Logistics Challenges in Sub-Saharan Africa and Opportunities ...
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA: CHALLENGES AND OPPORTUNITIES
-
Upload
independent -
Category
Documents
-
view
3 -
download
0
Transcript of INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA: CHALLENGES AND OPPORTUNITIES
INCREASING RICE PRODUCTION IN
SUB‐SAHARAN AFRICA: CHALLENGES
AND OPPORTUNITIES
V. Balasubramanian,1,� M. Sie,2 R. J. Hijmans1 and K. Otsuka3,4
1International Rice Research Institute (IRRI), Metro Manila, Philippines2Africa Rice Center (WARDA), 01 BP 2031 Cotonou, Benin, West Africa
3Foundation for Advanced Studies on InternationalDevelopment (FASID), 7‐22‐1 Roppongi, Minatoku,
Tokyo 106‐8677, Japan4National Graduate Institute for Policy Studies, 7‐22‐1 Roppongi, Minatoku,
Tokyo 106‐8677, Japan
�PrPost, C
I.
esent
oimb
I
A
a
ntroduction
ddress: Freelance consultant, Ramya Illam, 42 Thadagam Road, Velandip
tore 641025, India.
55
Advances in Agronomy, Volume 94Copyright 2007, Elsevier Inc. All rights reserved.
0065-2113/07 $35.00DOI: 10.1016/S0065-2113(06)94002-4
II.
R ice Demand and SupplyIII.
W etlands: The Potential Resource for Rice Production in SSAA. D
efinition, Area, and Distribution of WetlandsB. T
ypes and Characteristics of WetlandsIV.
R ice Soil ResourcesA. D
ryland Soils and Their CharacteristicsB. W
etland Soils and Their CharacteristicsV.
A groclimatic Zones and Rice EcosystemsA. D
ryland Rice EcosystemsB. W
etland Rice EcosystemsVI.
R ice Production Constraints in SSAA. P
hysical, Biological, and Management ConstraintsB. H
uman Resource ConstraintsC. S
ocioeconomic and Policy ConstraintsVII.
R ice Research and Technology Development Duringthe Past 20 Years
A. R
ice Germplasm, Breeding, and Variety DevelopmentB. R
ice Seed Production and Distribution ServicesC. C
rop EstablishmentD. N
utrient ManagementE. W
ater Management for Rainfed and Irrigated AreasF. W
eeds, Insect Pests, and Diseases and Their ManagementG. G
rain Quality Management: From Breeding to Millingalayam
56 V. BALASUBRAMANIAN ETAL.
H. D
iversification of Rice Farming SystemsI. IC
M for RiceV
III. R ice Intensification Issues and Thoughts for the FutureA. R
ice Intensification in Relation to Vector‐Borne Human DiseasesB. E
nvironmental Issues Related to Rice Intensification in SSAC. P
reparing for the Impact of Climate ChangeD. T
echnology Delivery and Deployment IssuesE. P
olicy Support for Rice Intensification in SSAIX.
C onclusions: Challenges to and Opportunities for EnhancingRice Production in SSA
A
cknowledgmentsR
eferencesSub‐Saharan Africa (SSA) faces multiple problems. The main one is
improving the lives of the 30% of its population that suVers from extreme
poverty and food insecurity. As more than 70% of the population lives oVfarming and related activities, agricultural development will have to play a
major role in improving this situation. Fortunately, Africa has an abundant
supply of natural resources that can support a huge expansion in food,
specifically rice production. Because of strong demand, rice area expansion
in SSA is larger than for any other crop. Total milled rice production
increased from 2.2 millionMg in 1961 to 9.1 millionMg in 2004. Rice imports
into SSA also increased from 0.5 million Mg of milled rice in 1961 to
6.0 million Mg in 2003 and SSA currently accounts for 25% of global rice
imports, at a cost of more than US$1.5 billion per year. Therefore, many
African governments accord high priority to developing their local rice sector
as an important component of national food security, economic growth, and
poverty alleviation. The abundant supply of agroclimatically suitable
wetlands (�239 million ha) and water resources can support a large expan-
sion in rice area and productivity. Currently, less than 5% of the potentially
suitable wetlands are planted with rice because of various constraints.
Expansion and intensification of rice cultivation in SSA will not compete
with other crops in terms of land and water resources because, during the
rainy season, only rice can be grown on low‐lying wetlands, including inland
valleys. In addition, the labor‐intensive nature of rice cultivation will provide
additional sources of work and income to the rural poor, especially women.
Should labor shortages become acute, however, appropriate mechanization
can be considered. Small farmers want to earn money from rice farming, but
lack modern inputs and capital to fully exploit their rice lands as these items
are limited or not available. This is where an innovative public–private
partnership is desirable to support the intensification of rice farming.
Rice is cultivated in four ecosystems of SSA: dryland (38% of the
cultivated rice area), rainfed wetland (33%), deepwater and mangrove
swamps (9%), and irrigated wetland (20%). Many abiotic stresses (drought,
flood, and variable rainfall; extreme temperatures; salinity; acidity/alkalinity
and poor soils, soil erosion, and high P fixation) and biotic constraints
[weeds, blast, Rice yellow mottle virus (RYMV), and African rice gall
midge (AfRGM)] limit rice production on the continent. The changing
climate is expected to further aggravate the abiotic constraints and reduce
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 57
rice yields in all ecosystems. Rice production is also restricted by many
technical, management, socioeconomic, health, and policy constraints.
The constraints to irrigated wetland rice in the Sahel of SSA are similar to
those faced by Asian farmers in the 1960s; therefore, well ‐tested irrigated ricetechnologies from Asia and elsewhere are being introduced and adapted to
local conditions to obtain fast returns on investment. For rice in irrigated
wetlands in the humid and moist savanna zones, rainfed wetlands, and
drylands, locally developed NERICA (new rice for Africa) varieties and
production technologies are being tested in target environment s. The proge-
nies of Oryza glaberr ima and O. sativa subspecies indica are better adapted to
rainfed and irrigated wetlands, while those of O. glaberrima and O. sativa
subspecies japonica are more suited to rainfed drylands. In addition, research
is ongoing to tackle SSA‐ specific problems such as RYMV and AfRG M and
to develop e Ycient crop management technologies. Currently available best
management practices (integrated crop management options) for di Verentrice ecosys tems are shown in Table XV. Additional support through the
provision of technical advice through revamped national R&D services; a
supply of good‐quality seed and other inputs, including farm credit; and
enabling policy are needed for profitable and sustainable intensification of
rice cultivation in SSA. It is also critical to organize preventive health
measures for farmers against wetland‐related diseases (malaria, bilharzia,
and so on), protect certain natural wetlands (e.g., with bird sanctuaries),
preserve mangrove forests in strategic coastal belts and rich peats in inlands,
and use chemical inputs eYciently to minimize pollution and maintain
environmental quality while intensifying rice production. Anticipatory
research is needed to tackle the impacts of changing climate on rice farming
and the environment. Modern information and communication technologies
(ICTs) can be exploited to reach out to farmers in remote areas and to deploy
technologies eVectively. In addition, the development of innovative private–
public partnerships and the organization of farmers into user‐groups will
enhance the training, farmer education, and technology adoption required
for intensive commercial rice farming. # 2007, Elsevier Inc.
I. INTRODUCTION
Sub‐Saharan Africa (SSA) is the world’s poorest region. More than 30%
of the 900 million people living in SSA suVer from pernicious hunger and
malnutrition (Farm‐Africa, 2004; IAC, 2005; Rosegrant et al., 2005; Sanchez
and Swaminathan, 2005). The prevalence of diseases such as malaria and
AIDS is very high (FAO, 2006b) and many countries have been disrupted by
civil war. Although the current situation is grim, there is also much potential
to improve this situation.
More than 70% of the population of SSA is rural, and agricultural
development is essential to achieve economic growth, poverty alleviation,
58 V. BALASUBRAMANIAN ETAL.
and food security. The adoption of more productive agricultural practices
coupled with the development of rural infrastructure and local markets and
supportive agricultural policy is crucial to improving both rural and urban
living conditions. Achieving this requires that donor funds and a significant
part of national income be strategically invested in scientific agriculture and
farming methods and revamping national agricultural research and exten-
sion systems (NARES) as a means of improving rural livelihoods and
national food security.
African agriculture consists of a diverse set of farming systems that have
arisen in response to the large variations in ecological, social, and economic
conditions. Dixon et al. (2001) delineated 15 broad farming systems, includ-
ing forest‐based systems; systems dominated by livestock, cereals, and root
crops; and mixed systems. It is thus evident that improving the production of
several crops and livestock will have a role in agricultural development in
SSA. In this chapter, we focus on the role of a single crop, rice, that we think
will have a particularly important role in this process. From a low base, rice
consumption and production have increased tremendously in SSA over the
past decades, and this trend is expected to continue. Moreover, rice can be
very productive and sustainable and be produced in areas where other crops
cannot be grown. However, unlike in Asia, where the unique and uniform
rice‐based farming system benefited from Green Revolution technologies
during 1965–2000, productivity gains in African rice farming will come in
small increments due to Africa’s diverse nature of cropping systems. Yet the
potential for growth in the African rice sector is enormous.
In this chapter, we first assess the potential resources for rice production
and characterize the rice‐growing environments in SSA. We then discuss the
current status of rice farming and related production constraints in four
major ecosystems, assess the progress of rice research and development in
the past 20 years, and finally indicate the challenges and opportunities for
rice intensification in SSA.
II. RICE DEMAND AND SUPPLY
Rice is a traditional staple food in parts of West Africa and Madagascar,
and it is increasingly becoming an important food in East, Central, and
Southern Africa. In recent years, the relative growth in demand for rice is
faster in SSA than anywhere else in the world. Demand for rice has been
growing due to population growth and a shift in consumer preference for rice,
especially in urban areas. Annual per capita milled rice consumption in SSA
has increased from 11 kg in 1961 to 22 kg in 2003 (Fig. 1). Rice consumption
increased steadily in all countries, except Madagascar. Mean per capita
0
20
40
60
80
100
120
140
160
180
200
1966 1971 1976 1981 1986 1991 19961961 2001
Year
Mill
ed r
ice
con
sum
pti
on
(kg
per
cap
ita)
Côte d’Ivoire Kenya MadagascarNigeria Senegal South AfricaSub-Saharan Africa
Figure 1 Trends in per capita rice consumption in SSA, 1961–2001 (source: FAO‐STAT,
FAO, 2006a).
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 59
rice consumption is high in Madagascar (122 kg), Guinea Bissau (103 kg),
Cote d’Ivoire (Ivory Coast; 100 kg), Senegal (100 kg), Sierra Leone (93 kg),
the Gambia (90 kg), Guinea (73 kg), and Gabon (72 kg).
Local rice production cannot meet the increasing demand for rice in many
African countries (Hossain, 2006). Although milled rice production increased
from 2.2 million Mg in 1961 to 8.7 million Mg in 2004, rice imports also
increased from 0.5 million Mg in 1961 to 7.4 million Mg in 2004 (Fig. 2)
(FAO ‐STA T, FAO, 2006a ; IRRI, 2006). In 2002, fou r of the six largest rice
importers were Cote d’Ivoire, Nigeria, Senegal, and South Africa. Currently,
SSA accounts for 25% of global rice imports at a cost of more than US$1.5
billion per year. Projected rice imports intoWest Africa alone will be between
6.5 and 10.1 million Mg in 2020 (Lancon and Erenstein, 2002). Declining
global rice stocks and the predicted doubling of the rice price by 2008 will put
additional strains on rice‐importing countries in SSA. Therefore, national,
regional, and international agencies are now placing a high priority on devel-
oping the local rice sector in SSA as an important component of food security,
national economic growth, and poverty alleviation.
Rice is grown and consumed in 38 countries of SSA (Table I). Figure 3
shows the distribution of rice areas in SSA. Of the total of 8.46 million
hectares (ha) of harvested rice area in SSA (5.5% of the global rice area) in
2
4
6
8
10
12
0
14
1966 1971 1976 1981 1986 1991 19961961 2001
Pad
dy
rice
pro
du
ctio
n (
mill
ion
Mg
)
Year
Côte d’Ivoire Kenya MadagascarNigeria Senegal South AfricaSub-Saharan Africa
0
1
2
3
4
5
6
7
8
1966 1971 1976 1981 1986 1991 19961961 2001
Year
Ric
e im
po
rt (
mill
ion
Mg
)
Côte d’Ivoire Kenya MadagascarNigeria Senegal South AfricaSub-Saharan Africa
A
B
Figure 2 Local paddy rice (unmilled rice) production, 1961–2005 (A) and imports of milled
rice, 1961–2004 (B) into selected countries and SSA (source: FAO‐STAT, FAO, 2006a).
60 V. BALASUBRAMANIAN ETAL.
Table I
Harvested Rice Area, Percent of Irrigated Area, Production, and Yield of
Unmilled Rice (Paddy) in African Countries (2004)
Country Area (�103 ha)aFully/partially
irrigated (%)bProduction
(�103 Mg)aYield
(Mg ha�1)a
Angola 20.0 0 16.0 0.80
Benin 33.0 2 70.0 2.12
Burkina Faso 51.0 18 95.2 1.87
Burundi 19.5 21 64.5 3.31
Cameroon 20.0 95 62.0 3.10
Central African Republic 14.5 – 29.7 2.05
Chad 80.0 9 109.0 1.36
Comoros 14.0 – 17.0 1.21
Congo, DR 415.0 0 315.1 0.76
Congo Republic 2.0 0 1.5 0.75
Cote d’Ivoire 500.0 7 1150.0 2.30
Ethiopia 8.4 85 15.5 1.86
Gambia 16.0 7 22.0 1.38
Ghana 119.4 4 241.8 2.03
Guinea 525.0 10 900.0 1.71
Guinea Bissau 65.0 1 127.0 1.95
Kenya 11.0 100 50.0 4.55
Liberia 120.0 2 110.0 0.92
Madagascar 1222.7 52 3030.0 2.48
Malawi 30.0 28 49.7 1.66
Mali 451.0 22 877.0 1.94
Mauritania 17.0 100 77.0 4.53
Mauritiusc 0.0 0.0 0.0 0.0
Mozambique 179.0 2 201.0 1.12
Niger 27.8 80 76.5 2.75
Nigeria 3704.0 15 3542.0 0.96
Reunion 0.04 100 0.08 2.00
Rwanda 13.0 8 46.2 3.55
Senegal 95.0 50 264.5 2.78
Sierra Leone 210.0 – 265.0 1.26
Sudan 4.8 75 15.8 3.28
Swaziland 0.05 100 0.17 3.40
Tanzania 330.0 27 647.0 1.96
Togo 35.0 1 68.1 1.95
Uganda 93.0 2 140.0 1.51
Zambia 10.0 0 12.0 1.20
Zimbabwe 0.25 – 0.6 2.40
Total/mean for SSA 8459.4 19.8 12,714.0 1.50
aSource of basic data: FAO‐STAT (FAO, 2006a).bBest estimates from data (1995–2004) obtained from FAO‐Aquastat (2005) at www.fao.org/
WAICENT/FAOINFO/AGRICULT/AGL/aglw/aquastat/countries/index.stm (accessed May
10, 2006) and FAO‐CORIFA (2005).cMauritius has a few hectares of rice now; however, it plans to replace some sugarcane fields
with irrigated rice in coming years to reduce rice imports.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 61
Each dotrepresents 2500 ha
15�
0�
0� 15� 30� 45� 60�
0 1000
(km)
−15�
−15�
−30�
30�
Figure 3 Distribution of rice areas in SSA.
62 V. BALASUBRAMANIAN ETAL.
2004, Nigeria and Madagascar accounted for 60% of the rice land; nine
countries cultivated rice on more than 100,000 ha each; four countries grew
rice on 50,000–100,000 ha each; and others were small rice producers with
less than 50,000 ha each (Table I). All these countries together produced
only 3% of the global rough rice output of more than 600 million Mg
per annum. In 2004, rice yields varied from 0.76 Mg ha�1 in Congo DR to
4.53–4.55Mg ha�1 inMauritania and Kenya, with an average of 1.50Mg ha�1
for the whole of SSA (FAO, 2006a).
In 1960–2000, rice area increased in SSA, but rice yields stagnated at a low
level or decreased (Fig. 4). From 2000 to 2005, average rice yields continued to
decline in West Africa and Nigeria, and increased only slightly in East Africa
and Madagascar. Therefore, any increase in national rice production came
from an expansion in area rather than a substantial increase in productivity.
This is a disturbing trend given the escalating demand for rice in many
SSA countries.
10
8
6
4
2
0
Ric
e ar
ea (
mill
ion
ha)
Ric
e yi
eld
(Mg
ha−1
)
Sub-Saharan Africa
Western Africa
Eastern Africa
Madagascar
Nigeria
Sub-Saharan Africa
Western Africa
Eastern Africa
Madagascar
Nigeria
Year1960 1970 1980 1990 2000
1960 1970 1980 1990 2000
2.5
2.0
1.5
1.0
0.5
Year
Figure 4 Trends in rice area and yield in Nigeria, Madagascar, West Africa, East Africa, and
SSA, 1960–2005 (source: FAO‐STAT, FAO, 2006a).
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 63
III. WETLANDS: THE POTENTIAL RESOURCE FORRICE PRODUCTION IN SSA
The rice plant is adapted to growing in swampy conditions and is sensitive
to water stress. This makes wetlands the most suited areas for rice cultiva-
tion. The types and distribution of wetlands and their potential for rice
cultivation in SSA are discussed in this section.
64 V. BALASUBRAMANIAN ETAL.
A. DEFINITION, AREA, AND DISTRIBUTION OF WETLANDS
For the purpose of agricultural land‐use planning, wetlands can be defined
as areas subject to periods of completely water‐saturated soils with possibi-
lities of flooding during part of the crop‐growing period. The depth and
duration of flooding of these soils depend on the position of the land on the
catena and the extent of drainage available; the flooding may be for part of
the growing season or throughout the year.
Wetlands are found not only in low‐lying areas (river and coastal flood-
plains, deltas, and depressions) but also on upper river terraces, foot slopes,
and hilltops. For example, in West and Central Africa, wetlands are a part of
the inland valley system, which is a continuum of drylands on upper and
mid‐slopes and wetlands in valley bottoms. On lower slopes, the boundary
between wetland and dryland is often gradual. The distribution of normal
and salt‐aVected (saline) wetlands along with major rivers and lakes in SSA
is shown in Fig. 5.
Our discussion on area, types and characteristics, and distribution of wet-
lands in SSA is derived from Andriesse (1986). On the basis of the 1:5 million
soil map of FAO (1977), the estimated wetland area in SSA is 239 million ha
(Table II) (Andriesse, 1986). About 40% of the total wetland area is found
in the equatorial region and the rest in the Guinea savannas (Fig. 5). About
36 million ha of wetlands are located in the tropical highlands of East and
Central Africa and Madagascar. These estimates of wetland area must be
used cautiously because of the possible large margins of error in estimates
derived from small‐scale maps.
B. TYPES AND CHARACTERISTICS OF WETLANDS
Wetlands in SSA are grouped into four categories: inland basins and
drainage depressions, inland valleys, river floodplains, and coastal wetlands.
A brief description of each category is given below.
1. Inland Basins
Inland basins are the largest area of wetlands potentially suitable for rice
cultivation, they occupy 108 million ha (45% of the total wetland area in
SSA) (Table II). They comprise drainage depressions and inland deltas
of rivers, with imperfectly to poorly drained and potentially acidic soils
(Ultisols, Oxisols, Alfisols, Entisols, and Vertisols). Examples are the Upper
Nile, Sokoto, Lake Chad, and Congo basins; the shallow swamps around
Lake Bangweulu and the Kafue flats; and the lacustrine deposits of Lake
Rukwa and Lake Eyasi.
30�
15�
0�
Wetlands
Saline wetlands
Rivers and lakes
30� 45� 60�
0 1000
(km)
−15�
−15� 0� 15�
−30�
Figure 5 Distribution of normal and saline wetlands along with major rivers and lakes in
SSA.
Table II
Estimated Areas of Four Types of Wetlands in SSAa
Type of wetlands Area (million ha)
Percentage of
total wetlands
Percentage of
total area
Inland basins/depressions 107.5 45 9.0
Inland valleys 85.0 36 7.0
River floodplains 30.0 12 2.5
Coastal wetlands 16.5 7 1.5
Total 239.0 100 20.0
aAdapted from Andriesse (1986).
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 65
66 V. BALASUBRAMANIAN ETAL.
2. Inland Valleys
Inland valley wetlands occupy �85 million ha (36% of the total wetland
area in SSA) (Table II); only 10–15% of the inland valley area is used for
agriculture. Narrow inland valleys are located upstream from river flood-
plains that are much wider. Each inland valley represents a toposequence of
a valley bottom with its hydromorphic edges, and the contiguous dryland
slopes and crests that contribute runoV and seepage to the valley bottom.
They are known as dambos or boliland in eastern and central Africa, fadamas
in northern Nigeria and Chad, bas‐fonds or marigots in francophone Africa,
and inland valley swamps in Sierra Leone. Most inland valley wetlands are
concentrated in the intertropical zone where rainfall is superior to 700 mm,
and their catchment sizes generally range from 100 to 2000 ha. The soils
(Entisols) in the valley bottoms are flooded during the rainy season, whereas
the soils (Ultisols, Oxisols, Alfisols, and Inceptisols) on adjacent drylands
are aerobic and erosion‐prone.
3. River Floodplains
An estimated area of 30 million ha (12% of the total wetland area) is under
river floodplains in SSA (Table II). A floodplain is awide, flat plain of alluvium
bordering streams and rivers that flood it periodically. Well‐developed flood-
plains extend from tens of meters to tens of kilometers on either side of
large rivers such as the Gambia, Niger, Benue, Zaire, Zambezi, Limpopo,
Tana, White and Blue Nile, and Chari. Soils of the floodplains (Entisols,
Inceptisols) are moderately well to poorly drained and medium‐ to fine‐textured with moderate to high fertility. Soils can be saline and/or alkaline
in drier regions.
4. Coastal Wetlands
Coastal wetlands cover an estimated area of 16.5 million ha (7% of the
total wetland area in SSA) (Table II). They comprise
� Deltas (e.g., Niger in Nigeria, Rufiji in Tanzania, and Zambezi in
Mozambique)
� Estuaries (at the mouths of the Zaire, Cross, Gambia, and Corubal rivers)
� Intertidal flats or lagoons along the West and East African coasts.
Soils (Entisols, Inceptisols, Histosols) are poorly drained and nonsaline in
freshwater swamps, acid sulfate (Entisols, Inceptisols) in mangrove swamps,
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 67
poorly drained and saline (Inceptisols) in lagoons, coarse‐textured (Entisols,
Inceptisols) in sand bars and dunes, and organic (Histosols) in permanently
flooded areas.
IV. RICE SOIL RESOURCES
Soils provide the base for crop production. The physical and chemical
characteristics of soils reflect the parent materials from which they are
derived. For example, coarse infertile soils are developed from poor and
acid rocks (sandstones, granites, quartzites, rhyolites), coarse‐ to medium‐textured soils of low to moderate fertility from intermediate rocks (horn-
blende, granites, quartz‐feldspar gneisses, diorites, andesites), and medium
to fine, relatively fertile soils from rich parent rocks (amphibolites, dolerites,
basalts, hornblende gneisses, shales, siltstones). Soils derived from volcanic
ash are rich in most nutrients except N and P. In addition, the climate, the
extent of weathering, the position of soils on landscape, vegetation cover,
and soil moisture regimes aVect the physical, chemical, and biological prop-
erties of soils. In this section, we briefly discuss the two broad categories of
soils—dryland and wetland—and their characteristics.
A. DRYLAND SOILS AND THEIR CHARACTERISTICS
Dryland soils on the upper and middle slopes of the catena are generally
well drained, deep to very deep, and coarse‐ to medium‐textured or gravelly
in the humid forest and Guinea savanna zones. Soils in the upper parts of the
toposequence are moderately to well drained, shallow to medium in depth,
and coarse‐ to medium‐textured or gravelly in the Sudan savanna transition
zones (Andriesse and Fresco, 1991). Most of the dryland soils are highly
weathered and greatly leached with low‐activity clays (silicious, kaolinitic,
and halloysitic minerals) in the humid zones and less leached and fairly rich
in weatherable minerals in the Sudan savanna transition zones.
