ABSTRACT
ABSTRACT
Alternate wetting and drying (AWD) is a water-saving irrigation technique in a paddy field that can reduce the emission of methane, a potent greenhouse gas (GHG). It is being adopted to Asian countries, but different results are reported in literatures on methane, nitrous oxide emission, and rice productivity under AWD. Therefore, the local feasibility needs to be investigated before its adoption by farmers. The current study carried out a 3-year experiment in an acid sulfate paddy field in Prachin Buri, Thailand. During five crops (3 dry and 2 wet seasons), three treatments of water management were compared: continuous flooding (CF), flooding whenever surface water level declined to 15 cm below the soil surface (AWD), and site-specific AWD (AWDS) that weakened the criteria of soil drying (AWDS). Methane and nitrous oxide emissions were measured by a closed chamber method. Rice grain yield did not significantly (p < 0.05) differ among the three treatments. The amount of total water use (irrigation + rainfall) was significantly reduced by AWD (by 42%) and AWDS (by 34%) compared to CF. There was a significant effect of treatment on the seasonal total methane emission; the mean methane emission in AWD was 49% smaller than that in CF. The seasonal total nitrous oxide emission and the global warming potential (GWP) of methane and nitrous oxide did not differ among treatments. The contribution of nitrous oxide to the GWP ranged 39–62% among three treatments in dry season whereas 3–13% in wet season. The results indicate that AWD is feasible in terms of GHG emission mitigation, rice productivity, and water saving in this site, especially in dry season.
1. Introduction
In Thailand, agriculture sector contributes to around 17% of the total national greenhouse gas (GHG) emission (Office of Natural Resources and Environmental Policy and Planning (ONEP) Citation2015). Within the agriculture sector, methane (CH4) emission from paddy field is the major source that accounts for 72% of the sector’s totals. Mitigating emission of CH4 from paddy field therefore potentially contributes to emission mitigation target proposed to the United Nations Framework Convention on Climate Change under the Intended Nationally Determined Contribution schemes (ONEP Citation2015). However, it needs to be confirmed that any mitigation options will not negatively affect rice yield and consequently national food production.
Due to flooded conditions during its growth, rice cultivation is the major irrigation water consumer in Thailand. As water shortage especially in the dry season is common in Thailand, available water for rice field irrigation is becoming limited. In addition, extreme climatic events, such as drought and shift in rainfall distribution pattern, have been widespread in recent decades and imposed a significant threat to water resources management, especially for rice cultivation in the future (IPCC Citation2012; Thailand Research Fund (TRF) Citation2016). As a consequent, effective mitigation options for GHG emissions and increasing water use efficiency in rice cultivation are needed for a sustainable rice cropping system.
There have been various measures proposed for mitigating the emission of CH4 from paddy field through water management. Drainage at appropriate timing during rice growth can reduce CH4 emission (e.g., Minamikawa and Yagi Citation2009) because it introduces oxygen into the soil and thereby inhibiting methanogenesis and promoting methanotrophic activity (Conrad Citation1996; Conrad and Rothfuss Citation1991; Bender and Conrad Citation1992). In central Thailand, for example, mid-season drainage was able to reduce CH4 emission by 35% compared to the conventional continuous flooding (CF; Towprayoon et al. Citation2005).
