IJEI-James SN

International Journal of Economic Issues, Vol. 9, No. 1 (January-June, 2016) : 1-21 © International Science Press

CLIMATE CHANGE PERCEPTIONS AND ADAPTATIONS FOR LIVESTOCK FARMERS IN BOTSWANA JAMES S. JUANA, FRANCIS N. OKURUT, PATRICIA M. MAKEPE & ZIBANANI KAHAKA1

Abstract: Livestock production plays an important role in the promotion of rural livelihoods, and economic development in many African countries. National governments have implemented various livestock development programmes to address food security, rural livelihood and improved productivity. However, the success of the outcomes of interventions aimed at promoting livestock production in Africa is critically challenged by the threat of climate change and climate variability. This study has been designed to investigate how livestock farmers in Botswana perceive climate change, the adaptation strategies farmers have adopted to cope with the challenges and the barriers to the adoption of these adaptation strategies. The study also examines the socio-economic determinants of adapting specific strategies. The study used primary data collected from the sample livestock farmers in the ten agricultural districts in Botswana. A multinomial probit method is used to model the determinants of the adaptation strategies. The results show that livestock development interventions have not addressed Climate change concerns. The four main adaptation strategies for livestock farmers include selling livestock, seeking help from veterinary officers, livestock diversification, water harvesting. Factors that influence adaptation for livestock farmers include: being married, belonging to farmers’ organization, farming experience, public employment, education level, access to agricultural extension and household size.Mainstreaming of climate change in all livestock development interventions is recommended. Keywords: climate change, livestock, adaptation, barriers, agro-ecological zones

1. INTRODUCTION Globally, it has been observed that climate change has consequences for dairy, meat and wool production, mainly arising from its impact on grassland and rangeland (Thornton & Herrero, 2008). Lack of water and increased frequency of drought in drought-prone countries will lead to increased animal death, loss in livestock weight, outbreak or increased livestock diseases; and consequently loss in the income of rural populations who depend on livestock farming for their economic survival. These impacts are bound to exacerbate rural poverty (Thornton & Herrero, 2008). 1

Corresponding author: James S. Juana, Department of Economics, University of Botswana, P/Bag UB00705, Gaborone, Botswana, E-mails: [email protected]/[email protected]/[email protected]

2 / JAMES S. JUANA, FRANCIS N. OKURUT, PATRICIA M. MAKEPE & ZIBANANI KAHAKA

Botswana is a semi-arid country in Southern Africa, with a total area of 58,173,000 hectares. Botswana is a water scarce country, with an average annual precipitation rate of 416 mm/year, which ranges from 650 mm/year in the northwest to 250 mm/ year in the southwest with the total internal renewable water resources of 2.4 km3/ year. Of this total, surface water produced internally accounts for 0.8 km3, while groundwater accounts for 1.7km3/year and an overlap of 0.1 km3/year (FAO, 2005). Apart from the perennial rivers and wetlands in the north, and the over utilized Limpopo River and its tributaries in the east, Botswana lacks enough surface water for both socio-economic and environmental sustainable development. Botswana’s agro-environment is drought-prone, has large expanses of infertile sandy soils (Chanda et al. 2009). Livestock products contribute 80% to agricultural GDP in Botswana (Republic of Botswana, 2009). The contribution of the agricultural sector (mainly livestock) to Botswana’s GDP declined from around 40% at independence to less than 2% in 2011/2012 (Republic of Botswana, 2012). Although agriculture’s contribution to GDP is minimal as compared to the mining sector, the former sector remains to be an important source of food, income and employment for the majority of rural population (Juana et al, 2014). Therefore, any adverse decline in the agriculture sector has consequences for rural income, employment and livelihood. Livestock production is the main driver of the rural economy of Botswana. The country has been exporting beef to the EU under Economic partnership agreement, Norway and South Africa (ODI 2007). Livestock exports from Botswana to EU including live animals and animal products were worth 45 Million Euros in 2010 (EC Trade 2013). Meat and Meat products contributed 3% and 1.5% to GDP and export revenue respectively in 2004(ODI 2007). However, droughts have affected the beef exports to EU (ODI 2007). In a bid to commercialize livestock production and conserve rangelands, the Botswana government formulated the tribal land grazing policy (TLGP) in 1975 (Darkoh and Mbaiwa 2002). The need for the reform arose from the perception that pastoralist production in the communal areas was ecologically destructive and unproductive (Mazonde 1994). As a result of the policy, large chucks of traditional communal grazing land1 were converted into commercial ranches and distributed to individuals on a 50-year lease period (Darkoh and Mbaiwa 2002). This has led to two livestock production systems2 namely tribal/communal and commercial with the former accounting for majority. The communal systems are comprised of the village grazing areas and cattle post areas3. This has implications for climate change adaptation with increase in the occurrence of drought in semi-arid Botswana. Over the years, Botswana had experienced climate variability and change. There is an increase in erratic rainfall patterns which have contributed to frequency and intensity of the droughts with negative implications to the livestock industry (Kenabatho et al., 2010). Temperature is projected to rise between 1 and 3 degrees Celsius by 2050, resulting in higher evapotranspiration rates. Future trends in rainfall are uncertain, but the majority of General Circulation Models (GCM) predict a decrease in precipitation, possibly with more intense rains in some parts of the country (Masike and Urich, 2008). Desertification, land degradation and biodiversity loss due to climate

