Vulnerability of Peasant Farmers to Climate Variability and Change in Semi-Arid Ethiopia

Vulnerability study identifies the most vulnerable systems, regions, peoples, and the contributing factors to the vulnerability. Vulnerability study in climate change context is crucial to effectively and efficiently reduce the impacts of climate variability and change. This study assessed the vulnerability of peasant farmers to climate variability and change in semi-arid Ethiopia. The semi-arid zone was categorized into southern, central, and northern semi-arid. From each semi-arid zone, sample districts and sample peasant associations were selected respectively by manual lottery and purposive sampling techniques. 396 household surveys, 12 focus group discussions, and rainfall and temperature data were used for analysis. Principal Component Analysis (PCA) was used as the main analysis technique to construct the vulnerability indices. The central semi-arid zone with the vulnerability index of -3.07 was highly vulnerable at χ²(2) =43.9986, P ≤ 0.05, while the northern semi-arid zone with the vulnerability index of 4.83 was the least vulnerable. The lack of access to a clean drinking water source, main road, and market center were among the factors that contributed to farmers’ vulnerability to climate variability and change in central semi-arid Ethiopia. The lack of access to information, small farmland holding size, and lack of multipurpose trees on the farmland are among the factors that have to be addressed in northern semi-arid even if it is relatively the least vulnerable. The level of farmers’ vulnerability to climate variability and change and the contributing factors to farmers’ vulnerability varies in semi-arid Ethiopia. Vulnerability reduction measures need to be specific to each semi-arid zone and priority needs to be given according to their degree of vulnerability.

square kilometers, Kola Tembien has a population density of 52.92 persons per square kilometer (CSA, 2007). Cereals such as, teff, wheat and sorghum are the major crops that are commonly grown in the area. Livestock rearing is also part of the agricultural practice. Degraded hills and valleys mainly characterize the geomorphology of the area.

Semi -arid areas
Other dry land areas

Sampling and data collection
Simple random sampling technique (manual lottery method) was used to select the sample districts from the three Journal of Environment and Earth Science www.iiste.org ISSN 2224-3216 (Paper) ISSN 2225-0948 (Online) Vol.10, No.5, 2020 semi-arid zones of the country (southern, central and northern semi-arid zones). Hagere Mariam district from the southern semi-arid, Ginir district from the central semi-arid and Dehana and Kolla Tembien district from the northern semi-arid zone were selected. Two sample districts (Dehana and Kolla Tembien) from the northern semi -arid were selected due to its large area coverage. Three Peasant Associations (PAs) near the district center from each sample district were selected purposely to compare the three semi-arid zones based on their minimum vulnerability. Stratified random sampling technique based on wealth ranking was used to select the respondents from each peasant association proportionally from the different wealth classes (rich, medium and poor). This is due to the fact that, the way a rich household respond to climate change is different from that of medium or poor household. The number of respondents at each peasant association was determined based on their household population proportion. Including 10% nonresponsive rate, a total of 421 semi-structured questionnaires were administered to collect data about the household and farm characteristics, income source technology source, infrastructure, cereal yield, and access to information and about other identified adaptive capacity and sensitivity indicators. However, only 396 questionnaires were used for analysis. Furthermore, one Focus Group Discussion (FGD) at each sample peasant association, making a total of 12 focus group discussions were conducted to address the issues that were not easily quantifiable in this study, such as the quality and adequacy of supply and service provision of infrastructure, institutions and other identified socio-economic indicators. Each focus group consist 6-10 members from the two gender classes. Rainfall and temperature data of the last 21 years  were gathered from the National Meteorological Agency (NMA), Combolicha, Mekele, Bale and Hawassa branch.

