Received: 15.12.2020, Revision: 02. 01.2021, Accepted: 19. 01.2021, Published: 30. 01.2021
KIBETU, Dickson Kinoti1 ,HUHO, Julius Mburu2, OUNA, Tom Odhiambo1
1Karatina University, Department of Humanities, P.O. Box 1975, Karatina, Kenya.
2Garissa University, Department of Arts and Social Sciences, P. O. Box 1801, Garissa, Kenya
Abstract: Marketisation of food crops is a pathway to poverty alleviation and food insecurity mitigation in most rural agrarian societies. Income generation and development of food trade entrepreneurial skills positively impacts the livelihoods of mixed small-scale farmers. To understand reasons for the decision to sell food crops by food insecure households within semi-arid Tharaka constituency, data obtained through household survey across the Rainfed Cropping (RFC), Mixed Marginal Farming (MMF) and Marginal Farming (MF) livelihood zones was used. Within Tharaka, malnutrition rates are relatively as high as 30% compared to the national average of 20.9%. Despite various food security and nutrition interventions, majority of the households still lack enough food to meet their dietary needs. This is contributed by the household’s tendency to sell off food crops early in the season after harvesting. Descriptive data presentation and Tobit regression were applied in data analysis. Results showed that majority of the households in the Rain Fed Cropping livelihood zone (47%) engaged in marketisation of their food crops compared to those in the Mixed Marginal Farming (31%) and Marginal Farming livelihood zones (22%). Besides marketisation, the findings indicate that food production and consumption is also skewed across the three livelihood zones. This is explained by demographic distribution which show that 33% of the population live in RFC, a zone where high levels of food crop marketisation was reported with the rest 67% of the population found in MMF and MF zones. Education levels and market price were to a large extent found to influence decision to sell food crops by household heads in the study area. To upscale the current food crop commercialisation levels in the constituency, agro-policy interventions especially regulating food crop prices through market access fee subsidization for produce farmers and the improving of market related information communication methods are important.
Keywords: Food crops, House hold, Marketisation, Livelihood, policy intervention, Semi-arid.
Market-oriented production in rural areas is gaining momentum owing to the fact that livelihoods in these areas depends on agriculture for sustenance. Dynamisms introduced by devolution of resources and climate variability are partly influencing local food production systems and consumption patterns as well. Rural households consume about half of their produce and sell the rest mainly in retail open-air markets within the rural areas (Omiti et al., 2011). Agriculture provides the main source of income for rural households even as it employs close to 60 % of all population in sub Saharan Africa (Thornton, 2011). Due to the promise agriculture has in addressing food insecurity, market-oriented production strategies of addressing food and nutrition security are being fronted by various countries in Sub Saharan Africa. One such is the national food and nutrition security policy of Kenya which retaliates the need for local market-oriented food production and provision of reliable market to local farmers (GoK, 2011). Increased population growth and urbanisation has pushed the demand for local foods up leading to intensification of agriculture in both rural and peri-urban areas. Policy makers and researchers have too advocated for market-oriented agricultural production to ensure the welfare of farmers and sustainable household food security are promoted (Godfray et al., 2010; Pingali 1997).
Most studies done in the field of market-oriented production have clearly shown the potential of commercialised food crop production towards realisation of food security for small holder households. A study by Misselhorn et al., 2012 on effects of transitioning to market production by subsistence households found out that low market supply and limited access causes food insecurity amongst resource poor households. Ntakyo & Berge, 2009 in their study on commercial production of rice in Uganda revealed evidence that market production promotes dietary diversity for smallholder households by increasing access to different food stuffs. Other studies have shown that for majority of small holder farmers in rural Africa, food crops comprise also the major market crops (Carletto et al., 2017; Von Braun 1995).
