Quantifying and Comparing Greenhouse Gas Emissions in Monsoon Rice Production: A Comprehensive Analysis of Transplanting and Broadcasting Sowing Methods in Myanmar

2.1. Estimating On-farm and Off-farm Emissions of Greenhouse Gases

Based on the emission factors for agricultural inputs [23] (Table 1) and CF-Rice’s formulation, default values, and scaling factors [22] (Table 2), this study determined the amount of GHGs emitted from monsoon rice production. According to the Life Cycle Assessment approach guidelines, CF-Rice estimates the carbon footprint, a functional unit of kilogram carbon dioxide equivalent per kilogram of rice (kg CO₂ eq. kg^-1) based on the IPCC Tier 1 methodology [12]. The functional unit used to estimate GHG emissions in this study was area-based, quantified as kilogram carbon dioxide equivalent per hectare (kg CO₂ eq. ha^-1). Therefore, all GHGs were converted to CO₂ equivalent. Fig. (1) shows the types of GHGs and the range of estimates for GHGs released during rice production, from land tilling to straw burning.

2.2. GHG Emissions from Machine Operations

Machine operations included preparing land for seedbeds in the TP and fields in both the BC and TP methods for the main rice fields. Seedbed preparation is required to raise seedlings using the TP method. In the BC method, seeds are directly broadcasted in the field. Fuel represents a direct source of CO₂ emissions, the quantity of which depends on the plowing depth, tractor speed, and types of equipment used. Machinery depends on the duration of working hour for preparation. Machine hours are converted into energy (MJ h^-1) [31], which is then further converted into carbon dioxide emissions using an emission factor for energy (kg CO₂ eq. MJ^-1) [23]. The amounts of CO₂ released during the preparation of seedbeds and land were determined using Eqs. (1 and 2), respectively [23, 31]:

(1)

(2)

where is the CO₂ emissions from seedbed preparation, Fuel_Rate_Seedbed is the amount of fuel used to prepare seedbeds (L ha^-1), EF_Fuel is the emission factor of fuel, Machine_Time_Seedbed is the duration of machine use for seedbed preparation (h ha^-1), CF_Machine is the conversion factor of machine hours to a unit of energy (MJ h^-1), and EF_Machine is the emission factor of machinery usage (kg CO₂ eq. MJ^-1), is the CO₂ emissions from field preparation, Fuel_Rate_Field is the amount of fuel used in rice field preparation (L ha^-1), and Machine_Time_Field is the duration of machine use for rice field preparation (h ha^-1).

2.3. GHG Emissions from Seed Input in Rice Production

Emissions from the production of seed inputs differed between conventional and hybrid seed production. CO₂ emissions from seed input were calculated according to Eq. (3) [22]:

(3)

where is the CO₂ emission from the production of seed input, Seed_Rate is the rate of seed input (kg ha^-1), and EF_Seed is the emission factor from the production of seed input (kg CO₂ eq. kg^-1).

2.4. GHG Emissions from Fertilizer Inputs in Rice Production

N, P, and K fertilizers are the main fertilizers used in rice production. Farmers also use S, Mg, Fe, Zn, and Bo. However, this study only considered emissions from N, P, and K fertilizers. It is noteworthy that N fertilizer contributes to both direct N₂O emissions upon application to rice fields and indirect CO₂ emissions due to its production, transportation, formulation, storage, distribution, and application. P and K fertilizers exclusively cause indirect CO₂ emissions stemming from their manufacturing and utilization processes [23]. CO₂ emissions from fertilizer inputs in rice production were calculated according to Eq. (4) [22, 23]:

(4)

where is the CO₂ emissions from the fertilizer inputs, NFer_Rate is the rate of N fertilizer application (kg N ha^-1), EF_CO2-N2O is the emission factor for direct N₂O emission from N fertilizer (kg CO₂ eq. kg^-1 N), EF_{CO2_N} is the emission factor for indirect CO₂ emissions associated with N fertilizer (kg CO₂ eq. kg^-1 N), PFer_Rate is the rate of P fertilizer application (kg P ha^-1), EF_{CO2_P} is the emission factor for P fertilizer (kg CO₂ eq. kg^-1 P), KFer_Rate is the rate of K fertilizer application (kg K ha^-1), and EF_{CO2_K} is the emission factor for K fertilizer (kg CO₂ eq. kg^-1 K).

2.5. GHG Emissions from Biocides for Protecting Rice Crops

The production, transportation, formulation, storage, distribution, and application phases of pesticides and herbicides cause the emissions of CO₂ and other GHGs. The CO₂ emissions associated with the use of biocides for the protection of crops during rice production were calculated according to Eq. (5) [23]:

(5)

where is the CO₂ emission from the production of pesticides and herbicides, Pesticide_Rate is the rate of pesticide application (kg active ingredient (a.i). ha^-1), EF_Pesticide is the emission factor linked to the production of pesticides (kg CO₂ eq. kg^-1 a.i.), Herbicide_Rate is the rate of herbicide application (kg a.i. ha^-1), and EF_Herbicide is the emission factor for the production of herbicides (kg CO₂ eq. kg^-1 a.i.).

