Int. J. Agron. Agri. R.

International Journal of Agronomy and Agricultural Research (IJAAR)

ISSN: 2223-7054 (Print) 2225-3610 (Online)
http://www.innspub.net
Vol. 18, No. 2, p. 8-16, 2021
RESEARCH PAPER OPEN ACCESS

Wheat crop responds to climate change in rainfed areas of

District Mansehra, Khyber-Pakhtunkhwa- An Econometric
analysis
Rizwan Ahmad1, Muhammad Zulfiqar*1, Himatullah Khan1, Jawad Ali2, Arjumand
Nizami2, Dilawar Khan3

1
The University of Agriculture, Peshawar, Pakistan
2
Helvetas Swiss Intercooperation, Pakistan
3
University of Science and Technology, Kohat, Pakistan

Article published on February 28, 2021

Key words: Climate change, Rainfed, Wheat, Productivity, Mansehra

Abstract

Agriculture in many ways is affected by climate change and has impact for productivity of crops particularly in
rainfed areas. Climate change related research remained a poorly investigated area in KP and instant study filled
that gap by investigating impacts of change in climate on farm productivity. The secondary data, spread over 30
years from 1984 to 2013 pertaining to temperature, precipitation, area under cultivation and yield of crops was
collected. Analytical models used are ARDL Model. The results pertaining to impact of temperature and
precipitation on wheat yield suggest long run relationship among the variables. Temperature is positively and
significantly related in Mansehra. The precipitation is positively and significantly related. Short run relationship
implies that around 100% deviations from long-term equilibrium are adjusted every year in case of Mansehra.
The results wheat areas suggest long run relationship among the variables based on F Statistics value. Both
temperature and precipitation are positively and significantly related to the area under wheat in the long run in
case of Mansehra. Based on objectives of the research study and field findings recommendations offered include;
farmers awareness drive, policies to promote adaptation measures, enhancing farmers’ adaptive capacity to
strengthen local resilience, participation of farming community in formulation of policies, making meteorological
information available to farmers, Design research plans to evolve crops varieties addressing changing climatic
challenges, construct water harvesting structures for high efficiency irrigation and further research to estimate
range of temperature and precipitation within which crops under study perform better.
* Corresponding Author: Muhammad Zulfiqar  [email protected]

Ahmad et al. Page 8

 Int. J. Agron. Agri. R.

Introduction The climatic changes have led to increased

Agriculture in many ways is affected by climate vulnerabilities to agriculture, forestry and water
change and has impact for productivity of crops resources upon which a large part of the economy and
(Ziervogel 2009; Falco 2011 and Chandrasiri 2013). livelihood depend. Pakistan being a developing country
The negative effects of climate change are considered is likely to face severe challenges on account of
too high such as recurrent droughts, flooding, heat economic and social development, environmental
waves, cold waves and land deterioration (Adger sustainability and land degradation. The adverse
2003 and Rosenweig and Hillel 1998). In rainfed outcome of climatic changes are already been felt in
areas, climate is key to crop productivity, thus Pakistan because of recurring droughts, increased
droughts, floods or extreme temperatures could intensity of floods and un predictable weather patterns
devastate agriculture sector. This devastating and changes in behavior of agricultural production
situation could result into poor and unsustainable system. Within the country, regions with arid land will
livelihood of the communities depending on be on the top of affected list. Such arid regions will face
agriculture in the rainfed areas (Calzadilla 2009 and a whole brunt of adverse impacts in terms of their
Ahmad and Zulfiqar 2019). According to Climate socio-economic conditions, physical, environmental as
Change Synthesis Report (2014), the limit of global well as biological resources (Khan 2012).
surface temperature increases to 1.5 degrees Celsius
and continues to rise beyond 2100 in all scenarios Agriculture is an important source of livelihood for
except lower emissions scenario. To come out with the rural communities of rainfed areas. The majority
evidence about adverse effects of climatic changes on of the 35.50 million masses living in Khyber
agriculture including ecosystem services became a Pakhtunkhwa are relying on rain-fed agriculture
critical challenge for development practitioners system. The food availability for such areas is
around the globe. It is internationally recognized that dependent on agricultural productivity. The crop
decreased productivity of agriculture sector means productivity/yield is dependent on timely rainfall
GDP level loss, decrease in income and consumption specially in rainfed areas. District Mansehra selected
for the most vulnerable population and deterioration for instant research study is one of the 34 districts of
in households’ welfare. Thus influence of climate Khyber Pakhtunkhwa with almost 80% rainfed area
changes continuously poses major threat to rural out of its total 80,747 ha of cultivated area (GoKP
livelihoods (Kangalawe and Lyimo 2013). 2017). In view of the fact that a change in climatic
patterns has been occurring and will continue even in
Climatic variations and resultant weather patterns future, underlines the importance of understanding
have already marked as negative effects on as how farmers perceive these changes and how they
agricultural resources, food production and food adapt to these variations. The perceptions of climate
security globally (Diao, 2010). Climate change may change play a vital role in farmers’ decisions of
alter rainfall patterns, drought cycles and more adopting protective measures against climate
frequent severe weather patterns and increased extremes and therefore, are important factors which
diseases and agricultural pests (Yanda P.Z. 2010 and need to be considered. Minimize the negative impacts
Hewitson B.C. 2010). Moreover, consequent to of change in climate needs awareness of the farmers
climatic changes, productivity of farms decreased in and policy makers. Therefore, through instant study,

