Available online at www.sciencedirect.

com

Livestock Science 112 (2007) 240 251

www.elsevier.com/locate/livsci

A review of farm level modelling approaches for mitigating

greenhouse gas emissions from ruminant livestock systems
R.L.M. Schils a,, J.E. Olesen b , A. del Prado c , J.F. Soussana d
a

Animal Sciences Group, P.O. Box 65, 8200 AB Lelystad, The Netherlands
University of Aarhus, Faculty of Agricultural Sciences, Tjele, Denmark
c
Institute for Grassland and Environmental Research, North Wyke, UK
d
INRA, UR874 Grassland Ecosystem Research, Clermont-Ferrand, France
b

Abstract
Ruminant livestock systems contribute to global warming through the emission of nitrous oxide (N2O), methane (CH4) and
carbon dioxide (CO2). This paper discusses a general framework for a whole-farm approach to develop cost-effective GHG
mitigation strategies. A dairy farm is a complex system with different interacting components. Generally, whole-farm approaches
distinguish at least an animal component and a soilcrop component. Whole-farm models should be able to give an accurate
representation of the internal cycling of materials and its constituents as well as the exchange between the farming system and
its environment. The paper gives an overview of current whole-farm models that are able to simulate GHG emissions for dairy
farms. These models are DairySim, FarmGHG, SIMSDAIRY and FarmSim. All models are able to calculate CH4 and N2O
emissions, but differences appear in the ability to calculate CO2 emissions, economics and other parameters. The effects of
selected mitigation strategies are demonstrated with some of the models. It is concluded that a whole-farm approach is a
powerful tool for the development of cost-effective GHG mitigation options as it reveals relevant interactions between farm
components. Model calculations underlined the relationship between farm gate N surplus and GHG emissions, and thus the
possibility to use N surpluses as an indicator for GHG emissions.
2007 Elsevier B.V. All rights reserved.
Keywords: Dairy farm; Greenhouse gases; Model; Nitrogen

1. Introduction
Ruminant livestock systems contribute significantly
to global warming through the emission of nitrous oxide
(N2O) and methane (CH4). In the European Union (EU15) the total emission of these greenhouse gases (GHG)

Corresponding author. Tel.: +31 320 293450; fax: +31 320

238050.
E-mail address: rene.schils@wur.nl (R.L.M. Schils).
1871-1413/$ - see front matter 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.livsci.2007.09.005

was 435 Tg CO2-equivalents in the reference year 1990

(EEA, 2006), which was 9% of the total GHG emission.
Emissions from ruminant animal husbandry are roughly
estimated to account for 55% of the total agricultural
emissions (Freibauer, 2003). By 2004, the agriculturalrelated emissions of N2O and CH4 were reduced to
393 Tg CO2-equivalents, or with nearly 10%. This reduction did not result from specific GHG policies, but
was driven mainly by reduced cattle populations and
lower nitrogen (N) fertiliser inputs. Although the EU has
not yet implied direct GHG reduction targets for the

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

agricultural sector, it may be assumed that GHG emissions will be further reduced due to the indirect effects of
other EU policies. Projections for the year 2010 show a
further reduction for the agricultural sector, due to the
continuing effects of the Common Agricultural Policy
(CAP) reform and the Nitrate Directive (EC, 1991), again
mainly through reduced cattle numbers and lower Ninputs. The agricultural sector is way ahead of other
sectors in the reduction of GHG emissions. Due to the
potential for the implementation of cost-effective measures, there is even scope for further reductions so that the
agricultural sector could share a larger part of the burden.
In contrast to industry, the emissions from agriculture
are not confined to relatively few large sources, but are
diffusely spread across Europe. On each individual
holding the farm manager is responsible for the actions
taken to achieve the farmers' goals. The objectives can
be general, like income or continuity, or specific,
depending on personal drive, conviction and style
(Oenema et al., 2004). Consequently, farms are very
different and thus require an individual approach when
mitigation options are developed. In the past years many
mitigation options have been suggested (Mosier et al.,
1998a,b; Oenema et al., 2001). Many options focus on a
single gas only and are often treated as isolated
activities, independent of the farming system. However,
a dairy farm is a complex system with different
interacting components as soils, crops, feeds, animals
and manures. A whole-farm approach ensures that
possible interactions with other GHGs are taken into
account (Janzen et al., 2006). Next to the issue of GHGs,
dairy farmers are constantly challenged to adapt to other
environmental and societal needs as well as changing
market conditions. This requires an integrated approach
in which whole-farm models are indispensable.

241

The objective of this paper is to offer a framework for

a farm level approach, integrating GHG emissions with
other environmentally relevant emissions. Our paper
describes four current whole-farm GHG models and
discusses potential applications of these models.
2. Whole-farm modelling
2.1. General framework
A whole-farm approach for ruminant livestock
systems requires at least the definition of two essential
farm compartments (Fig. 1). The utilisation of homegrown roughage by animals and the return of excreta to
the soilcrop system is a unique feature for ruminant
livestock systems. This distinguishes them from intensive pig or poultry production systems where compound
feeds are imported and animals and manure are
exported, and from arable systems where fertilisers
and manures are imported and crops are exported.
Inputs for ruminant livestock systems comprise those
of biogenic origin like manures or biological N-fixation
by legumes, but also industrially manufactured inputs
like compound feeds and fertilisers. The outputs are
generally milk and meat products. However, in some
intensive systems manure surpluses have to be exported
from the farm as well. Emissions occur at several stages
within the nutrient cycle. The level of detail depends on
the objective of the research. Choices to be made are
related to:
The number of farm components. As stated above
both animal and a soilcrop components are essential
for ruminant livestock systems. A typical cycle used
in N studies comprises components animalmanure

Fig. 1. Basic elements of modelling greenhouse gas (GHG) emissions in a whole-farm approach.

242

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

Table 1
General characteristics of whole-farm GHG models
DairyWise FarmGHG SIMSDAIRY
Model type

Empirical

Empirical

CH4 and N2O

emissions
CO2 emissions
C sequestration
NH3 and NO3
emissions
P cycling
Pre-chain
emissions
Animal welfare
Economics
Biodiversity
Product quality
Soil quality
Landscape
aesthetics

x
x

2.2. Current GHG models

FarmSim

SemiSemimechanistic mechanistic
x
x

x
x
x

x
x
x
x
x
x
x
x

soilcrop (Aarts et al., 2000), but further subdivisions are possible.

