Content uploaded by Patikorn Sriphirom
Author content
All content in this area was uploaded by Patikorn Sriphirom on Oct 10, 2019
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tssp20
Soil Science and Plant Nutrition
ISSN: 0038-0768 (Print) 1747-0765 (Online) Journal homepage: http://www.tandfonline.com/loi/tssp20
Evaluating the effects of alternate wetting and
drying (AWD) on methane and nitrous oxide
emissions from a paddy field in Thailand
Amnat Chidthaisong, Nittaya Cha-un, Benjamas Rossopa, Chitnucha
Buddaboon, Choosak Kunuthai, Patikorn Sriphirom, Sirintornthep
Towprayoon, Takeshi Tokida, Agnes T. Padre & Kazunori Minamikawa
To cite this article: Amnat Chidthaisong, Nittaya Cha-un, Benjamas Rossopa, Chitnucha
Buddaboon, Choosak Kunuthai, Patikorn Sriphirom, Sirintornthep Towprayoon, Takeshi Tokida,
Agnes T. Padre & Kazunori Minamikawa (2018) Evaluating the effects of alternate wetting and
drying (AWD) on methane and nitrous oxide emissions from a paddy field in Thailand, Soil Science
and Plant Nutrition, 64:1, 31-38, DOI: 10.1080/00380768.2017.1399044
To link to this article: https://doi.org/10.1080/00380768.2017.1399044
View supplementary material Published online: 06 Nov 2017.
Submit your article to this journal Article views: 245
View related articles View Crossmark data
Citing articles: 1 View citing articles
Evaluating the effects of alternate wetting and drying (AWD) on methane and
nitrous oxide emissions from a paddy field in Thailand
Amnat Chidthaisong
a
, Nittaya Cha-un
a
, Benjamas Rossopa
b
, Chitnucha Buddaboon
b
, Choosak Kunuthai
a
,
Patikorn Sriphirom
a
, Sirintornthep Towprayoon
a
, Takeshi Tokida
c
, Agnes T. Padre
d
and Kazunori Minamikawa
c
a
The Joint Graduate School of Energy and Environment (JGSEE) and Center of Excellence on Energy Technology & Environment, King Mongkut’sUniversity
of Technology Thonburi (KMUTT), Bangkok, Thailand;
b
Prachin Buri Rice Research Center, Rice Department, Ministry of Agriculture and Cooperatives,
Ban Sang, Thailand;
c
Institute for Agro-Environmental Sciences, National Agriculture and Food Research Organization (NARO), Tsukuba, Japan;
d
Crop and
Environmental Sciences Division (CESD), International Rice Research Institute (IRRI), Los Baños, Philippines
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 contribu-
tion 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.
ARTICLE HISTORY
Received 22 May 2017
Accepted 27 October 2017
KEY WORDS
Rice; alternate wetting and
drying (AWD); methane;
nitrous oxide; acid sulfate soil
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)
2015). Within the agriculture sector, methane (CH
4
) emission
from paddy field is the major source that accounts for 72% of
the sector’s totals. Mitigating emission of CH
4
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 2015). 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 cultiva-
tion 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 2012; Thailand Research Fund (TRF) 2016). 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 CH
4
from paddy field through water manage-
ment. Drainage at appropriate timing during rice growth can
reduce CH
4
emission (e.g., Minamikawa and Yagi 2009) because
it introduces oxygen into the soil and thereby inhibiting metha-
nogenesis and promoting methanotrophic activity (Conrad
1996; Conrad and Rothfuss 1991;BenderandConrad1992). In
central Thailand, for example, mid-season drainage was able to
reduce CH
4
emission by 35% compared to the conventional
continuous flooding (CF; Towprayoon et al.2005).