Although marketability of food crops amongst small scale farmers provides them opportunity to raise better income and improve their livelihoods, issues to do with market prices and access to market information continue to be key challenges (Nyambok, 2014; Rutto, 2015). An analysis of the performance of agricultural markets in the country show how poor participation by small scale farmers contributes to inefficient and imperfect food markets (Wangare, 2014; Vermeulen et al., 2012). It is against this background our study sought to analyse factors that informed decision by households in some of the most poverty stricken as well as food insecure semi-arid areas to sell their subsistence crops. Market channels used were also explored to understand their role in marketisation of food crops. In particular, maize and millet crops were studied as they form main staple crops consumed by majority of people in the region and even traded for cash income by most households.
This would inform causes likely to drive commercialisation of principal food crops in rural food deficit arid and semi-arid areas. In an effort to promote food availability through food crop trade, findings from this study could shed more light on intra household food utilisation variations. This is an important starting point for formulation of policy on local food crop trade development and how open-air food markets can be harnessed to address local food deficits through regulation by counties as enshrined in the Fourth Schedule, Part 2 and paragraph 7 of the Kenya constitution 2010.
Tharaka constituency is located along the equator on latitude 0007’ and 0026’ South and longitude 370 19’ and 370 46’ East (Figure 1). The constituency has a population of 130,098 persons and 27493 households (KNBS, 2009). This area is a semi-arid land and receives less than 700mm of rainfall per annum. Based on agro-ecological characteristics, the area is divided into three livelihood cluster zones of; mixed farming (MF), marginal mixed farming (MMF) and rain-fed cropping (RFC). Farmers in this region are small scale subsistence crop producers owning on average of 2.9 hectares. Major food crops grown are also traded as cash crops and include maize, cowpeas, green grams, pearl millet, sorghum and pigeon peas. Besides crop production, households also engage in livestock herding (Smucker & Binsey, 2008). Principal food crops are mostly maize and millet grains valued as important food security crops in the region although also traded for household cash income.
Figure is Available in PDF Format
Figure1: Map showing location of Tharaka Constituency
MATERIALS AND METHODS
A two-stage cluster sampling was used where sub locations were purposively selected in the first stage based on road network and agro-ecological cropping potential. In the second stage, villages were selected from the sub locations using probability proportionate to household size sampling technique. Participant households were selected at the village level through systematic random sampling method. Questionnaires were then administered by trained research assistants across the sampled 395 households stratified according to the Marginal Mixed Farming (MMF), Rain Fed cropping (RFC) and Mixed Farming (MM) livelihood zones.
A total of 25 sub locations were identified in the first stage and 27 villages proportionately sampled across the three livelihood zones in the second stage from which the participating 395 households were selected.
Data analysis and presentation Methods
Due to the nature of quantitative and qualitative data acquired from the field surveys, mixed analysis method was applied. Descriptive analysis was used where bar graphs and tables were used to present results of socio-economic variables from interviewed households. Quantitative analysis of explanatory variables for marketing food crops was done using Tobit regression in STATA version 16.0. Tobit regression is a strong kind of regression analysis which gives a binary outcome for dependent variable. Nine explanatory variables were identified and tested in this particular study.
DECISION TO SELL = 3.71- (0.094) EDUCATION LEVEL + (-0.228) HEADSHIP + (-0.236) ICT USED+ (-0.239) LAND OWNED + (-0.205) QUANTITY HARVESTED + (0.033) TRANSPORT + (0.106) PRICE + (-0.132) DISTANCE + (-0.065) HOUSEHOLD SIZE
RESULTS AND DISCUSSION
Findings on determinants to sell in the context of household socio-economic attributes as well as the marketing channels used are discussed here below.
Socio-economic attributes of households
The study had more female participants compared to their male counterparts at 64.1 % and males at 35.9% respectively. This parity could be explained by the fact that women are majorly engaged in home making and other reproductive roles like nursing babies, cooking and tending to livestock and crops. These responsibilities mostly tend to confine women at home unlike their male counterparts who engage in productive roles away from home. Therefore, this observation explains why more females participated in the study unlike male counterparts who at the time could not be available for the interviews. These findings closely mirror the 2009 Kenya National Population and Housing Census report which put the females at 51.7% and males at 48.3% of the total population in Tharaka district.