general (Makungwa, 2010 and Parry 1999). the impact of climate change in terms of precipitation
and temperature on area and yield of wheat crop is
Being an agricultural based economy, Pakistan is investigated using time series data. The outcome and
under immediate risk because of global climatic recommendation of the research study will have
variability. The country ranked 12th amongst the important implications for future policies addressing
most vulnerable states expected to have brunt of the climate change within the country and farmers
climatic changes (Global climate risk index 2015). behavior towards crop husbandry.

Ahmad et al. Page 9

 Int. J. Agron. Agri. R.

trict Mansehra is located in the North Eastern part of The total cultivated area of district Mansehra is 80,747
Khyber Pakhtunkhwa with 4,579 sq km of area. The hectares. The area brought under cultivation during
district is mainly mountainous, with an altitude winter season is 40,029 hectares out of which 37,374
ranging from 200 meters in the south upto more than hectares is under wheat i.e. 93.37%.
4,500 meters in the north. The well-known Babosar
pass is located on the northeastern border of the Data Collection
district. The winter climate is cold and the summer The secondary data, time series data, spread over 30
climate is pleasant. In the Kaghan Valley which is years from 1984 to 2013 was collected. The time
situated in the northern, the conditions in summer series data included temperature, precipitation, area
are cool whereas in the season of winter extremely under cultivation and yield of wheat. The data related
cold and during this period there is snow fall of heavy to temperature and precipitation was collected from
magnitude. Summer and winter are dominant Meteorological Department, Government of Pakistan,
seasons of the district. The summer season spread while data pertaining to area and yield was collected
over April to September, while that of winter from from the Crop Reporting Wing of the Agriculture
October to March. Department, Government of Khyber Pakhtunkhwa.

There are two farming seasons. The one is summer Analytical Modeling

locally called as Kharif season and it starts from May A 30 years meteorological data pertaining to

and ends in September. Seasonal vegetables, rice and precipitation and temperature was analyzed against

maize are the main crops of this season. The other is area and productivity. The data was then subjected to

winter season locally called as Rabi season. The an analysis using Auto Regressive Distributed Lag

winter growing season starts from October and ends (ARDL), Co-Integration Technique (Engle 1987) and

in March. Important winter season crops are wheat, Bound Test of Co Integration (Pesaran and Shin 1999

seasonal vegetables and peas. and Pesaran et al. 2001 and Johnsen and Juselius
1990). The Autoregressive Distributed Lag (ARDL)

Material and methods model also called bound testing cointegration model

Selection of Study Area was advanced by M.H. Pesaran and Y. Shin (1999).

The study has been planned and carried out in district The rationale behind use of ARDL model is that the

Mansehra of Khyber Pakhtunkhwa. The selection is combination of time series variables in the study is

based on rainfed nature of the district. The cultivated stationary at level and integration at order 1. Thus in

area of district Mansehra is 80,747 hectares out of such a situation ARDL approach is most suitable

which 77% is rainfed (GoKP 2015-16). Mansehra is econometric tool compared to some of the other

situated at 34.33° North latitude, 73.2° East longitude econometric models. Additionally, it was also aimed

and 1067 meters (https://www.maps- to determine impact of diverse independent variables

streetview.com/Pakistan/) dated 2.11.2017. Monthly on the dependent variable both, for the short-run and

average annual minimum temperature in Mansehra for the long-run for which ARDL model is best suited.

ranges from 2.6 to 21.3 oC, while monthly average The ARDL produces the short run and long run

annual maximum temperature in Mansehra ranges coefficient concurrently along with followings OLS

from 13.4 to 34.8 oC. Average annual precipitation in process for cointegration amid the involved variables.

