See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/225445835

Quantifying the Reduction of Greenhouse Gas

Emissions as a Resultof Composting Dairy and
Beef Cattle Manure

Article in Nutrient Cycling in Agroecosystems · June 2005

DOI: 10.1007/s10705-005-1268-5

CITATIONS READS

65 86

3 authors, including:

Elizabeth Pattey R. Desjardins

Agriculture and Agri-Food Canada Ottawa Research and Development Center,…
154 PUBLICATIONS 3,813 CITATIONS 348 PUBLICATIONS 6,952 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Measuring and assessing Canadian rangeland and other agricultural best management
practices with the enhanced whole-farm model Holos View project

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: R. Desjardins
letting you access and read them immediately. Retrieved on: 22 September 2016
Nutrient Cycling in Agroecosystems (2005) 72:173–187
Springer 2005
DOI 10.1007/s10705-005-1268-5

Quantifying the reduction of greenhouse gas emissions as a result

of composting dairy and beef cattle manure

E. Pattey1,*, M.K. Trzcinski2 and R.L. Desjardins1

1
Research Branch, Agriculture and Agri-Food Canada, 960 Carling Ave., Ottawa, Ontario, K1A 0C6,
Canada; 2Department of Biology, Dalhousie University, Halifax, NS, B3H 4J1, Canada; *Author for
correspondence (email: [email protected]; phone: +613-759-1523; fax: +613-759-1724)

Received 17 December 2003; accepted in revised form 14 January 2005

Key words: Beef cattle, Dairy cattle, Emission factors, Greenhouse gas emissions, Manure, Methane,
Nitrous oxide

Abstract

Greenhouse gas emissions from the agricultural sector can be reduced through implementation of improved
management practices. For example, the choice of manure storage method should be based on environ-
mental decision criteria, as well as production capacity. In this study, greenhouse gas emissions from three
methods of storing dairy and beef cattle manure were compared during the summer period. The emissions
of CH4, N2O and CO2 from manure stored as slurry, stockpile, and compost were measured using a flow-
through closed chamber. The largest combined N2O–CH4 emissions in CO2 equivalent were observed from
the slurry storage, followed by the stockpile and lastly the passively aerated compost. This ranking was
governed by CH4 emissions in relation to the degree of aerobic conditions within the manure. The radiative
forcing in CO2 equivalent from the stockpiled manure was 1.46 times higher than from the compost for
both types of cattle manure. It was almost twice as high from the dairy cattle manure slurry and four to
seven times higher from the beef cattle manure slurry than from the compost. The potential reduction of
GHG was estimated, by extrapolating the results of the study to all of Canada. By composting all the cattle
manure stored as slurry and stockpile, a reduction of 0.70 Tg CO2-eq year
1 would be achieved. Similarly,
by collecting and burning CH4 emissions from existing slurry facilities, a reduction of 0.76 Tg CO2-
eq year
1 would be achieved. New CH4 emission factors were estimated based on these results and
incorporated into the IPCC methodology. For North-America under cool conditions, the CH4 emission
factors would be 45 kg CH4 hd
1 year
1 for dairy cattle manure rather than 36 kg CH4 hd
1 year
1, and
3 kg CH4 hd
1 year
1 for beef cattle manure rather than 1 kg CH4 hd
1 year
1.

Introduction The primary sources of these emissions are enteric

fermentation, chemical fertilizers, and manure
On-farm agricultural practices account for almost management. Based on direct effects, the 20- and
10% of the anthropogenic greenhouse gas (GHG) 100-year global warming potentials for CH4 are 62
emissions in Canada (Desjardins et al. 2001). Over and 23 times, and for N2O 275 and 296 times that
half of the total N2O emissions, and over a third of of CO2 on a mass basis, respectively (IPCC 2001).
the total CH4 emissions come from agriculture. The major reasons why changes in agricultural
174

practices have the potential to significantly impact manure was produced by a small herd equivalent
climate change are the high emission rates and to 20 animal units. The dairy cow manure con-
high warming potential of CH4, and N2O. Several tained about 10% bedding material, consisting of
authors have suggested that altering the storage dry straw and wood shavings, while the beef cattle
and handling of manure could reduce GHG manure contained about 50%. The manure was
emissions, as well as the overall contribution of mixed with a spreader to produce a homogenous
agriculture to global warming (Sommer et al. 2000; material before transferring it to experimental
Amon et al. 2001; Hao et al. 2001; Kulling et al. storage bins.
2003).
Between 5 to 30% of global CH4 emissions are
derived from livestock manure (Svenson et al. Manure storage experiment
1991; Sommer et al. 2000; Kulling et al. 2002). In
Canada, manure accounts for approximately 20% The GHG emissions from three different methods
of the total CH4 emissions from agriculture of manure storage – slurry (S), stockpile (P), and
(Desjardins and Riznek 2000). CH4 emissions compost (C) were quantified. Overall, the slurry
from livestock manure are affected by the type of was mostly anaerobic (except the crust that formed
livestock (pig, dairy cattle, etc.), storage conditions at the surface), the compost was mainly aerobic,
(slurry, solid, pasture), and temperature (Husted and the stockpile was in between. The emissions
1994; Amon et al. 2001; Su et al. 2003). Because from three replicates of each storage type were
CH4 is produced by anaerobic decomposition of monitored for 14 weeks for the dairy cattle man-
organic matter (Gaudy and Gaudy 1988) it is ure and for 11 weeks for the beef cattle manure.
expected that conditions that promote aerobic Each manure storage bin, made of a wood frame,
decomposition will result in less CH4 emissions. was 1.00 m deep, 1.22 m wide and 2.13 m long
Observations of N2O emissions should also be (internal dimensions). In order to minimize the
included in studies of livestock manure because changes in the water content of the manure and
manure accounts for approximately 18% of the simulate the environment of the core of the stored
total N2O emissions. Sound recommendations for manure while keeping the bin size to a minimum,
manure storage should take into account the different film types were installed around the sides
tradeoff between CH4 and N2O, which depends on of the bins. For the slurry treatment, a heavy black
the level of oxygen in the manure during the plastic was used to retain water, and the walls of
storage. The objectives of this study are the fol- the bins were reinforced. The slurry treatment was
lowing: (1) investigate the effect of manure storage designed to mimic the environmental conditions of
types on the emission of GHGs, (2) compare GHG above-ground tanks. For the stockpiled manure
emissions from dairy cattle manure to emissions treatment, white polyethylene sheeting backed by
from beef cattle manure, (3) determine the envi- chicken wire was used to retain moisture in the
ronmental- and chemical variables that account manure. For the compost treatment, house wrap, a
for most of the variations in GHG emissions, and porous synthetic material used in building houses,
(4) provide management recommendations for was used to maintain air exchange while mini-
reducing GHG emissions. mizing evaporation. In addition, two polyvinyl
chloride pipes with two rows of perforations were
placed on the bottom of the bins to increase aer-
Materials and methods ation and promote the composting of manure (as
in the Passively aerated Windrow System of
Study site Agriculture and Agri-Food Canada; Mathur 1991,
1992; Hayes et al. 1993). In the passively aerated
The measurements were carried out at the Central system, exothermic reactions create a chimney
Experimental Farm of Agriculture and Agri-Food effect within the windrow, which in turn causes
Canada, in Ottawa Ontario (latitude 4522¢ N, oxygen to move into the pipes and through the
longitude 7543¢ W, 79 m above sea level) during composting material. To minimize ammonia vol-
two consecutive summers; first over dairy cattle atilization, the manure being composted had a
manure and then over beef cattle manure. The 15 cm layer of pre-composted manure added to
 175

