Abstract
This paper tries to explore the dynamic relationship between the yield of major crops in Bangladesh (aus, amas, boro, wheat and potato) and climate change (average rainfall, average maximum temperature and average minimum temperature). Graphical visualization shows that, there is increasing trend in average annual minimum temperature and average annual minimum rainfall, while there is decreasing trend of average annual maximum temperature. We applied autoregressive distributive lag (ARDL) cointegration method in order to study the dynamic relationship. The paper finds long run dynamic link between yield of aus rice and average rainfall and average minimum temperature. We have found short run dynamic relationship for yield of aman rice, boro rice and potato. However, we do not find any dynamic relationship has been found for yield of wheat with the variables of climate change. After estimation the dynamic relationship, we ensure there is no serial correlation among residuals using Breusch-Godfrey test and models are normally distributed using histogram and Jarque-Bera test. Finally, we ensure the stability of the models applying CUSUM test.
Keywords: Climate change, Food crops, Time series, Crops yield, Dynamic relation, Temperature, Rainfall, Bangladesh.
Received: 21 September 2020 / Revised: 19 October 2020 / Accepted:30 October 2020/ Published: 16 November 2020
Contribution/ Originality
This study contributes to the existing literatures of climate change and its impact on yield of major food crops in Bangladesh with more recent and extended dataset and more logical statistical analysis based on global literatures. This is our original work.
1. INTRODUCTION
World’s climate is changing. It is experiencing era of global warming (IPCC). The average temperature of the world is rising and 2016 is the warmest year in last 70 year (NASA). Changing climate has impact in many aspects, environment, agriculture, health, natural resources, eco system and most commonly weather. Nonetheless, climate change is the threat on global food security (Easterling, 2007). Study shown that, global top four produced food’s (rice, maize, wheat, soybean) yield is affected through climate change since 1980 (Lobell, Wolfram, & Justin, 2011). Climate change affects Global food security in four dimensions, availability, stability, utilization and security (Schmidhuber & Tubiello, 2007).
There are many empirical evidences of relationship between different crop production and different variables of climate. They are mainly considering change in temperature and rainfall. Considering temperature change, Kalra et al. (2008) studies the consequence of increasing temperature in Northwest India on yield of wheat, mustard, barley and chickpea. Schlenker and Roberts (2009) shows the impact of US crops yield due to temperature. There are some other evidences, which consider rainfall is the one of the determinant of crops yield.
There are few studies on Bangladesh about rice yield and climate change. Sarker, Khorshed, and Jeff (2012) consider three variables of climate change; maximum temperature, minimum temperature and rainfall. They use ordinary least square (OLS) method. Tesso, Emana, and Ketema (2012) uses Johansen cointegration approach to assess the impact of climate change on crops yield. In their paper, considering cointegration approach, they tried to observe long run relationship of climate change with yield of most three important food; rice wheat and potato of Bangladesh. To estimates three variables of climate change; maximum temperature, minimum temperature and rainfall are taken under concern (Lobell & Field, 2007). The later parts are background of the study, reviewing the existing literature, methodology, result and at the end the paper is concluded.
2. TREND OF CLIMATE CHANGE
The mean temperature in Bangladesh varies from 7.2 degree Celsius to 12.8 degree Celsius throughout the winter and 23.9 degree Celsius to 31.1 degree Celsius throughout the summer.
Figure-1. Trend of climate change.
Source: BBS statistical yearbook.
Moreover, there is an increase in trend of the average minimum temperature and a decrease trend of the maximum temperature over the period 1988-2015 Figure 1 In winter and pre monsoon season minimum temperature is increasing rapidly than the maximum temperature. But maximum temperature of monsoon and post monsoon season are expanding faster. The mean temperature is increased by 0.64 degree Celsius in last 64 years (Van Scheltinga & CTHM Terwisscha, 2015).
Temperature has a notable impact on the productivity, growth duration and pattern of the rice crops. In every growth stage of rice production, it needs proper temperature for development. Deviation of temperature will affect the rice production. According to Bangladesh National Adaptation Program of Action 2000 a hike of 10C to 20C, in combination with lower solar radiation, causes sterility in rice spikelet, and excessive temperature was found to reduce yields of HYVs of Aus, Aman and Boro rice in all study locations and in all seasons in Bangladesh (Talukder, 2007). Higher temperature can deteriorate the grain quality. 1o% increase in peak temperature can decrease the Aman rice production by 2.94, 53.06 and 17.28 tons in three growth stage Bangladesh (Islam, Baten, Hossain, & Islam, 2008).
