Climate and Development

ISSN: 1756-5529 (Print) 1756-5537 (Online) Journal homepage: https://www.tandfonline.com/loi/tcld20

Multidimensional livelihood vulnerability analysis

in Dinki watershed, central highlands of Ethiopia

Mengistu Asmamaw, Seid Tiku Mereta, Embialle Mengistie Beyene & Argaw
Ambelu

To cite this article: Mengistu Asmamaw, Seid Tiku Mereta, Embialle Mengistie Beyene & Argaw
Ambelu (2019): Multidimensional livelihood vulnerability analysis in Dinki watershed, central
highlands of Ethiopia, Climate and Development, DOI: 10.1080/17565529.2019.1698405

To link to this article: https://doi.org/10.1080/17565529.2019.1698405

Published online: 10 Dec 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tcld20
CLIMATE AND DEVELOPMENT
https://doi.org/10.1080/17565529.2019.1698405

RESEARCH ARTICLE

Multidimensional livelihood vulnerability analysis in Dinki watershed, central

highlands of Ethiopia
a,b
Mengistu Asmamaw , Seid Tiku Meretaa, Embialle Mengistie Beyene a
and Argaw Ambelua
a
Department of Environmental Health Science and Technology, Jimma University, Ethiopia; bDepartment of Biology, Debre Berhan University, Ethiopia

ABSTRACT ARTICLE HISTORY

This study assesses the multidimensional livelihood vulnerability of smallholder farmers to climate Received 20 February 2019
change-induced shocks in Dinki watershed, northcentral highlands of Ethiopia. The data were collected Accepted 22 November 2019
through a cross-sectional survey conducted on 288 households, six focus group discussions, and 15
KEYWORDS
key informant interviews, complemented with 30 years of rainfall and temperature data obtained from Agro-ecology; climate
the National Metrological Agency. The Livelihood Vulnerability Index (LVI) framed within the change; climate change-
Intergovernmental Panel on Climate Change and vulnerability (LVI-IPCC) approach was used to induced shock; Dinki;
measure households’ livelihood vulnerability on the agro-ecology unit of analysis. In each agro-ecology, Vulnerability index
the LVI and LVI-IPCC were calculated as well as one-way analysis of variance was used to test
differences between agro-ecological zones. The findings indicate that the vulnerability dimensions and
major components varied between agro-ecological zones (p < 0.001). The result also reveals that
increased sensitivity to water and health facilities, recurrent exposure to climate change-induced
shocks, poor technology utilization, and limited livelihood diversification practices are principal factors
contributing to mounted sensitivity, exposure, and overall livelihood vulnerability of the lowland agro-
ecology. The finding suggests that designing resilience-building adaptation strategies based on the
agro-ecology unit of analysis is sound to reduce the vulnerability of smallholder farmers to climate
change-induced shocks.

1. Introduction
Foltz, 2014). Unless management interventions towards enhan-
Climate refers to the average measure of weather conditions cing resilience are effective, the current trends of climate change
documented often within 30 years period. The average extremes impacts will shrink the Ethiopian gross domestic product by 2–
of weather conditions are known as climate shocks (Bohra-Mis- 6% by 2015, and up to 10% by 2045, with a reduction by 0.5–
hra, Oppenheimer, & Hsiang, 2014). Climate change-induced 2.5% per annum (Conway, Lisa, & Schipper, 2011). Their
shocks are climate change-associated disasters (for example, reliance on climate-sensitive livelihood (agriculture); tra-
floods, disease outbreaks, market fluctuation, drought, land ditional farming practices; lack of essential household capitals
resources degradation, etc.), which are the principal livelihood to invest in adaptation strategies and poor complementary ser-
threats challenging humankind globally. Developing countries vices (like extension, credit, market, etc.), make Ethiopians vul-
are disproportionately suffering from multifaceted shock nerable to climate change-induced shocks (Deressa, Hassan, &
impacts despite their minimal contribution to their causes Ringler, 2011; IPCC, 2007).
(Green, Jackson, & Morrison, 2009). The upcoming effects of Vulnerability in this study was defined following the Inter-
global warming adversely affect Africans with hydrologic governmental Panel on Climate Change as ‘the degree to
extremes (flood and drought), which further amplify stress on which a system is susceptible to or unable to cope with adverse
human health, food security, and water resources. Conse- effects of climate change including variability and extremes’
quently, about 350–600 million people in the region are (IPCC, 2001, 2007). The IPCC approach perceives vulnerability
expected to be impacted by 2050s (IPCC, 2001, 2007). to be a function of both ‘internal’ factors (sensitivity and adap-
Historical records in Ethiopia show a continuous rising of tive capacity) and ‘external’ factors (exposure to shocks) (Ger-
mean atmospheric temperature in a range of 1.7–2.1°C by litz et al., 2016). On the other hand, as humans are known to be
2050 and 2.7–3.4°C by 2080 (NMA, 2007). Most models are the principal actors to adapt to and mitigate with shock
consistent with increasing patterns of temperature. However, impacts, an integrated assessment of the social vulnerability
depending on the type of models used, both increasing and (adaptive capacity) with the biophysical dimensions (exposure
decreasing trends are reported in precipitation projections. and sensitivity) is found to be plausible to draw comprehensive
Hydrologic extremes, mainly flood and drought, are occurring information and widely applicable in recent studies (Piya, Joshi,
every three to five years in Ethiopia (World Bank, 2006), and & Maharjan, 2012).
their frequency and magnitude are projected to increase in Vulnerability is a crucial concept in ecology, sustainability,
the future, adding stress on livelihoods (Simane, Zaitchik, & disaster management, as well as in climate change researches

CONTACT Mengistu Asmamaw [email protected] Department of Environmental Health Science and Technology, Jimma University, P. O. Box:
378, Ethiopia; Department of Biology, Debre Berhan University, P. O. Box: 445, Ethiopia
© 2019 Informa UK Limited, trading as Taylor & Francis Group
2 M. ASMAMAW ET AL.

Figure 1. Location of Dinki watershed in the central highlands of Ethiopia (Asmamaw et al., 2019).

to identify key drivers and impacts. If it is to address adaptation net effect remaining after adaptation matters. Accordingly, sys-
limits and development needs for the vulnerable poor (Tanner tems’ adaptive capacity affects their vulnerability by modulat-
et al., 2015), as well as to understand the community as active ing exposure and sensitivity (Gallopin, 2006).
agent to make decisions, comprehensive vulnerability analysis In Ethiopian highlands, spatial variability in biophysical and
considering exposure, sensitivity and adaptation is critical to socioeconomic settings is evident within small kilometre dis-
measure the state of vulnerability (Tanner et al., 2015). In tances, contributing to differences in both states of vulnerability
this study, the Livelihood Vulnerability Index (LVI) developed and adaptation mechanisms. Within such dramatic spatial het-
by Hahn, Riederer, and Foster (2009) and the Intergovernmen- erogeneity, national scale vulnerability measures are inadequate
tal Panel on Climate Change and vulnerability frame (LVI- to capture local conditions and highly problematic for decision
IPCC) that integrates the three vulnerability dimensions: and policy-makers (Simane et al., 2014). Accordingly, agroecol-
exposure, sensitivity and adaptive capacity were developed to ogy-based analysis is identified to be instrumental in capturing
measure vulnerability (Dechasa, Simane, Alamirew, & Azadi, local level variability. It has been tested in other parts of Ethio-
2016; Etwire, Al-Hassan, Kuwornu, & Osei-Owusu, 2013; pia and suggested to be extrapolated (Dechasa et al., 2016;
Panthi et al., 2015). Simane et al., 2014). Thus, this study used the agro-ecological
In the climate change literature, exposure denotes the extent zone as a unit of analysis to explore the livelihood vulnerability
to which a system is ‘exposed’ to climate change and variability of households to climate change-induced shocks. The findings
(IPCC, 2001). Sensitivity is defined as the ‘degree to which a of the study would help to prioritize intervention measures for
system is affected, either adversely or beneficially, by climate livelihood resilience by identifying areas of susceptibility and
change and variability’ (IPCC, 2007). Both exposure and sensi- adaptation limits in Dinki watershed, northcentral highlands
tivity jointly reflect the impact that a change in climate may of Ethiopia. In effect, three key research questions were formu-
worsen systems’ vulnerability (Gallopin, 2006). Whereas, adap- lated as follows: (i) is there a statistically significant difference
tive capacity is the ability of a system to moderate potential in vulnerability dimensions and major components between
impacts, take advantage of opportunities, and cope with the agro-ecological zones? (ii) Is there variation in vulnerability
adverse effects of climate change and variability (IPCC, level (score) between agro-ecological zones? (iii) what is the
2007). A system exposed or sensitive to climate change and implication of livelihood vulnerability index in Dinki watershed
variability does not necessarily mean vulnerable; instead, the socio-ecological systems?
 CLIMATE AND DEVELOPMENT 3

Table 1. Vulnerability dimensions, major components, sub-components, and hypothesized relationships.

