Archivos Latinoamericanos de Producción Animal. 2023. 31 (4)
Repeatability and variability of measurements of methane and carbon
dioxide production in cattle housed in opencircuit respiration chambers
Recibido: 20220821. Revisado: 20221017. Aceptado: 20231005.
1Southeast University Regional Unit, Autonomous University Chapingo, km 7.5 Carr. TeapaVicente Guerrero, Teapa, Tabasco, Mexico
2Statistics and Calculation Program. Postgraduate College, Campus Montecillo. Carr. MexicoTexcoco. Km. 35.5. Montecillo, Edo. of Mexico,
3Corresponding author:
Ever del Jesús FloresSantiago1
Abstract. Gas recovery tests are necessary when the respiration chamber technique is employed for the
measurement of greenhouse gases exhaled by domestic animals. A dataset of 98 individual measurements of
methane and carbon dioxide production from cattle housed in two respirations chambers was used to assess
variability and repeatability of the measurements performed. Analysis of variance was carried out to assess if
statistically significant differences existed between chambers and between animals (P < .0001). Results showed the
occurrence of a moderate but acceptable variability in methane production measurements between the chambers
Key words: greenhouse gas, indirect calorimetry, uncertainty.
Laboratory of Climate Change and Livestock Production, Department of Animal Nutrition, Faculty of Veterinary Medicine and
Animal Science, University of Yucatan, Merida, Yucatan, Mexico.
Humberto VaqueraHuerta2
Repetibilidad y variabilidad de las mediciones de produccn de metano y
dióxido de carbono en bovinos alojados en cámaras de respiración de
circuito abierto
pero aceptable en las mediciones de producción de metano entre las cámaras de respiración evaluadas.
Palabras clave: gases de efecto invernadero, calorimetría indirecta, incertidumbre.
Repetibilidade e variabilidade de medições de produção de metano e dióxido
de carbono em bovinos alojados em câmaras de respiração de circuito aberto
Resumo. Experimentos de recuperação de gás são necessários quando a técnica da câmara respiratória é utilizada
para medição de gases de efeito estufa exalados por animais domésticos. Um banco de dados de 98 medições de
metano entérico (CH4) e dióxido de carbono produzido por bovinos alojados em duas câmaras respiratórias foi
usado para estimar a variabilidade e repetibilidade das medições realizadas. Uma análise de variância dos dados foi
realizada para avaliar se havia diferenças estatisticamente significativas entre as maras e entre os animais (P < 0,0001).
Os resultados mostraram a ocorrência de variabilidade moderada, mas aceitável, nas medidas de prodão de metano
entre as maras respirarias avaliadas.
Palavraschave: gases de efeito estufa, calorimetria indireta, incerteza.
Jesús Miguel CalzadaMarín3
Paulina Vásquez Mendoza1 Roberto González Garduño Juan Carlos KuVera3
FloresSantiago et al.
Greenhouse gas (GHG) emissions reached a record
51.5 gigatons of CO2 equivalent (GtCO2e) in 2019
excluding landuse change (LUC) emissions and 58.1
GtCO2e including LUC (United Nations Environmental
Programme, 2021). The main GHGs that determine
climate change are carbon dioxide (CO2), methane (CH4)
and nitrous oxide (N2O) (Olivier, 2022), these gases
contributed 73, 19 and 5 % of global total GHG
emissions respectively, excluding land use, with Fgases
accounting for the remaining 3 % (Olivier & Peters,
2020). Methane has a global warming potential of 28 to
36 times that of CO2 over a 100year period and 80 times
that of CO2 over a 20year period (IPCC, 2021).
Livestock production contributes 14.5 to 19 % of
global GHG emissions (Gerber et al., 2013; Johnson &
Johnson, 1995). Enteric methane is a major source of
greenhouse gas emissions from milk and beef
production systems that contributes to global warming
(Tricarico et al., 2022). Cattle are estimated to produce
between 250 to 500 L of CH4 per day (Johnson &
Johnson, 1995) with up to 90 % of the CH4 from
ruminants is produced in the process of rumen
microbial methanogenesis (McAllister et al., 2015).