The red clayey or clay‐loam soils dominant in the humid forest andGuinea
savanna zones are classified as Ultisols and Oxisols, and some as Alfisols; the
brown sandy or sandy loam soils of the Sudan savanna transition zones
belong to Alfisols, Inceptisols, or Entisols; and the organic matter (OM)‐rich, relatively fertile soils of the highlands of East and Central Africa and
Madagascar are mostly Inceptisols. Volcanic ash soils found in some dryland
rice areas (e.g., in parts of Cameroon) are also Inceptisols (Kawaguchi and
Kyuma, 1977; Sanchez and Buol, 1985). Gravelly soils (Entisols) and soils on
steep slopes are highly prone to erosion and not suitable for dryland rice.
68 V. BALASUBRAMANIAN ETAL.
The inherent fertility of most dryland soils is low (low pH, cation
exchange capacity, and base saturation) in warm and humid zones due to
intense weathering of parent materials, and medium to high in semiarid
zones due to less intense weathering and, in West Africa, due to deposition
of Ca, Mg, and K from dust deposits from the Sahara Desert that occur
in the dry season (DS) (Andriesse and Fresco, 1991). Subsoil acidity and Al
and Mn toxicity are common problems in some soils. Soil OM and N
contents decrease as we move from humid to subhumid to semiarid zones.
Total soil P is generally low, with only 2–4% of the total P available to plants
(Table III) (Breman, 1998). This is because of the high P‐fixing capacity of fine‐textured soils found in humid and subhumid zones (Abekoe and Sahrawat,
2001; Juo and Fox, 1977). Therefore, P fertilization is an important requisite
to get a crop response to other nutrients such as N. K deficiency is more
severe in humid regions than in drier areas.
B. WETLAND SOILS AND THEIR CHARACTERISTICS
Rice soils converted from natural wetlands are fairly rich in exchangeable
bases (Ca, Mg, and K), slightly acidic to neutral (pH 6–7), low in P‐fixationcapacity, and not Al toxic. The soils are often mineral in seasonally dry
wetlands and peaty in permanently flooded wetlands. Most mineral wetland
soils are fluvial, lacustrine, estuarine, or marine alluvial materials. In soil
taxonomy (Soil Survey StaV, 1998), soils with an aquic moisture regime and
specified morphologic characteristics of wetness are distinguished at the
suborder level (e.g., Inceptisols: Aquepts; Alfisols: Aqualfs). At the subgroup
level, aquic subgroups with signs of wetness only in lower horizons are not
considered as wetland soils, except the aquic subgroups of flooded Vertisols,
Table III
Mean Chemical Characteristics of Dryland Soils (0–0.3 m) in Four
Agroecological Zones of West Africa
Soil properties Equatorial forest Guinea savanna Sudan savanna Sahel
pH (water) 5.7 5.7 6.7 5.7
Org. C (g kg�1) 20 12 6 3
Total N (g kg�1) 2.0 1.3 0.5 0.2
Total P (mg kg�1) 260 340 210 100
Bray‐1 P (mg kg�1) 9 7 4 4
CECa (cmol kg�1) 8.7 8.5 8.1 2.5
Base saturation (%) 28 59 69 28
aCEC, cation exchange capacity.
Adapted from Breman (1998).
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 69
such as Chromoxererts, Chromuderts, Pelloxererts, and Pelluderts, with
distinct mottles in the upper 0.5 m of the soil profile. Histosols are mostly
wetland soils. In another classification of soils, the FAO–UNESCO legend for
the soil map of the world (FAO, 1977), the Gleysol, Fluvisol, Planosol, and
Histosol and gleyic units of Acrisol, Arenosol, Cambisol, Luvisol, Podsol,
and Regosol and periodically wet Ferralsol refer to wetland soils. Major
wetland rice soils of SSA, their classification in soil taxonomy and the
FAO–UNESCO system, and a third system based on fertility capability
classification (FCC) and their potential for rice cultivation are given in
Table IV. The FCC (Buol et al., 1975) is a technical system for grouping
soils according to the kinds of problems they present in agronomic manage-
ment of crops. It is based on quantitative topsoil and subsoil (up to 0.5 m)
parameters directly relevant to plant growth. Additional locally important
features can be considered and added to the FCC as primes (0) or asterisks (�).The adapted FCC of wetland soils (Sanchez and Buol, 1985) is used for the
FCC in Table IV.
Rice soils from converted natural wetlands include alluvial (Entisols),
hydromorphic (Inceptisols), black cracking clay (Vertisols), and red clayey
(Ultisols and Oxisols) or loamy soils (Oxisols or Alfisols). Except for the
black clay soils, they are acidic, moderately to highly P‐fixing (Abekoe and
Sahrawat, 2001), and low in potentiality, with less than 10% weatherable
minerals; they are also diYcult to puddle. These soils scattered all over SSA
are moderately to highly suitable for rice cultivation (Table IV).
Organic soils (OM> 20%) orHistosols include peat andmuck soils found in
continuously flooded wetlands in low‐ and high‐altitude regions (forest zonesof West Africa and highlands of East and Central Africa and Madagascar).
There are two classes of organic soils: deep (organic in top 0.5m of the profile)
and shallow organic over (clay, loamy, or sandy) mineral soils within the
top 0.2 m of the profile. Shallow organic soils over clay are the most common
in SSA and elsewhere, and they are low in bearing capacity (mechanization
not possible) and deficient in many nutrients (Sanchez and Buol, 1985).
Acid sulfate soils are organic soils over clay with marine sediments found in
some coastal wetlands of SSA.
Many soils with adequate water retention capacity but lacking an aquic
soil moisture regime are converted into artificial (or anthropic) wetlands
through terracing and irrigation. Two important soil types of artificial wet-
lands are oxic soils (Oxisols, OxicUltisols/Alfisols) andAndepts (volcanic ash
soils). Oxic soils have more than 35% clay, low pH and cation exchange
capacity, high levels of free iron that can result in Fe toxicity when flooded,
and high P‐fixing capacity and Al toxicity (Abekoe and Sahrawat, 2001;
Sahrawat, 2004a); they are generally diYcult to puddle and have low to
moderate potential for rice cultivation (Sanchez and Buol, 1985). These
soils are found in rainforest and derived savanna zones. Volcanic ash soils
Table IV
Wetland Rice Soils, Their Classification, and Their Potential for Rice Production in Africa
Soil type
Soil Survey StaV
(1998)
FAO–UNESCO Soil
Map (FAO, 1977) FCCa
Potential for
rice cropping
A. Wetland rice soils converted from natural wetlands
Young alluvial
soils
Entisols:
Ustifluvents,
Torrifluvents,
Xerofluvents,
Ustipsamments
Fluvisols: Dystric,
Eutric
SLgek, SCghek High
Arenosols: Gleyic
Black or peaty
soils
Inceptisols:
Epiaquepts,
Humaquepts
Gleysols: Humic Lgak (coastal),
Cgaik (highlands)
Moderate
Hydromorphic
Vertisols
Vertisols:
Calciaquerts,
Duraquerts,
Dystruderts,
Calcixererts,
Durixererts
Vertisols: Pellic Cvg High
Inceptisols:
Epiaquepts
Red clay soils:
humid zone
Ultisols: Udults,
Aquults
Acrisols: Ferric,
Gleyic
Cgaeik0 Moderate
to low
Red clay‐loamsoils:
subhumid and
semiarid zones
Oxisols: Aquox,
Ustox
Ferralsols: Orthic,
humic
(LC)ghk0d,Lgh(e)k
Moderate
Alfisols: Aqualfs,
Udalfs, Ustalfs
Luvisols: Orthic,
Ferric
B. Wetland rice soils in anthropic (artificial) wetlands on terraced landscapes
Oxic soils
(>35% clay)
Oxisols, Oxic
Ultisols/Alfisols
Ferralsols, Acrisols,
Luvisols
Cigae Low/
moderate
Tropical brown
soils/volcanic
ash soils
Inceptisols:
Epiaquepts,
Sulfaquepts,
Halaquepts
Cambisols: Gleyic,
Eutric
Andosols: Ochric,
Humic
Lxd, Lxg High/very
high
Entisols:
Hydraquents
C. Problem soils
Acid sulfate soils Entisols:
Sulfaquents
Fluvisols: Thionic OCcg�ak Low/
moderate
Inceptisols:
Sulfaquepts
Peats (deep) Histosols (>20%
organic matter)
Histosols: (>20%
organic matter)
Og�akk0 Low/very low
Peats (shallow) OCg�akk0,OLg�akk0,OSg�akk0
Low/
moderate
Saline soils Inceptisols:
Epiaquepts,
Aridisols:
Natrargids
Solonchaks: Gleyic Ssgbkk0, SCsgbkk0 Limited
70 V. BALASUBRAMANIAN ETAL.
Table IV (continued)
Soil type
Soil Survey StaV
(1998)
FAO–UNESCO Soil
Map (FAO, 1977) FCCa
Potential for
rice cropping
Alkaline soils Inceptisols:
Natraquepts
Solonets: Gleyic Cngbgkk0,SCngbkk0
Limited
Alfisols: Natraqualfs
Vertisols:
Natraquerts
Degraded gray
soils
Ultisols:
Plinthaquults
Acrisols: Plinthic SLgkk0ea,CLgkea
Moderate
(H2S
toxicity)
Skeletal soils
(>35% gravel)
Ultisols:
Endoaquults
Planosols: Eutric,
Dystric
SL00gkk0i, LC00gkk0i,SRgk, LRgk
Very low
aFCC, fertility capability classification (Sanchez and Buol, 1985): S, sand; C, clay; L, loam;
O, organic; R, rocky; c, acid sulfate; x, X‐ray amorphous allophonic; g, gleic; g�, pergleic;gd, wetland soils dry for 60 days or more; k, low potential productivity with <10% weatherable
minerals; k0, low in K; e, low cation exchange and buVering capacity; a, acid pH < 5 þ Al toxic
þ P‐fixing; h, acid, pH 5–6 but not Al toxic; b, basic pH > 7.3; v, vertic; s, saline; n, alkaline;
%, steep slope; i, high levels of free iron, highly P‐fixing.Note: FCC ratings for diVerent soils are first approximations based on available information,
and they can be modified or further improved with additional data at local level.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 71
(Inceptisols) are dominant in X‐ray amorphous allophoneminerals, rich in all
nutrients except N and P, and high in P‐fixation capacity; they pose moderate
diYculty to puddling and regenerate the structure easily on drying. They are
good for rice dryland crop rotations with high productivity (Kawaguchi and
Kyuma, 1977).
DiVerent types of problem soils occur in SSA: acid sulfate soils (Inceptisols:
Sulfaquepts) in mangrove swamps, deep and shallow peats (Histosols) in low‐and high‐altitude areas, saline (Inceptisols: Halaquepts) and alkaline (Incep-
tisols: Natraquepts; Vertisols: Natraquerts) soils in the Sahel and coastal
wetlands, degraded soils (Ultisols: Aquults) in the humid forest zones, and
skeletal soils (Ultisols: Aquults) on adjacent sloping drylands of inland
valleys. Soil pH decreases drastically and Al toxicity increases only when
acid sulfate soils are drained. Remediation of adverse conditions and the
combined use of salt‐tolerant rice varieties and integrated nutrient applicationare needed to grow rice on saline and alkaline soils.
V. AGROCLIMATIC ZONES AND RICE ECOSYSTEMS
The suitability of land for various crops is determined by climate and
weather variables (agroclimatic zones), landscape moisture regimes (physio‐hydrographic positions), and soil characteristics. Rice thrives in areas with
72 V. BALASUBRAMANIAN ETAL.
warm temperatures (above 20�C during the cropping season), annual rain-
fall ranging from 0.5 to >1.5 m, and growing periods of 90þ days. Rainfall
distribution is generally monomodal with a distinct humid period in areas
with a rainfall range of 500–1000 mm per annum (e.g., northern and south-
ern Guinea savanna zones) and bimodal with two growing seasons toward
the equator (Andriesse, 1986).
The classification of land and water resources into (agro)ecosystems is
helpful for the planning, development, management, and monitoring of
these resources. The choice of classification scheme is always subjective to
some extent and will depend on the purpose of classification. Andriesse and
Fresco (1991) defined 18 rice environments by combining six landscape or
physio‐hydrographic positions with three agroecological zones (AEZs) for
West and Central Africa. For this chapter, we use surface‐water regimes to
define four major rice ecosystems: dryland, rainfed wetland, deepwater and
mangrove swamps, and irrigated wetland (Fig. 6). The distribution of rice
areas into four ecosystems and the related environmental/human disease
constraints for diVerent African countries are shown in Table V. For 16
large rice‐growing countries, actual mean rice areas (1995–2004) in diVerentecosystems are provided in Table VI.
A. DRYLAND RICE ECOSYSTEMS
Dryland rice is also known as ‘‘upland’’ or ‘‘pluvial’’ rice. It is cultivated
on level or sloping lands and on hydromorphic fringes (Fig. 6) in fields that
do not have bunds to retain surface water (Sie, 1991). Flooding is rare in this
ecosystem, and dryland rice depends solely on rainfall and should have a
water table that remains at 0.5 m or more below the soil surface. Similar to
Plateau
Slopes
Valley bottom FloodplainsHydromorphic edge
Rainfed wetlandIrrigated wetland
Dryland
DW/Mangrove
Figure 6 Major rice ecosystems of SSA as defined by surface‐water regimes.
Table V
Rice Area by Ecology/Ecosystem and Related Environmental and Human Disease Problems
for Countries of SSA
Country
Mean area
(1995–2004)
(�103 ha)a
Percentage of area by
ecology/ecosystema
Environmental and human
disease problemsIRb RWRc DRd DWRþe
Angola 21.2 0 100 0 0 HIV, malaria, cholera
Benin 21.7 2 7 91 – Malaria
Burkina Faso 47.9 46 50 4 0 Bacteria/N/Cl�1
pollution, salinity
Burundi 17.7 21 74 4 0 Malaria, bilharzia
Cameroon 16.9 95 5 0 0 Malaria, bilharzia
Central Africa
Republic
13.1 0 3 97 0 Malaria
Chad 83.7 9 6 0 85 Poor water and sanitation
Comoros 14.0 – – – – Malaria/diarrhea/dengue
Congo, DR 445.2 0 11 89 0 Petroleum, mineral
pollution
Congo Republic 1.8 0 100 0 0 –
Cote d’Ivoire 500.9 7 15 78 0 Malaria/worms
Ethiopia
(2001–2004)
8.4 85 15 0 0 Salinity, malaria
Gabon 0.5 100 0 0 0 –
Gambia 13.9 7 63 16 14 Salinity, deforestation
Ghana 116.9 24 21 55 0 Water pollution, malaria
Guinea 495.5 10 23 44 23 Malaria/diarrhea/bacterial
diseases
Guinea Bissau 65.5 1 27 27 45 Salinity/silting,
malaria/diarrhea
Kenya 11.8 100 0 0 0 Land/wildlife park
conflicts, malaria
Liberia 121.0 2 6 92 0 –
Madagascar 1195.7 52 18 29 1 Malaria, bilharzia
Malawi 43.8 28 72 0 0 Malaria
Mali 365.3 22 13 1 64 Bilharzia
Mauritania 18.1 100 0 0 0 Salinity, soil degrading
Mauritius 0 0 0 0 0 –
Mozambique 164.6 2 59 39 0 Salinity, HIV/malaria/
diarrhea/bilharzia
Niger 22.7 80 0 0 20 Malaria/diarrhea/bilharzia
Nigeria 2466.4 17 47 30 6 Pollution, malaria/worms
Reunion 0.04 100 0 0 0 –
Rwanda 5.1 8 92 0 0 Deforestation/soil erosion/
peat drying
Senegal 79.2 50 40 0 10 Acid/alkalinity/pollution
Sierra Leone 236.6 0 28 68 4 Deforestation/peat harvest
Sudan 4.7 75 – – – Silting/pollution
(continued)
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 73
Table V (continued)
Country
Mean area
(1995–2004)
(�103 ha)a
Percentage of area by
ecology/ecosystema
Environmental and human
disease problemsIRb RWRc DRd DWRþe
Swaziland 0.05 100 0 0 0 Malaria, bilharzia
Tanzania 439.9 4 73 23 0 Salinity/pollution/silting
Togo 38.1 1 19 80 0 Pollution, malaria
Uganda 71.0 2 53 45 0 Salinity/silting, malaria
Zambia 10.9 0 100 0 0 Malaria/diarrhea
Zimbabwe 0.2 – – – – Civil strife, malaria/
bilharzia/diarrhea/
agrochemical poisoning
Total/mean for
SSA
7180.0 19.8 33.5 37.8 8.9 –
aBest estimates from data (1995–2004) obtained from FAO‐Aquastat (2005) at www.fao.org/
WAICENT/FAOINFO/AGRICULT/AGL/aglw/aquastat/countries/index.stm (accessed May
10, 2006), FAO‐CORIFA (2005), Defoer et al. (2002), Adegbola and Singbo (2005), and
Ezedinma (2005).bIR, irrigated wetland rice.cRWR, rainfed wetland rice.dDR, dryland rice.eDWRþ, deepwater and mangrove rice.
74 V. BALASUBRAMANIAN ETAL.
the situation in Asia, dryland rice is generally a subsistence crop in Africa.
It is of critical importance for the local food security of poor communities
that do not have access to wetland fields. In Cote d’Ivoire, Guinea, Guinea
Bissau, and Sierra Leone, dryland rice is the only staple available between
the maize and cassava harvests or between sweet potato and cassava
(McLean et al., 2002). In the Zambezi region of Mozambique, dryland rice
is the main source of food for the poor.
1. Dryland Rice Area
Of the global dryland rice area of 14 million ha in the 1990s, 2.7 million ha
are planted to dryland rice in Africa (Table III); it accounts for an estimated
38% of the total rice area in SSA. Nearly 97% of the cultivated rice area
in the Central African Republic is in drylands. Other countries with domi-
nant dryland rice ecosystems (more than 65% of the cultivated rice area) are
Liberia, Benin, Congo DR, Togo, Cote d’Ivoire, and Sierra Leone (Table V).
Countries with more than 100,000 ha of dryland rice are Congo DR, Guinea,
Table VI
SSA Countries with Large Areas in Rice Ecologies/Ecosystems (Mean of 1995–2004)
Country
1995–2004 Mean rice area (�103 ha)a
IRb RWRc DRd DWRþe Total
Benin 0.4 1.5 19.8 0.0 21.7
Burkina Faso 22.0 24.0 1.9 0.0 47.9
Chad 7.5 5.0 0.0 71.2 83.7
Congo, DR 0.0 49.0 396.2 0.0 445.2
Cote d’Ivoire 35.1 75.1 390.7 0.0 500.9
Ghana 28.1 24.5 64.3 0.0 116.9
Guinea 49.5 114.0 218.0 114.0 495.5
Liberia 2.4 7.3 111.3 0.0 121.0
Madagascar 621.8 215.2 346.8 12.0 1195.7
Mali 80.4 47.5 3.6 233.8 365.3
Mozambique 3.3 97.1 64.2 0.0 164.6
Nigeria 419.3 1159.2 739.9 148.0 2466.4
Senegal 39.6 31.7 0.0 7.9 79.2
Sierra Leone 0.0 66.2 160.9 9.5 236.6
Tanzania 17.6 321.1 101.2 0.0 439.9
Togo 0.4 7.2 30.5 0.0 38.1
Uganda 1.4 37.6 32.0 0.0 71.0
aBest estimates from data (1995–2004) obtained from FAO‐Aquastat (2005) at www.fao.org/
WAICENT/FAOINFO/AGRICULT/AGL/aglw/aquastat/countries/index.stm (accessed May
10, 2006), FAO‐CORIFA (2005), and FAO‐STAT (FAO, 2006a).bIR, irrigated wetland rice.cRWR, rainfed wetland rice.dDR, dryland rice.eDWRþ, deepwater and mangrove rice.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 75
Cote d’Ivoire, Liberia, Madagascar, Nigeria, Sierra Leone, and Tanzania
(Table VI).
2. Cropping Systems
Dryland rice systems range from shifting to permanent cultivation. Shift-
ing or slash‐and‐burn cultivation is common in humid forest zones of West
Africa where farmers cut and burn the bush‐fallow vegetation and plant rice
as the first crop to exploit the soil fertility built up during the fallow period
and the ash from burning. They may apply a little manure or compost, but
no chemical fertilizers are used. As this practice depletes the soil, weeds build
up and rice yields decline drastically after the second crop, when farmers
plant cassava on old plots and move to a new area for rice cultivation.
76 V. BALASUBRAMANIAN ETAL.
Nutrient mining degrades the soil in this type of slash‐and‐burn system
(Fernandez et al., 2000; Oldeman et al., 1991).
Farmers plant rice as a sole crop or mixed with maize, beans, yam,
cassava, or plantains (mixed cropping) to avoid risks. In areas with a long
growing season and suYcient rainfall, dryland rice is rotated with maize,
cowpea, beans, soybean, or sweet potato. In inland valleys, dryland rice and
cash crops are often grown on hydromorphic fringes while fodder crops and
trees grow on upper slopes and crests. In the highlands of East Africa,
dryland rice is rotated with wheat, maize, potato, or sweet potato.
3. Cultivation Practices and Yields
In this ecosystem, the fields are not bunded, there is no flooding, and the
soil remains aerobic (not saturated with water) for most of the growing
season. Rice seeds are sown by broadcasting or dibbling in hand‐hoed fields.
An adequate supply of soil water is critical for good plant growth and yield.
This can be achieved by in situ rainwater harvesting (RWH) through im-
proved infiltration of rainwater by proper tillage, reduced water loss from the
soil surface by proper mulching or plant cover, and improved crop‐water useby selecting adapted varieties and following moisture‐conserving cultivation
practices (Hatibu et al., 2000). As dryland rice farmers are generally poor,
they may apply some household wastes and other organic manure but do not
generally apply purchased inputs to their rice crops. At harvest time, mature
panicles are collected, dried, and threshed manually.
Dryland rice yields range from less than 0.5 Mg ha�1 on subsistence farms
to 2 Mg ha�1 in well‐managed permanent cropping of rice in rotation with
legumes and other crops. Yields are low on subsistence farms because of poor
cultivation methods, low‐input use, excessive weeds, and depleted soils. With
the introduction of dryland NERICA varieties in SSA, farmers in Uganda
obtained average rice yields of 2.2 Mg ha�1 with moderate inputs (Kijima
et al., 2006).
4. Production Constraints
Both abiotic and biotic constraints limit rice production in drylands.
Serious abiotic constraints include variable rainfall, low temperature in
high‐altitude areas, and poor soils. The total rainfall of 0.9–2.0 m is ade-
quate in the humid and subhumid areas where dryland rice is generally
grown in SSA, but the rainfall distribution can be poor, with unpredictable
dry spells. On the other hand, the temperature during the growing season is
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 77
relatively favorable in most parts, except in the tropical highlands of East
and Central Africa and Madagascar.
Degradation of soil structure and surface sealing constrain crop emergence
and growth in semiarid areas (Andriesse and Fresco, 1991). Dryland rice
requires extensive weeding to get a decent yield and, in such systems, topsoil
erosion can be a serious problem, especially on slopes. Drought is another
serious problem for dryland rice due to the inadequate quantity and/or poor
distribution of rainfall and shallow depth of and surface crusting in some soils.
Among biotic factors, weeds are the most serious, followed by blast and
brown spot diseases. Weeds are generally more competitive than rice in
infertile dryland soils. Striga is a parasitic weed of dryland cereals, including
rice. Estimated yield losses due to weeds range from 30 to 100%. Weed
infestation and loss of N reduce yields by 25% on intensive dryland rice
farms of West Africa (Becker and Johnson, 2001). Stem borers and rice bugs
are the major insect pests. Nematodes are serious problems in continuously
monocropped dryland rice fields (Coyne et al., 2004; Plowright and Hunt,
1994) and can reduce yields by up to 30%. Termites are problems in some
areas. Rodents and birds damage rice crops in all ecosystems.
B. WETLAND RICE ECOSYSTEMS
Unlike the dryland ecosystem, rice fields in the wetland ecosystem are
flooded during the growing season. We distinguish three types of wetland
rice ecosystems—rainfed wetland, deepwater, and irrigated—as determined
by the surface‐water regime. The ecosystem is considered to be rainfed
wetland when the water supply to crops is from rainfall and groundwater.