Bangladesh receives plenty of rainfall, but it is not evenly distributed across region or season. In our country rainfall differs from 1400 mm in the west to more than 4400 mm in the east. The rainfall pattern has been changed over the years. Though the annual rainfall is increased by 6%-11% last 50 years (Shelley, Takahashi-Nosaka, Kano-Nakata, Haque, & Inukai, 2016).
In recent years annual rainfall has a decreasing trend with very narrow margin Figure 1. But during the monsoon there are back-to-back 8-10 rainy days in West and 30-40 day to Northeast (Ahmed & Kim, 2003).
Rainfall is one of the key factors that determine the agronomy of Bangladesh. Bangladesh is a flood prone country. On the other hand, drought is a frequent phenomenon in the Northwest region. Both are the result of rainfall. Heavy rainfall in monsoon of 2007 destroyed the 2.47 million of hectares of agricultural land. On the other hand, in 1994-1995 drought led to a reduction in rice and wheat yield by 3.5 × 106 MT (Rahman & Biswas, 1995). In addition, during the winter season less rainfall will decrease the moisture content in topsoil and it will lead to less recharging in ground water. It adversely affects the agricultural production of Bangladesh (Talukder, 2007).
3. LITERATURE REVIEW
There is extensive literature by many researchers about the dependency of agriculture on the climate. Most of literature is focused on the food sufficiency rather than the food security which is the emerging problem in our country.
Gregory, Ingram, and Brklacich (2005) linked the climate change with food system in several ways from food production to adjustments in market, prices of food and supply cycle framework. A study of Campbell et al. (2016) suggest that the climate change has far reaching impact on crop yield as well as livestock and fisheries. We are interested to assess the current scenario of climate change and food availability, accessibility, stability and utilization for reducing the risk to food supply from climate change in the context of Bangladesh.
Schmidhuber and Tubiello (2007) considered the notability of the different measurements and the comprehensive effect of climate alter on food supply will vary over districts and over time and will depend on the generally socio-economic status. This finding of the studies indicates that changes in climate will increase the dependency of food import of emergent nations. In this study, there is lack of empirical analysis and mainly dependent on theoretical analysis. So, we want to conduct a research to find an empirical relationship between climate change and food security.
Lobell and Field (2007) showed the measure of harvesting season temperature and precipitation spatial average based on the location of crops for the period 1961 to 2002. They took climate variables like temperature and rainfall and used distinct econometrics tools i.e. first difference time series, multiple linear regression and causality test. Climate is inversely related with wheat, maize and barley. They also found that warming temperature trends had significant effect on yield rather than precipitation. Schlenker and Roberts (2009) conducted a research pairing a group of state-level production for corn, soybeans and cotton with a weather dataset of United State for 1950 to 2005. They found that temperature up to 29° C for corn, 30° C for soybeans, and 32° C for cotton helped to increase but that temperatures above these thresholds are very deleterious. Precipitation for corn and soybeans has a statistically significant inverted-U shape but not statistically significant for cotton.
Climate change and agriculture is related in two way relationship for developing countries as they are dependent on agriculture mostly and they have lack of infrastructure for adaptation than developed country. Yohannes (2016) stated in his paper that agriculture in developing country affected due to direct dependency on climate. Actually they are interrelated, using fertilizer in agriculture affects climate through emission of greenhouse gases on the other hand widespread adaptation of mitigation and adaptation actions can be solutions climate change. De (2017) analyzed the impact of climate on the agricultural output using panel data of last six decades for ten zones of Assam. They have used various econometrics tools (panel Co-integration and VECM) and found the long run correlation among CPI, Khariff/Rabi maximum temperature, Fertilizer, GCA and annual rainfall. It also exhibits that negative relationship of Khariff/Rabi maximum temperature on agriculture productivity and positive relationship with rainfall.
Tesso et al. (2012) analyzed the climate fluctuation and its impact on food production of New Shewa area in Ethiopia over three decades. For estimating the impact of climate variability on food production they used VAR and VECM and showed that co integration technology, manure usage, Meher rain and temperature have significant impact and fertilizer and Belg rain has less significant impact on food production. They also used Johannes’ approach to show the dependency of dependent variable on the independent variables.