CLIMATE CHANGE PERCEPTIONS

AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 3

change have negative impacts on the economy in general and agricultural productivity in particular (McDonald, 2000). The need to adapt to the adverse consequences of climate change is two sided; the vulnerable population’s awareness of the risks of climate change and their adaptive capacity on the one hand and planned adaptations on the other (FAO, 2007). The former can be achieved by providing relevant information to the vulnerable groups about the risks and consequences of climate change, while the latter includes technical developments and government insurance provided by public and private agents (FAO 2007). For public and private interventions in climate change impact adaptations to be successful, there is the need to understand farmers’ perceptions about climate change, what adaptation strategies they have adopted and what the barriers to these adaptations are. Against this background, this study was designed to assess how livestock farmers in Botswana perceive climate change and the adaptations methods they have adopted to deal with the adverse consequences of climate change. Specifically, the study was designed to: i) Assess the existing knowledge or experience livestock farmers have on climate change. ii) Explore the adaptation methods livestock farmers have adopted to cope with the adverse consequences of climate change. iii) Identify the perceived barriers to the adopted adaptation methods. iv) Analyze the socio-economic, institutional and ecological factors that determine livestock famers’ adaptations to climate change. v) Recommend both private and public policies that match with livestock farmers’ perceptions, adaptations and coping mechanisms. 2. METHODOLOGY The study used a structured questionnaire to elicit information from 350 livestock farmers on their perceptions about climate change, adaptations and barriers to climate change, and the socio-economic and environmental factors that determine livestock farmers’ adoption of a particular adaptation strategy against no adaptation. To complement the structured questionnaire, the study also had focus group discussions with selected farmer organizations in each agricultural district. The responses from these group discussions were compared with the individual livestock farmer’s responses to their perceived changes in climatic variables, adaptation strategies and barriers to adaptation. 2.1. Target population The population of interest in the study was all registered livestock farmers who are above eighteen years old. According to the Central Statistics Office (CSO), (2010) there are 108, 017 registered farmers (comprising of 107, 482 subsistence farmers and 535 commercial farmers) in Botswana. Of this figure there are 85, 346 livestock farmers, 487 of whom are commercial livestock farmers. The rest are communal or small-scale livestock farmers (CSO, 2010). This served as the reference population in the study.


The study covered all the ten agricultural administrative districts in Botswana. Table 1 presents the detailed agricultural districts and the number of farm households in each district. This is further broken down into livestock farm households (commercial and small-scale). 2.2. Sample size and sampling design The study used a sample size calculator, which shows that for the population of 85, 346 livestock farmer households given above, with a statistically appropriate sample size, allowing a margin of error of 5% and at a 95% confidence level, the response from the sample that will be the same as that of the population, is 320 livestock farmers (Raosoft, Inc., 2004). However, to allow for replacement in the sample of those who might back out of the study, 20% of the statistically selected sample will be included giving a total study sample of approximately 384 farmers. Each agricultural district is sub-divided into designated agricultural area/subdistricts. One designated agricultural area/sub-district was randomly selected from each district for the study. However, because the Central District is much larger than the rest of the districts, for Central District two areas were selected. The statistically selected sample size of livestock farmers was allocated to the selected sub-districts using proportional allocation to size (PPS), where the size represents the number of livestock farmer households in the sub-district (see Table 1). However, since the average number of cattle heads per commercial farmer exceeded that of small-scale farmers, 30% of the allocated sample size was commercial livestock farmers and 70% were small-scale livestock farmers. The numbers of livestock farmers, selected sample size, sample of small-scale and commercial farmers are presented in columns 4, 5, 6 and 7 respectively of Table 1. A list of farmer households was obtained from the District Agricultural Coordinators of the selected areas. The method of systematic sampling Table 1 Sample Size Allocation of Farmers District(1)

Area/SubDistrict (2)

Southern

Ngwaketse Central (Kanye) South East Ramotswa Kweneng Kweneng South Kgatleng Mochudi Central Serowe Selibe Phikwe North East Tati/Masunga Chobe Kasane (Pandamaatengaentenga) North West Ngamiland East (Maun) Gantsi Ghanzi Kgalagadi Hukuntsi Total Source: CSO (2010)

No. of farm households (3)

No. of livestock farmers (4)

Sample size (5)

Small-scale farmers (6)

Commercial farmers (7)

16400

12958

58

42

16

7105 25438 9682 14019 7181 7063 2283

5614 20099 7650 11077 5674 5581 1804

25 90 34 50 26 25 8

18 63 24 35 18 18 6

8 27 10 15 8 8 2

10161 4724 3962 108018

8028 3733 3130 85348

36 17 14 384

25 12 10 270

11 5 4 114


AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 5

was then used to identify the farmer households that were interviewed. One farmer per selected household (18 years and above) was interviewed. 2.3. Data Collection and analysis Tools The main data collection tool was the structured questionnaire which captured information on the socio-economic and demographic characteristics of livestock farmers; the ecological zone in which they keep their animals, their perceptions about climate change and how climate change impacts livestock farming; whether or not they have adopted any adaptation method in response to the adverse consequences of climate change and barriers to adaptation. This was complimented by focus group discussions with farmer organizations or selected group of farmers in the area or subdistrict. The data collection tool was pre-tested after the training on a sub-population similar to the target population. This was done to check for content ambiguity, clarity, data quality and time needed for the assignment before being used for the main study. Households were selected using systematic sampling method from sub-districts belonging to different strata according to the stratification in the sample design.The collected data was captured and coded using the SPSS software, and exported to STATA for analysis. The analysis used descriptive measures (such as percentages, means and measures of associations), and the multinomial logit model was used to estimate the factors that influence livestock farmers’ choice of adaptation method to cope with the consequences of climate change. 2.4. Theoretical framework The study follows a standard discrete choice model in which an individual who is randomly drawn from the population of livestock farmers in Botswana makes a choice from a set of ‘J’ mutually exclusive climate change adaptation alternatives, where j = 0, 1, 2, 3, 4 and ‘0’ means no adaptation against all the other adaptation options available to the livestock farmer. Following McFadden (1989), this study assumed that the utility individual livestock farmer ‘n’ gets from adopting adaptation method ‘j’ is given by Unj. Therefore, the general form of this relationship is given as: U nj � XnjT � ij j � 0, 1, 2, ...., J

(1)

Where X is a vector of explanatory variables and �ij is a vector of tastes or preferences of individual ‘n’, which result from a function of attributes related to the jth alternative adaptation method chosen by livestock farmer ‘n’. With ‘j’=0 implies no adaptation. The tastes or preferences of the individual livestock farmer ‘n’ may be further divided into ‘�’, which shows the mean of the population of livestock farmers, and ‘ô’ which is the individual deviation from the mean. This preference may be determined by the individual socio-economic factors like income, household size, farming experience, gender, number of livestock heads etc, or institutional or ecological, like wetland, grassland and desert. In this study, the explanatory variables were classified T


into socio-economic, institutional and environmental. These are explained in the next sub-heading. The above theory implies that the utility function of adaptation takes on the form: U ij � XijT (� � �n )� � � �nj

