Europ. J.

Agronomy 20 (2004) 247–264

Environmental impact assessment of agricultural production

systems using the life cycle assessment methodology
I. Theoretical concept of a LCA method tailored to
crop production
F. Brentrup∗ , J. Küsters, H. Kuhlmann, J. Lammel
Centre for Plant Nutrition, Hanninghof, Hydro Agri, Hanninghof 35, 48249 Duelmen, Germany

Received 19 July 2002; received in revised form 10 March 2003; accepted 27 March 2003

Abstract

A new life cycle assessment (LCA) method is presented, which is specifically tailored to plant nutrition in arable crop produc-
tion. Generally, LCA is a methodology to assess all environmental impacts associated with a product or a process by accounting
and evaluating its resource consumption and emissions. In LCA studies the entire production system should be considered, i.e.
for crop production systems the analysis includes not only the on-field activities, but also all impacts related to the production of
raw materials (minerals, fossil fuels) and farm inputs like fertilizers, plant protection substances, machinery or seeds. The LCA
method developed in this study evaluates the impact of emissions and resource consumption associated with crop production
on the following environmental effects: depletion of abiotic resources, land use, climate change, toxicity, acidification, and
eutrophication. In order to enable conclusions on the overall environmental impact of alternative crop nutrition systems, an ag-
gregation procedure to calculate indicators for resource depletion (RDI) and environmental impacts (EcoX) has been developed.
The higher the EcoX value, the higher is the overall environmental burden associated with the product under investigation. An
environmental analysis of arable crop production systems based on this LCA method is especially appropriate in order to: (1)
detect environmental hot spots in the system; (2) trace back environmental impacts of arable farming products to their sources and
on that basis to suggest options for improvement; and (3) contribute to the debate on the environmental preference of alternative
cropping systems in an informed way.
© 2003 Elsevier B.V. All rights reserved.

Keywords: Life cycle assessment; Nitrogen fertilizer; Arable farming; Plant nutrition; Environmental impact; Resource depletion; Ecosystem
health

1. Introduction to be environmentally benign (Commission of the

European Communities, 1999; UN-DSD, 2000). To
Agriculture is expected to be competitive, to pro- evaluate the sustainability of agricultural production
duce high quality food in sufficient quantities and systems, it is necessary to have appropriate indicators
∗ Corresponding author. Tel.: +49-2594-798-137; in place.
fax: +49-2594-7455. The environmental impacts of N-use in agricul-
E-mail address: [email protected] (F. Brentrup). ture have been analyzed in numerous investigations.

1161-0301/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S1161-0301(03)00024-8
248 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

Frequently, these focus only on individual effects pacts relevant to agriculture (e.g. land use, resource
such as nitrate leaching or ammonia volatilization consumption; Brentrup et al., 2001).
(e.g. Bach and Becker, 1995; ECETOC, 1988, 1994; This paper describes an LCA method that was
Engels, 1993; Sommer, 1992). However, agricultural developed to integrate the environmental effects,
production systems contribute to a wide range of envi- which are relevant to crop production with a special
ronmental impacts (e.g. climate change, acidification, focus on plant nutrition, and to combine the best
eutrophication etc.). The analysis of individual effects available procedures within the impact assessment
does not permit an overall conclusion from an envi- phase. In addition, a normalization and weighting
ronmental point of view on the overall preference of procedure is suggested, which enables the aggrega-
one or another production strategy. Different environ- tion of the environmental impacts into two summa-
mental management tools such as Eco Management rizing indicators, one for impacts on eco-systems
and Audit Scheme (EMAS; Spindler, 1998) or Krite- and human health, and the other one for resource
rien umweltverträglicher Landbewirtschaftung (KUL depletion.
[criteria for an environmentally compatible agricul- A subsequent paper will describe the application
ture]; Eckert et al., 1999) have been developed to of this methodology to investigate the environmen-
investigate the overall environmental performance of tal impact of different production systems of winter
farms. Such systems are used (1) to detect options for wheat.
improvement and (2) to compare or to monitor the
environmental impact of farms. In order to analyze
agricultural products, the environmental impact of the 2. General introduction of the LCA methodology
product itself and the entire production system to pro-
duce it should be investigated. The life cycle assess- LCA is a methodology to assess all environmental
ment (LCA) methodology was designed to study all impacts associated with a product, process or activity
environmental impacts connected to an entire produc- by accounting and evaluating the resource consump-
tion system (Consoli et al., 1993). For crop production tion and the emissions. According to ISO (ISO, 1997)
not only on-field activities but also all impacts related LCA is divided into four steps, which are (1) goal and
to the production of farm inputs, such as emissions scope definition, (2) inventory analysis, (3) impact as-
and resource consumption due to the production of sessment, and (4) interpretation.
fertilizers, are included. All impacts are related to
one common unit (e.g. 1 tonne of wheat grain) and 2.1. Goal and scope definition
summarized into environmental effects (such as cli-
mate change or acidification) or even aggregated into The first step in LCA is the definition of the goal
a summarizing environmental index. Such an index and scope of the study. This step defines the reasons
allows the ranking of different product or production for the LCA study and the intended use of the results.
alternatives according to their overall environmental For LCA studies in the agricultural sector this could
performance. be for instance to investigate the environmental im-
International Organization for Standardization pacts of different intensities in crop production or to
(ISO) and Society for Environmental Toxicology analyze the advantages and disadvantages of intensive
and Chemistry (SETAC) (ISO, 1997; Consoli et al., or extensive arable farming systems. Furthermore,
1993; Udo de Haes et al., 1999a,b) provide a general this step describes the system under investigation, its
description of the LCA methodology. However, the function, and boundaries. Subsequently, a reference
impact assessment procedure, the aggregation meth- unit (functional unit; ISO, 1998) is defined, to which
ods for the different impact categories and the final all environmental impacts are related to, and which
calculation of a summarizing environmental index should represent the function of the analyzed system.
are still in debate. Furthermore, if currently available In Fig. 1 an illustration of an arable farming system
LCA applications are used to investigate agricul- with the primary function to produce winter wheat is
tural products or processes, the methods reveal some presented. The appropriate functional unit (FU) for
shortcomings, such as the missing integration of im- this system is one tonne of grain.
 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 249

Fig. 1. System boundary, relevant in- and outputs, and functional unit of a wheat production system.

2.2. Life cycle inventory (LCI) analysis tribute to the respective impact category (ISO, 2000).
An example for such an indicator is the global warm-
The inventory analysis compiles all resources that ing potential (GWP) expressed in CO2 -equivalents,
are needed for and all emissions that are released which is derived from the rate of CO2 , CH4 , N2 O and
by the specific system under investigation and relates CFC emissions multiplied by their respective charac-
them to the defined functional unit (ISO, 1998). terization factor (e.g. 1 for CO2 , 310 for N2 O). Ac-
cording to ISO the aggregation of inventory results to
2.3. Life cycle impact assessment (LCIA) impact categories is mandatory in LCIA (ISO, 2000).
The list of impact category indicator values for a sys-
The impact assessment aims at a further interpreta- tem under investigation is called its environmental pro-
tion of the LCI data. The inventory data are multiplied file. Table 1 gives a list of the impact categories as
by characterization factors (CF) to give indicators proposed by the SETAC-Europe Working Group on
for the so-called environmental impact categories LCIA (WIA-2) (Udo de Haes et al., 1999a,b).
(Eq. (1)).
impact category indicatori Table 1

List of environmental effects (impact categories) treated in LCA
= (Ej or Rj ) × CFi,j (1)
j General distinction Impact category

where: impact category indicatori = indicator value Input related categories Depletion of abiotic resources
per functional unit for impact category i; Ej or Land use
Output related categories Climate change (global warming)
Rj = release of emission j or consumption of resource Stratospheric ozone depletion
j per functional unit; CFi,j = characterization factor Human toxicity, ecotoxicity
for emission j or resource j contributing to impact Photo-oxidant formation
category i. (‘summer smog’)
The characterization factors represent the potential Acidification
Nutrification (eutrophication)
of a single emission or resource consumption to con-
250 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