Household Headship Roles
Who headed a given household influenced how decisions made impacted on production, consumption and sale of food stuffs by the said household. Results indicated that majority of households within Tharaka are headed by males at 73.2 % as compared to 26.8% which had a female acting as the head. These findings are interesting given that in this region despite female population being slightly higher than males at 67,215 compared to males at 62,883 persons, males still head and control key household decisions. This is a clear indication that Tharaka community is a patriarchal society where a man heads the family besides owning and controlling family assets especially land and livestock.
A large proportion of population is between 31- 40 years (37.2%) followed by those of 41-50 years (29.1%) as indicated in Figure 2. These are economically active and productive age groups engaged in agricultural production, marketing and related agri-businesses activities. Although these age groups also comprised the largest population associated with out migration, they often returned home from off-farm employment during tilling and planting seasons to engage in farming activities. These in-migrations were mostly observed during the October-November-December long rain season which is reliable for cropping. Those in the age bracket of 21-30 accounted for about 22% while those less than 20 years and above 50 years were 3.8% and 7.8 % respectively. These results agree with findings of a study by National Council for Population and Development, 2017 which established that Tharaka-Nithi County has a high proportion of population in the working ages (52.0 %).
Figure is Available in PDF Format
Figure 2: Age distribution of the interviewed population
Levels of Education
In order to assess literacy levels of participant respondents; primary education, secondary education, college/ university education and No education categories were used. Survey results as shown in Table.1 shows that a significant majority of interviewed respondents had primary education (42.5%). 28.6 % of the respondents had attained secondary education while those without any formal education accounted for 22.3%. Only 9.6 % of the respondents had attained tertiary education which comprised of either college or university education. Education of household members influence farm management and marketing skills which results in improved decisions (Makhura et al.,, 2001). Several studies have shown that education levels can improve competitiveness of the farmers (Moyo, 2010; Enitan, 2010; Mohammed and Ortmann, 2005). Therefore, education is a determinant in decision making for households. The findings are consistent with those of a study by KNBS and SID in 2013 which puts illiteracy levels in Tharaka constituency at 22.0%.
Table 1: Proportion of education levels for participant respondent
Level of Education
Proportion in Percent (%)
In order to understand effects of household size on quantity of foodstuffs sold in local markets, information on the number of individuals living with a household head was collected. As indicated in Figure 3 below, about 2.0% of households reached by this survey said they lived with at least a member of the family, 22.8% had between 2 to 3 members living with them. Most households in the study area had between 4 to 6 members. This accounted for 42.3% of the sampled household population. Only 32.9% of the households had more than 7 members living with them. Findings from this survey reveal that most households in the study area have between four and six family members living with the house head at any time. Although it means cheap labour is available, a large household size may negatively affect marketisation as less food stuff is retained by the said household owing to increased consumption. In case of limited food stocks within the family especially during the planting and subsequent harvesting period, huge financial resources are also required to procure food for a large household compared to a lean one.
Figure is Available in PDF Format
Figure 3: Household membership size as sampled in the study area
Marketing Channels mostly used by households
Close analysis of marketing channels gives an opportunity to understand market functions, linkages and intermediaries acting within a market chain (Zeberga, 2010). Survey results revealed that most households produce maize and millet food crops for subsistence although surplus was sold in the local markets. Marketing models examined showed formal and informal channels were mostly used (Table 2). Formal channels comprised of licensed agent traders who bought maize and millet for grain handling co-operatives operating in Meru town. These traders offered fixed prices across all buying centers where they owned a stall in selected periodic markets. Large scale grain millers identified local traders who established satellite buying centers within local markets to target households selling surplus foodstuffs during harvesting season. Identified formal channels operated optimally above peak during harvesting season when there was surplus production. During this period, market prices for the cereals were relatively low due to oversupply in the local markets. Most cereal buying centers were located within RainFed Cropping Zone and a few in the Mixed Farming Zone. Informal market channels especially selling to others households in the village were also another common avenue by which households accessed food during food deficit seasons.
