Agroforestry is increasing being recognition of a viable option for overcoming
a large portion of the current global challenges of climate change mitigation
and adaptation, food security and household income. The recent recognition of
agroforestry as a greenhouse gas mitigation strategy under the Kyoto Protocol
has earned it added attention (Nair et al., 2009;
Syampungani et al., 2010). Efforts to increase
the adoption of agroforestry and as a land management strategy are therefore
highly needed, especially for small-holder farmers in sub Saharan Africa where
small land holdings and high cost of inputs and poor market structures (Mukadasi
and Maxwell, 2008).
Okia et al. (2009) expressed the urgent need
to test the technological robustness of agro forestry through on-farm testing
with farmers as a principle step in up-scaling the adoption of the technologies.
Efforts by National Agricultural Research Organization (NARO) are geared towards
improving farmer adoption of agro forestry technologies. Although, several efforts
by government and Non-Government Organizations (NGOs) have be implemented to
ensure the contribution of agro forestry to livelihoods of small-holder farmers,
the levels of adoption of these technologies has been limited. Several factors
accounting for this low uptake include land and labour shortage, lack of adequate
planting materials for preferred species, gender differences at household level,
limited knowledge about specific technologies (Franzel,
1999; Noordin et al., 2001; Mercer,
2004; Ogunlana, 2004). Kiptot
et al. (2007) showed that the process of adoption of improved fallow
agorforestry systems in Western Kenya was highly dynamic and variable with farmers
planting and discontinuing or re-adopting them due to a whole range of factors
of which soil fertility improvement is just one. These factors included incentives
from projects, the tying of adoption to credit programs, prestige, participation
in seminars/ tours and the availability of a seed market from projects promoting
Against this background, this baseline study was carried out in the LVCZ to establish the status and potential of promoting improved agro forestry technologies. The objectives of the study were to determine the level of awareness of the various agro forestry technologies for livelihood improvement, opinions about the usefulness and willingness to adopt these technologies if introduced.
MATERIALS AND METHODS
The study was conducted in the districts of Kayunga, Luwero and Mubende in
the Lake Victoria Crescent Agro-ecological zone of Uganda (Fig.
1). The contemporary climate in this area is wet tropical with a mean annual
precipitation of 1200 mm (distinctly bimodal distribution) and a mean annual
temperature of 23°C at an elevation over 1 km above sea level. Due to the
range in K-feldspar content and variable texture contrast, the soils are classified
as a mixture of oxisols, ultisols and inceptisols (Fungo
et al., 2011). Black and grey clays are also found in the flat, poorly
drained, dambos (flat, channel-less poorly drained valley bottoms) with yellow
sands on the sloping dambo margins. The topography is characterized by hills
and ridges that are highly dissected by streams and drainage ways. The main
economic activity of the people in the sampled districts is subsistence farming
of bananas, beans, maize, rice, potatoes, cassava among other crops land use
types include annual crops, plantation forestry, perennial cropping such as
bananas, coffee and agro forestry. Large expanses of grazing lands are common
in Luwero and Kayunga districts.
The districts of Kayunga, Luwero and Mubende were selected for this study because
they ranked lowest among that least experienced agro forestry extension in the
past decade (Okia et al., 2009). From each of
these districts, two sub counties were selected using key informants. The sampling
frame consisting of households was generated using the Local Councils of the
villages in the selected sub counties. Using a list of random numbers, 80 farmers
were selected from each district. Household interviews were held with the selected
households using a structured questionnaire.
The socio-economic characteristic of the households were presented as frequencies
and percentages in tables. Chi-square test of was used to establish the dependence
of some of the agro forestry practices to their benefits.
||Map of Uganda showing the location of sampled districts in
the Lake Victoria Crescent agro-ecological zone of Uganda
Binary logistic regression was used to assess the socio-economic factors affecting
the adoption of the three major agro forestry practices (home gardens, scattered
plots and boundary planting).
RESULTS AND DISCUSSION
Socio-economic characteristics of the respondents: Majority (~84%) of the sampled households were male-headed. Approximately 57% of the respondents were aged between 30 and 50 years (Table 1). The 83% of the respondents were married while the rest were not married, divorced or widowed. Almost half of the respondents had not exceeded primary level education (about 7 years of formal education). The average number of people per household ranges between four and seven. Close to one half of the respondents have between 4 and 10 acres of land and 64% of the land is under the Mailo tenure system.
It is reasonable to say that where 33% of the respondents had received training
on at least one aspect of agroforestry, the awareness is relatively inadequate.
