may not result in a net increase in
food production, although it will
result in higher total incomes from
increased off-farm earnings (Low

Draft Power Shortages
Draft power can be especially
important when labor is in short
supply, especially if land must
be prepared rapidly to take advan-
tage of scarce rainfall at the begin-
ning of the growing season. As
maize based production systems
have become more intensive, the
adoption of mechanized technolo-
gies has increased, beginning in
those parts of eastern and southern

Africa characterized by a conver-
gence of facilitating technical
factors (appropriate soils, flat and
open land, absence of trypano-
somiasis), economic factors (stable
and remunerative markets for
maize), historical factors (introduc-
tion of draft animals and tractors
by settlers), and institutional fac-
tors (availability of maintenance
and repair services, veterinary
services). Mechanization has not
been widely adopted in areas where
some or all of these facilitating
factors are absent, particularly in
the forest zone of western and
central Africa.

Animal traction is relatively recent
in Africa, dating back at most 100
years (except in parts of Ethiopia
where the use of draft animals goes
back several millennia). The use of
animals in African agriculture ex-
panded rapidly with the arrival of
European settlers, who practiced
animal traction and frequently ex-
propriated large tracts of relatively
flat, open land suitable for the
plow. In most of eastern Africa, use
of draft animals spread relatively
slowly to smallholders, who tended
to be unfamiliar with animal
traction technologies, frequently
lacked access to enough land to

Table 7. Returns to labor for traditional and recommended
practices for maize, Mampong-Sekodumasi area, Ghana
are-a =,w: :" tji ":pded
.. li~;t7

Yield (t/ha,
Returns to land (cedis/ha)
Total labor inputs (days/ha)
Returns to labor (cedis/day)

make them profitable, and in any
case were often prohibited from
growing the export crops best
suited to the use of draft animals.
But further to the south, small-
scale farmers were more quick to
adopt animal traction technologies,
particularly the moldboard plow,
which by the 1920s was used in
parts of Botswana, Lesotho,
Swaziland, and Zimbabwe
(Yudelman 1964).

In western and central Africa,
adoption of animal traction tech-
nologies was stimulated by the
intensification of agriculture associ-
ated with population growth and
access to markets. In the northern
Guinea savanna of Nigeria, al-
though ox-plow technology had
been available since colonial times,
it did not become widespread until
the 1970s, when improved maize
varieties were introduced and
fertilizer use became common.
Extended fallow periods also
disappeared from many areas
around that time.

The use of draft animals continues
to spread in Africa. Efforts to eradi-
cate the tsetse fly, the vector of try-
panosomiasis, have succeeded in re-
ducing the incidence of the disease
in some areas ald opening them up
to animal traction. As with any
technological innovation, the rate
and degree of adoption is influenced
not only by technical considera-
tions, but also by economic factors.
Animal traction technologies have
often proved profitable in zones
where the supply of agricultural
labor is limited. In zones where
land is abundant, greater use of
animal draft power has increased


Source: Bruce et al. (1980).

the area that can be worked by
each farming household, leading to
higher rural incomes (Figure 19).
On the other hand, since many
farmers do not own oxen and must
wait to use hired animals, plowing
and planting may be delayed, so
that the effect of the larger area
planted may be offset in part by
lower yields. In a few more densely
populated zones, expansion in total
cultivated area has reduced the
availability of land for pasture, and
the resulting shortage of animal
feed resources has created economic
incentives for farmers to seek ways
to reduce their use of draft animals.

Tractors, which first appeared in
Africa in large numbers immedi-
ately following World War I, were

used almost exclusively by Euro-
pean settlers and on large govern-
ment farms; not until after World
War II did their use spread to
African farmers. In eastern and
southern Africa, tractors are now
used for maize production primarily
on large commercial farms, al-
though more smallholders have
started to rent tractors, particularly
where labor is scarce. However,
lack of local manufacturing capac-
ity and maintenance facilities con-
tinues to keep operating costs high,
restricting adoption in many areas.
In western Africa, tractor use for
maize production remains much
more limited, partly because of
technical factors (small fields, fields
with stumps) and partly because of
economic factors (low profitability of
maize, high operating costs).