the bottom and another one at the top of the bin. sampling was completed as described for the dairy
To limit rain and the drying of stored manure, a cattle experiment. Since the slurry treatment was
roof made of white plastic was installed over each still emitting high levels of CH4 on August 15, the
bin. Each bin was aerated naturally by the wind. sampling for this treatment continued until
October 25 (CD 298).
Emissions of CH4, N2O and CO2 from the
GHG emission measurements storage bins were measured using a flow-through
closed chamber (Figure 1). The enclosure was
Because manure carbon originates from auto- placed over the to be measured bin. Four fans were
trophic fixation which is released as CO2 to the used to ensure good air mixing within the enclo-
atmosphere over a relatively short period of time, sure. These were positioned so that air was forced
manure CO2 emissions are not considered to away from the surface of the manure. In order to
contribute to global warming and are not included measure CH4 and N2O concentrations, 10-ml air
in the calculation of manure storage contribution samples were taken through a rubber septum
to GHG emissions. However, CO2 emissions were within 2 min after the bins were enclosed, and
measured concomitantly to CH4, N2O emissions subsequently every 2.5 min over a 10 min period.
for monitoring the aerobic decomposition and for A sample of ambient air was taken 3 cm above the
evaluating if the enclosure was tightly installed manure surface prior to enclosure. All these CH4
over the bin, as explained further. and N2O concentration samples measured over
In the dairy cattle manure experiment, the time were used to determine the initial slope (at
sampling of slurry, stockpile and compost treat- t=0), required to calculate the flux. The air sam-
ments for CH4, N2O and CO2 emissions was ples were stored at room temperature and ana-
rotated among the three storage methods so that lyzed within 48 h using a gas chromatograph
two of the three storage methods were sampled (Varian, Walnut Creek, CA).
each day and all storage methods were sampled An infrared CO2 analyzer (LI-6251 from
twice per week. GHG emissions were measured for LI-COR, Lincoln, NE) was put in parallel with the
a total of 14 weeks from May 21 to August 18 tube that was equipped with the septum (Fig-
(Calendar Day: CD, 141–230). For the beef cattle ure 1). Air was circulated in the closed loop using
manure experiment, the sampling of each storage a diaphragm pump located downstream of the gas
bin was completed in one day, twice weekly for sampling port. The slopes of CO2 concentrations
11 weeks from June 2 to August 15 (CD 153–227). over time were measured using the LI-6251 ana-
However, from June 6 to 22 (CD 157–173), lyzer in three series of 1-min readings recorded at

Figure 1. Top view of the flow-though closed-loop gas sampling enclosure assembly placed over each manure storage bin during GHG
emissions determination.
176

1-s intervals on a CR-21X data logger (Campbell ionization detector for CH4 and an electron cap-
Scientific, Logan, UT). Measurements began ture detector for N2O. For CH4, the following
within 2 min after the bins were covered. The conditions were used: (i) 10-ml air sample injection
change of CO2 concentration over time was dis- at ambient temperature through a 2-ml sample
played in real time on a laptop computer for loop; (ii) the detector temperature, was at 390 C
diagnostic purposes. A steady CO2 slope over time and oven temperature was at 60 C; (iii) the col-
indicated that the system was leak free, and that a umn was made of 80/100 mesh Porapak-N packed
thorough mixing of the air within the enclosure into a 4.8 m long · 3.18 mm internal diameter
was being achieved. The LI-6251 analyzer was stainless steel column (Superlco, PA,USA) and (iv)
flushed with ambient air between measurements the gas flow rates were 30 ml min
1 for nitrogen,
over the different bins. 30 ml min
1 for hydrogen and 300 ml min
1 for
GHG emissions (FG, mg kg
1 DM s
1; DM: air. Methane (retention time 1.8 min) was quan-
dry manure) were calculated using the following tified by comparing the peak area with that of
equation: 1.36, 51.0 and 100.1 ppmv custom standards pre-
pared by Matheson Gases (Ottawa, Canada). For
@CG ð0Þ VE MG 1
FG ¼ N2O, an electron capture detector was used with
@t VG mm Ar-CH4 as carrier gas, a 2 ml-sample loop, and a
Where @CG(0)/@t is the rate of change in GHG column packed with Poropak Q (Superlco,
concentration at time t=0 (mmol mol
1 s
1), VE PA,USA). The two N2O standards used were
is the chamber headspace volume (m3), MG is the 1.17 ppmv and 10.0 ppmv (Matheson Gas,
molecular mass of the GHG (CO2 and N2O: Ottawa, Canada). For both gases, the minimum
44 g mol
1, CH4:16 g mol
1), VG is the volume of detection level was determined to be between 10
the gas at 21 C (0.024 m3 mol
1), and mm is the and 20 ppbv. A standard was run between each
initial dry mass of manure (kg). series of five air samples.
The fluxes of CH4 and N2O across the manure-
atmosphere interface are a function of the diffusion
coefficients and concentration gradients between Measurement of CO2 concentration
sites of generation (or transformation) and the
manure surface. Within a flow-through closed The infrared CO2 analyzer was operated in abso-
chamber, the gradient of the gas concentration lute mode within the 0 to 1500 ppmv CO2 range. A
changes over time, hence the rate of change of the scrubber that contained soda lime was used to
concentration of the gas within the enclosure provide CO2-free air to the reference cell. During
decreases continuously (Rochette et al. 1997). A non- regular operation, the scrubber was changed
linear model is used for describing the change of biweekly. The LI-6251 analyzer was calibrated
GHG concentration with time when the period is not before each series of measurements in the field
short enough to have a linear change (Hutchinson using a reference concentration at 618 ppmv of
and Livingston 1993). Since the gas concentration CO2 in nitrogen gas (Matheson Gas, Ottawa,
within the manure is much higher than in the head- Canada).
space atmosphere, it takes some time before the
gradient between manure and the atmosphere above
is affected significantly. A linear approximation was Chemical composition of the manure
adequate for about 75% of the GHG concentration
slopes measured over 10 min. For the cases when it Carbon, nitrate and ammonium concentrations in
was not adequate, a quadratic model was used. dairy cattle manure were measured on CD 141,
148 and 176. Initial moisture content was also
measured.
Beef manure was sampled seven times from all
Determination of CH4 and N2O concentrations the bins on CD 150, 152, 164, 178, 192, 220, 227
and three additional times from slurry bins on CD
Methane and N2O concentrations were measured 251, 279 and 300. Moisture determination was
using a gas chromatograph coupled to a flame done gravimetrically. The pH was measured after
 177

