You are on page 1of 10

Agriculture, Ecosystems and Environment 190 (2014) 94–103

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment

journal homepage:

Nitrous oxide emission factors for urine and dung patches

in a subtropical Brazilian pastureland
André Sordi a , Jeferson Dieckow a,∗ , Cimélio Bayer b , Márcio Amaral Alburquerque a ,
Jonatas Thiago Piva c , Josiléia Acordi Zanatta d , Michely Tomazi e ,
Carla Machado da Rosa b , Anibal de Moraes f
Universidade Federal do Paraná (UFPR), Programa de Pós-Graduação em Ciência do Solo, Departamento de Solos e Engenharia Agrícola, 80035-050
Curitiba, PR, Brazil
Universidade Federal do Rio Grande do Sul (UFRGS), Departamento de Solos, 91501-970 Porto Alegre, RS, Brazil
Universidade Federal de Santa Catarina (UFSC), Campus Curitibanos, 89520-000 Curitibanos, SC, Brazil
Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA), Centro Nacional de Pesquisa em Floresta, 83411-000 Colombo, PR, Brazil
Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA), Agropecuária Oeste, 79804-970 Dourados, MS, Brazil
Universidade Federal do Paraná (UFPR), Departamento de Fitotecnia e Fitosanitarismo, 80035-050 Curitiba, PR, Brazil

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

Article history: Cattle urine and dung (faeces) patches are nitrous oxide (N2 O) sources in pasturelands with impacts
Received 30 January 2013 in the global N2 O budget, but specific information about those emissions are still missing for Brazilian
Received in revised form 6 September 2013 subtropical and tropical regions. We conducted a sequence of 3 field-trials (summer, winter and spring,
Accepted 9 September 2013
90 days each) to evaluate the N2 O emission and emission factor (EF) after the deposition of 3 volumes of
Available online 16 October 2013
cattle urine or 3 weights of dung (½, 1 and 1½ time the mean urination volume or defecation weight of
Friesian cows) on a free-drained Cambisol of a subtropical pastureland of Brazil. The N2 O emission peaks
(3198 ␮g N2 O-N m−2 h−1 after urine in summer was the highest) occurred on average 17 ± 9 days after
Dung weight
application (DAA), both for urine and dung, and dropped to the background levels 41 ± 10 DAA of urine
Emission factor and 49 ± 10 DAA of dung. The highest contents of NH4 + -N in soil (200–250 mg N kg−1 ) occurred one day
Nitrate after urine application and 10–14 days later for dung (100–200 mg N kg−1 ). Nitrate peaks occurred from
Season 23 to 26 DAA in urine patches (∼40–50 mg N kg−1 ) and 19–50 DAA in dung patches (∼40–50 mg N kg−1 ).
Urine volume The N2 O emission peaks for urine coincided with soil NH4 + -N peak in winter but with soil NO3 − -N peak
in spring. For dung, the emission peak seemed to be more associated with soil NO3 − -N than to NH4 + -N,
either in winter or spring (inorganic-N was not assessed in summer). It was not possible to conclude
whether nitrification or denitrification was the dominant process in N2 O production, but it seemed that
both played relevant roles. The EF for urine, averaged across the seasons, diminished with increments in
urine volume, from 0.33% in ½ volume to 0.19% in 1½ volume, possibly because urine percolated deeper
into the soil and proportionally less N remained available for N2 O production in the top layer. The EF
for dung was 0.19%, 0.12% and 0.14% for ½, 1 and 1½ weight, respectively, showing no clear trend with
increment in dung weight. The lowest EFs for urine and dung occurred in winter, possibly because of
lowest temperatures and soil water-filled pore space. The average EF for dung (0.15%) was lower than
that of urine (0.26%), because urea-N of urine is more readily available for the hydrolysis than organic
N forms of dung. This result suggests that these two excreta should be addressed separately in national
greenhouse gases inventories or communications. Our results suggest that the default 2% EF proposed in
IPCC Guidelines for cattle excreta are overestimated for subtropical Brazil.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Nitrous oxide (N2 O) has radiative forcing properties that make

it a potent greenhouse gas with a global warming potential of about
300 times greater than that of carbon dioxide (CO2 ) (IPCC, 2007).
∗ Corresponding author. Tel.: +55 41 33505608; fax: +55 41 3350 5673.
It is also the most important ozone-depleting agent, being the pri-
E-mail addresses:,
mary source of stratospheric NOx that catalytically destroys ozone (J. Dieckow). (Ravishankara et al., 2009). Since pre-industrial time up to 2005,

