Drought is one of the most damaging climate–induced threats impacting the lives of many people every year. Proactive and reactive drought coping and adapting mechanisms enable farmers to be resilient against climate induced drought and improve the current disaster preparedness and early warning systems in the drylands. The aim of this study was to assess proactive and reactive farmers’ drought adaptation & coping strategies at household level in Raya valley in southern Tigray, Ethiopia. Agro–climatological based 246 households were sampled from the lowlands, midlands and highlands. Multinomial logit model was used to identify best drought coping strategies. Respondents had mild to extremely severe drought in the last three decades. The association between the various drought severity and household heads were significant (chi2= 9.861, df=3, p < 0.05). Pasture collection and saving, soil and water conservation, and weather prediction information to adjust saving and farming system were proactive drought adaptation and coping strategies. Livestock Feeding with roasted cactus cladode, small business loans, livestock selling, productive safety–net programme, and food consumption reduction were the major reactive drought adaptation and coping strategies. The proactive and reactive drought coping and adaptation strategies identified in this study should be used to improve the current disaster preparedness and early warning systems monitoring and evaluation in the face of climate and weather extreme related impacts of climate variability and change.
Research Article
Analysis of Drought Coping Strategies in Northern Ethiopian Highlands
https://doi.org/10.21203/rs.3.rs-2741157/v1
This work is licensed under a CC BY 4.0 License
You are reading this latest preprint version
climate variability
disaster preparedness
drought adaptation
early warning
multinomial logit model
Tigray
- Proactive drought adaptation strategies were pasture management, land conservation and weather forecast information
- Reactive drought adaptation strategies were livestock feed and selling, loans, safety–net, and food consumption reduction
- Climate adaptation strategies determine disaster preparedness and early warning systems monitoring and evaluation
Drought is one of the most devastating climate induced risks that affect the lives of many people annually [1–3]. Drought is originated primarily from a precipitation deficiency and shortage of soil moisture due to the rising of surface and air temperature. Droughts occurred across all parts of the world during the periods of 2001–2010 [3] caused decline in agricultural production and productivity [1]. The response to drought has been poorly managed, coordinated and ineffective [3]. There have been also no rigorous attempts at the global level to initiate dialogue on formulating and implementing national drought policies that include a basis for proactive, risk–based management to resolve drought impacts [3]. As a result, the effects of drought in Africa are getting worst because most of the farmers are relying on rain–fed agriculture and limited capacity to cope–up with the impacts due to high levels of poverty, lack of skilled human resources and technologies. It is also highly anticipated that agricultural production and productivity of the continent to be more affected in the near future due to shortage of rainfall. In this paper, the term coping strategies referred to the short & medium–term responses to the adverse impacts of drought. The effect may differ in each agro–climatic zone [4]. At present, several parts of the sub–Saharan Africa have faced with a 10–40% chance of failed seasons during the main cropping period [5]. For example, drought has severely affected maize crop which is one of the continent's most important crops [6] and deteriorates a significant number of livestock. While agriculture has been the first and most affected sector, many other sectors have also experienced substantial decreases [3]. Agriculture in Africa is highly sensitive to drought due to increasing climate variability and change [5, 7–8]. Shiferaw et al. [5] reported that the regions of Eastern and Southern Africa are primarily characterized by semi–arid and sub–humid climates with an extreme dry season. For example, the prolonged dry season has resulted in wide spread crop failure between 2013 and 2015 across Namibia [3] and Ethiopia. Twongyirwe et al. [9] reported that drought induced food insecurity is a critical problem affecting about 6.3% of the households in Uganda. Speranza [7] stated that there are also extensive drought impacts not only on crops, but also on livestock in Kenya and 80 per cent of the area is arid and semi–arid, with pastoralism and agro–pastoralism being the dominant rural livelihoods. For decades, Tigray has been one of the most seriously affected regions in Ethiopia by a prolonged drought due to climate change and variability, environmental degradation and other related factors. The changes in the climate have brought direct impacts on livelihood assets, health, food, and water security [10]. The production and productivity of major crops such as teff, barley, maize, wheat, sorghum and others yields have truncated. The economic and social growth of the country has also been greatly affected by this tragedy. Approximately 98% of farmland relies on rainfall and this has an immense effect on both agriculture and households food security [11].
Tesfamariam et al. [12] reported that during the 1984–1985 droughts, which are recorded as one of the most devastating drought periods in Ethiopia, the national GDP was dropped by 9.7%, while agricultural output was further declined by 21%. The Tigray National Regional State in northern Ethiopia has been one of the most vulnerable regions to drought. Gidey et al. [13] reported that the arid and semi–arid area of the region has been severely affected by the recurrent droughts due to late rainfall onset, early cessations, erratic and high dry spells. In the event of drought, there will be a possibility of losses of 21–100% of crop yield during the main crop cycle in the region in general and the study area in particular. There may be minimal capacity for smallholder farmers to cope–up with the current climate fluctuations and potential changes. The impacts of drought seem to be increasing and it is a good indication of unsustainable resource use [3]. The lack of adequate rainfall or soil moisture affect crop production and productivity as a result of drought impacts would be severe in the area because of the dependency on rainfall. Drought does not only affect crop production, but also livestock production and their products due to lack of fodder, water and diseases [14]. Speranza [7] reported that production of milk decreased during the drought, putting an end to the market for livestock products in Kenya. Kassie [15] pointed out that loss of pasture and depletion of water sources, loss of vegetation cover, loss of livestock production, and resources depletion are good examples of drought impacts.
