Spatial Differentiation of Small Holder Farmers’ Vulnerability to Climate Change in the Kyoga Plains of Uganda

The paper assessed the variation in the level of vulnerability to climate change among small holder farmers in the Kyoga plains of Uganda. It was hypothesized that there is no spatial variation in the level of vulnerability to climate change among the small holder farmers of different socioeconomic characteristics in the Kyoga plains. It improves the understanding of the different dimensions of vulnerability. This can help to design practical policies and intervention strategies that are specific to the communities’ spatial strata to reduce development imbalances and empower the most vulnerable small holder farmers. The conceptual framework is based on the three elements of vulnerability that is, exposure, sensitivity and adaptive capacity. The cross-sectional survey research design was used to collect both quantitative and qualitative data. Household data were acquired by using a structured questionnaire supported by focussed group discussions while meteorological data were collected using data base review. The study was done in the Kyoga plains agro ecological zone of Uganda comprising of several districts out of which Tororo and Pallisa were picked. Indicators for the components of vulnerability (Exposure, Sensitivity and Adaptive Capacity) were selected by Principle Component Analysis (PCA) and Vulnerability Indices constructed at household level then aggregated at sub county level for correlation using ANOVA. Inter sub county vulnerability index correlation revealed a spatial variation in the level of vulnerability between the different sub counties with Kasodo Sub County in Pallisa being the most vulnerable and Rubongi in Tororo being the least vulnerable. Policy measures and development efforts should therefore focus on place specific strategies of adapting to climate change rather nationwide or region wide strategies. There is also need to refocus policy to nonfarm activities which are less susceptible to climate change and enhance farmers’ income. How to cite this paper: Chombo, O., Lwasa, S. and Makooma, T.M. (2018) Spatial Differentiation of Small Holder Farmers’ Vulnerability to Climate Change in the Kyoga Plains of Uganda. American Journal of Climate Change, 7, 624-648. https://doi.org/10.4236/ajcc.2018.74039 Received: September 14, 2018 Accepted: December 11, 2018 Published: December 14, 2018 Copyright © 2018 by authors and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/


Introduction
It has been noted that in Uganda, some climate change induced outcomes have been as integral components of the overall constraints to agricultural productivity for instance up to 34% of crop damage in the country is caused by climate induced stimuli [1] such as rainfall shortage, crop diseases and insect damage [2]. Whereas this is accurate, the scale of scoping is macro hence we cannot be able to appreciate the area specific climate change impacts.
It has further been noted that Uganda is vulnerable to climate change as most of its agriculture is rain fed yet agriculture is the backbone of the economy. Any slight variation in climate may therefore be reflected in the productivity of agricultural systems across the country and pronounced variability may result in adverse physical, environmental and socio-economic impacts. Rainfall across the country is currently unreliable and highly variable in terms of its onset, cessation, amount and distribution, leading to either low crop yields or total crop failure [2] [3] [4].
Studies have been done on vulnerability and climate change among small holder farmers in Africa and Uganda in particular for instance; [4] [5]. These studies deal with vulnerability to climate change among small holder farmers but they are either exclusively biophysical [4] [6] or socioeconomic [5] [7] [8], in their approach to vulnerability assessment with attention given on crop yields, animal production or the socioeconomic characteristics of the farmers. However, it is increasingly being recognized that vulnerability cannot be fully understood by exclusively assessing either biophysical or socioeconomic conditions [9] [10], hence the need for more comprehensive and holistic approaches to the assessment of vulnerability that integrates both socioeconomic and biophysical aspects of vulnerability to climate change among small holder farmers.
The current study explores both the biophysical and socioeconomic conditions associated with vulnerability to climate change among small holder farmers in the Kyoga plains of Uganda. Whereas it is known that farmers in the Kyoga plains are generally vulnerable to climate change, there seems to be a knowledge gap regarding the spatial dimension of the vulnerability to climate change in relation to the socio-economic characteristics of small holder farmers.
The study therefore examines the spatial differentiation in the extent to which the level of vulnerability to climate change varies among small holder farmers of different socioeconomic characteristics in the Kyoga plains of Uganda.
The main objective of the paper is to assess the spatial variation in the level of

