Chemosphere 50 (2003) 237–246

www.elsevier.com/locate/chemosphere

Methane emission from fields with differences in

nitrogen fertilizers and rice varieties in Taiwan paddy soils
a,*
Rey-May Liou , Shan-Ney Huang b, Chin-Wei Lin c

a
Department of Environmental Engineering and Health, Chia-Nan University of Pharmacy and Science, Tainan 717, Taiwan, ROC
b
Food and Agriculture Department, Council of Agriculture, Taipei 100, Taiwan, ROC
c
Tainan District Agricultural Improvement Station, Tainan 701, Taiwan, ROC
Received 18 June 2001; received in revised form 24 January 2002; accepted 19 February 2002

Importance of this paper: Methane is an important greenhouse gas, and successful mitigation strategies for agricultural
sources of CH4 require overall understanding of agricultural practices on fluxes of CH4 and on controlling mechanisms in
flooded rice fields. In the previous studies, the methane fluxes were affected greatly with fertilization, rice plant, soil type
and environmental conditions. The effects of nitrogen fertilizer and rice variety on methane fluxes were determined in this
paper, methane fluxes were observed by the static chamber method once per two weeks during the overall growing period.
We attempt to search for promising mitigation candidates to reduce CH4 emission in wetland rice agriculture. Results
indicated that combination of (NH4 )2 SO4 application and Japonica rice plantation would be appropriate methane miti-
gation strategy in Taiwan paddy soils.

Abstract

Flooded rice fields are one of the major biogenic methane sources. In this study, methane emission rates were
measured after transplanting in paddy fields with application of two kinds of nitrogen fertilizers (ammonium sulfate,
NHþ 
4 –N and potassium nitrate, NO3 –N) and with two kinds of rice varieties (Japonica and Indica). The experiment
was conducted in fields located at Tainan District Agricultural Improvement Station in Chia-Yi county (23°250 0800 N,
120°160 2600 E) of southern Taiwan throughout the first and the second crop seasons in 1999. The seasonal methane flux
in the first crop season with NHþ  2
4 –N and NO3 –N ranged from 2.48 to 2.78 and from 8.65 to 9.22 g CH4 m ; and the
2
values ranged 24.6–34.2 and 36.4–52.6 g CH4 m in the second crop season, respectively. In the first crop season, there
were significantly increased 3.1–3.7-fold in methane emission fluxes due to plantation of Indica rice. In comparison of
two rice varieties, the Indica rice variety showed a tendency for larger methane emission than the Japonica rice variety
in the second crop season. Moreover, ammonium sulfate treatment significantly reduced CH4 emissions by 37–85%
emissions compared to potassium nitrate plots. It was concluded that the CH4 emission was markedly dependent on the
type of nitrogen fertilizer and rice variety in Taiwan paddy soils.
Ó 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Methane emission; Paddy field; Nitrogen fertilizer; Rice variety

1. Introduction

Methane is an important greenhouse gas due to the

recent worldwide atmospheric increase at the rate of
*
Corresponding author. Tel.: +886-6-2660035; fax: +886-6- 0.5–0.8% year1 , as well as its strong absorption of in-
2667323. frared–red radiation and heat, which is contributes to
E-mail address: mrmliou@mail.chna.edu.tw (R.-M. Liou). changes in atmospheric chemistry and may cause global
0045-6535/03/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 2 ) 0 0 1 5 8 - 3
238 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246

warming (Bouwman, 1991). Irrigated rice fields are es-
timated to contribute between 25.4 and 54 million ton
per year (Cole et al., 1995) of the total 410–660 million
ton per year emitted globally (Houghton et al., 1995).
Rice (Oryza sativa L.) is the major food crop for
people living in Asia, and about 80% of it is grown under
irrigated wetland conditions (Roger and Watanabe,
1986). Furthermore, the world’s annual rice production
must increase from 518 million tons in 1990 to 760
million tons in 2020 (International Rice Research In-
stitute, 1989). Flooding of irrigated rice fields produces
anaerobic soil conditions conducted to production of
methane (Neue, 1993). Demand for rice over the next 30
years may require an additional 50 million ha to be
cultivated (International Rice Research Institute, 1989).
Owing to conversion of upland soils to wetland rice
cultivation, global methane emissions to the atmosphere
may increase 20% over the next decade. Rice is an im-
portant food crop for people living in Taiwan; the cul-
tivation area in 1997 was 202 010 ha for the first crop
season and 162 202 ha for the second crop season in
Taiwan. There has been considerable research world-
wide on developing farmer-friendly, eco-friendly and
cost-effective technologies to mitigate methane emission
from rice paddies. Changes in water management, nu-
trient management, cultural practices, and cultivars se-
lection may have potential to increase rice production
and productivity, as well as reduce CH4 fluxes in Taiwan
(Yang et al., 1994; Yang and Chang, 1998, 1999; Liou
et al., 1999; Huang et al., 2000). Thus, successful miti-
gation strategies for agricultural sources of CH4 require
overall understanding of agricultural practices on fluxes
of CH4 and on controlling mechanisms. The objectives
of this study were to investigate the effects of various
nitrogen fertilizers and rice varieties on methane emis-
sion throughout the growth stages of rice from the
paddy soils of southern Taiwan, and to provide the data
for estimation of the mitigation potentials in wetland
rice agricultural systems.
2. Materials and methods
2.1. Experimental site
Experimental paddy fields were located at the Tainan
District Agricultural Improvement Station (23°250 0800 N,
120°160 2600 E), Chia-Yi county, southern Taiwan, and the
methane emission data were collected between February
and November in 1999. Climatic properties of test paddy
fields are listed in Table 1, and the properties of test
paddy soils are listed in Table 2. All experiments were
performed in triplicate.
2.2. Rice varieties and nitrogen fertilizers
Taichung sen 10, a Japonica rice and Taikeng 8, an
Indica rice (O. sativa L.) were planted in 1999. Taifei
No. 5 (N:P2 O5 :K2 O ¼ 16:8:12 wt.%, Taiwan Fertilizer
Corporation, Taiwan) was applied as basal fertilizer
(400 kg ha1 ) before rice transplanting. In the 1999 field
experiment, methane emission from fields with 140
þ
kg N ha1 for NO 3 –N (potassium nitrate) and NH4 –N
(ammonium sulfate) were applied as the first and second
top-dressing fertilizers at the active tillering and the
booting stages, respectively. There were four plots, plot
A was Japonica rice and ammonium sulfate, plot B was
Indica rice and ammonium sulfate, plot C was Japonica
rice and potassium nitrate, plot D was Indica rice and
potassium nitrate. Agricultural practices of test paddy
fields are shown in Table 3. Rice field operations in the
first crop season were as follows: plowing on 7 February,

