Impact of Climate Variability on Cool Weather Crop Yield ...

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

415

Impact of Climate Variability on Cool Weather CropYield in Ethiopia

Arega Shumetie 1*, Belay Kassa 2, Degye Goshu 3 and Majaliwa Mwanjalolo 4

Keywords: climate variability, crop yield,technical efficiency, small-holder, rainfall, temperature

Received: 07 November 2016,Accepted: 01 January 2017 The research examined effect of climate variability on yield

of the two dominant cool weather cereals (wheat andbarley) in central highland and Arssi grain plough farmingsystems of Ethiopia using eight round unbalanced panel data(1994-2014). The stochastic frontier model result revealed thatproduction inputs for producing wheat and barley in the twofarming system had significant effect. Crop season rainfall in-crement had negative and significant effect on technicalefficiency of smallholders to produce wheat as to the modelresult. Technical efficiency of two crops responded differentlyfor cropping season rainfall variability, in which wheat hadnegative and significant interaction with it while barley hadpositive. Given this, cropping season temperature had significantand positive effect on technical efficiency of both wheat andbarley. Having this into account, yield of the two cropsresponded similarly for changes in production inputs likeworking capital, human labor and fertilizer. In general, rainfallinconsistency at the different stages of the production periodhad strong effect on yield of the two crops. Given this, thestudy forwarded an assignment to plant scientists in order tohave further investigation on how the two crops responded dif-ferently to temperature variability.

Abstract

International Journal of Agricultural Management and Development (IJAMAD)Available online on: www.ijamad.iaurasht.ac.irISSN: 2159-5852 (Print)ISSN:2159-5860 (Online)

1 PhD Candidate in Haramaya University of Ethiopia and Makerere University of Uganda2 Professor Pan African University, Africa Union, Addis Ababa, Ethiopia3 PhD, Haramaya University, College of Agriculture and Environmental Science: School of Agricultural Economics andAgri-business: Department of Agricultural Economics, Ethiopia4 PhD, Makerere University, College of Agriculture and Environmental Science. School of Forestry, Environmental andGeographical Science: Department of Geography, Geo Informatics and Climatic Sciences, Uganda * Corresponding author’s email: [email protected]

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

416

INTRODUCTIONBy the end of the 21st century, climate change

would have a substantial impact on agriculturalproduction and aggravate poverty level (Slateret al., 2007). Intergovernmental Panel ClimateChange (IPCC), 2007) argued that there havebeen widespread changes in extreme temperaturesand longer droughts and frequency of heavyprecipitation over most land areas. Climatechange affects developed and developing nationsdifferently. For example, the latter ones sufferfrom serious food shortage and starvation as ofthe problem while developed ones face relativelysimple problems (FAO, 2008). Land productivityand crop yield have been expected to declinebecause of climate change more or less contin-uously in recent times, and have experiencedsharp decline in some places (Asha et al., 2012).

Agriculture is the most sensitive sector for cli-mate change and variability (Cruz et al., 2007).Climate variability is currently the dominantcause of short-term fluctuation in rain-fed agri-cultural production of developing regions. Insemi-arid and sub-humid areas, rainfall deficiencyreduces crop yields and productivity drastically.Households pursuing crop production as theonly source of livelihood are more affected byclimate variability (Corinne et al., 2004). Theproblem is severe in Sub Saharan African (SSA)region specifically in which crop production isthe main source of generating food and incomefor households. A study done by Wolfram andDavid (2010) argued that yield of many cerealcrops will decline by 10% in 2050 nearly in allSSA countries due to rainfall inconsistency.Fractional variation in one of the climate elementswould have devastating impact on smallholderproduction system as well as overall futurity ofagriculture in most of African nations. SinceEthiopian agriculture is purely rain-fed typeeach variation in either rainfall or temperature,or combination of them would have huge effecton crop yield. Variation in national yields dueto climate variability is common in most of therural areas of Ethiopia (Getnet & Mehrab, 2010).

Ethiopia is one of SSA nations that suffermost from climate change and variability. Agri-culture contributes 40% to the GDP, 84% of

labor force employment and more than 90% tothe export of the country. Agriculture survey in2010 showed that cereal crop production takesthe lion’s share (85.94%) in the overall graincrop production of the country followed bypulses (10.5%). In the same production year,wheat and barley respectively took 19.8% and11.27% proportion from the overall cereal pro-duction. Nearly 66% of the cereals produced inEthiopia were used for household consumption(Central Statistics Agency (CSA), 2010), whichmeans variability in yield of those crops mayhave huge impact on livelihood.

Since agriculture is the crucial sector in em-ployment and livelihoods of Ethiopia, loss ofagricultural productivity due to climate changemay affect the entire economy strongly. Cerealcrop damage sourced from rainfall inconsistencyshowed increment from 2006 to 2008 productionyears. Pest infestation and crop diseases outbreakwere also high in the previously specified timeinterval. Major causes of crop damage in Ethiopiaas to the response of smallholder farmers werehailstone, too much rain, frost and flood all ofwhich are related to climate variability. CSA (2010)reported that hailstone was the major cause ofcrop damage followed by too much rain. Rain-fed agriculture of the country encountered dif-ferent problems sourced from the ongoingclimate variability. Each change in one or moreof the climate variability elements results inhuge production loss and starvation in mostparts of the country. Researchers conductedbefore have not examined crop level responsesto climate variability elements by employingtime series or panel type of data. Thus, this re-search identified the impact of climate variabilityon the cool weather cereal crops (wheat andbarley) yield in Central Highlands and ArssiGrain Plough farming systems of Ethiopia.