Profile of the Study Area
The study was done in the Kyoga plains agro ecological zone of Uganda which is one the ten agro ecological zones in the country according to the National Agricultural Research Organization (NARO) classification [11]. The zone comprises the districts of Kayunga, Kamuli, Iganga, Bugiri, Busia, Tororo, Manafwa, Mbale, Pallisa, Kumi, Soroti, Kaberamaido, Lira and Apac ( Figure 1). The area is characterized by the following climatic features: The average rainfall is 1215 -1238 mm, two rainfall seasons in the southern part of the zone (Tororo) (Figure 1), that is, March-May and August-November and one dry season, between December to February. Evaporation exceeds rainfall during the dry months while during the rainy season; rainfall is greater or equal to evaporation.
In the northern part of the agro-ecological zone (Pallisa) (Figure 1), there is one rainy season, that is, March to November and one dry season, December to March. Evaporation exceeds rainfall during dry months and rainfall is greater or equal to evaporation during dry months.
The temperature ranges between 24˚C -36˚C. The altitude ranges from 914 -1800 metres above sea level. The land is mainly flat and swampy and the soils range from low to moderate productivity [12].
The Kyoga plain was chosen because it is an important focal area for Uganda given its importance in the Nile basin. The area has important resources for production for example fresh water, vegetation, soil, to mention but a few and yet there are significant differences in human welfare indicators such as health, population, poverty, food security and others [13]. The Kyoga plains have a fast-growing population with a growth rate of 4% -6% with the poverty and  food security situation worse than the national average. Northern Uganda (Lira, Apac, among others), which form part of the Kyoga plains, is the poorest region in the country, with a poverty level of 75.8% of the population. One district was chosen from the northern sub zone (Pallisa) and one from the southern sub zone (Tororo), respectively. This was done purposively because these two districts are particularly vulnerable to climate change and given their proximity to other districts, it is hoped that results from these will reflect the conditions in the others.
Tororo district has a total land area of 1193.8 square kilometres and is estimated to have 103,585 households with a total population of 517,080, out of which 250,830 are males and 266,250 are female, giving a sex ratio of 94.2. In addition, Tororo has a total of 22 sub counties including the two divisions of Tororo Municipality. The district has over the years witnessed environmental degradation manifested in deforestation, poor garbage management and wetland encroachment. Most of the shallow wetlands have been drained for rice growing.
On the other hand, Pallisa district has 63.4 square kilometres, 66,668 households with a total population of 386,890 people, 186,125 of whom are male and 198,765 are female and a sex ratio of 94.6. In addition, Pallisa has a total of 19 sub counties including Pallisa Town Council. Pallisa is one of the poorest districts in eastern Uganda located in the plains of the Lake Kyoga system. The district is one of the poverty hotspots in the country and region with an index of 63% [14]. The district is also home to the extensive wetland of Mpologoma river with numerous lakes that form part of the Kyoga system. The area is in a natural sink with lakes, large ponds and permanent wetlands, the nearest and biggest being Nakuwa that is also a Ramsar Site [14]. The area is also characterized by low-lying grasslands with soil types that are of medium to low productivity [12].

