Enhancing Household Livelihood Outcomes through Resilient Agri-Food Systems among Small-scale farmers in Kitui County, Kenya

The practice of resilient agri-food systems, especially in arid and semi-arid areas, is considered key to attaining sustainable livelihoods. Changes in climatic conditions have forced small-scale farmers to develop adaptive capacities related to food security. This study was done in Nguni ward Kitui County involving 246 small-scale farmers, two focus groups discussion, and four key informants. The sampled participants had practiced resilient agri-food systems for the last five years from the time of data collection. A gender dynamic was established after 71.5% of the respondents were found to be female and 28.5% male. More females practiced resilient agri-food systems but did not own productive assets like land. A statistically significant relationship between resilient agri-food systems and household livelihood outcomes was determined where the probability value was less at a 95% confidence level (0.01<0.05). Further analysis showed a positive association between household income and sex χ2=9.551 where p (0.008) was less than 0.05. In conclusion, resilient agri-food systems practices such as conservation agriculture positively influence the household livelihood outcomes like diversification of livelihoods, reduced prevalence of human, crop, and livestock diseases, and improved ability to cope with shocks and trends. The study recommends the provision of incentives, insurance, and social safety nets to encourage small-scale farmers to practice resilient agri-food systems.


INTRODUCTION
Globally, the primary source of livelihood for nearly four billion people is food systems (United Nations, 2020). Food is vital to human beings, and this has been emphasized in the Sustainable Development Goal two on zero hunger, achieving food security, and promoting sustainable agriculture. The demand for safe and adequate food is on the rise since it is estimated that the global population will grow from 7.8 billion in 2020 to around 9.9 billion people by 2050 (Population Reference Bureau, 2021). The transformation of agri-food systems is discussed almost in all the development forums by scholars and policymakers as a subject of interest (Béné et al., 2019). The major concern is the sustainability of modern food systems, including how and where they are produced, packaged, processed, distributed, consumed, and disposed.
Oxford Dictionary of Human Geography defines the agri-food system as "the totality of actors involved in the production, distribution, and consumption of food, the relations between them, and the regulatory apparatus governing these arrangements" (Castree et al., 2013 pg. 2). Through food systems, livelihoods are created. However, to realize this sustainably, there is a need for a global food revolution failure to which a twin burden will be created. The burden will entail feeding humanity and struggling to live within the polluted environment (Steffen et al., 2015). Food systems are affected by environmental, technological, economic, political, and socio-cultural factors. These drivers are vital to ensuring the sustainability of food systems. Drivers like economic and political factors have allowed the domination of food systems by capitalists worldwide. Large agribusinesses focus on maximizing profits by buying food at low prices from the producers. To meet the demand of these large corporations, the producers, in return, use extractive farming practices without considering human and planet health (Béné et al., 2019). A shift to resilient food systems indicates a sustainable change and introduction of practices geared towards positive outcomes such as reducing food wastage, food loss, equity in food access, creation of value chains, and biodiversity protection (Chaboud & Daviron, 2017).
Research, development, governance, and regulations are among the key activities embraced in agri-food systems. For instance, Hudson et al. (2020) found that resilience to extreme weather in Europe was attributed to recovery, resistance, and adaptive capacity. In general, Lipper et al. (2021) posit that adaptation and adoption of innovative methodologies in agriculture call for transformation in food systems. In Uttarakhand hills-India, Bisht (2021) opines that there are small-scale farmers embracing resilience to climate change through various ecological farming methods such as organic farming and polycultures. According to Lu et al. (2022), China has been trying to achieve food security not only within its borders but also across the globe. The country has food security measures through transformative food systems. Some of the functional areas of China's resilient agri-foods systems include land consolidation, creation of food reserves, value addition, and diverse nutrition foods innovations.
In the African context, Sitko and Jayne (2018) argue that challenges such as declining productivity, soil degradation, and weak African food systems, irresponsiveness to climate change need new dimensions in sustainable and innovative food systems methods. Makate et al. (2019) posit that in South Africa, multiple factors affect the level of resilience in agricultural production, including credit access, education, training, and market deficits.
There are still many gaps in the practice of resilient agri-food systems (Gillespie et al., 2019). For instance, the way food is produced, packaged, processed, distributed, consumed, and disposed of is ecologically and socially unsustainable. Food's ecological and social meaning is slowly being lost, leading to changes that endanger human and planetary health (Aude, 2020). This has posed complex challenges such as chronic diseases, environmental degradation, inequality, and malnutrition (Duncan et al., 2021). In addition, the practice of extractive farming by small-scale farmers has contributed to climate change, affecting agri-food systems' sustainability. Amicable solutions to these challenges are inadequate. Therefore, there is a need to change the patterns in the agri-food systems for resilient food pathways to be realized. The effects of these on household livelihood outcomes also need to be investigated. Kitui County is among the arid and semi-arid areas in Kenya, and over the years, disarray concerning food systems has been witnessed. However, smallscale farmers in this region survive through resilient practices. Together with the paucity of empirical data on resilient agri-food systems and their influence on household livelihood outcomes, these fundamental reasons necessitated carrying out this study.

