East African Journal of Forestry & Agroforestry

This study investigated the contributions of agroforestry to household food security in Kwale County. The study objectives were to determine the status of agroforestry, the status of household food security, and the relationship between agroforestry and household food security. Descriptive cross-sectional survey research design was applied, with multistage sampling procedure, which comprised of purposive and random sampling for selection of the study sites and respondents from different villages in Kubo south and Samburu wards, Kwale County. The sample size for the study comprised of 270 smallholders and 10 key informants


INTRODUCTION
Agroforestry has been promoted in Africa as a low input land use practice to facilitate farmers improve soil fertility, increase income, and enhance food security (Oduol et al. 2006). Since the early 1990s, the World Agroforestry Centre and its partners in Eastern and Southern Africa have developed a wide range of technologies for improving soil fertility and significantly boost crop yields by providing high returns on labour and land (Makumba et al., 2007). According to Thangata et al. (2002), agroforestry protects farmers against socioeconomic shocks due to the diversification of species and products that ensure that risks are spread and farmers are well cushioned in case of extreme events. Over 480,000 farmers are practicing agroforestry in Southern Africa and have achieved various benefits including increases in crop yields, income, savings, wealth creation, and soil improvement (ICRAF, 2007). Those who practice agroforestry have had an increase in maize yields from 1 t/ha to 3 t/ha (Garitty et al., 2010). Agroforestry also plays an important role of increasing the yields of vegetables and fruit, and also provide varied and nutritionally balanced diets rather than calories alone (Susila et al., 2012). The agroforestry sector is often impeded by legal policy and institutional arrangements, leading to its environmental benefits being unrewarded and investments discouraged . Most agricultural policies in Africa have more support for external than internal technologies, inputs, and practices (FAO, 2001). Globally, more than 1.3 billion people practice agroforestry ranging from linear planting, mixed, home gardens, hedges, boundary planting, woodlots, and alley planting . Over the years agroforestry system have evolved and become more complex and some examples include Cacao system in West Africa, the rotational woodlots in Kenya, home gardens in Mt Kilimanjaro area of Tanzania, and the parkland system of the Sahel (Beddington et al., 2012;de Foresta et al., 2013).
In Kenya, ICRAF and other development partners have made concerted efforts to promote agroforestry in different parts of the country through partnership with Government and communitybased organizations as a solution to increasing agricultural food productivity for food insecurity and environmental degradation (Place, 2009). The Kenya Agricultural Research Institute (KARI) conducted several agroforestry experiments with farmers on the adoption of agroforestry technologies from 1988 to 1992, to determine their suitability for the coastal lowlands. These adoption studies showed that the intercropping system had an advantage for smallholder farmers (ILCA, 1994). In a study conducted by Jamnadass et al. (2013), in Kenya and Malawi, exotic and indigenous fruits dominate the rural agroforestry systems. Agroforestry contributes to household food security directly by providing edible products and indirectly through fodder, soil fertility improvement, and conducive micro-climate for crop productivity (Maliki et al., 2012). In North Lampung, Sumatra, home gardens, averaging 0.75 ha, contain as many as 21 tree species excluding the understory component (Tukan et al., 2002). In a survey conducted around eastern Mount Kenya, approximately 200 different tree species were identified on farms (Oginosako et al., 2003). In one central market in Ghana, it was found that there are approximately 700 people involved in the forest product trade on a full-time basis and that trading was a dominant economic activity in even the remotest villages of Southern Ghana (Weibe & Maxwell, 1998). Tropical home gardens and agroforests have shown to improve diversification and resulted in better food security for smallholder farmers and that research should be intensified on how such systems can be improved for better food security particularly in areas far from the markets (Hagos et al., 2002;Michon & De Foresta, 1995;Kumar & Nair, 2004). Agricultural growth is very crucial for improving the living standards of the poor and alleviating poverty as this helps increase food productivity and income (Thirtle et al., 2001). In this research, we sought to investigate the contribution of agroforestry to food security in Kwale County, a region which is relatively poor and has extremely low agricultural activities.