Dimension Major component Sub-components(indicators) Hypothesis, vulnerable if:
Exposure Natural disaster and Change in temperature and precipitation since 1987; the increase in temperature, precipitation variability, the
climate variability (NDCV) occurrence of climate change-induced shocks (drought, occurrence of shocks
flood, erosion, etc.) during the last 12 months
Sensitivity Income and food access Annual income per capita, HFIAS, and dietary diversity a HH has below national average income per capita; higher
(IFA) values in the range of 0–27
Health Illness score, access to improved toilet Higher values in the range of 0–24; no improved toilet or
shared or open pit
Water security Access to improved water, water sufficiency, water conflict HH has no access to improved drinking water or cannot be
collected within 30 min’ walk from home (round trip),
water shortage during the last 12 months, conflict due to
water
Stability Steep slope topography, infertile soil, SWC practices, Most land is steep slope and infertile, minimal SWC and
perception of CC impacts perception to CC impacts
Adaptive Sociodemographic profile The proportion of female-headed households, dependency Female-headed, higher dependency ratio, illiterate HH
capacity (SDP) ratio, education status, family size head, and large family size
Assets Farm size, livestock holding, access to communication small land and livestock holding and no device
devices
Livelihood strategy Diversity of income sources, number of coping strategies, depend on non-diversified income sources, minimal CC
access to CC information, technology utilization information, low technology utilization
Social capital Membership to community-based organizations (CBOs), minimal CBOs membership and SSS
social support score (SSS)
Access to basic services Access to market, all-weather road, primary school, has no access or above 5 km from home
(ABS) electricity, saving and credit facilities

2. Materials and methods information obtained from meteorological stations, and the
Ankober district agricultural office, respectively. The study
2.1 The study area
populations were the agricultural experts in the district, and
Dinki watershed is located in the Ankober district in the central household heads/wives or any household member above 18
highlands of Ethiopia. Ankober is bounded between 9° 23′ years living in Dinki watershed at least for a year.
19.99′′ to 9° 31′ 3.90′′ N latitude and 039° 44′ 43.47′′ to 039°
44′ 46.47′′ E longitude (Figure 1.) (Asmamaw, Mereta, &
2.3. Study design and data collection techniques
Ambelu, 2019). Most of the District’s topography are hills
and mountainous (75%), where rugged terrains and plains A cross-section study design was used to collect data viz: high-
account for 17% and 8%, respectively. More than half of the land, midland, and lowland agro-climatic zones (AEZs). A par-
District (53%) has sub-tropical (woinadega) climatic condition ticipatory rural appraisal sampling strategy involving interview
followed by tropical (kola); where temperate (dega) and cool focus group discussions, and a household survey were used to
(wurich) constitute 10.5 and 1.5 percent, respectively (Lulekal, collect the data (Boafo et al., 2015). The agro-ecological zones
Zemede, Ensermu, & Van Damme, 2013). were classified based on the Ethiopian agro-ecology zonation
The 30 years (1987–2016) of meteorological data showed a system, and information from the district agricultural experts.
mean annual rainfall of 1,187.72 mm (Table 1), where its Accordingly, villages (the smallest administrative unit in Ethio-
mean minimum, and mean maximum values were recorded pia), and gots (smaller administrative units than village)
in 2012 (692 mm), and in 2004 (1568 mm), respectively included in the highland agro-ecology were the Ememihret,
(Figure 2), indicating inter-annual variability. In addition to Lay-Gorobela, and upper parts of the Mehal-Woniz. Whereas
inter-annual variability, seasonal variability was also evident the lower parts of Mehal-Wonz (Lik-Marefiya), Gedamoch,
where peak precipitation was recorded in different months and some portion of the Hagere-Selam were included in the
(July, August, September, December, and March) across the midland agro-ecology. Likewise, the lower parts of the Zego,
study period. The Rainfall pattern is bimodal, where some and Hegere-Selam, as well as parts of the Aliyu-Amba (Addis
short and long-term rainy periods are recorded in March and Alem got), were included in the lowland agro-ecological zone.
in late June to September, respectively. The Dinki watershed Besides, the meteorological data recorded from Debre Ber-
is warmer in the winter (December to February), and rainier han, Ankober, and Aliyu-Amba stations were received from
during summer (June to August), presenting a warm and tem- the National Meteorological Agency (NMA, 2016), Ethiopia.
perate climate. In effect, it is classified as Cwb (subtropical The qualitative data were collected during November 2017
highland or monsoon-influenced temperate climate) by Köp- using key informant interviews and focus group discussions,
pen and Geiger classification system (https://en.climate-data. whereas the questionnaire survey was collected during Febru-
org/africa/ethiopia/amhara/debre-birhan-3657). ary 2018.

2.3.1. Qualitative data collection

2.2. Data sources, and study population
Focus group discussion (FGD): A total of four FGDs (one FGD
In this study, the data sources are from primary and secondary in each agro-ecology, and another female-segregated FGD in
sources. The primary sources include field observation, individ- the midland). It was designed to conduct six gender-segregated
ual interviews, group interviews, and household surveys. The FGDs; however, it was unsuccessful in the highland and low-
secondary sources are the climatic, and agro-ecology related land, where groups of females in these agro-ecological zones
4 M. ASMAMAW ET AL.

Figure 2. Mean annual rainfall (mm), mean maximum, and mean minimum monthly temperature (°C) of the Dinki watershed during the last 30 (from 1987 to 2016) years.
RF = rainfall, Max Temp = maximum temperature, and Min Temp = minimum temperature data.

were converted into individual interviews due to their small were proportionally distributed into each agro-ecology, and
size. The FGDs constituted in a range of 8–12 participants, their selection involves receiving of N lists of households
and issues on livelihood vulnerability of households to climate from respective village administrators, calculating of the sample
change-induced shocks were discussed. Some of the questions size n, providing of numeric number for each N household, and
asked were: Did you notice any change in weather conditions finally the household from the list of population was selected
of this locality during the last two to three decades? Which randomly using a lottery method (Taherdoost, 2016). The
environmental and/or socioeconomic shock was most recur- sample size determination followed the formula of small
rent and severe in this locality? What coping strategies does sample size population correction to improve the power of
the community use to manage these disturbances? What factors the statistical test following the prescription by Kothari
influence households’ vulnerability to shock impacts? (2004), and Daniel and Cross (2013) were used to determine
Individual interview: The same guiding questions were used sample size as follows:
to conduct 15 face-to-face interviews on the community mem-
N∗p∗q∗Z 2
bers selected using a snowball technique. The FGD and inter- n=
view guids were used, where information redundancy was an e2 (N − 1) + Z 2 ∗p∗q
assurance for information saturation. The total time ranged where n = sample size, Z = 95% confidence interval under the
25–30 min was used to complete an individual interview. normal curve that is 1.96, p = 0.5 (proportion of the population
Where as 60 to a maximum of 120 min, was required to com- to be included in the sample that is 50%), q = None occurrence
plete a focus group discussion (Gibson, 2012; Nyumba, Wilson, of event = 1–0.5; that is 0.5, N = size of population, e = Margin
Derrick, & Mukherjee, 2018). of error or degree of accuracy (acceptable error term) (0.05)
The household data were collected by enumerators selected
from agricultural extension workers in respective agro-ecology.
2.3.2. Quantitative data collection
A total of six enumerators (two in each agro-ecology) were
Based on the feedback and information from qualitative data, a
involved in the household survey data collection. Before the
standardized questionnaire was developed. In addition to the
actual survey, enumerators and supervisors were trained for
questions used in the interviews, a sample of questions asked
six days on the general data collection techniques, and context
in the questionnaire survey was: Was there any environmental
of the questions. Moreover, pilot tests were made on 5% of the
and/or socioeconomic shocks during the last 12 months? Do
total sample, which was not included in the final survey around
you think climate change-induced shocks affect your livelihood
the Gorobela village communities to familiarize the enumer-
strategies? What coping strategies do you use to prevent the
ators with the questionnaire.
negative impacts of shocks? What adjustment strategy
(example: livelihood diversification, farming practice, social
networking, etc.) do you apply to sustain system functioning
2.4. Data processing and statistical analysis
even during crises?
A simple random sampling technique was employed to Data processing is simply the conversion of raw data to mean-
select respondents for the household survey. The respondents ingful information through a process. Data is manipulated to
 CLIMATE AND DEVELOPMENT 5