METHODOLOGY
The study was done in Nguni ward, Kitui County, in Kenya. The area receives minimal rainfall and is characterized by rocks and underground salty water. Nguni ward was purposively sampled because despite the ecological challenges, some farmers use resilient practices to produce food. The study employed quantitative and qualitative approaches during data collection between September to December 2021. A list of 915 small-scale farmers was generated purposively by identifying those who practiced resilient agri-food systems consistently for the last five years from the data collection period. Simple random sampling was used through a random number generator to sample 278 smallscale farmers, but 246 formed the response rate as 32 respondents were not available in their households. A household questionnaire was designed to capture respondents' demographics, resilient agri-food systems, and household outcomes. The household survey was supplemented with two qualitative methods: focus group discussion and interview. A total of two Focus Group Discussions were conducted, each with ten participants. Key informant interviews were conducted with the following participants: Nguni ward county administrator, community research leader, agricultural extension officer, and a project coordinator from Caritas-Kitui region-an NGO focusing on resilience livelihoods in dry lands. Quantitative data was analysed using SPSS and hypothesis tested through Chi-square and Analysis of Variance (ANOVA). Qualitative data were transcribed and used to support the quantitative data during the interpretation of the findings.

Demographic Characteristics of the Respondents
Age, sex, and income variables had a significant impact on the study. Table 1 presents data on these demographics.  Table 1, 71.5% of the respondents who practiced resilient agri-food systems were female, while 28.5% were male. The variance in gender distribution is based on the fact that women, especially in Africa, are major suppliers of labour on the farm. Danso-Wiredu (2018) in her study on gendered dynamics and reciprocity in fishing communities in Ghana, confirms that work is highly divided on a gender basis. As in the study area, Kitui, such divisions have become entangled in peoples' everyday lives. This also corresponds with findings from Zimbabwe, which underscores greatly that after the dilapidation of the Zimbabwe British relations, women have played an important role in the revival of the country's agricultural sector. In particular, many women could own land, which allowed them to be involved in food production and tobacco export (Bungu, 2019).
Age is also considered an important demographic variable when analysing agri-food systems. It sheds light on the productive categories that should be given more attention when improving food systems. The majority of the respondents (48.8%) were between 38-47 years. The study findings also revealed that most of the respondents (45.9%) earned a monthly income in Kes 0-5000, while a small percentage (20.7%) earned above Kes 10,001. These earnings were largely from the practice of resilient agri-food systems considering that the study area is dry and receives minimal rainfall.

