Gaseous emissions from weaned pigs raised on different floor systems

Jean-Francois Cabaraux a,*, Francois-Xavier Philippe a, Martine Laitat b, Bernard Canart a,Marc Vandenheede a, Baudouin Nicks a

a Veterinary Ecology and Ethology Unit, Department of Animal Production, Bat. B43, Faculty of Veterinary Medicine, University of Liege, Boulevard de Colonster 20, 4000 Liege, Belgiumb Department of Production Animals Clinic, Bat. B42, Faculty of Veterinary Medicine, University of Liege, Boulevard de Colonster 20, 4000 Liege, Belgium

Agriculture, Ecosystems and Environment 130 (2009) 86–92

A R T I C L E I N F O

Article history:

Received 28 March 2008

Received in revised form 24 November 2008

Accepted 26 November 2008

Available online 20 January 2009

Keywords:

Weaned pigs

Deep litter

Slatted floor

Ammonia

Greenhouse gases

Water vapour

A B S T R A C T

Gaseous emissions from agriculture contribute to a number of environmental effects. Carbon dioxide

(CO2), methane (CH4) and nitrous oxide (N2O) are greenhouse gases taking part to the global problem of

climate change. Ammonia (NH3) emissions are responsible of soil acidification and eutrophication and

contribute also to indirect emissions of N2O. This work evaluated the influence of the type of floor on the

emissions of these gases in the raising of weaned pigs. Two trials were carried out. In the first trial, the

animals were kept either on fully slatted floor or on straw-based deep litter and, in the second one, either

on fully slatted floor or on sawdust-based deep litter. For each trial and on each type of floor, 2 successive

batches of weaned pigs were raised without changing the litter or emptying the slurry pit between the 2

batches. The rooms were automatically ventilated to maintain a constant ambient temperature.

The performance of the animals was not significantly different according to the floor type. In trial 1,

the nitrogen contents of the straw deep litter (including the substrate) and slurry were respectively 276

and 389 g pig�1. In trial 2, the sawdust deep litter and slurry nitrogen contents were respectively 122 and

318 g pig�1.

Raising pigs on straw deep litter produced proportionately around 100% more NH3 than raising pigs

on slatted floor (0.61 g NH3-N d�1 pig�1 vs. 0.31 g NH3-N d�1 pig�1; P < 0.05). Differences in CO2, H2O

and CH4 emissions were not significant between systems. Raising pigs on sawdust deep litter produced

also proportionately more NH3 (+52%; 0.55 g NH3-N d�1 pig�1 vs. 0.36 g NH3-N d�1 pig�1; P < 0.01) but

also more CO2 (+25%; 427 g d�1 pig�1 vs. 341 g d�1 pig�1; P < 0.001) and H2O (+65%; 981 g d�1 pig�1 vs.

593 g d�1 pig�1; P < 0.001) and less CH4 (�40%; 0.52 g d�1 pig�1 vs. 0.86 g d�1 pig�1; P < 0.001) than

raising pigs on slatted floor. Practically no N2O emission was observed from rooms with slatted floor

while the N2O emissions were 0.03 and 0.32 g N2O-N d�1 pig�1 for the straw and sawdust deep litter

respectively. The warming potential of the greenhouse gases (N2O + CH4), were about 22, 34 and 168 g

CO2 equivalents per day and per pig on fully slatted floor, straw or sawdust deep litter respectively.

In conclusion, pollutant gas emissions from rearing of weaned pig seem lower with fully slatted

plastic floor system than with deep litter systems.

� 2008 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment

journa l homepage: www.e lsev ier .com/ locate /agee

1. Introduction

Agriculture contributes importantly to pollutant gaseousemissions such as ammonia (NH3), carbon dioxide (CO2), methane(CH4) and nitrous oxide (N2O) (Monteny et al., 2006; Aneja et al.,2007). These gases have a number of environmental bad effects.

NH3 contributes to the acidification and eutrophication of soilsand waters (Brink et al., 2001) and to indirect emissions of N2O(Intergovernmental Panel on Climate Change, 2006a). Further-more, NH3 is well known as a toxic gas, irritating the respiratorytract at concentrations exceeding 15 ppm (Urbain, 1997). In

* Corresponding author. Tel.: +32 4 366 59 03; fax: +32 4 366 41 22.

E-mail address: jfcabaraux@ulg.ac.be (J.-F. Cabaraux).

0167-8809/$ – see front matter � 2008 Elsevier B.V. All rights reserved.

doi:10.1016/j.agee.2008.11.016

Europe, approximately 80% of NH3 production originated fromanimal production facilities (Brink et al., 2001). According to vander Peet-Schwering et al. (1999), approximately 50% of theammonia emissions from pig production are from pig housingand slurry storage. The other 50% is emitted during surfaceapplication of the slurry.

