Ammonia and greenhouse gas emissions from slatted and solid floors in dairy cattle houses: A scale model study

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Research Paper

Ammonia and greenhouse gas emissions from slattedand solid floors in dairy cattle houses: A scale model study

Jose Pereira a,e,*, David Fangueiro b, Tom H. Misselbrook c, David R. Chadwick c,Joao Coutinho d, Henrique Trindade e

aEscola Superior Agraria de Viseu, Instituto Politecnico de Viseu, Quinta da Alagoa, 3500-606 Viseu, PortugalbUIQA, Instituto Superior de Agronomia, TU Lisbon, Tapada da Ajuda, 1349-017 Lisboa, PortugalcRothamsted Research, North Wyke, Okehampton, Devon, EX20 2SB, UKdChemistry Centre, Department of Biology and Environment, Universidade de Tras-os-Montes e Alto Douro, Apartado 1013,

5001-801 Vila Real, PortugaleCITAB e Centre for the Research and Technology of Agro-Environment and Biological Sciences, Department of Agronomy,

Universidade de Tras-os-Montes e Alto Douro, 5001-801 Vila Real, Portugal

a r t i c l e i n f o

Article history:

Received 15 July 2010

Received in revised form

18 February 2011

Accepted 28 February 2011

Published online 30 March 2011

Dairy cattle are usually housed in naturally ventilated houses where removal of excreta is

periodically performed. The aim of this controlled study was to compare the effect of two

floor designs and three air temperatures (5, 15 and 25 �C) on NH3, N2O, CH4 and CO2

emissions arising from cattle excreta deposition to the floor. Two scale models were built

to simulate a level solid floor without urine drainage, and a slatted concrete floor. Following

application of a mixture of urine and faeces, these two floor type models were subjected to

a constant airflow rate (12.5 exchanges h�1) and gaseous emissions were measured over

a 72-h period.

Emissions of NH3, N2O, CO2 and CH4 increased significantly with air temperature with

both floor type models and emissions of NH3, N2O and CO2 were significantly greater from

the solid floor relative to the slatted floor at all temperatures considered. The cumulative

NH3 (27e66% of total N applied) and CO2 (<19% of total C applied) emissions were greater

from the solid floor than from the slatted floor (by 36% and 44%, respectively). The

cumulative N2O (<0.1% of total N applied) and CH4 (<0.4% of total C applied) emissions

were relatively low and CH4 values did not differ significantly between treatments.

Cumulative greenhouse gas emissions (as CO2-equivalents) increased significantly with

temperature but did not differ between the floor types.

ª 2011 IAgrE. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Livestock production is the major source of gaseous emissions

from agriculture (Snell, Seipelt, & van denWeghe, 2003) and, in

Portugal, dairy farming is the main source of ammonia (NH3)

emission (PNIR, 2009). Ammonia emissions lead to acidification

and nutrient-N enrichment of ecosystems and can also be

consideredasa secondaryparticulatepollutantassociatedwith

health hazards (Erisman, Bleeker, Galloway, & Sutton, 2007).

Dairy farming isalsoa sourceofother environmentallyharmful

* Corresponding author. Escola Superior Agraria de Viseu, Instituto Politecnico de Viseu, Quinta da Alagoa, 3500-606 Viseu, Portugal.Tel.: þ351 232 446 600; fax: þ351 232 426 536.

E-mail addresses: [email protected], [email protected] (J. Pereira).

Avai lab le at www.sc iencedi rect .com

journa l homepage : www.e lsev i er . com/ locate / i ssn /15375110

b i o s y s t em s e n g i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7

1537-5110/$ e see front matter ª 2011 IAgrE. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.biosystemseng.2011.02.011

Author's personal copy

gases, suchasnitrousoxide (N2O)andmethane (CH4), twoof the

main greenhouse gases (GHG) contributing to global warming

(IPCC, 2007). Ammonia is not considered to be a direct GHG but

indirectly contributes to global warming since deposited NH3

may be converted to N2O in the soil after nitrification and

subsequent denitrification (IPCC, 2007).

Excreta (urine and faeces) deposited in cattle houses results

in significant gaseous emissions.Ammonia emissions induced

by the hydrolysis of urea present in the urine (Ni, Vinckier,

Hendriks, & Coenegrachts, 1999; Sommer et al., 2006) depend

onseveral parameters, suchas theair velocityover themanure

surface, rate of urea hydrolysis, pH of excreta and air temper-

ature (Cortus, Lemay, Barber, Hill, & Godbout, 2008; Elzing &

Monteny, 1997a; Morsing, Strøm, Zhang, & Kai, 2008). The

formation of ammoniacal N can subsequently lead to N2O

emission from nitrification and/or denitrification processes

(Misselbrook, Webb, Chadwick, Ellis, & Pain, 2001). Methane

may be emitted due the existence of anaerobic conditions in

excreta deposited on floors (Misselbrook et al., 2001) and

carbon dioxide (CO2) should be predominantly a product of

aerobic decomposition of organic compounds in slurry based

cattle buildings (i.e. in open ventilated systems), and will also

result from urea hydrolysis (Møller, Sommer, & Ahring, 2004).

In Portugal, dairy cattle are usually housed in naturally

ventilated buildings with slatted or solid floors on which urine

and faeces remain for a period of a few hours to several days

(Pereira, Misselbrook, Chadwick, Coutinho, & Trindade, 2010).

Previous studies have shown that deposited excreta may lead

to significant emissions of N2O, CO2, CH4 and NH3

(Misselbrook et al., 2001; Snell et al., 2003) and that increasing

temperature leads to increases in NH3 (Elzing & Monteny,

1997a), N2O (Sommer et al., 2006), CO2 (Ni et al., 1999) and

CH4 emissions (Kashyap, Dadhich, & Sharma, 2003). The

emissions of these four gases from cattle buildings have been

previously studied in countries fromNorthern Europe (Braam,

Smits, Gunnink, & Swierstra, 1997; Braam & Swierstra, 1999;

Ngwabie, Jeppsson, Nimmermark, Swensson, & Gustafsson,

2009; Ni et al., 1999; Snell et al., 2003). However, there is

a lack of information for gaseous emissions from cattle

housing in Mediterranean countries, where climatic condi-

tions are significantly different to those of Northern Europe

and where it may be hypothesised that higher air tempera-

tures will contribute to increased gaseous emissions.