0167-8809/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103 95

atmospheric N2 O concentration increased by 18% (270–319 ppb) et al., 2005b). In some temperate regions, higher N2 O emissions
and agriculture played a significant role in that increase. Agricul- from excreta patches have been reported to occur in winter (Allen
tural N2 O emission is responsible for 60% of global N2 O emission, et al., 1996; Luo et al., 2008), mainly due to the wet conditions
mainly due to application of fertilizer-N and deposition of animal of the season in those regions. In subtropical Brazil, seasons are
excreta onto pasturelands; and it is projected to increase by 35–60% distinguishable mainly by temperature, a climatic factor shown to
until 2030 (Smith et al., 2007). In Brazil, where about 175 million influence N2 O emissions in pasturelands (Uchida et al., 2011). We
hectares support 205 million cattle (IBGE, 2009), 40% of the esti- hypothesized that because of lower temperatures in winter, the EF
mated national N2 O emission would be derived from urine and for urine and dung would be lower in winter than for spring or
dung (faeces) excreted onto pastureland soils (Brazil, 2010). summer under subtropical conditions.
Much of the nitrogen (N) ingested by cattle as feed (80–95%) Nitrous oxide emission and EF might also be affected by the
is returned to soil as urine or dung (Haynes and Williams, 1993; volume of urine and the weight of dung voided by animals. With
Bolan et al., 2004). Considering that equivalent rates of 800 and higher urination volume, preferential flows may occur through the
2000 kg N ha−1 are applied in urine patches and dung pats, respec- soil and carry the urea-N deeper (Haynes and Williams, 1993) and
tively (Oenema et al., 1997), and that this N can be hydrolyzed to so reduce the availability of N for N2 O production. We hypothesized
ammonium (NH4 + ) and subsequently nitrified to nitrate (NO3 − ), that the higher the urination volume, the more N is leached into
urine and dung patches have been regarded as important localized soil, the less N is available to N2 O production and the lower the EF
N2 O sources in pasturelands, capable of impacting the N2 O global is. For dung, we hypothesized that the larger the dung weight, the
budget (Mosier et al., 1998). When NO3 − undergoes denitrification, bigger the pat, the longer it remains moist, the more favourable the
the N is reduced to the gaseous forms of nitric oxide (NO), N2 O conditions are for N2 O production and the higher the EF is.
or dinitrogen (N2 ). Nitrous oxide may also be produced by nitri- This study aimed at assessing N2 O emissions from cattle urine
fier denitrification, where the intermediate nitrite (NO2 − ) may be and dung patches in a subtropical pastureland of Brazil to evaluate
reduced to NO, N2 O or N2 instead of being oxidized to NO3 − (Wrage (i) how appropriate the default 2% EF is for the region; (ii) how
et al., 2001). excreta type (urine vs. dung) affect the emission and if EF is different
The N2 O emission factor (EF) for a given N source is the per- between them; (iii) how seasons affect emission patterns; and (iv)
centage of the applied N that is emitted as N2 O, and so it allows how the volume of urine and the weight of dung voided by animals
comparison between studies carried out under different agronomic affect emission and EF.
and environmental conditions. According to the IPCC, the default
EF for cattle excreta deposited by grazing animals onto pastures is
2. Materials and methods
2% (no distinction between urine and dung), with an uncertainty
range from 0.7% to 6% (IPCC, 2006). If information about N2 O emis-
2.1. Site description
sion is available for a country, specific EFs can be used in national
inventories or communications. For example, New Zealand uses a
The study was conducted on cattle pastureland comprising
country-specific EF of 1% for urine and considers a separate and
mainly Paspalum paniculatum L., Axonopus compressus (Sw.) P.
lower one for dung (0.25%) (New Zealand, 2012). In Australia, the
Beauv and Pennisetum clandestinum Hochst. ex Chiov., at Can-
country-specific EF for urine is 0.4% and a higher one (0.5%) is
guiri Experimental Farm (dairy sector), near Curitiba-PR, Brazil
for dung (Australia, 2012). However, such specific information is
(25◦ 23 55 S; 49◦ 07 29 W; 912 m altitude). The clayey Cambisol has
lacking for subtropical and tropical conditions of South America.
a 10% slope and contains 439 g kg−1 of clay, pH of 4.9 and 25 g kg−1
Moreover, the default EF of IPCC is based on studies carried out
of organic carbon in the 0–20 cm layer, and is free drained. The
primarily in temperate conditions and may not be appropriate for
humid mesothermic subtropical climate (Cfb, Köppen) has a mean
tropical and subtropical regions. Despite Brazil holding the second
precipitation of 1408 mm per year (around 75 mm in August and
largest cattle herd of the world, it does not have specific N2 O EFs
165 in January) and a mean monthly temperature varying from
for cattle urine and dung yet. In the second and most recent Brazil-
12.2 ◦ C in June to 19.9 in February (Brasil, 1992). Frosts are frequent
ian Communication of greenhouse gases (Brazil, 2010), estimates
in winter.
of N2 O emissions from animal excreta in pastureland were based
on the IPCC 2% default EF (Alves, 2010). However, most of the sub-
tropical Brazilian pasturelands are in well drained soils, where N2 O 2.2. Experimental design and excreta characteristics
production is not so favourable because of better aeration condi-
tions. Thus, we hypothesized that the EFs for urine and dung are We measured N2 O fluxes during 3 separate 90-day trials
lower than the default 2%. which included summer (January 17th–April 17th), winter (June
Despite differences in N contents and forms between dung and 3rd–September 1st) and spring (September 16th–December 15th)
urine, the 2% default EF is applied indistinguishably for both excreta seasons of 2011 (South Hemisphere). Although the trial periods did
types (IPCC, 2006). However, studies have found that N2 O EF for not coincide exactly with regular seasons, we decided to name them
dung is generally lower than that for urine (Flessa et al., 1996; according to the predominant season each trial encompassed.
Yamulki et al., 1998; van der Weerden et al., 2011). Moreover, Three volumes of urine and 3 weights of dung were applied over
van der Weerden et al. (2011) recently recommended the disag- the soil in microplots delimitated by circular metal bases of 32.5-
gregation of urine and dung EFs for national greenhouse gases cm internal diameter (area = 0.083 m2 ) inserted 2.5 cm into the soil.
inventories (in the specific case, for New Zealand), but stated that Bases remained inserted during the entire period of sampling in
further research needs to be undertaken on the subject. We hypoth- each season. A treatment of soil without excreta was used as con-
esized that, under subtropical Brazil conditions, the EF is lower for trol. No fertilizer-N had been used in the area during the last 2 years.
cattle dung than for urine, because dung N is not as readily available When herbage reached 20-cm high, it was cut to 10-cm high and
to N2 O production as urea-N of urine; and this difference should be removed from the area during the course of the trials, simulating
considered in national inventories or communications. grazing.
The N2 O emission patterns after dung or urine application may Urine volumes were equivalent to half (U0.5 ), full (U1.0 ) and one
differ considerably by season, and studies have highlighted the and a half times (U1.5 ) the mean volume of 20 urine samples of
importance of considering this factor in dung and urine N2 O emis- Friesian milking cows (live weight ∼500 kg) fed on diets based on
sion studies (Allen et al., 1996; Yamulki et al., 1998; van Groenigen grazing (adjacent pasture with the same botanical composition of
96 A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103

Table 1
Nitrogen concentration of urine and volumes applied with half (U0.5 ), full (U1.0 ) and one and a half times (U1.5 ) the mean volume per urination (1.97 L obtained in summer
and replicated in winter and spring).

Season Urine N (g L−1 ) N application (g m−2 )

U0.5 U1.0 U1.5

Summer 11.0 ± a
0.4 131 ± 8 261 ± 16 392 ± 25
Winter 7.5 ± 0.1 89 ± 19 178 ± 39 267 ± 58
Spring 9.3 ± 0.5 110 ± 6 221 ± 12 331 ± 18
Mean 9.3 ± 1.8 110 ± 21 220 ± 41 330 ± 63
Values after ±signs are standard deviations.

Table 2
Dry matter (DM), C and N concentrations and C:N ratio of dung and weights applied with half (D0.5 ), full (D1.0 ) and one and a half times (D1.5 ) the mean weight per defecation
(3.37 kg obtained in summer and replicated in winter and spring).