Table 2: Channels used by households to sell and buy their food crops across the livelihood zones
RFC1 MMF2 MF3
Number % Number % Number %
Farm gate Assemblers 143 67 23 18 30 58
Institutions 37 17 45 34 9 17
Millers 2 1 10 8 1 2
Cereal traders 18 9 24 18 7 13
Other Households 12 6 29 22 5 10
212 100 131 100 52 100
1. RainFed Cropping; 2. Marginal Mixed Farming; 3. Mixed Farming
As seen, the number of cereal grain buyers in the villages varied spatially across the three cluster livelihood zones. For instance, in the RainFed Cropping zone, farm gate retail assemblers were dominant accounting for 67%, followed by institutions at 37 %. The same situation was observed within Marginal Mixed Livelihood where the proportion of farm gate assemblers comprised of 23 % while institutions buying from farmers accounted for 45%. This scenario was also mirrored in the Mixed Farming zone where 58 % of small retail traders were main farm gate assemblers followed by local institutions which accounted for 17% of farmers marketing channel outlets. It is therefore factual to claim that farm gate assemblers are the major buyers and sellers of cereal produce in semi-arid rural Tharaka Sub County. Their dominance advantage is because they buy millet and maize in small quantities preferably in a 1kg tin, quantities by which most households sell their surplus. This makes them more preferred than other marketing outlets for their availability.
The high presence of aggregators in the RainFed Cropping zone is due to high commercialization of farm produce by households who engage in green grams and peanuts farming as alternative cash crops. This means that many households can dispose maize and millet held in surplus since they have other crops to sell for income during lean season. Prices of Millet and Maize fluctuated across seasons with Maize retailing at an average of Kshs 25/Kg during harvest season and kshs50/kg during planting season. Millet was sold at Kshs 15/Kg during harvest season and at Kshs 100/Kg during Planting season. Price seasonal variation was attributed to supply and demand forces where plenty of the two crops in the local markets during surplus seasons (February, March, April for long rains and May, June, July for short rains) attracted low market prices while lean seasons of January, August September and October pushed the market prices of millet and maize to almost double.
When inquired about the quantities by which households sold maize and millet which they produced, 48.5% of sampled households sold their produce in a 1kg tin specially to farm gate brokers. About 32.3% of households said they sold their produce in Debes to cereal traders operating stores in open air markets of Mukothima, Gatunga, Tunyai and Marimanti. Few households (9.9%) sold maize and millet in a 90kg bag. Those households selling in 90kg baggage cropped over 5acres of land and were mostly engaged in contract farming. They were supplying to schools, brewers, millers and community-based organisations. Institutions were another important marketing channel outlet significantly influencing levels of foodstuff commercialization. Close to 68% of participants in the study expressed importance of institutions as an outlet in millet-maize agri-food marketing and distribution value chain, an observation also made by Christopher, 2010 who reiterated institutions play a pivotal role in the success of value chains.
Direct reference was made to schools which topped this category as it was observed that most poor farming households exchanged farm produce for their children’s school fees. Although mention was made indirectly to the schools feeding program in day primary schools by World Food Program in the area, most schools especially days mixed and boarding secondary schools allowed parents to bring in maize, green grams and beans in place of monies. This occurrence was cited by most respondents within MF (34%), RFC (17%) and MMF (17%) who affirmed this was oftenly practiced during second and third term coinciding with lean seasons when most households had limited off farm income generating activities.
Generally, more assemblers were found in RFC zone compared to MF and MMF due to high production of maize crop within Nkondi, Mwanyani, Thiiti and Mukothima sub locations. Maize a staple crop in high demand could be sourced easily from farmers in these areas and in large quantities compared to millet which could be gotten from other livelihood zones. In the Marginal Mixed Farming livelihood zone for instance, families buying from other households were more (22%) when compared to the same in RainFed Cropping Zone (6%) and Marginal Mixed Farming (10%). This is explained by constant rain failure in this area resulting to crop failure and over reliance with livestock. Interesting to note that 18% of wholesale and retail cereal traders were mostly concentrated within Marginal Mixed Farming livelihood zone. It was revealed that these cereal traders established retail stores to sell produce procured in the RFC and MF zones because most households were food deficit and therefore sold livestock to purchase food.