Drought stress can be characterized by the reduction of moisture level. During the acute drought period, various adaptation and coping strategies have been adopted by farmers. Reducing food sales, reducing food intake, removing children from school and liquidating productive assets such as livestock, land, trees are some of the drought coping and adaptation strategies among others [5]. There is no single solution to cope–up with drought as it takes an integrated strategy composed of several practices for exchanging knowledge, designing response strategies and implementing solutions [16]. The ability to cope with drought varies widely from one country to another and from one area, community or population group to another [3]. Menghistu et al. [17] and Twongyirwe et al. [9] pointed out that there is limited study on the socio–economic impacts of recurrent drought and coping strategies. Selling of wood and charcoal, involving in small business activities, reduction of food consumption, livestock sales, and migration were some coping strategies used by farmers [18]. The coping strategies vary from place to place based on the community's exposure to drought risk and socio–economic context. Hadgu et al. [19] have used multinomial logit model to comprehend the best adaptation strategies and the findings indicated that farming in the midlands declines irrigation by 23% as compared to lowland. Irrigation reduces the risk of crop failure during periods of prolonged drought [9]. Selection of an improved crop varies may increase crop production and productivities and also diminish farmers’ susceptibility to drought risks. In addition, household with poor resource endowments would face a higher risk of being food insecure [9]. Impact of drought could be reduced through involving the smallholder farmers in a wide range of on–farm and off–farm practices and/or drought coping strategies and mitigation programs [1]. Thus, it is highly important to analyze on how the smallholder farmers have been responded to drought and its implication. The aim of this paper was to analyze the proactive and reactive farmers’ drought adaptation and coping strategy at household level in the drylands of Tigray, Ethiopia.
2.1 Study area
The study area is located in southern Tigray, Ethiopia. It constitutes both Raya Chercher and Raya Azebo districts. It is located at 39°20′0″ and 40°0′0″ longitude east and 12°20′0″ and 12°50′0″ latitude north (Fig. 1). The total land mass of the area is 1777 km2 where 58% is in Raya Azebo district and 42% is in Raya Chercher district. The area receives an annual average rainfall of 558 mm [1]. The mean maximum and minimum temperatures were about 30.5 and 15.9°C, respectively [1]. The altitude was between 913 to 3110 m above sea level. Both crop production and livestock rearing were the major livelihoods and main source of income of the community. Cultivated land 1061 Km2 (59.7%) shrub/bush land 519 Km2 (29.2%) and barren lands or exposed rock surface 197 Km2 (11.1%) were the major land cover types. The population of the area was 192,287 people. The male population was roughly 93,948 (48.9%), while the female population was 98,338 (51.1%). The maximum, minimum and average family size of the study area was 13, 2, and 6.2, respectively.
2.2 Sampling procedure
A multistage sampling technique was applied to determine an appropriate sample size. Based on the Ethiopian Ministry of Agriculture and Rural Development agro–climatic zonation, the study area was stratified in to Kolla or lowland (500–1500 m.a.s.l), Weyna–Dega or midland (1500–2300 m.a.s.l) and Dega or highland (2300–3200 m.a.s.l) (Table 1). A cluster sampling technique based on the homogeneity or agro–climatology similarity of the area was employed to determine the number of households residing in each similar agro–climatology zones and then the sample size was estimated based on [20] as follows (Eq. 1):
$$n=\frac{{{z^2}\left( p \right)\left( q \right)}}{{\left( {{d^2}} \right)}}$$
1
Where n = desired sample size, z = confidence level, 95per cent, p = proportion of the households to be included in the sample (20per cent), q = 1–0.2 i.e. 0.80, d = acceptable error 5 per cent.
The equation was used to sample 20 per cent of households. A systematic random sampling method was then applied to proportionally sample after the total number of households residing in each of the similar agro–climatic zones were multiplied by the required sample size (i.e., 246) divided into the total number of households of the study area (i.e., 41,803) (Table 1).
