Journal of Environmental Management 276 (2020) 111275
Contents lists available at ScienceDirect 
Journal of Environmental Management 
journal homepage: http://www.elsevier.com/locate/jenvman 
Research article 
Assessment of smallholder farmers’ adaptive capacity to climate change: 
Use of a mixed weighting scheme 
Yves C. Zanmassou a,b,*, Ramatu M. Al-Hassan b, Akwasi Mensah-Bonsu b, Yaw B. Osei-Asare b, 
Charlemagne B. Igue a 
a Faculté des Sciences Economiques et de Gestion (FASEG), Université d’Abomey-Calavi (UAC), Benin 
b Department of Agricultural Economics and Agribusiness (DAEA), CBAS, University of Ghana, Legon   
A R T I C L E  I N F O   A B S T R A C T   
Keywords: Weighting scheme definition represents an important step in assessment of adaptive capacity to climate change 
Weighting scheme with indicator approach since it defines the trade-offs among indicators or components and can be source of 
Adaptive capacity uncertainty. This study aims to assess smallholder farmers’ adaptive capacity to climate change by using a mixed 
Smallholder farmers weighting scheme that reflect farmers’ perceived importance of adaptive capacity components to inform policy 
Climate change 
makers. To achieve that objective, the sustainable livelihood framework was adopted and indicator approach was 
used for the assessment. The mixed weighting scheme were defined by using both equal weights and experts 
judgement methods during the assessment process. The mixed weighting scheme index is compared to the case 
where equal weights are applied in the assessment process and an uncertainty analysis was performed on relative 
standard deviation through a Monte Carlo simulation. Primary Data were collected from 450 farmers in two 
communities in northern Benin with a structured questionnaire and through focus groups discussion. The results 
show that smallholder farmers in both communities do not have the same perceived importance of adaptive 
capacity components. The index scores show that farmers have in majority low adaptive capacity. When 
weighted product aggregation method is used, there is more uncertainty related to the index computed with the 
mixed weighting scheme, but it leads to the same characterisation when compared with the index computed with 
the equal weights. It is recommended that mixed weighting scheme should be preferred for the assessment of 
adaptive capacity and weighted product aggregation method should be used.   
1. Introduction system to adjust to a current or future threats (Smit and Wandel, 2006). 
Because of the nature of adaptive capacity which is latent and does not 
Assessment of adaptive capacity to climate change has received great have indicator to measure it (Engle, 2011; Engle and Lemos, 2010), 
attention from scholars and policy makers last decade because of the composite indices are generally used to quantify and characterise the 
necessity for both developing and developed countries to reduce the concept. The use of composite indices to evaluate adaptive capacity 
negative impact of climate change through adaptation. The aims of these requires for aggregation a weighting scheme to highlight the relative 
attention were to define the concept (Hinkel, 2011; Smit and Wandel, importance of sub-components (Gan et al., 2017; Nardo et al., 2005). 
2006; Yohe and Tol, 2002), provide framework, methods and tools for Moreover, when there is some compensation among components used to 
its assessment (Adger and Vincent, 2005; Brooks et al., 2005; Bryan define adaptive capacity, weighting scheme can express the trade-off 
et al., 2015) to better inform adaptation policy. Adaptive capacity is one ratios among the dimensions (Böhringer and Jochem, 2007). Thus, 
of the important factors that define vulnerability to climate change inappropriate definition of weighting scheme can make adaptive ca-
(Adger et al., 2007; Engle, 2011; Gallopín, 2006) and it assessment pacity index misleads vulnerability assessment and then adaptation 
represents an important step in vulnerability evaluation and provides policy (Gan et al., 2017). However, during index development there is 
means to develop adaptation policy since factors that underlies and generally no information on components or indicators contribution to 
enables adaptation activities are analysed (Adger et al., 2007; Juhola the final index that are defined. That leads to a definition of weighting 
and Kruse, 2013). Adaptive capacity is generally defined as ability of a scheme which does not reflect the importance of each sub-components 
* Corresponding author. Faculté des Sciences Economique et de Gestion (FASEG), Université d’Abomey-Calavi (UAC), Benin. 
E-mail address: yzanmassou@outlook.com (Y.C. Zanmassou).  
https://doi.org/10.1016/j.jenvman.2020.111275 
Received 20 May 2020; Received in revised form 1 August 2020; Accepted 19 August 2020   
Available online 4 September 2020
0301-4797/© 2020 Elsevier Ltd. All rights reserved.
Y.C. Zanmassou et al.                                                                                                                                                                 J  o  u r  n a  l  o  f  E  n  v i r o  n  m   e n  t a  l  M   a n  a  g e  m   e n  t  276 (2020) 111275
and constitute a source of uncertainty for the composite index (Burgass information on farmers’ adaptive capacity so that adaptation option that 
et al., 2017; Marzi et al., 2018). will be proposed could be easily adopted by farmers and will not lead to 
Methods generally use to define weights for composite indices to maladaptation among farmers due to lack of adaptive capacity. 
assess adaptive capacity to climate change can be grouped into three In order to assess farmers’ adaptive capacity to climate change using 
categories: equal weighting, data driven weighting and experts’ opinion- a composite index with a weighting scheme that reflect farmers’ 
based weighting (Gan et al., 2017; Nardo et al., 2005). Although, there is perceive importance of each components, we use a mixed method that 
no agreement among scholars on which methods seems the best, equals combine equal weight and expert judgment methods to define weights. 
weighting represents the most used method because of its simplicity and The equal weight method is applied during aggregation of indicators to 
ability to give the same level of importance to each component (Greco obtain components because of its simplicity for application and ability to 
et al., 2019). However, this might reveal the lack of knowledge on the deal with a large number of variables while the experts’ judgement is 
importance of components to the final index or the causal relationship used to determine weight that is applied to each component for aggre-
among indicators (Nardo et al., 2005; Wang and Fu, 2019). For data gation into a composite index. Specifically, budget allocation method is 
driven weighting methods, weights are derived based on statistical tools used to reveal farmers’ perceived importance of adaptive capacity 
which rely on characteristic of data and correlation analysis to help components during a focus groups using farmers as experts. The index 
avoid double counting. However, this could lead to a missing of obtained is useful for policy implementation because it gives the current 
important variables who are correlated and definition of inappropriate level of adaptive capacity and also which components policy interven-
composite indicator since weights are based on values and may not tion should prioritise regardless of the values of the sub-component. The 
highlight the importance of information variables conveyed (Gan et al., efficiency of the mixed weighting method is assessed through an un-
2017; Greco et al., 2019). Equals and data driven weights definition certainty analysis using Monte Carlo simulation and compared with the 
methods may mislead policy intervention if the composite index been case where equal weight method is used through the hall process. 