produce results that lead to a resolution of a problem or stability major components makes this study different from
improvement of an existing situation. The collected data were Hahn et al. (2009). Moreover, the perception of climate change
entered into an excel spreadsheet, and to the computer soft- impact was added to enable the community to prepared to, and
ware, cleaned for errors, coded, verified, and checked for con- strengthen future stability (Tambo, 2016). It is argued that a
sistency. Hence, the climatic data mainly from the Debre household with a low perception of climate change impacts is
Berhan station had minimal missing values, and sufficient hypothesized to be more sensitive.
sample size following deletion, a listwise deletion technique Adaptive capacity is considered as a function of livelihood
was used to handle missing values. Finally, useful and informa- capital (natural, human, physical, financial and social capi-
tive outputs were presented in texts, tables, graphs, charts. tals), and indicators, such as sociodemographic profile, assets,
The Analysis of Variance (ANOVA) tests were carried out livelihood strategy, access to basic services, and social capital
using the IBM Statistical Product and Service Solutions package were used with certain modification from Hahn et al. (2009),
(IBM SPSS ver. 23, 2015). A Mann-Kendall trend test was com- Simane et al. (2014), Tambo (2016), and Gerlitz et al. (2016).
puted to test for the presence of a monotonic trend in meteor- In this study, the social safety net major component
ological data. Furthermore, a linear regression graph was drawn (measured based on the number of households who access
to illustrate the state of the maximum and minimum annual to loans, and political voice which in turn linked to
temperature during the last 30 years. The summary statistics inadequate social inclusion) was replaced by social capital
and Mann-Kendall trend test of the climatic data were per- where membership to community-based organizations
formed using a Past ver.3 software. (measured by social participation score), and social support
score (measured based on the sum of the number of non-
relative friends, and number of people other than the neigh-
2.5. Livelihood vulnerability index (LVI) calculation
bourhood a household can list to assist him during accidents)
Livelihood strategy is a practical tool developed to monitor were used as possible indicators to aggregate the major
the vulnerability of a geographic region in terms of its component.
exposure, sensitivity to shocks effects, and adaptation strat- Membership in community-based organizations and other
egies (Hahn et al., 2009). The index development followed traditional associations like idir, debo, etc. is a means to
techniques used by Hahn et al. (2009), with the replacement exchange information, build trust and stability. Besides, work-
of some indicators to be applicable in the Ethiopian rural ing in groups with friends, relatives, and neighbours for agricul-
context. Tools in this study were structured into three dimen- tural works, and natural resource management activities is a
sions, ten major components, and 37 sub-components. In common trend in rural Ethiopia, including in the study area.
this study, the multidimensional livelihood vulnerability inte- This supporting system is functional during emergencies,
grated exposure, sensitivity, and adaptive capacity (IPCC, including in climatic extremes. As a result, in the Ethiopian
2001, 2007). Exposure is the nature and amount to which a context, membership in community-based organizations and
system is exposed to climate change and variability. Occur- social support scores are more plausible and preferable to
rences of historical changes and frequency of extreme cli- ‘access to loans’ and ‘political voice’, which were used by Gerlitz
matic shocks are considered as possible indicators of et al. (2016). Table 1 summarizes details of the different vulner-
exposure in this study. ability dimensions.
Sensitivity is the extent to which a system is impacted by cli- The LVI uses a balanced weighted technique (Sullivan,
mate-related stimuli (Gallopin, 2006). In this study, income, Meigh, & Fediw, 2002) where each sub-component (indicator)
and food access, health, and sanitation, water security, and contributes equally to the index. As each major component is
stability were considered as possible indicators to measure sen- composed of a different number of indicators measured on
sitivity (Gerlitz et al., 2016; Tambo, 2016). Income and food different scales, two methods of standardization were used for
access were measured by per capita income, household food normalization. Indicators that are expected to have a direct
insecurity, and access score (HFIAS), and dietary diversity. A relationship with vulnerability, such as occurrences of socioe-
household with below the national average income per capita conomic and environmental shocks, illness score, Household
(3,781 Birr/adult/year, (CSA, 2010)), higher HFIAS in the Food Insecurity and Access Score (HFIAS), etc. were standar-
range of 0–27 (Coates, Swindale, & Bilinsky, 2007), and lower dized using equation (1):
dietary diversity scores in the range of 0–32 (Swindale &
Bilinsky, 2006) are hypothesized to be more vulnerable. Cli- Sr − Smin
Ia = (1)
mate change exacerbates the sensitivity of households by ampli- Smax − Smin
fying food insecurity, illness, and water scarcity (Gregory,
whereas indicators expected to have inversely related to vulner-
Ingram, & Brklacich, 2005).
ability, such as income and food access, diversity of income
In addition, Besides, environmental fragility may favour soil
sources, access to safe drinking water, etc., were standardized
erosion and hydrologic disturbance, where communities
using:
experiencing extreme weather events exhibit severe impacts.
Therefore, households with limited income, living in steep Smax − Sr
topography, as well as whose most farmlands are infertile are Ia = (2)
Smax − Smin
expected to be more sensitive to climatic shocks (Gerlitz
et al., 2016). Furthermore, the perception of climate change where Ia is the standardized value for the indicator a, Sr is the
impacts is added as an indicator of stability. The inclusion of observed (average) value of the indicator for agro-ecology r,
6 M. ASMAMAW ET AL.

min and max are the minimum and maximum values of the 3. Results and discussion
indicator across all the agro-ecology, respectively. Once each
3.1. Exposure
indicator has been standardized, the average value of each
major component was computed using equation 3: The exposure analysis showed that the lowland (0.485) was the

most exposed agro-ecology followed by the highland (0.474)
Iai
Mr = (3) and the midland (0.214) (Table 2). The mean values (mean ±
N standard deviation SD) of exposure to climate change-induced
where Mr is one of the ten major components for agro-ecology shocks varied from 0.22 ± 0.03 (midland), to 0.47 ± 0.00 (low-
r, Iai is the indicator indexed by i, that make up each major land) and 0.49 ± 0.00 (lowland) as well as the mean values of
component, N is the number of indicators in each major com- exposure to climate change and variability were statistically
ponent. After values for each of the ten major components for different between agro-ecological zones (F = 8760.15, p <
each agro-ecology were calculated, they were averaged using 0.001) (Table 2).
equation (4) to obtain AEZ-level LVI: The mean (mean ± standard deviation (SD)) annual rainfall

10 value of the Dinki watershed during the last 30 years was
p=1 WMi Mri 1,187.72 ± 211.84. Although the precipitation was not statisti-
LVIr =
10 (4)
p=1 WMri cally significant, the distribution was not uniform between
the study periods. Similarly, the mean (mean ± standard devi-
Equation 4 can be further expanded as:

