The role of life cycle assessment in supporting sustainable agri-food [622492]

The role of life cycle assessment in supporting sustainable agri-food
systems: A review of the challenges
Bruno Notarnicolaa, Serenella Salab,*, Assumpci /C19o Antonc, Sarah J. McLarend,
Erwan Saouterb, Ulf Sonessone
aIonian Department of Law, Economics and Environment, University of Bari Aldo Moro, Via Duomo, 259, 74123 Taranto, Italy
bEuropean Commission, Joint Research Centre, Institute for Environment and Sustainability, Via Enrico Fermi 2749, T.P. 290, 21027 Ispra, VA, Italy
cIRTA, Institute for Food and Agricultural Research and Technology (IRTA), Carretera de Cabrils, km 2, Cabrils, Barcelona 08348, Spain
dIAE, Massey University, Private Bag 11222, Palmerston North 4442, New Zealand
eSP Technical Research Institute of Sweden, Food and Bioscience, Box 5401, SE-402 29 G €oteborg, Sweden
article info
Article history:
Received 3 May 2016Received in revised form10 June 2016
Accepted 11 June 2016
Available online 15 June 2016
Keywords:
Food supply chains
Food LCAFood wasteSustainable production and consumptionAgri-food productsabstract
Life cycle thinking is increasingly seen as a key concept for ensuring a transition towards more sus-
tainable production and consumption patterns. As food production systems and consumption patterns
are among the leading drivers of impacts on the environment, it is important to assess and improve food-related supply chains as much as possible. Over the years, life cycle assessment has been used extensively
to assess agricultural systems and food processing and manufacturing activities, and compare alterna-
tives “from field to fork ”and through to food waste management. Notwithstanding the efforts, several
methodological aspects of life cycle assessment still need further improvement in order to ensureadequate and robust support for decision making in both business and policy development contexts. This
paper discusses the challenges for life cycle assessment arising from the complexity of food systems, and
recommends research priorities for both scienti fic development and improvements in practical imple-
mentation. In summary, the intrinsic variability of food production systems requires dedicated modelling
approaches, including addressing issues related to: the distinction between technosphere and ecosphere;
the most appropriate functional unit; the multi-functionality of biological systems; and the modelling ofthe emissions and how this links with life cycle impact assessment. Also, data availability and inter-
pretation of the results are two issues requiring further attention, including how to account for consumer
behaviour.
©2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).
1. Introduction
Food production addresses one of the most important and basic
human needs and has developed in parallel with the evolution of
humanity to ensure steady provision ( Diamond, 2002 ), safety and
variety of food as well as improved nutritional composition.
Currently, food production responds to a basic need and also to a
plethora of social, cultural and even aesthetic needs and wants.
However, with the requirement to feed seven billion people, this
food production comes with a huge environmental cost ( Tilman
et al., 2001; Garnett, 2011 ). Farming approaches have been
depleting the Earth's resources and contributing signi ficantly togreenhouse gas emissions, to soil fertility and biodiversity loss, to
water scarcity, and to the release of large amounts of nutrients and
other pollutants that affect ecosystem quality ( McMichael et al.,
2007 ). If nothing changes in the way we produce and consume
food, and in light of the need to increase food production by more
than 60% by 2050 ( FAO, 2006; FAO et al, 2015 ), the environmental
impacts associated with food production systems will become even
more severe and will increasingly surpass the planetary
boundaries.
Improving food production and consumption systems is at the
heart of every discourse on sustainable development from both
environmental and socio-economic perspectives. Recent studies
have suggested a research agenda for food sustainability. For
example, Soussana (2014) , who addressed the European context
speci fically, prioritised for the production side: i) the sustainable
intensi fication of European agriculture, ii) the operationalisation of*Corresponding author.
E-mail address: serenella.sala@jrc.ec.europa.eu (S. Sala).
Contents lists available at ScienceDirect
Journal of Cleaner Production
journal homepage: www.elsevier.com/locate/jclepro
http://dx.doi.org/10.1016/j.jclepro.2016.06.071
0959-6526/ ©2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).Journal of Cleaner Production 140 (2017) 399 e409

agriculture within limits for greenhouse gases, energy, biodiversity
and contaminants, and iii) the improvement of resilience to climate
change in agricultural and food systems. For the consumption side,
Soussana prioritised: i) the identi fication of the determinants of a
healthy diet including physical activity, ii) the development of
healthy, high-quality, safe and sustainable foods, and iii) the fight
against diet-related chronic diseases.