For further interpretation of the environmental pro- 3. Life cycle impact assessment tailored to plant
file, a normalization step relates the indicator values nutrition in crop production
to reference values. The resulting normalized indi-
cator values give the share of the analyzed system This study describes an LCIA method, which has
in the defined reference, e.g. European values for been refined to evaluate the environmental impact of
the respective impact categories. For a system un- plant nutrition in arable crop production and which in-
der investigation this would mean the division of the cludes a new combination of impact assessment pro-
GWP calculated for this specific system by the total cedures. In addition this LCIA approach includes new
Global Warming Potential for a defined region, e.g. normalization values and weighting factors in order
Europe. to enable a conclusion on the overall environmental
In the following weighting step the normalized preference of different plant nutrition systems.
indicator values are multiplied by weighting factors,
which represent the potential of the different environ- 3.1. Characterization
mental impact categories to harm natural ecosystems,
human health and resources. For example the normal- For most impact categories various methods for the
ized indicator value for global warming for a product aggregation of LCI data to impact category indicators
or system under analysis is multiplied by a specific are described in the literature. For climate change, hu-
weighting factor for global warming. Subsequently, man toxicity, eco-toxicity and acidification appropri-
the weighted indicator values can be summed up ate characterization methods are available. However,
to one overall environmental indicator. ISO (2000) for depletion of abiotic resources and land use impact
describes both, normalization and weighting as op- categories, a need for improvement has been identified
tional elements of LCIA. Fig. 2 gives an overview (Brentrup et al., 2002a,b). Since in crop production
of the general LCIA procedure and its different systems there are no emissions, which contribute to
elements. the depletion of the stratospheric ozone layer (such as

Fig. 2. The general life cycle impact assessment procedure.

 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 251

chlorofluorocarbons), this impact category has been do not consider that many resources have different
excluded from the suggested LCA approach. Recent functions and are not equivalent to each other. By
LCA studies on crop production systems have shown contrast, emissions are usually aggregated based on
that the contribution of arable farming to the for- their ‘function’, i.e. the effect on the environment
mation of tropospheric photo-oxidants (e.g. ozone, (e.g. N2 O, CH4 , CO2 →climate change). Therefore,
‘summer smog’) due to emissions of nitrogen ox- a new aggregation method has been developed to
ides (NOx) and volatile organic compounds (VOC) separate the use of different abiotic resources into
is negligible compared to the contribution of other different impact sub-categories and aggregate them to
human activities like traffic or industrial production indicators according to the primary function of the re-
(Brentrup et al., 2001; Küsters and Brentrup, 1999; sources (e.g. coal, natural gas, oil→depletion of fossil
Küsters and Jenssen, 1998). Therefore, the ‘formation fuels; Brentrup et al., 2002a). For LCA studies on
of photo-oxidants’ impact category can be regarded as arable crop production the consumption of fossil fuels
not relevant to arable crop production systems. How- and minerals such as phosphate, potash and lime are
ever, if this impact category is to be considered in an sub-categories of particular importance. Table 2 gives
LCA study, a characterization method developed by the characterization factors for abiotic resources typi-
Hauschild et al. (2000) can be recommended, as this cally consumed in an agricultural production system.
is the only method that considers the impact of both, The characterization result (RDI, resource depletion
NOx and VOC emissions. indicators) for the sub-categories can be calculated
The chosen aggregation methods are described sep- according to Eq. (1).
arately for each impact category in the following sec-
tions. 3.1.2. Land use
The ‘land use’ impact category describes the en-
3.1.1. Depletion of abiotic resources vironmental impacts of utilizing and reshaping land
The issue related to the depletion of abiotic re- for human purposes (Heijungs et al., 1997; Lindeijer
sources, such as fossil fuels or minerals is their et al., 1998; Müller-Wenk, 1998; Köllner, 2000).
decreasing availability for future generations. In cur- The environmental consequences of land use such
rently available LCIA methods (e.g. Goedkoop and as arable farming or urban settlement are the de-
Spriensma, 1999; Guinée, 2001), the consumption creasing availability of habitats and the decreasing
of different abiotic resources is aggregated to one diversity of wildlife species (EEA, 1998; BfN, 1999;
summarizing indicator for resource depletion within Statistisches Bundesamt, 1999). Current approaches
the characterization step. However, these methods to assess the impact of human land use are mainly

Table 2
‘Depletion of abiotic resources’ impact category: characterization factors for the aggregation of single resources to resource depletion
indicators for each impact sub-category
Resource Unit CFa Impact sub-category (unit of resource depletion indicator, RDI)

Oil kg OEb 42.868 Depletion of fossil fuels (RDIfossil fuels in MJ)

Natural gas m3 31.736
Hard coal kg 29.704
Lignite kg 8.506
Phosphate rock kg 0.25 Depletion of phosphate rock (RDIphosphate rock in kg P2 O5 )
Raw phosphate kg 0.32
Potash, potassium chloride kg 0.105 Depletion of potash (RDIpotash in kg K2 O)
Limestone/lime kg 0.54 Depletion of lime (RDIlime in kg CaO)
Dolomite kg 0.30
a CF, characterization factor (heat values [in MJ/kg or m3 ] for fossil fuels taken from BMWi, 1995; P O , K O, CaO contents [in
2 5 2
kg/kg] taken from Patyk and Reinhardt, 1997 and www.dolomit.de)
b OE, crude oil equivalents.
252 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

based on empirical investigations of species diversity the assessment of land use should be regionalized.
(e.g. of vascular plants; Goedkoop and Spriensma, Consequently, the impact category ‘land use’ is sub-
1999; Köllner, 2000). There are two main problems divided into different sub-categories. The definition
associated with these approaches. First, it is difficult of these sub-categories should be based on ecologi-
to determine a natural reference situation, to which cally homogenous land units as for example the bio-
the situation of land under use can be compared, geographic regions for Europe (EEA, 1998). Fig. 3
because the diversity of species per area within Eu- shows the biogeographic regions of Europe. The land
rope varies naturally by a factor of 10–15 (BfN, use indicator (NDI, naturalness degradation indicator)
1999). Furthermore, not only the number, but also for a system under investigation can be calculated ac-
the composition of species (e.g. share of neophytes) cording to Eq. (1). If land is used in more than one
is important for the assessment of the influence of biogeographic region, this calculation should be done
human land use (Kretschmer et al., 1997). Conse- separately for each biogeographic region.
quently, a new method for the assessment of land
use impacts has been developed (Brentrup et al., 3.1.3. Climate change
2002b). Emissions of gases with specific radiative char-
This new method treats ‘natural land’ like a re- acteristics like carbon dioxide (CO2 ) and nitrous
source and it is assumed that the utilization of land oxide (N2 O) lead to an unnatural warming of the
leads to a reduced availability of this resource. Natural Earth’s surface, which in turn will cause global and
land can be defined as the sum of actually uninflu- regional climatic changes. This environmental impact
enced area and the accumulated remaining naturalness is commonly described as ‘global warming’. The
of the land under use. To determine the remaining term ‘climate change’ indicates that the possible con-
naturalness of land under use, the Hemeroby concept sequences of global warming concern more elements
(Kowarik, 1999) can be applied. Hemeroby is a mea- of the global climate than only the temperature (e.g.
sure for the ‘human influence on ecosystems’ and precipitation, wind). The main anthropogenic contrib-
is therefore used to characterize the environmental utors to the enhanced greenhouse effect are (sorted
impact of different land use types. The characteriza- according to their contribution): CO2 (65%), methane
tion factors are described as ‘naturalness degradation (CH4 , 20%), halogenated gases (e.g. CFCs, 10%)
potential’ (NDP) and given in Table 3 for selected and N2 O (5%; EEA, 1998). The different potential
land use types. NDP values for additional land use of these emissions to contribute to climate change
types and a detailed description of this aggregation is represented by their GWP (Table 4). The climate
method can be found in Brentrup et al. (2002b). change indicator for a system under investigation can
On a larger geographical scale the ‘natural area/ be calculated according to Eq. (1).
naturalness’ resource is not homogeneous. Therefore
3.1.4. Toxicity
This impact category includes all direct toxic ef-
Table 3
fects of emissions on humans (human toxicity) and
‘Land use’ impact category: characterization factors (NDP, natu-
ralness degradation potentials) for selected land use types
Land use type (in ha year) NDPa Table 4
‘Climate change’ impact category: characterization factors (global
Continuous urban area 0.95 warming potentials) for selected greenhouse gases (Houghton et al.,
Industrial/commercial units 0.95 1993)
Road/rail networks 0.90
Discontinuous urban area 0.85 Substance (in kg) Global warming potential (GWP
Intensive arable land 0.80 100)a (in kg CO2 -equivalents per kg)
Extensive arable land 0.70 CO2 1
Green urban area 0.70 CH4 21
Intensive permanent pasture 0.60 N2 O 310
Extensive permanent pasture 0.50 a GWP 100, global warming potential for the time horizon of
a NDP, naturalness degradation potential. 100 years.
 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 253