A precise water controlling technique known as safe alter-
nate wetting and drying (AWD) has been originally developed
to save irrigation water use in paddy field (Bouman and Tuong
2001). This technique controls surface water level not to fall
below a soil depth of 15 cm. Water consumption and CH
4
emission were reported to be effectively reduced (e.g., LaHue
et al.2016). However, there are different results in the litera-
tures on the effects of AWD on nitrous oxide (N
2
O) and rice
CONTACT Amnat Chidthaisong amnat_c@jgsee.kmutt.ac.th The Joint Graduate School of Energy and Environment (JGSEE) and Center of Excellence on Energy
Technology & Environment, King Mongkut’s University of Technology Thonburi (KMUTT), 126 Prachauthit Rd, Bangmod, Tungkru, Bangkok 10140, Thailand
Supplemental data for this article can be accessed here.
SOIL SCIENCE AND PLANT NUTRITION, 2018
VOL. 64, NO. 1, 31–38
https://doi.org/10.1080/00380768.2017.1399044
© 2017 Japanese Society of Soil Science and Plant Nutrition
grain yields. LaHue et al.(2016) reported a reduction of
60–87% in CH
4
emission while maintaining a low N
2
O emis-
sion 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.
(2016) reported in a 2-year study in Italian paddy field that
70% water consumption and 33% yields, and 97% CH
4
emis-
sions were reduced by implementation of AWD as compared
to permanent flooded fields. N
2
O emissions were increased by
more than 5-fold under AWD. In the second year, with 40%
water saving, the reductions of rice yields and CH
4
emissions
(13% and 11%, respectively) were not significant, but N
2
O
emissions were more than doubled. Carrijo et al.(2017) con-
ducted 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 mitiga-
tion, 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.2000)asit
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 ear-
lier, 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 (H
2
O) 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 Table 1. 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.
2.2. Experimental design
The experimental plots were laid out following the rando-
mized 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 sur-
face water depth of 10 cm during 15–90 DAS. For the AWD
treatment, the fields were flooded (5-cm water depth) when-
ever 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
,37.5kgP
2
O
5
ha
−1
, and 37.5 kg K
2
Oha
−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.
2.3. Measurements
The closed chamber method was used to collect gas sample. A
square-shape chamber was made from acrylic, with the dimen-
sion 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
Table 1. Physical and chemical properties of acid sulfate paddy soil at Prachin
Buri Rice Research Center, not determined.
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 (H
2
O) 4.8 4.7 4.7 4.5 4.5 4.4
Organic matter (g kg
−1
)30853 3 4
Available P (mg kg
−1
)11321 1 1
Available K (mg kg
−1
) 115 66 71 99 110 119
CEC (cmol kg
−1
) 37312728 29 30
32 A. CHIDTHAISONG ET AL.
sampling was conducted every week and 1–5 days after fertili-
zer 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, CH
4
concentration was analyzed
by a gas chromatograph equipped with FID and Custom
Packed SG 804 column (Shimadzu GC 8A, Japan), using N
2
as
carried gas. The N
2
O concentration was analyzed by gas chro-
matograph (Shimadzu GC-14B, Japan) equipped with ECD and
Porapak column. Specific operation conditions of the gas chro-
matographs were described in Cha-un et al.(2016). For the
subsequent seasons, the Agilent 7890B (Agilent Technologies,
Inc., USA) gas chromatograph equipped with FID operated at
300 °C for CH
4
and ECD at 300 °C for N
2
OandHaySepQpacked
column was used. N
2
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.(2015).
Seasonal cumulative CH
4
and N
2
O emissions were estimated
by linear interpolation and numerical integration between
sampling times. Due to the malfunction of ECD-GC in WS1,
the N
2
O data were excluded from the analysis. The global
warming potential (GWP), CO
2
-equivalent cumulative emission
of CH
4
and N
2
O, was calculated using the GWPs for 100-year
time horizon with inclusion of climate-carbon feedbacks (IPCC
2013; 34 for CH
4
and 298 for N
2
O).