Research Design and Data Collection Procedure
The cross-sectional research design was used in the study. Structured questionnaires and focus group discussions were used to elicit data from the small holder farmers on their level of exposure, sensitivity and adaptive capacity to climate change, from which vulnerability indices were computed.
A reconnaissance was initially done with a view of establishing the main climate risks that are common in the area and the livelihood issues that could constitute the indicators of adaptive capacity and sensitivity. This was to validate the indicators that had been identified from literature. The respondents were consulted on the major climate risks that occur in their community, how these affect their livelihoods and how they deal with these impacts. The outcome of the reconnaissance helped to strengthen the survey tool.
Following the reconnaissance, a survey was conducted with a total of three hundred and eighty-four respondents, chosen across the two districts of study.
Two sub counties were chosen per district and from each of these, two parishes were chosen then three villages were chosen from each parish. Sixteen respondents were then chosen from each village basing on the list of residents from the LC one chairperson. This was first stratified to male and female headed households and from each substratum, an appropriate number randomly picked to make the sixteen. These were then interviewed using a semi structured questionnaire with the help of five local interviewers. These interviewers were able to speak the main languages used in Palisa and Tororo districts.
Prior to conducting the survey, the interviewers were trained in data collection methods of surveys and use of Global Positioning System (GPS) for coordinate readings of household locations. A part from this, observation and focused group discussions were held thereafter to generate views on community-wide climate risks experienced and livelihood issues that constituted indicators of vulnerability. For all the surveyed households GPS readings for location were taken.
This paper also makes use of raw monthly minimum and maximum temperature and monthly precipitation data obtained from the Uganda National Meteorological Authority (UNMA) in Kampala Uganda for 30 years (1984-2014), the period recommended by the World Meteorological Organisation (WMO) [15]. Temperature and precipitation data were obtained for four meteorological stations distributed across the agro-ecological zone namely, Tororo, Soroti, Lira and Jinja. Data from more stations from within the region like Serere, Kamuli and Mbale were anticipated but these were reported to be having so much gaps hence the four were relied upon since it is all that was available. The temperature and precipitation at the household level was interpolated for each year from the weather stations using the latitude-longitude-altitude information of each

Conceptual Framework of Vulnerability
Vulnerability is the degree to which geophysical, biological and socio-economic systems are susceptible to and unable to cope with adverse impacts of climate change, including climate variability and extremes [16]. According to [17], three main models for conceptualizing and assessing vulnerability can be distinguished. The first is the risk-hazard framework (biophysical vulnerability assessment), which is characteristic of the technical literature on risk and disaster management. It conceptualizes vulnerability as the dose-response relationship between an exogenous hazard to a system and its adverse effects. This notion of vulnerability corresponds most closely to "sensitivity" in IPCC terminology.
The second is the social constructivist framework (social vulnerability assessment), which prevails in political economy and human geography. It regards vulnerability as a priori condition of a household or a community that is determined by socio-economic and political factors. Vulnerability according to this view is seen as the socioeconomic causes of differential sensitivity and exposure.
The third is integrated framework (both bio physical and social vulnerability assessment), which is related to the IPCC definition of vulnerability (to climate change), that is, the degree to which a system is susceptible to, or unable to cope with adverse effects of climate change, including climate variability and extremes [18]. Vulnerability is a function of the character, magnitude, and rate of climate variation to which a system is exposed, its sensitivity, and its adaptive capacity. Vulnerability, according to this definition, is an integrated measure of the expected magnitude of adverse effects to a system caused by a given level of certain external stressors.
Vulnerability, according to this school, includes an external dimension, which is represented here by the "exposure" of a system to climate variations, as well as an internal dimension, which comprises its "sensitivity" and its "adaptive capacity" to these stressors [19]. This study uses the integrated framework approach of vulnerability assessment, in line with the IPCC 2001 frame work [18].
According to [10], two approaches are generally used in literature to assess vulnerability: vulnerability variable assessment and the indicator approach. The vulnerability variable assessment approach is an econometric approach that measures the welfare loss for selected variables of concern for example house hold consumption, agricultural yield, among others, in relation to specific sets of stressors, such as climate change. Several generic vulnerability metrics have been proposed in economic and agricultural studies. While these metrics can provide an indication of the vulnerability of a given place, they are not sufficient to fully capture all the three dimensions of vulnerability [20].
On the other hand, the indicator approach uses a specific set or combination of indicators (proxy indicators) and measures vulnerability by computing indices, averages or weighted averages for those selected variables or indicators. This approach can be applied at any scale (household, county/district, or national level).