Wndcv NDCVr + Wifai IFAr + Wh Hr + Ww Wr + Wsb SBr + Wsdp SDPr + Wlvs LVSr + Was ASr + Wsc SCr + Wabs ABSr
LVIr =
WNDCV + WIFA + WH + WW + WSB + WSDP + WLVS + WA + WSC + WABS

where LVIr is the Livelihood Vulnerability Index for agroecolo- ation (SD)) value of the average monthly temperature was
gical zones, Mri = the number of indicators of the major com- 13.25 ± 0.87°C, where the mean minimum, and mean maxi-
ponent, WMi = weight of major component i, NDCV = mum values ranged from 4.80°C to 11.22°C (6.46 ± 1.11°C),
natural disaster and climate variability, IFA = income and and 18.84°C to 23.58°C (20.04 ± 0.79°C), respectively (Table 3).
food access, H = health, W = water, Sb = stability, SDP = socio- The Mann-Kendall trend test showed a monotonic trend in
demographic profile, LVS = livelihood strategy, A = assets, SC maximum, and average temperature (p < 0.05), both of which
= social capital, ABS = access to basic services. presented an increasing trend. However, no monotonic trend
The weights of each major component, Wmi, are dependent was observed in the mean annual rainfall, and mean minimum
upon the number of indicators that constitute the major com- temperature values (Table 3). In this study, the linear regression
ponent and are included to confirm that all the indicators con- analysis showed a rising of the average minimum, and maxi-
tribute equally to the overall LVI (Sullivan et al., 2002). mum temperature by 0.038, and 0.052°C per annum during
Following equations (1) to (4), the livelihood vulnerability the last three decades, respectively (Figure 2). The finding of
index was computed based on IPCC vulnerability (LVI- the meteorological data also agrees with people’s perception
IPCC) framework viz: Exposure (E), Sensitivity (S) and Adap- of climate change, and variability, where rainfall variability, ris-
tation (A) using equation (5): ing of temperature in the lowland, and recurrent frost in the
highland, recurrent dry periods, and associated crop yield

n reduction, erratic rainfall often leading with accelerated soil
i=1 WMi Mri
CFr =
n (5) erosion mainly in the highland were noticed, and reported by
i=1 WMri the key informant, and discussants.
In agreement with the current finding, a study by the
where CFr is the LVI-IPCC defined by contributing factor for National Meteorological Agency (NMA) of Ethiopia showed
agro-ecology of r, Mri is the number of indicators in the an increase of mean annual temperature by 0.37°C every ten
major components for the agro-ecology of r, indexed by i, Wi
is the weights of each major component. Following the pre-
scriptions by Hahn et al. (2009), Etwire et al. (2013), Panthi Table 2. The one-way ANOVA result of vulnerability dimensions in each agro-
et al. (2015), and Dechasa et al. (2016), the LVI-IPCC was ecological zone (mean ± standard deviation in parenthesis).
developed using equation (6): Statistics
Vulnerability class Agro-ecology Mean ± SD df F-value
Exposure Highland (HL) 0.47 ± 0.00 2 8760.15a
LVI − IPCCr = (Er − Ar )∗Sr (6) Midland (ML) 0.22 ± 0.03
Lowland (LL) 0.49 ± 0.00
Sensitivity Highland 0.59 ± 0.00 2 8289.36a
where LVI-IPCCr is the LVI for agro-ecology r presented based midland 0.53 ± 0.01
lowland 0.60 ± 0.00
on the IPCC vulnerability framework; Er, Ar and Sr and the cal- Adaptive capacity highland 0.54 ± 0.00 2 108484.82a
culated exposure, adaptation and sensitivity score for agro- midland 0.50 ± 0.001
ecology r, respectively. The LVI-IPCC value ranges from −1 lowland 0.49 ± 0.00
(least vulnerable) to 1 (most vulnerable). Significance at p < 1%.
a
 CLIMATE AND DEVELOPMENT 7

Table 3. Summary statistics, and Mann-Kendall test results of rainfall, and temperature data of Dinki watershed, Ethiopia.
Variables Mean Minimum Maximum Standard deviation Mann-Kendall statistics (S) p-value
Annual rainfall 1187.72 692.07 1568.3 211.84 −59 0.39
Maximum temperature 20.04 18.84 23.58 0.79 218 0.000
Minimum temperature 6.46 4.80 11.2 1.11 59 0.29
Average temperature 13.25 12.24 17.4 0.87 135 0.02

years, and a rising by 0.1°C was reported to the mean maximum the lowland agro-ecology showed more sensitivity than the
temperature. Moreover, the country was reported to exhibit highland and midland agro-ecological zones.
great rainfall variability during the last 50 years (NMA, The undulating topography makes the highland with mini-
2007). Studies also estimate future climate variabilities, where mal irrigation potential leading to limited livelihood diversifica-
the temperature is projected to continue rising in the range tion options. Furthermore, the expansion of agriculture and
of 0.9–1.1°C by 2030, in the range of 1.7–2.1°C by 2050, and settlement into mountainous areas amplified its vulnerability
in the range of 2.7–3.4°C by 2080 over Ethiopia. An increase to soil erosion and associated income loses. Among others,
in precipitation trends is also projected in the country (NMA, these factors might have contributed to higher average
2007). Although the precipitation did not show a statistical HFIAS (10.67) and minimal annual per capita income (16.51
difference, both the meteorological records and reports from USD/month) scores in the highland compared to the midland
the respondents agree on rainfall fluctuation resulting in shift- (18.93 USD/month) and lowland (20.54 USD/month). Simi-
ing of the agricultural calendar. In effect, while most models are larly, studies argue that households with limited access to
consistent with an increasing trend of temperature, both food, health, and water facilities as well as living in steep topo-
increasing and decreasing trends are reported in precipitation graphy are more sensitive to climatic shocks (Gerlitz et al.,
patterns depending on the type of model used (NMA, 2007). 2016).
During the last 55 years, some years were known with dry con- In this study, the illness index was computed based on the
ditions resulting in drought and famine as well as there were number of days a household member unable to work for a
wet conditions as well (Deressa, Ringler, & Hassan, 2011). period of six months. The value ranges from 0(healthy) to 4
On the other hand, the finding in this study is consistent (major illness, sick for more than 25 days), and the values of
with previous results that state the Earth’s surface warms, cli- all the household members were aggregated to generate the
mate change and variability increases resulting in episodes of index at the household level. The higher average score was
both excessive and deficient precipitation, elevated tempera- recorded in lowlands (8.83) followed by midland (6.58) and
ture, increased rate of flooding, which could add further stress highlands (6.38). It might be the effects of its exposure to elev-
on health facilities, access to food and water resources (IPCC, ated temperature and drought, sensitivity to health and water
2007). In Ethiopia, environmental problems, such as land facilities were mounted in the lowland agro-ecology. In line
degradation, soil erosion, and climate-related hazards including with this finding, studies state that climate change and variabil-
recurrent drought, floods, heavy rains, frost, heat waves or ity exacerbate shock impacts by amplifying illness, water, and
increased temperature, among others are limiting its potential food scarcity, especially in vulnerable regions (Gregory et al.,
to generate ecological services, creating extra stress on access 2005).
to and use of livelihood resources (NMA, 2007). Furthermore, In terms of water resources, 42.1, 42.86 and 45.26% of
scholars argue that recurrent environmental (example: households in the highland, midland, and lowland, respectively
drought. flooding, heating, erosion, etc.) and socioeconomic access improved water within 30 min of a round trip from
(example: unemployment, market failure, illness, loss of social home; where a considerable proportion of households collect
networking, etc.) shocks could mount households’ exposure water at a distant impacting school attendance and time bur-
to extra climatic disasters (Gerlitz et al., 2016). den, especially for women. On the other hand, less than a
third (32.99%) of the study communities, specifically only
18.94% of the lowland residents have sufficient water for
3.2. Sensitivity year-round. This has caused water to be a source of conflict
in more than half (55.90%) of the households, especially during
The sensitivity analysis showed more sensitivity in the highland an extended period of drought. In agreement with this finding,
(0.588) than the lowland (0.484) and the midland (0.322). The studies state that climate change is the critical factor that chal-
highest sensitivity in the highland was highly attributed to its lenges water security in recent periods, causing water to be a
highest sensitivity in ecological stability (0.60) than the other source of conflict (Barnett & Adger, 2007).
ecological zones (Table 4). The mean sensitivity values varied
from 0.53 ± 0.01 (ML), to 0.59 ± 0.00 (HL), to 0.60 ± 0.00
(LL) and statistically differed between agro-ecological zones 2.3. Adaptive capacity
(F = 8289.36, p < 0.001) (Table 2). Likewise, the mean values
2.3.1 . Sociodemographic profile (SDP) and social
in IFA, health, water, and stability were statistically different
capital (SC)
between agro-ecological zones (p < 0.001) (Table 4). Discus-
sants disclose that its highest sensitivity instability (steep- In sociodemographic profile analysis, it is hypothesized that
slope) and in income and food access (mainly due to small male-headed households with minimal dependency ratio (if
farm size and soil erosion) makes the highland more vulner- one productive age cares for ≤1 dependent member) and an
able. However, in terms of access to health and water facilities, average family size that does not exceed the national average
8 M. ASMAMAW ET AL.