The system boundaries. In whole-farm approaches
the system within the farm gate is the minimum that
should be studied. However, emissions related to the
production of inputs or the consumption of outputs is
not accounted for. Efficiency indicators derived from
a farm gate approach may therefore not represent the
complete picture (Schroder et al., 2003). Therefore it
can be justified to include pre- and post-chain effects
into the whole-farm approach.
The simulation methodology. Whole-farm models are
usually a diverse mix of empirical and mechanistic
modelling, with more or less reliance on one of them.
Whole-farm models are often developed through the
combination of existing submodels, which may have
different underlying simulation methodologies. With
respect to GHG, emissions can be calculated with
emission factors, comparable to the IPCC methodology, or simulated with mechanistic (sub)models.
The aspects to be studied. In this paper we focus on
GHG emissions, mainly in relation to N cycling.
However, this can be extended with aspects such as
phosphorus, energy use, heavy metals, landscape,
biodiversity, animal welfare and milk or meat quality.
Inclusion of financial evaluation of mitigation options
is a must, when it comes to potential implementation
by farmers. Some models generate farm plans in which
the financial margin is maximised (Evert et al. 2003;
Gibbons et al., 2006), enabling identification of
economically optimal methods for emission reduction.

The development of whole-farm approaches for the

mitigation of GHG emissions has been taken up recently
by several research groups in Europe (Table 1). A
common feature of all four models is the ability to
calculate CH4 and N2O emissions from dairy farms.
However, the calculation procedures for the GHG
emissions are not always the same. Furthermore, the
models vary considerably on many other aspects. Next
sections give a short description of the models in general,
and more specific of GHG simulation procedures.
2.2.1. DairyWise
DairyWise is an existing empirical model integrating
all major subsystems of a dairy farm into one whole-farm
model (Schils et al., 2007). The model is used
extensively in research, consultancy and teaching for
technical, environmental and financial simulations of
dairy farms. To operate DairyWise, input data are
required which are classified into several categories
such as general farm management, soil characteristics,
herd type and feeding management, cropping plan, grass
and forage management, buildings and machinery. The
key submodels of DairyWise are the GrassGrowth model
and the DairyHerd model. The GrassGrowth model
predicts the daily growth and quality of grass as a
function of soil type, N application and previous
management. The DairyHerd model predicts daily feed
intake and milk production of a complete dairy herd,
including young stock. In the next step the FeedSupply
model combines the herd requirements in terms of
energy and protein with the supply of home-grown grass
and other forage crops and imported compound feeds.
The output of the FeedSupply model contains all data for
a year-round feeding plan. Together with additional
inputs, the output of the FeedSupply model serves as the
input for all technical, environmental and economic
submodels. The technical and economic submodels
calculate the required housing, milk installation, feed
and manure storages, paved roads and yards, fencing, as
well as diesel, gas and water use, medicine use and
required services from contractors, vets and accountants.
The technical farm plan is translated into a financial farm
budget showing all costs and incomes. The prices of
services and commodities are stored in an annually
updated database. The submodel Nutrient cycling
simulates all internal and external flows of nitrogen,
potassium and phosphorus (P). The nitrogen farm gate
surplus is partitioned among ammonia, nitrate and
nitrous oxide losses. Other ecological submodels involve
the calculation of energy use, the integration of nature

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

protection schemes or the implementation of organic

farming.
Recently, a GHG module has been added in which
CH4, N2O and CO2 emissions are calculated with
refined emission factors (Schils et al., 2006a; Schils
et al., 2006b). The CH 4 emission from enteric
fermentation is calculated with emission factors per kg
DM uptake, distinguishing between concentrates, grass
products and maize silage. Emission from manure is
calculated separately for stored manure and manure
excreted during grazing. According to the IPCC guidelines, DairyWise distinguishes between direct and
indirect N2O emissions, related to later denitrification
of leached nitrate. Direct emissions are calculated with
emission factors for N-inputs through fertiliser, biological fixation, manure application, urine excreted during
grazing, crop residues and peat oxidation. Furthermore,
the model calculates an additional N2O emission when
grassland is ploughed. The emission factors are refined
for three categories of ground water level and for two
soil types (mineral and organic). The indirect N2O
emissions are calculated with emission factors from the
modelled nitrate and ammonia emissions. The direct
carbon dioxide (CO2) emission are calculated from the
on-farm use of diesel and gas, while the indirect CO2
emissions are calculated for the use of electricity, the
import of feed and fertiliser, and all other indirect energy
consumption associated with buildings, machinery and
contractor services.
2.2.2. FarmGHG
FarmGHG is a model of carbon (C) and N flows on
dairy farms (Olesen et al., 2006). The model is designed
to allow quantification of direct and indirect gaseous
emissions of CH4 and N2O from dairy farms, so that the
model can be used for assessment of mitigation
measures and strategies. The pre-chain emissions
included in the model comprise use of energy, fertilisers,
pesticides and feedstuffs. However, energy costs
associated with farm buildings and machinery are not
included. The imports, exports and flows of all products
through the internal chains on the farm are modelled.
The model thus allows assessments of emissions from
the production unit and all pre-chains. The model
includes N balance, and allows calculation of environmental effect balances for greenhouse gas emissions
(CO2, CH4 and N2O) and eutrophication (nitrate and
NH3). The model includes not only the farm gate budget
components (input/output), but also the internal flows in
the system (Watson and Atkinson, 1999). These internal
flows are represented as flows between compartments in
the farm system. The model also explicitly includes all C

243

and N losses except for soil respiration and N2 emissions

from soils. The energy use is calculated for each
compartment and is converted to pre-chain emissions of
CO2 and other greenhouse gases.
Milk production and herd size are given from the
description of the farms. The imports of C and N in
fertiliser, bedding, feed, seed and irrigation resulted
from the desired milk production, and from the
specification of farms, but are distributed between the
model compartments in the simulations. This results in
an export of C and N in milk and meat. The farms are
assumed to be operated by best management, and not to
export manure. Net crop production that is not used for
feed is either exported or added to the farm manure
storage.
The model uses a basic time step of 1 month for
inputs and outputs. However, in order to properly reflect
the flows and emissions and the effects of management
changes that operate at sub-monthly scale, daily time
steps are used to simulate the flows between the animals,
house and manure storage compartments. The model
includes a particular manure management component,
which also allows the effects of anaerobic digestion to
be estimated (Weiske et al., 2006).
The model allows different methodologies for
emissions estimations to be used. The tier 1 and tier 2
methodologies of the IPCC (1997) and the IPCC Good
Practice Guidance (IPCC, 2000) were implemented. In
addition, a default FarmGHG methodology can be used
to estimate flows and emissions of CH4 and N2O. This
methodology has the following main characteristics:
The methane emission from enteric fermentation is
based on Kirchgessner et al. (1995), which includes
the effects of feed composition on methane emissions.
The methane emission from slurry systems is
simulated dynamically and influenced by methane
production potential, storage time and temperature.
The methane emission from solid manures depends
on storage time and temperature (Amon, 1998).
The nitrous oxide emissions from slurry stores
depend on slurry tank surface area and storage time,
and from solid manure stores the emissions depend
on storage time.
Ammonia volatilisation is modelled by a combination of emission factors and empirical relationships
for each part of the flow chain (Hutchings et al.,
1996; Sgaard et al., 2002).
Nitrogen fixation depends on legume type and
yield (Hgh-Jensen et al., 2004).
Nitrate leaching depends on N-input in fertilisers,
manure and N-fixation.