O. Chombo et al.
The major limitation of the approach is that the application of indices is limited by subjectivity in the selection of variables. However, the indicator approach is valuable for monitoring trends and exploring conceptual frameworks. The composite indices can capture the multi-dimensionality of vulnerability in a comprehensible form. Vulnerability indicators are needed for practical decision-making processes, to provide policy makers with appropriate information about where the most vulnerable individuals are located.
This study assessed the spatial variation in the level of vulnerability of small holder farmers to climate change and therefore adopted the indicator approach in line with the integrated vulnerability conceptualisation approach [17] highlighted above, which provides a multi-dimensional or integrated outlook of vulnerability and information about where the most vulnerable individuals are located.
As noted by [18] vulnerability comprises of three elements, that is, exposure, sensitivity and adaptive capacity, which are conceptualised below.
Exposure is the nature and degree to which a system is exposed to significant climatic variations [21]. It refers to the stress that impacts a system; the extent to which a system is exposed to the climate hazard [9]. The indicators of exposure here include historical changes in climate variables (temperature and rainfall) and the frequency of occurrence of extreme climatic events (drought and floods).
Sensitivity is the degree to which a system is modified or affected by an internal or external disturbance or set of disturbances [21]. Sensitivity represents the system's condition that can reduce or worsen the impact. Frequency of disease out breaks, income structure and loss of properties (viz livestock, house hold property and crop) due to climate related disasters over the last thirty years represent the sensitivity for the purpose of this study.
Adaptive capacity is the ability of a system to adjust to climate change including climate variability and extremes, to moderate the potential damage from it, to take advantage of its opportunities, or to cope with its consequences [9].

Choice of Indicators and Calculation Vulnerability Indices/Data Analysis
Following the definition of vulnerability given by the IPCC 2001 frame work [18], vulnerability in this study is taken to be a function of exposure, sensitivity, and adaptive capacity. Exposure is the nature and degree to which a system is exposed to significant climatic variations. Sensitivity is the degree to which a system is affected, either adversely or beneficially by climate-related stimuli. Adaptive capacity is the ability of a system to adjust to climate change including climate variability and extremes, to moderate the potential damage from it, to take advantage of its opportunities, or to cope with its consequences. Selection of indicators for adaptive capacity is based on the British Department for Foreign Development (DFID) sustainable livelihoods framework, whereby adaptive capacity is taken to be a function of asset possession by the households [13] [21].

Exposure
For this study, historical changes in climate variables and occurrence of extreme climatic events constitute the indicators of exposure ( Table 1) of these extreme events for the last ten years was obtained for each household from the household survey. It was hypothesized that the higher the rate of change of the climate variables and higher the frequency of extreme climate events, the higher would be the exposure of the households to climate change and extremes.

Sensitivity
Sensitivity is the degree to which a system is modified or affected by an internal or external disturbance or set of disturbances [21]. Accordingly, indicators of sensitivity in this study include; frequency of disease out breaks, fatalities, that is, death of family members due to flood or drought and loss of properties (viz livestock, house hold property and crops) due to climate related disasters over the last ten years (Table 2). It was hypothesized that higher rates of change in the past climatic hazards would increase the frequency of disease out breaks hence the sensitivity of the households to such events. The other indicator is fatality of members of the household. It was hypothesized that the higher the frequency of extreme climate events like floods and drought, the higher would be the rate of death of family members as a result of these events hence higher sensitivity of the community to climate change. In the same vain, a higher rate of change in past climatic events would lead to more damage to property namely, livestock, household property and crops hence more sensitivity.

