Agroforestry Systems 59: 173–186, 2003.

© 2003 Kluwer Academic Publishers. Printed in the Netherlands.

173

Agroforestry research and development in southern Africa during the

1990s: Review and challenges ahead

F. Kwesiga1,∗ , F.K. Akinnifesi2, P.L. Mafongoya3, M.H. McDermott1 and A. Agumya1

1 InternationalCentre for Research in Agroforestry (ICRAF), Southern Africa Regional Agroforestry Programme,
P.O. Box MP 128, Mt. Pleasant, Harare, Zimbabwe; 2 ICRAF Agroforestry Programme, P.O. Box 134, Zomba,
Malawi; 3 ICRAF Agroforestry Programme, Chipata, Zambia (∗ Corresponding author, e-mail: fkwesiga@
africaonline.co.zw)

Key words: Biomass transfer, Improved fallows, Indigenous fruits, Rotational woodlots, Scaling up, Soil fertility

Abstract
The International Centre for Research in Agroforestry (ICRAF) initiated in 1987 the Southern Africa Regional
Agroforestry Programme in partnership with the national research systems in Malawi, Zambia, Zimbabwe and
Tanzania to address the problems of low soil fertility and consequent low crop and livestock production, low cash
income, and shortages of fuelwood and timber that are common to most rural households in the region. This paper
synthesizes agroforestry research and development during the 1990s from a regional perspective and emphasizes
scaling-up of promising technologies. The problem of nutrient-depletion can be overcome using nitrogen-fixing
and fast growing tree/shrub species, such as Sesbania sesban, Tephrosia vogelii and Gliricidia sepium, as short-
duration planted fallows in rotation with crops. Intercropping of food crops with coppicing trees, annual relay
intercropping and biomass transfer technologies were found to be appropriate for soil fertility improvement under
specific conditions. Tree fodder banks greatly increase fodder production and enrich livestock diets with protein
supplements. Rotational woodlots were developed to meet fuelwood and timber demand and reduce pressure on
natural woodlands. Research has identified the potential for a number of indigenous fruits in the region to supple-
ment the food needs of rural families, improve their nutritional status and generate cash income. Achieving the
ultimate goals of ensuring food security, alleviating poverty and sustaining the environment will require a massive
scaling up of adoption of these agroforestry technologies. Key strategies to accomplish this include: increasing the
benefits and targeting of agroforestry technologies (for example, through providing substitutes for costly inputs,
improving diversification, marketing and processing of agroforestry products, and employing GIS-based targeting
techniques) and information-sharing, training and collaborative partnerships in implementation and dissemination
of agroforestry options with all major stakeholders (farmers, non-government organizations, extension services,
educational institutions, and policy-makers).

Introduction for development aid). Whereas an economic growth

rate of around 7% is needed to maintain a favorable
The southern Africa region faces a worsening crisis as economic climate for a population growing at about
a result of poverty, food shortages and the HIV/AIDS 3% per annum, regional economic growth falls below
pandemic. These problems are caused by a complex of 2% (IMF 2000). Agriculture is the main source of live-
interrelated factors that are agro-ecological (declining lihood for approximately 80% of the rural population
soil fertility and crop yields, droughts, floods, environ- in the region, hence its potential to be a major driving
mental degradation), social (illiteracy, class and ethnic force in the recovery of regional economies. However,
disparities), and politico-economic (unfavorable do- agricultural production is currently constrained by un-
mestic policies, a massive debt burden, corruption, affordable inputs, especially fertilizers, lack of access
distorted international trade policies and skewed terms
174

to credit, and the minimum involvement of smallhold- The Food and Agricultural Organization (FAO) has
ers in the market economy. The development of the estimated that over 50% of future gains in food crop
agricultural sector is essential for attaining food secur- yields will come from fertilizers (FAO 1994) and ge-
ity, economic recovery and poverty alleviation in the netically modified crops. Such a scenario presupposes
region. massive increase in fertilizer consumption, expansion
Declining soil fertility is one of the root causes of of irrigated lands, agricultural modernization and in-
low crop productivity and consequently of deforest- creased investment in the agricultural sector. Is this
ation, as natural miombo forests are cleared for the feasible? Alternative, low-cost technologies and prac-
expansion of farmland. Traditionally, croplands were tices are needed. Agroforestry provides one such set
allowed to rest after cultivation for three years or more. of options.
This fallow phase is necessary to enable natural veget- This paper synthesizes the progress that has been
ation to regenerate and restore soil fertility. However, made by the regional research programme over the
due to rapid population growth and inequitable land past eleven years from 1991 to 2001 in developing and
distribution, farmers now are forced to cultivate the evaluating these agroforestry options. It also discusses
same piece of land more frequently, and in some cases approaches to scaling-up adoption, further research
continuously every year, thereby exhausting the soils. needs and future trends.
Furthermore, economic structural adjustment pro-
grammes in the region have led to the removal of sub-
sidies on agricultural inputs, especially inorganic fer- Viable agroforestry options
tilizers. This development, coupled with the frequent
devaluation of local currencies, high transportation Since 1987, the International Centre for Research
costs due to poor road networks, and lack of access in Agroforestry (ICRAF), recently designated as the
to credit, has rendered fertilizers unaffordable by the World Agroforestry Centre, and national agricultural
vast majority of smallholders. Without fertilizers or research and extension systems (NARES) have ex-
long fallow periods, productivity of food crops, re- amined traditional agroforestry practices and tested
mains low, with yields of the staple maize (Zea mays new innovations at research stations and on farms.
L.) usually falling under 500 kg ha−1 . Given the small This research has led to the development of many
farm sizes (often under one hectare), many farm fam- improved agroforestry options, which are now be-
ilies cannot produce enough to feed themselves even ing disseminated and used by tens of thousands of
during years of favorable rainfall. Most smallholders resource-poor farmers in southern Africa.
face food deficits during the periodic droughts affect- In the area of integrated soil fertility management,
ing the region. Drought and lack of dry-season fodder the focus has been on growing trees in situ or on
also constrain livestock production. Alternative cash- the transfer of biomass from one part of the farm
generating options are very limited in the rural areas. to another. Improved fallow systems as alternatives
At the same time, the native woodlands that provide to the traditional fallow practices were developed us-
timber, fuelwood, fruits, medicines, saleable products ing either coppicing tree species, which are grown in
and ecological services are overexploited or destroyed permanent association with crops, or non-coppicing
for agricultural expansion, as well as for supplying species, which are grown in rotation with crops. These
fuelwood and charcoal for domestic use and for curing systems are effective at restoring degraded lands and
tobacco, a popular cash crop for smallholders. maintaining productivity in the absence of mineral
Farmers in southern Africa are looking for ways fertilizers in a number of situations. Moreover, they
to break out of the poverty cycle. They are searching are economically attractive to smallholders in many
for techniques to improve productivity of their lands situations in southern Africa. ICRAF has also done
and sustain their environment. They are also seeking significant work on biomass transfer and green ma-
alternative means of supporting their families, gen- nures. In a number of situations, particularly where
erating cash and building a future for their children. soils are P deficient, fallow and biomass technolo-
Available options to respond to the challenges outlined gies are not sufficient to restore the degraded soils.
above include: dependency on food imports, contin- ICRAF has been working on combining inorganic
ued expansion of farming to marginal lands, increas- amendments into these systems.
ing the use of unaffordable inputs, or development and Other options that have been developed include do-
adoption of new technologies, including agroforestry. mestication and processing of indigenous fruits to en-
 175