Focusing on livestock, Scollan et al. (2011) identi fied priorities
related to balancing the need for increased production of animal
products coupled with a lower environmental footprint and
addressing societal needs in terms of product quality as perceived
by the consumer. Indeed, there are synergies between research
needs and strategies dedicated to improving food quality and
safety, and those dedicated to reducing the environmental impact
of ruminant livestock production. The main research priorities
identi fied were related to the capitalization of knowledge to ensure
optimization of production, application of life cycle based assess-
ment, and use of an adaptation strategy based on selection of
profitable animals under different production systems. Putting this
into practice, the Livestock Environmental Assessment and Per-
formance (LEAP) Partnership of FAO (2015) has developed a multi-
stakeholder partnership for benchmarking and monitoring the life
cycle-based environmental performance of the livestock sector. The
initiative aims to promote adoption of life cycle thinking as a way to
understand and improve the environmental pro file of livestock
production systems.
Several authors have proposed that the coupled assessment of
environmental- and human health-related concerns together
should be the building block of future research activities (e.g Tukker
et al., 2011; Adams and Demmig-Adams, 2013 ). Bridging the con-
servation of natural capital, on the one hand, and human health
issues on the other hand, using a life cycle perspective may lead to
breakthroughs in the sustainability assessment of food systems
(Soussana, 2014 ).
In recent research, dietary shifts have been identi fied as one of
the most powerful ways to increase the sustainability of our food
systems (as in the recent review of Hallstr €om et al., 2015 ; and in
studies focusing on combining environmental and nutritional as-
pects such as R€o€os et al., 2015 ; as well as in studies addressing the
complementary role of technological innovation and demand-side
changes as in Bryngelsson et al., 2016 ), and are usually associated
with reducing meat consumption. However, any dietary shift may
imply burden shifting (from one stage to another in the life cycle of
different foods, or from one impact category to another). Therefore,
there is a strong need for use of life cycle-based methods to ensure
that dietary shifts are coupled with improved sustainability of food
systems.
In view of the above-mentioned research priorities, this paper
describes some of the main research challenges that need to be
addressed in order for Life Cycle Assessment (LCA) to more fully
support decision-making and the transition towards sustainable
food systems. Given the importance of the environmental conse-
quences of food production and consumption, the added value of
this study is related to the identi fication of the needs and the means
for better accounting those, adopting life cycle assessment. Spe-
cifically, section 2further discusses the additional advances needed
in food-related LCA. Section 3focuses on the intrinsic high vari-
ability of food systems and the related challenges for food-related
LCA. Section 4deals with the main modelling issues speci fict o
LCAs of food systems. The importance of further developing reliable
LCA databases is discussed in Section 5. The role of consumers and
of the industry in driving the development of food-related LCA is
discussed in Section 6, while the role of LCA in assessing food waste
is addressed in Section 7. The final Section 8deals with the inter-
pretation of food-related LCA studies, including the necessity ofunderstanding how to best use LCA study results to avoid
misleading conclusions.
2. State of the art and challenges for agri-food LCA in
answering sustainability and food security issues
Increasingly, Life Cycle Thinking (LCT) is recognised as funda-
mental for addressing current challenges and research needs
related to the sustainability of food production and consumption
systems. However, progress towards environmentally sustainable
systems requires improving the methods for quantitative, inte-
grated assessment and promoting the use of these methods in
different domains. Indeed, LCT and the different life cycle-based
methodologies, such as Life Cycle Assessment (LCA) ( ISO,
2006a,b ), Life Cycle Costing (LCC), Social Life Cycle Assessment
(sLCA) and the overall Life Cycle Sustainability Assessment (LCSA)
may support a transition toward increasing the sustainability of
current patterns of production and consumption. Given the
importance of adopting a life cycle approach, literature on appli-
cation of LCA to food system has been thriving ( Notarnicola et al.,
2012a; van den Werf et al., 2014; Notarnicola et al., 2015;Nemecek et al., 2016 ).