Adaptive Capacity
Adaptive capacity is indicated by the five types of livelihood assets viz. physical, human, natural, financial, and social indicators as explained and hypothesized here (Table 3). Physical assets: type of house, ownership of devices to access information (mobile phone, radio and television) and walking distance to the nearest motor road. Possession of better-quality house would improve the capacity to with stand the risks from extreme climate events. Type of house was indicated by a value of 1 -3, 3 indicated the most durable type of house (see Table 3). Ownership of mobile phone, radio and television would increase the adaptive capacity through access to weather related information which would enable households  to plan proactive adaptation measures against climate risks. Walking distance to the nearest motor road, which in this case is also equivalent to the nearest market place, was assumed to be inversely related to adaptive capacity as households located far away from the markets would be in a disadvantageous position for lacking the opportunity of income generation from alternative sources like non-farm labor, which help in securing livelihoods during the periods of food shortage or crop failure. Human asset: highest qualification in the family and the number of household members with trainings or vocational courses attended. It was supposed that development of human capabilities through vocational trainings or formal education would enable households to increase their income by undertaking skilled non-farm activities, which are less climate-sensitive compared to farming, thereby helping the households to avert climate risks. Furthermore, it would also diversify household livelihood sources which help to buffer the risks posed by climate on farm income.
Natural assets: The ownership of the total land utilized by the households, size of land owned and ownership of oxen. In the Kyoga plains there is a practice of farmers using land beyond what one owns through renting to make up for land shortage. Households owning all the land that they use were assumed to be more flexible in production decision making and have lower production costs hence would suffer less from climate disasters. The other indicator was the size of land owned. Households that own big sizes of land were assumed to be more adaptive than those who own small sizes of land. Apart from that, the other indicator was possession of oxen. Households that possess oxen, which are the main means of plowing fields, were assumed able to have more timely preparation of gardens and production hence suffer less from climate disasters as compared to those without.
Financial assets: Gross household annual income, household savings, and owner ship of livestock (goat, poultry, cattle) were taken as the indicators of financial assets. Higher income means greater availability of resources at disposal to maximize positive livelihood outcomes. In addition to income at disposal, households that were able to make some savings out of their income would be able to make productive investments like family education or use the savings as buffer during the times of need. Households which own livestock were able to dispose them off during crisis and so were less vulnerable than those without.
Apart from owning livestock, the number of these livestock that a household O. Chombo et al.
owned, was another indicator of adaptive capacity. Households with more livestock were considered more adaptive and hence less vulnerable than those with fewer.
Social asset: Having relatives who can support one during difficult times and the number of such relatives that a household would have. Having relatives who could support the household was considered a strong safety net against climate change challenges. Therefore, the households which were having relatives who support them were considered more adaptive than those who didn't. In addition, the higher the number of such relatives a household had, the more adaptive such a household was considered to be.
Having selected the appropriate indicators of farmer vulnerability, they were normalized or standardised so as to bring them within a comparable range [21] [22]. Normalisation was done by the general formula below as used by [10]. Step wise PCA was run for the indicators of exposure and adaptive capacity and overall indices calculated using the weights (loadings) obtained from second step PCA.
The normalised variables were then multiplied with the assigned weights to construct the indices for exposure, sensitivity and adaptive capacity, each separately using the equation below: where, "I" is the value of the respective index (exposure, sensitivity and adaptive capacity), "b" is the loadings from the first components from PCA (PCA1), taken as weights from respective indicators, "a" is the indicator value, "x" is the mean indicator value and "s" is the standard deviation of the indicators. The vulnerability index for each house hold was then calculated as: Household vulnerability indices were also cross tabulated against the ten socio economic characteristics of the households namely, gender, age, farming experience, farm size, income level, access to credit, access to extension services, household size, educational level and membership to a support groups so as to determine the variation of vulnerability of smallholder farmers within these socio-economic groups.

Results and Discussion
The index scores for sensitivity, exposure, adaptive capacities and overall vulnerability for each individual household were aggregated at the village, parish, Sub County and district levels. This study uses the sub county level scores of exposure, sensitivity, adaptive capacity and overall vulnerability for analysis because the Sub County is an administrative unit of government authorized to plan and implement development plans.

Indicators of Exposure
The weights obtained from PCA analysis for the indicators of exposure (  [24] and [25] showed that incidents of meteorological, hydro meteorological and agricultural droughts lead to massive crop failures thereby increasing the exposure of the farmers. Following the mean values of the indicators of exposure shows that they varied with sub counties for instance, minimum temperature was highest in Kasodo and lowest in Apopong, both in Pallisa. Maximum temperature on the other hand was highest in Apopong and lowest in Kasodo. This implies that temperature was slightly higher in Pallisa than Tororo and this is perhaps because Pallisa is at a lower altitude than Tororo. Rainfall was slightly higher in Rubongi, Tororo though the sub county with the lowest amount of rainfall was Kisoko, also in Tororo. The highest frequency of extreme climate events in the last ten years occurred in Kasodo sub county, Pallisa district while the lowest was reported in Apopong, also in Pallisa.