measured is to capture information at individual level such as farmers’ 
adaptive capacity to climate change where weights have to reflect 2. Methods 
farmers’ perceived importance of each components to inform policy 
makers. By contract, experts’ opinion weighting methods which define 2.1. Choice of indicators 
weights based on value judgment of experts can be appropriate to define 
weights for composite index at individual level if theirs results were not Indicators approach is adopted to assess smallholder farmers’ 
subject to the number of indicators assessed (Greco et al., 2019; Nardo adaptive capacity. However, the theory-driven approach was adopted 
et al., 2005). and the sustainable livelihood framework (Ellis, 2000; Scoones, 1998) is 
Empirical studies that assessed smallholder farmers’ adaptive ca- followed to define the components of farmers’ adaptive capacity to 
pacity to climate change with indicator approach generally formulate climate change. In this context, farmer’s adaptive capacity represents a 
policies interventions to improve only components of adaptive capacity stock of capital available to formulate his adaptations strategies to 
that have low score (Abdul-razak and Kruse, 2017; Egyir et al., 2015). In climate change within an institutional context with the purpose to 
so Doing, they do not exploit information provided by the weighing achieve a sustainable livelihood. Like previous studies on adaptive ca-
scheme and may fail to have farmers’ adhesion to the policy. Moreover, pacity at household or individual levels (Bryan et al., 2015; Defiesta and 
in a context of developing countries where policy makers have limited Rapera, 2014; Ruiz Meza, 2015; Shah and Dulal, 2015), the five capitals 
resources for interventions, all components of adaptive capacity could (social, human, physical, natural and financial) essentials for a sus-
end up with low scores. Choice have to be made on which component to tainable livelihood are used as components of farmers’ adaptive ca-
prioritise for policy intervention to have support of farmers. pacity since they can determine actions a farmer can undertake against 
In Benin, like most west African country, the agricultural sector climate change impact within a policy and institutional context 
contributes up to 30% of GDP and employs nearly 75% of labour force (Scoones, 1998). Indicators for each components were selected 
(INSAE, 2012; PAM, 2009). Agricultural production depends heavily on following the theory-driven approach and empirical studies on assess-
rainfall and is characterized by smallholder farmers who farm on ment of adaptive capacity to climate. A total of twenty-one indicators 
average 1.7 ha of land (MAEP, 2011). Challenges facing the agricultural were used to assess smallholder farmers’ adaptive capacity as sum-
sector include access to markets, low use of production inputs and soil marised in Table 1. 
infertility (MAEP, 2011; PAM, 2009). The sector is also affected by 
climate change which is observed through high concentration of rainfall 2.2. Data analysis 
over a short period, fluctuation of the onset and duration of rainfall 
season, drought, floods, decrease in rainfall and increased of tempera- 2.2.1. Normalization of indicators 
ture (Boko et al., 2012). In fact, it was observed an increased in the Indicators selected were normalised before indices were computed 
average temperature by more than 1 ◦C and a decrease in the average because they do not have the same measurement unit and cannot be 
rainfall by 5.5 mm per year between 1960 and 2008 (Gnanglè et al., aggregated. Zhou, Ang, and Poh (2006), approach was used to normalise 
2011). Climate simulation studies by the horizon of 2100 indicate that indicators to satisfy “the larger the better” interpretation. Let Vijk be a 
temperature could increase between +2.6 ◦C and +3.2 ◦C and the variable representing indicator k of jth component of adaptive capacity 
northern part of Benin will experience a decrease in rainfall between for the ith individual and let x be the observed measurement of that 
13% and 15% (MEHU, 2011). ijk indicator. The normalization was done using equation (1) where r is 
The northern part of Benin, which represents the basket of foods and ijk
crops, plays an important role in the country’s food security and ma- the resulting measure after transformation. 
jority of dwellers livelihood in this area depends on agriculture (Yeg- ⎧ x⎪ {ijk } if Vijk satisfies "the larger the better"
bemey et al., 2013). Assessments of the impact of climate change reveal ⎪⎨ max xijki
that the northern Benin is the most vulnerable agro-ecological area to rijk = { } (1)  
climate change and has been affected by all extreme climatic events that ⎪min x⎪ ijk
⎩ i if Vijk satisfies "the smaller the better"occurred in Benin from 1984 to 2010 (Boko et al., 2012; MEHU, 2011; xijk
Yabi and Afouda, 2012). There is therefore a need to design and 
implement policies that support smallholder farmers in northern Benin 2.2.2. Mixed weighting schemes definition 
to reduce their vulnerability to climate change. However, policies to Two weights schemes were used to combine indicators into com-
support farmers to reduce their vulnerability to climate change requires posite index, equal weight and expert judgement. At the first stage, the 
2
Y.C. Zanmassou et al.                                                                                                                                                                 J  o  u r  n a  l  o  f  E  n  v i r o  n  m   e n  t a  l  M   a n  a  g e  m   e n  t  276 (2020) 111275
Table 1 
Components and indicators of farmers’ adaptive capacity to climate change.   
Components of adaptive Indicators of components Supporting sources 
capacity 
Adaptive capacity to Social capital  - Member of FBO  Jacob et al. (2015); Ruiz Meza (2015); Egyir et al. (2015); Bryan et al. (2015);  
climate change - Number of relatives in the Lockwood et al. (2015); Abdul-razak and Kruse (2017) 
community  
- Participation to community 
activities 
Human capital  - Year of schooling  Yohe and Tol (2002);Jacob et al. (2015); Gbetibouo et al. (2010); Ruiz Meza (2015);  
- Experience in agriculture  Defiesta and Rapera (2014); Lockwood et al. (2015);  
- Number of crops grown  Bryan et al. (2015). 