Previous studies (e.g. Braam et al., 1997; Morsing et al., 2008;

Swierstra, Smits, & Kroodsma, 1995) have shown that floor

design,namely slattedversus solidfloors, has a strong influence

on NH3 emissions. Swierstra et al. (1995) observed that a cubicle

house with solid concrete floors with a 3% slope to the centre,

a central gutter andno finish gave a reduction of 48% of theNH3

emission compared to an identical cow house with a standard

slatted floor. Furthermore, Braam et al. (1997) reported that

scrapinganon-slopingconcretefloorwillhave littleeffectonthe

NH3 emission because a thin layer of slurry is retained by the

floor,whichwillbeasignificantsourceofNH3.ThepotentialNH3

emission from a slatted floor was estimated to be three times

higher than for a V-shaped solid floor (Monteny & Erisman,

1998). However, the effect of floor design on GHG emissions

has been poorly studied. Furthermore, results obtained in some

previousstudiesdidnotdifferentiateemissionsoriginating from

animal respiration (CO2), enteric fermentation (CH4) or excreta

deposition on the floor (NH3, N2O, CO2 and CH4) (Cabaraux,

Philippe, Laitat, & Canart, 2009; Ngwabie et al., 2009; Ni et al.,

1999). Additionally, results from in-situ measurements of

gaseous emissions in cattle housingmaybe affected by building

ventilation (temperature variation andnon-constant flux of air).

The aim of this present study was to evaluate, using a scale

model, the effect of floor design (solid or slatted floor) under

different controlled indoor air temperatures on NH3 and GHG

(N2O, CH4 and CO2) emissions from cattle excreta deposition.

2. Materials and methods

2.1. Experimental set-up

2.1.1. Scale model of concrete floorsScale models of two types of concrete floor e slatted and solid

(the latter with no slope or urine drainage), were built. The

Nomenclature

Symbols

A Exposed surface area of the floor, m2

CO2-equivalents Cumulative N2O and/or CH4 emissions

expressed as CO2 using the conversion factors,

g m�2 or %

F Ammonia emission rates, mg m�2 h�1

NH4þ Ammonium, g kg�1

NO3� Nitrate, mg kg�1

t Time lenght of the sampling period, h

t0 Sample of urea solution taken at 0 min

t1 Sample urea solution taken at 30 min

TAN Total ammoniacal N concentration, mg l�1 or

g kg�1

V Volume of the acid trap solution, l

Abbreviations

ANOVA Analysis of variance

DM Dry matter

EMEP-CORINIAR Air pollutant emission inventory

guidebook

EN European normalization

GC Gas chromatography

GHG Greenhouse gas emissions

N Number of replications

ND Not determined

NIR Near-infrared detection

PVC Polyvinyl chloride

TGA Trace gas analyser

TOC Total organic carbon

Total N emissions Total cumulative N (NH3 þ N2O)

emissions

Total C emissions Total cumulative C (CO2 þ CH4)

emissions

b i o s y s t em s e ng i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7 149

Author's personal copy

floors were provided by a commercial company (SOCIDIAS -

Estruturas e Revestimentos Metalicos, Lda, Portugal). Each

model had an experimental area of 1.0 m � 1.0 m and the

slatted floor was equipped with a drainage channel, as would

be found in a dairy house with a slatted floor. The excess of

excreta applied to the slatted floor drained directly to the

drainage channel located under the slats and was retained

there. Details of the scale models are shown in Fig. 1. Each

scale model was covered with hermetic PVC containers

(1.0 m � 1.0 m � 1.0 m) leaving a headspace of 115 l above the

floors. The internal surfaces of the containers were sprayed

with a Teflon coating (Loctite�, provided by Fisher Scientific,

Loures, Portugal) to minimise adsorption of NH3 on container

walls. Airflow through the headspace of each container was

achieved using six individual pumps, with a 4 lmin�1 flow rate

regulated by a needle valve coupled to a flowmeter (model GD

100; KDG-Mobrey, Crawley, West Sussex, UK), located before

each pump. The exact volumetric flow was measured using

a gas meter (model Gallus 2000 G1.6; Schlumberger Elaborate,

Reims, France) located after each pump. Testswere performed

prior to experimental measurements to check the hermetic

condition of the systems. The air flow through the scale

models was equivalent to 12.5 head space exchanges h�1,

within the range of typical air exchange rates for a naturally

ventilated dairy house of between 4 and 15 exchanges h�1

(Snell et al., 2003). In our study, such air exchange rate cor-

responded to an air velocity (60 mm above concrete floors) of

4 mm s�1.

2.1.2. Experimental model preparationOver a period of one year prior to experiments, the model

slatted and solid floors were installed in the walking area of

a cattle house in a typical dairy farm from NW Portugal. This

procedure allowed the development of the urease activity over

that period. Immediately before the beginning of the experi-

ments, the two floors were removed from the farm and then

carefully washed with water to remove the deposited excreta.

Later, the floors were inserted in the respective scale model.

At the start of the experiments the urease activity was

measured in triplicate for both floors at 15 �C, using the

procedure described in detail by Braam and Swierstra (1999).

Briefly, 50 ml of a urea solution (10 g [N] l�1) was incubated in

a PVC cylinder (inside diameter ¼ 80 mm, deep ¼ 60 mm) on

each floor for 30 min. A 10 ml sample was taken at t0 (0 min)

and t1 (30 min) for analysis of ammoniacal N (NH4þ þ NH3)

content using automated segmented-flow (SanPlus, Skalar,

Breda, The Netherlands) molecular absorption spectropho-

tometry (Houba, Van der Lee, & Novozamsky, 1995). Urease

activity was expressed as the increase in total ammoniacal N

(TAN) relative to the initial content.

2.2. Urine and faeces samples

One month before the start of the experiments, urine and

faeces were collected separately from a subgroup of 10

lactating cows randomly selected from cows housed in

a typical commercial dairy farm located in NW Portugal. Urine

Fig. 1 e Schematic plan of the slatted (left) and solid floors (right) of the scale models (dimensions m).

b i o s y s t em s e n g i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7150

Author's personal copy

and faeces were collected directly under the tail of the

lactating cows using plastic containers. During the entire

collecting period (96-h), the urine and faeces were stored in

plastic containers at 4 �C. Composite samples, respectively, of

urine and faeces were created, and immediately frozen until

required. Lactating cows were fed with a diet of maize silage,

straw and concentrates with 19% crude protein and had

a mean milk production of 29.3 kg cow�1 day�1.