Various strategy can be taken for the susceptibility of Bangladesh’s agriculture due to the climate change. Participating in the international negotiations process for realization of the objectives under the Bali Action Plan (2007) and planning itself at home for fundamental domestic action-these two approaches will help Bangladesh to deal with the climate issues (Sikder & Xiaoying, 2014). Islam, Tasnuva, Sultana, and Rumana (2014) carried out their research based on local perception of climate change and food production in Khulna district for the year 2013. They observed that high temperature, rainfall and disaster and low food production affect the food supply in the study area. According to their study that the higher knowledge vulnerability is presented for who have US $ 340-390 and US $280-$330 per capita income. They also divided the people who have US $100- $ 270 per capita income into 3 groups and from their perception climate change and food production are negatively correlated. So we can say perception from different people can give us different results.
Numerous studies have investigated the climate change-food production nexus in Bangladesh. Sarker et al. (2012) scrutinized that climate variable have had cogent effects on the rice yield for the period of 1972-2009 years in Bangladesh using the ADF test, OLS and QR models. Though these effects varied among three types of rice. Highest & lowest temperature and rainfall are the climate variables. Highest temperature has impacted positively on Aus and Aman rice but negatively on Boro rice. Opposite is true for the minimum temperature. Finally rainfall has statistically significant effect on Aus and Aman rice. Mamun, Ghosh, and Islam (2015) examined and found significant effect of climate variables on production of rice using time series data from 1972-2010 for Rajshahi. Three climate variables: - Temperature, Humidity and Rainfall have significant effect on three different rice varieties (Aus, Aman, Boro). In their empirical analysis, they found that temperature has inverse correlation with Boro and rainfall has negative correlation with Aman and Aus. About 23%,91% and 89% variance of Aus,Amna and Boro respectively can be explained by the climate variables. Chowdhury and Mohammad (2015) observed the relationship of climate vulnerability and yield of rice crops. They used OLS method for the period 1972-2014 on the basis of country level and the outcomes of the study shows that all four climate variables have substantial effects on the yields of Aus, Aman and Boro rice. In their study they have evaluated the relationship of climate change and yield of rice crops but every year yield of rice is increasing whether climate change or not. Basak (2009) analyzed the effects of climate change on production of Boro rice using DSSAT (Decision Support System for Agro-technology Transfer), for six major rice-growing locales in Bangladesh in 2008. The production varied place to place because of climate condition and hydrological properties of the region. Rajhshahi has the lowest production due to the inconsistency of day and night temperature in winter. Because of 2 degree increase in temperature Boro production can decline by 2.6 % to 13.5%. Minimum temperature affect the Boro production as well. So, from this study we can’t find the accurate dependency of food production on climate change. So, we will conduct this research from another dimension that is growth of yield of crops.
Many researchers have found that fluctuation of climate has most devastating effect on the Coastal area of Bangladesh. Ismail (2016) and Dasgupta, Hossain, Huq, and Wheeler (2014) both proclaimed in their study that because of rise in sea levels and tidal flooding, soil salinity has been rapidly expanded and it will affect the agriculture production. Scientists predict that sea levels could rise 50 to 130cm by 2100 which is a threat for coastal population as two third of Bangladesh is less than 5 meters above sea levels. Hossain and Noman (2018) examined the impact of climate factor in agriculture transformation and food security in Northern Bangladesh. They have shown how rice cultivation land has been transformed into mango and guava orchard because of climate change and economic factor. As production of our major crops is decreasing it is a threat for our food security.
On the other hand, Kabir, Ahmed, and Khan (2016) investigated and found that climate change doesn’t have significant effect on food availability and food security. Which is contradictory with other literature. They worked on the Patuakhali district and demonstrated recent climate change scenarios for Kalapara, analyzed the relationship between climate change scenarios and adaptation of relevant measures and strategies by the cultivators which adequately helps in eliminating the risks of climate change and progresses vocations. As different school of thought have different opinion so we will conduct our research to find out the relationship of climate change with growth of yield of food crops (rice, wheat and potato). To overcome the above mentioned shortcomings of the literature, we will analyze the relation between climate change and food production system from another dimension.
4. METHODOLOGY
We try to identify the long relationship between three major food crops (Rice, Wheat, Potato) and three important indicators of climate (Rainfall, Maximum Temperature, Minimum Temperature).