Study Site
The research on which this paper was conducted was in Kwale County. Kwale County has 122,047 households out of which 74.9 % live below poverty line. The county receives rainfall ranging from 400 to 1600 mm per annum (Kwale County Government, 2013). Agriculture is the main economic activity for the households with most farmers practicing mixed farming.

Research Design
The study used descriptive survey design with the total target population of 122,047 households (Kwale County Government, 2013). The sample frame population of 10,513 households, was drawn from smallholder farmers in Kubo South and Samburu Wards, Kwale County where 400 small scale farmers were selected using simple random sampling and 10 key informants randomly selected from ministry of agriculture, Kenya Forestry Service, Kenya Agricultural and Livestock Research Organizations, County government staff, staff of NGOs, community leaders and communitybased organizations.

Demographic Profile of Households
The socio-economic characteristics of the household heads as respondents that were studied included age, gender, education level, and household and farm size. These characteristics were chosen in order to get an overview of the composition and status of the respondents for the purpose of drawing implication in relation to agroforestry practices and household food security in the study area. The results are as shown in Table  1. The results in Table 1 shows that among the 270 respondents interviewed, majority (62%) were within the age bracket of 36 to 60 with 23% and 15% being above 60 years and between ages 15-35 years respectively. This implies that generally there were very few respondents within the youth age bracket of 15 to 35 years. The findings are in line with those of FAO (2015) who asserted that rural youths have developed negative perception on agriculture hindering them from fully becoming active in farming. Study finding in table 1 show that more than three quarters (78%) of the respondents were male while the rest (22%) were female. This implies that probably there were more male-headed households who participated in agroforestry than female-headed households. It could also be due to patriarchal nature of most African societies. The findings agree with those of Hildreth (2008) which also showed that males were major players in the practice of agroforestry. Similarly, (Herrero et al., 2014) agreed with survey results and concluded that that, in some African societies it is a taboo for women to plant trees creating the perception that planting and managing trees in agroforestry is the responsibility of the masculine gender. These cultural beliefs may hinder the adoption of agroforestry practices. A study by Kiptot and Franzel (2012) asserted that, women are actively involved in agroforestry although the level of participation is low relative to men. The majority (50.74 %) of the respondents had acquired primary level of education with very few (3.7%) having attained college level education ( Table 1).
According to Oino and Mugure (2013)household size is an influencing factor to practice agroforestry in that when the family size is big, members tend to plant more trees because the need of tree products are higher and also labour is available. Results of the study further show that majority (64.7%) of the households had land sizes of six acres and above with a very negligible (1.9%) percentage having land size less than 1 acre. The farmers' decision to practice agroforestry and what to grow on farms for food for their household and the market is dependent on how much land is owned by the farmers. This result is supported by studies conducted in Nyeri, which revealed that farmers with relatively larger parcels of land allocated more land for tree planting (Nyambari, 2008;. Studies done in Zambia also support the claim that farmers with more land are likely to adopt agroforestry technology and can afford to experiment with new technology and still have land to spare for food production without compromising on food production and food security (Waha et al., 2016). However, the results contradict those from Kakamega which indicate that majority of farmers have planted trees despite their small holdings Similar studies by Keil et al., (2005) noted that most agroforestry farmers were willing to expand their farm size but cited limited labour as constraining factors to expansion