Table 1
Climatic properties of Lu-Tsao rice fields in 1999
Growing season Mean temperature (°C) Total precipitation Rainy days Sunshine duration
(mm season1 ) (d season1 ) (h season1 )
First crop 23.3 337.4 39 667.5
Second crop 26.4 1153.5 65 724.7

Table 2
Properties of test paddy soils in four plots
Plot Treatment Soil characteristics Organic matter Total nitrogen
pH Sand (%) Clay (%) Silt (%) (g kg1 ) (g kg1 )

A Japonica rice and 7.7  0.1 59.9 28.8 11.3 11.5  0.2 0.82  0.04
(NH4 )2 SO4
B Indica rice and (NH4 )2 SO4 7.5  0.2 59.6 27.3 13.1 12.5  0.8 0.94  0.02
C Japonica rice and KNO3 6.6  0.1 58.4 28.6 13.0 14.1  0.6 1.08  0.07
D Indica rice and KNO3 6.5  0.1 61.4 26.4 12.2 14.7  0.9 1.31  0.02
 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246 239

Table 3
Properties of Lu-Tsao rice fields during cultivation
Growing Season length Plot Basal fertilizer Rice variety Topdressing fertilizer
season (day) Type kg ha 1
Type kg N ha1
First crop 128 A Taifei No.5 400 Taichung sen 10 Japonica (NH4 )2 SO4 140
1999 128 B Taifei No.5 400 Taikeng 8 Indica (NH4 )2 SO4 140
128 C Taifei No.5 400 Taichung sen 10 Japonica KNO3 140
128 D Taifei No.5 400 Taikeng 8 Indica KNO3 140
Second 112 A Taifei No.5 400 Taichung sen 10 Japonica (NH4 )2 SO4 140
crop 1999 112 B Taifei No.5 400 Taikeng 8 Indica (NH4 )2 SO4 140
112 C Taifei No.5 400 Taichung sen 10 Japonica KNO3 140
112 D Taifei No.5 400 Taikeng 8 Indica KNO3 140

flooding on 14 February, transplanting on 18 February,
and harvesting on 25 June in the first crop season (total
cultivation period of 128 days). The second crop season
schedule was plowing on 5 July, flooding on 12 July,
transplanting on 19 July and harvesting on 10 Novem-
ber in 1999 (total cultivation period of 112 days). Rice
straw was removed from the field after each harvest.
2.3. Gas sampling period and method
Methane flux methodology followed the recommen-
dation of previous studies (Yang et al., 1994; Buendia
et al., 1998; Yang and Chang, 1998). Gas samples were
collected using closed-chamber technique (IAEA, 1992;
Yang and Chang, 1997) every 2 weeks at 6:00–7:00 a.m.
and 12:00–13:00 p.m. over the whole growth stage.
Stainless steel chambers (25  25  20 cm3 ) were pushed
into the soil approximately 7–15 cm depending on soil
moisture content at the time of the sampling, and each
one was equipped with a custom-made acrylic chamber
(23  23  40, or 23  23  100 cm3 ). Each chamber en-
closed a single plant, four chambers were used in each
measurement in the test field (Yang and Chang, 1997).
After the chamber was pushed into the soil, the electric
fan on the lid was turned on for 10–15 min to homog-
enize the air inside the chamber. Portions (15 ml) of gas
sample were withdrawn by syringe from the headspace
immediately after the mixing (t0 ) and again at 30 (t1 ) and
60 (t2 ) min after the initial sample were taken, four gas
samples were taken from each chamber. Samples were
then stored in 2 ml vacuumed (0.08 mbar) crimp-sealed
vials, and the analysis was performed within 2 weeks
after the sample was taken (Yang and Chang, 1997).
2.4. Analysis of methane concentration of gas samples
Methane emission was determined at 0.5 h intervals
for 1.0 h by examining the changes of methane con-
centration in the chamber. The gas sample was injected
into a Shimadzu 14A gas chromatography equipped with
a flame ionized detector and a glass column (0:26 mm 
2 m) which was packed with Porapak Q (80/100 mesh).
The column temperature was set at 100 °C, and the in-
jection and the detector temperature were set at 130 °C.
Methane concentration was calculated with a standard
curve from 0.1 to 1000 mg kg1 (vol.) (Yang and Chang,
1997).
2.5. Estimation of methane emission
Methane emission from paddy soil was estimated by
the following equation of Rolston (1986), based on the
following equation:
f ¼ ðV =AÞðDC=DtÞ
where f is the methane emission rate (mg m2 h1 ), V the
volume of chamber above the soil (m3 ), A the cross-
section of chamber (m2 ), DC the concentration difference
between zero and t times (mg m3 ), and Dt is the time
duration between the two sampling periods (h). Daily
mean of methane flux was calculated as arithmetic mean
of measurements taken at 6 a.m. and 12 p.m. Annual
amounts of fluxes were computed using the summation
of methane emission in the different sampling stages of
rice plants.
2.6. Analytical methods
Soil pH was determined at 1:1 (w/w) soil to water
suspension using a pH meter. Total nitrogen was mea-
sured by the modified Kjeldahl method (Yang et al.,
1991). Soil organic matter was determined by the wet
oxidation of Walkey–Black method (Nelson and Som-
mer, 1982). Air, water and soil temperatures were mea-
sured with a thermometer. Plant heights and tiller
numbers were determined after drainage in the crop
fields. Crop yields were measured at harvest. Treatments
were replicated three times and flux data subjected to
analysis of variance and Duncan’s multiple range test
(p ¼ 0:05) using the Statistical Analysis System (SAS,
1990).
240 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246