244

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

2.2.3. SIMSDAIRY
Sustainable and Integrated Management Systems for
Dairy Production (SIMSDAIRY) is a new modelling
framework which integrates existing models for N flows,
transformations and losses (NGAUGE: Brown et al.,
2005; NARSES: Webb and Misselbrook, 2004), P losses
(PSYCHIC: Davison et al., in press) and farm economics
(A. Butler, pers. comm), equations to simulate NH3
losses from manure application (Chambers et al., 1999),
predict CH4 losses (Chadwick and Pain, 1997; GigerReverdin et al., 2003) and cows' nutrient requirements
[Feed into Milk (FiM) (Thomas, 2004)]. SIMSDAIRY
simulates through score matrices sustainability farm
attributes of biodiversity, landscape, product quality, soil
quality and animal welfare (del Prado et al., 2006).
SIMSDAIRY scope focus on strategic and tactical
management levels and is capable of optimising dairy
management factors in order to find more sustainable
systems (del Prado and Scholefield, 2006).
SIMSDAIRY operates at the farm level. Although prechain emissions are not implicitly included, SIMSDAIRY
produces a qualitative index that accounts for the
amount of bought-in manufactured fertiliser and concentrates. Emissions of CO2 are not predicted within the
current version of the model. A future version is
intended to be developed incorporating CO2 emissions,
energy use and uncertainty analysis.
The model is very sensitive not only to management
but also weather, topography and soil characteristics and
is capable of optimising farm management practices to
meet user multi-weighted criteria and to explore the
possible impact of application of mitigation options on
(i) pollutants such as: N2O, CH4, NH3, NOx, NO3 and P;
(ii) economic profitability; (iii) milk quality; (iv)
biodiversity; (v) landscape; (vi) soil quality and (vii)
animal welfare.
The effect of management practices on N, P and CH4
losses is predicted within different components and
through different processes in the soilplantanimal
system using a monthly time step. These practices can
be defined in terms of management for instance of: (i)
manure (i.e. straw or slurry-based system, storage type,
application method, incorporation time and technique,
timing of application, rate, manure dry matter % and
spatial distribution), (ii) mineral fertiliser (i.e. rate, type
and spatial distribution), (iii) animal (i.e. milk target/
cow, fat content target in milk, protein content target in
milk, calving month, grazing time, diet profile, animal
breeds) and (iv) forage production (i.e. spatial distribution, sward age, history, tillage, plant varieties, silage
making technique). SIMSDAIRY is an advance to other
approaches that study agricultural sustainability as it is

capable of optimising several objectives (advance

towards linear programming) and it can allow the
expression of impacts according to several reference
units (i.e. per kg milk, per ha) (advance towards Life
Cycle Analysis).
The simulation of farm GHG losses includes those
emissions from the soil (as N2O), from animal excreta as
manure or urine and dung deposited during grazing (as
N2O and CH4) and those emissions from the rumen (as
CH4). Nitrous oxide losses from soil are simulated to occur
through denitrification and nitrification processes (Brown
et al., 2005). Denitrification is modelled as a function of
soil inorganic N, soil moisture (as WFPS), net mineralisation rate and temperature. Subsequently, the partition
between N2O and N2 from denitrification is modelled
based on the approach by Parton et al. (1996) and the holein-the-pipe conceptual model described by Firestone and
Davidson (1989). Nitrous oxide emissions are modelled as
a function of nitrification rate [zero-order kinetics:
Gilmour (1984)], soil temperature and soil moisture.
Losses from N2O of total ammonium nitrogen (TAN) in
manure N are according to different emission factors (Efs)
for different manure management stages and managements according to Chambers et al. (1999), Webb and
Misselbrook (2004) and Webb et al. (2006). SIMSDAIRY
simulates the effect of diet characteristics on the % of
initial TAN and mineralisation of manure during different
manure management stages. For instance, increases in
dietary protein or N intake lead to substantial increases in
urinary loss (high proportion of TAN) with almost all N
ingested in excess of animal requirements excreted in
urine.
Methane losses from rumination are simulated
according to the approach by Giger-Reverdin et al.
(2003), by which CH4 output is proportional to the DM
intake and the degree of unsaturation of the fatty acids in
the diet. Methane losses from excreta are predicted
using Efs (per animal) derived from Chadwick and Pain
(1997) and Yamulki et al. (1999) for applied manure and
dung excreted during grazing, respectively. A sensitivity
analysis of SIMSDAIRY indicated that GHG emissions
were largely sensitive to edapho-climatic conditions (i.e.
soil texture) and farm management parameters (i.e. milk
yield/cow, fat and protein content in milk, housing
period, calving system, manure system and diet),
showing in most cases non-linear relationships between
these parameters and these GHG emissions (as CO2equivalents of global warming potential).
2.2.4. FarmSim
FarmSim (for FARM SIMulation) is a model of
greenhouse gas emissions at the livestock farm scale,

