Science of the Total Environment 472 (2014) 112–124

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Nitrous oxide and methane emissions from optimized and alternative

cereal cropping systems on the North China Plain: A two-year field study
Bing Gao a, Xiaotang Ju a,⁎, Fang Su a, Qingfeng Meng a, Oene Oenema b, Peter Christie a,c,
Xinping Chen a, Fusuo Zhang a
a
College of Resources and Environmental Sciences, China Agricultural University, Beijing 100193, China
b
Wageningen University and Research, Alterra, Wageningen, The Netherlands
c
Agri-Environment Branch, Agri-Food and Biosciences Institute, Belfast BT9 5PX, UK

H I G H L I G H T S

• Yields, resource use efficiency and N2O + CH4 emission differ among cropping systems.
• An alternative cropping system is recommended for the North China Plain region.
• A rotation cycle of it is in sequence winter wheat, summer maize and spring maize.
• This alternative is better than an optimized winter wheat–summer maize system.
• It is sustainable due to saving water and N, maintaining yields and reducing GHG.

a r t i c l e i n f o a b s t r a c t

Article history: The impacts of different crop rotation systems with their corresponding management practices on grain yield,
Received 16 June 2013 greenhouse gas emissions, and fertilizer nitrogen (N) and irrigation water use efficiencies are not well
Received in revised form 1 November 2013 documented. This holds especially for the North China Plain which provides the staple food for hundreds
Accepted 1 November 2013
of millions of people and where groundwater resources are polluted with nitrate and depleted through
Available online 27 November 2013
irrigation. Here, we report on fertilizer N and irrigation water use, grain yields, and nitrous oxide (N2O)
Keywords:
and methane (CH 4) emissions of conventional and optimized winter wheat–summer maize double-
N2O emission cropping systems, and of three alternative cropping systems, namely a winter wheat–summer maize
CH4 uptake (or soybean)–spring maize system, with three harvests in two years; and a single spring maize system
Nitrogen with one crop per year. The results of this two-year study show that the optimized double-cropping system
Cropping systems led to a significant increase in grain yields and a significant decrease in fertilizer N use and net greenhouse
Grain yield gas intensity, but the net greenhouse gas N2O emissions plus CH4 uptake and the use of irrigation water did
not decrease relative to the conventional system. Compared to the conventional system the net greenhouse
gas emissions, net greenhouse gas intensity and use of fertilizer N and irrigation water decreased in the
three alternative cropping systems, but at the cost of grain yields except in the winter wheat–summer
maize–spring maize system. Net uptake of CH4 by the soil was little affected by cropping system. Average
N 2O emission factors were only 0.17% for winter wheat and 0.53% for maize. In conclusion, the winter
wheat–summer maize–spring maize system has considerable potential to decrease water and N use and
decrease N2O emissions while maintaining high grain yields and sustainable use of groundwater.
Crown Copyright © 2013 Published by Elsevier B.V. All rights reserved.

1. Introduction 24% of total anthropogenic radiative forcing (Denman et al., 2007).

Nitrous oxide (N2O) and methane (CH4) are two important green-
house gases with global warming potentials 298 and 25 times that of
carbon dioxide (CO2) on a 100-year time scale. The atmospheric con-
centrations of N2O and CH4 are increasing at rates of 0.26% and 0.4%
per year, respectively (Forster et al., 2007). Together they account for
⁎ Corresponding author. Tel.: +86 10 62732006; fax: +86 10 62731016.
E-mail address: juxt@cau.edu.cn (X. Ju).
Moreover, N2O is involved in the destruction of stratospheric ozone
(O3) (Forster et al., 2007). There is an urgent need to reduce N2O and
CH4 emissions from different sources because of climate change.
Agriculture is an important emitter of N2O and CH4 (Robertson et al.,
2000; Reay et al., 2012), accounting for 60% and 50% of global anthropo-
genic N2O and CH4 emissions, respectively (Forster et al., 2007).
Emissions of N2O from croplands are strongly influenced by nitrogen
(N) fertilizer and animal manure applications (Bouwman et al., 2002;
Qin et al., 2012), soil temperature and soil moisture (Hu et al., 2013),

0048-9697/$ – see front matter. Crown Copyright © 2013 Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.scitotenv.2013.11.003
 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124
tillage and straw management practices (Dendooven et al., 2012) and
crop rotations (Huang et al., 2013a). Upland soils act as a sink for atmo-
spheric CH4, depending on soil temperature (Whalen and Reeburgh,
1996), soil moisture (Dunfield, 2007) and ammonium-N fertilization
(Glatzel and Stahr, 2001). Agriculture is rightly recognized as a large
source of greenhouse gas emissions with various opportunities for
mitigation (Burney et al., 2010). Recently van Groenigen et al. (2010)
proposed the concept of yield-scaled N2O emissions (emissions
expressed as g N2O-N per kg aboveground N uptake) and found that
the lowest emissions occur at the optimum range of N application
rates without sacrificing yield. Yield-scaled N2O emissions will be rela-
tively high when insufficient N application rates are accompanied by
low yields (as in many parts of Africa) and very high when excessive
N is applied (as in most intensive agricultural systems in China).
The North China Plain (NCP) is the largest agricultural plain in China,
comprising about 39% of the arable land (Wei et al., 2007). The common
winter wheat–summer maize double-cropping system in this area
accounts for 61% of the total winter wheat yield and 33% of the maize
yield in China (Cui et al., 2008). The double cropping system and espe-
cially current farming practices are under scrutiny because of the over-
use of N fertilizers (Ju et al., 2009), the pollution of groundwater with
nitrate (Ju et al., 2006), and the depletion of groundwater resources
due to irrigation (Foster and Garduno, 2004). Several studies have ex-
amined yields and use of water and N when one season of winter
wheat is replaced by fallow due to the high consumption of groundwa-
ter by winter wheat (e.g., Liu et al., 2008; Meng et al., 2012a). It has been
suggested that an integrated management system comprising a winter
wheat–summer maize–spring maize rotation, with three crop harvests
over two years, has considerable potential to reduce N and water use
and maintain similar net economic return as the conventional winter
wheat–summer maize double-cropping system and this new cropping
system should be considered for economic and sustainable agricultural
development on the North China Plain (Meng et al., 2012a).
Previous studies have indicated that N2O emissions occur mainly
within the one-to-two-week period after N fertilization in conventional
cropping systems in this region (Ju et al., 2011; Bao et al., 2012; Hu et al.,
2013). Emissions are controlled mainly by the chemical form and appli-
cation rate of the fertilizer N (Ju et al., 2011; Bao et al., 2012) but irriga-
tion and precipitation events (Ju et al., 2011) and straw management
practices (Huang et al., 2013b) have also shown some effects on N2O
emissions. The mean N2O emission factors ranged from 0.11 to 0.21%
and 0.45 to 0.59% for winter wheat and summer maize, respectively,
in different fertilizer N treatments (Ju et al., 2011; Hu et al., 2013). The
recommend integrated management practices significantly reduced
N2O emissions by about 30–50% and maintained grain yields relative
to the conventional winter wheat–summer maize system. Management
strategies include optimum fertilizer N and water inputs, use of nitrifi-
cation inhibitors and careful crop management (Ju et al., 2011; Bao
et al., 2012; Hu et al., 2013). The agricultural soils represented a weak
sink for atmospheric CH4 with an annual CH4 uptake of only about
1 kg CH4-C ha− 1 yr− 1 in this double-cropping system and different
fertilizer N rates and other management practices had no significant
effect on CH4 uptake in one cereal rotational cycle (Hu et al., 2013; Shi
et al., 2013). Clearly, N fertilization and irrigation regimes, other crop
management practices, and soil temperature and soil moisture will
change after conversion to alternative cropping systems. However,
no studies have yet been conducted on N2O emissions or CH4 uptake
in alternative cropping systems on the North China Plain and it still un-
clear which cropping system might maintain or increase grain yields
and achieve sustainable use of groundwater at the same time as produc-
ing low N2O emissions. We therefore hypothesized that a winter
wheat–summer maize (or summer soybean)–spring maize rotation
with three crop harvests over two years and a single spring maize sys-
tem with one crop per year will produce lower N2O emissions than
the conventional double-cropping system due to lower fertilizer N and
irrigation water inputs, and the optimized double-cropping system
113
will produce the highest yields due to, inter alia, careful management
of fertilizer N and irrigation, tillage and straw together with other
agronomic aspects of the management package. The fertilizer N and
irrigation regimes will be the main factors controlling N2O emissions,
and straw return will accelerate N2O emissions but increases carbon
storage by the soil.
The objectives of the present study were therefore to evaluate
the performance of an optimized double-cropping system (involving
different fertilizer and irrigation practices but the same crop rotations)
and alternative cropping systems in terms of N2O emissions, CH4
uptake, fertilizer N and irrigation water use, and grain yields over a
two-year period, to further understanding of the factors controlling
N2O emissions and CH4 uptake in different cropping systems, and to
elucidate which cropping systems can produce low N2O emissions
while maintaining or increasing grain yields and the sustainable use
of groundwater.
2. Materials and methods
2.1. Experimental site
A long-term field experiment was set up at Quzhou experimental
station (36.87°N, 115.02°E) in Hebei province in October 2007. The
site is at an altitude of 40 m and has a temperate monsoon climate.
The annual mean temperature was 13.2 °C and the annual mean precip-
itation was 494 mm from 1980 to 2010 (range 213–840 mm), with 68%
of the annual precipitation falling between June and September (Meng
et al., 2012a). The surface 30 cm of the Fluvo-aquic soil has a bulk
density of 1.37 g cm−3, a pH of 7.72 (soil:water ratio 1:2.5), an organic
matter content of 12.6 g kg− 1, total N of 0.7 g kg− 1, Olsen-P of
4.8 mg kg−1 and available K of 72.7 mg kg−1. These values were deter-
mined on one composite soil sample across the whole experimental
field before the start of the long-term field experiment.
2.2. Experiment treatments and field management
A completely randomized design was employed with five treat-
ments and four replicates. Each plot is 1800 (30 × 60) m2. The control
is the conventional winter wheat–summer maize double-cropping sys-
tem (Con.W/M). The other four cropping systems were designed with
optimized crop, water and fertilizer N management, namely (i) an
optimized winter wheat–summer maize double-cropping system
(Opt.W/M), (ii) a winter wheat–summer maize–fallow–spring maize
system (W/M-M), (iii) a winter wheat–summer soybean–fallow–
spring maize system (W/S-M), with three harvests in two years, and
(iv) a single spring maize system with one crop per year (M). Detailed
field management practices in the five different treatments over the
two-year study period are shown in Table 1.
The three cultivars used, all commonly used in this region, were
winter wheat Shijiazhuang 8 sown at the beginning of October, sum-
mer maize Zhengdan 958 and summer soybean Yudou 25 planted in
mid-June. Spring maize cultivar Denghai 605 was planted at the end
of May. In the Con.W/M treatment, wheat (225 kg ha− 1) and maize
(75,000 plants ha−1) were seeded in rows at distances of 15 and 60 cm,
respectively. In the alternative systems, wheat (188 kg ha−1), summer
maize (75,000 plants ha− 1) and spring maize (85,000 plants ha− 1)
were seeded in rows at distances of 15, 60 and 60 cm, respectively,
and soybean had a planting density of 235,000 plants ha− 1 and was
seeded in rows at a distance of 40 cm.
Nitrogen application and the irrigation rate were system-dependent
(Table 1). In Con.W/M the management followed current farming prac-
tice on the North China Plain. Total N input was 300 kg ha−1 yr−1 for
winter wheat and 250 kg ha−1 yr−1 for summer maize, and the ratios
of basal and topdressed applications were 1:1 and 1:1.5, respectively
(Cui et al., 2010; Meng et al., 2012a). The basal fertilizer for winter
wheat was surface broadcast just before tillage and the topdressed
114 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124