Table 4. Indexed major components, sub-components and overall LVI of Dinki watershed socio-ecological system.
Major component Sub-component Highland Midland Lowland
Value for Value for major Value for Value for major Value for Value for major
indicator component indicator component indicator component
Natural disaster and %HH reporting >1 environmental 0.853 0.474 0.367 0.214 0.832 0.485
climate variability shock during the last 12 months
%HH reporting >1 socioeconomic 0.800 0.306 0.8737
shock during the last 12 months
%HH with injury/death due to 0.547 0.224 0.547
climatic shocks during the last 12
months
Rate of change in RF since 2007 0.1664 0.166 0.166
Rate of change in temperature 0.004 0.004 0.004
since 2007
Income and food access Annual per capita income 0.726 0.588 0.675 0.322 0.642 0.484
HFIAS 0.395 0.289 0.325
Dietary diversity 0.642 0.561 0.484
Health Illness score 0.199 0.505 0.208 0.492 0.310 0.524
Improved toilet 0.811 0.776 0.737
Water Improved water 0.579 0.456 0.571 0.535 0.547 0.639
Water sufficiency 0.474 0.674 0.811
Water conflict 0.316 0.36 0.56
Stability %unsuitable land slopes 0.682 0.600 0.531 0.575 0.44 0.550
(topography)
%infertile soil 0.530 0.482 0.518
%land under SWC 0.368 0.828 0.927
perception of CC impacts 0.821 0.459 0.316
Sociodemographic profile %female headed households 0.263 0.402 0.337 0.386 0.253 0.396
Age of household head 0.638 0.415 0.648
Dependency ratio 0.57 0.53 0.525
Education status 0.263 0.337 0.253
Family size 0.275 0.31 0.300
Assets Farm size 0.597 0.639 0.527 0.593 0.459 0.531
Livestock ownership 0.753 0.722 0.713
access to a communication device 0.569 0.531 0.421
Livelihood strategy Diversity of income sources 0.685 0.619 0.56 0.478 0.675 0.594
Coping strategies 0.540 0.505 0.4475
Access to CC information 0.821 0.592 0.516
Technology utilization 0.432 0.255 0.737
Social capital CBO membership 0.782 0.566 0.760 0.506 0.745 0.499
Social support score 0.350 0.253 0.252
Access to basic services Market 0.600 0.526 0.510 0.565 0.221 0.475
health services 0.337 0.439 0.642
primary school 0.137 0.133 0.179
all-weather road 0.789 0.949 0.716
saving and credit 0.463 0.408 0.347
Electricity 0.832 0.949 0.747
Where HH = household, RF-rainfall, HFIAS = Household food Insecurity, and Access Score, CC = climate change, CBOs = Community-based organizations.

are less vulnerable to climate change and variability. In this of the sociodemographic profile and social capital were statisti-
study, half of the households in the highland (50.53%), more cally different between agro-ecological zones (p < 0.001)
than a third of households in midland (45.92%) and lowland (Table 5).
(44.21%) have a family size beyond the national average. Like- In terms of education, more than half (53%) of the respon-
wise, the proportion of households having the above optimal dents were unable to read and write, making them more vulner-
range of dependency was higher in the highland (35.79%), able to climate change-induced shocks. Studies show that the
making it more vulnerable to climate change-induced shocks. education status of the household head is a basis for strategic
In agreement with this finding, scholars highlight that large planning, household-level decisions (in agriculture, education,
family size of working age could be insurance to satisfy human health care, etc.) and wellbeing of the family. Because schooling
capital that would contribute to enhancing livelihood resilience improves the potential of the household to better understand
through access to the labour force, remittance returns and risk extension services and to apply alternative options in times of
management (Deressa, Ringler, et al., 2011; Weldegebriel & shocks (Etwire et al., 2013). On the other hand, illiterate house-
Amphune, 2017). However, large family size is more likely to holds are minimal in their ability to access climate change
have large dependent peoples making the household more sus- information, making them extremely susceptible to climatic
ceptible to climatic shocks (Nkondze, Masuku, & Manyatsi, shocks. These households are incapable to attend training, lim-
2013). ited in diversification experiences and solely rely on climate-
On the other hand, the mean adaptive capacity values ran- sensitive sectors like livestock raring and rainfed agriculture
ged from 0.49 ± 0.00 (LL) to 0.50 ± 0.00 (ML) to 0.54 ± 0.00 (Dumenu & Obeng, 2016).
(HL) and statistically varied between agro-ecological zones (F Likewise, the social capital was higher in the highland
= 108484.82, p < 0.001) (Table 2). Similarly, the mean values (0.566) followed by the midland (0.506) and the lowland
 CLIMATE AND DEVELOPMENT 9

Table 5. ANOVA results of the LVI major components between agroecological zones.
Descriptive statistics
Major component Agro-ecology Mean SD SE df F-value
Natural disaster and climate variability (NDCV) Highland 0.474 0.000 0.000 2 8760.15a
Midland 0.217 0.027 0.0027
Lowland 0.485 0.000 0.000
Income and food access (IFA) Highland 0.588 0.000 0.000 2 18810.95a
Midland 0.324 0.016 0.002
Lowland 0.484 0.000 0.000
Health Highland 0.505 0.000 0.000 2 6879.99a
Midland 0.492 0.003 0.000
Lowland 0.524 0.000 0.000
Water Highland 0.456 0.000 0.000 2 21286.99a
Midland 0.536 0.011 0.002
Lowland 0.639 0.000 0.000
Stability Highland 0.600 0.000 0.000 2 27355.09a
Midland 0.575 0.003 0.000
Lowland 0.550 0.000 0.000
Sociodemographic profile (SDP) Highland 0.402 0.000 0.000 2 17946.41a
Midland 0.386 0.01 0.00
Lowland 0.396 0.00 0.00
Asset Highland 0.639 0.000 0.000 2 20881.85a
Midland 0.592 0.006 0.001
Lowland 0.531 0.000 0.000
Social capital (SC) Highland 0.566 0.000 0.000 2 760612.11a
Midland 0.506 0.001 0.000
Lowland 0.499 0.000 0.000
Livelihood strategy (LVS) Highland 0.619 0.000 0.000 2 11530.29a
Midland 0.479 0.012 0.001
Lowland 0.594 0.000 0.000
Access to basic services (ABS) Highland 0.526 0.000 0.000 2 6841.45a
Midland 0.564 0.009 0.000
Lowland 0.475 0.000 0.002
a
significant at p < 1%.

(0.499) (Table 4). The social capital was assessed using mem- for farming communities designed to manage risks as well as
bership to community-based organizations (CBOs) and social to alleviate poverty (Gautam & Andersen, 2016). Although
support scores (SSS). The index showed that the highland on-farm (livestock-farming integrated agriculture) livelihood
(0.566) was the most vulnerable agro-ecology followed by dominates the majority of income sources, households in
the midland (0.506) and lowland (0.499) (Table 4). Study Dinki watershed are engaged in wide range of livelihood strat-
participants disclosed that social capital is a basis for agricul- egies, such as agroforestry practices through home garden, tree
tural works and contributes to exchange information, build garden, perennial crops; collecting natural resources; small-
social networking and manage crises. For instance, one of a scale trading and others like seasonal migration of household
62-year key informant in Mehal-Weniz village stated that, members into other areas for wage labour. Following Barrett,
although their social values are declining in recent periods, Reardon, and Webb (2001), these income sources in which
social networking is a basis in both crises as well as festival the households are engaged in for survival and for the better-
events as a human cannot survive without human beings. ment of standard of living were clustered into on-farm, non-
Despite its valuable contribution to strengthening social farm, and off-farm.
bonds and trust, social capital in the highland was highly On-farm livelihood strategies are activities such as farming,
vulnerable compared to the midland and lowland agro-ecologi- livestock production, and irrigation practices. Except for irriga-
cal zones. Likewise, studies argue that local-level associations tion, it is often practiced by almost all households. The non-
are established by friendship, financial exchange, kinship farm activities include agroforestry practices like growing of
or related cultural and spiritual values (Armah, Yawson, perennial crops, home garden, tree garden, non-farm employ-
Yengoh, Odoi, & Afrifa, 2010). This networking is effective ment, rental income, remittance, small-scale trade, ecosystem-
in supporting livelihoods as well as in managing natural generated goods, etc. Nearly half (48.96%) of the households
resources and crises, including climate change-induced shocks rely on these strategies with variable degrees of involvement.
(Etwire et al., 2013). Whereas the off-farm activities include income gained from
labour wage working from other farms, both skilled and non-
3.3.2. Livelihood strategy and asset skilled employment, etc. Less than a quarter (23%) of respon-
In this study, the highland was more vulnerable (0.619) in live- dents experience this livelihood strategy.
lihood strategy compared to the lowland (0.594) and midland On the other hand, although more than half (57%) of house-
(0.478) agro-ecological zones (Table 4). The mean values of holds use more than one livelihood strategy at least some time
the livelihood strategy ranged from 0.48 ± 0.02 (ML) to 0.59 of a year, 90%, 14% and 8% of income sources are from on-
± 0.00 (LL), to 0.62 ± 0.00 (HL), and statistically varied between farm, non-farm, and off-farm, respectively. Besides, the liveli-
agro-ecological zones (F = 11530.29, p < 0.001) (Table 5). Live- hood diversification index showed that the majority (95%) of
lihood strategy is a strategy for survival or asset accumulation households experience single income dominated income
10 M. ASMAMAW ET AL.