Household Food Security among Agroforestry Farmers
The first objective of the study was to determine the status of food security among smallholder agroforestry farmers. Household food security in this context is the ability of the household to produce food from agroforestry practices in enough quality and quantity all year round. Three dimensions of food security were considered namely food availability, food access and food stability. The direct and indirect contributions to household food security were considered in the study. The findings and discussions are presented in the following subsections.  year round, and had to reduce the quantities and number of meals they eat for their food production to last the whole year hence. This implies that the agroforestry systems practiced by the households do not adequately support food security either directly for food or indirectly through income to purchase food. The other possibility is that the food production and income from agroforestry practices is not enough to meet the demand of the households due to high dependency ratio and low productivity due to low application of agroforestry technology. Further analysis was done on the findings on status of food security to categorize households in the levels of food insecurity according to international practice. The results of the categorization of the smallholder farmers into food insecurity levels are as shown in Table 3. The results in Table 3 Show that households were placed into 4 categories as per the procedure applied by FAO and USAID for determining household food insecurity among households (Coates, 2013). The responses indicated that majority of the households are moderately food-insecure (54.8%), followed by severely food -insecure households at (27.4%). less than 20% were food -secure (14.8%) and mildly food-insecure households accounted for (3.0%).
Only 14.8 % had access to enough quality and quantity of food all year round with the rest being at different levels of food insecurity. This implies that households practicing agroforestry in the coast region of Kenya are not producing adequate food from their farms and have to meet their food needs through diversification of farming activities and also through venturing into various off-farm activities to meet the short fall. The inability to produce enough food is probably due to the poor soil conditions, erratic and unreliable rainfall, and destruction by animals. The results of the study are consistent with other studies which indicate that farmers have to diversify their income sources by engaging in different activities both on-farm and off-farm (Silvestri et al., 2015). Focus group discussions and key informant interviews with different stakeholders revealed that households have not had good crop years over the last two years due to drought and unreliable rainfall, inadequate extension services, and lack of credit to access farm inputs which have all worked in combination to reduce farm productivity and hence contribute to food insecurity for the majority of the households. This study is consistent with studies conducted in India using the HFIAS tool and found that 77 % of the households were food insecure and also further indicated that households were spending up to 50 % of their income on food (Chinnakali et al., 2014). The results of this study are also comparable to those conducted in Sabah, Malaysia that applied the HFIAS tool and indicated that 64.7 % of the households were food insecure (Farhadian et al., 2015).The food insecurity among the households was found to be mainly due to inadequate funds and other resources to purchase food. Silvestri et al. (2015) studied the link between agriculture and food security among farming households in selected sites in Kenya, Uganda and Tanzania and revealed that there is high food insecurity among the farming households, 62 % in Rakai (Uganda), 80 % in Lushoto (Tanzania) and 85 % in Wote (Kenya).

The Status of Agroforestry Practice among Smallholder Farmers
Objective two sought to establish the status of agroforestry practice among smallholder farmers.
Respondents were asked questions on the purpose of the trees, acreage under tree, types of plants intercropped with trees and the average number of trees planted. The results are presented in the subsequent sub sections.

Table 4: Purposes of planting trees
The results in Table 4 show that trees on the farms with commercial purpose scored the highest count 33.1%, food (32.9%), domestic use (10%), seeds production (7.1%), shade (4.1%), building materials (3.5%), fruits (2.7%), timber (1.6%) and firewood (1.5%). The results imply that there are diverse tree species being grown in the farms. Diversification of trees planted on the farms ensures continuous income streams and helps the framers spread the risks due to fluctuating markets and unpredictable weather conditions. The results of this study are in agreement with studies by Regmi (2003) and Baul et al. (2015), which revealed that trees grown through various agroforestry practices play vital roles in supporting households in many diverse ways, including provision of food, income, fuel wood, building materials, and fodder for livestock. The results of this study are also consistent with the results of a study by Gibreel (2013), who asserted that preferences in many small-scale agroforestry farmers on different tree species on farms are driven mainly by economic factors and less on environmental management. The results of this study are also in tandem with those of Nzilano (2013), in Tanzania on the choice of trees in agroforestry home gardens by households revealed that majority of the farmers planted tree based on their economic benefits.