Notwithstanding the positive and peculiar features of LCA-
based methodologies compared to other environmental assess-
ment methodologies ( Sala et al., 2013a,b ), a number of food-related
challenges need to be addressed in order to further advance the
currently available approaches and methods. Consider for example,
the comparisons between more or less intensive agriculture, and
organic versus non-organic agriculture; many LCA studies find that
agricultural intensi fication leads to less overall environmental im-
pacts per functional unit ( Sonesson et al., 2016a ;Kulak et al., 2013 )
(but see Chobtang et al., 2016a (this volume ) for an exception to this
generalisation). The rationale is that higher yields per hectare of
land or per animal are bene ficial if the level of used resources does
not increase to the same extent. In the face of increasing pressure
on agricultural land for other purposes such as bioenergy, and
pressure from urbanisation and deserti fication, increased ef ficiency
of land use seems a logical way forward. However, the current LCA
method is incomplete and does not comprehensively assess some
aspects that are critical for long-term sustainable food production
e.g. decreased soil quality and fertility, increased erosion, reduced
ecosystem services due to intensi fication, biodiversity loss. The
challenge here is that the missing aspects are dependent on syn-
ergies between many factors that are not all “captured ”in current
LCA methods. Also, the information needed to describe these as-
pects is often at the landscape level – and landscape attributes are
only partly dependent upon production management at the field
level. LCA studies usually focus on the field level and therefore, by
not acknowledging such emergent aspects, the conclusions from an
LCA study might support less preferred policies and actions from a
sustainability perspective. LCA-based methods that combine boththefield and landscape perspectives are needed in order to capture
environmental impacts at these different scales. For example,
several agricultural measures for mitigating impacts (such as hav-ingfield margins acting as buffer zones) are not captured by current
land use inventories, whereas reporting their presence may be
regarded as a “credit ”in terms of reduction of environmental im-
pacts. Besides, the landscape pattern (e.g. the patchiness index,
Weissteiner et al., 2016 ) is fundamental to understand the level of
threat to biodiversity posed by an agricultural system. This is in-
formation at landscape level which is completely missing in LCA.
Besides, two of the most incomplete modelling challenges in LCAs
of food systems are the signi ficant inconsistencies between emis-
sion inventory modelling and impact assessment of pesticides
(Rosenbaum et al., 2015 ), and assessment of land use changeB. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 400

associated with off-farm inputs to agricultural production systems.
Compared with other economic sectors, food systems are
inherently more variable in the inventory data (e.g. in the same
area, on the same crop, two different active ingredients could be
applied for the same purpose) and in the reliability of impact
assessment (e.g. the impact on biodiversity due to a particular land
use may change dramatically from one ecoregion to another). Yet,
current data available in food LCA databases and life cycle impact
assessment (LCIA) models, are mostly non-spatially and temporally
resolved ( Hauschild et al., 2012 ). This results in severe limitations
when agricultural systems are being evaluated. This is further
exacerbated by the fact that, in a globalised world, consuming a
food product in one particular location may be associated with
environmental impacts occurring in many other countries ( Lenzen
et al., 2012 ).
In addition, other sustainability aspects considered highly
relevant by many consumers, such as working conditions and an-
imal welfare, are largely neglected in LCA. In moving from just the
environmental to inclusion of more socio-economic aspects, so far,
several Social LCAs studies have been conducted, especially in
developing countries (e.g. Feschet et al., 2013 on bananas,
Lemeilleur and Vagneron, 2010 on coffee, and Kruse et al., 2009 on
salmon). A review of social and economic tools combined with LCA
of food products can be respectively found in Settanni et al. (2010)
and Kruse (2010) .
Ultimately, cultural-related aspects are a fundamental compo-
nent of food supply chains and heavily affect patterns of con-
sumption ehence generating another source of variability.
Inclusion of cultural aspects in LCSA has been discussed recently by
Pizzirani et al. (2014) , and addressed in a case study of forestry
products ( Pizzirani et al., 2016 ). Cultural values in fluence the way
we produce and consume food, and also in fluence assessment of
the environmental impacts associated with food systems.