Daily air temperature (minimum and maximum) and pre-
cipitation 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 grow-
ing 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 repli-
cations. Variance components were estimated by the
restricted maximum-likelihood method with the ‘nobound’
option, and the denominator degrees of freedom were esti-
mated by the Kenward–Roger approximation (see the
Supplemental Material for the SAS scripts). Because emissions
of CH
4
and N
2
O showed highly skewed distributions and
violated normality and homoscedasticity assumptions, and
Box-Cox transformation was conducted for CH
4
,N
2
O, GWP,
and yield-scaled GWP using the ‘powerTransform’function in
the ‘car’package of R (Box and Cox 1964; Fox and Weisberg
2011). To test differences among water managements, Tukey's
Table 2. Key crop cultivation practices and dates in the dry seasons(DS) and wet seasons (WS).
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)
P
2
O
5
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)
K
2
O 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
SOIL SCIENCE AND PLANT NUTRITION 33
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–Cand 2A,B). During DS, relatively small rain-
fall 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 gen-
erally lower during WS than during DS (Figs. 1D–Fand 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.
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 com-
pared to CF and significantly differed among the five cropping
seasons (Table 3). 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.
3.2. GHG emission
3.2.1. CH
4
The CH
4
fluxes varied spatially as indicated by large error bars
for each sampling date (Figs. 1J–Land 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 CH
4
emission (Table 3).
Although the result of Tukey’sHSDtestwasmarginal
(p<0.1),thetotalCH
4
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 CH
4
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.
020406080
0
50
100
150
200
020406080 0 20 40 60 80 100
-15
-10
-5
0
5
10
10
20
30
40
7DAS
CF: 5.30
0
1
2
3
4
5
59DAS
CF: 306
o
Days after sowing
DS1 DS2 DS3
N2O (µg N m-2h-1)CH
4 (mg CH4 m-2 h-1)Water table (cm) Temperature(°C)
0
20
40
60
80
100
Rainfall (mm)
j
g
d
a
m
h
e
b
k
n
i
f
c
l
CF
AWD
AWDS
Max Min
CF
AWD
AWDS
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), CH
4
flux (j, k, l), and
N
2
O flux (m, n) for three water management practices in dry season (DS1, DS2, and DS3). Error bars for CH
4
and N
2
O 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.
34 A. CHIDTHAISONG ET AL.
3.2.2. N
2
O
The relatively high N
2
O fluxes were observed in DS1 and DS2
among all the seasons (Figs. 1M–Oand 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 N
2
O emission was signifi-
cantly different among cropping seasons but not significantly
different among treatments (Table 3). Keeping flooded con-
ditions 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 CH
4
and N
2
O
The GWP was significantly different among cropping seasons
and marginally different among treatments (Table 3). The GWP
was highest in AWDS among the three treatments mainly due
to the exceptionally large CH
4
emission in WS3. The GWP in
AWD was 25% smaller than that in CF, although not signifi-
cant. The N
2
O emission accounted for 39–62% (the range of
three treatments) of the GWP in DS whereas 3–13% in WS. The
highest N
2
O’s contribution was from AWD due to the smaller
CH
4
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 (Table 3). On
the other hand, there was significant effect of cropping season
020406080
0
50
100
150
200
0 20 40 60 80 100
10
20
30
40
80 90
10
20
30
40
50
70 80 90
10
20
30
40
50
-15
-10
-5
0
5
10
0
1
2
3
4
5
0
20
40
60
80
100
Rainfall (mm)
Days after sowing
N2O (µg N m-2 h-1)CH
4 (mg CH4 m-2 h-1)Water table (cm) Temperature (°C)
WS3
Max Min
a
c
e
g
f
d
b
h
107 mm
CF
AWD
AWDS
84DAS
AWDS:
7.93
7DAS
AWDS: 8.16
CH488DAS
AWD: 9.99
AWDS: 8.24
i
WS1
CF
AWD
AWDS
Figure 2. Seasonal variations in daily rainfall (a, b), daily maximum and minimum air temperature (c, d), mean surface water level (e, f), CH
4
flux (g, h), and N
2
O flux
(i) for three water management practices wet seasons (WS1 and WS3). Error bars for CH
4
and N
2
O 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.
SOIL SCIENCE AND PLANT NUTRITION 35
due mainly to the low yield in DS3 by the blast disease and
significant interaction due to the contradictory results of treat-
ments among cropping seasons (see Supplementary Table S1).