|
Agro–climatological zone classification (ACZ) |
Total Household (HH) |
Collected sample size (n = 246) |
Total area coverage |
||||
|---|---|---|---|---|---|---|---|
|
Values (m.a.s.l) |
ACZ |
HH |
% |
N |
% |
Km2 |
% |
|
500–1500 |
Lowlands (Kolla) |
6109 |
14.6 |
36 |
15 |
454.5 |
25.6 |
|
1500–2300 |
Midlands (Weyna–Dega) |
34,324 |
82.1 |
202 |
82 |
1297.2 |
73 |
|
2300–3200 |
Highlands (Dega) |
1370 |
3.3 |
8 |
3 |
24.8 |
1.4 |
|
Total |
41,803 |
100 |
246 |
100 |
1,777 |
100 |
|
2.3 Data collection
The socio–economic data were collected from 246 households. Each household took an average of one and half hours to respond to the semi–structured questions. Focus group discussion consisting of eight team members per group were conducted to comprehend how the farmers were responding to drought, its possible causes, coping strategies and impacts on the natural resources, livelihood and livestock. A total of eleven group members have been formulated to discuss openly. Field observation was undertaken by the researchers to observe and gather information pertaining to the biophysical and socioeconomic conditions of the study area. Moreover, earth observation (elevation, land cover, vegetation) at moderate spatial resolution have been obtained from Water and Land Resources Information System (WALRIS) archive at www.wlrc–eth.org the National Geospatial Database System, Ethiopia.
2.4 Data analysis
The socioeconomic datasets were analyzed using Stata SE version 13. A multinomial (polytomous) logit model (MNL) was applied to evaluate the best proactive and reactive drought adaptation and coping strategies. The MNL model is the best model to predict discrete choices because of the computational drawbacks of the multinomial probit (MNP) model [18, 21]. This model has been used in several studies to select best coping practice or choices for climate change adaptation [19]. This study employed a multinomial logit (MNL) model to examine the various drought coping strategies that influences the smallholder farmers. To describe the MNL model, assume Y represents the response random variable taking on the values {1, 2.., j} for choices j, a positive integer, and let X denote a set of explanatory variables. In this case, Y representing the coping strategies measure chosen by the household heads. Here, we assumed that each farmer household selected a set of categorical discrete, disjoint choices of drought coping strategies and these strategies are assumed to depend on one or more factors of X (Gender, Age, Family size, farm size). The question shows how other variables changes in the factors of X that affects the response probabilities,\(P\left(Y=j|X=x\right)\), for j=1, …., J. Since the probabilities must sum to unity,\(P\left(Y=j|X=x\right)\), is determined once we know the probabilities for j = 1,2,…., J. Let X is a 1xK vector with first element unity, the MNL model has response probabilities given by (Eq. 2):
$${P}\left({Y}={j}|{X}={x}\right)=\frac{{{e}}^{{x}{{\beta }}_{{j}}}}{1+\sum _{{k}=1}^{{J}}{{e}}^{{x}{{\beta }}_{{k}}}}, {j}=1,...,{J} \left(2\right)$$
Where \({\beta }_{j}\) is K x 1 vector of coefficients, j = 1, ….., J
The response variables Y’s for the on–farm drought coping strategies are: (I) Selection of drought tolerant crops (II) Irrigation Construction of flood (III) Land preparation and use of compost (IV) Soil and Water Conservation and (V) Use of weather predictions information. In addition the response variables for the off–farm drought coping strategies are: (I) Looking for government support (II) Migration both animal and human (III) Borrow loans from microfinance (IV) Selling of household assets (V) Daily Laborer (VI) Reduction of food consumption and (VII) feeding of roasted cactuses. The unbiased and consistent parameter estimate of the model Eq. (2) requires the assumptions of Independence of Irrelevant Alternatives (IIA) to hold. In addition, the IIA assumption requires that the probability of using a certain coping strategies (Collection and saving of pastures) selected by a given household head needs to be independent from the probability of choosing another coping strategies (i.e., \({\text{P}}_{\text{j}}\!\left/ \!{\text{P}}_{\text{k}}\right.\)is independent of the remaining probabilities). The parameter estimates of the MNL model provide only the direction of the effect of the explanatory factor effects variables on the response variable, but estimates do not represent either the actual magnitude of change nor probabilities (Greene, 2003). To interpret the effects of explanatory variables on the probabilities, the marginal effect has been computed. Differentiating equation–2 partially with respect to the explanatory variables provides marginal effects of the explanatory variables as shown below (Eq. 3):
$$\frac{\partial {p}_{j}}{\partial {x}_{k}}={P}_{j}\left[{\beta }_{jk}-{\sum }_{j=1}^{j-1}{P}_{j}{\beta }_{jk}\right] \left(3\right)$$
The marginal effects or marginal probabilities are functions of the probability itself and measure the expected change in probability of a particular choice being made with respect to a unit change in an independent variable from their average e.g., proactive and reactive coping strategies or on–off response). The explanatory variables of the study were households socio–economic, environmental, and climatic and others, as hypothesized to have associations with the households’ farmers coping strategies. Based on theory, empirical literature, and researcher’s best knowledge of the contextual setting, nineteen explanatory variables (Table 2) were used to determine the best coping strategies to drought both for on and off–response farm activities. Following the model building, Variance Inflation Factor (VIF) and Contingency Coefficient (CC) were employed to detect multicollinearity for continuous variables and discrete variables, respectively. Breusch–Pagan