These periodic traders were agents contracted by whole sale cereal dealers in Meru, Chuka and Nanyuki towns. Small scale cereal traders were more due to the bulking function of aggregating produce brought in by households from RainFed cropping and Mixed Farming as they perceived demand and better prices in the Marginal mixed farming livelihood zones.
Determinants of intensity to sell by households
Different factors were identified through preliminary analysis as influencing decision to market food crops. These factors were regressed in STATA software Version 16.0 to get statistical significance for each factor identified. (Table3).
Market Outlet Coef. Std. Err. t P>|t| Marginal Effect(dy/dx)
Education Level 0.094 0.060 1.55 0.121** 0.093
Headship -0.228 0.160 -1.42 0.156** -0.227
ICT used -0.236 0.109 -2.16 0.031*** -0.236
Land owned -0.239 0.094 -2.55 0.011 -0.239
Quantity Harvested -0.205 0.091 -2.26 0.024 -0.205
Transport -0.033 0.057 -0.58 0.561*** -0.033
Price 0.106 0.163 0.65 0.514** 0.106
Distance -0.132 0.066 -2.01 0.045*** -0.131
Household Size -0.065 0.077 -0.84 0.402*** -0.064
Constant 3.710 0.326 11.38 0.000
Table 3: Explanatory factors regressed in Tobit model
Market price is a significant factor influencing decision to sell by rural households in this area. It was found that a slight variation in the price of food stuff influence decision on how much to sell and buy from markets. Majority of the population depends on open air markets for most of their food needs. From the analysis, a unit increase in price change will positively influence a household to use available market outlet by 10.6%. Therefore, an upward increase in food stuff prices in the local open-air markets will result to more households selling their foodstuffs. This is attributed to the fact that most households depend on food for income and lack of adequate food storage facilities also triggers disposal of food stuffs at low prices.
Level of education of a household head was statistically significant (p<0.05) factor positively influencing decision to sell food stuff by the said household. This implies that an educated family head was likely to make informed decision on who, when and where to sell household produce by about 9.3%. Those with at least secondary education were found to understand merits and demerits of various market channels available and could decide on which marketing outlet was preferable. These findings are consistent with those of Makhura et al.,, 2001 who noted that formal education increases household understanding of market dynamics improving decision about the amount of output sold. This also means educated household heads can read, understand and utilise information related to marketing as communicated through mass media, phones and famer magazines.
Coefficient of household headship gender was significant (P<0.05) and negatively correlated with the decision by household to sell their foodstuffs. This is indicative that although headship is an important factor influencing family decisions, that of to sell food crops or not was however dependent on the whether the household was headed by a male or female. In this study, household headship was the dummy variable with male headed household taking value 0 while female headed household took value 1. Out of 395 households sampled, 288 households were male headed and 107 female headed. This means male headed households were likely to sell food stuffs than those families headed by a female. Findings from this study correlate to those by Cunningham et al., 2008 who found that men are likely to sell more when prices are high unlike their female counterparts who prefer to store more food for household self-sufficiency.
Access and use of ICT
The kind of ICT used by household to access marketing information was found to influence decision on marketing of foodstuff. Coefficient of ICT used was however significant at p<0.1 but negative. This implies that although ICT use increases awareness on marketing options available to a household, its use does not necessary translate into more food stuff being sold or choice of a particular market channel. This negative coefficient could be attributed to the fact that most households do not use formal sources of information like mobile phones or Radio &TV sets to access market related information in particular. This is partly attributed to lack of ICT skills amongst majority of the respondents. Similarly, ICT infrastructure in the County of Tharaka-Nithi is also not well developed as noted by Ameru, 2019. The few who can access ICT skills do not use it frequently due to high cost associated with accessing market information via mobile phone supported apps. The general indication is that informal sources of information especially neighbours are widely used in the study area to get market related information.
