Economic Impact Assessment of Climatic Change Sensitivity in Rice-Wheat Cropping System of Pakistan

Objectives: To measure the impacts of climate change sensitivity and how it is affecting economic conditions of farmers in current rice wheat cropping system. Methods/Statistical analysis: Cross-sectional data of 210 farmers from the seven different strata were collected from Punjab, Pakistan. Climate data of baseline (1980-2010) and future (2039-2040) under representative concentration pathways 4.5 and 8.5 for five global circulation models were collected from secondary sources. The climate scenarios were used in two crop simulation models, i.e., DSSAT and APSIM. Tradeoff Analysis Model for Multidimensional Impact Assessment (TOA-MD) was used for the economic analysis. Findings: The crop modeling results of the study using different GCMs and RCPs show that there was negative impact of climate change on the yields of both major crops i.e., rice and wheat. The comparison of both CSMs given the insight that the percent losses were higher in APSIM as compared to DSSAT. The economic analysis endorsed the negative impacts of climate change on farming community. The major economic indicators (net returns, per capita income and poverty) of the study area expressed the declining trend in both RCPs (4.5 and 8.5) and all five GSMs. The observed household vulnerability to climate change percentage was more intense in RCP 8.5 as compared to RCP 4.5, however, among GCMs the figures shown higher vulnerability in hot dry climate conditions and lower in cool wet. The poverty of the study area increased with climate change and it was more prominent while using RCP 8.5 as compared with RCP 4.5.The highest increase in poverty was observed using APSIM crop model for hot-dry conditions. Application/Improvements: The study concluded that to ensure food security, poverty alleviation and to minimize climatic risks there is the need to update agronomic practices and develop adaptation strategies.