- Visits of extension services  
- received a training in agriculture 
Physical capital  - total area cropped  Jacob et al. (2015); Ruiz Meza (2015); Defiesta and Rapera (2014); Bryan et al. 
- distance house to farm  (2015);  
- distance house to market  Juhola and Kruse (2013) 
- distance house to financial 
institution  
- distance house to extension 
services office 
Natural capital  - Rainfall variability  Jacob et al. (2015); Ruiz Meza (2015); Egyir et al. (2015); Dannevig et al. (2015);  
- soil Fertility  Bryan et al. (2015) 
- Land ownership  
- Experienced with natural hazard 
events on farm 
Financial capital  - Off farm Income  Ruiz Meza (2015); Defiesta and Rapera (2014); Gbetibouo et al. (2010); Bryan et al. 
- value of livestock  (2015); Lockwood et al. (2015); Jacob et al. (2015) 
- crop income  
- value of agricultural equipment  
- Credit sources diversification 
Source: Authors compilation 
Fig. 1. Study area and villages visited. Source: Topographical Map IGN-Bénin.  
3
Y.C. Zanmassou et al.                                                                                                                                                                 J  o  u r  n a  l  o  f  E  n  v i r o  n  m   e n  t a  l  M   a n  a  g e  m   e n  t  276 (2020) 111275
equal weight method, because of its simplicity for application and solution proposed in the literature to overcome this limitation is the 
ability to treat all indicators in the same (Wang and Fu, 2019), is applied characterisation of the index to present results in visual form to highlight 
during the first stage of aggregation where the normalised indicators are distribution and heterogeneity among farmers (Gupta et al., 2010; 
combined into components (social, human, natural, physical and Juhola and Kruse, 2013). This approach was adopted and the commonly 
financial capital) of adaptive capacity. At this stage, it was assumed that use method that transforms quantitative scores into qualitative values by 
each indicator has the same importance in defining the component that dividing the range of scores into three groups with equal interval was 
they explained. However, this implied that indicators that have higher applied. Following Asante et al.(2012), farmers’ estimated adaptive 
value have more impact on the value of the component compared to capacity ACi was characterised into three (03) groups. A farmer is 
indicators with lower value. At the second stage, to combine compo- considered having low capacity if 0 < ACi < 0.33 , moderate capacity if 
nents into a final composite index, expert judgement is used to define 0.33 < ACi < 0.66 and high capacity if 0.66 < ACi ≤ 1. 
weighting schemes through focus group discussions. Since components 
of adaptive capacity are related to livelihood, weights will be easier to 2.2.4. Uncertainty analysis 
assign by local expert (Below et al., 2012) and allows the involvement of An uncertainty analysis was performed on the adaptive capacity 
farmers in the assessment process. Community leaders who are farmers indices to find out wether the use of mixed weighting scheme is subject 
were used as experts to define weights. Farmers’ perceived importance to less uncertainty compared to the common case where the euqal 
of adaptive capacity’s components in their adaptation process were used weighting scheme is applied. The relative standard deviation (RSD) was 
as weights for each component. The importance of each components was used as indicateur of incertainty. In order to analyse the consistency of 
elicited through focus groups discussion with community leaders in the the uncertainty, a Monte Carlo simulation was performed to appreciate 
study area (Malanville and Karimma). the value of RSD related to the adaptive capacity index from 10000 
samples. 
2.2.3. Aggregation methods 
A combination of both simple additive and weighted product 
methods were adopted to combine indicators into adaptive capacity 2.3. Data 
index. This choice is made to take advantage of the properties of each 
method. The simple additive method of aggregation was used to The study is carried out in the northern part of Benin, in Alibori re-
combine indicators into indices that represent components of farmers’ gion because it is the most productive in food crop production and is also 
adaptive capacity (social, human, natural, physical and financial exposed to climate change events (Vodounou and Doubogan, 2016; Yabi 
capital). and Afouda, 2012; MAEP, 2011). A multi-stage sampling technique was 
LetSC be the index that captures farmer’s i adaptive capacity used to select smallholder farmers. At the first stage, two municipalities ij 
component j. SC are obtained from equation (2) which is a simple mean were selected within Alibori. The vulnerability to weather risks criteria ij 
of the normalised measured (r ) of n indicators that define the was used at this stage and then Malanville and Karimama municipalities ijk
component j. were selected because they are the most vulnerable to climate change 
(MEPN, 2007). At the second stage, two districts were selected within 
1 ∑n each municipality based on density of agricultural population and the 
SCij = rijk (2) n same criteria were used to select two villages within each district at the k=1
third stage. Lists of farmers per village were obtained from extension 
After the indices that define adaptive capacity’s components been offices at Malanville and Karimama and random selection of farmers was 
computed, farmers’ adaptive capacity was calculated using the weighted performed. Fig. 1 shows the location of Alibori region on Benin’s map, 
product method because it takes into account the complementarity Karimama and Malanville districts and locations of villages where data 
among components, penalises components with poor values and is the were collected. 
best among the multiple attributes decision making (MADM) methods The minimum required sample size for the study was obtained using 
since it has the minimum loss of information (Zhou et al., 2010, 2006). equation (4) and proportionality sampling technique is used to deter-
Equation (3) indicates how adaptive capacity index (ACi) of a farmer i is mine the number of smallholder farmers to interview in each village. 
obtained with the weighted product method where wj the weight 
assigned to component j and t the number of components that defined Table 3 
adaptive capacity. Socioeconomic characteristic of respondents.  