Data analysis
Descriptive statistics such as frequency, percentage and mean was used to summarize and describe the findings at each semi-arid zone. The vulnerability index of the three semi-arid zones was constructed by using the weight of the indicators determined by the Principal Component Analysis (PCA) technique together with the normalized value of the indicators. PCA was employed to reduce the original variables that were possibly correlated variables into orthogonal or uncorrelated lower number of principal components that most successfully capture the largest amount of information common to all the variables, and at the same time PCA was used to determine the weight or the contribution of each indicator to the vulnerability of peasant farmers to climate variability and change by using SPSS software version 16.0.
The dataset was first normalized or standardized by using centering and standard deviation normalization technique in order to adjust the values measured in different scales and units into comparable scale and unit and was checked for the suitability to run PCA by using the Kaiser-Meyer-Olkin (KMO) test for sampling adequacy and Bartlett`s test of sphericity.
The weight and normalized value of the indicators were used in the following principal component model to construct the adaptive capacity index, sensitivity index and exposure index for each semi-arid zone (southern, central and northern). = + ⋯ + , = 1 … …………………..1 Where, Y= Adaptive capacity index or sensitivity or exposure index a = The weight or coefficient of the adaptive capacity or sensitivity or exposure indicator x = The normalized value of the adaptive capacity or sensitivity or exposure indicator For instance, the adaptive capacity index of each semi-arid zone was constructed by using the weight and the normalized value of the adaptive capacity indicators at each semi-arid zone in equation1. Similar procedure was followed to construct the sensitivity and exposure index.
Vulnerability was conceptualized in this study as the function of adaptive capacity, sensitivity and exposure, and vulnerability index was calculated as: = − + ……………………2 Where V= vulnerability index, =Adaptive capacity index, = Sensitivity index, = Exposure index.

Results and discussion 3.1. Adaptive capacity indicators
Various adaptive capacity indicators were identified from the different aspects of the literature and were adapted to the Ethiopian context in order to measure the adaptive capacity of peasant farmers to climate variability and change.

Wealth
Wealth is among one of the key determinants of the nations, communities, and individuals adaptive capacity (IPCC, 2001;Smit & pilifosova, 2001;Yohe & Tol, 2002). The quality of the residential house and the size of farmland were taken as an indicator to measure the wealth status of the farmers.

Quality of the residential house
In Ethiopia, the residential house of the farming community was mainly two types. 1) Thatch roofing with wood/stone wall and mud floor. 2) Corrugated iron sheet roofing with wood/stone wall and mud floor. The two types of residential houses differ only by their roofing material. In Ethiopian context, a residential house with corrugated iron sheet roofing material belongs to the wealthier farmers and considered as a quality residential house in the community. Hence, in this study, the house roofing material was taken as the measure of the quality of the residential house, and the farmers with corrugated iron sheet roofing material residential house were considered as relatively wealthier. However, the measures of the quality of the residential house may vary from community to community and overtime.
House roofing materials As presented in table 1, 55.4% of the respondents from the southern semi-arid zone, 46.4% of the respondents from the central semi-arid zone and 27.9% of the respondents from the northern semi-arid zone were with a residential house with corrugated iron sheet roofing material. This indicates that farmers in the southern semi-arid zone are relatively the wealthiest than the farmers in the central and northern semi-arid zone in terms of quality residential house.

The size of farmland
Farmers with the farmland size equal or greater than the average land holding size of the country, one hectare per household (EEC/EEPRI, 2002) were considered as wealthier. 87.1% of the respondents from the southern semiarid zone, 60.8% of the respondents from the central semi-arid zone and 37.1% of the respondents from the northern semi-arid zone were with the farmland ≥ 1ha. This indicates that the farmers in the southern semi-arid zone are relatively the wealthiest than the farmers in the central and northern semi-arid zone in terms of the size of farmland. Farmers with the larger arable land holding size may have a better productive soil as a result of the long fallow period which can help them to better adapt to climate variability and change.

Dependency Ratio (DR)
Total dependency ratio is the measure of the ratio of dependents, aged under 14 and above 65 to the working-age (labor force), aged from 15 to 64 (World Bank, 2015).
We found the greatest (82%) dependency ratio per household in the southern semi-arid zone, followed by the central semi-arid zone (80%) while the least (63.5%) in the northern semi-arid zone. Higher dependency ratio reflects the higher burden on the economically active population/producers to support and provide the social services needed by children and older persons who are economically dependent i.e., the communities with the greater dependency ratio have the lesser adaptive capacity. Dependency ratio (DR) suggests that all children aged under 15 and persons aged above 65 as an economically dependent portion of the population. However, in many populations, people do not stop being economically active at age 65, and it is not absolutely true that all persons aged 15-64 are economically active. Thus, the dependency ratio is the estimation not an absolute measure.