4.1. Data
As our research area is Bangladesh, annual data of from 1987 to 2015 are collected from Bangladesh Bureau of Statistics (BBS).
4.2. Model Specification
The selected crops sowing to harvesting time are (Yearbook of Agricultural Statistics-2017).
- Aus – March to August.
- Aman – June to November.
- Boro – December to May.
- Wheat – November to April.
- Potato – September to March.
So, models are
4.3. Unit Root Test
To examine data are stationary or not we checked the unit root of data. Non-stationary contains unit root. To check the unit root, we apply Augmented Dickey-Fuller (ADF) and Phillips Perron (PP) ADF model with intercept and no trend:
4.4. Co-Integration Test
To examine cointegration we apply Autoregressive Distributed Lag (ARDL) method. ARDL incorporate both lag value of regressand and regressors (Greene, 2008). It is a cointegration method when variables are stationary at both; level and first difference (Pesaran, Shin, & Smith, 2001). To estimates the models, we use maximum lag 1 for both regresands and regressor as we take only 28 years data. Moreover, we take constant as trend specification. We consider automatic selection process in order to select model.
After performing ARDL, we check the significance of the estimated model by Bound test (Pesaran et al., 2001). If the test statistics of bound test (F statistics and t statistics) are significant, model will be considered as long run cointegrated. If the test statistics are insignificant, model will be considered as not long run cointegrated. If model is long run cointegrated, we estimate error correction model (ECM), where we check the speed of adjustment of both short run and long run to on another. Before checking the ECM, we identify lag length criteria following Schwarz information criterion (SIC). To do so, we adopt vector autoregressive (VAR) method. Later, we identify the residual series for the long run cointegrated model and ECM model. We use ordinary least square (OLS) method.
For not significant long run cointegration, short run ARDL is estimated using OLS method.
4.5. Diagnostic Test
We check the diagnostic of residuals of estimated model. To study the normality, we observe the histogram and apply Jarque-Bera test. Moreover, to examines the autocorrelation we apply Breusch-Godfrey Serial Correlation LM test. In case of ECM and short run ARDL model we also check the stability of the model using Cusum test.
5. RESULT
5.1. Unit Root Test
Addressing level of significance 5%, yield of Aman, Aus, Boro, Wheat and Potato are stationary at first difference, which means integrated order 1 (I (1)), Rainfall and Min_temp are stationary at level and Max_temp is stationary at first difference Table 1
Table-1. Unit root test result (ADF & PP).
| Variables | Level |
First Difference |
||||||
t stat (ADF) |
p value (ADF) |
t stat (PP) |
P value (PP) |
t stat (ADF) |
p value (ADF) |
t stat (PP) |
p value (PP) |
|
| aman | -0.784029 |
0.8076 |
-0.324044 |
0.9087 |
-5.358397 |
0.0002 |
-6.893000 |
0.0000 |
| aus | 0.773477 |
0.9915 |
-1.841660 |
0.3534 |
-9.452017 |
0.0000 |
-10.49737 |
0.0000 |
| boro | -0.249541 |
0.9198 |
-0.464847 |
0.8835 |
-8.625585 |
0.0000 |
-8.625585 |
0.0000 |
| wheat | -1.558606 |
0.4893 |
-1.459260 |
0.5384 |
-5.681142 |
0.0001 |
-6.140013 |
0.0000 |
| potato | 0.495467 |
0.9830 |
0.447217 |
0.9813 |
-9.352221 |
0.0000 |
-13.16608 |
0.0000 |
| rainfall | -6.156032 |
0.0000 |
-6.142354 |
0.0000 |
-6.716668 |
0.0000 |
-26.81666 |
0.0001 |
| max_temp | -2.000072 |
0.2850 |
-4.445422 |
0.0016 |
-11.11705 |
0.0000 |
-15.54201 |
0.0000 |
| min_temp | -1.521472 |
0.5072 |
-4.260515 |
0.0026 |
-12.71418 |
0.0000 |
-15.35659 |
0.0000 |
5.2. Best Fitted ARDL Model Selection
By evaluation 8 models, the best fitted ARDL model for yield of Aus rice is (1,1,0,1), which provides the lowest criterion value and ranks highest Figure 2. This means, 1 lag for Aus, 1 lag for Rainfall no lags for Max_temp and 1 lag for Min_temp
One same basis (1,1,0,1) is the best fitted model for Aman and (1,1,1,1) is the best fitted model for Boro rice. Meanwhile, (1,0,0,0) is the best fitted model for both yield of Wheat and Potato.