However, it is important to understand the various aspects in which they were
trained and the relevance of these aspects to their farming practices. It is
also important to understand whether the exposure to the train in any way affects
the adoption and sustained use of the technologies. More importantly, it is
necessary to further understand the areas of training that would be of more
relevance for the livelihood of the farmers. By providing the relevant training,
farmer adoption and of the technologies is more likely and.
Existing agro forestry technologies: The agro forestry technologies
practices by farmers in the study area are shown in Table 2.
Trees scattered on farmland represent the most frequently encountered practice
by approximately 94% of the households in the area. Between 30 and 40% of the
respondents practice boundary planting and establish home gardens. Several technologies
are known to be practices in the zone but are not mentioned by farmers. These
include apiary, sericulture and aquaforestry (Agea et al.,
2007). The absence of these technologies is attributed to the limited samples
but also on their rare occurrence.
The major agro forestry species grown on farms are shown in Table
3. The major objectives for which farmers grow trees on farm include food,
timber and poles and bio-environmental improvement (soil fertility, shade, wind
|| Socio-economic characteristics of the respondents
|| Agro forestry technologies practiced by farmers in Kayunga,
Luwero and Mubende, Lake Victoria Crescent zone of Uganda
|| Agro forestry tree species and the technologies in which
they are use in Kayunga, Luwero and Mubende, Lake Victoria Crescent zone
The most common tree species include planted for food include Autocapus
heterophylus, Mangifera indica and Persea americana but others
like Carica papaya, Gavira gujava and oranges are also grown by
some few farmers. Makhamia lutea, Maesopsis eminii and Albizia
sp. are the most common species for timber and poles. There also exist some
farmers who maintain Milicea excelsa that are found growing on their
farmers. Although, the later species is liked by many people for its high quality
timber, no farmer was found planting the tree.
The partner of arrangement of the trees or farm is usually home gardens, tree
scattered on farm land or boundary planting. Case of rangeland trees fodder
banks and allays were also report but with very cases. Trees more commonly found
in home gardens are those meant for food while those for timber and environmental
modification are predominantly scattered or planted on boundaries and in range
lands. Okia et al. (2009) attributed the patterns
to respond to the problem or opportunity domain of the LVCZ; available market
of farm produce due to urban population, low soil fertility due to highly weathered
and acid soils and poles and field wood. The dominance of scattered tree practice
is attributable to the limited labour required to establish and manage geometrically
organized trees on farm. As Adesina et al. (2000)
noted, it is also probably due to the random manner in which small-scale farmers
plant the crops in space and time so that the predictability of future crop-tree
arrangement is difficult to determine. It is important also to note that many
farmers do not actually plant but tend trees that regenerate by natural means.
In this way, scattered trees are likely to dominate. The arrangement also largely
depends on the farmers knowledge of the interaction between the target tree
and the crops mostly planted on the plot (Feder et al.,
1985; Besley and Case 2003; Franzel,
1999; Ajayi et al., 2003). Many fruit trees
are not known to have allelopathetic effects and so many times famers plant
them in other than at the plot boundary. However, some species like Eucalyptus
sp. are rarely if ever, planted inside the plot because of their perceived effect
A farmers choice to practice an agroforestry technology will depend on
the objective (s) he/she has for so doing. Economic benefits seem to override
other benefits of agroforestry according to some studies, especially if the
farmers income is comparatively low. Income from sale of agroforestry
products attracts many farmers to adopt a net technology. For example, scattering
of Autocapus heterophylus on farm is a common practice by many farmers
as they usually sale the fruits quickly to earn cash. They also use it as a
food security. Okia et al. (2009) reported that
scattered tree practices were common in the LVCZ because farmers usually retain
rather than plant the trees and this makes mechanized farm operations difficult.
Knowledge and awareness of improved tree species: Table
4 shows that only 19 farmers (~8%) have received training in at least one
aspect of agroforestry. Soil management is the aspect where most farmers (10)
have received training followed by tree management.
|| Level of training in various aspects of agroforestry by farmer
in the Lake Victoria Crescent of Uganda
|| Socio-economic determinates of adoption of planting patters
of improves agro forestry tree species in the Lake Victoria Crescent Zone
|*, ** and *** means the variable is significant at 0.1, 0.05
and 0.001, respectively; home gardens: N = 96, LR χ2 (11)
= 68.69; Prob>χ2 = 0.0000; Log likelihood = -31.862781,
Pseudo R2 = 0.5188; Scattered: N = 83, LR χ2
(10) = 26.22, Prob>χ2 = 0.0035, Log likelihood = -15.377167,
Pseudo R2 = 0.4602; Boundary: N = 96, LR χ2
(11) = 23.50, Prob>χ2 = 0.0150, Log likelihood = -53.768741,
Pseudo R2 = 0.1793
This lack of training implies that farmers continue to use only indigenous
knowledge or the knowledge learned from neighbours to practice agroforestry.