In contrast, in the deepwater ecosystem, most of the water on the fields is from
the lateral flow of water onto the land. In the irrigated wetland ecosystem, a
significant part of the water supply is from irrigation.
1. Rainfed Wetland Rice Ecosystem
Rainfed wetland rice is grown on lower parts of the toposequence and in
valley bottoms (Fig. 6) in level to slightly sloping bunded fields that are
flooded by rainwater for a part of the growing season to water depths that
may exceed 1.0 m for not more than 10 consecutive days. Both rainwater
and stored groundwater support rainfed wetland rice. Four types of flooded
wetlands are recognized in Africa: riverine shallow, riverine deep, boli-
land (grassy inland swamps), and mangrove (Andriesse and Fresco, 1991;
Buddenhagen, 1986).
78 V. BALASUBRAMANIAN ETAL.
Rainfed wetlands are characterized by a lack of water control, with
droughts and floods being potential problems (Hatibu et al., 2000; McLean
et al., 2002). On the basis of the constraints, rainfed wetlands can be divided
into four subecosystems: (1) favorable, (2) drought‐prone, (3) submergence‐prone, and (4) drought‐ and submergence‐prone. All four subecosystems
occur in SSA. Most inland valley bottoms represent the favorable rainfed
wetlands which could be the next frontier for intensification of rice and rice‐based cropping systems. Rice varieties and production technologies devel-
oped for the irrigated ecology can be easily adapted for rice in favorable
rainfed wetlands. Other rainfed wetland (drought‐, flood‐, and drought‐ andflood‐prone) subecosystems are found in riverine shallow, riverine deep,
hydromorphic edges of inland valleys, and mangrove ecologies. They are
highly diverse, with often variable rainfall patterns, adverse soils, and many
abiotic and biotic constraints. In addition, farmers are poor and have to
adapt their cropping practices to the complex risks, potentials, and problems
characteristic of such ecosystems.
a. Rainfed Wetland Rice Area. Of the global rainfed wetland rice area
of 54 million ha, an estimated 2.4 million ha are found in SSA; this is equal to
an estimated 33% of the total rice area in SSA (Table V). The rainfed wetland
rice ecosystem is generally found in the humid and subhumid forest andmoist
savanna zones of SSA. During 1995–2004, Nigeria had the largest area under
rainfed wetland rice (1,159,208 ha), followed by Tanzania (321,127 ha),
Madagascar (215,226 ha), Guinea (113,965 ha), Mozambique (97,114 ha),
Cote d’Ivoire (75,135 ha), and Sierra Leone (66,248 ha) (Table VI).
b. Cropping Systems. The number and choice of rice and other crops
to grow on rainfed wetlands depend on the length of the rainy season and
water availability. Farmers generally grow one crop of rice during the wet
season (WS) and leave the land fallow for the rest of the year. Vegetables,
sweet potato, or taro may be grown on mounds or bunds adjacent to flooded
rice fields (Andriesse and Fresco, 1991). If the rains extend to 5 months or
more, farmers can grow a post‐rice crop such as corn, soybean, vegetables,
or wheat (in highlands). Beans, cowpea, and vegetables are grown with
residual moisture in many inland valley bottoms and swamps of West Africa
(WARDA, 2002).
c. Cultivation Practices and Yields. Farmers sow seeds by broadcasting
on plowed land and fields are bunded to collect rainwater. They weed,
redistribute seedlings to ensure uniform crop stand, and harvest by hand,
but they add few or no purchased inputs (fertilizers or biocides). Harvested
rice is threshed on dry ground in the field or near the house. Rice is dehusked
by hand pounding in a mortar or by small village mills. Rice is stored in
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 79
large baskets for home consumption or in community warehouses for sale.
In West Africa, rice husks are mixed with clay and used for plastering house
walls or for making mud bricks (Murray, 2005).
Small farmers often plant traditional rice varieties and apply very few
external inputs. As a result, they obtain yields from 1 to 2 Mg ha�1 in SSA
in contrast to the world average yield of 2.3 Mg ha�1 for rainfed wetland
rice. However, WARDA (West Africa Rice Development Association, The
Africa Rice Center) has identified 46 promising rice varieties with target
yields of 3 Mg ha�1 or more in rainfed lowlands when moderate inputs are
applied (WARDA, 2002). These high yields have been recorded in some
areas in West Africa (Sakurai, 2006).
d. Production Constraints. Poor water control is a major constraint to
rice intensification in this ecosystem. Many abiotic and biotic constraints
aVect wetland rice production. Abiotic stresses include variable rainfall,
with drought and flood occurrences in the same season. Rainfed wetland
rice is adversely aVected by Fe, Al, and Mn toxicity in wet forest zones
(Buddenhagen, 1986; Sahrawat, 2004a) and in poorly drained soils of coastal
wetlands. Inland salinity and alkalinity are problems in drier and desert
areas (Van Asten et al., 2003). Weeds are the principal biotic constraint,
followed by insect pests (stem borers, African rice gall midge, AfRGM, and
rice bugs) and diseases (blast and brown spot) (WARDA, 1998, 2000). Rice
yellow mottle virus (RYMV) is a major scourge of wetland rice and can
sometimes lead to total crop failure (WARDA, 2000). In addition, rats and
birds are serious problems in all ecosystems.
2. Deepwater and Mangrove Rice Ecosystems
The deepwater ecosystem (Fig. 6) covers several environments where rice
is planted, which is adapted to increasing water depths of 1.0 m or more for
durations of 10 days to 5 months. These rice plants must have the ability to
elongate rapidly to stay above the water surface. ‘‘Floating rice’’ can elon-
gate up to 5 m and form adventitious roots that can absorb nutrients directly
from the floodwater in addition to regular roots grounded in the soil. No
varieties are available that are adapted to rapid or irregular rise of floodwa-
ter or sediment‐laden floodwater that can cover crops for longer than
10 days in some deepwater areas (McLean et al., 2002).
In low‐lying coastal areas, we can diVerentiate perennially fresh, seasonally
saline, and perennially saline tidal wetlands where rice plants are subject to
daily tidal submergence. Plants in tidal lands do not elongate greatly, but
tillering and tiller survival may be reduced in saline soils. In problem soils
(acid sulfate and sodic or alkaline soils), excess water accumulates in fields due
80 V. BALASUBRAMANIAN ETAL.
to poor drainage, but no prolonged submergence occurs (McLean et al., 2002).
Al or Fe toxicity is a serious risk when acid sulfate soils of coastal wetlands
are drained (Sahrawat, 2004a), whereas deep peat soils constrain rice
production in high‐altitude areas.
a. Deepwater and Mangrove Swamp Rice Area. Worldwide, about
11–14 million ha will come under deepwater ecosystems. In SSA, about 0.63
million ha are estimated to be aVected by excess flooding, tidal submergence,
saltwater intrusion, and salinity and acid sulfate soils; these ecosystems cover
an estimated 9% of the total rice area in SSA (Table V). Some parts of the
floodplains of the Niger River, the low‐lying wetlands of Madagascar, and
the poorly drained inland basins of Chad, Guinea, Mali, Niger, and Nigeria
have deep flooding, whereas the low‐lying coastal wetlands of East and West
Africa are aVected by salinity and alkalinity due to seawater intrusion.
Mangrove swamps constitute about 49% of the rice land in Guinea Bissau,
14% in the Gambia, and 13% in Guinea (Defoer et al., 2002).
b. Cropping Systems. Only rice is grown in the rainy season. The type,
depth, and duration of flooding determine the rice varieties grown. Specific
rice varieties with diVerent elongation ability are selected for deepwater
(about 1.0 m) and very deepwater (1–5 m) conditions. When flooded, deep-
water rice varieties can elongate 0.02–0.03 m per day, while floating
rice varieties elongate rapidly up to 0.2 m per day (McLean et al., 2002).
Non‐elongation‐type rice varieties with submergence tolerance (Xu et al.,
2006) are needed for freshwater tidal wetlands where flash floods are of
short duration (less than 2 weeks), whereas salt‐tolerant varieties should be
selected for coastal saline lands, including mangrove swamps. In the drier
mangrove areas (e.g., Casamance in Senegal, Guinea Bissau, and the
Gambia) ofWest Africa, farmers plant rice on ridges tomitigate the problems
of Fe toxicity and salinity (WARDA, 2002). During the dry or winter season,
vegetables or legumes can be grown in deepwater areas. In salt‐aVected areas,
the lands are too dry or saline for cropping in the DS.
c. Cultivation Practices and Yields. Deepwater rice and floating rice
aremainly grown in deepwater areas.With the onset of rains, the land is plowed
and harrowed and dry rice seeds are sown by broadcasting. With the moisture
from rainfall, the seeds germinate and the seedlings start growing. If early
rainfall is regular and adequate for land preparation by puddling, seedlings
raised in nurseries are transplanted.Generally, flooding occurs in the later stages
of plant growth and can last for several months. Crop survival and productivity
depend on the age of the rice crop when inundation starts, the rate of rise of
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 81
the water, and the depth and duration of the flood. Sudden or flash flooding
can completely destroy the crop at any stage. In some areas, the floodwater
may be loaded with sediments that can cover the leaves and obstruct plant
photosynthesis (McLean et al., 2002). Only minimal or no inputs are applied to
deepwater rice crops due to the high uncertainty of harvests. Once mature,
panicles are harvested, dried, and threshed. Farmers go by canoe or wait until
the water recedes to harvest the panicles of floating rice (Murray, 2005).
Tidal rice is cultivated during the rainy season in the coastal wetlands of
East and West Africa. Tidal rice can tolerate submergence or flash floods to
a great extent. Areas with saltwater intrusion from the sea require salinity‐tolerant rice varieties. About 35‐ to 40‐day‐old rice seedlings raised in
nurseries are transplanted in prepared main fields at 3 or more seedlings
per hill (Gibba, 2003). In areas of poorly drained and acid sulfate soils,
seedlings are transplanted on ridges to reduce Al and Fe toxicity (WARDA,
2002). It is recommended to apply 68‐22‐22 kg ha�1 of N‐P‐K to mangrove
rice in the Gambia (Gibba, 2003). Other practices are similar to those for rice
in the rainfed wetland ecosystem.
NonirrigatedWS rice yields range from 0 to 4.0Mg ha�1 depending on the
season, location, and rice type. Floating rice yields are low, 1.0–2.5 Mg ha�1.
Yields of tidal rice vary widely and complete crop failure can occur in salt‐aVected coastal wetlands. Although the mean rice yield is 1.0 Mg ha�1 in
mangrove swamps of the Gambia, improved rice varieties with adequate
fertilization can yield 3.2–3.3 Mg ha�1 in research plots (Gibba, 2003).
WARDA has identified 341 promising rice varieties with target yields of
3 Mg ha�1 or more in mangrove swamps (WARDA, 2002).
d. Production Constraints. Major production problems in deepwater
areas are submergence, salinity, acid sulfate soils, and Fe and Mn toxicity in
coastal wetlands (Sahrawat, 2004a), and peat soils and cold injury at the
seedling stage in high‐altitude areas (Balasubramanian et al., 1995). Farmers
obtain low and extremely variable yields from nonirrigated WS rice crops
due to the use of low‐yielding but adapted traditional varieties, the applica-
tion of few or no inputs, and multiple environmental stresses—soil problems
and unpredictable combinations of drought and flood. There is a lack of
suitable plant types tolerant of submergence, salinity, acidity, Fe and Mn
toxicity, and cold at the seedling stage (in highlands).
3. Irrigated Wetland Rice Ecosystems
Irrigated rice is grown in bunded fields with assured irrigation for one or
more crops per year. Usually, farmers try to maintain 0.05–0.1 m of water
in rice fields. Irrigated rice areas are concentrated mostly in the humid,
82 V. BALASUBRAMANIAN ETAL.
subhumid, semiarid, and high‐altitude tropics of the continent. Dams across
rivers, diversion of water from rivers, or tube wells provide water for irriga-
tion. We can distinguish three types of irrigated rice ecologies in SSA: the
irrigated rice in the arid and semiarid Sahel, the irrigated rice in the humid
forest and savanna zones (Defoer et al., 2002), and the irrigated rice of the
tropical highlands (Balasubramanian et al., 1995).
� Irrigated wetland rice in the Sahel is akin to DS fully irrigated rice in Asia.
In the Sahel, solar radiation is high, water control is good, mechanization
is widespread, pests are less prevalent, and input use is relatively high.
Farmers are organized and infrastructure is well developed. About half of
the irrigated rice area in the Sahel is direct seeded and the rest transplanted.
Only 10% of the area is double cropped (Defoer et al., 2002). Per‐season rice
yields are high, 5–8 Mg ha�1 in research plots and 4–5 Mg ha�1 in farmers’
fields vis‐a‐vis potential yield of 8–12 Mg ha�1 (Haefele and Wopereis,
2004). Mean rice yields in the OYce du Niger in Mali have increased from
2 Mg ha�1 in 1977 to 6 Mg ha�1 in 2002 (Defoer et al., 2002). However, in
the DS, evapotranspiration is high and water consumption is considerably
greater than in the WS. In addition, extreme temperatures limit rice yields
in the WS and DS in some parts of the Sahel.
� Irrigated wetland rice in the humid forest and savanna zones is mostly
transplanted. Irrigation schemes are small and located near main roads
and towns. In the WS, rainfall is the main source of water for crop growth,
and irrigation is used as a supplement during crop establishment and early
crop growth periods as well as during mid‐season dry spells. Both the
potential yields of 5–8 Mg ha�1 per season and the actual yields of around
3 Mg ha�1 per season are relatively lower for irrigated rice in the forest
and savanna zones than in the Sahel (Defoer et al., 2002), mainly due
to lower solar radiation, poorer water control, and higher pest incidence.
In addition, iron toxicity is a constraint in irrigated wetland rice in the
savanna and humid zones of West Africa, where tolerant varieties and
nutrient management showed promising results on a long‐term basis
(Sahrawat, 2004a; Sahrawat et al., 1996).
� Irrigated wetland rice in tropical highlands, in East and Central Africa and
Madagascar, is grown on higher elevation marshlands or inland valley
swamps at 700–900 m above msl (low altitude) and 900–>1600 m above
msl (medium altitude). In well‐developed marshlands, rice land is terraced
and fields are fairly level; small earth dams built across streams and small
rivers at the higher level provide good water control and irrigate from
50 to >1000 ha. There are two seasons: the WS is from July to December
and the DS from January to June. Water supply from rainfall and irriga-
tion is good during the WS and erratic during the DS; flooding may be a
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 83
problem in the WS in undeveloped marshlands. All wetland areas (100%)
are planted to rice in the WS and only 60–70% of the area in the DS; in the
remaining area,WS rice is rotated with wheat, maize, potato, or vegetables.
Low temperature or cold injury often aVects the seedlings in the nursery
or main field in June–July. As elsewhere in SSA, P is the most limiting
nutrient to rice production in highlands. Achievable rice yields are high
(7–9 Mg ha�1) in higher elevation marshlands because of warm days
(20–30�C) and cool nights (10–20�C) and longer duration (150–180
days). Rice yields decrease with increasing OM content: 7.3 Mg ha�1
per season in soils with OM < 40‐g kg�1 soil, 6.1 in soils with OM
between 40‐ and 80‐g kg�1 soil, and 4.7 in soils with OM > 80‐g kg�1
soil (Balasubramanian et al., 1995).
a. Irrigated Wetland Rice Area. Globally, the harvested irrigated rice
area of 79 million ha produces more than 75% of the world’s rice output of
600 million Mg or more. In SSA, an estimated 19.8% of the cultivated rice
area was irrigated (1.42 million ha) during 1995–2004 (Table V). Most of the
WS rice areas are located in the rainforest and moist savanna zones of SSA.
About 60% of the DS irrigated rice area of West and Central Africa is found
in the Sahel, Sudan, and savanna zones (WARDA, 2002). High‐altitudeirrigated rice is found in East and Central Africa andMadagascar. Countries
with significant irrigated rice areas are Madagascar (621,764 ha), Nigeria
(419,288 ha), Mali (80,366 ha), Guinea (49,550 ha), Senegal (39,600 ha), and
Cote d’Ivoire (35,063 ha) (Table VI).
b. Cropping Systems. Rice is the main crop during the rainy season
and the land is left fallow for the rest of the year. The rice–rice‐fallowcropping system is practiced in double‐cropped irrigated wetlands of the
Sahel (Defoer et al., 2002). Rice may be rotated with maize, soybean, or
vegetables in inland valley bottoms of West Africa. In double‐cropped areas
of tropical highlands (East and Central Africa and Madagascar), two rice
crops are grown or WS rice is rotated with potato, wheat, soybean, or
vegetables in the winter–spring season (Balasubramanian et al., 1995).
c. Cultivation Practices and Yields. Rice fields are soaked with water,
plowed, puddled, and leveled before crop establishment. Compost and
animal manure, if applied, are incorporated into the soil during land prepa-
ration. Farmers use hand tools, animal‐drawn implements, or hand tractors
to prepare the land. Large tractors are used only in large public and private
sector irrigation schemes. Most of the WS irrigated rice is transplanted in
84 V. BALASUBRAMANIAN ETAL.
rainforest and savanna zones and coastal plains, whereas 50% of the irri-
gated rice in the Sahel is direct seeded (Defoer et al., 2002). Seedlings are
prepared in nurseries and 20‐ to 30‐day‐old seedlings are manually trans-
planted in main fields. Farmers weed their rice fields manually and apply
some amount of fertilizers and pesticides, depending on availability and cost
in local stores. Herbicide use to control weeds is minimal. At maturity, rice
is harvested, dried, and threshed manually. In the Sahel of West Africa,
mechanization is widespread, especially for land preparation and threshing
(Defoer et al., 2002).
With good water control and crop management, potentially irrigated DS
rice yields in the semiarid Sahel zones can be as high as 8–12Mg ha�1 (Haefele
and Wopereis, 2004). WARDA has identified 21 varieties with target yields
higher than 5 Mg ha�1 for irrigated rice in the Sahel and 426 varieties with
yields higher than 4 Mg ha�1 for irrigated lands in humid and subhumid
forest zones (WARDA, 2002). However, in most of the irrigated areas of
SSA, farmers obtain 2–5 Mg ha�1 due to irregular irrigation, poor soil and
crop management, and inadequate input use (Miezan and Sie, 1997). The gap
between attainable and average farmers’ yields is 2–6 Mg ha�1 in the Sahel,
2–5 Mg ha�1 in the rainforest and savanna zones (Defoer et al., 2002),
and 2–6 Mg ha�1 in the tropical highlands (Balasubramanian et al., 1995).
d. Production Constraints. Many of the irrigated rice production con-
straints are similar to those in Asia: poor land preparation, leveling, and
irrigation management; inadequate drainage leading to the development of
salinity and alkalinity (inland basins and coastal wetlands); poor manage-
ment of production inputs; yield instability due to weeds (in direct‐seededrice), insect pests, and diseases; and deteriorating irrigation infrastructure,
especially in large public irrigation schemes (Defoer et al., 2002). P deficiency
in all soils, N and P deficiency in mineral and slightly organic hydromorphic
soils, Fe toxicity at poorly drained sites, and high OM content in peat soils
limit irrigated rice yields in the tropical highlands (Balasubramanian et al.,
1995; Sahrawat, 2004a; Sahrawat et al., 1996). K or Si deficiency increases
the susceptibility of rice crops to diseases. In addition to blast disease,
Africa‐specific biotic constraints include AfRGM, RYMV, and glume dis-
coloration. Stem borers, rats, and birds are other pests that attack rice in all
ecosystems.
Many large public sector irrigation projects have not been successful due
to a combination of factors (Defoer et al., 2002). As a result, irrigated rice
yields have declined from more than 7 Mg ha�1 at the start of many irri-
gation projects to less than 3 Mg ha�1 after a few years. Smaller farmer‐managed irrigation schemes may be a viable alternative for sustainable
irrigated rice production inmanyAfrican countries. At the same time, rigorous
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 85
socioeconomic studies exploring the causes for failure of large‐scale irrigationschemes are called for.
VI. RICE PRODUCTION CONSTRAINTS IN SSA
Biophysical, management, human resource, and socioeconomic/policy
constraints plague rice farming in SSA.
A. PHYSICAL, BIOLOGICAL, AND MANAGEMENT CONSTRAINTS
Physical, biological, and management constraints vary with rice ecosystems
as discussed above in Section V.
B. HUMAN RESOURCE CONSTRAINTS
Particularly serious is the lack of researchers. As is pointed out by
Evenson and Golin (2003), the ratio of researchers to extension workers is
much lower in SSA than in Asia. This is truly a serious problem because the
lack of profitable technology, but not the lack of extending it, is the most
basic constraint to improving farming eYciency in SSA. Once new profitable
technologies are developed, demand for extension services will increase.
In such a situation, it is expected that capacity enhancement programs for
extension workers, which will have high payoVs, will be undertaken.The lack of education among rice farmers is another major constraint, as
better‐educated farmers are more willing to adopt new technologies (Schultz,
1975). According to the Green Revolution experience in Asia, however, the
role of education becomes less important over time as the new technology
is widely adopted (David and Otsuka, 1994). Given the substitutability
between farmers’ education and extension services, and the lack of education
among farmers in SSA, strengthening the extension system is likely to be an
appropriate strategy once new technologies become available. In the longer
run, the continuous development of new technologies will attract educated
farmers to engage in scientific rice farming and stimulate investments in
schooling of children.
Other human resource‐related constraints are as follows:
� Weak or nonexistent research‐extension‐farmer linkage
� Poor or no farmer organizations
� Lack of public–private partnerships
86 V. BALASUBRAMANIAN ETAL.
� Reduced labor availability due to poor nutrition and/or diseases such as
AIDS, cholera, malaria, bilharzia, and so on.
C. SOCIOECONOMIC AND POLICY CONSTRAINTS
In addition to biophysical and human resource constraints, rice produc-
tion in SSA is aVected by socioeconomic and policy constraints:
� Unfavorable input and output pricing policies at the national level. Low
output prices vis‐a‐vis high and rising input prices reduce profit and
the competitiveness of smallholder farms in local, regional, and global
markets.
� Limited access to credit, inputs (seed, fertilizers, pesticides, implements,
and so on), markets, and market information.
� Poor rural infrastructure and transportation.
This unfavorable price structure reflects the ineYcient marketing systems
in SSA. The establishment of eYcient marketing systems requires trust
between local traders and farmers and between local and urban traders,
because dishonest behavior, such as cheating on product quality and late
delivery, can easily occur in any transaction (Hayami and Kikuchi, 2000).
To prevent such behavior, trust must be developed through long‐termand repeated transactions. Prerequisites for such development are (1) the
improvement of rural infrastructure and transportation systems and (2) the
availability of fertilizer‐responsive‐improved varieties and eYcient tech-
nologies that enhance the profitability of long‐term transactions between
farmers and traders. Experience shows that socioeconomic institutions are
not rigid, but are subject to change as new profitable opportunities arise not
only in Asia (Hayami and Kikuchi, 1982; Hayami and Ruttan, 1985) but
also in SSA (Otsuka and Place, 2001). How such development may take
place and whether any major obstacles to change exist in SSA need to be
analyzed through collaborative research between social scientists and
researchers engaged in the development of new rice technologies.
VII. RICE RESEARCH AND TECHNOLOGYDEVELOPMENT DURING THE PAST 20 YEARS
The two leading international rice R&D institutions—International Rice
Research Institute (IRRI) and Africa Rice Center (WARDA)—complement
each other in developing the rice sector in SSA and bringing concrete
benefits to its rice farmers and consumers. During the past two decades,
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 87
both centers have generated many rice research outputs and technologies
through individual and collaborative research with national agricultural
research institutes (NARIs). These research findings and technologies are
discussed briefly in this section.
A. RICE GERMPLASM, BREEDING, AND VARIETY DEVELOPMENT
1. Plant Types and Traditional Rice Varieties of SSA
The genus Oryza has two cultivated species, O. glaberrima Steud (African
rice) and O. sativa (L.) (Asian rice), and 21 wild taxa of Asian and African
origin. Of the 21 wild rice species, seven originated in SSA and the rest in
Asia. The seven African wild rice species and their characteristics are given
in Table VII. These wild species carry genes for specific traits, for example,
resistance to biotic and abiotic stresses, and these genes are valuable in
interspecific crossing (Khush, 1997).