Season DM (g kg−1 ) C (g kg−1 DM) N (g kg−1 DM) C:N C application (g m−2 ) N application (g m−2 )(a)

D0.5 D1.0 D1.5 D0.5 D1.0 D1.5

Summer 127 ± 11a 375 ± 4 20.4 ± 0.1 18 ± 3 967 ± 9 1934 ± 18 2901 ± 27 53 ± 1 105 ± 1 158 ± 2
Winter 142 ± 21 386 ± 1 18.0 ± 0.5 21 ± 6 1113 ± 5 2225 ± 11 3338 ± 17 52 ± 2 104 ± 3 156 ± 5
Spring 118 ± 1 414 ± 4 26.2 ± 2.4 16 ± 2 992 ± 6 1984 ± 11 2975 ± 17 63 ± 1 126 ± 2 188 ± 3
Mean ± S.D. 129 ± 12 392 ± 20 22.0 ± 4.2 18 ± 3 1024 ± 78 2048 ± 233 3071 ± 388 56 ± 6 111 ± 12 167 ± 18
Values after ±signs are standard deviations.

the trials area) and about 5 kg of concentrate (∼20% total protein) comparable to the range of 20–28 g kg−1 reported by Whitehead
per day per cow. Dung weights were equivalent to half (D0.5 ), full (1970). The N application rates with dung varied from 52 to
(D1.0 ) and one and a half times (D1.5 ) of the mean weight of 20 188 g m−2 , or 520 to 1880 kg ha−1 (Table 2). Total C was determined
defecations of the same cows. Urine and dung were collected with in dung by dry combustion and ranged from 375 to 414 g kg−1 while
rigid plastic bucket held manually below the perineum of the cows, the C:N ratio varied from 16:1 to 21:1 (Table 2).
when they were kept stabled during a 2–3 h-period after the morn-
ing milking. The amount of urine in each of the 20 urinations was
measured separately and then composited and homogenized. The 2.3. N2 O flux measurements and emission factors
same was done for dung. The mean volume per urination obtained
in summer was 1.97 L, within the range of 1.6–2.2 L reported by We used the closed static chamber technique (Mosier, 1989)
Haynes and Williams (1993). The mean weight of the 20 fresh dung to collect air samples. Polyurethane chambers of 35-cm height
samples was 3.37 kg, higher than the 1.5–2.7 kg reported by Haynes were covered with aluminium mantle (for insulation). Chambers
and Williams (1993). The mean urine volume and dung weight were deployed on the circular metal bases at the beginning of
obtained in summer were replicated in winter and spring trials, to each sampling event (9:00 am). Rubber belt was used to seal the
avoid the effect of seasons (mainly temperature) being confounded chamber-base. Chambers were equipped with a thermometer, a
with effect of different rates of urine and dung, although in real- 12 V fan (for headspace homogenization), and output valve for sam-
ity urine and dung depositions may vary seasonally due to various ple removal. Air samples were taken with 10 mL polypropylene
conditions. syringes at 0, 15 and 30 min after chamber closure. The duplicate
The experimental design was a randomized complete block, samples of each plot were transferred to a 20-mL polypropylene
with 3 replicates. Each of the 7 treatments included 2 microplots syringe for homogenization and then transferred into 12-mL pre-
(duplicates) per plot (6 microplots in total), for N2 O measurement. evacuated vials (Exetainer, Labco). Samples were analyzed by gas
A third microplot was added in each plot exclusively for soil NH4 + chromatography (Shimadzu 2014) under the following conditions
and NO3 − evaluations, except in the summer trial, when inor- to measure N2 O: injector at 250 ◦ C, column at 70 ◦ C, carrier gas was
ganic N was not assessed because in that season we unfortunately N2 (30 mL min−1 ) and electron capture detector (ECD) at 325 ◦ C.
had technical problems that prevented assessment. The distance The N2 O fluxes (␮g m−2 h−1 ) were calculated considering the
between microplots was about 3.5 m. After ending each season slope of the linear increase of gas concentration during the chamber
trial, microplots were relocated into a place without the influence enclosure, the air temperature and pressure, the chamber volume
of previous excreta. The experiment area was fenced in to avoid and area of the metal bases (Gomes et al., 2009). The cumulative
approaching of animals. Gaseous measurements always began one emission (kg ha−1 ) in each 90-day season was calculated by inte-
day after application (DAA) and extended for 90 DAA, including 11 grating the hourly fluxes over time.
measurements in summer and 13 in winter and spring. The mea- The N2 O-N emission factor for urine or dung, for each of the
surement interval was 2–3 days in the first two weeks that followed 90-day seasons, was calculated according to Eq. (1).
excreta application and was increased up to 40 days, as between
the penultimate and last measurement in summer.
(N2 O − Nemitted ) − (N2 O − Ncontrol )
Urine samples were analyzed one day after collection (4 ◦ C- EF(%) = × 100 (1)
storage meanwhile) for N determination by Kjeldahl digestion Napplied
procedure. Urine N concentration varied from 7.5 to 11.0 g L−1 ,
practically within the range of 8–15 g L−1 reported by Whitehead where EF is the emission factor (percentage of the urine or dung
(1970), and N application rates varied from 89 to 392 g m−2 (or applied-N emitted as N2 O), N2 O-Nemitted is the cumulative N2 O-N
890–3920 kg ha−1 ) (Table 1). emission from urine or dung treated plot during the 90-day period
Nitrogen concentration was determined in dung samples by dry (g m−2 ), N2 O-Ncontrol is the cumulative N2 O-N emission from the
combustion (Vario EL III – Elementary Analysensysteme GmbH, control plot during the 90-day period (g m−2 ), and Napplied is the
Germany) and varied from 18.0 to 26.2 g kg−1 (Table 2) which is urine or dung N application rate (g m−2 ).
A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103 97

100 30
Summer Winter Spring
80 25

Precipitation (mm)

Temperature (oC)

0 0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Precipitation Temperature

Fig. 1. Daily rainfall precipitation (vertical lines) and mean daily temperature (continuous line) during the season trials, which are indicated by the grey-coloured strips.