A precise water controlling technique known as safe alternate wetting and drying (AWD) has been originally developed to save irrigation water use in paddy field (Bouman and Tuong Citation2001). This technique controls surface water level not to fall below a soil depth of 15 cm. Water consumption and CH4 emission were reported to be effectively reduced (e.g., LaHue et al. Citation2016). However, there are different results in the literatures on the effects of AWD on nitrous oxide (N2O) and rice grain yields. LaHue et al. (Citation2016) reported a reduction of 60–87% in CH4 emission while maintaining a low N2O emission level in a California paddy field by AWD. Grain yield was not affected by AWD or higher in AWD treatment compared to the flooding control. On the other hand, Lagomarsino et al. (Citation2016) reported in a 2-year study in Italian paddy field that 70% water consumption and 33% yields, and 97% CH4 emissions were reduced by implementation of AWD as compared to permanent flooded fields. N2O emissions were increased by more than 5-fold under AWD. In the second year, with 40% water saving, the reductions of rice yields and CH4 emissions (13% and 11%, respectively) were not significant, but N2O emissions were more than doubled. Carrijo et al. (Citation2017) conducted a meta-analysis on the effects of AWD on rice yield and found that when water level is controlled not to drop below a soil depth of 15 cm, the AWD’s (negative) effect on rice yield is not significant. They also indicate that soil properties such as pH and organic carbon content can interact with AWD in such a way that the effects of AWD on yield are more pronounced in soil with pH higher than 7 or carbon content less than 1%.
The varying effects of AWD on GHG emissions and grain yields mentioned above highlight the needs for more research in key rice cultivation regions to improve our understanding of links among cultivation practice, local environmental factors, rice growth, and GHG emissions. Such information will be important for adoption of AWD by agricultural extension agency and by local farmer. The objective of the current study was to evaluate the potential of AWD for GHG mitigation, and its effects on rice productivity and water saving in an acid sulfate paddy field soil in Thailand.
2. Materials and methods
2.1. Experimental site
The experiments were conducted during 2013–2016 in an experimental paddy field of Prachin Buri Rice Research Center (PRRC), Bansang District, Prachin Buri, Thailand (14.01°N, 101.22°E). The experimental site is situated at approximately 3 m above mean sea level. The site is under monsoon climate with the mean annual precipitation of 1700 mm and the mean annual air temperature of 28 °C. The field at the site is usually flooded annually with water depth of more than 1 m in the late rainy season (i.e., in October, Chareonsilp et al. Citation2000) as it is in a low-lying area and received outflow and runoff from Khao Yai mountainous areas in the northern side. The local farmers have adapted to avoid flooding by planting rice earlier, allowing harvest before the water comes.
The soil is originated from marine sediment, poorly drained, and classified as Rangsit soil series; an acid sulfate soil (very-fine, mixed, active, acid, isohyperthermic Sulfic Endoaquepts). The soil texture is heavy clay with pH (H2O) between 4.4 and 4.8, the carbon and nitrogen content of 2% and 0.2%, and the active iron and manganese content of 1.57% and 0.007%, respectively. Other soil properties to 150 cm below the soil surface are given in . The upper 50-cm layer is strongly acidic (pH 4.5–5.0), very dark gray (10YR3/1), and is separated into the Apg (0–20 cm) and Bg (20–50 cm) layers. The brown mottles are common. The main difference between Apg and Bg layers is the existence of fine roots which are much more abundant in Apg than in Bg layers. Below 50 cm, several sub-layers (Bjg1-4) are identified. They are commonly characterized by gray color (10YR5/1), very strongly acid (pH 4.0–5.0) with common fine yellow (2.5Y7/8) jarosite mottles.
Table 1. Physical and chemical properties of acid sulfate paddy soil at Prachin Buri Rice Research Center, not determined.
2.2. Experimental design
The experimental plots were laid out following the randomized completed block design. The study covered five crops: three crops during dry seasons (DS1, DS2, DS3) and two crops during wet seasons (WS1 and WS3). Rice was not planted in the second wet season (WS2) due to high flooding. Due to the occurrence of blast disease at the beginning of DS3, rice was replanted 15 days after the first sowing. Rice (Oryza sativa) seed of RD41 variety (non-photosensitive) was sown by pre-germinated broadcasting method at a rate of 125 kg ha−1. The RD41 normally requires about 100 days for maturity, with the vegetative growth stage up to 35 days after sowing (DAS), and flowering around 65 DAS.