Figure-2. Best fitted model for dependent variables.
5.3. Aus Rice
So, the estimated equation for yield of Aus –
Table-2. Estimation for dependent variable Aus.
Long run ARDL model for yield of Aus rice
| Variable | Coefficient |
t-stat |
Probability |
| Aus(-1) | 0.667491 |
7.665745 |
0.0000 |
| Rainfall | -0.000152 |
-1.608862 |
0.1233 |
| Rainfall(-1) | -0.000174 |
-1.752608 |
0.0950 |
| Max_Temp | -0.008989 |
-0.220259 |
0.8279 |
| Min_Temp | 0.016774 |
0.961372 |
0.3478 |
| Min_Temp(-1) | 0.014939 |
1.757831 |
0.0941 |
| C | 0.999812 |
0.564839 |
0.5785 |
| R-squared | 0.919113 |
||
| F-statistic | 37.87655 |
0.000000 |
|
| DW statistics | 1.286575 |
Bound test result for ARDL model of yield of Aus rice
Test statistics |
Significant value (10%) I(0) |
Significant value (10%) I(1) |
|
F-Bound test |
4.080808 |
2.72 |
3.77 |
t-Bound test |
-3.818677 |
-2.57 |
-3.46 |
Lag length identification of yield of Aus rice
Lag |
SIC |
0 |
0.829026 |
1 |
-1.578743* |
2 |
-1.565932 |
3 |
-1.570548 |
4 |
-1.441893 |
Note: *Indicate selected lag length.
ECM model of yield of Aus rice
5.4. Aman Rice
The estimated equation-
So, average rainfall significant for yield of Aman in the short run. Change in yield of Aman rice will increase by .044026 unit if there is no change in average yearly rainfall in previous year meanwhile if change in average annual rainfall increase by 1 unit in the previous year the change in yield of Aman rice will be -.000104 unit.
Table-3. Estimation for dependent variable Aman.
Long run ARDL model for yield of Aman rice
| Variable | Coefficient |
t-Statistic |
Prob.* |
| Aman(-1) | 0.689239 |
5.430356 |
0.0000 |
| Rainfall | -0.000102 |
-1.129192 |
0.2722 |
| Rainfall(-1) | -0.000169 |
-1.733215 |
0.0984 |
| Max_Temp | -0.039847 |
-1.028399 |
0.3160 |
| Min_Temp | -0.002104 |
-0.129090 |
0.8986 |
| Min_Temp(-1) | 0.013909 |
1.663201 |
0.1119 |
| C | 2.293591 |
1.329778 |
0.1986 |
| R-squared | 0.892894 |
||
| F-statistic | 27.78838 |
0.00000 |
|
| Durbin-Watson stat | 1.895409 |
Bound test result for ARDL model of yield of Aman rice
Test statistics |
Significant value (10%) I(0) |
Significant value (10%) I(1) |
|
F-Bound test |
1.689561 |
2.72 |
3.77 |
t-Bound test |
-2.448410 |
-2.57 |
-3.46 |
Lag length identification of yield of Aman rice
Lag |
SIC |
0 |
0.245946 |
1 |
-1.113329* |
2 |
-0.990323 |
3 |
-1.041080 |
4 |
-0.968001 |
Note: *Indicate selected lag length.
Short run ARDL model of yield of Aman rice
Note: ** Dependent Variable: ∆(AMAN)
5.5. Boro Rice
The estimated equation:
So, only yield of Boro itself is significant in the short run for yield of Boro. Change in yield of Boro rice will increase by .0939 unit if there is no change in yield of Boro rice in previous year and if change in yield of Boro rice increase by 1 unit in one year lag the change in yield of Boro rice would be -.5337014 unit in current year.
Table-4. Estimation for dependent variable Boro.