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

structured into 9 interacting modules (Saletes et al.,

2004). FarmSim includes imports, exports and internal
flows of products between the different components of
the farm system. FarmSim includes the PASIM model
(Riedo et al., 1998; Vuichard et al., 2007) for the
greenhouse gases exchanged over the different grassland types on the farm and integrates the IPCC
methodology Tier 1 and Tier 2 (IPCC, 1997) for the
CH4 and N2O emissions coming from croplands and
cattle housing. FarmSim uses detailed data inputs
concerning the farm structure (area and type of crops
and of grasslands, herd types), the herd (number of
animals per type each fortnight), the grasslands (grazing
and cutting dates, stocking rates, organic and inorganic
fertiliser applications), the crops and the feeding and
waste management systems.
From these data, FarmSim calculates inputs needed to
run the pasture simulation model (PASIM) for each of the
grassland plot in the farm. The PASIM model allows to
simulate the average net annual balance of greenhouse
gases (CO2, N2O, CH4) exchanged by the managed
grassland plots. Other outputs from FarmSim are used to
estimate CH4 and CO2 emissions in the cowshed,
according to liveweight, feed composition and level of
feed intake, and CH4 and N2O emissions from the
manure management using the IPCC methodology
Tier 2.
The PASIM model is driven by hourly or daily weather
data for temperature, precipitation, vapour pressure,
radiation, and wind speed. Site-specific model parameters
include the N-input from mineral and/or organic fertilisers
and atmospheric deposition, the fractional clover content
of the grass/clover-mixture, the depth of the main rooting
zone, and some other soil physical parameters. The model
is fully dynamic and is used to simulate above- and belowground dry matter production of a perennial sward in
relationship to fluxes of C, N, water, and energy. PASIM
can be used for either grazing or cutting. The animal
submodel calculates intake by cattle, milk and meat
production, and losses of CO2 via respiration and CH4 by
enteric fermentation at the grazed grassland level. The
production and transport of the main inputs were
accounted for into the input fluxes category, which
includes fuel, electricity, N-fertilisers and feedstuffs. The
data for production and transport of inputs were based on
a full accounting scheme as used in Life Cycle Analysis
(Kramer et al., 1999; Kaltschmitt and Reinhardt, 1997;
Dalgaard et al., 2001). Energy use results into direct and
indirect emissions of carbon dioxide, nitrous oxide and
methane. The energy use consists of electricity and fuel
use. For electricity, country-specific systems are used,
depending on the specific power mix of fossil energies

245

(coal, oil and natural gas), atomic, hydraulic and wind

sources. Furthermore, the energy use and associated GHG
emissions during the production and transport of imported
resources have been accounted especially for purchased
N-fertilisers and feedstuffs. Energy and materials for
buildings and machinery were not accounted for. The total
GHG budget of the farm was calculated as the sum of the
internal GHG fluxes (at the farm gate) and of the GHG
emissions that result from input fluxes (pre-chain
emissions plus direct use of fossil fuel energy on the
farm).
3. Model application
3.1. Whole-farm evaluation of mitigation options
Recent research has delivered a wide range of
mitigation options. Mitigation options for enteric
fermentation are either aimed at an increased animal
productivity, like higher milk yields per cow or lower
replacement rates or are aimed to affect rumen
functioning like increased levels of starch or the use of
additives (Monteny et al., 2006). Options to reduce CH4
emissions from manure storage are anaerobic digestion,
but also manure cooling or simple things like proper
management of bedding to avoid anaerobic conditions.
There is a wide range of mitigation options for N2O that
can roughly be divided into two strategies, (i) increasing
the N use efficiency and (ii) reducing the N2O
production per unit of N (Oenema et al., 2001). Nitrous
oxide emissions can be reduced by implementing
practices aimed at enhancing the ability of the crop to
compete with processes that lead to the escape of N from
the soil-plant system (Freney, 1997). For instance, there
are several methods for increasing the efficiency of the
crop to remove mineral N from the soil. These include
improving fertiliser efficiency (Brown et al., 2005),
optimising methods and timing of applications (Dosch
and Gutser, 1996), using ammonium-based fertilisers
rather than nitrate-based ones (Eichner, 1990; Dobbie
and Smith, 2003) and employing chemical nitrification
inhibitors (Dittert et al., 2001; Merino et al., 2002;
Macadam et al., 2003).
Increasing the soil aeration may significantly reduce
N2O emissions. Improving drainage would be particularly beneficial on grazed grassland (Monteny et al.,
2006). Hence, avoiding compaction by traffic, tillage
(Pinto et al., 2004) and grazing livestock may help to
reduce N2O emissions. Housing system and management will also influence N2O emissions, e.g. strawbased manures result in greater N2O emissions than
slurry-based ones (Groenestein and Van Faassen, 1996).

246

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

Fig. 2. GHG emissions (kg CO2-equiv. per kg milk) for 13 dairy farms
either with anaerobic digestion of manure or without digestion (Baseline).

Minimizing the grazing period is likely to reduce N2O

emissions as long as the slurry produced during the
housing period is uniformly spread, and less fertiliser N
is used to compensate the higher N utilisation from
slurry.
Carbon dioxide emissions, especially the indirect
emissions, are linked to the energy and resource use. So
any type of management that increases the energy
efficiency will also reduce CO2 emissions. These options
can range from technological solutions in the milking
parlour to the use of white clover to replace fertiliser N.
Digestion of manure and the subsequent production of
heat and electricity also contribute to lower fossil energy
use and thus lower CO2 emissions. Carbon sequestration
frequently occurs in grasslands (Soussana et al., 2004)
and can be increased at the farm level by changes from
arable land to grassland, and from short duration leys to
permanent grasslands (Soussana et al., 2004), or with
other types of measures like reduced tillage and use of
crop residues and sewage (Smith et al., 1998).
In the following sections the whole-farm models are
applied to calculate the effect of several mitigation
options on GHG emissions. We then discuss these
modelling results in view of the current limitations of
each modelling approach.
3.1.1. Anaerobic digestion
Biogas can be produced either from manure digestion
alone, or by co-digestion of manure with farm-grown or
imported products. In this scenario analysis we compare
the GHG emissions of a baseline farm without digestion
with both a digestion and co-digestion scenario,
simulated with FarmGHG (Weiske et al., 2006). The
calculations were performed for a wide range of dairy
farm types in Europe. The 14 farms consisted of
combinations of 5 regions (Atlantic, continental, prealpine, boreal and Mediterranean), several stocking
densities (0.7 to 2.7 LU ha 1), 3 cropping plans (grass,

maize and mixed), 2 manure types (slurry, FYM) and 2

farm types (conventional and organic).
The two scenarios presented here consisted of
digestion of manure only or co-digestion of manure and
potato residues. The scenarios were assessed on the effect
of CH4 and N2O emissions, but also on the effect of CO2
emissions, including that of the substitution of fossil fuel
energy with bio-energy. The potato residues were seen as
a waste product, so no emissions are assigned to the
production of this material for co-digestion, although
transport related emissions were included.
All scenarios showed a clear reduction in GHG
emissions, with the best results for the co-digestion scenarios (Fig. 2). The cost of electricity generation varied
from 8 to 55 cents per kWh, and in many cases digestion
could only be made financially viable with the help of
subsidies or by increasing the farm size to more than
100 ha.
3.1.2. Reduced grazing and increased proportion of
maize
In the Netherlands, forage maize is grown on most
dairy farms on mineral soils. On clay soils, approximately 10% of the area is in maize, while on sandy soils
the average proportion of the maize area is 23%
(Table 2). On organic soils, no maize is grown, although
some maize is grown on peat soils with a clayey top
layer. The increased use of maize is seen as one of the
potential mitigation options for the CH4 emission from
Table 2
Effect of increased proportion of maize and reduced grazing on CH4
and N2O emissions of dairy farms on three soil types in the Netherlands