Mansehra ranges from 34-302 mm (https://en. Another advantage of ARDL is that it offers flexibility

climate- data.org/location/1299/). about order of the integration pertaining to variables

used. Further, it possesses the suitability regardless of
Selection of Crop the fact as whether the variables included are purely
The crop selected for the study is wheat. The crop I(1), purely I(0) or whether mutually cointegrated but
selected is grown on vast areas of the rainfed district and the same is not true if there is any variable having 2nd
have direct impact on the livelihoods of the people. order difference.

Ahmad et al. Page 10

 Int. J. Agron. Agri. R.

The common formula of ARDL model with n lags for analytical model are without any bias included; ADF
variable Y and m lag for variable X stands as given Unit Root Test to decide on analytical model
below: (Kwiatkowski, Phillips, Schmidt and Shinn 1992;
Maddala and Kim 1998; and Phillips and Xiao 1998);
𝑌𝑡 = 𝛼0 + ∑𝑛𝑖=1 𝛼𝑖𝑌𝑡 −1+ ∑𝑚
𝑖=0 𝛽𝑖𝑋𝑡 − 𝑖 + 𝑈𝑡………..1 Autocorrelation Test - Brush Godfrey LM Test
(Breusch, T. S. 1978 and Godfrey, L. G. 1978);
While general format of the ARDL ECM runs as below: Heteroscedasticity Test - Breusch-Pagan-Godfrey
Test (Breusch, T.S. and A.R. Pagan, 1979) and
𝛥𝑌𝑡 = 𝛼0 + ∑ 𝛽𝑗𝑌𝑡 − 𝑖 + ∑ 𝛽𝑗𝑋𝑡 − 𝑗 + 𝜓𝐸𝐶𝑀𝑡 − 1
Stability check Test - CUSUM Test & CUSUMSQ Test
+ ɛ𝑡 … … … … .2
(Brown, R.L., J. Durbin and J.M. Evans 1975).
Therefore given equation reveals the speed with
which adjustment of parameter is taking place. It is Results and discussions
also pertinent to note that for Error Correction (EC) Wheat Productivity
Model to be significant; its value must be in negative Before econometric analysis using ARDL Model,
form. The EC term states that any of the divergence certain assumptions were necessary to be satisfied.
occurring in long-run equilibrium amongst the Therefore, a number of tests were conducted which
variables is to be corrected in each period as well as are detailed below along with their results.
the time period that is to be taken to come yet again
to the long-run equilibrium point. ECMt-1 shows the ADF Unit Root
residuals which are attained out of estimated co- For the econometric analysis of time series data, it is
integration mode. necessary that the data should be stationery.
Therefore, the data was transformed to logarithm and
A regression analysis showed attribution of climatic then in order to determine stationarity of the data,
changes to crop area and productivity. The ADF test was used. The test checked integration
productivity regression model used is as under: order. The test included extra lagged length of
P = β0 + β1R+β2T+ e variable (dependent) in order to remove the problem
Where of autocorrelation in the model. Justification of using
P = productivity per hectare ARDL model is based on ADF test for the stationarity
R = rain/precipitation of the data. The results of the ADF unit root test are
T = temperature given in table 1.
βs = coefficients
e = error term Table 1. Results of the ADF Unit Root Test.
ADF Unit Root Test
Variable Order
The area under cultivation regression model has been T-Statistics Prob. Values
Logarithm of Wheat
as under; -6.722 0.0000 I(1)
yield
A = β0 + β1R+β2T+ e Logarithm Mean
-4.339 0.002 I(0)
Temperature
Where
Logarithm
-5.824 0.0000 I(0)
A = Area in hectare Mean precipitation
R = rain/precipitation
T = temperature The results in table 1 shows that data of variable
βs = coefficients ‘wheat yield’ is stationary at first difference while data
e = error term of variables ‘temperature’ and ‘precipitation’ is
stationary at level. The value of t-statistics is -6.772
Diagnostic Tools Used and Probability value is 0.0000. Thus p value is less
The diagnostic tools used to satisfy various than 5% which means at 5% mean variable is
assumptions and to ensure that the results of the stationary at level.