Fig. 3. The biogeographic regions of Europe (adopted from EEA, 1998).

ecosystems (eco-toxicity). Emissions, which may be ated with specific weather conditions together with
potentially toxic and are released by arable farming high emission rates particularly of SO2 and suspended
systems, are (1) inorganic air pollutants like NH3 , particles, leading to respiratory problems. However,
SO2 and NOx, (2) plant protection substances, and own investigations have shown that in arable farming
(3) heavy metals. systems at least 70% of the SO2 , NOx, NH3 , CO and
Inorganic air emissions (e.g. SO2 , NOx, CO, particle emissions are released during on-field activ-
NH3 , particles) are potentially toxic to humans due ities (e.g. tractor use, fertilizer application) in spring
to their contribution to winter smog episodes with and summer. Also, because of the short atmospheric
high concentrations of air pollutants in urban areas residence time of these substances, the exclusion of
(Stanners and Bourdeau, 1995). The contribution of these emissions as far as arable farming systems are
these emissions to other environmental problems like concerned from the ‘toxicity’ impact category can be
acidification or eutrophication is accounted for in the justified, because they are unlikely to contribute to
respective impact categories. Winter smog is associ- the winter smog problem.
254 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

Plant protection substances are applied in or- the environmental fate, the probable exposure of hu-
der to control certain organisms (e.g. weeds, fungi, mans or ecosystems, and the potential toxic effects.
and insects) in order to improve the productivity For human toxicity Goedkoop and Spriensma
of arable farming. However, via wind drift, evapo- (1999) use the concept of disability-adjusted life
ration, leaching, and surface run-off, a part of the years (DALYs), which was developed on behalf of
applied agro-chemicals may impact upon terrestrial WHO and World Bank (Murray, 1994). In the DALY
and aquatic ecosystems or even humans (Hauschild, concept, weights for the different severity of human
2000). To estimate the rate of unintended emissions health effects have been established. These weights
of toxic substances, their fate in the environment allow for comparisons between time lived with a cer-
and the final effects on ecosystems and humans dif- tain limitation and time lost due to premature mortal-
ferent models have been developed (Goedkoop and ity. The human toxicity potential (HTP) for emissions
Spriensma, 1999; Guinée et al., 1996; Huijbregts, of toxic substances is therefore expressed in DALYs.
2001; Jolliet and Crettaz, 1997). These models con- For eco-toxicity, the eco-toxicity potential (ETP)
centrate on toxic substances other than pesticides and has been calculated by Huijbregts (2001) for 5
therefore do not include the currently available plant different types of ecosystems: (1) terrestrial; (2)
protection agents. Since the environmental impact of fresh water; (3) sea water; (4) fresh water sedi-
plant nutrition within arable crop production was in ment; and (5) sea water sediment. The ETP are
focus of this study, it was out of scope of this study expressed relative to a reference substance, which
to develop a new impact assessment approach for is 1,4-dichlorobenzene (1,4DCB) and are there-
potential toxic impacts of plant protection measures. fore called 1,4DCB-equivalents. The human and
Other ‘non-toxic’ environmental impacts, which are eco-toxicity potentials for cadmium emissions to soil
due to the production, packaging, transport and appli- are given in Table 5. Toxicity potentials for other
cation of plant protection agents (e.g. consumption of toxic substances like other heavy metals or persistent
fossil fuels, emissions related to energy use), are in- organic pollutants can be found in Goedkoop and
cluded in the relevant impact categories. Spriensma (1999) and Huijbregts (2001). The toxic-
The agricultural use of mineral phosphate fertiliz- ity potentials for a system under investigation can be
ers and organic materials like slurry, sewage sludge calculated according to Eq. (1).
or compost may lead to emissions of heavy metals to
soils. The contamination of these materials with heavy 3.1.5. Acidification
metals varies substantially depending on the origin Acidification is mainly caused by air emissions of
of the raw material (P rock, industrial and household sulfur dioxide (SO2 , share: 36% for EU15), nitrogen
waste). For heavy metal emissions to soil, models de- oxides (NOx, 33%) and ammonia (NH3 , 31%; EEA,
veloped by Goedkoop and Spriensma (1999) for hu- 2001). SO2 primarily originates from combustion of
man toxicity and Huijbregts (2001) for eco-toxicity sulfur-containing coal and oil, NOx from combustion
are most suitable for the estimation of their toxic po- processes in motor vehicles, whereas NH3 predomi-
tential. Both models take into account information on nantly originates from animal husbandry (EEA, 1998).

Table 5
‘Human and eco-toxicity’ impact category: characterization factors (toxicity potentials) for cadmium (Cd) emissions to soil
Sub-category Unit of sub-category indicator Toxicity potential per kg Cd to soil

Human toxicity DALYa 3.98 × 10−3

Terrestrial ecotoxicity kg 1,4-DCB-equiv.b 1.7 × 102
Aquatic ecotoxicity, fresh water kg 1,4-DCB-equiv. 7.8 × 102
Aquatic ecotoxicity, marine kg 1,4-DCB-equiv. 1.1 × 105
Sediment ecotoxicity, fresh water kg 1,4-DCB-equiv. 2.0 × 103
Sediment ecotoxicity, marine kg 1,4-DCB-equiv. 1.1 × 105
a DALY, disability adjusted life-years (Goedkoop and Spriensma, 1999).
b 1,4-DCB-equiv., 1,4-dichlorobenzene-equivalents (Huijbregts, 2001).
 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 255

SO2 , NOx and NH3 are also released during arable aquatic plants (i.e. mainly algae) respond differently
crop production. In particular the use of organic and to an additional supply of nutrients. Therefore, in this
mineral fertilizers can result in important emissions of LCIA approach the eutrophication impact category is
NH3 due to volatilization during and after application separated into terrestrial and aquatic eutrophication.
of urea and ammonium-containing fertilizer (Brentrup
et al., 2000). 3.1.6.1. Terrestrial eutrophication. Huijbregts
Acid deposition has negative effects on terrestrial (2001) developed a characterization method for ter-
and aquatic ecosystems. restrial eutrophication that considers atmospheric
The effect of potentially acidifying emissions de- pathways, deposition patterns and eutrophication ef-
pends on the deposition pattern (fate) and the sus- fects of NOx and NH3 emissions. Since for terrestrial
ceptibility of the receiving area to acidification (e.g. ecosystems nitrogen is the major limiting nutrient,
buffer capacity, CaCO3 -content). Comparing the acid- NOx and NH3 depositions are the most important
ification potential of NH3 emissions (expressed in contributors to terrestrial eutrophication (Finnveden
SO2 -equivalents) from Sweden and Greece illustrates and Potting, 1999; Potting et al., 2000). Huijbregts
this effect. Whereas 1kg NH3 released in Greece (2001) calculated regionalized terrestrial eutrophica-
results in only 0.13 kg SO2 -equivalents, the same tion potentials (TEP) expressed in NOx-equivalents
emission released in Sweden has an acidification po- (Table 6). The terrestrial eutrophication potential
tential of 4.4 kg SO2 -equivalents (Huijbregts, 2001). for any system under evaluation can be calculated
This difference is due to the different deposition pat- according to Eq. (1).
tern of the emission and to the different sensitivity of
the receiving area (e.g. buffer capacity of soils and 3.1.6.2. Aquatic eutrophication. Important anthro-
surface waters). This illustrates the importance of a pogenic N and P emissions to surface waters are: (1)
site-specific characterization approach. deposition of airborne NOx and NH3 on surface wa-
A method developed by Huijbregts (2001) includes ters from combustion processes and livestock farm-
this kind of information and has therefore been se- ing; (2) direct effluents of N and P (point sources, e.g.
lected for this LCIA approach. As a result separate municipalities, industries); and (3) diffuse losses of N
characterization factors for acidifying emissions re- via leaching (non-point sources, e.g. arable farming)
leased in different European countries are proposed. (Klepper et al., 1995). In LCIA it should be taken into
In addition Huijbregts (2001) calculated average char- account that not all of the nutrients initially released
acterization factors for Western, Eastern and total Eu- to air and soil actually end up in surface waters and
rope, which should be used, if the source region of an contribute to eutrophication.
emission in not known in more detail. Table 6 gives Fate factors developed by Huijbregts and Seppälä
the CFs for SO2 , NOx and NH3 emissions released (2000) enable the approximation of N deposition on
in Western European countries, which can be used to surface waters caused by airborne NOx and NH3 emis-
calculate the acidification indicator for a system under sions. The factors indicate which fraction of a NOx
analysis according to Eq. (1). or NH3 emission released in different European coun-
tries reaches marine ecosystems (Table 7). The calcu-
3.1.6. Eutrophication lation is based on a modified RAINS model, which
Eutrophication can be defined as an undesired has been originally developed as a tool for the inte-
increase in biomass production in aquatic and ter- grated assessment of acidification, terrestrial eutroph-
restrial ecosystems caused by high nutrient inputs, ication and photo-chemical ozone creation in Europe
which result in a shift in species composition. In (Huijbregts, 2001). Freshwater systems are not con-
surface waters eutrophication is particularly serious sidered, because they only have a small fraction of the
because it can lead to algal blooms and the sub- total surface water area and are mainly limited by P
sequent oxygen-consuming degradation processes, and not by N.
which finally may result in the death of the total For direct effluents of N and P into surface water
aquatic biocoenosis (EEA, 1998; Potting et al., 2000). it is assumed, that these nutrients are either directly
Terrestrial vegetation (i.e. mainly higher plants) and available for eutrophication (P in fresh water systems)
256 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