dilution with deionized water (1:5 ratio). Around 1993). Therefore, because GHG emissions were
10 g of fresh manure were extracted with 100 ml measured from the same nine storage bins for many
2 M KCl after shaking for 1 h. The extracts were weeks, we expected a lack of independence in
filtered and analyzed (Keeney and Nelson 1982) emissions over time. A repeated measures analysis
for mineral N (NH4+–N and NO3
–N) using a partitions the variation into a treatment effect, time
TRAACS 800 Auto Analyzer (Method 325.2 of effect, and time by treatment interaction.
Bran-Luebbe, Inc., Technicon Industrial Systems The data were divided into three periods based on
Corp., Elmsford, NY). Total soluble and inorganic the level of emissions, which corresponded to initial
C in the KCl extracts were measured using a Total (P1), intermediate (P2), and final (P3) phases of
Carbon Analyzer (Shimadzu TOC-5050). Total GHG emission as illustrated in Figures 2–4. We
soluble organic C was determined by subtracting used repeated measures analysis to study the effect
the total soluble inorganic C from total soluble of manure storage, measurement week, and their
C. Total C and N were determined by combustion interaction on GHG emissions for the three key
on a Carlo Erba (NA 1500, Nitrogen/Carbon/ periods. A weekly estimate of GHG emissions was
Sulfur Analyzer). Wet manure bulk density was calculated by averaging emissions from samples
determined on g l
1 basis by using 1 l of wet within a week. If only one sample was taken in a
manure. week, it was used as the weekly estimate. A repeated
measures analysis cannot be done when there are
missing values unless additional assumptions are
Environmental conditions of the storage bins made about the data. Therefore, if after averaging
within each week there were still missing estimates
Hourly temperatures at the surface of the manure of GHG emissions, the week was excluded from the
and in the middle of the bin were recorded over the analysis. In the dairy cattle manure experiment,
duration of the experiment. Observations of N2O emissions from the third week, and CO2
ambient temperature and pressure (PX961-16A5V emissions from the second week were unavailable.
pressure transducer, Omega Engineering Inc., In the beef cattle manure experiment, the initial CH4
Stamford, CT), relative humidity (HMP 35A emissions, and CO2 emissions from week five were
Humidity Probe, Vaisala, Helsinki, Finland), wind unavailable. Separate repeated measures analyses
speed and wind direction were also recorded at the were performed for each period and for each gas
site of the experiment, and averaged hourly. When using SAS version 6 PROC GLM with the
meteorological data were missing due to instru- REPEATED statement. This procedure provides
ment failure, data from the Ottawa airport both univariate and multivariate analyses of
(7.4 km away) were used. Three- and six-hour repeated measures. Following the recommendation
mean ambient temperature and pressure values of Potvin et al. (1990) conclusions were based on the
were calculated over four time intervals univariate tests when the significance levels adjusted
(9:00–12:00, 6:00–9:00, 6:00–12:00, 0:00–6:00 am by the Greenhouse-Geisser correction and the
EDT) prior to emission measurements to deter- Huynh-Feldt correction agreed. In this case, we
mine which interval influenced GHG emissions the report the Huynh-Feldt adjusted significance levels.
most. The pressure trend was calculated as being When the corrected significance levels did not agree
the difference in ambient pressure between the we used the multivariate tests with the level of sig-
emission measurement time and 3, 6, 12, and 24 h nificance a=0.1, since we had a large number of
prior to it. weeks compared to the number of replicates
(Stevens 1992). Stepwise multiple regressions were
used to determine if GHG emissions were correlated
Statistical analyses with environmental variables. Specifically, we tested
for relationships between GHG emissions and the
Repeated measures analysis (O’Brien and Kaiser mean ambient temperature (Ta), mean ambient
1985) was used to analyze the effect of manure pressure (P), pressure change (DP), relative
storage on GHG emissions. Repeated measures of humidity (RH), wind speed (U), wind direction (H),
an experimental unit over time are rarely inde- surface temperature (Ts) and core temperature (Tc)
pendent (O’Brien and Kaiser 1985; Von Ende of the manure. We also used stepwise multiple
178

Figure 2. Methane emissions from (a) dairy cattle manure and (b) beef cattle manure stored during the summer period as compost,
stockpile and slurry. Significant differences between storage methods resulting from the repeated measures analysis are indicated for
periods delineated with the dashed lines. *P=0.05, **P=0.01, ***P=0.001.