CO2, CH4 and N2O are the most important greenhouse gases(GHG) associated with livestock production. These gases take partto the global problem of climate change. However, agriculture isalso a CO2-consumer through plant photosynthesis and thecontribution of CO2 to the greenhouse effect is less importantthan that of CH4 and N2O, whose warming potentials over a 100-year period are, respectively, 21 and 310 times that of CO2

(Intergovernmental Panel on Climate Change, 2007). N2O alsocontributes to the destruction of the ozone shield. Approximately

J.-F. Cabaraux et al. / Agriculture, Ecosystems and Environment 130 (2009) 86–92 87

40% and more than 50% of the anthropogenic emissions of CH4 andN2O originate from agriculture. Most important agriculture relatedCH4 sources are animals and their excreta (manure), whereas, mostof the N2O is produced in the field (faeces and urine excretedduring grazing, nitrogen (N) chemical fertilisers, land appliedanimal manure) and from animal houses where straw or litter isused (Monteny et al., 2006). The gaseous emissions from livestockhouses are thus dependent among others from the housing andfloor systems.

The collection of wastes in the form of solid manure with litterpresents various environmental advantages compared to liquidslurry such as a reduction in the weight of wastes collected, adecrease in the amount of N in the wastes and a decreasedolfactory nuisance (Nicks et al., 2003, 2004). Furthermore, the useof straw has also the asset to improve the pig welfare (Tuyttens,2005). However, the diminution of the N content of the wastesfrom litter results in the higher emissions of atmospheric N (N2),NH3 and N2O (Nicks et al., 2003, 2004).

The Kyoto protocol specifies that each complying countryshould provide adequate methods and instruments to quantify,monitor and verify GHG emissions and their reductions (Montenyet al., 2006). It is thus important to know precisely the emissionsassociated with different production techniques. However, veryfew experiments have compared in standardized conditionsgaseous emissions of weaned pigs according to floor systems andtherefore, few data are available in the literature. The aim of thisstudy was thus to quantify gaseous emissions in the raising ofweaned pigs according to the type of floor (fully slatted floor,straw-based deep litter or sawdust-based deep litter).

2. Material and methods

Two trials were carried out successively in experimental roomslocated at the Faculty of Veterinary Medicine of Liege University(Belgium). In the first trial, the gaseous emissions were measuredwith piglets kept either on fully slatted floor or on straw-baseddeep litter and in the second one, with piglets either on fully slattedfloor or on sawdust-based deep litter.

2.1. Animals and feed

For each trial and on each type of floor, 2 successive batches of40 weaned pigs were raised without changing the litter oremptying the slurry pit between the 2 batches. All the pigs wereoriginated from the same farrowing herd and were divided into 2homogeneous groups according to the sex and the body weight.

The pigs were fed ad libitum. Upon their arrival, they were givena transition feed (baby starter) which after 4 d was graduallyreplaced by a post-weaning feed (starter). Crude protein, lysineand crude fibre contents measured for the baby starter were 17.5,1.3 and 3.9% and those for the starter, 17.6, 1.2 and 4.3%respectively. The pigs were weighted individually at the beginningand at the end of the experimental period. The quantities of feedingested and water consumed were determined per batch.

2.2. Experimental rooms

Two identical rooms with an area of 30 m2 and a volume of103 m3 were arranged to house simultaneously a group of 40weaned pigs on a fully slatted floor in the first one and on a deeplitter in the second one. The slatted floor of plastic panels had avoid percentage of 37%. The available floor area for the pigs was12.2 m2 (0.31 m2 pig�1). The slurry pit was 50 cm deep. Before thearrival of the first animals, 600 l water was poured into the pit tohave a 5 cm water layer. The available floor space for animals inthe room with deep litter was 21.6 m2 (0.54 m2 pig�1). Straw deep

litter was realized with a 30 cm layer before the arrival of theanimals. Thereafter, supplementary quantities of straw wereprovided depending on the cleanliness of the litter. The totalamount of wheat straw was 642 kg (8 kg pig�1 on average for the 2batches). For the sawdust deep litter, 2000 kg of sawdust was usedto have a 20 cm layer before the arrivals of the pigs. Nosupplementary amount was provided thereafter, so the averageamount for the 2 batches was 26.7 kg pig�1. The sawdust had a drymatter (DM) content of 65% and was mostly composed of particleswith a diameter of 0.2–2.0 mm, which represented 80% of theweight. At the end of each trial, slurry and deep litter wereweighed and their DM content and N content, analysed by theKjeldah method, were determined. About every 10 d, but notduring the gaseous concentration measurements, wastes from thepigs raised on sawdust were dispersed over all the area of the penand incorporated manually, which was not done with the straw-based litter. In addition, to avoid an excessively high concentra-tion of dust in the air, the sawdust-based litter was moistened the8th and the 13th days after the arrival of the second batch withrespectively 24 and 48 l of water.