(2)

From equation 2, ‘µ’ is the mean of the alternative –specific errors and ‘�nj’is the random deviation from the mean (Bunch, 1991). From equation 2, the probability of choosing alternative yi is the probability that the utility from alternative yi exceeds the utility from other alternatives. Pr( yi � 1) � Pr(ui1 � ui 2 ) � Pr(� i1 � �i 1 � �i 2 ) � Pr(�i 1 � �i 2 � �i 2 �� i1 )

(3)

There are four most common climate change adaptation strategies practiced by livestock farmers in Botswana. If none of the alternatives is selected then the farmer has not adopted any climate change adaptation strategy. These include: i) ‘0’ No adaptation ii) ‘1’ Selling livestock iii) ‘3, Help from veterinary officers iv) ‘4’ Livestock diversification v) ‘5’ Water harvesting Since there are four alternatives, the probability of choice ‘m’ is given as;

Pr(yij � m) � Pr(� m � � j for all j � m ; )

(4)

Using the above information the probability that yi = j is given as: exp( xij' �)

Pr( yi � 1) �

1 � exp( xi 2 ' �) � exp(x i 3� ') � exp(xi 4 ')

(5)

Similarly; Pr( yi � 2) �

exp(xij' �) 1 � exp( xi1 ' �) � exp( xi 3 � ') � exp(x i 4 ')

Pr( yi � 3) �

Pr( yi � 4) �

exp(xij' �) 1 � exp(xi 1 ' �) � exp( xi 2� ') � exp(x i 4 ') exp( xij' �) 1 � exp( xi1 ' �) � exp(x i 2� ') � exp(x i 3 ')

(6)

(7)

(8)

A positive � implies that farmers attach a utility to the particular adaptation strategy.


AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 7

2.5. Empirical model As discussed in the previous sub-section, four adaptation strategies were most common among livestock farmers in Botswana. These were selling livestock, help from veterinary officers, livestock diversification and water harvesting in order of importance. A multinomial logit regression was used to determine the factors affecting livestock farmers’ adoption of particular adaptation methods.

Yi � �Xij � �

(10)

Where Yi is the limited dependent variable and takes the values of 0, 1, 2, 3, and 4; Xi is the vector of explanatory variables which are discussed in Table 2, and � is the error term. The explanatory variables are placed into three categories; households’ socioeconomic characteristics, institutional factors and ecological factors. Households’ socioeconomic characteristics include level of education, age, gender, household size, Table 2 Description and measurement of the explanatory variables Category of variable

Variable

Type of variable

Variable measurement and expected sign

Households’ Characteristics

Years of education Age Gender Household size

continuous Continuous Dummy Continuous

Farming experience Farm income

Continuous Continuous

Non-farm income

Continuous

Marital status

Dummy

No of years of formal education (+) No of years since birth (+) ‘1’ if male and ‘0’ otherwise (+ or –) Number of individuals taken care of by the farmer (+) No of years of farming (+) Income derived from farming activities (+) Income derived from other sources besides farming (+) ‘1’ if married and ‘0’ otherwise (+ or – ) ‘1’ if farmer owns radio/TV and ‘0’otherwise (+) ‘1 if farmer has access and ‘0’ otherwise (+ ) ‘1’ if farmer ha s information on climate change and ‘0’ otherwise (+) ‘1’ if farmer has access to credit and ‘0’ otherwise (+) ‘0’ if farmer easy access to farm inputs and ‘0’ otherwise (+) ‘1’ if farmer has easy access to output market and ‘0’ otherwise (+) ‘1’if farmer lives in the Kalahari desert and ‘0’otherwise (+/-) ‘1’if farmer is located in the Okavango Delta and ‘0’ otherwise (+/-) ‘1’ if farmer is located in the Savannah Grassland and Woodland (+/-)

Ownership of radio or TV Dummy Institutional factors

Ecological Factors

Access to extension services Information on climate change Access to credit/loan facilities Access to input market

Dummy Dummy Dummy Dummy

Access to output market

Dummy

Kalahari Desert

Dummy

Okavango Delta

Dummy

Savannah Grasslands and Woodland

Dummy


farming experience, farm income, non-farm, income, marital status and ownership of radio or television set. The institutional factors considered in the study are; access to extension services, information on climate change, access to credit/loan facilities and access to input and output markets. Botswana has three distinguishable ecological zones; Kalahari Desert, Savannah Grassland and Woodland and the Okavango Delta. i) The Kalahari Desert is predominantly covered with sandy soil and is not suitable for cultivation, but supports considerable numbers of livestock and wildlife. This zone is also the most water stressed in the country. ii) The Savannah Grasslands and woodland with less forested area. This zone is mostly found in the eastern part of the country, with less harsh climate and more fertile soil than the Kalahari, and it receives in excess of 400 mm rainfall (average rainfall in the country). This zone supports both livestock production and perennial crops. iii) The Okavango Delta, which is the northwest of the country, has vast areas of open lush water, green wetlands, with abundance of wildlife (FAO, 2005). This ecocologial zone supports arable and livestock farming, and is conducive for wildlife. It is expected that climate change adaptation stratetgies practiced by livestock farmers is significantly influenced by ecological factors. Table 2 shows the detailed description of the variables, their measurement and expected signs. A positive sign shows that there is positive relationship between the particular variable and the probability of choosing the related climate change adaptation strategy and a negative sign shows otherwise. Where both signs are given the implication is that the explanatory variable may either increase or decrease the probability of choosing the related climate change adaptation technique. 3. PRESENTATION AND DISCUSSION OF EMPIRICAL RESULTS This section presents the demographic and socio-economic characteristics of livestock farmers in Botswana, their perceptions about climate change, adaptation strategies adopted to respond to the challenges of climate change and barriers to the adoption of these strategies. The section also presents the socioeconomic, institutional and ecological factors that determine livestock farmers’ adaptations to climate change. 3.1. General characteristics of livestock farmers in Botswana As indicated in the methodology, 384 livestock farmers in Botswana were interviewed on their perceptions and adaptations to climate change in Botswana. Of this figure, 76% of the respondents were males and only 24% were females. This implies that the majority of livestock farmers in Botswana males; hence, more males than females are engaged in livestock farming activities. The analytical results also show that 55% of the respondents were engaged in livestock farming on a full time basis, while 45% farmed on part time basis. This implies that there are a significant portion of farmers (smallholder or otherwise) in Botswana who are engaged in other activities, especially paid employment, besides farming.


AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 9

Results also show that 15% of the respondents never went to formal schools, 19% had primary school certificates, 23% had junior secondary school certificates, 22% had senior secondary school certificates, 13% have diplomas from recognised institutions in the country and only 8% have acquired university degree(s). This has consequences for farmers’ perceptions about and adaptations to climate change, which are further discussed under the econometric results. About 27% of the livestock farmers are unemployed, while 22% are employed either in the public or private sector and 49% are self-employed. These findings are consistent with the education status analysed above in which only 21% of livestock farmers have post-secondary education. There is an equal split between married and single livestock farmers in Botswana, with about 12% respondents who are either divorced or widowed. 3.2. Farmers perceptions and adaptations to climate change This sub-section presents the perceptions livestock farmers have about climate change and the adaptation strategies they have used to cope with the adverse consequences of climate change. Table 3 presents farmers awareness of climate change. The first row shows that all the respondents are aware that the climatic variables are changing. While all the livestock farmers agree that temperature has changed over the last twenty years, an insignificant number (9%) perceived a decrease in temperature. As with temperature, the majority of livestock farmers in Botswana reported that both precipitation and rainfall duration have changed. Responding to the question of a decrease or an increase, 95% of those who perceived a change in precipitation reported that precipitation has significantly decreased in the country. Similarly, 98% of the respondents reported a decrease in rainfall duration. These responses indicate that the majority of farmers in Botswana have perceived that both precipitation and rainfall duration have decreased. These perceptions are consistent with the Intergovernmental Panel Climate Change Working Group II AR5 report which suggests that while temperature has increased, precipitation and rainfall duration has decreased in most parts of Africa (IPCC, 2014). All the respondents indicated that there has been a perceived change in the frequency of droughts and floods. However, there is almost an equal split between farmers perceiving an increase and those perceiving a decrease in the frequency of droughts in the country. Similarly, there is an equal split between farmers who have perceived increased flooding and those who have perceived decreased flooding in Botswana. These responses are consistent with the findings of Tsheko (2003) which shows that while some parts of Botswana (example Mahalapye) are drought prone areas, others like Shakawe and Maun are prone to flooding, which confirms the half split in responses between increase and decrease droughts and flooding. The majority of livestock farmers (97%) reported that there has been no change in humidity. 3.3. Perceived climate change impact on livestock Livestock farmers were asked about their perceived impact of climate change on livestock in Botswana. The analysed responses are presented on Table 4. Out of the

Source: Authors’ computations

384

Flooding

341

Rainfall Duration 11

341

Precipitation

384

384

Temperature

Drought

384

Climate Change

Humidity

Change

Climate variables

0

0

373

43

43

0

0

No Change

384

384

384

384

384

384

384

Total

198

195

11

8

16

348

N/A

Increased

186

189

0

333

325

36

N/A

Decreased

384

384

11

341

341

384

N/A

Total

100

100

3

89

89

100

100

% Changed

Table 3 Climate change awareness by livestock farmers in Botswana

0

0

97

11

11

0

0

% No Change

52

51

100

2

5

91

N/A

% Increased

48

49

0

98

95

9

N/A

% Decreased



AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 11

384 livestock farmers interviewed 100% perceived that climate change leads to livestock death, while the majority of them perceived loss of weight of farm animals, increased distance of watering livestock, decreased pasture yield and outbreak of livestock diseases as adverse consequences of climate change. Details are presented in Table 4. These findings are consistent with the findings of Juana et al (2013), Nkondzeet al (2014), and Mendleni and Anim (2011). Table 4 Perceived impact of Climate change on livestock in Botswana Impact of Climate change Death of livestock loss of weight of farm animals Increased distance of watering livestock Decreased pasture yield Increased Livestock diseases

Frequency

%

384 348 325 256 350

100 91 85 67 91


3.3. Adaptation strategies adopted by livestock farmers in Botswana As one of the main objectives of this study, livestock farmers were asked about the adaptation strategies that they have adopted to respond to the adverse consequences of climate change. The question was two folded; i) the adaptation strategies adopted by the livestock farmers and ii) the main adaptation strategy that the farmer adopted. The analysed answers to the first question are presented in Table 5 and the second are presented in Figure 1. Table 5 Adaptation strategies adopted by livestock farmers in Botswana Adaptation Strategy Selling livestock Water Harvesting Help from Veterinary Officers Livestock diversification 318 Buying feeding supplements Switching from farm to non-farm activities Keeping only drought resistant livestock Moving livestock from drier to wetter region

Frequency

%

356 298 325 83 82 104 60 68

93 78 85 21 27 16 18


Table 5 presents the number of farmers that have adopted each of the adaptation strategies. From the table, the most common adaptation strategy adopted by livestock farmers is “selling livestock”. This strategy has been adopted by 93% of the livestock farmers in Botswana. This is followed by “seeking help from veterinary officers” which has been adopted by 85% of the farmers and “livestock diversification” and “water harvesting”, which were adopted by 83% and 78% of the livestock farmers respectively. The other adaptation strategies were practiced by an insignificant number of livestock farmers.


