Greenhouse gas emissions from rice straw burning and straw-mushroom cultivation in a triple rice cropping system in the Mekong Delta

Abstract

The Mekong Delta produces 21 Mt of rough rice (Oryza sativa L.) and an estimated 24 Mt of straw (dry weight) annually. Approximately one fourth of the straw is burned on the field, which is a common practice in intensive rice cultivation systems in this region because there is limited time to prepare the field for the next crop. The spread of intensive rice production in the Delta may increase the total biomass of burning crop residues, significantly impacting greenhouse gas (GHG) emissions in Vietnam. In this study, GHG emissions from the major uses of straw (burning and mushroom beds) were monitored in a triple rice cropping system located in the central Mekong Delta. Between September 2011 and November 2012, both wind tunnel and closed chamber methods were used to measure the emissions of major GHGs from straw-burning and straw-mushroom cultivation systems, respectively. The global warming potential (GWP) was then determined. Methane (CH₄) and non-methane volatile organic carbon emissions (NMVOC) increased with lower modified combustion efficiency [MCE: emissions ratio of Carbon composing carbon dioxide (CO₂-C) and carbon monooxide (CO-C) (CO₂-C/(CO-C + CO₂-C))]. Furthermore, higher moisture straw stacks generated lower nitrous oxide (N₂O) emissions. Small straw stacks (5 or 10 kg dry straw) with higher moisture content emitted more carbon monoxide (CO), CH₄ and NMVOC. These results suggest that factors that increase the straw moisture content, such as rainfall, can cause smoldering combustion in small straw stacks or when straw is scattered on the ground, thereby inhibiting N₂O emissions but enhancing CO, CH₄ and NMVOC. The measured N₂O emissions contributed negligible amounts to the GWP compared with measured CO and CH₄, which are relatively intense GHG emissions; this was likely a result of the slow and inefficient burning that was observed from the smaller straw stacks with higher moisture content. In this study, rice straw burning threatened to generate more GHGs than straw-mushroom (Volvariella volvacea (Bul. ex Fr.) Singer) cultivation under the studied agroecosystems.

Key words:

• greenhouse gases
• Mekong Delta
• straw burning
• straw-mushroom cultivation
• triple rice cropping

INTRODUCTION

In Asia, approximately 730 Mt of biomass are burned every year by both human and natural influences. On-field crop residue burning was estimated to occupy 250 Mt of the biomass burning (Streets et al. 2003). Rice straw is one of the main residues produced in the region, and its management varies widely. Open burning of straw is a common practice in rice straw management in Asia (Gadde et al. 2009).

Vietnam was the world’s fourth largest rice producer in 2012 and the second largest rice exporter in 2011 (FAOSTAT 2013). The Mekong Delta has played a central role in sustaining Vietnam’s high level of rice production. Although the delta (3.9 Mha) only accounts for approximately 10% of Vietnamese area, half of the national rice production and approximately 90% of annual rice exports originate from this region (Can Tho Statistical Office 2010). In the past, rice production in the Mekong Delta was constrained by severe floods during the rainy season, and other hydrological limitations (Kaida 1974). During the rainy season, almost 50% of the total delta area (2 Mha), particularly the northern portion, was flooded due to overflow from the upper reaches of the Mekong River (Chiem 1994; Minh et al. 1997). Since 1981, drastic water management alterations have been undertaken in the Vietnamese Mekong Delta to improve rice production (Duong and Cho 1994; Xuan and Matsui 1998; Hashimoto 2001; Minh and Kawaguchi 2002). Accompanied by the introduction of non-photoperiod-sensitive high-yield rice varieties, infrastructure development for water resource management (e.g., the large-scale establishment of canals, irrigation/drainage systems and water sluices) has enabled rural farmers to cultivate rice two or three times annually (Kono 2001; Fujii et al. 2003; Cho 2005).

With these improved conditions for rice production, the Mekong Delta yields 21 Mt of rough rice (in 2008; Can Tho Statistical Office 2010) and an estimated 24 Mt of straw (dry weight of the total aboveground biomass; Hong Van et al. 2014) annually. Approximately 6.1 Mt of crop residue is burned annually on-field in Vietnam (the sixth largest amount in Asia; Streets et al. 2003). In one triple rice cropping system in the central Mekong Delta, most of the rice straw harvested during the dry season was burned on-field; the straw harvested during the rainy season was removed from paddies, utilized for straw-mushroom (Volvariella volvacea (Bul. ex Fr.) Singer) cultivation, and then sun-dried for several days outdoors; consequently, the well-dried mushroom beds were burned by indigenous farmers to remove the mushroom beds and to sell the ash. Consequently, 23.0% of the total aboveground straw biomass was burned annually in the triple rice cropping system (Hong Van et al. 2014). As for the straw mushrooms, it became one of the most extensively cultivated mushrooms in tropical and sub-tropical regions and currently ranks fifth in terms of annual production worldwide (Chen et al. 2003; Chang 1999); however, its biological efficiency is considerably lower compared with other widely cultivated species (i.e., conversion of growth substrate into mushroom fruit bodies; Chen et al. 2003). Hong Van et al. (2014) suggested that on-field burning of rice straw may be commonly practiced in intensive rice production systems when there is a short time to prepare the field for the next crop. With the large amount of rice straw in the delta, expansion of this type of production system may significantly increase the amount of crop residue burned. Consequently, this change in practice could considerably affect greenhouse gas (GHG) emissions from biomass burning in Vietnam.

The magnitudes of gas and particle emissions during open burning depend strongly on the type of crop and the burning method (US EPA 1995; Turn et al. 1997; Hays et al. 2005). However, according to Gadde et al. (2009), nitrous oxide (N₂O) emissions factors for rice straw burning are not yet available. Furthermore, emissions factors for each gas species cited from different papers were to calculate the gaseous emissions from open field burning of rice straw in India, Thailand and the Philippines in Gadde et al. (2009). These results indicated that few straw burning experiments integrally quantified major GHG emissions factors [e.g., carbon dioxide (CO₂), carbon monoxide (CO), methane (CH₄), N₂O and non-methane volatile organic carbon emissions (NMVOC)]. Furthermore, there has been limited study of the effect of straw burning conditions (e.g., moisture of rice straw or size of rice straw stacks) on the amount of each gas emitted during straw burning. Since the moisture content of burned biomass affects carbon (C) and nitrogen (N) emissions (Chen et al. 2010) and emissions factors—for example, residue moistness has been shown to be positively correlated with particle emissions (Darley et al. 1974; Oanh et al. 2011)—the effects on the burned straw condition also need to be taken into account when considering regional GHG emissions. High moisture content was shown to enhance the emissions of gases that originated from incomplete combustion, typically CO (Hayashi et al. 2014). Although many studies have been conducted to obtain emissions factors from crop residue burning (e.g., US EPA 1995; Andreae and Merlet 2001; Hays et al. 2005; Sahai et al. 2007; Zhang et al. 2008; Oanh et al. 2011), few of these studies considered the effects of residue moistness on emissions factors; therefore, current emissions inventories do not explicitly incorporate the effects of residue moistness on gas and particle emissions. Even in the Mekong Delta, where huge amounts of rice straw are produced annually, few straw burning experiments have been conducted to quantify GHG emissions from indigenous straw burning (burning straw scattered on soils or piled up as stacks) and from conventional straw management. Furthermore, few studies have been conducted worldwide to integrally calculate regional GHG emissions from a rice-based agricultural system by quantifying both the biomass of rice straw produced and/or burned annually and the amount of GHGs emitted from straw burning.