Ahmad et al. Page 11

 Int. J. Agron. Agri. R.

Similarly t-statistics and p-values of temperature and sum (CUSUM) and the cumulative sum of squares
precipitation reflect that both variables are stationary (CUSUM) test were applied. As shown in fig.s 1 and 2,
at level. The data stationarity was determined using both the CUSUM and the CUSUMQ plot residual are
same test by a number of researchers (Peter et al., found inside the boundaries. That is to say that the
1988). stability of the parameters has remained within its
critical bounds of parameter stability at 5%.
Autocorrelation 12

In order to know the existence of autocorrelation

8
problem or otherwise, Brush God fray LM test was
applied. The results of the test are presented in table 2. 4

0
Table 2. Results of Brush God fray LM Test.
F-statistic 0.920149 Prob. F(2,11) 0.4271 -4

Prob. Chi-
Obs*R-squared 3.869696 0.1444
Square(2) -8

-12
Table 2 shows that the p-values associated with test 19 20 21 22 23 24 25 26 27 28 29 30 31

statistic is greater than the standard significant level CUSUM 5% Significance

[i.e. 0.144> 0.05]. Thus, Brush’s LM test results Fig. 1. CUSUM Result.
reveal that the data is free of autocorrelation problem
i.e. there is no autocorrelation. The same test to check 1.6

autocorrelation was used Pervez et al., 2010.

1.2

Heteroscedasticity
0.8
The presence of Heteroscedasticity in the data could
result in biased results. Therefore, it was important to 0.4

check its presence or otherwise. For this purpose,

Breusch-Pagan-Godfrey test was used. The results of 0.0

the Breusch-Pagan-Godfrey (BPG) test are presented

-0.4
in the table 3. 19 20 21 22 23 24 25 26 27 28 29 30 31

CUSUM of Squares 5% Significance

Table 3. Results of BPG test.
F-statistic 0.994834 Prob. F(13,13) 0.5037 Fig. 2. CUSUM of Square Result.
Prob. Chi-
Obs*R-squared 13.46504 0.4126
Square(13)
Scaled explained Prob. Chi- Both the figures clarify that the CUSUM as well as the
2.741691 0.9987
SS Square(13) CUSUMQ tests confirm the stability of the long-run
coefficients along with the short-run dynamics. These
The diagnostic test for Heteroscedasticity using
tests are in line with the many other researchers work
Breusch-Pagan-Godfrey test given in the table 3
such as Sahbi F. (2012) and Stephen E. M. (1982).
shows at 5%, p-value associated with the test statistics
is greater than the standard significance level [i.e.
Analytical model-ARDL Bounds Test Results
0.4126> 0.05]. The results reveal that the data is
After satisfying all the necessary assumptions, the
homoscedastic and there is no problem of
relationship between wheat yield as a dependent
Heteroscedasticity.
variable and temperature and precipitation as
Stability Check independent variables was estimated using ARDL
For analyzing the stability of the long-run coefficients (bounds) test. The result of the test as depicted in table
together with the short-run dynamics, the cumulative 4 shows long run relationship amongst the variables.

Ahmad et al. Page 12

 Int. J. Agron. Agri. R.

Table 4. ARDL Bound Test Results. In the estimation shown in table 6 indicate for model
Test Statistic Value K to be fit, the sign and value of ECT must be negative
F-statistic 8.460 2
Critical Value Bounds and significant. The coefficient of ECT is (-1.00)
Significance I0 Bound I1 Bound means the speed of adjustment is (-1.00) which
10.0% 3.17 4.14
05.0% 3.79 4.85 implies that around 100% deviations from long-term
02.5% 4.41 5.52
01.0% 5.15 6.36 equilibrium are adjusted every year. The coefficient of
determination (R2) is 84, indicating that 84% of the
The value of F Statistics in table 4 is 8.46 which is
dependent variable’s variation i.e. wheat yield is
higher than upper boundary that shows long run
explained by the explanatory variables presented in
relationship among the variables. Akike Info Criterion
the linear model.
(AIC) is used to select the optimal lag length of
variables included in the ARDL model. Table 6
Wheat area
presents the results of long-run relationship of the
selected ARDL model (3,4,4) using SBC. The long run After satisfying all the necessary assumptions, the

estimation results reflected in table 6 shows that both relationship between wheat area as a dependent variable