3. Variability in food LCA
The importance of distinguishing between variability and un-
certainty in LCA studies has been highlighted by a number of re-
searchers (e.g. Hauck et al., 2014; Huijbregts, 1998, 2001;
Steinmann et al., 2014 ). Uncertainty may be reduced by addi-
tional research but variability describes actual differences amongst
alternative processes and/or products (and thus cannot be reduced
unless there are changes in the systems under analysis) ( Huijbregts,
1998; Steinmann et al., 2014 ).
For agricultural systems in particular, there is potential for
considerable variability in inventory data between individual
agricultural enterprises. Some of the aspects that underlie this
variability include different management practices, soil types and
climates, seasonality, the life cycle of perennial crops, and distances
(and related transportation modes) between locations of activities
in the life cycle of product systems.
Regarding management practices, farmers may utilise different
practices based on their own preferences and expertise. Sets of
practices may be grouped together and labelled as “organic ”,
“biodynamic ”,“integrated ”,“heated greenhouse ”production.
However, in reality there is usually a continuous spectrum of
practices both within and between these categories. A number of
LCA studies compare these categories of activities for speci fied
products or production systems e.g. dairy systems ( O'Brien et al.,
2012; Thomassen et al., 2008; van der Werf et al., 2009 ); pig pro-
duction systems in France ( Basset-Mens et al., 2006 ), greenhouse
production in Europe ( Torrellas et al., 2012 ), bean cultivation in
Greece ( Abeliotis et al., 2013 ), tomato production in France ( Boulard
et al., 2011 ), wheat production in the USA ( Meisterling et al., 2009 ).
However, relatively few LCA studies have actually focused on thevariability within these categories. Notable studies include Fenollos
et al. (2014) on tigernuts in Spain, Mouron et al. (2006) on Swiss
apple production systems, da Silva et al. (2010) on Brazilian soy-
bean production, and a number of studies on dairy systems (e.g.
Chobtang et al., 2016b ;Thomassen et al., 2009 ).
Beyond human-controlled variability, differences in manage-
ment practices and in yields may be related to soil types and cli-
mates e.g. some soils require regular application of lime to raise the
soil's pH, and dryer climates require use of irrigation systems.
However, in their review of LCA studies of vegetable products,
Perrin et al. (2014) note that many of these studies fail to specify the
representativeness of data with respect to soil and climate
conditions.
Some sources of variability are related to the timescale adopted
for the study. Within a single year, seasonality may contribute to
differences in LCA results for food products. For example, Hospido
et al. (2009) found differences in environmental impacts for
Spanish and English lettuce consumed in the UK ebutFoster et al.
(2014) found relatively small differences in environmental impacts
for Spanish and English raspberries consumed in the UK – at
different times of the year. Between years the environmental im-pacts of a single crop may vary due to differences in yields related to
variable weather conditions. Furthermore, over a period of several
years, perennial crops exhibit a cycle of increasing and then
decreasing yields and this is often not accounted for in LCA studies
(Bessou et al., 2013 ).
Finally, variability may be related to the different transport
distances (and modes) between the locations of agricultural pro-
duction in relation to production of inputs used in agricultural
production (such as fertilisers and compost), and subsequent pro-
cessing, retailing and consumption activities. LCA studies on this
aspect that discuss a range of food products include those by
Michalský and Hooda (2015), Rothwell et al. (2016), Webb et al.
(2013) , and Wiedemann et al. (2015) .
There is also potential for signi ficant variability to arise at other
stages in the life cycle of food products. In particular, this variability
may be related to storage time, packaging and food preparation
(including related wastage). For example, Meneses et al. (2012)
found that the climate change and acidi fication potential in-
dicators were higher for plastic bottles than for aseptic cartons (of
various sizes) for Spanish milk packaging; Keyes et al. (2015) found
that storage activities contributed the majority of the result for four
out of eleven environmental indicators studied in an LCA of Nova
Scotian apple production and delivery to a retailer in Halifax,
Canada. And Schmidt Rivera et al. (2014) compared home- and
ready-prepared meals in the UK, and found that home-prepared
meals had lower impacts for ten out of eleven environmental
indicators.
LCA studies vary hugely with respect to discussing and quanti-
fying these sources of variability. When addressed, the variability
may be represented as a range of values, a speci fic metric to
quantify variability (e.g. Hauck et al., 2014; Steinmann et al., 2014 )
or using statistical analysis (e.g. Mouron et al., 2006 ).