test was conducted to assess the presence of heteroscedasticity or constancy of variance in the model. Finally, the model was tested for the validity of the Independence of Irrelevant Alternatives (IIA) assumptions by using Hausman’s test. This test failed to reject the null hypothesis of independence of the drought coping strategies, suggesting that the multinomial logit (MNL) specification is appropriate to model drought coping strategies of smallholder farmers in the study area. MNL model specification was used by several researchers to model drought coping strategies of smallholder farmers in Africa [23–24]. The response variables in the empirical estimation were coping strategies that were chosen by the household farmers during the field survey (Table 2). In this case, various best coping strategies were selected to examine farmers’ response to drought incidence, coping and impacts on natural resources (vegetation cover status), agricultural productions (crop and livestock)
|
S.no |
Selected variables |
Category |
Expected Sign for on–farm |
Expected Sign for off–farm |
|---|---|---|---|---|
|
1 |
Gender (Male and Female) |
Dummy |
+ |
+ |
|
2 |
Age (household head) |
Continuous |
+ |
+ |
|
4 |
Educational status (literate and illiterate) |
Categorical |
+ |
+ |
|
5 |
Family size |
Continuous |
+ |
+ |
|
6 |
Farm size |
Continuous |
+ |
– |
|
7 |
Harvest times per annual |
Categorical |
+ |
– |
|
8 |
Farming system use (summer, spring & irrigation) |
Categorical |
+ |
– |
|
9 |
Production status (increased, decreased) |
Categorical |
+ |
+ |
|
10 |
Drought frequency (every year, once in every two years, once in every three years, once every four years) |
Categorical |
+ |
+ |
|
11 |
Drought severity (mild, moderate, severe, extremely severe) |
Categorical |
+ |
+ |
|
12 |
Drought causes (perceived by the household head such as lack of rainfall, high temperature, God punishment ) |
Categorical |
+ |
+ |
|
13 |
On–farm drought response (coping) activities |
Categorical |
+ |
– |
|
14 |
Off–farm drought response (coping) activities |
Categorical |
– |
+ |
|
15 |
Effects of drought on natural resources |
Categorical |
+ |
+ |
|
16 |
Effects of drought on livestock |
Categorical |
+ |
+ |
|
17 |
Effects of drought on crop production |
Categorical |
+ |
+ |
|
18 |
Access to credit |
Categorical |
+ |
+ |
|
19 |
Access to early warning information |
Categorical |
+ |
+ |
3.1 Households’ characteristics
The age of household head is essential variable that can influence selection of drought coping strategy. The respondent’s age ranged from 31 to 77 with an average age of 47. The results indicates that about 36% of the male headed households were in their middle age adults group, 9% were old adults of female headed households and 6% of the respondents were elderly. The age of household head is a proxy for farming experience, on the assumption that the household's knowledge of drought coping strategies and food security issues increase as the household head grows older and become more experienced [9].
3.2 Educational status at household level
Literacy is the most important factor or instrument for understanding facts (e.g., about drought) and changing our lifestyle and/or strengthening our resilience, coping and mitigation to various climate hazards. It is explained in terms of contribution on working efficiency, competency, income, adopting technologies and becoming visionary in creating conducive environment to educate dependents with long term target to ensure better living condition than illiterate ones [9]. Studies indicated that the level of literacy could influence selection of drought coping strategies due to the behaviors of respondents. In the study area, about 61.38% of the respondents are illiterate, while 38.62% of the respondents are literate. An illiterate male headed household respondent which covers about 68.8% are higher than female headed illiterate (31.4%) households (Table 3). Similarly, the literate male headed households covers only 88.4% and 11.6% female headed. Therefore, literate households can help to generate new ideas or methods towards understanding the causes of drought, and improving the coping strategies. Besides, literate households can reduce the chance of becoming food insecure [9].
|
` Household head |
Literacy status |
Total |
chi–square 1.82 |
P–value 0.401 |
||||
|
Illiterate |
Literate |
|||||||
|
1–4 |
5–8 |
9–12 |
Others |
|||||
|
Female |
48 (31.4%) |
6 (12.8%) |
2(5.4%) |
3(33.3%) |
0.0 |
11(11.6%) |
||
|
Male |
103 (68.2%) |
41(87.2%) |
35(94.6%) |
6(66.7%) |
2(100%) |
84(88.4%) |
||
|
Total |
151 (100%) |
47(100%) |
37(100%) |
9(100%) |
2(100%) |
246(100%) |
||
3.1.3 Farm size at household level
Male headed households owned a maximum of 3.5 ha, while females owned 2.5 ha. Farm size is important proxy for determining farmers’ economic status and coping strategies [4]. The mean farm size owned by male headed household (0.93 ± 0.74ha) was higher than female (0.54 ± 0.75 ha) (Table 4). For example, in the study area, about 2.03 per cent of females and 6.5 per cent of male headed households were landless farmers. Farmers were renting plots for cultivation from other farmers who have serious health problems and economic challenges such as access to get or own oxen or labor to cultivate the land on time. Farmers who have their own farm have a high probability of investing in coping strategies compared to landless farmers [4]. Studies indicates that if the cultivated land size increases the possibility that the household gets more yield is also high during the wet season ([9]. In this case, the household may have better drought resilience mechanism. However, those households who live in the marginal areas and whose livelihoods are highly dependent on natural resources are more susceptible to drought because they have limited drought coping capacity [10].