Table 6
‘Acidification’ and ‘terrestrial eutrophication’ impact categories: regionalized characterization factors (acidification potentials for SO2 , NOx,
NH3 and terrestrial eutrophication potentials for NOx, NH3 ; Huijbregts, 2001; modified)
Emission source Acidification Potential Terrestrial Eutrophication
region (in kg SO2 -equivalents Potential (in kg NOx-
per kg emission) equivalents per kg emission)
SO2 NOx NH3 NOx NH3

Switzerlanda 1.00 0.28 1.30 1.00 5.00

Austria 1.00 0.27 1.30 0.89 4.20
Belgium 1.00 0.49 1.00 1.20 2.90
Denmark 1.80 0.88 1.50 1.60 2.50
Finland 5.00 1.90 6.40 3.50 11.50
France 1.10 0.43 2.00 1.30 6.40
Germany 1.30 0.53 1.50 1.50 4.60
Greece 0.066 0.037 0.13 0.27 1.50
Ireland 0.57 0.34 0.79 0.52 1.00
Italy 0.46 0.13 0.59 0.60 2.80
Luxembourg 1.30 0.50 1.50 1.40 4.40
Netherlands 0.92 0.51 1.00 1.10 2.30
Norway 3.80 1.20 6.00 1.60 6.20
Portugal 0.18 0.08 0.28 0.49 2.40
Spain 0.22 0.10 0.27 0.52 2.00
Sweden 3.80 1.30 4.40 2.10 5.70
UK 0.86 0.43 1.50 0.76 1.70
Western Europe, average 0.79 0.41 1.30 0.99 3.70
Eastern Europe, average 1.60 0.70 1.80 1.70 5.00
Europe, average 1.20 0.50 1.60 1.20 4.30
a All acidification and terrestrial eutrophication potentials are calculated relative to the potential of 1 kg SO and 1 kg NOx, respectively,
2
released in Switzerland.

or will be transported to places where they poten- tion methods proposed for aquatic eutrophication by
tially contribute to nutrient enrichment (N in sea water Potting et al. (2000) and Huijbregts (2001) ignore
systems). Therefore the total N and P emission rates the strong dependency on the conditions of any agri-
from point sources (e.g. from fertilizer production or cultural production system under investigation. They
wastewater treatment plants) are considered. suggest the application of fixed national factors for
The main pathway for diffuse N emissions from nutrient losses due to fertilizer and manure applica-
soil to aquatic ecosystems is via nitrate (NO3 ) leach- tion. Since such a general procedure is inadequate for
ing. There is no linear relationship between N-input to an LCA method tailored to arable farming systems, a
the soil (e.g. as fertilizer) and nitrate load of ground- site- and study-specific estimation of N losses is sug-
and surface water. Nitrate losses to groundwater via gested in this LCA method. Methods, which enable an
leaching are strongly dependent on agricultural man- estimation of fertilizer N reaching groundwater as ni-
agement (e.g. fertilization rates, N removal with har- trate, are described in detail by Brentrup et al. (2000).
vested crops) as well as site-specific soil and climate According to Potting et al. (2000) the calculated NO3
conditions (e.g. field capacity, rate of drainage wa- immission to groundwater should be further reduced
ter; Brentrup et al., 2000). Therefore, NO3 leaching by 30% assuming denitrification losses on the way
losses from soil to groundwater are highly variable from groundwater to surface water and, finally, the sea.
and should be carefully estimated considering all rel- Subsequent to this fate analysis, the estimated rates
evant parameters determining the NO3 content in the of the N and P immissions to surface waters due to
soil at the beginning of the leaching period in autumn airborne emissions, effluents and diffuse losses can
as well as site-specific soil and climate characteristics be finally aggregated to a total aquatic eutrophication
determining NO3 load in the groundwater. Aggrega- potential. For this aggregation characterization factors
 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 257

Table 7 trient emissions, the second direct effluents from point

Regionalized fate factors for airborne NOx and NH3 emissions sources and the third immissions of nitrate to surface
to approximate the related immissions to marine surface wa-
ters (Huijbregts and Seppälä, 2000; modified) for the ‘Aquatic
waters due to on-field N losses.
eutrophication’ impact sub-category  


Emission source region Fate factors AEP = EAi,j × FFi,j × CFi
i
NOx NH3  


Austria 0.088 0.048 + EWi × CFi
Belgium 0.240 0.230
i
Denmark 0.290 0.430  
Finland 0.200 0.250 + ISNO3 × RFNO3 × CFNO3 (2)
France 0.230 0.250
Germany 0.170 0.140 where: AEP = aquatic eutrophication potential [in kg
Greece 0.180 0.230 PO4 -equivalents/FU]; EAi,j = airborne emission of
Ireland 0.470 0.460
Italy 0.190 0.210
nutrient i released in region j [in kg/FU]; FFi,j = fate
Luxembourg 0.170 0.110 factor for airborne emission i released in region j
Netherlands 0.280 0.260 (Table 7); CFi = characterization factor for nutri-
Norway 0.280 0.450 ent i [in kg PO4 -equivalents/kg]; EWi = effluent of
Portugal 0.160 0.230 nutrient i [in kg/FU]; ISNO3 = nitrate immission to
Spain 0.170 0.160
Sweden 0.240 0.330
groundwater [in kg NO3 /FU]; RFNO3 = reduction
Switzerland 0.094 0.051 factor for nitrate immission to surface water (30%);
UK 0.390 0.430 CFNO3 = characterization factor for nitrate [in kg
Western Europe, average 0.250 0.240 PO4 -equivalents/kg].
Eastern Europe, average 0.089 0.072
Europe, average 0.210 0.160
3.2. Normalization

Even after aggregation of the inventory data to im-

can be employed, which are based on the typical nutri-
pact categories (Section 3.1) it is not possible to con-
ent ratio of a phytoplankton (Redfield ratio; Heijungs,
clude on the relative importance of these values. A
1992). The factors given in Table 8 are calculated rela-
high indicator value may represent only a small con-
tive to the eutrophication potential of phosphate (PO4 ).
tribution to the total environmental effect, whereas a
The AEP for a system under analysis can be cal-
several times smaller indicator value may represent an
culated according to Eq. (2). Eq. (2) consists of three
important contribution to the respective environmen-
main terms, the first of which comprises airborne nu-
tal effect. Thus, ‘the aim of the normalization of in-
dicator results is to better understand the magnitude
Table 8 for each indicator result of the product system under
‘Aquatic eutrophication’ impact sub-category: characterization fac- study’ (ISO, 2000). During the normalization the indi-
tors (aquatic eutrophication potentials) for N and P emissions cator results per functional unit (i.e. a tonne of grain)
Substance Aquatic Eutrophication Potential (in are related to the respective indicator results for a de-
(in kg) kg PO4 -equivalents per kg emission) fined reference area according to Eq. (3).
N 0.42 Ii
NH3 0.35 Ni = (3)
NH4 0.33 NVi
NOx 0.13
NO3 0.10 where: Ni = normalization result per functional unit
NO3 –N 0.42 for impact category i; Ii = indicator value per func-
P 3.06 tional unit for impact category i; NVi = indicator value
P2 O5 1.34
for a reference situation (e.g. per person in Europe)
PO4 1.00
for impact category i = normalization value.
258 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