According to Hong Van et al. (2014), rice (Oryza sativa L.) grains in a triple rice cropping system in the Mekong Delta are collected by hand or with combine-harvesters, and the straw is treated by one of the following practices: (1) the aboveground biomass of rice is hand-reaped and put into a rice-threshing apparatus; then, the straw is extracted and piled in a stack for burning; (2) the aboveground biomass of rice is collected and threshed with combine-harvesters and the threshed straw is scattered on the soil; or (3) the hand-reaped straw is burned after being utilized as mushroom beds for straw-mushroom cultivation. These different straw burning practices may influence the GHG emissions. However, few previous studies have assessed the exhaust gas emissions and composition regarding several indigenous straw-burning conditions in tropical-intensive rice cropping systems, such as the size of straw stacks and their moisture contents.

This study was conducted to quantify the GHG emissions under several straw-burning conditions in a triple rice cropping system of the Mekong Delta. The GHG emissions from the primary straw management practices in this cropping system were measured using both wind-tunnel and closed-chamber methods. The GHG emissions from straw burning were compared to the emissions from mushroom cultivation using rice straw, and the different burning conditions of the straw piles, and the water content, were also tested. Furthermore, the emissions from each GHG species were reported based on the direct and/or indirect global warming potential (GWP) as defined by the Intergovernmental Panel on Climate Change (IPCC 2007). The indirect GWP of CH₄, CO and NMVOC were also discussed because these gases are linked to ozone formation or destruction, enhancement of stratospheric water vapor, decrease of the hydroxide (OH) radical concentration with the main effect of extending the lifetime of CH₄.

Hong Van et al. (2014) quantified the dry-weight of rice straw that is annually burned or utilized for mushroom beds in a triple rice cropping system of the central Mekong Delta. In this study, the GWP from our straw burning and straw-mushroom cultivation experiment, which were conducted with a wind tunnel and by the closed chamber method, were integrated with the results reported by Hong Van et al. (2014).

MATERIALS AND METHODS

Field site description

This study was conducted from September 2011 to November 2012 in a typical triple rice cropping system in Tan Loi 2 Hamlet, Thot Not District in Can Tho City, Vietnam (10^○18ʹN, 105^○54ʹE) located in the central Vietnamese Mekong Delta. The silty-clay Fluvisol soils (clay 52%, silt 48%, sand 0.3%; FAO Soil Classification) of the area were formed by a Holocene complex of river levees and basins. Based on temperature and precipitation data (Can Tho Statistical Office, 2013), the climate of the Can Tho City region is classified as a tropical savannah in the Koppen climate classification (annual mean temperature 27.7°C, annual precipitation 1227 mm in 2012). Seasonal precipitation changes caused by monsoons divide the year into dry and rainy seasons. Approximately 90% of the annual precipitation falls during the rainy season, which generally begins in April or May and lasts until November or December (Hong Van et al. 2014).

Irrigation water is supplied by the Hau River, a tributary of the Mekong River, and a triple rice cropping system has been developed in the area. Although the study site is 100 km upstream from the estuary, the altitude of the paddies in the study area ranges from −1 to 2 m above mean sea level (a.m.s.l). Surrounded by irrigation canals and dykes, irrigation is managed with sluices and pumping apparatuses attached to the paddy dykes. The ebb and flow of the tide causes daily water level fluctuations of approximately 1 m in the irrigation canals. As a result, farmers commonly use tidal irrigation when the water is high enough in the canals. Farmers also use irrigation water pumps when the water levels are lower than the soil surface. Paddies are flooded for about 2 months each year, starting at the beginning of September because the canal water overflows the surrounding dykes. These floods generally moderate in the early part of November, allowing farmers to begin rice cultivation. Farmers generally pump paddy water to irrigation canals and immediately start leveling fields and directly broadcasting seeds on wet fields in mid-November. Triple rice cropping is performed over approximately 10 months between the flooding seasons. The cropping seasons are designated winter–spring (approximately November–February), spring–summer (March–May) and summer–autumn (June–September).

The lowest precipitation rates were observed during winter–spring, the driest season, with increasing precipitation during later cropping seasons. It rained most intensely in the summer–autumn cropping season, and the rice was harvested just prior to potentially devastating floods (Hong Van et al. 2014). Straw management, wet-land tillage with a hand-tractor and direct wet seeding for the next cropping season were performed immediately after the winter–spring and spring–summer rice harvests. A limited fallow period of about 1 week between cropping seasons constrained the rice harvest, with the exception of the 2-month flooding season. Several conventional Indica rice cultivars were grown at the study site during each rice-growing season. These consisted of Jasmine 85, an aromatic long-grain rice cultivar developed in Thailand; OM4218 and OM1490 rice cultivars developed at the Cuu Long Delta Rice Research Institute, Vietnam; and IR50404, a rice cultivar developed by the International Rice Research Institute in the Philippines.

Straw burning experiment using a wind-tunnel method

In February 2012, the rice was harvested at the end of the winter–spring season in the Tan Loi 2 hamlet. Rice straw was air dried immediately after harvest and used for mushroom beds (Jasmine 85, 370 g C per kg dry straw, 9.4 g N per kg dry straw, 12.6% moisture). In this experiment, two types of rice straw were used: (1) rice straw harvested directly from the paddies, which was adjusted for several moisture content variations; and (2) rice straw from mushroom beds (342 g C per kg dry straw, 19.8 g N per kg dry straw, 18.2% moisture) that was burned in a specialized burning apparatus. Several size variations of conical straw stacks were prepared. Due to the variable moisture content of rice straw just after harvest, the moisture content was adjusted with tap water before ignition as follows: (A) fixed amounts of rice straw (5, 10, 20 or 40 kg oven dried) were prepared in stacks; (B) each stack was piled up conically by hand with a bottom diameter of 2 m, within 30 minutes, up to heights of 25 cm (5 kg), 50 cm (10 kg), 100 cm (20 kg), and 200 cm (40 kg) (Table 1); and (C) each pile of straw piled by hand in a stack during procedure B was splayed uniformly and mixed well with tap water. Subsamples of straw were then collected to measure the moisture content at ignition time. A rodlike (1 m length) K-type sheathed thermocouple (TMBDS-KS48Ⅱ/316L-1000-WX13-4000, Yamari Industries, Ltd., Japan), connected with a TM20 Thermo-Collector (Yokogawa Meters & Instruments Corporation, Japan), was suspended with metallic wires and inserted into the conical straw stacks from the top of the stacks to the center of the bottom circle vertically until the lower end of the thermocouple reached the ground. The internal temperature was logged every 10 seconds during burning.