The yield-scaled GWP was significantly affected only by
cropping season (Table 3). On the other hand, water produc-
tivity was significantly improved by the implementation of
AWD and AWDS (Table 3). 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 manage-
ment technique to effectively mitigate CH
4
emission in paddy
field (Pandey et al.2014; LaHue et al.2016; Liang et al.2016). If
effectively drained, the emission of CH
4
is expected to be
substantially reduced. However, the effectiveness of drainage
practices on CH
4
emission reduction depends on the efficiency
of water control, soil type, and other cultivation practices (Yan
et al.2005; Liang et al.2016). In the current study, practicing
AWD in this soil resulted in marginal (p< 0.1) reduction in the
CH
4
emission compared to CF by Tukey’s HSD test (Table 3).
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 CH
4
emission.
Contrary to our expectations, the effectiveness of CH
4
emis-
sion reduction by AWD and AWDS treatments was occasion-
ally limited (Table 3 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 CH
4
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 CH
4
emission in
WS3, especially under AWDS (Fig. 2H), could be partly
explained by heavy rainfall after rice sowing. These excep-
tional events would have masked the effects of AWD on CH
4
emission reduction.
One common characteristic of CH
4
emission in this site is its
large spatial variation among three replications: the coeffi-
cients 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
CH
4
emissions are well known and have been reported in
many studies (e.g., Wagner and Pfeiffer 1997; Wachinger
et al.2000; Sey et al.2008). The presence of microsite and
microbial activity within the microsite was reported to be a
hot spot of CH
4
production even in oxic soil and a source of
emission spatial variability (Peters and Conrad 1995; Parkin
1987; Sey et al.2008; Plaza-Bonilla et al.2014). In addition,
the high spatial variability in CH
4
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 CH
4
would be
Table 3. Statistical analysis results of combined seasonal means of CH
4
,N
2
O, grain yield, and water use, with the effects of both treatment (Trt), growing season (s), and a combination of treatment and season (S × Trt).
Treatment CH
4a
(kg ha
−1
)N
2
O (kg ha
−1
) GWP
a
(kgCO
2
eha
−1
) Grain yield (t ha
−1
) Yield-scaled GWP (kgCO
2
e ton
−1
) Water use
a
(m
3
ha
−1
) Water productivity
a
(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 pvalue
Trt * 0.554 †0.398 0.140 *** **
S *** * *** *** *** *** ***
S × Trt 0.502 * 0.129 * 0.221 *** **
a
Means 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 pvalue of <0.10, <0.05, <0.01, and <0.001,
respectively.
36 A. CHIDTHAISONG ET AL.
released and it would be also possible that anaerobic micro-
sites 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 CH
4
emission was also another distinct charac-
teristic at the site. High acidity and slowly developed anaero-
bic conditions as described by Chareonsilp et al.(2000) would
explain such low emission of CH
4
. Further study is necessary to
elucidate the effects of soil type and soil property on the
magnitude of CH
4
emission and on the effectiveness of drai-
nage practices on reducing CH
4
emission.
Similar to the case of CH
4
, large spatial variations in N
2
O
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 N
2
O emissions among treatments. Another reason
for the insignificant differences in N
2
O emissions among treat-
ments 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 (Table 2). 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 ≥3cmabovethesoilsurface).It
is well known that such flooded conditions can reduce nitrifica-
tion of NH
4+
and the resultant N
2
O emissions by preventing soil
aeration during drainage (Cai et al.1997;Butterbach-Bahlet al.
2013). These suggest that managing the timing of N fertilization
along with practicing AWD would help minimize the emissions
of N
2
Oduringdrainingperiod.
4.2. Effects of AWD on grain yield and water use
Bouman et al.(2007) 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
(Table 3). 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 (Table 3),
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 2001; Carrijo et al.2017). 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.(2005) reported that, although
the emissions of CH
4
and N
2
O 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 (Table 3). 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 CH
4
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 reduc-
tion), 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 CH
4
emission by 49% compared
to CF and did not affect the seasonal N
2
O 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, espe-
cially in DS. The large spatial variability in CH
4
and N
2
O
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 accep-
table level is the requirement for the adoption of AWD to
the current local farmers because the environmental conser-
vation 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.