Three treatments with regard to water management were made with three replicate plots each with 5 × 7 m area: CF, AWD, and site-specific AWD (AWDS). For the CF treatment, the field was continuously flooded by maintaining a constant surface water depth of 10 cm during 15–90 DAS. For the AWD treatment, the fields were flooded (5-cm water depth) whenever the water level was decreased to 15 cm below the soil surface. For AWDS treatment, the similar water managements to that of AWD were made but with the water depth of 10 cm when the field was reflooded.
All treatments received the same chemical fertilizer rate of 70 kg N ha−1, 37.5 kg P2O5 ha−1, and 37.5 kg K2O ha−1. Fungicides were applied occasionally, especially during the dry seasons. The detailed timing and application rates together with other key cultivation practices are shown in .
Table 2. Key crop cultivation practices and dates in the dry seasons(DS) and wet seasons (WS).
2.3. Measurements
The closed chamber method was used to collect gas sample. A square-shape chamber was made from acrylic, with the dimension of 60 cm long, 60 cm wide, and 60 cm high. When gas samples were collected, the chamber was placed on the base that was inserted permanently in the field during the growing period. The gas sampling was started at 20 DAS in DS1 and WS1, and at 7 DAS in the subsequent seasons. The gas sampling was conducted every week and 1–5 days after fertilizer application. The sampling time was during 9–11 am local time. Five gas samples at 0, 6, 12, 20, 30 min were taken from the chamber headspace by a plastic syringe. For the samples collected during DS1 and WS1, CH4 concentration was analyzed by a gas chromatograph equipped with FID and Custom Packed SG 804 column (Shimadzu GC 8A, Japan), using N2 as carried gas. The N2O concentration was analyzed by gas chromatograph (Shimadzu GC-14B, Japan) equipped with ECD and Porapak column. Specific operation conditions of the gas chromatographs were described in Cha-un et al. (Citation2016). For the subsequent seasons, the Agilent 7890B (Agilent Technologies, Inc., USA) gas chromatograph equipped with FID operated at 300 °C for CH4 and ECD at 300 °C for N2O and HaySep Q packed column was used. N2 and He were used as carrier gases for FID and ECD, respectively, at a flow rate of 20 mL min−1.
The gas fluxes were calculated based on a linear regression using the equations given in Minamikawa et al. (Citation2015). Seasonal cumulative CH4 and N2O emissions were estimated by linear interpolation and numerical integration between sampling times. Due to the malfunction of ECD-GC in WS1, the N2O data were excluded from the analysis. The global warming potential (GWP), CO2-equivalent cumulative emission of CH4 and N2O, was calculated using the GWPs for 100-year time horizon with inclusion of climate-carbon feedbacks (IPCC Citation2013; 34 for CH4 and 298 for N2O).
Daily air temperature (minimum and maximum) and precipitation were recorded through the experimental period. The daily water level was automatically monitored by a water level sensor (TruTrack SE-TR/WT500, Senecom Inc., Tokyo, Japan). Tiller number and plant height of rice were regularly measured, and grain yield was measured at a 2 × 3 m area in each plot at the harvest stage. The yield-scaled GWP was calculated by dividing GWP over grain yields. The water productivity, the weight of grain yield over the total volume of water used (irrigation and rainfall) during the growing period, was calculated.
2.4. Data analysis
Analysis of variance (ANOVA) was performed with a mixed model (‘proc mixed’) in the SAS software ver. 9.00 (SAS Institute Inc., Cary, NC, USA) to assess the main effects of cropping season (DS1, WS1, DS2, DS3, and WS3), treatment (CF, AWD, AWDS), and their interactions. A split-plot model was applied in which cropping season was treated as the main factor, and treatment as the split-plot factor with three replications. Variance components were estimated by the restricted maximum-likelihood method with the ‘nobound’ option, and the denominator degrees of freedom were estimated by the Kenward–Roger approximation (see the Supplemental Material for the SAS scripts). Because emissions of CH4 and N2O showed highly skewed distributions and violated normality and homoscedasticity assumptions, and Box-Cox transformation was conducted for CH4, N2O, GWP, and yield-scaled GWP using the ‘powerTransform’ function in the ‘car’ package of R (Box and Cox Citation1964; Fox and Weisberg Citation2011). To test differences among water managements, Tukey's honest significant difference (HSD) test was performed when significant main effect was detected in ANOVA.