The composite urine and faecal samples were analysed by

standard laboratory methods to assess the physico-chemical

properties. Urine pH values were determined directly and

faecal pH values were determined after 2-h of occasional

agitation in a faeces/deionised water suspension (1:5 w/v),

according to the European standard EN 13037 (ES, 1999). Dry

matter (DM) content of the faeces was determined by drying

to a constant weight at 102 �C. Total C of the faeces was

determined by dry combustion followed by near-infrared

detection (NIR) using an elemental TOC analyser (Primac,

Skalar, Breda, The Netherlands). The total N content of the

urine and faeces was determined using a modified Kjeldahl

method (Novozamsky, Houba, Van Eck, & Van Vark, 1983).

Mineral N content of the urine and faeces was extracted with

2 M KCl in a 1:10 (excreta:extractant) ratio (Mulvaney, 1996).

Total ammoniacal N (NH4þ-N þNH3eN) and NO3

� contents of

the extracts were determined by automated segmented-flow

molecular absorption spectrophotometry (Houba et al., 1995).

The segmented flow analyser (SanPlus, Skalar, Breda, The

Netherlands) was equipped with dialysers to prevent inter-

ferences from colour or suspended solid particles in the

extracts. Urea N content of the urine was determined by

reacting with diacetyl monoxime in the presence of thio-

semicarbazide (to intensify the colour) in acid conditions and

reading the absorbance at 525 nm (Sullivan & Havlin, 1991).

2.3. Measurement of ammonia and greenhouse gasemissions

The two scale models were housed in a large room with

controlled climate (climatic room) (model EVK 211, EVCO SPA,

Belluno, Italy). A mixture of 0.8 l of urine and 1.2 kg of faeces

was homogenised and spread uniformly over each floor type

leading to a 2 mm layer. Since the solid floor was not sloped,

the urine did not drain off. For the slatted floor, the mixture of

urine/faeces was applied using a modified bricklayer’s trowel

to carefully spread the mixture above the slat area while

excess mixture drained to the floor of the channel below the

slats. The urine/faeces ratio of 2:3 corresponds to the ratio

produced by dairy cows (Morse, Nordstedt, Head, & Van Horn,

1994). The quantity of mixture spread over the floors was

based on an average stocking rate of one livestock unit per

6.1 m2 within the house and took into account the mean

cleaning intervals that are usually made in a typical Portu-

guese cattle house, as observed in a previous study made by

Pereira et al. (2010).

This procedure was performed in triplicate at three

different temperatures: 5, 15 and 25 �C ( � 0.5 �C). Prior to

application, sub-samples of urine and faeceswere thawed and

their temperatures increased to that of the experimental

conditions (5, 15 or 25 �C). Immediately after the urine/faeces

application to the floor, each scale model was closed and

hermetically sealed and the airflow through the system star-

ted. The outlet air from the two scale models was exhausted

out of the climatic room.

Ammonia and GHG emissions were measured for 72-h

after excreta application. For the measurement of the NH3

emissions, acid traps (containing 150ml of H3PO4 0.02 M) were

used with exposure periods of 1-h. Six acid traps (4 l min�1

each) were placed before the scale model to remove NH3 from

the inlet air. A sub sample of the outlet air was drawn

(4 l min�1) through a second acid trap to collect the NH3

emitted inside the scale model during the measurement

period. Details of the system used for NH3 measurement are

shown in Fig. 2. The ammoniacal N content of the solution

was analysed by automated segmented-flow molecular

absorption spectrophotometry (Houba et al., 1995). In order to

estimate the daily flux, NH3 emissions were measured

continuously during the first 24-h and then 20 h day�1 in the

next 48-h. Ammonia emission rates (F, mg [N] m�2 h�1) for

each sampling period were calculated as Eq. (1):

F ¼ ½TAN�VA t

(1)

Where, [TAN] was the total ammoniacal N concentration of

the acid trap (mg l�1), V the volume of acid trap solution (l), A

Fig. 2 e Schematic plan of the system used for gaseous emission measurement in the scale models.

b i o s y s t em s e ng i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7 151

Author's personal copy

the exposed surface area of the floor (m2), and t the duration of

the sampling period (h). The total emission for the period (mg

[N]) was determined as 6 � [TAN] � V (accounting for the total

volumetric flow of air through the chamber being 6 times that

of the volume sub-sampled), and total NH3 emission for the

duration of the experiment (72-h) was derived by summing

emissions for each sampling period. In the periods that NH3

wasmeasured for only 20 h day�1, gas losses for the remaining

period (4 h day�1) were derived by interpolation.

The fluxes of N2O, CO2 and CH4 were measured directly

through a sampling point located immediately before and

after the scale model (Fig. 2), with a trace gas analyser (TGA)

(1412 Photoacoustic Field Gas-Monitor, Innova AirTech

Instruments, Ballerup, Denmark), equipped with internal

filters for particles and water and optical filters for N2O (filter

type UA0985), CO2 (UA0982) and CH4 (UA0969). The TGA was

run manually in a mode including corrections for cross-

interferences between CO2 and N2O and between CH4 and

N2O. The detection limits were 0.40 and 1.50 ppm for N2O and

CH4, respectively. The concentrations of these gases were

measured in each replication at 0, 0.5, 1 and every 2-h during

the first 24-h and every 6-h in the following 48-h. Cumulative

emissions of N2O, CO2 and CH4 from each chamber were

determined by averaging the flux between two sampling

occasions and multiplying by the time interval between

sampling occasions.

To assess the effects of the treatments (temperature and

floor type) on total GHG emissions, N2O and CH4 emissions

were converted to CO2-equivalents using the conversion

factors of 298 and 25 for N2O and CH4, respectively (Forster

et al., 2007). At the end of the experiment (72-h), samples of

the mixture that remained on the solid floor or slats and also

the mixture deposited in the channel under the slatted floor

were collected. These sampleswere analysed for pH, DM, TAN

and NO3� contents. Before the beginning of a new replication,

the excreta mixture remaining on the models was scraped

manually and each model carefully cleaned with deionised

water.

2.4. Statistical analysis

The experimental runs were carried out in triplicate. Results

were analysed by ANOVA, considering the time after excreta

deposition as a split factor over the two experimental factors

(temperatures and floors), and Tukey testswere carried out for

comparison of means between treatments and their interac-

tions. The statistical software package used was STATISTIX

7.0 (Analytical Software, Tallahassee, USA). Statistical differ-

ences, referred to in the text as significant, correspond to

P < 0.05 unless otherwise stated.