Table 1
Detailed field management practices in the five different treatments over the two-year study period.

Crop season Management practices Con.W/Ma Opt.W/M W/M-M W/S-M M

2009–2010 wheat Chemical fertilizer application 300:120:100 100:120:100 140:120:100 140:120:100 –b

rate (N:P2O5:K2O, kg ha−1)
Split N application rate 150:150 40:60 110:30 110:30 –
Seeding rate (kg ha−1) 225 188 188 188 –
Irrigation rate (mm) 180 120 120 120 –
Straw management Removal Return Return Return –
Tillage after harvest No-tillage No-tillage No-tillage No-tillage Rotary tillage (20 cm)
2010 maize or soybean Chemical fertilizer application 250:45:90 185:45:135 185::45:135 45:45:45 150:135:150
rate (N:P2O5:K2O, kg ha−1)
Split N application rate 100:150 45:80:60 45:80:60 45:0 30:30:30:30:30
Split P2O5 application rate 45:0 45:0 45:0 45:0 90:45
Split K2O application rate 45:45 45:90 45:90 45:0 90:60
Plant density (plants ha−1) 75,000 75,000 75,000 235,000 85,000
Irrigation rate (mm) 60 60 60 60 110
Straw management Removal Return Return Return Return
Tillage after harvest Rotary tillage (15 cm) Deep plowing (30 cm) Deep plowing (30 cm) Deep plowing (30 cm) Deep plowing (30 cm)
2010–2011 wheat Chemical fertilizer application 300:120:100 139:120:100 – – –
rate (N:P2O5:K2O, kg ha−1)
Split N application rate 150:159 32:107 – – –
Seeding rate (kg ha−1) 250 188 – – –
Irrigation rate (mm) 240 275 – – –
Straw management Removal Return – – –
Tillage after harvest No-tillage No-tillage Rotary tillage (20 cm) Rotary tillage (20 cm) Rotary tillage (20 cm)
2011 maize Chemical fertilizer application 250:45:90 185:45:135 162:135:150 178:135:150 150:135:150
rate (N:P2O5:K2O, kg ha−1)
Split N application rate 100:150 45:80:60 30:30:30:42:30 30:30:30:58:30 30:30:30:30:30
Split P2O5 application rate 45:0 45:0 45:0 45:0 90:45
Split K2O application rate 45:45 45:90 45:90 45:0 90:60
Plant density (plants ha−1) 75,000 75,000 85,000 85,000 85,000
Irrigation rate (mm) 70 70 60 60 60
Straw management Removal Return Return Return Return
Tillage after harvest Rotary tillage (15 cm) Deep plowing (30 cm) Deep plowing (30 cm) Deep plowing (30 cm) Deep plowing (30 cm)
Total over 2 years Total N application rate 1100 609 487 363 300
(kg N ha−1)
Irrigation rate (mm) 550 525 240 240 170
a
Con.W/M, Opt.W/M, W/M-M, W/S-M and M represent conventional and optimized winter wheat–summer maize in one year, winter wheat–summer maize (or summer soybean)–
spring maize three harvests in two years and single spring maize system in one year.
b
Represents no data in the fallow season.

fertilizer N was surface broadcast at the shooting stage followed by
irrigation. The basal 15-15-15 NPK fertilizer for summer maize was
applied in bands near the plant rows at a depth of 0–5 cm at the same
time as the maize seeding, and the surface broadcast urea topdressings
were applied just before irrigation or precipitation. In Opt.W/M, the N
application rate for wheat was determined according to target N values
(80 kg N ha−1 from planting to shooting and 160 kg N ha− 1 from
shooting to the mature stage) corrected for soil nitrate N content in
the root zone which was measured just before the planned topdressings
as described by Cui et al. (2008). In summer maize, 45 kg N ha−1 was
band applied near the plant rows as a basal application and 80 and
60 kg N ha−1 were band applied in the middle of the rows at a depth
of 5 cm at the six- and ten-leaf stages, respectively. Soybean received
a broadcast basal application of 45 kg N ha−1 only. In W/M-M, W/S-M
and M, the calculation method of N application rate for wheat and
spring maize was the same as for winter wheat in the Opt.W/M treat-
ment. The target N values for spring maize were 80, 130, 130, 140 and
120 kg N ha−1 from planting to the six-leaf stage (0–30 cm), from the
six-leaf to the ten-leaf stage (0–60 cm), from the ten-leaf stage to
anthesis (0–90 cm), from anthesis to the blister stage (0–90 cm), and
from the blister stage to physiological maturity (0–90 cm), respectively
(Chen et al., 2011).
Winter wheat was irrigated two or three times and maize once or
twice with an amount of water ranging from 60 to 100 mm depending
on previous farming practice as shown in Table 1. Irrigation was opti-
mized in the Opt.W/M treatment and the three alternative cropping
systems by testing soil water content before critical growing stages of
wheat, soybean and maize. The standard is to maintain the soil water
content between 45 and 80% plant available water content, which was
determined by measuring field water holding capacity and the corre-
sponding wilting point (Meng et al., 2012a). Irrigation water was ap-
plied via a 15 cm plastic hose and the rate of irrigation was recorded
using a flow meter.
Wheat and maize straw were removed after harvest in Con.W/M
and then the soil was tilled using rotary tillage (15 cm depth). In the
other systems wheat straw was left on the soil surface after being
chopped into pieces of 5–10 cm and the chopped summer and spring
maize and soybean residues were plowed down with rotary tillage
into the upper 30 cm of the soil profile.
Weeding was done manually and chemical crop protection was
applied three times per wheat growing season and three times per
maize growing season. At harvest time the grain and straw yields were
measured in each plot and the dry matter contents of grain and straw
were determined following drying at 65 °C to constant weight.
2.3. Gas sampling and measurements
We measured N2O emissions and CH4 uptake in every plot from
October 2009 to September 2011, representing two rotation cycles
of Con.W/M, Opt.W/M and M, and one rotation cycle of W/M-M and
W/S-M. Emissions were measured manually using the closed static
chamber method as described by Zheng et al. (2008). Each chamber
was composed of a 60 × 50 × 20 cm stainless steel frame with a
water-filled groove on top which sealed the upper chamber
(60 × 50 × 50 cm) to the frame. The upper chamber was covered
with insulating material to minimize the increase in air temperature
in the chamber during summer to less than 3 °C during closure for
45 min. The chambers were equipped with two ventilators at
 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124
opposite angles to ensure complete mixing of air inside the chamber.
The frames were inserted 20 cm into the soil and remained there
throughout a crop growing season. Three rows of wheat and two
maize plants were sown in the chamber. Maize plants were cut
50 cm above the soil surface when their main stalks became too
high at the beginning of August as discussed in the Discussion section
below.
During closure of the chambers a total of four gas samples were
taken at 15-minute intervals using 60 ml plastic syringes through a
three-way stopcock and a Teflon tube connected to the chamber. The
first sample was taken immediately after enclosure. Measurements
of N2O and CH4 emissions were carried out daily for 10 days after fertil-
ization events and 3–5 days after irrigation or precipitation events
(N 10 mm) depending on when the gas fluxes decreased to the normal
level. During the remaining periods emissions were measured twice
per week and once a week when the soil was frozen (Hu et al., 2013).
Measurements were made between 08:30 and 11:00 am.
Gas samples were analyzed for N2O and CH4 using a gas chromato-
graph (Shimadzu GC-14B, Kyoto, Japan) equipped with an electron
capture detector (ECD) and flame ionization detector (FID) within
24 h of sampling. We used high-purity di-nitrogen (N2) (99.999%) as
carrier gas for N2O analysis and 10% CO2 in pure N2 as a buffering gas
for the ECD (Zheng et al., 2008). The detection limits of the emission
measurements were 2 μg N m−2 h−1 for N2O and 3.4 μg C m−2 h−1
for CH4 when the chamber height was 50 cm. The gas samples were cal-
ibrated using known concentrations of compressed air (333 μL m− 3
N2O and 197 μL m−3 CH4) during each measurement cycle.
Following recent reports by Kutzbach et al. (2007) and Kroon et al.
(2008), emissions of N2O were estimated in two ways using Eq. 1
(see also Zheng et al. (2008)) where dc / dt was derived from Eq. 2
(linear increase) or from Eq. 3 (nonlinear increase), depending on the
pattern of change in the concentration in the headspace (Kroon et al.,
2008). The parameter b in Eqs. 2 and 3 represents the initial change
rate in concentration (dc / dt) after chamber enclosure. Uptake of CH4
was estimated using Eqs. 1 and 2 (linear regression method) and set
to zero if the linear determination coefficient R2 b 0.8, with n = 4
(Barton et al., 2008; Hu et al., 2013). Our calculation procedure does
not correct for differences in air pressure, humidity and density with
assumptions that the effects of these factors are not important given
the low mean humidity and flatness of our semi-arid study area at a
near sea-level altitude.
273 M dc
F ¼ k1   H
273 þ T V dt
 