3. Results
3.1. Weather and the irrigation water
DS and WS had distinct rainfall amounts and temperature ranges (Figs. 1A–C and 2A,B). During DS, relatively small rainfall amount or almost no rainfall was observed, while during WS intermittent rainfall was observed throughout the growing period. The total amounts of rainfall during DS1, DS2 and DS3 were 0.4, 193, 102 mm, and during WS2 and WS3 were 1006 and 877 mm, respectively. The WSs (WS1 and WS3) received 58–59% of the annual total rainfall. Air temperature was generally lower during WS than during DS (Figs. 1D–F and 2C,D). The temperature ranges during DS1, DS2, and DS3 were 22.6–34.3, 23.2–34.6, 26.3–37.4 °C and during WS1 and WS3 were 25.3–33.7 and 25.6–33.9 °C, respectively.
Published online:
06 November 2017Figure 1. Seasonal variations in daily rainfall (a, b, c), daily maximum and minimum air temperature (d, e, f), mean surface water level (g, h, i), CH4 flux (j, k, l), and N2O flux (m, n) for three water management practices in dry season (DS1, DS2, and DS3). Error bars for CH4 and N2O fluxes indicate the standard error (n = 3). Vertical dotted lines indicate the application timing of nitrogen fertilizer. Gray areas in a and d indicate the lack of data observation.
Figure 1. Seasonal variations in daily rainfall (a, b, c), daily maximum and minimum air temperature (d, e, f), mean surface water level (g, h, i), CH4 flux (j, k, l), and N2O flux (m, n) for three water management practices in dry season (DS1, DS2, and DS3). Error bars for CH4 and N2O fluxes indicate the standard error (n = 3). Vertical dotted lines indicate the application timing of nitrogen fertilizer. Gray areas in a and d indicate the lack of data observation.
Published online:
06 November 2017Figure 2. Seasonal variations in daily rainfall (a, b), daily maximum and minimum air temperature (c, d), mean surface water level (e, f), CH4 flux (g, h), and N2O flux (i) for three water management practices wet seasons (WS1 and WS3). Error bars for CH4 and N2O fluxes indicate the standard error (n = 3). Vertical dotted lines indicate the application timing of nitrogen fertilizer. Inserted panels in e and f show the magnified y-axis for the high values.
Figure 2. Seasonal variations in daily rainfall (a, b), daily maximum and minimum air temperature (c, d), mean surface water level (e, f), CH4 flux (g, h), and N2O flux (i) for three water management practices wet seasons (WS1 and WS3). Error bars for CH4 and N2O fluxes indicate the standard error (n = 3). Vertical dotted lines indicate the application timing of nitrogen fertilizer. Inserted panels in e and f show the magnified y-axis for the high values.
Effective water control with minimal water level reaching 15 cm below soil surface was achieved only during DS (Fig. 1G–I), while in WS (Fig. 2E,F) the water level in most cases was between 0 and 10 cm above the soil surface. The total numbers of drained day (water level < 0 cm) in the AWD treatments were 20, 28, and 28 for DS1, DS2, and DS3, and 3 and 0 for WS1 and WS2, respectively. For AWDS, these were 8, 16, and 18 for DS1, DS2, and DS3, and 5 and 0 for WS1 and WS3, respectively.
The seasonal total volume of water use (irrigation + rainfall) was significantly (p < 0.05) reduced by AWD and AWDS compared to CF and significantly differed among the five cropping seasons (). Their interaction was also significant due to the efficiency of water saving by AWD and AWDS was higher in WS than in DS. The seasonal water use in AWD and AWDS was 42% and 34% smaller than that in CF, respectively.