2.4. Soil and meteorological parameters time, data were analyzed independently for each sampling event,
being subjected to analysis of variance and means were compared
For each air sampling event in summer, soil samples of the by the LSD of Tukey test, at P < 0.10. The cumulative N2 O emission
0–5 cm layer (2.5-cm cores) were randomly collected within each and EFs were also subjected to analysis of variance and Tukey test,
experimental block, out of treatment microplots, for measurement at P < 0.10.
of gravimetric water content (105 ◦ C) and determination of water-
3. Results
filled pore space (WFPS). Evaluations were not done separately for
each treatment in summer. 3.1. Temperature and precipitation
In winter and spring, soil samples were collected from within
the additional microplot that was installed for each treatment The average daily temperature was 20.3 ◦ C in the summer trial, decreased to
13.9 ◦ C in winter and rose to 17.1 ◦ C in spring (Fig. 1). The lowest daily temperature
(dung and urine), for measurement of gravimetric water content
was 5.8 ◦ C (June 27th), 25 DAA in winter’s trial, and the highest was 23.7 ◦ C (February
(105 ◦ C) and NH4 + and NO3 − contents. Soil bulk density in the 5th), 20 DAA in summer’s trial. Summer was the rainiest season (687 mm during the
0–5 cm layer was also measured, by using 56-mm diameter and 90-day trial), compared to winter (555 mm) and spring (365 mm) (Fig. 1). The first
30-mm height cylinders. The WFPS was calculated after consider- major rainfall with at least 20 mm day−1 precipitation occurred 2, 5 and 24 DAA in
ing the gravimetric water content, the bulk density and a particle summer, winter and spring trials, respectively.
density of 2.65 Mg m−3 .
3.2. N2 O emission from urine
For inorganic N analysis, extraction with 2 M KCl was per-
formed with field moist samples (water content correction was Urine markedly increased N2 O emission, which occurred in similar patterns
done after 105 ◦ C drying). For NH4 + -N determination, ultraviolet across the 3 volumes, with background fluxes 1 DAA increasing to maximum peaks
at 12, 8 and 30 DAA, in summer, winter and spring, respectively (Fig. 2a–c). The
absorption spectrometric measurement was carried out at 640 nm,
highest flux recorded was 3198 ␮g N2 O-N m−2 h−1 , after application of U1.0 in sum-
after addition of phenol, sodium nitroprusside and an oxidant mer. Fluxes returned to background 30–50 DAA and remained so until the end of
solution (sodium citrate and sodium hypochlorite) (APHA, 1995). each trial.
Nitrate-N determination was carried out by ultraviolet absorption The cumulative N2 O emission and the EF were affected by volumes of urine and
spectrometry at 210 nm, after addition of H2 SO4 (Heinzmann et al., by seasons. The cumulative emission, averaged across the seasons, increased with
increment in urine volume, but the EF diminished from 0.33% in U0.5 to 0.19% in U1.5
1984; Mulvaney, 1996).
(Table 3). The effect of season on N2 O emission was shown by the highest cumulative
Data of daily rainfall precipitation and mean daily temperature emissions in summer and spring, which were about 3 times of that recorded in
of the 3 seasons were obtained in a meteorological station (Instituto winter (183.5 mg N m−2 ) (Table 3). Summer and spring also had the highest EFs for
Simepar) located 10 km away. urine (0.31% and 0.32%, respectively), compared to winter (0.15%).

3.3. N2 O emission from dung

2.5. Statistical analysis
Similar to urine, application of dung increased N2 O emission. Peaks occurred
from 7 to 26 DAA and the return to background fluxes was 30–72 DAA, depending
For data of urine and dung characteristics the standard deviation on the season (Fig. 2d–f). The highest flux recorded for dung was 1037 ␮g N2 O-
was calculated. For N2 O fluxes, soil WFPS, NH4 + and NO3 − over N m−2 h−1 and occurred in spring, in treatment D1.5 (Fig. 2f).

Table 3
Cumulative emission of N2 O and emission factor (EF) for urine, with half (U0.5 ), full (U1.0 ) and one and a half times (U1.5 ) the mean volume per urination, applied in summer,
winter and spring seasons (90 days each).

Control U0.5 U1.0 U1.5 Mean

Accumulated N2 O emission in 90 days (mg N2 O-N m−2 )

Summer 9.4 Aca 590.4 Ab 817.2 Aa 686.5 Bb 525.9 A
Winter −2.3 Ac 152.4 Cb 332.3 Ca 251.7 Cab 183.5 B
Spring 3.9 Ad 427.4 Bc 610.5 Bb 960.3 Aa 500.5 A

Mean 3.7 c 390.1 b 586.7 a 632.8 a 403.3

N2 O-N EF (%)
Summer 0.45 Aa 0.31 Ab 0.17 Bc 0.31 A
Winter 0.17 Bab 0.19 Ba 0.10 Bb 0.15 B
Spring 0.38 Aa 0.28 ABb 0.29 Ab 0.32 A

Mean 0.33 a 0.26 ab 0.19 b 0.26 ± 0.11 (s.d)

Uppercase letters compare seasons, within the same urine volume; and lowercase letters compare urine volumes within the same season, according to Tukey test
(P < 0.10).
98 A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103

Fig. 2. Nitrous oxide fluxes after urine application in summer (a), winter (b) and spring (c); and after dung application in summer (d), winter (e) and spring (f). Urine volumes
were half (U0.5 ), full (U1.0 ) and one and a half times (U1.5 ) the mean volume per urination. Dung weights were half (D0.5 ), full (D1.0 ) and one and a half times (D1.5 ) the mean
weight per defecation. Note the different y-axes scales for urine and dung. Vertical bars denote the LSD according to Tukey test (P < 0.10).

Increments in dung weight increased the cumulative N2 O emission in summer Comparing the two excreta types, the mean EF for dung (0.15%) was 41% lower
and winter (but not so clearly in spring), so that the mean cumulative emission, than the EF for urine (0.26%), and such trend was repeated across the 3 seasons
in mg N2 O-N m−2 , increased from the control treatment (3.7) to D0.5 (117.3) < D1.0 (Tables 3 and 4).
(143.4) < D1.5 (247.6) (Table 4). The EF for dung showed a certain tendency to
decrease with increment in dung weight but that was not so clear.
Spring was the season in which the highest cumulative N2 O emissions from 3.4. Soil parameters
dung were recorded (more than twice the emissions of summer and winter),
and this trend occurred for every dung weight (Table 4). Spring also had the 3.4.1. Water-filled pore space
highest EF for dung (0.25%), and that was more than twice those reported in The WFPS was very similar after either urine or dung, and also across the 3 levels
summer (0.09%) and winter (0.11%) (Table 4). Compared to urine (Table 3), of excreta (Fig. 3b–e). In summer, the average WFPS was above the critical value of
the trend was different for dung because of the low emission in summer 60% (Linn and Doran, 1984) during the entire season, except for the 3rd sampling
(Table 4). (Fig. 3a); but was below 60% during the first half of the winter season (Fig. 3b and d)

Table 4
Cumulative emission of N2 O and emission factor (EF) for dung, with half (D0.5 ), full (D1.0 ) and one and a half times (D1.5 ) the mean weight per defecation, applied in summer,
winter and spring seasons (90 days each).