Table 1 Ignition loss of carbon (C) and nitrogen (N) at each straw burning condition (straw of Jasmine 85 harvested in February 2012)

Straw burning condition					Straw-ash after burning			Ignition loss
	Dry weight	Moisture	Carbon content	Nitrogen content	Dry weight	Carbon content	Nitrogen content	Calculated carbon loss*	Calculated nitrogen loss*
Condition	(kg)	(%)	(kg)	(kg)	(kg)	(kg)	(kg)	(g C kg dry straw^−1)	(g N kg dry straw^−1)
Harvested straw	5	12.6	1.85	0.05	1.01	0.04	0.002	361	9.1
	5	23.1	1.85	0.05	1.07	0.06	0.003	357	8.9
	5	30.9	1.85	0.05	1.24	0.09	0.004	352	8.7
	5	52.2	1.85	0.05	1.50	0.06	0.003	358	8.8
	10	19.8	3.70	0.09	1.83	0.10	0.005	360	8.9
	10	23.6	3.70	0.09	2.11	0.08	0.005	362	8.9
	10	32.0	3.70	0.09	2.26	0.12	0.005	358	8.9
	10	50.9	3.70	0.09	2.40	0.08	0.003	362	9.1
	20	12.6	7.40	0.19	3.74	0.31	0.013	354	8.7
	20	20.6	7.40	0.19	3.69	0.20	0.009	360	8.9
	20	33.7	7.40	0.19	3.88	0.13	0.006	363	9.1
	20	65.0	7.40	0.19	4.53	0.21	0.008	360	9.0
	40	17.1	14.80	0.38	7.58	0.38	0.016	360	9.0
	40	29.0	14.80	0.38	8.02	0.23	0.010	369	9.1
	40	38.1	14.80	0.38	6.89	0.18	0.008	384	9.2
	40	61.4	14.80	0.38	9.54	0.17	0.004	362	9.3
Mushroom beds	10	18.2	3.42	0.20	4.62	1.16	0.086	225	11.2
	20	18.2	6.83	0.40	6.81	0.90	0.076	297	16.0

*See text (Eq. 1).

The pipe house-shaped straw-burning tunnel (4.5 m wide × 15 m long × 5 m high; 308 m³) was constructed with iron sheets galvanized and coated with zinc and oxidized primer (hydro-desulfurized heavy naphtha) as described by Miura and Kanno (1997). A gas-discharging fan (465 mm in diameter) was attached at a height of 3.2 m on the outlet side of the wall, and the center of the straw stack bottoms were placed 3 m from the outlet. To maintain a constant air-flow rate of 95 m³ minute⁻¹, discharging inner air through the outlet and letting outside air in, the inlet aperture was adjusted during each straw-burning episode according to the tunnel’s inner air pressure. Furthermore, a gas sampling tube was attached to the wall in front of the fan blades, allowing for the collection of air in the tunnel just prior to discharge through the fan.

For each sample, 500 mL of air was collected at the outlet with the sampling tube through 100-mL plastic syringes. Ambient air samples were collected at the tunnel inlet through 100-mL plastic syringes. The samples were collected in 3-L Tedlar bags and 10-mL of evacuated air glass vials with butyl rubber stoppers. From ignition to 1 hour after ignition, the air samples were collected every minute in Tedlar bags and every 5 minutes in evacuated bottles. From 1 to 2 hours after ignition, the air samples were collected every 2 minutes in Tedlar bags and every 10 minutes in evacuated vials. Thereafter, air samples were collected every 3 minutes in Tedlar bags and every 20 minutes in evacuated vials; the collection continued until differences between the inlet and outlet concentrations of CO₂, CO and NMVOC in the Tedlar bags were undetectable with a gas sensor (Multi RAE IR plus, RAE Systems, Inc. USA). Air samples collected at the inlet and the outlet using Tedlar bags were tested immediately after air sampling to measure the concentrations of CO₂, CO, oxygen (O₂) and NMVOC immediately after air sampling. The air samples in the evacuated vials were analyzed using a gas chromatograph (Shimadzu 14B, Japan) equipped with a flame ionization detector and an electron capture detector to quantify the concentrations of methane and nitrous oxide, respectively.

After burning, the straw-ash was collected, and its chemical properties were analyzed. The ignition loss of C and N was calculated using Eq. 1 below, and the calculated C and N loss was assumed to be equivalent to the ignition losses:

(1)

where IL is the ignition loss (kg C or N), dw.str is the dry weight of the straw before burning (kg dry straw), CNstr. is the C or N content of the straw (kg C or N kg dry straw⁻¹), dw.ash is the dry weight of straw-ash (kg dry straw ash), and CNash is the C or N content of straw-ash (kg C or N kg dry straw ash⁻¹).

The gas fluxes were calculated using Eq. 2 below, and the cumulative emissions (emission factors, g C or N kg dry straw⁻¹) over the experimental period were determined with a linear interpolation and numerical integration using the trapezoidal rule (Whittaker and Robinson 1967):

(2)

where F is the gas flux (in mg C kg straw⁻¹ second⁻¹ for CO₂, CO, CH₄ and NMVOC; and mg N kg straw⁻¹ second⁻¹ for N₂O), ρ is the gas density (ρ(CO₂-C, CO-C, CH₄-C and NMVOC) = 5.36 × 10⁵ mg C m⁻³ and ρN₂O-N = 1.25 × 10⁶ mg N m⁻³ at 273 K and 1013 hPa), ae is the air exchange rate of the tunnel (the value of ae is assumed to be 1.58 m³ second⁻¹ as described at the fan), dw is the dry weight of the straw before burning (kg dry straw), Δc is the difference in each gaseous species concentration between the tunnel outlet and inlet (m³ m⁻³), T is the air temperature in the tunnel (in °C), and P is the air pressure (in hPa; the value of P was assumed to be 1013 hPa). CFCil is the conversion factor of C emissions required to adjust the cumulative carbon fluxes (CO₂-C + CO-C + CH₄-C + NMVOC) to the equivalent ignition loss of C (calculated C loss), whose values varied from 0.46 to 0.76. This conversion factor was employed to adjust the difference between the assumed volume of exchanged air and the actual volume of exchanged air, because the actual ae was not stable or slower than the assumed value (1.58 m³ second⁻¹). This value is supposed to be close to 1.00. However, it actually decreased to below 1.00. The reason for this might be that the actual ae is not stable, or it may be slower than 1.58 m³ second⁻¹. In this study, it was assumed that CFCil was applicable to adjust the difference not only for the cumulative C emission, but also for each component of the air (i.e. CO₂, CO, CH₄, NMVOC and N₂O) consistently. Therefore, N₂O emissions were calculated using the CFCil as well as the emission of C gas species.