Table 3. Statistical analysis results of combined seasonal means of CH4, N2O, grain yield, and water use, with the effects of both treatment (Trt), growing season (s), and a combination of treatment and season (S × Trt).
3.2. GHG emission
3.2.1. CH4
The CH4 fluxes varied spatially as indicated by large error bars for each sampling date (Figs. 1J–L and 2G,H). Large flux peaks during the first week of seed sowing were observed in DS2 and WS3. Since gas samples were not taken during the first week of other seasons, such large peaks were not ruled out.
There were significant effects of cropping season and treatment on the seasonal total CH4 emission (). Although the result of Tukey’s HSD test was marginal (p < 0.1), the total CH4 emission from AWD was 49% smaller than that from CF. In case of removing the data in WS3 with the exceptional high emission from AWDS, the total emission from AWDS was 21% smaller than that from CF, although not significant. Supplementary Table S1 shows the total CH4 emission data for each season. With exceptions for DS3 with the blast disease and WS3, the highest total emission was usually found in CF, followed by AWDS and AWD.
3.2.2. N2O
The relatively high N2O fluxes were observed in DS1 and DS2 among all the seasons (Figs. 1M–O and 2A). However, several temporal peaks after N fertilizer topdressing and after the final drainage in other seasons as well as DS1 and DS2 were found. The resultant seasonal total N2O emission was significantly different among cropping seasons but not significantly different among treatments (). Keeping flooded conditions after N fertilizer application caused no difference among treatments. The significant interaction was due to the different results of treatments among cropping seasons (see Supplementary Table S1).
3.2.3. GWP of CH4 and N2O
The GWP was significantly different among cropping seasons and marginally different among treatments (). The GWP was highest in AWDS among the three treatments mainly due to the exceptionally large CH4 emission in WS3. The GWP in AWD was 25% smaller than that in CF, although not significant. The N2O emission accounted for 39–62% (the range of three treatments) of the GWP in DS whereas 3–13% in WS. The highest N2O’s contribution was from AWD due to the smaller CH4 emission, although not significant.
3.3. Rice productivity and yield-related indices
Rice growth and the grain yield were normal in the current experiment, except for DS3. The number of tillers tended to be higher during the early growth stage in AWD compared to CF and AWDS, but not significantly different (Supplementary Fig. S1). The maximum tiller number ranged between 900 and 1200 tiller m−2 in all the treatments. Plant aboveground biomass was also comparable among treatments (Supplementary Fig. S2). As a result, there was no significant difference in rice grain yield among treatments (). On the other hand, there was significant effect of cropping season due mainly to the low yield in DS3 by the blast disease and significant interaction due to the contradictory results of treatments among cropping seasons (see Supplementary Table S1).
The yield-scaled GWP was significantly affected only by cropping season (). On the other hand, water productivity was significantly improved by the implementation of AWD and AWDS (). The significant effects of cropping season and interaction with treatment were found due to high water productivity under AWD and AWDS in WS.
4. Discussion
4.1. Effects of AWD on GHG emissions
The AWD irrigation has been suggested as a water management technique to effectively mitigate CH4 emission in paddy field (Pandey et al. Citation2014; LaHue et al. Citation2016; Liang et al. Citation2016). If effectively drained, the emission of CH4 is expected to be substantially reduced. However, the effectiveness of drainage practices on CH4 emission reduction depends on the efficiency of water control, soil type, and other cultivation practices (Yan et al. Citation2005; Liang et al. Citation2016). In the current study, practicing AWD in this soil resulted in marginal (p < 0.1) reduction in the CH4 emission compared to CF by Tukey’s HSD test (). The AWDS had a lower ability to reduce the GWP compared to AWD in this site, suggesting that the soil drying under AWDS was not enough to reduce CH4 emission.