Status of Tree Growing and Acreage under Agroforestry
The researcher sought to establish the changes in tree growing over the last five years. The farmers were asked about the trends in tree growing in the area. The results are as presented in Table 5 below: The results in Table 5 show that the majority (59.6%) of the respondents indicated that trees within the farms were reducing while 40.4% of respondents stated that trees were increasing on the farms. The results imply that the tree cover in the study area is reducing and therefore the benefits that farmers obtain from the trees are shrinking hence need to intensify planting of trees for enhanced benefits. During the Focus Group discussions, farmers emphasized the role of markets and prices in determining which trees they grow. Fruit trees have always been planted as there is ready market and that they also provide the household with food while other trees such as Bixa and Casuarina has mainly been planted for commercial use and the land dedicated to the trees have been proportional to the market availability. Farmers also rationalize tree planting in response to the markets and plant more of a particular tree depending on the market availability. The reduction in the number of trees could probably be due to small land sizes that force the farmers to prioritize between planting crops and trees. The results of this study are in agreement with those of Hitchcock (2004), that indicated that food production among smallholders takes the highest priority and farmers allocated more land to food production with the remaining land allocated to commercial tree and crop production which is further scaled down when land becomes increasingly smaller. The results also concur with a study by LeRoy et.al (2017), that asserted that farmers preferred cutting down of old trees on their farms to give way to crop production as they were not able to get ready market for poles due to the slump in the tourism industry which have been major market sink for poles for construction of hotels. The results in Table 6 indicate that the mean land size for private ownership was (3.71±0.286 acres) and for communal land tenure system was 3.80±0.107 for those who practiced tree and crops agroforest system. The respondents who practiced tree/crops/livestock had low size for private (2.20±0.200 acres) and in relation to communal land tenure system (1.70 ±0.153 acres) under the same agroforestry system. The results imply that there was more land allocated to agroforestry practice under the communal land tenure system than under private land tenure system. The results support studies by Quisumbing (2001) and Holden et al. (2009), which indicated that land tenure system is a critical determinant on adoption of agroforestry among rural households. However, a similar study by Mugure et al. (2013), in Busia County contradicts these results as it found that allocation of land by household head to the sons' hindered tree planting because trees claimed a huge portion of the land therefore reducing the land available for the planting trees.

Tree Planted Under Different Agroforestry Systems
In order to establish the average number of trees planted, farmers were asked the number of trees they plant under the different agroforestry systems.
The results were as shown in Figure 2.

Figure 2: Trees planted on the farms under different agroforestry systems
The respondents indicated in Figure 2 that the mean number of trees planted by the 128 (47.39%) respondents who practiced tree/crops agroforestry system were 92.25±15.878 trees while 141 (52.17%) respondents who practiced tree/crops/livestock agroforestry system were 67.31±9.133 trees. The results imply that majority of farmers prefer growing trees and crops alongside keeping livestock. The results are in agreement with studies done by Jerneck and Olsson (2014) showed that farmers plant more tree species on the farm as an insurance against uncertainty in climate and food shortage and to generate income to meet their domestic.

Income Under Agroforestry
The information from the respondents showed that majority of the farmers who practiced tree/crops mostly grew maize (56.2%), beans (18.8%) and cassava (16.1%) while those who practiced tree/crops and livestock also grow maize (69.3%), Cassava (17%) and Beans (5.9%) among other crops as shown in Table 7. The findings depict that majority of the households grew different types of crops. The reason for the crop diversification could be to enhance the capacities of the households to cope with drought adversities. Further inquiry by the FGD revealed that in recent years, rainy seasons had become shorter and often unpredictable a situation that calls for diverse crop production mostly drought resistant crops. The results of this study support other studies by Zomer et al. (2014) and Coulibaly et al. (2015) The results in Figure 3 indicate that yield for oranges (Ksh 4,086) were highest followed by coconuts (Ksh 2,795), mangoes (Ksh 2,295) and Timber (1,200). Fruit trees generated more income followed by makuti and farmers responded by planting more fruit trees and casuarina to generate more income to meet their domestic demands such as food.

Figure 3: Income from Agro forestry products' yield
The results of the study is consistent with other studies that indicated that fruits provide farmers with income and also double up as food therefore enhancing the food security of the farmers (Oduol et al., 2006). The results also correspond to another study that showed that the markets for agroforestry products have been very unreliable and have affected what farmers can produce and sale to generate income (Lindell, 2010). Lindell (ibid) further add that through the markets, farmers have been able to exchange various agroforestry products for income that is then used to purchase food.

