Ethiopian Journal of Environmental Studies & Management 7 Suppl.: 832 – 839, 2014. ISSN:1998-0507 doi: http://dx.doi.org/10.4314/ejesm.v7i2.3S Submitted: September 17, 2014 Accepted: November 17, 2014

FARMERS’ PERCEPTION AND ADAPTATION TO CLIMATE CHANGE: HECKMAN’S TWO STAGE SAMPLE SELECTION MODEL

*URGESSA TILAHUN 1 AND AMSALU BEDEMO 2 1Haro Sabu Agricultural Research Center, P.O.Box 10, Dale Sadi, Kellem Wollega, 2 Assosa University, P.O.Box 018, Asossa, Ethiopia

Abstract This study examines farmers’ perception and adaptation to climate change in Guto Gidda and Sasigga districts of Regional National State in Ethiopia. Primary data for the study were collected from 142 farm household heads drawn from five kebeles of district and four kebeles of district through structured questionnaire. Heckman two stage sample selection model was employed to examine farmers’ perception and adaptation strategies to climate change. Estimation result shows that farmers level of education (p=0.040), household nonfarm income (p=0.030), livestock ownership (0.001), extension on crop and livestock (p=0.000), household’s credit accessibility (p=0.001), perception of increase in temperature (p=0.000) and perception of decrease in precipitation (p=0.000) significantly affect the adaptation to climate change. Similarly, farmers’ perception of climate change was affected significantly by information on climate (p=0.000), farmer to farmer extension (p=0.009), local agro \–ecology (p=0.002), number of relatives in development group (p=0.015) and perception of change in duration of season (p=0.002).

Key Words: Climate Change, East Wollega, Western Ethiopia

Introduction steep slopes and erodible soils, low Ethiopia, one of the developing countries, vegetation cover of the soil, burning of dung is facing serious natural resource degradation and crop residues, declining fallow periods, problems (Anemut, 2006). The diversity in and limited application of organic or altitude accompanied with climatic and inorganic fertilizers. ecological variations which affect production Climate is a primary determinant of is among the features of the country (Shibru agricultural productivity (Apata et al ., 2009). and Kifle, 1998). One of Ethiopia's principal The rate and magnitude of change in climate natural resources is its rich endowment of characteristics determines agronomic and agricultural land. Agriculture is the backbone economic impacts from climate change of the Ethiopian economy and is given (Bruce et al ., 2001). Though climate change special attention by the government to is a threat to agriculture and non-agricultural spearhead the economic transformation of the socio-economic development, agricultural country. However, land degradation, production activities are generally more especially soil erosion, soil nutrient depletion vulnerable to climate change than other and soil moisture stress, is a major problem sectors (Ayanwuyi et al ., 2010). The purpose confronting Ethiopia. The proximate causes of this study is therefore, to examine the of land degradation include cultivation of farmers’ perceptions and adaptation to

*Corresponding Author: Urgessa Tilahun 832 Email: [email protected]

