Reprint of Alternative cropping systems for greenhouse gases mitigation in rice field: a case study in Phichit province of Thailand

Journal of Cleaner Production 134 (2016) 547e562

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Reprint of Alternative cropping systems for greenhouse gases mitigation in rice field: a case study in Phichit province of Thailand Noppol Arunrat a, b, Can Wang a, c, Nathsuda Pumijumnong b, * a

State Key Joint Laboratory of Environment Simulation and Pollution Control (SKLESPC), School of Environment, Tsinghua University, Beijing 100084, China Faculty of Environment and Resource Studies, Mahidol University Nakhon Pathom 73170, Thailand c Ministry of Education Key Laboratory for Earth System Modeling, Center for Earth System Science, Tsinghua University, Beijing 100084, China b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 14 May 2015 Received in revised form 17 May 2016 Accepted 22 May 2016 Available online 4 August 2016

Seeking new practical approaches as alternative ways for greenhouse gas (GHG) mitigation to incentivize farmers’ behaviors regarding farming is a challenging point. Thus, five alternative crop rotations (rice, corn, mung bean, soybean and watermelon) under ten alternative cropping systems were investigated. Data were collected using field surveys and structured questionnaires at the same 144 farms (in irrigated and rainfed areas of 72 farms equally) covering two crop years. GHG emissions were evaluated based on the concept of the life cycle assessment of the greenhouse gas emissions (LCA-GHG) of products. Economic analyses of each cropping system were preformed: cost-benefit analysis, net profit and marginal abatement cost (MAC). Results revealed that rice cultivation was the major source of GHG emissions, particularly due to the planting and burning of rice residue stages, while GHG emissions of crop rotation systems were generated mainly from the land preparation stage. On a per area basis, large farms show significantly higher GHG emissions than small and medium farms. Conversely, for per kg of crops produced, small farms generated the highest GHG emissions, compared to the other sizes of farms. This study strongly supports the implementation of a triple cropping system in irrigated areas, which suggests that crop rotations after the first and second rice harvesting with mung bean gained the highest B/C ratio at 1.48; the negative abatement potential was
5.47 ton CO2eq ha
1, the negative abatement cost was
2378.31 Baht ha
1 and the negative MAC was
434.86 Baht ton
1 CO2eq. For rainfed areas, the double cropping system with selecting mung bean is recommended because it is the most profitable as the B/C ratio is 1.52, the negative abatement potential is
7.34 ton CO2eq ha
1, the negative abatement cost is
2161.11 Baht ha
1 and the negative MAC is
294.48 Baht ton
1 CO2eq. Alternative cropping systems with selecting crop rotation not only reduce GHG emissions in the rice field but also increase the benefits to farmers. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Rice field Cropping systems Greenhouse gas emissions Life cycle assessment Marginal abatement cost

1. Introduction Thailand is an agricultural country. Of the total land area in Thailand, 21.28 million ha or 41% is engaged in the agricultural sector. In this regard, 10.88 million ha or 21% is accounted for by rice fields, and 10.4 million ha or 20% is accounted for by other croplands (maize, sugarcane, cassava, mung bean and soybean) (Singhapreecha, 2014; OAE, 2014). However, rice cultivation is an important emitter of greenhouse gases (GHG) into the atmosphere,

DOI of original article: http://dx.doi.org/10.1016/j.jclepro.2016.05.137. * Corresponding author. E-mail addresses: [email protected], [email protected] (N. Pumijumnong). http://dx.doi.org/10.1016/j.jclepro.2016.08.015 0959-6526/© 2016 Elsevier Ltd. All rights reserved.

especially methane (CH4), nitrous oxide (N2O) and carbon dioxide (CO2) (Zheng et al., 2000; Xu et al., 2002; Ghosh et al., 2003; IPCC, 2001, 2007). All have the potential to contribute to global warming since their atmospheric concentrations have been increasing. Rice fields are a possible area for reducing emissions through changes in cultivation practices (Cai et al., 2003; Scheehle and Kruger, 2006; Connor and Comas, 2008). Alternative approaches to crop rotation can greatly increase yield and soil health and sustain efficient farming (Doran et al., 1998; Collins et al., 2000) by decreasing nitrogen fertilizer use, substantially lowering related N2O emissions (Mosier et al., 1998; Zou and Huang, 2007; Maelinda and Noorlidah, 2008) and allowing the fields to dry out occasionally, which can decrease CH4 emissions (Smith and Almaraz, 2004; Pattey et al., 2008).

548

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

Meanwhile, today, the climate change situation can severely affect farmers. A shorter period of rainfall means a long period of drought, causing water scarcity. Phichit is one of the provinces where water distribution for farming has been stopped by Thailand’s government almost every year because the water level in major dams and reservoirs remains critically low and poses an immediate risk of water shortages. Farmers should refrain from planting a second rice crop. This is a major cause of crop failure, income volatility, liability increase and poverty for farmers. Alternatively, giving farmers the incentive to plant crops that consume less water can be an effective way to solve these problems. However, there is no information available, and there is a lack of research to convince farmers of the economic benefits of introducing a multiple cropping system into rice fields or to convince government agencies of the practical guidelines for GHG mitigation in rice fields. These issues raise the question of which cropping systems are beneficial and efficient for farmers, the environment and society as a whole and which are appropriate under different cropping systems and with limitations in different areas, in particular, for irrigation systems. It is necessary to implement a joint concept of the life cycle assessment of the greenhouse gas emissions (LCA-GHG) of products whose results serve to pinpoint and improve the practices and conditions behind GHG emissions, along with a cost-benefit analysis to propose an economically worthwhile farm management method that produces a smaller amount of GHG. In Thailand, LCAGHG has been assessed in various publications in recent years, e.g. for cassava (Nguyen et al., 2007; Moriizumi et al., 2012), palm oil (Silalertruksa et al., 2012; Saswattecha et al., 2015), sugarcane (Yuttitham et al., 2011) and rice (Kasmaprapruet et al., 2009). Unfortunately, no studies have investigated farmers’ activities for the whole cropping calendar. No studies have evaluated GHG emissions and farmers’ profit from rice production and crop rotation patterns after rice harvesting. Therefore, the aim of this study is to evaluate GHG emissions and farmers’ profit from five alternative crop

rotations (rice, corn, mung bean, soybean and watermelon) under ten alternative cropping systems.

2. Materials and methods 2.1. Site selection Multi-stage sampling was employed for this study as follows. Firstly, at the provincial level, purposive sampling was used, focusing on farmers who have grown rice and had dominant types of crop rotation after harvesting rice each year. They voluntarily participated and provided their information and opinions. This method depends on the study’s specific objective, what, exactly, needs to be investigated and finding people who can and are willing to provide information (Bernard, 2002). Therefore, Phichit province was chosen as a case study as the topography can be divided into irrigated and rainfed areas. This can reflect drought and flood events unequivocally. Frequently, the areas are far away from the Yom and Nan rivers, which have faced drought events and in which flood events have damaged the nearby areas. Secondly, at the district and sub-district levels, cluster sampling was used to determine two clusters: irrigated areas and rainfed areas, which may have different impacts on GHG emissions (Lee et al., 2015) and farmers’ income and benefits (Mushtaq et al., 2015). Moreover, farmers’ net household income (calculated by subtracting expenses from total revenue) of each district and sub-district were set as the criterion, based on the assumption that money is what people seek to improve their livelihood and in what convinces farmers to change their behavior. Four districts (Bang Mun Nak, Taphan Hin, Bueng Na Rang and Pho Prathap Chang districts) with the highest net incomes and four other districts (Sam Ngam, Wachira Barami, Wang Sai Phun and Thap Khlo districts) with the lowest net incomes in Phichit province were selected as samples (Fig. 1). Lastly, at the farm level, purposive sampling considering different farm

Fig. 1. Study area.

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

sizes in both clusters was employed because of possibilities of different impacts on GHG emissions (Sefeedpari et al., 2013). 2.2. Data collection Data were drawn from participatory observation, in-depth interviews and a questionnaire survey at the same 144 farms (in irrigated and rainfed areas of 72 farms equally) for two crop years (2012/2013 and 2013/2014). Data throughout the crop years from each crop, consisting of cultivation practices, agriculture inputs (e.g., fossil fuel, fertilizer, insecticide, herbicide), yields and transportation costs and benefits were collected from farm owners. Data were also obtained from the record book for the standards for good agricultural practices (GAP) for farm owners, which was disseminated to the farmers by the Department of Agricultural Extension, Ministry of Agriculture and Cooperatives. The number of farms surveyed and the life cycle inventory of crop cultivation practices are shown in the Supplementary Material (Section S1, Table S1 and S2). In this study, an alternative cropping system refers to a series of options to systematically grow different crops in the same field each year. There were ten alternative cropping systems as follows: 1) a single cropping of rice cultivated only once (first rice) and fallow land (hereinafter referred to as RF); 2) a double cropping of rice cultivated twice (first rice and second rice) (hereinafter referred to as RR); 3) a double cropping of rice cultivated once (first rice) and corn (hereinafter referred to as RC); 4) a double cropping of rice cultivated once (first rice) and mung bean (hereinafter referred to as RM); 5) a double cropping of rice cultivated once (first rice) and soybean (hereinafter referred to as RS); 6) a double cropping of rice cultivated once (first rice) and watermelon (hereinafter referred to as RW); 7) a triple cropping of rice cultivated twice (first rice and second rice) and mung bean (hereinafter referred to as RRM); 8) a triple cropping of rice cultivated twice (first rice and second rice) and soybean (hereinafter referred to as RRS); 9) a triple cropping of rice cultivated twice (first rice and second rice) and watermelon (hereinafter referred to as RRW) and 10) a triple cropping of rice cultivated three times (first rice, second rice and third rice) (hereinafter referred to as RRR). The cropping calendar for the cropping system each year is shown in Fig. 2. 2.3. System boundary, functional unit and limitations The concept of the LCA-GHG of products based on cradle to gate