dc
c ¼ a þ bt ¼b
dt
 
2 dc
c ¼ a þ bt þ dt ¼b
dt
where F (μg N2O-N m− 2 h− 1 or μg CH4-C m− 2 h− 1) is the net flux,
k1 is a coefficient (0.001) for dimensional conversion, T (°C) is the
mean air temperature in the chamber, M (28 g N 2 O-N mol− 1 for
N2O, 12 g CH4-C mol−1 for CH4) is the molecular weight of N2 in the
N2O molecule and of C in CH4, V is the mole volume (22.4 L mol−1) at
273 K and 1013 hPa, H (m) is the chamber height, c (μL L−1) is the con-
centration of N2O or CH4 in volume mixing ratio, t (h) is the time of
chamber closure, dc / dt (μL L− 1 h− 1) is the initial rate of change in
the N2O or CH4 concentration in the chamber enclosure (Zheng et al.,
2008), and a, b, and d are parameters derived by fitting the function of
linear or nonlinear changes in concentration over time.
Emissions of N2O and CH4 are presented as the mean values of four
replicated measurements on four different plots. We estimated the
emissions on non-sampling days by linear interpolation between
115
every two adjacent intervals of the measurements (Mosier et al.,
2006) and seasonal and annual totals by summing up the daily fluxes.
2.4. Calculation of direct N 2 O emission factor, net greenhouse gas
emission and intensity of net greenhouse gas emission in terms of N2O
plus CH4 fluxes
The direct N2O emission factors were determined by percentage of
the slope after establishing the linear relationships between seasonal
N2O emission and N application rate for winter and summer crops,
respectively. We calculated the net greenhouse gas (NGHG) emission
in terms of N2O plus CH4 fluxes in CO2 equivalents (CO2-eq) to evaluate
the combined effects of different crop rotations on N2O emissions and
CH4 uptake. The global warming potentials of 1 kg N2O and CH4 being
equivalent to 298 and 25 kg CO2, respectively, at the 100-year time
horizon (Forster et al., 2007) were applied to measure the NGHG in
CO2-eq. We also calculated the intensity of net greenhouse gas intensity
(NGHGI), which was expressed as the magnitude of NGHG to produce
the same crop yield (Mosier et al., 2006).
2.5. Soil sampling and measurements
Samples for soil moisture and soil mineral N (Nmin , NH+ 4 -N +
NO −
3 -N) determinations were taken from each plot every 8 days
during the summer half year and once in 2 weeks during the winter
half year. Following fertilization and irrigation events samples were
taken every other day for 10 days and 5 days, respectively. Soil samples
were taken to a depth of 20 cm using a 3-cm-diameter gage auger. Each
sample was a composite of two subsamples which were taken random-
ly from each replicate after fertilizer N was surface broadcast in the
wheat season and without fertilizer application in the fallow season.
When fertilizer N was band applied in row crops, a concentration gradi-
ent of the fertilizer formed between the application bands and the crop
rows because the fertilizer diffused rapidly following irrigation or pre-
cipitation. In order to represent the average condition, soil samples
were taken between the fertilizer bands and the maize rows. Soil sam-
ples were transported to the laboratory in an ice box and immediately
sieved through a 3-mm mesh for thorough mixing. A 12.0-g
−
sub-sample was used to extract Nmin (NH+ 4 -N + NO 3 -N) with
0.01 mol L− 1 CaCl2 solution (Cui et al., 2008; Ju et al., 2011), and a
20-g soil sample was oven-dried at 105 °C for 24 h to determine soil
water content and to calculate water filled pore space (WFPS) using
Eq. 4. Extracts were stored at − 18 °C prior to analysis of NH+ 4 -N and
ð1Þ NO−
3 -N with an automated ammonium and nitrate analyzer (TRAACS
2000, Bran and Luebbe, Norderstedt, Germany).
ð2Þ Soil water contentð% Þ  Soil bulk density
WFPS ¼  100% ð4Þ
Soil bulk density
1−
2:65
ð3Þ
2.6. Determination of temperature and precipitation
The soil temperature at 5 cm depth and the air temperature in
the headspace of the chambers were measured at the start and end of
gas sampling using digital thermometers (JM624, Tianjin Jinming
Instrument Co., Ltd., China). The mean value represents the temperature
during gas sampling. The average air temperature in the chambers was
used to calculate the fluxes. Soil temperature at 5 cm depth was
measured during the gas sampling period and precipitation data were
obtained from an automatic weather station at the experimental site
as shown in Fig. 1.
2.7. Statistical analysis
The differences between treatments in grain yields and seasonal and
annual N2O and CH4 emissions were determined by t-test for least
116 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124
significant differences (LSD) at P b 0.05. All statistical analysis was per-
formed using the SAS version 9.2 software package (SAS Institute Inc.,
Cary, NC, USA). Relationships between N2O emissions and soil temper-
ature to 5 cm depth and WFPS to 20 cm depth were analyzed using
SigmaPlot 12.0 software (Systat Software Inc., Erkrath, Germany). We
performed stepwise multiple linear regressions to evaluate the effects
−
of soil temperature at 5 cm depth, soil WFPS, soil NH+ 4 -N and NO3 -N
content to 20 cm depth on N2O emissions and CH4 uptake. N2O emis-
sion data were natural logarithm-transformed before fitting the models
to improve the normality of the data. CH4 uptake data were used direct-
ly without transformation. Variables introduced into the models were
significant at P b 0.05 level. Multicollinearity was determined according
to variance inflation factor and influential points were determined
according to Student's residual.
3. Results
3.1. N2O emission
Emissions of N2O were larger during the warm and relatively moist
maize growing seasons than during the relatively cold and dry winter
wheat growing seasons, the mean values of N2O emission ranging
from 34 to 86 μg N2O-N m−2 h−1 in the maize growing seasons of the
different treatments but only 15 to 30 μg N2O-N m−2 h−1 in the winter
wheat growing seasons. Peaks of N2O emission were observed for sever-
al days after N fertilization and irrigation events, especially during
the maize growing seasons (Fig. 2). Relatively large peaks occurred
in treatments Con.W/M (1097 ± 402 μg N2O-N m−2 h− 1 in 2010;
1249 ± 975 μg N2O-N m−2 h−1in 2011, the type of errors is standard
deviation, hereafter is the same) and Opt.W/M (748 ± 347 μg
N2O-N m− 2 h− 1 in 2010; 171 ± 136 μg N2O-N m− 2 h− 1 in 2011)
after sowing of maize in mid-June (Fig. 2a, b) and this may be
attributed to the combined effects of N fertilization, soil tillage,
irrigation, and incorporation of crop residues. An additional peak
occurred in Con.W/M during the summer maize growing season
at the ten-leaf stage (end of July) following top dressings of fertil-
izer N. Two smaller peaks occurred in Opt.W/M during the summer
maize growing season at the six- (mid-July) and ten-leaf (end of July)
stages. Surprisingly, the sum of N2O emissions within ten days follow-
ing two times top dressing of fertilizer N was significantly larger from
treatment Opt.W/M (707 g N2O-N ha− 1 on average) than from
Con.W/M (546 g N2O-N ha−1 on average) during the corresponding
period in 2010 (but not in 2011), while the total N application was
Fig. 1. Soil temperature to 5 cm depth (°C) and precipitation (mm) during the
field experiment.
less (Table 1). Emission peaks of N2O were also noted in treatment
Con.W/M at the shooting stage of winter wheat, and following N fertil-
ization and irrigation in April 2011 the peaks were lower in Opt.W/M
than in Con.W/M. No peaks occurred after N fertilization in the cold
spring of 2010, suggesting large annual variability. Peaks were also rel-
atively small following winter wheat sowing and basal fertilizer appli-
cations in early October of each year.
Emission peaks in spring maize in the alternative cropping systems
were lower than the corresponding period in the double-cropping
system except for a peak that occurred on 3rd July 2011 (Fig. 2c, d, e).
No peaks occurred in the fallow season and only one minor peak during
the soybean growing period in the W/S-M treatment in 2010. Several
smaller peaks (b200 μg N2 O-N m− 2 h− 1 ) occurred in the spring
maize growing period in the W/M-M and W/S-M treatments in
2011 and in the M treatment in 2010 and 2011. In addition, large
peaks of N 2O emission occurred in the W/M-M (447 ± 215 μg
N2O-N m− 2 h− 1), W/S-M (332 ± 119 μg N2O-N m− 2 h− 1) and M
(359 ± 140 μg N2 O-N m− 2 h− 1) treatments after 70 mm rainfall
on 1st July 2011 following top dressing of fertilizer N. However, the
numbers of peaks in the winter wheat and summer maize growing
seasons were similar in treatments W/M-M, W/S-M and Opt.W/M.
3.2. CH4 uptake
Mean uptake of atmospheric CH4 by soil was about 15 μg CH4-C
m− 2 h− 1 with a range of − 6 to 36 μg CH4 -C m− 2 h− 1. Uptake of
CH 4 showed significant seasonal variation which corresponded
with the seasonal temperature (Figs. 1, 3), tending to increase
from March to June and then slowly decreasing to a minimum of
about 5 μg CH4-C m− 2 h− 1 in winter. Uptake of CH4 increased sig-
nificantly within a short period after soil tillage in early October.
Uptake of CH4 was inhibited within a short time after N fertilization
coupled with irrigation or irrigation only (Fig. 3). In addition, CH4
uptake was inhibited by heavy rainfall, e.g., on 1st July 2011 after
70 mm rainfall.
3.3. Dynamics of soil mineral N and moisture
The NH+ 4 -N content in the top 20 cm of the soil profile ranged from 0
to 12 mg N kg−1 dry soil in all treatments and there were significant
difference among different treatments within a few days after N fertili-
zation with mean values of 1 ± 2, 1 ± 1, 0.9 ± 0.9, 0.9 ± 0.8 and
0.9 ± 0.8 mg N kg−1 dry soil in Con.W/M, Opt.W/M, W/M-M, W/S-M
and M throughout the study period (Fig. 4a). The highest NH+ 4 -N con-
tent occurred in treatment Con.W/M after top-dressing of fertilizer N
at the shooting stage of wheat in 2010. NH+ 4 -N content in treatment
M was relatively stable and lower than in the other treatments. In con-
trast, the NO−3 -N content in the top 20 cm of the soil was relatively high
and ranged from 2 to 84 mg N kg−1 dry soil (Fig. 4b), depending on
treatment and rainfall intensity. Compared to Con.W/M the alternative
cropping systems had significantly lower NO− 3 -N contents (P b 0.05)
except for a short period after winter wheat and spring maize were
sown and top-dressing of fertilizer N in summer maize. During the
fallow season, NO− 3 -N contents were around 18 mg N kg
−1
dry soil
−1
and in the alternative systems b 36 mg N kg dry soil after each
crop harvest.
Soil moisture ranged from 20% to 89% WFPS (Fig. 4c). Five treat-
ments maintained the same WFPS from sowing to the shooting stage
of wheat in 2009–2010 since the plots were not irrigated during this
period, and the high WFPS in winter was due to heavy snow on 14th
November 2009. However, treatments Con.W/M and Opt.W/M had
higher WFPS than the three alternative cropping systems during
the winter of 2010 because the former included winter wheat, which
was irrigated.
 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124 117