regressions to determine if variations in GHG to be about 90% of the total emissions over the
emissions could be explained by changes in the year.
chemical composition of beef manure. We tested for
relationships between emissions and the following
twelve chemical variables: NH4+, NO3
, inorganic Conversion of GHG emissions from a dry manure
nitrogen (INN), total carbon (TC), inorganic car- basis to a per head basis over a year
bon (IC), total organic carbon (TOC), percent
moisture (%W), pH, %N, %C, bulk density (BD), The average dairy cattle (IPCC, 2001,Table B-1
and ash content (ASH). p4.39) and non-dairy cattle (IPCC, 2001,Table
B-3 p4.43) dry manure produced per day and
per head for North America were converted to
Estimation of annual GHG emissions volatile solids (VS) by dividing them by 92%
(considering a proportion of 8% ash (IPCC,
The overall GHG emissions from various storage 2001)). The daily VS production was multiplied
methods were quantified by measuring GHG by 365 d to estimate the yearly production to be
emissions from a known mass of manure. No 1909 kg VS hd
1 year
1 for dairy cattle and to
fresh manure was added in the storage facility as be 858 kg VS hd
1 year
1 for non-dairy cattle.
is the usual practice on animal farms. This ap- The average non-dairy cattle manure production
proach was selected in order to be able to assign was applied to beef cattle production. The yearly
the overall measured GHG emissions to a known GHG emissions in CO2-eq estimated from the
mass of manure. The cumulative GHG emissions measurements were divided by the proportion of
measured over the sampling period were considered VS (92%) and then multiplied by the yearly
 179

Figure 3. Nitrous oxide emissions from (a) dairy cattle manure and (b) beef cattle manure stored during the summer period as
compost, stockpile and slurry. Significant differences between storage methods resulting from the repeated measures analysis are
indicated for periods delineated with the dashed lines. *P=0.05, **P=0.01, ***P=0.001.

averaged production of VS to express them on a population in Canada on manure storage systems

per head and year basis. and the proportion of manure storage facility rel-
evant to the experiment (for dairy cattle: 33%
liquid/slurry and 40% solid; for beef cattle: 6%
Total GHG emissions in Canada from manure liquid/slurry and 57% solid) (Statistics Canada
storage facility relevant to the experiment 2003) to obtain the total emissions in Canada from
manure storage facility relevant to the experiment.
The Canadian dairy cattle population is totaling The total emissions for dairy and beef cattle
1.06 million cows and 0.72 million heifers (Statis- manure by fully adopting a given storage system
tics Canada 2003). Because the manure produced was obtained by multiplying the yearly CH4 plus
by a heifer is about half that of a dairy cow, the N2O emissions in CO2-eq from manure produced
effective dairy cattle population was estimated to per cattle by the product of the effective cattle
be 1.48 million heads. The beef cattle population population in Canada times the proportion of all
on farms equipped with manure storage systems the manure storage facilities relevant to the
was estimated at 6.80 millions (Statistics Canada experiment, which is totaling 1.09 million dairy
2003). Because the proportion of beef calves is cattle and 2.59 millions beef cattle.
about 40% of the beef cattle and the proportion of
manure produced by heifers is similar to that of
beef cows, the effective beef cattle population was Reduction in emissions
estimated to be 4.09 millions. The yearly CH4 plus
N2O emissions in CO2-eq from manure produced The reduction in GHG emissions was calculated
per cattle were multiplied by the effective cattle by removing from the actual total emissions in
180

Figure 4. Carbon dioxide emissions from (a) dairy cattle manure and (b) beef cattle manure stored during the summer period as
compost, stockpile and slurry. Significant differences between storage methods resulting from the repeated measures analysis are
indicated for periods delineated with the dashed lines. *P=0.05, **P=0.01, ***P=0.001.

Canada from either dairy cattle manure storage in the intermediate period at 5.7 lg kg
1 DM s
1,
(0.57+0.35) Tg CO2-eq year
1 or beef cattle and then declined in the final period. Methane was
manure storage (0.08+0.15) Tg CO2-eq year
1 the still being emitted at the end of the experiment.
total emissions by fully adopting a given storage Methane emissions from the stockpiled manure
method. No results were reported for slurry since were significantly higher than from the slurry and
adopting this storage method would generate a from composting manure in the initial period (P1
surplus when methane emissions are not recycled. P< 0.001) and higher than from the composting
manure in the intermediate and final periods (P2
P<0.01, P3 P<0.05).
Results and discussion Methane emissions from the composting beef
manure were very low (Figure 2b). Methane
Methane emissions emissions from the stockpiled beef cattle manure
showed a similar pattern to the stockpiled dairy
Methane emissions from composting dairy manure cattle manure, but the emissions were significantly
were initially low (<2 lg kg
1 DM s
1) and de- lower (P=0.0012), being a third of those of the
creased throughout the experiment (Figure 2a). dairy cattle manure. Seasonal CH4 emissions from
Large episodic CH4 emissions from the stockpiled the beef cattle manure slurry were similar to CH4
manure were measured in the initial period with emissions from the dairy cattle manure slurry
peaks ranging from 2 to 5 lg kg
1 DM s
1, which (Figure 2a and b).
then decreased in the intermediate and final peri- Methane emissions from the composting beef
ods. Methane emissions from the slurry were low manure were positively correlated with total
initially but did increase in the first period, peaked inorganic nitrogen (Table 1), which consisted
 181

Table 1. Effects of beef manure chemical composition on GHG emissions.

Treatment Gas Variable 1 Partial R2 Variable 2 Partial R2 Variable 3 Partial R2 Other sig. variables R2

Compost CH4 +INN 0.71 0.71

N2O +NH4+ 0.64
%N 0.13 0.77
CO2 +NH4+ 0.52
%N 0.21
BD 0.09 0.81
Stockpile CH4 +IC 0.89 +Ash 0.04 0.92
N2O +pH 0.35 +IC 0.19 0.53
CO2 +IC 0.81
Ash 0.09 +%W 0.06
%N 0.98
Slurry CH4 +pH 0.27 0.27
N2O No significant variables
CO2 No significant variables

Models were created using Proc Stepwise in SAS. Only variables which were significant at a=0.05 using the Type II SS are included.
%W – water content; INN – Inorganic nitrogen; TC – Total carbon; IC – Inorganic carbon; BD – Bulk density; TOC – Total organic
carbon.

mostly of NH+ 4 in the first period and of NO3 in As temperature increased, higher rates of metha-
the second period. Ammonium is a primary nogenesis occurred in all manure storage types, al-
source of N for methanogenic bacteria (Knowles though overall increases in slurry temperature
1993). Consequently, ammonium promotes bac- resulted in the largest increase in CH4 production.
terial growth and CH4 production. In the stock- The methanogenic bacteria responded to the in-
piled manure, CH4 emissions were positively creased temperature and enhanced anaerobic
correlated with the concentration of inorganic environment. Anaerobic microsites were less com-
carbon, which was initially high in P1 and tended mon in the stockpiled manure, thus increases in
to decrease over time. In the slurry, CH4 emis- temperature resulted in less increase in CH4
sions were positively correlated with pH. How- production than in the slurry. In the initial
ever, the low R2 of the regression model indicates period, small amounts of CH4 were produced
that different descriptors are needed to better in the composting manure when core tempera-
predict CH4 emissions. tures increased and some anaerobic sites remained.
Methane emissions were positively correlated In the stockpiled manure and compost, CH4 pro-
with core and surface temperatures (Tables 2 and 3). duction was low in the intermediate and final

Table 2. Effects of environmental conditions on GHG emissions from dairy cattle manure.