O. glaberrima is the African indigenous species, propagated from its
original center, the upper‐middle delta of the Niger River, and extended
toward the Senegal, Gambia, Casamance, and Sokoto basins (Carpenter,
1978; Murray, 2005). It has been cultivated for 3500 years (Dembele, 1995;
Porteres, 1956). O. glaberrima has no subspecies. It is adapted to African
environments but prone to lodging and grain shattering when compared
with O. sativa (Sie, 1991). It possesses early plant vigor and resistance to
African stresses such as drought, blast, RYMV, nematodes, and insects
(Adeyemi and Vodouhe, 1996). These qualities are valuable in interspecific
crossing.
O. sativa has two subspecies—indica and japonica—with a continuous
array of intermediates. The many agroecotypes of O. sativa are adapted to
various growing conditions and improved varieties derived from them are
highly productive. The japonica is adapted to rainfed drylands and indica to
the aquatic ecology of SSA. But they are susceptible to many abiotic and
biotic stresses in SSA (Sie, 1991).
O. glaberrima is now being replaced by the Asian rice O. sativa introduced
by European traders into SSA around the sixteenth century (Buddenhagen,
1986; Porteres, 1970). Some farmers grow both Asian and African rice side
by side to meet their varied needs and to tackle adverse conditions. They like
the African rice for its fast early growth that can suppress weeds, its shorter
duration, its resistance to diseases, and the nutty flavor of its grains (Nyanteng
et al., 1986).
Varietal classification is done based on the characteristics of the two
cultivated species and cultural practices in the field—time of inundation,
Table VII
African Wild and Cultivated Rice Species and Their Origin, Distribution, Plant Type, Grain Features, and Valuable Genes for Interspecific Crossing
Species 2n Genome
Origin and
distribution
Plant type and grain
features
Valuable donor
genes References
I. Wild species
O. barthii
(O. breviculata)
A. Chev. and
Roehr. (ancestor
of O. glaberrima)
24 AgAg West African
Savanna and Sahel
(swamps and
waterholes)
An annual grass; self‐fertile;long grains w/awns,
shattering; dormancy
Resistance to GLH,
bacterial blight;
drought avoidance
Khush (1997);
Vaughan (1994)
O. longistaminata
A. Chev. and
Roehr. (possible
ancestor of
O. barthii)
24 AgAg Tropical Africa and
Madagascar
A tall, erect perennial
w/rhizomes; cross‐pollinated; long thin
grains w/awns; shedding
Resistance to
bacterial blight;
drought avoidance
Besancon et al.
(1978); Khush
(1997); Vaughan
(1994)
O. punctata Ktoschy
ex Steud.
24
48
BB
BBCC
East and West
Africa;
Madagascar
(forest and
waterholes)
Both annual and perennial;
long, narrow grains
Resistance to BPH,
leafhopper
Katayama (1990);
Khush (1997);
Vaughan (1994)
O. brachyantha
Chev. and Roehr.
24 FF West, Central and
South‐easternAfrica
A short, slender annual;
small and very narrow
grains w/long awns
Resistance to yellow
SB, leaf folder,
whorl maggot, of
tolerance to
lateritic soil
Khush (1997);
Vaughan (1994)
88V.BALASUBRAMANIA
NETAL.
O. stapfii Roschev. 24 AgAg West Africa A perennial weedy species – Bradenas and Chang
(1966)
O. eichingeri 24
48
CC
BBCC
East, Central and
S.E. Africa;
Sri Lanka
Short, sturdy annual and
perennial; short grains
– Vaughan (1994)
O. schwein furthianaa 48 BBCC Tropical Africa A perennial –
II. Cultivated species
O. glaberrima Steud. 24 AgAg West Africa An annual; dryland erect
photoinsensitive; and
floating photosensitive;
no second or third
branching in panicles;
short grains, shedding
Cultigen; fast early
growth, weed
suppression
Bradenas and Chang
(1966); Khush
(1997); Vaughan
(1994)
O. sativa (L.) 24 AA Asia Indica and japonica
subspecies w/
intermediates; erect to
floating; traditional and
improved
Cultigen; forked
branches in
panicle, no grain
shedding
Khush (1997);
Vaughan (1994)
INCREASIN
GRIC
EPRODUCTIO
NIN
SUB‐SAHARAN
AFRIC
A89
90 V. BALASUBRAMANIAN ETAL.
flood duration, maximum water depth, and level of soil fertility. On the basis
of these criteria, five varietal groups are recognized in SSA:
1. ‘‘Rainfed’’ varieties called ‘‘mountain rice’’ or ‘‘dryland rice’’ are often
cultivated in watershed areas and along forest galleries and in upper parts
of inland valleys where flooding is not common. These plants are low in
tillering and suited to direct seeding.
2. ‘‘Early‐duration erect varieties adapted to a submerged environment’’ are
often found on valley fringes with hydromorphic soils having variable
moisture regimes.
3. ‘‘Season‐erect varieties’’ are often planted on lower parts of the topose-
quence than the previous ones and are subject to flooding (0.5‐ to 0.8‐mwater depth). They are suitable for rainfed wetlands.
4. ‘‘Late‐duration erect varieties’’ are mostly adapted to river floodplains
and deepwater areas.
5. ‘‘Floating’’ varieties capable of surviving floods remain above water level
even at significant water depths (1 to>3 m).O. glaberrima is predominant
in this category.
2. Glaberrima � sativa Crosses and the Development of NERICAVarieties for African Drylands
In 1991, Dr. Monty Jones of WARDA led a team of scientists in a new
breeding eVort to unlock and combine the genetic potential of Asian and
African rice. Key to the eVort was WARDA’s rice Gene Bank with 16,000
rice varieties preserved in cold storage, and duplicated at the International
Institute for Tropical Agriculture (IITA) in Nigeria and at IRRI in the
Philippines. Among the preserved varieties are 1500 O. glaberrima lines.
Jones and his group used molecular biology to overcome sterility, the
main problem in interspecific crossing, and to speed up the breeding cycle.
They made ‘‘wide crosses’’ of the African and Asian rice, then removed the
fertilized embryos by embryo rescue and grew them in artificial media. They
then backcrossed the progeny twice to the Asian parent to recombine the
genetic backgrounds of the two distinctly diVerent species. Backcrossing
allowed the introgression (merging) of useful genes such as the wide, droopy
leaves from the rugged O. glaberrima into the more productive O. sativa
subspecies japonica background. Anther culture helped breeders ‘‘fix’’ prog-
eny lines rapidly and retain recombinant lines—with the combined traits of
both the African and Asian parents (Fig. 7). With conventional breeding
where the progeny of a cross segregates into diVerent plant types, it takes 5–7generations to isolate, purify, and select a line with a desired combination of
genetic traits. For dryland rice, this could mean 5–7 years because usually
only one crop can be grown per year. With anther culture, a line can be
Hybridization scheme for the production of NERICAs
Oryza glaberrima African rice
Oryza sativa Asian rice
O. sativa
O. sativa
F1
BC1F1
BC2F1
BC2F1BC2F6
Anther culture
Ho Fixed lines(new plant type)
NERICA
Pedigree selection
×
×
×
Figure 7 Hybridization scheme for producing the new rice (NERICA) varieties for SSA
drylands.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 91
selected after only one generation, and a new variety developed in 18–24
months. By the mid‐1990s, WARDA scientists were testing the new rice for
Africa (NERICA) in rainfed drylands (WARDA, 2001–2002).
Traits of dryland NERICA varieties: The genetic diVerences of the two
distant species crossed gave NERICA varieties high levels of heterosis or
hybrid vigor for faster growth, higher yield, and more resistance to stresses
than either parent. The NERICA varieties have raised the ‘‘yield ceiling’’ of
dryland rice by 50%, from the current level of 4 Mg ha�1, due to longer
panicles with forked branches bearing 2–3 grains each and more grains per
panicle (Table VIII). The new varieties are taller than O. glaberrima, which
makes harvesting easier—especially if the woman farmer has a baby
strapped to her back.
The shorter duration (90–100 days) of NERICA varieties allows farmers
to grow two crops during one rainy season—one rice crop and a post‐ricedryland crop such as grain or fallow legumes that can smother weeds and
add up to 60 kg of biologically fixed N per hectare. Each hectare of well‐managed rice–legume rotation could save 4 ha of fallow land from clearing.
Projections on the spread of NERICA varieties and rice‐fallow legume
rotations in West Africa indicate that by 2010 the total area of land saved
will be about 15,000 ha (WARDA, 2001).
O. glaberrima is highly tolerant of drought. When faced with drought, the
thin leaves ‘‘roll’’ quickly to retain water and the thin roots grow deep into
the soil to explore water. The total length per gram of O. glaberrima roots
can reach about 150 m in contrast to 100 m for O. sativa roots. When the
rains come after drought, the O. glaberrima recovers faster because the
replacement of the thin leaves and roots requires less water and nutrients.
Table VIII
Improved Traits of the O. glaberrima � O. sativa Crosses (NERICAa)
Traits O. glaberrima parent O. sativa parent
O. glaberrima � O. sativa crosses
(NERICA lines)
Inherited advantages of
NERICA lines
Plant height Short Short Taller than both parents Facilitates harvesting of panicles
without bending
Leaves Droopy, wide Erect Droopy, wide lower leaves at
early vegetative growth and
erect upper leaves at
reproductive phase
Early weed suppression and
higher photosynthesis at
reproductive phase > higher
grain yield
Stems Strong, sturdy Weak, thin Strong Help bear heavy panicles,
no lodging
Tillering Low High High More productive tillers,
higher yield
Panicle type Medium long with no
forked branches
Long with forked
branches
Longer than either parent with
forked branches
Higher yield
Grains per branch Single 3–4 3–4
Grains per panicle Low (75–180) High (100–250) Very high (up to 400)
Grain shattering
at maturity
High No No No loss of grain at harvest
Duration (days) 150–170 120–140 90–100 Higher yield per day, allows
double cropping
Drought tolerance
traits
Thin leaves and long thin
roots (150 m g�1), thin
leaves and roots recover
fast with rains after
drought
Medium thick leaves,
short roots (100 m g�1),
slow recovery after
drought
Thin leaves and long thin roots,
thin leaves and roots recover
fast with rains after drought
Good drought tolerance and
avoidance, fast recovery with
rains after drought
aNERICA, New Rice for Africa.
92V.BALASUBRAMANIA
NETAL.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 93
The NERICAs have inherited the thin leaves and roots of O. glaberrima
(WARDA, 2001–2002, 2002–2003).
3. Breeding of New NERICA Varieties for African Rainfed Wetlands
Rainfed wetland rice in Africa suVers from variable rainfall, unpredictable
drought and flooding, AfRGM, RYMV, and blast. Most of the traditional
wetland rice varieties grown in the region have a narrow genetic base, which
leads to their vulnerability to drought, diseases, and pests. Some diseases,
such as RYMV, are spreading fast in the region because of the predominant
cultivation of susceptible rice varieties. Therefore, Dr. Sie of WARDA and
his partners crossed specific RYMV‐resistant African rice varieties with
popular—but susceptible—Asian rice (O. sativa subspecies indica) varieties.
As can be envisaged, the initial problem was hybrid sterility (infertile
oVspring of the crosses) because the two rice species have evolved separately
over millennia and are so diVerent that often attempts to cross them do not
lead to reliable variety development (WARDA, 2003–2004, 2004–2005).
The sterility problem is greater when we cross the African rice with indica
than with japonica. The sterility blockage was overcome by backcrossing
(crossing the hybrid with an O. sativa parent) to restore fertility. Some of the
progeny combined the best features of both parents: the droopy leaves and
vigorous early growth (associated with weed competitiveness) typical of the
African rice and the high number of spikelets (indicating productivity) of
the Asian rice. A major scientific milestone was achieved when the screening
for resistance to RYMV under artificial infestation showed that the crosses
had successfully transferred resistance to RYMV into some of the progeny.
A new plant type with high yield potential is now available for wetlands,
endowed with resistance to local stresses, particularly to the dreaded
RYMV. The progeny of O. glaberrima and O. sativa subspecies indica
are better adapted to rainfed and irrigated wetland conditions, while those
of O. glaberrima and O. sativa subspecies japonica are more suited to rainfed
dryland conditions in SSA (WARDA, 2003–2004, 2004–2005).
The shuttle‐breeding approach used between WARDA breeders and
national programs facilitated the fast exchange and evaluation of breeding
lines under diVerent conditions, accelerated the selection process and increased
its eYciency, and helped achieve wide adaptability of new plant types. For
example, in Burkina Faso, about 600 new plant‐type lines were tested in the
wetlands of the Banfora research station for 4 years (2000–2003) and the 20
most promising lines were selected based on yield and resistance to stresses,
especially to RYMV. Lines of the new plant type were also evaluated in other
important rice‐growing countries of West Africa—Mali, Burkina Faso, Togo,
and Senegal—and more than 70 promising lines were selected.
94 V. BALASUBRAMANIAN ETAL.
In addition, farmers were involved in the early varietal selection process
through farmer participatory variety selection (PVS). The PVS approach
helped farmers choose varieties that meet their needs and breeders obtain
feedback from farmers regarding their preferences for plant type and grain
characteristics and speed up the fine‐tuning, adoption, and dissemination of
new varieties. The PVS exercise showed clearly that men farmers gave
importance to short growth duration and plant height, whereas women
preferred traits such as good emergence, seedling vigor, and droopy leaves
that indicate weed competitiveness, since they are mostly involved in sowing
and weeding operations.
The three most preferred new plant‐type lines are WAS 122‐IDSA‐1‐WAS‐B‐FKR‐B‐1, WAS 122‐IDSA‐1‐WAS‐2‐FKR‐B‐1, and WAS 122‐IDSA‐1‐WAS‐6–1‐FKR‐B‐1. They have a yield potential of 6–7 Mg ha�1,
good tillering ability, growth duration of 120 days, and an acceptable plant
height; all three varieties showed good resistance to major wetland stresses
and also responded well to N application. Four new wetland NERICA vari-
eties, now oYcially known as the ‘‘WetlandNERICAs,’’ have been released in
Burkina Faso, two in Mali, and three in the Gambia (WARDA, 2003–2004,
2004–2005). The wetland NERICA varieties oVer a powerful new weapon to
rice farmers to manage their complex wetland rice stresses. However, to be
most eVective, they should be used as part of the integrated crop management
(ICM) approach developed by WARDA.
Work continues with genetic engineering and molecular tools to identify
and incorporate traits for various abiotic stresses into elite rice germplasm
and to develop suitable wetland varieties through both interspecific crossing
(crosses between the two cultivated species of rice) and intraspecific cross-
ing (crosses within the species, i.e., between O. sativa varieties). Table IX
provides a summary of ongoing conventional and molecular‐assisted breed-
ing eVorts at IRRI and WARDA to develop rice varieties suitable for
diYcult rice environments in SSA (Gregorio et al., 2006).
4. Development of Improved Irrigated Rice Varieties
Since 1971, IITA has been involved in varietal improvement research for
irrigated rice in Africa and in the national and international varietal‐testingnetwork in the region. The strategy for irrigated rice improvement was to
incorporate resistance to or tolerance of Africa‐specific stresses such as
RYMV, blast, AfRGM, and so on, into promising introductions from Asia
and Latin America (Buddenhagen, 1986). Through this program, IITA has
identified or developed high‐yielding rice varieties suitable for irrigated sys-
tems: ITA 212, ITA 222, and ITA 306, with a duration of 120–130 days, a
Table IX
Priority for Genetic Engineering and Conventional Breeding Approaches for Incorporating
Resistance to/Tolerance of Abiotic Stresses in Rice
Traits for abiotic
stresses
Priority
Comments1 2 3
P deficiency MASa CONb – G � E interactions favor use of MAS
Zn deficiency MAS CON – Genes introgressed from wild species into
O. sativa and MAS under progress
Salinity MAS CON GMRc MAS allows pyramiding genes of QTLs
for diVerent mechanisms of tolerance
Drought MAS CON GMR MAS and GMR under development
through functional genomics
Submergence MAS CON GMR MAS in full implementation with the
identified Sub1A gene
Fe toxicity MAS CON – G � E interactions favor MAS
Al toxicity MAS CON – Genes introgressed from wild species into
O. sativa and MAS under progress
Cold tolerance MAS CON – G � E interactions favor MAS
Elongation
ability
MAS CON – Genes introgressed from wild species into
O. sativa and MAS under progress
aMAS, DNA marker‐aided selection (includes use of linked markers, and candidate genes or
identified genes).bCON, conventional breeding.cGMR, genetically modified rice.
Adapted from Gregorio et al. (2006).
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 95
plant height of 0.97–1.06 m, and a mean yield of 4.7–5.4 Mg ha�1 in national
trials in Nigeria and Cameroon; two Fe‐toxicity‐tolerant varieties, ITA 247
and ITA 249; and two cold‐tolerant varieties, B2161‐C‐MR‐51‐1‐3‐1 and
IR7167‐33‐2‐3 for high‐altitude areas. Donors identified for incorporating
resistance to African stresses were Moroberekan, LAC 23, ITA 235, and CT
19 for RYMV; Cisadane and Eswarakora for AfRGM; and ITA 121 and DJ
12‐539‐2 for stalk‐eyed fly (Masajo et al., 1986).
The NARES introduced a large number of irrigated rice varieties from
abroad—Japan, United States, Thailand, China, Portugal, Spain, Egypt,
and Madagascar. However, most of the current irrigated rice varieties were
introductions of the 1970s through WARDA–INGER (International Net-
work for Genetic Evaluation of Rice) coordinated regional trials in West
Africa (WARDA, 1996). The local irrigated rice‐breeding program initiated
in the mid‐1970 exploited both African and exotic germplasm to develop
improved irrigated rice varieties (Sie, 1994).
96 V. BALASUBRAMANIAN ETAL.
5. Germplasm Exchange by INGER‐Africa
In 1975, IRRI launched the global International Rice Testing Program
(IRTP) for the systematic collection, distribution, and testing of rice genetic
materials. Later, it became INGER, with the objectives of promoting global
exchange, evaluation, and use of improved breeding materials originating
from sources worldwide. INGER‐Africa was part of the global program
until it was transferred to WARDA in April 1997. INGER‐Africa had 16
types of regional rice evaluation trials/nurseries targeted for dryland, irri-
gated, rainfed wetland, and mangrove ecosystems and biological stresses of
blast and RYMV. From 1985 to 1996, INGER‐Africa distributed 3726
nursery sets to African countries. Only a few rice varieties adapted to local
conditions were selected by national breeders from the INGER‐Africa nur-
series due to the mismatch between national capacity and needs and the
supply from the program.
A new germplasm exchange mechanism began in 1991 and was formalized
in 1994 at WARDA to modify certain operational aspects of INGER‐Africa
to make it more eYcient and responsive to national needs in SSA. With the
new approach, nurseries are designed to (1) fit NARES’ needs and avoid
overloading their capacity, (2) provide genetic diversity and variability
for key rice ecosystems, and (3) target a supply of valuable germplasm.
In addition, the new mechanism oVers NARES:
� Improved germplasm from a wide range of sources for national breeding
and direct selection of varieties—nursery entries come from several
countries or institutions in Africa (73%), Asia (19%), and Latin America,
Europe, the United States, and other countries (8%)
� A mechanism to screen their own genetic materials for resistance to or
tolerance of specific stresses at reliable hot spot locations in the region
� The means to test the agronomic stability and adaptability of their elite
varieties in regional multilocation trials
� The means to handle segregating populations—F3 populations nominated
by breeders are grown and harvested in bulk for distribution to NARES
on request for in situ selection and advancement nationally.
With the creation of the WARDA research task forces in 1991, members
of task forces who are also members of the new INGER‐Africa meet every
year to report on their results over the previous year and plan their activities
for the following year. This arrangement has increased interactions among
scientists and improved the rate of return of trial results. The NARES’
share of varieties in regional nurseries increased from <10% in 1985 to
about 60% in 2000.
Although most of these rice germplasm materials were bred or selected in
Africa, their parents came from wide sources: India, Sri Lanka, Bangladesh,
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 97
Mozambique, Indonesia, Philippines, Thailand, Taiwan, China, Colombia,
Argentina, Brazil, Zaire, Sierra Leone, Cote d’Ivoire, Nigeria, Guinea, Mali,
Senegal, the Gambia, and international research centers such as IRRI, IITA,
WARDA, and the International Center for Tropical Agriculture (CIAT).
Thus, despite limited regional resources invested annually in varietal impro-
vement, 197 improved varieties had been released up to 1999, with more
than 122 during the next 5 years (2000–2004). These results involve only
7 out of 17 WARDA member countries and indicate the eYciency and
eVectiveness of the new approach to germplasm exchange and evaluation
in SSA. In addition to germplasm exchange, INGER‐Africa activities include
(degree, group, and in‐country) training of NARES scientists and technicians
in the region.
B. RICE SEED PRODUCTION AND DISTRIBUTION SERVICES
One of the principal strategies for improving food security in SSA is to
strengthen the seed supply sector (FAO, 1997). Unlike fertilizers and pesti-
cides, which are often available and used, particularly on cash crops, good‐quality seeds are rarely used by farmers in SSA. Guaranteeing farmers access
to good‐quality seed can be achieved only if there is a viable seed supply
system to multiply and distribute seeds of improved varieties and if mechan-
isms to assist farmers in emergency situations have been established. As long
as the seed varieties oVered are well adapted to small‐farm environments and
low‐input crop management practices, other inputs such as fertilizers are less
critical compared to the benefits derived from improved seed.
1. Status of Existing Seed Sector Services and Seed Research
To design realistic strategies for the future development of the seed supply
sector in SSA, it is imperative to assess in‐depth the existing seed sector.
There are formal and informal seed sectors in all countries.
Formal seed supply systems include public sector institutions, such as para-
statal (quasigovernment) seed agencies providing seed certification and
quality control, and the private sector. The formal seed production and
supply programs are organized mostly by the public sector and commonly
assisted by donor agencies. Private companies are also involved in the formal
seed sector, especially in Ethiopia, Madagascar, Malawi, Mozambique,
Nigeria, South Africa, Zambia, and Zimbabwe. However, despite these
eVorts, formal seed supply systems currently meet not more than 5–10% of
the seed needs of farmers in the region.
98 V. BALASUBRAMANIAN ETAL.
Most formal seed production projects in SSA failed because of
� The use of the criteria and seed standards of the seed industry of developed
countries, which are diYcult to implement and sustain under developing
country conditions
� The lack of follow‐up of donor‐funded programs to ensure continuity of
activities after a program or project ended
� The concentration of donor‐funded projects exclusively on the formal seed
sector, ignoring the already well‐established informal sector
� The concentration of almost all projects on major crops of commercial
value, neglecting local crops of economic and social importance to small
farmers.
Informal seed supply systems are composed of indigenous strategies to
improve the quality and quantity of seed used by farmers. About 90–95%
of the seed production in almost all countries of SSA is still in the informal
seed supply system.
Seed research: The average investment by SSA countries in agricultural
research is estimated to be less than 3% of gross domestic product (GDP).
Thus, the primary seed research activities in most countries involve variety
development and testing, variety release and registration, variety mainte-
nance, and breeder seed production. Since variety development is time con-
suming and costly, many governments in the region tend to prioritize variety
development research alone according to urgent national needs. As a result,
research strategies are directed to a few cereals and staple food crops only.
2. National Seed Laws and Variety Release Process
Only 25% of the African countries have passed a seed act that stipulates
specific seed regulations that must be satisfied. The remaining 75% of the
countries in SSA do not have legislation governing the production, distribu-
tion, and sale of seeds. In addition, even in most of those countries where a
seed act has been passed, putting the various laws and policies into practice
has been impeded by inadequate enforcement mechanisms and a lack of
logistical, financial, and human resources.
In addition, the variety release process is weak inmany countries, and itmay
take 10 years to release and register a new variety in some countries. The
documentation of new varieties with their origin and agronomic and technical
characteristics is again poor due to the lack of trained staV and resources.
The breeder holds an exclusive right tomultiplication and distribution of seeds.
There is no national strategywith a good awareness campaign to encourage the
production, distribution, and use of good‐quality seeds widely.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 99
C. CROP ESTABLISHMENT
Timely crop establishment is critical for good plant growth and high
yields. In irrigated DS rice areas, optimum planting time can be maintained
by a regular water supply. In rainfed drylands and wetlands, the arrival of
rains decides the time of planting.
Rice is mostly transplanted in irrigatedwetlands and some rainfed wetlands.