3. Results

Initial characteristics of the urine and faeces used in the

experiments are given in Table 1. No NO3�-N was determined

in either urine or faecal samples. Initial values of urease

activity for the slatted (1.35 g [NH4þ-N] m�2 h�1) and solid floor

(1.40 g [NH4þ-N] m�2 h�1) were not significantly different and

were in the range (1.0e2.4 g [NH4þ-N] m�2 h�1) reported by

Leinker, Reinhardt-Hanisch, von Borell, and Hartung (2007).

3.1. Nitrogen emissions

3.1.1. NH3 emissionsThe increase in temperature induced a significant increase in

NH3 flux from both floor types (Fig. 3a). At all temperatures,

NH3 flux increased significantly immediately after excreta

deposition on the floors, peaking within the first 24-h, fol-

lowed by a progressive decrease until the end of the experi-

ment (72-h). Ammonia fluxes peaked at different times

depending on the temperature considered, with an inverse

correlation between temperature and peak time. At 5 and

15 �C, NH3 flux peaked between 3 and 12-h, while at 25 �C flux

peaked early, between 1 and 6-h after deposition (Fig. 3a).

There was a significant interaction (P < 0.001) between floor

type and time after excreta deposition on NH3 flux, with peak

emission rate earlier from the slatted floor (Fig. 3a). There was

no significant interaction between temperature and floor type

on NH3 flux, although at 15 and 25 �C, up to 6-h after deposi-

tion of the excreta until the end of the experiment, the abso-

lute values of NH3 fluxeswere numerically greater for the solid

floor than the slatted floor.

The increase in temperature from 5 to 15 �C, and from 15 to

25 �C, led to higher cumulative (72-h) NH3 emissions of about 25

and 45%, respectively, for both floor types. Increasing temper-

ature from5to25 �C increased thecumulativeNH3emissionsby

70and104%for theslattedandsolidfloor, respectively (Table 2).

At 5 �C, total NH3 emission represented about 30% of the total N

contained in excreta. However, at 15 and 25 �C cumulative

emission of NH3 varied between 32 and 47% of the total N

deposited on the slatted floor and between 43 and 66% for the

solidfloor (Table2).Thesedifferencesare inagreementwith the

significantly higher TAN contents observed in the excreta 72-h

after deposition on the floors at 5 �C relative to that at the other

temperatures (Table 3). Cumulative NH3 emissions from the

solid floor were significantly higher (about 36%) than from the

slatted floor, with no significant interaction between tempera-

ture and floor type (Table 2).

3.1.2. N2O emissionsEmissions of N2O were low throughout the experiment.

However, a significant effect of temperature, time after

excreta deposition, and the interaction between these two

Table 1 e Composition of urine (on a fresh weight basis)and faeces (on a dry matter basis) used in the study,collected from lactating cows.

Parameters Urine Faeces

pH 8.14 (0.02) 7.58 (0.04)

Dry matter (%) ND 14.43 (0.18)

Total C (g kg�1) ND 550.6 (21.8)

Total N (g kg�1) 7.04 (0.49) 13.35 (1.77)

Urea-N (g l�1) 6.35 (0.18) ND

TAN (g [N] kg�1) 0.13 (0.00) 1.98 (0.10)

Values between parentheses represent standard error of the mean

(N ¼ 8). ND ¼ Not determined.

b i o s y s t em s e n g i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7152

Author's personal copy

factors was observed on N2O fluxes. Fluxes at 25 �C were

significantly higher than those at 5 or 15 �C, except during the

first hour. At all temperatures, N2O fluxes peaked 1-h after

excreta deposition and then decreased significantly over the

72-h (Fig. 3b). There was a significant effect of floor type on

N2O flux, with higher fluxes from the solid floor. In addition,

there was a significant interaction (P < 0.001) between floor

type and time after excreta deposition, but not between

temperature and floor type, (although at 15 and 25 �C, fluxes ofN2O were numerically higher from the solid floor relative to

the slatted floor (Fig. 3b)).

Cumulative N2O emissions, while being very low, were

significantly influenced by temperature. The increase in

temperature from 5 to 15 �C, and from 15 to 25 �C, increasedcumulative (72-h) N2O emissions for both floor types by about

142 and 89%, respectively. There was also a significant inter-

action between floor type and temperature (P< 0.001), with no

significantdifference in cumulativeN2Oemissionsover 72-hat

5 �C, while at 15 and 25 �C cumulative emissions were signifi-

cantly higher from the solid floor (by ca. 185%) (see Table 2).

3.1.3. Total N emissionsAmmonia emissions represented >99% of total cumulative N

(NH3 þ N2O) emissions (Table 2), therefore the treatment

effects for N lost as NH3 and N2O emissions were the same as

described in Section 3.1.1 for NH3 emissions.

Fig. 3 e Average fluxes of NH3 (A), N2O (B), CO2 (C) and CH4 (D) following excreta deposition on concrete floors. Vertical bars

represent standard error of the mean (N [ 3).

Table 2 e Cumulative emissions of NH3 and N2O following excreta deposition on concrete floors (N [ 3).

Floors Temp.(�C)

NH3 emissions N2O emissions NH3 þ N2O emissions

mgN m�2

As % urea-Napplied

As % totalN applied

mgN m�2

As % urea-Napplied

As % totalN applied

mgN m�2

As % urea-Napplied

As % totalN applied

Slatted 5 2140 (50) 42 (1) 27 (1) 0.86 (0.09) 0.00 (0.00) 0.0 (0.0) 2141 (50) 42 (1) 27 (1)

Solid 5 2560 (11) 50 (0) 32 (0) 0.94 (0.22) 0.02 (0.00) 0.0 (0.0) 2561 (11) 50 (0) 32 (0)

Slatted 15 2570 (330) 50 (6) 32 (4) 2.13 (0.22) 0.04 (0.00) 0.0 (0.0) 2572 (329) 51 (6) 32 (4)

Solid 15 3390 (593) 67 (12) 43 (7) 4.16 (0.19) 0.08 (0.00) 0.1 (0.0) 3394 (593) 67 (12) 43 (7)

Slatted 25 3720 (11) 73 (0) 47 (0) 3.93 (0.14) 0.08 (0.00) 0.0 (0.0) 3724 (11) 73 (0) 47 (0)