∏t ( )w Variables All Malanville Karimama T-test or ACi = SCij (3) Chi2 
j=1
Sex 
The index scores obtained are comprised between 0 and 1. Present Man 0.91 0.87 0.96 − 3.38*** 
directly, the scores obtained for the index may limit its efficacy of the Age 48.63 48.39 48.95 − 0.83 
index to communicate with policy makers (Juhola and Kruse, 2013). The Education 
Not educated 0.60 0.57 0.63 2.21 
Primary 0.30 0.33 0.26  
Secondary 0.10 0.10 0.11  
Table 2 Experience in Agriculture 24.75 25.95 23.13 3.16*** 
Numbers of farmers interviewed per village.  Main Crop 
Municipalities districts villages Sample size Maize 0.29 0.33 0.23 9.69** 
Rice 0.32 0.32 0.31  
Malanville Malanville Wollo 57 Millet 0.20 0.16 0.26  
Bodjécali 67 Sorghum 0.19 0.19 0.20  
Guéné Koara-Tédji 67 Farm size 2.4 2.7 2.05 5.53*** 
Goungoun 66 FBO 0.67 0.68 0.66 0.48 
Karimama Karimama Karimama 47 Extension services 0.75 0.77 0.72 1.29 
Gouroubéri 46 Access to credit 0.62 0.60 0.65 − 0.99 
Birni-Lafia Birni-Lafia 50 Perception of climate change 0.90 0.87 0.92 − 1.75* 
Karigui 50 Experience with natural 0.31 0.30 0.33 − 0.81 
Total   450 hazard event 
Source: Authors Source: Authors’ Field work (2016) 
4
Y.C. Zanmassou et al.                                                                                                                                                                 J  o  u r  n a  l  o  f  E  n  v i r o  n  m   e n  t a  l  M   a n  a  g e  m   e n  t  276 (2020) 111275
Table 4 important component. For the second activity, the budget allocation 
Component of farmers’ adaptive capacity to climate change.  technique is applied and ten stones were giving to farmers to share on 
Component of All sample Karimama Malanville Prob(t- the components of adaptive capacity respect to their contribution to 
Adaptive test) adaptation process from their experience. More stones on a component 
capacity 
Mean Std Mean Std Mean Std means that component is important in the adaptation process in the area 
dev dev dev  and must have higher weight. The same number of stones on component 
Social capital 0.30 0.13 0.33 0.13 0.28 0.13 0.000*** reveal that components of adaptive capacity have the same importance 
Human capital 0.24 0.09 0.25 0.09 0.24 0.09 0.383 in adaptation process and then deserve the same weight. The weight of 
Natural capital 0.38 0.08 0.36 0.09 0.40 0.07 0.000*** each component is obtained by dividing the number of stone put on the 
Physical capital 0.20 0.08 0.24 0.09 0.17 0.05 0.000*** component by the total stones given. 
Financial 0.27 0.11 0.28 0.11 0.28 0.11 0.893 
capital 
3. Results 
Source: Authors’ Field work (2016) 
3.1. Socioeconomic characteristics of respondents 
Table 5 
Weighting scheme used to compute adaptive capacity index.  Table 3 summarizes socio-economic characteristics of farmers 
interviewed. It shows that man are more involved in crop production 
Adaptive capacity component Malanville Karimama than woman in the study area and this is observed in the two commu-
rank weight rank weight nities. Farmers’ population is dominated by non-educated persons and 
Social capital 2 0.2 1 0.3 they have on average twenty four years of experience in agriculture with 
Human capital 3 0.2 3 0.2 farmers in Malanville more experienced than farmers in Karimama. 
Physical capital 4 0.2 4 0.2 Maize and rice are the most grown crop by cereal farmers in Malanville 
Natural capital 5 0.1 5 0.1 compared to rice and millet for farmers in Karimama. Majority of 
Financial capital 1 0.3 2 0.2 
farmers interviewed have access to institutional support such as exten-
Source: Authors’ Field work (2016) sion service and credit (either formal or informal) for agricultural pro-
duction and are also member of farmers based organization (FBO). 
t2p(1 − p) Climate change is observed in the study area by majority of farmers in 
n= 2 (4)  E term of increase in temperature, decrease of annual rainfall and late 
onset of rains. However only 31% of farmers experienced natural events 
Where t represents the t-value corresponding to the level of confidence such as flood and drought on their farms in the last five years. 
(α) chosen by the researcher, P the proportion of population having 
attributes of interest and E the margin of error. Since agriculture is the 
main source of income and employs 70% of labour force in Benin 3.2. Components of adaptive capacity to climate change 
(INSAE, 2012), it is assumed that 70% of population in Alibori region 
Benin are engaged in agriculture, thus P 0.70 and a 5% (E 5%) Assessment of indices that captured components of smallholder = =
margin error was adopted. The minimum required sample size was 323 farmers’ adaptive capacity are summarised in Table 4. It shows that 
farmers. Because of non-responses to some questions which will lead to farmers adaptive capacity to weather risk in the study area are built with 
missing data and reduce the sample size for analysis, the sample size was 
over-estimated to be able to meet the minimum sample size requirement 
after data cleaning. A total of 450 smallholder farmers were randomly 
selected and interviewed with structured questionnaire. Table 2 shows 
the repartition of farmers interviewed per villages. 
2.3.1. Focus group discussion 
One focus group discussion with community leaders was carry out in 
each community. The group of community leaders was composed by 
chef of village visited in a community assisted by 3 or 4 notables that are 
well known and are involved in agricultural production. They were in 
total 8 leaders that participated in the discussion for Malanville 
compared to 7 leaders for Karimama. Two main activities were executed 
during the discussion. For the first activity farmers were asked, after the 
meaning of the concept being explained to them, to rank the five capi-
tals, social, natural, physical, human and financial, respect to their 
contribution to build adaptive capacity from their experience. The Fig. 2. Characterisation of farmers’ adaptive capacity based on weighted 
ranking was performed from one, most important to five, the least product index. Source: Authors’ Field work (2016). 
Table 6 
Farmers’ adaptive capacity to climate change index.   