This knowledge may be limited as there have been tremendous developments in
agroforestry that farmers may not be exposed to. The traditional practices of
scattering tree on farm have long faced challenges of land shortage can no longer
make significant contribution to desired benefits of agroforestry. Further farmer
training and sensitization may be urgently required to improve farmers
access to adequate knowledge of potentially useful technologies.
The models estimating the determinants of adoption of agro forestry practices
are shown in Table 5. The factors that significantly affect
adoption home gardens include land size and level of income derived from agro
forestry, land tenure, exposure to technology, training in any agroforestry
technology and exposure presence demonstration sites in the area. For scattered
trees only land tenure significantly affected adoption.
Mailo land was fixed at zero. Therefore, the impact reported is that of changing from mailo land to another type of tenure. The model indicates that changing from mailo land to private, Freehold and leasehold tenure reduces chances of adopting scattered trees by 0.5, 1.2 and 21.6 times, respectively. Farmers with larger land holdings have 1.2 more chances of adopting home gardens compared to those with smaller ones. However, chances of adopting home gardens are fewer if the farmers have higher income. This could be due to the high opportunity cost of the home gardens compared to other specialized income agro-enterprises that highly educated farmers focus on. For example, growing high value crops like tomatoes would be preferred to scattered trees on farm. Home gardens are also common among low-income groups because they are insurance means for food security and income diversification. Similarly, exposure to an improved technology also decreases the chances of adopting home gardens.
Kiptot et al. (2007) reported that farmers in
areas with a long history of exposure to agroforestry research had higher adoption
levels than those with recent history. However, other variables such as gender,
age, household type, type of housing, education, farm size, adults working on
the farm, livestock ownership and improved cows were not found to influence
adoption of improved fallows. The findings support these findings because we
found that exposure to a technology is the most important determinant of adoption
among all agroforestry systems. Several other studies however, show the influence
of these factors on adoption (Lapar and Pandey, 1999;
Baidu-Forson, 1999; Doss, 2002;
Naagula and Buyinza, 2009; Mazvimavi
and Twomlow, 2009; Peterman et al., 2010).
Important to note is that adoption is not a straightforward process. It is a
continuous process and the categories are therefore only relevant at a specific
point in time (Williamson, 1985; White,
2002). Farmers may oscillate between testing, adoption, discontinuation
and re-adoption. Adoption is complex and influenced by many factors that do
not lie solely within the household (Keil et al.,
2005; Kiptot et al., 2007; Mazvimavi
and Twomlow, 2009). These factors may include socio-economic, biophysical,
institutional and even political ones (as in the case of farmers refusing to
pay back credit). To classify farmers into two groups, adopters and non-adopters
is often an oversimplification. It is not easy to classify farmers into various
adoption categories, such as the four defined by Kiptot
et al. (2007) as the boundaries are often blurred. Nevertheless,
such classification provides a framework for understanding the perceptions of
different categories of farmers. Seeing such differences may in turn improve
understanding of the obstacles preventing initial adoption of a technology.
There is a difference between the decision to discontinue a technology that
one has tried and that of not adopting it at all. Similarly, discussions with
farmers who discontinue the use of a technology may provide information on the
features of the technology that proved unappealing to them under prevailing
field conditions and bring out other issues that had not been anticipated at
all, such as lack of benefits or inaccessibility to credit.
The study has revealed that the level of awareness of the various agro forestry technologies for livelihood improvement in the sampled districts is relatively low (~30%). Farmers acknowledge the usefulness of planting trees on farm for economic, environmental and even social benefits. Despite this interest in tree planting, several constraints such as land shortage, scarcity of planting material, inadequate knowledge and skills to establish and manage trees appropriately and high labour requirements for some technologies such as hedge-raw intercrops. There is generally high level of willingness to adopt these technologies if introduced. This will however, depend on socio-economic and environmental drivers.
Training and sensitization are highly recommended approaches for promoting adoption of technologies that the National agricultural Advisory services should include as an integral ingredient of the advisory role. Germplasm of candidate species is should be developed and availed to farmers possibly through local nursery operators that are periodically supervised by NARO and NAADS.
This study was funded by the Government of Uganda through the National Agricultural Research Organization (NARO). The researchers thank the Mr. Peter Lusembo, the director MuZARDI for coordinating the funding, district agricultural officers and NAADS coordinators for helping us with the site selection sampling of rice-farming households. The farmers are greatly appreciated for sparing time to provide the valuable information we used for this analysis. Research assistants who participated in the collection and entry of the data are highly appreciated.