Solid 25 5230 (32) 103 (1) 66 (0) 6.87 (0.07) 0.14 (0.00) 0.1 (0.0) 5237 (32) 103 (1) 66 (0)

PTemp <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05

PFloor <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05

PTemp � Floor >0.05 >0.05 >0.05 <0.05 <0.05 >0.05 >0.05 >0.05 >0.05

Values between parentheses represent standard error of the mean.

b i o s y s t em s e ng i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7 153

Author's personal copy

3.2. Carbon emissions

3.2.1. CO2 emissionsThe fluxes of CO2 emitted from both floor types increased

significantly with the temperature, and decreased with time

after excreta deposition. In addition, results showed signifi-

cant effects of floor type and a significant interaction between

floor type and time after excreta deposition (P < 0.001), but

there was no significant interaction between temperature and

floor type. Hence, at all temperatures studied, CO2 emissions

peaked between 0 and 1-h after application, followed by

a progressive decrease until the end of the experiment. Fluxes

of CO2 were always significantly higher from the solid floor

than the slatted floor (Fig. 3c).

The increases in temperature from 5 to 15 �C and 15e25 �Cled to increases in cumulative CO2 emissions of about 60 and

50%, respectively, and the increase in temperature from 5 to

25 �C led to an increase in cumulative CO2 emissions of 140%.

Cumulative CO2 emissions represented less than 19% of total

C deposited in the excreta (Table 4). Cumulative CO2 emis-

sions were significantly affected by temperature and floor

type as well as the interaction between these two factors. At

5 �C, the cumulative CO2 emissions did not differ significantly

between the two types of floors. However, at 15 and 25 �C,cumulative CO2 emission from the solid floor was about 44%

higher than from the slatted floor (Table 4).

3.2.2. CH4 emissionsMethane fluxes were generally low throughout the experi-

ment. Fluxes of CH4 were affected by temperature, time after

excreta deposition, and the interaction between these two

factors was significant (P < 0.001). Floor type did not signifi-

cantly influence CH4 fluxes and there was no significant

interaction between temperature and floor type. Hence, at 15

and 25 �C, CH4 emissions peaked between 3 and 6-h after

deposition of the excreta, but, at 5 �C, CH4 emissions were

below the analyser detection limit (1.50 ppm) (Fig. 3d).

Cumulative CH4 emissions were significantly higher at

25 �C relative to others temperatures studied. Cumulative CH4

emissions represented less than 0.4% of total C deposited on

the floors (Table 4).

3.2.3. Total C emissionsTotal C (CO2 þ CH4) emissions were dominated by CO2 emis-

sions, so treatment effects for total C emissionswere the same

as described in Section 3.2.1 for CO2 emissions.

3.3. Greenhouse gas emissions

Greenhouse gases (N2O and CH4) emissions, expressed as

CO2-equivalents, increased significantly with temperature.

A greater increase in GHG emissions was observed when

temperature increased from 5 to 15 �C (14.8 times) than from

15 to 25 �C (2.3 times). Greenhouse gas emissions did not differ

significantly between the two types of floor and there was no

significant interaction between temperature and floor type. At

15 and 25 �C for both floor types, N2O emissions (as CO2-

equivalents) were the dominant GHG (>50%). However, at 5 �C,CH4 emissions represented 100% of the total GHG emissions

for both floor types (Table 5).

4. Discussion

4.1. Nitrogen emissions

The pattern of NH3 emission, with a peak within the first 24-h

following excreta deposition, is typical of those observed by

others (e.g. Elzing & Monteny, 1997a). The low NH3 emissions

Table 3 e Excreta characteristics 72-h after deposition onconcrete floors (TAN on a fresh weight basis) (N [ 3).

Floors Temp.(�C) pH Drymatter(%)

TAN(g [N] kg�1)

Slatted (slats) 5 8.70d 14.7cd 0.88b

Slatted (channel) 5 8.62d 11.6d 1.68a

Solid 5 9.05abc 26.3abc 0.62bc

Slatted (slats) 15 9.01bc 22.5bcd 0.34d

Slatted (channel) 15 8.92c 11.2d 0.34d

Solid 15 9.06abc 28.0ab 0.41cd

Slatted (slats) 25 9.25a 25.2abc 0.22d

Slatted (channel) 25 9.08abc 17.2bcd 0.18d

Solid 25 9.20ab 37.1a 0.34d

Values with different superscripts within columns are significantly

different (P < 0.05).

Table 4 e Cumulative emissions of CO2 and CH4 following excreta deposition on concrete floors (N [ 3).

Floors Temp.(�C) CO2 emissions CH4 emissions CO2 þ CH4 emissions

mg C m�2 As % totalC applied

mg C m�2 As % totalC applied

mg C m�2 As % totalC applied

Slatted 5 5200 (241) 5 (0) 0 (0) 0.0 (0.0) 5200 (241) 5 (0)

Solid 5 7700 (6) 8 (0) 0 (0) 0.0 (0.0) 7700 (6) 8 (0)

Slatted 15 8600 (954) 9 (1) 143 (21) 0.2 (0.0) 8743 (960) 9 (1)

Solid 15 12300 (1007) 13 (1) 138 (17) 0.1 (0.0) 12438 (1008) 13 (1)

Slatted 25 12600 (434) 13 (0) 358 (93) 0.4 (0.1) 12958 (341) 14 (0)

Solid 25 18300 (357) 19 (0) 398 (106) 0.4 (0.1) 18698 (463) 20 (0)

PTemp <0.05 <0.05 <0.05 <0.05 <0.05 <0.05

PFloor <0.05 <0.05 >0.05 >0.05 <0.05 <0.05

PTemp � Floor <0.05 <0.05 >0.05 >0.05 <0.05 <0.05

Values between parentheses represent standard error of the mean.

b i o s y s t em s e n g i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7154

Author's personal copy

observed fromboth floor types at 5 �Cduring the first fewhours

after application of excreta (Fig. 3a) were probably due to the

inactivation of the urease enzyme present on the floors and/or

reduction in thehydrolysis of theurea (Braametal., 1997;Moyo,

Kissel, & Cabrera, 1989). Urine is generally acknowledged to be

the predominant source of NH3 emissions from cattle excreta

(Petersen, Sommer, Aaes, & Søeggard, 1998; Saarijarvi, Mattila,

& Virkajarvi, 2006). The increase in NH3 emission with

temperature is related to two processes: the formation of NH4þ

in aqueousphase (fromureahydrolysis), and the release ofNH3

to the gaseous phase. The NH4þ formation increases with the

increase in urease activity between 10 and 40 �C, whereas at

temperatures lower than 10 �C, urease activity is reduced

markedly (Sommer et al., 2006). The NH3 release from the

aqueous to gaseous phase is controlled by chemical and phys-

ical factors influencing diffusion, dissociation, evaporation and

volatilisation, and will increase with increasing temperature

(Elzing & Monteny, 1997b).