549

was employed, which covers raw material production, transport of agricultural inputs (diesel, gasoline fuel, chemical fertilizer, insecticide and herbicide) to the farm, land preparation, planting, harvesting, storing and burning crop residues (post-harvest) (Fig. 3). Transportation of agricultural inputs from the manufacturer to the retailer was not included because there was very little information available. The transportation data were considered for two distances: the average distance from the farms to the retailer in the municipality of each sub-district and the average distance from the farms to the retailer in the community of each farm. To assess the combined GWP, CH4 and N2O were calculated as CO2 equivalents over a 100-year time scale using a radiative forcing potential relative to CO2 of 28 for CH4 and CO2 of 265 for N2O (IPCC, 2013). The functional unit, which is related to GHG emissions estimation for each cropping system, is expressed as kg CO2eq ha
1 yr
1. 2.4. GHG emissions calculation The GHG emissions were calculated for each farm and in each crop year. Upstream emissions were accounted for in terms of raw materials production and the transportation of agricultural inputs to the farm. Fossil fuel, chemical fertilizer, insecticide and herbicide productions were estimated using specific emission factors as characterized in Ecoinvent 3.2 (Ecoinvent Centre, 2015). Emissions from the transportation of agriculture inputs to the farm were collected based on diesel fuel consumption, accounted for by using the emission factors from the National Technical Committee on Product Carbon Footprinting (Thailand) (2011). In some cases, the specific emission factors of gasoline and of insecticides and herbicides were not available in Ecoinvent 3.2, so we used Thailand’s country specific emission factors from the National Technical Committee on Product Carbon Footprinting (Thailand) (2011). Detailed input quantities and life cycle inventory (LCI) dataset names are provided in the Supplementary Material (Section S2, Table S3). Field CH4 emissions from rice cultivation were used as the model for the calculations according to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC, 2006). The baseline emission factor was taken from Yan et al. (2003), who adjusted the region-specific emission factors for rice fields in east, southeast and south Asian countries, and all scaling factors used were from the IPCC (2006). The direct and indirect N2O emissions and CO2 emissions from urea applications were also estimated using the methodology proposed by the IPCC (2006). The GHG emissions calculation and parameters and emission factors for the diesel and gasoline usage of stationary combustion were taken from the IPCC (2006). Meanwhile, GHG emissions from the mobile combustion of the diesel fuel of farm tractors and harvesters were estimated from the emission factors of Maciel et al. (2015), and GHG emissions from gasoline fuel were estimated from the EPA (2014). Figures for insecticides and herbicides were provided by the emissions factors from Lal (2004). The equations, parameters and emission factors for the calculation of GHG emissions are presented in the Supplementary Material (Section S2, Table S4). 2.5. Economic analysis: benefit-cost ratio (B/C), net profit and marginal abatement cost (MAC)

Fig. 2. Cropping calendar for cropping system in each year.

The B/C ratio was used to calculate the overall benefit value versus the cost value in the lifetime of the cropping system investment. If B/C > 1, the investment cropping system would be accepted. The net profit of each cropping system is calculated by subtracting the total production cost (water, tillage, labor, fertilizer, insecticide, herbicide, harvest and land rental) from the farmer’s income. The net profit must be higher than or equal to

550

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

Fig. 3. System boundary for LCA from cradle to farm gate of the study.

zero. MAC is usually used to assess the economic potential for GHG emissions reduction (Moran et al., 2011). In this study, MAC is referred to as a cost for implementing a cropping system to reduce GHG emissions to an anticipated level. The RR cropping system is taken as the business as usual (BAU) case. It was compared with the other cropping systems because the method is traditionally practiced and widely found in the study area as reported in the Supplementary Material (Section S1, Table S1). The MAC (Baht ton
1 CO2eq) of each cropping system was calculated by dividing the total abatement cost (Baht ha
1) (TAC) by total abatement potential (ton CO2eq ha
1) (TAP), and each TAC and TAP were obtained by subtracting each cropping system by the BAU case. The equations and parameters for the B/C ratio, net profit and MAC of each cropping system are shown in the Supplementary Material (Section S3).

2.6. Statistical analysis Statistical analyses of the data were carried out using SPSS (Version 20.0, USA). The mean and standard deviation (SD) values were used to represent GHG emissions and cost-benefit values in each cropping system. Differences in GHG emissions and costbenefit values among irrigated and rainfed areas and farm sizes were analyzed by a t-test and least significant difference (LSD) test (p < 0.05). The effect of different GWP values on the GHG emissions from each crop production was estimated. The IPCC (2013) GWP values were assigned in this study, and the GHG emissions for IPCC 1996 (21 for CH4 and 310 for N2O) and IPCC 2007 (25 for CH4 and 298 for N2O) were compared.

3. Results and discussion 3.1. GHG emissions based on life cycle stage in different cropping systems 3.1.1. Raw materials production There were significant differences among nitrogen (N), phosphorus (P2O5), potassium (K2O) fertilizer production (p < 0.05) and total GHG emissions from raw materials production (p < 0.10), with the exception of diesel, gasoline fuel, insecticide and herbicide production and transportation between irrigated and rainfed areas (Table 1). The GHG emissions rate from the raw materials production varied across all areas from 160.98 to 4364.27 and from 132.25 to 4259.72 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. The average GHG emissions from raw materials production was 1011.81 and 912.53 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively (Table 1). Diesel and gasoline fuel were mainly used for tillage, water management, sowing, spraying insecticide and herbicide, harvesting and storing during farm operations. The N, P2O5 and K2O fertilizers were applied in the land preparation and crop planting stages. The overall GHG emissions from raw materials production showed that the highest GHG emissions came from N fertilizer production with an average of 555 and 443.03 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. The RRW cropping system showed the highest emissions, particularly for watermelon planting after the first and second rice, with total GHG emissions of 4364.27 and 4259.72 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. This was due to the high consumption of agricultural inputs, compared with the case of other cropping systems. On the

Table 1 GHG emissions (kg CO2eq ha
1 yr
1) from production of raw material used in each crop and cropping system. Cropping systems

RR RC RM RS RW RRM

RRS

RRW

RRR

Average

1st Rice Fallow 1st Rice 2nd Rice 1st Rice Corn 1st Rice Mung bean 1st Rice Soybean 1st Rice Watermelon 1st Rice 2nd Rice Mung bean 1st Rice 2nd Rice Soybean 1st Rice 2nd Rice Watermelon 1st Rice 2nd Rice 3rd Rice

Rainfed area

(a)

(b)

(c)**

(d)**

(e)**

(f)

(g)

(h)*

(a)

(b)

(c)**

(d)**

(e)**

(f)

(g)

(h)*

69.53 0.00 76.96 79.09 74.85 190.93 71.50 15.14 70.38 26.12 80.03 384.43 76.06 72.66 15.34 75.96 74.23 29.97 79.50 74.67 441.00 82.82 82.38 81.48 96.88

23.68 0.00 30.78 34.14 30.15 88.55 26.56 3.39 27.24 5.99 35.57 185.50 33.66 32.73 2.77 34.22 33.56 7.95 37.68 34.82 239.69 39.93 41.71 39.57 44.58

471.15 0.00 481.46 472.79 466.39 1087.74 465.21 118.06 448.25 202.46 469.41 2126.94 447.45 422.46 125.47 441.59 430.51 222.87 445.85 423.75 2204.39 458.53 438.66 448.55 555.00

33.06 0.00 40.24 43.35 39.30 109.51 35.76 5.79 36.07 10.11 44.67 226.23 42.34 40.91 5.35 42.76 41.89 12.46 46.24 42.98 280.91 48.71 50.03 48.15 55.28

25.30 0.00 32.42 35.74 31.73 92.18 28.15 3.80 28.77 6.71 37.15 192.55 35.16 34.14 3.22 35.70 35.00 8.73 39.16 36.23 246.82 41.45 43.15 41.06 46.43

135.73 0.00 143.64 143.99 139.39 338.77 136.40 12.08 132.67 20.89 144.22 671.68 137.28 130.33 12.08 136.23 132.96 24.25 139.89 132.22 731.70 144.75 141.11 141.99 167.68

41.91 0.00 39.45 26.64 41.29 77.50 32.38 2.72 35.07 5.32 38.95 212.56 49.95 19.88 3.27 45.67 25.79 6.00 47.29 31.91 219.77 53.66 23.66 22.59 45.97

800.35 0.00 844.97 835.74 823.11 1985.16 795.96 160.98 778.43 277.61 850.00 3999.89 821.90 753.11 167.50 812.12 773.94 312.24 835.61 776.60 4364.27 869.84 820.70 823.41 1011.81

62.82 0.00 69.50 73.42 68.08 199.80 68.36 14.64 67.80 26.27 77.43 392.57 78.82 74.13 17.96 78.19 76.33 33.50 81.61 76.66 539.57 e e e 103.69

17.32 0.00 23.93 27.95 23.66 83.34 21.29 2.03 22.34 3.83 30.60 172.10 37.88 36.15 4.55 38.13 37.33 10.91 41.50 38.46 308.06 e e e 46.73

360.95 0.00 368.10 371.43 359.14 962.95 376.83 97.26 365.75 173.35 384.31 1837.28 289.23 269.38 86.89 284.12 276.78 149.62 287.78 272.97 1729.46 e e e 443.03

24.82 0.00 31.44 35.45 30.98 102.55 29.05 4.11 29.83 7.53 38.33 208.45 35.36 33.81 3.73 35.66 34.93 9.52 39.03 36.11 293.80 e e e 50.69

20.10 0.00 26.71 30.72 26.37 90.45 24.16 2.80 25.11 5.20 33.46 185.56 30.84 29.61 2.25 31.24 30.62 7.03 34.60 31.89 268.24 e e e 44.62

122.89 0.00 129.66 133.47 126.72 353.57 130.52 10.65 127.84 19.16 139.27 747.09 141.70 132.48 15.21 139.73 136.24 28.88 143.23 135.33 895.18 e e e 181.37

33.89 0.00 31.34 19.84 33.51 69.19 26.32 1.12 29.46 2.70 33.19 197.00 43.47 13.93 1.67 39.11 19.74 3.09 40.65 25.84 225.41 e e e 42.40

642.80 0.00 680.67 692.28 668.45 1861.85 676.54 132.62 668.14 238.04 736.60 3740.05 657.29 589.49 132.25 646.18 611.96 242.55 668.39 617.25 4259.72 e e e 912.53

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

RF

Irrigated area

(a) Diesel fuel; (b) Gasoline fuel; (c) Nitrogen (N); (d) Phosphorus (P2O5); (e) Potassium (K2O); (f) Insecticide and herbicide; (g) Transportation; and (h) Total GHG emissions. *, ** ¼ Significant at 10% and 5% level, respectively, between irrigated and rainfed areas.