Table 9
Normalization values (NV) for the different impact categories (indicator values per person in Europea
Impact category Impact sub-category Unit Yearb NVc Europe

Abiotic resources Lime consumption kg CaO –d

Phosphate consumption kg P2 O5 1999 7.66
Potash consumption kg K2 O 1999 8.14
Fossil fuel consumption MJ 1999 1.33 × 105
Land use Land use, Alpine region ha year 1998 1.22 × 104
Land use, Anatolian region ha year 1998 –d
Land use, Arctic region ha year 1998 –d
Land use, Atlantic region ha year 1998 1.79 × 104
Land use, Black sea region ha year 1998 1.57 × 104
Land use, Boreal region ha year 1998 1.58 × 104
Land use, Continental region ha year 1998 1.86 × 104
Land use, Macaronesian region ha year 1998 1.00 × 104
Land use, Mediterranean region ha year 1998 1.61 × 104
Land use, Pannonian region ha year 1998 2.09 × 104
Land use, Steppic region ha year 1998 2.11 × 104
Climate change kg CO2 -equiv. 1999 9.73 × 103
Toxicity Human toxicity DALY 1995/99 7.50 × 10−3
Terrestrial eco-toxicity kg 1,4 DCB-equiv. 1995/99 1.15 × 102
Freshwater aquatic eco-toxicity kg 1,4 DCB-equiv. 1995/99 1.24 × 103
Marine aquatic eco-toxicity kg 1,4 DCB-equiv. 1995/99 2.88 × 105
Freshwater sediment eco-toxicity kg 1,4 DCB-equiv. 1995/99 1.28 × 103
Marine sediment eco-toxicity kg 1,4 DCB-equiv. 1995/99 2.65 × 105
Acidification kg SO2 -equiv. 1999 47.7
Eutrophication Terrestrial eutrophication kg NOx-equiv. 1999 60.7
Aquatic eutrophication kg PO4 -equiv. 1999 8.56
a European indicator values were divided by European population figure for 1994 (727 Mio; EEA, 1998) in order to keep resulting
numbers manageable.
b Most recent data available were chosen.
c NV, normalization value for Europe (Europe = Albania, Austria, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia,

Czech Republic, Denmark, Estonia, Finland, France, FYR of Macedonia, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia,
Lithuania, Luxembourg, Moldova, Netherlands, Norway, Poland, Portugal, Romania, Russian Federation, Slovakia, Slovenia, Spain, Sweden,
Switzerland, Ukraine, UK, Yugoslavia).
d No data available.

The decision about which reference situation shall Data for a worldwide normalization are currently only
be used depends on the subsequent weighting proce- available as crude estimates based on extrapolations
dure as well as on the availability of normalization (Guinée, 1996; Van Oers, 2001).
data (Lindeijer, 1996). Table 9 gives the European nor-
malization values suggested in this LCIA method. The 3.3. Weighting
European reference situation has been chosen as an
example due to data availability and because the cal- The weighting step is necessary to conclude on the
culation of weighting factors, given in the following overall environmental preference of one or the other
section, is also mainly based on environmental targets products or processes under investigation. Weighting
for Europe. The European normalization values sug- means an evaluation of the different effects such as
gested in this study have been calculated per person global warming or acidification according to their po-
(Table 9), in order to keep the resulting numbers in tential to harm the environment. In LCA the so-called
a manageable range. Normalization values could be safeguard subjects ‘human health, natural ecosystems,
calculated for any reference region as long as reliable and resources’ represent the environment (Consoli
data for the different impact categories are available. et al., 1993; Lindfors et al., 1995). Weighting allows
 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 259

the further interpretation of complex environmental egories, no overall international targets for human-
profiles in order to support users of LCA studies or eco-toxicity could be found. Consequently, tox-
with clear and aggregated results. Weighting factors icity is only considered within characterization and
represents the environmental weight of each impact normalization, but excluded from the weighting step.
category. The higher the weighting factor for an im- However, in contrast to other LCA methods (e.g.
pact category, the higher is the potential of that im- Goedkoop, 1995; Goedkoop and Spriensma, 1999;
pact category to harm the environment. For this LCA Steen, 1999), this approach suggests two separate fi-
method weighting factors were derived by using au- nal indicators for (a) resource depletion and (b) im-
thorized environmental goals like the Kyoto protocol pacts on natural ecosystems and human health. This
for climate change in the so-called ‘distance-to-target’ separation is important, because the problems related
principle (Müller-Wenk, 1996; Lindeijer, 1996). to the depletion of abiotic resources are substantially
‘Distance-to-target’ means a comparison of the cur- different to those related to the other impact cate-
rent level of an environmental effect in a certain gories. The impact categories other than the depletion
region and time to a target level of the same effect. of abiotic resources have direct effects on either natu-
The ratio between both values gives the weighting ral ecosystems (land use, acidification, eutrophication,
factor for the environmental effect (Eq. (4)). eco-toxicity), or human health (human toxicity), or
CIi,j,k even on both (climate change). In contrast, the deple-
WFi,j,k = (4) tion of abiotic resources, i.e. the decreasing availabil-
TIi,j,k
ity of raw materials for future generations has no direct
where: WFi,j,k = weighting factor for impact category impact on human health or the shape of natural ecosys-
i, valid for region j and year k; CIi,j,k = current indi- tems. The environmental impacts associated with the
cator value for impact category i for region j and year extraction and processing of resources (e.g. land use,
k; TIi,j,k = target indicator value for impact category emissions or effluents) are considered in the respec-
i for region j and year k. tive impact categories. Resources such as fossil fuels
The selection of accepted environmental targets or phosphate rock rather have an intrinsic value for hu-
makes it possible to consider that the evaluation of mans, as they substantially contribute to development
environmental problems may differ substantially be- and wealth creation (e.g. through mobility and nutri-
tween regions and societies. Accepted environmental tion). Therefore, the availability of abiotic resources
goals implicitly include a range of criteria for the for future generations is more an economic and social
evaluation of environmental impacts like the magni- issue than an environmental problem. This is the rea-
tude, the reversibility, and the geographical extent of son why this LCA method separates the aggregated re-
the ecological damage, the uncertainty of the dam- source depletion indicator (RDI) from the aggregated
age and the substitutability of the damaged item environmental indicator (EcoX). The aggregation of
(Müller-Wenk, 1996). Furthermore, accepted envi- these two indicators would necessitate the calculation
ronmental goals consider not only the environmental of a kind of a sustainability indicator. The develop-
but also economic and social aspects of the respective ment of a sustainability indicator would be an ambi-
impacts (e.g. Kyoto protocol, UN-FCCC, 1998). The tious goal for further research, but should certainly
application of such authorized targets can be regarded comprise more economic and social aspects besides
as the most objective and justified way to include resource depletion (e.g. income, employment, prices,
value judgments within the weighting of different en- food security and quality, rural development, etc.).
vironmental impacts. Therefore, the distance-to-target The environmental index ‘EcoX’ can be calculated
principle with accepted environmental goals as targets for a specific product or system under examination by
for the calculation of weighting factors is suggested multiplying the normalization result for each impact
for this LCA method. Table 10 gives the weight- category by the respective weighting factor and sum-
ing factors for the environmental effects and their ming up the weighted results (Eq. (5)).
sub-categories. Due to the fact that emissions of nu-

merous substances to air, water and soil contribute to EcoX = Ni × WFi (5)
the ‘human toxicity’ and ‘eco-toxicity’ impact cat- i
260 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

Table 10
Weighting factors for the impact categories and data used for the calculation (current status, target value)
Impact category and Indicator value (for unit see Basis for target Weighting
impact sub-category Table 10) factor
Current status (year) Target value