Infrastructure
The availability and access to key infrastructures determines the individuals and communities adaptive capacity to climate variability and change (IPCC, 2001;Smit & Pilifosova, 2001;Kelly & Adger, 1999). In this study, access to the main road and access to the drinking water source were taken as the key indicators of infrastructure to measure the adaptive capacity.

Access to a drinking water source
In Ethiopia, more than 80% of the country`s population lives in rural areas (WPP, 2015). To make those people have an access to drinking water, both international and local NGOs and the government was putting efforts by mainly focusing on building at least one common hand pump in each peasant association at or near to the peasant association center. Access to drinking water in this context refers to access to the hand pump water which has no measured water quality parameters. Hence, the term access to drinking water was used instead of access to clean drinking water, and the percentage of farmers with access to hand pump drinking water source was taken in this work as one of the measures of the ability of the community to adapt and cope with the effects of climate variability and change.
76.7% of the respondents from the northern semi-arid zone, 64.4% of the respondents from the southern semiarid zone and 35% of respondents from the central semi-arid zone were with access to hand pump drinking water source. The percentage of farmers with access to drinking water source indicates the proportion of the population with access to an improved drinking water source in a dwelling or located within a convenient distance, in this case, 6 kilometer from the user`s dwellings. The percentage of population with access to improved drinking water have been used as an indicator by many international institutions such as, WHO, UN,UNICEF, world bank and by many individual nations. However, this indicator has been critiqued for its lack of information on the adequacy and quality of service provision. To have an understanding of the quality and adequacy of water supply, focus group discussions were held at each sample PAs. The majority of the FGD members in all sample PAs noted that hand pumps were only visible at the PA centers and the issue of maintenance and the quality of the water was no longer the concern (no one was taking into account) once the pump built. Also, they have mentioned that People who were inhabited far from the PA centers usually collect their drinking water from the rivers and creeks, and they drunk without any pre-treatment. The statement offered by one of the participants from the central semi-arid, Ginir district, Ebisa PA summed up the comments of the majority of the participants concerning the issue of drinking water.
We used to collect drinking water from the creeks and rivers which were clean and not too far from our home. Now they are no longer available, we have lost our creeks and rivers. Our wives and children travel a long distance to collect water or they spend too long time waiting for their turn to collect water from the hand pump. The hand pump is for us who lives near to the PA center, but for our peoples who live far from the PA centers, finding clean drinking water is not an easy task. When we found the water we do not care about its quality, we just drink by saying the local proverb called "no ugly mother no bad water. The finding indicates the need for all concerned stakeholders to work hard to make the farmers to have an access to improved drinking water source, and to work in improving the quality of the water for those who have already with access. A community without access to clean drinking water faces a relatively long list of social and economic challenges than a community which has access to it (UNDP, 2006). For instance, unhealthy labor force due to water-related health problems, wastage of productive time and energy by traveling a long distance to collect water, high medical expense, poor hygiene, low school attendance, especially adolescent girls etc.

Access to the main road
In Ethiopia, the majority of the population lives in rural areas and their livelihood is completely dependent on agriculture (FAO, 2015). Building networks of roadways can help to link those rural communities (the majority of the population) to the market and other service and facility centers. A road with a gravel surface type was the most common road throughout the country. Hence, in this study, the percentage of farmers with access to gravel road rather than asphalt road was taken as one of the relative measure of the ability of a community to adapt and cope with the effect of climate variability and change. 63.3% of the respondents from the southern semi-arid zone and 33.5% of the respondents from the northern semi-arid zone were with access to gravel surface road type while all of the respondents from the central semi-arid zone were with no access to the main road. The physical access to the gravel surface road could not tell us the current condition and challenges associated with it. The following FGD result summarized the condition and challenges associated with the main road. The gravel surface road was the common type of road throughout the country. However, all most all of the focus group members noted that the road networks with this surface type were not passable by all vehicle types and during the rainy season. The challenges associated with the road were summed up in the words of one participant from Dehana district as follows: A car comes once in a week in a dry season, so that we need to register for our turn prior a week. Besides, many of us are not able to afford the transportation cost; as a result, we use animals and/or human as a means to transport agricultural goods to the market center or we just sell to the middleman by the lower price.