MATERIALS AND METHODSDescription of the Data Source

ERHS was conducted in the seven dominantfarming systems of the country except pastoralism.The survey considered 15 villages and covereddifferent aspects like demographic, consumption,production, asset holding, purchases and sales,

Impact of Climate Variability on Cool Weather Crop Yield in Ethiopia ... / Shumetie et al.

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

417

landholdings and livestock ownership. This studyconsidered central highland and Arssi grainplough farming systems of Ethiopia in whichfive villages (out of six for ERHS) were samplesdue to focusing only on wheat and barley yield.

Sampling Unit and Design Farming system was an important stratification

base in selecting sample villages in ERHS(Dercon & Hoddinott, 2011). The survey collecteddata for seven rounds from 1994 to 2009 havingunbalanced period gap. Each survey followed asimilar format of sample units that were alreadydetermined in the first survey in 1994. Thisstudy directly accepted all sample householdsfrom the two sample farming systems, consideredby the previous survey in 2009 from eachvillage, to collect the eighth round data in 2014.To have representative sample households fromthe farming systems, the research consideredthe whole households (502) in the five villages(Table 1), which were considered by ERHS in

2009, to collect the eighth round data in 2014. This farming system took 40% of the total

population of the country based on CSA (1994)and 56.4% sampling share of ERHS in 1994(Dercon & Hoddinott, 2011).

Type and Methods of Data Collection Both primary and secondary data were con-

sidered in this study, in which the former onewas from sampled villages listed in Table 1.The survey in 1994a1 considered about 1,477households and these households have been re-interviewed in 1994b2 as well as in 1995, 1997,1999, 2004 and 2009. The eighth round of datain 2014 could help understand the recent cir-cumstances of sample smallholders from selectedvillages. The primary data collected in 2014adopted the questionnaire (after having amend-ment as to the objective of this research) usedby the previous agency in collecting the seventhround in 2009. Including the previous sevenrounds (1994a-2009) and the recent one in 2014,


Name of village Sample households in 2009 Current sample households Percentage

Debre BerhanKorordegagaYetmen Sirbana Turufe Total

168106518295

502

168106518295

502

33.4721.1210.1616.3318.92100

Table 1Sample Distribution in Each Village

Figure 1. Geographical location of sample villagesSource: Community survey Ethiopian Rural Household Survey 1994

1 There were two round data collection in 1994 and this study consider the first one as 1994a2 The second round data collected in 1994 is named as 1994b

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

418

the research had eight rounded unbalanced paneldata for sample households of the five villages.In addition to the panel data collected by ERHSand this research, the study also used other sec-ondary data collected by different national andinternational organizations.

Methods of Data Analysis Crops have different optimal temperature and

precipitation level for the well performance ofthem. Movement away from these levels maybe damaging for crops, especially in countrieswhere the current temperature and precipitationlevels are already close to the tolerance limit.Production per unit of land yield was the de-pendent variable to analyze the impact of climatevariability on agricultural productivity in Greek(Nastis et al., 2012). Gupta et al. (2012) adoptedCobb-Douglas production function for investi-gating impact of climatic variability on rice,sorghum and millet productivity utilizing paneldata. Thus, this research considered wheat andbarley yield in analyzing impact of climate vari-ability and adopted Cobb-Douglas productionsystem where the model assumes that agriculturalproduction is a function of many variables likecultivated area, oxen power, fertilizers, labors,working capital, rainfall and temperature.

Functional form of the model is:

TPiht=F{ASiht,IFAiht,ALiht,WCiht,DPiht,Rirt,Tirt} (1)

where:-i=1, 2,…., 5 (for the five cereals)h=1, 2, ….,H (the sample household)t=1, 2, ….,8 (the eight round sample years) TPiht = Total Production of sample cereal i for

smallholder h at year t ASiht = crop-wise Area Sown by each small-

holder h at year tIFAiht = Inorganic Fertilizer Application for

each crop by smallholder h at year tALiht = Agricultural Labor of each smallholder

h for each crop at year tDPiht = Draught Power of each smallholder h

for each crop at year tWCiht = the Working Capital budgeted by

smallholder h for each crop at year tRirt = Cropping season rainfall level for each crop

i from the nearest meteorological site r at time tTirt = Average cropping season temperature

for crop i from the nearest meteorological site rat time t

Thus, total production (TP) should be dividedby area sown (AS) of each sample cereal i toget yield, then the above equation (1) wouldbecome: (TP⁄AS)iht=Qiht=F{IFAiht,ALiht,WCiht,DPiht,Rirt,Tirt}(2)

(TP⁄AS)iht=Qiht is yield of crop i for each house-hold h at year t. After computing yield of eachcereal, then the above equation would take thefollowing form:

Qiht=F{IFAiht, ALiht,WCiht,DPiht,Rirt,Tirt} (3)

Thus, equation 3 can be rewritten followingBattesse & Coelli, (1992):

lnQiht =ln(f(Xit,β) exp(Vit-Uit) (4)

And

Uit=ηitUi={exp(-η[(t-Ti)]}Ui (5)

Qiht represents yield of cereal i for the hth

household at the tth period; f(Xit,β) is a suitablefunction of a vector, Xit, of factor inputs associatedwith production of ith cereal in the tth period,and a vector, β, of unknown parameters. Therandom effect, Ui is independent and identicallydistributed (iid) non-negative truncations withN(0,σμ2), Vit is also iid N(0,σV2) random errorsand independent of μi. η is an unknown scalarparameter that shows the type of yield variation.Household’s technical efficiency (TE) tends toimprove, remain constant if not decrease inproducing cereals towards the base year, 2014if η > 0, η = 0 or η < 0, respectively. The timeperiod t = 1, 2……, Ti periods for which obser-vations for the ith household are obtained.