3. Results and discussion

The study was conducted during the first and the

second crop seasons from February to December 1999.
Flux measurements were initiated on 2 March, 1999, a
week after transplantation, and continued until 9 De-
cember, 1999, a week before harvesting for the second
crop season.

3.1. Effect of temperature on methane emission

Methane is produced by methanogens, a group of

anaerobes which require very low redox potentials; use
CO2 , H2 or acetate; and rely on a common set of en-
zymes (Schimel and Gulledge, 1998). Methane is an end-
product of anaerobic decomposition of organic residues
in flooded paddy (Conrad et al., 1989), methanogens
using the products of fermentation to produce CH4 . Soil
Fig. 1. Temporal variation of methane emission rate and soil
temperature is known as an important factor for CH4
temperature from two nitrogen fertilizers and two rice varieties
production in paddy soils, therefore the air, soil and on at 6:00 a.m. during the first crop season in 1999 in the Lu-
water temperature were determined in the early morning Tsao rice fields: plot A (NH4 )2 SO4 and Japonica rice d–d; plot
and at noon at each sampling date. There are two crop B (NH4 )2 SO4 and Indica rice s–s; plot C KNO3 and Japonica
seasons for paddy rice in Taiwan; the first of which is rice r–r; and plot D KNO3 and Indica rice }–}. Each data
longer than the second due to lower temperature in the point is an average of four replicate plots. Bars indicate SD.
first crop season. Total growth period of paddy rice for
the first crop season (February–July) was 128 days and it
was 112 days for the second crop season (July–Decem-
ber). Air temperature was 22.8, 23.5, 25.3, 30, and 30 °C
at the transplanting, active tillering, booting, flowering
and ripening stages, respectively in the first crop season.
In the second crop season, it was 31.8, 30.5, 29, 26.6, and
25.5 °C, respectively. The fluctuation of water and soil
temperatures was narrower than air temperature. Water
and soil have high heat capacity that adjusted the tem-
perature variation. Soil temperature in different crop
seasons can show different variation, being low in the
early morning, and high at noon. Soil temperature in-
creased during cultivation in the first crop season (from
16 to 29 °C, Figs. 1 and 2) and decreased in the second
crop season (from 34 to 23 °C, Figs. 3 and 4). Methane
emission at different growth stages of the rice plant is
shown in Table 4. The integrated seasonal average of
methane emission rate was 1.78 and 13.7 mg m2 h1 for
the first and the second crop season, respectively. The Fig. 2. Temporal variation of methane emission rate and soil
correlation coefficient between methane emission rate temperature from two nitrogen fertilizers and two rice varieties
and temperature was high in the active tillering stage, on at 12:00 p.m. during the first crop season in 1999 in the Lu-
and was moderate in other growth stages (Table 5). Tsao rice fields: plot A (NH4 )2 SO4 and Japonica rice d–d; plot
From the statistical analysis, soil temperature had the B (NH4 )2 SO4 and Indica rice s–s; plot C KNO3 and Japonica
highest correlation coefficient with methane emission rice r–r; and plot D KNO3 and Indica rice }–}. Each data
among the test temperatures in the active tillering stage, point is an average of four replicate plots. Bars indicate SD.
while the value was low between the booting and rip-
ening stages. In the present study, results indicated that afternoon, then declined gradually until leveling off at
methane emission at noon was about 1.7–2-fold higher night in the Philippines. Methane emission patterns in
than that in the early morning in the second crop season. the present studies at different growth stages with two
Neue and Sass (1994) showed that methane emission cropping seasons were different from those found
rate increased rapidly after sunrise, peaked early in the in Texas, Korea and Japan (Yagi and Minami, 1990;
stage of rice due to the low temperature and the small
season, methane emission was low at the early growth
Sass et al., 1991; Shin et al., 1995). In the first crop

dicate SD.
Each data point is an average of four replicate plots. Bars in-
Japonica rice r–r; and plot D KNO3 and Indica rice }–}.
plot B (NH4 )2 SO4 and Indica rice s–s; plot C KNO3 and
Lu-Tsao rice fields: plot A (NH4 )2 SO4 and Japonica rice d–d;
on at 12:00 p.m. during the second crop season in 1999 in the
temperature from two nitrogen fertilizers and two rice varieties
Fig. 4. Temporal variation of methane emission rate and soil

point is an average of four replicate plots. Bars indicate SD.

rice r–r; and plot D KNO3 and Indica rice }–}. Each data
B (NH4 )2 SO4 and Indica rice s–s; plot C KNO3 and Japonica
Tsao rice fields: plot A (NH4 )2 SO4 and Japonica rice d–d; plot
on at 6:00 a.m. during the second crop season in 1999 in the Lu-
temperature from two nitrogen fertilizers and two rice varieties
Fig. 3. Temporal variation of methane emission rate and soil