Half of the irrigated rice in the Sahel is direct seeded. In transplanted rice areas
of West and Central Africa, research data recommend not to transplant in
August or from November to December to avoid low‐temperature stress
(WARDA, 2002). Good nursery management and transplanting of young
and vigorous seedlings at optimum spacing will increase irrigated rice yields
(Balasubramanian et al., 2005). In mangrove swamps, it is recommended to
plant rice on ridges to improve drainage and to reduce Al and Fe toxicity.
Farmers in Africa prefer direct seeding if suitable methods are available.
In addition to seeding techniques, research must develop technologies to pre-
pare smooth and level seedbeds for direct sowing and control weeds eVectivelyin direct‐seeded rice (Buddenhagen, 1986; Defoer et al., 2002).
In rainfed drylands and wetlands, rice seeds are broadcast, drilled, or
dibbled in prepared dry‐to‐moist soil. All three methods are equally eVectivewhen the optimum seed rate of 50–80 kg ha�1 is used (WARDA, 2002).
In drilled and dibbled rice fields, eVective weed control is possible with
interrow cultivation.
D. NUTRIENT MANAGEMENT
Unlike dryland soils, which are mostly acidic, infertile, and highly prone
to erosion, it is easier to keep soils fertile and stable in wetlands. Naturally,
water and nutrients accumulate in wetlands; in addition, N is made available
through microbial N fixation, P through increased solubility under reduced
conditions, and K and other bases are supplied by rain and irrigation water.
Of the 16 known elements required for plant growth, N, P, K, S, and Zn
are important for rice; Si is becoming deficient in intensively cultivated rice
areas of Asia (e.g., Sri Lanka). Other elements taken up by rice crops are
Ca, Mg, Fe, and Mn. The importance of boron (B) to flooded rice is not
established. To produce 1 Mg ha�1 of grain yield, rice crops remove on
average 19.1 kg ha�1 of N, 2.8 of P, 23.6 of K, 3.9 of Ca, 5.2 of Mg, 0.50 of
Fe, 0.55 of Mn, and 0.03 of Zn. Nutrient uptake by irrigated and dryland
rice is similar for N, Mn, and Zn; irrigated rice removes more P, K, Ca, and
Mg than dryland rice per megagram per hectare of grain yield and the
reverse is true for Fe only (Table X) (Sahrawat, 2000). Dryland rice seems
to be more eYcient in using P for grain production. In this study, the
Table X
Nutrients Removed (in Grain þ Straw) by Irrigated Wetland and Rainfed Dryland Rice Crops
per Megagram per Hectare of Grain Yield in Cote d’Ivoire, West Africa, 1994 and 1996
Nutrient
Nutrients removed Nutrient element harvest index (%)
Irrigated ricea Dryland riceb Irrigated ricea Dryland riceb
N (kg ha�1) 18.9 19.4 57 60
P (kg ha�1) 3.4 2.1 67 71
K (kg ha�1) 28.7 18.5 9 11
Ca (kg ha�1) 4.8 3.0 17 15
Mg (kg ha�1) 6.8 3.6 28 30
Fe (g ha�1) 384 616 48 21
Mn (g ha�1) 561 541 21 8
Zn (g ha�1) 30 25 50 38
aIrrigated rice: Mbe site; Ultisol; cv Bouake 189; 1996 DS; Yield: 6.77 Mg ha�1.bDryland rice; Man site, Alfisol; cv WAB 56–50; 1994 WS; Yield: 3.14 Mg ha�1.
Data derived from Sahrawat (2000).
100 V. BALASUBRAMANIAN ETAL.
nutrient element harvest index [amount in grain/amount in (grain þ straw)]
was the highest for P, followed by N and Zn, indicating the presence of a
higher proportion of these three elements in the grain. About 90% of K, 85%
of Mn, 84% of Ca, 71% of Mg, and 65% of Fe were retained in the rice straw
(Sahrawat, 2000). Thus, the return of straw to rice fields will help reduce the
depletion of K, Ca, and Mg from rice soils.
For any system to be sustainable, the export of nutrients from the system
through harvests must be balanced by nutrient additions. Integrated nutrient
management considers all sources of nutrients interacting in a system.
Essentially, it involves three principles (Smaling et al., 2002):
� Add new nutrients to the system—composts and manures from outside the
field, fertilizers, N fixation in wetland rice fields, and biological N fixation
by legumes
� Save nutrients from being lost from the system—soil erosion control, return
of crop residues, planting deep‐rooted crops to reduce leaching losses
� Recycle nutrients within the system tomaximize productivity and nutrient‐useeYciency.
Using the three principles, several nutrient management approaches
have been developed: ecosystem‐based and toposequence‐based nutrient
management developed by WARDA, Benin; site‐specific nutrient manage-
ment (SSNM) developed by IRRI, Philippines; integrated soil fertility man-
agement (ISFM) developed by the International Fertilizer Development
Center (IFDC)‐Africa, Togo; and integrated plant nutrient systems (IPNS)
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 101
developed by the FAO, United Nations, Rome, Italy. We shall briefly look
at the first three approaches for rice in SSA.
1. Ecosystem‐Based Nutrient Management
Dryland rice ecosystem: Dryland rice soils (Oxisols and Ultisols) are
invariably acidic, infertile, and highly P‐fixing (Sahrawat et al., 2003). The
application of rock‐phosphate (rock‐P) or other P fertilizers is critical before
crops respond to N and other nutrients. Finely ground and moderately to
highly reactive rock‐P can be applied to acid dryland soils (Diamond, 1985).
Rock‐P applied to legume fallows benefits the following rice crops more than
its direct application to rice crops (Somado et al., 2003). Research recom-
mends an application of magnesium limestone at 150–200 kg ha�1 for
slightly acidic soils and 2–3 Mg ha�1 for highly acidic soils before fertilizers
are applied to rice crops. Recommended fertilizer rates for dryland rice
range from 40‐36‐64 to 90‐27‐36 kg ha�1 of N‐P‐K (WARDA, 2002).
P and K are applied at planting, while N is to be applied in 2–3 splits at
basal, mid‐tillering, and panicle initiation (PI) stages. The field must be
weeded before applying N fertilizers. However, farmers apply little or no
fertilizer and thus mine the soil continuously. At least, they should be
encouraged to apply available crop residues, composts, and animal manures
(Erenstein, 2003); improve short‐season fallows by planting legume cover
crops (Akanvou et al., 2001a; Carsky et al., 2001; Somado et al., 2003);
rotate or intercrop rice with legumes (Akanvou et al., 2001b); and mulch
soils with crop residues to preserve soil fertility (Erenstein, 2003).
Rainfed wetland ecosystem: P is deficient in most wetland (Oxisols, Ulti-
sols, Vertisols, and certain Inceptisols) and acid sulfate soils (Sulfaquepts,
Sulfaquents). P availability is controlled by P sorption–desorption kinetics
and solubility rates of various Al, Fe, and Ca phosphates in soils (Diamond,
1985). P adsorption by soils is related to their mineralogy: quartz ¼ Al‐freeOM < 2:1 clays < 1:1 clays < crystalline Fe‐ and Al‐oxides < amorphous
Al‐ and Fe‐oxides (Juo and Fox, 1977). When the soil is submerged, P
availability increases from ferric and aluminum phosphates due to increasing
pH in acid to neutral soils and from calcium phosphates due to decreasing pH
in basic calcareous and alkaline soils (Ponnamperuma, 1972). The incor-
poration of P fertilizers during land preparation or surface application until
28 days after transplanting or dipping of seedling roots into a soil‐P slurry
before transplanting is a good practice for P application to flooded rice
crops. In degraded wetland soils with high P‐fixing capacity, P should be
applied in two splits—at planting and PI (WARDA, 2002).
Recommended N rates for wetland rice range from 40 to 90 kg ha�1,
applied in 2–3 splits (at planting, mid‐tillering, and PI). For rice crops in
102 V. BALASUBRAMANIAN ETAL.
mangrove swamps and tidal wetlands, 40–80 kg ha�1 of N in 3–4 splits and
40 kg ha�1 of P as basal are recommended for they are the most limiting
nutrients in this ecosystem (WARDA, 2002). Rice straw incorporation into
soils may maintain soil K at low‐yield levels; but K fertilizer needs to be
applied along with N and P to get high rice yields (Kyuma et al., 1986).
In addition to NPK, plants require low amounts of S and micronutrients
such as Zn, Fe, and so on. Soil submergence increases the availability of Fe,
Mn, and Mo, but decreases that of S, Zn, and Cu; it also reduces Fe and Mn
toxicity in acid soils, including acid sulfate soils (Neue and Mamaril, 1985).
In addition, the slow decomposition of OM accumulated in flooded soils
(Sahrawat, 2004b) may supply some of the micronutrients to rice plants.
Thus, the need for the application of S and micronutrients to rice crops must
be based on the knowledge of their established deficiency in soils. Nutrient
forms in soils and sources for application, critical values for their deficiency
and toxicity, and application methods are summarized in Table XI.
Irrigated rice ecosystem: Inadequate nutrient application is among the key
constraints to achieving high yields in irrigated rice where fertilizer use is
most profitable (Haefele et al., 2000, 2001, 2002). Although rice varieties and
production technologies, including fertilizer rates, were introduced from
Asia, they were gradually adapted and modified to the local biophysical
and economic conditions (Haefele and Wopereis, 2004). Farmers’ N rates
varied from 37 to 251 kg ha�1, P from 9 to 66 kg ha�1, and K from 0 to
7 kg ha�1 in Senegal and Mauritania, which represent the irrigated rice
systems in the Sahel zone of West Africa. Average grain yields were
4.1–5.6 Mg ha�1 vis‐a‐vis the simulated potential yields of 8.0–10.0 Mg ha�1
for the Senegal River Valley (Table XII). An estimated 46–70% of the applied
N is lost from the soil‐plant‐water system due to variable rates and poor timing
of N application (Haefele and Wopereis, 2004).
Haefele and Wopereis (2004) improved irrigated rice fertilizer recommen-
dations for five AEZs of the Senegal River Valley, based on a simple three‐quadrant model, farmer surveys, on‐station experiments, and simulated
potential yield for the WS and DS. The blanket recommendation of
120‐26‐50 kg ha�1 of N‐P‐K for all five AEZs was changed to 161‐20‐0 kg ha�1
(AEZ I, II) and 135‐20‐0 kg ha�1 (AEZ III, IV, V) for the WS. For the DS,
they recommended 187‐25‐0 kg ha�1 (AEZ I), 161‐20‐0 kg ha�1 (AEZ II),
and 135‐20‐0 kg ha�1 (AEZ III); no DS recommendations were derived for
AEZ IV because of a lack of weather data and for AEZ V because of low‐yield potential in the DS. These AEZ‐based fertilizer recommendations were
embedded into ICM recommendations, with emphasis on weed control and
soil fertility management. In addition to optimum fertilizer use, the selection
of nutrient‐eYcient rice varieties, adequate radiation and temperature, good
water control, and optimum crop management are critical to maximize rice
yields and nutrient‐use eYciency (Balasubramanian et al., 2004).
Table XI
Forms in Soils, Sources, Critical Values for Deficiency and Toxicity, and Application Methods for Essential Nutrients in Rice Cropping
Nutrientsa Occurrence in soil
Critical value:
deficiency Critical value: toxicity Nutrient sourcesb Methods of applicationc
N NH4þ, NO3
�; organicmatter
1.4‐mg m�2 leaf area – Urea, DAP, AS, CAN,
coated urea; organic/
green manures
Soil two to four splits; LCC
based; deep placement;
0.5% urea spray
P Solution P $ Labile
P $ Nonlabile P;
organic P
Bray‐1 P: 7 mg kg�1
soil; Olson‐P:5 mg kg�1 soil
Excess P induces Zn
deficiency in basic soils
Rock‐ or single/double/triple super‐phosphate,DAP
Basal, seedling root‐dippingin soil‐P slurry; OPM‐/APM‐based rates
K Solution K $ Ex.
K $ Non‐ex.K $ Mineral K;
organic K
Ammonium acetate‐K:
20‐cmol kg�1 soil
– KCl, straw or straw
compost
Basal at low yield level; two
splits at high yield level;
OPM‐/APM‐based rates
S SO42�, S�, elemental S;
organic S
Ca(H2PO4)2 S: 9‐mg
kg�1 soil (combined
w/Zn deficiency in low
OM, high‐allophanematerials and oxides,
and sandy soils)
>0.07 mg liter�1 of H2S
(mainly in sandy, peat,
and acid sulfate soils)
Rain or irrigation water;
AS, single super‐phosphate; gypsum,
elemental S
Soil application of
S‐containing materials
Zn Zn in clay minerals;
ZnS; Zn adsorbed
on amorphous
oxides and
mg‐carbonates;humic Zn
1‐N NH4‐acetate (pH4.8)‐Zn: 0.6‐mg kg�1
soil; 0.05‐N HCl‐Zn:1.0‐mg kg�1 soil
Nil ZnSO4, ZnO 25‐kg ha�1 ZnSO4 applied
on soil surface/
floodwater; seedling
root‐dipping in 2% ZnO
solution
Fe Fe3þ, Fe2þ 2 mg liter�1 (dryland
rice only)
300‐mg liter�1 Fe2þ
(aggravated by poor
supply of K, P, Zn)
Fe‐sulfate, Fe‐citrate Foliar spray of 0.1–0.5%
solution; apply fast‐decomposing GM
(continued)
INCREASIN
GRIC
EPRODUCTIO
NIN
SUB‐SAHARAN
AFRIC
A103
Table XI (continued)
Nutrientsa Occurrence in soil
Critical value:
deficiency Critical value: toxicity Nutrient sourcesb Methods of applicationc
Mn Acid soil: 10‐ to 100‐mg
liter�1 Mn2þ in soil
solution; Calcareous:
0.5‐mg liter�1 Mn2þ
DTPA þ CaCl2‐Mn:
<1 mg liter�1
>100‐mg liter�1 Mn2þ Mn‐oxide (flooded rice);
Mn‐sulfate (dryland rice)
Soil; foliar spray; soaking
of seeds in Mn solution
Cu Cu‐oxides, carbonates,silicates, and sulfides;
chelated‐Cu
0.05‐N HCl‐Cu:0.1‐mg kg�1 soil
(mostly in peat soils)
– Cu2O, CuSO4 Soil or spray application of
Cu2O or CuSO4 once
in 3 years
Mo Mo on silicates;
Mo‐sesquioxides;organic Mo; MoO4
2�
soluble ions
0.15‐mg liter�1 soil
solution (rarely
deficient)
– Ammonium molybdate Rarely applied for rice
B H3BO3 or borates;
B‐sesquioxide andillite/vermi‐culiteclay; organic B
Hot water‐B:0.5‐mg kg�1 soil
(rarely deficient)
>5‐mg kg�1 soil (only in
saline, alkaline soils)
B‐rich groundwater
(>2 mg B liter�1);
borates
Rarely applied for rice
(applied for wheat in
rotation with rice)
Si Silicates; clay minerals mainly in acid
sandy soils
– Burnt rice hull; silicates Soil application
aN, nitrogen; P, phosphorus; K, potassium; S, sulfur; Zn, zinc; Fe, iron; Mn, manganese; Cu, copper; Mo, molybdenum; B, boron; Si, silica.bAS, ammonium sulfate; CAN, calcium ammonium nitrate; DAP, diammonium phosphate; KCl, potassium chloride (muriate of potash); rock‐P, rock‐phosphate.cAPM, addition plot method; LCC, leaf color chart; OPM, omission plot method.
Information sources: Balasubramanian et al. (1999), Diamond (1985), Dobermann and Fairhurst (2000), Dobermann et al. (2004), Fairhurst and Witt
(2002), and Neue and Mamaril (1985).
104V.BALASUBRAMANIA
NETAL.
Table XII
Indigenous N Supply, Applied Fertilizer Rates, Grain Yield, and N Recovery EYciency
of Irrigated Rice in Farmers’ Fields in the Senegal River Delta of West Africa,
Based on Field Surveys 1995–1997
Parameter
Sites
Tiagar 1,
Senegal
Tiagar 2,
Senegal
Guede,
Senegal
Location,
Mauritania
Season (WS)a 1995b 1995c 1996 1997
No. of cases 10 10 20 37
Mean INSd 72 60 31 32
Applied N (kg ha�1) 101 80 117 115
Applied P (kg ha�1) 22 15 21 20
Applied K (kg ha�1) 0 7 0 0
Grain yield (Mg ha�1) 4.9 4.1 5.6 4.4
RENe (%) 30 38 44 33
aWS, wet season.bSingle‐rice cropping.cDouble‐rice cropping.dINS, indigenous N supply.eREN, recovery eYciency of applied N (% of applied N recovered in grain þ straw).
Adapted from Haefele and Wopereis (2004).
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 105
For irrigated wetlands with adverse soil conditions such as salinity, alka-
linity, acidity, or Fe and Al toxicity, an integrated approach is needed—by
combining resistant or tolerant rice varieties with balanced plant nutrition
and crop management. For example, ridge planting of rice and the applica-
tion of balanced (N, P, K, Zn) fertilizers can reduce Fe toxicity significantly in
the poorly drained wetlands of rainforests andGuinea savanna zones ofWest
and Central Africa (Sahrawat, 2004a). In semiarid areas, the salinity problem
can be tackled by planting salt‐tolerant rice varieties and applying a full dose
of recommended fertilizers (WARDA, 2002).
2. Toposequence‐Based Nutrient Management in an Inland Valley
Soil fertility problems vary according to position along the inland valley
toposequence. For coarse infertile soils in the upper and middle parts, ferti-
lizers must be combined with soil amendments such as composts, manure, or
rock‐P to optimize rice yields and nutrient‐use eYciency. On the other hand,
soils in the valley bottom are relatively high in clay and OM contents and
water‐holding capacity, but poor in P and N; therefore, they generally
respond to P and N fertilizers, with less additional response to soil amend-
ments. For example, in an inland valley of theGambia, Jobe (2003) noted that
106 V. BALASUBRAMANIAN ETAL.
rice responded mainly to fertilizers but not to soil amendments such as
manure, rock‐P, and phosphogypsum in the valley bottom, whereas rice in
the upper and middle parts benefited most from the application of 50% of
recommended nutrients through fertilizers and the rest through 4Mg ha�1 of
animal manure. Thus, the combined use of organic and inorganic sources
of nutrients is recommended by the ISFM strategy to maximize cereal yields,
especially in nutrient‐poor soils of the upper reaches (Vanlauwe et al., 2002).
3. Site‐Specific Nutrient Management
In Asia, a strategy called SSNM has been developed for precision nutrient
management in irrigated rice (Buresh et al., 2003; Dobermann et al., 2004).
SSNM provides an approach for ‘‘feeding’’ rice crops with nutrients as and
when needed. Farmers dynamically adjust the application and management
of nutrients to crop needs according to location and season. SSNM advocates
� The optimal use of existing indigenous nutrient sources, including crop
residues and manures
� Timely fertilizer application to meet the deficit between rice demand for
nutrients and the supply of nutrients from soil and organic inputs (Fig. 8).
In SSNM, the leaf color chart (LCC) is used to apply N fertilizer as
per the crop’s need during the growing season (real time N management)
(Balasubramanian, 2004; Shukla et al., 2004). Rice leaf color is monitored in
the field with an LCC from 15 days after transplanting to the booting stage
at 7‐ to 10‐day intervals, and N fertilizer is applied whenever the leaf color
falls below the chosen critical value.
Figure 8 SSNM approach: ‘‘apply fertilizers as and when needed to fill the deficit between
crop need and indigenous nutrient supply’’ (source: Buresh et al., 2003).
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 107
An omission plot technique is used for crop need‐based P and K applica-
tion. A P‐omission plot—a plot with no added P and the full rate of other
nutrients—visually demonstrates to farmers the deficit of P. Similarly, a
K‐omission plot—a plot with no added K and the full rate of other nutri-
ents—demonstrates the deficit of K (Buresh et al., 2003; Dobermann et al.,
2004). The diVerence in grain yield between a nutrient‐omission plot and a
full NPK plot is used to make P and K recommendations, using the formula:
‘‘To produce one metric ton of paddy over the yield of the nutrient omission
plot, farmers should apply 15‐ to 20‐kg P2O5 and 30‐kg K2O per ha’’
(Balasubramanian et al., 2003).
Other nutrients such as S and Zn are applied as per local recommendations
at sites deficient in such elements.
E. WATER MANAGEMENT FOR RAINFED AND IRRIGATED AREAS
1. Managing Water for Rainfed Rice Farming
Water is the single most critical constraint in rainfed rice farming. For all
rice varieties, the water requirement is the highest during the reproductive
phase, that is, from PI to heading; therefore, during this period, rice fields
must have enough soil moisture in rainfed drylands or standing water (about
0.05‐m depth) in rainfed or irrigated wetlands. Any moisture deficit (due to
drought) at this critical period will seriously aVect spikelet formation and
grain filling, with drastic reductions in yield or complete crop failure.
For drylands and rainfed wetlands, management options are needed to
tackle extreme drought and flood events. First, the duration of rice varieties
must be matched with the duration of the wet period of the growing season.
Second, the planting date is adjusted in such a way that rainfall is adequate
during the period of maximum water requirement (PI to heading). Third,
minimum tillage and mulching can be practiced in drylands to conserve
moisture at critical periods or to overcome unexpected dry spells (Hatibu
et al., 2000; WARDA, 2002).
RWH is defined as ‘‘concentrating, collecting, and storing rainwater for
diVerent uses at a later time in the same area where the rain falls, or in
another area during the same or later time’’ (Hatibu, 2000). There are two
types of RWH: (1) in situ RWH where rain is captured where it falls and
(2) micro‐ and macrocatchment RWH where the runoV from the catchment
area is collected and stored for use in the same or downstream areas during
the same or later time. Building farm ponds, small reservoirs, or earth dams
across water courses is necessary to store peak floods (Hatibu et al., 2000).
RWH and the development of microcatchment storage structures must be
given top priority in drought‐ and flood‐prone rainfed wetlands. Water from
108 V. BALASUBRAMANIAN ETAL.
farm ponds can be used to provide 1–2 life‐saving irrigations during periods
of drought at critical stages to increase and stabilize rice yields. Once
supplementary irrigation is available, farmers will tend to apply additional
inputs to enhance rice productivity with reduced risks. In addition, farm
ponds will help recharge groundwater in the vicinity, which in turn will
revive rainfall‐dependent streams and increase water availability in rural
areas. If enough pond water is available in the DS, farmers can diversify
into growing small plots of vegetables or other cash crops to diversify their
income sources and improve their diets. Further diversification into poultry
and livestock is possible if enough water is available to take care of them.
Farm ponds can also be used to grow fish.
In flood‐prone areas, using reservoirs to store and regulate flash floods
can minimize crop damage due to sudden floods (Hatibu et al., 2000).
In addition, the provision of drainage is critical to reduce Fe toxicity and
stabilize rice production. It is diYcult to drain fields during periods of
heavy rains. At other times, farmers are reluctant to drain their rice fields
for fear of losing applied N, worries about the arrival of the next rain to
reflood their fields, and increased weed infestation under nonflooded condi-
tions. In saline and mangrove swamps, ridge planting is recommended to
improve drainage and reduce the adverse eVect of excess Fe and Al and salt
(WARDA, 2002).
Developing stress‐tolerant rice varieties will help increase and stabilize
rice yields in poorly drained wetlands with problem soils. IRRI has devel-
oped a breeding line (IR73678‐6‐9‐B) derived form O. sativa cv IR64 �O. rufipogon that can tolerate acid sulfate conditions in coastal wetlands;
Vietnam released it as national variety AS‐996 in 2002. It can be evaluated in
acid sulfate soils of Africa. Genes have been tagged for submergence toler-
ance (Sub1A) (Xu et al., 2006) and other abiotic stresses (Table IX) for
incorporation into popular varieties (Gregorio et al., 2006).
2. Water Use and Water Productivity in Irrigated Rice Systems
Irrigation is the best means of increasing rice production with modern
varieties. Improved water conveyance at the systems level and judicious
water use at the farm level are critical to derive maximum benefits from
irrigation schemes.