Each room was ventilated with an exhaust fan and theventilation rate was adapted automatically to maintain a constantambient temperature. Fresh air entered through an opening of0.34 m2 which was connected to the service corridor of thebuilding; the outside air was thereby preheated before entering theexperimental rooms. Moreover, a radiator and 2 heat lamps wereplaced in each room to obtain the piglets required temperatureduring the first part of the stay. The air temperatures of the 2 roomsand the corridor were measured automatically every hour. Theventilation rates were measured continuously and the hourlymeans were recorded with an Exavent apparatus (Fancom1) withaccuracy as specified by the manufacturer, of 35 m3 h�1, i.e. 1% ofthe maximum ventilation rate of the fan.

2.3. Gas emissions measurement

The concentrations of gases in the 2 experimental rooms andthe corridor supplying fresh air were measured with an apparatusfrom Innova Air Tech Instruments (1312 Photoacoustic Multi-gasMonitor) equipped and calibrated for the measurement of NH3,N2O, CH4, CO2 and water vapour (H2O). The air in the experimentalrooms was sampled upstream of the exhaust fan and that of thecorridor, at 1 m from the air inlet. For each batch, the concentra-tions were measured 3 times at about 1-week intervals and for 6consecutive days respectively, i.e. during approximately 45% of thestay. The Multi-gas monitor was programmed by conducting acycle of 3 measurements every half-hour, once every 10 min, theair being sampled successively in the room with the fully slatedfloor, the room with the deep litter and the corridor.

The emissions were calculated on an hourly basis taking thehourly concentration as the average of the 2 measurementsperformed per hour at each location. The emissions were expressedin mg h�1 utilizing the following formula: E = D � (Ci � Ce) with D,the hourly mass flow (kg air h�1); Ci and Ce, the concentrations of gasin the air of the room and corridor respectively (mg kg�1 dry air). Themean emissions per day and per pig were calculated for each seriesof measurements.

2.4. Statistical analyses

Statistical analyses were realized for each trial separately.About performance, initial and final body weights and averagedaily gains data were collected individually and tested per trialaccording to the floor system using a general linear model -procGLM- (SAS, 1999). About gaseous emissions, for each batch andeach gas and for the combined data obtained with the 2 batches,

J.-F. Cabaraux et al. / Agriculture, Ecosystems and Environment 130 (2009) 86–9288

the differences of the emissions with regard to the floor systemwere tested in the form of a mixed model for repeated mea-surements -proc mixed- (SAS, 1999). However, in each trial,because the litters were not changed and the slurry pit not emptiedbetween the 2 batches, the second batch cannot be considered as areal replication of the first batch.

3. Results

3.1. Animal performance

The performance is presented in Table 1. The staying duration ofa batch was around 6 weeks. During trial 2, accidental lossesoccurred, respectively two in batch 1 and three in batch 2. In orderto maintain equal the pig number in the 2 groups raised either ondeep litter or on slatted floor, one healthy pig was removed fromone of the experimental room the same day that one mortality wasobserved in the other room.

The mean initial and final weights were respectively 7.14� 0.2 and 23.0 � 1.6 kg (S.D. between 8 groups). The average dailyweight gains (ADG) were not significantly different with pigs kepteither on slatted floor or on deep litter, with an average of382 � 17 g d�1. The mean feed conversion ratio (kg feed kg�1 gain)was 1.66 � 0.15.

3.2. Climatic characteristics of the rooms

In trial 1, the average temperatures of the air were 23.9 8C in theroom with the straw deep litter and 26.4 8C in the room with theslatted floor (Table 2). The mean ventilation rates were 216 and238 m3 h�1 for the 2 rooms respectively. In trial 2, the averagetemperatures of the air were 22.6 8C in the room with the sawdustdeep litter and 24.2 8C in the room with the slatted floor. Themean ventilation rates were 414 and 483 m3 h�1 for the 2 roomsrespectively.

Table 1Animal performance.

Batch Trial 1

Straw deep litte

Pigs number 1 40

2 40

1 and 2 80

Staying duration (d) 1 39

2 46

1 and 2 85

Initial weight (kg) 1 7.4a � 1.1

2 7.0a � 1.2

Mean � S.D. 7.2a � 1.2

Final weight (kg) 1 21.9a � 2.7

2 25.5a � 3.8

Mean � S.D. 23.7a � 3.3

Average daily gain (g d�1) 1 373a � 59

2 401a � 72

Mean � S.D. 387a � 65

Feed conversion ratio (kg feed kg�1 gain) 1 1.37

2 1.76

Mean 1.57

Drinking water (l) per day 1 1.6

2 1.6

Mean 1.6

Drinking water (l) per kg ingested feed 1 3.1

2 2.3

Mean 2.7

Data within a trial and a line with different superscripts (a, b) are significantly differen