Contrary to our expectations, the effectiveness of CH4 emission reduction by AWD and AWDS treatments was occasionally limited ( and Supplementary Table S1) although several cycles of drying and wetting were achieved during DS (Fig. 1G–I). In addition to the effects of blast disease, the unexpected small CH4 emission in DS3 (Fig. 1L) could be attributed to the low availability of carbon substrates (i.e., rice stubble and straw) for methanogenesis in the soil due to the cancelation of WS2. The relatively large CH4 emission in WS3, especially under AWDS (Fig. 2H), could be partly explained by heavy rainfall after rice sowing. These exceptional events would have masked the effects of AWD on CH4 emission reduction.
One common characteristic of CH4 emission in this site is its large spatial variation among three replications: the coefficients of variation ranged from 43% to 120% in CF, 11% to 79% in AWD, and 21% to 61% in AWDS. This high spatial variability would be another reason that masks the emission differences when comparing the effectiveness of AWD and AWDS against that of CF. In fact, high spatial variations in CH4 emissions are well known and have been reported in many studies (e.g., Wagner and Pfeiffer Citation1997; Wachinger et al. Citation2000; Sey et al. Citation2008). The presence of microsite and microbial activity within the microsite was reported to be a hot spot of CH4 production even in oxic soil and a source of emission spatial variability (Peters and Conrad Citation1995; Parkin Citation1987; Sey et al. Citation2008; Plaza-Bonilla et al. Citation2014). In addition, the high spatial variability in CH4 fluxes commonly found at this site could be partly attributed to the fine texture and the high heterogeneity of this acid sulfate soil (heavy clay, Supplementary Fig. S3). During the dry period, it was observed that soil cracks were common. The trapped CH4 would be released and it would be also possible that anaerobic microsites were developed in the soil aggregates and methanogenic activity was protected by such microsite despite water was drained in AWD and AWDS. In addition, it is noted that a relatively low CH4 emission was also another distinct characteristic at the site. High acidity and slowly developed anaerobic conditions as described by Chareonsilp et al. (Citation2000) would explain such low emission of CH4. Further study is necessary to elucidate the effects of soil type and soil property on the magnitude of CH4 emission and on the effectiveness of drainage practices on reducing CH4 emission.
Similar to the case of CH4, large spatial variations in N2O fluxes were also observed, especially in DS (Fig. 1M–O, Table S3). The CV for DS2, DS2, and DS3 ranged from 35% to 182% for CF, 24% to 50% for AWD, and 50% to 55% for AWDS, respectively. This would have contributed to the insignificant difference in N2O emissions among treatments. Another reason for the insignificant differences in N2O emissions among treatments could be the effects of the timing of N application. In the current study, N fertilizer as di-ammonium phosphate was applied within the first month of DAS, during which the water levels were relatively high and remained at similar level among treatments (Figs. 1 and 2). Urea was later applied during 40–60 DAS (). The water level for each urea application date was analyzed and it was found that in most cases (85% of all application dates) urea application event occurred when the field was flooded (water level ≥ 3 cm above the soil surface). It is well known that such flooded conditions can reduce nitrification of NH4+ and the resultant N2O emissions by preventing soil aeration during drainage (Cai et al. Citation1997; Butterbach-Bahl et al. Citation2013). These suggest that managing the timing of N fertilization along with practicing AWD would help minimize the emissions of N2O during draining period.
4.2. Effects of AWD on grain yield and water use
Bouman et al. (Citation2007) found that, if the field is reflooded when the water level reaches 15 cm below the soil surface, rice yield was not reduced. Our results are in agreement with theirs (). Practicing AWD at 15-cm water level below soil surface thus did not induce plant water stress. Furthermore, as shown by no difference between AWD and AWDS (), the timing and the numbers of rewetting-drying cycles did not affect the grain yield. Similar results were also reported by other studies (Bouman and Tuong Citation2001; Carrijo et al. Citation2017). In Thailand, there are only a few studies that investigated directly the impacts of water management on grain yield and GHG emissions. Towprayoon et al. (Citation2005) reported that, although the emissions of CH4 and N2O were reduced by multiple or single mid-drainage, the grain yield was also reduced by 8–11% compared to the locally conventional practice. Draining with fewer drain days during the flowering period was recommended by the authors as a compromise between emissions and yield. Therefore, although this study confirmed no negative impact on rice yield in an acid sulfate soil, it is necessary to investigate whether AWD has the negative impact under different soil types in Thailand.