Weighted product method Simple Additive method 
Mixed weights Equal weights Difference Mixed weights Equal weights Difference 
All sample 0.255 (26) 0.261 (23) − 0.006*** 0.274 (24) 0.282 (21) − 0.008*** 
Karimama 0.275 (25) 0.277 (23) − 0.002** 0.291 (24) 0.294 (21) − 0003*** 
Malanville 0.241 (25) 0.250 (22) − 0.008*** 0.261 (22) 0.273 (20) − 0.011*** 
***p < 0.01, **p < 0.05. Value in parentheses are the relative standard deviation (RSD). 
Source: Authors’ Field work (2016) 
5
Y.C. Zanmassou et al.                                                                                                                                                                 J  o  u r  n a  l  o  f  E  n  v i r o  n  m   e n  t a  l  M   a n  a  g e  m   e n  t  276 (2020) 111275
same importance to the five capitals (social, financial, human, natural 
and physical) to build their adaptive capacity and how they affect the 
adaptation process. However, there is inconsistency between the 
ranking and the weights proposed by leaders in both communities 
through the budget allocation exercise. Component that were ranked at 
second, third and fourth for their importance in building adaptive ca-
pacity received the same (number of stones) weight for their contribu-
tion to adaptation process. Despite that, the weights proposed by leaders 
during the budget allocation exercise were used to compute the com-
posite index of adaptive capacity since they reflect the contribution of 
adaptive capacity components to adaptation process. 
3.4. Smallholder farmers’ adaptive capacity to weather risk 
Regardless of the methods and the weighting schemes used to 
Fig. 3. Characterisation of farmers’ adaptive capacity based on simple additive compute adaptive capacity indices, the results show that, on average 
index. Source: Authors’ Field work (2016). smallholder farmers have low adaptive capacity to weather risks in the 
study area (Table 6). This result is consistent with the claim that ma-
a low level of social, human, physical and financial capital and moderate jority of smallholders farmers in developing country have low adaptive 
level for naturel capital. However, there is disparity between the two capacity as shown by previous studies (Egyir et al., 2015; Mekonnen and 
municipalities for some components of adaptive capacity. Farmers in Kassa, 2019; Shirima et al., 2017). Results also show that, there is a small 
Karimama municipality appear to have a relatively higher social, difference between the average values of indices estimated with mixed 
naturel, and physical capital compared to farmers in Malanville mu- weighting scheme compared to the average value of indices estimated 
nicipality and the differences are statistically significant. with equal weighting schemes respect to methods used to compute the 
indices. This could suggest that, respect to the method use, mixed 
3.3. Weighting scheme proposed by smallholder farmers for adaptive weights and equal weights leads to the same results. 
capacity components However, the differences in the average value of indices for each case 
are statistically different to zero and the relative standard deviation 
The results of focus groups discussion with leaders in the two com- obtained for indices computed with mixed weights scheme are slightly 
munities are presented in Table 5. Community leaders in both Malan- higher than those obtained for indices computed with equal weights 
ville and Karimama, ranked social and financial capital as important scheme. These results suggest that there is more uncertainty related to 
components to build adaptive capacity to climate change whilst natural indices computed with mixed weights scheme compared to indices 
capital is considered the least important component. The same results computed with the equal weights scheme. 
were also observed during the budget allocation exercise. However, The characterisation of the indices (Figs. 2 and 3) shows that, respect 
leaders in Karimama indicate social capital as the most important capital to the methods used to compute the index (weighted product or simple 
to build adaptive capacity within the adaptation process compared to additive), mixed weighting and equal weighting schemes lead to the 
leaders in Malanville who point out financial capital. From farmers’ same results. In fact, with the weighted product method, both mixed 
perspective in both communities, although financial resources are weights and equals weights indices indicate that majority of smallholder 
crucial to respond to weather risks, collective action, support from farmers (84%) in the study area have low adaptive capacity to weather 
friends and relatives and networking are also important to avoid or risk while with the simple additive method both methods weighting 
overcome the negative impact of weather risks events. The main reason scheme show that 76% of smallholder farmers that low adaptive 
pointed out by community leaders during the discussion was that “when capacity. 
an individual pass through a difficult moment due to weather risks, reliefs 
from friends and family boost their morale to surmount the problem psy- 3.5. Uncertainty analysis on adaptive capacity indices 
chologically before financial resources play their roles”. These results are 
consistent with others studies who demonstrate that financials capital The results of uncertainty analysis using Monte Carlo simulation are 
(Monterroso et al., 2018), relationship and network (Keys et al., 2016) summarised in Table 7. It shows that over 10000 simulations, small-
are keys factors to build adaptive capacity to climate change. The results holder farmers have low adaptive capacity in the study area. The aver-
provide also empirical support to the view that social factors are ages values of adaptive capacity indices obtained with mixed weights 
important as financial factors to build farmer’s adaptive capacity to and equals weights are closed when the indices are computed with the 
climate change (Clay and King, 2019; Hogan et al., 2011; Williges et al., weighted product method although the difference is statistically signif-
2017). icant. Similar pattern is observed for uncertainties (RSD) related to the 
These results show a similarity between the ranking and weighting indices. When adaptive capacity indices are computed with simple ad-
proposed by leader un both communities and suggest leaders have the ditive method, simulation shows that there is large difference between 
same view of adaptive in two communities, but they do not attribute the the average value of mixed weights index and the average value of equal 
Table 7 
Adaptive capacity scores and associated RSD from Monte Carlo simulation.   