3.3. Amounts and composition of manure

In trial 1, the amounts of straw manure and slurry removed atthe end of the experiment were 27.5 kg pig�1 at 326 g DM kg�1 and37.0 kg pig�1 at 163 g DM kg�1 respectively. For trial 2, there were24.0 kg pig�1 sawdust manure (438 g DM kg�1) and 32.0 kg pig�1

slurry (124 g DM kg�1) removed.In trial 1, the deep litter and slurry N contents were 30.7 and

64.5 g kg�1 DM respectively and in trial 2, 11.6 and 80.1 g kg�1 DM.Taking into account the amounts of deep litter and slurry producedand their DM content, their N content per pig raised could becalculated and were 276 g pig�1 for the straw deep litter and389 g pig�1 for slurry in trial 1 and 122 g pig�1 for the sawdustdeep litter and 318 g pig�1 for slurry in trial 2.

3.4. Gas emissions

Table 3 presents the mean emissions observed for each batchand Fig. 1 shows the evolution of the emissions from the beginningto the end for each post-weaning period.

In trial 1, over the 2 post-weaning periods altogether, raisingpigs on straw deep litter produced proportionately around 100%more NH3 (0.61 g NH3-N d�1 pig�1 vs. 0.31 g NH3-N d�1 pig�1;P < 0.05) than raising pigs on slatted floor. Differences betweenCO2, H2O and CH4 emissions were not significant. No N2O wasmeasured from the slurry while the N2O emissions from the strawdeep litter were observed only during the stay of the second batch(0.06 g N2O-N d�1 pig�1).

Gas emissions increased regularly from the beginning to theend of each post-weaning period whatever the floor system (Fig. 1).

In trial 2 and over the 2 periods altogether, raising pigs onsawdust deep litter produced also per day proportionately moreNH3 (52%; 0.55 g NH3-N d�1 vs. 0.36 g NH3-N d�1; P < 0.01), CO2

(25%; 427 g d�1 vs. 341 g d�1; P < 0.001) and H2O (65%; 981 g d�1

vs. 593 g d�1; P < 0.001) and less CH4 (40%; 0.52 g d�1 vs.

Trial 2

r Fully slatted floor Sawdust deep litter Fully slatted floor

40 38 38

40 37 37

80 75 75

39 40 40

46 40 40

85 80 80

7.4a � 1.0 7.0b � 1.1 6.9b � 0.9

7.0a � 1.3 7.2b � 1.1 7.2b � 1.1

7.2a � 1.2 7.1b � 1.1 7.1b � 1.0

21.7a � 2.9 22.3b � 2.4 21.5b � 3.0

25.0a � 3.6 22.1b � 3.0 23.7b � 3.6

23.4a � 3.3 22.2b � 2.7 22.6b � 3.3

368a � 62 381b � 52 362b � 66

391a � 67 372b � 74 410b � 74

379a � 64 377b � 63 386b � 70

1.66 1.51 1.64

1.81 1.77 1.73

1.74 1.64 1.69

1.6 1.7 1.3

1.3 1.8 1.7

1.5 1.8 1.5

2.6 3.0 2.2

1.9 2.7 2.4

2.3 2.9 2.3

t (P < 0.05).

Table 2Mean air temperatures of experimental rooms and service corridor, and mean ventilation rates during the 2 periods.

Batch Trial 1 Trial 2

Straw deep litter room Fully slatted floor room Service corridor Sawdust deep litter room Fully slatted floor room Service corridor

Air temperatures (8C)

1 24.4 � 1.8 27.1 � 1.8 16.5 � 1.8 17.7 � 1.5 22.7 � 0.8 23.8 � 1.4

2 23.4 � 1.3 26.7 � 1.0 15.4 � 1.3 14.9 � 1.8 22.5 � 1.0 24.5 � 1.0

Mean � S.D. 23.9 � 1.6 26.4 � 1.6 15.9 � 1.7 16.3 � 1.6 22.6 � 0.9 24.2 � 1.2

Ventilation rates (m3 h�1)

1 229 � 80 253 � 87 – 421 � 158 454 � 219 –

2 203 � 60 224 � 89 – 407 � 59 511 � 71 –

Mean � S.D. 216 � 71 238 � 88 – 414 � 109 483 � 145 –

J.-F. Cabaraux et al. / Agriculture, Ecosystems and Environment 130 (2009) 86–92 89

0.86 g d�1; P < 0.001) than raising pigs on slatted floor. Practicallyno N2O emission was observed from the room with slatted floor(0.01 g N2O-N d�1 pig�1) while the N2O emissions were 0.08 and0.57 g N2O-N d�1 pig�1 respectively for the first and the secondbatches on sawdust deep litter. CO2 and H2O emissions increasedregularly from the beginning to the end of each period whateverthe floor system.