The analysed results of the other question, the main adaptation strategy practiced by the farmers is presented in Figure 1. The figure shows that a significant proportion of livestock famers practiced “selling of livestock”, “seeking help from veterinary officers”, “water harvesting” and “livestock diversification” as the main climate change adaptation strategies. From the figure, it is shown that 20% , 20%, 19% and 18% of respondents have respectively adopted these adaptation options as their main adaptation strategies. These findings are consistent with those presented in Table 5. Figure 1: Main adaptation strategy adopted by livestock farmers


An insignificant number of livestock farmers practiced “keeping only drought resistant livestock, “switching from farm to non-farm activities”, “buying animal feed supplements” and “moving livestock from drier to wetter regions” as the main adaptation strategies. These may be related to both ecological factors and economic/ livelihood opportunities where the farmers are located. While we expect farmers in the Kalahari Desert to move livestock to wetter regions, those in the Okavango Delta may switch from livestock farming to eco-tourism. This is elaborated on in the econometric analysis. 3.4. Barriers to climate change adaptations There are certain barriers to climate change adaptations strategies. Only one percent of the interviewed livestock farmers reported that they have not experienced any barriers to climate change adaptation strategies. This implies that 99% of the interviewed farmers had experienced certain barriers to climate change adaptation strategies. Table 6 shows the results of farmers’ responses on barriers to climate change adaptation strategies. From the farmers’ responses, the main barriers to climate change


AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 13

adaptation strategies in Botswana are “access to improved technology”, “limited land for livestock farming activities” and “government restrictions on land use”. Other barriers like “lack of access to extension workers”, “lack of access to loan facilities”, “lack of access to output market” and “lack of access to input market” are not are not perceived by the majority of farmers in Botswana as barriers. Table 6 Barriers to climate change adaptations Barrier

Yes

No

% Yes

%No

No Barrier Inadequate or no access to extension services Inadequate or no information on climate change Lack of access to loan facilities Lack of access to input market Lack of access to output market Inadequate or lack of access to improved technology Limited land for livestock farming activities Government policy restrictions on land use lack of capital

5 66 5 18 10 10 348 332 289 22

379 313 374 361 369 369 31 47 90 357

1 17 1 5 3 3 92 88 76 6

99 83 99 95 97 97 8 12 24 94


Livestock farming is seen as one of the major economic activities that leads to sustainable rural livelihoods and makes a significant contribution to the country’s export earnings. Therefore, the Government of Botswana has instituted a number of programmes that assist the development of livestock farming and grant loans to farmers in the country. These include Youth Development Fund (YDF), and Citizen Entrepreneurial Development Agency (CEDA), National Agriculture Master Plan for Arable Agriculture and Dairy Development (NAMPAAD) and Livestock Management and Infrastructure Development (LIMID). In addition to these, the government also offers grants and subsidies especially to the rural farmers. There is also the Botswana Meat Commission (BMC) that provides access to both inputs and output (NDP10, 2010). Therefore, farmers in Botswana do not perceive lack of access to loan facilities and input markets as barriers to climate change adaptation strategies. 3.5. Determinants of adaptations to climate change The study estimated the multinomial logit model for livestock farmers’ responses to climate adaptations strategies in Botswana. From the responses, there are eight adaption strategies that livestock farmers in Botswana have practiced. However, the main adaptation method they adopted was “Selling livestock”, which is generally practiced by 356 (83%) of livestock farmers with 76 (20%) of the livestock farmers using it as their main adaptation strategy. “Help from veterinary officers”, was generally practiced by 325 (85%) of the farmers, but 76 (20%) used it as the main adaptation method. “Livestock diversification”, was generally practiced by 318 (83%) livestock farmers, but used as the main adaptation method by 70 (18%) livestock farmers .”Water Harvesting”, was generally practiced by 298 (78%) livestock farmers, and used by 73(19%) as their main adaptation method (see Table 5 and Figure 1). In the


multinomial logit regression analysis for each of the four main adaptation strategies, the base category was “no adaptation”. 3.5.1. Pre-estimation econometric tests The Hausman Test was carried out to determine whether the assumption of Independent Irrelevant Alternatives (IIA) was fulfilled. This implies that there is no misspecification of the estimation model. The null hypothesis is that there is no model misspecification. The result is presented on Table 7. The results indicate that there is no model misspecification; hence the model was adequately specified for all the adaptation strategies. Table 7 Results of the Hausman test for IIA Omitted

Chi Square

P-Value

Evidence

Selling Livestock Help from veterinary officers Livestock diversification Water Harvesting

0.87 1.43 0.68 0.75

0.8769 0.9832 0.7951 0.8217

Fail Fail Fail Fail

to reject HO to reject HO to reject HO to reject HO


The Variance Inflation Factor (VIF) and the Breusch-Pagan Lagrange tests were performed to test for the presence of multicolinearity and heteroscedasticity respectively. Since none of the VIF values exceeded 5, the conclusion was that there is no severe problem of multicolinearity. Also, the result of the Breusch-Pagan Lagrange test showed that no two independent variables are heteroscedastic. The Chi square (�2) scores for all the four alternative adaptation strategies were 35.58, 28.56, 35.08 and 64.92 respectively each with �-value of 0.000 implying that the model has a high probability predictive value. 3.5.2. Multinomial Logit Results The results of the multinomial logit estimation are presented on Table 8. For each of the adaptation options, the estimated coefficients are presented in the first column, the Z-scores in column 2, and the marginal effects are presented in column 3. The estimated coefficients show the direction of the variables impact on a particular adaptation option. The marginal effect shows both the direction and the magnitude of the impact of the variable on an adaptation option. Selling Livestock “Selling livestock” is identified to be one of the main adaptation methods practiced by livestock farmers in Botswana. The variables that significantly determine the choice of this adaptation method are employment status (self-employment), marital status, membership of farmer organization and agro- ecological zone. Being self-employed increases the probability of choosing this adaptation option by 27.5% as opposed to not been employed. Similarly, being married increases the probability of choosing “selling livestock” as an adaptation option. Belonging to a farmer organization increases

0.950 0.320 -1.140 -0.940 1.140 1.910 -1.050 0.000 0.960 -1.100 -0.780 2.040 0.380 -0.060 3.250 -0.030 5.414 2.783 16.727 -0.880 76 31.24 0.0521 0.2656