Weighted product method Simple additive method 
Mixed weights Equal weights Difference Mixed weights Equal weights Difference 
Indices 0.24 0.25 − 0.01*** 0.41 0.27 0.14*** 
RSD (%) 22 20 0.02*** 46 16 30*** 
Number of simulations 10000 
***p < 0.01, **p < 0.05. 
Source: Authors’ Field work (2016) 
6
Y.C. Zanmassou et al.                                                                                                                                                                 J  o  u r  n a  l  o  f  E  n  v i r o  n  m   e n  t a  l  M   a n  a  g e  m   e n  t  276 (2020) 111275
weighted index and the same pattern is observed between uncertainties unit of analysis and then fail to provide relevant policy information 
related to each index.The results suggest that the mixed weighting except the level of adaptive capacity that will be provided. This study 
scheme performed better when applied on the weighted product ag- assesses smallholder farmer’s adaptive capacity to climate change by 
gregation method and produced almost the same results as the equal using a mixed weighting scheme. The mixed weighting scheme is 
weights. defined through the combination of equal weights and expert judgment 
methods to define a weighing scheme that reflect farmers’ perceived 
4. Discussion importance of each components of their adaptive capacity since farmers 
were used as experts. The results show that the mixed weighing scheme 
The assessment shows that smallholder farmers do not attribute the adaptive capacity index lead to the same characterisation as the equal 
same importance to the five capital of sustainable livelihood framework weighting scheme index when weighted product aggregation method is 
to build their adaptive capacity and that change from one community to used to combine components. This provides empirical support to the 
another. These results suggest that weighting scheme should be specific growing demand to integrate the quantitative assessment framework 
to each community and are consistent with other studies who found that with elements of the qualitative approach such stakeholder interactions 
communities of farmers are heterogeneous groups (Hogan et al., 2011; and use of experts. However, the results should be taken with caution 
Smit and Wandel, 2006) and one-size-fits-all weighting scheme (equal since the framework used for the analysis focus mostly on socioeco-
weights) methods are inappropriate to reflect the socio-economic and nomic factors compared to environmental factors which also define 
environmental conditions under which farmers mediate their livelihood adaptive capacity. Moreover, the selection and the number experts used 
decision (Hinkel, 2011) and the trade-offs among components. during the focus group and the method used to elicit leaders perceived 
The analysis shows that mixed and equals weighting schemes leads importance of adaptive capacity may influence the quality of informa-
to similar results when the composite indices is computed with weighted tion provided. The definition of an assessment framework that can 
product methods. However, they have different results when simple provide the appropriate method of selection of experts and their number 
additive aggregation method is used to compute the composite indices. with the way in which their opinions will be taken into account, without 
These results indicate how important are the choice of weighting scheme increase uncertainty related to the indices, would advance knowledge 
and aggregation methods in index construction and are consistent with on integrated methods of adaptive capacity assessment.Though, there is 
those obtained by Marzi et al. (2018), who show that adaptive capacity more uncertainty related to the mixed weighting method, it should be 
index score is sensitive to the aggregation and weighting methods. Re- prioritised for weights definition when adaptive capacity is assessed at 
sults also suggest that weighted product method is the most appropriate individual level with indicators approach. 
aggregation method when the mixed weighting scheme is adopted. 
Both indices lead to the conclusion that smallholder farmers in both CRediT authorship contribution statement 
communities, Karimama and Malanville, have in majority low adaptive 
capacity. With equal weights index, policy makers may not know which Yves C. Zanmassou: Conceptualization, Methodology, Formal 
components to prioritise for policy interventions or apply the same analysis, Writing - original draft. Ramatu M. Al-Hassan: Writing - re-
strategy for both communities since most of the component indices view & editing, Investigation, Visualization, Supervision. Akwasi 
suggest low level of capital. This could be misleading since the two Mensah-Bonsu: Writing - review & editing. Yaw B. Osei-Asare: 
municipalities are not homogenous groups. With mixed weights index, Writing - review & editing. Charlemagne B. Igue: Investigation, 
interventions that create incentive for social capital improvement could Supervision. 
be applied in Karimama while interventions to improve farmers’ access 
to financial capital could be implemented in Malanville since farmers 
reveal those capitals as most important components for their adaptive Declaration of competing interest 
capacity. These additional information provides by the mixed weighting 
scheme index were obtained thought the weights definition process The authors declare that they have no known competing financial 
which make that index more relevant for policy formulation compared interests or personal relationships that could have appeared to influence 
to the index computed with equal weighting scheme. the work reported in this paper. 
This suggests that improvement in methods used to define weights of 
composite index to evaluate adaptive capacity can help have addition Acknowledgment 
information useful for policy implementation while maintaining the 
same level of information provided by the final results. The combination Thanks to anonymous reviewers for their helpful comments. Authors 
of weights definition methods throughout the index development the would like to thank Dr. Armel Nonvidé for comments one earlier drafts. 
process can help define mixed weighting scheme that will provide We are grateful to AGRA for providing financial support for Data 
additional information for policy formulation if weighted product ag- collection. 
gregation method is used. Though, several combinations could be 
defined, association of expert judgment methods with others methods Appendix A. Supplementary data 
should be prioritised since it can provide venues to take into account the 
perceived importance of indicators and component. Supplementary data to this article can be found online at https://doi. 
org/10.1016/j.jenvman.2020.111275. 
5. Conclusion 
References 
The need to quantify adaptive capacity to climate change at various 
levels, to help define vulnerability and inform policy makers, has led Abdul-razak, M., Kruse, S., 2017. The adaptive capacity of smallholder farmers to climate 
scholars to use composite indices which requires the use of weighting change in the Northern Region of Ghana. Clim. Risk Manag. https://doi.org/ 
scheme for aggregation. Various methods (equal weights, data driven 10.1016/j.crm.2017.06.001. 
Adger, W.N., Agrawala, S., Mirza, M.M.Q., Conde, C., O’Brien, K., Pulhin, J., 
weights and experts’ judgement) have been proposed to define Pulwarty, R., Smit, B., Takahashi, K., 2007. Assessment of adaptation practices, 
weighting scheme for adaptive capacity assessment. However, those options, constraints and capacity. In: Parry, M.L., Canziani, O.F., Palutikof, J.P., Van 
weights definitions methods when applied at household or individual Der Linden, P.J., Hanson, C.E. (Eds.), Climate Change 2007: Impact, Adaptation and 
Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of 
level to assess adaptive capacity to climate change provide weighting the Intergovernmental Panel on Climate Change. Cambridge University Press, 
scheme that do not reflect perceived importance of components to the Cambridge, UK, pp. 717–743. 