Farmers’ Perception and Adaptation to Climate Change ...... URGESSA & AMSALU climate change, measures taken by farmers in to 2,450 meters above sea level which adapting climate change and causes of non reflects the existence of three agro-ecological adaptation in climate change. zones i.e. Kolla, Woina Dega and Dega. As a Description of the Study Area result this area is experienced mean annual Guto Gida is among the 17 districts of temperature of slightly greater than 15 0c and East Wollega Zone that has 21 farmers mean annual rainfall of 1,600 mm to 2,000 associations and 1 urban center possessing a mm (GGDFEDO, 2011). Most part of the total area of 901.80 km 2. It is contiguous with land of Sasiga district has an elevation above Gidda Ayyana, Abe Dongoro and Gudaya 1000 meters and characterized by sub Bila through the North, and Wayyu tropical climatic condition with a mean Tuka through the East, through annual temperature between 26.5°C and 27°C the south and Digga, Sasiga and Benshengul and mean annual rainfall of 800 mm to 1,200 Gumuz Regional State through the west mm (SDFEDO, 2011). (GGDFEDO, 2011). Sasiga district is also found in East Methodology Wollega Zone. It is located at about 18 Sampling Procedure kilometers west of town, zonal This paper used both primary and town, possessing a total area of 980.70 km 2. secondary data. Primary data was collected This district is contiguous with Benshengul by structured questionnaire. Detailed Gumuz Regional State in the North and information on household and farm West, Digga district in the south, and Guto characteristics, household socio-economic Gida district in the east. It is divided in to 27 and demographic characteristics, location farmers associations and one urban center characteristics and farm management having the capital town named Gallo practices and other related information were (SDFEDO, 2011). collected through interview of sample Climate household heads. Climate is the long-term effect of the sun's The study was conducted in Guto Gida radiation on the rotating earth's varied surface and Sasiga districts, East Wollega Zone of and atmosphere. Agroecological zones are Oromia Regional State. These districts were areas where predominant physical conditions purposefully selected due to the fact that in guide relatively homogeneous agricultural these areas the environment has been land use options. Because of Ethiopia’s degraded largely and the occurrence of location near the equator, elevation has very climate change that affect agricultural strong influence on temperature and, to lesser production during the year 2010 and 2011 in extent, on rainfall. Elevation is the basis for three kebeles of Guto Gida district. traditional agroecological divisions, which Systematic random sampling technique have long been used to characterize different was employed to draw sample of household environments in Ethiopia (CSA and IFPRI, heads. From a total of 50 peasant 2011). Climate can be understood most easily associations in these districts nine peasant in terms of annual or seasonal averages of associations were selected randomly. From temperature and precipitation. Guto Gida these sampled peasant associations based on district is situated between altitude of 1,350 formula by Kothari (2004) 142 households were selected proportionally.

833

Ethiopian Journal of Environmental Studies and Management Vol. 7 Suppl. 2014 East Wollega Zone Ü

Ebantu Kiremu

Gida Ayana Haro

Limu Sasiga District Guto Gida District

Legend East Wellega_District W_NAME Gudeyabila Kiremu Wama Hagelo Boneya Bushe Sasiga Guto Gida

Sibu Sire Ebantu Gida Ayana Nekemte Town Diga Gobu Seyo Gudeyabila Wayu Tuqa Guto Gida Leqa Dulecha Boneya Bushe Haro Leqa Dulecha Wama Hagelo Limu Jimma Arjo Nunu Qumba Nekemte Town Nunu Qumba km Sasiga 010,000 20,000 40,000 60,000 80,000 100,000 Sibu Sire Wayu Tuqa Source: East Wollega Zone Finance and Economic Development Office (2013) Figure 1: Map of the study area

Econometric Model (Maddison, 2006). The model for sample Adaptation to climate change involves a selection assumes that an underlying two-stage process: first, perceiving change relationship exists, the latent equation given and, second, deciding whether or not to adapt by by taking a particular measure. This leads to * y j = x j β + u j ------(1) a sample selectivity problem, since only 1 those who perceive climate change will such that we observe only the binary outcome adapt, whereas we need to make an inference given by the probit model as probit * about adaptation by the agricultural y j = (y j > 0) ------(2) population in general, which implies the use of Heckman’s sample selectivity model 834