To determine each gas species’ contribution to global warming, the direct and indirect GWP (CO₂ equivalent) was used as described by the IPCC (2007): on a mass basis, 1.0 for CO₂, 1.9 for CO, 25 for CH₄, 0.81 (0.5 for Dimethyl ether (CH₃OCH₃))–7.47 (2.8 for methanol (CH₃OH)) for NMVOC (mass of C basis), 298 for N₂O and 0 for NO_x. As an indicator of biomass burning characteristics, the modified combustion efficiency (MCE; Koppmann et al. 2005) was calculated using Eq. 3 as follows:

(3)

where CO₂-C is the emissions factor of CO₂-C (g C kg dry straw⁻¹) and CO-C is the emissions factor of CO-C (g C kg dry straw⁻¹).

Straw-mushroom cultivation experiment using the closed-chamber method

Straw was fermented thermophilically (60–80°C) outdoors (on the dyke surrounding the studied paddies) for a week by daily watering after rice harvest. Ridge-shaped mushroom beds (60 cm wide, 30 cm high) were prepared outdoors with the fermented straw; 1 kg of straw-mushroom spawn (Volvariella volvacea spp.) and 10 kg of moist rice-husk per 1 t rice straw were inoculated into the fermented rice straw. During cultivation, the young straw-mushroom fruiting bodies were harvested every morning (when there were suitable ones for harvest), and the weight and moisture content of the fruiting bodies were immediately measured.

Acrylic chambers (60 × 30 cm horizontal cross section × 65 cm height with open bottoms) were vertically inserted 5 cm deep into mushroom beds for air sampling [60 cm × 30 cm × (65–5) cm = 108,000 cm³ = 108 L]. The gas samples were collected from the triplicate chambers 5 and 20 minutes after the chamber lids were closed. Five hundred milliliters of air (0.5–1.0% of each chamber’s head space) in each chamber was transferred within 1 minute into a 3-L Tedlar bag with a 100-mL plastic syringe. The air samples were transferred through a three-way stopcock that was connected with a teflon tube (4 and 6 mm inner and outer diameters, respectively) that passed through the top plate of the chamber lid at its center and that had an inlet 5 cm below the top plate. Next, 20 mL of each chamber-air sample was collected with the syringe through the same path and transferred into a 10-mL evacuated glass vial with a butyl rubber stopper.

The concentration of each gas species was analyzed in the same way as in the above-mentioned straw-burning experiment. The following materials, collected throughout the straw-mushroom cultivation process, were analyzed for their C and N contents: (1) harvested straw from rice paddies (before use for mushroom beds); (2) mushroom-bed straw after cultivation; (3) harvested straw-mushrooms; and (4) straw-ash after straw-mushroom cultivation and burning. Using these values, the calculated C and N losses of the mushroom beds during the cropping process were calculated using Eq. 4. These losses, combined with ignition C and N losses from mushroom-bed burning after straw-mushroom cultivation, amounted to the calculated C and N loss of the straw-mushroom cultivation system.

(4)

where CNcrop is the calculated C and N loss of the mushroom beds during cropping (kg C or N), dw.har is the dry weight of the straw harvested from rice paddies (kg dry straw), CNhar is the C or N content of the harvested straw before use for mushroom beds (kg C or N kg dry straw⁻¹), dw.beds is the dry weight of straw after use for mushroom beds just before burning (kg dry straw), and CNbeds is the C or N content of the straw after use for mushroom beds just before burning (kg C or N kg dry straw⁻¹).

The gas fluxes were calculated using Eq. 5 (Rolston 1986), and the cumulative emissions over the experimental period were determined using the same method as the straw-burning experiment:

(5)

where F is the flux (in mg C kg dry straw⁻¹ h⁻¹ for CO₂ and CH₄, and mg N kg dry straw⁻¹ h⁻¹ for N₂O), ρ is the gas density (ρ(CO₂-C and CH₄-C) = 536 × 10³ mg C m⁻³ and ρN₂O-N = 125 × 10⁴ mg N m⁻³ at 273 K and 1013 hPa), V is the volume of the chamber (in m³), A is the cross-sectional area of the chamber (in m²), Δc/Δt is the change in gas concentration inside the chamber as a function of time (m³ m⁻³ h⁻¹), T is the air temperature inside the chamber (in °C), P is the air pressure (in hPa; the value of P was assumed to be 1013 hPa) and CFCmd is the conversion factor of C emissions (m² kg dry straw⁻¹) to convert area-basis cumulative C fluxes (CO₂-C + CH₄-C, mg C m⁻²) to mass-basis C emissions during mushroom cultivation (mg C kg dry straw⁻¹); these values varied from 0.061 to 0.080 (dry weight of mushroom beds in each chamber: 12.5–16.3 kg dry straw m⁻²). CO-C and NMVOC were not considered here because the emissions of these gases were negligible. Irrespective of gas species, CFCmd was assumed to be consistent in each chamber. N₂O emissions were also calculated using the CFCmd as well as the emission of C gas species.

The CO₂ emissions from mushroom-bed burning in each rice cropping season were calculated using Eq. 6:

(6)

where CO₂Emb is the CO₂-C emissions from mushroom-bed burning in each cropping season (in g C kg dry straw⁻¹ cropping season⁻¹), Cil is the C ignition loss of the mushroom bed in each cropping season (in g C kg dry straw⁻¹ cropping season⁻¹) and ERmbCO₂ is the emissions ratio of CO₂–C per calculated C loss (%) whose value was obtained from a straw burning experiment (using the method described above) conducted with 20 kg (18.2% moisture) of mushroom bed collected after actual mushroom cultivation. The mushroom-bed straw-burning experiment was conducted for 10-kg (scattered mushroom beds on the ground) and 20-kg stacks (piled) in this study. We employed data obtained for 20-kg stacks for Eq. 5 because 20-kg stack burning was closer to conventional mushroom-bed burning practiced in the study area. Most of the mushroom beds after mushroom cultivation in the area are conventionally piled up and burned. The other gas species (CO, CH₄, NMVOC, N₂O) emissions from mushroom-bed burning in each cropping season (in g C or N kg dry straw⁻¹ cropping season⁻¹) were also calculated as mentioned above.

Statistical analysis

All statistical analyses were carried out using SPSS 11.0 software for Windows (SPSS, Chicago, IL, USA). The means and standard deviations of the data were calculated. A mean comparison was completed using Tukey’s honestly significant difference (HSD < 0.05) test.

RESULTS

The effects of straw burning conditions on C and N emissions

The time series of each gas flux are shown in Fig. 1. The ignition loss of C and N in each straw-burning experiment is summarized in Table 1. Ignition loss of C in harvested straw varied from 352 to 384 g C kg straw⁻¹ and was greater than the ignition losses from burned mushroom beds that had been used for mushroom cultivation (225–297 g C kg straw⁻¹). The highest peak of each gas species’ flux was observed before the internal temperature peaked. This finding indicated that the burning temperature might not directly affect most gas emissions factors and that temperature might not be useful as an indicator of burning characteristics or smoldering. With lower moisture straw, the internal temperature increased to higher than 100°C sooner after ignition (Fig. 1).