2.6. GHG Emissions Converted from CH₄ Emissions from Soil and Water Management

CH₄ emissions during water and soil management in rice production primarily stem from soil flooding and can be calculated based on the daily emission factors, encompassing water management, pre-season management, and organic amendments during the cultivation period. The assessment of CO₂ emissions from soil and water management involved the conversion of CH₄ emissions to CO₂ emissions using a conversion factor. Eq. (6) was employed for the quantification of CO₂ emissions from soil and water management, based on the scaling and emission factors from the IPCC Tier 1 methodology [12] and CF-Rice [22].

(6)

where is the CO₂ emissions converted from CH₄ emissions in rice fields due to soil and water management, EF_CO2-CH4 is the CH₄ emission from soils converted to CO_2e (kg CO₂ eq. ha^-1 d^-1) [22], Cult_Per is the cultivation period of the rice (days), S_cF_W is the scaling factor for water regimes during the cultivation period, S_cF_p is the scaling factor for water regimes in the season before the cultivation period, ROA_Straw is the application rate of straw (t ha^-1), CFOA_Straw is the conversion factor for straw amendment in terms of time before cultivation, ROA_{Add_Org} is the application rate of compost, farm yard manure and green manure (t ha^-1), and CFOA_{Add_Org} is the conversion factor for compost, farm yard manure, and green manure. The exponent of 0.59 refers to an uncertainty range of 0.54-0.64 as an average.

2.7. GHG Emissions from Harvesting

Combined harvesters are a widespread and popular harvesting method in Myanmar. To determine the GHG emissions linked to harvesting operations, the quantities of fuel and machinery utilized were multiplied by the corresponding emission factors described in Eq. (7) [23]:

(7)

where is the CO₂ emission from harvesting.

2.8. GHG Emission from Straw Burning

CO₂ is released as a greenhouse gas when rice straw is burned after harvesting. The calculation of CO₂ emissions resulting from straw burning was done using Eq. (8):

(8)

where is the CO₂ emission from straw burning, Straw_Rate is the amount of rice straw, and EF_Straw is the emission factor for straw burning (kg CO₂ eq. t^-1). Approximately 66% of the farmers initially harvested the straw for use as cattle feed but later burned the remaining portion, totaling 1.50 t ha^-1 of rice straw (default straw amount used in the Philippines as per CF-Rice). The remaining farm households (about 34%) burned a total of 3 t ha^-1 of rice straw.

2.9. Description of the Study Area and Data Collection

Nay Pyi Taw Union Territory, the administrative capital of Myanmar, was selected as the study area due to previous reports [32, 33] highlighting the excessive use of N fertilizer, the second largest source of GHG emissions, above rates recommended by the Land Use Division of the Ministry of Agriculture, Livestock, and Irrigation for rice fields, namely 102 Kg N ha^-1, 90 Kg N ha^-1, and 57 Kg N ha^-1 for low, medium, and high fertility conditions, respectively. In lowland areas, the main rice production area in Myanmar [32], < 70 Kg N ha^-1 of N fertilizer is used. By comparison, in the Nay Pyi Taw and Taungoo Regions of central Myanmar, nitrogen application rates of > 100 Kg N ha^-1 [33] and > 115 N ha^-1 have been reported, respectively. Despite being situated in a dry zone, this region exhibits a subtropical climate with a high annual drought index, an annual rainfall range of 859 to 1,273 mm, an annual temperature range of 23.4 to 34.4°C, and relative humidity between 72% and 75% [34, 35].

The Nay Pyi Taw Union Territory comprises two districts, Ottarathiri and Dekkhinathiri, each further divided into four townships. Ottarathiri District consists of Tatkon, Zeyarthiri, Ottarathiri, and Pobbathiri Townships, while Dekkhinathiri District encompasses Pyinmana, Lewe, Zabuthiri, and Dekkhinathiri Townships. For data collection, Tatkon and Lewe Townships were selected as they had large sown areas. Zabuthiri and Zeyarthiri Townships were selected based on their comprehensive adoption of all cultivation practices (Fig. 2). Rice is the major crop in these townships, covering an extensive area of approximately 69,623 ha, roughly 48% of the total cropland. The average rice yield in the year 2020 was 4.70 t ha^-1. Monsoon rice is cultivated in 90% of the total rice sown area, with summer rice accounting for the remaining 10%. Cultivation practices for rice production include broadcasting (69%), direct seeding with a seeder (16.8%), transplanting by seedlings raised in the field (8.75%), transplanting by seedlings raised in the bed (3.24%), line sowing (1.91%), rice transplanting machine (0.15%), and the system of rice intensification (0.02%) [36].