Farmers’ Perception and Adaptation to Climate Change ...... URGESSA & AMSALU The dependent variable is observed only if j awareness of farmers to climate change or is observed in the selection equation their risk perceptions. The variables included select in the analysis are level of education, farmers y j = (z jδ + u2 j > 0)------(3) information on climate change, farmer to u ~ N ( 1,0 ) 1 farmer extension, household’s number of u ~ N ( 1,0 ) 2 relatives in development group, household corr u , u = ρ ( 1 2 ) farm income, total family size, sex of Where x is a k-vector of regressors or household, household non-farm income, independent variables that is affect farmers household’s livestock ownership, extension perception and adaptation to climate change, on crop and livestock, household’s credit z is an m vector of regressors, u 1and u 2 are availability, farm size in hectares, distance to error terms. When ρ≠ 0, the standard probit input market, distance to output market, local techniques applied to equation (1) yield agro ecology, increase in temperature, biased results (Deressa et al ., 2008). Thus, decrease in temperature, no change in the Heckman probit provides consistent, temperature, increase in precipitation, asymptotically efficient estimates for all decrease in precipitation, and no change in parameters in such models. Thus, the precipitation. Heckman two stage selection model was employed to analyze the perception and Results and Discussions adaptation to climate change in the Guto Farmers’ Perception and Adaptation to Gida and Sasiga districts. Climate Change For this study, the first stage of the Adaptation to climate change is a two- Heckman probit model considers whether the step process which requires that farmers farmer perceived a climate change; this is the perceive climate change in the first step and selection model. The second-stage model respond to changes in the second step looks at whether the farmer tried to adapt to through adaptation. The analysis of farmers’ climate change, and it is conditional on the perceptions of climate change indicates that first stage, that is, a perceived change in most of the farmers in this study are aware of climate. This second stage is the outcome the fact that temperature is increasing and the model (Deressa et al ., 2008). level of precipitation is declining (Table 1). Variables used in the analysis Different socio-economic and environmental The explanatory variables for the factors affect the abilities to perceive and selection equation include different socio- adapt to climate change. demographic and environmental factors based on the literature on factors affecting the

Table 1: Farmers Perception of Change in temperature and precipitation List of variables Districts Guto Gida Sasiga Total No Yes No Yes No Yes Increase_Temperature 18 51 31 42 49 93 Decrease_Temperature 68 1 67 6 135 7 Nochange_Temperature 66 3 67 6 133 9 Decrease_Precipitation 28 41 24 49 52 90 Increase_Precipitation 60 9 71 2 131 11 Nochange_Precipitation 64 5 70 3 134 8

835

Ethiopian Journal of Environmental Studies and Management Vol. 7 Suppl. 2014 Among the explanatory variables used in the model, 7 variables were significant with respect to outcome equation with less than 10 percent of the probability level while 5 variables were significant with respect to selection equation (Table 2). The variables having a significant effect on adaptation to climate change in the study area are discussed below.

Table 2: Result of Heckman two stage sample selection model (n=142) Farmers’ Perception to Climate Change (Selection Farmers’ Adaptation to Climate Change equation) (Outcome equation) Marginal Effect Marginal Effect List of variables List of variables dy/dx P-value dy/dx P-value Education 0.00349 0.180 Education 0.03757 0.040** Sex 0.04491 0.265 Family_Size 0.01327 0.152 No_of_relatives 0.00160 0.015** Sex 0.00286 0.965 Farm income 0.00000 0.802 Non-Farm Income 0.00001 0.030** Local agroeco -0.16196 0.002*** Livestock_Ownership 0.19014 0.001*** Information 0.18417 0.000*** Extension_on_Crop 0.20033 0.000*** Season_change 0.03442 0.002*** Credit 0.12781 0.001*** Farmer_extension 0.50799 0.009*** Farm_Size -0.01752 0.337 Distance_ Input 0.00153 0.580 Distance_ Output 0.00153 0.431 Increase_Temperature 0.42288 0.000*** Decrease_Temperature 0.13554 0.190 Decrease_precipitation 0.64763 0.000*** District -0.10635 0.036** -0.08341 0.233 ***, ** and * significant at 1%, 5% and 10% level respectively.

The number of relatives is one of the precondition for farmers to take up social capitals which increase the awareness adaptation measures (Maddison, 2006). of the households on their environment. As The agro-ecological setting of farmers expected, households’ number of relatives in influences the perception of farmers to development group was positively related climate change. As expected, different with perception of climate change. One farmers living in different agro-ecological increase in number of relative of household settings perceive the occurrence of climate head raises the probability of perceiving change differently. The result of this study climate change by 0.16 percent. shows as one moves from Kolla to Woina Having access to farmer to farmer dega local agro-ecology the probability of extension increases the likelihood of perceiving the occurrence of climate change perceiving occurrence of climate change by decreases by 16.19 percent. Contrary to 50.79 percent. Information on temperature Deressa et al., (2011) farmers living in Kolla and rainfall has a significant and positive (lowland) perceived more change in climate impact on the likelihood perceiving climate than farmers in Woina dega (mid-land) or change and access to information on climate Dega (high land). change increases the probability of The farmers’ perception of change in perceiving the occurrence of change in duration of season significantly affects climate by 18.42 percent. Access to climate perception of climate change. As an change Information is an important individual farmer observe change in duration of season, his/her probability of perceiving