Reference scenario
Farm area (ha)
Maize area (%)
Grazing system

Alternative scenario
Maize area (%)
Grazing system

Change in emissions
CH4 (kg ha 1)
N2O (kg ha 1)
GHG
(CO2-equiv. ha 1)
GHG
(CO2-equiv. kg milk 1)
Costs
(1000 kg CO2-equiv.)

Sand

Clay

Peat

35
23
Daytime
only (5 h)

42
10
Daytime
only (10 h)

39
6
Day-and-night
(20 h)

30
30
No grazing Daytime
only (5 h)

6
Daytime
only (10 h)

+21.4
3.0
468

4.9
3.9
1482

+4.3
10.3
3110

0.04

0.11

0.26

591

108

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

enteric fermentation. The grazing intensity on dairy

farms is often related to the proportion of maize, i.e. the
higher the proportion of maize, the lower the grazing
time. Reduced grazing is considered as a mitigation
option for N2O emissions, so the combination of an
increased proportion of maize and reduced grazing
could be a promising mitigation strategy for GHG
emissions. However, an increased proportion of arable
land reduces soil C sequestration, leading to higher CO2
emissions. Changes in CH4 and N2O emissions and
farm income were calculated with the DairyWise model
for farms that simultaneously increased the proportion
of maize and reduced the grazing time of the dairy herd.
On the mineral soil types, the proportion of maize was
increased to 30%, which is the maximum proportion of
maize allowed under the current nitrate policy. The

247

farms on peat soils did not increase the area of maize,

but increased maize silage imports so that they could
double the amount of maize in the cows ration. The
grazing time was reduced by 5 h a day on the mineral
soils and 10 h a day on the organic soil.
In order of importance the main effects were a
decreased nitrous oxide emission, an increased methane
emission from manure management and a decreased
methane emission from enteric fermentation. For all soil
types, the proposed strategy reduces the total GHG
emissions, but the results depend on the initial reference
situation, and do not account for a reduced C sequestration. On sandy soils, where the proportion of maize is
already relatively high and the grazing time relatively low,
the achieved emission reductions are much lower
compared to the farms on clay and peat soil. This effect

Fig. 3. Effect of housing time on predicted losses per hectare of CH4 and N2O on UK dairy farms under sandy loam (a) and clay loam (b) soils.

248

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

Fig. 4. Relationship between GHG emission (N2O and CH4) and N surplus. DairyWise: conventional NL farms on mineral and peat soils. FarmGHG:
organic and conventional farms in several regions of Europe. SIMSDAIRY: farms in UK with either conventional grazing or extended grazing, and
either under normal (720 mm/year; 1112 C during growing season) or wet conditions (1493 mm/year; 1011 C during growing season). FarmSim:
farms in United Kingdom, the Netherlands, France, Denmark and Switzerland with contrasting soil, climate and farming system.

of soil type is also reflected in the cost effectiveness,

which decreases in the order sand, clay and peat.
Similar results were obtained for a typical intensive
dairy farm in the South West of England.
SIMSDAIRY was used to simulate the effect of housing
days, proportion of maize area and soil type (sandy loam
and clay loam) on losses of N, P and CH4. The maize area
fraction was proportional to the number of housing days
ranging from 3 to 14% and 4 to 17% in sandy loam and
clay loam soils, respectively. The predicted results
indicated that increasing housing days generally decreased the impact of global warming potential (GWP) of
CH4 and N2O. This is a result of a balance between the
decrease, up to a certain level of housing time, in N2O
emissions from soils through less grazing and decreasing
the proportion of urine N in excreta through a rich starchbased (maize) diet (Kebreab et al., 2001), and the small
increase in CH4 emissions due to possible an increase
impact of CH4 emissions from rumination through an
increased animal intake and larger volumes of manure
generated (Fig. 3). The non-linear relationship suggests
that for intensive systems there may be a housing time per
soil type and probably per management choice where the
positive effect on N2O emissions of replacing grazing
with increasing manure production (and subsequently
apply it to land) and increasing maize area is counteracted
by excessive manure per ha to be applied.
Increasing housing time also had a remarkable effect
on aqueous losses (NO3 leaching and P losses) and
losses that affect the acidification of environments (NH3
and NOx). Increasing housing time leads to a decrease,
up to a certain number of housing days, and an increase

in NO3 leaching losses for both soils. Phosphorus losses

tended to increase with increasing housing days.
Results, however, showed that this tendency is subject
to some degree of variation, possibly due to diet
management. On average, increasing housing time in
intensive dairy systems had a negative effect on losses
leading to eutrophication of water systems.
Losses of NH3 and NOx generally increased with
increasing housing days through an increased amount of
losses from manure and silage making.
SIMSDAIRY predicts that increasing housing time and
reliance on maize could have an increased risk of
negative effects on issues such as the proportion of
polyunsaturated fatty acids (PUFA) in the milk content
(through the reduction of ingested PUFA through
reduced fresh grass feeding), animal welfare (housing
may have a negative effect on animal social relationships) and plant biodiversity (through more even
distribution of excreta in the land).
3.2. Relating GHG emissions to N surpluses
In recent years, N policies have had a substantial effect
on dairy farming. In the Netherlands for instance, farms
have improved the N utilisation through improved
manure management, less fertiliser and feed import,
reduced grazing and less young stock per cow. Unintended, the improved N management also reduced the GHG
emissions. Modelling exercises with DairyWise,
FarmGHG, SIMSDAIRY and FarmSim for a range of
farm types across Europe confirm the positive relationship
between N surplus and GHG emissions (Fig. 4). The