Ammonia emissions from floor surfaces are mainly influ-

enced by (i) the characteristics of the surface layer of the floor

affecting the urea hydrolysis rate due to the contact area

between the urine and the enzyme urease present on floor,

and (ii) by the characteristics of the emitting layer, particularly

the overall emitting surface area of urine remaining on the

floor (Elzing & Monteny, 1997a; Hamelin, Godbout, Theriault,

& Lemay, 2010). In our study, the surface area of contact for

the solid floor was 22% higher than for the slatted floor,

contributing to the significantly higher NH3 emissions at 15

and 25 �C (Fig. 3A and Table 2). For the slatted floor, significant

drainage of urine to the below-floor channel may also have

occurred prior to hydrolysis, leading to lower emissions

(Cortus et al., 2008; Hamelin et al., 2010). The differences

between the texture and porosity of the slatted and solid floors

could also have led to different infiltration rates and absorp-

tion of the urea in the concrete pores of each floor (Cortus

et al., 2008). The porosity and surface roughness of the

slatted floor was decreased by a compressive stress during

construction relative to solid floor. Nevertheless, this should

have not been a limiting factor considering the values of

urease activity (>1.0) determined here for the two floor

surfaces (Braam & Swierstra, 1999).

The increase in cumulative N2O emissions with tempera-

ture is probably related to the enhanced activity of nitrifying

bacteria at higher temperatures (Sommer et al., 2006).

However, emissions of N2O were small even if they were

significantly different from zero, with cumulative emissions

representing less than 0.1% of the total N deposited on the

floors in all treatments (Table 2). These low N2O emissions

were probably due to the fact that themain source of N2O was

the nitrification process since no NO3� was initially present in

urine and faeces deposited on the floors and the aerobic

conditions of the experiment did not favour the denitrification

processes. Similarly, previous studies have shown that in

cattle houses (Ngwabie et al., 2009) and outdoor concrete

yards (Misselbrook et al., 2001) with systems of liquid manure

management and daily or frequent removal of the excreta to

outdoor stores, N2O emissions are very low or negligible.

The lower N2O emissions observed from the slatted floor

could be related to the lower surface area,which promoted the

drainage of urine from the slats thereby reducing the avail-

ability of NH4þ for nitrification/denitrification processes.

4.2. Carbon emissions

Carbon dioxide emitted by excreta deposition does not

contribute to the greenhouse effect because it is part of the so-

called short C cycle. However, rates of CO2 emission give an

indication of biological activity and the rate at which the

decomposition processes are occurring. The CO2 emissions

that occurred during the 72-h of experiment had a similar

pattern toNH3 emissions. However, CO2 emissionsweremuch

higher, and peaked earlier, than NH3 emissions because the

CO2 was released more quickly from excreta deposited on

floors due to the lower solubility of the CO2 (Ni et al., 1999).

Since most of the C content of excreta is derived from faeces

rather than urine (Sommer et al., 2006), processes such as

organic matter decomposition and bacterial respiration under

aerobic conditions are likely to be the main source of CO2

emissions,withsomeemissionsderiving fromureahydrolysis.

As biological processes, these will be highly influenced by

temperature, in the sameway as urease activity (Fig. 3a and c).

In the present study, CH4 emissions were very low

(Table 4), as might be expected for a thin film of excreta where

generally aerobic conditions will exist. Misselbrook et al.

(2001), measuring from a concrete yard used by cattle, repor-

ted that significant CH4 emissions were associated with the

presence of dung pats, within which anaerobic conditions

may occur, with little or no emission from other areas. In the

present study, CH4 emissions were not detected at 5 �C,because of the detection limits of the gas analyser being used.

Misselbrook et al. (2001), usingmore sensitive GC analysis, did

report low CH4 emission rates under the winter conditions of

the UK.

4.3. Practical considerations regarding floor type

In this study, NH3 emissions were higher from the simulated

solid floor (level without urine drainage) compared to the

slatted floor at all temperatures considered. In addition, these

two floors did not differ significantly in terms of GHG emis-

sions. Hence, in terms of gaseous emissions, slatted floor use

Table 5 e Cumulative greenhouse gas emissions fromexcreta deposited on concrete floors (N [ 3).

Floors Temp. (�C) GHG emissions

a g CO2-eq m�2 b N2O (%) b CH4 (%)

Slatted 5 0.1 (0.1) 100 (0) 0 (0)

Solid 5 0.9 (0.2) 100 (0) 0 (0)

Slatted 15 6.7 (0.9) 29 (1) 71 (1)

Solid 15 8.5 (0.4) 46 (4) 54 (4)

Slatted 25 15.6 (3.2) 24 (5) 76 (5)

Solid 25 19.7 (3.6) 33 (6) 67 (6)

PTemp <0.05 <0.05 <0.05

PFloor >0.05 <0.05 <0.05

PTemp � Floor >0.05 >0.05 >0.05

Values between parentheses represent standard error of the mean.

a Cumulative (72-h) GHG emissions expressed in CO2-equivalents

m�2

b Percentage of total emission of GHG as N2O or CH4.

b i o s y s t em s e ng i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7 155

Author's personal copy

would be recommended for dairy cattle houses. However,

there are other considerations to take into account when

selecting floor type for a cattle house, including animal health

andwelfare, slurry removal management and cost. Compared

to a solid floor, a slatted floor has the following characteristics:

(i) a potential increase in cattle foot problems (Hinterhofer,

Ferguson, Apprich, Haider, & Stanek, 2006); (ii) animals tend

to remain cleaner (Trevisi, Bionaz, Piccioli-Cappelli, & Bertoni,

2006); (iii) improved slurry removal via an under floor col-

lecting pit or channel below the slats, although emissions will

occur from this channel (Sommer et al., 2006); and (iv) more

expensive. However, a cattle house with a solid floor needs

cleaning more frequently and the presence of excessive

moisture on the floor could enhance foot diseases and the

development of mastitis (Trevisi et al., 2006). Hence, from the

animal health and welfare point of view it may be preferable

to use a solid floor than a slatted floor in cattle houses. An

alternative to both floor designs may be the use of a central

gutter in a double-sloped solid floor that reduces NH3 emis-

sions by ca. 50% relative to a slatted floor (Braamet al., 1997), is

less expensive than a slatted floor and address the animal

health and welfare point of view.