hance family nutrition and income, and fodder banks Sesbania fallows produced 15 Mg ha−1 of fuel wood
for supplementary feeding of dairy cattle. In addi- (Franzel et al. 2002; Place et al. 2002). Current evid-
tion, the programme has been identifying the factors ence suggests that high maize yields following such
that constrain or favour adoption of these innovations, fallows are primarily due to increased organic matter
and has been engaging in dialogue on the policy im- input into the soil and nitrogen supply to crops (Bar-
plications with local and national decision-makers. rios et al. 1996). Table 1 summarizes the impact of
Identification of appropriate niches for the various improved tree fallow-based options without fertilizer
agroforestry options has improved the targeting of on maize yields at five sites in southern Africa. Yield
scaling up efforts, and thus its efficiency. To this end, increases ranged from 40% in Shinyanga to 317% in
biophysical, social and economic data have been as- Chipata.
sembled and combined in a geographic information A new provenance of Tephrosia candida (Mad-
system (GIS). agascar) that produces very high biomass has been
identified. It is suitable for a two-year fallow and does
not produce seeds in the first year. Sesbania fallows
Improved fallows were also found to reduce or completely eliminate
‘witch weed’ (Striga asiatica (L.) Kuntze) populations
Performance and species selection in colonized fields, from 1532 counts over a 100 m2
plot in the continuously cultivated unfertilized plot
Long-term fallowing (20 to 30 years) is no longer feas- to zero in the plot after Sesbania fallow (Kwesiga
ible and shorter natural fallows (up to 10 years) do et al. 1999). Rao et al. (1999) report similar effects
not adequately replenish soil fertility. ICRAF and its of Sesbania fallows on this parasitic weed.
partners have initiated research on the development of In Zimbabwe, several planted tree-fallow options,
short-term (1 to 3 year) improved or managed fallows including Acacia angustissima (Mill) Kuntze, Cajanus
to allow for rapid replenishment of soil fertility. A cajan (L.) Millsp. and S. sesban were compared with
range of leguminous trees and shrubs was screened grass fallow and continuous cropping with or without
to identify promising species that would add high fertilizers (Mafongoya and Dzowela 1998). The tree
amounts of nitrogen and organic matter into the soil fallows all increased the subsequent maize yields over
while also producing fuelwood. Sesbania sesban (L.) the control crops, maize after grass fallow and continu-
Merr., Tephrosia vogelii Hook.f., Gliricidia sepium ous maize without inorganic fertilizers. There was a
(Jacq.) Walp. and Leucaena leucocephala (Lam.) De significant interaction between tree species and fallow
Wit. are the most promising N-fixing trees identified duration on grain yields of unfertilized maize in three
for soil fertility replenishment (Kwesiga et al. 1999; years following fallows (Table 2). In addition, applic-
Rao et al. 1999). The techniques for integrating these ation of N fertilizer to maize following the fallows
species as short-duration planted fallows in rotation increased grain yields. In order to determine whether
with crops to build up N-capital in farmers’ fields are improved fallow systems are financially viable, given
now in place. Several thousand farmers have collabor- the labour requirements and the time land is out of
ated in developing these technologies by testing them production (i.e., tied to fallows), financial analysis was
on their farms in the region (Kwesiga et al. 1999). undertaken for Sesbania fallows in eastern Zambia
Improved fallows benefit farmers in the form of (Table 3). Results proved that these improved fallow
greater food crop yields, representing increased re- systems were profitable, nearly doubling the net bene-
turns to land and labour. For example, in experiments fits on a per hectare basis five years after planting the
conducted over the last 15 years at Chipata, Zam- fallow (Place et al. 2002). Over a six-year time hori-
bia, maize yields in three normal rainfall years after zon, discounted profits per unit land area were much
two years of Sesbania fallow averaged 5.6 Mg ha−1 higher for improved fallow–crop rotation system than
compared to 2.0 Mg ha−1 in unfertilized continuous for continuous maize without fertilizer (Franzel et al.
maize and 4.1 Mg ha−1 when maize was fertilized 2002). Planting S. sesban for two years emerged as the
with 112 kg N ha−1 , 20 kg P and 16 kg K ha−1 most profitable fallowing option (Kwesiga et al. 1999).
(Kwesiga et al. 1999). Over a 6-year period from 1988 These initial results encouraged scientists to set up tri-
to 1993, the Sesbania improved fallows required less als on farms and provide opportunities for progressive
than half the amount of labour needed for one hectare farmers to begin experimenting on their farms. By the
of continuously cropped maize. Furthermore, 2-year end of 1998, over 14 000 farmers were experimenting
176

Table 1. Impact of improved tree fallow-based options on maize yields at different sites in southern Africa

Site Rainfall (mm), Planted fallow Control Yield increase due

soil type to improved fallow
over the control
(Mg ha−1 )
(% increase)

1. Chipata, 950, Alfisols 2-year S. sesban Maize 3.8 (317%)

Zambia∗ 2-year C. cajan without 1.7 (155%)
2-year T. vogelii fertilizer 2.1 (191%)
2. Makoka, 980, Alfisols 2-year Sesbania Grass 2.8 (255%)
Malawi∗∗ sesban fallow
3. Domboshawa, 750, Alfisols 2-year S. sesban Grass 3.0 (188%)
Zimbabwe† 2-year C. cajan fallow 1.8 (113%)
2-year A. angustissima 1.0 (63%)
4. Tabora, 700, Ultisols 2-year S. sesban Maize (120%)
Tanzania†† 2-year C. cajan without 0.5 (50%)
fertilizer
4. Shinyanga, Tanzania†† 800, Vertisols 2-year S. sesban Maize 0.5 (40%)
without
fertilizer

Sources:
∗ (ICRAF-Zambia Annual reports for 1994, 1995 and 1996).
∗∗ (ICRAF Annual report for 1995; ICRAF-Malawi Annual report for 1994).
† (ICRAF-Zimbabwe Annual report for 1995; Mafongoya and Dzowela 1998).
†† (ICRAF-Tanzania Annual reports for Shinyanga and Tabora in 1995).