R.-M. Liou et al. / Chemosphere 50 (2003) 237–246

Table 4
Methane emission from paddy soil at different growth stage of rice plant over the first and the second crop seasons in 1999a
Stage First crop season Second crop season
Japonica rice Indica rice Japonica rice Indica rice Japonica rice Indica rice Japonica rice Indica rice
(NH4 )2 SO4 (NH4 )2 SO4 KNO3 KNO3 (NH4 )2 SO4 (NH4 )2 SO4 KNO3 KNO3
Transplanting 0.62  0.23b 0.18  0.05b 0.07  0.04b 0.05  0.01b 0.34  0.03b 0.29  0.17b 0.53  0.11b 9.42  5.42a
0.04  0.03c 0.01  0.00c 0.03  0.02c 0.03  0.00c 0.39  0.06c 0.39  0.02c 11.85  1.05b 49.43  6.43a
Active tillering 1.24  0.88c 0.44  0.11c 1.50  0.20c 1.57  0.66c 18.50  6.64b 18.97  4.43a;b 21.59  4.84a 21.97  6.44a
0.95  0.72d 0.37  0.06d 1.57  0.26d 1.67  0.79d 19.21  7.63c 32.81  4.89b 29.33  2.74b 39.45  5.60a
Booting 0.47  0.14c 0.67  0.03c 4.09  0.51b 4.46  2.2b 0.90  0.73c 4.80  1.18a;b 3.12  0.45b 6.77  3.17a
0.73  0.06c 1.00  0.02c 6.01  0.10a;b 8.60  6.15a 0.92  0.43c 4.50  1.20a;b;c 3.40  0.48b;c 7.00  3.93a;b
Flowering 1.04  0.68c;d 0.36  0.05d 6.92  1.01a 6.69  1.04a 2.08  0.20c 1.71  0.88c 3.49  0.71b 6.82  0.05a
1.68  0.58d 0.74  0.28d 7.10  0.94b 7.19  2.44b 3.66  0.15c 3.98  1.04c 7.21  1.78b 9.84  0.46a
Ripening 0.09  0.03a;b 0.05  0.02b 0.17  0.04a 0.15  0.03a 0.16  0.06a 0.15  0.05a 0.14  0.10a 0.15  0.01a
0.07  0.01a 0.05  0.04a;b 0.06  0.03a 0.04  0.02a;b 0.01  0.01b 0.05  0.02a;b 0.05  0.03a;b 0.07  0.00a
Average 0.90  0.52 0.41  0.08 2.82  0.36 3.00  1.18 9.17  3.28 12.70  2.43 13.56  2.08 19.54  3.78
a
Rice cultivation was described in the text. All values are given in mg m2 h1 . Means  SD (n ¼ 3), in the same row that do not share the same alphabetic superscript are
significantly different at 5% level according to Duncan’s multiple range test. The data showed range of methane emission in the early morning (6 a.m., low value) and at noon (12
p.m., high value) in each measurement.