From equation 4 and 5 TE of household h forproducing the ith cereal at time t would be:

TEiht= Qiht/(Qiht*)= (f(Xit,β)exp(Vit-Uit))/ (f(Xit,β)exp(Vit))=exp(-Uit) (6)

To deal with factors affecting smallholder’s


Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

419

efficiencies in cereal production, different so-cio-economic and climate variability elementswere considered on the regression using Gener-alized Least Squares (GLS). This model wasemployed because the TE scores of all the ob-servations were between 0 and 1, but not 0 or 1,which allowed to have linear type of dependentvariable. Since the dependent variable was basedon panel data, the appropriate regression modelwas GLS that consider hetroscedasticity andpanel-autocorrelation problems. The functionalform of the efficiency equation could be:

TEiht = ∑i Xht + vi + εit (7)

where: TEiht = the TE scores of smallholder hfor cereal i at time t

αi = a vector of unknown parameters ought tobe estimated

Xht = a vector of explanatory variables i (i =1, 2, ..., k) for household h

υi= an error term that is iid, N (0, συ2), and it isthe time invariant part of error term

εit = the time variant error term that should beiid, N (0, σε2) independently of υi

RESULTS AND DISCUSSIONCrop yield trend in the two farming systems

Cereal crops that are produced in differentagro-ecology of the country contributed around86% of the national grain production. Majorgrain crops of the country including teff 3, wheatand barley (categorized as primarily cool weathercrops), maize, sorghum and millet are warmweather cereals (Dereje & Yilma, 2003).

Table 2 shows that overall average yield ofeach crop was better in Arssi grain plough as

compared to central highland farming system.The test results also revealed that yield of wheathad a significant difference in the two farmingsystems (p<0.01), but there was no significantdifference in yield of barley (p-value = 0.317)for the sample production years.

There was no wider gap in the overall yield ofsampled crops, and ups and downs of themhappened mostly on the same production year.For instance, in 1999, there was yield reductionfor both crops and this overall reduction alsohappened in 2009 exactly after ten years forwheat. Rainfall inconsistency was the main reasonfor those drastic yield reductions as to the viewof respondents.

Yield of the two crops showed drastic incrementin central highland for production years of 2009and 2014. After 1999, yield of the two cropsshowed better and successive improvement incentral highlands than Arssi grain plough farmingsystem. Table 3 reveals that there was a significantdifference in yield of wheat and barley for mostof the production years.

Average level difference of the two farmingsystems in yield of the two crops was low forsome of the production years. Variability in pro-duction of the two crops revealed that wheat andbarley respond similarly, though it is not statisticallyproved, for each production input variation.

Trend of crop season rainfall and temperaturein the two farming systems

Meteorological data collected from the nearestsites of each sample village of the two farmingsystems revealed that starting from 1994 pro-duction year rainfall condition of some sitesshowed similar trend in the ups and downs, es-


Description Crop type Central highland Arssi p-value Combined Average

Area in hectare

Production in kg

Yield in kg/haYield in kg/ha

Barley Wheat Barley Wheat BarleyWheat

0.910.43

736.50405.78939.56

1068.56

0.300.48

321.58673.38978.72

1262.20

0.0000.0000.0000.0000.3170.000

0.720.44

617.30497.57950.811134.97

Table 2Average Crop Yield in the Two Farming Systems

3 It is an Ethiopian indigenous crop produced only in the country

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

420

pecially in those that have geographic proximity.Rainfall trend of sites in Arssi grain ploughfarming system was lower as compared withothers in central highlands.

Crop seasonal rainfall in central higland farmingsystem was higher in all of the production yearscompared with the other farming system, and itshowed greater inconsistency in some of theproduction years registering the highest amountin 2010 (see Figure 2). Rainfall gap betweenthe two farming systems became wider in recentyears, mostly after 2006. There was cyclicaltype of movement in rainfall of central highlandsfarming system having reduction after three tofive years of interval. Though it was not clearenough like the central highlands’, but rainfalltrend of Arssi grain plough also had cyclicaltype of movement to some extent and exhibited

extrem reduction in few years like 2002 and2012 exactly after ten years.

Crop season temperature of Arssi grain ploughfarming system was consistently higher thancentral highland, and some ups and downs happenedmore frequently in the latter farming system. Testof equality of variances showed significant dif-ference in crop season temperature of the twofarming systems having p-value of 0.004.