Treatment Gas Variable 1 Partial R2 Variable 2 Partial R2 Variable 3 Partial R2 Other significant variables R2

Compost CH4 +TS 0.58
Ta 3 0.07 0.65

N2O +TC 0.36
D P 1 0.05 0.40
CO2 +TC 0.54 +RH 0.03 0.57
Stockpile CH4 +TS 0.50 +TC 0.08
H 0.06
Ta 1;
D P 3 0.70
N2O NONE
CO2 +TS 0.50
U 0.04 +TC 0.04 +RH 0.62
Slurry CH4 +TC 0.52 +TS 0.03 0.55
N2O +TS 0.20
D P 3 0.05 0.25
CO2 +TC 0.45
U 0.05 0.50

Models were created using Proc Stepwise in SAS. Only variables which were significant at a=0.05 using the Type II SS are included.
Ta – Mean ambient temperature averaged over: 1=9:00–12:00, 2=6:00–9:00, 3=6:00–12:00, 4=0:00–6:00
P – Mean ambient pressure averaged over: 1=9:00–12:00, 2=6:00–9:00, 3=6:00–12:00, 4=0:00–6:00
DP=Pressure change. Change calculated over: 1=3 h, 2=6 h, 3=12 h, 4=24 h
RH – Relative humidity
U – Wind speed
h – Wind direction
TS – Surface temperature
TC – Core temperature
182

Table 3. Effects of environmental conditions on GHG emissions from beef cattle manure.

Treatment Gas Variable 1 Partial R2 Variable 2 Partial R2 Variable 3 Partial R2 Other significant variables R2

Compost CH4 TS 0.60 U 0.16 0.75

N2O TS 0.68 H 0.06 DP 1 0.02 RH 0.80
CO2 TS 0.74 U 0.11 H 0.001 0.86
Stockpile CH4 TC 0.43 U 0.05 0.48
N2O TC 0.08 Ta 3 0.07 0.20
CO2 TC 0.68 U 0.10 TS 0.02 0.81
Slurry CH4 Ta 1 0.20 U 0.05 0.25
N2O No significant variables
CO2 U 0.37 Ta 1 0.10 H 0.07 0.54

Models were created using Proc Stepwise in SAS. Only variables which were significant at a=0.05 using the Type II SS are included.
Ta – Mean ambient temperature averaged over: 1: 9:00–12:00, 2: 6:00–9:00, 3: 6:00–12:00, 4: 0:00–6:00
P – Mean ambient pressure averaged over: 1:9:00–12:00, 2: 6:00–9:00, 3: 6:00–12:00, 4: 0:00–6:00
DP – Pressure change calculated over: 1: 3 h, 2: 6 h, 3: 12 h, 4: 24 h
RH – Relative humidity
U – Wind speed
h – Wind direction
TS – Surface temperature
TC – Core temperature

periods. As temperatures decreased and organic than from the stockpiled manure and slurry in the
material decomposed, the manure became more initial period (P<0.001), whereas they were simi-
porous and aerated, which caused CH4 emissions to lar from the stockpiled manure and slurry. In the
decline. intermediate and final periods, the N2O emissions
from both the composted and stockpiled manure
were higher than from the slurry (P<0.01,
Nitrous oxide emissions P<0.05).
Chemical changes in composted and stockpiled
Nitrous oxide emissions from composted beef cattle manure indicate that initial N2O emis-
and stockpiled dairy cattle manure were vari- sions were a result of nitrification, while later N2O
able in the first period, ranging from 0.05 emissions were a result of low levels of denitrifi-
to 0.16 lg kg
1 DM s
1 (Figure 3a), and cation (Table 4). During composting NH+ 4
decreased considerably during the final period reached a peak value after 12 days at 4.2 mg g
1
(<0.08 lg kg
1 DM s
1). Nitrous oxide emissions then decreased to 0.2 mg g
1 (data not shown).
from the composting manure were significantly Inorganic C concentration followed a similar trend


higher than emissions from stockpiled manure in to that of NH+ 4 concentration. NO3 peaked at

1
the initial and intermediate periods (P<0.01, 1300 mg g , right after ammonium decreased and
P<0.01). Nitrous oxide emissions from slurry stabilized at 700 mg g
1. Ammonium concentra-
were relatively low (<0.05 lg kg
1 DM s
1) and tion remained low in the stockpiled manure, while
significantly lower than from the composted and inorganic C increased from 40 to 78 ppmv in
stockpiled manure (P<0.05). 12 days then decreased to 19 ppmv. Shortly after
Nitrous oxide emissions from composted the decrease in inorganic C, NO3
concentration
beef cattle manure were highly variable ini- increased, indicating that inorganic C was the
tially (0.03–0.11 lg kg
1 DM s
1), and were primary source of N for nitrification
low in the intermediate and final periods Because there were limited N2O emissions from
(<0.03 lg kg
1 DM s
1) (Figure 3b). Nitrous the slurry, both the chemical and environmental
oxide emissions from the stockpiled manure and statistical models were ineffective at partitioning
slurry were low throughout the experiment the variance and had little predictive value
(<0.03 lg kg
1 DM s
1). Nitrous oxide emis- (Tables 1, 2 and 3). For both types of composted
sions from the composting manure were higher cattle manure, measurements of temperature
 183

Table 4. Changes in the chemical composition of manure stored under different conditions.

1
Storage Condition Sampling time NH+
4 lg g NO3
lg g
1 pH C/N Bulk Density g l
1 % Moisture

Dairy cattle manure

Start 1383 9 N/A 20.1 N/A 74.2
Compost 37 days 2010 1790 N/A 16.5 N/A 74.2
Stockpile 37 days 850 679 N/A 14.8 N/A 74.3
Slurry 37 days 940 0 N/A 19.9 N/A 82.2
Beef cattle manure
Start 1670 4 9.01 35.0 69.2 58.0
Compost End 197 722 9.15 17.6 113.8 57.7
Stockpile End 265 1078 8.65 17.4 79.0 63.1
Slurry End 492 53 7.31 21.5 89.7 86.2

Measurements were taken at the start, after 37 days (end of first period) for dairy cattle manure, and at the end of the beef cattle
manure experiment. N/A indicates where a measurement was not taken.