Figure 1 Straw stack dynamics: (A) internal temperature and (B) emission factors of carbon dioxide (CO₂), (C) carbon monoxide (CO), (D) methane (CH₄), (E) non-methane volatile organic carbon (NMVOC) and (F) nitrous oxide (N₂O) during the combustion of (a) 5 kg, (b) 10 kg, (c) 20 kg and (d) 40 kg (oven-dried weight) straw stacks with different moisture contents (12.6, 23.1, 30.9 and 52.2%; mass basis).

Figure 1 (Continued).

The dynamics of each GHG species’ flux (CO₂, CO, CH₄, NMVOC and N₂O), as shown in Fig. 1, were used to calculate cumulative gas emissions (CO₂-C, 204–322 g C kg dry straw⁻¹; CO-C, 26–124 g C kg dry straw⁻¹; CH₄-C, 2.7–21.0 g C kg dry straw⁻¹; NMVOC, 0.15–8.60 g C kg dry straw⁻¹; N₂O-N, 0.006–0.163 g N kg dry straw⁻¹; Table 2; Eq. 1, 2). Regardless of the straw stack size, the time after ignition to reach an internal temperature of 100°C increased when the burning straw contained more moisture (Fig. 2). This increase in time for the low-temperature period (lower than 100°C) was most dramatic in the 40-kg straw stacks (Fig. 2a). The MCE demonstrated a significant negative relationship with the moisture content in the 5-kg straw stacks (Fig. 2b). The MCE also demonstrated a significant negative relationship with CH₄ and NMVOC emissions as a proportion of the calculated C loss (Fig. 3a, 3b). Although the MCE did not exhibit any significant relationship with N₂O emissions (Fig. 3c), straw moisture demonstrated a significant negative relationship with the ratio of N₂O emission to calculated N loss (Fig. 3f).

Table 2 Greenhouse gas emissions derived from straw burning. Emission factors are expressed as the amounts of carbon (C) or nitrogen (N) emitted per 1 kg oven-dried weight of straw or mushroom beds. Values in parentheses show relative gas emission per calculated C or N loss

Straw burning condition			Carbon emission									Nitrogen emission
	Dry-weight	Moisture	CO2-C		CO-C		CH4-C		NMVOC		MCE*	N2O-N
Burned straw	(kg)	(%)	(g C kg^−1)		(g C kg^−1)		(g C kg^−1)		(g C kg^−1)		(%)	(g N kg^−1)
Harvested straw	5	12.6	322	(89%)	36	(11%)	2.7	(0.8%)	0.15	(0.04%)	90	0.076	(0.84%)
	5	23.1	302	(84%)	52	(15%)	2.8	(0.8%)	1.27	(0.35%)	85	0.092	(1.04%)
	5	30.9	273	(78%)	65	(19%)	12.3	(3.5%)	0.95	(0.27%)	81	0.026	(0.30%)
	5	52.2	204	(57%)	124	(35%)	21.0	(5.9%)	8.60	(2.40%)	62	0.073	(0.83%)
	10	19.8	317	(88%)	38	(10%)	4.8	(1.3%)	0.24	(0.07%)	89	0.067	(0.75%)
	10	23.6	300	(83%)	55	(15%)	6.5	(1.8%)	0.79	(0.22%)	84	0.043	(0.48%)
	10	32.0	292	(81%)	57	(16%)	8.6	(2.4%)	0.83	(0.23%)	84	0.015	(0.17%)
	10	50.9	295	(81%)	57	(16%)	9.3	(2.6%)	1.18	(0.33%)	84	0.014	(0.15%)
	20	12.6	314	(89%)	26	(7%)	14.4	(4.1%)	0.38	(0.11%)	92	0.163	(1.86%)
	20	20.6	290	(81%)	59	(16%)	10.2	(2.8%)	0.73	(0.20%)	83	0.037	(0.41%)
	20	33.7	283	(78%)	60	(16%)	19.7	(5.4%)	1.08	(0.30%)	83	0.023	(0.25%)
	20	65.0	302	(84%)	48	(13%)	8.4	(2.3%)	1.22	(0.34%)	86	0.006	(0.07%)
	40	17.1	300	(83%)	53	(15%)	6.8	(1.9%)	0.59	(0.16%)	85	0.030	(0.33%)
	40	29.0	293	(81%)	57	(16%)	12.8	(3.5%)	0.92	(0.25%)	84	0.078	(0.85%)
	40	38.1	305	(84%)	48	(13%)	10.7	(2.9%)	1.05	(0.29%)	86	0.025	(0.27%)
	40	61.4	313	(86%)	43	(11%)	9.2	(2.5%)	0.62	(0.17%)	88	0.030	(0.32%)
Mushroom beds	10	18.2	199	(87%)	21	(9%)	8.7	(3.8%)	0.00	(0.00%)	90	0.035	(0.31%)
	20	18.2	268	(90%)	25	(8%)	3.7	(1.2%)	0.00	(0.00%)	92	0.031	(0.20%)

*See Eq. 3.

Notes: [Carbon composing carbon dioxide (CO₂-C), carbon monoxide (CO-C) and methane (CH₄-C), non-methane volatile organic carbon (NMVOC), modified combustion efficiency (MCE: emissions ratio of CO₂-C/(CO-C + CO₂-C)) and nitrogen composing nitrous oxide (N₂O)].

Figure 2 Correlations between straw moisture content (x) and (A) hours after ignition to reach a 100°C straw stack internal temperature; (B) modified combustion efficiency (MCE). Lines of significant regressions are given for straw stacks with different masses (5, 10, 20 and 40 kg dry straw; p < 0.05, n = 4 each).

Figure 3 Correlations between straw moisture content or modified combustion efficiency (MCE) and emission ratios of (A, D) methane (CH₄), (B, E) non-methane volatile organic carbon (NMVOC) and (C, F) nitrous oxide (N₂O) per calculated carbon (C) or nitrogen (N) loss (mass basis). Solid lines of significant regressions are given (p < 0.05, n = 16).

C and N emissions from straw-mushroom cultivation

The C and N contents in straw immediately after harvest from rice paddies, mushroom beds after straw-mushroom cultivation, harvested straw-mushrooms, and ash of burned mushroom beds are summarized in Table 3. Gaseous fluxes from straw-mushroom cultivation and mushroom yields were also quantified. Rice straw in the study paddies was collected at each rice harvest: 3.72 t ha⁻¹ rice straw (Jasmine 85) during the winter–spring cropping season; 3.10 t ha⁻¹ of rice straw (OM 4218) in the spring–summer cropping season; and 2.22 t ha⁻¹ of rice straw (OM 4218) in the summer–autumn cropping season (oven dry weight, Table 3). The dry weight and N content of the harvested straw from rice paddies were largest in the winter–spring season and tended to become smaller in later cropping seasons (Table 3). Regarding these C and N contents, C and N emissions from mushroom beds during cropping were calculated using Eq. 4, and C and N emissions from mushroom-bed burning (ignition loss of mushroom beds) were measured using Eq. 1 (Table 4). As a result, relatively high N emissions from mushroom beds during cropping were measured in the winter–spring season, when straw N content was also relatively high (Table 4). As for the C emissions from mushroom beds, the decomposition of straw during rice-mushroom cultivation (190–199 g C kg dry straw⁻¹, 0.6–2.0 g N kg dry straw⁻¹; Table 4) and burning of the mushroom beds (138–157 g C kg dry straw⁻¹, 5.9–7.7 g N kg dry straw⁻¹; Table 4) were recorded, but a seasonal difference was not found (Table 4).