7
Y.C. Zanmassou et al.                                                                                                                                                                 J  o  u r  n a  l  o  f  E  n  v i r o  n  m   e n  t a  l  M   a n  a  g e  m   e n  t  276 (2020) 111275
Adger, W.N., Vincent, K., 2005. Uncertainty in adaptive capacity. Compt. Rendus Geosci. INSAE, 2012. Enqête modulaire integrée sur les conditions de vies des ménages. 2ème 
337, 399–410. https://doi.org/10.1016/j.crte.2004.11.004. édition (EMICoV, 2011). Institue Nationale de le Statistique et de l’Analyse 
Asante, F.A., Boakye, A.A., Egyir, I.S., Jatoe, J.B.D., 2012. Climate change and farmers’ Economique, Cotonou, Bénin.  
adaptive capacity to strategic innovations : the case of northern Ghana. Int. J. Dev. Jacob, B, Nelson, R, Kuruppu, N, Leith, P, 2015. An Adaptive Capacity Guide Book: 
Sustain. 1, 766–784. Assessing, building an evaluating the capacity of community to adapt in changing climate, 
Below, T., Mutabazi, K., Kirschke, D., Franke, C., Sieber, S., Siebert, R., Tscherning, K., Southern Slopes Climate change Adaptation Research Partnership (SCARP). 
2012. Can farmers’ adaptation to climate change be explained by socio-economic University of Technology Sydney and University of Tasmania, Hobart, Tasmania, 
household-level variables? Global Environ. Change 22, 223–235. https://doi.org/ pp. 1–27. 
10.1016/j.gloenvcha.2011.11.012. Juhola, S., Kruse, S., 2013. A framework for analysing regional adaptive capacity 
Böhringer, C., Jochem, P.E.P., 2007. Measuring the immeasurable-A survey of assessments : challenges for methodology and policy making. Mitig. Adapt. 
sustainability indices. Ecol. Econ. 63, 1–8. https://doi.org/10.1016/j. Strategies Glob. Change. https://doi.org/10.1007/s11027-013-9481-z. 
ecolecon.2007.03.008. Keys, N., Thomsen, D.C., Smith, T.F., 2016. Adaptive capacity and climate change : the 
Boko, M., Kosmowski, F., Vissin, E., 2012. Les Enjeux du changement climatique au role of community opinion leaders leaders. Local Environ. 24, 432–450. https://doi. 
Benin. Programme pour le Dialogue Politique en Afrique de l’Ouest. Konrad org/10.1080/13549839.2014.967758. 
Adenauer Stiftung. Lockwood, Michael, Raymond, Christopher M., Oczkowski, Eddie, Morrison, Mark, 2015. 
Brooks, N., Adger, W.N., Kelly, P.M., 2005. The determinants of vulnerability and Measuring the dimensions of adaptive capacity: a psychometric approach. Ecol. Soc. 
adaptive capacity at the national level and the implications for adaptation. Global 20 https://doi.org/10.5751/ES-07203-200137. 
Environ. Change 15, 151–163. https://doi.org/10.1016/j.gloenvcha.2004.12.006. MAEP, 2011. Plan Stratégique de Relance du Secteur Agricole (PSRSA)-Benin. Ministère 
Bryan, B.A., Huai, J., Connor, J., Gao, L., King, D., Kandulu, J., Zhao, G., 2015. What de l’Agriculture, de l’Elevage et de la Pêche (MAEP). Cotonou, Benin.  
actually confers adaptive capacity? insights from agro-climatic vulnerability of Marzi, S., Mysiak, J., Santato, S., 2018. Comparing adaptive capacity index across scales : 
australian wheat. PloS One 10, 1–20. https://doi.org/10.1371/journal. the case of Italy. J. Environ. Manag. 223, 1023–1036. https://doi.org/10.1016/j. 
pone.0117600. jenvman.2018.06.060. 
Burgass, M.J., Halpern, B.S., Nicholson, E., Milner-gulland, E.J., 2017. Navigating MEHU, 2011. Deuxieme communication nationale de la Republique du Benin sur les 
uncertainty in environmental composite indicators. Ecol. Indicat. 75, 268–278. changements climatiques. Ministère de l’Ennvironnement, de l’Habitat et de 
https://doi.org/10.1016/j.ecolind.2016.12.034. l’Urbanisme. Direction Génerale de l’Environnement, Cotonou, Benin.  
Clay, N., King, B., 2019. Smallholders’ uneven capacities to adapt to climate change amid Mekonnen, Z., Kassa, H., 2019. Living with climate change: assessment of the adaptive 
Africa’s ‘green revolution ’: case study of Rwanda’s crop intensification program. capacities of smallholders in central rift valley, Ethiopia. Am. J. Clim. Change 
World Dev. 116, 1–14. https://doi.org/10.1016/j.worlddev.2018.11.022. 205–227. https://doi.org/10.4236/ajcc.2019.82012, 08.  
Dannevig, Halvor, Bay-Larsen, Ingrid, Van Oort, Bob, Keskitalo, E. Carina H., 2015. MEPN, 2007. P. rogramme d’Action National d’Adaptation aux Changements 
Adaptive capacity to changes in terrestrial ecosystem services amongst primary Climatiques du Benin (PANA-BENIN). Ministère de l’Environnement et de Protection 
small-scale resource users in northern Norway and Sweden. Pol. Geogr. 38, 271288. de la Nature(MEPN). Cotonou, Benin.  
https://doi.org/10.1080/1088937X.2015.1114533. Monterroso, A., Conde, C., Monterroso, A., 2018. Adaptive capacity : identifying the 
Defiesta, G., Rapera, C.L., 2014. Measuring adaptive capacity of farmers to climate challenges faced by municipalities addressing climate change in Mexico. Clim. Dev. 
change and variability: application of a composite index to an agricultural 10, 729–741. https://doi.org/10.1080/17565529.2017.1372264. 
community in the Philippines. J. Environ. Sci. Manag. 17, 48–62. Nardo, M., Saisana, M., Saltelli, A., Tarantola, S., 2005. Tools for composit indecators 
Egyir, I.S., Ofori, K., Antwi, G., Ntiamoa-baidu, Y., 2015. Adaptive capacity and coping building. Joint Research Center. European Commision, Ispra, Italy.  
strategies in the face of climate Change : a comparative study of communities around Pam, 2009. Analyse Globale de la Vulnérabilité de la Sécurité Alimentaire et de la 
two protected areas in the coastal savanna and transitional zones of Ghana. Nutrition (AGVSAN): Benin. Rome, Italy.  