Introduction
The most threatening concern of this century for the coming generations is climate change (CC), and the expected consequences of it would be considerable [1][2][3] .The climate variability and change have substantial impacts on all biological and human systems 4 . The problem of climate change gets worse because its impacts could be felt at the places far beyond its origin 5 . Climate change intensity and effects vary in different regions, countries, sectors and communities according to the prevailing environmental conditions 7 . Cool temperate regions will observe the positive impacts on climatic factors such as temperature and precipitation while the tropical regions with already hot climatic conditions will face further rise in average temperature due to CC over the period of time 5,6,7 . The potential climate change risks disturb the whole economic system. Among all major sectors of an economy agriculture is more prone and susceptible to climatic changes. As a result, the developing countries get worse off due to climate change because of the high dependence of their economies on agriculture sector 8,9 . Agricultural farming systems are diverse because of their inherent link to climate and natural resources (water and soil), which make it most susceptible to the changes in climate [10][11][12] . The global food system is at risk due to one of the most concerning issues of agricultural vulnerability to climate change 13,14 . Worldwide, millions of households depending upon agriculture for the livelihood are fluctuating above and below poverty line because of climate variability. These climatic variabilities and changes are a constant threat to the food security and stable food supply by impacting availability, accessibility and utilization of food [15][16][17] .
Climate change excessively affects farmers with small landholding and limited financial stability by further worsening the risks that they face 18 .Adverse impacts of CC on agricultural production and the linked livelihoods have been observed especially in recent two-three decades [19][20][21][22][23] . The identified impacts as stunted crop growth and increase in pest attacks lower the crop yields, hence, reducing the crop revenue worsening the situation of food insecurity [24][25][26][27] .
In South Asia, the rise in temperature more than the global average is a major concern for the existing ecological, economic systems and especially for the sensitive sectors; water, biodiversity and agriculture 8,28 . The increasing climatic concern for the region is due to less adaptive behavior of the countries. Therefore, the food security situation is also very poor in this region. South Asia will be home to highest figure of food insecure masses in the coming years 29,30 . Pakistan is one of the most affected countries in South Asia to climate change 31,32 . Global Climate Risk Index and the World Bank report have placed Pakistan at 7 th position in the index of the countries facing climatic extremes in the time period 1998-2012 33 .
By the year 2100, the increase in temperature will decline the yields of cereal crops 25 to 30 percent and the water availability will decline to 37 percent in South Asian region 34,35 . The other threat to the agricultural production systems of the region is uneven rainfall patterns, risks of floods and droughts which will lower the crop production. Different studies in Pakistan have revealed that cereals and other crop productions are expected to decline due to rise in temperature 36,37 . Wheat production in arid, semiarid and sub-humid regions of Pakistan would decline by 6 to 9 percent while it is expected to increase in the humid areas 38 . In the northern areas of Pakistan, for swat district, the increase in temperature by 1.5 to 3 º C would decrease the wheat production by 7 to 21 percent, and for district Chitral, the decrease would be 14 to 23 percent 39,40 . The declining effect of the rise in temperature on rice yield for semi-arid regions of Pakistan could decline by 15 percent for early midcentury 2012 to 2039 and 36 percent for late century 2070-2099 4142 . Decreasing rainfall effects on crop production are also negative. The net irrigation water requirements in Pakistan will increase by 30 percent by 6 percent decrease in rainfall. The negative effects of decreasing rainfall would affect 1.3 million rural farm households in Pakistan for cereal crops, fruits, and vegetables 43 .
Climate models suggest that temperature will increase up to 0.5-2ºC by 2030 and between 1-7ºC by 2070 in the Asian Pacific region 44 . Wheat crop is sensitive to rise in temperature at the early stages of crop growth. The higher temperature than 30ºC can accelerate senescence by damaging leaf photosynthetic system which results in a reduction of grain filling [45][46][47] . Rice crop is little less sensitive to a high temperature before microsporogenesis, and at tilling stage of crop growth, the temperature range between 27-32 º C is optimal 48 . Temperature above this may lead to pollen unavailability, reduced pollen disposition, embryo abortion, and spikelet sterility ultimately lowering grain yield [48][49][50] . High night time temperature is also a concern for Rice. However, challenges other than rise in temperature for both rice and wheat due to climate are increase in floods, soil salinity, pest attack, weed competition, though these issues vary with geographical location and crop management practices [51][52][53] .
Indo-Gangetic Plains are the hub of rice-wheat cropping system (RWCS) and almost cover about 13 million hectares from Pakistan to Bangladesh. Rice and wheat are the key global food crops which are vital to ensure food security. Rice and wheat are the two main staple cereal crops of Pakistan and are grown almost in all agroecological zones of the country in different climatic and hydrological conditions 54 . The concern about the productivity of Rice and Wheat crops is very crucial because these two crops contribute about 20 and 75 percent in average daily calorie intake of Pakistanis' 55 . The study area was chosen for the significance of both major crops of agriculture sector wheat and rice in the food security situation of the country.
There is a plethora of research on climate change and its impacts on agriculture. And recent literature in last two decades has evolved from research on mitigation Vol 12 (37) | October 2019 | www.indjst.org Tayyaba Hina, Sultan Ali Adil, Muhammad Ashfaq and Ashfaq Ahmad strategies 56-59 to climate change impact assessment [60][61][62][63][64] . The studies on climatic change sensitivity on the integrated agricultural production system for Pakistan are rare and few 3,63,[65][66][67] .The focus of the previous studies was either on crop modelling or econometric modeling. This study is unique and innovative in the sense that it uses an integrated approach using climate, crop, and economic modeling.
The study also included representative agricultural pathways (RAPs) for the non-modelled activities (minor crops and livestock) as RAPs are the climate, economic and social environment, or socio-economic settings in which production systems operate. These are basically qualitative storylines, which are developed with the help of a team of a transdisciplinary scientist by following the nested approach as was used by IPCC for SSPs 68 . The impact of climate change in this study is calculated on current integrated rice-wheat cropping system (including major, minor crops and livestock) for both mild and harsh RCPs (4.5 and 8.5), using APSIM and DSSAT crop simulation models. Undertaking the integrated climatecrop-economic modelling assessment is important to enable a wide-ranging investigation of climate change impacts on agriculture sector, to characterize the actual situation of food security and poverty of the study region and also to highlight how the climatic changes affect the future agricultural productions 69,70 .
Specifically, the basic objectives of the study are; what is the impact of climatic change sensitivity on the integrated RWCS across different RCPs and GCMs? And what is the impact of climatic changes on the socioeconomic conditions (Net Returns (NR), Per Capita Income (PCI), and poverty) of the farmers? The remainder of the paper is followed by the materials and methods in section 1. Results of the analysis are provided in section 2. Section 3 provides the discussions of the results and conclusions are provided in section 4.