Resilient Agri-Food Practices, Opportunities, and Interventions
The study also analysed the practices, opportunities, and interventions for the small-scale farmers practicing resilient agri-food systems. The findings on this are presented in Table 2.  Table 2 shows that most of the respondents (57.7%) were trained by non-profit organizations (NPOs) on resilient agri-food practices. Only a small percentage of the respondents (6.1%) used social media to learn the same practices. This could be attributed to a lack of network in the study area and respondents' ignorance of social media usage. About 28.9% of the respondents were trained by their peer farmers, while government extension officers taught 7.3%.
Following the training, 38% of the respondents started practicing conservation agriculture as their main source of livelihood. Some of these conservation agriculture practices include; the use of organic fertilizers, minimum tillage of the land, and cover crops and sunken beds. Poultry farming was also practiced by 31.7% of the respondents, followed by apiculture (beekeeping) at 17.1%, kitchen garden at 11.8%, agroforestry at 6.1%, and livestock integration at 2.8%. Few respondents (2.4%) practiced aquaculture, which could be attributed to the lack of water in the area.
The respondents also needed extra training to improve their livelihood outcomes at the household level. The majority of the respondents (28.9%) wanted to be taught how to market and sell their produce without intermediaries. Availability of water being a big challenge in the study area, 20.3% of the respondents requested to be trained on rain harvesting techniques to enable them to do farming. Considering that most of the respondents practiced conservation agriculture, 17.5% requested training on how to produce organic fertilizers in large quantities. Other areas that needed further training include the incubation process (9.3%), record keeping (8.9%), handling of chicks (8.1%), and vaccination of chickens (6.9%). One of the members of the Focus Group Discussion added that extra training was needed because they faced a lot of challenges on their farms: "…. practicing resilient agri-food systems is not easy, we need more training on it. We lack consistent agricultural extension services and this affect the way we produce food. Market prices within our locality are also discouraging and our wish has always been to sell our produce in big towns or even export to outside countries. Some of our crops wither because the water available is salty. How we wish government could do something and provide reliable water…" (Participant 005 in FGD 1).
The respondents also made general recommendations to support their food systems. For example, 34.1% of the respondents requested nonsaline water for irrigation, while 19.1% called for development agencies to intervene and provide good infrastructures such as roads and mobile networks. Provision of aggregation centres was also considered very urgent by 18.7% of the respondents, as 18.3% needed linkages to markets, training, and farm inputs. A small percentage of the respondents (9.8%) wanted to be guided on how to get certification for the packaging of their produce. This also came out strongly from an apiculture farmer in the Focus Group Discussion and Nguni ward county administrator: The study findings depict the existing literature that describes resilience to maintain structures around a lifestyle. The same description is shown in the works of Ahmed et al. (2022) who conducted a study in the Rift Valley region of Ethiopia. The researchers point out the drought periods that have left the community without food, weakening their livelihoods. In efforts to estimate the level of vulnerability among the communities, the study used spatial and metrological data to determine the level of exposure, adaptive capacities of the communities, and sensitivity. The study findings showed that adaptive capacity significantly varied from one region to another, where the regions with higher vulnerability in livelihoods were aggravated by drought. The study recommended climate-smart farming as an intervention measure. However, the study only showed the levels of vulnerability rather than creating an intervention like the one in the current study on the use of Resilient Agri-Food Systems (RAFS).
The findings reflect the posting of Ortiz and Peris (2022) who argued that family farmers have been innovating ideas that would create sustainability in food systems. Using an association of farmers to practice resilient methods in farming in Guatemala, the study was guided by transformative capacity building among the farmers in rural settings. The study used Focus Group Discussions and interviews to collect data from farming stakeholders. The findings indicated that farmers' organizations and associations were key in cultivating innovations in the agriculture sector. Under the umbrella of farmers' organization, the farmers attracted technical transformation in innovative agriculture, creating sustainability. The study also supported that, partnerships with institutions and other relevant networking were key interventions in resilient agri-methods in Guatemala. The study also identified gaps ascribed around generational dimensions.