|
Group |
Obs. |
Mean |
Std. Err. |
Std. Dev. |
[95% Conf. Interval] |
|
|---|---|---|---|---|---|---|
|
Female |
61 |
. 5491803 |
. 0569067 |
. 4444558 |
. 4353499 |
. 6630108 |
|
Male |
185 |
. 932973 |
. 0548457 |
. 7459815 |
. 8247658 |
1.04118 |
|
Combined |
246 |
. 8378049 |
. 0448072 |
. 7027741 |
. 7495483 |
. 9260614 |
|
diff |
–. 3837926 |
. 1010283 |
–0.5827915 |
–0.1847938 |
||
|
diff |
= mean (Female) – mean (Male) |
t = |
–3.7989 |
|||
|
Ho: diff |
0 |
degrees of freedom = |
244 |
|||
|
Ha: |
diff < 0 |
Ha: diff ! = 0 |
Ha: diff > 0 |
|||
|
Pr (T < t) |
= 0.0001 |
Pr (|T| > |t| = 0.0002 |
Pr (T > t) = 0.9999 |
|||
Furthermore, we compared the average farm size of male and female headed households for assessing the difference. Farm size was significantly different between the male and female headed households in the study area (p < 0.001). One of the possible reasons for the difference is that demographic composition of the respondents.
3.2 Drought status at household level
Understanding the status of drought at early stage and household level is imperative for formulating effective mitigation measures [2] and coping strategies. The mitigation measures help to reduce the adverse effects, especially for smallholder farmers who grow crops under unfavorable conditions, such as low and erratic precipitation and poor soils. In the study area, about 24.8% of female headed and 75.2% of male headed respondents had experienced mild to extremely severe drought in the last three decades (Table 5). The severity may vary depending on the households demand for water and other natural resources. The association between the various severity level in both of the household heads shows that there was significant association between the different drought severity levels and household heads (chi–square = 9.861, df = 3, p–value < 0.05). Tembo et al [25] reported that about 92% of the farmers in the south east of Tigray perceived that rainfall is significantly decreasing annually in the last three decades. This aggravates the incidence of drought and its impacts on natural resources, human and livestock.
|
Severity level |
Household type |
Total |
chi–square |
P–value |
|
|---|---|---|---|---|---|
|
Female headed |
Male headed |
||||
|
Mild |
4 (1.63%) |
8 (3.25%) |
12 (4.88%) |
9.8608 |
0.02 |
|
Moderate |
20 (8.13%) |
77 (31.30%) |
97 (39.03%) |
||
|
Severe |
31 (12.60%) |
97 (39.43%) |
128 (52.03%) |
||
|
Extremely severe |
6 (2.44%) |
3 (1.22%) |
9 (3.66%) |
||
|
Total |
61 (24.80%) |
185 (75.20%) |
246(100.00%) |
||
3.3 Smallholder farmers’ drought coping strategies
Since 1983/85, the recurrent drought has regularly affected southern Tigray, particularly the natural resources (e.g., vegetation cover) and agricultural production and productivity (e.g., maize, teff) have declined both in Raya Azebo and Raya Chercher districts. This affects the education, health, and livelihoods that lead to food insecurity and poverty. The major triggering factors were the lack of optimum rainfall during the belg or short rainy season which usually lasts from March–June and summer or the main rain season (July–early September). Macon et al. [16] also reported that California has been severely affected by droughts due to high temperatures and dry summer seasons that have significantly affected the agricultural sector (e.g., seasonal decline in the quality of dryland forage). The change in rainfall patterns and wind erosion, increased salinization and decreased carbon mineralization as a result it aggravates the incidence of drought impacts [26]. Numerous attempts have been made to establish coping strategies to overcome drought stress in the study area. Drought was one of the most important challenges to ensure food security in the study area. Both the proactive approach and the reactive approach were essential components of coping strategies [16]. The smallholder farmers have identified multiple strategies to cope–up with the impact of drought. The selected strategies depend on the household heads farm size, family size, education, age, awareness and severity of drought. The gender of respondents can influence the decision to select best drought coping strategies because they have different roles [4, 27].