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

range of N surpluses was generated by differences in

stocking rates in combination with several specific farm
management practices. Although the effect of climate and
soil conditions on GHG emissions are sometimes greater
than the effect of management system choice, there is
certainly scope to use N surplus as a proxy for GHG
emissions. Each kg of N surplus corresponds with a GHG
emission of approximately 30 to 70 kg CO2-equivalents.
A similar relationship is found if the GHG emissions are
calculated with the IPCC defaults (data not shown), albeit
that the GHG emission level is generally lower due to the
exclusion of pre-chain emissions.
FarmSim also includes C sources for arable lands and
C sinks for grassland, changing the total farm scale
picture. Including the simulated C sequestration,
changed the total GHG balance from 8.7 to + 0.5
CO2-equivalents ha 1.
4. Discussion and conclusions
4.1. Whole-farm approach
A dairy farm is a complex system with several
interacting subsystems. Whole-farm models of dairy
systems should therefore be able to give an accurate
representation of the internal cycling of materials and their
constituents as well as the exchange of materials and
nutrients between the farming system and its environment. Moreover, such models should reliably predict the
effects of changes in management. For dairy farming
systems in Europe, the described models are able to
simulate internal and external flows of materials and
nutrients, and calculate the related GHG emissions. The
whole-farm approach, adopted in these models, ensures
that the interactions between the relevant themes are taken
into account. Effects of changes in one farm component
are not confined to that particular subsystem, but are
transferred throughout the whole-farm system.
In order to limit potential bias in GHG balance
estimates, farm scale models need to include all CO2,
CH4 and N2O fluxes as well as pre-chain emissions.
Consequently, such an approach prevents that issues
are viewed in isolation and management options are
proposed without reflecting the effect on other relevant
issues. The whole-farm approach should not be seen as
a replacement of the IPCC methodology. The choice
depends on the objective. The IPCC method is used to
prepare transparent and consistent inventories for national emission reporting. Calculations on farm scale
are useful to explore mitigation options for individual farms. However, for fulfilling national reduction
targets there is a need to ensure that such mitigation

249

options are also reflected in the national emission

inventories.
4.2. Model applications
Generally, whole-farm models are suitable to assess
the technical, environmental and financial implications of
alternative farm management strategies, under changing
external conditions. The whole-farm approach ensures
that potential negative trade-offs are taken into account
and that positive synergies are identified. On the other
hand, when analysing effects on farm scale, higher
integration levels should not be ignored. It is possible that
certain strategies, developed within a whole-farm approach, are considered to have a positive effect, while the
effect on a regional, national or even higher scale might be
neutral or even negative (Oborn et al., 2003). In the case of
energy use, DairyWise and FarmGHG (Olesen et al.,
2006) take into account the pre-farm energy consumption
for the production and transport of materials used on the
farm. For other parameters and especially for downstream
post-farm processes, most models do not take into account
possible transfer of effects.
The models discussed in this paper simulate average
emissions, disregarding the uncertainty associated with
both the activity data and the emission factors. In order
to analyse the uncertainty in a Monte Carlo framework,
the ranges and the probability density functions of the
model parameters are needed (Gibbons et al., 2006;
Payraudeau et al., 2007).
Considering GHGs, the described models are mainly
used in research for the exploration of cost-effective
mitigation strategies. Potential other applications are the
use in extension and teaching. In extension, DairyWise
is already used to explore future farm strategies for
individual farms or for study groups of farmers working
under similar environmental conditions. As the model
operates on farming level, it is relatively easy for
farmers to learn and understand the underlying processes on their own farm. Therefore, a whole-farm approach
is also useful for the communication of mitigation
options to farmers, especially if the model also evaluates
additional costs and benefits. This is a crucial step in the
implementation of mitigation options since it is the
farmers who themselves decide and judge the effectiveness in the context of a whole-farm system.
4.3. Conclusions
A whole-farm approach is a powerful tool for the
development of cost-effective GHG mitigation options.
It reveals relevant interactions between farm components

250

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

and is useful for integrated scenario development and

evaluation. Moreover it enables farmers to address the
issue of GHG emissions in their operational, tactical
and strategic management. Model calculations confirmed the relationship between farm gate N surplus and
GHG emissions. It is reasonable to further explore the
possibility to use N surpluses as an indicator for GHG
emissions as data on N surpluses are becoming
increasingly available for farms in Europe.
References
Aarts, H.F.M., Habekotte, B., Van Keulen, H., 2000. Nitrogen (N)
management in the De Marke dairy farming system. Nutr. Cycl.
Agroecosyst. 56, 231240.
Amon, B., 1998. NH3 -, N2O- und CH4-Emissionen aus der
Festmistanbindehaltung fr MilchviehStallLagerungAusbringung. Dissertation. Forschungsbericht Agrartechnik 331 (in
German). Universitt fr Bodenkultur, Institut fr Land-, Umwelt
und Energietechnik, Wien.
Brown, L., Scholefield, D., Jewkes, E.C., Lockyer, D.R., del Prado, A.,
2005. NGAUGE: a decision support system to optimise N
fertilization of British grassland for economic and environmental
goals. Agric. Ecosyst. Environ. 109, 2039.
Chadwick, D.R., Pain, B.F., 1997. Methane fluxes following slurry
applications to grassland soils: laboratory experiments. Agric.
Ecosyst. Environ. 63, 5160.
Chambers, B.J., Lord, E.I., Nicholson, F.A., Smith, K.A., 1999.
Predicting nitrogen availability and losses following application of
organic manures to arable land: MANNER. Soil Use Manage. 15,
137143.
Dalgaard, T., Halberg, N., Porter, J.R., 2001. A model for fossil energy
use in Danish agriculture used to compare organic and conventional farming. Agric. Ecosyst. Environ. 87, 5165.
Davison, P.S., Withers, P.J.A., Lord, E.I., Betson, M.J., Strmqvist, J.,
in press. PSYCHIC a process-based model of phosphorus and
sediment mobilisation and delivery within agricultural catchments
in England and Wales. Part 1 Model description and
parameterisation. J. Hydrol.
Del Prado, A., Scholefield, D., Chadwick, D., Misselbrook, T.,
Haygarth, P., Hopkins, A., Dewhurst, R., Davison, P., Lord, E.,
Turner, M., Aikman, P., Schrder, J., 2006. A modelling
framework to identify new integrated dairy production systems.
EGF: 21st General Meeting on Sustainable Grassland Productivity, Badajoz, Spain. 36 April.
Del Prado, A., Scholefield, D., 2006. Use of SIMSDAIRY modelling
framework system to specify sustainable UK dairy farms.
Delivering sustainability within profitable farming systems is
it possible? Aspects Appl. Biol. 80, 7380.
Dittert, K., Bol, R., King, R., Chadwick, D., Hatch, D., 2001. Use of a
novel nitrification inhibitor to reduce nitrous oxide emission from
N-15-labelled dairy slurry injected into soil. Rapid Commun. Mass
Spectrom. 15, 12911296.
Dobbie, K.E., Smith, K.A., 2003. Impact of different forms of N fertilizer
on N2O emissions from intensive grassland. Nutr. Cycl. Agroecosyst.
67, 3746.
Dosch, P., Gutser, R., 1996. Reducing N losses (NH3, N2O, N2) and
immobilization from slurry through optimized application techniques. Fert. Res. 43, 165171.