Sensitivity Indicators
Indicators of sensitivity are contributing to the sensitivity index in the direction as hypothesized (Table 2)

Adaptive Capacity Indicators
Mean average values of the indicators for adaptive capacity (Table 3)  protection that people will get from the direct impact of extreme climate events especially floods. Social assets did not get any weight. Thus as noted by [26] in a study in central and western Uganda, improving adaptive capacity would require that farmers engage in off farm self-employment.

Vulnerability Indices
Examining the vulnerability index (Table 4)  An inter district variation in the level of vulnerability is shown in (Figure 2) to complement the findings above. Accordingly, Pallisa district was found to be more vulnerable than Tororo district.

Vulnerability in Relation to Farmers' Socioeconomic Characteristics
Apart from examining the level of vulnerability among farmers in the different sub counties to determine the spatial extent, vulnerability index was also cross tabulated with the socioeconomic characteristics of the farmers across the study area. To this end it was found that vulnerability levels vary with different socioeconomic characteristics of the respondents as follows.

Age
Vulnerability levels varied across the age groups with age group less than 30 years showing the highest vulnerability level (2.4) and the lowest being age

Farming Experience
The highest level of vulnerability was noted among farmers with farming experience of between 20 -30 years (2.

Gender
The males were reported to have a higher level of vulnerability (1. land, and other resources due to traditional social barriers. Nonetheless, [32] have contrary results to the effect that female headed households are more likely to adopt different methods of climate change adaptation than male headed households hence are less vulnerable than their male counterparts. Gender variation may equally have different effects on the level of vulnerability since the concepts are closely related and one aspect leads to the other. Some studies on the other hand found that household gender was not a significant factor influencing farmers' level of vulnerability to climate change through decisions to adopt conservation measures [33]. In this study, it was hypothesized that the gender of the household head would have a significant influence on the household's level of vulnerability to climate change. Our findings show that the male headed households are more vulnerable to climate change than the females. However, the relationship is not significant meaning that the gender of the household head is not significant enough to influence the level of vulnerability.

Level of Education
The highest level of vulnerability was reported amongst farmers with secondary level education (3.1) followed by those with primary level (1.1) and then those with higher education level (−0.6). Those with no formal education were reported to be least vulnerable (−1.8

Farm Size
Vulnerability levels were found to be highest among farmers with farms ranging from 4.1 -6 acres (3.0) followed by those with farms between 2 -4 acres (1.0).

O. Chombo et al.
Those with farms of over 6 acres were the least vulnerable (−5.5) while those with less than two acres (0.3), were second last. The P-value was 0.222. Available literature indicates mixed effects of farm size on vulnerability to climate change for example a study on soil conservation measures in South Africa showed that farm size was not a significant adoption factor and therefore does not influence the level of vulnerability to climate change [42]. Other studies, however, found that farmers with larger farms were found to have more land to allocate for constructing soil bunds (embankments), improved cut-off drains in Haiti and grow both perennial and annual crops which would improve their adaptative capacity, thereby reducing their vulnerability [35] [39] [43]. On the other hand, [44] found that farmers with a small area of land were more likely to invest in soil conservation than those with a large area and therefore be less vulnerable. In this study it was hypothesized that farmers with larger farms are less vulnerable to climate change than those with smaller farms. Our findings indicate that there was a positive relationship between farm size and the level of vulnerability implying that farmers with larger farms are more vulnerable to climate change but this correlation is not significant, that is, the size of the farm is not significant enough to determine the level of vulnerability of a farmer to climate change.
This finding appears contradictory to the general truth that would be expected.
However, it is true because many of the farmers with large size of land were lacking the requisite resources and other capacities that could enable them use the land effectively for example some who did not have oxen for plowing found it expensive to grow sufficient crops and so only have the little, they could do with their house hold labour.