Model Result for Stochastic Frontier The dependent variable was in physical quantity,

yield, then explanatory variables should be alsoin quantity but not dummy type non-input vari-ables. This research employed different equationsto regress yield of wheat and barley that aremost common cereals in the two farming systems.The coefficient of μ (mu) was significant for


YearWheat (kg/ha)

p-valueBarley(kg/ha)

p-valueCentral highland Arssi Central highland Arssi

1994a199519971999200420092014

Average

901.91803.40

1216.251112.001181.621236.582091.491220.46

895.57970.39

1555.67921.361104.091308.431460.481173.71

0.0000.0000.0180.0290.0000.0200.000

1187.201169.93995.06

1028.69987.76

1373.192322.041294.84

1043.74942.36

1242.191003.68869.021185.561178.981066.50

0.0120.0000.0560.0030.3140.0000.000

Table 3Crop Yield Trend of the Two Farming Systems

Source: Data from Ethiopian Rural Household Survey (1994a-2009) and researcher in 2014

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

YearFigure 2. Crop season rainfall of sample farming systemsSource: Ethiopian National Meteorology Agency, 2014.

1991

1992

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

421

the time variant frontier model of the two cereals(Table 4), which showed presences of timevariant efficiency variability as well as fitnessof the model, Stochastic frontier. The returns toscale of producing those cereals in the twofarming systems were not in constant returnssince the coefficient of μ was significant.

The relatively large value of μ for wheattestified strong variability of the returns to scalefor the across time (Table 4). The value of η(Eta) in the following model result revealedsignificant production inefficiency reduction to-wards the base year, 2014. The inefficiency

level decays towards the base year, since valueof η (Eta) is positive for the two cereals.Technical inefficiency reduction for barley wasthe higher than wheat (Table 4). The inefficiencyfor those cereals had significant time variantcomponent that was normally distributed witha constant mean and variance level. Householdsin the two farming systems were highly inefficientin producing wheat, wherein they had unutilizedpotential of 61.50%. Households in the twofarming systems could increase wheat yield bymore than 60% from the given inputs at the ex-isting technology.


Description Wheat Barley

Coefficient Standard error Coefficient Standard error

Ln man equivalentLn Fertilizer Ln Oxen Ln Capital ln Cropping season rainfallln Cropping season temperConstant Mu (μ)Eta (η)Ln sigma2Ilgt gammasigma2Gammasigma_u2sigma_v2Log likelihood P-value

0.129***

0.098***

0.050*

0.041***

-0.124*

0.984***

3.576***

0.901***

0.023***

-0.161***

-0.919***

0.8520.2850.2430.609

-2870.60.000

0.0470.0120.0280.0070.0670.2140.8460.1910.0040.0440.154

0.105**

0.082***

0.0270.028***

0.332***

0.564***

2.764***

0.466***

0.056***

-0.637***

-2.774***

0.5290.0590.0310.498

-2089.60.000

0.0430.0110.0260.0070.1080.1621.0430.1510.0110.0370.317

Table 4Time Variant Frontier Model Result

Note: ***, ** and * implies 1%, 5% and 10% level of significance, respectively Ln represents natural logarithm Ilgt implies inverse logit Source: Model result, 2016

Crop season

temperature (°C)

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

Year

1991

1992

Figure 3. Crop season temperature of the two farming systems Source: Ethiopian National Meteorology Agency, 2014

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

422

The dependent variable in the random effectGLS regression model was generated based on thestochastic regression result presented below. P-value of each model result, on the last row of Table5, indicated that the models for regressing technicalefficiency were best fit. The independent variableswere different from zero and had partial contributionin the technical efficiency variation. The test valueenable to reject the null hypothesis with high con-fidence level of above 99.99% and each of the co-variate had individual effect, either significant orinsignificant, on the dependent variable.

Estimation result of technical efficiency 4The technical efficiency scores computed from

frontier model result were between 0 and 1, andthis revealed that none of the sample smallholderwas either at the frontier level or producing 0from the given input combination. The wholesample households were technically inefficientin producing the two crops. Since there werenot censored observations, the study rejectingtobit model instead the best alternative modelfor linear type of observation GLS was employed.

The regression result presented in Table 5 con-sidered three broadly categorized types of ex-planatory variables that affect smallholders’ technicalefficiency in producing the two crops. Wheat andbarley responded similarly for variation in someof the covariates considered as an explanatory.

DISCUSSION Interaction of crop yield with crop seasonrainfall and temperature

The main crop season harvest for some house-holds in Arssi grain plough in 1994a was verylow and it was the worst when compared withprevious five years (Dercon & Hoddinott, 2011).The study also showed that in 2009 productionyear there was a poor distribution of rain formany of the households in the two farming sys-tems, which resulted in huge reduction in yieldof some crops. Households in Arssi grain ploughfarming system were suffering from rainfall in-consistency in a great extent and they tried tocollect crops they needed by covering largercropland. Some of the problems related torainfall condition got extremity and drastic in-crement in their frequency of happening in eachsample village of the two farming systems. Forinstance, many households in the two farmingsystems were suffering from untimed rainfallnear harvesting period in 2009 production season.