OpenCircuit Respiration Chamber (OCRC) is the
gold standard technique for measuring methane in
ruminants provided that their gas recovery rates are
close to 100 % (Garnsworthy et al., 2019). Charmley et al.
(2016) conducted a metaanalysis of 1034 individual
observations generated by experiments using OCRC
and where foragebased diets (> 70 %) were used,
records obtained from dairy cattle fed warm forages
(220 records), beefproducing bovines fed with
temperate forages (680 records) and meatproducing
bovines fed with tropical forages (113 records). The
authors reported CH4 emissions g/d on the range,
237─623 (average 421 g CH4/d) for dairy cattle in
temperate regions, 78.9─241 (average 133 g CH4/d) for
beef cattle in temperate regions, and 32.2─184 (average
94.7 g CH4/d) for cattle in tropical regions. Based on
these results, they suggest using the value of 20.7 g CH4/
kg DM to estimate methane emissions in Australia.
In Mexico KuVera et al. (2020) analyzed 125 indivi
dual methane yield (CH4/kg DMI) data for Bos taurus ×
Bos indicus crosses that were fed lowquality tropical
forages (> 70 %) and evaluated at OCRC, their results
indicate a CH4 production of 17 g/kg of DMI under
those conditions. Which is comparable with the results
presented by Charmley et al. (2016). However, although
OCRC determinations are considered the gold standard
technique to determine CH4 emissions, it is necessary to
make significant improvements that contribute to
reducing variability and repeatability to improve the
values obtained. This can be achieved by homogenizing
the weight of the animals used, stabilize daily
consumption and calibrate the OCRC on a routine basis
(Fernández et al., 201Gerrits et al., 201Hristov et al.,
2018).Dhumez et al. (2022) reported that determination
of the gas recovery rate in respiration chamber facilities
is a central prerequisite to assess the accuracy of
methane emission quantification. However, data of
recovery tests are seldom reported (Gerrits et al., 2017).
Therefore, the objectives of this trial were to evaluate the
interanimal variability and repeatability of CH4, and
CO2 production measurements carried out in crossbred
(Bos taurus × Bos indicus) heifers housed in the open
circuit respiration chambers at the Laboratory of
Climate Change and Livestock Production of the
University of Yucatan, Mexico.
Ethical considerations
The experiment was
approved by the Bioethics
Committee and Manual for Research with Living
Organisms and Environmental Conservation of the
Faculty of Veterinary Medicine and Animal Science,
University of Yucatan, Mexico.
The experiment was carried out at the Laboratory of
Climate Change and Livestock Production (LACCLIGA)
of the Faculty of Veterinary Medicine and Animal
Science of the University of Yucatan (21°15´N 83°32´W)
in Mérida, México. The region has a warm subhumid
climate (Aw0) with rains in the summer. The average
annual temperature is 26.8 °C and the average rainfall is
984.4 mm (García, 1981).
Six crossbred heifers (Bos taurus × Bos indicus)
cannulated in the dorsal sac in the rumen (10 cm Bar
Diamond Inc.), 43 ± 4.4 months old and with an average
body weight (BW) of 426 ± 56.1 kg, were used. The
heifers were housed during the experimental period in
individual metabolic crates equipped with feeders and
drinkers; located in a roofed building, with a concrete
floor and no walls. The heifers were dewormed and
vitaminized 15 days before starting the experimental
period. The dewormer used was an oral anthelmintic
suspension (Oxfenil® Virbac México), 5 mL was
administered for every 100 kg of BW (equivalent to 4.5
mg of oxfendazole/kg of BW). In addition, 5 mL of
Vitafluid® Virbac México was administered individually
intramuscularly (each mL contains = vitamin A, 500,000
IU; Vitamin D3, 50,000 IU; vitamin E, 50 IU).