836

Farmers’ Perception and Adaptation to Climate Change ...... URGESSA & AMSALU change in climate increases by 3.44 percent. (2010) who reported that access to extension District dummy variable negatively and facilities of households had positive and significantly affected perception of climate significant relationship with perception and change. This shows that respondent adaptation of climate change. households in Guto Gida district perceived Resource availability is generally the occurrence of climate change than those expected to positively influence farmers’ to in Sasiga district. adapt climate change. Hence, access to credit Level of education of household took the is expected to have positive relationship with expected sign and its coefficient was farmers’ adaptation to change in climate. significant at less than 10 percent probability Credit availability is one factor that leads level. It had a positive and strong relationship household to adapt climate change. An with the dependent variable showing that increase in access to credit raises the literate household heads were more probability of adaptation to climate change probability to adapt climate change on by 12.78 percent. Similar with this finding average. One level increase in education Charles and Rashid (2007) and Apata et al . raises the probability of adaptation to climate (2009) showed farmers with access to credit change by 3.75 percent. This result is in line have higher chances of adapting to changing with Ayanwuyi et al . (2010) who reported climatic conditions. This result is also in line that education level of households had with Ayanwuyi et al . (2010) who reported positive and significant relationship with that access to credit facilities of households perception of climate change. had positive and significant relationship with Nonfarm income increases the probability perception of climate change and adaptation of adapting the climate change. One birr options. increase in household nonfarm income leads Increment in temperature and adaptation to the increment of the probability of to climate change were hypothesized to be adaptation to climate change by 0.001 related positively. As expected the result of percent. This implies that households with this study shows the direct relationship income may get capital, land and labour. between adaptation to climate change and These factors serve as important factors for perception of increase in temperature. coping with adaptation (Apata et al ., 2009). Perceived change in temperature has So, adaptation to climate change depends on significant effect in the likelihood of availability of income. employing climate change adaptation Livestock ownership is a sign of wealth strategies (ACCCA, 2010). For increment in to farmers (Sofoluwe et al ., 2011). The perception of increase in temperature raises ownership of livestock is also positively the probability of adaptation to climate related to the adaptation of climate change. change by 42.28 percent. Adaptation to An increase in access to livestock ownership climate change and precipitation were raises the probability of adaptation to change negatively related as expected. Perceiving in climate by 19.01 percent. decrease in precipitation raises the As expected, access to crop and livestock probability of adapting climate change by extension has a positive and significant 64.76 percent. impact on adaptation to climate change. Climate Change Adaptation Measures and Having access to crop and livestock Causes of Non Adaptation production increases the probability of From all respondents who adapted the adapting climate change by 20.03 percent. change in climate majority of them (45.33 This result is in line with Ayanwuyi et al . percent) adapted the change through taking