Table 3 Carbon (C) and nitrogen (N) contents in straw immediately after being harvested from rice paddies, straw after straw-mushroom [Volvariella volvacea (Bul. ex Fr.) Singer] cultivation, harvested straw-mushroom and ash of straw that was used for straw-mushroom cultivation as mushroom beds and then burned. Contents of carbon and nitrogen are expressed as the amounts of C or N contained in 1 kg of oven-dried straw (n = 3, mean ±standard deviation)

	Harvested straw from rice paddies			Straw after straw-mushroom cultivation (mushroom beds)			Harvested straw-mushroom			Straw-ash after straw-mushroom cultivation and burning**
Straw harvested Season*	Dry weight(t ha^−1)	Total C(g C kg^−1)	Total N(g N kg^−1)	Dry weight(t ha ^−1)	Total C(g C kg ^−1)	Total N(g N kg ^−1)	Dry weight(t ha^−1)	Total C(g C kg^−1)	Total N(g N kg^−1)	Dry weight(t ha^−1)	Total C(g C kg^−1)	Total N(g N kg^−1)
Winter- Spring	3.72 (± 0.71)	352 (± 15)	11 (± 1)	1.67 (± 0.51)	342 (± 27)	20 (± 3)	0.005 (± 0.001)	347 (± 20)	21 (± 5)	0.43 (± 0.08)	133 (± 18)	11 (± 3)
Spring- Summer	3.10 (± 0.80)	366 (± 87)	8 (± 1)	1.52 (± 0.29)	352 (± 14)	15 (± 2)	0.007 (± 0.001)	348 (± 20)	14 (± 3)	0.35 (± 0.07)	139 (± 23)	11 (± 3)
Summer- Autumn	2.22 (± 0.67)	354 (± 48)	7 (± 2)	1.10 (± 0.19)	331 (± 23)	14 (± 2)	0.005 (± 0.001)	345 (± 21)	18 (± 4)	0.25 (± 0.04)	129 (± 19)	9 (± 4)

*Winter–spring: First cropping season after flooding (the driest season, November–February; rice cultivar: Jasmine 85).

Spring–summer: Second cropping season after flooding (March–May; rice cultivar: OM4218).

Summer–autumn: Third cropping season after flooding (the rainy season, June–September; rice cultivar: OM4218).

**After straw-mushroom cultivation, mushroom beds were burned using farmers’ conventional practice. Mushroom beds were left on the upland for a couple of weeks to be air-dried, and then piled up and burned.

Table 4 Carbon (C) and nitrogen (N) emitted from straw-mushroom (Oryza sativa L.), cultivation. Emission rates are expressed as the amounts of C or N emitted per 1 kg oven-dried weight of straw

Straw harvested season*	Emission from mushroom beds during the cropping process		Ignition loss of mushroom beds		Total emission
	Carbon	Nitrogen	Carbon	Nitrogen	Carbon	Nitrogen
	(g C kg^−1)	(g N kg^−1)	(g C kg^−1)	(g N kg^−1)	(g C kg^−1)	(g N kg ^−1)
Winter–spring (Jasmine 25)	199	2.0	138	7.7	337	9.7
Spring–summer (OM4218)	194	0.6	157	6.1	351	6.7
Summer–autumn (OM4218)	190	0.6	149	5.9	339	6.5

*Winter–spring: First cropping season after flooding (the driest season, November–February; rice cultivar: Jasmine 85).

Spring–summer: Second cropping season after flooding (March–May; rice cultivar: OM4218).

Summer–sutumn: Third cropping season after flooding (the rainy season, June–September; rice cultivar: OM4218).

Regarding the temporal patterns of GHG fluxes, relatively high emissions of CO₂, CH₄ and N₂O were detected during fermentation in each cropping season (Fig. 4). After fermentation, CO₂ emissions tended to decline as the days passed, while the CH₄ emissions became negligible in every following season. Regarding the N₂O flux, it was similar to the temporal pattern of the CH₄ flux, except for during the spring–summer season. The N₂O flux in spring–summer was kept relatively high after inoculation of the straw-mushroom spawn (Volvariella volvacea spp.); however, the N₂O flux during other seasons and the CH₄ flux in all seasons were negligible (Fig. 4). The cause might have been the following: (1) rice-mushroom cultivation with straw harvested in the spring–summer season occurred in the rainy season; or (2) anaerobic fermentation before inoculation did not progress enough and a substantial amount of protein might have been left in the mushroom beds and been decomposed by denitrifying bacteria or fungi (Shoun et al. 1992), which grows in beds of Volvariella volvacea spp. or might be contained in the Volvariella volvacea spp. itself during the cultivation period. Using gas flux quantification via the closed-chamber method, the emissions factor for each gas species during mushroom cultivation was calculated (CO₂-C, 190–199 g C kg dry straw⁻¹; CH₄-C, 0.19–0.67 g C kg dry straw⁻¹; NMVOC, 0 g C kg dry straw⁻¹; N₂O-N, 0.11–0.17 g N kg dry straw⁻¹; Fig. 4; Table 4).

Figure 4 Dynamics of (A–C) carbon dioxide (CO₂), (D–F) methane (CH₄) and (G–I) nitrous oxide (N₂O) emitted by microbial decomposition of mushroom beds during straw-mushroom [Volvariella volvacea (Bul. ex Fr.) Singer] cultivation, which is conducted with straw harvested in (A, D, G) winter–spring, (B, E, H) spring–summer and (C, F, I) summer–autumn rice cropping seasons. Error bars show standard deviations (n = 3).

The emissions factors from the burning of the mushroom beds were also calculated (CO₂-C, 125–142 g C kg dry straw⁻¹; CO-C, 11.4–13.0 g C kg dry straw⁻¹; CH₄-C, 1.71–1.94 g C kg dry straw⁻¹; NMVOC, 0 g C kg dry straw⁻¹; N₂O-N, 0.01–0.02 g N kg dry straw⁻¹; Eq. 1, 2) from the C ignition loss from the mushroom beds (Table 4) and from the straw-burning experiment with 20-kg stacked mushroom beds (Table 2). The total gas emissions from straw-mushroom cultivation (the sum of the gas emissions from straw decomposition during cultivation and burning of the mushroom beds) amounted to 337–351 g C kg straw⁻¹ (CO₂-C, 324–335 g C kg dry straw⁻¹; CO-C, 11.4–13.0 g C kg dry straw⁻¹; CH₄-C, 1.90–2.61 g C kg dry straw⁻¹; NMVOC, 0 g C kg dry straw⁻¹; Eq. 4–6) and 6.7–9.7 g N kg straw⁻¹ (N₂O, 0.12–0.18 g N kg dry straw⁻¹) (Table 4).