Human capital
The human capital of the community in this study was represented by the adult literacy rate. 80.2% of the respondents from southern semi-arid, 57.8% of the respondents from the northern semi-arid and 50.5% of the respondents from the central semi-arid were illiterates. Illiterates lack the ability to get, understand and interpret up-to-date information about the market chain, market price, consumer demands, product quality and standards, efficient way of production, weather and climate information etc., which are the key in production decision. Furthermore, illiterates may not readily adopt new agricultural technologies such as, drought resistant crop varieties, soil fertility amendment technologies, etc which are highly important in climate change adaption.

Social capital
Social capital refers to the social networks with shared norms and goals that enable the society to function or act collectively and effectively (Adger, 2003). It was measured by the percentage of farmers who participates in at least one government or community organization including, cooperatives. 64.6% of the respondents from the northern semi-arid zone, 57.7% of the respondents from the central semi-arid and 49.5% of the respondents from the southern semi-arid zone were participants at least in one government or community organization. In climate change and adaptation context, the better the ability of the society to act collectively against climate change impact, the higher the adaptive capacity (Adger, 2003;Wall & Marzall, 2006).

Engagement in alternative economic activities
Alternative economic activities refer to the nonfarm income generation activities such as, employment, small-scale trade, petty-trade, gift and remittance and others, and the percentages of farmers engaged in at least one nonfarm Journal of Environment and Earth Science www.iiste.org ISSN 2224-3216 (Paper) ISSN 2225-0948 (Online) Vol.10, No.5, 2020 activity were taken as one of the indicators of adaptive capacity. 4.1% of the respondents from the central semiarid, 8.9% of the respondents from the southern semi-arid and 9.85% of the respondents from the northern semiarid were engaged at least in one nonfarm activity. Farmers' engagements in profitable nonfarm activities can provide additional income and asset that can help them to better adapt and quickly recover from climate change related impacts (O'Brien et al., 2004;Byrne, 2014). 7. Access to information It was measured by the percentage of farmers with a radio and/or television. 38.1% of the respondents from the central semi-arid zone, 40.6% of the respondents from the southern semi-arid and 31.3% of the respondents from the northern semi-arid zone were with a radio and/or television. Only ownership of radio and/or television could not tell us that farmers were really getting and understanding the production information including weather and climate information. As a result, during FGD farmers` were assisted to share the production information that they got and understood from the radio and/or television.
The majority of the participants mentioned that they did not understand and/or use production information including weather and climate information from the radio and/or television. This implies the need to work hard on the ways how to disseminate and make information from the radio and/or television understandable by the farmers. Access to critical agricultural information's such as, about improved seeds and seedlings, market chain, demand and price, crop and livestock disease treatment and control, soil fertility amendment technologies and its application, new herbicides and pesticides and its application, weather and climate forecasts and early warning system etc., can determine the adaptive capacity, adoption of new technologies and adaptation choice of farmers to climate variability and change (Eakin & Lemos 2006;O'Brien & Vogel, 2003;Nhemachena & Hassan, 2007;Baethgen et al., 2003). 8. Access to agricultural technologies 45.5% of the respondents from the southern semi-arid zone, 59.7% of the respondents from the northern semi-arid zone and 47.4% of the respondents from the central semi-arid zone were with access to at least one "hard" agricultural technology source such as, improved seed and seedlings, new herbicides and pesticides and fertilizers. The quantitative result about the population with access with at least one agricultural technology does not reflect the timely availability, affordability, its impact on the production system and the challenges related to it. These issues were addressed during the FGD. Majority of the participants shared the idea that the price of the technologies was not easily affordable, technologies mainly reach to the farmers through development agents after passing through several government administrative stages, as a result, farmers get the technologies on their hand often times very late after their scheduled sowing date. Furthermore, they have mentioned that they lack knowledge and skills on how to apply or use the technologies, poor quality of the technologies especially, the seed and the lack of locally validated technologies, i.e., they are applying technologies that were developed outside their system. Access to a wide range of production technologies and the ability of the community to develop new technologies can determine the adaptive capacity and effective adaptation.