Large irrigation projects, especially those in the public sector, have often
failed in Africa because of several factors: the lack of a well developed and
coherent irrigation subsector policy and strategy; high capital and operating
costs (especially for irrigation schemes based on pumping); poor cost‐recoveryand lack of funds for management; poor operation, repair, and maintenance
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 109
of irrigation infrastructure; inadequate farm support services; poor ownership
by users; and low economic viability (FAO‐Aquastat, 2005). Owing to inade-
quate drainage, salinity and alkalinity have developed in many irrigated areas
in the semiarid savanna, Sudan, and Sahel zones of West Africa (Massoud,
1977; WARDA, 1999). Large investments needed to rehabilitate such irriga-
tion schemes and to improve systems level water delivery may not be available
from governments. This is where strong public–private partnerships need to
be developed to encourage initial investments in reconstructing irrigation
infrastructure and providing processing and marketing facilities for the
commercialization of irrigated rice production.
Farmer‐controlled tube‐well irrigation systems are highly successful in
Asia (e.g., in Bangladesh) (Hossain et al., 2006). Such small irrigation
schemes have been developed for rice cultivation in the high‐altitude marsh-
lands of Rwanda and Burundi. It is predicted that Africa will have more
small‐scale irrigation schemes in the next 25 years. The development of small
irrigation schemes, the construction of related rural infrastructure, and the
formation of viable water‐user associations must go hand‐in‐hand to realize
the full potential of these irrigation projects and to improve farmers’ liveli-
hood. In addition, both land and water managements must be integrated to
maintain productivity at a high level and environmental quality at an
acceptable level. We need to consider cropping systems instead of individual
crops for enhancing water productivity. Barker et al. (1998) show how
diVerent practices reduce various components of water losses in irrigated
rice systems (Table XIII).
Salinity management is critical for improving the productivity of irrigated
rice in semiarid areas of West Africa. Among the management practices that
minimize salinity problems, crop selection and rotations, land leveling, and
salt leaching have long‐term eVects, while tillage, fallowing and mulching,
landform and planting method, and the application of manures, fertilizers,
and amendments have only a short‐term eVect (Massoud, 1977). Crops such
as rice that require frequent irrigation can reduce salinity eVectively if
adequate drainage is available. With land leveling and tillage, precautions
must be taken not to bring a salt layer, if present, to the crop root zone; any
tillage that improves internal drainage will help reduce salt accumulation.
Evaporation from exposed soil surface during fallow periods will lead to salt
accumulation, and, in this case, mulching can reduce evaporation and
thus salt accumulation on the surface. Although the application of organic
residues and manures improves soil physical and chemical properties, they
carry some salts with them. Proper fertilizer use will not aVect salinity.
Amendments are added mainly to sodic soils to reduce the exchangeable
sodium percentage and improve permeability and drainage (Massoud, 1977).
In addition, eVective management of waste water and saline water for
irrigation is also important.
Table XIII
Practices for Increasing Water Productivity in Irrigated Rice Farming
Practice Ta Eb S&Pc SROd RCLe
Improved short‐duration rice varieties þImproved agronomic management þChanging schedules to reduce evaporation þReducing water for land preparation þ þ þChanging rice planting practices þ þ þReducing crop growth water þ þ þMaking more eVective use of rainfall þ þWater distribution strategies þ þ þWater recycling and conjunctive use þaT, transpiration.bE, evaporation.cS&P, seepage and percolation.dSRO, surface runoV.eRCL, recycled water use.
Adapted from Barker et al. (1998).
110 V. BALASUBRAMANIAN ETAL.
Among the water‐saving technologies, intermittent irrigation or alterna-
tive wetting and drying (AWD) is the most promising because it oVers highwater productivity coupled with a low penalty on grain yield. In China and
the Philippines, AWD is reported to save from 13 to 30% of irrigation water
at the field level, with no significant reduction in yield (Cabangon et al.,
2001).
The development of rice varieties with short duration, early seedling vigor,
and tolerance of submergence will increase water‐use eYciency. Several
medium‐duration (120–130 days) high‐yielding rice varieties are available
for irrigated wetlands: BG 90‐2 and 380‐2; BKN 7033 and 7167; BR
51‐46‐5; IR8, 22, 46, 54, 58, and 841; ITA 123; Sahel 108, 201, and 202;
TOX 728‐1; and WITA 1, 3, and 7. However, improved very early maturing
rice varieties are needed for irrigated lands in northern Guinea, Sudan
savanna, and Sahel zones and early maturing and Fe toxicity‐tolerant vari-eties for irrigated rice in rainforest zones of SSA (WARDA, 2002). In Asia,
hybrid rice varieties were found to be more adapted to water‐deficit condi-tions because of their early seedling vigor and vigorous root system that favor
the eYcient use of available water (Virmani, 1996). But suitable hybrid rice
varieties have yet to be developed for SSA. Salinity‐tolerant high‐yieldingvarieties such as ITA 212 (FARO 35) and ITA 222 (FARO 36) are available
for irrigated areas with saline soils.
Aerobic rice is another alternative for water‐deficit areas (Atlin, 2005); in
this system, aerobic rice varieties grow in nonflooded soils and yield as high
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 111
as 5–6 Mg ha�1, using 40–50% less water than flooded rice systems. Water
productivity was as high as 0.6–0.8 g grain kg�1 water for aerobic rice
varieties in North China (Yang et al., 2005). Promising aerobic rice varieties
are Apo and Magat for the Philippines (Bouman et al., 2005) and HD 502
and HD 297 for China (Bouman et al., 2006). Research is ongoing for
further improvement of rice varieties, crop and water management, and
sustainability for aerobic rice systems. In SSA, aerobic rice varieties can be
tried in water‐deficit irrigated areas or rainfed wetlands where supplementary
irrigation is available.
F. WEEDS, INSECT PESTS, AND DISEASES AND THEIR MANAGEMENT
1. Weeds, Their Management, and Their Multiple Uses
Weeds are plants out of place. They are the most serious biological
constraint to rice production in SSA (Johnson, 1997). The major rice‐fieldweeds that occur in SSA are listed in Table XIV.
Weed infestation is serious wherever water control is poor as in most
rainfed and poorly managed irrigated areas. Improper land preparation and
poor leveling lead to poor spread of water and more weeds in irrigated rice
fields. Becker and Johnson (1999) observed that Echinochloa species are the
most common weeds in fully irrigated systems and Panicum laxum in poorly
irrigated fields in West Africa. More weeds germinate and compete with rice
under direct seeding than in transplanting conditions (Balasubramanian and
Hill, 2002). When soil is depleted with reduced fallow periods, weeds become
more competitive than rice, especially in drylands. The parasitic weed, Striga
spp. (S. aspera and S. hermonthica), is a major problem in dryland rice
mainly in the savanna zone but not in the forest zone (Adagba et al., 2002).
Rice yield loss due to weeds can be as high as 25–40%, with total crop loss
in extreme cases (WARDA, 1998). Manual weeding is the most common, as
herbicides are beyond the reach of poor farmers and, hence, basically
suitable for the low‐income countries of SSA. In direct‐seeded irrigated
rice trials in the Senegal River delta, weeds had to be controlled until
32 days after seeding (DAS) in the WS and 83 DAS in the DS to get 95%
of the rice yield of weed‐free fields (Johnson et al., 2004). Manual weeding,
however, is time‐consuming and labor‐intensive. An estimated 27–37%
of the total labor input goes for weeding alone (WARDA, 1998). However,
timely hand weeding not only controlled weeds eVectively but also
maintained higher levels of pest predators such as spiders and beetles in
rice fields than in herbicide‐treated plots. Placing the uprooted weeds in piles
within rice fields resulted in larger populations of spiders than any other
method of weed disposal, including removal from fields (Afun et al., 2000).
Table XIV
Major Rice‐Field Weeds Reported from Africaa
Grasses Sedges Broad‐leaf typesParasitic
(dryland rice)
Cynodon dactylon Cyperus diVormis Alternanthera sessilis (L.) Striga asiatica
(L.) Pers.,
Dactyloctenium
aegyptium (L.) Willd.,
Digitaria spp.,
Echinochloa crus‐galli(L.) P. Beauv., E.
colona (L.) Link., E.
stagnina (Retz.) P.
Beauv., E. pyramidalis
L., Eleusine indica (L.)
Gaertn., Imperata
cylindrica (L.)
Raeuschel, Ischaemum
rugosum Salisb.,
Leersia hexandra Sw.,
Oryza sativa (red rice)
(L.), Panicum repens
L., Panicum laxum
Sw., Paspalum
scrobiculatum L.,
P. polystachum R. Br.,
Rottboellia
cochinchinensis
(Lour.) Clayton,
R. exaltata, Setaria
pumila (L.) P. Beauv.,
Spermacoce ruelliae
L., C. happen L.,
C. iria L.,
Fimbristylis
ferruginea (L.)
Vahl., F. pilosa
(L.) Vahl., Pycreus
macrostachyos,
Kyllinga
squamulata
(Thon),
Abildgaardia
hispidula
R. Br. ex DC.,
Aeschynomene indica
L., Ageratum
conyzoides L.,
Amaranthus spinosus L.,
Ammania baccifera L.,
Commelina benghalensis
L., C. diVusa Burm. f.,
Euphorbia heterophylla
L., Diplazium sammatii
(Kuhn), Eclypta
prostrata (L.) L.,
Ludwigia octovalvis
(Jacq.), Trianthema
portulacastrum L.,
Portulaca oleracea L.,
Zaleya pentandra L.
L. Kuntze, Striga
aspera (Willd.)
Benth.
aData sources: Adagba et al. (2002), Johnson (1997), and Rao et al. (2007).
112 V. BALASUBRAMANIAN ETAL.
Developing weed‐competitive rice varieties is one strategy for cost‐eYcient weed control in rice crops. In West Africa, O. glaberrima was
found to be more tolerant of Striga (WARDA, 1998) and other weeds
(Johnson et al., 1998) than O. sativa. As discussed earlier, the dryland
NERICA (a cross between O. glaberrima and O. sativa) varieties are highly
competitive with weeds due to their early seedling vigor, fast canopy devel-
opment, and droopy lower leaves that shade out weeds (Futakuchi and
Jones, 2005; Haefele et al., 2004; WARDA, 2003–2004). The newly devel-
oped wetland NERICA varieties have weed‐suppressing ability similar to
that of their dryland counterparts (WARDA, 2003–2004, 2004–2005). The
second strategy is planted legume fallows, which are reported to smother
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 113
weeds and fix N for the following rice crops. The rice–legume fallow rotation
was more productive than the rice–natural fallow rotation (Akanvou et al.,
2001a; Carsky et al., 2001; WARDA, 1998). The third strategy of rice seed
priming for fast germination and early canopy development in dry direct‐seeded rice and their eVect on weed suppression is yet to be assessed in the
field (WARDA, 1998).
Although weeds compete with rice crops, some of them are useful to
subsistence farmers. For example, weeds such as Echinochloa colona and
E. pyramidalis have multiple uses as fodder, thatching material, and grains;
Paspalum scrobiculatum as fodder, poisonous raw grains, and medicine
(Burkhill, 1994); Setaria pumila as fodder and thatching material; Tri-
anthema portulacastrum as food and medicine; and Zaleya pentandra as
fodder for horses and donkeys (Burkhill, 1985). Other rice‐field weeds such
as Rottboellia exaltata and Spermacoce ruelliae provide fodder or grazing
matter for livestock.
2. Major Insect Pests and Diseases and Their Management
Pink stem borer (Sesamia calamistis) is the main pest of dryland rice,
whereas AfRGM (Orsylia oryzivora) is common in rainfed and irrigated
wetlands. Other pests are striped (Chilo zacconius), white (Maliarpha spp.),
and yellow (Scirpophaga spp.) borers and stalk‐eyed fly (Diopsis macro-
phthalma); grain‐sucking bugs (Aspavia sp., Stenocoris claviformis); case
worm (Nymphula depunctalis); and whorl maggot (Hydrellia sp.) (John
et al., 1986; WARDA, 2002). Incidence of stem borers and stalk‐eyed fly is
severe in all humid and dry zones, while that of grain‐sucking bugs, case
worm, whorl maggot, and AfRGM is more severe in humid forest and
Guinea savanna zones than in the Sudan savanna zone (John et al., 1986).
The major diseases of rice in Africa are blast (caused by Pyricularia oryzae),
glume discoloration (fungal complex: caused by Sarocladium sp. and
Curuvularia sp.), RYMV, sheath rot (caused by Sarocladium sp.), leaf scald
(caused by Rhyncosporium oryzae), sheath blight (caused by Thanetophorus
cucumeris), and bacterial leaf blight (caused by Xanthomonas campestris pv.
oryzae) (John et al., 1986; WARDA, 2002). Other pests such as rodents and
birds attack rice in all ecosystems. Nematodes and termites are a serious
problem in dryland rice in some areas.
Integrated pest management (IPM) is a decision‐support system for the
selection and use of pest control strategies that minimize dependence on
chemical pesticides and improve human health and environmental quality.
Growing a healthy crop is the key to good IPM. Other IPM technologies
for rice are: (1) deployment of pest‐resistant varieties, (2) no early spraying
against leaf folders and thrips, (3) an active barrier system for rat control,
114 V. BALASUBRAMANIAN ETAL.
(4) silica application for blast control, and (5) timely and judicious use
of fast‐acting bio or synthetic pesticides when pest infestation is serious,
threatening the crops.
Developing resistant varieties has been the main focus of research at
WARDA. Two rice varieties, WITA 8 and WITA 9 (WARDA, 2002), and
three wetland NERICAs have been selected for resistance to RYMV
(WARDA, 2003–2004, 2004–2005). Breeders employ biotechnology tools
to incorporate specific genes into high‐yielding rice varieties to make them
resistant to or tolerant of various pests and diseases. For example, two
brown planthopper (BPH ) genes for BPH, six Gm genes for gall midge,
eight Xa genes for bacterial blight, and eight Pi genes for blast have been
tagged at IRRI for incorporation into high‐yielding rice varieties through
MAS. Similar work is ongoing at WARDA to identify and tag resistance
genes from cultivated and wild rice species to deploy against Africa‐specificinsect pests and diseases.
G. GRAIN QUALITY MANAGEMENT: FROM BREEDING TO MILLING
As early as 1985, Buddenhagen (1986) pointed out the importance of
having good grain quality for local rice in Africa. He gave four reasons:
� Rice consumers, especially in urban areas, used to high‐quality and low‐costimported rice will demand the same level of quality and cost‐eVectivenessfrom local rice markets
� The earlier introduced rice varieties in the Sahel of West Africa were high‐quality, long‐grain rice varieties from Southeast Asia
� Although poorly milled and maybe not of long‐grain type, many of the old
land races of West and East Africa have very good taste (some with good
aroma) and cooking quality preferred by local consumers
� African rice traders are as sophisticated as any in the world to exploit
consumer preferences for quality rice.
Buddenhagen (1986) recommended the breeding of medium‐grain types
with good taste and cooking quality such as Sierra Leone’s Ngovie or LAC
23, or Tanzania’s aromatic rice for rainfed drylands and wetlands, and long‐grain types for irrigated rice in SSA. Proper plant nutrition and good water
and pest control during crop growth are important for producing healthy,
well‐filled grains.
Rice farmers in Africa lose 15–50% of the market value of grains because
of improper handling of rice during and after harvest. To reduce grain losses
andmaintain grain quality, rice must be harvested immediately after maturity
(95% mature), threshed soon after harvest, cleaned, and dried to <14%
moisture content (MC). Simple dryers and sealed storage options are
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 115
available for drying and storing the grain properly. Airtight storage is recom-
mended to kill storage pests without pesticides, preserve grain quality, and
maintain seed viability. Rice must be milled at 12–14% MC depending on
grain types in modern mills to maximize head‐rice recovery. IRRI has devel-
oped appropriate and cost‐eVective tools such as a low‐cost moisture meter to
monitor grain moisture while processing, super bags for sealed grain storage,
a rice‐milling chart to guide optimum milling of rice, and a portable grain
quality test kit (IRRI, 2005). An improvement in harvest and postharvest
processing will enhance the market value of rice and thus not only improve
farmers’ income and livelihood but also enhance profit to millers.
H. DIVERSIFICATION OF RICE FARMING SYSTEMS
Generally, rice is not highly profitable in many areas. However, in the
WS, rice is the most suitable crop for wetlands. In the DS, dryland crops
such as grain legumes, vegetables, onion, garlic, watermelon, and cotton
are grown in Asia and Africa to enhance farmers’ income. In a 5‐year(1996–2001) crop rotation trial in the Accra Plain of Ghana, Nyalemegbe
et al. (2003) found that rice–vegetable (rice–okra and rice–eggplant) systems
are more profitable than rice–rice, rice–cowpea, or rice–watermelon systems.
Vegetable yields were low but the prices were high in the DS and vice versa in
the WS. In this study, deep‐rooted crops such as okra and eggplant tapped
nutrients from the subsoil, aerated the soil, and added OM to the system,
while grain legumes such as cowpea fixed N and added crop residues to
improve soil fertility. The rice–rice system reduced sedges considerably, but
not the grasses, due to continuous submergence, while the rice–cowpea
and rice–okra systems reduced noxious water‐loving grass weeds such as
E. crus‐galli and Ischaemum rugosum.
Crop‐livestock integration with a cut‐and‐carry fodder system will
improve cash flow and generate income for dryland farmers (Otsuka and
Yamano, 2005). This system will also improve the use of crop residues as
fodder for animals and animal manure to improve soil fertility in rice fields.
I. ICM FOR RICE
In the literature, there is no universal definition of ICM. In practical terms,
it means good agronomy or good crop management. We use the ICM
approach to promote the combined use of adapted rice varieties and
location‐specific crop management technologies to increase land and water
productivity and profit to farmers in irrigated rice farming (Balasubramanian
et al., 2005; Chandrasekaran et al., 2004). From 2004, the FAO of the United
116 V. BALASUBRAMANIAN ETAL.
Nations has adopted the ICM approach to identify and deploy technologies
for increasing rice productivity in Asia and elsewhere.
ICM has two sets of options: core and location specific. Core options are
those that perform similarly in multiple locations, for example a locally
adapted variety, good‐quality seed, robust young seedlings, crop need‐based nutrient application, and IPM. Location‐specific options are those
that perform diVerently at diVerent locations or that can be practiced only in
certain locations; plant spacing, intermittent irrigation, the use of organic
manure, and so on are location specific. Both types of ICM options have to
be selected to meet farmers’ needs in each location; they must also be
updated periodically to incorporate new research findings as they become
available. Additionally, timely harvest and threshing, proper cleaning and
drying of grain to 14% MC, the use of sealed storage systems, and proper
milling with the help of a rice‐milling chart will reduce grain losses and
increase the market value of milled rice.
Key components of ICM successfully validated in Asia are: (1) the selec-
tion of locally adapted rice varieties, (2) the use of good‐quality seed at low
seed rates, (3) the preparation and planting of young seedlings at the four‐leafstage, (4) transplanting of 1–2 young seedlings per hill in a square pattern
(0.2 m � 0.2 m to 0.25 m � 0.25 m), (5) 2–3 mechanical weedings þ soil
stirring at 10‐day intervals from 15 days after transplanting, (6) intermittent
irrigation during vegetative and grain‐maturing phases and continuous shal-
low flooding from PI to flowering, (7) crop need‐based nutrient management
(SSNM), and (8) IPM.
ICM pilot studies conducted in Asia during 2000–2004 demonstrated that
ICM fields produced 1.1–2.5 Mg ha�1 more rice yield than rice crops under
recommended conventional practices; additional profit due to ICM ranged
from US$115 to US$265 ha�1 per season (Balasubramanian et al., 2005).
Growing healthy crops through ICM will also help reduce pesticide use and
pesticide‐related health risks to farmers. ICM will be especially beneficial to
smallholders with limited land and suYcient labor. Seed producers will
benefit more than grain producers when they adopt the ICM approach.
ICM for irrigated rice in West Africa focuses on land preparation, seed
rates and sowing time, cultivar choice, and crop establishment techniques,
and provides a farming calendar for a given combination of site, sowing
date, cultivar choice, and crop establishment technique (WARDA, 2000).
Additional ICM options are fertilizer rates for specific target yields, weed
and water management techniques, and postharvest techniques (Defoer
et al., 2002). In on‐farm evaluations in Senegal and Mauritania, mean rice
yield increased from 3.8 Mg ha�1 for the farmers’ practice to 5.5 Mg ha�1 for
ICM plots and the average net benefit rose from US$215 to US$525 ha�1,
despite the slightly increased cost for ICM plots. Moreover, the adoption of
ICM options tends to reduce risks (Haefele et al., 2000). The ICM options
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 117
most attractive to farmers were improved fertilizer and weed management,
adapted high‐yielding varieties, and mechanized harvesting and postharvest
technologies. Adding proper land leveling; the preparing and planting
of young, vigorous seedlings; intermittent irrigation with soil stirring; and
improved postharvest management to rice‐ICM recommendations in SSA
may further enhance rice yields and profitability to African farmers, as
evidenced in Asia (Balasubramanian et al., 2005). The available best manage-
ment technologies for rice in diVerent ecosystems of SSA are listed in
Table XV. These technologies can be deployed as a basket of options to
farmers using the ICM approach.
VIII. RICE INTENSIFICATION ISSUES ANDTHOUGHTS FOR THE FUTURE
A. RICE INTENSIFICATION IN RELATION TO VECTOR‐BORNE
HUMAN DISEASES
Farmers consider flooded wetland rice cultivation risky for health reasons—
the fear of contracting wetland‐related human diseases such as malaria
and bilharzia. The distribution of these two diseases in diVerent countries ofSSA is shown in Table V.
Malaria caused by mosquitoes (Anopheles gambiae s.l.) is the single
biggest killer disease in SSA; about half a million children die of malaria
ever year. The severity of malaria incidence depends on the susceptibility of
local people and the preventive measures they use. Research findings indi-
cate that the expansion of rice cultivation to a new area or conversion of
single‐rice to double‐rice cropping will not further increase the incidence of
malaria in the already endemic humid and savanna zones. However, in new
irrigation schemes or with the expansion of existing irrigation schemes to
new areas in the Sahel, the risk of malaria may increase in the short term as
new people and laborers settle the area. As the settlers acquire immunity, the
danger of malaria goes down steeply (WARDA, 1996).
Of the more than 200 million people infected by bilharzia or schistoso-
miasis worldwide, about 10% of them are seriously incapacitated—80% of
which are in SSA. The disease is caused by worms known as schistosomes—
urinogenital schistosomiasis by Schistosoma haematobium and intestinal
schistosomiasis by S. mansoni. Aquatic snails and humans provide alternative
hosts for the worms. The disease is prevalent where people are in contact
with snail‐infested water—slow‐moving rivers and streams and vegetation
banks of lakes. Preventive health measures such as the use of boots and the
destruction of snail‐infested floating vegetation as well as timely treatment of
Table XV
Available Best Management Technologies for Rice in DiVerent Ecosystems of SSA
Practice
Irrigated
Rainfed wetland Mangrove swamps DrylandTransplanted Direct seeded
Variety Short‐duration high‐yielding varieties;
salt‐tolerant forsaline areas
Weed competitive,
early vigor, fast
canopy
development,
nonlodging
Short duration, weed
competitive,
drought‐ and flood‐tolerant
Tolerant of Fe and
Al toxicity,
submergence,
salinity
Weed competitive,
acid‐tolerant, andP‐eYcient
Seed quality Good seed (certified) Good, primed seed
Nursery Good seedbed and
thin sowing
– Good seedbed and thin
sowing
Good seedbed and
thin sowing
–
Seedling quality Young, robust (four‐leaf stage)
– Young, robust, tall Robust, healthy –
Seed rate (kg ha�1)a 20–30 60–80 20–30 for TPR and
60–80 for DSR
20–30 60–100
Landform Good puddling and
leveling
Plowing or puddling
and leveling
Plowing or puddling
and leveling
Plowing, forming
ridges
Plowing, leveling, or
terracing
Planting Two to three seedlings
per hill at 0.2 m �0.2 m spacing
Broadcast or row‐seeded on puddled
or moist soil
Transplanted or
broadcast or row‐seeded on wet soil
Transplanted on
ridges with two to
three rows per
ridge
Broadcast, drilled in
rows, or dibbled
Water/irrigation
managementbIntermittent irrigation
up to PI and then
after flowering;
0.05‐m water level
from PI to flowering
Saturated soil first 10
days; then
intermittent up to PI
and after flowering;
0.05‐m flood from
PI to flowering
Rainwater harvesting
and farm pond,
reservoir, earth dam;
one to two life‐saving irrigations at
critical stages
Good drainage to
mitigate tidal
flooding; ridges to
facilitate drainage
Terracing on slopes;
mulching to
conserve moisture;
contour live
hedges
118V.BALASUBRAMANIA
NETAL.