Education level 0.057 age 0.035 Gender -0.486 pub employment -0.626 Priv. employment 0.697 self- employment 0.975* Household size -0.131 Farming experience 0.000 fulltime 0.407 Farm income 0.000 nonfarm income 0.000 Marital status 1.017** Media 0.232 Extension worker -0.027 Farmer org 2.033*** peers -0.012 Workshop Kalahari Desert 0.314** Okavango Delta -0.275** Savannah Grassland 0.184** _cons -2.340 Number of obs LR chi2(20) Prob > chi2 Pseudo R2

1.030 -0.130 0.310 1.880 1.650 1.950 -0.920 1.740 0.460 -0.540 -1.430 -0.740 0.430 -0.890 2.870 1.430 0.330 11.133 2.635 0.514 -0.680 76 16.36 0.6941 0.1663

Z 0.018 -0.004 0.041 0.480 0.388 0.318 -0.033 0.017 0.057 0.000 0.000 -0.166 0.067 -0.118 0.156 0.225 0.097 0.095 0.031 0.008

Mfx (dy/dx)

Help from Vet Officers Coef.

0.018 0.061 0.011 -0.013 -0.141 0.137 -0.169 1.350* 0.252 1.094 0.275 0.997* -0.042 -0.111 0.000 0.058* 0.134 0.191 0.000 0.000 0.000 0.000 0.309 -0.559 0.070 0.249 -0.009 -0.433 0.069 0.470** -0.004 0.676 0.301 0.230 0.167*** -0.263 0.303** 0.060 0.113 -1.837

Mfx (dy/dx) 0.110* -0.121 -0.275 1.402* 1.246* 0.734 0.000 0.028 1.271*** 0.000 0.000 -0.188 0.333 1.062** 0.872* 0.286 0.153*** 0.175*** 0.211* -0.600

1.810 -1.170 -0.570 1.840 1.790 1.370 0.000 0.770 2.650 -0.840 -1.040 -0.220 0.610 2.370 1.913 0.580 7.321 14.583 1.991 -0.230 70 38.44 0.0078 0.3316

Z 0.029 -0.032 -0.069 0.495 0.430 0.214 0.000 0.007 0.287 0.000 0.000 -0.051 0.078 0.326 0.072 0.082 0.253 0.092 0.031

mfx (dy/dx)

Livestock Diversification Coef. 0.002 -0.088 0.850 1.271 0.469 0.536*** -0.008 0.176 0.000 0.000 -0.336 -0.331 0.133 0.531** 0.190 1.771** 0.503** -0.122** 0.139** -1.949

Coef. 0.030 -0.670 1.510 1.470 0.840 2.690 0.180 0.350 0.990 0.930 0.610 0.390 0.250 2.953 0.310 2.190 2.634 2.711 2.439 -0.610 73 31.8 0.033 0.3306

z

0.000 -0.018 0.227 0.375 0.100 0.112 -0.002 0.038 0.000 0.000 -0.073 -0.079 0.029 0.074 0.043 0.280 0.043 -0.093 0.069

mfx (dy/dx)

Water Harvesting

AND

ADAPTATIONS FOR

Source: Adapted from the econometric results Note: - Base Category is no adaptation - P- values are the Z-scores - dy/dx for discrete change for dummy variable show a change from 0 to 1. - Significant at 10% level (*) - Significant at 5% level (**) - Significant at 1% level (***)

Z

Selling Livestock Coef.

Variable

Table 8 Determinants of livestock farmers’ adaptations to climate change

CLIMATE CHANGE PERCEPTIONS LIVESTOCK FARMERS IN

BOTSWANA / 15


the probability of selling livestock by about 7%. Agro-ecological zones also significantly determine the probability of selling livestock as an adaptation option. Livestock farmers in the Kalahari Desert are 26.3% more likely to sell livestock as an adaptation option than those in the other agro-ecological zones. Similarly, livestock farmers in Savannah Grassland are 6% more likely to sell livestock. On the other hand, farmers in the Okavango Delta are 26% less likely to sell livestock as an adaptation option. Changes in climatic variables have more negative impact on livestock farmers in the Kalahari Desert and the Savannah Grassland than for those in the Okavango Delta. Climate change impact on livestock has more adverse consequences on selfemployed livestock farmers than their private or public employed counterparts, who have other sources of stable non-farm income. Therefore, selling livestock is viewed as a climate change adaptation option. Also, belonging to a farmer organization may increase the awareness of the adverse consequences of farmers; hence, choosing to sell livestock to a manageable quantity. Help from veterinary officers The estimation results show that choosing “ help from veterinary officers” as a climate change adaptation method relative to “ no adaptation” significantly increases with employment in the public sector, self -employment, farming experience, and agroecological zones. Relative to “no adaptation” public and self-employment increases the probability of seeking help from veterinary officers by 24% and 31% respectively. Similarly increasing farming experience by one year from an average of 14 years increases the probability of seeking help from veterinary officers by 2%. This implies that the more the number of years of farming experience, the greater the probability of choosing this adaptation method as the main adaptation strategy as compared to the other adaptation measures. As with “Selling livestock”, agro-ecological zones significantly determine the choice of “seeking help from veterinary officers” as an adaptation option. Livestock farmers living in the Kalahari Desert are 9.5% more likely to choose this option than those who do not live in this zone. Similarly, those living Okavango Delta are 3.1% more likely to use this adaptation method than those who living in other agro-ecological zones. All the other variables do not significantly determine “seeking help from veterinary officers”. Livestock diversification Level of education, public and private employment, extension workers and agroecological zones significantly increase the probability of adopting “livestock diversification” as an adaptation method, while practicing livestock farming on fulltime basis significantly decreases the probability. Increasing the level of education by one year from an average of 10 years increases the probability of adopting livestock diversification by 3%. Also, being employed in the public or private sector, relative to unemployment, increases the probability of choosing livestock diversification by 49.5% and 43% respectively. Also having adequate


AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 17

access to extension workers increases the probability of choosing livestock diversification by 33%. Furthermore, being a fulltime livestock relative to part-time livestock farmer increases the chances of choosing livestock diversification as the main adaptation method by 28.7%. This implies that full-time farmers are more likely to diversify livestock than their part-time partners. The regression results also suggest that livestock diversification, as a main adaptation method, relative to no adaptation is significantly increased by all the three agro-ecological zones. Living in the Kalahari Desert, Okavango Delta and the Savannah Grassland increases the probability of diversifying livestock by 25%, 9% and 3% respectively. This implies that regardless of the agro-ecological zone, livestock farmers view diversifying livestock as a main climate change adaptation technique in Botswana. This is supported by the notion that farmers in Botswana raise cattle, goats, sheep and donkeys together. Water harvesting Water harvesting is the practice of collecting and storing rain water from rooftops and runoffs, and used to water farm animals when there is acute shortage of rain. This practice is regarded as a main climate change adaptation method. The regression results indicate that the size of livestock farmers’ households and the attendance of workshops significantly determine the probability of choosing water harvesting as the main climate change adaptation method. Attending workshops increases the probability of choosing water harvesting as the main climate change method by 28% relative to farmers who do not attend workshops. Similarly, increasing household size by one member increases the probability of choosing water harvesting as the main climate change method by 11%. This implies that farmers with more family members are more likely to practice water harvesting as the main climate change adaptation method than those with less number of family members. The possible reasons for these are that in workshops, farmers as well as extension workers share knowledge on the practice of water harvesting as a climate change adaptation method. Also, farmers with more household members find it easier to engage in water harvesting than those with fewer family members. The results also suggest that agro-ecological zones significantly determine the probability of choosing water harvesting as the main adaptation method in Botswana. Farmers in the Kalahari Dessert and Savannah Grassland are more likely to practice water harvesting as the main adaptation technique than those living in the Okavango Delta. As shown in Table 8, living in the Kalahari Desert and the Savannah Grasslands increases the probability of choosing water harvesting by 4% and 7% respectively, while living in the Okavango Delta decreases the probability of adopting water harvesting by about 9%. While both the Kalahari and Savannah Grasslands are relatively dry, the Okavango Delta is wetter and has an expansive wetland that supports livestock grazing even during the driest period in Botswana. Brief discussions of the econometric results Various studies have been undertaken in different countries in sub-Sahara Africa to estimate the determinants of farmers’ choice of main climate change adaptation


methods. The empirical findings of this study can be compared to the results of other empirical findings on the continent. Unlike the findings of Komba and Muchapondwa (2012), Juana et al (2013), Nhemachena and Hassan (2007) the regression results of the current study indicate that the level of education, gender and age are insignificant determinants of the probability to choose climate change adaptation methods in Botswana. The most important determinant of livestock farmers’ choice of adaptation methods in Botswana is agro-ecological zones. In each of the main climate change adaptation options agro-ecological zones significantly determine the choice probability. For example, while living in Okavango Delta decreases the probability of selling livestock, living in the Kalahari Desert or Savannah Grassland increases the probability of selling livestock as the main climate change adaptation method. Similarly, living in the Okavango Delta or Kalahari Desert increases the probability of seeking help from veterinary officers, while living in the Savannah Grassland does not significantly determine the probability of choosing this option. These and other findings suggest that the main adaptation option selected by livestock farmers differs in the level of significance, magnitude and direction with agro-ecological zones in Botswana. These findings are consistent with the findings of Komba and Muchapondwa (2012), Ishaya and Abaje (2008) and Deresa et al. (2011). The other empirical finding that needs further clarification is being a member of a farmer organization. The estimated results show that this variable significantly determines the probability of choosing each of the main adaptation methods in Botswana. In literature, most studies either ignore this variable or the variable is found to be insignificant. However, in Botswana, it is not surprising that this variable plays an important role in determining the probability of choosing each of the main climate change adaptation techniques. The majority of livestock farmers in Botswana have not gone beyond secondary school, but gain information on climate change through farmer organizations and workshops. Even the media is not a significant determinant of the probability of choosing each of the climate change options. Even farming experience is not as strong a determinant of farmers’ choice of adaptation techniques as does membership of a farmer organization. 4. CONCLUSION AND POLICY RECOMMENDATIONS This study was designed to assess the perceptions livestock farmers in Botswana have about climate change, the main adaptation options they have used to address the adverse consequences of climate change, the barriers inhibiting the implementation these adaptation techniques and to examine the determinants of choosing each main adaptation strategy. The study used a highly structured questionnaire, together with focus group interviews to elicit information from livestock farmers in all the agricultural districts in Botswana. The analysed results show that all the livestock farmers of Botswana are aware of changes in climate variables and they perceive that temperature and precipitation have changed over the past years. The frequency of droughts and floods has increased in country. This has led to increased livestock death, loss of weight, long distances to


AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 19

watering livestock and increased livestock diseases. To address these adverse consequences, farmers have resorted to selling livestock to reduce the numbers to manageable levels, diversified livestock to include those that can withstand adverse climate conditions, water harvesting and seeking help from veterinary officers as the main adaptation methods. However, the families reported that limited land for livestock grazing because of government policy restriction on land use, and inadequate access to improve livestock farming technologies are the main barriers to adopting climate change adaptation techniques. Because of government policy programmes on livestock, the farmers have adequate access to credit facilities, input and output markets, and capital. The results of the multinomial logit regression indicate that the probability of livestock farmers choosing a main climate change adaptation strategy is significantly determined by membership of a farmer organization and the agro-ecological zone in which the farmers live. These have consequences for both private and public policies that address climate change adaptation in Botswana. Policy implications of the research findings Given the above research findings and the farmers’ recommendations, the role of the Government of Botswana in helping livestock farmers to adapt to the adverse consequences of climate change should include: i) Since living in the Kalahari Desert and Savannah Grassland increases the probability of farmers choosing water harvesting as the main adaptation technique, this technique should be facilitated through the provision of water storage equipment for farmers living in these agro-ecological zones. ii) Livestock diversification is a common and significant climate change adaptation technique in Botswana. Therefore, the need exists for the dissemination of this information and encouragement of livestock farmers to practice this method in the country. iii) Farming experience is not a significant determinant of the probability to choose any of the main adaptation methods. However, belonging to a farmer organization significantly determines the probability to choose each of the main climate change adaptation methods. This implies that the government and non-governmental and multilateral institutions should encourage the formation of farmers’ organizations and that agricultural extension workers should organize frequent workshop to discuss climate change issues with farmers through the farmer organizations. iv) In their responses, livestock farmers stated that lack of access to improved technology, limited land for livestock farming activities and government policy restrictions on land use are the main barriers to climate change adaptation. These could be addressed by government improving on the existing research capacity on innovative climate change adaptation technologies and letting farmers to have access to these technologies and include farmers’ representatives in all land use policy issues.