Fig. 2. Mean N2O emissions from five cropping systems; error bars represent standard deviation of the mean (n = 4); W, M1, M2, S and F represent winter wheat, summer maize, spring
maize, summer soybean and fallow, respectively; short dash, solid and dotted line arrows indicate fertilization with irrigation, fertilization only and irrigation only, respectively.

3.4. Relationships between soil mineral N, moisture and temperature,
and emissions of N2O and CH4
Seasonally cumulative N2O emissions were positively correlated
with fertilizer N application rate (Fig. 5); the relationship explained
48% of the variance during the winter wheat season and 57% of the var-
iance during the maize and soybean growing seasons. Emission factors
were also relatively low, ranging from a mean of 0.17% for winter
wheat to 0.53% for maize and soybean (Fig. 5).
The boundary line approach was used to study the relationships
between N2O emissions and soil surface temperature and WFPS in the
topsoil (Schmide et al., 2000). N2O emissions were clearly low when
the soil temperature to 5 cm depth was below 10 °C, even after N appli-
cation (Fig. 6a). Maximum emissions of 431 μg N2O-N m−2 h−1 on av-
erage occurred at 23.8 °C; at higher temperatures N2O emissions tended
to decrease again. Emissions of N2O were also low when soil moisture in
the top 20 cm of soil was below 30%WFPS. Highest emissions were
measured when WFPS was around 60%. The mean N2O emissions in
W/M-M, W/S-M and M increased more than 8 times compared with
the values before fertilizer application (Fig. 6b).
Stepwise multiple linear regressions indicate that soil temperature,
−
WFPS, and NH+ 4 -N and NO3 -N contents explained 92–95% of the
variance in the logarithm-transformed values of N2 O emissions
(P b 0.05; Table 2). The contribution of factors to explaining variation in
N2O emissions followed the order WFPS N soil temperature N NO− 3 -N
content. The NH+ 4 -N content had no significant effect on N2O emissions.
Soil temperature, WFPS, and NH+ 4 -N content explained up to 35% of
the variance in all CH 4 uptakes of the five treatments (P b 0.05;
Table 2). The contribution to explain variation in CH4 uptake followed
the order soil temperature N WFPS N NO− 3 -N content. Uptake of CH4
118 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124

Fig. 3. Mean atmospheric CH4 uptake by soil in five cropping systems; error bars represent standard deviation of the mean (n = 4); abbreviations for crop phases as in the footnote of
Fig. 2; short dash, solid and dotted line arrows indicate fertilization with irrigation, fertilization only and irrigation only, respectively.

was positively correlated with soil surface temperature and negatively
correlated with soil WFPS.
3.5. Effects of cropping systems on grain yields, N 2 O emissions and
CH4 uptake
Grain yields were system-dependent (Table 3). The mean winter
wheat and summer maize yields were roughly 5 and 8 Mg ha−1 in
Con.W/M. They were significantly higher by 40 and 9%, respectively,
in Opt.W/M than in Con.W/M. Winter wheat yields increased further
to 8 Mg ha−1 in treatments W/M-M and W/S-M. Summer maize yields
were not significantly different in Opt.W/M and W/M-M but the grain
yield was only 44% of the summer maize yield in Con.W/M when sum-
mer soybean was planted in W/S-M. The mean spring maize yields of
W/M-M and W/S-M increased significantly by 40% relative to the
corresponding summer maize in Con.W/M. The mean spring maize
yield of M was 9 Mg ha− 1 with a large range of 8–11 Mg ha− 1. The
low spring maize yield in 2010 was the result of rough dwarf disease
at the seedling stage and stalk rot before harvest. In order to evaluate
the impacts of the cropping systems on grain yields, we calculated
two-year total grain yields as a rotation cycle. Total grain yields were
21% higher in treatment Opt.W/M compared to treatment Con.W/M.
Treatment W/M-M also had slightly higher total grain yields than
Con.W/M but the differences were not statistically significant. However,
grain yields in W/S-M were 11% lower and in treatment M were 30%
lower than in treatment Con.W/M (Table 3).
Seasonal cumulative N2O emissions from Opt.W/M and Con.W/M
treatments were not significantly different except for winter wheat
in 2010–2011. However, seasonal cumulative N2O emissions were
reduced significantly in W/M-M, W/S-M and M treatments compared
 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124 119

−
Fig. 4. Dynamics of NH+
4 -N and NO3 -N contents and water filled pore space (WFPS, %) in the top 20 cm of the soil profile during the period of measurement.