Table 6 shows best on–farm drought coping strategies estimated using the MNL model that the smallholder farmers of the study area frequently used to cope–up with drought. Results of the multinomial logit (MNL) model indicates that age of the household head, which represents experience, affecting the coping to drought positively and significantly in strategies number II and V (p–value < 5%). The relative risk ratio, which is the exponent of the coefficient of the estimate (\({\text{e}}^{{\beta }}\)) shows that a change in the age of the household head result in a 3.5 and a 2.4 higher probability of choosing strategy number II and V as compared to the base choice strategy (i.e., land preparation and use of compost (manure). The marginal effect of the age estimate shows that the probability of choosing strategy number II and V increased by 0.04% and decreased by 0.17%, respectively, as compared to choosing the base choice strategy. Likewise, households family size, which usually represents the size of the family living within the family affecting the coping to drought positively and significantly both strategies numbers IV and V (p–value < 10%). The relative risk estimate of the family size estimate shows that a unit change in the family size of the household result in a 4 and 6.4 higher likelihood of choosing strategy number IV and V, respectively, as compared to the choice of the base outcome. The average marginal effect of the family size estimate also shows that the likelihood of choosing strategy number IV and V decreased by 0.06% and increased by 0.04%, respectively, as compared to the base outcome. Besides, farm size of the household which represents the share of cultivated land measured affecting the coping to drought positively and significantly the strategy number II (p–value < 10%). The relative risk estimate shows that a unit change in farm size of the household in choosing coping strategy II produces to have the same effect on coping to drought as compared to the base outcome. The average marginal effect of the estimate of the farm size indicates that the probability of choosing strategy number II decreased by 6.63% as compared to the base outcome. Furthermore, soil and water conservation which represents the involvement of the farmers in preserving the cultivated land of the area under study affecting the coping to drought positively and significantly the choice of strategy II (p–value < 1%). The average marginal effect of the estimate shows that the probability of choosing strategy number II increases by 5.4% as compared to the base outcome. Moreover, the effects of drought on production, measured both in terms of complete loss, partial loss and pest and disease, the result indicates that complete loss is found to have significant effect in choosing the coping strategy number II (p–value < 5%). The marginal effect of this estimate also shows that the probability of choosing strategy number II increases by 18.1% as compared to the base outcome. Access to accurate weather information also reduces the impacts of drought on both humans and livestock as ranchers can prepare for potential emergencies and serve as a valuable source of knowledge for researchers, land managers and policymakers. This may support to improve the existing response capacity to drought in the form of preparedness strategies when drought strikes [7]. However, the farmers of the study area are not getting timely information about the future drought due to lack of scientific drought monitoring and early warning systems. Now–a–days, high–quality information sources (e.g., weather forecast information) and peer–to–peer knowledge sharing on incoming drought risk improves the ability to respond to drought [3, 16]. Therefore, the community of the study area should get timely information to strengthen their coping strategies and reduce the impact felt from drought.
Table 6. Best proactive drought coping strategies and on–farm response in the study area
In addition, Table 7 shows that the MNL model estimates for the reactive drought coping strategy activities. The result reveals that male headed household, affects the choice of reactive drought coping positively and significantly for strategy number V (p–value < 10%). The relative risk ratio of the estimate indicates that a male headed household have a 3.5 times higher in favoring the choice for strategy number V as compared to the base choice strategy (i.e., migration to adjacent and remote areas). The marginal effect estimate of the male headed households shows that the probability of choosing strategy V increases by 0.77% as compared to choosing the base outcome strategy. Beside, age of the household head, affecting the reactive drought coping strategy number I (p–value < 5%) and its relative risk estimate indicates that the odds of choosing this strategy number I is same as compared to the base strategy. The marginal effect of the estimate of age shows that the likelihood of choosing strategy number I increased by 0.7% as compared to choosing base outcome strategy. Moreover, the family size of the household affecting the reactive drought strategy number III, positively and significantly (p–value < 10%) and strategy number V (p–value < 5%). The relative risk of the estimate indicates that a unit change in size of the family have a 1.3 times higher effects in choosing strategy number III and a 0.63 times less effect as compared to the base outcome. The marginal effect of the estimate for the family size tells us that the probability of choosing strategy number III is less likely to be selected as compared to the base outcome. This fact is true for all factors that have significant influence on strategy number III.
On the other hand, farm size affects the choice of strategy number I affecting the coping to drought positively and significantly (p–value < 10%) and strategy number IV (p–value < 1%). The relative risk estimate of the farm size shows that a unit change in farm size has a 0.45 and 0.31 times higher in selecting strategy number I and IV, respectively as compared to the base outcome. The marginal effect of the estimate shows that the probability of choosing strategy IV decreases by 9.7% and increases by 5.6% for strategy number I as compared to the base outcome. Moreover, the sub–category of belg and irrigation affects the choice of the strategy number II positively and significantly (p–value < 5%). The relative risk of the estimate indicates that a unit change of the farming system use of belg and irrigation by the farmers have a 16.6 times higher in choosing strategy I as compared to the base outcome. The marginal effect of the estimate shows that the probability of adopting strategy II decreases by 3.6% as compared to the base outcome. Furthermore, the effect of drought on crop production of complete loss affects the farmers choice of strategies I and V (p–value < 5%). The relative risk of the estimate indicates that a change in the estimate changes the choice of strategy I and V by 5.5% and 5%, respectively as compared to the base outcome. The average marginal effect of the model, it is found out that in the present study the likelihood of choosing strategy number I and strategy number V would decrease (increase) the probability of selecting strategy I by 9.8% as compared to the base outcome.