Materials and Methods
ISSNL 10221301. Archivos Latinoamericanos de Producción Animal. 2023. 31 (4): 319329
Uncertainty of CH4 and CO2 emissions of cattle in respiration chambers
Table 1. Proportion of ingredients in the ration and chemical composition.
Duration of the experimental period
The experimental period lasted 45 days, divided into
five measurement subperiods. Each subperiod was nine
days, during which heifers were interspersed in the two
opencircuit respirations chambers (OCRC) for three
days for determination of enteric methane emissions
(one animal per chamber, starting on the day one to
three with heifer one and two, ending each
measurement subperiod with heifer five and six) and the
other six days they remained in their respective
metabolic crates.
Experimental ration
Chemical composition of the experimental ration is
shown in Table 1, which consisted of 83 % Pennisetum
purpureum (regrowth of 120 days) fresh and chopped
(2.5 cm) and 17 % concentrate consisting of ground corn,
soybean paste and a commercial mineral premix,
covering the nutritional requirements for the
maintenance of growing heifers (National Academies of
Sciences & Medicine, 2021). The heifers had access to
clean, fresh water always during the experiment.
Item Treatment
Ingredients, g/kg of DM
Pennisetum purpureum 830
Ground corn 73
Soybean meal 37
Minerals 60
Chemical composition, g/kg of DM
Dry matter 942 ± 1.39
Organic matter 941 ± 1.22
Crude protein 52.7 ± 3.18
Neutral detergent fiber 709 ± 4.36
Acid detergent fiber 437 ± 13.88
Ether extract 8.13 ± 0.17
Ash 59.5 ± 1.22
Gross energy (MJ/kg DM) 15.4 ± 0.22
1 Mineral premix contained (minimum values per kg) = 40 g of phosphorus, 120 g of calcium, 0.74 g of iron, 10 g of magnesium, 400 g of sodium chloride, 1.5 g of
manganese, 1.5 g of zinc, 0.15 g of copper, 0.0018 g of iodine and 0.001 g of cobalt.
Experimental design
The experimental design used was completely
randomized where the study factor was the chamber
(levels = chamber 1 and 2), the number of replicates was
six (randomly assigned, three to chamber 1 and three to
chamber 2). The variables evaluated were dry matter
intake (DMI), enteric methane (CH4) production, and
carbon dioxide (CO2) production. These values were
determined when the animals remained in the chamber.
Voluntary intake
Individual dry matter intake (DMI) was measured
daily as the difference between the amount offered and
that rejected the following day. The full ration was
offered once a day at 8:00 h. The ration was adjusted
every third day and a 10 % excess over the expected
daily intake was offered. The rejects were withdrawn at
7:45 h. the following day. The samples of food offered
and rejected for each day were kept at 4 °C until the
end of the experiment for further analysis. The samples
were dried in a forcedair oven at 60 °C for 72 h and then
ground through a 2mm mesh in a Wiley® mill (Arthur
H. Thomas Co., Philadelphia, PA, USA). USA) and sent
to the Animal Nutrition Laboratory the Faculty of
Veterinary Medicine and Animal Science, University of
Yucatan, Mexico for chemical analysis.
Chemical analysis
Dry matter content of ration and refusals were
determined by drying subsamples in a forcedair oven
at 105 °C for 48 h (constant weight; # 7.007; AOAC
International, 2016). Nitrogen concentration in the
samples was analyzed (CP; N × 6.25) by the Dumas
combustion procedure using a LECO CN2000 3740
series equipment (LECO, Corporation, #2057; AOAC
International, 2016). OM and ashes were determined by
incineration in a muffle furnace at 550 °C for 6 h (AOAC
International, 2016; # 923.03) and, the NDF content was
determined using sodium sulfite without alpha amylase
(Van Soest et al., 1991). Ether extract (EE) was obtained
by the acid hydrolysis method using petroleum ether as
solvent (#920.39; AOAC International, 2016)). GE
concentration was determined in a calorimetric bomb
(C200, IKA Works® Inc., Staufen, Germany).