Control D0.5 D1.0 D1.5 Mean

Accumulated N2 O emission in 90 days (mg N2 O-N m−2 )

Summer 9.4 Ada 55.0 Bc 106.0 Bb 159.3 Ca 82.4 B
Winter −2.3 Ad 44.5 Bc 103.4 Bb 211.8 Ba 89.4 B
Spring 3.9 Ac 252.4 Ab 220.7 Ab 371.6 Aa 212.2 A

Mean 3.7 c 117.3 b 143.4 b 247.6 a 128.0

N2 O-N EF (%)
Summer 0.09 Ba 0.09 Ba 0.10 Ba 0.09 B
Winter 0.09 Bb 0.10 Bab 0.14 ABa 0.11 B
Spring 0.40 Aa 0.17 Ab 0.19 Ab 0.25 A
Mean 0.19 a 0.12 b 0.14 b 0.15 ± 0.10 (s.d)
Uppercase letters compare seasons, within the same dung weight; and lowercase letters compare dung weights within the same season, according to Tukey test (P < 0.10).
A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103 99

Fig. 3. Water-filled pore space (WFPS) in the experimental area during the summer trial (a), after urine application in winter (b) and spring (c); and after dung application
in winter (d) and spring (e). Urine volumes were half (U0.5 ), full (U1.0 ) and one and a half times (U1.5 ) the mean volume per urination. Dung weights were half (D0.5 ), full
(D1.0 ) and one and a half times (D1.5 ) the mean weight per defecation. Vertical bars denote the LSD according to Tukey test (P < 0.10), but in summer (a), bars denote standard
deviation. Evaluations in summer were not separated for treatments.

and during most of the spring season, except for some sampling events in the first after a rainfall event >20 mm day−1 . This explains the peak delay in
quarter of the trial (Fig. 3c and e). spring (30 DAA) (Fig. 2c), because the first rainfall >20 mm day−1
occurred 24 DAA (Fig. 1). The occurrence of N2 O peaks in urine
3.4.2. Inorganic N patches after heavy rainfall has been already reported (Yamulki
One day after urine application, soil NH4 + -N rose to about 200–250 mg kg−1 and
et al., 1998, 2000). However, in our study no relationship was
fell to background concentrations (<50 mg kg−1 ) within 26–36 DAA (Fig. 4a and b).
Peaks in NO3 − -N occurred later (23–26 DAA) and were lower (about 40–50 mg kg−1 ) found between emission peaks and WFPS within the same season
(Fig. 5a and b). For dung, the peaks of NH4 + -N were lower (100–200 mg kg−1 ) and (Figs. 2a–c and 3a and b). This would suggest that WFPS did not
came later (10–14 DAA) compared to the NH4 + peaks in urine (Fig. 4); and NO3 − closely reflect the rainfall pattern, maybe because of changes in soil
peaks occurred between 19 and 50 DAA and were also lower than NH4 + peaks (Fig. 5). water content caused by evapo-transpiration or drainage between
rainfall events and soil samplings. Similarly, the occurrence of the
4. Discussion emission peak in dung (Fig. 2d–f) was related neither with WFPS
(Fig. 3d and e) nor with rainfall events (Fig. 1).
4.1. Temporal patterns of N2 O fluxes after excreta application The emission peak for urine and dung occurred on average 17 ± 9
DAA and no trend was observed of which excreta peaked first. Emis-
The N2 O fluxes from urine and dung occurred in typical emis- sions dropped to the background levels 41 ± 10 DAA of urine and
sion peaks, as already reported in the literature (Allen et al., 1996; 49 ± 10 DAA of dung, which suggests a slower process in N2 O pro-
Flessa et al., 1996; Yamulki et al., 1998; de Klein et al., 2003; Hoeft duction in dung pats. In the literature, emission peaks after cattle
et al., 2012). For urine, there was a relationship between rainfall urine or dung application occurs within 5–45 DAA, and fall to back-
and the occurrence of the emission peaks, generally at 3–10 days ground levels within 90 days or earlier (Allen et al., 1996; Flessa
100 A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103

Fig. 4. Soil ammonium (NH4 + -N) concentration after urine application in winter (a) and spring (b); and after dung application in winter (c) and spring (d). Urine volumes
were half (U0.5 ), full (U1.0 ) and one and a half times (U1.5 ) the mean volume per urination. Dung weights were half (D0.5 ), full (D1.0 ) and one and a half times (D1.5 ) the mean
weight per defecation. Vertical bars denote the LSD according to Tukey test (P < 0.10). Ammonium was not evaluated in summer.

Fig. 5. Soil nitrate (NO3 − -N) concentration after urine application in winter (a) and spring (b); and after dung application in winter (c) and spring (d). Urine volumes were
half (U0.5 ), full (U1.0 ) and one and a half times (U1.5 ) the mean volume per urination. Dung weights were half (D0.5 ), one time (D1.0 ) and one and a half time (D1.5 ) the mean
weight per defecation. Vertical bars denote the LSD according to Tukey test (P < 0.10). Nitrate was not evaluated in summer.
A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103 101