Long run ARDL model for yield of Boro rice
| Variable | Coefficient |
t-Statistic |
Prob.* |
| Boro(-1) | 0.929771 |
16.43755 |
0.0000 |
| Rainfall | 1.40E-05 |
0.152617 |
0.8801 |
| Max_Temp | 0.045050 |
1.118386 |
0.2755 |
| Min_Temp | 0.031471 |
1.754183 |
0.0933 |
| C | -1.787097 |
-1.076539 |
0.2934 |
| R-squared | 0.956741 |
||
| F-statistic | 121.6404 |
0.000000 |
|
| Durbin-Watson stat | 2.758056 |
Bound test result for ARDL model of yield of Boro rice
Test statistics |
Significant value (10%) I(0) |
Significant value (10%) I(1) |
|
F-Bound test |
1.008663 |
2.72 |
3.77 |
t-Bound test |
-1.241582 |
-2.57 |
-3.46 |
Lag length identification of yield of Boro rice
Lag |
SIC |
0 |
1.404424 |
1 |
-1.331337* |
2 |
-1.275219 |
3 |
-1.177557 |
4 |
-1.061960 |
Note: *Indicate selected lag length.
Short run ARDL model of yield of Boro rice
5.6. Wheat
The estimated equation:
Bound test-
Considering the test statistics value (F statistics and t statistics) at 10% level of significant, the null hypothesis cannot be rejected), which means, long run cointegration relationship does not exists Table 5.
Short run ARDL- The appropriate lag length is 1 Table 5 and the short run ARDL model is insignificant. So, no short run cointegration exists.
Table-5. Estimation for dependent variable Wheat.
Long run ARDL model for yield of Wheat
| Variable | Coefficient |
t-Statistic |
Prob.* |
| Wheat(-1) | 0.809729 |
5.844071 |
0.0000 |
| Rainfall | 0.000270 |
1.096746 |
0.2846 |
| Max_Temp | 0.141894 |
1.300154 |
0.2070 |
| Min_Temp | 0.090399 |
1.929035 |
0.0667 |
| C | -6.426551 |
-1.415960 |
0.1708 |
| R-squared | 0.648132 |
||
| F-statistic | 10.13084 |
0.000083 |
|
| Durbin-Watson stat | 1.829913 |
Bound test result for ARDL model of yield of Wheat
Test statistics |
Significant value (10%) I(0) |
Significant value (10%) I(1) |
|
F-Bound test |
2.120446 |
2.72 |
3.77 |
t-Bound test |
-1.373249 |
-2.57 |
-3.46 |
Lag length identification of yield of Wheat
Lag |
SIC |
0 |
1.904740 |
1 |
1.141363* |
2 |
1.253283 |
3 |
1.379750 |
4 |
1.404417 |
Note: *Indicate selected lag length
Short run ARDL model of yield of Wheat
Note: ** Dependent Variable: ∆ (WHEAT).
5.7. Potato
The estimated equation-
So, average rainfall and yield of Potato has impact on yield of Potato in the short run. Change in yield of potato will increase by .794245 unit if there is no change in yield of Potato in previous year and no change in average annual rainfall in two years lag. If change in yield of Potato increase by 1 unit in previous year the change in yield of Potato will be -.688477 unit in current year. The change in average annual rainfall is two years lag increases by 1 unit will cause change in yield of Potato in current year by -.001482.
Table-6. Estimation for dependent variable Potato.
Long run ARDL model for yield of Potato
| Variable | Coefficient |
t-Statistic |
Prob.* |
| Potato(-1) | 0.909199 |
9.165206 |
0.0000 |
| Rainfall | -0.000273 |
-0.268212 |
0.7910 |
| Max_Temp | -0.171262 |
-0.381917 |
0.7062 |
| Min_Temp | 0.039042 |
0.197401 |
0.8453 |
| C | 6.853697 |
0.372281 |
0.7132 |
| R-squared | 0.858370 |
||
| F-statistic | 33.33357 |
0.000000 |
|
| Durbin-Watson stat | 2.788806 |
Bound test result for ARDL model of yield of Potato
Test statistics |
Significant value (10%) I(0) |
Significant value (10%) I(1) |
|
F-Bound test |
0.418736 |
2.72 |
3.77 |
t-Bound test |
-0.915321 |
-2.57 |
-3.46 |
Lag length identification of yield of Potato
Lag |
SIC |
0 |
5.204510 |
1 |
3.840328 |
2 |
3.645734* |
3 |
3.693741 |
4 |
3.800531 |
Note: *Indicate selected lag length.
Short run ARDL model of yield of wheat
Note: ** Dependent Variable: ∆(POTATO).