So, it is often unclear as to whether or not it is important to
represent these different sources of variability in LCA study results
(and, in particular, those intended to support decision-making). Theexamples given above show that the choice of system boundaries,
temporal and spatial, and choice of agricultural management
practices as well as activities at other life cycle stages, can make big
differences to the LCA results. Yet, for example, current Environ-
mental Product Declaration (EPD) programmes do not generally
provide detailed requirements or guidelines on representation of
variability in LCA results. Instead, they simply require data to be
“representative ”or be calculated as averages or weighted averages,
and EPDs generally present single values for differentB. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 401

environmental indicator results. This could potentially lead to
misrepresentation of products in the marketplace if comparisons
are made between alternative food products on the basis of what is,
effectively, biased data. It may also lead to overlooking of potential
improvement options if the inferior environmental performance of
individual enterprises in the supply chain of food products is “lost ”
in the calculation of average data for the different life cycle stages of
a food product. The challenge for LCA researchers and practitioners
concerns how to represent relevant variability in LCA study results
without having to collect such a huge range of data that these
studies become infeasible.
4. Modelling issues speci fic to agricultural systems
4.1. Distinction between technosphere and ecosphere in relation to
modelling of environmental impacts
Traditionally, LCA has been used for the assessment of industrial
systems where processes are located in the technosphere and
environmental emissions are assessed in the ecosphere. Following
this approach, when LCA has been applied to agricultural systems,soil has been de fined as part of the technosphere ( Audsley, 1997 )
and is regarded as merely a physical support for plants and a me-
dium for delivery of inputs by farmers. This is actually the vision
that prevailed during the early stage of the green revolution
(1950e1970) where agricultural lands were managed like indus-
trial production sites: soils were regarded as only a physical asset.
As a result, impacts on soil fertility, soil structure, soil hydrology
balance, and soil biodiversity are currently not included in the
majority of food LCA studies despite being essential elements for
ensuring the conservation of the natural capital as well as long-
term security of food supply. A possible solution to overcome this
flaw is to include agricultural soil in the ecosphere or to include the
evaluation of these impacts under the land use impact category.
Mainly for impact categories where the use of fertilizers and
pesticides is particularly relevant (i.e. eutrophication and toxicity),
a further dif ficulty is the de finition of the boundary between the
technosphere and ecosphere. If some fate modelling is included at
inventory analysis and it includes degradation of substances, then
any environmental impacts caused by these substances prior to
degradation are omitted from the analysis. Current eutrophication
or toxicity LCIA models account for emissions rather than the
amounts applied, and because there is no agreement on emission
models to be used at inventory analysis, different modelling ap-
proaches will lead to different results (as shown by Perrin et al.,
2014; Rosenbaum et al., 2015; Van Zelm et al., 2014 ) (see section
4.4).
4.2. De finition of an appropriate functional unit
In studies involving agricultural systems, yield in kg or area used
are the most popular functional units (FU). However, even if these
may be appropriate as a reference flow or unit of analysis at in-
ventory level, this choice does not really represent the true function
of agricultural products. Sticking to mass or area is actually not in
line with usual LCA practices where the performance of a productmust be included in the FU.
A number of researchers have proposed alternatives that
include e.g. the nutritional value of the food in the FU ( Heller et al.,
2013 ). It is also important to be aware that not all the food products
could be considered as nutritional per se as they may address needs
that are beyond the basic ones ( Notarnicola et al., 2016a,b this
volume ) and involve social dimensions like drinking wine, beers or
coffee. So, a more sophisticated way of de fining FU would be to
include the cultural function provided by the hedonistic value offood and drink (e.g. Notarnicola et al., 2003 ). However, de finition of
a cultural function is not straightforward since it is not always
defined objectively, and is thus not always feasible for the de fini-
tion of FUs.
From a farmer perspective, it could be argued that economic
value best represents the main function of farmer activities. Severalauthors ( van der Werf and Salou, 2015; Notarnicola et al., 2015 )
support this approach as a way of including the quality of the
product i.e. using a product's price as a measure of the product's
quality. However, prices are usually determined by external factors
that are not necessarily linked to the quality of the product (e.g. out
of season products).