Household size had a coefficient of-0.065 at significant of (P<0.1) indicating that it negatively correlates to the type of market outlet chosen by a household. Household size influences production and consumption levels as observed by Alene et al., 2008. Most households totalling to 166 in number from the study area had between four to six family members. These comprised the households with large family members. The negative coefficient value implies large household produces less output owing to the small parcels of land but ends up consuming a higher proportion decreasing proportions left for sale. For this very reason, large households often prefer to conserve their produce for family use and if they decide to sell then it is in small quantities and at farmgate. From the results the likelihood to choose a given market channel decreased when the number of the family members increased. The above findings are consistent with those of Temesgen et al., (2017) who found that as the number of families increased the probability to participate in onion production decreased. In contrary to this, Efa and Tura (2018) indicated that large family size enables better labour endowment so that households are in a position to look for profitable market outlets.
Transport was found to be statistically significant (P<0.1) factor in explaining household decision to sell or not. However, transport negatively correlated with market outlet chosen by a household. It implies that households near food retail markets will incur less transport costs in moving their produce to markets than those located far away from available market. In rural Tharaka, human potage is a common transport typo used by farmers and households to take their produce in the market. Due to this, only small quantities are brought to the markets by households which are far located. As a consequence, most households prefer selling at farm gate and in near open-air markets due to problems associated with transportation. It simply means choosing far away markets translates into high transport costs of moving produce from farms to the market. The poor road network in the study area hinders mobility and more so spatial interaction across villages. This is a constraint which lessens marketisation of farm produce in the study area.
Distances negatively influence decision to sell produce by households. Distance factor had a negative coefficient of -0.132 at a statistical significance of P<0.1. Geographic isolation of households due to distance from markets adversely affects quantity sold by the concerned families and by extension dictates final market price of commodities. In the study area 40% of the households live within 5-10km distance from an open-air market. Distance also determines transport cost associated with moving farm commodities to the market. In most instances transport cost per unit distance covered increases with the size of marketable load (Omiti et al., 2009). These results corroborate with those of Key et al., (2000) who found that distance to the market negatively influences the decision to participate in markets by farming households. Since most households in the study area are peasant farmers, the cost of moving their small surplus to the markets especially transportation costs discourage them from using formal market channels
Households in rural Tharaka constituency consume almost half of their food crops and sell the remaining through open air markets, retail cereal traders and farm gate assemblers. Spatially, more households in the Rain Fed cropping zone engage in marketisation of their produce as compared to those in the mixed marginal farming and marginal farming livelihood zones. Decision to sell food crops by households was influenced positively by the education levels of household heads as well as prevailing market prices of foodstuffs. Regression analysis showed that transport, household size, distance to nearest market, household headship and ICT type used to access market related information had negative sign implying that these factors least influenced marketing decisions. Although land size and quantity harvested by household were considered to impact decision to sell, they were found to be statistically insignificant in influencing household’s decision to sell.
Conclusively, it is evident that higher market price for food stuffs and education increases the propensity to market food crops by households across the constituency' livelihood zones. Policy interventions which should be developed to promote marketisation of local crops involve; setting and regulating food stuff prices by the concerned county government. This should safeguard households from undue exploitation. Secondly, on-farm demonstration training centres should be established and equipped to educate households on innovative farming systems.
Thirdly, it is extremely important to enhance local market integration and access through improvement of communication and transport infrastructure in order to enhance food trade and agri businesses.
Lastly, there is a need for a spatially infused research to unearth trends and motivation for the heightened land use and land cover (LULC) changes in the study area. This will explore the implications of these changes on sustainable food production in the wake of increased land sub-divisions amongst small holder families.