et al., 1996; de Klein et al., 2003; van Groenigen et al., 2005b; Luo of sheep excreta compared to cattle excreta. These results diverge
et al., 2008; Hoeft et al., 2012). from the view that with even urine distribution the EF should be
The increase in NH4 + -N in soil one day after urine application, up lower, as in the case of the lower EF for sheep (1%) than for cattle
to 200–250 mg N kg−1 (Fig. 4a and b), evidenced the rapid hydroly- urine (2%) (IPCC, 2006).
sis of urea and was within the common range of 100–250 mg N kg−1 For dung, the increment in weight increased the cumulative N2 O
reported to the 0–10 cm layer after urine application (Haynes and emission but tended to reduce EF, although not as clear as in urine
Williams, 1993). The highest NO3 − -N contents occurred at 23–26 (Table 4). This result disagrees with our hypothesis that the larger
DAA, which corresponds to the time required for nitrification to the dung weight, the bigger the pat, the longer it remains moist,
occur, but values (∼40–50 mg kg−1 ) were smaller than those of the more favourable the conditions are for N2 O production and the
NH4 + . The lower NO3 − -N contents may be due to a higher N loss higher the EF is. It could be possible however that the bigger pat
via ammonia (NH3 ) volatilization or a higher leaching of NO3 − - remained in such a moisture condition, associated with high O2
N below the sampled 0–5 cm layer. Several studies have already consuming microbial activity related to decomposition, that anaer-
reported NO3 − contents being several times lower than those of obiosis was very intense. In that condition, instead of the N being
NH4 + after urine application (Allen et al., 1996; Koops et al., 1997; denitrified to N2 O, it could be converted to N2 .
Yamulki et al., 1998; Hoeft et al., 2012).
In dung, the delay and the lower contents of NH4 + -N peaks, com- 4.3. Seasonal changes
pared to those of urine, relates to the lower amount of N applied
per area (Tables 1 and 2) and to the organic-N forms of dung, which The lowest cumulative emission of N2 O and lowest EF for urine
are not as readily available for hydrolysis as the urea-N of urine. in winter relates to the lower temperatures (Fig. 1) and lower
Another possibility is that greater amount of N was still kept inside WFPSs in this season (Fig. 3), a combination that certainly reduced
the dung pat, while in practice all urine-N enters the soil immedi- the microbial processes on N2 O production. The dry period of
ately after urination. Accordingly, NO3 − -N peaks of dung were also May–June (Fig. 1) was possibly the cause of the lower WFPS in the
delayed compared to those of urine. first half of the winter trial. Our findings of lower emission in winter
The emission peak after application of urine in winter (8 DAA, differ from those of temperate regions, where emission is generally
Fig. 2b) coincided with the highest NH4 + -N levels (Fig. 4a), but higher in winter because of wet conditions, which favour anaero-
not with NO3 − -N (Fig. 5a), and these results suggest that nitri- biosis, and low N uptake by plants, which increases N availability
fication could be the main process to N2 O production. But then to N2 O production (Allen et al., 1996; Luo et al., 2008). Seasonal
in summer, the emission peak (30 DAA, Fig. 2c) coincided with changes were also reported by van Groenigen et al. (2005b) but
the highest NO3 − -N availability (Fig. 5b), not with NH4 + (Fig. 4b), they believed that those changes were more related to the fertil-
suggesting that denitrification could be the main process in sum- ization schedule than to the emissions from urine, and stated that
mer. For dung, emission peaks (Fig. 2e and f) seemed to be more in overall no consistent seasonal pattern can be established.
associated with NO3 − than with NH4 + , either in winter or spring For dung, not only winter, but summer also showed low cumu-
(Figs. 4c and d, 5c and d), which also would suggest a greater contri- lative emission and EF (Table 4). This low emission in summer could
bution of denitrification to N2 O production in dung pats. Our results be associated with the high rainfall precipitation, especially in the
did not allow a separation of whether nitrification or denitrifica- first half of the summer trial (Fig. 1). Perhaps under heavy rain-
tion was the dominant process in N2 O production in urine or dung fall the fresh dung pat remained saturated and a highly anaerobic
patches, but indicated that both might have played relevant roles. condition allowed the reduction of N2 O into N2 .
Some studies have suggested that N2 O production after urine or
dung application occurred both due to nitrification and denitrifica- 4.4. Differences between urine and dung
tion (Allen et al., 1996; Flessa et al., 1996; Carter, 2007), while others
suggest that nitrification is the main N2 O producing process (Koops The lower N2 O EF obtained for dung than urine is almost cer-
et al., 1997; Bol et al., 2004), and even others, that denitrification tainly related to the N forms present in each excreta type. More
is the main process (Williams et al., 1998; Yamulki et al., 2000; than 70% of urine N is in urea form (Haynes and Williams, 1993)
van Groenigen et al., 2005a). These inconsistencies indicate that a and thus readily available for hydrolysis to convert it into NH4 + .
combination of several factors, which might change from experi- The sharp increase of soil NH4 + concentration to maximum peaks
ment to experiment, are involved in determining nitrification and just one day after urine application (Fig. 4a and b) supports the
denitrification processes. view of rapid hydrolization of urea-N into NH4 + in urine patch. In
contrast, in dung most of the N is present in organic-NH2 forms of
4.2. Effects of urine volume and dung weight undigested feed and metabolic products of digestion, as well as in
bacterial cells presumably in the form of amino sugars of the cell
Although cumulative N2 O emission, averaged across seasons, wall (Haynes and Williams, 1993). This N in dung is not as readily
increased with increment in urine volume, the average EF for urine available for hydrolysis as urea-N, so that the increase in NH4 + after
showed the opposite trend and diminished from 0.33% in U0.5 to dung application was not as rapid as urea with maximum peaks at
0.19% in U1.5 (Table 3). At the highest volume, urine possibly per- 10–14 days after dung application (Fig. 4a and b). Hence, less N was
colated deeper into the soil and thus proportionally less N remained proportionally available for N2 O production in dung pats and there-
for N2 O production in the topsoil. Haynes and Williams (1993) fore EF was lower. An additional explanation for lower EF for dung
suggested that under larger urine volumes preferential flows may is the higher C availability either in or beneath pats, considering
occur through the soil macropores immediately following urina- that an equivalent rate of 9.7–33.4 Mg C ha−1 was applied (Table 2).
tion, carrying urea-N deeper in the profile (below 15 cm), from Carbon availability to microorganisms leads to greater O2 depletion
where the N2 O produced may not reach the atmosphere because and further reduction of N2 O into N2 (Yamulki et al., 1998; Bolan
of its absorption during the upward path (Neftel et al., 2000). Thus, et al., 2004).
larger but less frequent urinations might be better to reduce N2 O Several studies have already reported lower EFs for dung than
emission in grasslands than smaller and more frequent urinations. for urine (Flessa et al., 1996; Yamulki et al., 1998; van der Weerden
In practice this would be difficult to manage. Hoeft et al. (2012) et al., 2011). The fact that EF for dung was half of that of urine
showed that the EF for cattle urine (0.39%) was smaller than for suggests that these two excreta should be addressed separately in
sheep urine (0.48%) and attributed this to the more even spread greenhouse gases inventories or communications, which is in line
102 A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103