Household Livelihood Outcomes
Data on household livelihood outcomes resulting from practicing resilient agri-food systems were analysed descriptively using frequencies and percentages. The questionnaire was designed with statements measuring the extent the respondents had experiences with each item. These were measured on a Likert scale of 1-5 where 1 = none at all, 2=little extent, 3=moderate extent, 4=large extent, 5=very large extent. The results are presented in Table 3. Source: Survey data Table 3 shows that the increase of assets was moderate among the households at 48%. In comparison, 43.1% of the respondents acquired more income to a larger extent after engaging in resilient agri-food systems. The findings also shows that about 52.5% had improved their food security to a little extent, and 50.8% were, to a larger extent, able to participate in farmers' groups, unlike before. Despite the consistent problem of markets for their produce, 56.9% of the respondents felt that there was little improvement in access to the market and pricing of farm inputs. Shuaibu and Nchake (2021) studied the use of innovative marketing of agricultural products in Sub-Saharan Africa. Using two-stage least square instrumental methods, the study considered the concept of reverse causation for products. The study found that creating a market credit pool would enhance the agricultural products market as direct intervention. This shows that there are still innovative ideas for enhancing the market for agricultural products.
In another dimension of marketing, Taufiq et al. (2022) studied the association between prices of agricultural products at the farm versus consumer retailing in Indonesia. Using rice as the study product, the regression analysis indicated an association between prices at the two levels of the product processing. The study revealed that while there was horizontal integration of the prices at each level in the long run, the process in the short run had a vertical integration. There were differentiated market prices at the farm, and the consumer sales points attributed to distribution costs. The study recommended government intervention in enabling cheaper transport of the rice to retail markets. From the rating of 1-5, a very large extent had the greatest impact and depicted the most improved household livelihood outcomes. Within this scale, it is evident that diversification and intensification of resilient agri-food systems had the greatest impact (58.5%), followed by the reduced prevalence of human, crop, and livestock diseases (55.7%). Improved coping with various shocks and trends (drought, floods, disease, death, job loss) came third at 44.7%. In contrast, increased investments in social capital networks, e.g., farmers' cooperatives and members' associations, came fourth at 37%. The study findings reflect the existing literature, which points out the outcomes of food systems. According to Delgado and Tschunkert (2022), systems, in this case, food systems, aim to create outcomes that they are designed to produce. This, therefore, means that we should analyse and alter food systems if the outcomes are negative.

Bivariate Analysis
Data analysis using inferential statistics was done to test the association between variables. Chi-square and ANOVA tests examined whether the relationship between variables was statistically significant.  Table 4 shows the calculated Chi-square statistic was 9.551, the degree of freedom 2 and the reported significance level 0.008, which is less than the set alpha of 0.05. This means a positive statistical association between the increases in household income by sex. The findings reflect the existing literature from the United States, where women were keen on resisting gender orientation to farming to attain food justice and sustainability. Using a feminist and queer approach, Leslie et al. (2019) indicate that relational agriculture was a tool for stabilizing food supplies through re-orienting gender roles in the United State farms. In a related study in Tanzania, Masamha et al. (2018) found a weak link between sexes in cassava farming. The study stated that women were the majority in the processes of harvesting and sales. On the other hand, more men were in control of finances and resource planning around cassava farming. Women were lowly involved in ownership, cross-border marketing, and control of high-value nodes. Adam et al. (2021) recognizes the importance of women's empowerment and gender equality in the agri-sector. Using fish agri-food systems, the researchers opined that gender equality would be a factor in food security and livelihoods across the globe. Guided by the UN Sustainable Development Goal of zero poverty, the study proposed a systemic gendered tool for reducing inequities and barriers in resilient agri-food systems.