Figure 1. Options available to farmers in different management variables in establishing the planted fallows.
 177
Table 2. Maize yields over three years subsequent to planted fallows, as influenced by fallow species, fallow
duration and N fertilization of maize in Zimbabwe

Species Duration (years) Unfertilized maize yield Maize yield increase with
(Mg ha−1 ) fertilizer (Mg ha−1 )∗
Year 1 Year 2 Year 3 Year 1 Year 2 Year 3

Acacia angustissima 1 2.7 3.1 1.7 1.4 0.2 0.6

2 2.9 1.3 2.5 1.1 2.3 0.1
3 2.2 3.2 2.5 2.0 1.6 0.6
Cajanus cajan 1 3.2 2.7 1.5 0.9 1.1 0.5
2 3.4 3.0 1.4 0.4 1.0 0.5
3 4.0 3.6 3.0 −0.5 1.1 0.7
Sesbania sesban 1 4.7 3.1 1.5 −0.7 1.0 1.3
2 4.9 3.7 1.5 −0.7 1.6 1.5
3 3.1 3.6 2.9 2.1 2.9 0.9
SED 0.2 0.6 0.3 0.5 0.8 0.3

Sources of variation Significance level of F test

Species <0.001 0.04 0.26 <0.001 0.21 0.001
Duration <0.001 0.07 <0.001 0.01 0.05 0.006
Species × duration <0.001 0.05 0.02 <0.001 0.16 0.003

Adapted from (Mafongoya and Dzowela 1998)

∗ Maize grain yield increase is due to addition of 50% of the full fertilizer rate after planted fallows. The full
fertilizer rate is 76 kg N, 42 kg P and 28 kg K ha−1 .

Table 3. Sum of discounted net benefits over time for five maize cropping options examined
in a researcher managed, on-station trial at Chipata in eastern Zambia.

Option Sum of discounted net benefits (US$ ha−1 )

1988 1989 1990 1991 1992 1993

Continuous maize + no fertilizer 119 201 235 310 299 307

Continuous maize + fertilizer∗ 483 844 1054 1195 1153 1303
One year sesbania fallow −155 −6 222 453 457 545
Two year sesbania fallow −171 −151 175 475 488 588
Three year sesbania fallow −171 −171 −147 182 205 313
∗ Fertilizer rate is 112 kg N, 20 kg P and 16 kg K ha−1 yr−1 .
(Source: Kwesiga et al. 1999)

with improved fallows in eastern Zambia. This number (e.g. Tephrosia spp., C. cajan and S. macrantha). Ses-
is estimated to have grown to 80 000 across southern bania sesban can be planted using bare-root seedlings
Africa by the end of 2001. These encouraging results produced on raised nursery beds and inoculated, or by
have laid the ground for a regional goal of scaling up producing seedlings in polythene tubes (Kwesiga et al.
the technology to benefit 400 000 farmers by 2006. 1999).
Of the various species, T. vogelii was most pre-
Management of planted fallows ferred by farmers because it is both a soil-improver
and a pesticide. Cajanus cajan was also ranked high
Farmers have a range of options to choose from with among the preferred species because it provides food
regard to management of planted fallows (Figure 1). in addition to improving the soil. A number of spe-
These options vary with respect to fallow duration (1, cies that could be established once and then managed
2, or 3 years), establishment method of fallow spe- for a long duration (>15 years) through coppicing,
cies, fallow management and cropping system harvest such as Gliricidia sepium, Leucaena leucocephala,
schedule depending on land and labour availability. Senna siamea Lam., Calliandra calothyrsus Meiss.,
Some of the species can be planted by direct sowing
178
Table 4. Aboveground biomass at the end of 3-year fallow period, maize yields and Striga counts
following planted tree fallows compared with control plots at Msekera, eastern Zambia in the
1994–1995 season.

Fallow species Biomass Maize yield Striga (plants plot−1 )

(Mg ha−1 ) (Mg ha−1 ) (plot size = 10 m × 10 m)

Sesbania sesban 23.5 5.6 0

Gliricidia sepium 20.5 3.8 712
Flemingia macrophylla 17.8 3.5 448
Leucaena leucocephala 29.0 3.7 0
Calliandra calothyrsus 11.5 2.6 44
Senna siamea 59.0 2.1 0
Grass fallow 17.2 2.2 130
Groundnut–maize rotation – 3.1 130
Continuous maize (−Fertilizer) – 2.0 1532
Continuous maize (+ Fertilizer)∗ – 4.1 157
SED 8.8 0.33 334

∗ Fertilizer rate is 112 kg N, 20 kg P and 16 kg K ha−1 yr−1 .

and Flemingia macrophylla Roxb. were compared to recommended fertilizer (i.e., 112 kg N ha−1 , 20 kg P
non-coppicing S. sesban. These options were tested and 16 kg K ha−1 ). There were no significant differ-
against farmer-controls such as grass fallows, continu- ences between the first-year yields with 0, 1/4 and 1/2
ous cropping with or without application of inorganic recommended rates of fertilizer application following
fertilizers, and legume–cereal rotations. Sesbania tri- the Sesbania fallow (Table 5). However, application
umphed over all the coppicing options as well as the of 1/4 and 1/2 fertilizer rates to the second post-fallow
controls with respect to increases in maize yields and crop significantly increased maize yields compared to
suppression of weeds (Table 4). Senna siamea fallows fallows without any fertilizer applied to them in the
were outstanding in wood and leaf-litter production, first cropping season. A similar trend was recorded in
as well as coppicing ability. However, the residual the third year. The combined effect of improved fal-
yield-augmenting effect did not last beyond 3 years, lows and inorganic fertilizers is due to the fact that the
as it later produced comparable yields to those of amount of nutrients provided by fallows are not suf-
mono-cropped maize without fertilizer (2.0 Mg ha−1 ). ficient to exploit the genetic potential of the improved
varieties and good conditions of the site. The inorganic
Synergy between residual fertility from improved fertilizer used above bridges this gap by providing ad-
fallows and fertilizers ditional N, as well as P and K that the fallows do not
provide. A stochastic model was developed to determ-
Farmers have been interested in experimenting with ine which option is more profitable and resilient to
the application of modest amounts of nitrogen fertil- risk. The results from this model indicate that two year
izer to maize after improved fallows to enhance yields T. vogelii and S. sesban fallows supplemented with
and sustain the increase beyond two years of crop- 1/4 dose of the recommended fertilizer rate provide
ping after improved fallows. They developed interest cash-constrained farmers with better returns and risk
in this innovation following research demonstrations resilience than fallows alone or fallows and 1/2 the
at farmer training centres in eastern Zambia and the recommended rate.
implementation of the ‘starter pack programme’ in
Malawi. Biophysical limits of improved fallows
An on-farm trial involving 48 farmers was initiated
in four pilot project areas (Feni, Kalichero, Kalungu In order to determine the biophysical conditions under
and Katete) in Eastern Zambia from 1999 to 2002. which the planted fallow technology is appropriate, a
Sesbania sesban was grown for two years and was multi-site study was conducted in Zambia, Malawi,
followed by test crop maize (hybrid MM 604) in the Tanzania, Zimbabwe and Kenya. Sesbania sesban and
subsequent two years fertilized with 1/4 and 1/2 of the T. vogelii were planted at over 100 sites, both on-farm
 179
Table 5. Maize yields following 2-year Sesbania sesban fallows in combination with
25% or 50% of recommended inorganic fertilizer compared with continuous maize
with and without recommended fertilizer in eastern Zambia.