No significant differences were observed in the present

study in terms of cumulative N emissions until 6-h after

deposition of the excreta on the two floors. After this period,

cumulative N emissions increased significantly with temper-

ature, which suggests that the floors should be cleaned at

shorter time intervals. In a controlled study, scraping solid

concrete areas at 2-h and 6-h after excreta deposition was

associated with reductions in NH3 emissions of 49 and 27%,

respectively (Misselbrook, Pain, & Headon, 1998). However,

more frequent scraping of the excreta from a level solid floor

may not reduce NH3 emissions as much as expected because

a thin emitting layer remains above the floor (Braam et al.,

1997; Kroodsma, Huis in ’t Veld, & Scholtens, 1993) and

measurements on commercial farms have shown scraping to

be less effective in reducing emissions (Misselbrook et al.,

1998; Misselbrook, Webb, & Gilhespy, 2006). Cleaning floors

by washing/hosing is more effective than scraping, but has

the disadvantage of increasing the amount of slurry at farm

level (Kroodsma et al., 1993; Misselbrook et al., 1998, 2006).

Caution should be exercised in extrapolating the results

from this controlled study to full scale cattle housing, where

distribution of urine and faeces are likely to be more hetero-

geneous and actual air movements and ventilation rates will

influence emissions (Elzing & Monteny, 1997a; Morsing et al.,

2008; Groenestein, den Hartog, & Metz, 2006; Sommer et al.,

2006). Nevertheless, our experimental set-up allows a con-

trolled comparison of the fluxes from the two floor types.

The current Portuguese National NH3 emissions inventory

employs one generic emission factor for all types of cattle

houses (with solid and/or slatted floors) obtained from EMEP-

CORINIAR (0.12 kg NH3eN per kg N excreted in buildings)

(PNIR, 2009). Since the mean air temperature in Portugal is

about 16 �C and varies between 17 and 28 �C during the

warmest month, the present laboratory study suggests that

significantly greater NH3 emissions would be expected from

cattle houses with solid floors (level, without urine drainage)

than for houses with slatted floors. However, recent

measurements of NH3 emissions from dairy cattle houses in

NW Portugal with solid and/or slatted floors did not find floor

type to be a main factor for differences between housing

(Pereira et al., 2010), although the number of farms studied

was limited. Further on-farm studies are required for condi-

tions in Portugal (and also for other Mediterranean countries)

to assess if the cattle houses with solid or concrete floors

should have different NH3 emission factors.

5. Conclusions

Based on measurements using scale models under controlled

conditions where the urine was mixed with the faeces, this

study suggests that in naturally ventilated cattle houses,

emissions of NH3, N2O, CO2 and CH4 were significantly

increased with air temperature (5e25 �C) irrespective of floor

type (solid, level and without urine drainage or slatted).

Emissions of NH3, N2O and CO2 were significantly greater from

the solid floor design (by 36, 185 and 44%, respectively) relative

to the slatted floor. Cumulative GHG emissions (as CO2-

equivalents) increased significantly with temperature but did

not differ between the floor types.

Acknowledgments

This work was supported by a grant to the first author (SFRH/

BD/32267/2006) from the Fundacao para a Ciencia e a Tecno-

logia (FCT/MCTES, Portugal). Rothamsted Research, North

Wyke is sponsored by the Biotechnology and Biological

Sciences ResearchCouncil (UK).We thankDr. A.M.D. Silvestre

and Dr. J. Louzada for statistical advice as well as the two

reviewers of the first draft of this submission for their

constructive comments and suggestions.

r e f e r e n c e s

Braam, C. R., Smits, M. C. J., Gunnink, H., & Swierstra, D. (1997).Ammonia emission from a double-sloped floor in a cubiclehouse for dairy cows. Journal of Agricultural EngineeringResearch, 68, 375e386.

Braam, C. R., & Swierstra, D. (1999). Volatilization of ammoniafrom dairy housing floors with different surfacecharacteristics. Journal of Agricultural Engineering Research, 72,59e69.

Cabaraux, J.-F., Philippe, F.-X., Laitat, M., & Canart, B. (2009).Gaseous emissions from weaned pigs raised on different floorsystems. Agriculture. Ecosystems & Environment, 130, 86e92.

Cortus, E. L., Lemay, S. P., Barber, E. M., Hill, G. A., & Godbout, S.(2008). A dynamic model of ammonia emission from urinepuddles. Biosystems Engineering, 99, 390e402.

Elzing, A., & Monteny, G. J. (1997a). Ammonia emissions in a scalemodel of a dairy-cow house. Transactions of the ASAE, 40,713e720.

Elzing, A., & Monteny, G. J. (1997b). Modelling and experimentaldetermination of ammonia emission rates from a scale modeldairy-cow house. Transactions of the ASAE, 40, 721e726.

ES (1999). Soil improvers and growing media. European Standards(ES) 13037. Determination of pH. Brussels: EuropeanCommittee for Standardization.

b i o s y s t em s e n g i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7156

Author's personal copy

Erisman, J. W., Bleeker, A., Galloway, J., & Sutton, M. S. (2007).Reduced nitrogen in ecology and the environment.Environmental Pollution, 150, 140e149.

Forster, P., Ramaswamy, V., Artaxo, P., Berntsen, T., Betts, R.,Fahey, D. W., et al. (2007). Changes in Atmospheric Constituentsand in Radiative Forcing. Cambridge, United Kingdom:Cambridge University Press.

Groenestein, C. M., den Hartog, L. A., & Metz, J. H. M. (2006).Potential ammonia emissions from straw bedding, slurry, pitand concrete floors in a group-housing system for sows.Journal of Agricultural Engineering Research, 95, 235e243.

Hamelin, L., Godbout, S., Theriault, R., & Lemay, S. P. (2010).Evaluating ammonia emission potential from concrete slatdesigns for pig housing. Biosystems Engineering, 105, 455e465.

Hinterhofer, C., Ferguson, J. C., Apprich, V., Haider, H., &Stanek, C. (2006). Slatted floors and solid floors: stress andstrain on the bovine hoof capsule analysed in finite elementanalysis. Journal of Dairy Science, 89, 155e162.