to Con.W/M when summer soybean and spring maize were planted

Fig. 5. Relationships between N fertilization and N2O emissions for wheat and maize plus
soybean during the period of measurement.
or no crop was planted in the fallow season. Seasonal cumulative
N2 O emissions were ranked in the order summer maize N spring
maize N summer soybean, and winter wheat N fallow in the corre-
sponding growing season. Total N 2O emissions in the two-year
study period were 22, 47 and 39% lower in W/M-M, W/S-M and M,
and 5% lower in Opt.W/M than that in Con.W/M (5.57 kg N2O-N ha−1
on average) (Table 3). Seasonal CH4 uptake was similar in the Con.W/M
and Opt.W/M treatments and not significantly different in the
Opt.W/M, W/M-M, W/S-M and M treatments. Total CH4 uptake was
not significantly affected by different cropping systems over the
two-year study period.
We analyzed further the effects of different crops and rotations on
NGHG or NGHGI (Table 4). The effects on N2O emissions dominated
the affections on NGHG in terms of N2O plus CH4. The NGHG was not
significantly different between Con.W/M and Opt.W/M, and was 23,
48 and 40% lower in W/M-M, W/S-M and M compared to Con.W/M.
NGHGI declined by 22, 27, 44 and 14% in Opt.W/M, W/M-M, W/S-M
and M, respectively, compared to Con.W/M.
120 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124
Fig. 6. Relationships between (a) soil temperature to 5 cm depth and N2O emissions, and
(b) water filled pore space (WFPS, %) and N2O emissions.
4. Discussion
4.1. Benefits of alternative cropping systems
The main objective of the alternative cropping systems was to
achieve drastic decreases in fertilizer N and groundwater use and
N2O emissions while maintaining or increasing grain yields. Results
of this two-year study show that optimizing the current winter
wheat–summer maize double-cropping system increased total
grain yields by 21% and decreased NGHGI by 22% and total fertilizer
N use by 45% (Opt.W/M versus Con.W/M; Table 3). The strategies in-
cluded converting rotary tillage (15 cm depth) and straw removal
into deep plow (30 cm depth) incorporation with straw for sowing
wheat, decreasing sowing density of wheat, and optimizing N and
water management in the wheat and maize seasons. It has already
been shown that the fertilizer N application rate can be reduced by
30–60% without sacrificing yield in this double-cropping system by
optimizing N management based on farming practices (Ju et al.,
2009). If this optimized winter wheat–summer maize system was
adopted across the North China Plain the total grain yields would in-
crease by 21%, fertilizer N consumption and NGHGI would decline by
45% and 22%, respectively, and SOC content might increase as a result
of straw return (Huang and Sun, 2006). However, irrigation water
use and NGHG from N2O emissions and CH4 uptake were not reduced
relative to conventional farming practice. It has been estimated that
the annual mean groundwater exploitation should not exceed
150 mm yr− 1 in order to avoid decline in the groundwater table in
this area (Chen et al., 2003; Meng et al., 2012a). The mean irrigation
rate ranged from 263 to 275 mm yr− 1 in Con.W/M and Opt.W/M,
which would result in a groundwater table decline. Compared to
Con.W/M, fertilizer N use (56 to 73%), irrigation (56 to 69%), N2O
emissions (22 to 47%), and NGHG (23 to 48%) were significantly
reduced in the three alternative cropping systems. This likely oc-
curred because the alternative cropping systems lacked one or two
crops during the two-year period. The irrigation rates in the three
alternative cropping systems W/M-M, W/S-M and M were 120, 120
and 85 mm yr− 1 , respectively, suggesting sustainable use of the
groundwater reservoir, but at the expense of a grain yield reduction
(−11 to − 30%) except in W/M-M (Tables 1, 3). Perhaps surprisingly,
the total grain yield in W/M-M was comparable with that in Con.W/M
in two cycles of rotation in this study, mainly due to the relatively
higher yield in each crop of W/M-M treatment by increasing the plant
density of spring maize, deep plowing, straw return, optimum nutrient
and water management and other agronomic factors.
Treatment W/M-M lowered fertilizer N by 56%, groundwater use by
56%, N2O emissions by 22% and NGHG by 48% while maintaining grain
yields and achieving sustainable use of the groundwater reservoir
relative to two cycles of treatment Con.W/M. In addition, the nitrate
accumulation in treatment W/M-M was usually significantly lower
than in Con.W/M. The large decline in soil nitrate and low water in-
puts in treatment W/M-M helped to reduce the high risk of nitrate
contamination of the groundwater (Diez et al., 2000; Ju et al., 2009).
The winter wheat–summer maize–spring maize cropping system with
three harvests over two years seems to offer the best compromise of
high grain yields, low N inputs and balanced use of groundwater with
lower N2O emissions and NGHG than the conventional winter wheat–
summer maize system. Total fertilizer N inputs, groundwater use, N2O
emissions and NGHG would decline by 14, 14, 6 and 12%, respectively,
without sacrificing total grain yield but would greatly protect the envi-
ronment if 25% of the conventional winter wheat–summer maize sys-
tem was converted to the new alternative winter wheat–summer
maize–spring maize cropping system on the North China Plain. The re-
duction in the wheat yield of 5% over two years would have little impact
on the food supply and the diet of the local population.
Recent studies have indicated that spring maize has considerable
potential to reach a yield of 16.5 Mg ha−1 in this area, which could be
achieved with high yield cultivars combined with adequate supply of
nutrients and water and effective control of pests and diseases (Meng
et al., 2012b). However, we achieved only 47–70% of the potential
yield in the current study. The large yield gap between current and
potential spring maize yields is related mainly to inefficient crop man-
agement practices including low efficiency use of light and heat re-
sources due to low plant density, unsuitable crop cultivars and sowing
dates, early harvest with incomplete grain filling, inefficient water and
fertilizer management, poor sowing quality and ineffective plant-
protection practices (Zhang et al., 2011; Meng et al., 2012b). The yield
gap can be greatly narrowed by combining the suite of management
practices in the management package. There have been many recent
successful field trials across China in which the yields of spring maize
achieved 13 Mg ha−1, 86% of the potential yield stimulated by use of
hybrid maize cultivars and nearly double the yield in local farming prac-
tice without any increase in N use (Chen et al., 2011). Therefore, the
winter wheat–summer maize–spring maize cropping system with
three harvests over two years has substantial potential to increase
grain yields by narrowing the yield gap of spring maize compared to
the current winter wheat–summer maize double-cropping system
(Meng et al., 2012a).
4.2. Factors controlling N2O emissions and CH4 uptake
More than 90% of the variance in N2O emissions and up to 35% of the
statistical variance in CH4 uptake were explained by soil temperature,
soil moisture and soil mineral N (Table 2). This high percentage was
 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124 121

Table 2
Multiple linear regression models for N2O emissions and CH4 uptake.

Treatmenta Regression equation (P b 0.0001) Standardized estimated regression Number of observationsc Adjusted R2
coefficientb(P b 0.05)

N2O T W A N
All ln(FN2O) = 0.05T + 0.04W + 0.004N 0.35 0.57 ns 0.08 700 0.93
Con.W/M ln(FN2O) = 0.05T + 0.04W + 0.004N 0.34 0.56 ns 0.11 140 0.92
Opt.W/M ln(FN2O) = 0.07T + 0.03W + 0.01N 0.41 0.48 ns 0.12 140 0.95
W/M-M ln(FN2O) = 0.07T + 0.04W 0.46 0.53 ns ns 140 0.94
W/S-M ln(FN2O) = 0.04T + 0.04W + 0.01N 0.33 0.58 ns 0.10 140 0.94
M ln(FN2O) = 0.04T + 0.05W − 0.004N 0.30 0.74 ns −0.07 140 0.95

CH4
All FCH4 = 13.36 + 0.31T − 0.10W + 0.01N 0.54 −0.23 ns 0.07 700 0.35
Con.W/M FCH4 = 15.26 + 0.40T − 0.14W 0.57 −0.26 ns ns 140 0.44
Opt.W/M FCH4 = 13.00 + 0.40T − 0.10W 0.58 −0.21 ns ns 140 0.42
W/M-M FCH4 = 10.37 + 0.25T − 0.06W + 0.03N 0.53 −0.15 ns 0.26 140 0.33
W/S-M FCH4 = 14.01 + 0.23T − 0.10W 0.54 −0.27 ns ns 140 0.35
M FCH4 = 13.69 + 0.23T − 0.14W + 0.05N 0.48 −0.29 ns 0.26 140 0.39

Units for N2O emissions and CH4 uptake are g N2O-N ha−1 d−1 and g CH4-C ha−1 d−1, for temperature (T), soil water content (W), NH+ −
4 -N (A) and NO3 -N (N) contents are °C, WFPS (%)
and kg N ha−1, respectively.
a
Treatment data used for multiple linear regressions; abbreviations for the treatments are shown in the footnotes of Table 1.
b
ns, not significant at P b 0.05 level.
c
Each observation is the mean of four replicates in each treatment.