with the proposition of van der Weerden et al. (2011). Currently, 10/0054-7) and scholarships to J. Dieckow, C. Bayer, M. Tomazi, C.
urine and dung are accounted together under the same default EF Rosa, and A. Moraes. To CAPES (Coordenação de Aperfeiçoamento
of 2%, according to the IPPC Guidelines (IPCC, 2006). de Pessoal de Nível Superior), for scholarships to A. Sordi, J. Piva
and M. Alburquerque. To M. Pergher, A. Ivatiuk, E. Nogas-Neto (in
4.5. Emission factors for dung and urine memoriam), B. Andrade-Silva, D. Rocha and A. Maoski for the tech-
nical support in field and laboratory.
Compared to the range of 0.1–0.7% of EFs for dung presented
by Oenema et al. (1997) after a global literature review, the aver- References
age 0.15% found in the current study falls in the bottom part of
that range. However, in the study of Yamulki et al. (1998), EFs for Allen, A.G., Jarvis, S.C., Headon, D.M., 1996. Nitrous oxide emissions from soils due
dung as low as 0.04% or 0.07% were reported for dry conditions of to inputs of nitrogen from excreta return by livestock on grazed grassland in the
UK. Soil Biol. Biochem. 28, 597–607.
summer in England, while the average EF of 7 trials across the year Alves, B.J.R., 2010. Emissões de óxido nitroso de solos agrícolas e de manejo de deje-
was 0.19%. Recently, van der Weerden et al. (2011) and Hoeft et al. tos – Série Relatórios de Referência (Segundo inventário brasileiro de emissões
(2012) reported EFs for cattle dung of 0.04% and 0.05%, respectively, antrópicas de gases de efeito estufa). MCT, Brasilia.
APHA, 1995. Phenate method. In: Eaton, A.D., Clesceri, L.S., Greenberg, A.E. (Eds.),
which is even lower than what we found. Based on those and our Standard Methods for the Examination of Water and Wastewater. American
findings, indications are that the default 2% EF for cattle dung may Public Health Association, Washington, DC, pp. 4.80–84.81.
be overestimated. Australia, 2012. Australian National Greenhouse Accounts: National Inven-
tory Report 2012. Department of Climate Change and Energy Efficiency,
The observed EF of 0.26% for urine in this subtropical clayey Australia.
Cambisol was at the bottom of the global range of 0.1–3.8% pre- Bol, R., Petersen, S.O., Christofides, C., Dittert, K., Hansens, M.N., 2004. Short-term
sented by Oenema et al. (1997), of 0.3–2.5% reported by de Klein N2 O, CO2 , NH3 fluxes, and N/C transfers in a Danish grass-clover pasture
after simulated urine deposition in autumn. J. Plant Nutr. Soil Sci. 167,
et al. (2003) over 5 pastoral soils in New Zealand, and of the 568–576.
default 2% proposed by IPCC (IPCC, 2006). In the study of de Klein Bolan, N.S., Saggar, S., Luo, J.F., Bhandral, R., Singh, J., 2004. Gaseous emissions
et al. (2003), the lowest EFs were reported for well drained soils of nitrogen from grazed pastures: processes, measurements and modelling,
environmental implications, and mitigation. In: Sparks, D.L. (Ed.), Advances in
while the highest for moderately or poorly drained soils. How-
Agronomy, 84. Elsevier Academic Press Inc., San Diego, pp. 37–120.
ever, Vermoesen et al. (1997) found EF for urine as low as 0.13% Brasil, 1992. Normais climatológicas (1961–1990). Departamento Nacional de Mete-
which was attributed to low temperatures and low soil water con- orologia, Brasília.
tent (Belgium), while Yamulki et al. (1998) and Luo et al. (2008) Brazil, 2010. Second National Communication of Brazil to the United Nations Frame-
work Convention on Climate Change. Ministry of Science and Technology,
reported even lower values, 0.02%, in dry conditions of summer Brasilia.
(England). In the recent studies of van der Weerden et al. (2011) Carter, M.S., 2007. Contribution of nitrification and denitrification to N2 O emissions
and Hoeft et al. (2012), the average EFs of 0.29% and 0.39% were from urine patches. Soil Biol. Biochem. 39, 2091–2102.
de Klein, C.A.M., Barton, L., Sherlock, R.R., Li, Z., Littlejohn, R.P., 2003. Estimating a
reported, respectively. Thus, our result of lower EF for urine is in nitrous oxide emission factor for animal urine from some New Zealand pastoral
line with those of many previous studies. soils. Aust. J. Soil Res. 41, 381–399.
One explanation for the low EFs for urine found in the Cambisol Flessa, H., Dorsch, P., Beese, F., Konig, H., Bouwman, A.F., 1996. Influence of cattle
wastes on nitrous oxide and methane fluxes in pasture land. J. Environ. Qual. 25,
of our study is the free-drained condition, that allows a deeper flow 1366–1370.
of urine into the soil as well as a higher O2 diffusion, combined with Gomes, J., Bayer, C., Costa, F.D., Piccolo, M.D., Zanatta, J.A., Vieira, F.C.B., Six, J., 2009.
subtropical mild temperatures (mean of 12.2 ◦ C in June and 19.9 in Soil nitrous oxide emissions in long-term cover crops-based rotations under
subtropical climate. Soil Till. Res. 106, 36–44.
February). Thus, the adoption of a 2% EF would be inappropriate Haynes, R.J., Williams, P.H., 1993. Nutrient cycling and soil fertility in the grazed
for this region. Considering that most of pasturelands in Brazil are pasture ecosystem. Adv. Agron. 49, 119–199.
over Ferralsols (Latossolos), which by nature are well-drained soils, Heinzmann, F.X., Miyazawa, M., Pavan, M.A., 1984. Determinação de nitrato em
extratos de solos ácidos por espectrofotometria de absorção ultravioleta. Rev.
we can speculate that the 2% EF would not be appropriate for those
Bras. Ci. Solo 8, 159–163.
soils, but more studies are needed to confirm that and to provide Hoeft, I., Steude, K., Wrage, N., Veldkamp, E., 2012. Response of nitrogen oxide emis-
additional data towards the establishment of country-specific EFs. sions to grazer species and plant species composition in temperate agricultural
grassland. Agric. Ecosyst. Environ. 151, 34–43.
IBGE, 2009. Censo Agropecuário 2006: Brasil, Grandes Regiões e Unidades da
5. Conclusions Federação. IBGE, Rio de Janeiro.
IPCC, 2006. In: Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., Ngara, T., Tanabe, K.
(Eds.), 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Prepared
The N2 O emission from dung and urine patches in subtropical by the National Greenhouse Gas Inventories Programme. IGES, Japan [Vol. 4 –
Brazilian pastureland occurs on average within 7 weeks after depo- Chapter 11: de Klein et al. N2 O emissions from managed soils, and CO2 emissions
sition and is seasonally affected, being lower in winter. The volume from lime and urea application].
IPCC, 2007. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B.,
of urine voided per urination affects the N2 O EF, being lower with Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis.
higher urine volumes, possibly because more N is lost by leaching Contribution of Working Group I to the Fourth Assessment Report of the Inter-
and less is available to N2 O production; but the effect of weight of governmental Panel on Climate Change. Cambridge University Press, Cambridge,
United Kingdom and New York, NY, USA [Technical Summary].
dung on EF was not clear. Results also suggest that dung and urine Koops, J.G., vanBeusichem, M.L., Oenema, O., 1997. Nitrous oxide production, its
should be accounted separately in national greenhouse gas inven- source and distribution in urine patches on grassland on peat soil. Plant Soil
tories or communications, because dung has a lower EF (0.15%) than 191, 57–65.
Linn, D.M., Doran, J.W., 1984. Effect of water-filled pore space on carbon dioxide
urine (0.26%), and this is related to the form N is present in each of and nitrous oxide production in tilled and nontilled soils. Soil Sci. Soc. Am. J. 48,
these excreta types. Finally, the default EF of 2% proposed in IPCC 1267–1272.
Guidelines for cattle excreta seems to be overestimated for sub- Luo, J., Lindsey, S.B., Ledgard, S.F., 2008. Nitrous oxide emissions from animal urine
application on a New Zealand pasture. Biol. Fertil. Soils 44, 463–470.
tropical Brazilian conditions and more studies need to be carried Mosier, A.R., 1989. Chamber and isotope techniques. In: Andreae, M.O., Schimel,
out towards the establishment of country-specific EFs for animal D.S. (Eds.), Exchange of Traces Gases Between Terrestrial Ecosystems and the
dung and urine. Atmosphere: Report of the Dahlem Workshop. Wiley, Berlin, pp. 175–187.
Mosier, A., Kroeze, C., Nevison, C., Oenema, O., Seitzinger, S., van Cleemput, O., 1998.
Closing the global N2 O budget: nitrous oxide emissions through the agricultural
Acknowledgements nitrogen cycle. Nutr. Cycl. Agroecosyst. 52, 225–248.
Mulvaney, R.L., 1996. Nitrogen – inorganic forms. In: Sparks, D.L., Page, A.L., Helmke,
P.A., Loeppert, R.H., Soltanpour, P.N., Tabatabai, M.A., Johnston, C.T., Sumner, M.E.
To CNPq (Conselho Nacional de Desenvolvimento Centífico e (Eds.), Methods of Soil Analysis: Part 3 Chemical Methods. Soil Science Society
Tecnológico), for financial support (grants 476613/2009-4 and of America, Madison, pp. 1123–1184.
A. Sordi et al. / Agriculture, Ecosystems and Environment 190 (2014) 94–103 103