Global warming potential of straw burning and straw-mushroom cultivation

Using the GWP of each gas species and the above-mentioned GHG emission factors derived from straw burning and straw-mushroom cultivation, including mushroom-bed burning after cultivation, the total GHG emissions were calculated (straw burning, 1469–2098 g CO_2-eq. kg dry straw⁻¹; straw-mushroom cultivation, 1362–1461 g CO_2-eq. kg dry straw⁻¹, Fig. 5). The GHG emissions factors (average ± standard deviation, g CO_2-eq. kg dry straw⁻¹, using a temporary GWP value of 7.47 for NMVOCs) were determined to be 1688 ± 171 (n = 16, Fig. 5) for straw burning and 1400 ± 54 (n = 3) for straw-mushroom cultivation (1362 in the winter–spring, 1461 in the spring–summer and 1375 in the summer–autumn rice cropping season, Fig. 5).

Figure 5 Greenhouse gas emissions derived from straw-use [(A) straw burning and (B) straw-mushroom [Volvariella volvacea (Bul. ex Fr.) Singer] cultivation] from different straw stack masses, and their moisture contents (moisture contents of burned mushroom beds were 16% in the winter–spring season, 17% in the spring–summer season and 22% in the summer–autumn season). Each gaseous species’ emissions factor was integrated with respect to each direct and indirect global warming potential [1.0 for carbon dioxide (CO₂), 1.9 for carbon monoxide (CO), 25 for methane (CH₄), 0.81-7.47 for non-methane volatile organic carbon (NMVOC], and 298 for N₂O; mass basis referring to IPCC 2007). Error bars illustrate the ranges of the NMVOC’s global warming potential.

Using these GHG emissions values and the straw-production rates for the study area as described in Hong Van et al. (2014), the annual GHG emissions from straw use in the region amounted to 7.60 ± 1.50 t CO_2-eq. ha-paddy⁻¹year⁻¹ from straw burning and 6.65 ± 1.12 t CO_2-eq. ha-paddy⁻¹year⁻¹ from straw-mushroom cultivation (Table 5).

Table 5 Straw production, amount of harvested straw, its use and greenhouse gas emissions from the harvested straw-use in a triple rice (Oryza sativa L.) cropping system of Tan Loi 2 hamlet, Thot not, Can Tho city, Viet nam from 2007 to 2011

	Straw production		Use of harvested straw			Greenhouse gas emission from the harvested straw-use
Rice cropping season*	Total straw production (including stumps)(t dry-straw ha-paddy^−1)	Harvested straw (except stumps)(t dry-straw ha-paddy^−1)	Burned straw with scattered on the paddy soil(t dry-straw ha-paddy^−1)	Burned haystack of straw(t dry-straw ha-paddy^−1)	Used as mushroom beds for mushroom cultivation(t dry-straw ha-paddy^−1)	Straw burning(t CO2-eq. ha-paddy^−1)	Straw-mushroom cultivation(t CO2-eq. ha-paddy^−1)
Winter–spring	7.05 ± 1.64	3.78 ± 0.88	3.78 ± 0.88	0	0	6.39 ± 1.49^a	0^d
Spring–summer	6.59 ± 1.25	3.14 ± 0.60	0.10 ± 0.03	0.58 ± 0.11	2.47 ± 0.47	1.16 ± 0.22^c	3.53 ± 0.67^b
Summer–sutumn	5.82 ± 1.67	2.32 ± 0.67	0	0.03 ± 0.01	2.30 ± 0.66	0.06 ± 0.01^d	3.11 ± 0.89^b
Annual	19.46 ± 2.66	9.24 ± 1.26	3.88 ± 0.88	0.61 ± 0.11	4.82 ± 0.81	7.60 ± 1.50	6.65 ± 1.12

*Winter–spring: First cropping season after flooding (the driest season, November–February; rice cultivar: Jasmine 85).

Spring–summer: Second cropping season after flooding (March–May; rice cultivar: OM4218).

Summer–autumn: Third cropping season after flooding (the rainy season, June–September; rice cultivar: OM4218).

The quantitative information on straw production and use of harvested straw is from Hong Van et al. (2014). Referring to Fig. 8, greenhouse gaseous emission factors (t CO_2-equivalent t dry-straw⁻¹) were calculated as 1.69 ± 0.17 for straw burning (n = 16), 1.36, 1.46 and 1.38 for straw-mushroom cultivation in the winter–spring, the spring–summer and the summer–autumn rice cropping season respectively.

Greenhouse gas emissions from the use of harvested straw are not significantly different from each other when they are marked with the same letter above the bars, based on Tukey–Kramer honestly significant difference (HSD) test (p < 0.05, n = 35).

DISCUSSION

Ignition losses of C from harvested straw (352–384 g C kg dry straw⁻¹) were higher than those from mushroom beds (225–297 g C kg dry straw⁻¹). At the ignition, easily oxidizable organic matter which composes the mushroom beds would have been decomposed by microbes to CO₂ already. Therefore, the mushroom beds would have consisted mostly of recalcitrant organic matter which is relatively harder to oxidize. In contrast, ignition losses of N from harvested straw (8.7–9.3 g N kg dry straw⁻¹) were lower than those from mushroom beds (11.2–16.0 g N kg dry straw⁻¹) in Table 1. The reason for this would have been the difference in the C/N ratio of straw (38.9 for harvested straw and 17.1 for mushroom beds, Tables 1 and 3).

Gadde et al. (2009) compiled GHG emissions factors from straw burning (CO₂: 1460 g kg dry straw⁻¹ in Jenkins and Bhatnagar (1991), CO: 347 g kg dry straw⁻¹ in Kadam et al. (2000), CH₄: 1.2 g kg dry fuel⁻¹ in US EPA (1992), NMVOC: 4 g kg dry straw⁻¹ in US EPA (1992), N₂O: 0.07 g kg dry fuel⁻¹ in Andreae and Merlet (2001)). Using these values and the GWP of each gas species, the total GHG emissions amounted to 1607 g CO_2-eq. kg dry straw⁻¹, which is commensurate with the value obtained in our study (1688 ± 171 (1469–2098) g CO_2-eq. kg dry straw⁻¹, Fig. 5). However, these values were relatively higher than the amount of GHGs emitted from straw burning that has been evaluated in Japan (CO₂-C, 194–303 g kg dry straw⁻¹; CO-C, 18.9 – 39.0 g kg dry straw⁻¹; CH₄-C, 1.6 – 3.1 g kg dry straw⁻¹; N₂O-N, 0.06–0.08 g kg dry straw⁻¹; Miura and Kanno 1997). Because the experiment was conducted at only one of the representative triple rice cropping systems located in the central Vietnamese Mekong Delta, our results cannot be used to estimate the GHG emissions derived from straw burning throughout the Mekong Delta. However, our results would illustrate the amount of GHGs emitted from straw burning and the cultivation of straw-mushrooms in a representative triple rice cropping systems in the study area.