Revenue (CFA 000)
100 -Rmm m

Mali Central Central Mali Central Central
Sud Burkina Burkina Sud Burkina Burkina
(Zone 1) (Zone 2) (Zone 1) (Zone 2)

Source: Pingali et al. (1987).

Figure 19. Effect of animal traction on farm income in the Sahel: change
in revenue per hectare and per laborer.

Inadequate Incentives
for Producers and
Marketing Problems
One reason for the low yields of
maize in many parts of Africa is
that producers are not always given
adequate price incentives. Maize
producer prices vary across Africa:
some countries maintain prices
above the world price, whereas
others maintain them below the
world price. These policies often
change over time as well. Govern-
ments that attempt to increase the
profitability of maize production by
supporting producer prices are
often forced to abandon this policy
in favorable years, when large
marketed surpluses quickly ex-
haust the funds available to buy
and store grain; as a result, pro-
ducer prices fall precipitously (see
"Maize Price Cycles in Eastern and
Southern Africa," p. 36). Low
producer prices for maize do not
necessarily discourage production,
since much depends on the prices of
alternative crops, but they decrease
the likelihood that farmers will look
to maize production as an attractive
source of income.

In addition to supporting producer
prices, African governments have
also attempted to stimulate maize
production by facilitating market-
ing. Unfortunately, although public
sector participation in maize mar-
keting sometimes improves per-
formance, more often it places a
severe financial and administrative
burden on the state while improv-
ing market performance only
marginally. Poorly equipped,
inadequately staffed, and under-
funded government marketing
organizations frequently fail to sup-
port and stabilize producer prices
or provide guaranteed market
outlets. Pan-seasonal and pan-
territorial pricing policies, suppos-
edly introduced to raise and stabi-

lize producer incomes, are often
counterproductive because they
disrupt incentives for private
dealers to transport and store
grain, with the result that the state
assumes these functions at great
expense. Economies of scale in
transporting, storing, and
processing grain often do not mate-
rialize because centralized market-
ing facilities are inefficient and
wasteful. Given all of these prob-
lems, the results of government
participation in maize markets
have been mixed at best.

Recent policy changes introduced in
many countries to increase the
productivity of resources devoted to
maize production are intended to
bring about the increasing commer-
cialization of smallholder produc-
tion. If these new policies are to
succeed, it will be increasingly
important that small-scale farmers
have reliable access to market
outlets and that governments
follow a consistent policy towards
marketing and pricing. This will
require considerable rethinking of
the respective roles of the public
and private sectors in maize

In many African countries, the
most desirable level of public sector
participation in maize marketing
remains an open question. Govern-
ment marketing organizations
address real political and economic
needs in countries where the uncer-
tainty of maize production and the
difficulties of trade make policy
makers reluctant to rely on unregu-
lated private trading to ensure
adequate supplies and stable prices.
That is why countries of such
different political persuasions as
Kenya and Tanzania rely exten-
sively on state participation in
maize marketing. However, the
experience of the past 25 years
suggests that the political benefits

of public sector participation in
maize marketing must be weighed
carefully against the high economic
costs. Efforts to place maize mar-
keting entirely in the hands of
government organizations have
often proved inefficient and ineffec-
tive, confirming that private
traders are better able to perform
certain marketing functions. Thus,
the challenge facing policy makers
is to design effective marketing
systems that allow private traders
sufficient freedom to exercise their
considerable marketing skills while
at the same time ensuring the
stability that can be provided only
by the active participation of public
sector organizations.

Realizing the
Potential of Maize in

Although it is risky to make precise
projections, general orders of
magnitude can be estimated for the
factors most likely to contribute to
future growth in demand for maize
in Africa. Population growth will
stimulate demand at an annual
rate of approximately 3% well into
the next century. Income growth is
more difficult to predict, especially
in view of the disappointing per-
formance of many African econo-
mies during the past two decades,
but it is perhaps not too optimistic
to hope that economic growth of
1-2% per year will resume in many
countries before the end of the
century. Rising incomes will
translate into increasing demand
for maize, both as human food and
especially as livestock feed. At
higher income levels, the demand
for maize for food should eventually
diminish, but this effect will be
more than offset by an increase in
demand for maize used as feed and/
or in industry.