The optimum and maximum crop season tem-perature for producing the two cool weathercrops (wheat and barley) respectively were 25cand 30.2c in the sampled farming systems ofthis research. Given this, maximum crop seasonrainfall of the two farming systems was 1321mm.For better yields, water requirements should be350-500mm depending on climate and length of


Variables description Wheat Barley

Coefficient Standard error Coefficient Standard error

Summer rain on timeHarvesting time rainPestsShocks Family size Age of headLn TLUCredit access Off -farmConstantP-value

0.004-0.142***

-0.090***

-0.050***

0.029***

0.051***

-0.009***

0.036***

-0.034***

-0.878***

0.000

0.0080.0080.0090.0070.0080.0110.0030.0080.0060.053

0.159***

-0.178***

-0.123***

0.0098-0.081***

0.061***

0.033***

0.095***

-0.040**

-0.640***

0.000

0.01190.0120.0150.0120.0120.0160.0080.0120.0140.080

Table 5Determinants of technical inefficiency, GLS regression result

Note: The dependent variable was technical efficiency in percentThe coefficients showed two dimensional changes

Source: Model result, 2016

4 To avoid complications for readers the dependent variable was technical efficiency but not inefficiency

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

423

growing period, which is an amount lower thanthe average level (676.42mm) in the sampledfarming systems of this research. Temperatureincrement may reduce yield and thereby exacerbatevulnerability in food supply (Joshi et al., 2011).

Rainfall inconsistency in the onset and cessationperiod showed greater variation from 1999 to2009 within 10 years as per the data collected.This inconsistency and other elements of climatevariability may be the main reasons for thelower crop production. In 1995 households inthe two farming systems suffered from problemsrelated to climate and weather variability inwhich 34.4% of the households who producewheat were affected moderately or drastically.This resulted in a drastic reduction in yield forsome households in central highlands farmingsystem. Spearman correlation test of rainfall withwheat yield showed a significant interactionbetween them having p-value of 0.094 in centralhighland farming system. Spring rainfall is becominghighly unreliable for households in the two farmingsystems because of greater variability and sometimescomplete absence, which frequently result in lossof yield (Dercon & Hoddinott, 2011). Samplehouseholds replied that spring season productionis usually risky because of variability, delay orabsence of rain.

Findings of Asha et al. (2012) revealed thatreduction in rainfall was the major reason foryield reduction followed by pests and diseasesoutbreak; and changes in temperature andseasonal patterns of rainfall quoted as otherreasons for yield reduction. In the same fashion,many households in Arssi grain plough farmingsystem faced insufficient rainfall for their cropproduction and this resulted in serious cropdamage in some production years. Overall rainfall

shortage and poor distribution were the problemsthat happened in the two farming systems fre-quently miffed many of the sampled households(Table 6).

Discussion with focal persons in Arssi grainplough farming system indicated that there wasproduction of short seasoned crops in autumnin the last 15 to 20 years, but recently it is notpossible due to rainfall inconsistency. Currently,households of the farming system produce onlyin the main cropping season because of autumnrainfall inconsistency. The number of householdsmiffed by inconsistency in rainfall increased in2009 production year especially those who pro-duce wheat.

Model Result Discussion Discussion on determinants of yield

Inorganic fertilizer had significant and positiveeffect on yield of the two crops as to the modelresult (Table 4), which means small unit incrementin its application could enhance yield significantly.Sibiko et al. (2013) had similar finding aboutfertilizer usage on bean productivity. Any fertilizerusage increment could enhance wheat, maize, barleyand sorghum productivity (Ajay & Pritee, 2013),which was a finding corroborated with the abovemodel result for all cereal crops. The modelresult in Table 4 shows that inorganic fertilizerand working capital had significant and positiveeffect on cereals’ yield for smallholders in bothcentral highlands and Arsi grain plough farmingsystems of Ethiopia. Rahman & Umar (2009)and Ehsan et al. (2012) also claimed that thereis positive relationship between wheat yieldand fertilizer application, which was in supportof the model result in Table 4. Increasing theavailable capital for purchasing inputs like seed,


Variables description

Number of households Percentage change

In 1999 In 2009Central highland Arssi

Central highland Arssi Central highland Arssi

Wheat Barley Total

2026

150

151

117

2212

156

103

44

10.00-53.85

-33.33200.00

Table 6Number of Households Reported About Rainfall Inconsistency

Source: Own data in 1999 and 2009

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

424

pesticides and insecticides resulted in significantcereal yield increment as to the model result.

Oxen power and human labor had similareffect on cereals yield in the study area. Theircomplement behaviors may be the reason behindhaving similar interaction between the twosources of power and cereals yield in the twofarming systems. Cereals are laborious cropsand smallholders in the two farming systemsmanage them by human and animal powerstarting from plowing to harvesting, thus eachchange in the available labor had significanteffect on yield of the two crops. Smallholders’crop yield variation may not purely climateforcing that is determined solely by rainfall, butalso by the use of yield-improving externalinputs (Amikuzuno & Donkoh, 2012). Similarly,Akinseye et al. (2013) confirmed that optimumyield performance of any crop do not dependonly on climatic parameters. Smallholder farmersin less drought-stricken regions may improveland productivity through budgeting more laborand other production inputs (James et al., 2008),which was a finding similar to Table 4 regardingeffect of human labor and draught power. Sincesmallholders of the study area are purely dependenton animal and human power for producing theircrop, this positive and significant effect of theinputs on cereal yield was expected.

Cropping season temperature increment hadpositive and significant effect on yield of wheatand barley as to the model result (Table 4). Thisindicated that temperature variation significantlychanged yield of the two crops. If croppingseason temperature of the area increased by1%, yield of wheat and barley would increaseby 0.98% and 0.56%, respectively, wherein theformer crop was more sensitive for each tem-perature variation. This finding was similar toJoshi et al. (2011), wherein cropping seasonmaximum temperature increment has positivecontribution on wheat and barley yield variation.Similarly, Sommer et al. (2012) found thatwheat yield would increase in the future withthe projected climate change in the semiaridzone of Tajikistan.