temperature and precipitation are positively and and temperature and precipitation as independent

significantly related to the wheat yield in the long run. variables was estimated using ARDL (bounds) test. The
result of the test as depicted in table 7 shows long run
Table 5. Long Run Estimation Results of ARDL relationship amongst the variables.
(3,4,4) Model.
Variables Coefficients Std. Error t-Statistic Prob. Table 7. ARDL Bound Test Results.
Lwheat_temp 3.714 0.782195 4.748261 0.0004
Lwheat_precip 1.583 0.586459 2.698739 0.0182 Test Statistic Value K
C -6.39 2.913093 -2.194763 0.0469 F-statistic 3.83 2
Lwheat_yield=- Critical Value Bounds
6.39+3.714lwheat_temperature+1.583lwheat_precipitation Significance I0 Bound I1 Bound
10% 3.17 4.14
5% 3.79 4.85
The scenario reveals that temperature and 2.5% 4.41 5.52
1% 5.15 6.36
precipitation are positively and significantly related to
wheat yield in the long run. The equation reflects that if
The ARDL (Bound test) shows long-run relationship
1% change occur in precipitation wheat yield will
among the variables. F Statistics value is 3.83 which
increase by 1.583% and if temperature increases by 1%
lies between lower and upper boundaries. It shows
the wheat yield will increase by 3.714% in the long run.
long-run relationship since we take decision from
ECT. ECT must be negative and significant for the
Short Run Estimation
model to be fit. When Error correction term is
The table 6 indicates the short run relationship between
negative and significant it shows long run
wheat yield and temperature and precipitation.
relationship. Akike Info Criterion (AIC) was used to
select the optimal lag length of variables included in
Table 6. Short Run Estimation Results of
the ARDL model. Table 4.11 presents the results of
Cointegration Form of ARDL (3,4,4) Model.
Variable Coefficient Std. Error t-Statistic Prob.
long-run relationship of the selected ARDL model
D(LWHEAT_YIELD(-1)) -0.025488 0.196943 -0.129416 0.8990 (1,0,0) using AIC.
D(LWHEAT_YIELD(-2)) -0.216262 0.160628 -1.346355 0.2012
D(LWHEAT_TEMP) 0.378930 0.463448 0.817631 0.4283
D(LWHEAT_TEMP(-1)) -0.320502 0.496660 -0.645315 0.5299
D(LWHEAT_TEMP(-2)) -1.538222 0.534321 -2.878836 0.0129
D(LWHEAT_TEMP(-3)) -1.648004 0.582195 -2.830671 0.0142
Long Run Estimation
D(LWHEAT_PRECIP) 0.341336 0.212161 1.608857 0.1317
D(LWHEAT_PRECIP(-1)) -0.295750 0.235570 -1.255463 0.2314
The long run estimation results are reflected in table
D(LWHEAT_PRECIP(-2)) -0.139246 0.202482 -0.687699 0.5037
D(LWHEAT_PRECIP(-3)) -0.452239 0.194169 -2.329105 0.0366
8 shows that both temperature and precipitation are
ECT(-1) -1.001 0.224848 -4.454906 0.0006
Cointeq = LWHEAT_YIELD - (3.7141*LWHEAT_TEMP + 1.5827
positively and significantly related to the area under
*LWHEAT_PRECIP -6.3935 )
wheat in the long run.

Ahmad et al. Page 13

 Int. J. Agron. Agri. R.