Alene, A., Manyong, V., Omanya, G., Mignouma, H., Bokanga, M., & Odhiambo, G. (2008). Smallholder market participation under transactions costs: Maize supply and fertilizer demand in Kenya. Food Policy, 33, 318–328. DOI: https://doi.org/10.1016/j.foodpol.2007.12.001
Carletto, C., Corral, P., & Guelfi, A. (2017). Agricultural commercialization and nutrition revisited: Empirical evidence from three African countries. Food Policy, 67, 106–118. DOI: https://doi.org/10.1016/j.foodpol.2016.09.020
Cunningham, L., Brown, B., Anderson, K., & Tostao, E. (2008). Gender differences in marketing styles. Agricultural Economics, 38(1),1-7. DOI: https://doi.org/10.1111/j.1574-0862.2007.00225.x
Enitan, F. (2010). Influence of socio-economic characteristics on use of modern cassava processing technologies among women processors in Ogun State, Nigeria. Journal of Social Sciences, 24(1), 43-50.DOI: https://doi.org/10.1080/09718923.2010.11892835
Godfray, H., Beddington, J., Crute, I., Haddad, L., Lawrence, D., Muir, J., Pretty, J., Thomas, S., & Toulmin, C. (2010). Food security: The challenge of feeding 9 billion people. Science, 327(5967), 812–818. DOI: 10.1126/science.1185383
Kenya National Bureau of Statistics. (2009). Tharaka District Multiple Indicator Cluster Survey 2008
Key, N., Sadoulet, E., & De janvry, A. (2000). Transactions Costs and Agricultural Household Supply Response. American Journal of Agricultural Economics, 82(2), 245-259 DOI: https://doi.org/10.1111/0002-9092.00022
Makhura, M. (2001). Overcoming transaction costs barriers to market participation of smallholder farmers in the Northern Province of South Africa. Unpublished PhD thesis. Pretoria: University of Pretoria.
Mohammed, M., & Ortmann, G. (2005). Factors Influencing Adoption of Livestock Insurance by Commercial Dairy Farmers in Three Zobatat of Eritrea. Aggreko, 44(2),106-113 DOI: https://doi.org/10.1080/03031853.2005.9523708
Moyo, T. (2010). Determinants of participation of smallholder farmers in the marketing of small grains and strategies for improving their participation in the Limpopo River Basin of Zimbabwe. MSc thesis. University of Pretoria: Pretoria.
Nyambok, L.O. (2014). An evaluation of market information system of finger millet farmers in Teso south district, Kenya. Unpublished MSc Thesis, University of Nairobi
Omiti, J., Otieno. D., Nyanamba. T., & Mcclough, E. (2009). Factors influencing the intensity of market participation by smallholder farmers: A case study of rural and peri-urban areas of Kenya. Afjare, 3(1), 57-82
Pingali, P. L. (1997). From subsistence to commercial production systems: The transformation of Asian agriculture. American Journal of Agricultural Economics, 79: 628–634.DOI: https://doi.org/10.2307/1244162
Rutto, C.P (2015). Factors influencing marketability of maize produce among smallholder farmers’ in soy sub-county Uasin Gishu- Kenya. Unpublished MA dissertation, University of Nairobi.
Smucker. A., & Wisner, B.(2008). Changing household responses to drought in Tharaka, Kenya: vulnerability, persistence and challenge. Blackwell Publishing, USA. pp. 190-215.DOI: https://doi.org/10.1111/j.1467-7717.2007.01035.x
Thornton, P. K., Jones, P. G., Ericksen, P. J., & Challinor, A. J. (2011). Agriculture and food systems in sub-Saharan Africa in a 4 C+ world. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 369(1934), 117–136. DOI: https://doi.org/10.1098/rsta.2010.0246
Vermeulen, J., Aggarwal, P., Ainslie,A., Angelone, C., Campbell, B., Challinor, A., & Kristjanson, P. (2012). Options for support to agriculture and food security under climate change. Environmental Science & Policy, 15(1), 136–144. DOI: https://doi.org/10.1016/j.envsci.2011.09.003
Von Braun, J. (1995). Agricultural commercialization: Impacts on income and nutrition and implications for policy. Food Policy, 20(3), 187-202. DOI: https://doi.org/10.1016/0306-9192(95)00013-5
Wangare, L.N (2009). An analysis of the efficiency of rural food markets in the eastern central highlands of Kenya. Unpublished Ph.D. Thesis, University of Nairobi
Zeberga, A. (2010). Analysis of poultry marketing chain: The case of Dale and Alaba “special” Woredas in SNNPRS of Ethiopia. Unpublished MSc Dissertation, Haramaya University, Ethiopia.