The result in Table 4 was in contradiction

with Geethalakshmi et al. (2011), who identifiedthat mean temperature increment reduced barleyyield. Liangzhi et al. (2009) argued that impactof growing season temperature on wheat pro-ductivity is region specific, wherein effect of theproblem is relatively lower in China as comparedto the other regions. Successive temperature in-crement would not affect crop yield if there wereextensive rainfall or water supply to curve downthe moisture loss of crops (Kutcher et al., 2010).This indicated that positive interaction betweencropping season temperature and yield of thosecereals might be due to the excessive croppingseason rainfall of the study area. The rampantrainfall of cool and moderate agro-ecology ofEthiopia may reduce negative effect of the tem-perature rise. In the other direction, successivetemperature increment may eradicate waterlog-ging problem of the heavy summer rainfall andenhance wheat and barley yield. Thus, it maybe easy to conclude that yield of the two coolweather crops would increase due to continuouscropping season temperature increment.

Discussion on determinants of technicalefficiency (TE) 5

The regression result presented in Table 5considered three broadly categorized types ofexplanatory variables that affect smallholders’TE in producing cereal crops. Wheat and barleyresponded similarly for variation in some of thecovariates considered as an explanatory. Timelyonset of summer rainfall in the two farmingsystems significantly increased smallholders’TE of producing barley, which was in supportwith Inoussa (2010) who claimed that precipi-tation intensity affected sorghum yield signifi-cantly. Consistent availability of rainfall startingfrom the beginning may create auspicious envi-ronment for producing the crop efficiently. Theadequate rainfall at the beginning of summermay enable smallholders to sow cereals on time,which would enhance their TE in producingcool weather crops.

TE of smallholders in producing those cerealsresponded similarly to harvesting season rainfall.Frequent happening of this problem in harvesting


4 To avoid complications for readers the dependent variable was technical efficiency but not inefficiency

Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

425

season could result in serious yield variabilitythat might decrease smallholders’ TE as to themodel result (Table 5). The two cereal cropswere strongly responsive to this problem. Small-holders’ TE in producing barley and wheat re-spectively, decreased by 17.8% and 14.2% ifthere was harvesting season rainfall. The coef-ficients indicated that barley and wheat productionwere sensitive to one of the dimensions ofclimate variability, harvesting season rainfall.

Climate change will reduce crop productivityup to 14% by 2020 (Srivastava et al., 2010),and yields are likely to be affected even morein 2050 and 2080, which was a finding highlyrelated to Table 5. Outbreak of pests and diseasessignificantly decreased TE of households in thestudy area for producing the two cereals, whereinbarley was more sensitive (Table 5). In identifyingimpacts of current climate variability, Kettlewellet al. (1999) showed that heavy rainfall couldresult in fungal disease outbreak that wouldfinally reduce yield.

The demographic related factors including ageof the household head had similar effects on small-holders’ TE of producing the two cereals. Experi-enced household heads would have significanteffect on TE of smallholders in the two farmingsystems in producing cereals. Having experiencedhousehold head would significantly increase TEof households in producing the two crops, whichwas a finding in support of Hardwick (2009),Zenebe et al. (2012) and Sibiko et al. (2013).One of the critical crop production input, humanlabor, may be easily available if the householdhad larger family size, which would improveTE of producing cereals. Family size incrementmay be a source of labor power for the cropproduction that would enhance TE of wheat,which was a finding corroborated with Rahmanand Umar (2009). Family size increment hasboth productive and consumption effect, whereinTE of households in producing cereals wouldbe depending on magnitude of the two effects.The latter effect may be the reason for havingnegative effect of the covariate on TE of pro-ducing barley in the study area.

Given those covariates, the study also consideredother economic factors as an explanatory variable

that have significant effect on smallholders’ TEin producing the two cereals. Household’s TEof producing barley responded positively to thelivestock holding level. Farmers with more live-stock units, which are ready to convert intocash, can be able to buy modern inputs thanthose that own fewer (Fantu et al., 2011). Pos-session of large livestock could significantlyincrease TE, which may be due to having addi-tional income and labor for the crop production(Table 5). Participation of households in off/non-farm activities had negative and significanteffect on TE of producing wheat and barley incentral highlands and Arsi grain plough farmingsystems of Ethiopia, which may be due to laborcompetition between the crop production andthose activities.

CONCLUSIONS AND RECOMMENDATIONSLong run crop season rainfall of Arssi grain

plough farming system was becoming lowerwith the passing of time and there were frequentups and downs recently in very short periodgaps. Similarly, crop season temperature ofcentral highland farming system showed con-tinuous increment with the passing of time. Ex-treme low and high rainfall as well as severedroughts also became common in the twofarming systems. These all notify that the globalproblem, climate variability, is also happeningin Ethiopia and it would have devastating effecton poor smallholders who have nature dependentlivelihood.

Data from focus group discussion indicatedthat wheat and barley production showed con-tinuous improvement in the cool weather areasof central highland farming system especiallythe former one exhibited successive incrementimplying that the weather condition is becomingauspicious for its production.

Agricultural inputs like inorganic fertilizerand human labor are critical in improving pro-ductivity of the two crops in the two farmingsystems. However, successive input price in-crement was a strong threat to apply the inorganicfertilizer as to the standard level. Sample small-holders in the study area affirmed that thecurrent time sky rocketing input price rise is


Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

426

going out of their capacity to purchase the pro-duction inputs.