4. Discussion

In each trial, the manure N content was higher in the slurrysystem than in the two litter systems. These data confirmedprevious results with weaned (Nicks et al., 2000) and fattening pigs(Philippe et al., 2007a).

The mean NH3 emissions from the slatted floor rooms weresimilar in the 2 trials (around 0.33 g NH3-N d�1 pig�1). Theseemissions were lower than those in the 2 rooms with deep litter(around 0.58 g NH3-N d�1 pig�1), which can partially explain thehigher N content in slurry manure. The results with pigs on litterare in agreement with data of Nicks et al. (2003) but NH3 emissionfrom the slurry system was low compared to data from theliterature. The reference values for determining the best availabletechniques to reduce NH3 emissions are 0.50–0.66 kg NH3-N perweaner place, i.e. 1.35–1.80 g NH3-N d�1. Some techniques definedas best available can reduce this level by 60% (EuropeanCommission, 2003). The low NH3 emissions of slurry observed

Table 3Gas emissions in the raising of 2 batches of weaned pigs on deep litter with straw or

Trial 1

Straw deep litter Fully slatted floor S.E.

Batch 1

NH3 (g N d�1 pig�1) 0.66 0.29 0.15

N2O (g N d�1 pig�1) 0.00 0.00 –

CH4 (g d�1 pig�1) 0.56 1.03 0.06

CO2 (g d�1 pig�1) 327 308 38.1

H2O (g d�1 pig�1) 694 634 104

Batch 2

NH3 (g N d�1 pig�1) 0.57 0.34 0.15

N2O (g N d�1 pig�1) 0.06 0.00 0.03

CH4 (g d�1 pig�1) 0.95 0.79 0.09

CO2 (g d�1 pig�1) 340 298 28.6

H2O (g d�1 pig�1) 685 556 99.8

Batches 1 and 2

NH3 (g N d�1 pig�1) 0.61 0.31 0.07

N2O (g N d�1 pig�1) 0.03 0.00 0.01

CH4 (g d�1 pig�1) 0.75 0.91 0.1

CO2 (g d�1 pig�1) 334 303 19.7

H2O (g d�1 pig�1) 689 595 48.1

S.E.: mean standard error; Sig.: significance; NS: not significant.* P < 0.05.** P < 0.01.*** P < 0.001.

in this study could be related to the presence of a water layer in thebottom of the slurry pit at the beginning of the experiment and tothe use of a plastic floor which is generally cleaner than a concretefloor.

Higher NH3 emissions with litter compared to slurry were everobserved in trials with fattening pigs (Philippe et al., 2007a).According to Andersson (1996) a combination of high temperatureand high pH, both observed in litter, could explain this difference.However, other factors like feed management – especially proteinand fibre levels – and interior conditions can influence the level ofNH3 emissions from pig housing.

N2O is produced during the nitrification and denitrificationprocesses which normally convert NH3 into inert N2 gas.Nitrification requires aerobic conditions and denitrificationrequires anaerobic conditions (Monteny et al., 2006). Bothconditions can be found in deep litter but not in slurry. However,emissions from manure on the floor can occur in pig houses withslatted floors. So, values of about 2–4% of excreted N wereobserved with concrete slatted floor system for fattening pigs(Kermarrec, 1999).

According to our results there were no N2O emissions with thefully slatted floor probably in relation with the use of plastic floorwhich is generally cleaner than concrete floor. The emissions fromdeep litter were lower with straw compared with sawdust(0.03 g N2O-N d�1 pig�1 vs. 0.32 g N2O-N d�1 pig�1). This differ-ence between straw and sawdust was already observed by Nicks

sawdust or on fully slatted floor.

Trial 2

Sig. Sawdust deep litter Fully slatted floor S.E. Sig.

NS 0.55 0.29 0.04 *

NS 0.08 0.01 0.05 NS* 0.43 0.75 0.02 **

NS 425 314 11 *

NS 889 528 57.8 *

NS 0.55 0.44 0.07 NS

NS 0.57 0.00 0.06 **

NS 0.61 0.96 0.01 **

NS 429 367 7.65 *

NS 1074 659 26.4 **

* 0.55 0.36 0.04 **

NS 0.32 0.01 0.08 **

NS 0.52 0.86 0.02 ***

NS 427 341 7.15 ***

NS 981 593 37 ***

Fig. 1. Gas emission (per day and per pig) during the raising of weaned pigs on fully slatted floor (open bars) and on deep litter (closed bars) with 3 series of measurements per

batch at the beginning (B), middle (M) and end (E) of each period.