irrespective of cropping system, suggesting indeed that temperature,
moisture and soil mineral N were the key variables controlling N2O
emissions. Previous studies have found that the soils on the North
China Plain have high nitrification potential (Ju et al., 2004; Wan et al.,
2009) and the N2O emissions following urea application were mainly
attributed to nitrification (Ju et al., 2011; Bao et al., 2012). N2O emission
fluxes reached about 300–1200 μg N2O-N m−2 h−1 after N fertilization
and were comparable with the magnitude of the N2O fluxes attributed
to nitrification by Bao et al. (2012). However, NH+ 4 -N content was not
significantly correlated with N2O emissions in the present study and
this contrasts with the positive relationship between NH+ 4 -N content
and N2O emissions found by Hu et al. (2013) who collected soil samples
after fertilizer N was surface broadcast coupled with irrigation. The ab-
sence of a significant statistical relationship between NH+ 4 -N content
and N2O emissions might be explained in two ways. First, N2O emissions
were very low even at high NH+ 4 -N levels after fertilizer N application
at some growth stages due to the restriction of emissions by low soil
temperatures or soil moisture. Second, we did not sample on the
fertilization bands. Our sampling procedure may not show accurately
the dynamics of soil NH+ 4 -N content.
The calculated mean N2O emission factors for winter and summer
seasons across treatments were 0.17 and 0.53%, respectively and fall
within the ranges 0.11–0.21% and 0.45–0.59%, respectively, obtained
by Ju et al. (2011) and Hu et al. (2013) in the same study area. These
emission factors are relatively low compared to the default N2O emis-
sion factor of 1% (IPCC, 2006). One limitation of the study was the
cutting of corn stalks during the growing season to allow the gas cham-
bers to fit the frames. This approach may have affected N2O emissions in
two ways, first by increasing the amount of N available to microbes
since the competition from plant N uptake will be affected after cutting
stalks (Smith and Tiedje, 1979), and second by decreasing rhizo-
deposition carbon which could fuel denitrification rates (Hanne and
Soren, 1987). The chamber height can be increased easily to 1.5 m but
the N2O concentration will then be too low for accurate measurement.
We therefore cut the maize when the main stalks were too high to fit
the chambers. This may have affected the C and N transformations

Table 3
Mean grain yields (water content, 15.5%), cumulative N2O emissions and CH4 uptake in each growing season and over two years (mean ± SD, n = 4).

Treatmenta 2009–2010Wb 2010M or S 2010–2011W 2011M Total over 2 years

Mg ha−1
Con.W/M 5.17 ± 1.06bc 8.18 ± 0.48b 4.82 ± 0.38b 7.96 ± 0.31d 26.13 ± 2.07b
Opt.W/M 7.12 ± 0.35a 8.75 ± 0.03a 6.84 ± 0.31a 8.90 ± 0.18c 31.63 ± 0.55a
W/M-M 8.01 ± 0.52a 8.85 ± 0.29a –d 10.76 ± 0.23b 27.63 ± 0.60b
W/S-M 7.99 ± 0.41a 3.62 ± 0.38c – 11.56 ± 0.23a 23.18 ± 0.95c
M – 7.76 ± 0.34b – 10.56 ± 0.66ab 18.31 ± 0.60d

kg N2O-N ha−1
Con.W/M 0.73 ± 0.12a 1.67 ± 0.19a 1.42 ± 0.06a 1.76 ± 0.58a 5.57 ± 0.63a
Opt.W/M 0.79 ± 0.04a 2.01 ± 0.35a 1.20 ± 0.14b 1.28 ± 0.15ab 5.29 ± 0.51a
W/M-M 0.74 ± 0.09a 1.72 ± 0.25a 0.58 ± 0.03c 1.29 ± 0.12ab 4.33 ± 0.17b
W/S-M 0.72 ± 0.05a 0.72 ± 0.09c 0.53 ± 0.07c 1.01 ± 0.17b 2.97 ± 0.14c
M 0.51 ± 0.06b 1.13 ± 0.02b 0.64 ± 0.14c 1.11 ± 0.31b 3.40 ± 0.15c