Cropping season temperature increment posi-tively affected yield of producing wheat andbarley. Thus, successive increment of this inputimproved yield of the tow cereals significantlyin the two farming systems of Ethiopia. Furtherincrement of cropping season rainfall from thecurrent level resulted in mixed effect on technicalefficiency of the two cereals for households inthe two farming systems of Ethiopia.

The rainfall inconsistencies especially in har-vesting time, and outbreak of pests significantlyaffected smallholders’ technical efficiency inproducing wheat and barley. Yield of the twocrops significantly affected by those climatevariability elements.

Source of additional working capital like creditaccess have positive and significant effect onyield of the two cereals in the study area.

Smallholders should have easy access to agri-cultural inputs like inorganic fertilizer anddraught power at an affordable price to enhanceapplication of them as to the recommendedlevel for improving yield of the two crops.

The current level of crop season rainfall andtemperature may be above optimum of wheatand barley production in the near future eventhough yield of the two crops currently improved.Thus, there should due attention to keep thetemperature on the optimum level.

There should be continuous effort from boththe government and private stakeholders to reducenegative and significant effect of pest and diseaseoutbreak in order to have better yield from thetwo crops in the two farming systems.

To have a better yield of the two crops in thesample farming systems there should be due at-tention from the concerned body to keep themoisture on its normal level through reducingany form of rainfall inconsistencies at thedifferent stages of the crop production.

Mechanisms like credit access and diversifiedsource of income should be expanded in thestudy area to have alternative source of capitalfor producing wheat and barley.

Since rainfall inconsistencies were criticalbottlenecks in producing the two crops, there

must be brainwashing and awareness creationfor smallholders to make them clairvoyant aboutprecautionary measures.

There should be further investigation on thewheat yield both in laboratory and extendedfield work to examine why it had negative in-teraction with the crop season rainfall.

ACKNOWLEDGEMENTThe principal author has strong appreciation

for the co-authors’ unreserved effort to improvethe quality of the article. Additionally, HaramayaUniversity of Ethiopia and Makerere Universityof Uganda have significant contribution, whichshould be acknowledged.

REFERENCESAjay, K., & Pritee, S. (2013). Impact of climate

variation on agricultural productivity and food security in Rural India. Discussion Paper No. 2013-43: http://www. economicsejournal. org/economics/discussionpapers/

Akinseye, F.M., Ajayi, V.O., & Oladitan, T.O. (2013). Assessing the impacts of climate variability on crop yield over Sudano- Sahelian zone in Nigeria. International Journal of Agricultural Science,1(7), 91-98. Retrieved from http:// www.accessinterjournals.org/aijas

Amikuzuno, J., & Donkoh, S.A. (2012). Climate variability and yields of major staple food crops in Northern Ghana: Tamale, Ghana. African Crop Science Journal, 20 (2), 349 – 360.

Asha, L.K.V., Gopinath, M., & Bha, A.R.S. (2012). Impact of climate change on rain-fed agriculture in India. International Journal of Environmental Science and Development, 3(4), 368-371.

Baltagi, B. H. (2008). Econometrics. 4th Edition. Springer-Verlag Berlin Heidelberg: Syracuse University. USA.

Battesse, G.E., & Coelli, T.J. (1992). Frontier Production Functions, Technical Efficiency and Panel Data: With Application to Paddy Farmers in India. The Journal of Productivity Analysis, 3, 153-169

Corinne,V., Jere, L. G., Christian, J., Robert, Q., & Rigoberto, E. (2004). Coping and adapting to climate variability: The Role of Assets, Networks, Knowledge and Institutions.


Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

427

http://citeseerx.ist.psu.edu/viewdoc/download; jsessionid=1255C45E22E18E1CBE48CDC283BE36AF?doi=10.1.1.201.7357&rep=rep1&type=pdf

Cruz, R.V., Harasawa, H., Lal, M., Wu, S., Anokhin, Y., Punsalmaa, B., Honda, Y., Jafari, M., Li, C., & Huu, N. (2007). Asia: Climate Change 2007: Working Group II: Impacts, Adaptation and Vulnerability. IPCC (Inter-governmental Panel on Climate Change) Contribution of Working Group II to the Fourth Assessment Report. https://www.ipcc.ch/ publications _ and_data/ar4/wg2/en/ch10.html

CSA (Central Statistics Agency) (2010). Agricultural sample survey 2009/10 (2002 e.c.) Volume VII Report on Crop and livestock product utilization (private peasant holdings, meher season. Statistical bulletin 468 Addis Ababa, Ethiopia.

Dercon, S., & Hoddinott, J. (2011). The Ethiopian Rural household surveys 1989-2009: Introduction. International Food Policy Research Institute. Addis Ababa: Ethiopia.

Dejene A., & Yilma S. (2003). Causes and variability of Ethiopian agriculture: modeling the relative importance of environment factors, external shocks, and state policies 1980–1997. Regional and Local Development Studies (RLDS), Addis Ababa University, Ethiopia

Ehsan, N., Mohammad, B., Mohammad, R. J., Paul, S., & Alireza, F. (2012). Agro-ecological zoning for wheat (Triticum aestivum), sugar beet (Beta vulgaris) and corn (Zea mays) on the Mashhad plain, Khorasan Razavi Province. The Egyptian Journal of Remote Sensing and Space Sciences, 15, 99–112.