Table 8. Long Run Estimation Results ARDL (1,0,0). Keeping in view the importance of the issue and to fill
Variable Coefficient Std. Error t-Statistic Prob. research gap current study has been designed to
Wheat_precip 104.962 93.003323 1.128584 0.2694
Wheat_temp 317.849 1627.611914 0.195286 0.8467 assess the potential impacts of change in climate on
C 25345.874 14685.665593 1.725892 0.0962
farm productivity. The result of the research shows
Wheat_Area=25345.874+104.962whet_Mean_Precipitation+317.84
9wheat_Mean_Temperature long run relationship among the variables. The Akike
Info Criterion (AIC) was used to select the optimal lag
The table 8 reveals that temperature and precipitation length of variables and showed that both temperature
is positively and significantly related to area under and precipitation are positively and significantly
wheat cultivation in the long run. The equation shows related to the wheat yield in the long run. The results
that if 1 percent increase in precipitation will increase reveal that if 1% change occur in precipitation wheat
by 104.962 unit and if temperature increases by 1 yield will increase by 1.583% and if temperature
unit, wheat area under cultivation will increase by increases by 1% the wheat yield will increase by
317.849 unit in the long run. 3.714% in the long run. For the short run relationship,
the coefficient of ECT is (-1.00) which implies that
Short Run Estimation around 100% deviations from long-term equilibrium
The table 9 indicates the short run relationship between are adjusted every year. The R2 is 84, indicating that
wheat area and temperature and precipitation. 84% of the dependent variable’s variation i.e. wheat
yield is explained by the explanatory variables.
Table 9. Short Run Estimation Results of Cointegration
Form. The ARDL (Bound test) shows long-run relationship
Variable Coefficient Std. Error t-Statistic Prob. among the area related variables based on F Statistics
D(wheat_precip) 51.936661 35.977296 1.443595 0.1608
D(wheat_temp) 157.276433 798.399548 0.196990 0.8454 value. The ECT is negative and significant showing
Ect(-1).cointeq(-1) -0.494814 0.181729 -2.722803 0.0114
Cointeq = wheat_area - (104.9621*wheat_precip + 317.8499 long run relationship. AIC result shows that both
*Wheat_temp + 25345.8738 )
temperature and precipitation are positively and
In the estimation, the sign and value of ECT must be significantly related to the area under wheat in the
negative and significant respectively. The coefficient long run. The result shows that 1 percent increase in
of ECT is -0.49 means the speed of adjustment is - precipitation will increase 104.962% increase in area
0.49 which implies that around 49% deviations from and if temperature increases by 1%, wheat area under
long-term equilibrium are adjusted every year. The cultivation will increase by 317.849% in the long run.
coefficient of determination (R2) is 24 indicating that The short run relationship between wheat area and
24% of the dependent variable’s variation i.e. wheat temperature and precipitation shows that ECT is (-
area is explained by the explanatory variables 0.49) which implies that around 49% deviations from
presented in the model applied. long-term equilibrium are adjusted every year. The R2
is 24 indicating that 24% of the dependent variable’s
Conclusion and recommendations
variation i.e. wheat area is explained by the
Agriculture is an important source of livelihood for
explanatory variables.
the rural communities of rainfed areas. More than
80% of the 35.6 million people of Khyber
The study recommends that Government should
Pakhtunkhwa are living in rural areas and depend on
workout with public and private research organizations
agriculture. Out of 1.621 million hectares cultivated
to design research plans for rainfed areas aiming at
area of KP, 47% is rain-fed. The food availability for
such areas is mainly dependent on agricultural evolving crops varieties which offer high yields along

productivity which is further dependent on timely with suitability to changing climatic scenarios such as

rainfall and temperature that has now been affected heat and cold resistance, short duration maturity,

by climate change. needing less water, etc.

Ahmad et al. Page 14

 Int. J. Agron. Agri. R.

Further research is required to estimate range of Diao X. 2010. “Economic Importance of Agriculture
temperature and precipitation within which wheat for Sustainable Development and Poverty Reduction:
productivity is optimum. Findings from a Case Study of Ghana. Discussion
paper”. International Food Policy Research Institute.
References
Engle F, Granger WJ. 1987. Co-Integration and
Adger WN, Huq S, Brown K, Conway D,
Error Correction: Representation, Estimation and
Hulme M. 2003. “Adaptation to climate change in
Testing. Econometrica 55(2), 251-276.
the developing world”. Progress in Development
Studies 3, 179-195.
Falco DS, Yesuf M, Kohlin G, Ringler C. 2011.
“Estimating the Impact of Climate Change on
Ahmad R, Zulfiqar M. 2019. Climate change-farmers’
Agriculture in Low-Income Countries: Household
perception, adaptation and impact on agriculture in the
Level Evidence from the Nile Basin, Ethiopia”,
Lakki Marwat district of Khyber Pakhtunkhwa. Sarhad Environ Resource Econ, DOI 10. 1007/s10640-011-
Journal of Agriculture 35(3), 880-889 9538-y, Springer Science+Business Media B.V. 2011.

Breusch TS. 1978. "Testing for Autocorrelation in Godfrey LG. 1978. "Testing Against General
Dynamic Linear Models”. Australian Economic Autoregressive and Moving Average Error Models
Papers 17,334–355. when the Regressors Include Lagged Dependent
DOI:10.1111/j.1467-8454.1978.tb0 Variables". Econometric 46, 1293-1301.