Abiotic resource depletion

Lime consumption 1.16 × 1011 (1999) –a 100 years availabilityb 0.00
Phosphate consumption 4.32 × 1010 (1999) 3.60 × 1010 1.20
Potash consumption 2.56 × 1010 (1999) 8.40 × 1010 0.30
Fossil fuel consumption 3.25 × 1014 (1999) 3.08 × 1014 1.05
Land use
Alpine region 8.84 × 106 (1998) 8.84 × 106 Maintenance of current land 1.00
use intensity in Europec
Atlantic region 4.34 × 107 (1998) 4.34 × 107 –
Black sea region 5.62 × 105 (1998) 5.62 × 105 –
Boreal region 8.77 × 106 (1998) 8.77 × 106 –
Continental region 7.71 × 107 (1998) 7.71 × 107 –
Macaronesian region 2.30 × 105 (1998) 2.30 × 105 –
Mediterranean region 4.40 × 107 (1998) 4.40 × 107 –
Pannonian region 7.22 × 106 (1998) 7.22 × 106 –
Steppic region 2.00 × 106 (1998) 2.00 × 106 –
Climate change 3.50 × 106 (1998) 3.32 × 106 UN-FCCC (1998)d 1.06
Toxicity
Human toxicity, eco-toxicity – – No target definede –
Acidification 1.42 × 104 (1999) 1.06 × 104 UN-ECE/CLRTAP (1999)d 1.34
Eutrophication
Terrestrial eutrophication 2.46 × 104 (1999) 1.95 × 104 UN-ECE/CLRTAP (1999)d 1.26
Aquatic eutrophication 1.10 × 106 (1995) 8.07 × 105 OSPAR (1995) –
HELCOM (2001)f 1.37 – –
a No problem expected due to very large lime reserves (USGS, 2001).
b Target based on assumption that an availability of the resource for at least 100 years is sufficient for the development of substitution
or recycling techniques (Brentrup et al., 2002a).
c Target based on the assumption that the current land use intensity in each biogeographic region of Europe is tolerable and should be

maintained (Brentrup et al., 2002b).

d Values based on emission rates and reduction targets for Western European countries.
e Only separate targets for specific groups of toxic substances (e.g. heavy metals to air), but no overall international target for the

reduction of toxic emissions available.

f Values based on emission rates and reduction targets for Western European signatory states of OSPAR and HELCOM conventions.

where: EcoX = environmental index per functional where: RDI = resource depletion index per functional
unit; Ni = normalization result per functional unit for unit; Ni = normalization result per functional unit for
impact category i; WFi = weighting factor for impact impact category i; WFi = weighting factor for impact
category i. category i.
The aggregation of the normalized indicator values
for the different resource categories (e.g. P rock, fossil
fuels) into the summarizing resource depletion index 4. Discussion and conclusions
‘RDI’ can be performed equivalently (Eq. (6)).
The LCA method described in this paper is based


RDI = Ni × WFi (6) on the general LCA methodology given by ISO (1997)
i and SETAC (Consoli et al., 1993) and adapted to the
 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 261

study of plant nutrition in crop production systems. available characterization approaches (Goedkoop and
Different ready-to-use LCA approaches, primarily Spriensma, 1999; Huijbregts, 2001). In particular the
designed for industrial applications, have been pub- inclusion of possible toxic impacts due to the use
lished (BUWAL, 1998; Goedkoop, 1995; Goedkoop of fungicides would be desirable, because there is a
and Spriensma, 1999; Guinée, 2001; Heijungs, 1992; positive relationship between increased N fertiliza-
Steen, 1999). It is therefore not surprising that there tion and application rate of fungicides (Newe and
are problems concerning their application to arable Volk, 2000). So far, the presented LCA model solely
crop production systems (Brentrup et al., 2001). In considers energy use and energy-related emissions
particular, some important environmental impacts due to the production, packaging, transportation and
are not included (e.g. land use, resource depletion application of pesticides.
in Goedkoop, 1995, nutrient emissions in Goedkoop However, the proposed LCA method enables a com-
and Spriensma, 1999). Some methods do not use prehensive analysis of all other environmental impacts
state-of-the-art aggregation procedures (e.g. for acidi- related to arable farming products. This is currently
fication in Goedkoop, 1995). Others are not transpar- of particular interest because of an increasing public
ent in their weighting procedure (Steen, 1999), do not awareness and interest in the environmental effects of
calculate an overall environmental indicator (Guinée, food production. The life-cycle perspective in LCA
2001; Heijungs, 1992) or may be valid for only one studies of arable farming products like wheat grain, i.e.
country (BUWAL, 1998, for Switzerland). the consideration of sub-processes such as raw mate-
Therefore, in this paper a new LCA approach is rial extraction or fertilizer production together with the
described that has been developed to study the envi- on-farm processes allows the detection of environmen-
ronmental impact of arable crop production. A major tal hot-spots in the total production system. For exam-
advantage of this approach is the integration of all ple, own investigations on the environmental impact of
impact categories relevant to arable crop production. wheat production have shown that the main environ-
New impact assessment procedures, including ag- mental impact of the production system is related to
gregation, normalization and weighting, have been on-field activities (e.g. fertilizer application), whereas
developed for ‘land use’ and ‘resource consumption’ the production and transport of farm inputs has a much
impact categories (Brentrup et al., 2002a,b). For smaller effect (Küsters and Brentrup, 1999). Other en-
the ‘climate change’, ‘toxicity’, ‘acidification’ and vironmental problems, which depend on nitrogen (N)
‘eutrophication’ impact categories, the currently fertilizer management, can be eutrophication (if N ap-
best available aggregation methods have been cho- plication rates exceed the crop demand), acidification
sen and refined to include new normalization values (if urea or ammonium-containing fertilizers are used)
and weighting factors. The ‘depletion of the strato- or climate change. The application of this LCA method
spheric ozone layer’ and ‘formation of tropospheric provides an insight into the contribution of different
photo-oxidants’ impact categories have been shown sub-systems, e.g. the transport, production and appli-
to be unimportant for arable crop production systems cation of farm inputs to the environmental impact and
because usually no (stratospheric ozone depletion) enables the suggestion of measures to improve the
or only negligible emissions (photo-oxidants) are overall environmental performance of arable farming
released from crop production. With regard to the systems. Furthermore, the proposed method can be
‘toxicity’ impact category only heavy metal emissions used to support the choice of alternative products or
to soil resulting from the application of contaminated processes to reduce environmental effects.
organic and inorganic fertilizers are considered. None To this end, the weighting step should be seen as a
of the available characterization approaches for toxic valuable and objective interpretation tool, which pre-
impacts includes a sufficient number of currently rel- vents LCA users from deriving their own subjective
evant pesticides. In particular fungicides are clearly conclusions on the overall environmental preferences
underrepresented in those methods (e.g. in Huijbregts, of different alternatives. However, the main challenge
2001). The presented LCA model could be improved, with weighting is that the evaluation of environmen-
if toxicity potentials for a greater number of rele- tal impacts on humans, ecosystems and resources is
vant pesticides would be developed based on already not only a matter of natural science. Natural science
262 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