241
242 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246
Table 5
The correlation coefficient between methane emission rate at
different rice growth stagesa and temperature of air, water, and
soil
Crop growth Air Water Soil
season temperature temperature temperature
Transplanting 0.36 0.32 0.27
Active tillering 0.72 0.67 0.81
Booting 0.21 0.24 0.24
Flowering 0.20 0.06 0.26
Ripening 0.10 0.11 0.12
a 2
Rice cultivation was described in the text.
size of rice plants; similar results were also found in the
northern Taiwan (Yang et al., 1994; Yang and Chang,
1997; Yang and Chang, 2001). The amount of methane
emission in the second crop season was about 4–13 times
higher than that at the first crop season in the southern
Taiwan, compare to about 2–6 times higher in northern
Taiwan (Yang and Chang, 1997, 1999). Methane emis-
sion was high during the booting and flowering stages in
the first crop season and the values were also high during
the early (from transplanting stage to active tillering
stage) growth stage in the second crop season. The
means of CH4 fluxes in the first crop season were smaller
than those in the second crop season. In the early stage
(3 weeks after transplantation), the largest CH4 emission
rate (74.1 mg m2 h1 ) was obtained in the plot D (In-
dica rice amended with NO 3 –N fertilizer) in the second
crop season. Yang and Chang (1998) found that meth-
ane production rate increased with temperature and had
a linear relationship between 15 and 37 °C in laboratory
studies. High temperature would increase the degrada-
tion of organic matter, and enhance the activities of
methanogens. Sensitivity analysis showed that the tem-
perature–soil moisture interaction is critical (Cao et al.,
1996). CH4 emission was enhanced by a temperature
increase of less than 2 °C, but higher temperature caused
reduced CH4 emissions due to the associated reduction
in soil moisture content. In Taiwan paddies, intermittent
irrigation from the late booting to the ripening stages
(based on the tillering stage of rice growth) is a common
practice to lower ineffective tillers and the accumulation
of organic acids in rhizospheric soils. After drainage, the
methane fluxes would decrease sharply. This strategy of
water management is an important factor in rice pad-
dies; a short aeration period at tillering and before
harvest has been shown to increase yield and reduce
CH4 emission (Huang et al., 2000; Yang and Chang,
2001).
3.2. Effect of nitrogen fertilizer on methane emission
Nitrogen fertilizers are commonly used in rice culti-
vation to increase crop yields. Urea and ammonium
sulfate account for 80–90% of the total nitrogen fertilizer
required in rice cultivation (FAO, 1997). The effects of
nitrogen fertilizers depend on form and amount of the
fertilizer, as well as on mode and time of application.
The effects of NHþ 
4 –N and NO3 –N fertilizers on meth-
anogenesis are not clearly understood and are often
contradictory (Wassmann et al., 1993). Figs. 1–4 show
the methane emission rates in the early morning and at
noon throughout the growth period from the Lu-Tsao
paddy field. The seasonal methane emission in the first
crop season with NHþ 
4 –N and NO3 –N ranged 1.2–2.6
and 8.3–8.8 g CH4 m ; and the value ranges were 34.3–
36.7 and 36.6–58.6 g CH4 m2 in the second crop season,
respectively. In the first crop season, methane emission
rates from NHþ 4 -plots increased slightly from the active
tillering stage to the flowering stage of rice growth, while
the NO 3 -plots fluctuated with time, having a peak in the
flowering stage and being generally larger than those
from NHþ 4 plots. In the second crop season, application
of potassium nitrate did not enhance methane emission
significantly (p < 0:05) over that of the NHþ 4 –N plots
(shown in the Figs. 3 and 4). The methane emission rates
were high at the transplanting and the active tillering
stages, and was low at the flowering and the ripen-
ing stages due to the lower temperature and drainage
treatment in the second crop season. Plot D (Indica rice
amended with potassium nitrate) had the largest meth-
ane emission rates in the all treatments. Methane emis-
sion rates from the NO 3 -plots were 1.5–3.7-fold higher
than the NHþ 4 -plots throughout the growth period.
Researchers have reported that the major controlling
factors for CH4 emission from paddies are water man-
agement, organic amendments, fertilization, rice culti-
vars, and cultural practices (Neue et al., 1996; Minami
and Takata, 1997). However, the results from numerous
studies on the application of nitrogen fertilizer relation
to methane emission have so far been inconsistent,
ranging from stimulation (Banik et al., 1996; Cicerone
and Shetter, 1981; Lindau, 1994) to inhibition (Neue and
Sass, 1994; Sch€ utz et al., 1989) on methane emission.
The influence of various nitrogen fertilizers on methane
emission to the atmosphere was reported by Kimura
(1992), who indicated that the CH4 emission rates from
a pot experiment with ammonium sulfate treatment was
the lowest, followed by ammonium chloride and then
urea fertilization by broadcasting and foliar spray. Such
an inhibition effect of ammonium sulfate was also ob-
served in this study. It seems that the forms of nitrogen
fertilizer application strongly influenced the methane
emission in this study. Based on these results, it was
shown that there was significant inhibition of the
methane fluxes with addition of ammonium sulfate, es-
pecially in the first crop season. Methane production is
the final step in a series of sequential reduction pro-
cesses, with nitrate as the electron acceptor, that occurs
the first after the depletion of oxygen followed by
 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246
manganese (IV), iron (III), sulfate, and carbon dioxide,
a series that can be explained with thermodynamic the-
ory (Zehnder and Stumm, 1988). The electron acceptors
are ordered according to their redox potential, which