ISSNL 10221301. Archivos Latinoamericanos de Producción Animal. 2023. 31 (4): 319329
Dry matter intake (DMI), CH4 and CO2 production
are shown in Table 2. According to the analysis of
variance performed on the data, DMI was statistically
different between chambers (P=0.014) and between
animals (P=0.007). Similarly, CH4 production (g/d and
g/kg DMI), CO2 production [g/d and g/kg DMI
(methane yield)] and the CH4/CO2 ratio showed highly
significant differences between chambers (P < .0001;
Table 3), adjusted for the effect of animal.
On the other hand, the CV and residual CV for CH4/
d, CO2/d, CO2/DMI and CH4/CO2 ratio were more
uniform in chamber 2 (Table 2), indicating less
variability when we compared the values obtained
with those of chamber 1. In addition, DMI and CH4/
DMI, the CV and residual CV were better in chamber 1.
Intraclass correlation coefficient or repeatability for
chambers 1 and 2, is in a range of 0.43 to 0.57 (Table 2);
R values for DMI, CH4/DMI, CO2/DMI were higher in
chamber 1, on the contrary, in chamber 2 CH4/d, CO2/
d and CH4/CO2 presented the highest value of R.
Methane production
Measurements of CH4 and CO2 production were
carried out in two opencircuit respiration chambers
(OCRC) for periods of 23 hours. In total, 98 daily
individual measurements of heat production in three
heifers were analyzed. Heifers were fed a basal ration of
chopped fresh Pennisetum purpureum grass and a
supplement (ground corn + soybean meal); the level of
feeding was slightly above maintenance. Construction,
description, operation and calibration of the chambers is
described in CanulSolis et al. (2017) and ArceoCastillo
et al. (2019). The chambers (9.97 m3 volume) were built
from metalsheet panels with doublelayer insulation,
equipped with concrete floor, internal cage of tubular
steel, feeder, automatic waterer, and a lock for air intake.
Acrylic windows (9 mm thick) were installed at both
sides of the chambers so that cattle had visual contact
between them in the adjacent chamber as well as with
their surroundings. Chambers are equipped with air
conditioning units to guarantee comfort [temperature:
23 ±1 °C and relative humidity (RH) = 55 ±10 %].
Chambers are fitted with a small fan to provide
movement of air in the closed environment. To measure
the concentration of CH4 in air samples, an infrared
analyzer (MA10, Sable Systems International, Las
Vegas, Nevada, USA) was used. The apparatus was
calibrated before each run by zeroing it with pure N2
(99.999 Praxair, Mexico) following the methodology
described by ArceoCastillo et al. (2019). Subsequently, a
known concentration of CH4 (1000 μmol/mol; Praxair®
Gases Industrial Inc., Mexico) was released until the
equipment stabilized at 0.1 ±0.03 and the measurements
were then started. The respiration chambers had been
previously calibrated by infusing a known amount of
high purity methane CH4 (99.997 % purity) to assess
recovery rates that ranged from 97102 %, similar to
those reported by Gardiner et al. (2015) and Machado et
al. (2016). Carbon dioxide concentration in chamber air
samples was determined with an infrared analyzer
(Sable Systems, Las Vegas, Nevada, USA). The air inside
the chambers was removed using two mass flow
generators (Flow Kit 50500; Sable Systems, Las Vegas,
USA) at a rate of 1.0 L/min for each kg of animal live
weight (Machado et al., 2016), generating an internal
pressure of 276 Pa. The air samples passed through a
drying column filled with Drierite (WA Hammond
Drierite Company LTD®, USA) before being sent to the
CH4 analyzer through a multiplexer. The values
obtained (μmol/mol) in the ExpeData® software (Sable
Systems International®, USA) were extrapolated to 24 h.
Statistical analysis
To analyze the experimental data, the PROC GLM
procedure of SAS 9.4 (SAS, 2012) was used. Mean
separation was made by using the Tukey test with an
alpha of 0.05. The data was analyzed under the
following model:
Yij = µ + i + Aj(i) + ij
Yij = response variable in question taken in the i─th
chamber and the j─th cow.