Nutrient
management
Crop need‐based SSNM; adequate P in saline
soils; S and Zn in deficient soils
Ecosystem and
toposequence based
for wetland rice
Ecosystem based for
mangrove swamps
P before N; rock‐P to
legume fallows,
then rice
Weed control Herbicide, manual Herbicide, manual;
weed‐competitive
variety
Weed‐competitive
variety; manual
weeding
Manual, timely
weeding
Weed‐competitive
variety; mulching;
manual weeding
Pest controlc Resistant variety, IPM
Mechanization Animal power and two‐wheel tractor for land preparation; simple manual seeders; rotating hoes for weeding; simple threshers
and cleaners
Preharvest Drain 2 weeks before harvesting –
Harvest Harvest soon after maturity (95% mature)
Threshing Thresh immediately after harvest; do not leave in the field for drying
Drying, cleaning Clean and dry to <14% moisture content (low‐cost moisture meter to manage grain moisture)
Storage Store in airtight containers (e.g., painted mud pots, metal drums, super bags)
Millingd Mill at right moisture content (12–14% MC) (rice‐milling chart; portable grain quality test kit)
aTPR, transplanted rice; DSR, direct‐seeded rice.bPI, panicle initiation.cIPM, integrated pest management.dMC, moisture content.
INCREASIN
GRIC
EPRODUCTIO
NIN
SUB‐SAHARAN
AFRIC
A119
120 V. BALASUBRAMANIAN ETAL.
infected people with a single dose of an appropriate anthelmintic drug will
control the disease eVectively in all areas, including wetland rice areas
(WARDA, 1999).
B. ENVIRONMENTAL ISSUES RELATED TO RICE INTENSIFICATION IN SSA
Rice intensification in SSA must take preventive measures to protect
the environment and ensure the long‐term sustainability of rice farming.
IRRI has developed five environmental indicators to monitor production,
biodiversity, pollution, land degradation, and water (IRRI, 2004).
� Production: For rice production to be sustainable in the long run, it is
important to preserve the natural resources—land, water, and soil fauna
and flora—that support rice production. Combining balanced fertilizer
use with adequate weed control will enhance rice yields and maintain soil
fertility in rainfed and irrigated wetlands of SSA (WARDA, 1998). Con-
servation agricultural methods such as zero tillage, planted fallows with
legumes, and dryland rice‐cover crop rotations are being developed to
reduce soil erosion, conserve soil moisture, and improve soil fertility in
drylands (Akanvou et al., 2001a; Carsky et al., 2001; Erenstein, 2003;
Somado et al., 2003).
� Biodiversity: Preserving wild rice species and diverse cultivated rice vari-
eties in situ in fields and outside in Gene Banks is critical to guard against
pest outbreaks and to supply desirable genes for future needs. Many of the
hardy O. glaberrima lines and varieties would be extinct had their seeds
not been collected and preserved in Gene Banks, because African farmers
had abandoned them to grow their distant Asian cousins (WARDA, 1999,
2001–2002).
We need careful planning based on scientific research to seek a balance
between the economic return of land conversion and social value of
preserving marshlands. It is argued that the conversion of marshlands to
wetland paddies will reduce biodiversity and release a huge amount of
carbon accumulated under the ground. Thus, only high‐potential man-
grove areas must be identified and developed for rice farming and the
remaining mangrove forests preserved in their natural state to save the
richmangrove biodiversity and protect local communities against cyclones
and tsunami‐induced invasion of coastal lands.
Similarly, there is a danger of the harvesting and irreversible drying of peats
in inland valleys and basins. In addition, acid sulfate soils must be kept
flooded all the time to prevent irreversible drying into a nonusable resource.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 121
� Pollution: Herbicides are eVective in controlling weeds in all systems, but
they have to be applied at the right time and at the right dose to destroy
weeds with minimal impact on the environment. The use of IPM for pest
control and SSNM and the LCC for precision nutrient management in rice
will help reduce pesticide‐ and fertilizer‐related pollution of water sources
(Balasubramanian, 2004; Buresh et al., 2003; Dobermann et al., 2004).
Methane emissions from flooded rice fields are an environmental issue
(IRRI, 2004). An estimated 5–10% of global methane emissions are attrib-
uted to flooded rice systems in Asia (Wassmann et al., 2000). Methane
mitigationmeasures developed during the past 20 years include crop residue
management, the addition of various soil amendments, and midseason
drainage (Wassmann et al., 2000); these measures can be proactively
applied in areas of rice intensification in SSA.
� Land degradation: In irrigated rice areas of the semiarid savanna and Sahel
zones, the danger of soil salinization and alkalization must be addressed
through proper drainage and the application of adequate soil amendments
and nutrients to minimize or prevent these types of land degradation
(Massoud, 1977; WARDA, 1999). Conservation agricultural methods
are needed to prevent soil erosion and soil degradation on drylands.
� Water: Water is the most precious natural resource on Earth, and large
amounts of fresh water are diverted to rice cultivation all over the world.
It is therefore important to optimize water productivity by using water‐saving technologies such as AWD, and water‐eYcient short‐duration and
aerobic rice varieties (Bouman et al., 2005, 2006) and hybrid varieties
(Virmani, 1996). The use of less water does not have to obligate farmers
to use more chemical inputs if we develop suitable row‐seeding and line‐transplanting methods and mechanical weed control options. In rainfed
wetlands, RWH is critical to recharge groundwater, revive small streams
and rivers, stabilize rice yields, and diversify farming and income sources.
C. PREPARING FOR THE IMPACT OF CLIMATE CHANGE
Climate change has become a major concern worldwide. SSA will also
face new burdens due to climate change. In SSA, major symptoms of such
a change include the disruption of normal climate patterns over large areas,
increasing incidences of drought and temperature extremes, heavy flooding
and soil erosion/landslides, and increasing levels of salt stress in both inlands
and coastal areas. For example, in Mali, the rainy season is now too short
for rice and the DS is too hot for potatoes. With time, these problems occur
more frequently and with increasing severity, threatening the livelihoods of
122 V. BALASUBRAMANIAN ETAL.
resource‐poor farmers. Therefore, underlying processes and the short‐ andlong‐term directions of such changes need to be well understood to formulate
appropriate adaptive strategies and policies for farming under the impact of
changing climate. Solutions require multidisciplinary and integrated appro-
aches to develop a combination of improved rice germplasmwith tolerance of
prevailing and anticipated stresses and adapted crop management as well as
mitigation and amendment strategies that could help simultaneously protect
farmers, consumers, and the environment.
D. TECHNOLOGY DELIVERY AND DEPLOYMENT ISSUES
The new knowledge and technologies must be evaluated in producers’
fields and promoted widely for real impact on productivity and livelihood
of the poor. The technology deployment part is as important as the supply
side of the research‐innovation system. It is important to train, equip, and
motivate field and extension staV of the public and private sectors, nongov-
ernmental organizations, and civil society as well as farmer leaders, and to
strengthen the research‐extension‐farmer linkage to eVectively move new
research findings and technologies from research stations to producers.
Training of and technical support to farmers on new technologies, field
demonstrations, the organization of farmers’ days at harvest time, and
encouragement of farmer‐to‐farmer communication are eVective technologydeployment tools. With the advent of information and communication tech-
nology (ICT), we can now reach out to farmers scattered far and wide. IRRI
has developed a Rice Knowledge Bank (www.knowledgebank.irri.org/)—a
virtual digital extension and training information tool for the extension of
rice technologies. The scheme for the ICT‐based knowledge delivery system
shown in Fig. 9 will help to eVectively deploy rice technologies to producers
and collect farmers’ feedback. The development of a region‐specific Rice
Knowledge Bank for SSA is important for the eVective deployment of rice
technologies.
The voluntary organization of farmers into groups is essential to eVectivelycoordinate farming operations and allocate community resources (water,
grazing land), shape supportive farm policy (crop insurance, minimum
support price, and so on), develop processing and value addition enter-
prises, and take up direct marketing. In addition, training, extension, and
delivery of knowledge are more eVective with farmer groups than with
individual farmers. Private sector‐mediated and supported farmer groups
will enable the timely delivery of the latest technologies with required inputs
and technical support; guarantee farm credit through commercial banks;
develop processing, storage, and marketing facilities in strategic locations;
assure buyback of produce from farmers at a predetermined price in each
Riceknowledgebank
Delivery mode Pilot sites
Feedback/experiences/derivativeproducts
Production
Fact sheetsDiagnosticsTraining
resourcesE-learningDatabasesImage library
RuralCyber units- Internet- CD-based
Village 1
HAM radioSMS phone
Mobile ICTvans
Village 2
Village 3
Village n
guides
Figure 9 An example of the ICT‐based knowledge delivery system.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 123
season; and ensure a good‐quality rice supply to national, regional, and
global markets.
E. POLICY SUPPORT FOR RICE INTENSIFICATION IN SSA
It must be emphasized here that the advent of high‐yielding rice varieties
has triggered subsequent changes in supporting policies, such as investments
in irrigation, initiation of credit programs, and the establishment or
strengthening of national research and extension systems in Asia as the
rates of return to such investments increased significantly (Barker and
Herdt, 1985; Hayami and Kikuchi, 1982; Otsuka and Kalirajan, 2005,
2006). The national governments in the major rice‐producing countries of
SSA must design supportive policies for science‐based intensive rice produc-
tion systems based on estimated rates of return to various investments.
In particular, urgent analysis is needed on the benefits of rehabilitation of
irrigation facilities, initiation of farm credit, creation of rural infrastructure
and marketing facilities, and development of public–private partnerships.
The NARIs in many African countries must be revamped and supported to
do relevant research on the emerging issues of intensification of rice farming.
Similarly, the public and private sectors, NGOs, and community extension
services must be trained, equipped, and enabled to provide focused advice
and technical support to rice farmers, small and large. We believe that if
124 V. BALASUBRAMANIAN ETAL.
technological breakthroughs are supported by political will, all these
required changes for a Green Revolution in SSA will take root, as occurred
earlier in Asia.
IX. CONCLUSIONS: CHALLENGES TO ANDOPPORTUNITIES FOR ENHANCING
RICE PRODUCTION IN SSA
The cost of irrigated rice production is high in many SSA countries
(FAO‐CORIFA, 2005), mainly because of the high initial investment in
irrigation infrastructure and the poor operation of many irrigated rice
schemes (FAO‐Aquastat, 2005). In addition, irrigated rice farmers have
not realized the full potential of improved irrigated rice varieties due to
less‐than‐optimum input use and crop management. On the other hand,
rainfed wetland rice production by smallholders is often constrained by
many biotic and abiotic stresses as well as inadequate crop management—
all resulting in low yields, less than 1.5 Mg ha�1. Moreover, postharvest
losses are high and the quality of milled rice is often poor in many countries.
Therefore, the production of locally preferred rice at a competitive price is
the biggest challenge to African farmers.
The most important challenge to developing high‐yielding rice varieties
with acceptable grain quality and resistance to or tolerance of local pests in
rainfed and irrigated wetland ecosystems is being addressed progressively.
Although O. glaberrima Steud (African rice) is native to SSA, Asian rice
(O. sativa L.) introduced during the early sixteenth century by European and
Asian traders is cultivated in many irrigated areas of SSA (Buddenhagen,
1986). Breeders are trying to incorporate tolerance of drought and local
biotic stresses into the introduced high‐yielding‐irrigated rice varieties.
Of course, combining high yield with grain quality is critical to transforming
subsistence agriculture into commercial rice farming.
The second challenge lies in identifying, branding, and promoting high‐quality locally adapted rice varieties in national, regional, and international
markets. Although increasing crop yields is very important, a major problem
faced by all SSA rice producers is to reduce postharvest losses, which
presently account for 15–50% of the market value of production.
We believe that to address these challenges, it is critical to improve the
weak rice R&D capacity in SSA. In Mozambique, for example, only 2–3
national scientists are available to address the problems of 500,000 rice
farmers who grow rice on 200,000 ha of land; in Tanzania, only 23 national
scientists are involved in rice research for 322,000 ha of rice land; and, in
Kenya, only 5 researchers are employed to examine the problems of more
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 125
than 100,000 rice farmers. Even these limited human and financial resources
tend to be spent on irrigated rice schemes.
The third challenge, the most critical one, is the absence of a coherent and
comprehensive policy, plan, and program to tackle the many constraints and
deficiencies of the national rice sector in African countries. It must be
emphasized here that the development of high‐yielding rice varieties and
profitable production technologies is a prerequisite to trigger changes in
supporting policies, such as investments in irrigation, initiation of credit
programs, revamping of national rice R&D systems, and development of
rural infrastructure and market systems for local rice.
Great opportunities exist to increase rice production and strengthen both
household and national food security in SSA. First, national governments
are trying to increase local rice production to reduce rice imports. Simulta-
neously, the donor community has doubled its aid to SSA for reducing
poverty and improving food security and nutrition. In such an environment,
the vastly underexploited rice sector oVers a tremendous opportunity to
substantially increase local food production and improve food security and
farmers’ income and livelihood.
Donors such as the Rockefeller Foundation and the USAID through
Borlaug LEAP (Leadership Enhancement in Agriculture Program) Fellow-
ships are willing to support the training of a large number of new African
plant breeders and crop production scientists to help develop the agricultural
sector on the continent. In addition, the NEPAD’s (New Partnership for
African Development) CAADP (Comprehensive African Agricultural
Development Program) initiative and various other multilateral facilities
can help develop the rice sector in SSA.
The WARDA and IRRI have a wide array of international expertise in
modern rice breeding, biotechnology, and production and postproduc-
tion management. These two centers are joining hands to help improve
the rice sector in SSA. This is a great opportunity for African agricultural
R&D institutions to improve their rice science and technology‐development
capability through focused training and joint development, and the
implementation of collaborative rice R&D projects.
ACKNOWLEDGMENTS
Earlier review of this chapter and critical comments and suggestions by
Dr. K. L. Sahrawat of ICRISAT, India are gratefully acknowledged. The
authors thank Dr. Bill Hardy for editorial assistance and improvement of
the chapter.
126 V. BALASUBRAMANIAN ETAL.
REFERENCES
Abekoe, M. K., and Sahrawat, K. L. (2001). Phosphate retention and extractability in soils of
the humid zone in West Africa. Geoderma 102, 175–187.
Adagba, M. A., Lagoke, S. T. O., Singh, B. N., and Ndahi, W. B. (2002). Potential of
cinosulfuron and CGA152005 seed treatment for control of Striga hermonthica in dryland
rice. Acta Agron. Hung. 50, 7–18.
Adegbola, P. Y., and Singbo, A. G. (2005). Impact de l’importation du riz sur la competitivite et
la rentabilite de la production nationale au Benin. Paper Presented at the WARDA
Workshop on ‘‘Policies and Strategies for Promoting Rice Production in Sub‐SaharanAfrica.’’ November 7–9, 2005. WARDA, Cotonou, Benin.
Adeyemi, P., and Vodouhe, S. R. (1996). Amelioration de la productivite des varietes locales de
Oryza glaberrima Steud. par des croisements intra et interspecifiques avecOryza sativa Linn
in Hybridations interspecifiques au Benin. Report, pp. 159–175. WARDA, Cotonou,
Benin.
Afun, J. V. K., Johnson, D. E., and Russell‐Smith, A. (2000). Weeds and natural enemy
regulation of insect pests in dryland rice: A case study from West Africa. Bull. Entomol.
Res. 89(5), 391–402.
Akanvou, R., Becker, M., Chano, M., Johnson, D. E., Gbaka‐Tcheche, H., and Toure, A.
(2001a). Fallow residue management eVects on dryland rice in three agro‐ecological zonesof West Africa. Biol. Fertil. Soils 31, 501–507.
Akanvou, R., Becker, M., Bastiaans, L., and KropV, M. (2001b). EVects of sowing dates on
cover legumes‐dryland rice intercropping systems in West Africa. Biol. Fertil. Soils 33,
15–21.
Andriesse, W. (1986). Area and distribution. In ‘‘The Wetlands and Rice in Sub‐Saharan Africa:
Proceedings of an International Conference on Wetland Utilization for Rice Production in
Sub‐Saharan Africa’’ (A. S. R. Juo and J. A. Lowe, Eds.), pp. 15–30. November 4–8, 1985.
International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria.
Andriesse, W., and Fresco, L. O. (1991). A characterization of rice‐growing environments in
West Africa. Agric. Ecosyst. Environ. 33, 377–395.
Atlin, G. (2005). A dry vision. In ‘‘Rice Today,’’ Vol. 4, pp. 28–29. IRRI, Los Banos,
Philippines.
Balasubramanian, V. (2004). EYcientNmanagement technologies for irrigated rice inAsia. Poster
Paper Presented at the Fourth International. Crop Science Congress, Brisbane, Australia,
September 26 to October 1, 2004. IRRI, Los Banos, Philippines.
Balasubramanian, V., and Hill, J. E. (2002). Direct seeding of rice in Asia: Emerging issues and
strategic research needs for the 21st century. In ‘‘Direct Seeding: Research Strategies and
Opportunities’’ (S. Pandey, M. Mortimer, L. Wade, T. P. Tuong, K. Lopez, and B. Hardy,
Eds.), pp. 15–39. IRRI, Los Banos, Philippines.
Balasubramanian, V., Rabesan, R., Razafinjara, L., and Ratsimandresy, J. (1995). Rice soil
constraints and fertility management in the highlands of Madagascar. In ‘‘Fragile Lives in
Fragile Ecosystems: Proceedings of the International Rice Research Conference,’’
pp. 313–324. February 13–17, 1995. IRRI, Los Banos, Philippines.
Balasubramanian, V., Ladha, J. K., and Denning, G. L. (Eds.) (1999). ‘‘Resource Management
in Rice Systems: Nutrients,’’ 355 p. Kluwer Academic Publishers Dordrecht, The
Netherlands.
Balasubramanian, V., Ladha, J. K., Gupta, R. K., Mehla, R. S., Singh, B., and Singh, Y. (2003).
Technology options for rice in the rice‐wheat system in South Asia. In ‘‘Improving the
Productivity and Sustainability of Rice‐Wheat Systems: Issues and Impact’’ (J. K. Ladha,
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 127
et al., Eds.), ASA Special Publication 65, pp. 115–147. ASA, CSSA, & SSSA, Madison,
Wisconsin, USA.
Balasubramanian, V., Alves, B., Aulak, M. S., Bekunda, M., Cai, Z., Drinkwater, L., Mugendi,
D., van Kessel, C., and Oenema, O. (2004). Crop, environmental and management factors
aVecting fertilizer nitrogen use eYciency. In ‘‘Agriculture and the Nitrogen Cycle: Assessing
the Impacts of Fertilizer Use EYciency on Food Production and the Environment’’
(A. R. Mosier, J. K. Syers, and J. R. Freney, Eds.), SCOPE Series, Vol. 65, pp. 19–33.
Island Press, Washington, DC.
Balasubramanian, V., Rajendran, R., Ravi, V., Chellaiah, N., Castro, E., Chandrasekaran, B.,
Jayaraj, T., and Ramanathan, S. (2005). Integrated crop management for enhancing yield,
factor productivity, and profitability in Asian rice farms. Int. Rice Comm. (FAO) Newsl.
54, 63–72.
Barker, R., andHerdt, R.W. (1985). ‘‘TheRice Economy ofAsia.’’ The JohnHopkinsUniversity
Press, Baltimore, Maryland, USA. 324 p.
Barker, R., Dawe, D., Tuong, T. P., Bhuiyan, S. I., and Guerra, L. C. (1998). The outlook for
water resources in the year 2020: Challenges for research on water management. Paper
Presented at the 19th Session of the International Rice Commission (IRC), Cairo, Egypt,
September 7–9, 1998. FAO, Rome, Italy.
Becker, M., and Johnson, D. E. (1999). Rice yield and productivity gaps in irrigated systems of
the forest zone of Cote d’Ivoire. Field Crops Res. 60, 201–208.
Becker, M., and Johnson, D. E. (2001). Cropping intensity eVects on dryland rice yield and
sustainability in West Africa. Nutr. Cycl. Agroecosyst. 59, 107–117.
Besancon, G., Bozz, J., Kofi, G., and Second, G. (1978). Genetic diversity of indigenous rice in
Africa. In ‘‘Rice in Africa’’ (I. W. Buddenhagen and G. J. Persley, Eds.), pp. 45–60.
Academic Press, London, New York, San Francisco.
Bouman, B. A. M., Peng, S., Castaneda, A. R., and Visperas, R. M. (2005). Yield and water use
of irrigated tropical aerobic rice systems. Agric. Water Manage. 74, 87–105.
Bouman, B. A. M., Yang, X., Wang, H., Wang, Z., Zhao, J., and Chen, B. (2006). Performance
of aerobic rice varieties under irrigated conditions in North China. Field Crops Res. 97,
53–65.
Bradenas, E. A., and Chang, T. T. (1966). Morpho‐taxonomic studies of Oryza glaberrima
Steud. and its related wild taxa, O. breviligulata A. Chev. et Roehr. and O. stapfii Roschev.
Botanical Magazine of Tokyo 79, 791–798.
Breman, H. (1998). Soil fertility improvement in Africa, a tool for or a byproduct of sustainable
production? African Fertil. Market 11(5), 2–10.
Buddenhagen, I. W. (1986). Strategies and approaches to wetland rice improvement. In ‘‘The
Wetlands and Rice in Sub‐Saharan Africa: Proceedings of an International Conference on
Wetland Utilization for Rice Production in Sub‐Saharan Africa’’ (A. S. R. Juo and
J. A. Lowe, Eds.), pp. 97–106. November 4–8, 1985. International Institute of Tropical
Agriculture (IITA), Ibadan, Nigeria.
Buol, S. W., Sanchez, P. A., Cate, R. B., Jr., and Granger, M. A. (1975). Soil fertility capability
classification: A technical soil classification for fertility management. In ‘‘Soil Management
in Tropical America’’ (E. Bornemisza and A. Alvarado, Eds.), pp. 126–145. North Carolina
State University of Agriculture and Engineering, Raleigh, USA.
Buresh, R. J., Witt, C., Balasubramanian, V., Peng, S., Dobermann, A., and Ladha, J. K.
(2003). The principles of site‐specific nutrient management for rice. Poster Paper Presented
at the IFA‐FAO Agriculture Conference on ‘‘Global Food Security and the Role of
Sustainable Fertilization.’’ March 26–28, 2003. Rome, Italy, IRRI, Los Banos, Philippines.
Burkhill, H. M. (1985). ‘‘The Useful Plants of West Tropical Africa,’’ Vol. 1, Families A–D.
Royal Botanic Gardens, Kew, UK.
128 V. BALASUBRAMANIAN ETAL.
Burkhill, H. M. (1994). ‘‘The Useful Plants of West Tropical Africa,’’ Vol. 2, Families E–I.
Royal Botanic Gardens, Kew, UK.
Cabangon, R. J., Castillo, E. G., Bao, L. X., Lu, G., Wang, G. H., Cui, Y. L., Tuong, T. P.,
Bouman, B. A. M., Li, Y. H., Chen, C. D., and Wang, J. Z. (2001). Impact of alternate
wetting and drying irrigation on rice growth and resource‐use eYciency. In ‘‘Proceedings of
the International Workshop on Water Saving Irrigation for Rice’’ (R. Barker, R. Loeve,
Y. H. Li, and T. P. Tuong, Eds.), pp. 55–79. March 23–25, 2001, Wuhan, China. IWMI,
Colombo, Sri Lanka.
Carpenter, A. J. (1978). The history of rice in Africa. In ‘‘Rice in Africa’’ (I. W. Buddenhagen
and G. J. Persley, Eds.), pp. 3–11. Academic Press, London, New York, San Francisco.
Carsky, R. J., Becker, M., and Hauser, S. (2001). Mucuna cover crop fallow systems: Potential
and limitations. In ‘‘Sustaining Soil Fertility in West Africa’’ (R. J. Carsky et al., Eds.),
SSSA Special Publication No. 58, pp. 111–135. Soil Sciences Society of America,
Wisconsin, USA.