Similarly, partial loss of the effect of drought on crop production affects the choice of the strategy numbers I and V (p–value < 2%). The relative risk of the estimate shows that a change in the estimate changes the choice of the strategy I and V by 3.4% and 2%, respectively as compared to the base outcome. The average marginal effect reveals that the probability of adopting strategy I and V would decrease the probability of selecting these two strategies as compared to the bases outcome. In addition, the effect of the drought on livestock, result indicates that it affects positively and significantly for the choice of the strategy number I and V (p–value < 5%). The relative risk of the estimate of the factor changes the farmers’ choice in strategies number I and V by 10.23 and 6.43, respectively compared to the base outcome. The average marginal effect of the estimate reveals that the probability of making a choice for the strategy number I and IV would decrease (increase), respectively as compared to the base outcome. Finally, access to credit affects the selection of strategy number II positively and significantly (p–value < 10%). The relative risk of the estimate shows that a change in the estimate would result in a change by about 7% in selecting strategy II as compared to the base outcome. The average marginal effect result reveals that the probability of choosing strategy II increases by 0.1% as compared to the base outcome.
Table 7. Best reactive drought coping strategies and off–farm response in the study area
Crop failures caused by drought could affect labor supply [6]. Farmers in Raya Azebo and Raya Chercher districts were migrating during the drought episodes to save their lives and livestock. This attitude severely affected the agricultural sector and most of the households depend themselves on governments aid. Providing drought relief may decrease self–reliance and rise dependency on government and donor organizations [3], however, technological support may advance the farmers drought coping strategy and farming system. Therefore, the attitude of farmers must be changed to build self–resilience and improve their drought coping. Besides, there is a need of coordinated national drought policy that includes comprehensive monitoring, early warning and information systems, impact assessment procedures, risk management measures, drought preparedness plans, and emergency response programs to respond to drought effectively [3].
Similarly, Macon et al. [16] stated that the recurrent droughts pose a risk to health both economically and in rangeland. In the study area, a number of smallholder farmers sell their crop and buy livestock assets as a form of savings or insurances during the non–drought period. This practice may give relief to the households to cope–up with drought. However, during the periods of dry season and drought, most of the farmers reduce their food intake and sell their livestock at lower price due to lack of fodder and livestock diseases [28]. For instance, diarrhea and mumps, skin infections, trypanosomiasis, worms and parasites, coughs and lung infections were some of the major diseases observed in the study area. Mehar et al. [4] and Speranza [7] were also reported similar cases in Bihar, India and Kenya. This entails a substantial decrease in the existing production systems. It will also increase the livestock deaths, poor fertility and breeding, flea infestations and retarded growth [7]. Therefore, drought risk reduction policies and effective measures should be developed to diminish the impacts associated with droughts in the study area. Moreover, access to irrigation is an important strategy to reduce the vulnerability of agriculture to climate risks such as drought. Therefore, the farmers of Raya Azebo and Raya Chercher should have equitable access to irrigation to enhance their strategies for coping with drought.
This study evaluated farmers’ proactive and reactive drought coping strategies at household level. Collection and saving of pasture and hay, soil and water conservation practices, and use of weather prediction information were the main proactive drought coping strategies identified by farmers. In addition, feeding of roasted cactus for livestock, borrowing loans for running small business, selling of household assets such as livestock, food for work and reduction of food consumption both in quantity and quality were identified as reactive or off–farm drought coping strategies in the study area. This result may give good insights to improve the current drought preparedness and early warning systems monitoring and evaluation in the area. Besides, it helps to develop impact based drought mitigation measures to save the lives of local community and their livestock assets to improve the production and productivity of the area.
Acknowledgements
This research was supported by Mekelle University. We are also indebted to the support of development agents and community of the study area for their assistance during the field work activities of the research.
Funding
“The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.”
Competing interest
The authors have no relevant financial or non-financial interests to disclose
Author Contributions
“All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Eskinder Gidey, Paidamwoyo Mhangara, Tirhas Gebregergs,Woldegebriel Zeweld, Haftamu Gebretsadik, Oagile Dikinya, Said Mussa, Amanuel Zenebe, Atkilt Girma, Girmatsion Fisseha, Alemu Addisu, Jemal Nasir, Tesfaye Zeleke, Emiru Birhane. The first draft of the manuscript was written by Eskinder Gidey and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.”
Data availability statement
The authors have agreed to share any data used in the manuscript upon a request.
- Gidey, E., Dikinya, O., Sebego, R., Segosebe, E., & Zenebe, A. (2018a). Modeling the spatio–temporal meteorological drought characteristics using the standardized precipitation index (SPI) in raya and its environs, northern Ethiopia. Earth Systems and Environment, 2(2), 281–292.
- Velpuri, N. M., Senay, G. B., & Morisette, J. T. (2016). Evaluating new SMAP soil moisture for drought monitoring in the rangelands of the US high plains. Rangelands, 38(4), 183–190.
- Wilhite, D. A., Sivakumar, M. V., & Pulwarty, R. (2014). Managing drought risk in a changing climate: The role of national drought policy. Weather and Climate Extremes, 3, 4–13.
- Shiferaw, B., Tesfaye, K., Kassie, M., Abate, T., Prasanna, B. M., & Menkir, A. (2014). Managing vulnerability to drought and enhancing livelihood resilience in sub–Saharan Africa: Technological, institutional and policy options. Weather and Climate Extremes, 3, 67–79.
