By Raziah Mat Lin; Engku Elini Engku Ariff; Tapsir Serin; Mohd Zabawi Abdul Ghani

Climate change is used to describe the changing climate patterns that can be attributed to human activity that alters the earth's atmosphere. Climate change could threaten lives, sustainable agriculture production, fresh water supplies and the survival of native species and ecosystems. The environmental threats such as rising sea levels and inadequate water resources would lead to food shortages. Food shortages, especially rice, experienced by many countries two years ago led to food riots and political instability around the globe. Malaysia was not spared from the crisis.

Although food consumption in Malaysia has shifted away from starchy staples toward wheat-based and livestock products, rice is still the staple food for Malaysians. Per capita consumption of rice constituted about 11 per cent of the total per capita food consumption in the country. Malaysia is a net importer of rice. The total production of rice in 2008 was about 1.5 million tons. The per capita consumption of rice was 82 kilograms and the total consumption of rice in Malaysia was estimated to be around 2.3 million tons yearly.

The current self sufficiency level (SSL) is about 72 per cent. Due to the food crisis and for food security reason, the SSL for rice in Malaysia has been revised to achieve 90 per cent by 2010 (Mid-term review of the Ninth Malaysia Plan, RMKe9, 2006-2010).

Strategies to achieve the 90 per cent SSL level are being debated; increase productivity or open new areas for rice cultivation. Improving rice productivity by increasing farms' efficiency and mitigating the effect of climate change would be feasible options.

The general objective of the study was to evaluate the impact of climate change on food production. The specific objectives are: to forecast the impact of climate change on rice production based on advanced crop modelling and economic regression models and to formulate policies to minimize the negative impacts of climate change on rice production.

Muda Agriculture Development Authority (MADA)
Muda Agriculture Development Authority (MADA) area was selected for the study. The total area covered by the Muda Irrigation Scheme is about 125,155 hectares contributing 60 per cent of the total rice production in the country. About 76 per cent of the land is under paddy cultivation (96,558 ha) and approximately 48,500 farm families reside in the area. The average farm size for each household is 2.2 ha.

The productivity of rice in MADA has been increasing over the years due to the improvement in farms' infrastructure and technologies. New rice varieties have been introduced from time to time by the Malaysian Agriculture Research and Development Institute (MARDI). Almost all farmers in the area are adopting the recommended varieties and the crop husbandry technologies proposed by MARDI.

In this study the process-based Decision Support System for Agro-technology Transfer (DSSAT) model and economic regression model were utilized to assess climate change impacts on rice production in MADA area. The implication of climate change to the farmers' well-being as well as the long-term consequences on the self sufficiency and food security of the country were also analysed.

Secondary data were gathered from MADA, National Hydrology Research Institute of Malaysia (NAHRIM), Malaysian Meteorological Services (MMS) and soil survey report. The data considered included yearly total production (tons), productivity (tons/ha), site, weather, soil characteristics, soil type and management. The relevant data from the study on productivity (2008-2009) were selected to model the rice farms production function.

Model specification
Changes in yield of rice under different climate scenarios can be predicted using DSSAT 4.5. The econometric regression models were utilized to test the consistency and validity of the DSSAT results. A relationship between rice production and climate change (temperature, rainfall and relative humidity (RH)) was estimated over the period studied (1988-2008). The model to be estimated is in Equation (1).

Equation (1)
Yit= α + β1 Wit + β2 W2t + β3 W3t
Y1 = Production (tons),
Y2 = Yield ( kg/ha),
α= constant,
W1 = Temperature (0C),
W2 = Rainfall (mm),
W3 = RH (%)
t = years such that t = 0,1,2,3,..n
βi = parameter to be estimated

Ordinary Least Square (OLS) was performed on the data to determine whether there were significant relations between the endogenous and exogenous variables. The magnitude of the climate change impacts on production were also determined using the model.