The Relationship between Resilient Agri-food Systems and Household Livelihood Outcomes
The regression analysis between resilient agri-food systems and household livelihood outcomes was computed, and the results are shown in Table 5. The hypothesis was also tested using ANOVA and coefficient. The study's null hypothesis was that "there is no statistically significant relationship between resilient agri-food systems and household livelihood outcomes". The ANOVA statistics presented in Table 5 show the regression model significance with an F-statistic of 6.775 at a .010 significance level. The significance value of p=0.010 is less than 0.05. This implies that the observed aggregate variability in the resilient agri-food systems had a statistical influence on a household livelihood outcome while the random factors did not. The findings can be compared with the results from South Africa, where multiple factors were found to affect the level of resilience in agricultural production (Makate et al., 2019). Smart innovations adoption, access to credit, land, education, and information were factors that affected the level of agricultural productivity. There was also a differentiated impact based on geographical locations and gender. The study recommended the institutionalization of gendersensitive policy to address gender disparities in the practice of resilient agriculture.
The regression analysis process yielded coefficients presented in Table 6.  Table 6 shows that the standardized coefficient was positive (.164). This coefficient was statistically significant at a 5% significance level because the pvalue (0.01) is less than 0.05. This shows that the independent variable had a positive and significant relationship with the dependent variable. Therefore, the null hypothesis "there is no statistically significant relationship between resilient agri-food systems and household livelihood outcomes" was rejected. The findings imply that agri-food systems significantly influence household livelihood outcomes among farmers in Kitui County. In a modelled context, the following would be the results; Y = 48.88 + 0.164X +/-0.084 Where: Y is the Household Livelihood Outcomes; X is the level of Resilient Agri-food Systems The statistics imply that with zero level of resilient agri-food systems (X=0 units), there would still be household livelihood outcomes equivalent to 48.88 +/-.084 units. However, introducing one unit of resilient agri-food systems would increase household livelihood outcomes by .164. This change would be significant at .01.
The findings concur with Bisht (2021) who opines that there are small-scale farmers in Uttarakhand hills in India who are embracing resilience to climate change through various ecological farming methods. Interventions that have benefitted the farmers include economic incentives, eco-nutrition education, and support. To champion innovative and improved agriculture on the hilly farms, the practices have been enhanced with support for biodynamic farming, improving markets, value chain development, and school meal programs adoption. Still in India, Priyadarshini and Abhilash (2021) evaluated the impact of COVID-19 on agrifood systems. India was reported to have had the second-largest number of COVI-19 infections by the World Health Organization. The researchers found that the country had to incorporate the most innovative ways into the agri-food systems to remain afloat in household food supplies. The country used strategies including 'Planetary Healthy Diets', an adoption to overcome food insecurity that faced the country during the peak of the COVID-19 pandemic. The findings depict the importance of preparedness for pandemics affecting household food supply. The current study in Kitui County-Kenya, did not have an objective on the effect of the pandemic. The COVID-19 pandemic accelerated the practice of agri-food systems in the county, creating a surge in food supplies in the country (CABI, 2020).
In a Chinese study by Lu et al. (2022), it is reported that China has been in efforts to achieve food security within its borders and across the globe. China has been practicing food security measures through transformative food systems. Some of the effective areas of China's resilient agri-foods systems include land consolidation, creation of food reserves, value addition, and diverse nutrition foods innovations. However, numerous challenges face the practices. The reduced diversity in foods, reduced agricultural lands and unsustainable environments are some of the pointed-out challenges. Similarly, the Kitui County scenario in resilient agri-food practices is not unique nor void of challenges.
In the African context, Sitko and Jayne (2018) used the challenges facing agricultural production such as declining productivity, soil degradation, weak African food systems, and irresponsiveness to climate change to call out for new dimensions in sustainable and innovative methods. Among the proposed methods include transformative subsidies for agricultural productivity, social safety through direct investment, the creation of alternative food systems, and organic farming. These are the concerns raised in the justification for the current study for the small-scale farmers in the larger Kitui County. The use and adoption of innovative mechanisms for agri-food systems are long overdue.

CONCLUSION
The study concluded that resilient agri-food systems practices influence household livelihood outcomes in Kitui County. Most of the farmers trained were facilitated by non-profit organizations (NPOs) on resilient agri-food practices and practiced conservation agriculture as their primary source of livelihood. Despite the training, the majority felt that they needed extra training to improve their livelihood outcomes. To enhance resilient agri-food systems in Kitui County, the provision of aggregation centres was considered very urgent (18.7% of the respondents). The majority of the study participants (43.1%) indicated having acquired more income to a more significant extent after engaging in resilient agri-food systems. Moreover, most of the respondents (56.9%) felt that there was little improvement in access to the market and pricing of farm inputs. Household livelihoods had improved as a large portion of the participants (61.8%) pointed out that their level of dependency on other people had reduced, and they were selfreliant.
The study also showed that most of the respondents (42.7%) had reduced their consumption of savings and assets to a larger extent. This was supported by statistics where a positive statistical association between increased household income and sex was found (χ 2 =9.551). The study found that aggregate variability in the resilient agri-food systems had a statistical influence on a household livelihood outcome while the random factors did not.

RECOMMENDATIONS
The study recommends a clear focus on agri-food systems with diversified pathways. This will enable the adoption of the most feasible methods for supporting individual household livelihoods. The use of training, providing incentives, and introducing insurance and social safety nets like it is in European countries may be a turnaround that will increase the benefits of agri-food systems practices on household livelihood outcomes in Kitui County.