Figure 3. Coping strategies practiced in the Dinki watershed socio-ecological system.

sources making them more vulnerable to shock impacts. In line (ANOVA) indicated statistically significant differences in
with this finding, studies state that diversity of income sources fertilizer utilization (F = 4.04, p = 0.018), irrigation potential
is insurance against the effects of shocks. Households who (F = 3.36, p = 0.036), in using improved seeds (F = 9.42,
depend on a single source of income are more vulnerable to p = 0.000), and pesticide application (F = 3.36, p = 0.036),
adverse impacts of hazards and less likely to satisfy their across agro-ecological zones. The Tukey post-hoc analysis
needs during crises. Particularly, climate-sensitive sectors like also revealed that the mean differences of all components of
rain-fed agriculture are extremely susceptible to climatic technology utilization in the midland were statistically signifi-
extremes (Eriksen, Klein, Ulsrud, Næss, & O’Brien, 2007). cant with the highland agro-ecological zone. The topographic,
In terms of coping strategies, the highland and midland soil and climatic conditions in the midland might have sup-
agro-ecologies are better off in natural resources management ported households to experience intensive crop production by
(soil and water conservation and enclosure) and agronomic using agricultural inputs. This finding agrees with a study in
(irrigation, applying fertilizer and improved seed verities) Dedesa watershed, Ethiopia, where farmers’ technology utiliz-
measures, respectively. Whereas it might be the contribution ation was maximum in a midland agro-ecological zone
of their improved access to communication device (57.89%) (Dechasa et al., 2016).
coupled with their exposure to recurrent shocks, households The asset index showed that the highland (0.639) agro-
in lowland agro-ecological zone have improved access to cli- ecology was more vulnerable than the midland and lowland
mate change information (48.42%) and are better off in econ- (Table 4). It might be the contributions of large average farm
omic (trade, growing of market-oriented crops like mung size (1.95 ha) and livestock (2.39 TLU) holdings; the lowland
bean, credit, and migration for remittance) and management agro-ecology is relatively less vulnerable in terms of asset own-
(notification and preparedness) measures (Figure 3). This state- ership. On the other hand, although the average landholding in
ment agrees with a study that states improving access to each agro-ecology is above the national average (1.22 ha),
weather event prediction and coping strategies through tele- (CSA, 2012), 37.89, 29.59 and 28.42% of households in the
communication devices significantly empowers households to highland, midland, and lowland, respectively own less than
prepared to, mitigate with and adapt to climatic extremes one hectare, making them vulnerable to climate change-
(Dumenu & Obeng, 2016). induced shocks. In Ethiopian farming community, land and
The proportion of households who use technology livestock are two of the most basic assets where their livelihoods
(improved seeds, fertilizer, pesticides, and irrigation) were depend. Accordingly, Households with large farm sizes are
74.49, 56.84, and 27.37% in the midland, highland, and low- more likely to be engaged in more diversified livelihood options
land, respectively. The outputs of one-way Analysis of Variance and more likely to empower their adaptive capacities. This
finding was similar to Tesso, Emana, and Ketema (2012) and
Dechasa et al. (2016).
Table 6. Access to basic services within 5 km distance from home and indicators,
as reported by households in Dinki watershed socio-ecological system.
Indicators n Percent (%) 2.3.2. Access to basic services (ABS)
Access to an all-weather road 52 18.06 Although the midland (0.565) agro-ecology is more vulnerable
Access to market 160 55.56
Access to primary school 245 85.07
than the lowland and highland agro-ecological zones (Table 4),
Access to health facilities 152 52.78 the overall accessibility of basic infrastructures in Dinki water-
Access to saving and credit institution 171 59.38 shed socio-ecological system is highly minimal (Table 6), indi-
Access to electricity 45 15.63
cating its vulnerability to climate change-induced shocks.
 CLIMATE AND DEVELOPMENT 11

Figure 4. Vulnerability Spider Diagram of the Major Components for Dinki watershed socio-ecological system.

Table 7. Calculated indices for vulnerability dimensions and the Livelihood terms of climate variability followed by the highland and mid-
Vulnerability Index under the LVI-IPCC frame for the Dinki watershed socio- land agro-ecological zone (Table 7).
ecological system.
Assessment of local level livelihood vulnerability is essential
Vulnerability dimensions
to identify areas of vulnerabilities and basis for management
Agro-ecology Exposure Sensitivity Adaptive capacity LVI_IPCC
options. In the Dinki watershed socio-ecological system, a con-
Highland 0.474 0.595 0.535 −0.036
Midland 0.214 0.532 0.501 −0.153 siderable proportion of households (32.29%) own less than one
Lowland 0.485 0.602 0.489 −0.002 hectare of farmland, and nearly half (47.22%) of the households
Average 0.391 0.576 0.508 −0.068 have less than two livestock units. Increased vulnerability of the
highland in terms of income and food access, livelihood strat-
Underdeveloped infrastructure makes the households isolated egy and asset, might be associated with underdeveloped infra-
from regional centres. This physical isolation from market structure, rugged topography, and small asset holding, mainly
and information centres amplifies their vulnerability to a mul- due to small farm size and livestock ownership. On the other
titude of climate change-induced shocks. In agreement with the side, increased the vulnerability of the lowland to water and
current finding, studies state that the physical isolation of com- health facilities might be attributed to its extreme exposure to
munities from economic centres hinders their access to inputs, climatic extremes such as drought.
market exchange, information, and opportunities to diversify A study in other part of Ethiopia showed that minimal
livelihood options (Oparinde & Hodge, 2011). Hence, house- access to infrastures causes to lack of access to information
holds having market access are more likely to exchange infor- and opportunities to diversify livelihoos strategies, contributing
mation, diversify livelihood strategies, and mitigate with as to mount communities’ vulnerability to climate change-
well as to adapt to climate change-induced shocks (Tesso induced shocks (Tesso et al., 2012). As a result, households
et al., 2012). Other studies added that insufficient physical with small farm sizes, and livestock holding as well as minimal
structures significantly limit access to basic services like health livelihood strategies have limited adaptive capacity and are vul-
and credit facilities, contributing socioeconomic marginaliza- nerable to climate change-induced shocks (Nyamwanza, 2012).
tion (Gerlitz et al., 2014). Hence, adaptive capacity is the dimension of vulnerability
which has direct policy implication. For example, intensive
soil and water conservation measures and technology utiliz-
3.4. Livelihood vulnerability index ation could contribute to decreasing soil erosion and the house-
The aggregated results of all the ten major components were hold’s sensitivity to income and food accesses. Likewise,
presented in a spider diagram (Figure 4) and in LVI-IPCC diversifying income sources through non-farm activities
dimensions: exposure, sensitivity, and adaptive capacity reduces the sole dependency of households on farming liveli-
(Table 7). The vulnerability spider diagram ranged from 0 hood, contributing to their adaptive capacity to climate
(least vulnerable) to 0.7(most vulnerable). Accordingly, the change-induced shocks (Piya et al., 2012).
highland agro-ecology was more vulnerable in terms of stab-
ility, income and food access, social capital, livelihood strat-
3. Conclusion
egies, and assets. Whereas the lowland was highly exposed to
natural disasters and climate vulnerability as well as more sen- The result of LVI and LVI-IPCC showed that the lowland agro-
sitive in health and water resources (Figure 4). ecology is found to be higher in exposure and sensitivity but
Likewise, the LVI-IPCC analysis showed that the highland have the relatively higher adaptive capacity. Despite its higher
was more vulnerable in adaptive capacity, and the lowland adaptive capacity, the lowland agro-ecology is higher in overall
was the most exposed and sensitive agro-ecology. The LVI- livelihood vulnerability to climate change-induced shocks. On
IPCC estimates for the highland, midland, and lowland were the other hand,
−0.036, −0.153, and −0.002, respectively. The highest score Such agro-ecology based variation in the level of exposure
among the negative values is indicating an agro-ecology with and livelihood vulnerability suggests agro-ecology specific
more vulnerability to climate variability and vice versa. As a intervention actions, which could contribute to ameliorate cli-
result, the lowland is the most vulnerable agro-ecology in mate change-induced shocks and build resilience. In general,
12 M. ASMAMAW ET AL.