551

552

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

other hand, the RM cropping system in both irrigated and rainfed areas where mung bean was planted after the first rice had the lowest total GHG emissions, 160.98 and 132.25 kg CO2eq ha
1 yr
1 respectively (Table 1).

respectively for the irrigated area, while GHG emissions for the rainfed area were 562.00, 538.55 and 185.75 kg CO2eq ha
1 yr
1 respectively (Table 2). Water management revealed that rainfed areas generated more GHG emissions from drainage water into the field than did irrigated areas. The average GHG emissions from diesel and gasoline fuel utilization were 123.24 and 33.83 kg CO2eq ha
1 yr
1 for the irrigated area respectively while average GHG emissions were 341.32 and 72.60 kg CO2eq ha
1 yr
1 for rainfed areas respectively (Table 2). This is because the farmers in the rainfed areas needed more agricultural machinery for the drainage of water into the field than did farmers of irrigated areas where water could flow by gravity-fed irrigation systems. The RRW and RM cropping systems were seen as producing the highest and lowest GHG emissions in both irrigated and rainfed areas. For tillage practice, the amount of GHG emissions varied across all areas from 10.23 to 2400.77, and 7.70 to 2709.45 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. The average GHG emissions from tillage practice was 322.09 and 400.71 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively (Table 2). In irrigated areas, the RW cropping system had the highest GHG emissions, consisting of 166.18 and 2400.77 kg CO2eq ha
1 yr
1 for first rice and watermelon planting respectively. For rainfed areas, the highest GHG emissions were found for the RRW cropping system with first rice, second rice and watermelon planting at 85.06, 68.08 and 2709.45 kg CO2eq ha
1 yr
1 respectively. From the survey information acquired, the tillage practice for watermelon and corn production was carried out two times per crop. The farmers performed the primary tillage, followed by soil drying of 10e15 days; then, the secondary tillage was practiced before planting. Meanwhile, only one tillage was conducted for mung bean and soybean. After this, water drainage into the field and seeded broadcasting were conducted without soil drying. These practices caused the highest energy use (diesel and

3.1.2. Land preparation stage The rainy season in Phichit province lasts from about June to November every year and is when the farmers usually start their land preparation for the first rice cultivation (Fig. 2). In this stage, there were significant differences among water management (p < 0.05), basal fertilizer and total GHG emissions from the land preparation stage (p < 0.10) while there was no significant difference among tillage and insecticide and herbicide utilization between irrigated and rainfed areas (Table 2). The range in the amount of total GHG emissions from the land preparation stage was also broad, being from 261.60 to 8687.72 and 306.17 to 10,133.75 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. The average GHG emissions from the land preparation stage were 1712.73 and 1902.34 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively (Table 2). Basal fertilizer was the main source of GHG emissions from the land preparation stage in both irrigated and rainfed areas, with average values of 1233.34 and 1087.60 kg CO2eq ha
1 yr
1 respectively. Basal fertilizer is added to increase soil nutrients for growing seedlings. The farmers usually apply manure during the land preparation stage of the rice cultivation season only and do not apply it for crop rotations. In both irrigated and rainfed areas, the RW cropping system, consisting of first rice, had the highest GHG emissions, 1137.24 and 978.96 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively, while GHG emissions for watermelon planting were 5688.40 and 5207.56 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas, respectively. Conversely, the RRM cropping system, first rice, second rice and mung bean, had the lowest GHG emissions, 733.64, 710.38 and 218.25 kg CO2eq ha
1 yr
1

Table 2 GHG emissions (kg CO2eq ha
1 yr
1) from land preparation stage in each crop and cropping system. Cropping systems

Irrigated area Water management** Diesel

RF

Rainfed area Tillage

Basal* fertilizer Insecticide Total* and herbicide

19.22 0.00 13.56 16.59 121.49 1691.91 146.24 10.23 131.56 27.09 166.18 2400.77 102.93 136.80 26.07 147.55 148.78 66.52 134.18 115.18 1980.43 42.99 39.11 44.71 322.09

481.63 0.00 704.23 785.36 1072.93 3088.86 1011.55 307.28 934.37 474.38 1137.24 5688.40 733.64 710.38 218.25 813.49 749.99 421.33 835.57 690.24 5172.79 1119.62 1176.89 1271.71 1233.34

Gasoline

1st Rice 15.11 4.15 Fallow 0.00 0.00 RR 1st Rice 4.89 1.34 2nd Rice 9.16 2.52 RC 1st Rice 67.18 18.44 Corn 354.75 97.39 RM 1st Rice 83.02 22.79 Mung bean 5.21 1.43 RS 1st Rice 71.45 19.62 Soybean 9.33 2.56 RW 1st Rice 97.08 26.65 Watermelon 469.62 128.93 RRM 1st Rice 94.20 25.86 2nd Rice 87.03 23.89 Mung bean 13.55 3.72 RRS 1st Rice 89.84 24.67 2nd Rice 71.29 19.57 Soybean 39.22 10.77 RRW 1st Rice 94.37 25.91 2nd Rice 73.61 20.21 Watermelon 1154.14 316.85 RRR 1st Rice 23.24 6.38 2nd Rice 9.11 2.50 3rd Rice 21.39 5.87 Average 123.24 33.83

0.45 0.00 0.34 0.22 0.38 0.00 0.36 0.00 0.26 0.00 0.43 0.00 0.28 0.23 0.00 0.34 0.30 0.00 0.36 0.32 0.00 0.41 0.32 0.35 0.22

520.56 0.00 724.36 813.85 1280.42 5232.91 1263.96 324.16 1157.25 513.36 1427.59 8687.72 956.90 958.34 261.60 1075.89 989.94 537.84 1090.40 899.56 8624.22 1192.63 1227.92 1344.03 1712.73

*, ** ¼ Significant at 10% and 5% level, respectively, between irrigated and rainfed areas.

Water management** Diesel

Gasoline

171.67 0.00 151.76 158.97 247.90 614.06 282.57 47.10 284.26 90.40 283.06 1313.23 246.99 222.43 63.62 264.02 241.68 110.38 258.08 222.13 1893.37 e e e 341.32

36.52 0.00 32.28 33.81 52.73 130.62 60.11 10.02 60.47 19.23 60.21 279.34 52.54 47.31 13.53 56.16 51.41 23.48 54.90 47.25 402.74 e e e 72.60

Tillage

Basal* fertilizer Insecticide and Total* herbicide

25.15 0.00 7.70 17.74 103.61 2048.55 109.70 16.32 95.49 49.73 103.89 2452.77 96.50 103.71 43.27 114.36 82.17 81.57 85.06 68.08 2709.45 e e e 400.71

286.36 0.00 498.97 603.15 863.53 2927.08 840.61 265.49 777.39 411.44 978.96 5207.56 562.00 538.55 185.75 644.47 585.04 339.87 666.67 528.51 5128.19 e e e 1087.60

0.24 0.00 0.20 0.14 0.23 0.00 0.28 0.00 0.19 0.00 0.15 0.00 0.14 0.11 0.00 0.26 0.21 0.00 0.14 0.20 0.00 e e e 0.12

519.93 0.00 690.92 813.81 1268.01 5720.31 1293.27 338.92 1217.79 570.80 1426.26 9252.91 958.16 912.12 306.17 1079.27 960.50 555.30 1064.85 866.17 10,133.75 e e e 1902.34

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

553

(p < 0.10), GHG emissions from rice cultivation, fossil fuel usage by machines and total GHG emissions from the planting stage (p < 0.05) were found whilst there was no significant difference for insecticide and herbicide utilization between irrigated and rainfed areas. The average GHG emissions from the planting stage was 6145.30 and 4362.71 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively (Table 3). Rice cultivation showed the highest GHG emissions for planting stage, which varied from 5085.9 to 8546.31 and 2620.41 to 6363.16 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. The average GHG emissions from rice cultivation was 4785.73 and 3020.94 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. As for our results, the RR cropping system generated the highest GHG emission for first rice and second rice, 8519.71 and 8546.31 kg CO2eq ha
1 yr
1 for irrigated areas respectively, and generated 6065.50 and 6363.16 kg CO2eq ha
1 yr
1 for rainfed areas respectively. Considering all GHG emissions in this stage between irrigated and rainfed areas, GHG emissions in irrigated areas were higher than those in rainfed areas (Table 3). This is consistent with Singh et al. (1998), Jain et al. (2000), Majumdar et al. (2000), Majumdar (2003) and Bhatia et al. (2013). These findings are also supported by several studies that reported that CH4 is produced in soils by methanogen bacteria, which function under strictly anaerobic conditions that are permanently flooded and receive manure application (Dubey, 2001; Yan et al., 2009). GHG emissions from the application of chemical fertilizer varied from 118.30 to 7113.79 and 66.68 to 6743.73 with an average of 1305.13 and 1191.68 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. Results point out that the RW cropping system generated the highest GHG emissions in both irrigated and rainfed areas (Table 3). Comparing rice cultivation and crop rotation, growing watermelon and corn as crop rotation contributed to higher GHG emissions from chemical fertilizer than did growing rice. Meanwhile, growing soybean and mung bean produced lower

gasoline fuels) for RW and RRW cropping systems, followed by the RC cropping system. The use of insecticides and herbicides is necessary to protect the seeding or sprout against pests and to control weeds. In this stage, GHG emissions associated with machine usage (gasoline machine) to spray insecticides and herbicides were calculated along with the amounts of insecticide and herbicide application. The farmers usually applied insecticide and herbicide in this stage for rice cultivation (first, second and third rice) but not for crop rotations. The average GHG emissions from insecticide and herbicide applications was 0.22 and 0.12 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively (Table 2). Water management, tillage, and insecticide and herbicide applications can generate GHG emissions from fuel combustion, so policy makers should consider measures to encourage vehicle efficiency improvement to reduce emissions from combustion (IPCC, 2007). Several studies have mentioned that the increased use of agricultural inputs has resulted in increasing the energy consumption for land preparation, planting and harvesting (Cetin and Vardar, 2008; Mohammadi and Omid, 2010; Banaeian and Namdari, 2011; Pishgar-Komleh et al., 2012; Khoshnevisan et al., 2013). However, changing tillage practices has indicated a change in fossil fuel usage (West and Marland, 2002) and could provide additional opportunities for GHG emissions reduction (Brodt et al., 2014). This is consistent with our findings that the RW and RRW cropping systems had the highest GHG emissions for tillage practice (Table 2). It is important to note that using crop rotation can lead to dramatic increases in N use efficiency and soil fertility and reduce N fertilizer input (Plaza-Bonilla et al., 2016). These results are in complete agreement with our findings that a low amount of basal fertilizer was applied for RM, RRM and RRS cropping systems, especially for mung bean and soybean planting (Table 2). 3.1.3. Planting stage Significant differences among chemical fertilizer utilization