Treatments Maize yields (Mg ha−1 )

1997–1998 1998–1999 1999–2000

Sesbania fallow + no fertilizer 3.6 2.0 1.6

Sesbania fallow + 50% fertilizer 3.6 4.4 2.7
Sesbania fallow + 25% fertilizer 3.6 3.4 2.3
Continuous maize + 100% fertilizer 4.0 4.0 2.2
Continuous maize + no fertilizer 1.0 1.2 0.4
LSD (0.05) – 0.6 1.1

∗ Full fertilizer rate is 112 kg N, 20 kg P and 16 kg K ha−1 (n = 48).

sites and research stations, from 1994 to 1997. The tri- to overcome the problem of re-planting S. sesban
als produced a large variation in aboveground biomass every year as with relay fallow-cropping (see below),
(0 to 43 Mg ha−1 ), with a mean of 9.7 Mg ha−1 . Part G. sepium, a leguminous tree species with an im-
of this variation was due to the failure of S. sesban pressive coppicing performance, was chosen as an
to survive the first dry season after establishment in alternative (Kwesiga 1994). Gliricidia was selected on
1994. The factors that led to low survival were hy- the basis of its high foliage N content (4%) and its ease
pothesized to include two consecutive droughts (1994 of management, as the system involves severe cutting
and 1995), frost occurrence in Zimbabwe, shallow soil back of the coppice to reduce competition with the
that prevented trees from reaching the subsoil water maize crop. The prunings are incorporated in the soil
during the dry season, and nematode infestation, par- as manure during crop planting and at first weeding.
ticularly on sandy soils, which damaged root systems The advantage of this system is that farmers can
(Franzel et al. 2001). Tree biomass was used as the maintain trees on farm as a ‘fertilizer factory’ for
main indicator of the residual effects of S. sesban fal- over 15 years, without replanting. With respect to
lows on maize. Hence, initial soil conditions such as scientific principle, management requirement, func-
clay, pH and soil nutrients and tree biomass production tional objective and species choice alike, tree/crop
could be used for screening fallows. Results from this mixed intercropping can be legitimately considered
trial and other socioeconomic sudies show that Ses- as a variant of and an improvement over hedgerow
bania sesban fallows perform satisfactorily in areas intercropping (syn. alley cropping). It is a variant be-
that receive more than 600 mm of rainfall; are free of cause, although Gliricidia is established at the same
frost and nematode infestation; have deep soils with density per hectare as in hedgerow intercropping, it
clay content of at least 20% and PH of at least 4.5; is evenly spaced (0.9 × 1.5 m) throughout the field
where land is relatively abundant to allow fallowing; and cut at 30 cm above ground such that no area is
in maize/cotton (Gossypium herbaceum L./sorghum forfeited to the hedgerow. The tree is managed during
(Sorghum bicolor (L.) Moench farming systems; and the dry season to enhance production of more green
for households that cannot afford fertilizer. foliage rather than woody stems and green manure
is incorporated into the soil rather than the surface
application common to hedgerow intercropping. This
Gliricidia/maize intercropping arrangement seems to result in higher biomass pro-
duction and less hedgerow competition than in an
In the densely populated Shire highlands of south- alley-cropping arrangement, especially in unimodal
ern Malawi, farm sizes are extremely small (0.1 to rainfall environments.
0.5 ha), precluding the possibility of leaving a part of The performance of this soil fertility replenish-
the farm fallow. The traditional system is continuous ment technology depends on efficient pruning man-
maize cultivation without fertilizers and intercropping agement. Recent on-farm assessment in Malawi has
with herbaceous legumes, more recently with C. ca- shown that some farmers have been pruning the trees
jan (Ikerra et al.1999; Rao et al. 1999). In order more than twice during the cropping season. Out of
180
Table 6. Maize grain yield in a Gliricidia + maize intercropping system with different levels of fertilizer from 1992
to1997 at Makoka, Malawi.

Fertilizer as % 1992–1993 1993–1994 1994–1995 1995–1996 1996–1997

of recommended SM G/M SM G/M SM G/M SM G/M SM G/M
rate

————————————(Mg ha−1 )——————————————–

0 2.0 1.60 1.20 2.50 1.10 2.10 1.07 4.72 0.56 3.28
25 3.4 3.10 1.60 3.00 2.20 2.90 3.49 6.34 2.11 4.23
50 4.2 4.00 2.40 3.20 2.40 2.90 4.23 6.70 1.89 4.39
SED 0.25 0.24 0.17 0.33 0.33
Rainfall (mm) 975 539 600 963 1605

SM = sole continuous maize; G/M = Gliricidia/maize intercropping. Recommended fertilizer rate in Malawi: 96 kg N
and 40 kg P ha−1 .