Neftel, A., Blatter, A., Schmid, M., Lehmann, B., Tarakanov, S.V., 2000. An experimental van Groenigen, J.W., Kuikman, P.J., de Groot, W.J.M., Velthof, G.L., 2005a. Nitrous
determination of the scale length of N2 O in the soil of a grassland. J. Geophys. oxide emission from urine-treated soil as influenced by urine composition and
Res. Atmos. 105, 12095–12103. soil physical conditions. Soil Biol. Biochem. 37, 463–473.
New Zealand, 2012. New Zealand’s Greenhouse Gas Inventory 1990–2010. Ministry van Groenigen, J.W., Velthof, G.L., van der Bolt, F.J.E., Vos, A., Kuikman, P.J., 2005b.
for the Environment, Wellington. Seasonal variation in N2 O emissions from urine patches: effects of urine con-
Oenema, O., Velthof, G.L., Yamulki, S., Jarvis, S.C., 1997. Nitrous oxide emissions from centration, soil compaction and dung. Plant Soil 273, 15–27.
grazed grassland. Soil Use Manage. 13, 288–295. Vermoesen, A., van Cleemput, O., Hofman, G., 1997. Contribution of urine patches to
Ravishankara, A.R., Daniel, J.S., Portmann, R.W., 2009. Nitrous Oxide (N2 O): the the emission of nitrous oxide. In: Jarvis, S.C., Pain, B.F. (Eds.), Gaseous Nitrogen
dominant ozone depleting substance emitted in the 21st Century. Science 326, Emissions from Grassland. CAB International, Oxon, pp. 189–194.
123–125. Whitehead, D.C., 1970. The Role of Nitrogen in Grassland Productivity: A Review
Smith, P., Martino, D., Cai, Z., Gwary, D., Janzen, H., Kumar, P., McCarl, B., Ogle, S., of Information From Temperate Regions. Commonwealth Agricultural Bureaux,
O’Mara, F., Rice, C., Scholes, B., Sirotenko, O., 2007. Agriculture. In: Metz, B., England.
Davidson, O.R., Bosch, P.R., Dave, R., Meyer, L.A. (Eds.), Climate Change 2007: Williams, P.H., Jarvis, S.C., Dixon, E., 1998. Emission of nitric oxide and nitrous
Mitigation. Contribution of Working Group III to the Fourth Assessment Report oxide from soil under field and laboratory conditions. Soil Biol. Biochem. 30,
of the Intergovernmental Panel on Climate Change. Cambridge University Press, 1885–1893.
Cambridge, pp. 497–540. Wrage, N., Velthof, G.L., van Beusichem, M.L., Oenema, O., 2001. Role of nitrifier deni-
Uchida, Y., Clough, T.J., Kelliher, F.M., Hunt, J.E., Sherlock, R.R., 2011. Effects of bovine trification in the production of nitrous oxide. Soil Biol. Biochem. 33, 1723–1732.
urine, plants and temperature on N2 O and CO2 emissions from a sub-tropical Yamulki, S., Jarvis, S.C., Owen, P., 1998. Nitrous oxide emissions from excreta applied
soil. Plant Soil 345, 171–186. in a simulated grazing pattern. Soil Biol. Biochem. 30, 491–500.
van der Weerden, T.J., Luo, J.F., de Klein, C.A.M., Hoogendoorn, C.J., Littlejohn, R.P., Yamulki, S., Wolf, I., Bol, R., Grant, B., Brumme, R., Veldkamp, E., Jarvis, S.C., 2000.
Rys, G.J., 2011. Disaggregating nitrous oxide emission factors for ruminant urine Effects of dung and urine amendments on the isotopic content of N2 O released
and dung deposited onto pastoral soils. Agric. Ecosyst. Environ. 141, 426–436. from grasslands. Rapid Commun. Mass Spectrom. 14, 1356–1360.