To understand further on other factors that influenced rice production, a production function was estimated using log linear estimation model. A total elasticity of production was estimated to determine the status of rice farms' operations, i.e., whether it was increasing, constant or decreasing returns to scale (Equation 2).

Equation (2)
lnYt = α + β1 lnLand + β2 lnLab + β3 lnInput + β3 lnDeprec + ui
Y1 = Value production (RM),
Y2 = Productivity index,
α = constant,
Land = Cost of land,
Lab = Cost of labour,
Input = Cost of variable inputs,
Deprec = Cost of capital depreciation and
βi = parameters to be estimated

The EXCEL, SPSS and E-views software were used to run the econometric modelling.

Results and discussions
The impact of temperature on rice yield using DSSAT model
In Malaysia, the actual farm yields vary from 3 - 5 tons per ha, while the potential yield was estimated to be around 7.2 tons per ha.

The effects of climate change on rice production depend on the variability in temperature and rainfall. The optimum temperature for rice cultivation is between 240C - 340C, while the optimum rainfall is not less than 2,000 mm per year. Presently, the average temperature in rice growing areas in Malaysia is about 260C. In general, an increase in temperature above the tolerance limit would reduce the potential yield due to the reduction in photosynthesis, increase respiration and shorten vegetative and grain-filling period.

The average temperature in MADA was 27.40C in 2008. The optimum temperature requirement for MR219 and MR232 is 340C. Table 1 reports the effects of climate change on rice production in MADA using DSSAT.

When the temperature increased from 27.40C to 30.50C, the rice yield was expected to reduce by 17.82 per cent. For every 10C increase in temperature, the rice yield reduced by an average of 6 per cent. Compared to Thailand, the forecasted yield loss in the country was higher, a reduction of 10 per cent yield when the temperature increased by 10C.

The International Panel on Climate Change (IPCC) forecasts that temperature could rise between 2.4-6.40C by the end of the century. However, the drastic increase in temperature in MADA area is very unlikely to happen in the near future. As mentioned before, the average yearly temperature in MADA area from 1988-2007 was 27.50C with a minimum temperature of 26.60C and maximum of 29.670C (MADA, 2008). The average percentage change in temperature for the period was only 0.15 per cent. Even with human intervention, MADA would be resilient enough to sustain the temperature in the area with its mitigation initiatives.


Climate change impact on farmers' income
The price of rice at harvest is predetermined by the government. In 2008, the price quoted was RM1,250.00 per ton (MADA, 2008). At the average temperature of 27.50C in 2008, the average revenue received by the household was estimated to be around RM7,625.00 per ha per season (Table 2). At a constant price, the revenue received tends to decrease as the yield decreases when the temperature increases based on the DSSAT modelling. At a constant cost of production of about RM1,807.00 per ha per season (MADA, 2008), the projected net income received by the farmers also decreases. The poverty line for rural household in Peninsular Malaysia is RM760.00 per month or RM9,120.00 per year (RMKe 9, 2006-2010). Based on an average of 2.2 hectares farm size owned by the households in MADA area, the majority of farmers in the area would likely fall within the poverty group only when the temperature increases to 36.50C, when the projected household incomes would come to about RM4,259.00 per year based on this simulation. As said before, the drastic increase in temperature is unlikely to happen in the near future.

The impact of climate change on rice yield using economic regression models
Ordinary least square (OLS) procedure was performed on time series data (1988-2008) using the log linear function to determine whether there was significant change in temperature, rainfall and RH over the periods. Results of the OLS indicated that there was significant increase in temperature (Adjusted R square 0.344, F value 22.54) and significant decrease in RH (Adjusted R square 0.316, F value 19.83) in MADA but insignificant changes in rainfall.