Table 3 GHG emissions (kg CO2eq ha
1 yr
1) from planting stage in each crop and cropping system. Cropping systems Irrigated area Machinery**

Rainfed area Fertilizer* Flooding** Insecticide and herbicide Total**

Diesel Gasoline RF

1st Rice Fallow RR 1st Rice 2nd Rice RC 1st Rice Corn RM 1st Rice Mung bean RS 1st Rice Soybean RW 1st Rice Watermelon RRM 1st Rice 2nd Rice Mung bean RRS 1st Rice 2nd Rice Soybean RRW 1st Rice 2nd Rice Watermelon RRR 1st Rice 2nd Rice 3rd Rice Average

7.92 0.00 3.75 8.06 21.56 14.95 67.02 0.29 19.60 0.37 76.86 43.58 54.69 56.49 0.31 44.72 65.85 0.40 87.12 67.18 44.10 18.54 6.38 9.94 29.99

2.43 0.00 1.69 1.09 1.36 1.25 1.70 0.09 1.21 0.12 0.83 4.84 1.20 1.02 0.09 1.63 0.81 0.13 1.23 0.85 4.90 1.13 0.93 1.00 1.31

Machinery**

Fertilizer* Flooding** Insecticide and herbicide Total**

Diesel Gasoline 455.00 0.00 770.55 1042.27 897.76 2985.54 1128.12 118.30 1144.21 283.08 1190.03 7113.79 892.01 1042.31 164.85 905.17 781.27 290.11 1038.96 796.23 5407.02 897.53 988.61 990.49 1305.13

5085.90 0.00 8519.71 8546.31 8487.16 0.00 5360.75 0.00 6317.09 0.00 5849.00 0.00 7457.17 6386.19 0.00 7744.45 7030.02 0.00 7193.49 6803.34 0.00 8341.28 7660.30 8075.46 4785.73

9.42 0.00 29.53 53.21 2.55 0.37 1.30 0.00 3.01 0.35 1.66 162.08 10.00 7.27 0.46 56.32 43.90 11.39 12.25 8.59 126.42 7.07 1.92 6.07 23.13

5560.67 0.00 9325.22 9650.95 9410.39 3002.11 6558.88 118.68 7485.13 283.92 7118.38 7324.29 8415.06 7493.28 165.71 8752.29 7921.85 302.04 8333.05 7676.18 5582.44 9265.55 8658.13 9082.96 6145.30

*, ** ¼ Significant at 10% and 5% level, respectively, between irrigated and rainfed areas.

158.48 0.00 150.58 158.42 186.73 0.00 227.08 41.29 215.57 0.00 201.36 0.00 207.51 181.02 0.00 210.85 214.57 0.00 205.91 187.08 0.00 e e e 121.26

3.30 0.00 2.24 0.96 1.51 2.72 1.38 0.12 1.28 0.11 0.97 7.74 1.25 1.46 0.13 1.99 0.87 0.14 1.35 0.75 3.74 e e e 1.62

236.16 0.00 538.60 836.32 669.86 2774.54 932.37 66.68 964.95 205.30 1009.67 6743.73 695.06 841.41 117.60 709.93 594.15 198.73 847.97 613.86 5428.47 e e e 1191.68

2620.41 0.00 6065.50 6363.16 6141.82 0.00 3255.23 0.00 4358.58 0.00 3799.42 0.00 5446.01 4409.05 0.00 5747.65 5082.54 0.00 5228.45 4921.85 0.00 e e e 3020.94

9.06 0.00 28.40 53.06 2.46 0.40 1.31 0.00 3.07 0.38 1.69 172.62 9.97 7.12 0.50 56.20 43.78 11.94 12.24 8.56 148.55 e e e 27.21

3027.42 0.00 6785.32 7411.93 7002.38 2777.66 4417.37 108.08 5543.45 205.78 5013.11 6924.09 6359.81 5440.07 118.24 6726.62 5935.92 210.81 6295.92 5732.09 5580.76 e e e 4362.71

554

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

GHG emissions than did growing rice. This was due to the differences in the quantity of chemical fertilizer applied during the planting stage. Mosier et al. (2006) and Alluvione et al. (2010) have reported that direct N2O emissions can be mitigated by reducing nitrogen inputs into soils. Machines were used for sowing, planting, water management and spraying insecticide and herbicide. More diesel and gasoline fuel consumption was found in rainfed areas than in irrigated areas (Table 3). The information collected from surveys revealed that the farmers in rainfed areas must pump water to the rice fields about 3e4 times throughout rice cultivation. 3.1.4. Harvesting and storing stage There was no significant difference in GHG emissions from harvesting and storing between irrigated and rainfed areas (p < 0.05; Table 4). This might be because farmers usually used the same harvest machinery in their communities or nearby. However, harvesting and storing rice led to more GHG emissions than did harvesting the crop rotation of soybean and mung bean, while corn and watermelon had no GHG emissions from this stage due to employing manpower for harvesting. 3.1.5. Burning crop residue There was a significant difference in GHG emissions from burning crop residue in the post-harvest stage between irrigated and rainfed areas (p < 0.05; Table 4). Concerning burning rice residue (straw and stubble), there were four kinds of practices: (1) no burning, (2) burning less than 50% of the harvested area, (3) burning more than 50% of the harvested area and (4) burning all rice residue (100% burnt), carried out by 12.24%, 22.68%, 17.73% and 47.35% of all farmers respectively. Meanwhile, 47.35% of the farmers burnt all rice residue, amounting to 28.65% and 18.70% of farmers in rainfed and irrigated areas respectively. There were two main reasons for the farmers to burn rice residues: 1) the ease and

convenience of tillage to prepare for the next crop and 2) to kill diseases and pests remaining in the rice fields. It is also reported by Hrynchuk (1998) that, although rice straw incorporation into the soil can enhance soil fertility, at the same time, it also leads to crop diseases from old crops being transferred to new crops. Ahmed and Ahmad (2013) reported that the inconvenience faced by machinery usage for land preparation is the main reason for the farmers to burn residues. In addition, no farmers burned residues after harvesting all types of crop rotation, and they left crop residue in the fields and conducted tillage with soil to prepare for the next crop. Our results highlighted that rainfed areas contributed higher GHG emissions from burning crop residue than did irrigated areas, where a large amount of crop residue was left in the fields, while most of the crop residue left in rainfed areas was often burned after being slightly grazed by cattle. The average GHG emissions from burning crop residue were 2109.16 and 3013.00 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. Many studies confirmed that not burning crop residue can enhance soil organic matter and, after crop residue decomposition, can generate nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg) and sulfur (S) (Eagle et al., 2000; Dobermann and Fairhurst, 2002) and also reduce GHG emissions to the atmosphere (Wassman and Vlek, 2004). Overall, the dominance of GHG emissions in rice fields indicates that reductions in GHG emissions from rice cultivation must be the main target, particularly in terms of the planting (especially fertilizer application) and crop residue burning stages. Throughout the life cycle stages it was found that the rainfed areas had higher GHG emissions than the irrigated areas because burning rice residues was found as the main source of emissions. For crop rotation systems, GHG emissions were generated mainly from the land preparation stage (Fig. 4). Therefore, the mitigation of GHG emissions from rice fields could be achieved by reducing burning and fertilization practices.

Table 4 GHG emissions (kg CO2eq ha
1 yr
1) from harvesting and storing, and burning crop residues stages in each crop and cropping system. Cropping systems Irrigated area

RF

1st Rice Fallow RR 1st Rice 2nd Rice RC 1st Rice Corn RM 1st Rice Mung bean RS 1st Rice Soybean RW 1st Rice Watermelon RRM 1st Rice 2nd Rice Mung bean RRS 1st Rice 2nd Rice Soybean RRW 1st Rice 2nd Rice Watermelon RRR 1st Rice 2nd Rice 3rd Rice Average

Rainfed area

Harvesting and storing

Burning crop residues Total GHG* emission from burning CH4* N2O* CO2*

Harvesting and storing

Burning crop residues CH4*

N2O*

CO2*

70.64 0.00 150.90 128.66 169.49 0.00 169.70 0.00 233.73 48.32 133.28 0.00 207.05 166.98 43.54 224.23 187.72 61.90 157.29 172.06 0.00 100.93 147.66 89.29 110.97

115.94 0.00 117.92 116.51 110.12 0.00 114.22 0.00 108.19 0.00 114.58 0.00 134.23 128.97 0.00 133.30 132.67 0.00 134.51 129.99 0.00 142.58 135.95 135.92 83.57

71.66 0.00 36.61 45.19 135.38 0.00 177.25 22.15 158.58 43.02 221.64 0.00 214.60 217.27 36.81 212.34 220.39 76.92 222.06 234.14 0.00 e e e 111.71

270.50 0.00 271.17 276.52 261.89 0.00 289.49 0.00 279.09 0.00 290.78 0.00 248.04 230.77 0.00 243.07 239.24 0.00 244.08 234.31 0.00 e e e 160.90

189.71 0.00 190.10 191.98 184.37 0.00 199.97 0.00 192.94 0.00 200.00 0.00 177.39 164.83 0.00 173.73 170.11 0.00 174.21 166.61 0.00 e e e 113.14

4411.02 0.00 4410.84 4416.34 4365.16 0.00 4580.47 0.00 4468.45 0.00 4547.61 0.00 4623.52 4280.52 0.00 4348.04 4227.81 0.00 4583.30 4255.09 0.00 e e e 2738.96

15.34 0.00 15.89 16.13 13.84 0.00 15.25 0.00 14.09 0.00 15.70 0.00 51.37 49.36 0.00 51.01 50.77 0.00 51.48 49.75 0.00 54.57 52.03 52.02 23.69

3033.65 0.00 3081.97 3033.70 3029.97 0.00 3063.96 0.00 2899.55 0.00 3030.36 0.00 2945.58 2789.60 0.00 3058.16 3001.06 0.00 3119.67 2865.43 0.00 3153.66 2962.34 3112.92 2007.57

3164.93 0.00 3215.77 3166.33 3153.92 0.00 3193.43 0.00 3021.83 0.00 3160.63 0.00 3131.18 2967.93 0.00 3242.48 3184.50 0.00 3305.66 3045.16 0.00 3350.81 3150.32 3164.93 2109.16

*, ** ¼ Significant at 10% and 5% level, respectively, between irrigated and rainfed areas.