95 farmers surveyed, 51% had pruned gliricidia twice Annual relay planting of tree fallow
and 20% did not prune. Pruning two times has been
shown to be adequate when plots are supplemented This system involves planting of nitrogen-fixing trees
with small doses of fertilizer, while the third pruning 3 to 5 weeks after food crops, notably maize, have
would be important in order to optimize grain yields been sown. Whereas species such as S. sesban can be
in non-fertilized Gliricidia plots. transplanted into the standing crop using bare-rooted
Long-term results from the maize/Gliricidia inter- seedlings, others such as T. vogelii, S. macrantha,
cropping system at Makoka show that maize yields Crotalaria spp. or perennial pigeonpea can be sown
began to exceed those of maize planted alone with and directly. Relay planting ensures minimum competi-
without fertilizer by the third season after tree estab- tion between the shrub/tree component and crops. The
lishment, often doubling those of the control by the trees mostly grow and develop a full canopy in the
fourth year and subsequently (Table 6). The unfertil- post-rainy season using the residual water after maize
ized Gliricidia plot was superior to the control with is harvested. Trees are clear-felled at the beginning
25% dosage of inorganic fertilizers, except for the year of the subsequent rainy season; wood is separated
of tree establishment, and was comparable to the con- and taken out from the field for fuelwood; and green
trol supplied with 50% dosage in three out of every foliage litter, pods and twigs are incorporated into
five years. In the Gliricidia/maize intercrop plots, the the soil in the course of land preparation. The cycle
differences between 50% and 25% fertilizer dosages is repeated annually, which means trees have to be
were not significant after the initial two-year period. replanted every year.
During years of drought, the maize yields from Foliar biomass produced by trees in this system is
unfertilized Gliricidia plots were not significantly dif- usually smaller than pure fallows and ranges from 1.0
ferent from sole maize plots supplied with a 50% to 2.5 Mg ha−1 for T. vogelii and 1.8 to 4.0 Mg ha−1
fertilizer dosage. Hence, it can be said that the for S. sesban due to the shorter growing period of
maize/Gliricidia system demonstrates greater resili- only 8-months (Boehringer et al. 1999). Consequently,
ence against drought than farmers’ standard practice amounts of nitrogen added through tree foliage are not
(Table 6). Gliricidia maintained foliar biomass at 2 sufficient to realize yields similar to those with com-
to 5 Mg ha−1 season−1 during an eight-year period, mercial fertilizers. This results in an average maize
without the need for replanting (Rao et al. 1999). Re- grain yield of only about 2 Mg ha−1 depending on
cent findings have also shown that the inclusion of rainfall received, soil type and on the position of the
pigeonpea (C. cajan) into Gliricidia/maize intercrop field in the landscape. The main limitations of the tech-
can increase overall production without negatively af- nology are labour for relay planting and harvesting of
fecting the advantages of the technology (Chirwa et al. trees, and the requirement of late-season rainfall for
2003). good tree growth. The labour requirement is, however,
reduced for directly sown species such as Tephrosia.
In seasons when maize is planted late due to delays
 181

in early rains, it is suggested that trees be planted evaluated in the region have shown potential to con-
within 1 to 2 weeks after maize is sown. During the tribute towards overcoming this problem when grown
drier years, this technology exposes farmers to greater in densely planted fodder banks. Browse species
risk because biomass production and hence its residual evaluated include Leucaena pallida, L. leucocephala,
benefit to the maize crop would be diminished. L. diversifolia, C. calothyrsus, A. angustissima and
G. sepium. Fodder bank technologies are targeted to
dairy cows and draught-oxen.
Biomass transfer In Tanzania, preliminary studies on supplement-
ation of dairy animals at a low level of 1 to 4 kg
Biomass transfer refers to mulching or green-leaf ma- animal−1 day−1 of dry Leucaena leaves increased
nuring using foliage of trees and shrubs cut and carried milk yields significantly (ICRAF 1996). In Zimb-
to cropping areas. In parts of Zimbabwe, Tanzania abwe, where the main emphasis of ICRAF’s work
and northern Zambia, biomass transfer is a traditional has been on the development of fodder bank techno-
practice. Farmers gather litter from the miombo wood- logy, species screening has identified A. angustissima
land and incorporate it in the food production fields and C. calothyrsus as sources of good quality fodder
(Nyathi and Campbell 1993). To improve this tradi- (Dzowela, 1998). Others found to have high yield po-
tional system, studies were conducted on a range of tential are provenances of psyllid-resistant leucaena
multipurpose trees (MPT) to optimize crop response species (L. esculenta, L. pallida and L. diversifolia
to their foliage. Green manure can be produced outside sub species stenocarpa). Presently, hundreds of farm-
the cropping area, for example in swampy fields (dim- ers in Tanzania and Zimbabwe have established fodder
bas), on contour bunds, sloping lands, field bunds or banks.
on land that is unsuitable for annual crops. However,
as many organic materials contain low nutrient con-
centrations, very large amounts of organic material are Rotational woodlots
required to obtain reasonable increases in crop yields.
The large amount of labour needed to transport and Rotational woodlots are meant primarily for fuelwood
apply the biomass justifies its use only for high-value production, but also serve to improve soil fertility,
crops, such as vegetables grown around towns. and are therefore grown in rotation with crops. The
It is important to address issues concerning ap- emphasis so far has been on Australian acacias such
propriate niches for biomass production, the effects as Acacia leptocarpa A. Cunn. ex Benth., A. crassi-
of biomass banks on nutrient mining, labour require- carpa A. Cunn. ex Benth., A. julifera Benth., and other
ments and economics before wide-scale promotion fast-growing species in the Tabora region of Tanzania,
of this technology in southern Africa. Currently, re- where demand for wood to cure tobacco is high. Sev-
searchers in the region are examining alternative uses eral farmers around Shinyanga have also embarked on
of leaf biomass for fodder. Recent results from Zambia testing woodlots on their farms, using L. leucocephala,
indicate that the application of Gliricidia biomass to A. polyacantha Willd., S. siamea, Azadirachta indica
cabbages and onions in dimbas followed by growing a Adr. Juss. and Albizia lebbeck (L.) Benth.. At the age
maize crop during the dry season is a profitable form of five years, A. crassicarpa woodlots produced nearly
of biomass transfer (Kuntashula et al. 2003). Tithonia 100 Mg ha−1 of biomass (Ramadhani et al. 2002).
is a colonizer of wastelands in Malawi and use of its Farmers are managing Australian Acacia spp. to
biomass for crop production could be turned into a maximize seed production on-farms. The potential
profitable venture. beneficiaries of this technology are about 35 000 to-
bacco farmers in the Tabora region. Ramadhani et al.
(2002) estimate that if all tobacco farmers in Tabora
Fodder trees district alone adopt rotational woodlots, approxim-
ately 8500 ha of miombo woodlands would be saved
One of the land use problems that small-scale livestock from deforestation every year. However, wider ad-
farmers face in the region is shortage of fodder, espe- option of the woodlots is constrained by inadequate
cially during the 7 to 8 month-long dry season every supply of tree seeds and training of farmers in nursery
year. Several multipurpose protein-rich browse trees techniques to achieve better germination rates and the
narrow range of species and provenances. Therefore,
182