Level of Income
Farmers with a monthly income of 10,000 -100,000 were reported to be the most vulnerable (2.1) followed by those with between 101,000 -200,000 (1.1) and then those with 201,000 -300,000. Those with less than 10,000 (−6.5) and those with over 300000 (−3.1), were the last and second last respectively, with a P-value of 0.017***. This is in line with the available literature which seems unanimous to the effect that income of the farmers, whether farm or nonfarm, represents the wealth of individual households. Empirical evidence by [45] and [46] indicate that farmers' income has a positive relationship with the uptake of farming technologies and therefore reduction in vulnerability since any adoption/adaptation process requires that the farmer has sufficient financial wellbeing. Higher income positively affects public perception of climate change [47] [48].

Group Membership
Vulnerability was reported to be highest among the farmers belonged to support groups (1.5) as compared to those who did not belong to any group (−0.  [55]. The findings above are in line with this.

Access to Credit
An examination into the vulnerability index of farmers according to their access to credit revealed that those had access had a higher vulnerability index (1.8) than those without (0.4) and the P-value was (0.335) This seems to be in line with available literature which notes the role of credit in the uptake of farming technologies. [36] [56] and [57] observe that a positive relationship exists between the level of adoption and the availability of credit since credit eases the cash constraints and allows farmers to buy inputs such as fertilizer, improved crop varieties and irrigation facilities which reduces vulnerability. Our findings however indicate that the relationship is not significant though positive.

Household Size
Vulnerability was found to be highest amongst those with families of fifteen and above members (6.0) followed by those with families of ten to fifteen members (1.9) then families of five to 9 members (0.8) and lastly families with less than five members were the least vulnerable (−0.6) and the P-value was (0.074). As for the household size, [48], argue that larger households have a larger pool of labour and as a result, they are more likely to adopt agricultural techniques than smaller households. Moreover, [36] notes that the size of the household influences individuals' adaptation to climate change in two perspectives. In the first perspective, households with large families may be forced to divert part of the labour force from farm to off-farm activities in an attempt to earn some income that can ease the consumption pressure imposed by a large family in the face of climate change. In the second perspective, households with a large family size are considered to have a larger pool of cheap labour resource, which can readily O. Chombo et al.
be employed on the farm for crop and/or livestock production, unlike families with smaller household size. The findings affirm the assumption that farmers with large households may be more vulnerable to climate change due high dependency ratio. However, the correlation is not significant.

Conclusions and Policy Recommendations
The analyses brought up three main ideas of the vulnerability of farmers of the Kyoga plains to climate change sector to climate change. First and foremost, overall vulnerability to climate change is spatially differentiated across the agro ecological zone. Vulnerability indices differ in the different sub counties that were picked for analysis, for instance, Kasodo was the most vulnerable sub county with the highest level of exposure and second lowest level of adaptive capacity while Apopong with the least level of adaptive capacity was the second least vulnerable with the least exposure implying that the three elements of vulnerability occur at different levels in different areas in the agro-ecological zone. Thus, although national and regional/zonal climate change adaptation policy is necessary, policymakers should develop area-specific policies and address climate change at the lower level (zones and sub zones) depending on their unique characteristics.
Secondly, whereas results show that exposure of a locality to long-term changes in climate variables and occurrences of natural disasters are the most important component to determine the overall vulnerability of the locality, biophysical elements determining the exposure like temperature, rainfall and natural disasters are beyond the immediate influence of the policy makers. Out of the three components of vulnerability, adaptive capacity is the component having direct policy implications. Thus, improving the adaptive capacity of these vulnerable households should be the main focus of policy formulation and implementation since improved adaptive capacity reduces their sensitivity and finally decreases their overall vulnerability. Among the various components of adaptive capacity, policy emphasis should be placed to create opportunities for non-farm livelihoods options which will reduce the dependence of the community on natural resources. Besides, relief measures to support the community during emergencies must be put in place for all the sub counties having both higher exposure as well as lesser adaptive capacity.