Description of the Study Area
In Pakistan, Punjab is the most populated province and contributes the largest share in national agriculture production sector. There are five major agro-ecological zones of Punjab province namely, Cotton-Wheat zone, Rice-Wheat zone, Mixed-Cropping zone, Low-Intensity zone, Rain-Fed zone. Rice-Wheat zone is specifically chosen for the study because of its importance in ensuring food security of the country and the importance of export revenue earned through both crops. The rice-wheat cropping system is the major one which account for total 2.2mha of area, supporting the livelihood of 1.1 million farm families 71 . In Pakistan, the rice-wheat cropping areas are mainly located in central Punjab (main districts include Sheikhupura, Nankana Sahib, Hafizabad, Gujranwala, Sialkot, Gujrat and Mandi Bahauddin) followed by Sindh. The study covers RWCS of Punjab province comprising on the seven famous strata mentioned earlier, forming a heterogeneous sample size.

Collection of Farm Surveyed Data
Both primary and secondary data were collected and used in this study. Primary data were collected from farmers after taking their consent to provide information using a well-structured questionnaire. For secondary data, different government sources and surveys, i.e., Soil Surveys, Economic Surveys, Pakistan Meteorological Department and Pakistan Bureau of Statistics were used. The farming population is heterogeneous, so, multistage stratified random sampling technique was employed to collect the primary data following Naseer, Ashfaq 72 . In the first stage, the RWCS was chosen for this study due to its importance of both major crops rice and wheat, used as staple food. In the second stage, seven rice-producing districts were chosen from the RWCS, i.e., Sheikhupura, Nankana Sahib, Hafizabad, Gujranwala, Sialkot, Gujrat and Mandi-Bahauddin which form the seven strata of the study ( Figure 1). In the third stage, three villages from each stratum were chosen randomly. In the last stage, ten respondents from each village were chosen randomly which makes the total sample size of 210 respondents.

Climate Change Projections
A baseline daily weather dataset (1980-2009) was collected from Pakistan Metrological Department (PMD) and calibrated for future scenarios by using a well-developed climatic methodology following Coupled Model Intercomparison Project (CMIP5) 73 . Statistical downscaling and climate change scenarios were produced by Pakistan Metrological Department (PMD), a method described by Ruane, Goldberg 74 . Future climatic projections of the midcentury 2040-2069 were made for both RCP 4.5 (mild climatic conditions) and RCP 8.5 (harsh climatic conditions). The carbon dioxide concentration of 499ppm were used for RCP 4.5 and 571ppm for RCP 8.5 75 .
Future climate scenarios were developed by using GCMs, representing physical processes in the atmosphere, ocean, cryosphere and land surface. GCMs are the most advanced tools currently available for simulating the response of the global climate system to increasing greenhouse gas concentrations. For this study five best GCMs were used. These models were same for both RCPs (4.5 and 8.5) and are namely; BCC-CSM (cool wet), CCSM4 (cool dry), BNU-ESM (middle), CMCC-CM (hot dry) and MIROC-ESM (hot wet).

Crop Modeling
In this study, two famous Crop Simulation Models (CSMs); the Agricultural Production Systems Simulator (APSIM) 76 version 7.5 and the Decision Support System for Agrotechnology Transfer (DSSAT) 77,78 version 4.6 were used and economic results were evaluated using the simulated yields of both CSMs (for both RCPs and all five GCMs). Both CSMs used four data files for simulation run; (i) weather file with daily solar radiation, maximum and minimum air temperature and precipitation; (ii) soil file of the study area having physical and chemical properties of soil; (iii) crop management file including all input use and application dates; (iv) genetic coefficient file. The detailed information can be found in earlier studies 41,79 . For the analysis, the average crop yields of the farm activities over the time period were also used. For the non-modeled activities (minor crops and milk production) regional representative agricultural pathways (RAPs) were used for the future projections.