Total GHG* emission from burning 4871.24 0.00 4872.12 4884.84 4811.42 0.00 5069.92 0.00 4940.48 0.00 5038.39 0.00 5048.95 4676.12 0.00 4764.84 4637.16 0.00 5001.58 4656.00 0.00 e e e 3013.00

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

555

Fig. 4. Contribution by life cycle stage to GHG emission sources in each crop and cropping system: (a) irrigated areas; and (b) rainfed areas.

3.2. Crop yields and GHG emissions from different cropping systems Significant differences in each cropping system in terms of crop yields and total GHG emissions between irrigated and rainfed areas are presented in Table 5. Considering all first and second rice cultivation, the results showed that irrigated areas gained more rice yield than did rainfed areas. For crop rotations, rainfed areas generated more yields than irrigated areas, particularly for corn, soybean and watermelon, while the mung bean yield showed no difference between irrigated and rainfed areas (Table 5). This indicates that the development of irrigation systems has contributed to an improvement in the rice production efficiency of Thai farmers. Meanwhile, nutrients and water remaining in the soil from previous rice cultivation have become available for crop rotation planting. This could be highly effective in rainfed areas with inadequate irrigation systems and dry seasons. The average GHG emissions, considering the two crop years of rice production, varied from 10,117.16 to 14,837.33 and 9133.05 to 13,885.65 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas

respectively and 1.81 to 2.87 and 1.72 to 2.70 kg CO2eq kg
1 rice for irrigated and rainfed areas respectively. Soybean varied from 1123.22 to 1214.02 and 1057.64 to 1085.57 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively and 0.45 to 0.46 and 0.39 to 0.40 kg CO2eq kg
1 soybean for irrigated and rainfed areas respectively (Table 5). This is compared with the study of Kasmaprapruet et al. (2009) who applied the LCA concept to determine GHG emissions of milled rice production and reported that, on average, one kg of rice production in terms of GWP was 2.93 kg CO2eq kg
1 rice. Also, Thanawong et al. (2014) found that the rice production of rainfed systems in the rainy season had the lowest GWP on average, at 2.97 kg CO2eq kg
1 rice; this was followed by irrigated systems in the dry and rainy season with an average of 4.87 and 5.55 kg CO2eq kg
1 rice respectively. The average GHG emissions of soybean production in the state of mato grosso, Brazil were 0.186 kg CO2eq kg
1 soybean (Raucci et al., 2015). However, further research is required for corn, mung bean and watermelon. In this study, the average GHG emissions of corn

556

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

Table 5 Summary of crop yield and GHG emissions in each crop and cropping system. Cropping systems

RF

RR

RC

RM

RS

RW

RRM

RRS

RRW

RRR

1st Rice Fallow Overall 1st Rice 2nd Rice Overall 1st Rice Corn Overall 1st Rice Mung bean Overall 1st Rice Soybean Overall 1st Rice Watermelon Overall 1st Rice 2nd Rice Mung bean Overall 1st Rice 2nd Rice Soybean Overall 1st Rice 2nd Rice Watermelon Overall 1st Rice 2nd Rice 3rd Rice Overall

Irrigated area

Rainfed area

Crop yield (kg ha
1 yr
1)

GHG emissions kg CO2eq ha
1 yr
1

kg CO2eq kg
1 crop yield

5580 0.00 5580 5620 5470 11,090 5440 12,460

10,117.16 0.00 10,117.16 14,116.22 14,495.53 28,611.75 14,837.33 10,220.18 25,057.51 11,981.94 603.81 12,585.74 12,676.38 1123.22 13,799.59 12,689.88 20,011.90 32,701.78 13,532.09 12,339.64 638.35 26,510.08 14,107.02 13,057.95 1214.02 28,378.99 13,722.00 12,569.56 18,570.93 44,862.50 14,779.76 14,004.73 14,640.54 43,425.04

5470 1430 5250 2450 5410 24,210 5160 4860 1530 5080 4950 2680 5090 4850 24,500 5220 4950 5100 15,270

Sig. Diff.

Crop yield (kg ha
1 yr
1)

GHG emissions kg CO2eq ha
1 yr
1

kg CO2eq kg
1 crop yield

1.81a 0.00 1.81 2.51a 2.65a 5.16 2.73a 0.82b

5321 0.00 5321 5329 5318 10,648 5194 13,620

1.72a 0.00 1.72 2.45a 2.60a 5.06 2.67a 0.76b

i*, j*, k** e

2.19a 0.42c

5505 1474

2.11a 0.41c

j*, k** ns

2.41a 0.46d

5317 2625

2.36a 0.40d

ns i*, k*

2.35a 0.83e

5477 25,785

2.27a 0.77e

j*, k** i**

2.62a 2.54a 0.42c

5090 4723 1667

2.60a 2.51a 0.36c

ns i*, j**, k* ns

2.78a 2.64a 0.45d

4982 4849 2809

2.70a 2.55a 0.39d

j**, k** i**, j**, k** i**, k**

2.70a 2.59a 0.76e

4988 4749 28,788

2.66a 2.55a 0.69e

i**, j**, k** i**, j**, k** i**, j**, k**

2.83a 2.83a 2.87a 8.53

e e e e

9133.05 0.00 9133.05 13,065.64 13,848.05 26,913.69 13,885.65 10,359.83 24,245.48 11,634.35 601.77 12,236.12 12,528.44 1057.64 13,586.08 12,436.00 19,917.05 32,353.05 13,238.81 11,835.07 593.47 25,667.34 13,429.25 12,365.93 1085.57 26,880.76 13,252.80 12,105.64 19,974.22 45,332.66 e e e e

e e e e

e e e e

i*, j**, k* i*, j**, k* i*, j*, k* i**

i ¼ Significant difference of crop yield between irrigated and rainfed areas; j ¼ Significant difference GHG emissions in the unit of kg CO2eq ha
1 yr
1 between irrigated and rainfed areas; k ¼ Significant difference GHG emissions in the unit of kg CO2eq kg
1 crop yield between irrigated and rainfed areas. *, ** ¼ Significant at 10% and 5% level, respectively. ns ¼ No significant difference. a kg CO2eq kg
1 rice. b kg CO2eq kg
1 corn. c kg CO2eq kg
1 mung bean. d kg CO2eq kg
1 soybean; and e kg CO2eq kg
1 watermelon.

production were 10,220.18 and 10,359.83 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively and 0.82 and 0.76 kg CO2eq kg
1 corn for irrigated and rainfed areas respectively. Mung bean varied from 603.81 to 638.35 and 593.47 to 601.77 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively and 0.42 and 0.36 to 0.41 kg CO2eq kg
1 mung bean for irrigated and rainfed areas respectively. Watermelon varied from 18,570.93 to 20,011.90 and 19,917.05 to 19,974.22 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively and 0.76 to 0.83 and 0.69 to 0.77 kg CO2eq kg
1 watermelon for irrigated and rainfed areas respectively (Table 5). Overall, the RRW cropping system had the highest GHG emissions, 44,862.50 and 45,332.66 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively. Meanwhile, the RF cropping system showed the lowest GHG emissions, 10,117.16 and 9133.05 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively, followed by the RM cropping system of 12,585.74 and 12,236.12 kg CO2eq ha
1 yr
1 for irrigated and rainfed areas respectively (Table 5). The difference of GHG emission values in our study in terms of other studies is due to the different values of emission factors used

and GWP assigned. Wang et al. (2016), who used the GWP values of IPCC (2013) to calculate GHG emissions per kg of fat and protein corrected milk (FPCM), have reported that the GHG emission was higher than the values of IPCC (2007) and IPCC (1996) for about 4% and 12% respectively. In this study, the difference in GWP values was estimated in GHG emissions expressed per kg CO2eq ha
1 yr
1 and kg CO2eq kg
1 crop yield as presented in Table 6. Based on the IPCC (2013) GWP values, GHG emissions from rice production increased by 9.63% and 23.92% as compared with IPCC (2007) and IPCC (1996) GWP values respectively. GHG emissions from corn, mung bean, soybean and watermelon productions decreased by 12.45% and 16.98% respectively due to CH4 emission from flooding and burning crop residue were not occurred for crop rotations. 3.3. GHG emissions from different farm sizes Classification was done for 144 different farm sizes, divided into small (1.28e8.13 ha), medium (8.13e14.99 ha) and large (14.99e21.84 ha) farms. Results were on a per area basis, and the significant differences among small, medium and large farms are presented in Table 7.

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

557

Table 6 GHG emissions expressed per kg CO2eq ha
1 yr
1 and kg CO2eq kg
1 crop yield based on different GWP values. Crop types

Range

Rice Corn Mung bean Soybean Watermelon a b c d e

kg kg kg kg kg

CO2eq CO2eq CO2eq CO2eq CO2eq

Min. Max. e Min. Max. Min. Max. Min. Max.

kg
1 kg
1 kg
1 kg
1 kg
1

Irrigated area

Rainfed area

GHG emissions based on IPCC 1996

GHG emissions based on IPCC 2007

GHG emissions based on IPCC 2013

GHG emissions based on IPCC 1996

GHG emissions based on IPCC 2007

GHG emissions based on IPCC 2013

kg CO2eq ha
1 yr
1

kg CO2 eq kg
1 crop yield

kg CO2eq ha
1 yr
1

kg CO2 eq kg
1 crop yield

kg CO2eq ha
1 yr
1

kg CO2 eq kg
1 crop yield

kg CO2eq ha
1 yr
1

kg CO2eq kg
1 crop yield

kg CO2eq ha
1 yr
1

kg CO2 eq kg
1 crop yield

kg CO2eq ha
1 yr
1

kg CO2 eq kg
1 crop yield

7378.99 10,948.01 11,955.57 706.34 746.74 1313.94 1420.16 21,724.27 23409.92

1.36a 2.19a 0.96b 0.49c 0.49c 0.53d 0.54d 0.89e 0.97e

8824.73 13,068.26 11,492.59 678.98 717.82 1263.06 1365.17 20,883.01 22,503.38

1.62a 2.58a 0.92b 0.47c 0.47c 0.51d 0.52d 0.85e 0.93e

10,117.16 14,837.33 10,220.18 603.81 638.35 1123.22 1214.02 18,570.93 20,011.90

1.81a 2.87a 0.82b 0.42c 0.42c 0.45d 0.46d 0.76e 0.83e

6640.91 10,234.25 12,118.93 694.24 703.95 1237.23 1269.90 23,298.97 23,365.84

1.29a 2.03a 0.89b 0.42c 0.48c 0.46d 0.47d 0.81e 0.90e

7946.02 12,218.51 11,649.63 667.36 676.69 1189.32 1220.72 22,396.72 22,461.01

1.54a 2.41a 0.85b 0.40c 0.46c 0.44d 0.45d 0.78e 0.87e

9133.05 13,885.65 10,359.83 593.47 601.77 1057.64 1085.57 19,917.05 19,974.22

1.72a 2.70a 0.76b 0.36c 0.41c 0.39d 0.40d 0.69e 0.77e

rice. corn. mung bean. soybean; and watermelon.