determines the energy yield of the process. Previous
studies have shown that nitrate reducers, ferric iron re-
ducers and sulfate reducers successfully compete with
methanogens for H2 , thus inhibiting CH4 production
when nitrate, ferric iron or sulfate is added to metha-
nogenic rice field soil (Achtnich et al., 1995a,b; Kluber
and Conrad, 1998). The main mechanism for nitrate
inhibition of methanogenesis was the formation of toxic
denitrification intermediates rather than competition
between denitrifiers and methanogens for substrate (Roy
and Conrad, 1999). In the Chia-Yi county, the rainy
season concentrated from April to September, the rain-
fall was 337.4 mm (the late growth stage, from April to
June) in the first crop season and was 1153.5 mm (the
early growth stage, from July to September) in the sec-
ond crop season. In this study, the inhibition effect of
nitrate on methane emission was not observed probably
due to the NO 3 –N was easier leaching to soil profiles.
Therefore, addition of ammonium sulfate to paddy soils
in the first crop season would significantly inhibit the
methane emission caused by an immediate decrease
in the hydrogen partial pressure and sulfate reducers
competing with methanogens for common substrates
(Achtnich et al., 1995b). The similar result was also
observed that the rate of the induced sulfate reduction
in the methanogenic soil slurry was 32 nmol g1 h1
(Chidthaisong and Conrad, 2000). It was concluded that
the amendment of ammonium sulfate was probably a
very useful strategy for methane mitigation for the first
crop season in Taiwan paddy fields.
3.3. Effect of rice varieties on methane emission
Rice plants play an important role in the flux of
methane, and there are three processes by which CH4 is
released into the atmosphere from rice paddies. Methane
loss as bubbles from paddy soils is a common mecha-
Table 6
Agronomic traits of rice plant in four plots in 1999
Crop season Treatment
Nitrogen fertilizer Rice variety
First crop (NH4 )2 SO4 Japonica
(NH4 )2 SO4 Indica
KNO3 Japonica
KNO3 Indica
Second crop (NH4 )2 SO4 Japonica
(NH4 )2 SO4 Indica
KNO3 Japonica
KNO3 Indica
243
nism; diffusion loss of CH4 across the water surface is
another process. Therefore, up to 90% of the methane
released from rice fields to the atmosphere may be dif-
fusive transport through the aerenchyma of the rice
plant (Sch€ utz et al., 1989). The production and transport
of methane to the atmosphere depend on the charac-
teristics of rice plant. Root exudates and degrading roots
are also important sources of carbon for methane pro-
duction, especially at the later growth stage. As shown in
Fig. 4, the maximum methane emission rate for Japonica
rice was 34:4  17:6 and 58:8  3:1 mg m2 h1 at noon
in the second crop season with ammonium sulfate and
potassium nitrate, respectively, whereas Indica rice was
68:6  11:8 and 74:1  11:1 mg m2 h2 . The Indica rice
variety showed a tendency for larger methane emission
than the Japonica rice variety, even though the growing
season was sometimes shorter. Larger methane emission
from the Indica rice variety was correlated with the tiller
number, which is larger than for the Japonica rice va-
riety (Table 6). (NH4 )2 SO4 application plots recorded a
higher crop yield than the KNO3 application plots, while
the reverse order was observed in the methane fluxes.
Aerenchyma in leaf blades, leaf sheaths, culm, and roots
provide an efficient gas exchange medium between the
atmosphere and the anaerobic soils. Nouchi et al. (1990)
reported that methane was mostly released from the
culms of rice plants but not from the leaf blades. Wang
et al. (1997) found that no methane was emitted via
transpiration, and leaves were the major release sites in
the early growth stage. The nodes became more im-
portant in the later growth stage. Aerenchyma system
development and/or root exudates and root litter pro-
duction have been suggested and may influence methane
production and transport from rice plants (Bouwman,
1991; Wassmann et al., 1993). Semi-dwarf varieties pro-
duced 36% less methane than tall varieties in the USA
(Lindau et al., 1995). In the continuous flooding treat-
ment, the average methane flux of IR-64 was 20
mg m2 h1 , greater than that of Cisadane variety (14
mg m2 h1 ). In the intermittent irrigation treatment, the
average methane flux of IR-64 was about equal to that
Plant height (cm) Tiller (no./hill) Crop production
(kg ha1 )
69  1.4 28  1.4 6460
69  1.3 34  1.4 6429
73  2.8 29  0.4 6238
69  3.5 30  1.4 6141
72  2.8 32  0.4 6310
67  2.5 24  2.1 5813
72  1.4 28  4.9 6063
71  0.7 32  2.8 5504
244 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246
of the Cisadane variety (both 8.7 mg m2 h1 ) in Indo-
nesia (Husin et al., 1995). The right selection of rice
variety that emits a small amount of CH4 may be an-
other effective mitigation strategy for CH4 emission
from paddy fields.
3.4. Seasonal methane emission during the growth period
Estimating of total methane evolution over the whole
growing season for each treatment, the seasonal meth-
ane flux during the first crop season was calculated to be
2.78 g m2 for plot A; 2.48 g m2 for plot B; 8.65 g m2
for plot C and 9.22 g m2 for plot D, respectively. There
was slightly increased in methane emission fluxes due
to plantation of Indica rice and more than 3.1–3.7-fold
increase due to application of nitrate fertilizer. In the
second crop season, the seasonal methane flux of the
four plots were estimated to be 24.6 g m2 for plot A;
34.2 g m2 for plot B; 36.4 g m2 for plot C and 52.6
g m2 for plot D. There were slight increases in methane
emission due to application of nitrate fertilizer in the
Japonica plots, thus significantly (p < 0:05) increased
methane emission flux in Indica plots due to application
of nitrate fertilizer in the Indica plots.
The harvested area of paddy rice worldwide has in-
crease from 86  106 ha in 1935 to 148  106 ha in 1990,
which means an annual average increase of 1.05%. The