EEA, 2006. Annual European Community Greenhouse Gas Inventory

19902004 and Inventory Report 2006. European Environment
Agency, Kopenhagen, p. 434.
EC, 1991. Directive of the Council of December 12, 1991 Concerning
the Protection of Waters Against Pollution Caused by Nitrates from
Agricultural Sources (91/676/EEC). European Commision, Brussels, pp. 18.
Eichner, M.J., 1990. Nitrous oxide emissions from fertilized soils
summary of available data. J. Environ. Qual. 19, 272280.
Evert, F.K., Ten Berge, H.F., Van Der Meer, H.G., Rutgers, B., Schut,
A.G., Ketelaars, J.J., 2003. FARMMIN: Modeling Crop-Livestock
Nutrient Flows. ASA-CSSA-SSSA, Madison, WI.
Firestone, M.K., Davidson, E.A., 1989. Microbial basis of NO and
N2O production and consumption in soil. In: Andreae, M.O.,
Schimel, D.S. (Eds.), Exchange of Trace Gases Between Terrestrial
Ecosystems and the Atmosphere: Report of the Dahlem Workshop.
Wiley and Sons, Berlin, New York.
Freibauer, A., 2003. Regionalised inventory of biogenic greenhouse
gas emissions from European agriculture. Eur. J. Agron. 19,
135160.
Freney, J.R., 1997. Strategies to reduce gaseous emissions of nitrogen
from irrigated agriculture. Nutr. Cycl. Agroecosyst. 48, 155160.
Gibbons, M., Ramsden, S.J., Blake, A., 2006. Modelling uncertainty
in greenhouse gas emissions from UK agriculture at the farm level.
Agric. Ecosyst. Environ., 112, 347355.
Gilmour, J.T., 1984. The effects of soil properties on nitrification and
nitrification inhibition. Soil Sci. Soc. Am. J. 48 (6), 12621266.
Giger-Reverdin, S., Morand-Fehr, P., Tran, G., 2003. Literature survey
of the influence of dietary fat composition on methane production
in dairy cattle. Livest. Prod. Sci. 82, 7379.
Groenestein, C.M., Van Faassen, H.G., 1996. Volatilization of
ammonia, nitrous oxide and nitric oxide in deep-litter systems
for fattening pigs. J. Agric. Eng. Res. 65, 269274.
Hutchings, N.J., Sommer, S.G., Jarvis, S.C., 1996. A model of
ammonia volatilization from a grazing livestock farm. Atmos.
Environ. 30, 589599.
Hgh-Jensen, H., Loges, R., Jrgensen, F.V., Vinther, F.P., Jensen, E.S.,
2004. Empirical model for quantification of symbiotic nitrogen
fixation in leguminous crops. Agric. Syst. 82, 181194.
IPCC, 1997. Greenhouse Gas Inventories. Revised 1996 IPCC
Guidelines for National Greenhouse Gas Inventories.
IPCC, 2000. IPCC Good Practice Guidance and Uncertainty
Management in National Greenhouse Gas Inventories.
Janzen, H.H., Angers, D.A., Boehm, M., Bolinder, M., Desjardins, R.L.,
Dyer, J.A., Ellert, B.H., Gibb, D.J., Gregorich, E.G., Helgason, B.L.,
Lemke, R., Mass, D., McGinn, S.M., McAllister, T.A., Newlands,
N., Pattey, E., Rochette, P., Smith, W., VandenBygaart, A.J., Wang,
H., 2006. A proposed approach to estimate and reduce net
greenhouse gas emissions from whole farms. Can. J. Soil Sci. 86,
401418.
Kaltschmitt, M., Reinhardt, G.A., 1997. Nachwachsende Energietrger - Grundlagen, Verfahren, kologische Bilanzierung. ViewegVerlag, Braunschweig/Wiesbaden, Germany.
Kebreab, E., France, J., Beever, D.E., Castillo, A.R., 2001. Nitrogen
pollution by dairy cows and its mitigation by dietary manipulation.
Nutr. Cycl. Agroecosyst. 60, 275285.
Kirchgessner, M., Windisch, W., Mller, H.L., 1995. Nutritional
factors for the quantification of methane production. In: Engelhardt, W.v., Leonhard-Marek, S., Breves, G., Giesecke, D. (Eds.),
Ruminant Physiology: Digestion, Metabolism, Growth and
Reproduction. Proceedings VIII International Symp. on Ruminant
Physiology, pp. 333348.