Elodie, B. (2012). The impact of climate change on crop yields in Sub-Saharan Africa. American Journal of Climate Change, 1, 1-13

FAO (Food and Agriculture Organization) (2008). The state of food insecurity in the world.Food and Agriculture Organization of the United Nations, Rome, Italy.

Geethalakshmi, V., Lakshmanan, A., Rajalakshmi, D., Jagannathan, R., Sridhar, G., Ramara, Bhuvaneswari, A. P., Gurusamy, K. L., & Anbhazhagan, R. (2011). Climate change impact assessment and adaptation strategies to sustain

rice production in cauvery basin of tamil nadu. Current Science, 101(03), 342-347.

Getnet, A. Z., & Mehrab, M. (2010). Implications of land policies for rural-urban linkages and rural transformation in Ethiopia. ESSP II Working Paper 15. Development Strategy and Governance Division, IFPRI Ethiopia Strategy Support Program (ESSP), Ethiopia.

Govinda, B. (2013). Effect of precipitation and temperature variation on the yield of major cereals in Dadeldhura District of Far Western Development Region, Nepal. International Journal of Plant, Animal and Environmental Science, 3 (1).

Gupta, S., Sen, P., & Srinivasan, S. (2012). Impact of climate change on Indian economy: Evidence from food grain yields. Centre for Development Economics Working Paper 218, New Delhi.

IPCC (Intergovernmental Panel Climate Change) (2007). Agriculture in climate change 2007: Mitigation. Working Group III Contribution to the Fourth Assessment Report of the IPCC. Cambridge: Cambridge University Press

Jaiswal, J.P. (2009). Climatic requirement for Wheat- temperature. Retrieved from http://agropedia. iitk.ac.in/content/climatic-requirement-wheat-temperature

Joshi, N., Maharjan, P., Keshav, L., & Piya, L. (2011). Effect of Climate Variables on Yield of Major Food-crops in Nepal: A Time-series Analysis. Journal of Contemporary India Studies,1:19-26

Kettlewell, P. S., Sothern, R. B., & Koukkari, W. L. (1999). UK wheat quality and economic value are dependent on the North Atlantic oscillation. Journal of Cereal Science, 29, 205–209.

Lobell, D.B., Burke, M.B., Tebaldi, C., Mastrandrea, M.D., Falcon, W.P., & Naylor, R. L. (2008). Prioritizing climate change adaptation needs for food security in 2030. Science, 319, 607-610.

Lott, N., Ross, T., Smith, A., Houston, T., & Shein, K. (2011). Billion dollar U.S. Weather disasters, 1980 - 2010. National Climatic Data Center, Asheville, NC, USA. Retrieved from http:// www.ncdc.noaa.gov/oa/reports/billionz.html

Morton, J. (2007). The impact of climate change on smallholder and subsistence agriculture. Proceedings of the National Academy of Sciences,


Inte

rnat

iona

l Jou

rnal

of A

gric

ultu

ral M

anag

emen

t and

Dev

elop

men

t, 7(

4), 4

15-4

28, D

ecem

ber 2

017.

428

How to cite this article:Shumetie, A., Kassa, B., Goshu, D., & Mwanjalolo, M. (2017). Impact of climate variability on coolweather crop yield in Ethiopia. International Journal of Agricultural Management and Development,7(4), 415-428.URL: http://ijamad.iaurasht.ac.ir/article_527242_67f679d52d7a3a105ca64ab0749e9065.pdf

USA 2007;104:19680-19685.Nastis, S.A., Michailidis, A., & Chatzitheodoridis,

F. (2012). Climate change and agricultural productivity. African Journal of Agricultural Research, 7 (35), 4885-4893.

Samarah, N.H., Alqudah, A.M., Amayreh, J.A., & McAndrews, G.M. (2009). The effect of late-terminal drought stress on yield components of four barley cultivars. Journal of Agronomy and Crop Science, 195 (6), 427-441

Sara, L. (2012). Climate change, weather variability and food consumption: A multidisciplinary study of Rural Uganda. Catholic University of the Sacred Heart. Via Emilia Parmense 84, 29122 Piacenza, Italy.

Schlenker, W., & Michael, J. R. (2009). Nonlinear temperature effects indicate severe damages to U.S. crop yields under climate change. Proceedings of the National Academy of Sciences, 106 (37),15594–15598.

Slater, R., Peskett, L., Ludi, E., & Brown, D. (2007). Climate change, agricultural policy and poverty reduction; how much do we know? Natural Resource Perspectives, 109.

Wolfram, S., & David, B.L. (2010). Robust negative impacts of climate change on African agriculture. Environmental Research Letters, 5 (1): IOP Publishing: http://iopscience.iop.org/1748-9326/5/1/014010

Wollenweber, B., Porter, J. R., & Schellberg, J. (2003). Lack of interaction between extreme high-temperature events at vegetative and reproductive growth stages in wheat. Journal of Agronomy and Crop Science, 189, 142–150. (doi:10.1046/j.1439-037X.2003.00025.x) World Bank, 2004. Memorandum May 24.

Zerihun, G. (2012). An analysis of the impacts of climate change on crop yield and yield variability in Ethiopia. PhD Dissertation: Graduate School of Social Sciences: Department of Economics and Management. University of Trento: Italy


Impact of Climate Variability on Cool Weather Crop Yield ...

Documents

Transcript of Impact of Climate Variability on Cool Weather Crop Yield ...