5.8. Diagnostics Test
We first check the srial correlation of the residuals. For all the residuals of respective models, excluding Boro rice, LM test is not significant and we can not reject the null hypothesis of residuals are not being serially correlated. However, in case of Boro rice residuals are seems serilally correlated Table 7.
Table-7. Breusch-Godfrey Serial Correlation LM Test for Residuals.
long run ARDL of yield of aus rice
Histogram chart and Jarque-Bera test confirm that models are normally distributed Figure 3 & Table 8. Meanwhile CUSUM test ensure the stability of the models Figure 4.
Figure-3. Histogram chart.
Table-8. Jarque-bera test for normality.
Figure-4. Stability Test (CUSUM).
6. CONCLUSION
After completing estimation, yield of aus rice is long run cointegrated with one lag of average annual rainfall and average minimum temperature, which supports many empirical evidence, discussed earlier. Yield of aman rice, yield of boro rice and yield of potato are not long run cointegrate with the climate variables. But yield of aman rice has short run cointegration with one lag of average annual rainfall. Yield of boro rice has short run cointegration with the one lag of yield of boro rice. Yield of potato has short run cointegration with one lag of yield of potato and two lags of average annual rainfall. Yield of wheat does not have any cointegration. The result shows us that the impact of climate on crops production is still not much noticeable. In defense we assume that improved technology, soil fertility and fertilizer can be good reason behind this. However, the narrow impact of climate on production may suggest a greater magnitude impact in the long run.
Funding: This study received no specific financial support. |
Competing Interests: The authors declare that they have no competing interests. |
Acknowledgement: The authors heartly appreciate the guidance from honorable Assistant Professor of Bangladesh University of Professionals, Nahin Rahman, Lecturer Adib Ahmed, Lecturer Abdul M. Khan and Lecturer Faryana Rafiq. |
REFERENCES
Ahmed, R., & Kim, I.-K. (2003). Patterns of daily rainfall in Bangladesh during the summer monsoon season: case studies at three stations. Physical Geography, 24(4), 295-318.Available at: https://doi.org/10.2747/0272-3646.24.4.295.
Bali Action Plan. (2007). Report of the Conference of the Parties on its thirteenth session, held in Bali from 3 to 15 December 2007, Retrieved from: http://unfccc.int/resource/docs/2007/cop13/eng/06a01.pdf . [Accessed 10 December 2013].
Basak, J. K. (2009). Climate change impacts on rice production in Bangladesh: Results from a model. Center for research and action on development. Dhaka, Bangladesh: Unnayan Onneshan (www. unnayan. org).
Campbell, B. M., Vermeulen, S. J., Aggarwal, P. K., Corner-Dolloff, C., Girvetz, E., Loboguerrero, A. M., . . . Thornton, P. K. (2016). Reducing risks to food security from climate change. Global Food Security, 11, 34-43.Available at: https://doi.org/10.1098/rstb.2005.1745.
Chowdhury, I. U. A., & Mohammad, A. E. K. (2015). The impact of climate change on rice yield in Bangladesh: A time series analysis. Russian Journal of Agricultural and Socio-Economic Sciences, 40(4), 12-28.Available at: https://doi.org/10.18551/rjoas.2015-04.02.
Dasgupta, S., Hossain, M. M., Huq, M., & Wheeler, D. (2014). Climate change, soil salinity, and the economics of high-yield rice production in coastal Bangladesh. Policy Research Working Paper 7147.
De, U. K. (2017). An application of vector error correction model to analyze the impact of climate change on agricultural productivity in India's north-eastern region Sub Title Author De, Utpal Kumar Mallik, Girijasankar. Keio Economic Studies, 53, 39-51.
Easterling, W. E. (2007). Climate change and the adequacy of food and timber in the 21st century. Proceedings of the National Academy of Sciences, 104(50), 19679-19679.Available at: https://doi.org/10.1073/pnas.0710388104.
Greene, W. (2008). Functional forms for the negative binomial model for count data. Economics Letters, 99(3), 585-590.Available at: https://doi.org/10.1016/j.econlet.2007.10.015.
Gregory, P. J., Ingram, J. S., & Brklacich, M. (2005). Climate change and food security. Philosophical Transactions of the Royal Society B: Biological Sciences, 360(1463), 2139-2148.Available at: https://doi.org/10.1098/rstb.2005.1745.