Population growth, rising incomes,
intensification of livestock produc-
tion, and policy-induced changes in
food consumption patterns could
contribute to future growth in
demand for maize in Africa at a
rate of 3-5% per year. Growth in
demand of this magnitude is well
above historical production growth
rates, implying that a significant
increase in future supply will be
necessary to maintain (or prefera-
bly improve) maize self-sufficiency.

Without a doubt, the potential
exists for significantly increasing
the productivity of resources
devoted to maize in Africa, despite
the imposing constraints reviewed
above. But will maize realize its

This report has shown that the
possibility exists for significantly
reducing the unit cost of producing
maize. If lower production costs are
translated into lower market
prices-that is, if governments do
not intervene to support prices at
high levels in an attempt to redis-
tribute income to producers-maize
production and consumption can be
expected to rise. The crucial impli-
cation for policy is that technologi-
cal change has the potential to
resolve both the supply and de-
mand problems, since higher
productivity will simultaneously
increase the profitability of maize
to producers and reduce the price
paid by consumers. Thus, techno-
logical change can lead to the inten-
sification of maize production
systems, generation of employment,
and growth in income.

Improved technologies are already
available to at least double maize
yields in many areas, and if these
technologies can be transferred to
farmers, maize production in Africa
could accelerate rapidly. However,
transferring technologies to farm-
ers and establishing the stable,

(continued on p. 37)

Maize Price Cycles in Eastern and Southern Africa

Under most state-controlled
maize marketing systems,
producer prices are announced
just prior to the growing season to
influence farmers' planting deci-
sions. However, producer prices
set in response to short-term
market conditions often have
undesirable consequences in the
long term. One striking example
of this process is the maize price
cycle observed in recent years in
several countries of eastern and
southern Africa.

The maize price cycle is a vari-
ation of the so-called "cobweb
model" of agricultural commodity
supply. The basic assumption of
the cobweb model is that produc-
ers respond to last year's price,
which reflected the previous
year's supply, which itself repre-
sented producers' response to the
price the year before, and so on
back in time. These delayed re-
sponses, which are a result of
agriculture's seasonal nature,
result in an unstable cyclical rela-
tionship between price and quan-
tity produced.

In eastern and southern Africa,
the maize price cycle typically
begins in a year when drought
severely limits production, leading
to maize shortages and necessitat-
ing high levels of imports to meet
domestic demand for food (see
Figure). Policy makers respond by
sharply raising the producer price
of maize. The higher price induces
farmers to increase their produc-
tion the following year, and a
large surplus results. In subse-
quent years, if rainfall is normal,
nominal producer prices are kept
virtually static, which in the pres-
ence of inflation translates into

declining real prices.
Declining real prices
eventually lead to
lower production as
farmers gradually
decrease area planted
to maize or reduce
their use of purchased
inputs. For a time, the
government can com-
pensate for declining
production by releas-
ing grain from its
stocks; thus, the effect
of lowered production
is concealed with the
help of the surplus
generated by the origi-
nal drought-induced
rise in prices. But
sooner or later a
serious drought
recurs, causing pro-
duction to slide again,
and shortages reap-
pear. Imports are again
to meet domestic food de
policy makers react by s
ing the producer price; a
cycle is repeated.

In Zimbabwe, the maize
has been particularly ev
Independence. When th,
government came to poi
maize producer prices h
declining for a number
and Zimbabwe was imp
maize for the first time
decades. In the face of d
production, the gover'im
the maize producer prior
higher than the previous
price and well above pri
for substitute crops.