J. Sustain. Dev. 8, 1–15. https://doi.org/10.5539/jsd.v8n1p1. Ruiz Meza, L.E., 2015. Adaptive capacity of small-scale coffee farmers to climate change 
Ellis, F., 2000. The determinants of rural livelihood diversification in developing impacts in the Soconusco region of. Clim. Dev. 7, 100–109. https://doi.org/ 
countries. J. Agric. Econ. 51, 289–302. https://doi.org/10.1111/j.1477-9552.2000. 10.1080/17565529.2014.900472. 
tb01229.x. Scoones, I., 1998. Sustainable rural livelihoods: a framework for analysis (No. 72), ids 
Engle, N.L., 2011. Adaptive capacity and its assessment. Global Environ. Change 21, working paper. Brighton. https://doi.org/10.1057/palgrave.development.1110037. 
647–656. https://doi.org/10.1016/j.gloenvcha.2011.01.019. Shah, K.U., Dulal, B.H., 2015. Household capacity to adapt to climate change and 
Engle, N.L., Lemos, M.C., 2010. Unpacking governance : building adaptive capacity to implications for food security in Trinidad and Tobago. Reg. Environ. Change. 
climate change of river basins in Brazil. Global Environ. Change 20, 4–13. https:// https://doi.org/10.1007/s10113-015-0830-1. 
doi.org/10.1016/j.gloenvcha.2009.07.001. Shirima, A.O., Mahonge, C., Chingonikaya, E., 2017. Smallholder farmers ’ levels of 
Gallopín, G.C., 2006. Linkages between vulnerability, resilience, and adaptive capacity. adaptive capacity to climate change and variability in manyoni district , Tanzania. 
Global Environ. Change 16, 293–303. https://doi.org/10.1016/j. Int. J. Redearch Methodol. Soocial Sci. 2, 19–29. 
gloenvcha.2006.02.004. Smit, B., Wandel, J., 2006. Adaptation, adaptive capacity and vulnerability. Global 
Gan, X., Fernandez, I.C., Guo, J., Wilson, M., Zhao, Y., Zhou, B., Wu, J., 2017. When to Environ. Change 16, 282–292. https://doi.org/10.1016/j.gloenvcha.2006.03.008. 
use what : methods for weighting and aggregating sustainability indicators. Ecol. Vodounou, J.B.K., Doubogan, Y.O., 2016. Agriculture paysanne et stratégies 
Indicat. 81, 491–502. https://doi.org/10.1016/j.ecolind.2017.05.068. d’adaptation au changement climatique au Nord-Bénin. CybergoEuropean J. Geogr. 
Gbetibouo, Glwadys, Ringler, Claudia, Hassan, Rashid, 2010. Vulnerability of the South Evironement , Nat. Paysage 0– 27 https://doi.org/10.4000/cybergeo.27836. 
African farming sector to climate change and variability: An indicator approach. Nat. Wang, T., Fu, Y., 2019. Constructing composite indicators with individual judgements 
Resour. Forum 34, 175–187. https://doi.org/10.1111/j.1477-8947.2010.01302.x. and best-worst method: an illustration of value measure. Soc. Indicat. Res. 14 
Gnanglè, C.P., Kakaï, R.G., Assogbadjo, A.E., Vodounnon, S., Yabi, J.A., Sokpon, N., https://doi.org/10.1007/s11205-019-02236-3. 
2011. Tendances climatiques passées , modélisation et adaptations locales au Bénin. Williges, K., Mechler, R., Bowyer, P., Balkovic, J., 2017. Towards an assessment of 
Climatologie 8, 27–40. adaptive capacity of the European agricultural sector to droughts. Clim. Serv. 7, 
Greco, S., Ishizaka, A., Tasiou, M., Torrisi, G., 2019. On the methodological framework of 47–63. https://doi.org/10.1016/j.cliser.2016.10.003. 
composite Indices : a review of the issues of weighting , aggregation , and robustness. Yabi, I., Afouda, F., 2012. Extreme rainfall years in Benin (west Africa). Quat. Int. 262, 
Soc. Indicat. Res. 141, 61–94. https://doi.org/10.1007/s11205-017-1832-9. 39–43. https://doi.org/10.1016/j.quaint.2010.12.010. 
Gupta, J., Termeer, C., Klostermann, J., Meijerink, S., Brink, M. Van Den, Jong, P., Yegbemey, R.N., Yabi, J.A., Tovignan, S.D., Gantoli, G., Kokoye, S.E.H., 2013. Farmers’ 
Nooteboom, S., Bergsma, E., 2010. The Adaptive Capacity Wheel : a method to assess decisions to adapt to climate change under various property rights: a case study of 
the inherent characteristics of institutions to enable the adaptive capacity of society. maize farming in northern Benin (West Africa). Land Use Pol. 34, 168–175. https:// 
Environ. Sci. Pol. 13, 459–471. https://doi.org/10.1016/j.envsci.2010.05.006. doi.org/10.1016/j.landusepol.2013.03.001. 
Hinkel, J., 2011. Indicators of vulnerability and adaptive capacity”: towards a Yohe, G., Tol, R.S.J., 2002. Indicators for social and economic coping capacity - moving 
clarification of the science-policy interface. Global Environ. Change 21, 198–208. toward a working definition of adaptive capacity. Global Environ. Change 12, 25–40. 
https://doi.org/10.1016/j.gloenvcha.2010.08.002. https://doi.org/10.1016/S0959-3780(01)00026-7. 
Hogan, A., Bode, A., Berry, H., 2011. Farmer health and adaptive capacity in the face of Zhou, P., Ang, B.W., Poh, K.L., 2006. Comparing aggregating methods for constructing 
climate change and variability. part 2: contexts, personal attributes and behaviors. the composite environmental index : an objective measure. Ecol. Econ. 59, 305–311. 
Int. J. Environ. Res. Publ. Health 8, 4055–4068. https://doi.org/10.3390/ Zhou, P., Ang, B.W., Zhou, D.Q., 2010. Weighting and aggregation in composite indicator 
ijerph8104055. Construction : a multiplicative optimization approach. Soc. Indicat. Res. 96, 
169–181. 
8