Houba, V. J. G., Van der Lee, J. J., & Novozamsky, I. (1995). Soilanalysis Procedures e Other procedures, Part 5B (6th ed.). TheNetherlands: Wageningen Agricultural University.

IPCC (2007). Climate change 2007. The physical science basis.available at. In S. Salomon, et al. (Eds.), Contribution of workingGroup I to the Fourth Assessment report of the IntergovernmentalPanel on climate Change (pp. 996). Cambridge, UK: CambridgeUniversity Press http://www.ipcc.ch/ipccreports/ar4-wg1.htm978-0-521-70596-7.

Kashyap, D. R., Dadhich, K. S., & Sharma, S. L. (2003).Biomethanation under psychrophilic conditions: a review.Bioresource Technology, 87, 147e153.

Kroodsma, W., Huis in ’t Veld, J. W. H., & Scholtens, R. (1993).Ammonia emission and its reduction from cubicle houses byflushing. Livestock Production Science, 35, 293e302.

Leinker, M., Reinhardt-Hanisch, A., von Borell, E., & Hartung, E.(2007). Application of urease inhibitors in dairy facilities toreduce ammonia volatilization. In G. J. Monteny, & E. Hartung(Eds.), Ammonia emissions in agriculture (pp. 105e107). TheNetherlands: Wageningen Academic Publishers.

Misselbrook, T. H., Pain, B. F., & Headon, D. M. (1998). Estimates ofammonia emission from dairy cow collecting yards. Journal ofAgricultural Engineering Research, 71, 127e135.

Misselbrook, T. H., Webb, J., Chadwick, D. R., Ellis, S., & Pain, B. F.(2001). Gaseous emissions from outdoor concrete yards usedby livestock. Atmospheric Environment, 35, 5331e5338.

Misselbrook, T. H., Webb, J., & Gilhespy, S. L. (2006). Ammoniaemissions from outdoor concrete yards used by livestock e

quantification and mitigation. Atmospheric Environment, 40,6752e6763.

Monteny, G. J., & Erisman, J. W. (1998). Ammonia emissions fromdairy cow buildings: a review of measurement techniques,influencing factors and possibilities for reduction. NetherlandsJournal of Agricultural Science, 46, 225e247.

Moyo, C. C., Kissel, D. E., & Cabrera, M. L. (1989). Temperatureeffects on soil urease activity. Soil Biology & Biochemistry, 21,935e938.

Møller, H. B., Sommer, S. G., & Ahring, B. K. (2004). Biologicaldegradation and greenhouse gas emissions during pre-storageof liquid animal manure. Journal of Environmental Quality, 33,27e36.

Morse, D., Nordstedt, R. A., Head, H. H., & Van Horn, H. H. (1994).Productions and characteristics of manure from lactatingdairy-cows in Florida. Transactions of the ASAE, 37, 275e279.

Morsing, S., Strøm, J. S., Zhang, G., & Kai, P. (2008). Scale modelexperiments to determine the effects of internal airflow andfloor design on gaseous emissions from animal houses.Biosystems Engineering, 99, 99e104.

Mulvaney, R. L. (1996). Nitrogen - inorganic forms. Number 5 inthe Soil Science Society of America Book Series. In D. L. Sparks(Ed.), Methods of soil analysis, Part 3, Chemical methods (3rd ed).(pp. 1123e1184). Madison, USA: SSSA.

Ngwabie, N. M., Jeppsson, K.-H., Nimmermark, S., Swensson, C., &Gustafsson, G. (2009). Multi-location measurements ofgreenhouse gases and emission rates of methane andammonia from a naturally-ventilated barn for dairy cows.Biosystems Engineering, 103, 68e77.

Ni, J. Q., Vinckier, C., Hendriks, J., & Coenegrachts, J. (1999).Production of carbon dioxide in a fattening pig house underfield condition. II. Release from manure. AtmosphericEnvironment, 33, 3697e3703.

Novozamsky, I., Houba, V. J. G., Van Eck, R., & Van Vark, W.(1983). A novel digestion technique for multi-element plantanalysis. Communications in Soil Science and Plant Analysis, 14,239e249.

Pereira, J., Misselbrook, T. H., Chadwick, D. R., Coutinho, J., &Trindade, H. (2010). Ammonia emissions from naturallyventilated dairy cattle buildings and outdoor concrete yards inPortugal. Atmospheric Environment, 44, 3413e3421.

Petersen, S. O., Sommer, S. G., Aaes, O., & Søeggard, K. (1998).Ammonia losses from urine and dung of grazing cattle: effectof N intake. Atmospheric Environment, 32, 295e300.

PNIR (2009). Portuguese National Inventory Report on GreenhouseGases, 1990-2007. Submitted under the United NationsFramework Convention on Climate Change and the KyotoProtocol, Portuguese National Inventory Report (PNIR),Portugal: Portuguese Environmental Agency http://cdr.eionet.europa.eu/.

Saarijarvi, K., Mattila, P. K., & Virkajarvi, P. (2006). Ammoniavolatilization from artificial dung and urine patches measuredby the equilibrium concentration technique (JTI method).Atmospheric Environment, 40, 5137e5145.

Snell, H. G. J., Seipelt, F., & van den Weghe, H. F. A. (2003).Ventilation rates and gaseous emissions from naturallyventilated dairy houses. Biosystems Engineering, 86, 67e73.

Sommer, S. G., Zhang, G. Q., Bannink, A., Chadwick, D.,Misselbrook, T., Harrison, R., et al. (2006). Algorithmsdetermining ammonia emission from buildings housing cattleand pigs and from manure stores. Advances in Agronomy, 89,261e335.

Sullivan, D. M., & Havlin, J. L. (1991). Flow injection analysis ofurea nitrogen in soil extracts. Soil Science Society of AmericaJournal, 55, 109e113.

Swierstra, D., Smits, M. C. J., & Kroodsma, W. (1995). Ammoniaemission from cubicle houses for cattle with slatted and solidfloors. Journal of Agricultural Engineering Research, 62, 127e132.

Trevisi, E., Bionaz, M., Piccioli-Cappelli, F., & Bertoni, G. (2006).The management of intensive dairy farms can be improvedfor better welfare and milk yield. Livestock Science, 103,231e236.

b i o s y s t em s e ng i n e e r i n g 1 0 9 ( 2 0 1 1 ) 1 4 8e1 5 7 157