µ = effect of the overall mean.
i = It is the effect of the i─th camera.
Aj(i) = Effect of the j─th cow within the i─th chamber.
ij = is the experimental random error, ij ~ N (0, δ2).
Residual coefficient of variation (CV) and Repeatability
(R) was determined according to those described by
Huhtanen et al. (2013); Residual coefficient of variation
was calculated as root mean square error divided by
mean, and repeatability was calculated as R = δ2Animal +
2Animal + δ2Residual).
FloresSantiago et al.
ISSNL 10221301. Archivos Latinoamericanos de Producción Animal. 2023. 31 (4): 319329
Table 2. Mean values, variability and repeatability of feed intake and gas emissions.
Item DMI(kg/d) CH4(g/d) CH4/DMI(g/kg) CO2(g/d) CO2/DMI(g/kg) CH4/CO2(g/kg)
OCRC 1 Mean 9.26a 265a 29.1a 2205a 243a 8.30a
CV 12.9 10.0 16.5 14.0 20.9 7.89
Residual CV (%) 45.7 57.1 49.1 56.9 51.7 53.9
Repeatability 0.47 0.52 0.43 0.51 0.44 0.53
OCRC 2 Mean 8.63b 320b 38.0b 2799b 332b 8.75b
CV (%) 15.3 7.50 17.1 10.6 19.5 6.76
Residual CV (%) 54.3 42.9 50.9 43.1 48.3 46.1
Repeatability 0.53 0.48 0.57 0.49 0.56 0.47
OCRC = Opencircuit respiration chamber; DMI = dry matter intake; CH4 = methane; CO2 = carbon dioxide.
Ruminants fed lowquality forages eructate
considerable amounts of CH4 gas to the atmosphere.
Methane is a GHG which leads to a decrease in the
energy available in feedstuffs for growth and milk
production (Palangi and Macit, 2021), representing
losses between 2 to 12 % of gross energy intake (Johnson
and Johnson, 1995), which may otherwise be used for
growth and production (Tapio et al., 2017). In the present
work, the emissions of CH4 represented on average losses
of around 16.3 MJ/d of energy intake or its equivalent
average Ym = 11.9 %, which agrees with the report by
Johnson & Johnson (1995). However, this value
Is above that reported by the Intergovernmental Panel
on Climate Change (IPCC, 2019) of 7 % for cattle fed
rations containing > 75 % forage and with digestibility
62 %. The average value of Ym reported by Niu et al.,
(2018) in a metaanalysis of an intercontinental database
is also lower than that found in the present trial (11.9 %
vs. 6.0 %). Ym’s higher to those usually reported (6.5 %
7.0 %) for grazing cattle are possible due to the fact that
the herd in a given region or country has inconsistent
production levels with the limits of feed quality as
defined in the categories of IPCC (2019). Therefore, this
organism recommends as a good practice a region or
OCRC = opencircuit respiration chamber CV = coefficient of variation; DMI = dry matter intake; CH4 = methane; CO2 = carbon dioxide. Means with different
letters denote significant difference at 5 % (Note Table 3).
Table 3. Statistical values for variation, F value and pvalue between chambers and between animals.
Item Variation FValue pValue
DMI OCRC 8.85 6.27 0.014
Animal (OCRC) 21.3 3.78 0.007
OCRC 68078 333 <.0001
Animal (OCRC) 38855 47.6 <.0001
OCRC 1764 88.0 <.0001
Animal (OCRC) 1180 14.7 <.0001
OCRC 7945215 471 <.0001
CO2 Animal (OCRC) 6657306 98.6 <.0001
OCRC 180544 98.8 <.0001
CO2/DMI Animal (OCRC) 145744 20.0 <.0001
OCRC 4.49 16.5 <.0001
CO2/CH4 Animal (OCRC) 11.5 10.6 <.0001
On the other hand, the CV and residual CV for CH4/
d, CO2/d, CO2/DMI and CH4/CO2 ratio were more
uniform in chamber 2 (Table 2), indicating less
variability when we compared the values obtained with
those of chamber 1. In addition, DMI and CH4/DMI, the
CV and residual CV were better in chamber 1. Intraclass
correlation coefficient or repeatability for chambers 1
and 2, is in a range of 0.43 to 0.57 (Table 2); R values for
DMI, CH4/DMI, CO2/DMI were higher in chamber 1,
on the contrary, in chamber 2 CH4/d, CO2/d and CH4/
CO2 presented the highest value of R.