Table 7 Summary of GHG emissions from different farm sizes. Cropping systems

Crop yield (kg ha
1 yr
1)

GHG emissions kg CO2eq ha
1 yr
1

RF RR RC RM RS RW RRM

RRS

RRW

RRR

1st Rice Fallow 1st Rice 2nd Rice 1st Rice Corn 1st Rice Mung bean 1st Rice Soybean 1st Rice Watermelon 1st Rice 2nd Rice Mung bean 1st Rice 2nd Rice Soybean 1st Rice 2nd Rice Watermelon 1st Rice 2nd Rice 3rd Rice

kg CO2eq kg
1 crop yield

Small

Medium

Large

Sig. Diff.

Small

Medium

Large

Sig. Diff.

Small

Medium

Large

Sig. Diff.

5724 0.00 5855 5308 5671 12,295 5328 1267 5487 2680 5254 24,148 5326 4794 1367 4916 5186 2519 4928 4984 24,336 5051 4786 4937

5681 0.00 5720 5571 5548 12,863 5578 1539 5351 2557 5713 25,314 5262 4969 1602 5883 5657 2743 5194 4958 24,600 5358 5608 5721

6332 0.00 6263 5916 5279 13,503 5936 1875 5295 3288 5457 26,650 6034 6102 1975 6024 6194 3127 5936 6212 25,344 5959 6094 6145

s**, t** e s*, t* s** ns r**, s**, t** s** s** ns s**, t** ns r**, s**, t** s*, t* s**, t** n** r**, s** s**, t* s** s**, t** s**, t** s**, t** s**, t** r**, s** r*, s**

4564.34 0.00 4575.56 4574.68 4434.31 2934.33 5887.36 1537.21 5765.47 1683.55 6106.07 4309.75 5418.88 5106.24 1664.7 6625.39 5487.42 1639.05 5817.49 5761.47 4205.36 4812.76 4630.82 4636.59

5761.92 0.00 6563.54 6255.04 5136.84 2905.37 5824.98 1145.47 4875.05 1337.44 6142.47 4228.5 6486.07 4735.75 1166.96 5790.84 5417.83 1315.73 6567.06 5424.75 3806.49 5025.73 4993.86 5016.31

7091.66 0.00 7214.28 6208.47 7074.63 2773.84 6525.88 1065.83 6407.64 1313.71 6722.77 4068.54 6465.07 5279.32 1135.81 6328.62 5385.05 1216.93 6514.47 5306.33 3702.54 5710.85 5289.01 5108.08

r**, s**, t** e r**, s**, t* r**, s** r*, s**, t** ns s*, t* ns r*, s*, t** ns s*, t* ns r**, s** t* s* r**, t** ns ns r*, s* ns s* s*, t* s* ns

0.80a 0.00 0.78a 0.86a 0.78a 0.24b 1.10a 1.21c 1.05a 0.63d 1.16a 0.18e 1.02a 1.07a 1.22c 1.35a 1.06a 0.65d 1.18a 1.16a 0.17e 0.95a 0.97a 0.94a

1.01a 0.00 1.15a 1.12a 0.93a 0.23b 1.04a 0.74c 0.91a 0.52d 1.08a 0.17e 1.23a 0.95a 0.73c 0.98a 0.96a 0.48d 1.26a 1.09a 0.15e 0.94a 0.89a 0.88a

1.12a 0.00 1.15a 1.05a 1.34a 0.21b 1.10a 0.57c 1.21a 0.40d 1.23a 0.15e 1.07a 0.87a 0.58c 1.05a 0.87a 0.39d 1.10a 0.85a 0.15e 0.96a 0.87a 0.83a

r**, s**, t* e r**, s** r**, s* s**, t** ns ns r**, s**, t* r*, s*, t** r*, s**, t* t* ns r**, t* r*, s** r**, s**, t* r**, s** r *, s * r*, s** t* s**, t** ns ns s* s*

r ¼ Significant difference between small and medium; s ¼ Significant difference between small and large; t ¼ Significant difference between medium and large. *, ** ¼ Significant at 10% and 5% level, respectively. ns ¼ No significant difference. a kg CO2eq kg
1 rice. b kg CO2eq kg
1 corn. c kg CO2eq kg
1 mung bean. d kg CO2eq kg
1 soybean, and e kg CO2eq kg
1 watermelon.

Large farms showed significantly higher GHG emissions than did small and medium farms (Table 7). This was mainly due to higher quantity inputs of chemical fertilizer, higher fossil fuel consumption and insecticide and herbicide use and burning crop residues. This means that large farms cause a severe environmental impact per ha to obtain high crop yields and economic returns. Table 7 highlights that large farms have the highest crop yields, followed

by medium and small farms respectively. Our findings are in complete agreement with Ullah et al. (2015), who reported the ecoefficiency of cotton-cropping systems in Pakistan. They mentioned that large farms used more crop-care inputs (e.g. weeding and pesticides, phosphate treatments) than did farms of other sizes. Conversely, for per kg of crops produced, small farms generated the highest GHG emissions, compared to farms of other sizes,

558

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

Fig. 5. Cost of each life cycle stage in irrigated and rainfed areas.

Table 8 Summary of total cost (Baht ha
1 yr
1), total benefit (Baht ha
1 yr
1), farmers’ profit (Baht ha
1 yr
1) and B/C ratio in each cropping system. Cropping systems

RF

RR

RC

RM

RS

RW

RRM

RRS

RRW

RRR

1st Rice Fallow Overall 1st Rice 2nd Rice Overall 1st Rice Corn Overall 1st Rice Mung bean Overall 1st Rice Soybean Overall 1st Rice Watermelon Overall 1st Rice 2nd Rice Mung bean Overall 1st Rice 2nd Rice Soybean Overall 1st Rice 2nd Rice Watermelon Overall 1st Rice 2nd Rice 3rd Rice Overall

Irrigated areas

Rainfed areas

Sig. Diff.

Total cost

Total benefit

Farmer’s profit

B/C

Total cost

Total benefit

Farmer’s profit

B/C

15,389.80 0.00 15,389.80 16,608.18 19,940.37 36,548.55 21,074.55 18,294.06 39,368.61 19,122.01 13,013.51 32,135.52 18,191.48 20,886.76 39,078.24 19,076.61 40,542.68 59,619.29 19,156.58 16,073.07 12,458.23 47,687.88 18,877.34 17,821.11 20,577.12 57,275.57 18,806.03 18,346.38 39,426.97 76,579.38 19,245.62 17,685.38 17,508.44 54,439.45

20,403.75 0.00 20,403.75 21,438.71 24,351.86 45,790.57 25,797.12 26,884.01 52,681.13 24,611.17 20,211.34 44,822.51 23,186.19 26,729.57 49,915.77 23,720.77 48,427.29 72,148.07 25,008.71 23,985.80 21,527.81 70,522.32 24,767.07 24,227.92 27,848.21 76,843.20 25,088.81 25,049.33 50,614.07 100,752.21 24,851.47 24,496.72 24,226.18 73,574.37

5013.95 0.00 5013.95 4830.53 4411.49 9242.02 4722.57 8589.95 13,312.52 5489.16 7197.83 12,686.99 4994.71 5842.82 10,837.53 4644.16 7884.62 12,528.78 5852.13 7912.73 9069.58 22,834.44 5889.73 6406.81 7271.09 19,567.63 6282.78 6702.95 11,187.10 24,172.83 5605.85 6811.34 6717.74 19,134.92

1.33 0.00 1.33 1.29 1.22 1.25 1.22 1.47 1.34 1.29 1.55 1.39 1.27 1.28 1.28 1.24 1.19 1.21 1.31 1.49 1.73 1.48 1.31 1.36 1.35 1.34 1.33 1.37 1.28 1.32 1.29 1.39 1.38 1.35

14,776.72 0.00 14,776.72 15,995.09 19,327.29 35,322.38 20,761.46 17,849.78 38,611.24 18,508.93 12,491.23 31,000.16 17,578.40 20,422.48 38,000.88 18,463.53 40,178.40 58,641.93 18,435.30 15,557.11 13,088.97 47,081.37 18,376.27 17,170.94 20,232.66 55,779.86 17,973.62 17,753.98 38,041.82 73,769.41 e e e e

18,729.52 0.00 18,729.52 19,964.48 23,235.58 43,200.06 26,322.89 27,677.64 54,000.53 25,136.95 22,004.97 47,141.92 23,711.97 28,523.21 52,235.17 24,246.55 51,220.93 75,467.48 24,226.28 23,908.70 20,086.45 68,221.44 23,404.85 23,716.60 27,341.66 74,463.12 24,395.25 24,595.78 47,306.83 96,297.87 e e e e

3952.81 0.00 3952.81 3969.39 3908.29 7877.68 5561.43 9827.86 15,389.29 6628.02 9513.74 16,141.76 6133.57 8100.73 14,234.30 5783.02 11,042.53 16,825.55 5790.99 8351.59 6997.49 21,140.07 5028.59 6545.67 7109.00 18,683.26 6421.64 6841.81 9265.01 22,528.46 e e e e

1.27 0.00 1.07 1.25 1.20 1.22 1.27 1.55 1.40 1.36 1.76 1.52 1.35 1.40 1.37 1.31 1.27 1.29 1.31 1.54 1.53 1.45 1.27 1.38 1.35 1.33 1.36 1.39 1.24 1.31 e e e e

m*, n**, p**, q* e m* , n * m* , n * ns p*, q** m*, n*, p*, q* m*, n**, p**, q** m*, n*, p*, q** n**, p**, q** m*, p*, q** n**, p**, q** m* m* m*, n*, p**, q** m* , n * m* ns m* m* m*, n**, p** e e e e

m ¼ Significant difference of total cost between irrigated and rainfed areas; n ¼ Significant difference of total benefit between irrigated and rainfed areas; p ¼ Significant difference of farmers’ profit between irrigated and rainfed areas; q ¼ Significant difference of B/C ratio between irrigated and rainfed areas *, ** ¼ Significant at 10% and 5% level, respectively. ns ¼ No significant difference.

because small farms obtained a lower crop yield than did medium and large farms (Table 7). It should be noted that GHG emissions of crop production could be affected by farm size and management

practices, especially in terms of chemical fertilizer, insecticide and herbicide application, fossil fuel consumption and crop residue burning.