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 respon-
sive to climate change’: Development of mitigation options for green-
house 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.
Funding
This study was funded by the Ministry of Agriculture Forestry and Fisheries
(MAFF) of Japan through the International Research Project ‘Technology
SOIL SCIENCE AND PLANT NUTRITION 37
development for circulatory food production systems responsive to cli-
mate change’: Development of mitigation options for greenhouse gas
emissions from agricultural lands in Southeast Asia 2 (MIRSA 2).
References
Bender M, Conrad R 1992: Kinetics of methane oxidation in oxic soils
exposed to ambient air and or high CH
4
mixing ratios. FEMS
Microbiol. Ecol.,101, 261–270.
Bouman BAM, Humphreys E, Yuong T, Barker R 2007: Rice and water. Adv.
Agron.,92, 187–237. doi:10.1016/S0065-2113(04)92004-4
Bouman BAM, Tuong TP 2001: Field water management to save water and
increase its productivity in irrigated rice. Agric. Water Manage,65,193–210.
Box GE, Cox DR 1964: An analysis of transformations. J. Royal Stat. Soc.
Series B.,26, 211–246.
Butterbach-Bahl K, Baggs EM, Dannenmann M, Kiese R, Zechmeister-
Boltenstern S 2013: Nitrous oxide emissions from soils: how well do
we understand the processes and their controls? Phi. Philos. Trans. R
Soc. Lond B Biol. Sci.,368,1–13.
Cai Z, Xing G, Yan X, Xu H, Tsuruta H, Yagi K, Minami K 1997: Methane and
nitrous oxide emissions from rice paddy fields as affected by nitrogen
fertilisers and water management. Plant Soil,196,7–14.
Carrijo DR, Lundy ME, Linquist BA 2017: Rice yields and water use under
alternate wetting and dryingirrigation: a meta-analysis. Field Crops Res.,
203, 173–180.
Chareonsilp N, Buddhaboon C, Promnart P, Wassmann R, Lantin RS 2000:
Methane emission from deepwater rice fields in Thailand. Nutr. Cycl.
Agroecosys,58, 121–130.
Cha-un N, Chidthaisong A, Yagi K, Sudo S, Towprayoon S 2016:
Greenhouse gas emissions, soil carbon sequestration and crop yields
in a rain-fed rice field with crop rotation management. Agric. Ecosyst.
Environ.,237, 109–120. 10.1016/j.agee.2016.12.025
Thailand Research Fund (TRF) 2016: Thailand’s second assessment report on
climate change 2016. In Trf, Eds. Chidthaisong A, Vannagovida P, Jewjiam
M, Limsakul A, Chinwanno S, Keansantisukmongkol C, Ring Co., Ltd.,
Bangkok, 527 (in Thai).
Conrad R 1996: Soil microorganisms as controllers of atmospheric trace
gases (H
2
, CO, CH
4
, OCS, N
2
O and NO). Microbiol. Rev.,60, 609–640. doi:
PMID: 8987358
Conrad R, Rothfuss F 1991: Methane oxidation in the soil surface layer of a
flooded rice field and the effect of ammonium. Biol. Fertil. Soils,12,28–
32. 10.1007/BF00369384
IPCC 2012: Summary for Policymakers. In Managing the Risks of Extreme
Events and Disasters to Advance Climate Change Adaptation, Eds. Field
CB, Barros V, Stocker TF et al., A Special Report of Working Groups I and
II of the Intergovernmental Panel on Climate Change. 1–19. Cambridge
University Press, Cambridge, UK, and New York, NY, USA
Fox J, Weisberg S 2011: An R Companion to Applied Regression. Sage,
Thousand Oaks, CA, USA.