Assuming temperature, rainfall and RH as exogenous and rice productivity as endogenous, a log linear OLS performed on the data indicated that variable temperature significantly influenced rice productivity in MADA in a positive direction (Table 3). Against the studies done by Lobell D.B. et al. (2008) and Peng et al. (2004), in this modelling, a 1 per cent increase in temperature tends to increase the rice productivity by 2.3 per cent. This can be explained by the fact that the temperature range in the MADA area was within the optimum level for the cultivated cultivars (26.600C -29.670C). The increase in temperature detected over the years could be due to increasing human interventions arising from the objective and strategy of MADA to increase rice productivity to a maximum possible level to meet the targeted SSL. Nevertheless, a minor increase in temperature did not affect the rice productivity; instead it increased the farms' productivity as proven by this model. On the other hand, both rainfall and RH were negatively associated with rice productivity but not significant. The model fits fairly well to explain the relationships between the dependent and independent variables with F value equal 5.345, R square equal 0.254, which means 25.4 per cent of the variation in productivity was attributed by temperature, rainfall and relative humidity and the rests were due to other factors.

Other factors that influence rice production
To understand further on other factors that affected the rice productivity, primary survey data set was used to estimate the rice farms' production function using log linear Cobb Douglas function. The dependent variable was specified as production (kg), while the independent variables were specified as land (ha), labour (man hour), inputs (RM) and depreciation (RM). The data was collected in 2009 involving 60 samples of households in MADA. Summary result of the OLS is presented in Table 4.

Variables land and labour significantly influenced the rice production in positive directions. Increase in land area and increase in labour would increase the rice production. However, variables input and depreciation did not influence the rice production. The model fit very well to explain the relationship between the dependent and independent variables with F value of 158.86 and adjusted R square of 0.83. About 83 per cent of the variation in production was attributed to the four input factors and the rest were caused by other factors, most likely climate factors. The total elasticity of production for this model was 1.04 which means production could still be increased by manipulation of all input factors since at present the farms in MADA generally were operating at increasing returns to scale.

Conclusions and policy recommendations
For the past two decades, there were indications of increasing temperature in MADA but the rice productivity in the area also increased. Improvement in technologies and human intervention in the area to increase productivity may contribute to a minor increase in temperature. MADA could continue manipulating farms' production factors and new technologies to increase productivity and efficiencies since the farms in the area generally were operating at increasing returns to scale. This would be a feasible strategy to increase the targeted SSL of 90 per cent in the RMK 10 (2011-2015) from the current level of 72 per cent.

Malaysia has been proactive in its efforts to mitigate the climate change impact on rice production in the granary areas in the country. One of its adaptation strategies is to increase water use efficiencies (WUE). Short- and mid-term strategies are applicable at the farm level with involvement of farmers and extension workers, and long-term strategies at national and regional levels which involve in changes of government policies.

The effective use of the available water resource through crop management can play a major role in sustainable production under limited water conditions. This can be done by optimizing water use pattern throughout the growing season by changing the cropping pattern. The choice of crop species and cultivars to be grown is critical and should reflect the availability of water. Utilization of the best adapted cultivar is important in maximizing yield under limited water. The choice of cultivar, such as drought-tolerant cultivar, is another means of adaptation to drought prone environments and of increasing WUE.

Development of varieties which are tolerant to water stress with high water use efficiency is also needed. Aerobic rice cultivation, which is a new way of growing rice in aerobic soil without flooding, will be the best alternative for producing rice using water saving technology. Under aerobic rice system, the potential yields vary from 4.5 to 6.5 tons per ha which is about 20-30 per cent lower than that of lowland varieties grown under flooded conditions.

The adoption of appropriate soil management practices is essential to conserve water, nutrients and soil, particularly the structure and drainage characteristics. Under climate change, additional crop water availability may be obtained by increasing the soil water storage capacity, reducing soil evaporation and increasing soil water extraction. The use of crop residues such as mulch is one means of reducing soil evaporation through a reduction of the amount of energy reaching the soil surface.

Apart from agronomic practices and policy, plant breeding and biotechnology could also play an important role to produce food under climate changes. The major focus in biotechnology should be to improve crop yield by increasing carbon gained during the crop cycle.