however, improving access to basic services, livelihood diver- project management on WaSH, Aquatic Ecology, Resilience Measurement,
sification, diversifying coping strategies, and technology utiliz- Bio-efficacy of chemicals, Biodiversity Data Mobilization, and related pro-
fessional activities. writes on both Environmental Health and Aquatic Ecol-
ation could contribute to lessen livelihood vulnerability and
ogy. Among the recent publications: A macroinvertebrate multi-metric
enhance resilience to climate change-induced shocks in Dinki index for Ethiopian highland streams (Hydrobiologia, 2019) and Responses
watershed socio-ecological system. This result also suggests of stream macroinvertebrates to hydrological disturbances in Dinki water-
that more emphasis needs to be given to investing in soil and shed, central highlands of Ethiopia (Limnologica, 2019).
water conservation measures to reduce farmers’ exposure to cli-
mate change-induced shocks like flooding as well as to reduce ORCID
their sensitivity to food insecurity, mainly in the highland agro-
ecology. Furthermore, it is imperative to give attention to build- Mengistu Asmamaw http://orcid.org/0000-0001-9656-748X
Embialle Mengistie http://orcid.org/0000-0002-0790-9087
ing social capital to reduce their exposure and sensitivity as well
as to enhance their adaptive capacity to climate change-induced
shocks. References
Armah, F. A., Yawson, D. O., Yengoh, G. T., Odoi, J. O., & Afrifa, E. K. A.
(2010). Impact of floods on livelihoods and vulnerability of natural
Acknowledgments resource dependent communities in Northern Ghana. Water, 2(2),
The authors would like to express their heartfelt gratitude to all who con- 120–139. doi:10.3390/w2020120
tributed to this study. This research was partly supported by Debre Ber- Asmamaw, M., Mereta, S. T., & Ambelu, A. (2019). Exploring households ‘
han and Jimma Universities as well as by the International Foundation resilience to climate change-induced shocks using climate resilience
for Science (IFS), Stockholm, Sweden, through a grant N0.-1-1-A-6055- index in Dinki watershed, central highlands of Ethiopia. PLoS ONE,
1 (by Asmamaw M.). We are very grateful for these institutions. All 14(7), 1–21. doi:10.1371/journal.pone.0219393
authors and sources of materials used in this paper are highly Barnett, J., & Adger, W. N. (2007). Climate change, human security and
acknowledged. violent conflict. Political Geography, 26(6), 639–655. doi:10.1016/j.
polgeo.2007.03.003
Barrett, C. B., Reardon, T., & Webb, P. (2001). Nonfarm income diversi
Disclosure statement cation and household livelihood strategies in rural Africa: Concepts,
dynamics, and policy implications. Food Policy, 26, 315–331. doi:10.
No potential conflict of interest was reported by the authors. 1016/S0306-9192(01)00014-8
Boafo, Y. A., Saito, O., Kato, S., Kamiyama, C., Takeuchi, K., & Nakahara,
M. (2015). The role of traditional ecological knowledge in ecosystem
Funding services management : the case of four rural communities in … The
This research was partly supported by Debre Berhan University and Jimma role of traditional ecological knowledge in ecosystem services manage-
University as well as by the International Foundation for Science (IFS), ment : the case of four rural communities in Northern Ghana. Int J
Stockholm, Sweden, through a [grant number 1-1-A-6055-1] (by Asma- Biodivers Sci Ecosyst Serv Manag, 12(December), 24–38. doi:10.1080/
maw M.). 21513732.2015.1124454
Bohra-Mishra, P., Oppenheimer, M., & Hsiang, S. M. (2014). Nonlinear
permanent migration response to climatic variations but minimal
Notes on contributors response to disasters. Proceedings of the National Academy of
Sciences, 111(27), 9780–9785. doi:10.1073/pnas.1317166111
Mengistu Asmamaw, a Ph.D. candidate in Applied Ecology, Department Coates, J., Swindale, A., & Bilinsky, P. (2007). Household food insecurity
of Environmental Health Science and Technology, Jimma University. He access scale (HFIAS) for measurement of household food access:
is a lecturer at Department of Biology, Debre Berhan University, Ethiopia. Indicator guide version 3. Washington, DC: Food and Nutrition
He has a research interest in Ecology, climate change, and natural resource Technical Assistance Project Academy for Educational
conservation fields. Among his most recent published works include Resi- Development.
lience of Households to climate change-induced shocks (PLoS One, 2019) Conway, D., Lisa, E., & Schipper, F. (2011). Adaptation to climate change
and Effects of Hydrological disturbances on macroinvertebrates (Limnolo- in Africa: Challenges and opportunities identified from Ethiopia. Global
gica, 2019). Environmental Change, 21(1), 227–237. doi:10.1016/j.gloenvcha.2010.
Seid Tiku Mereta is an associate Professor in Environmental Health and 07.013
Ecology. He is working on Ecological Monitoring, Assessment and man- CSA (Central Statistical Agency). (2010). Household income, consumption
agement of Freshwater resources. He has published more than sixty orig- and expenditure survey 2004 (HIES 2004-2005). Version 1.1, DDI-ETH-
inal articles in the area of biodiversity conservation, aquatic ecology, air CSA-HICE-2004-v1.1, Addis Ababa, Ethiopia.
pollution and control, occupational health and safety, vector ecology, CSA (Central Statistical Agency). (2012). The Federal Democratic Republic
water, and wastewater treatment. of Ethiopia central statistical agency the 2011 urban employment unem-
ployment survey. Retrieved from http://adapt.it/adapt-indice-a-z/wp-
Embialle Mengistie Beyene, an assistant professor in the Department of content/uploads/2015/01/survey-unemployment.pdf
Environmental Health Science and Technology, Jimma University. He is Daniel, W. W., & Cross, C. L. (2013). Biostatistics: A foundation for analy-
currently the head of the Department of Environmental Health Science sis in the health sciences (10th eds.). John Wiley & Sons, Inc, United
and Technology. Embialle research focuses on are pollution Prevention States of America, pp958. Journal of Chemical Information and
and Control Engineering, pollution remediation. Among his recent publi- Modeling, 53. doi:10.1017/CBO9781107415324.004
cations is Pollution remediation: Ultrasound assisted chrome tanning: Dechasa, C., Simane, B., Alamirew, B., & Azadi, H. (2016). Agro-ecological
Towards a clean leather production technology (Ultrasonic sonochemistry, based small- holder farmer’s livelihoods vulnerability to climate varia-
2016) and Impact of Tannery Effluent on the Self-purification Capacity bility and change in Didesa sub Basin of Blue Nile River, Ethiopia.
and Biodiversity Level of a River (Bulletin of environmental contamination Acad. J. Agric. Res, 4(May), 230–240. doi:10.15413/ajar.2016.0150
and toxicology, 2016). Deressa, T. T., Hassan, R. M., & Ringler, C. (2011). Perception of and adap-
Argaw Ambelu, a Professor of Environmental Health and Ecology, in the tation to climate change by farmers in the Nile basin of Ethiopia.
Department of Environmental Health Science and Technology, Jimma Uni- Journal of Agricultural Science, 149(1), 23–31. doi:10.1017/S002
versity. His main duties and responsibilities are teaching, research and 1859610000687
 CLIMATE AND DEVELOPMENT 13