Economic Modeling
Economic assessment of climatic change sensitivity was done with the tradeoff analysis model for multidimen-sional impact assessment (TOA-MD) version 6.1 in this study [55][56][57] . The economic analysis was done on the per farm basis. All farm-based activities; major crops (rice and wheat), minor crops (fodder) and livestock was included for the true representation of the existing socioeconomic conditions of the farming community of the surveyed farms. The analysis was done for both CSMs (APSIM and DSSAT) and both RCPs (RCP 4.5 and RCP 8.5) for each GCM simulation separately.
The model considers farmers as economically rational beings to make decisions on the predictable value and that's why uses binary codes 58 . The farmers may choose to stick to system 1, or they can choose to move to the alternative system 2. Generally, system 1 is described as the current production system (base technology) with current climate and system 2 as current production system (base technology) with changing climate. The productivity of the system depends largely on two factors technology and climate. Farmers decision making of whether to operate in system 1 or system 2 depends upon the opportunity cost (gains/losses) from switching.
In Equation 3, v 1 and v 2 are net returns from System 1 and 2, respectively.
Poverty line was set US$ 1.25/person/day (US$ 1= PKR 103) in the analysis according to international standards which was to check the vulnerability level of households with respect to climatic changes 59 .
For Climate Change Impact Assessment (CC-IA) analysis all the prices of inputs/outputs were site specific according to the production system(s) and net returns were accordingly. TOA-MD model parameters for system 1 and 2, for each farm in the survey data in future period, were calculated according to 60  = the year of data collection Ͳ ti = technology and management practices used for period, adapted to climate (t, i=H or F) γ jt = crop yield in year t (kg/ha) μ j (Ͳ ti, γ t ) = mean yield(s) of farm j using technology Ͳ ti with climate γ t Y 0 = observed mean yield of data Y H = historical mean yields used in current period (secondary data) β y0 = Y H /Y 0 = normalization factor of yields s j (Ͳ ti, γ t ) = simulated crop yield for farm j using technology Ͳ ti with climate γ t r j = relative yield for farm j used for analysis a jt = total crop area on the farm in period t (ha) R jt = revenue = p t * y jt * a jt (rupees per farm per time) R js = net returns in system s (rupees per farm) C jt = cost of production for period t (rupees per farm per time) C js =mean cost of production in system s (rupees per farm) C t = mean cost of production in the current period (t=H) β c0 = C H /C 0 = normalization factor for production cost (if β c can't be estimated, then use β c0 =β y0 ) G jt = C jt / R jt =production cost relative to revenue (unit free) V jt = R jt -C jt = crop net returns for the farm (rupees per time) V jqs = time-averaged net returns for part q and system s (rupees) RHO12 = correlation between μ j (Ͳ HH , γ F ) and μ j (Ͳ HH , γ H ) The TOA-MD incorporated the statistical correlation between environmental, social and economic impacts of technology adoption into the simulation of impacts on NR, PCI and poverty. The model simulates the impacts of the full range of adoption rates from 0 to 100 percent 61 .

Data Statistics
The surveyed area in all districts range from 129.66 to 192.31 hectares, smallest in Mandi Bahauddin and largest in Nankana Sahib (Table 1). Likewise, the average land for major crops rice and wheat was dedicated by farmers

Economic Assessment of Climatic Change Sensitivity
In this section, the isolated climate change impacts were assessed on the prevailing agricultural system. Both major cereal crops, i.e., rice and wheat were modeled and then economic analysis was performed using TOA-MD The observed poverty without CC was 8.5 percent, while the poverty with CC varied according to the effect of climatic conditions on relative yields obtained from both CSMs. In RCP 4.5, poverty fluctuated between 11.3 to 13.1 percent in APSIM and 10.6 to 11.9 percent in DSSAT. However, for RCP 8.5 poverty varied between 12.2 to 14.2 percent in APSIM and 11.0 to 12.5 percent in DSSAT (Table 2).  The observed poverty without CC was 3.3 percent, while the poverty with CC varied according to the effect of climatic conditions on relative yields obtained from both CSMs. In RCP 4.5, poverty fluctuated between 6.9 to 9.0 percent in APSIM and 5.7 to 7.9 percent in DSSAT. However, for RCP 8.5 poverty varied between 8.1 to 10.6 percent in APSIM and 6.2 to 8.8 percent in DSSAT (Table  3).  The observed poverty without CC was 5.0 percent, while the poverty with CC varied according to the effect of climatic conditions on relative yields obtained from both CSMs. In RCP 4.5, poverty fluctuated between 6.8 to 7.5 percent in APSIM and 5.9 to 6.8 percent in DSSAT. However, for RCP 8.5 poverty varied between 7.3 to 8.3 percent in APSIM and 6.0 to 6.8 percent in DSSAT (Table  4).