- Mehar, M., Mittal, S., & Prasad, N. (2016). Farmers coping strategies for climate shock: Is it differentiated by gender?. Journal of Rural Studies, 44, 123–131.
- Wossen, T., Abdoulaye, T., Alene, A., Feleke, S., Menkir, A., & Manyong, V. (2017). Measuring the impacts of adaptation strategies to drought stress: the case of drought tolerant maize varieties. Journal of Environmental Management, 203, 106–113.
- Speranza, C. I. (2010). Drought coping and adaptation strategies: understanding adaptations to climate change in agro–pastoral livestock production in Makueni district, Kenya. The European Journal of Development Research, 22(5), 623–642.
- Thompson, H. E., Berrang–Ford, L., & Ford, J. D. (2010). Climate change and food security in sub–Saharan Africa: a systematic literature review. Sustainability, 2(8), 2719–2733.
- Twongyirwe, R., Mfitumukiza, D., Barasa, B., Naggayi, B. R., Odongo, H., Nyakato, V., & Mutoni, G. (2019). Perceived effects of drought on household food security in South–western Uganda: Coping responses and determinants. Weather and Climate Extremes, 24, 100201.
- Akinnagbe, O. M., & Irohibe, I. J. (2014). Agricultural adaptation strategies to climate change impacts in Africa: a review. Bangladesh Journal of Agricultural Research, 39(3), 407–418.
- Aklilu, A., & Alebachew, A. (2008). Assessment of Climate Change–Induced Hazards. Impacts and Responses in the Southern Lowlands of Ethiopia.
- Tesfamariam, B. G., Gessesse, B., & Melgani, F. (2019). Characterizing the spatiotemporal distribution of meteorological drought as a response to climate variability: The case of rift valley lakes basin of Ethiopia. Weather and Climate Extremes, 26, 100237.
- Gidey, E., Dikinya, O., Sebego, R., Segosebe, E., & Zenebe, A. (2018b). Analysis of the long–term agricultural drought onset, cessation, duration, frequency, severity and spatial extent using vegetation health index (VHI) in Raya and its environs, northern Ethiopia. Environmental Systems Research, 7(1), 13.
- Webb, P., & Reardon, T. (1992). Drought impact and household response in East and West Africa. Quarterly journal of international agriculture, 31, 230–230.
- Kassie, B. T. (2014). Climate variability and change in Ethiopia: Exploring impacts and adaptation options for cereal production. Wageningen University.
- Macon, D. K., Barry, S., Becchetti, T., Davy, J. S., Doran, M. P., Finzel, J. A., … Lancaster, D. E. (2016). Coping with drought on California rangelands. Rangelands, 38(4), 222–228.
- Menghistu, H. T., Mersha, T. T., & Abraha, A. Z. (2018). Farmers’ perception of drought and its socioeconomic impact: the case of Tigray and Afar regions of Ethiopia. Journal of Applied Animal Research, 46(1), 1023–1031.
- Mohamed, A. A. (2017). Food security situation in Ethiopia: a review study. International Journal of Health Economics and Policy, 2(3), 86–96.
- Hadgu, G., Tesfaye, K., Mamo, G., & Kassa, B. (2015). Farmers climate change adaptation options and their determinants in Tigray Region, Northern Ethiopia. African Journal of Agricultural Research, 10(9), 956–964.
- Cochran, W.G. Sampling Techniques, 2nd ed.; John Wiley and Sons Inc.: New York, NY, USA, 1963
- Sosina Bezu, S., Holden, B., & Barrett, C. (2009). Activity Choice in Rural Non–farm Employment (RNFE): Survival versus Accumulative Strategy. Norwegian University of Life Sciences, Cornell University.
- Greene, W. H. (2003). Econometric analysis fifth edition. New Yersey: Prentice Hall.
- Temesgen, D., Hassan, R. M., Tekie, A., Mahmud, Y., & Ringler, C. (2008). Analyzing the determinants of farmers’ choice of adaptation methods and perceptions of climate change in the Nile Basin of Ethiopia. International Food Policy Research Institute (IFPRI).
- Hassan, R. M., & Nhemachena, C. (2008). Determinants of African farmers’ strategies for adapting to climate change: Multinomial choice analysis. African Journal of Agricultural and Resource Economics, 2(311-2016-5521), 83–104.
- Tembo, F. M., Tadesse, T., & Singini, W. (2018). Perceptions and choices of adaptation measures for climate change among teff (Eragrostis tef) farmers of Southeast Tigray, Ethiopia. Journal of Agricultural Extension and Rural Development, 10(1), 11–19.
- Thomas, R. J. (2008). Opportunities to reduce the vulnerability of dryland farmers in Central and West Asia and North Africa to climate change. Agriculture, Ecosystems & Environment, 126(1–2), 36–45.
- Sustainability, 11(10), 2991.
- Alemayehu, A., & Bewket, W. (2017). Smallholder farmers’ coping and adaptation strategies to climate change and variability in the central highlands of Ethiopia. Local Environment, 22(7), 825–839.
No competing interests reported.