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

559

Fig. 6. Comparison of B/C ratio between (a) irrigated areas and (b) rainfed areas.

Fig. 7. Comparison between abatement cost and abatement potential for each cropping system in irrigated areas.

3.4. Cost and benefit structure, B/C ratio and MAC 3.4.1. Total cost, total benefit, farmer’s profit and B/C ratio The cost structure was evaluated along with agricultural activities throughout the life cycle stages, as reported in detail in the Supplementary Material (Section S3, Tables S5 and S6, for irrigated and rainfed areas respectively). The highest cost came from chemical fertilizer in both irrigated and rainfed areas. There were significant differences for water management costs during the land preparation stage (p < 0.10) and chemical fertilizer cost during the planting stage (p < 0.05) between irrigated and rainfed areas while there was no significant difference for transportation, tillage, seeding, insecticide, herbicide and harvesting costs (Fig. 5). As results show in Table 8 and Fig. 6, the RRM cropping system was the most profitable with a B/C ratio of 1.48, followed by RM > RRR > RRS > RC > RF > RRW > RS > RR > RW cropping systems in irrigated areas. For rainfed areas, the most profitable was the RM cropping system as the B/C ratio was 1.52, followed by RRM > RC > RS > RRS > RRW > RW > RR > RF cropping systems. According to the area-based analysis, in irrigated areas, conducting a triple cropping system was more profitable than conducting a double cropping system while conducting a double cropping system in rainfed areas was more profitable than conducting a triple

cropping system. Moreover, mung bean was the crop rotation that was the most profitable for both irrigated and rainfed areas. 3.4.2. MAC of each cropping system Despite implementing the RF cropping system leading to high GHG reduction cost effectiveness with the negative MAC of
688.63 and
667.00 Baht ton
1 CO2eq for irrigated and rainfed areas respectively (Fig. 9), low profitability was seen (Fig. 6). Therefore, this study strongly supports the implementation of the RRM cropping system for irrigated areas as the negative abatement potential was
5.47 ton CO2eq ha
1, the negative abatement cost was
2378.31 Baht ha
1 (Fig. 7) and the negative MAC was
434.86 Baht ton
1 CO2eq (Fig. 9). In rainfed areas, the RM cropping system is recommended because of the negative abatement potential of
7.34 ton CO2eq ha
1, the negative abatement cost of
2161.11 Baht ha
1 (Fig. 8) and the negative MAC of
294.48 Baht ton
1 CO2eq (Fig. 9). All in all, crop rotations can decrease both costs and environment impacts. This is not only due to the reduction of GHG emissions but also to the reduction of insecticide, herbicide and chemical fertilizer used. The French Ministry of Ecology and Sustainable Development (2010) reported that crop rotation can decrease N fertilizer use by up to 100 kg N ha
1 yr
1, which leads to

560

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

Fig. 8. Comparison between abatement cost and abatement potential for each cropping system in rainfed areas.

Fig. 9. Summary of greenhouse gas reduction cost effectiveness by cropping systems.

reduced GHG emissions from the manufacturing process and transportation. Pretty and Waibel (2005) also mentioned that a reduction in insecticide and herbicide use can be conducive to reduced health care costs and hidden costs. The conservative estimation of these costs for rice cultivation in Germany, the UK, the US and China found that the amounts were from US$8e47 or 287e1683 Baht ha
1 of arable land, with an average of US$4.28 or 153 Baht kg
1 of insecticide and herbicide applied. 4. Conclusions The results revealed that rice cultivation was the major crop for greenhouse gas (GHG) emissions, particularly in terms of the planting and burning rice residue stages. Rice cultivation in rainfed areas led to higher GHG emissions than did rice cultivation in irrigated areas due to more burning of rice residues. GHG emissions from crop rotation systems were mainly generated from the land

preparation stage. The contributions of GHG emissions decreased sequentially, RRW > RRR > RW > RR > RRS > RRM > RC > RS > RM > RF, for the cropping systems in all areas. Large farms showed significantly higher GHG emissions than did farms of other sizes due to higher quantity inputs of chemical fertilizer, higher fossil fuel consumption, higher insecticide and herbicide use and burning crop residues. For per kg of crops produced, small farms generated the highest GHG emissions and obtained lower crop yields, compared to farms of other sizes. This study strongly supports the implementation of a triple cropping system in irrigated areas, suggesting crop rotations with mung bean after the first and second rice harvest (RRM). The RRM cropping system had the highest B/C ratio at 1.48. The negative abatement potential was
5.47 ton CO2eq ha
1, the negative abatement cost was
2378.31 Baht ha
1 and the negative marginal abatement cost (MAC) was
434.86 Baht ton
1 CO2eq. For rainfed areas, a double cropping system with selecting

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

mung bean (RM) was the most profitable with the B/C ratio at 1.52, a negative abatement potential of
7.34 ton CO2eq ha
1, a negative abatement cost of
2161.11 Baht ha
1 and a negative MAC of
294.48 Baht ton
1 CO2eq. Alternative cropping systems with selecting crop rotation are effective methods because of the water crisis situation, particularly in rainfed areas, and the dry season. Moreover, alternative cropping systems not only reduce GHG emissions but also increase benefits to farmers. We therefore conclude that a reduction in the application of chemical fertilizers and a reduction in crop residue burning and energy consumption are important for a reduction in GHG emissions from agricultural activities. Acknowledgments The first author wishes to express gratitude to the China Scholarship Council (CSC) for the opportunity to study at Tsinghua University. This study and article were also financially supported by a part of the research project entitled, “Developing incentive mechanisms to reduce greenhouse gas emissions of rice paddies: An Integrated Assessment Approach of the case study in Phichit province”, and was funded by National Research Council of Thailand (NRCT) (KO-BO-NGO/2557-68) and National Natural Science Foundation of China project (no. 71273153 and no. 71525007). Furthermore, the authors would like to thank the reviewers for their helpful comments to improve the manuscript. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jclepro.2016.08.015. References Ahmed, T., Ahmad, B., 2013. Why Do Farmers Burn Rice Residue? Examining Farmers’ Choices in Punjab, Pakistan. South Asian Network for Development and Environmental Economics (SANDEE), Kathmandu, Nepal. SANDEE Working Paper No. 76e13. Alluvione, F., Bertora, C., Zavattaro, L., Grignani, C., 2010. Nitrous oxide and carbon dioxide emissions following green manure and compost fertilization in corn. Soil Sci. Soc. Am. J. 74 (2), 384e395. Banaeian, N., Namdari, M., 2011. Effect of ownership energy use efficiency in watermelon farms a data envelopment analysis approach. Int. J. Renew. Energy Res. 1, 75e82. Bernard, H.R., 2002. Research Methods in Anthropology: Qualitative and Quantitative Methods, third ed. AltaMira Press, Walnut Creek, California. Bhatia, A., Jain, N., Pathak, H., 2013. Methane and nitrous oxide emissions from Indian rice paddies, agricultural soils and crop residue burning. Greenh. Gases Sci. Technol. 196e211. Brodt, S., Kendall, A., Mohammadi, Y., Arslan, A., Yuan, J., Lee, I.-S., Linquist, B., 2014. Life cycle greenhouse gas emissions in California rice production. Field Crops Res. 169, 89e98. Cai, Z.C., Tsuruta, H., Gao, M., Xu, H., Wei, C.F., 2003. Options for mitigating methane emission from a permanently flooded rice field. Glob. Change Biol. 9, 37e45. Cetin, B., Vardar, A., 2008. An economic analysis of energy requirements and input costs for tomato production in Turkey. Renew. Energy 33, 428e433. Collins, H.P., Elliott, E.T., Paustian, K., Bundy, L.C., Dick, W.A., Huggins, D.R., Smucker, A.J.M., Paul, E.A., 2000. Soil carbon pools and fluxes in long-term corn belt agroecosystems. Soil Biol. Biochem. 32, 157e168. Connor, D., Comas, J., 2008. Impact of smallholder irrigation on the agricultural production, food supply and economic prosperity of a representative village beside the Senegal River. Mauritania. Agric. Sys 96, 1e15. Dobermann, A., Fairhurst, T.H., 2002. Rice straw management. Better Crops Int. 16 (1), 7e11. Doran, J.W., Elliott, E.T., Paustian, K., 1998. Soil microbial activity, nitrogen cycling, and long-term changes in organic carbon pools as related to fallow tillage management. Soil Tillage Res. 49, 3e18. Dubey, S.K., 2001. Methane emission and rice agriculture. Curr. Sci. 81, 345e346. Eagle, A.J., Bird, J.A., Horwath, W.R., Linquist, B.A., Brouder, S.M., Hill, J.E., C., van Kessel., 2000. Rice yield and nitrogen utilization efficiency under alternative straw management practices. Agron. J. 92 (6), 1096e1103. Ecoinvent Centre, 2015. Ecoinvent Database v.3.2. Swiss Centre for Life Cycle Inventories, Duebendorf, Switzerland. Retrieved from. http://www.ecoinvent.org/ .