is necessary to describe and to quantify the single ef- (acidification, toxicity) result in a common unit for
fects and their impact on the environment during the all impacts (money). Provided that the varying meth-
impact assessment (e.g. the different potential of nu- ods to evaluate the different environmental impacts
trients to contribute to the eutrophication of aquatic lead to equivalent weighting and that the monetary
and terrestrial ecosystems). The challenge of the final values represent more than only the economic aspect
evaluation of the impacts on the different environmen- of the environmental impacts, a monetary weighting
tal compartments (fauna, flora, humans, resources) is approach would be possible.
the integration of natural science with subjective val- In a following publication (Brentrup et al., 2003)
ues and therefore needs consensus of the society. Fur- the proposed LCA method will be used to assess the
thermore, weighting procedures are expected to be environmental effects of different intensities of winter
transparent (Lindeijer, 1996). A set of generic weight- wheat production.
ing factors, based on the distance-to-target principle
with accepted environmental goals as targets, fulfills
these requirements. Environmental goals, such as the Acknowledgements
UNECE emissions reduction targets, to reduce acid-
The authors express appreciation to Dr Paul Seward
ification, eutrophication and photo-oxidant formation
for his critical review, discussion and suggestions in
(UN-ECE/CLRTAP, 1999) are a result of an inten-
preparing the manuscript.
sive debate between science, society’s economic goals,
and policy guidelines and reflect the society’s view on
these environmental problems. References
Other possible methods to derive weighting factors
such as expert panels, proxy approaches or monetary Bach, M., Becker, R., 1995. Regional differenzierte Abschätzung
methods reveal some specific problems. For exam- des Nitrateintrages aus diffusen Quellen in das Grundwasser-
ple, weighting factors based on expert panels (e.g. Untersuchung für die Bundesrepublik (Regionally differen-
tiated estimation of nitrate immissions to groundwater from
Landbank, 1994; Goedkoop and Spriensma, 1999)
non-point sources—investigations in Germany). Institut für
can be influenced by the personal priorities and per- Landeskultur, University of Gießen.
ceptions of the chosen panel members or by the way BfN (Bundesamt für Naturschutz) , 1999. Daten zur Natur 1999
they are interviewed (Landbank, 1994). Furthermore, (Environmental data 1999). Landwirtschaftsverlag GmbH,
the panel members may be not representative of all Münster.
BMWi (Bundesministerium für Wirtschaft), 1995. Energiedaten
social groups (Goedkoop and Spriensma, 1999). An-
’95, Nationale und internationale Entwicklung (Energy data
other option to derive weighting factors are proxy ’95, National and international developments). BMWi, Bonn.
approaches, which use for instance the accumu- Brentrup, F., Küsters, J., Lammel, J., Kuhlmann, H., 2000. Methods
lated energy consumption or the total material input to estimate on-field nitrogen emissions from crop production
(Giegrich et al., 1995) as a representative for the total as an input to LCA studies in the agricultural sector. Int. J.
LCA 5, 349–357.
environmental impact of a product or process under
Brentrup, F., Küsters, J., Kuhlmann, H., Lammel, J., 2001.
investigation. However, inputs (resources) need not Application of the life cycle assessment methodology to
to be representative for outputs (emissions), i.e. the agricultural production: an example of sugar beet production
output of toxic substances does not necessarily need with different forms of nitrogen fertilisers. Eur. J. Agronomy
high inputs. Therefore, this kind of weighting does 14, 221–233.
Brentrup, F., Küsters, J., Lammel, J., Kuhlmann, H., 2002a. Impact
not comply with the goal of LCA, since only part of
assessment of abiotic resources consumption—conceptual
the total environmental impact of a product or process considerations. Int. J. LCA 7, 301–307.
is considered. In monetary methods (e.g. Steen, 1999) Brentrup, F., Küsters, J., Lammel, J., Kuhlmann, H., 2002b. Life
cash values are assigned to environmental impacts by cycle impact assessment of land use based on the Hemeroby
applying for example market prices for resources like concept. Int. J. LCA 7, 339–348.
energy or willingness-to-pay surveys for externalities Brentrup, F., Küsters, J., Lammel, J., Barraclough, P., Kuhlmann,
H., 2003. Environmental impact assessment of agricultural
like decreasing biodiversity. Such real and virtual production systems using the life cycle assessment (LCA)
market prices together with costs for the technical methodology. II. The application to N fertilizer use in winter
avoidance or mitigation of environmental impacts wheat production systems. Eur. J. Agronomy (in press).
 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264 263

BUWAL (Bundesamt für Umwelt, Wald und Landschaft), 1998. Guinée, J., Heijungs, R., Van Oers, L., Van de Meent, D., Vermeire,
Bewertung in Ökobilanzen mit der Methode der ökologischen T., Rikken, M., 1996. LCA impact assessment of toxic releases.
Knappheit. Ökofaktoren 1997 (Evaluation in LCA using the Product Policy 1996/21. Ministry of Housing, Spatial Planning
Eco-scarcity method. Eco-factors 1997). Schriftenreihe Umwelt and Environment, The Hague.
Nr. 297. BUWAL, Bern. Hauschild, M., 2000. Estimating pesticide emissions for LCA
Commission of the European Communities, 1999. Directions of agricultural products. In: Weidema, B.P., Meeusen, M.J.G.
towards sustainable agriculture. COM (1999) 22 final. (Eds.), Agricultural Data for Life Cycle Assessments. Report
Commission of the European Communities, Brussels. 2.00.01, vol. 2. Agricultural Economics Research Institute
Consoli, F., Allen, D, Boustead, I., Fava, J., Franklin, W., Jensen, (LEI), The Haugue.
A.A., de Oude, N., Parrish, R., Perriman, R., Postlethwaite, Hauschild, M., Bastrup-Birk, A., Hertel, O., Schöpp, W.,
D., Quay, B., Séguin, J., Vignon, B., 1993. (Eds.) Guidelines Potting, J., 2000. The Danish LCA-methodology project.
for Life-Cycle Assessment: A ‘Code of Practice’. Society of Technical report chapter on photochemical ozone formation,
Environmental Toxicology and Chemistry (SETAC), Brussels. method development and consensus project. Department
ECETOC (European Chemical Industry Ecology and Toxicology of Manufacturing Engineering and Management, Technical
Centre), 1988. Nitrate and Drinking Water. Technical Report University of Denmark, Lyngby.
No. 27. ECETOC, Brussels. Heijungs, R. (Ed.), 1992. Environmental Life Cycle Assessment
ECETOC (European Chemical Industry Ecology and Toxicology of Products. Guide—October 1992. Centre of Environmental
Centre), 1994. Ammonia Emissions to Air in Western Europe. Science, Leiden University, Leiden, pp. 28–32.
Technical Report No. 62. ECETOC, Brussels. Heijungs, R., Guinée, J.B., Huppes, G., 1997. Impact categories
for natural resources and land use. CML report 138. Section
Eckert, H., Breitschuh, G., Sauerbeck, D., 1999. Kriterien
Substances and Products. Centre of Environmental Science
einer umweltverträglichen Landbewirtschaftung (KUL)—ein
(CML), Leiden University, Leiden.
Verfahren zur ökologischen Bewertung von Landwirtschaft-
sbetrieben (Criteria of Environmentally friendly land use HELCOM (Helsinki Commission), 2001. Evaluation of the
(KUL)—a method for the environmental evaluation of farms). Implementation of the 1988 Ministerial Declaration regarding
Agribiol. Res. 52, 57–76. Nutrient Load Reductions. Working Document. HELCOM,
Helsinki.
EEA (European Environment Agency), 1998. Europe’s Environ-
Houghton, J.T., Jenkins, G.J., Ephraums, J.J. (Eds.), 1993.
ment: The Second Assessment. EEA, Copenhagen.
Climate Change. The IPCC Scientific Assessment. Cambridge
EEA (European Environment Agency), 2001. Environmental
University Press, Cambridge.
Signals 2001. EEA regular indicator report. EEA, Copenhagen.
Huijbregts, M.A.J, 2001. Uncertainty and variability in
Engels, T., 1993. Nitratauswaschung aus Getreide- und Zucker- environmental life-cycle assessment. PhD thesis, University of
rübenflächen bei unterschiedlichem N-Angebot (Nitrate Amsterdam, Amsterdam.
leaching from cereal and sugar beet fields at different levels
Huijbregts, M.A.J., Seppälä, J., 2000. Towards region-specific,
of N supply). PhD thesis, University of Hanover.
European fate factors for airborne nitrogen compounds causing
Finnveden, G., Potting, J., 1999. Eutrophication as an impact aquatic eutrophication. Int. J. LCA 5, 65–67.
category. State of the art and research needs. Int. J. LCA 4, ISO (International Organization for Standardization), 1997. Envi-
311–314. ronmental management—life cycle assessment—principles and
Giegrich, J., Mampel, U., Duscha, M., Zazcyk, R., Osorio- framework. International Standard ISO 14040. ISO, Geneva.
Peters, S., Schmidt, T., 1995. Bilanzbewertung in produkt- ISO (International Organization for Standardization), 1998.
bezogenen Ökobilanzen. Evaluation von Bewertungsmethoden, Environmental management—life cycle assessment—goal and
Perspektiven (Weighting in product-related LCA studies. scope definition and life cycle inventory analysis. International
Evaluation of weighting methods, perspectives). In: Umwelt- Standard ISO 14041. ISO, Geneva.
bundesamt, 1995. UBA-Texte 23/95, Methodik der produkt- ISO (International Organization for Standardization), 2000.
bezogenen Ökobilanzen-Wirkungsbilanz und Bewertung (LCA Environmental management—life cycle assessment—life cycle
methodology—impact assessment and weighting). UBA, impact assessment. International Standard ISO 14042:2000.
Berlin. pp. 1–137. ISO, Geneva.
Goedkoop, M., 1995. NOH report 9523. The Eco-Indicator 95. Jolliet, O., Crettaz, P., 1997. Critical Surface-Time 95. A Life
Final Report. Pré Consultants, Amersfoort. Cycle Impact Assessment Methodology Including Fate and
Goedkoop, M., Spriensma, R., 1999. The Eco-Indicator 99. A Exposure École Polytechnique Fédérale de LausanneLaussane.
damage oriented method of life cycle impact assessment. Klepper, O., Beusen, A.H.W., Meinardi, C.R., 1995. Modelling
Methodology Report. Pré Consultants, Amersfoort. the flow of nitrogen and phosphorus in Europe: from loads to
Guinée, J., 1996. Data for the Normalization Step within Life coastal seas. Water Sci. Technol. 31, 141–145.
Cycle Assessment of Products. CML Paper no. 14 (revised Köllner, T., 2000. Species-pool effect potentials (SPEP) as a
version). CML (Centre of Environmental Science), Leiden. yardstick to evaluate land-use impacts on biodiversity. J.
Guinée (Ed.), J.B., 2001. Life cycle assessment: an operational Cleaner Prod. 8, 293–311.
guide to the ISO standards. Centre of Environmental Science, Kowarik, I., 1999. Natürlichkeit, Naturnähe und Hemerobie als
Leiden University, Leiden. Bewertungskriterien (Naturalness, proximity to nature and
264 F. Brentrup et al. / Europ. J. Agronomy 20 (2004) 247–264