v)

Help from veterinary officers has also been identified as one of the main climate change adaptation methods. This requires the inclusion of advanced training of veterinary officers, who can work in close collaboration with climate change control officers and give professional advice to livestock farmers on how to deal with climate change induced livestock diseases. vi) Selling livestock to reduce the number to a manageable level was also reported to be a main climate change adaptation technique for livestock farmers in Botswana. In periods of extreme climatic events, Botswana Meet Commission can increase its purchase of livestock, especially those that highly susceptible to climate, keep part of the sale amount to help restock after the extreme climatic event. Acknowledgements The authors are grateful to the Office of Research and Development, University of Botswana for funding the research. They would also like to thank the research assistants for their efforts to collect and enter country-wide data in a coherent and consistent way. All computational and grammatical errors are the responsibility of the authors.

Notes 1.

The acreages were 64km square and were revised to 36 square km.

2.

There were 2.554 million cattle, 1.769 million goats, 0.295 million sheep in 2011(Statistics Botswana 2013).

3.

Open access to the commons

References Chanda, R., Chipanshi, A., & Totolo, O. (2009), “Towards Climate Resilient Agriculture in Botswana”. Regional Conference on Strenhthening Local Agricultural Innovations to Adapt to Climate Change (pp. 109-124). Dar-Es-Salam: SADC. Deressa, T.; Hassan, R.; Ringler, C.; Alemu, T. and Yesuf, M. (2008), “Analysis of the determinants of farmers’ choice of adaptation methods and perceptions of climate change in the Nile Basin of Ethiopia”. IFPRI discussion paper no. 00798 International Food Policy Research Institute, Washington, DC. FAO. (2007), Food and Agricultural Organization of the United Nations. Retrieved 03 01, 2013, from www.fao.org/nr/water/aquastat/countries/botswana/index.stm: Government of Botswana. (2009), National Development Plan 10, 2009-2016. Gaborone: Government Printing. IPCC, (2014), Summary for policymakers. In: Climate Change 2014: Impacts, Adaptations, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of IPCC. Ishaya, S. and Abaje, I.B. (2008), “Indigenous people’s perceptions on climate change and adaptation strategies in Jema’a local government area in Kaduna State in Nigeria”. Journal of Geography and Regional Planning 1 (8): 138-143. Juana, J.S.; Makepe, P.M.; Mangadi, K.T. and Narayana, N. (2014), “The socio-economic impact of drought in Botswana”. International Journal of Environment and Development. 11(1): 4360.


AND

ADAPTATIONS

FOR

LIVESTOCK FARMERS

IN

BOTSWANA / 21

Juana, J.S.; Kahaka, Z. and Okurut, FN. (2013), “Farmers’ perceptions and adaptations to climate change in sub-Sahara Africa: A synthesis of empirical studies and implications for public policy in African Agriculture”. Journal of Agricultural Science 5(4): 121-135 Komba, C. and Muchapondwa, E. (2012), “Adaptations to Climate Change by smallholder farmers in Tanzania” ERSA working paper 299. Economic Research Southern Africa, UCT, Cape Town, South Africa Lapologang, M (2009), “The ‘Shrinking Commons’ in the Lake Ngami Grasslands, Botswana: The Impact of National Rangeland Policy.” Development Southern Africa 26(4),11-25. McDonald S (2000), “Drought in Southern Africa: A study for Botswana”. XIII International Conference on Input-Output Techniques, University of Macerata, Italy. August 21st-25th 2000. McFadden, D. (1989), “A method of simulated moments for estimation of discrete response models without numerical integration”. Econometrica 57(5): 995-1026. Mandleni, B., & Anim, F. (2011), Perceptions of Cattle and Sheep Framers on Climate Change and Adaptations in the Eastern Cape Province of South Africa. Journal of Human Ecology, 34 (2), 107112. Masike, S. & Urich, P. (2008), “Vulnerability of the traditional beef sector to drought, and the challenges of climate change: The case of Kgatleng District, Botswana”. Journal of Geography and Regional Planning 1(1): 012-018. Mengistu, D. K. (2011), Farmers’ perception and knowledge of climate change and their coping strategies to the related hazards: Case study from Adiha, central Tigray, Ethiopia, Agricultural Sciences, 2(2): 138-145. Nhemachena, C and Hassan, R. (2007), “Micro-level analysis of farmers’ adaptations to climate change in Southern Africa” IFPRI Discussion Paper no. 00714. International Food Policy Research Institute Washington, DC. Nyanga, P., Johnsen, F., Aune, J., & Kahinda, T. (2011), Smallholder Farmers’ Perceptions of Climate Change and Conservation Agriculture: Evidence from Zambia. Journal of Sustainable Development, 4 (4), 73-85. Rohde. R.F. , Moleele N.M. , Mphale. M., Allsopp. N., Chanda. R., Hoffman. M.T., Magole . L., Young. E., (2006), “Dynamics of Grazing Policy and Practice: Environmental and Social Impacts in Three Communal Areas of Southern Africa.” Environmental Science and policy, 9: 30 2 – 3 16. Statistics Botswana (2013), 2011 Annual agricultural survey report, Government Printing 2011. Tsheko, R. (2003), “Rainfall reliability, drought and flood vulnerability in Botswana” Water SA 29(4): 389-392.