Optimization of energy inputs and greenhouse gas emissions of wetland rice cultivation in Malaysia

Rice is currently being the most cultivated cereal crop and most consumable cereal in Malaysia. The country‘s annual total consumption of rice from 1981 to 2018 has increased 81.6% due to121.4% increase in the population. In order to meet its demand and reduce the importation, the local production of rice has to increase drastically. Increasing production may involve increasing energy consumption and at the same time results with increasing production cost and increasing environment contaminations that resulted from the inefficiency in productions. Ideally, significant increase in the local rice production could be made possible with increasing crop yield within the available crop planted areas through the optimum used of crop inputs to reduce the involved total production cost. Besides that, the total cost production could also be minimized with proper both in adoption and adaptation of the machines and equipment to perform the involved field operations efficiently and effectively. Extensive field evaluations were conducted in real field conditions from a total of 62.2 hectares (62 lots) of paddy plots at Sungai Burung, North-West Integrated Agricultural Development Authority Selangor under two rice growing seasons. The duration period for the main rice season lasted from 30rd June to 30rd November 2017 while the off-season lasted from 1st January to 1st June 2018. Analysis on both the crop inputs and outputs of the involved paddy plots were conducted to determine the energy and cost of rice production within the study area. The measured crop inputs were converted into the equivalent energy values and the greenhouse gas emissions (GHG) values using the appropriate conversion coefficients. Besides that, crops input costs for each field operations were calculated based on the prevailing market prices of the machineries, agricultural materials and services. Later, the energy inputs and GHG emissions were optimized using the multi objective genetic algorithm (MOGA) analysis techniques. In addition, evaluations were also being made on the field performance of the field machineries operating in the field plots and final calculations on the mechanization indexes of the respective field operations in wetland rice production. From the results, the recorded average crop yield in the main season were 7.3 ± 0.4 ton/ha for the transplanting method and 5.9±0.5 for the broadcast seeding method whereas in the off-season were 7.4± 0.3 ton/ha for the transplanting method and 6.2±0.6 ton/ha for the broadcast seeding method. Where else, the total input energy used and GHG emissions in the main season were 15345.73±375 MJ/ha and 465.56±10.45 kgCO2eq/ha in the transplanting method and 16811.98±1239.2 MJ/ha and 490.50±41.77 kgCO2eq/ha in the broadcast seeding method. While the total input energy used and GHG emissions in the off-season were 17571.9±1548.56 MJ/ha and 512.42±35.50 kgCO2eq/ha for the transplanting method and 19018.32±3601.3 MJ/ha and 527.952±99.08 kgCO2eq/ha in the broadcast seeding methods. The energy input in the off-season was higher than the main season due to the fact that in the offseason the farmers used 24.2% and 26.6% higher quantity of fertilizer in the transplanting and broadcasting fields respectively to achieve higher yield because they wanted to follow rice check. The average overall mechanization index was 0.75, and it varied from a lowest value of 0.19 in fertilizing operation to a highest value of 0.99 in harvesting operation for transplanting method. In broadcast seeding method, the average overall mechanization index was 0.63, and it varied from a lowest value of 0.25 in planting operation to a highest value of 0.99 in harvesting operation. Comparisons the on economic cost of production variables in transplanted method and broadcasted method, revealed that broadcasting method enabled farmers in main and off-seasons to save 16.75% and 14.2% of the total cost but the transplanting method showed 2.8 times mean greater net income than the broadcasting method in the main season and 79.3% in the off-season. The benefit-cost ratios of transplanting and broadcasting methods were found to be 1.3 and 1.1, respectively. The developed multiobjective genetic algorism MOGA Model for maximum yield, minimum energy inputs, and minimum greenhouse gas emissions (GHG) showed the energy inputs by the farmers were over excessed than the actual required energy. The accuracies of the models in predicting rice yield is from 90%% to 99.1% and in predicting GHG emissions is from 86.6% to 91.7%. Despite lower consumption of inputs by MOGA, the crop yield was estimated to be at 9.4 ton/ha in transplanting and 9.2 ton/ha in broadcast seeding methods, which are equal to the region’s maximum crop yield under current cultivation and weather conditions