Our comparison of GHG emissions under various straw-burning conditions (moisture and size of straw stacks) indicated that CH₄ and NMVOC emissions increased with lower MCE and N₂O emissions were higher from burning lower-moisture straw stacks (Fig. 3); however, under small straw stack conditions (e.g., 5 kg straw in a 2-m diameter circle), the increase in GWP emission caused by enhanced generation of CH₄ and NMVOC under higher moisture conditions was much larger than the simultaneous decrease in GWP emission caused by inhibited N₂O production (Fig. 5a).

Small straw stacks (5 or 10 kg) with higher moisture content emitted more CO, CH₄ and NMVOC than the small straw stacks with lower moisture content (Fig. 5). A similar result was described in Hayashi et al. (2014), the CO₂/CO ratios of the rice and barley (Hordeum vulgare L.) straw, which is a similar indicator of imperfect combustion to MCE, significantly decreased with the increase in the moisture content, which indicated that the increase in moisture enhanced the degree of imperfect combustion. In their study, moisture content was positively correlated with the emissions of carbonaceous gases, and the correlations were particularly strong for emissions of CO and CH₄ from rice and barley straw. The positive correlations supported the hypothesis that emissions of these gases during smoldering combustion were enhanced by the high moisture content, owing to imperfect combustion. The emissions of organic C under moist conditions were also larger than the emissions under dry conditions. As in the above-mentioned results from Hayashi et al. (2014), the results from our study also illustrated the effects of smolder-burning enhancing the CO, CH₄ and NMVOC emissions, because the moist conditions resulted in partial oxygen deficits during burning.

The N₂O emissions measured in our study became smaller as the moisture content of the straw stacks increased (Table 2), although higher N₂O emissions were observed as the straw’s moisture content increased in Hayashi et al. (2014). The reason the opposite phenomenon was found in these two studies might be a result of air samples for N₂O measurement being taken without a filter in our study, but the air samples were collected with a potassium chloride-impregnated filter for removal of sulfur dioxide (SO₂) in Hayashi et al. (2014). Because coexisting SO₂ and nitric oxide (NO) react to form N₂O during sample storage (Muzio and Kramlich 1988; Linak et al. 1990; Preto et al. 2004), the N₂O emissions calculated in our study might be overestimated, even though their contribution to global warming is negligible compared with other gas species (Fig. 5).

While imperfect combustion of straw inhibited N₂O emissions in our study, it enhanced the emissions of CO, CH₄ and NMVOC. Therefore, relatively intense GHG emissions, perhaps derived from imperfect combustion, were observed from the relatively small straw stacks (5 and 10 kg) with higher moisture content (Table 2). However, this same tendency was not observed in the relatively large straw stacks (20 and 40 kg). Notably, a considerable period of time after ignition was required for the internal temperature of the relatively large straw stacks to reach 100°C. With the internal temperature of the high-moisture-content straw stacks under 100°C for a relatively long time, the water in the straw could have evaporated without causing smoldering combustion. These results indicated that imperfect combustion effects of wet straw might become negligible when large straw stacks are burned on open fields. Hayashi et al. (2014) also found the effect of imperfect combustion of wet straw with a thinner layer of straw (5 cm height). Imperfect combustion may become significant only when the straw is wet and shallowly stacked, similar to actual situations in the paddy field after harvest with a combine harvester, in which rice straw is scattered on the ground.

Compared with the GHG emissions from the straw burning, GHG emissions from mushroom cultivation were smaller (straw burning, 1469–2098 g CO_2-eq. kg dry straw⁻¹; straw-mushroom cultivation, 1362–1461 g CO_2-eq. kg dry straw⁻¹, Fig. 5). Although relatively high CH₄ emissions were detected during the fermentation process under mushroom cultivation, the total amounts of CO, CH₄ and NMVOC emissions during mushroom cultivation and subsequent burning were lower than those for straw burning. The reason that GHG emissions from mushroom cultivation were smaller than GHG emissions from straw burning might have been the relatively low emissions of CO, CH₄ and NMVOC from mushroom cultivation, although there was no consistent tendency in the emissions of CO₂ and N₂O in these two types of straw-use. These results indicated that dissemination of straw mushroom cultivation in the rice-based agricultural systems would be effective to reduce the regional GHGs emissions. Because mushroom beds were conventionally air-dried for a couple of weeks and piled up before the ignition, these indigenous practices of the farmers might also contribute to the reduction of GHG emissions by suppressing the imperfect combustion of rice straw.

In the study area, combine harvesters were used to harvest rice in only 0–32.1% of the paddy fields from 2007 to 2011 (Hong Van et al. 2014); hand reaping was the most common rice harvesting method. Because hand-reaped rice is threshed in paddies, straw stacks are left in the paddies after threshing and are mostly burned immediately or removed from the paddies for straw-mushroom cultivation. According to Hong Van et al. (2014), while almost all of the straw was burned during the driest season (winter–spring), most straw was removed from the paddies to be used for straw-mushroom cultivation during the rainy season (summer–autumn). Because minimal straw was scattered conventionally on the soil for burning in the rainy season, smoldering combustion could not occur to any significant extent. However, the widespread use of combine harvesters during future rainy seasons could increase the regional level of imperfect straw combustion that may increase the GWP of rice paddies because combine harvesters shred and scatter straw on the paddy fields, making it difficult to remove. Because imperfect combustion effects of straw moisture were found only in relatively small straw stacks (5 or 10 kg), which resulted in a large amount of CO, CH₄ and NMVOC emissions, shredding and scattering rice straw on soils with combine harvesters may aggravate smoldering combustion to emit larger amount of GHGs. In addition, the straw shredding and scattering may also increase straw incorporation into paddy soils, further enhancing CH₄ emissions (Sass et al. 1991). In future research, the quantification of CH₄ and N₂O emissions from triple rice-cropping systems is required to fully assess the GHG emissions implications of different straw use.

In this study, the GHG emissions from different straw uses in a triple rice cropping system in the Mekong Delta were quantified. Relatively intense GHG emissions, which would have been derived from smoldering combustion, were observed from relatively small straw stacks under wet conditions. Straw used as mushroom beds for straw-mushroom cultivation exhibited lower GHG emissions than straw burning. Therefore, rice straw burning causes more GHG emissions than straw-mushroom cultivation does under the examined agroecosystems.

ACKNOWLEDGMENTS

We would like to thank the rice farmers of Tan Loi 2 Hamlet, including Mr. Ut, the chief manager of the experimental fields, for the field observation and management, and the students of Can Tho University for performing gas sampling in the straw-burning experiment.