DISCUSSION
Land availability is crucial in the practice of agroforestry as it is one of the major factors of production. The land allocated to tree planting affects the quantities of food that farmers will produce and the higher the land allocated to tree planting the higher the food produced. farmers with larger tracts of land allocated more land for tree planting (Nyambari, 2008). The results are also supported by studied done in Zambia where farmer with larger parcels of land were found to be more likely to adopt agroforestry technologies and were more food secure (Waha et al., 2016). The results of this study are in agreement with other studies by Tuffour (2013) andFAO (2014), that asserted that land availability and the trees farmers grow on the farm play a major role in contributing to household food security.
Access to inputs such as high-quality seeds, technologies and seedlings is also important in the practice of agroforestry. Access to latest technologies in agroforestry is limited to few farmers who can afford to reach out for extension services through visits to government offices and use of modern technologies. The low adoption of agroforestry technologies by the rural households has exacerbated food insecurity in Kwale County. The results are in agreement with other findings that revealed that incentives given to farmers by providing fertilizers and pesticides motivates farmers and improves their farming practices and results in high yields therefore enhancing food security (Segnon et. al. 2015). When farmers realize the full potential of agroforestry and intensify on the planting of trees that improve soil fertility, they will reduce reliance on inorganic fertilizers. Different trees provide different products at different times during the year therefore ensuring continuous streams of benefits to farmers all year round. Farmers plant different tree species for different purposes and adopt agroforestry technologies that give them the products that they desire. Majority of the farmers were found to plant trees for commercial purposes (33.1 %) and for food production (32.9 %). The results of the study are in agreement with other studies that showed that farmers plant trees that provide them with income, food, fodder, and wood for domestic and commercial uses (Regmi, 2003;Baul et al., 2015;Gibreel, 2013).
Farmers apply different tree planting arrangements for different products in the practice of agroforestry. The different tree planting arrangements results in different products for different farmers depending on the local conditions. The results are supported by other studies that found that farmers adopt different tree planting configurations to maximize the interactions between crops, livestock and for maximization of beneficial interactions and minimization of negative interactions (Lasco et al., 2016;Lewis, 2002).The farmers in Kwale have been practicing agroforestry for several decades and have been testing different technologies and adopting what works for them and reducing on the technologies that are less beneficial and changing their agroforestry practices depending on the market demand for different products. The case in point is the allocation of more land for planting trees for poles such as casuarina during the periods when the tourism industry is thriving to supply construction materials and cutting down the trees when there is a slump in tourism and allocating more land to crops and fruit trees to improve their food security situation.
Majority of farmers were found to prefer an agroforestry system that integrates trees crops and livestock (52.17%), trees/crops (47.39 %) and trees/livestock (0.44 %). The farmers prefer the agrosilvopastoral system as it provides a wide range of benefits all years round and greatly improving their food security situation. The results are in tandem with other studies conducted in Ethiopia and Tanzania that indicated that the agrosilvopastoral system is popular among farming communities especially when there is proper sequencing and integration of the different components to ensure optimal productivity (Lulandala, 2009).

RECOMMENDATION AND CONCLUSIONS
Majority of famers in Kubo South and Samburu wards practice some form of Agroforestry and are motivated by the various benefits they get including income, food, improvement of soil fertility, shade, fodder for livestock during drought. There is high level of food insecurity among most of the rural households as a result of several factors that interact at household level to affect each household in a very unique way. Based on this research work, the central government, regional and key stakeholders must engage in order to enhance household food security programs, farmers to improve and diversify by planting different types of trees within their farms, more land allocated to tree plantation establishment and finally the planted trees must have positive effect on improving soil productivity as opposed to degradation.

ACKNOWLEDGMENT
I have taken efforts in this project. However, it would not have been possible without the kind support and help of many individuals and organizations. First and foremost, let me that my Employer, Kenya Forestry Research Institute (KEFRI) for giving the opportunity to undertake the study. My supervisors, Dr. Hilda Ong'ayo and Professor Daniel Nyamai mentored and supervised me during the study diligently and with great devotion that made the study a success. The study would not have been possible without the financial support by the World Bank through the Kenya Coastal Development Project, which was implemented by Kenya Marine and Fisheries Research Institute in collaboration with other government agencies and non-Governmental agencies working in the coast Region of Kenya. I would like to express my special gratitude and thanks to Dr. Samson Okoth Ojunga of the Biometrics section for giving me such attention and time. My thanks and appreciations also go to my colleagues in KEFRI and the KCDP project for the great guidance and technical support in developing the project and people who have willingly helped me out with their abilities. My classmates, Masters in Environmental Studies (Community Development) Class and staff at Pwani University prepared a good foundation and learning environment for conducting the study and for this I thank them immensely both individually and collectively. May God bless you all.