average annual increase has been 1.23% between 1959
and 1985. About 90% of the world’s harvested area of
rice paddies are located in Asia. Rice has only a single
crop season in Texas, Korea and Japan (Yagi and Mi-
nami, 1990; Sass et al., 1991; Shin et al., 1995), while it
has two crop seasons in Taiwan. In 1999, the cultivated
areas were 197 123 and 155 942 ha for the first and the
second crop seasons, respectively. The crop productions
of unhulled rice were 1 225 711 and 690 511 tons for the
first and the second crop seasons, respectively. The
contribution of crop yields in the Chia-Yi county was
about 11.3% of the total crop yields in Taiwan. The
cultivated paddy fields were 19 488 and 17 761 ha for the
first and the second crop seasons in 1999.
In this study, methane emission rates in the second
crop season was higher than that in temperate locations
with only single crop season such as in Japan of 16.2
mg m2 h1 (Yagi and Minami, 1990), in Spain of 4
mg m2 h1 (Seiler, 1984), in California of 10 mg m2 h1
(Cicerone and Shetter, 1981; Cicerone et al., 1983), in
Australia of 3.8 mg m2 h1 (Denmead and Freney,
1990). However, other countries with two seasons such
as in Indonesia of 16.0–26.1 (rainy and dry seasons)
mg m2 h1 (Nugroho et al., 1996), in Philippines of 3.3
(wet season) and 7.9 (dry season) mg m2 h1 (Wass-
mann et al., 1994), in China of 5 (the first crop) and 30
(the second crop) mg m2 h1 (Khalil et al., 1998) were
also lower than the methane fluxes in second crop season
in southern Taiwan. Annual methane fluxes were 18.1
and 24.6 g m2 for Japonica and Indica rice in 1999, and
those were 16.0 and 26.7 g m2 for the application of
ammonium sulfate and potassium nitrate, respectively.
Total methane emission was calculated between 48 355
and 109 132 mg in southern Taiwan, which was higher
than 27 352–69 069 mg with intermittent irrigation in
northern Taiwan (Yang and Chang, 1999). The accu-
mulative methane emission was around 40% lower with
ammonium sulfate than with potassium nitrate, and
which was 26% lower due to plantation of Japonica rice.
In Taiwan, the cultivated area of Japonica rice is larger
than the Indica rice, and farmers often apply ammonium
sulfate as nitrogen fertilizer. Annual methane emission
would be 56% lower with the combination of (NH4 )2 SO4
and Japonica rice than with the combination of KNO3
and Indica rice.
Acknowledgements
The authors thank the assistant research fellow J.C.
Lin for her assistance in the statistical analysis, the
National Science Council of the Republic of China for
financial support (NSC87-2621-P-067F-001 and NSC88-
EPA-Z-067F-001), and the Editor-in-Chief for improve-
ments to the original draft of the paper.
References
Achtnich, C., Bak, F., Conrad, R., 1995a. Competition for
electron donors among nitrate reducers, ferric iron reducers,
sulfate reducers, and methanogens in anoxic paddy soil.
Biol. Fertil. Soils 19, 65–72.
Achtnich, C., Schuhmann, A., Wind, T., Conrad, R., 1995b.
Role of interspecies H2 transfer to sulfate and ferric iron-
reducing bacteria in acetate consumption in anoxic paddy
soil. FEMS Microbiol. Ecol. 16, 61–69.
Banik, A., Sen, M., Sen, S.P., 1996. Effects of inorganic
fertilizers and micronutrients on methane production from
wetland rice (Oryza sativa L.). Biol. Fertil. Soils 21, 319–
322.
Bouwman, C.M., 1991. Agronomic aspects of wetland rice
cultivation and associated methane emissions. Biogeochem-
istry 15, 65–88.
Buendia, L.V., Neue, H.U., Wassmann, R., Lantin, R.S.,
Javellana, A.M., Arah, J., Wang, Z., Wanfang, L., Marka-
rim, A.K., Corton, T.M., Charoenslip, N., 1998. An efficient
sampling strategy for estimating methane emission from rice
field. Chemosphere 36, 395–407.
Cao, M., Gregson, K., Marshall, S., Dent, J.B., Heal, O.W.,
1996. Global methane emissions from rice paddies. Chemo-
sphere 33, 879–897.
Chidthaisong, A., Conrad, R., 2000. Turnover of glucose and
acetate coupled to reduction of nitrate, ferric iron and
sulfate and to methanogenesis in anoxic rice field soil.
FEMS Microbiol. Ecol. 31, 73–86.
 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246
Cicerone, R.J., Shetter, J.D., 1981. Sources of atmospheric
methane measurements in rice paddies and a discussion.
J. Geophys. Res. 86, 7203–7209.
Cicerone, R.J., Shetter, J.D., Delwicje, C.C., 1983. Seasonal
variation of methane flux from a California rice paddy.
J. Geophys. Res. 88, 11022–11024.
Cole, V., Cerri, C., Minami, K., Mosier, A., Rosenberg, N.,
Sanerback, D., 1995. Agricultural options for mitigation of
greenhouse gas emissions. In: Houghton, R.A.J.T., Meria
Filho, L.G., Callander, B.A., Harris, N., Kattenberg, A.,
Maskell, K. (Eds.), Climate Change. Cambridge University
Press, Cambridge, UK, pp. 745–771.
Conrad, R., Mayer, H.P., Wust, M., 1989. Temporal change of
gas metabolism by hydrogen-syntrophic methanogenic bac-
terial associations in anoxic paddy soil. FEMS Microbiol.
Ecol. 62, 265–274.
Denmead, O.T., Freney, J.R., 1990. Micrometeorological
measurements of methane emission from flooded rice. In:
SCOPE/IGBP Workshop on Trace Gas Exchange in a
Global Perspective, 19–23 February, 1990, Sweden.
FAO (Food and Agriculture Organization), 1997. FAO Pro-
duction Yearbook. FAO, United Nations, Rome.
Huang, S.N., Lin, C.W., Liou, R.M., 2000. Effect of the
continuous flooding and intermittent irrigation on methane
emission of paddy soils in Taiwan. Soil Environ. 3, 217–226,
in Chinese with English abstract.
Husin, Y.A., Murdiyarso, D., Khalil, M.A.K., Rasmussen,
R.A., Shearer, S.S., Sunar, A., Adijuwana, H., 1995.
Methane flux from Indonesian wetland rice: the effects of
water management and rice variety. Chemosphere 31, 3153–
3180.
Houghton, J.T., MeiraFilho, L.G., Callander, B.A., Harris,
Kattenberg, N., Maskell, A.K., 1995. Climate change 1995.
In: The Science of Climate Change. Cambridge University
Press, Cambridge, UK.
International Atomic Energy Agency (IAEA), 1992. In: Manual