R.L.M. Schils et al. / Livestock Science 112 (2007) 240251

Kramer, K.J., Moll, H.C., Nonhebel, S., Wilting, H.C., 1999.
Greenhouse gas emissions related to Dutch food consumption.
Energy Policy (4), 203216.
Macadam, X.M.B., del Prado, A., Merino, P., Estavillo, J.M., Pinto,
M., Gonzalez-Murua, C., 2003. Dicyandiamide and 3,4-dimethyl
pyrazole phosphate decrease N2O emissions from grassland but
dicyandiamide produces deleterious effects in clover. J. Plant
Physiol. 160, 15171523.
Merino, P., Estavillo, J.M., Graciolli, L.A., Pinto, M., Lacuesta, M.,
Munoz-Rueda, A., Gonzalez-Murua, C., 2002. Mitigation of N2O
emissions from grassland by nitrification inhibitor and Actilith F2
applied with fertilizer and cattle slurry. Soil Use Manage. 18, 135141.
Monteny, G.J., Bannink, A., Chadwick, D., 2006. Greenhouse gas
abatement strategies for animal husbandry. Agric. Ecosyst.
Environ. 112, 163170.
Mosier, A.R., Duxbury, J.M., Freney, J.R., Heinemeyer, O., Minami,
K., 1998a. Assessing and mitigating N2O emissions from
agricultural soils. Clim. Change 40, 738.
Mosier, A.R., Duxbury, J.M., Freney, J.R., Heinemeyer, O., Minami,
K., Johnson, D.E., 1998b. Mitigating agricultural emissions of
methane. Clim. Change 40, 3980.
Oborn, I., Edwards, A.C., Witter, E., Oenema, O., Ivarsson, K.,
Withers, P.J.A., Nilsson, S.I., Stinzing, A.R., 2003. Element
balances as a tool for sustainable nutrient management: a critical
appraisal of their merits and limitations within an agronomic and
environmental context. Eur. J. Agron. 20, 211225.
Oenema, O., Velthof, G., Kuikman, P., 2001. Technical and policy
aspects of strategies to decrease greenhouse gas emissions from
agriculture. Nutr. Cycl. Agroecosyst. 60, 301315.
Oenema, O., Kuikman, P., Velthof, G., 2004. Assessment and
mitigation of greenhouse gas emissions at farm level. In:
Kaltschmitt, M., Weiske, A. (Eds.), Greenhouse Gas Emissions
from Agriculture. Mitigation Options and Strategies. Institute for
Energy and Environment, Leipzig, pp. 172178.
Olesen, J.E., Schelde, K., Weiske, A., Weisbjerg, M.R., Asman, W.A.H.,
Djurhuus, J., 2006. Modelling greenhouse gas emissions from
European conventional and organic dairy farms. Agric. Ecosyst.
Environ. 112, 207220.
Parton, W.J., Holland, E.A., Del Grosso, S.J., Hartman, M.D., Martin,
R.E., Mosier, A.R., Ojima, D.S., Schimel, D.S., 1996. Generalized
model for N2 and N2O production from nitrification and
denitrification. Glob. Biogeochem. Cycles 10 (3), 401412.
Payraudeau, S., Van der Werf, H.M.G., Vertes, F., 2007. Analysis of
the uncertainty associated with the estimation of nitrogen losses
from farming systems. Agric. Syst 94, 416430.
Pinto, M., Merino, P., del Prado, A., Estavillo, J.M., Yamulki, S.,
Gebauer, G., Piertzak, S., Lauf, J., Oenema, O., 2004. Increased
emissions of nitric oxide and nitrous oxide following tillage of a
perennial pasture. Nutr. Cycl. Agroecosyst. 70, 1322.
Riedo, M., Grub, A., Rosset, M., Fuhrer, J., 1998. A pasture simulation
model for dry matter production, and fluxes of carbon, nitrogen,
water and energy. Ecol. Mod. 23, 141183.
Saletes, S., Fiorelli, J., Vuichard, N., Cambou, J., Olesen, J.E., Hacala,
S., Sutton, M., Fuhrer, J., Soussana, J.F., 2004. Greenhouse gas
balance of cattle breeding farms and assessment of mitigation
options. In: Kaltschmitt, M., Weiske, A. (Eds.), Greenhouse Gas

251

Emissions from Agriculture. Mitigation Options and Strategies.

Institute for Energy and Environment, Leipzig, pp. 203208.
Schils, R.L.M., Oudendag, D.A., Van der Hoek, K.W., De Boer, J.A.,
Evers, A.G., De Haan, M.H.A., 2006a. Broeikasgasmodule BBPR.
Animal Sciences Group, Lelystad.
Schils, R.L.M., Verhagen, A., Aarts, H.F.M., Kuikman, P.J., ebek,
L.B.J., 2006b. Effect of improved nitrogen management on greenhouse gas emissions from intensive dairy systems in the Netherlands.
Glob. Change Biol. 12, 382391.
Schils, R.L.M., de Haan, M.H.A., Hemmer, J.G.A., van den Pol-van
Dasselaar, A., de Boer, J.A., Evers, A.G., Holshof, G., van
Middelkoop, J.C., Zom, R.L.G., 2007. DairyWise, a whole-farm
dairy model. J. Dairy Sci. 90. doi:10.3168/jds.2006-842.
Schroder, J.J., Aarts, H.F.M., ten Berge, H.F.M., van Keulen, H.,
Neeteson, J.J., 2003. An evaluation of whole-farm nitrogen
balances and related indices for efficient nitrogen use. Eur. J.
Agron. 20, 3344.
Smith, P., Powlson, D.S., Glendining, M.J., Smith, J.U., 1998.
Opportunities and limitations for C sequestration in European
agricultural soils through changes in management. In: Lal, R.,
Kimble, J.M., Follett, R.F., Stewart, B.A. (Eds.), Management of
Carbon Sequestration in Soil. CRC Press, New York, pp. 143152.
Soussana, J.F., Loiseau, P., Vuichard, N., Ceschia, F., Balesdent, J.,
Chevallier, T., Arrouays, D., 2004. Carbon cycling and sequestration opportunities in temperate grasslands. Soil Use Manage. 20,
219230.
Sgaard, H.T., Sommer, S.G., Hutchings, N.J., Huijsmans, J.F.M.,
Bussink, D.W., Nicholson, F., 2002. Ammonia volatilization from
field applied animal slurry The ALFAM Model. Atmos.
Environ. 36, 33093319.
Thomas, C., 2004. Feed into Milk (FiM): a New Applied Feeding
System for Diary Cows. Nottingham University Press, UK.
Vuichard, N., Soussana, J.F., Viovy, N., Calanca, P., Clifton-Brown, J.,
Ciais, P., 2007. Simulating the greenhouse gas budget of European
grasslands within a process driven approach: model evaluation
against in-situ measurements. Glob. Biogeochem. Cycles 21,
GB1004. doi:10.1029/2005GB002611.
Watson, C.A., Atkinson, D., 1999. Using nitrogen budgets to indicate
nitrogen use efficiency and losses from whole farm systems: a
comparison of three methodological approaches. Nutr. Cycl.
Agroecosyst. 53, 259267.
Webb, J., Ryan, M., Anthony, S.G., Brewer, A., Laws, J., Aller, M.F.,
Misselbrook, T.H., 2006. Cost-effective means of reducing
ammonia emissions from UK agriculture using the NARSES
model. Atmos. Environ. 40, 72227233.
Webb, J., Misselbrook, T.H., 2004. A mass-flow model of ammonia
emissions from UK livestock production. Atmos. Environ. 38,
21632176.
Weiske, A., Vabitsch, A., Olesen, J.E., Schelde, K., Michel, J.,
Friedrich, R., Kaltschmitt, M., 2006. Mitigation of greenhouse gas
emissions in European conventional and organic dairy farming.
Agric. Ecosyst. Environ. 112, 221232.
Yamulki, S., Jarvis, S.C., Owen, P., 1999. Methane emission and
uptake from soils as influenced by excreta deposition from grazing
animals. J. Environ. Qual. 28, 676682.