Uncertainty of CH4 and CO2 emissions of cattle in respiration chambers
ISSNL 10221301. Archivos Latinoamericanos de Producción Animal. 2023. 31 (4): 319329
countryspecific Ym, taking into account the quality of
the diet offered to the animal as a validation method. In
this respect, the quality of the diet (chemical
composition and digestibility (Garnsworthy et al. 2019)
will determine DMI (Congio et al., 2022), being these two
factors important in the production of enteric CH4
(Garnsworthy et al., 201 Hristov et al. 2022). In the
present work, average DMI was 8.95 kg and methane
yield was 33.6 g/kg DMI. This high methane yield (g
CH4/kg DMI) may have resulted from the positive
correlation existing between NDF content and methane
production (Niu et al., 2018). This has been confirmed by
Moraes et al. (2014) who reported that NDF may be
utilized as an attribute of chemical composition of a
ration to predict emissions of enteric methane. Recent
results confirm this, as daily CH4 emissions increased
linearly (p < 0.05) from 325.2 to 391.9, from 261 to 399.8
and from 241.8 to 390.6 g CH4/day in cows in the early
stage, intermediate and late lactation, respectively, as
NDF in the diet was increased (Dong et al., 2022).
Nonetheless, the methane yield of 33.6 g CH4/kg DMI in
the present work is 67.2 % above the value reported by
Niu et al. (2018) of 20.1 g CH4/kg DMI. In order to
explain this difference it is important to mention that
NDF content of the ration employed in the present work
was 200 % higher (709 ± 4.36 g/kg DM; Table 1)
compared to the 354 g NDF/kg DM reported by Niu et
al. (2018). The direct effect of a high content of NDF in
the diet is to induce a decrease in DMI (National
Academies of Sciences & Medicine, 2021) as a result of
the filling effect of the undigested fiber residues in the
rumen (Allen, 2000). Furthermore, retention time of
digesta in the rumen is increased, as well as the time for
fermentation as a result of the high NDF content and the
lignin present in mature forages (National Academies of
Sciences & Medicine, 2021) a situation which may have
occurred with the forage used in the present study (120
days regrowth). The fibrolytic bacteria may be affected
when the supply of protein is below the minimum 7 %
for optimal rumen function (National Academies of
Sciences and Medicine & Medicine, 2016) and therefore,
limit fiber digestibility (Firkins, 2021). In this respect,
apparent digestibility of NDF in the present work was
55.7 % on average (data not presented), and the supply
of protein per kg DM was 5.27 % (Table 1), which is not
enough for the appropriate function of fibrolytic bacteria
in the rumen. Additionally, the amount of energy
supplied to the rumen microorganisms is an important
factor which affects the amount of nitrogen incorporated
into microbial protein (Lu et al., 2019), in this study the
levels of digestible and metabolisable energy estimated
were 1.99 and 1.64 Mcal/kg DM (data not shown) which
may have been a limiting factor for microbial growth.
Thus, the associative effects of a high level of dietary
NDF, a low supply of crude protein and energy in the
diet, may have favoured a high rate of CH4 production
per kg dry matter intake, as microbial growth was
restricted and the apparent digestibility decreased.
Individual dry matter intake and methane production per
kg dry matter intake can be observed in Figs. 1 and 2.
Figure 1. Emission of enteric CH4 and DMI in heifers housed in Open Circuit Respiration Chamber 1.
FloresSantiago et al.
ISSNL 10221301. Archivos Latinoamericanos de Producción Animal. 2023. 31 (4): 319329