837

Ethiopian Journal of Environmental Studies and Management Vol. 7 Suppl. 2014 soil conservation measures. About 9.33 shows farmers with best access to crop and percent have taken measures of planting crop livestock extension adapt the impact of varieties in order to cope up climate change climate change more. Therefore, the problem. 32 percent of them were government and policy makers should participated on planting trees in reducing the encourage the way farmers get extension on problem caused by climate change. The crop and livestock than ever in order to remaining 13.34 percent of total households increase the farmer’s adaptability of climate who adapted the change in climate were change. participated in irrigation activities in order to The availability of resource affects solve problem faced them through climate farmers to adapt to climate change. The result change. of this research shows that availability of Out of all respondent households who credit affect adaptation to climate change perceived the change in climate, those who positively and significantly. So, the did not adapt the change was not free of government and policy makers should cause. Majority of them (47 percent) did not respond to the way through which farmers adapt because of shortage of labor while 32 get best access to credit and other resources percent of them did not adapt the change in from different saving and credit institutions climate because of lack of money to to increase their way of adaptation to climate undertake the adaptation measures. From all change. those who did not adapt the change in climate Accessibility of information on climate 12 percent of them were absent from change is necessary for the perception of adapting the change because of lack of climate change. Farmers with best information and poor potential for irrigation. accessibility of information on climate change perceive the occurrence of climate Conclusion and Policy Implications change than those without the accessibility of This study examined factors that the information. Therefore, the local influence farmer’s perception and adaptation government, NGOs and policy makers should to climate change and used Heckman two focus on the way farmers get more stage sample selection model to identify the accessibility of information on climate significant variables. change. Farmers level of education, household nonfarm income, livestock ownership, Acknowledgments extension on crop and livestock, household’s The authors thank the editor and anonymous credit availability, perception of increase in reviewers of this journal for their helpful temperature and perception of decrease in comments and suggestions on an earlier precipitation were those variables which version of this manuscript. significantly affect the adaptation to climate change. Similarly, the farmers’ perception of References climate change was affected significantly by ACCCA. (2010). Improving decision-making information on climate, farmer to farmer capacity of small holder farmers in extension, local agro-ecology, number of response to climate risk adaptation in relatives in development group and three drought-prone districts of Tigray, perception of change in duration of season. northern Ethiopia, Mekelle, Ethiopia Farmer’s access to crop and livestock Anemut, B. (2006). Determinants of extension is positively and significantly Farmers’ Willingness to Pay for the affects adaptation to climate change. This Conservation of National Parks: The

838

Farmers’ Perception and Adaptation to Climate Change ...... URGESSA & AMSALU Case of Simen Mountains National Analyzing the Determinants of Park, MSc Thesis in Agricultural Farmers’ Choice of Adaptation Economics, Haramaya University Methods and Perceptions of Climate Apata, T.G., Samuel, K.D. and Adeola, A.O. Change in the Nile Basin of Ethiopia (2009). Analysis of Climate Change GGDFEDO. (2011). Physical and Socio- Perception and Adaptation among Economic Profile of Guto Gida District Arable Food Crop Farmers in South Kothari, C.R. (2004). Research Western Nigeria, International Methodology; Methods and Association of Agricultural Techniques, 2nd Revised Edition, New Economists’ 2009 Conference, Beijing, Age International Publishers, New China Delhi, India Ayanwuyi, E. Kuponiyi, F.A., Ogunlade, and Maddison, D. (2006). The Perception of and Oyetoro, J. (2010). Farmers Perception Adaptation to Climate Change in of Impact of Climate Changes on Food Africa, CEEPA Discussion Paper Crop Production in Ogbomosho No.10, Centre for Environmental Agricultural Zone of Oyo State, Economics and Policy in Africa, Nigeria, Global Journal of Human University of Pretoria. Social Science , 10(7): 33-39 Shibru, T. and Kifle, L. (1998). Bruce, A.M., Richard, M.A., and Brian, H. Environmental Management in H. (2001). Global Climate Change and Ethiopia: Have the National Its Impact on Agriculture Conservation Plans Worrked? Charles, N. and Rashid, H. (2007). Micro- Organization for Social Science Level Analysis of Farmers’ Adaptation Research in Eastern and Southern to Climate Change in Southern Africa, Africa (OSSREA), Addis Ababa, IFPRI Discussion Paper 00714. Ethiopia CSA and IFPRI. (2011). Atlas of Agricultural Sofoluwe, N.A., Tijani, A.A. and Baruwa, Statistics, 2006/07-2010/11, Addis O.I. (2011). Farmers’ perception and Ababa, Ethiopia adaptation to climate change in Osun Deressa, T.T., Hassan, R.M., and Ringler, C. State, Nigeria, African Journal of (2011). Perception of and adaptation to Agricultural Research, 6(20): 4789- climate change by farmers in the Nile 4794 basin of Ethiopia, Journal of SDFEDO. (2011). Physical and Socio- Agricultural Science , 149: 23–31 Economic Profile of Sasiga District Deressa, T.T., Hassan, R.M. Tekie, A., Mahmud, Y. and Claudia, R. (2008).

839