Breusch TS, Pagan AR. 1979. “A simple test for GoKP. 2017. Crop Reporting Wing of the Agriculture
Heteroscedasticity and random coefficient variation”, Department, Government of Khyber Pakhtunkhwa
Econometrica 47, 1287-1294.
Hewitson BC. 2010. “Climate Change Scenario
Brown RL, Durbin J, Evans JM. 1975. Development in Sub-Saharan Africa” SARUA
“Techniques for Testing the Constancy of Regression Leadership Dialogue Series Vol. 2, No.4, pp. 46-67.

Relationships over Time,” Journal of the Royal

Statistical Society, Series B 35, 149-192. Johansen S, Juselius K. 1990. Maximum Likelihood
Estimation and Inference on Cointegration with

Calzadilla A. 2009. Economy-wide impacts of climate Applications to Demand for Money. Oxford Bulletin of
Economics and Statistics 52.
change on agriculture in sub-Saharan Africa. Working
Paper FNU-170. Hamburg: Hamburg University and
Kangalawe MYR, Lyimo JG. 2013. “Climate
Centre for Marine and Atmospheric Science.
Change, Adaptive Strategies and Rural Livelihoods in
Semiarid Tanzania”. Natural Resources 4, 266-278.
Chandrasiri WACK. 2013. “Farmers’ perception
and adaptation to climate change: A case study in
Khan S, Hasan M, Khan MA. 2012. “People
vulnerable areas of Kurunagala district”. Socio
Perception about Climate Change and Adaptation in
Economics and Planning Centre, Department of
the Arid Region of Pakistan. The 2nd World
Agriculture, Peradeniya, Sri Lanka. Annals of Sri
Sustainability Forum-2012”
Lanka Department of Agriculture 15, pp. 13-23.

Kwiatkowski D, Phillips PCB, Schmidt P, Shin

Crop Statistics. 2016-17. Crop Reporting Services,
Y. 1992. “Testing the Null Hypothesis of Stationarity
Agriculture, Livestock and Cooperation Department, Against the Alternative of a Unit Root,” Journal of
Government of KP. Econometrics 54, 159-178.

Ahmad et al. Page 15

 Int. J. Agron. Agri. R.

Kwiatkowski D, Phillips PCB, Schmidt P, Shin Phillips PCB, Xiao Z. 1998. “A Primer on Unit Root
Y. 1992. “Testing the Null Hypothesis of Stationarity Testing,” Journal of Economic Surveys 12, 423-470.
Against the Alternative of a Unit Root,” Journal of
Econometrics 54, 159-178. Rosenzweig C, Hillel D. 1998. Climate Change and
the Global Harvest: Potential Impacts of the Greenhouse
Maddala GS, Kim IM. 1998. “Unit Roots, Effect on Agriculture. Oxford University Press.
Cointegration and Structural Change”. Oxford
University Press, Oxford. Synthesis Report. 2014. “Climate Change 2014
Synthesis Report”. Summary for Policymakers (IPCC).
Makungwa S. 2010. “Adaptation, Agriculture and
Food Security” SARUA Leadership Dialogue Series Yanda PZ. 2010. “Climate Change Impacts,
Vol. 2, No. 4, pp. 68-80. Vulnerability and Adaptations in Southern Africa” SARUA
Leadership Dialogue Series Vol. 2, No.4, pp. 11-30.
Parry ML, Fischer C, Livermore M,
Rosenzweig C, Iglesias A. 1999. “Climate change Ziervogel G, Zermoglio F. 2009. Climate change
and world food security: a new assessment”. Global scenarios and the development of adaptation
environmental change 9, S51-S67. strategies in Africa: challenges and opportunities.
Climate Research 40, 133-146. doi: 10.3354/cr00804
Pesaran, Shin, Smith. 2001. Bounds testing
approaches to the analysis of level relationships. https://www. en.climate-data.org/location/1299/
Journal of Applied Econometrics 16, 289-326.

https://www.maps-streetview.com/Pakistan
Pesaran, Shin. 1999. An Autoregressive Distributed
Lag Modelling Approach to Cointegration Analysis. (https://en.climate-data.org/location/1299/
Econometrics and Economic Theory in the 20th
Century: The Ragnar Frisch Centennial Symposium. (https://www.maps-streetview.com/Pakistan/)
Cambridge University Press.

Ahmad et al. Page 16