With the help of except
vorable weather and f~i
lated to price, such as th
the civil war and impro


Year 7


The maize price cycle in eastern and southern

necessary bution of purchased inputs, farmers
demand; responded enthusiastically to the
harply rais- price increase. Sales to the official
nd the Grain Marketing Board surged
from 800,000 t in 1980 to over
2 million tons in 1981. This
price cycle quantity of maize overwhelmed
ident since marketing and storage facilities,
e new and the government was forced to
ver in 1980, export the surplus at a loss. For the
ad been next three years, producer prices
)f years, remained the same in nominal
sorting terms but actually decreased 27%
in several~ in real terms. Planted area conse
expressed quently fell,resulting in a steep
lent raised drop in production (which was
e 40% exacerbated by drought). In
s year'w reaction to plummeting maize
ces offered stocks, the government announced
a 30% nominal increase in the
1985/86 producer price. Again,
onally fE- producers responded by producing
tors ur a huge surptua
e end of
red distri&

remunerative markets that make
adoption profitable will not be easy.
Considerable effort-and public
sector resources-will have to go
into producer education programs
and market support activities. In
this regard, many problems with
maize in Africa appear institutional
and financial rather than purely
technical: Can farmers be provided
with the additional education they
need to manage more complex tech-
nologies? Can economic incentives
be created to increase production?
Can effective input delivery sys-
tems be designed and imple-
mented? Can stable markets be
assured without bankrupting state

At the same time, to say that many
obstacles to increased maize pro-
duction are institutional and
financial does not imply that all
technical problems have been
solved. Numerous technical chal-
lenges remain to be addressed,
especially with regard to sustaining
increased yields over the long term
in more marginal areas.

The relative paucity of technologies
"on the shelf," waiting to be trans-
ferred to farmers after minimal
modifications, emphasizes the need
to improve the quality of trained re-
search staff and to increase the
productivity of resources allocated
to research. During the past dec-
ade, agricultural research has been
relatively well funded in Africa, yet
technological progress has been
modest at best, particularly in food
crops (Eicher 1989). Many research
administrators have begun to
realize that money alone cannot
solve all problems; more effort must

go into carefully managing re-
sources and targeting expenditures
at research that will have a high
payoff. For example, it is not
enough to establish that drought is
a problem requiring additional
work. Instead, the implications for
research must be carefully thought
through. How can drought prob-
lems most effectively be addressed?
Should research resources be
invested in breeding programs
designed to develop drought-
resistant germplasm, or will higher
payoffs be attained if the resources
are used to finance crop manage-
ment research on practices for
conserving soil moisture?

The role of policy also requires
study. Resolving the constraints on
maize production in Africa will
certainly require improved tech-
nologies, but agricultural research
cannot be expected to produce
results in the absence of policy
reforms. Too often, technological in-
novations are blocked by inappro-
priate incentives facing producers,
unavailability of inputs, lack of in-
formation at the farm level, or
excessive delivery costs. If farmers
do not know about available tech-
nologies, cannot obtain improved
inputs, or have no incentives to
alter existing practices-or if the
government cannot afford to
sustain programs that promote
production-raising the low levels
of maize production will require
effective policy reform. Since
effective policy reform can only be
based on accurate knowledge,
research-not necessarily more,
but better, research-will be re-

quired. At the farm or local level,
research will be needed to identify
key production constraints and to
devise policies capable of resolving
them. At the national or global
level, research will also be needed
to help improve our understanding
of the factors affecting food con-
sumption patterns, so that policies
can be designed to manage produc-
tion, consumption, and trade.

Finally, the design of new technolo-
gies and policies will increasingly
have to be undertaken with a view
to their effects over more extended
time periods. The long-term impli-
cations of intensifying maize
production in Africa, with its often
fragile soils, limited rainfall, and
complex farming systems, remain
unclear. Past experience suggests
that "quick-fix" solutions designed
to raise cereal production in re-
sponse to pressing food shortages
often prove unsustainable over the
long run. Production technologies
must therefore be developed with
sustainability in mind. Institutions,
especially those governing land
tenure arrangements, will have to
be adapted as well to accommodate
the need to establish sustainable
systems for managing soil and
water resources.

The present challenge-and future
possibilities-should not be under-
estimated. Maize, more than any
other crop, offers the promise of
meeting Africa's food needs in the
years to come. If even a few of the
obstacles described in this report
can be overcome, there is no reason
to believe that the promise will not
be fulfilled.