The amount of total water use (irrigation + rainfall) was significantly improved by AWD and AWDS (). As a result, water productivity was much lower in CF compared to AWD and AWDS. The saving of water by practicing AWD at this site is thus an important benefit, especially for DS. However, the results also indicate that controlling water level as designed for AWD is difficult during the rainy season in Thailand, especially for flooding area like this site. Due to the heavy rainfall, the amount of the total water use during WS was more than double that of DS in this site (see Supplementary Table S1). Thus, the ideal implementation of AWD in the wet season would be difficult although the CH4 emission was reduced even under such conditions (Supplementary Table S1). In rice paddies distributed in the tropical region, the need to manage water in the dry season is more crucial than in the wet season because the drought is common and the water availability is limited. The application of AWD in WS would depend on the benefits obtained for farmers (rice productivity) and country (GHG emission reduction), which should be tested in future studies.
5. Conclusions
This study evaluated the feasibility of AWD in terms of GHG emission, rice productivity, and water saving in a paddy field with an acid sulfate soil in Thailand. The implementation of AWD reduced the seasonal CH4 emission by 49% compared to CF and did not affect the seasonal N2O emission. Rice grain yield did not differ among treatments and the total water use was substantially reduced by AWD. Although the effectiveness of AWD on reducing GHG emission was limited in the seasons with the exceptional events, the 3-year field experiment confirmed that AWD is feasible in this site, especially in DS. The large spatial variability in CH4 and N2O emissions was a characteristic of this site with an acid sulfate soil. The underlying mechanisms should be solved in future studies to improve the description of a process-based model for accurately simulating GHG emission and to revise the national GHG inventory. Simultaneous achievement of GHG emission reduction and maintaining grain yield at an acceptable level is the requirement for the adoption of AWD to the current local farmers because the environmental conservation is not reflected into the farmers’ income. Field trials for the demonstration of AWD’s ability should be carried out under various environmental conditions and soil types in Thailand.
| Depth (cm) | ||||||
|---|---|---|---|---|---|---|
| Properties of soil | 0–20 | 20–50 | 50–65 | 65–90 | 90–110 | 110–150 |
| Texture | Clay | Clay | Clay | Clay | Clay | Clay |
| Sand (%) | 16 | 20 | 22 | 20 | 20 | 20 |
| Silt (%) | 22 | 22 | 22 | 20 | 20 | 20 |
| Clay (%) | 62 | 58 | 56 | 60 | 60 | 60 |
| Bulk density (Mg m−3) | 1.2 | 1.4 | 1.3 | 1.2 | 1.0 | – |
| pH (H2O) | 4.8 | 4.7 | 4.7 | 4.5 | 4.5 | 4.4 |
| Organic matter (g kg−1) | 30 | 8 | 5 | 3 | 3 | 4 |
| Available P (mg kg−1) | 11 | 3 | 2 | 1 | 1 | 1 |
| Available K (mg kg−1) | 115 | 66 | 71 | 99 | 110 | 119 |
| CEC (cmol kg−1) | 37 | 31 | 27 | 28 | 29 | 30 |