Institutions
Institutions are required to maintain the economy, health, the social and human capital of a country, and are the key to hold the society together, and thereby to enable adaptation (O`Riordan & Jordan, 1999). In this work, the percentages of farmers with access to "soft" institutions, such as education, healthcare, veterinary, market, and microfinance were taken to measure the adaptive capacity of a community. 9.1. Access to agricultural market center 25.9% of the respondents from the northern semi-arid zone, 7.9% of the respondents from the southern semi-arid zone and only 1% of the respondents from the central semi-arid zone were with access to district level market center. Only physical access to the market center cannot allow the farmers to enjoy the market price; the quality of produce and up-to-date produce price information are also the key. During the focus group discussion many of the participants were repeatedly mentioned the idea about, lack of transport, storage and packing facility, and lack of up-to-date produce price information. Because of these factors, farmers who have even physical access to the market centers were not really enjoying the market price. The idea which was shared by the many participants in all FGD sites was expressed by one of the FGD member from Hagere mariam district as follows.
I usually, sell my agricultural goods to the middleman by the price determined by the middleman, because I am unable either to transport to the market center or to get and understand up-to-date product price information in order to reduce the unfair price burden imposed by the middleman. Access to agricultural market centers can enhance agricultural production and productivity, boost income and economic growth, reduce poverty and hunger etc. (Magesa et al., 2014;Mano et al., 2003;Lothoré & Delmas, 2009).

Access to microfinance services
Microfinance service involves the supply of the basic financial services to smallholder farmers (Visconti, 2015;Agrawala & Carraro, 2010;Hammill, Matthew, & Mccarter, 2008). 80.2% of the respondents from the southern Journal of Environment and Earth Science www.iiste.org ISSN 2224-3216 (Paper) ISSN 2225-0948 (Online) Vol.10, No.5, 2020 semi-arid, 34% of the respondents from the central semi-arid and 40.9% of the respondents from the northern semi-arid zone were the users of microfinance institutions. Providing microfinance services to small-scale farmers enable farmers to build assets, increase incomes and reduce their vulnerability to economic stress and external shocks including, climate change (Dhakal, 2010;Agrawala & Carraro, 2010;Hammill et al., 2008). 9.3. Access to veterinary service 64.9% of the respondents from the central semi-arid, 44.35% of the respondents from the northern semi-arid were with access to veterinary service, while respondents from the southern semi-arid zone were with no access to veterinary service. Climate change affects livestock production mainly by impairing the pasture and water availability and quality, and by facilitating the spread of livestock diseases (Hopkins & Del Prado, 2007;Thornton, et al., 2009). Thus, the farmers with access to veterinary center can have a healthy livestock and better production, so that they can easily cope and adapt to the effects of climate change than the farmers with no access to it.