on Measurements of Methane and Nitrous Oxide Emissions
from Agriculture, pp. 1–91.
International Rice Research Institute (IRRI), 1989. IRRI
Towards 2000 and Beyond. IRRI, Manila, Phillippines.
Kimura, M., 1992. Methane emission from paddy soils in Japan
and Thailand. In: Batjes, N.H., Bridges, E.M. (Eds.), World
Inventory of Soil Emission Potentials: WISE Report 2.
ISRIC, Wageningen, pp. 43–79.
Khalil, M.A.K., Rasmussen, R.A., Shearer, M.J., Chen, Z.L.,
Yao, H., Yang, J., 1998. Emissions of methane, nitrous
oxide, and other trace gases from rice fields in China.
J. Geophys. Res. 103, 241–250.
Kluber, H.D., Conrad, R., 1998. Effects of nitrate, nitrite, NO
and N2 O on methanogenesis and other redox processes in
anoxic rice field soil. FEMS Microbiol. Ecol. 25, 301–318.
Lindau, C.W., 1994. Methane emissions from Louisiana rice
fields amended with nitrogen fertilizers. Soil Biol. Biochem.
26, 353–359.
Lindau, C.W., Bollich, P.K., DeLaune, R.D., 1995. Effect of
rice variety on methane emission from Louisiana rice. Agri.
Ecosys. Environ. 54, 109–114.
Liou, R.M., Lin, C.W., Huang, S.N., 1999. Effect of organic
fertilizer on methane emission of paddy soils in Taiwan.
Chin. J. Agromet. 6, 139–148, in Chinese with English
abstract.
245
Minami, K., Takata, K., 1997. Atmospheric methane: sources,
sinks, and strategies for reducing agricultural emissions.
Water Sci. Technol. 36, 509–516.
Nelson, D.W., Sommer, L.E., 1982. Total carbon, organic
carbon and organic matter. In: Page, A.L. (Ed.), Methods
of Soil Analysis: Part 2. Chemical and Microbiological
Properties. The American Society of Agronomy, Wiscosin,
pp. 539–580.
Neue, H.U., 1993. Methane emission from rice fields. Bio-
science 43, 466–475.
Neue, H.U., Sass, R.L., 1994. Trace gas emissions from rice
fields. In: Prinn, R.G. (Ed.), Global Atmospheric–Bio-
spheric Chemistry. Plenum Press, New York, pp. 119–147.
Neue, H.U., Wassmann, R., Lantin, R.S., Alberto, M.A.C.R.,
Aduna, J.B., Javellana, A.M., 1996. Factors affecting
methane emission from rice fields. Atmos. Environ. 30,
1751–1754.
Nouchi, I., Mariko, S., Aoki, K., 1990. Mechanism of methane
transport from the rhizosphere to the atmosphere through
rice plants. Plant Physiol. 94, 59–66.
Nugroho, S.G., Lumbanraja, J., Suprapto, H., Sunyoto,
A..W.S., Haraguchi, H., Kimura, M., 1996. Three-year
measurement of methane emission from an Indonesian
paddy field. Plant Soil 181, 287–293.
Roger, P.A., Watanabe, I., 1986. Technologies for utilizing
biological nitrogen fixation in wetland rice: Potentialities,
current usage, and limiting factors. Fertil. Res. 9, 39–77.
Rolston, D.E., 1986. Gas flux. In: Klute, A. (Ed.), Methods of
Soil Analysis, second ed. American Society of Agronomy
and Soil Science Society of America, Wisconsin, pp. 1103–
1119.
Roy, R., Conrad, R., 1999. Effect of methanogenic precursors
on the methane production by nitrate in anoxic rice field
soil. FEMS Microbiol. Ecol. 49, 49–61.
Sass, R.L., Fisher, F.M., Tuner, F.T., Jund, M.F., 1991.
Methane emission from rice fields as influence by solar
radiation, temperature and straw incorporation. Global
Biogeochem. Cycles 5, 335–350.
Schimel, J.P., Gulledge, J., 1998. Microbial community struc-
ture and global trace gases. Global Change Biol. 4, 745–
758.
Sch€utz, H., Holzapfel-Pschorn, A., Conrad, R., Rennenberg,
H., Seiler, W., 1989. A three-year continuous record on the
influence of daytime, season, and fertilizer treatment on
methane emission rates from an Italian rice paddy. J.
Geophys. Res. 94, 16405–16416.
Seiler, W., 1984. Contribution of biological processes to the
global budget of methane in the atmosphere. In: Current
Perspectives in Microbial Ecology, pp. 468–477.
Shin, Y.K., Lee, Y.S., Yun, S.H., Park, M.E., 1995. Methane
emission measurement in rice paddy of Korea. In: The
Proceedings of International Symposium on Environment
and Biometeorology. China Agricultural Scientech Press,
Beijing, pp. 495–503.
SAS, 1990. SAS/ST in User’s Guide, version 6, fourth ed.
Staistical Analysis System, Cary, NC.
Wang, B., Neue, H.U., Samonte, H.P., 1997. Role of rice in
mediating methane emission. Plant Soil 189, 107–115.
Wassmann, R., Papen, H., Rennenberg, H., 1993. Methane
emission from rice paddies and possible mitigation strate-
gies. Chemosphere 26, 201–217.
246 R.-M. Liou et al. / Chemosphere 50 (2003) 237–246
Wassmann, R., Neue, H.U., Lantin, R.S., Aduna, J.B.,
Alberto, M.C.R., Andales, M.J., Tan, M.J., Denier van
der Gon, A.C., Hoffmann, H., Papen, H., Rennenberg, H.,
Seiler, W., 1994. Temporal patterns of methane emissions
from wetland rice fields treated by different modes of N
application. J. Geophic. Res. 16, 457–462.
Yagi, K., Minami, K., 1990. Effect of organic matter applica-
tion on methane emission from some Japanese paddy fields.
Soil Sci. Plant Nutr. 36, 599–610.
Yang, S.S., Chang, H.L., Wei, C.B., Lin, H.C., 1991. Reduce
waste production in the Kjeldahl methods. J. Biomass
Energy Soc. China 10, 147–155.
Yang, S.S., Lin, C.C., Chang, E.H., Chung, R.S., Huang, S.N.,
1994. Effect of fertilizer, soil type, growth season on
methane production and emission in the paddy soils of
Taiwan. J. Biomass Energy Soc. China 13, 68–87.
Yang, S.S., Chang, H.L., 1997. Effect of fertilizer application on
methane emission/production in the paddy soils of Taiwan.
Biol. Fertil. Soils 25, 245–251.
Yang, S.S., Chang, H.L., 1998. Effect of environmental
conditions on methane production and emission from
paddy soil. Agri. Ecosyst. Environ. 69, 69–80.
Yang, S.S., Chang, H.L., 1999. Diurnal variation of methane
emission from paddy fields at different growth stages of rice
cultivation in Taiwan. Agri. Ecosys. Environ. 76, 75–84.
Yang, S.S., Chang, H.L., 2001. Effect of green manure
amendment and flooding on methane emission from paddy
fields. Chemosphere––Global Change Sci. 3, 41–49.
Zehnder, A.J.B., Stumm, W., 1988. Geochemistry and biogeo-
chemistry of anaerobic habitats. In: Zehnder, A.J.B. (Ed.),
Biology of Anaerobic Microorganisms. Wiley, New York,
pp. 1–38.