further screening is necessary to identify species that trees as a strategy to cope with the famine period in
possess the following desirable features: high nitrogen 2001 (Akinnifesi et al. 2002).
fixation, coppicing ability and water-use efficiency. In spite of the popularity and importance of these
Farmers have compared the fast growing Australian fruits, they are not commonly seen in local super-
Acacia spp. to eucalyptus and voiced their concern markets due to the absence of value addition, lack
about water-use efficiency. Accordingly, studies on of improvement on their traits, the cultural notion
this subject have been initiated (Nyadzi et al. 2003). that indigenous fruits are free goods, and competi-
Experience from Thailand and Vietnam shows that tion from exotic fruits such as mangoes. However,
fast-growing Australian acacias like A. crassicarpa are growing urban markets are enhancing the economic
very shallow-rooted compared to eucalypts and mi- potential of wild fruits, and the prevalence of food in-
ombo species. They exploit water mostly from the security is simultaneously increasing the significance
top half-meter of the soil profile, and are vulnerable of their contribution to household nutrition. At the
to prolonged drought. Generally A. crassicarpa is re- same time, threats to the resource from deforestation
commended for sites with rainfall above 1000 mm in a are mounting.
unimodal environment. The performance of A. crassi- In the past, farmers seldom planted IFTs, be-
carpa in Tabora (where annual rainfall is 700 to cause miombo fruit trees were perceived to be both
800 mm under a unimodal distribution) contradicts slow growing and abundant. Limited understanding
the above boundary limit, since the species grows of natural variability, reproductive biology, propaga-
there very fast and produces approximately 100 Mg tion techniques and growth under field conditions
of woody biomass in 6 to 10 years (Ramadhani et al. hampered the domestication of these wild fruit trees.
2002). The fast growth has been attributed to the From 1989 to date, ICRAF and partners have been
combined effects of mychorrizae, rhizobia and a high carrying out research aimed at solving propagation and
water table. establishment problems. Other steps taken to advance
In order to increase awareness about the environ- the potential of IFTs include determining farmers’
mental problems caused by harvesting wood from nat- priority species, germplasm collection and screening,
ural forests and address constraints to the adoption of nursery and vegetative propagation studies and train-
interventions notably woodlots, dialogue with policy ing, market surveys, extension support for processing,
makers, tobacco buyers, politicians and technocrats in and identification of legal and institutional constraints
Tabora was initiated in 1998. Among the outcomes of to expanding market volume and value-addition.
this dialogue are recommendations for policies that fa- In order to develop a robust fruit tree domestication
vor production of wood on farms. For example, a levy programme, a survey was undertaken to determine the
on fuelwood harvests from the miombo for tobacco priority species for research and development efforts
curing has been proposed. in the region as well as to understand users’ needs
and preferences, market potential, and technological
opportunities and constraints. The survey involved
Domestication and commercialization of a total of 469 households spread over 20 districts
indigenous fruits in four southern African countries (Maghembe et al.
1998). The study identified Uapaca kirkiana (Euphor-
Indigenous fruits (IF) from miombo trees are an im- biaceae), Parinari curatellifolia (Chrysobalanaceae)
portant source of food and cash in southern Africa. and Strychnos cocculoides (Loganiaceae) as the three
They are widely consumed in both fresh and processed species most preferred by farmers, consumers and
forms. Extensive ethno-botanical surveys in southern experts across the region. Preferences for other spe-
Africa shows that more than fifty indigenous fruit tree cies, however, varied among countries (Kadzere et al.
(IFT) species provide food and income to rural people 1998), and considerable within-country differences
(Campbell 1987; Kwesiga et al. 2000). A recent sur- were evident by the variation observed across eight
vey in Malawi and Zambia revealed that as much as districts sampled in Malawi (Maghembe et al. 1998).
80% of the 323 rural households sampled had faced Farmers also identified quantitative traits that
severe food shortages, especially during the months needed improvement such as early fruiting, larger fruit
of November to January. About 50% and 26% from size and sweeter fruit taste, increase in the proportion
Malawi and Zambia, respectively, had adopted fruit of fruit pulp content and short tree height (Maghembe
et al. 1998). Tree precocity is by far the most critical
 183

attribute that needs urgent improvement, as some trees Technology adoption

have to grow for more than 10 years before first fruit-
ing. Findings from these surveys have been expanded Successful scaling up of agroforestry requires first
upon through on-farm research, village workshops and and foremost that technologies designed meet peoples’
other participatory rural appraisals (PRA) involving needs and match their circumstances. It also neces-
marketers, consumers and experts (Akinnifesi et al. sitates providing technical information and training
2002). to the practitioners, micro-finance, and formal credit
It is now evident that miombo IFTs are amen- systems, improving market access, and strengthen-
able to vegetative propagation. Graft takes of Adan- ing organizational linkages. The design of scaling up
sonia digitata (83%), Uapaca kirkiana (80%), Ziz- strategies must take into account appropriate biophys-
iphus mauritiana Lam. (90%), and Vangueria in- ical and socioeconomic conditions as well as policy
fausta (79%) compared favorably with exotics, such changes needed to support adoption.
as mango (97%). In Malawi, grafted Uapaca kirkiana ICRAF’s scaling up work began with on-farm test-
has started to produce fruits after two years, confirm- ing, that is, forging relations with farming communit-
ing that clonal propagation is a feasible improvement ies to implement problem diagnosis, field experiment-
approach (Akinnifesi et al. 2002). ation and on-farm evaluations. Building on this core
Currently, ICRAF is using vegetative propagation of farmer-experimenters and farm-tested technologies,
to select superior clones and cultivars of IFTs from ICRAF and partners are employing a number of meth-
the wild. Clonal selection offers a quicker and more ods for scaling-up (Boeringer 2001; Denning 2001).
efficient way of capturing the desired traits in IFTs. These include farmer-to-farmer exchanges, training of
Superior trees are identified directly with communities farmer-trainers, field days, distribution of extension
using PRA approaches involving traditional leaders, materials, provision of starter seeds, and training of
villagers, marketers and school children, based on partners.
fruit traits (sweetness, fruit size, fruit load, etc.). Effective partnerships – among farmers and in-
Trees having the desired criteria are named following dividuals and organizations engaged in research and
a set procedure aimed at crediting tree ownership to extension – are critical to success in scaling-up.
local communities and or the individual farmer who ICRAF has established enduring partnerships with
manages them government extension services, non-government or-
The addition of value to fresh fruits through pro- ganizations (NGOs) (e.g., World Vision and Plan
cessing, packaging and enterprise development has International), and local (e.g., chiefs) and national
been initiated to increase demand, utilization, and policymakers (e.g., parliamentarians). Complement-
profitability. Success stories have been recorded in arity in functions, resources, skills, and geographical
Malawi, Tanzania and Zambia, where local com- reach among these institutions has made possible the
munities were trained by ICRAF staff and partners current and growing expansion of agroforestry ad-
in pilot cottage processing of indigenous fruits into option. In addition, collaboration with universities,
juice, jam, wine, yogurt, and other products. Juices colleges and schools has strengthened agroforestry re-
and jams made from U. kirkiana and S. cocculoides search and helped build capacity in future researchers,
have been analyzed for their physicochemical prop- extensionists, and the institutions they will join.
erties and evaluated by a consumer panel composed ICRAF and its partners in extension and devel-
of 16 members. In order to impart skills in busi- opment have taken an approach to technology dis-
ness development, processing, packaging and enter- semination that rests not only on training farmers in
prise development to rural women’s groups, ICRAF’s techniques, but also in enhancing and encouraging
programme in southern Africa has now engaged the their ability to observe, adapt and innovate. Great
Commercial Products from the Africa Wild (CPWild value is placed on the farmers’ own knowledge of ag-
Consortium) thereby tapping expertise from the Uni- ronomic practices and their ecosystems, which is then
versities of Natal, Pretoria and Stellenbosch in South combined with new information introduced by the
Africa. The consortium plans to support improving technical staff. ICRAF and partners provide training
product quality and to propose mechanisms for qual- to groups of interested farmers, who are then expected
ity assurance and certification, thereby opening up to train their peers. We have experience in the region
marketing and enterprise development opportunities. that one farmer trainer can influence at least 10 new
farmers to adopt agroforestry in a season, if farmers
184