(surface and core) accounted for most of the were significantly higher than for the slurry
variations in N2O emissions (Tables 2 and 3). (P<0.001) in the initial period. In the intermediate
Composted dairy and beef cattle manure had the period, CO2 emissions from the composting man-
highest mean core temperatures in the initial per- ure and slurry were higher than from the stockpiled
iod (50–72 C). This result indicates that the Pas- manure (P<0.05). By the final period, CO2 emis-
sively Aerated Windrow System was effective at sions were low (<50 lg kg
1 DM s
1), and there
increasing the aeration of the manure and that were no significant differences among treatments.
there were high levels of exothermic reactions with Carbon dioxide emissions from the composting
O2. Stockpiled manure and slurry core tempera- and stockpiled beef cattle manure increased in the
tures were higher than ambient air temperature first week up to 400 lg kg
1 DM s
1 then
during the daytime by about 10–60 and 5–15 C decreased rapidly (Figure 4b). There was no sea-
respectively. sonal trend in CO2 emissions from the slurry.
Seasonal N2O emissions from dairy cattle Initially, CO2 emissions from the composting
manure and beef cattle manure were very different manure and stockpiled manure were higher than
(Figure 3a and b). There were more N2O emissions the slurry (P<0.001), but CO2 emissions from the
from stockpiled dairy cattle manure than from composting manure and stockpiled manure were
stockpiled beef cattle manure. Nitrous oxide not significantly different. Emissions were low for
emissions from composted dairy cattle manure the rest of the season (<50 mg kg
1 DM s
1), but
were similar to those of beef cattle manure in the emissions from the composting manure and the
initial period, but were higher in the intermediate stockpiled manure were still greater than from the
and final periods. Greater denitrification from slurry (P<0.01).
dairy cattle manure in the intermediate and final Carbon dioxide emissions from the composting
periods may account for the higher N2O emissions. beef cattle manure were correlated with NH4+
concentration. In the stockpiled manure, CO2
emissions were correlated with inorganic C con-
Carbon dioxide emissions centrations. There were little CO2 emissions from
the slurry, thus correlation between chemical
Carbon dioxide emissions from composted and variables and CO2 emissions were limited.
stockpiled dairy cattle manure were high initially Carbon dioxide emissions from the composting
(50–200 lg kg
1 DM s
1) but decreased with time and stockpiled manure were positively correlated
(Figure 4a). Carbon dioxide emissions from the with manure temperatures (surface and core,
slurry slowly increased to a maximum of Tables 2 and 3) which were much higher than
65 lg kg
1 DM s
1 on CD 189 and then ambient temperatures. These results indicate that,
decreased. Carbon dioxide emissions from the initially, there were high levels of aerobic decom-
composting manure and the stockpiled manure position. The decrease in CO2 emissions over time
184

gCO2-eq kg
1 DM
indicates that most of the organic material may
have been consumed by microbes.

Estimation obtained by multiplying the 3-month slurry dairy cattle manure emissions by the ratio of 5-month over 3-month of the slurry beef cattle manure emissions.
Trends in CO2 emissions were similar for both

CH4+N2O
Table 5. GHG emissions in g kg
1 and in CO2 equivalent over a 3-month summer storage period using global warming potentials of 23 for methane and 296 for N2O.
experiments, although initial emissions from beef

207.2
301.4
397.0
599.7

229.5
353.5
51.2
75.6
cattle manure were higher than from dairy cattle
manure. Beef cattle manure was better aerated due
to a higher proportion of bedding material. The
rapid respiration of organic carbon causing the

gCO2-eq kg
1 DM
peak in CO2 emissions was likely caused by higher
O2 levels. Carbon dioxide emissions from dairy
cattle manure slurry reached a peak in the middle
of the experiment, whereas CO2 emissions from

172.3
119.3
29.9
29.9

48.0
10.1
5.0
5.0
N2O
beef cattle manure slurry were consistently low.

gCO2-eq kg
1 DM
Management recommendations and potential
impact for Canada

Total emissions of CH4 and N2O (g kg
1 DM)

182.2
367.1
569.8

224.5
348.5
35.0

65.6
3.2
CH4
were calculated for each storage system for both
types of manure (Table 5). To obtain an estimate
of the effect of each manure storage treatment on
gCO2 kg
1 DM

global warming, CH4 and N2O emissions were

multiplied by their respective global warming
potential (IPCC 2001). The largest combined
398.58
440.00
208.13
231.25

340.06
358.11
141.69
157.43
CO2

N2O–CH4 emissions in CO2 equivalent were

observed from the slurry storage, followed by the
stockpile and lastly the passively aerated compost.
This ranking was governed by CH4 emissions in
gN2O kg
1 DM

relation to the degree of aerobic conditions within

the manure. The radiative forcing in CO2 equiva-
lent from the stockpiled manure was 1.46 times
0.582
0.403
0.101
0.101

0.162
0.034
0.017
0.017
N2O

higher than from the compost for both types of

cattle manure. It was almost twice as high from the
dairy cattle manure slurry and four to seven times
higher from the beef cattle manure slurry than
gCH4 kg
1 DM

from the compost. Negligible methane emissions

were found (Lessard et al. 1997) following field
application of stockpiled or composted dairy
15.96
24.77

15.15
1.52
7.92

0.14
2.85
9.76
CH4

manure at a rate of up to 100Mg ha
1 (equivalent

to 20 Mg DM ha
1), while N2O emissions were
10 times lower than the emissions during storage,
Slurry (5-month estimates)a

being equivalent to 1% of N added (Lessard et al.