CC-IA for District Sialkot
Results of Sialkot shown household vulnerability for RCP 4.5 in all five GCMs in both CSMs i.e., APSIM and DSSAT fluctuated between 79.0 to 86.4 percent and 72.0    The observed poverty without CC was 7.2 percent, while the poverty with CC varied according to the effect of climatic conditions on relative yields obtained from both CSMs. In RCP 4.5, poverty fluctuated between 9.5 to 11.8 percent in APSIM and 8.6 to 10.7 percent in DSSAT. However, for RCP 8.5 poverty varied between 10.5 to 13.2 percent in APSIM and 8.9 to 11.7 percent in DSSAT (Table 8).

Discussion
Most of the developing countries in the world are dependent on agriculture for the livelihood needs of its farmhands. The impacts of climate change are mostly negative in the case of the agricultural sector. Therefore, it is very important to predict these impacts on the farming community which can be dealt with economic modeling. This study is innovative in this way, as it uses an integrated approach of climate, crop, and economic modeling. In the case of Pakistan, it is in the pioneers of using such an integrated approach.
The crop modeling results of the study used for economic modelling that is relative yields show that there are negative impacts of climate change on the yields of both major cereal crops of the country i.e., rice and wheat for all GCMs and RCPs 41,63 . The results indicated that percentage losses were higher in RCP 8.5 as compared to RCP 4.5 for both CSMs. The comparison of both CSMs given the insight that the percent losses were higher in APSIM as compared to DSSAT, and DSSAT showed more gains. The empirical net impact of climate change was observed negative for both RCPs and CSMs in all districts.
The findings of the three main economic factors (NR per farm, PCI and poverty) depicted that the overall impact of climate change is negative for both RCPs and CSMs. The net returns per farm and per capita income shown the declining trend for both RCPs and CSMs, however, the intensity was higher in RCP 8.5 in comparison to RCP 4.5. Likewise, the poverty status was higher in RCP 8.5 (harsh climatic conditions) as compared to RCP 4.5 (mild climatic conditions) in both CSMs. The findings of the study re-endorsed the fact that climatic variations especially temperature and precipitation negatively affect the yields of both crops which in turn reduced net returns per farm, per capita income, and poverty rates 66,67 . Presently, Pakistan has very low adaptive capacity to climate change due to lack of extension services, infrastructure, required information to formulate and implement effective policy measures 3 . Therefore, the results of this study will act as a way forward in the formulation of current and future adaptation strategies.
The reduction in wheat productivity due to climatic factors is also evident from several other studies in Pakistan 3,66,67 . The decline in rice yield is also observed in this study. But there is less published literature on the impact of climate change on rice in Pakistan 37,63 . Naqvi, Asif 63 done a similar study in the rice-wheat zone of Punjab, Pakistan. But there were some limitations of that study which we have tried to incorporate. First, we took the whole RWCS and data was collected from all seven districts making the sample size most heterogeneous. Second, Naqvi, Asif 63 took only the rice-wheat crops in making the agricultural system, but this study used all farm-based activities; rice, wheat, minor fodder crops, and livestock. Therefore, the results of climate change impact assessment are clearly observed on the socioeconomic variables of the respondents because almost all sub-sections of agricultural income were considered in this study.

Conclusions
The socioeconomic conditions of farmers are vulnerable to both climatic and non-climatic risks in agriculture. The objective of this study was to measure the impact of climatic risks on currently integrated rice-wheat cropping zone of Pakistan considering both mild and harsh representative concentration pathways, i.e., RCP 4.5 and 8.5 using APSIM and DSSAT crop models. The climatic modelling reaffirmed the fact of increase in temperature for the study area. This increase in temperature resulted a decline in the relative yields for both CSMs of modeled activities (rice and wheat). The study concluded that the net economic impacts of climate change are negative for both RCPs and CSMs in the study area. The poverty of the study area will rise with climate change and it is more prominent while using RCP 8.5 as compared with RCP 4.5 in both crop simulation models. The highest increase in poverty was observed using APSIM crop model for hot-dry conditions. The study suggested that there is a dire need of adaptations strategies and to update agronomic practices to address the adverse impacts of climate risks and to ensure food security and livelihood of the people relying on agriculture in the study area. Therefore, the need of the hour is to put serious efforts in this aspect with a particular focus on the agricultural system at regional and national levels.

Acknowledgement
This study is the part of first author's (TH) doctoral dissertation. Authors' would like to acknowledge the Agricultural Model for Inter comparison and Improvement Project (AgMIP), global and Pakistan for providing the logistic and technical support to conduct this study. Authors' would also like to acknowledge the farmers and the data collection team for their time to be valued.