can witness the benefits of agroforestry and seed is An urgent need exists to understand the labour
available. demands of agroforestry and how the effects of
Many obstacles to the wider adoption of agro- HIV/AIDS may affect adoption of agroforestry. Meth-
forestry remain to be overcome. One intrinsic chal- ods have to be developed and implemented for track-
lenge lies in the long-term nature of the benefits to ing the impact of adoption of agroforestry practices
be derived from most agroforestry practices. Similar on the welfare of rural households, with costs and be-
experience is found with most natural resource man- nefits assessed and disaggregated by gender, age and
agement practices: farmers must be able to withstand socioeconomic status. In addition, research must also
initial years of low or negative profitability in or- address local, national and regional institutions and
der to reap longer-term economic profits. Access to key policy issues.
credit, reliable tree seed supply, and improved access In addition, a number of technical areas still re-
to agricultural markets offering adequate prices would quire scientific investigation. These include: the im-
minimize this constraint and greatly increase farm- pact of agroforestry practices on long-term nutrient
ers’ capacities and incentives to adopt agroforestry budgets, the optimal integration of organic and non-
technologies. Another significant source of limitation organic fertilizers, improving fodder for non-dairy,
lies with the national agricultural extension services, free-range livestock, and the identification of a wider
which need to be greatly strengthened by offering range of agroforestry species across all technologies.
more education and training to their personnel, and Finally, research must be directed to provide the
resources (e.g., for transportation and incentives). information needed to develop strategies to respond
As agroforestry takes root in the farming systems to projected trends in demand and opportunities for
of the region, several second-generation issues are agroforestry in the southern African region. Strategies
emerging. These include growing incidence of pest to make use of these opportunities include:
and diseases, weediness, increased net demand on wa- 1. Improving marketing and processing of agro-
ter by agroforestry trees, and reduced investment in forestry products. For such a strategy to be sustain-
the agricultural sector and related policy shifts such able, the involvement of the private sector would
as removal of agricultural subsidies. In addition, the have to be enhanced. New products that are in de-
HIV/AIDS pandemic has had serious consequences mand by rural and urban markets would have to be
for the agricultural sector, reducing the availability of developed.
labour, liquidating the assets of rural households, and 2. Diversification of agroforestry products and by-
seriously undermining their skill-base, productivity, products such as high value trees, indigenous and
incomes, and food security (Mutangadura et al. 1999; exotic fruit trees, medicinal plants, fodder for live-
IMF 2000). stock, organic vegetables production with green
manure from trees will increase opportunities for
small producers to develop their own processing
Research needs and future trends and marketing channels (cottage industries) and
add value to the raw products.
The rate of adoption of agroforestry technologies for 3. Development and promotion of substitutes and/or
soil fertility management remains low (Franzel et al. supplements for costly, imported external in-
2002; Place et al. 2002) despite their demonstrated puts. Examples of such options include fodder
potential. It is therefore important to understand the trees, which can supplement or replace expens-
constraints to adoption and identify strategies for over- ive dairy meal; and N-fixing trees together with
coming them. Previous studies have lacked farmer- their residues, which can supplement or serve as
derived data on profitability, feasibility and acceptabil- substitute for inorganic nitrogen fertilizers.
ity of these innovations for smallholders under a range 4. Options for mitigating the continuing degrada-
of conditions. It is essential to identify the biophys- tion of the environment and losses in biodiversity.
ical and socioeconomic factors that favour adoption Agroforestry strategies that address environmental
of new technologies in order to target systems where degradation and its consequences include replen-
they are likely to succeed. Techniques employing GIS ishing soils in degraded lands and reducing risk
and remote sensing for spatial targeting of soil fertility due to climatic change through tree-crop integ-
practices need to be validated and applied more widely ration; and promoting, protecting and planting
throughout the region. indigenous trees.
 185

5. The international discussion on global warming ity by devolving skills and knowledge to development
also offers opportunities for agroforestry research partners in the region
and development in southern Africa. This is linked The results of agroforestry research and devel-
to compensatory payments and clean development opment are now being widely promoted by govern-
mechanisms designed to reward tree planting and ment extension agencies, NGOs and other institu-
management, among other things. tions, working together with farmers. An encouraging
6. Development and implementation of strategies for portent for the future is the degree to which diffu-
large-scale dissemination of agroforestry techno- sion and adaptation of agroforestry technologies has
logies at the local level. The current thrust towards already occurred – spontaneously, through farmer-to-
decentralization policies, i.e., those leading to the farmer networks, and through NGO-led projects that
devolution of power and control from the central spotlight alternative technologies and provide support
state to lower levels, could provide an enabling for farmer-learning in the region. Further, widespread
environment and new avenues for agroforestry adoption of agroforestry technologies, supported by
dissemination. continued participatory research and dissemination,
7. Training and capacity building in agroforestry has the potential to achieve the goals of poverty al-
among all major stakeholders. Strategically, train- leviation, food security and environmental protection
ing (university, college, extension service, farmer- in the face of the manifold challenges besetting the
to-farmer) is the vehicle to achieve wide scale dis- southern African region.
semination of agroforestry technologies and ser-
vices. Policy dialogue to raise the awareness of
policy makers about the benefits of agroforestry Acknowledgements
and the constraints impeding its expanded adop-
tion is essential. We wish to acknowledge with gratitude the contin-
8. Co-operation and partnerships with a broad range ued financial support to the Southern African Regional
of actorsis a strategy that will enable the South- Agroforestry Programme by the Canadian Interna-
ern African Regional Agroforestry Programme, tional Development Agency (CIDA), the Swedish
despite limited resources, to succeed in the chal- International Development Agency (Sida), The Ger-
lenge of proactively seizing the opportunities for man Federal Ministry for International Development
agroforestry arising in the future. (BMZ) and the Rockefeller Foundation, and collabor-
ation of national scientists.