kg CH4-C ha−1
Con.W/M 0.79 ± 0.06a 0.40 ± 0.04a 0.76 ± 0.06a 0.58 ± 0.11a 2.53 ± 0.21a
Opt.W/M 0.71 ± 0.08ab 0.39 ± 0.04ab 0.70 ± 0.09a 0.53 ± 0.08a 2.33 ± 0.30a
W/M-M 0.67 ± 0.02b 0.34 ± 0.03b 0.69 ± 0.05a 0.55 ± 0.08a 2.24 ± 0.06a
W/S-M 0.68 ± 0.08ab 0.33 ± 0.04b 0.67 ± 0.03a 0.57 ± 0.04a 2.22 ± 0.19a
M 0.62 ± 0.04b 0.39 ± 0.03ab 0.69 ± 0.03a 0.51 ± 0.06a 2.20 ± 0.09a
a
Abbreviations for the treatments as in the footnote of Table 1.
b
W, M and S represent wheat, maize and soybean, respectively.
c
The same letter in the same column denotes no significant difference in grain yield, total emission or uptake in different treatments by LSD (P b 0.05).
d
As in the footnote of Table 1.
122 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124
Table 4
Net greenhouse gas (NGHG) emissions of N2O and CH4, and net greenhouse gas intensity (NGHGI) as a function of cropping system over a two-year period.
Treatmenta N2O (kg N2O-N ha−1) CH4 (kg CH4-C ha−1)
Con.W/M 5.57 ± 0.63ad 2.53 ± 0.21a
Opt.W/M 5.29 ± 0.51a 2.33 ± 0.30a
W/M-M 4.33 ± 0.17b 2.24 ± 0.06a
W/S-M 2.97 ± 0.14c 2.22 ± 0.19a
M 3.40 ± 0.15c 2.20 ± 0.09a
a
Abbreviations for the treatments are shown in the footnotes of Table 1.
b
NGHG (kg CO2-eq ha−1) = N2O-N × 44 / 28 × 298 + CH4-C × 16 / 12 × 25 (Forster et al., 2007).
c
NGHGI (kg CO2-eq Mg−1 grain) = NGHG / yield (Mosier et al., 2006).
d
The same letter in the same column denotes no significant difference in N2O emission, CH4 uptake, NGHG, yield and NGHGI in the different treatments by LSD (P b 0.05).
and thus C and N gases, but our former comparison between this type of
chamber with a two-part chamber (i.e., length × width × height =
60 × 30 × 20 cm, separated vertically into two parts and with a hole
(11 cm diameter) drilled in the center of the top of the chamber
(Liu et al., 2012) to allow the cornstalks to pass through so that only
the maize roots were covered) showed that total seasonal N2O emis-
sions were not significant because the emission peaks were driven
mainly by fertilization events and cutting the stalks was therefore un-
likely to have a marked effect. We did not consider the effect of cutting
stalks on subsequent reduction of transpiration potential, and thus
increasing soil moisture, on N2O emissions, but this effect would be
unlikely to affect the conclusions of our study.
Emission peaks were lower in Opt.W/M than in Con.W/M at the
shooting stage of winter wheat, following N fertilization and irrigation
in April 2011, likely because of lower N fertilization and irrigation appli-
cations. The threshold value for soil temperature at 5 cm depth is 10 °C
(Ma et al., 2010; Hu et al., 2013). N2O emissions were very low below
the threshold value. This may explain why no emission peaks occurred
after the top-dressing of 150 kg N ha−1 and 90 mm irrigation at the
shooting stage of winter wheat in 2010 because the temperature was
only 6.5–10.8 °C during the ten days after fertilization. N2O emissions
were also relatively small following winter wheat sowing and basal
fertilizer applications in early October of each year, likely because of
the relatively low soil moisture content (b 40%WFPS). Both low soil
temperatures after the top-dressing of fertilizer N and irrigation at the
shooting stage of winter wheat and low soil moisture at winter wheat
sowing explain why no significant differences in N2O emissions were
observed between Con.W/M and Opt.W/M during the 2009–2010 win-
ter wheat season, whereas fertilizer N application differed significantly
(Table 1). Favorable soil temperature and moisture conditions for N2O
emissions were roughly 25 °C and 60%WFPS, respectively, which is in
line with results of previous studies (e.g., Schmidt et al., 2000;
Davidson, 1991). Those data are not in accordance with the boundary
line between N2O emissions and soil WFPS after fertilization being
caused by the increase in soil moisture from 28% to 33%WFPS after rain-
fall on 1st August 2010.
Two applications of top dressing fertilizer N at the six- (80 kg N ha−1
and 50 mm rain following fertilization) and ten-leaf (60 kg N ha−1 and
20 mm rain following fertilization) stages of maize in Opt.W/M generat-
ed 707 g N2O-N ha−1 on average within 10 days after each fertilization
in 2010, but only 546 g N2O-N ha−1 on average in the corresponding
period in Con.W/M (150 kg N ha−1 at ten-leaf stage). This difference
accounted for about 50% of the difference in total N2O emissions in sum-
mer maize. Conversely, in 2011 there was no difference in N2O emissions
between the two treatments (Con.W/M, 534 g N2O-N ha−1 on average;
Opt.W/M, 525 g N2O-N ha−1 on average) because large N2O emissions
did not occur after fertilization (80 kg N ha−1) at the six-leaf stage
of maize in Opt.W/M treatment (16th July) due to a decrease in soil
moisture to 29–36%WFPS in the absence of rainfall about three weeks
before and after fertilization (from 3rd to 24th July). We found that
N2O emissions were low when soil moisture was close to 30%WFPS,
NGHGb Yield (Mg ha−1) NGHGIc
(kg CO2-eq ha−1) (kg CO2-eq Mg−1 grain)
2523 ± 297a 26.13 ± 2.07b 97 ± 15a
2399 ± 191a 31.63 ± 0.55a 76 ± 4bc
1950 ± 80b 27.63 ± 0.60b 71 ± 3c
1317 ± 68c 23.18 ± 0.95c 57 ± 10d
1517 ± 148c 18.31 ± 0.60d 83 ± 8b
even after fertilization. These differences are likely related to differences
in rainfall following the top dressings but it should be noted that
the top dressing was split into two portions in Opt.W/M. In addition,
crop residues were returned to the soil in Opt.W/M but removed in
Con.W/M. These crop residues are important for maintaining or
enhancing soil organic carbon stocks and they provide substrates for
heterotrophic microorganisms, including denitrifiers (Baggs et al.,
2003). Depending on the substrate composition, crop residues can
greatly increase N2O emissions from soils (Velthof et al., 2002). It was
found that total N2O emissions during the maize growing season
increased by 58% following the return of wheat straw to the soil in
North China (Liu et al., 2011). Moreover, the seasonal cumulative N2O
emissions from the straw return treatments increased by 28–37% com-
pared to straw removal treatments in two cycles of a winter wheat–
summer maize rotation (Huang et al., 2013b). In our study, crop residue
incorporation in treatment Opt.W/M was combined with changes
in N fertilization strategies (total amount and timing of the split
applications). Mean differences in N2O emissions between Opt.W/M
and Con.W/M were small in a two-year cycle, but it is impossible to
say how much was contributed by crop residue incorporation and
how much by the rainfall effect after changes in the N fertilization
strategies.
In the alternative cropping systems N2O emission peaks were signif-
icantly reduced compared to Con.W/M. This was the combined effect of
the increased number of fertilization events to 5 and the reduced fertil-
izer N rate to 30–50 kg N ha−1 at each fertilization event. The N fertili-
zation requirement and timing were calculated according to N target
values minus the nitrate N content in the root zone before side-
dressing (Cui et al., 2008). Thus, there was no large N surplus in the al-
ternative systems. It has been reported recently that high N surpluses
are conducive to high N2O emissions (e.g., Van Groenigen et al., 2010;
Grassini and Cassman, 2012). Total N applied for spring maize in the
three alternative cropping systems ranged from 150 to 178 kg N ha−1,
which is substantially less than in conventional practice (Con.W/M,
250 kg N ha−1) (Table 1). We found that N2O emissions did not vary
significantly within this range of N fertilization rates. This is in accor-
dance with the report of Snyder et al. (2009) who showed that N2O
emissions did not increase significantly when the fertilization rate
increased from 50 to 180 kg N ha − 1 . However, they reported a
sharp rise in N2O emissions when the N application rate exceeded
180 kg ha− 1.
Cropping system did not have a significant effect on average on CH4
uptake (Table 3). The uptake of CH4 by the soil ranged from 1.1 to 1.3 kg
CH4-C ha−1 yr−1 in the five cropping systems, slightly higher than the
mean uptake of 0.9 kg CH4-C ha−1 yr−1 obtained by Hu et al. (2013).
Seasonal changes in CH4 uptake were related to seasonal changes in
temperature (Figs. 1, 3), Soil temperature explained 48–57% of the
variance in CH4 uptake (P b 0.05). Uptake of CH4 was inhibited by N
fertilization, irrigation and heavy rainfall events within a short period.
One explanation might be that the oxidation of CH4 by methane
monooxygenase competed with the oxidation of ammonium-N after
 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124
N fertilization (King and Schnell, 1994). This indicates that soil
ammonium-N was negatively related to CH4 uptake (Goulding et al.,
1996). Another explanation might be that the high soil moisture
content inhibited CH4 and O2 diffusion into the soil (Dunfield, 2007).
Finally, a third explanation is that low CH4 uptake rates with higher
soil moisture may have been due to some CH4 production by
methanogens (van den Pol-van et al., 1998). Soil WFPS accounted
for 15–29% of the variance in CH4 uptake (P b 0.05). Soil NH+ 4 -N con-
tents were not related with CH4 uptake rates and the absence of a
negative relationship is likely a result of our soil sampling procedure
(as discussed above).
The annual maximum, mean and minimum air temperatures
increased significantly on the North China Plain from 1981 to 2005,
showing a significant warming trend with asymmetric increases in
maximum and minimum temperatures (Liu et al., 2010), which
shows that the increase in average air temperature ranged from 0.4
to 0.7 °C per decade from south to north and annual precipitation
did not increase significantly. It has been predicted that the warming
trend will continue in the coming decades (Ding et al., 2007). We
speculate that these changes in temperature and precipitation may
exert some effects on background emissions (no N fertilization) of
N2O and CH4 but may have little effect on emissions with N fertiliza-
tion because the N rate and irrigation are the main factors driving the
gaseous emissions.
5. Conclusions
NGHG in terms of N2O plus CH4 fluxes in CO2-eq was significantly
affected by the different cropping systems. Optimized winter wheat–
summer maize system can increase grain yields and decrease NGHGI
significantly but cannot reduce N2O emission and achieve sustainable
use of the groundwater reservoir. The alternative winter wheat–
summer maize–spring maize cropping system with three harvests
over two years has considerable potential to reduce N2O emissions,
NGHG and NGHGI and decrease the use of fertilizer N while maintaining
similar grain yields and achieving the sustainable use of groundwater.
Emissions of N2O occurred largely after the application of NH+ 4 -based
fertilizer and were also affected by temperature and soil moisture.
The direct N2O emission factors during the wheat and maize seasons
were both lower than the IPCC default values for upland crops. The
soil acted as a weak sink for atmospheric CH4 and the different cropping
systems had no significant effect on total CH4 uptake. Soil moisture and
temperature explained roughly one-third of the variance in CH4 uptake.
Conflict of interest
All authors have seen the manuscript and approved to submit to
your journal. Thank you very much for your attention and consideration.
Acknowledgments