J.-F. Cabaraux et al. / Agriculture, Ecosystems and Environment 130 (2009) 86–9290

et al. (2003) with 2 batches of weaned pigs (0.11 g N2O-N d�1 pig�1

vs. 0.69 g N2O-N d�1 pig�1). It is probably the state of litteraeration that accounts for this difference, being higher in thesawdust-based litter compared to that with straw. Indeed, the

sawdust litter was aerated every 10 d, except during themeasurements, in order to favour composting. Kermarrec (1999)has shown that the aeration of sawdust-based litter stronglystimulates the release of N2O. However, the N2O emissions might

J.-F. Cabaraux et al. / Agriculture, Ecosystems and Environment 130 (2009) 86–92 91

be underestimated due to the fact that they were not measured onthe days of sawdust litter aeration. Indeed, in a short laboratorytrial (15 d) with continuous measurements, Kermarrec (1999)observed more N2O emissions (+70%) when the sawdust litter wasmixed, all other conditions being similar.

From deep litters, the nitrification process leads to N2

emissions. To estimate these emissions, the N balance (differencebetween N input (ingested N + litter N content)/retained N/excreted N) was calculated on a basis of N retention by pigs equalto 53.6% of N consumption, as reported by Dourmad et al. (1999)for weaned pigs with initial and final weight of 7.5 and 26 kgrespectively. On this basis, N2 emissions from the straw andthe sawdust deep litter systems were 1.68 and 4.22 g N2-N d�1 pig�1respectively, representing 21 and 52% of excreted N.Applied to results from the slurry system, the N balance calculationled to a negative results (�0.3 g N2-N d�1 pig�1 on average for the 2trials) indicating a slight over-valuation either of retained N or/andexcreted N.

CH4 originates from enteric fermentation by animal anddegradation of organic components in manure. CH4 from manureis produced under anaerobic conditions and is enhanced by hightemperature (Sommer and Møller, 2000). The production by thepigs is related to the amount of gross energy intake of which about0.6% is eliminated in that form (Schneider and Menke, 1982;Crutzen et al., 1986) and to the level of dietary fibre (Ramonet et al.,2000; Noblet and Le Goff, 2001). The low levels of CH4 emissionsobserved in these experiments (mean: 0.76 g d�1) suggest thatproduction from the 2 experimental rooms came essentially fromthe digestive tract of animals. These results with pigs on litter arein agreement with data of Nicks et al. (2003) who observed anaverage CH4 emission with 2 successive batches of weaned pigsraised either on a straw or on a sawdust deep litter of 0.91 and0.48 g d�1 pig�1respectively. Results with pigs on slatted floorwere however more lower than data from Groot Koerkamp andUenk (1997) who reported a CH4 emission in commercial farms of11 g d�1 weaner�1 kept on partially slatted floor. According to theauthors, these values were however 3 times higher than expectedaccording to previous measurements.

The CO2 emissions are linked to animal metabolism and therelationship is about 688 g CO2 generated for each 100 W of energyproduced by a pig per day (Bruce, 1981). For piglets of 15 kg (themean weight in the present trials), the thermoneutral heatproduction is estimated at 66.2 W (Bruce, 1981; Baxter, 1984)and thus the CO2 production at 456 g d�1. This value is still higherthan the mean value observed in the present study (351 g d�1). PigCO2 production (PCO2

) is however dependent of other factors thanpig weight. It depends also of the animal respiratory quotient (RQ)defined as the volume of CO2 produced divided by the volume of O2

consumed. The lowest RQ, the lowest PCO2. In weaned pigs, RQ

ranges from 0.8 to 1.1 according to the level of feed intake and a RQdecrease from 1.1 to 0.8 decreases PCO2

by 22% (Van Ouwerkerkand Pedersen, 1994). Values of PCO2

observed in this study seemassociated with low RQ values. Furthermore, concerning the effectof the floor type, manure fermentation is known to increase CO2

release (Van Ouwerkerk and Pedersen, 1994).Because these gaseous emissions according to the type of floor

were different, the impact on environment will be also different.Indeed, the warming potential of the GHG, N2O and CH4 together,can be expressed in CO2 equivalents (CO2eq) using the followingequation: CO2eq ðg d�1 pig�1Þ ¼ 21ECH4

þ 310EN2O, with ECH4and

EN2O being the emissions of CH4 and N2O (g d�1 pig�1). For N2O,indirect emissions from atmospheric deposition of N from NH3 onsoils and water surfaces have been added to the direct emissions.The indirect emissions were calculated considering an emission of0.01 kg N2O-N kg�1 emitted NH3-N (Intergovernmental Panel onClimate Change, 2006b). The CO2eq emissions were about 22, 34

and 168 g d�1 pig�1 respectively on fully slatted floor, straw deeplitter or sawdust deep litter. The difference between the groupswas mainly due to the difference in N2O emissions.