Access to education
Access to education is among the key determinants of adaptive capacity (Yohe & Tol, 2001;IPCC, 2001), and was measured in this study by the percentage of the population with access to primary and secondary school.
64.9% of the respondents from the central semi-arid zone, 83.45% of the respondents from the northern semiarid and 72.3% of the respondents from the southern semi-arid were with access to primary school, while 23.4% of the respondents from the northern semi-arid, 9.9% of the respondents from the southern semi-arid and 2.06% of the respondents from the central semi-arid zone were with access to secondary school.
Education is the key instrument to pull families and communities out of the poverty cycle (Melin, 2001). In climate change context, by breaking illiteracy, education can make farmers able to get, understand and interpret production information such as, market and climate information, help farmers to make right production decision, enable farmers to become an experts on their farm, assist farmers to adopt and apply improved production technologies, develop farmers creativity in developing strategies to cope and adapt to the effects of climate change, make farmers to lead healthy and planned life etc. Moreover, providing access to education keep the countries away from putting surplus unproductive labor back to subsistence and climate sensitive agriculture.
During FGD, regarding education, almost all of the FGD members appreciated the efforts made so far by the government to make the primary school accessible by their children. However, they have pointed out that, they lack capital to send their students to secondary school which was somehow far from their settings. 9.5. Access to health care service In climate change and adaptive capacity context, access to improved health care services can enhance the adaptive capacity to climate change by maintaining healthier community and by reducing the burden from increasing climate-related disease such as, malaria, heat-related diseases, respiratory disease etc. (Campbell-Lendrum & Woodruff, 2006;WHO, 2003;IPCC, 2001). 64.95% of the respondents from the central semi-arid zone, 77.2% of the respondents from the southern semi-arid zone and 78% of the respondents from the northern semi-arid zone were with access to health post, while 2.06% of the respondents from the central semi-arid zone, 9.9% of the respondents from the southern semi-arid zone and 22.05% of the respondents from the northern semi-arid were with access to hospital or health center. Having access to health care services can provide: the overall physical, social and mental health status, prevention of disease and disability, detection and treatment of health conditions, quality of life, reduce life expectancy, reduce child marriage, reduce maternal deaths, combat HIV etc. (WHO, 2003). During the FGD, regarding the quality and adequacy of the health care service, the majority of the focus group discussion members have mentioned about the unavailability of experts, facilities and drugs, high health service bill and the poor transportation system to bring patients to the place where better health facilities were located.

Sensitivity indicators 1. Cereal yield
Climate variability and change including extreme events are greatly reducing the cereal yield more in small-scale agriculture and will continue to do so. Thus, in this study, it was assumed that, the farmers with the lesser cereal yield per hectare per year, the more sensitive to the effects of climate stimuli than the farmers with the higher cereal yield per year per hectare. Hence, the average cereal yield (kg) per household per hectare per year was taken in this work as one of the measures of the degree of sensitivity of a community to climate variability and change. High cereal yield help farmers to store surplus grain as a reserve for emergency time (FAO, 2011). The result of this study shows that the farmers in the central semi-arid zone produce relatively the highest cereal yield (202.9kg) per household per hectare per year, while the farmers in the southern semi-arid produce the least (131kg).

Livestock number
Livestock's are sensitive or has the potential to be affected by the climate change through its effect on livestock production mainly by impairing pasture and water availability and quality, and by facilitating the spread of livestock diseases (Hopkins & Del Prado, 2007;Thornton, et al., 2009). As a result, in this study, it was assumed that, the farmers with the greater number of livestock, the more sensitive than the farmers with the fewer numbers Journal of Environment and Earth Science www.iiste.org ISSN 2224-3216 (Paper) ISSN 2225-0948 (Online) Vol.10, No.5, 2020 of livestock. On the other hand, it is obvious that food crops that are grown on already poor soil under limited/no irrigation, like in Ethiopia, are more sensitive to climate variability and change than the livestock. For instance, during the drought period, animals can get their feed from drought resistant shrubs, trees and from already stored fodder for dry season, and can be driven to big natural water source for watering. Furthermore, animals have the natural ability to withstand high heat stress and flood effect than the common food crops. So that, livestock can serves as an important buffer for farmers in times of crop failure (Fafchamps et al., 1996). We found the respondents from the southern semi-arid zone owned the greatest average number of livestock per household (13.3) followed by the central semi-arid (9), while the respondents from the northern semi-arid zone owned the least number of livestock per household (6.8).