1996) or equivalent to 0.047 g N2O–N kg
1 DM.
Dairy cattle manure

Using the model of Sommer et al. (2004) and the

Storage conditions

Beef cattle manure

Slurry (5 months)

dairy cattle slurry chemical composition of Paul

and Zebart (1997), to estimate N2O emissions
Stockpile

Stockpile
Compost

Compost

following the field application of slurry provide

estimates in the range of 0.20 g N2O–N kg
1 DM,
Slurry

Slurry

which is twice as much as during storage. Rochette

a
 185

et al. (2000) derived an emission factor of 1.65% of mechanical aeration is required. Another mitiga-
N applied for pig slurry. Combining storage and tion option could be to collect and burn the CH4
field application, confirmed that the highest N2O– emitted from the existing slurry facilities. In this
CH4 emissions in CO2-eq kg
1 DM came from case a reduction of 0.76 Tg CO2-eq year
1 could
manure stored as slurry followed by stockpiled be achieved. If all the cattle manure stored in
and then compost. Indeed, emissions following facilities relevant to the experiment was stored as
manure application were estimated to be less or slurry and CH4 collected and burned, then a
equal to 7% of the combined storage and field reduction of 1.08 Tg CO2-eq year
1 could be
application N2O–CH4 emissions expressed in achieved. Collecting and using CH4 on farm in-
CO2-eq kg
1 DM. volve additional costs that can be offset to some
The results of this study (Table 5) were extrap- extend by the energy saving (e.g. Martin 2003;
olated for Canada (Table 6), as detailed in the Engler et al. 2003). Based on the 6% GHG
material and method section, using the dry man- reduction commitment by 2008–2012 in the Kyoto
ure produced per head per year and using the protocol, it is anticipated that the Canadian agri-
conversion of dry manure into volatile solids given cultural sector will need to reduce GHG emissions
by IPCC (2001) and the results from the 2001 by about 7.75 Tg CO2-eq year
1 (Boehm et al.
census and special survey (Statistics Canada 2003). 2004). If one assumes full adoption of the pro-
The extrapolation does not account for possible posed mitigation options, the cattle manure stor-
variations in weather conditions and manure age practice might contribute between 9–14% of
composition across Canada. The reduction of the objective. Manure stored as slurry could also
GHG emissions brought by several mitigation be applied to fields as fertilizer prior to crop
scenarios was estimated using the extrapolation growth to avoid the increase in temperature during
for Canada. A reduction of 0.70 Tg CO2- the summer, which would reduce GHG emissions.
eq year
1 could be achieved by composting all the If slurry must be stored during the summer,
cattle manure stored as slurry and stockpile in Sommer et al. (2000) suggest that surface covers,
Canada, using the passively aerated windrow sys- such as straw, could reduce GHG emissions. The
tem (Table 6). Composting would generate only dairy cattle manure had higher radiative forcing
minor additional costs (i.e. the perforated pipes) if than that of the beef cattle manure for all the
the technique used in this study is adopted since no storage methods (Table 5).

Table 6. Annual GHG emissions during storage expressed in t CO2 equivalent per head, which are relevant to cool conditions in North
America.

Dairy cattle Beef cattle

Slurry Stockpile Compost Slurry Stockpile Compost

1
1 a
CH4 emissions (t CO2-eq hd year ) 1.301 0.416 0.080 0.357 0.067 0.003
N2O emissions (t CO2-eq hd
1 year
1)a 0.068 0.272 0.393 0.005 0.010 0.049
Thousands of effective animals/storage 491 597 249 2345
system in Canadab
Total emissions for relevant storage 0.672 0.411 0.090 0.182
systems in Canada (Tg CO2-eq year
1)
Total Canadian emissions from 1.489 0.749 0.515 0.941 0.201 0.136
dairy/beef cattle manure by fully
adopting a given storage system (Tg CO2-eq year
1)
Reduction in Canadian emissions (Tg CO2-eq year
1)
0.334
0.568
0.071
0.136

Actual and potential annual GHG emissions for Canada during storage for systems relevant to the experiment and potential reduction
by converting to a given storage system, all expressed in Tg CO2 equivalent.
a
Assuming an increase of 10% in emissions to account for emissions beyond the sampling period. Using 100-year global warming
potentials, yearly dry manure production per cattle head and conversion of dry manure into volatile solids provided by IPCC (2001).
b
# of effective cattle animals multiplied by the proportion of storage facilities relevant to the experiment extracted from Statistics
Canada (2003).
186