Lagomarsino A, Agnelli AE, Linquist B, Adviento-Borbe MA, Agnelli A,
Gavina G, Ravaglia S, Ferrara M 2016: Alternate wetting and drying of
rice reduce CH4 emissions but triggered N2O peaks in a clayey soil of
Central Italy. Pedospere,46, 533–548. 10.1016/S1002-0160(15)60063-7
LaHue GT, Chaney RL, Adviento-Borbe MA, Linquista BA 2016: Alternate
wetting and drying in high yielding direct-seeded rice systems accom-
plishes multiple environmental and agronomic objectives. Agric.
Ecosyst. Environ.,229,30–39. 10.1016/j.agee.2016.05.020
Liang K, Zhong X, Huang N, Lampayan RM, Pan J, Tian K, Kiu Y 2016:Grain
yield, water productiviey and CH4 emission of irrigated rice in response to
water management in south China. Agric. Water Manage,163,319–331.
Minamikawa K, Tokida T, Sudo S, Padre A, Yagi K 2015: Guidelines for
Measuring CH
4
and N
2
O Emissions from Rice Paddies by a Manually
Operated Closed Chamber Method, 76. National Institute for Agro-
Environmental Sciences, Tsukuba, Japan.
Minamikawa K, Yagi K 2009: Possibility of water management for mitigat-
ing total emission of greenhouse gases from irrigated paddy fields. In
Climate Change and Crops, Eds. Singh SN, 307–328. Springer-Verlag,
Berlin, Heidelberg
Office of Natural Resources and Environmental Policy and Planning (ONEP)
2015: Thailand’s First Biennial Update Report Under the United Nations
Framework Conventionon Climate Change. 74 pp.
Pandey A, Mai VT, Vu DQ, Bui TPL, Mai TLA, Jensen LS, De Neergaard A
2014: Organic matter and water management strategies to reduce
methane and nitrous oxide emissions from rice paddies in Vietnam.
Agric. Ecosyst. Environ.,196, 137–146.
Parkin TB 1987: Soil Microsites as a source of denitrification variability. Soil
Sci. Soc. Am. J.,51, 1194–1199. 10.2136/sssaj1987.0361599
5005100050019x
Peters V, Conrad R 1995: Methanogenic and other strictly anaerobic
bacteria in desert soil and other oxic soils. Appl. Environ. Microbiol.,
61, 1673–1676.
Plaza-Bonilla D, Cantero-Martínez C, Álvaro-Fuentes J 2014: Soil manage-
ment effects on greenhouse gases production at the macroaggregate
scale. Soil Biol. Biochem.,68, 471–481.
Sey BK, Manceur AM, Whalen JK, Gregorich EG, Rochette P 2008: Small-
scale heterogeneity in carbon dioxide, nitrous oxide and methane
production from aggregates of a cultivated sandy-loam soil. Soil Biol.
Biochem.,40, 2468–2473.
IPCC 2013: Climate change 2013: the physical science basis. Contribution
of Working Group I to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change (Eds.: Stocker TF, Qin D,
Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V,
Midgley PM). Cambridge University Press, Cambridge, United Kingdom
and New York, NY, USA, 1535 pp.
Towprayoon S, Smakgahn K, Poonkaew S 2005: Mitigation of methane and
nitrous oxide emissions from drained irrigated rice fields. Chemosphere,
59, 1547–1556.
Wachinger G, Fiedler S, Zepp K, Gattinger A, Sommer M, Roth K 2000:
Variability of soil methane production on the micro-scale: spatial asso-
ciation with hot spots of organic material and Archaeal populations.
Soil Biol. Biochem.,32, 1121–1130.
Wagner D, Pfeiffer E 1997: Two temperature optima of methane produc-
tions in a typical soil of the Elbe marshland. FEMS Microbiol. Ecol.,22,
145–153.
Yan XY, Yagi K, Akiyama H, Akimoto H 2005: Statistical analysis of the
major variables controlling methane emission from rice fields. Glob.
Chang. Biol.,7, 1131–1141. 10.1111/j.1365-2486.2005.00976.x
38 A. CHIDTHAISONG ET AL.