Irrigation and water harvesting
The water and its whole cycle are highly sensitive or have the potential to be affected by the global warming. Nowadays, due to global warming, the timing of the rain has been changing, and agricultural drought is prevailing; which is pushing rainfed agriculture dependent farmer into destitute. Thus, it was assumed in this study that, the farmers who were the beneficiaries from the community irrigation scheme or with their own water harvesting technologies, the less sensitive than the farmers who were not the beneficiaries from the community irrigation scheme or without his/her own water harvesting technologies. 0.12% and 0.13% of the respondents from the southern and northern semi-arid respectively were the beneficiaries of irrigation and water harvesting scheme, while none of the respondents were the beneficiaries from the central semi-arid zone. Supplementing or replacing rainfed agriculture with irrigation has become one of the widely assumed effective measures of climate change adaption (IPCC, 2007). Improved irrigation reduces crop failure as a result of drought and enables farmers to grow food crops in a dry season.

The soil
Some of the physical, biological and chemical properties of the soil are very sensitive to climate change or have the potential to be affected by the climate change; which in turn affects the productivity of the soil. The degree of soil sensitivity can be measured and compared by considering its several physical, biological and chemical properties. However, it has been measured and compared by the amount of fertilizer applied per hectare per growing season. Unfortunately, there was no convincing information regarding fertilizer application in the study area. Thus, the presence of multipurpose trees on the farmland was taken to measure the sensitivity of the soil, by assuming that the farmlands without trees are the more sensitive. 94% of the respondents from the southern semiarid zone, 81% of the respondents from the central semi-arid zone and 16.7% of the respondents from the northern semi-arid zone had multi-purpose trees on their farmland. Trees on the farmer's farmland supply a wide range of benefits to the farm households, including food, medicine, livestock feed, timber, shade, nitrogen-fixing trees, as a substitute or complements for chemical fertilizer, erosion barrier, micro-climate amelioration, etc.

Exposure
Exposure was measured by the frequency of extreme events (drought and flood), long-term mean of surface air temperature and rainfall and by the variability of surface air temperature and rainfall measured by the coefficient of variation (CV).  Vol.10, No.5, 2020 plot techniques were used to determine the number of principal components that should be retained to explain the maximum variance. Accordingly, the first three components, which have explained 63.667% of the total variance in the dataset were retained for further analysis. Usually, the principal components are explained by the variables that have high component loadings after rotation. Verimax rotation was used to determine exactly which variable loads the most to which component.
Loadings are the correlation between variables and the principal component, and they are equivalent to the standardized regression coefficient in multiple linear regression (when variables are standardized, i.e., converted to Z-score) and are called β-weights (Beaumont, 2012). Beta weights represent the contribution of the variables to the multiple linear regression equation or to the component score while holding other variables constant (Johnson, 2012). Hence, the loadings are equivalent to the weights of the variables (OECD, 2008). Hence, in this study, loadings after rotation and weights have the same meaning and were used as a standardized coefficient in principal component analysis model to estimate the composite index. However, only the weights of the variables that loaded to the first three components (the retained components) were used to construct the indices. = + ⋯ + , = 1 … ………3 Where, a = the weight/coefficient of the indicator x = the normalized value of the indicator Y= vulnerability index Vulnerability index (V) = Adaptive capacity index (ACI) -(sensitivity index + exposure index). The vulnerability index of southern semi-arid was calculated as follows. V = 5(0.934 * 0.523) + (0.929 * −0.203) + (0.903 * −1.094) + (0.680 * −0.764) + (0.949 * −0.140) + (0.949 * −0.288) + (0.946 * −0.172)@ ˗ 5(−0.947 * 1.153) + (−0.947) * (1.126)@ = 0.38 The same procedure was followed to calculate the vulnerability indices for the central and northern semi-arid. 2.14 31.371 The null hypothesis was rejected at χ²(2) =43.9986, P ≤ 0.05, and concluded that the difference in the level of vulnerability of peasant farmers in three semi-arid zones of Ethiopia was not by chance.
The central semi-arid zone with the vulnerability index of -3.07 was highly vulnerable at χ²(2) =43.9986, P ≤ 0.05, while the northern semi-arid zone with the vulnerability index of 4.83 was the least vulnerable. Factors such as lack of access to clean drinking water source, main road, market center, education (both to primary and