Conclusions
References
Agroforestry research and development have made
great strides during the 1990s in southern African Akinnifesi F.K., Kwesiga F.R., Mhango J., Mkonda A., Chilanga
T. and Swai R. 2002. Domesticating priority miombo indigen-
countries. Many options have been developed and are ous fruit trees as a promising livelihood option for smallholder
being tested and adopted by thousands of farmers in farmers in southern Africa. Acta Horticulturae (in press).
this region. The vision for scaling-up agroforestry in Barrios E., Kwesiga F., Buresh R.J. and Sprent J.I. 1996. Nitro-
the region foresees enhanced agricultural productivity, gen mineralization in density fractions of soil organic matter
from maize and legume cropping systems. Soil Biology and
profitability and sustainability improving livelihoods Biochemistry 28: 1459–1465.
of rural people across the region. Realizing this goal Boehringer A., Maghembe J. and Phiri R. 1999. Tephrosia vogelii
will require the achievement of key objectives, includ- for soil fertility replenishment in maize-based cropping systems
ing: replenished soils through integrating a variety of of southern Malawi. Forest, Farm and Community Tree Research
Reports 4: 117–120.
organic options along with restricted use of inorganic Boehringer A. 2001. Facilitating the wider use of agroforestry for
fertilizers, the introduction of high value trees, mar- development in southern Africa. Development in Practice 11:
ket and product development, enabling policies, and 434–448.
effective and targeted scaling-up strategies. ICRAF Campbell B. 1987. The use of wild fruits in Zimbabwe. Economic
Botany 41: 375–385.
remains strategically positioned to be an effective fa- Chirwa P.W., Black C.R., Ong C.K. and Maghembe J.A. 2003.
cilitator and coordinator for agroforestry development Tree and crop productivity in gliricidia/maize/pigeonpea crop-
in the southern African region at the present time, ping systems in southern Malawi. Agroforestry Systems (this
issue).
with the long-term aim of creating greater sustainabil-
186

Denning G. 2001. Realising the potential of agroforestry: integ- Kwesiga F.1994. Performance of fifteen provenances of Gliricidia
rating research and development to achieve greater impact. sepium in eastern Zambia. Forest Ecology and Management 64:
Development in Practice 11: 407–416. 161–170.
Dzowela B.H. 1998 Environmental adaptation and performance of Kwesiga F.R., Franzel S., Place F., Phiri D. and Simwanza C.P.
various agroforestry tree species in Zimbabwe. Transactions of 1999. Sesbania sesban improved fallows in eastern Zambia: their
the Zimbabwe Scientific Association 72: 2–9. inception, development and farmer enthusiasm. Agroforestry
FAO 1994. Strategies for Sustainable Agriculture and Rural Devel- Systems 47: 49–66.
opment: New Directions for Agriculture, Forestry and Fisheries. Mafongoya P.L. and Dzowela B.H. 1998. Soil fertility replen-
Food and Agriculture Organisation (FAO), Rome. 64 pp. ishment through agroforestry systems in two contrasting agro-
Franzel S., Coe R., Cooper P., Place F. and Scherr S. 2001. ecological zones of Zimbabwe. Transactions of the Zimbabwe
Assessing the adoption potential of agroforestry practices in Scientific Association 72: 31–42.
sub-Saharan Africa. Agricultural Systems 69: 37–62. Maghembe J.A., Simons A.J., Kwesiga F. and Rarieya M. (eds)
Franzel S., Phiri D. and Kwesiga F. 2002. Assessing the adoption 1998. Selecting indigenous trees for domestication in southern
potential of improved fallows in eastern Zambia. In: S. Franzel Africa-priority setting with farmers in Malawi, Tanzania, Zam-
and S.J. Scherr (eds.), Trees on Farms: Assessing the Adoption bia and Zimbabwe. Nairobi: International Centre for Research in
Potential of Agroforestry Practices in Africa, CAB International, Agroforestry. 94 pp.
Wallingford, Oxon, UK, pp. 37–64. Mutangadura G., Jackson H. and Mukurazita D. (eds) 1999. AIDS
ICRAF 1996. Annual report for 1992. International Centre for and African Smallholder Agriculture. Southern Africa AIDS
Research in Agroforestry (ICRAF), P.O. Box 30677, Nairobi, Information and Dissemination Centre (SAFAIDS), Harare, Zi-
Kenya. mbabwe, 206 pp.
Ikerra S., Maghembe J., Smithson P. and Buresh R. 1999. Soil Nyathi P. and Campbell B. 1993. The acquisition and use of mi-
nitrogen dynamics and relationships with maize yields in Glir- ombo litter by small-scale farmers in Masvingo, Zimbabwe.
icidia-maize intercrop in Malawi. Plant and Soil 211: 155–164. Agroforestry Systems 25: 43–48.
IMF 2000. World Economic Outlook. International Monetary Fund, Nyadzi G.I., Janssen B.H., Otsyina R.M., Booltink H.W.G., Ong
Washington DC. USA. C.K. and Oenema O. 2003. Water and nitrogen dynamics in
Kadzere I., Chilanga T., Ramadhani T., Lungu S., Malembo L., rotational woodlots of five tree species in western Tanzania.
Rukuni D., Simwanza P., Rarieya M. and Maghembe J. 1998. Agroforestry Systems.
Choice of priority indigenous fruits for domestication in southern Place F., Franzel S., deWolf J., Rommellse R., Kwesiga F., Niang A.
Africa: summary of case studies in Malawi, Tanzania, Zambia and Jama B. 2002. Agroforestry for soil replenishment: evidence
and Zimbabwe. In: Maghembe J., Simons A., Kwesiga F. and on adoption processes in Kenya and Zambia. In: Barrett C.B.,
Rarieya M. (eds), Selecting Indigenous Trees for Domestication Place F. and Aboud A.A. (eds), Natural Resources Management
in Southern Africa: Priority Setting with Farmers in Malawi, in African Agriculture: understanding and improving current
Tanzania, Zambia and Zimbabwe. ICRAF, Nairobi, pp. 1–15. practices. CAB International, Wallingford, UK, pp. 155–168.
Kuntashula E., Mafongoya P.L., Sileshi G. and Lungu S. 2003. Ramadhani T., Otsyina R. and Franzel S. 2002 Improving house-
Potential of biomass transfer technologies in sustaining veget- hold incomes and reducing deforestation using rotational wood-
able production in the dambos of eastern Zambia. Experimental lots in Tabora district, Tanzania. Agriculture, Ecosystems, and
Agriculture (in press). Environment 89: 229–239.
Kwesiga F., Boehringer A. and Denning G. 2000. Scaling-up soil Rao M.R., Mafongoya P.L., Kwesiga F. and Maghembe J.A. 1999.
fertility management innovations in southern Africa: The im- Nutrient cycling in agroforestry systems of the semi-arid tropics
proved fallow case study in eastern Zambia. In: Going to Scale: of Africa. Annals of Arid Zone 38: 275–307.
Can We Bring More Benefits to More People More Quickly? In-
ternational Institute of Rural Reconstruction (IIRR), Y.C. James
Yen Centre, Silang, Cavite, Philippines, pp. 68–71.