Hossain, T., & Noman, A. (2018). Climate change, agricultural transformation and food security in Northern Bangladesh. Bangladesh Economic Association (BEA). Retrieved from: http://bea-bd.org/site/images/pdf/new17/47.pdf.
Islam, A., Tasnuva, A., Sultana, S., & Rumana, S. (2014). Climate change impact: Food production and local Perception. American Journal of Environmental Protection, 3(2), 45-50.Available at: https://doi.org/10.11648/j.ajep.20140302.11.
Islam, M., Baten, M., Hossain, M., & Islam, M. (2008). Impact of few important climatic parameters on aman rice production in Mymensingh District. Journal of Environmental Science and Natural Resources, 1(2), 49-54.Available at: https://doi.org/10.21474/ijar01/7272.
Ismail, H. (2016). Climate change, food and water security in Bangladesh (Vol. 5, pp. 202). Hampden Road, Nedlands WA 6009, Australia: Future Directions International Pty Ltd.,Suite.
Kabir, M. H., Ahmed, Z., & Khan, R. (2016). Impact of climate change on food security in Bangladesh. Journal Pet Environ Biotechnol, 7(6), 306.Available at: https://doi.org/10.4172/2157-7463.1000306.
Kalra, N., Chakraborty, D., Sharma, A., Rai, H., Jolly, M., Chander, S., . . . Mittal, R. (2008). Effect of increasing temperature on yield of some winter crops in northwest India. Current science, 94(1), 82-88.
Lobell, D. B., & Field, C. B. (2007). Global scale climate–crop yield relationships and the impacts of recent warming. Environmental Research Letters, 2(1), 014002.
Lobell, D. B., Wolfram, S., & Justin, C.-R. (2011). Climate trends and global crop production since 1980. Science, 333(6042 ), 616-620.Available at: 10.1126/science.1204531.
Mamun, A. M., Ghosh, B. C., & Islam, S. R. (2015). Climate change and rice yield in Bangladesh: A micro regional analysis of time series data. International Journal of Scientific and Research Publications, 5(2), 189-196.
Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics, 16(3), 289-326.Available at: https://doi.org/10.1002/jae.616.
Rahman, A., & Biswas, P. (1995). Devours resources. Dhaka Courier, 11(42), 7-8.
Sarker, M. A. R., Khorshed, A., & Jeff, G. (2012). Exploring the relationship between climate change and rice yield in Bangladesh: An analysis of time series data. Agricultural Systems, 112 11-16.Available at: https://doi.org/10.1016/j.agsy.2012.06.004.
Schlenker, W., & Roberts, M. J. (2009). Nonlinear temperature effects indicate severe damages to US crop yields under climate change. Proceedings of the National Academy of Sciences, 106(37 ), 15594-15598.Available at: https://doi.org/10.1073/pnas.0906865106.
Schmidhuber, J., & Tubiello, F. N. (2007). Global food security under climate change. Proceedings of the National Academy of Sciences, 104(50), 19703-19708.Available at: https://doi.org/10.1073/pnas.0701976104.
Shelley, I. J., Takahashi-Nosaka, M., Kano-Nakata, M., Haque, M. S., & Inukai, Y. (2016). Rice cultivation in Bangladesh: present scenario, problems, and prospects. Journal of International Cooperation for Agricultural Development, 14(4), 20-29.
Sikder, R., & Xiaoying, J. (2014). Climate change impact and agriculture of Bangladesh. Journal of Environment and Earth Science, 4(1), 35-40.
Talukder, B. (2007). Climate change and agriculture. The Daily Star. Retrived from: https://www.thedailystar.net/news-detail-5675.
Tesso, G., Emana, B., & Ketema, M. (2012). A time series analysis of climate variability and its impacts on food production in North Shewa zone in Ethiopia. African Crop Science Journal, 20 261-274.
Van Scheltinga, & CTHM Terwisscha. (2015). Recent changes in temperature and rainfall trends and variability over Bangladesh. Paper presented at the Gobeshona Conference on Research on Climate Change in Bangladesh.
Yohannes, H. (2016). A review on relationship between climate change and agriculture. Journal of Earth Science & Climatic Change, 7(2), 335.Available at: 10.4172/2157-7617.1000335.
Views and opinions expressed in this article are the views and opinions of the author(s), International Journal of Sustainable Agricultural Research shall not be responsible or answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content. |