hemeroby as evaluation criteria). In: Konold, W., Böcker, Patyk, A., Reinhardt, G.A., 1997. Düngemittel—Energie- und
R., Hampicke, U. (Eds.), 1999. Handbuch Naturschutz und Stoffstrombilanzen (Fertilizer—Energy and Substance Flow
Landschaftspflege (Handbook for conservation of nature and Analysis) Vieweg VerlagWiesbaden.
countryside). Ecomed-Verlag, Landsberg. Potting, J. (Ed.), Beusen, A.H.W., Øllgaard, H., Hansen, O.C.,
Kretschmer, H., Hoffmann, J., Wenkel, K.O., 1997. Einfluß de Haan, B., Hauschild, M., 2000. The Danish LCA-
der landwirtschaftlichen Flächennutzung auf Artenvielfalt methodology project. Technical report chapter on aquatic
und Artenzusammensetzung (Impact of agricultural land eutrophication, method development and consensus project.
use on diversity and composition of species. In: BML Department of Manufacturing Engineering and Management,
(Bundesministerium für Ernährung, Landwirtschaft und Technical University of Denmark, Lyngby.
Forsten), 1997. Biologische Vielfalt in Ökosystemen—Konflikt Sommer, S.G., 1992. Ammonia volatilization from cattle and pig
zwischen Nutzung und Erhaltung (Biological diversity in slurry during storage and after application in the field. PhD
eco-systems—conflict between use and conservation). BML- thesis, University of Copenhagen.
Schriftenreihe 465, pp. 266–288. Spindler, E.A. (Ed.), 1998. Agrar-Öko-Audit. Praxis und
Küsters, J., Jenssen, T., 1998. Life cycle analysis of different Perspektiven einer umweltorientierten Land- und Forstwirt-
fertilisers. In: Ceuterick, D. (Ed.), International Conference schaft. Springer-Verlag, Berlin.
on Life Cycle Assessment in Agriculture, Agro-Industry and Stanners, D., Bourdeau, P. (Eds.), 1995. Europe’s Environment—
Forestry, Brussels, pp. 131–139. The Dobris Assessment. European Environment Agency
Küsters, J., Brentrup, F., 1999. Life cycle analysis of (EEA), Copenhagen.
different fertilizing strategies in winter wheat production.
Statistisches Bundesamt, 1999. Gesamtrechnung für Bodennutzung
In: IFA (International Fertilizer Industry Association) (Ed.),
und Biodiversität, Abschlußbericht (Inventory of land use
Agricultural Conference. Managing Plant Nutrition Towards
and biodiversity. Final document). Verlag Metzler-Poeschel,
Maximum Resource Efficiency, Barcelona, pp. 42–51.
Stuttgart.
Landbank (Ed.), 1994. The Phosphate Report. Landbank
Steen, B., 1999. A systematic approach to environmental
Environmental Research and Consulting, London.
priority strategies in product development (EPS). Version
Lindeijer, E., 1996. Normalisation and valuation. In: Udo de Haes,
2000—models and data of the default method. CPM report
H.A. (Eds.), Towards a Methodology for Life Cycle Impact
1999:5. Centre for Environmental Assessment of Products
Assessment, SETAC, Brussels.
and Material Systems, Chalmers University of Technology,
Lindeijer, E.W., van Kampen, M., Fraanje, P.J., van Dobben,
Stockholm.
H.F., Nabuurs, G.J., Schouwenberg, E.P.A.G., Prins, A.H.,
Udo de Haes, H.A., Jolliet, O., Finnveden, G., Hauschild,
Dankers, N., Leopold, M.F., 1998. Biodiversity and life support
M., Krewitt, W., Müller-Wenk, R., 1999a. Best available
indicators land use impacts in LCA. Ministry of Transport,
practice regarding impact categories and category indicators
Public Works and Watermanagement, Delft.
in life cycle impact assessment. Part I. Int. J. LCA 4, 66–
Lindfors, L.G., Christiansen, K., Hoffman, L., Virtanen, Y., Juntilla,
74.
V., Leskinen, A., Hanssen, O.-J., Rønning, A., Ekvall, T.,
Finnveden, G., Weidema, B.P., Ersbøll, A.K., Boman, B., Ek, Udo de Haes, H.A., Jolliet, O., Finnveden, G., Hauschild, M.,
M., 1995. LCA-NORDIC Technical Report No. 10 and Special Krewitt, W., Müller-Wenk, R., 1999b. Best available practice
Reports Nos. 1–2. TemaNord 1995:503. Nordic Council of regarding impact categories and category indicators in life
Ministers, Copenhagen. cycle impact assessment, Part II. Int. J. LCA 4, 167–
174.
Müller-Wenk, R., 1996. Political and scientific targets in distance-
to-target valuation methods. In: Braunschweig, A., Förster, R., UN-DSD (United Nations Division for Sustainable Development),
Hofstetter, P., Müller-Wenk, R., Developments in LCA Valua- 2000. Agenda 21. Chapter 14: Promoting sustainable
tion. IWÖ-Diskussionsbeitrag Nr. 32. IWÖ-HSG (Institute of agriculture and rural development. United Nations (UN), New
Economy and Ecology), University St. Gallen, St. Gallen. York.
Müller-Wenk, R., 1998. Land use—the main threat to species. UN-ECE/CLRTAP (United Nations Economic Commission
How to include land use in LCA. IWÖ-Diskussionsbeitrag Nr. of Europe/Convention on Long-Range Transboundary Air
64. Institute for Economy and Ecology, St. Gallen University, Pollution), 1999. The 1999 Gothenburg Protocol to Abate
St. Gallen. Acidification, Eutrophication and Ground-level Ozone. UN-
Murray, C.J.L., 1994. Quantifying the burden of disease: the ECE, Geneva.
technical basis for disability-adjusted life years. Bull. World UN-FCCC (United Nations Framework Convention on Climate
Health Organ. 72, 429–445. Change), 1998. Report of the Conference of the Parties on its
Newe, M., Volk, T., 2000. Teilflächenspezifischer Einsatz von Third Session, held at Kyoto from 1 to 11 December, 1997.
Fungiziden und Wachstumsreglern im Getreide (Site-specific FCCC/CP/1997/7/Add.1. UN-FCCC, Bonn.
application of fungicides and growth regulators in cereals). USGS (US Geological Survey) Minerals Information, 2001.
Mitt. BBA 376, 286–287. Mineral Commodity Summaries. USGS, Reston.
OSPAR (Oslo and Paris Commissions), 1995. Nutrients in the Van Oers, L. (Ed.), 2001. LCA normalization data for the
Convention Area. Overview of Implementation of PARCOM Netherlands 1997/1998, Western Europe 1995 and the world
Recommendation 88/2. OSPAR, London. 1990 and 1995. Leiden University.