561

EPA, 2014. Emission Factors for Greenhouse Gas Inventories. United States Environmental Protection Agency. Retrieved from. https://www.epa.gov/sites/ production/files/2015-07/documents/emission-factors_2014.pdf. French Ministry of Ecology and Sustainable Development, 2010. Environmental and Agronomic Advantages of a Revival of Legumes in France. Le point sur Number 40, January 2010. Ghosh, S., Majumdar, D., Jain, M.C., 2003. Methane and nitrous oxide emissions from an irrigated rice of North India. Chemosphere 51, 181e192. Hrynchuk, L., 1998. In: McGuire, Terry (Ed.), Rice Straw Diversion Plan. California Air Resources Board, California. IPCC (Intergovernmental Panel on Climate Change), 1996. In: Houghton, J.T., Meira Filho, L.G., Callander, B.A., Harris, N., Kattenberg, A., Maskell, K. (Eds.), Climate Change 1995: The Science of Climate Change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). Cambridge University Press, Cambridge, United Kingdom. IPCC, 2001. In: Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., van der Linden, P.J., Xiaosa, D. (Eds.), The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the IPCC. Cambridge University Press, Cambridge, UK. IPCC, 2006. Agriculture, forestry and other land use. In: Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. (Eds.), 2006 IPCC Guidelines for National Greenhouse Gas Inventories, vol. 4. Institute for Global Environmental Strategies (IGES), Hayama, Japan. IPCC, 2007. Fourth Assessment Report on Climate Change: Climate Change Impacts, Adaptation and Vulnerability. Cambridge University Press, Cambridge, UK. IPCC, 2013. The Physical Science Basis: Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Jain, M.C., Kumar, S., Wassmann, R., Mitra, S., Singh, S.D., Singh, J.P., Singh, R., Yadav, A.K., Gupta, S., 2000. Methane emissions from irrigated rice fields in northern India (New Delhi). Nutr. Cycl. Agroecosys. 58, 75e83. Kasmaprapruet, S., Kasmaprapruet, W., Saikhwan, P., Phungrassami, H., 2009. Life cycle assessment of milled rice production: case study in Thailand. Eur. J. Sci. Res. 30 (2), 195e203. Khoshnevisan, B., Rafiee, B., Omid, M., Mousazadeh, H., 2013. Prognostication of environmental indices in potato production using artificial neural networks. J. Clean. Prod. 52, 402e409. Lal, R., 2004. Carbon emission from farm operations. Environ. Int. 30 (7), 981e990. Lee, J., Pedroso, G., van Kessel, C., Six, J., 2015. Potential regional productivity and greenhouse gas emissions of fertilizer and irrigated switchgrass in a Mediterranean climate. Agric. Ecosyst. Environ. 212, 64e74. Maciel, V.G., Zortea, R.B., Silva, da W.M., Cybis, L.F.A., Einloft, S., Seferin, M., 2015. Life Cycle Inventory for the agricultural stages of soybean production in the state of Rio Grande do Sul. Braz. J. Clean. Prod. 93, 65e74. Maelinda, D., Noorlidah, A., 2008. Antioxidant from maize and maize fermented by Marasmiellus sp. as stabiliser of lipid-rich foods. Food Chem. 107, 1092e1098. Majumdar, D., 2003. Methane and nitrous oxide emission from irrigated rice fields: proposed mitigation strategies. Curr. Sci. 84, 1317e1326. Majumdar, D., Kumar, S., Pathak, H., Jain, M.C., Kumar, U., 2000. Reducing nitrous oxide emission from an irrigated rice field of North India with nitrification inhibitors. Agric. Ecosyst. Environ. 81, 163e169. Mohammadi, A., Omid, M., 2010. Economical analysis and relation between energy inputs and yield of greenhouse cucumber production in Iran. Appl. Energy 87, 191e196. Moran, D., Macleod, M., Wall, E., Eory, V., McVittie, A., Barnes, A., Rees, R., Topp, C.F.E., Moxey, A., 2011. Marginal abatement cost curves for UK agricultural greenhouse gas emissions. J Agric. Econ. 62, 93e118. Moriizumi, Y., Suksri, P., Hondo, H., Wake, Y., 2012. Effect of biogas utilization and plant co-location on life-cycle greenhouse gas emissions of cassava ethanol production. J. Clean. Prod. 37, 326e334. Mosier, A., Kroeze, C., Nevison, C., Oenema, O., Seitzinger, S., vanCleemput, O., 1998. Closing the global N2O budget: nitrous oxide emissions through the agricultural nitrogen cycle: OECD/IPCC/IEA phase II development of IPCC guidelines for national greenhouse gas inventory methodology. Nutr. Cycl. Agroecosys. 52, 225e248. Mosier, A.R., Halvorson, A.D., Reule, C.A., Liu, X.J., 2006. Net global warming potential and greenhouse gas intensity in irrigated cropping systems in northeastern Colorado. J. Environ. Qual. 35 (4), 1584e1598. Mushtaq, S., Maraseni, T.N., Reardon-Smith, K., Bundschuh, J., Jackson, T., 2015. Integrated assessment of water-energy-GHG emissions tradeoffs in an irrigated Lucerne production system in eastern Australia. J. Clean. Prod. 103, 491e498. Nguyen, T.L.T., Gheewala, S.H., Garivait, S., 2007. Energy balance and GHGabatement cost of cassava utilization for fuel ethanol in Thailand. Energy Policy 35, 4585e4596. OAE, 2014. Agricultural Statistics of Thailand 2014. Office of Agricultural Economics, Bangkok, Thailand. Retrieved from. http://www.oae.go.th/download/ download_journal/2558/yearbook57.pdf (in Thai). Pattey, E., Blackburn, L.G., Strachan, I.B., Desjardins, R., Dow, D., 2008. Spring thaw and growing season N2O emissions from a field planted with edible peas and a cover crop. Can. J. Soil Sci. 88, 241e249. Pishgar-Komleh, S.H., Ghahderijani, M., Sefeedpari, P., 2012. Energy consumption and CO2 emissions analysis of potato production based on different farm size levels in Iran. J. Clean. Prod. 33, 183e191. Plaza-Bonilla, D., Nolot, J.-M., Passot, S., Raffaillac, D., Justes, E., 2016. Grain legumebased rotations managed under conventional tillage need cover crops to

562

N. Arunrat et al. / Journal of Cleaner Production 134 (2016) 547e562

mitigate soil organic matter losses. Soil Tillage Res. 156, 33e43. Pretty, J.N., Waibel, H., 2005. Paying the price: the full cost of pesticides. In: Pretty, J.N. (Ed.), The Pesticide Detox. Earthscan, London, pp. 39e54. Raucci, G.S., Moreira, C.S., Alves, P.A., Mello, F.F.C., Frazao, L. dA., Cerri, C.E.P., Cerri, C.C., 2015. Greenhouse gas assessment of Brazilian soybean production: a case study of Mato Grosso State. J. Clean. Prod. 96, 418e425. Saswattecha, K., Kroeze, C., Jawjit, W., Hein, L., 2015. Assessing the environmental impact of palm oil produced in Thailand. J. Clean. Prod. 100, 150e169. Scheehle, E.A., Kruger, D., 2006. Global anthropogenic methane and nitrous oxide emissions. Energy J. 22, 33e44. Sefeedpari, P., Ghahderijani, M., Pishgar-Komleh, S.H., 2013. Assessment the effect of wheat farm sizes on energy consumption and CO2 emission. J. Renew. Sustain. Energy 5, 023131. Silalertruksa, T., Bonnet, S., Gheewala, S.H., 2012. Life cycle costing and externalities of palm oil biodiesel in Thailand. J. Clean. Prod. 28, 225e232. Singh, J.S., Raghubanshi, A.S., Reddy, V.S., Singh, S., Kashyap, A.K., 1998. Methane flux from irrigated paddy and dryland rice fields, and from seasonally dry tropical forest and savanna soils of India. Soil Biol. Biochem. 30, 135e139. Singhapreecha, C., 2014. Economy and Agriculture in Thailand. Asia-Pacific Information Platform on Agricultural Policy (AP). Food and Fertilizer Technology Center for the Asian and Pacific Region. Smith, D.L., Almaraz, J.J., 2004. Climate change and crop production: contributions, impacts, and adaptations. Can. J. Plant Pathol. 26, 253e266. Thanawong, K., Perret, S.R., Basset-Mens, C., 2014. Eco-efficiency of paddy rice production in Northeastern Thailand: a comparison of rain-fed and irrigated cropping systems. J. Clean. Prod. 73, 204e217. The National Technical Committee on Product Carbon Footprinting (Thailand), 2011. The National Guideline on Product Carbon Footprint, third ed. Amarin Publishing, Bangkok, Thailand. Ullah, A., Perret, S.R., Gheewala, S.H., Soni, P., 2015. Eco-efficiency of cotton-

cropping systems in Pakistan: an integrated approach of life cycle assessment and data envelopment analysis. J. Clean. Prod. 1e10. Wang, X., Kristensen, T., Mogensen, L., Knudsen, M.T., Wang, X., 2016. Greenhouse gas emissions and land use from confinement dairy farms in the Guanzhong plain of China-using a life cycle assessment approach. J. Clean. Prod. 113, 577e586. Wassman, R., Vlek, P.L.G., 2004. Mitigating greenhouse gas emissions from tropical agriculture: scope and research priorities. Environ. Dev. Sustain 6, 1e9. West, T.O., Marland, G., 2002. A synthesis of carbon sequestration, carbon emissions, and net carbon flux in agriculture: comparing tillage practices in the United States. Agric. Ecosyst. Environ. 91, 217e232. Xu, H., Cai, Z.C., Jia, Z.J., 2002. Effect of soil water contents in the non-rice growth season on CH4 emission during the following rice-growing period. Nutr. Cycl. Agroecosys. 64, 101e110. Yan, X., Ohara, T., Akimoto, H., 2003. Development of region-specific emission factors and estimation of methane emission from rice fields in the East, Southeast and South Asian countries. Glob. Change Biol. 9, 237e254. Yan, X., Akiyama, H., Yagi, K., Akimoto, H., 2009. Global estimations of the inventory and mitigation potential of methane emissions from rice cultivation conducted using the 2006 Intergovernmental Panel on Climate Change Guidelines. Glob. Biogeochem. Cycles 23, GB2002. Yuttitham, M., Gheewala, S.H., Chidthaisong, A., 2011. Carbon footprint of sugar produced from sugarcane in eastern Thailand. J. Clean. Prod. 19, 2119e2127. Zheng, X.H., Wang, M.X., Wang, Y.S., Shen, R.X., Gou, J., Li, J., Jin, J.S., Li, L.T., 2000. Impacts of soil moisture on nitrous oxide emission from croplands: a case study on the rice-based agro-ecosystem in Southeast China. Chemosphere Glob. Chang. Sci. 2, 207e224. Zou, J., Huang, Y., 2007. Quantifying direct N2O emissions in paddy fields during rice growing season in mainland China: dependence on water regime. Atmos. Environ. 41, 8030e8042.