In order to estimate this impact of the study’s Desjardins R.L., Kulshreshtha S.N., Junkins B., Smith W.,
results for North America under cool conditions, Grant B. and Boehm M. 2001. Canadian greenhouse gas
mitigation options in Canada. Nutr. Cycl. Agroecosyst. 60:
new CH4 emission factors were calculated by first 317–326.
dividing the mass of the dry manure by 0.92 to Engler C.R., Jordan E.R., McFarland M.J. and Lacewell R.D.
convert it into volatile solids and then incorpo- 2003. Economics and environmental impact of biogas pro-
rating the measured emissions into the IPCC duction as a manure management strategy. http://tam-
methodology, which had proportions of storage mi.tamu.edu/Engler2.pdf 109–114.
Gaudy A.F. Jr. and Gaudy E.T. 1988. Elements of Bio-Envi-
practices already assigned. Emission factors were ronmental Engineering. Engineering Press, San Jose,
calculated (i) using the measured emissions as California.
input and (ii) assuming a 10% increase in the Hao X., Chang C., Larney F.J. and Travis G.R. 2001. Green-
measured emissions to account for emissions out- house gas emissions during cattle feedlot manure composting.
side the sampling period. For North-America J. Environ. Qual. 30: 376–386.
Hayes L.A., Richards R. and Mathur S.P. 1993. Economic
under cool conditions, the new CH4 emission Viability of Commercial Composting of Fisheries Waste by
factors were estimated to be 45.2 kg Passive Aeration at Shigawake Organics and Genesis
CH4 hd
1 year
1 for dairy cattle manure rather Organic. Proceedings of the Composting Council of Canada,
than 36 kg CH4 hd
1 year
1, and 2.7 kg 3rd Annual Meeting. Environment Canada, Montreal,
CH4 hd
1 year
1 for beef cattle manure rather Canada.
Husted S. 1994. Seasonal variation in methane emission from
than 1 kg CH4 hd
1 year
1. By using the pro- stored slurry and solid manures. J. Environ. Qual. 23: 585–
portion of manure storage practices relevant to 592.
Canada the emission factors become 35.8 kg Hutchinson G.L. and Livingston G.P. 1993. Use of chamber
CH4 hd
1 year
1 for dairy cattle manure and systems to measure trace gas fluxes. In: Harper L.A., Mosier
4.1 kg CH4 hd
1 year
1 for beef cattle manure. A.R., Duxbury J.M. and Rolston D.E. (eds), Agricultural
Ecosystem Effects on Trace Gases and Global Climate
This estimation shows how sensitive the emission Change. ASA Publication Number 55. Amer. Soc. Agron,
factors are to the allocation of the manure storage Madison, Wisconsin, pp. 63–78.
practices proportion. Keeney D.R. and Nelson D.W. 1982. Nitrogen-inorganic
forms. In: Page A.L., Miller R.H. and Keeney D.R. (eds),
Methods of Soil Analysis, Part 2, Chemical and Microbio-
Acknowledgements logical Properties. Amer. Soc. Agron. Madison, Wisconsin,
pp. 643–698.
Knowles and R. 1993. Methane: processes of production and
We thank Brendon McCormick and Mark Ed-
consumption. In: Harper L.A., Mosier A.R., Duxbury J.M.
wards for assisting in the data collection and and Rolston D.E. (eds), Agricultural Ecosystem Effects on
analyses, Théophile Paré for performing the Trace Gases and Global Climate Change. ASA Publication
chemical analyses, Lucie Bourque for assisting in Number 55. Amer. Soc. Agron, Madison, Wisconsin, pp.
the interpretation of Census and special survey 145–156.
Kulling D.R., Dohme F., Menzi H., Sutter F., Lischer P. and
data and Xavier Vergé for the suggestions while
Kreuzer M. 2002. Methane emissions of differently fed dairy
reviewing the manuscript. ECORC Contribution cows and corresponding methane and nitrogen emissions
number: 03-337. from their manure during storage. Environ. Monit. and
Assess. 79: 129–150.
Kulling D.R., Menzi H., Sutter F., Lischer P. and Kreuzer M.
References 2003. Ammonia, nitrous oxide and methane emissions from
differently stored dairy manure derived from grass- and
Amon B., Amon T., Boxberger J. and Alt C. 2001. Emissions of hay-based rations. Nutr. Cycl. Agroecosyst. 65: 13–22.
NH3, N2O and CH4 from dairy cows housed in a farmyard Lessard R., Rochette P., Gregorich E.G., Desjardins R.L. and
manure tying stall (housing, manure storage, manure Pattey E. 1997. CH4 fluxes from a soil amended with dairy
spreading). Nutr. Cycl. Agroecosyst. 60: 103–113. cattle manure and ammonium nitrate. Can. J. Soil Sci. 77:
Boehm M., Junkins B., Desjardins R.L., Kulshreshtha, S.N. 179–188.
and Lindwall W. 2004. Sink potential of Canadian agricul- Lessard R., Rochette P., Gregorich E.G., Pattey E. and
tural soils. Clim. Change 65(3): 297–314. Desjardins R.L. 1996. Nitrous oxide fluxes from manure-
Desjardins R.L. and Riznek R. 2000. Agricultural greenhouse amended soil under maize. J. Environ. Qual. 25: 1371–1377.
gas budget. In: McRae T., Smith C.A.S. and Gregorich L.J. Martin J.H.Jr. 2003. A Comparison of Dairy Cattle Manure
(eds), Environmental Sustainability of Canadian Agriculture. Management with and without Anaerobic Digestion and
Report of the Agri-Environmental Indicator Project. Agri- Biogas Utilization. Report for the AgSTAR Program U.S.
culture and Agri-Food Canada, Ottawa, Canada, pp 133– Environmental Protection Agency Contract #68-W7-0068
140. Task Order, 400: 58pp.
 187

Mathur S.P. 1991. Composting processes. In: Martin A.M. Rochette P., van BochoveE., Prévost D., Angers D.A., Côté D.
(ed.), Bioconversion of Waste Materials to Industrial Prod- and Bertrand N. 2000. Soil Carbon and Nitrogen dynamics
ucts. Elsevier, London, pp. 147–186. following application of pig slurry for the 19th consecutive
Mathur S.P. 1992. Agriculture Canada’s Passively Aerated year: II. Nitrous oxide fluxes and mineral nitrogen. Soil Sci.
Windrow System of Composting Farm, Food and Industrial Soc. Am. J. 64: 1396–1403.
Wastes. Proceedings of the Composting Council of Canada, Sommer S.G., Petersen S.O. and Moller H.B. 2004. Algorithms
2nd Annual Meeting. Environment Canada, Ottawa, for calculating methane And nitrous oxide emissions from
Canada. manure management. Nutr. Cycl. Agroecosyst. 69: 143–154.
O’Brien R.G. and Kaiser M.K. 1985. MANOVA method for Sommer S.G., Petersen S.O. and Sogaard H.T. 2000. Atmo-
analyzing repeated measures designs: an extensive primer. spheric pollutants and trace gases. J. Environ. Qual. 29: 744–
Psych. Bull. 97: 316–333. 751.
Overcash M.R., Humenik F.J. and Miner J.R. 1983. Livestock Statistics Canada 2003. Farm Environmental Management in
Waste Management II. CRC Press Inc, Boca Raton, Florida. Canada. Catalogue no. 21-021-MIE, 52pp. (http://
Paul J.W. and Zebarth B.J. 1997. Denitrification during the www.statcan.ca/english/freepub/21-021-MIE/free.htm).
growing season following dairy cattle slurry and fertilizer Stevens J. 1992. Applied Multivariate Statistics for the Social
application for silage corn. Can. J. Soil Sci. 77: 241–248. Sciences. 2nd ed. Lawrence Erlbaum, Hillsdale, New Jersey.
Potvin C., Lechowicz M.J. and Tardif S. 1990. The statistical Su J-J., Liu B-J. and Chang Y-C. 2003. Emission of greenhouse
analysis of ecophysiological response curves obtained from gas from livestock waste and wastewater treatment in
experiments involving repeated measures. Ecology 71: 1389– Tiawan. Agric. Ecosyst. Env. 95: 253–263.
1400. Svensson B.H., Lantsheer J.C. and Rodhe H. 1991. Sources and
SAS 1990. Version 6.09. SAS Institute, Cary, North Carolina. sinks of methane in Sweden. Ambio 20: 155–160.
Rochette P., Ellert B., Gregorich E.G., Desjardins R.L., Pattey von Ende and C.N. 1993. Repeated-measures analysis: growth
E., Lessard R. and Johnson B.G. 1997. Description of a and other time-dependent measures. In: Scheiner S.M. and
dynamic closed chamber for measuring soil respiration and Gurevitch J. (eds), Design and Analysis of Ecological
its comparison with other techniques. Can. J. Soil Sci. 77: Experiments. Chapman and Hall, New York, New York, pp.
195–203. 113–137.