The production of H2O by animals depends both on theirweight and the ambient temperature. The minimal production at15 kg can be estimated to be 177 g H2O d�1, with the maximumvalues being about 4 times higher (Baxter, 1984). The othersources of H2O emissions are those released from the manure andfrom animals playing with the drinker. The H2O emissions were inthe mean similar in the rooms with slatted floor for the 2 trials(594 g d�1 pig�1) and were higher in the rooms with deep litter.These observations were previously reported by Philippe et al.(2007a,b) with fattening pigs. Higher H2O emissions with weanedpiglets from sawdust deep litter rooms compared with straw deeplitter rooms were also observed by Nicks et al. (2003). Waterconsumption was relatively similar between the two systems intrial 1 (6% or 100 ml d�1 pig�1 higher for animals on straw vs.slatted floor) but was 17% higher (or 300 ml d�1 pig�1) for thepiglets on sawdust compared to those on slatted floor in trial 2.This could partially explain the difference. However, significantlygreater H2O emissions in sawdust deep litter could also beexplained by the higher temperature due to fermentationobserved in this litter. As H2O emissions are higher with thedeep litter system, this system needs higher ventilation rates in‘‘winter conditions’’ when air relative humidity is the key factordetermining the ventilation rate.

CO2 and H2O emissions increased regularly from the beginningto the end of the stay of the pigs whatever the floor system. In bothtrials, these emissions were clearly lower at the beginning of thestay of the second batch than at the end of the stay of the first one,indicating the decisive influence of pig weight on the gaseousemissions. NH3 emissions from the deep litters show a similarevolution and the mean emission at the beginning of the stay of thesecond batch was, on average of the 2 trials, 78% lower than theemission at the end of the stay of the first batch. This difference washowever lower, 24% on average, with pigs on slatted floor. NH3

emissions seem thus more dependent of the manure accumulationas slurry than as deep litter. Such a difference was not observedwith CH4 emissions.

5. Conclusion

Rearing pigs on deep litter (with straw or sawdust) has a goodbrand image for the consumer and, generally, a good welfare imagefor the public. However, raising weaned pigs on deep litterproduced more NH3 than raising on slatted floor. With sawdust aslitter, there were also more N2O and CO2 emissions but less CH4.According to the warming potentials of GHG gases (N2O and CH4),raising weaned pigs on deep litter, and mainly on sawdust-baseddeep litter, emitted also more CO2eq. On the other hand, the higherNH3 emissions induced lower N content in the litter compared tothe slurry.

In conclusion, the fully slatted plastic floor system is better thanthe deep litter system to limit NH3 and CO2eq emissions related tothe raising of weaned pigs.

Acknowledgment

The research was supported by the Ministry of the WalloonRegion of Belgium.

References

Andersson, M., 1996. Performance of bedding materials in affecting ammoniaemissions from pig manure. J. Agric. Eng. Res. 65, 213–222.

Aneja, V.P., Blunden, J., James, K., Schlesinger, W.H., Knighton, R., Gilliam, W., Niyogi,D., Cole, S., 2007. Ammonia assessment from agriculture: status and national

J.-F. Cabaraux et al. / Agriculture, Ecosystems and Environment 130 (2009) 86–9292

needs. In: Monteny, G.-J., Hartung, E. (Eds.), Ammonia Emissions in Agriculture.Wageningen Academic Publishers, Wageningen, pp. 44–51.

Baxter, S., 1984. Intensive pig production–environmental management and design.Granada publishing, London, United Kingdom.

Brink, J.C., Kroeze, C., Klimont, Z., 2001. Ammonia abatement and its impact onemissions of nitrous oxide and methane. Part 1. Method. Atmos. Environ. 35,6299–6312.

Bruce, J.M., 1981. Ventilation and temperature control criteria for pigs. In: Clark, J.A.(Ed.), Environmental Aspects of Housing for Animal Production. Butterworths,London, United Kingdom, pp. 197–216.

Crutzen, P.J., Aselmann, I., Seiler, W., 1986. Methane production by domesticanimals, wild ruminants, other herbivorous fauna, and humans. Tellus B 38,271–284.

Dourmad, J.Y., Seve, B., Latimier, P., Boisen, S., Fernandez, J., van der Peet-Schwering, C.,Jongbloed, A.W., 1999. Nitrogen consumption, utilisation and losses in pig pro-duction in France, The Netherlands and Denmark. Livest. Prod. Sci. 58, 261–264.

European Commission, 2003. Integrated pollution prevention and control (IPPC).Reference document on best available techniques for intensive rearing ofpoultry and pigs. http://www.jrc.es/pub/english.cgi/0/733169 (date of access:October 3, 2008).