| Practice | DS1 | WS1 | DS2 | DS3 | WS3 |
|---|---|---|---|---|---|
| Cropping period | 27 December 2013–4 April 2014 | 19 June–15 September 2014 | 1 January–9 April 2015 | 17 February–25 May 2016 | 30 June–26 September 2016 |
| Crop duration (days) | 97 | 88 | 99 | 97 | 89 |
| Herbicide application | |||||
| Butachlor + Propanil 70% | 6 January (10 DAS | 2 June 2014 | 11 January 2015 (10 DAS) | 27 February 2016 (10DAS) | 10 June 2016 (10 DAS) |
| Thiacloprid at 10 mL ha−1 | 14 January 2014 (18 DAS) | 18 July 2014 (29 DAS) | 20 January 2015 (19DAS) | – | – |
| Azoxystrobin + Difenoconazole at 250 mL ha−1 | 14 March 2014 (77 DAS) | – | 14 March 2014 (77 DAS) | – | – |
| Fertilizer application | |||||
| DAP at 20 kg N ha−1 | 16 January 2014 (20 DAS) | 10 July 2014 (21 DAS | 25 January 2015 (24 DAS) | 14 March 2016 (26 DAS) | 21 July 2016 (21 DAS) |
| P2O5 at 37.5 kg ha−1 | 17 January 2014 (20 DAS) | 11 July 2014 (21 DAS) | 25 January 2015 (24 DAS) | 14 March 2016 (26 DAS) | 21 July 2016 (21 DAS) |
| K2O at 37.5 kg ha−1 | 18 January 2014 (20 DAS) | 12 July 2014 (21 DAS) | 25 January 2015 (24 DAS) | 14 March 2016 (26 DAS) | 21 July 2016 (21 DAS) |
| First urea at 25 kg N ha−1 Second urea at 25 kg N ha−1 | 22 February 2014 (57 DAS) | 31 July 2014 (42 DAS) 20 August 2014 (62 DAS) | 14 February 2015 (44 DAS) 3 March 2015 (61 DAS) | 3 April 2016 (46 DAS) 19 April 2016 (62 DAS) | 4 August 2016 (35 DAS) 25 August 2016 (56 DAS) |
(WS)DAS: days after sowing; DAP: di-ammonium phosphate; not practiced
| Treatment | CH4a (kg ha−1) | N2O (kg ha−1) | GWPa (kgCO2e ha−1) | Grain yield (t ha−1) | Yield-scaled GWP (kgCO2e ton−1) | Water usea (m3 ha−1) | Water productivitya (ton m−3) |
|---|---|---|---|---|---|---|---|
| CF | 17.3 A | 0.785 | 857 B | 4.50 | 0.17 | 8805 a | 0.609 b |
| AWD | 8.8 B | 0.979 | 637 B | 4.19 | 0.14 | 5108 c | 1.086 a |
| AWDS | 21.0 A | 0.851 | 1097 A | 4.44 | 0.22 | 5811 b | 1.023 a |
| Source of variation | p value | ||||||
| Trt | * | 0.554 | † | 0.398 | 0.140 | *** | ** |
| S | *** | * | *** | *** | *** | *** | *** |
| S × Trt | 0.502 | * | 0.129 | * | 0.221 | *** | ** |
aMeans with different letters indicate significant difference at the 5% level for lower-case letter and at the 10% level for capital one. The asterisks †, *, **, and *** indicate the p value of <0.10, <0.05, <0.01, and <0.001, respectively.
Supplemenrat_files.zip
Download Zip (1531 KB)Related Research Data
Acknowledgments
This study was funded by the Ministry of Agriculture, Forestry and Fisheries (MAFF) of Japan through the International Research Project ‘Technology development for circulatory food production systems responsive to climate change’: Development of mitigation options for greenhouse gas emissions from agricultural lands in Southeast Asia 2 (MIRSA 2). We would like to thank Prof. Kazuyuki Inubushi (Chiba University, Japan), Dr Reiner Wassmann (IRRI, Philippines), and Dr Kazuyuki Yagi (NIAES, Japan) for their valuable comments on the earlier version of this article.
Supplemental data
Supplemental data for this article can be accessed here.