Deressa, T. T., Ringler, C., & Hassan, R. M. (2011). Factors affecting the Lulekal, E., Zemede, A., Ensermu, K., & Van Damme, P. (2013).
choices of coping strategies for climate extremes: The case of farmers Ethnomedicinal study of plants used for human ailments in Ankober
in the Nile Basin of Ethiopia. IFPRI Discussion Paper, 1032 district, North Shewa Zone, Amhara Region, Ethiopia. Journal of
(November), 25. Retrieved from http://www.ifpri.org/publication/ Ethnobiology and Ethnomedicine, 9(1), 63.
factors-affecting-choices-coping-strategies-climate-extremes%5Cnhttp: Nkondze, M. S., Masuku, M. B., & Manyatsi, A. (2013). Factors Affecting
//www.ifpri.org/sites/default/files/publications/ifpridp01032.pdf households vulnerability to climate change in Swaziland: A case of
Dumenu, W. K., & Obeng, E. A. (2016). Climate change and rural commu- mpolonjeni Area Development Programme (ADP). Journal of
nities in Ghana: Social vulnerability, impacts, adaptations and policy Agricultural Science, 5(10), 108–122. doi:10.5539/jas.v5n10p108
implications. Environmental Science and Policy, 55, 208–217. doi:10. NMA (Ethiopian National Meteorological Agency). (2007). Climate
1016/j.envsci.2015.10.010 change national adaptation programme of action (Napa) of Ethiopia.
Eriksen, S. E. H., Klein, R. J. T., Ulsrud, K., Næss, L. O., & O’Brien, K. (June), 1–73pp.
(2007). Climate Change Adaptation and Poverty Reduction: Key NMA (Ethiopian National Meteorological Agency). (2016). NMA
Interactions and Critical Measures. In Report prepared for the (National Metrological Agency), 2016. Ethiopian national metrological
Norwegian Agency for Development Cooperation (Norad). Retrieved agency. Addis Ababa, Ethiopia. http://www.ethiomet.gov.et/. Retrieved
from http://www.norad.no/en/tools-and-publications/publications/pu from http://www.ethiomet.gov.et/
blication/_attachment/106215?_download=true&_ts=11eb62af607 Nyamwanza, A. M. (2012). Resiliency and livelihoods inquiry in dynamic
Etwire, P. M., Al-Hassan, R. M., Kuwornu, J. K. M., & Osei-Owusu, Y. vulnerability contexts: Insights from northern Zimbabwe (Unpublished
(2013). Application of livelihood vulnerability index in Assessing vul- Doctoral Thesis).
nerability to climate change and variability in Northern Ghana. Nyumba, T. O., Wilson, K., Derrick, C. J., & Mukherjee, N. (2018). The use
Journal of Environment and Earth Science, 3(2), 157–170. of focus group discussion methodology: Insights from two decades of
Gautam, Y., & Andersen, P. (2016). Rural livelihood diversification and application in conservation. Methods in Ecology and Evolution, 9(1),
household well-being: Insights from Humla, Nepal. Journal of Rural 20–32. doi:10.1111/2041-210X.12860
Studies, 44, 239–249. Oparinde, A., & Hodge, I. (2011). Building livelihood resilience: A case
Gallopin, G. C. (2006). Linkages between vulnerability, resilience, and study of factors affecting farm households’ adoption of coping and adap-
adaptive capacity. Global Enviornmental Change, 16, 293–303. doi:10. tive strategies in rural Nigeria. Retrieved from https://mpra.ub.uni-
1016/j.gloenvcha.2006.02.004 muenchen.de/39162/
Gerlitz, J., Banerjee, S., Hoermann, B., Hunzai, K., Macchi, M., & Tuladhar, Panthi, J., Aryal, S., Dahal, P., Bhandari, P., Krakauer, N. Y., & Pandey, V.
S. (2014). Poverty and vulnerability assessment (PVA): A survey instru- P. (2015). Livelihood vulnerability approach to assessing climate change
ment for The Hindu Kush Himalayas. Kathmandu, Nepa: ICIMOD. impacts on mixed agro-livestock smallholders around the Gandaki
Retrieved from http://rds.icimod.org/Home/Data?group=4&&themek River Basin in Nepal. Regional Environmental Change, 16(4), 1121–
ey=Vulnerability 1132. doi:10.1007/s10113-015-0833-y
Gerlitz, J. Y. M., Macchi, M., Brooks, N., Pandey, R., B, S., & Jha, S. K. Piya, L., Joshi, N. P., & Maharjan, K. L. (2012). Vulnerability of Chepang
(2016). The multidimensional livelihood vulnerability index – an households to climate change and extremes in the Mid-hills of Nepal.
instrument to measure livelihood vulnerability to change in the Climatic Change, 135(3–4), 521–537. doi:10.1007/s10584-015-1572-2
Hindu Kush Himalayas. 5529(June). doi:10.1080/17565529.2016. Simane, B., Zaitchik, B. F., & Foltz, J. D. (2014). Agroecosystem specific cli-
1145099. mate vulnerability analysis: Application of the livelihood vulnerability
Gibson, J. E. (2012). Interviews and focus groups with children: Methods index to a tropical highland region. Mitigation and Adaptation
that match children’s developing competencies. Journal of Family Strategies for Global Change, 21(1), 39–65. doi:10.1007/s11027-014-
Theory & Review, 4(2), 148–159. doi:10.1111/j.1756-2589.2012.00119.x 9568-1
Green, D., Jackson, S., & Morrison, J. (2009). Risks from climate change to Sullivan, C. A., Meigh, J. R., & Fediw, T. S. (2002). Derivation and testing of
indigenous communities in the tropical North of Australia. Canberra: the water poverty index phase 1. In Deparment for International
Department of Climate change and Energy Effiency. Development (DFID) (Vol. 1). Retrieved from http://www.soas.ac.uk/
Gregory, P. J., Ingram, J. S. I., & Brklacich, M. (2005). Climate change and water/publications/papers/file38386.pdf
food security. Philosophical Transactions of the Royal Society B: Swindale, A., & Bilinsky, P. (2006). Household dietary diversity score
Biological Sciences, 360(1463), 2139–2148. doi:10.1098/rstb.2005.1745 (HDDS) for measurement of household food access: Indicator guide.
Hahn, M. B., Riederer, A. M., & Foster, S. O. (2009). The livelihood vulner- Food and Nutrition Technical Assistance … , 11. doi:10.1017/
ability index: A pragmatic approach to assessing risks from climate varia- CBO9781107415324.004
bility and change-a case study in Mozambique. Global Environmental Taherdoost, H. (2016). Sampling methods in research Methodology ; How
Change, 19(1), 74–88. doi:10.1016/j.gloenvcha.2008.11.002 to choose a sampling sampling methods in research methodology; how
IBM. (2015). IBM Corp. Released 2015. IBM SPSS statistics for windows, to choose a sampling technique for. International Journal of Acedemic
version 23.0. Armonk, NY: IBM Corp. Reasearch in Managemnet, 5(2), 2296–1747. doi:10.2139/ssrn.3205035
IPCC (Intergovernmental Pannel on Climate Change). (2001). Climate Tambo, J. A. (2016). Adaptation and resilience to climate change and
change 2001: Impacts, adaptation, and vulnerability. In J. J. variability in north-east Ghana. International Journal of Disaster Risk
McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White Reduction, 17, 85–94. doi:10.1016/j.ijdrr.2016.04.005
(Eds.), Contribution of working group II to the third assessment report. Tanner, T., Lewis, D., Wrathall, D., Bronen, R., Cradock-Henry, N., Huq,
Cambridge: Cambridge University Press. Retrieved from http://www. S., … Thomalla, F. (2015). Livelihood resilience in the face of climate
preventionweb.net/files/8387_wg2TARfrontmat%0Ater1.pdfdf change. Nature Climate Change, 5(1), 23–26. doi:10.1038/nclimate2431
IPCC (Intergovernmental pannel on Climate Change). (2007). Climate Tesso, G., Emana, B., & Ketema, M. (2012). Analysis of vulnerability and
change 2007: Impacts, adaptation and vulnerability. In M. L. Parry, resilience to climate change induced shocks in North Shewa, Ethiopia.
O. F. Canziani, J. P. Palutikof, P. J. van der Linden, & C. E. Hanson Agricultural Sciences, 3(6), 871–888.
(Eds.), Contribution of working group II to the Fourth assessment Weldegebriel, Z. B., & Amphune, B. E. (2017). Livelihood resilience in the
Report of the Intergovernmental Panel. Cambrideg University. face of recurring floods: An empirical evidence from Northwest
Cambridge: Cambridge University Press. doi:10.1256/004316502320 Ethiopia. Geoenvironmental Disasters, 4(1), 10. doi:10.1186/s40677-
517344. 017-0074-0
Kothari, C. R. (2004). Research methodology. Jaipur (India): New Age World Bank. (2006). Ethiopia: Managing water resources to maximize sus-
International (Pvt) Ltd, publishers (2nd ed.). New Delhi: New Age tainable growth. Washington, DC: Country water resources assistance
International Publisher Ltd. strategy.