Groot Koerkamp, P., Uenk, G., 1997. Climatic conditions and aerial pollutants in andemissions from commercial animal production systems in the Netherlands. In:Voermans, J.A.M., Monteny, G.J. (Eds.), Proceedings of the International Sym-posium Ammonia and Odour Control from Animal Production Facilities. Vin-keloord, The Netherlands p. 139.

Intergovernmental Panel on Climate Change, 2006a. 2006 IPCC Guidelines forNational Greenhouse Gas Inventories. Vol. 1: General Guidance and Reporting.Prepared by the National Greenhouse Gas Inventories Programme. In: Eggle-ston, H.S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. (Eds.), Institute for GlobalEnvironmental Strategies, Hayama, Japan.

Intergovernmental Panel on Climate Change, 2006b. 2006 IPCC Guidelines forNational Greenhouse Gas Inventories. Vol. 4: Agriculture, Forestry and OtherLand Use. Prepared by the National Greenhouse Gas Inventories Programme. In:Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. (Eds.), Institute forGlobal Environmental Strategies, Hayama, Japan.

Intergovernmental Panel on Climate Change, 2007. Climate Change 2007: The Phy-sical Science Basis. Contribution of Working Group I to the Fourth AssessmentReport of the Intergovernmental Panel on Climate Change. In: Solomon, S., Qin, D.,Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.),Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

Kermarrec, C., 1999. Bilan et transformations de l’azote en elevage intensif de porcssur litiere. Ph.D. Thesis, Universite de Rennes 1, France.

Monteny, G.J., Bannink, A., Chadwick, D., 2006. Greenhouse gas abatement strate-gies for animal husbandry. Agric. Ecosyst. Environ. 112, 163–170.

Nicks, B., Laitat, M., Vandenheede, M., Desiron, A., Canart, B., 2000. Water vapouremission and nitrogen balance from a sawdust deep litter system for weanedpigs. Ann. Zootech. 49, 119–128.

Nicks, B., Laitat, M., Vandenheede, M., Desiron, A., Verhaeghe, C., Canart, B., 2003.Emissions of ammonia, nitrous oxide, methane, carbon dioxide and watervapour in the raising of weaned pigs on straw-based and sawdust-based deeplitters. Anim. Res. 52, 299–308.

Nicks, B., Laitat, M., Farnir, F., Vandenheede, M., Desiron, A., Verhaeghe, C., Canart, B.,2004. Gaseous emissions from deep-litter pens with straw or sawdust forfattening pigs. Anim. Sci. 78, 99–107.

Noblet, J., Le Goff, G., 2001. Effect of dietary fibre on the energy value of feeds forpigs. Anim. Feed Sci. Technol. 90, 35–52.

Philippe, F.X., Laitat, M., Canart, B., Vandenheede, M., Nicks, B., 2007a. Comparisonof ammonia and greenhouse gas emissions during the fattening of pigs, kepteither on fully slatted floor or on deep litter. Livest. Sci. 111, 144–152.

Philippe, F.X., Laitat, M., Canart, B., Vandenheede, M., Nicks, B., 2007b. Gaseousemissions during the fattening of pigs kept either on fully slatted floors or onstraw flow. Animal 1, 1515–1523.

Ramonet, Y., van Milgen, J., Dourmad, J.Y., Dubois, S., Meunier-Salaun, M.C., Noblet,J., 2000. The effect of dietary fibre on energy utilisation and partitioning of heatproduction over pregnancy in sows. Br. J. Nutr. 84, 85–94.

SAS, 1999. SAS/STAT User’s Guide: Statistics (Version 8). SAS Inst. Inc., Cary, NorthCaroline, USA.

Schneider, W., Menke, K.H., 1982. Untersuchungen uber den energetischen Futter-wert von Melasseschnitzeln in Rationen fur Schweine. Z. Tierphysiol. Tierer-naehr. Futtermittelk. 48, 233–240.

Sommer, S.G., Møller, H.B., 2000. Emission of greenhouse gases during compostingof deep litter from pig production—effect of straw content. J. Agric. Sci. 134,327–335.

Tuyttens, F.A.M., 2005. The importance of straw for pig and cattle welfare: a review.Appl. Anim. Behav. Sci. 92, 261–282.

Urbain, B., 1997. Toxicite de l’ammoniac pour l’appareil respiratoire du porc. Ph.D.Thesis, University of Liege, Belgium.

van der Peet-Schwering, C.M.C., Aarnink, A.J.A., Rom, H.B., Dourmad, J.Y., 1999.Ammonia emissions from pig houses in The Netherlands, Denmark and France.Livest. Prod. Sci. 58, 265–269.

Van Ouwerkerk, E.N.J., Pedersen, S., 1994. Application of the carbon dioxide massbalance method to evaluate ventilation rates in livestock buildings. In: Pro-ceedings of the XII CIGR World Congress on Agricultural Engineering, vol. 1,Milano, Italy, pp. 516–529.