Chandrasekaran, B., Palanisami, K., Balasubramanian, V., Rajendran, R., Nandhagopal, A.,
and Sivakumar, S. D. (2004). Participatory irrigation management for equitable distribu-
tion of water and enhanced rice yield in the Cauvery Delta Zone, Tamil Nadu, India.
Int. Rice Res. Notes 29, 69–70.
Coyne, D. L., Sahrawat, K. L., and Plowright, R. A. (2004). The influence of mineral fertilizer
application and plant nutrition on plant‐parasitic nematodes in dryland and wetland rice in
Cote d’Ivoire and its implications in long‐term agricultural research trials. Exp. Agric. 40,
245–256.
David, C. C., and Otsuka, K. (1994). ‘‘Modern Rice Technology and Income Distribution in
Asia.’’ Lynne Rienner Publishers, Inc., Boulder, Colorado, USA.
Defoer, T., Wopereis, M. C. S., Jones, M. P., Lancon, F., and Erenstein, O. (2002). ‘‘Challenges,
Innovation and Change: Towards Rice‐Based Food Security in Sub‐Saharan Africa.’’
Paper Presented at the 20th session of the International Rice Commission, Bangkok. July
22–25, 2002. FAO, Rome, Italy.
Dembele, Y. (1995). Modelisation de la gestion hydraulique d’une retenue d’irrigation: Appli-
cation au perimetre rizicole de Mogtedo (Burkina Faso). These doctorat option genie rural
et des procedes, ENSA de Rennes (France). 156 p.
Diamond, R. B. (1985). Availability and management of phosphorus in wetland soils in relation
to soil characteristics. In ‘‘Wetland Soils: Characterization, Classification and Utilization,’’
pp. 269–283. IRRI, Los Banos, Philippines.
Dixon, J., Gulliver, A., and Gibbon, D. (2001). ‘‘Farming Systems and Poverty: Improving
Farmers’ Livelihoods in a ChangingWorld.’’ World Bank, Rome, Italy; FAO,Washington,
DC.
Dobermann, A., and Fairhurst, T. (2000). ‘‘Rice: Nutrient Disorders & Nutrient Management.’’
Switzerland: Potash and Phosphate Institute (PPI) and Canada: PPI of Canada (PPIC),
191 p. International Rice Research Institute (IRRI), Los Banos, Philippines.
Dobermann, A., Witt, C., and Dawe, D., (Eds.) (2004). ‘‘Increasing the Productivity of Intensive
Rice Systems Through Site‐Specific Nutrient Management.’’ 410 p. Science Publishers
Inc., and IRRI, Enfield, NH (USA) and Los Banos (Philippines).
Erenstein, O. (2003). Smallholder conservation farming in the tropics and sub‐tropics: A guide
to the development and dissemination of mulching with crop residues and cover crops.
Agric. Ecosyst. Environ. 100, 17–37.
Evenson, R. E., andGolin, D. (2003). ‘‘CropVariety Improvement and its EVect on Productivity:
The Impact of International Agricultural Research.’’ CABI Publishing, Wallingford, UK.
Ezedinma, C. I. (2005). Impact of trade on domestic rice production and the challenge of self‐suYciency in Nigeria. Paper Presented at the WARDA Workshop on ‘‘Policies and
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 129
Strategies for Promoting Rice Production in Sub‐Saharan Africa.’’ November 7–9, 2005.
WARDA, Cotonou, Benin.
Fairhurst, T. H., and Witt., C. (Eds.) (2002). ‘‘Rice: A Practical Guide to Nutrient Manage-
ment.’’ Switzerland: Potash and Phosphate Institute (PPI) and Canada: PPI of Canada
(PPIC); 87 p. þ annex 45 p. International Rice Research Institute (IRRI) Los Banos,
Philippines.
FAO (Food and Agriculture Organization of the UN) (1977). FAO‐UNESCO soil map of the
world, 1:5,000,000, Vol. 4, UNESCO, Africa, Paris, France; FAO, Rome, Italy.
FAO (Food and Agriculture Organization of the UN) (1997). Report of the World Food
Summit. November 13–17, 1996. FAO, Rome, Italy.
FAO (Food and Agriculture Organization of the UN) (2006a). FAO‐STAT data base, last
accessed on August 15, 2006. FAO, Rome, Italy.
FAO (Food and Agriculture Organization of the UN) (2006b). www.fao.org/hivaids/ (accessed
on August 30, 2006). FAO, Rome, Italy.
FAO‐Aquastat (2005). www.fao.org/WAICENT/FAOINFO/AGRICULT/AGL/aglw/aquastat/
countries/index.stm (accessed May 10, 2006). FAO, Rome, Italy.
FAO‐CORIFA (2005). Country Rice Facts (accessed on June 5, 2006). www.fao.org/ag/agp/
AGPC/doc/riceinfo/Corifa.htm. FAO, Rome, Italy.
Farm‐Africa (2004). ‘‘Reaching the Poor—ACall toAction: Investment in SmallholderAgriculture
in Sub‐Saharan Africa.’’ Farm‐Africa HarvestHelp, and Imperial College, London, UK.
Fernandez, P., Oliver, R., and Diatta, S. (2000). Changes in organic matter of a ferrallitic
tropical soil degraded by cropping systems: The case of southern Senegal. Acid Soil Res.
Rehabil. 14, 137–150.
Futakuchi, K., and Jones, M. P. (2005). Vegetative growth characteristics related to weed‐competitiveness in interspecific Oryza sativa � O. glaberrima progenies in wetland. Jpn.
J. Crop Sci. 74(Extra issue 2), 260–261.
Gibba, A. (2003). EVect of planting density on the grain yield of two local mangrove swamp rice
varieties. In ‘‘Proceedings of the Second Biennial Regional Rice Research Review’’
(S. Sanyang, A. Ajayi, and A. A. Sy, Eds.), pp. 11–14. WARDA, Cotonou, Beninpp.
Gregorio, G. B., Sie, M., Ndjiondjop, M. N., Tsunematsu, H., Wissua, M., Atlin, G., Brar,
D. S., Singh, R. K., Bennet, J., Ismail, A. M., andMackill, D. J. (2006). ‘‘Designing Abiotic
Stress‐Tolerant Rice Varieties for the Rainfed Wetlands of Sub‐Saharan Africa.’’ Paper
Presented at the African Rice Congress Dar es Salaam, Tanzania July 31 to August 4, 2006.
WARDA, Cotonou, Benin.
Haefele, S. M., and Wopereis, M. C. S. (2004). Combining field and simulation studies to
improve fertilizer recommendations for irrigated rice in the Senegal River Valley.
In ‘‘Increasing the Productivity of Intensive Rice Systems Through Site‐Specific Nutrient
Management’’ (A. Dobermann, C. Witt, and D. Dawe, Eds.), pp. 265–286. Science Publish-
ers, Inc., and IRRI, Enfield, NH (USA) and Los Banos (Philippines).
Haefele, S. M., Johnson, D. E., Diallo, S., Wopereis, M. C. S., and Janin, I. (2000). Improved
soil fertility and weed management is profitable for irrigated rice farmers in Sahelian West
Africa. Field Crops Res. 66, 101–113.
Haefele, S. M., Wopereis, M. C. S., Donovan, C., and Maubuisson, J. (2001). Improving produc-
tivity and profitability of irrigated rice production in Mauritania. Eur. J. Agron. 14, 181–196.
Haefele, S. M., Wopereis, M. C. S., and Wiechmann, H. (2002). Long‐term fertility experiments
for irrigated rice in the West African Sahel: Agronomic results. Field Crops Res. 78,
119–131.
Haefele, S. M., Johnson, D. E., Mbodj, D., Wopereis, M. C. S., andMiezan, K. M. (2004). Field
screening of diverse rice genotypes for weed competitiveness in irrigated wetland
ecosystems. Field Crops Res. 88, 29–46.
130 V. BALASUBRAMANIAN ETAL.
Hatibu, N. (2000). Introduction. In ‘‘Rainwater Harvesting for Natural Resource Management:
A Planning Guide for Tanzania’’ (N. Hatibu and H. F. Mahoo, Eds.), pp. 1–5. RELMA/
Sida, ICRAF House, Nairobi, Kenya.
Hatibu, N., Mahoo, H. F., and Kajiru, G. J. (2000). The role of RWH in agriculture and natural
resource management: From mitigating droughts to preventing floods. In ‘‘Rainwater
Harvesting for Natural Resource Management: A Planning Guide for Tanzania’’
(N. Hatibu and H. F. Mahoo, Eds.), pp. 58–83. RELMA/Sida, ICRAF House, Nairobi,
Kenya.
Hayami, Y., and Kikuchi, M. (1982). ‘‘Asian Village Economy at the Crossroads.’’ Johns
Hopkins University Press, Baltimore.
Hayami, Y., and Kikuchi, M. (2000). ‘‘A Rice Village Saga.’’ Macmillan, Houndmills.
Hayami, Y., and Ruttan, V. W. (1985). ‘‘Agricultural Development: An International Perspec-
tive.’’ Johns Hopkins University Press, Baltimore.
Hossain, M. (2006). Rice in Africa: Can rice help reduce hunger and poverty in sub‐SaharanAfrica? Rice Today 5(1), 41.
Hossain, M., Bose, M. L., and Mustafi, B. A. A. (2006). Adoption and productivity impact of
modern rice varieties in Bangladesh. Dev. Econ. 44, 149–166.
IAC (Inter Academy Council) Report. (2005). ‘‘Realizing the Promise and Potential of African
Agriculture: Executive Summary,’’ pp. xvii–xxx. IAC, Amsterdam, The Netherlands.
IRRI (International Rice Research Institute) (2004). ‘‘IRRI’s Environmental Agenda – an
Approach Toward Sustainable Development,’’ 38 p. IRRI Los Banos, Philippines.
IRRI (International Rice Research Institute) (2005). ‘‘Rice Fact Sheets—Postharvest Manage-
ment: Harvesting, Drying, Storage, and Grain Quality,’’ www.knowledgebank.irri.org/
factsheets/HowToGrowRice/default.htm (accessed on May 25, 2006). IRRI, Los Banos,
Philippines.
IRRI (International Rice Research Institute) (2006). ‘‘World Rice Statistics,’’ www.irri.org/
science/ricestat/index.asp (accessed on May 20, 2006). IRRI, Los Banos, Philippines.
Jobe, B. (2003). Integrated nutrient management for increased rice production in the inland
valleys of the Gambia. In ‘‘Proceedings of the Second Biennial Regional Rice Research
Review’’ (S. Sanyang, A. Ajayi, and A. A. Sy, Eds.), pp. 35–41. WARDA, Cotonou, Benin.
John, V. T., Alam, M. S., and Thottapilly, G. (1986). Diseases and insect pests of wetland rice in
tropical Africa. In ‘‘The Wetlands and Rice in Sub‐Saharan Africa: Proceedings of an
International Conference on Wetland Utilization for Rice Production in Sub‐SaharanAfrica’’ (A. S. R. Juo and J. A. Lowe, Eds.), pp. 141–150. November 4–8, 1985. International
Institute of Tropical Agriculture (IITA), Ibadan, Nigeria.
Johnson, D. E. (1997). ‘‘Weeds of Rice in West Africa.’’ 312 p. WARDA, Bouake, Cote
d’Ivoire.
Johnson, D. E., Dingkuhn, M., Jones, M. P., and Mahamane, M. C. (1998). The influence of
rice plant type on the eVect of weed competition on Oryza sativa and Oryza glaberrima.
Weed Res. 38, 207–216.
Johnson, D. E., Wopereis, M. C. S., Mbodj, D., Diallo, S., Powers, S., and Haefele, S. M.
(2004). Timing of weed management and yield losses due to weeds in irrigated rice in the
Sahel. Field Crops Res. 85, 31–42.
Juo, A. S. R., and Fox, R. L. (1977). Phosphate sorption characteristics of some bench‐mark
soils of West Africa. Soil Sci. 124, 370–376.
Katayama, T. C. (1990). ‘‘Considerationson Distribution of Cultivated Rice in Africa,’’
Occasional Papers, Vol. 18, pp. 193–200. Kagoshima University Research Center, South
Pacific.
Kawaguchi, K., and Kyuma, K. (1977). ‘‘Paddy Soils in Tropical Asia: Their Material Nature
and Fertility,’’ 258 p. Hawaii University Press, Hawaii, USA.
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 131
Khush, G. S. (1997). Origin, dispersal, cultivation and variation of rice. Plant Mol. Biol. 35,
25–34.
Kijima, Y., Serunkuuma, D., and Otsuka, K. (2006). How revolutionary is ‘‘NERICA revolu-
tion’’? Evidence from Uganda. Dev. Econ. 44, 252–267.
Kyuma, K., Kasaki, T., and Juo, A. S. R. (1986). Evaluation of the fertility of the soils. In ‘‘The
Wetlands and Rice in Sub‐Saharan Africa: Proceedings of an International Conference on
Wetland Utilization for Rice Production in Sub‐Saharan Africa’’ (A. S. R. Juo and
J. A. Lowe, Eds.), pp. 43–58. November 4–8, 1985. International Institute of Tropical
Agriculture (IITA), Ibadan, Nigeria.
Lancon, F., and Erenstein, O. (2002). Potential and prospects for rice production inWest Africa.
Paper Presented at Sub‐Regional Workshop on ‘‘Harmonization of Policies and Coordina-
tion of Programs on Rice in the ECOWAS Sub‐Region.’’ February 25–28, 2002. Accra,
Ghana.
Masajo, T. M., Alluri, K., Abifarin, A., and Janakiram, D. (1986). Breeding for high and stable
yields in Africa. In ‘‘The Wetlands and Rice in Sub‐Saharan Africa: Proceedings of an
International Conference on Wetland Utilization for Rice Production in Sub‐SaharanAfrica’’ (A. S. R. Juo and J. A. Lowe, Eds.), pp. 107–114. November 4–8, 1985. Interna-
tional Institute of Tropical Agriculture (IITA), Ibadan, Nigeria.
Massoud, F. I. (1977). Basic principles for prognosis and monitoring of salinity and sodicity. In
‘‘Proceedings of the International Conference on Managing Saline Water for Irrigation,’’
pp. 432–454. August 16–20, 1976. Texas Technical University, Lubbock, Texas, USA.
McLean, J. L., Dawe, D. C., Hardy, B., and Hettel, G. P., (Eds.) (2002). ‘‘Rice almanac,’’
3rd ed., 253 p. IRRI, Los Banos, Philippines; CIAT, Cali, Colombia; FAO, Rome, Italy.
Miezan, K. M., and Sie, M. (1997). Varietal improvement for irrigated rice in the Sahel.
In ‘‘Irrigated Rice in the Sahel: Prospects for Sustainable Development’’ (K. M. Miezan,
M. C. S. Wopereis, M. Dingkuhn, J. Deckers, and T. F. Randolf, Eds.), pp. 443–456.
WARDA, Cotonou, Benin.
Murray, S. S. (2005). ‘‘The rise of African rice farming and the economic use of plants in the
Upper Middle Niger Delta (Mali).’’ Ph.D. Thesis, University of Wisconsin‐Madison. 518 p.
Madison, Wisconsin, USA.
Neue, H. E., and Mamaril, C. P. (1985). Zinc, sulfur, and other micronutrients in wetland
soils. In ‘‘Wetland Soils: Characterization, Classification and Utilization,’’ pp. 307–319.
IRRI, Los Banos, Philippines.
Nyalemegbe, K. K., Oteng, J. W., Darkwa, E. O., and Oti‐Boeteng, C. (2003). Comparative
performance of diVerent rice‐based cropping patterns in the Vertisols of the Accra Plains
of Ghana. In ‘‘Proceedings of the Second Biennial Regional Rice Research Review’’
(S. Sanyang, A. Ajayi, and A. A. Sy, Eds.), pp. 18–26. WARDA, Cotonou, Benin.
Nyanteng, V. K., Samake, M., and Longabough, S. (1986). Socio‐economic study of rice
farming in Mali: The household, farm, labor characteristics and constraints. Occasional
Paper No. 8. WARDA, Monrovia, Liberia.
Oldeman, I. R., Hakkeling, R. T. A., and Sambrock, W. G. (1991). ‘‘World Map of the Status of
Human‐Induced Soil Degradation.’’ ISRIC and UNEP, Wageningen, The Netherlands.
Otsuka, K., and Kalirajan, K. P. (2005). An overview and prospects for a green revolution in
Sub‐Saharan Africa. J. Agric. Dev. Econ. 2, 1–6.
Otsuka, K., and Kalirajan, K. P. (2006). Rice green revolution in Asia and its transferability to
Africa: An introduction. Dev. Econ. 44, 2107–2122.
Otsuka, K., and Place, F. (2001). ‘‘Land Tenure and Natural Resource Management: A Com-
parative Study of Agrarian Communities in Asia and Africa.’’ Johns Hopkins University
Press, Baltimore.
Otsuka, K., and Yamano, T. (2005). Green revolution and regional inequality: Implications of
Asian experience for Africa. In ‘‘The African Food Crisis: The Lessons from the Asian
132 V. BALASUBRAMANIAN ETAL.
Green Revolution’’ (G. Djurfeldt, H. Holmen, M. Jistrom, and R. Larsson, Eds.), CABI
Publishing, Wallingford, UK.
Plowright, R. A., and Hunt, D. J. (1994). Plant parasitic nematodes of dryland, hydromorphic
and inland swamp rice ecosystems in Cote d’Ivoire. Afro‐Asian J. Nematol. 4, 61–67.
Ponnamperuma, F. N. (1972). The chemistry of submerged soils. In ‘‘Advances in Agronomy’’
(N. C. Brady, Ed.), pp. 29–97. Academic Press, New York, USA.
Porteres, R. (1956). Taxonomie agrobotanique des riz cultives O. sativa linne et O. glaberrima
Steudel. Compilations d’articles du JATBA. Museum National d’Histoire Naturelle, Paris,
France.
Porteres, R. (1970). Primary cradles of agriculture in the African continent. In ‘‘Papers in
African Prehistory’’ (J. D. Fage and R. A. Oliver, Eds.), pp. 43–58. Cambridge University
Press, UK.
Rao, A. N., Mortimer, A. M., Johnson, D. E., Sivaprasad, B., and Ladha, J. K. (2007). Weed
management in direct‐seeded rice. Advan. Agron. (in press).
Rosegrant, M. W., Sarah, A. C., Weibo, L., Timothy, B. S., and Rowena, A. V. (2007). Looking
ahead: Long‐term prospects for Africa’s agricultural development and food security. 2020
Discussion Paper No. 41. 60 p. International Food Policy Research Institute (IFPRI)
Washington, DC.
Sahrawat, K. L. (2000). Elemental composition of the rice plant as aVected by iron toxicity
under field conditions. Commun. Soil Sci. Plant Anal. 31(17–18), 2819–2827.
Sahrawat, K. L. (2004a). Iron toxicity in wetland rice and the role of other nutrients. J. Plant
Nutr. 27, 1471–1504.
Sahrawat, K. L. (2004b). Organic matter accumulation in submerged soils. Adv. Agron. 81,
169–201.
Sahrawat, K. L., Mulbah, C. K., Diatta, S, DeLaune, R. D., Patrick, W. H., Jr., Singh, B. N.,
and Jones, M. P. (1996). The role of tolerant genotypes and plant nutrients in the manage-
ment of iron toxicity in wetland rice. J. Agric. Sci. Cambridge 126, 143–149.
Sahrawat, K. L., Jones, M. P., Diatta, S., and Sika, M. (2003). Long‐term phosphorus fertilizer
eVects on phosphorus uptake, eYciency, and recovery by dryland rice on an Ultisol.
Commun. Soil Sci. Plant Anal. 14(7 and 8), 999–1011.
Sakurai, T (2006). Intensification of rainfed wetland rice production in West Africa: Present
status and potential green revolution. Dev. Econ. 44, 232–251.
Sanchez, P., and Buol, S. W. (1985). Agronomic taxonomy for wetland soils. In ‘‘Wetland
Soils: Characterization, Classification and Utilization,’’ pp. 207–227. IRRI, Los Banos,
Philippines.
Sanchez, P., and Swaminathan, M. S. (2005). Cutting world hunger in half. Science 307,
357–359.
Schultz, T. W. (1975). The value of ability to deal with disequilibria. J. Econ. Literature 13,
827–846.
Shukla, A. K., Ladha, J. K., Singh, V. K., Dwivedi, B. S., Gupta, R. K., Sharma, S. K.,
Balasubramanian, V., Singh, Y., Padre, A. T., and Yadav, R. L. (2004). Calibrating the
leaf color chart for nitrogen management in diVerent genotypes of rice and wheat in a
systems perspective. Agron. J. 96, 1606–1621.
Sie, M. (1991). Prospection et evaluation genetique des varietes traditionnelles de riz (Oryza
sativa L. et O. glaberrima Steud.) du Burkina Faso WARDA, Cotonou.
Sie, M. (1994). Rapport de synthese et amelioration varietal du riz (SAVARIZ)/riz irrigue.
Paper Presented at the Third Annual Meeting of the Sahel Task Force, Segou, Mali. March
21–23, 1994. WARDA, Cotonou.
Smaling, E. M. A., Stoorvogel, J. J., and de Jager, A. (2002). Decision making on
integrated nutrient management through the eyes of the scientist, the land‐user, and
the policy‐maker. In ‘‘Integrated Plant Nutrient Management in Sub‐Saharan Africa:
INCREASING RICE PRODUCTION IN SUB‐SAHARAN AFRICA 133
From Concept to Practice’’ (B. Vanlauwe, J. Diels, N. Sanginga, and R. Merckx, Eds.),
pp. 265–283. International Institute for Tropical Agriculture (IITA), Ibadan, Nigeria;
CAB International, Wallingford, UK.
Soil Survey StaV (1998). ‘‘Keys to Soil Taxonomy,’’ 8th ed. US Department of Agriculture
(USDA), Natural Resources Conservation Service, Washington, DC.
Somado, E. A., Becker, M., Kuehne, R. F., Sahrawat, K. L., and Vlek, P. L. G. (2003).
Combined eVects of legumes with rock phosphorus on rice in West Africa. Agron. J. 95,
1172–1178.
Van Asten, P. J. A., Barbiero, L., Wopereis, M. C. S., Maeght, J. L., and van der Zee,
S. E. A. T. M. (2003). Actual and potential salt‐related soil degradation in an irrigated
rice scheme in the Sahelian zone of Mauritania. Agric. Water Manage. 60, 13–32.
Vanlauwe, B., Diels, J., Sanginga, N, and Merckx, R. (2002). ‘‘Integrated Plant Nutrient
Management in Sub‐Saharan Africa: From Concept to Practice.’’ CABI Publishing,
Wallingford, UK.
Vaughan, D. A. (1994). ‘‘The Wild Relatives of Rice: A Genetic Resources Handbook.’’
International Rice Research Institute, Los Banos, Philippines.
Virmani, S. S. (1996). Hybrid rice. Adv. Agron. 57, 377–462.
WARDA (West African Rice Development Association—Africa Rice Center) (2001). ‘‘Bintu
and Her New Rice for Africa: Breaking the Shackles of Slash‐and‐Burn Farming in the
World’s Poorest Region.’’ WARDA, Bouake, Cote d’Ivoire.
WARDA (West African Rice Development Association—Africa Rice Center) (2002). ‘‘Promising
Technologies for Rice Production inWest and Central Africa,’’ 28 p.WARDA Bouake, Cote
d’Ivoire; FAO, Rome, Italy.
WARDA (West African Rice Development Association—Africa Rice Center) (1996, 1998,
1999, 2000, 2001–2002, 2002–2003, 2003–2004, 2004–2005). WARDA annual reports
WARDA, Benin, Cotonou.
Wassmann, R., Lantin, R. S., and Neue, H. U. (2000). ‘‘Methane Emissions from Major Rice
Ecosystems in Asia.’’ Special issue of the Nutr. Cycl. Agroecosyst. Vol 58.
Xu, K., Xu, X., Fukao, T., Canlas, P., Maghirang‐Rodriguez, R., Heuer, S., Ismail, A. M.,
Bailey‐Serres, J., Ronald, P. C., and Mackill, D. J. (2006). Sub1A is an ethylene‐response‐factor‐like gene that confers submergence tolerance to rice. Nature 442(10), 705–708.
Yang, X., Bouman, B. A. M., Wang, H., Wang, Z., Zhao, J., and Chen, B. (2005). Performance
of temperate aerobic rice under diVerent water regimes in North China. Agric. Water
Manage. 74, 107–122.