Detailed information on rice production was systematically collected from a randomly selected group of 317 farmers who willingly participated in interviews and completed structured questionnaires between January and May 2022. The questionnaires were designed to elicit in-depth information on rice production in the four townships, covering aspects such as the amount of fuel used for irrigation, seedbed and land preparation, as well as harvesting. They also addressed the number of seeds sown, the application rates of N, P, and K fertilizers, pesticides, and herbicides, details on water management, soil management of straw and organic matter, and practices related to straw burning. The operational duration of machinery used for tillage and harvesting was also determined. Owing to the limitations of this study, respondents from the sampled farm households were asked about the fate of the rice straw: whether it was used as cattle feed, incorporated into the soil, or burned. In cases where straw incorporation was reported, the time of incorporation was noted. Farmers could not precisely estimate the amount of straw left in the rice fields. Therefore, the present study used a predetermined value of straw amount (3 t ha^-1), specifically tailored to the Philippines [22].

2.10. Statistical Analysis

The mean and median values of fuel, machinery, seed inputs, biocides, and N, P, and K fertilizers used in rice production were determined using summary statistics. Frequencies and proportions were calculated to summarize categorical variables related to sowing techniques, soil and water management practices, training attendance, and attitudes toward alternate wetting and drainage practices. Statistical differences between the BC and TP sowing methods were calculated using the Student t-test.

3.1. Cultural Management Practices of Sample Farmers in Monsoon Rice Production

The management practices in monsoon rice production are listed in Table 3. The most employed tillage system in the study area involved a combination of machines (tractors or power tillers) and draft cattle (80.1%). The first and second steps involved plowing and harrowing by machines, respectively, while the third step involved leveling with draft cattle. A total of five rice varieties are grown in the area,

including four high-yielding varieties (Ayamin, Manaw- thukha, Thaihnankaut, and Thukhahmwe) and one hybrid variety (Pearlthwe), which is grown by about 1.89% of the farmers. About 34.7% of the farmers incorporated straw into the field for more than one month, while the other 65.3% burned the rice straw after harvesting. More than 50% of the sampled farmers did not use organic matter for soil management. Farmyard manure was the most used organic matter in the study area, with an average amount of 2.74 t ha^-1 in the TP and 2.21 t ha^-1 in the BC method. Due to the plant lodging, only five farmers (1.58%) manually harvested the rice fields in the study region.

3.2. Greenhouse Gas Emission Sources in Monsoon Rice Production

Table 4 displays the descriptive statistics of GHG emissions from monsoon rice production. Among the cultural management activities, soil and water management were the biggest contributors to GHG emissions, accounting for 64.8% of the total GHG emissions. Inorganic fertilizer management and straw burning accounted for 14.4% and 13.1% of total GHG emissions, respectively. Activities that accounted for < 5% of emissions included seedbed preparation, land preparation, seed input, biocide utilization, and harvesting. Together, these activities accounted for < 10% of emissions. Similar to the findings of Arunrat et al. [38], methane emissions from soil and water management were the highest. However, the second and third major sources of GHG emissions differed from their study. In this research, soil and water management accounted for about 72% of emissions, the burning of rice straw and stubble accounted for 14%, and the use of chemical fertilizer accounted for 11%. Other rice production practices accounted for < 5% of emissions [38]. The difference could be explained by the fact that 66.2% of farmers in this study used the harvested straw for cattle feed, with the amount assumed to be half of the total.

3.3. Comparison of Input Inventory, Greenhouse Gas Emissions, and Yield between BC and TP Sowing Methods

The comparison of input inventories and GHG emissions from monsoon rice production between BC and TP sowing methods is provided in Table 5. For the TP method, seedbed and field preparation emissions were 9.46 kg CO₂ eq. ha^-1 and 123.8 kg CO₂ eq. ha^-1, respectively. These emissions were still less compared to the BC method (162.1 kg CO₂ eq. ha^-1). The emission from crop establishment and protection under the BC method (102.4 kg CO₂ eq. ha^-1) was significantly higher than that of the TP method (49.0 kg CO₂ eq. ha^-1) at a 1% significance level. This difference can be attributed to the BC method using more seeds and herbicides than the TP method. It was found that farmers using the BC method employed twice as many herbicides as those using the TP method. This result aligns with the findings of Saha et al. (2021), who found that the BC method had a higher weed density compared to other direct-seeding methods such as line-seeding or drill seeding methods because of non-uniform crop stand in the former [39]. The total GHG emission from the BC method was significantly greater than that of the TP method, totaling 4,286.1 kg CO₂ eq. ha^-1, at a 1% significance level. This result is consistent with findings from a field experiment in 2020 where GHG emissions from plots with the line-sowing method increased by 21% for continuous flooding and 22% for the alternate wetting and drying method when compared to the TP method [28]. Rice yields were approximately 4,872.1 kg ha^-1 and 4,617.6 kg ha^-1 using the TP and BC methods, respectively. The differences between the two sowing methods were statistically significant at a 1% significance level. Thus, the TP method is a more effective sowing method than the BC method for reducing GHG emissions while increasing rice yield.