This work was funded by the Ministry of Science and Technology
(grant number: 2012CB417100), the Ministry of Agriculture (grant
number: 201103039), the National Natural Science Foundation
of China (grant numbers: 41230856, 31172033, 31121062), and the
State Key Laboratory of Atmospheric Physics and Atmospheric Chemistry
(LAPC).
Appendix A. Supplementary data
Supplementary data associated with this article can be found in the
online version, at http://dx.doi.org/10.1016/j.scitotenv.2013.11.003.
These data include Google map of the most important areas described
in this article.
123
References
Baggs EM, Stevenson M, Pihlatie M, Regar A, Cook H, Cadisch G. Nitrous oxide emissions
following application of residues and fertilizer under zero and conventional tillage.
Plant Soil 2003;254:361–70.
Bao QL, Ju XT, Gao B, Qu Z, Christie P, Lu YH. Response of nitrous oxide and corresponding
bacteria to managements in an agricultural soil. Soil Sci Soc Am 2012;76:130–41.
Barton L, Klese R, Gatter D, Butterbach-bahl K, Buck R, Hinz C, et al. Nitrous oxide
emissions from a cropped soil in a semi-arid climate. Glob Chang Biol 2008;14:
177–92.
Bouwman AF, Boumans LJM, Batjes NH. Emissions of N2O and NO from fertilized fields:
summary of available measurement data. Global Biogeochem Cycles 2002;16:
1080–8.
Burney JA, Davis SJ, Lobell DB. Greenhouse gas mitigation by agricultural intensification.
Proc Natl Acad Sci U S A 2010;107:12052–7.
Chen JY, Tang CY, Shen YJ, Sakura Y, Kondoh A, Shimada J. Use of water balance calcula-
tion and tritium to examine the dropdown of groundwater table in the piedmont
of the North China Plain (NCP). Environ Geol 2003;44:564–71.
Chen XP, Cui ZL, Vitousek PM, Cassman KG, Matson PA, Bai JS, et al. Integrated soil–
crop system management for food security. Proc Natl Acad Sci U S A 2011;108:
6399–404.
Cui ZL, Chen XP, Miao YX, Zhang FS, Sun QP, Schroder J, et al. On-farm evaluation of the
improved soil Nmin-based nitrogen management for summer maize in North China
Plain. Agron J 2008;100:517–25.
Cui ZL, Chen XP, Zhang FS. Current nitrogen management status and measures to improve
the intensive wheat–maize system in China. Ambio 2010;39:376–84.
Davidson EA. Fluxes of nitrous oxide and nitric oxide from terrestrial ecosystems.
In: Rogers JE, Whitman WB, editors. Microbial production and consumption of
greenhouse gases: methane, nitrogen oxides, and halomethanes. Washington,
D.C., US: Am Soc Microbiol; 1991. p. 219–35.
Dendooven L, Gutiérrez-Oliva VF, Patiño-Zúñiga L, Ramírez-Villanueva DA, Verhulst N,
Luna-Guido M, et al. Greenhouse gas emissions under conservation agriculture
compared to traditional cultivation of maize in the central highlands of Mexico. Sci
Total Environ 2012;431:237–44.
Denman KL, Brasseur G, Chidthaisong G, Ciais P, Cox PM, Dickinson RE, et al. Couplings
between changes in the climate system and biogeochemistry. In: Solomon S,
Goulding KWT, Willison TW, et al, editors. Climate Change 2007: The Physical Science
Basis. Contribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and
New York, NY, USA: Cambridge University Press; 2007.
Diez JA, Caballero R, Roman R, Tarquis A, Cartagena MC, Vallejo A. Integrated fertilizer and
irrigation management to reduce nitrate leaching in Central Spain. J Environ Qual
2000;29:1539–47.
Ding YH, Ren GY, Shi GY, Gong P, Zheng XH, Zai PM, et al. China's National Assessment
Report on climate change (I): climate change in China and the future trend. Adv
Clim Chang Res 2007;3:1–5.
Dunfield PF. The soil methane sink. In: Reay DS, Hewitt CN, Smith KA, Grace J, editors.
Greenhouse gas sinks. Cambridge, MA, USA: CABI; 2007. p. 152–70.
Forster P, Ramaswamy V, Artaxo P, Berntsen T, Betts R, Fahey DW, et al. Changes in
atmospheric constituents and in radiative forcing. In: Solomon S, Qin D, Manning M,
editors. Climate Change 2007: The Physical Science Basis. Contribution of Working
Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University
Press; 2007.
Foster S, Garduño H. China: towards sustainable groundwater resource use for irrigated
agriculture on the North China Plain. Counterpart organizations: Ministry of Water
Resources (MWR), Institute of Water Resources and Hydropower (IWRH) and
Guantao County Water Resources Bureau (GCWRB), the World Bank water resources
website (www.worldbank.org/gwmate) and the Global Water Partnership website
(www.gwpforum.org). Washington D.C., USA: The World Bank; 20041–16.
Glatzel S, Stahr K. Methane and nitrous oxide exchange in differently fertilized grassland
in southern Germany. Plant Soil 2001;231:21–35.
Goulding KWT, Willison TW, Webster CP, Powlson DS. Methane fluxes in aerobic soils.
Environ Monit Assess 1996;42:175–87.
Grassini P, Cassman KG. High-yield maize with large net energy yield and small global
warming intensity. Proc Natl Acad Sci U S A 2012;109:1074–9.
Hanne B, Soren C. The effect of grass maturing and root decay on N2O production in soil.
Plant Soil 1987;103:269–73.
Hu XK, Su F, Ju XT, Gao B, Oenema O, Christie P, et al. Greenhouse gas emissions from a
wheat–maize double cropping system with different nitrogen fertilization regimes.
Environ Pollut 2013;176:198–207.
Huang Y, Sun WJ. Changes in topsoil organic carbon of croplands in mainland China over
the last two decades. Chin Sci Bull 2006;51:1785–803.
Huang JX, Chen YQ, Sui P, Gao WS. Estimation of net greenhouse gas balance using crop-
and soil-based approaches: two case studies. Sci Total Environ 2013a;456–457:
299–306.
Huang T, Gao B, Christie P, Ju XT. Net global warming potential and greenhouse gas inten-
sity in a double cropping cereal rotation as affected by nitrogen and straw manage-
ment. Biogeosci Discuss 2013b;10:13191–229.
IPCC. Guidelines for national greenhouse gas inventories. Hayama, Japan: IGES; 2006.
Ju XT, Liu XJ, Zhang FS, Roelcke M. Nitrogen fertilization, soil nitrate accumulation, and
policy recommendations in several agricultural regions of China. Ambio 2004;33:
300–5.
Ju XT, Kou CL, Zhang FS, Christie P. Nitrogen balance and groundwater nitrate
contamination: comparison among three intensive cropping systems on the
North China Plain. Environ Pollut 2006;43:117–25.
124 B. Gao et al. / Science of the Total Environment 472 (2014) 112–124
Ju XT, Xing GX, Chen XP, Zhang SL, Zhang LJ, Liu XJ, et al. Reducing environmental risk by
improving N management in intensive Chinese agricultural systems. Proc Natl Acad
Sci U S A 2009;06:3041–6.
Ju XT, Lu X, Gao ZL, Chen XP, Su F, Kogge M, et al. Processes and factors controlling N2O
production in an intensively managed low carbon calcareous soil under sub-humid
monsoon conditions. Environ Pollut 2011;159:1007–16.
King GM, Schnell S. Effect of increasing atmospheric methane concentration on ammoni-
um inhibition of soil methane consumption. Nature 1994;370:282–4.
Kroon PS, Hensen A, van den Bulk WCM, Jongejan PAC, Vermeulen AT. The importance
of reducing the systematic error due to non-linearity in N2O flux measurements by
static chambers. Nutr Cycl Agroecosyst 2008;82:175–86.
Kutzbach L, Schneider J, Sachs T, Giebels M, Nykanen H, Shurpali NJ, et al. CO2 flux
determination by closed-chamber methods can be seriously biased by inappropriate
application of linear regression. Biogeosciences 2007;4:1005–25.
Liu M, Tao HB, Wang P, Lv LH, Zhang YJ, Zhang L. Analysis of different cropping systems
using optimizing water and nitrogen management on yield, water and nitrogen
utilization and economic benefits. J Chin Agric Univ 2008;13:12–8.
Liu Y, Wang EL, Yang XG, Wang J. Contributions of climatic and crop varietal changes to
crop production in the North China Plain, since 1980s. Glob Chang Biol 2010;16:
2287–99.
Liu CY, Wang K, Meng SX, Zheng XH, Zhou ZX, Han SH, et al. Effects of irrigation, fertiliza-
tion and crop straw management on nitrous oxide and nitric oxide emissions from a
wheat–maize rotation field in northern China. Agric Ecosyst Environ 2011;140:
226–33.
Liu CY, Wang K, Zheng XH. Responses of N2O and CH4 fluxes to fertilizer nitrogen addition
rates in an irrigated wheat–maize cropping system in northern China. Biogeosciences
2012;9:839–50.
Ma BL, Wu TY, Tremblay N, Deen W, Morrison MJ, Mclaughlin NB, et al. Nitrous oxide
fluxes from corn fields: on-farm assessment of the amount and timing of nitrogen
fertilizer. Glob Chang Biol 2010;16:156–70.
Meng QF, Sun QP, Chen XP, Cui ZL, Yue SC, Zhang FS, et al. Alternative cropping systems
for sustainable water and nitrogen use in the North China Plain. Agric Ecosyst Environ
2012a;146:93–102.
Meng QF, Hou P, Wu L, Chen X, Cui ZL, Zhang FS. Understanding production potentials
and yield gaps in intensive maize production in China. Field Crop Res 2012b;143:
91–7.
Mosier AR, Halvorson AD, Reule CA, Liu XJ. Net global warming potential and greenhouse
gas intensity in irrigated cropping systems in Northeastern Colorado. J Environ Qual
2006;35:1584–98.
Qin SP, Wang YY, Hu CS, Oenema O, Li XX, Zhang YM, et al. Yield-scaled N2O emissions in a
winter wheat–summer corn double-cropping system. Atmos Environ 2012;55:240–4.
Reay DS, Davidson EA, Smith KA, Smith P, Melillo JM, Dentener F, et al. Global agriculture
and nitrous oxide emissions. Nat Clim Chang 2012;2:410–6.
Robertson GP, Paul EA, Harwood RR. Greenhouse gases in intensive agriculture: contribu-
tions of individual gases to the radiative forcing of the atmosphere. Science 2000;289:
1922–6.
Schmidt U, Thöni H, Kaupenjohann M. Using a boundary line approach to analyze N2O
flux data from agricultural soils. Nutr Cycl Agroecosyst 2000;57:119–29.
Shi YF, Wu WL, Meng FQ, Zhang ZH, Zheng L, Wang DP. Integrated management practices
significantly affect N2O emissions and wheat–maize production at field scale in the
North China Plain. Nutr Cycl Agroecosyst 2013;95:203–18.
Smith MS, Tiedje MJ. The effect of roots on soil denitrification. Soil Sci Soc Am 1979;43:
951–7.
Snyder CS, Bruulsema TW, Jensen TL, Fixen PE. Review of greenhouse gas emissions from
crop production systems and fertilizer management effects. Agric Ecosyst Environ
2009;133:247–66.
van den Pol-van Dasselaar A, van Beusichem ML, Oenema O. Effects of soil moisture con-
tent and temperature on methane uptake by grasslands on sandy soils. Plant Soil
1998;204:213–22.
Van Groenigen JW, Velthof GL, Oenema O, Van Groenigen KJ, Van Kessel C. Towards an
agronomic assessment of N2O emissions: a case study for arable crops. Eur J Soil Sci
2010;61:903–13.
Velthof GL, Kuikman PJ, Oenema O. Nitrous oxide emission from soils amended with crop
residues. Nutr Cycl Agroecosyst 2002;62:249–61.
Wan YJ, Ju XT, Ingwersen J, Schwarz U, Stange F, Zhang FS, et al. Gross nitrogen transfor-
mations and related N2O emissions in an intensively used calcareous soil. Soil Sci Soc
Am 2009;73:102–12.
Wei YP, Chen DL, Li BG. Effectiveness of agricultural water pricing on solving the water
shortage of North China Plain: a case study in Fengqiu county of Henan province.
Resour Sci 2007;29:40–5. [in Chinese].
Whalen SC, Reeburgh WS. Moisture and temperature sensitivity of CH4 oxidation in
boreal soils. Soil Biol Biochem 1996;28:1271–81.
Zhang FS, Cui ZL, Fan MS, Zhang WF, Chen XP, Jiang RF. Integrated soil–crop system
management: reducing environmental risk while increasing crop productivity and
improving nutrient use efficiency in China. J Environ Qual 2011;40:1–7.
Zheng XH, Mei BL, Wang YH, Xie BH, Wang YS, Docng HB, et al. Quantification of N2O
fluxes from soil–plant systems may be biased by the applied gas chromatograph
methodology. Plant Soil 2008;311:211–34.