In addition to feeding humans and other animals, agriculture is
also a provider of social services (recreational) and is regarded as a
custodian of cultural and natural heritage ( Koohafkan and Altieri,
2011 ). It could be argued that these should be re flected in the
respective environmental and social assessments.
4.3. Multi-functional biological systems
Co-production is a common issue in food LCA with economic or
physical allocation being the most commonly approach in food
product studies due to ease of data collection. System expansion
should be preferred in order to be in line with ISO; however, the
method of system expansion is more complex and more
demanding on data collection. Schau and Fet (2008) suggested to
use biological rather than physical causality because most food
production systems include biological processes e.g. reduction of
CH
4outputs by changes in the fodder composition (input). Never-
theless, the high variability of biological processes could also
complicate assessment and comparisons. It is clear that different
allocation methods will provide different results; in this sense, the
PEF initiative and the development of Product Category Rules, as
well as Environmental Product Declaration schemes, could
contribute to de fine a consensus in establishing allocation criteria
for speci fic products.
4.4. Modelling emissions at inventory analysis (fertilizers, pesticides
and machinery)
As mentioned in section 4.1, a clear de finition of the boundaries
between the technosphere and ecosphere is needed at inventory
analysis in order to standardise the modelling at impact assess-
ment. However, the modelling at inventory analysis is further
complicated by a number of other factors peculiar to agricultural
systems. In particular, it is well known that emission flows are
closely related to not only site-speci fic soil and climate conditions
but also to the inputs of pesticides and fertilizers themselves.
Several guidelines for inventory modelling are provided in different
studies and reports, amongst which Nemecek et al. (2014) is an
important reference.
Different models used at inventory analysis provide different
environmental results for pesticides. In the ecoinvent database, for
example, there are no pesticide emissions to surface water leading
to no contribution of those chemicals to the aquatic toxicity impact
category, whatever impact assessment model is used. An agree-ment on a clear de finition of pesticide emission modelling is
necessary and is the objective of a current international effort for
finding a consensus ( Rosenbaum et al., 2015 ).
In relation to emissions from use of fertilizers, it is possible to
distinguish between synthetic and organic fertilizers, and between
emissions to different compartments (air, soil and water). Air
emissions are better de fined thanks to the IPCC ( IPCC, 2006 ), which
provides data for greenhouse gas emissions, and EEA guidelines
(EEA 2013 ) which provide data for a number of other air emissions.B. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 402

But a consensus is still missing on a globally applicable model for
calculating soil and water emissions (i.e leaching, erosion and run-
off) which are more dependent on soil conditions (e.g. pH, clay
content, slope, etc.).
Similar issues arise for emissions from the use of machinery
where fuel consumption is dependent not only on hours of work
but also on aspects such as tractor power, type of operation and soil
conditions ( Hansson and Mattsson, 1999; ASAE, 2003 ).
4.5. Impact categories such as land use, water use, biodiversity,
toxicity, particulate matter
Traditionally, LCIA methods have mostly relied on generic, non-
spatial, and steady state multimedia environmental models that
focus predominately on energy-related impacts. However, in the
agricultural sector, site dependent and closely related environ-
mental aspects, such as natural resources (i.e., water and land) and
ecosystems quality, acquire special relevance ( Ant/C19on et al., 2014 ).
Unlike the so-called global impact categories, such as climate
change and ozone depletion, regional impact categories (e.g. acid-
ification, eutrophication, toxicity) need to have spatially differen-
tiated models because evidence shows that differences in fate and
effect factors such as exposure mechanisms and sensitivity can vary
significantly in different geographical contexts ( Sala et al., 2011;
Ciuffo and Sala, 2013 ).
Although water and land use in agriculture could have major
environmental consequences, most LCA studies represent these
impacts as mere flows expressed in m2or m3and do not assess the
potential environmental damage arising from these uses. The In-
ternational Life Cycle Data System (ILCD) ( EC-JRC, 2011 ) recom-
mended two models to be applied with caution, respectively the
Swiss Ecoscarcity model for water and Mila i Canals et al. (2007a,b )
for land use. Those recommendations are currently under revision
to improve the robustness of the models.
At global level, much research has been undertaken in order to
provide operational, site speci fic and globally-applicable methods.
Under the efforts of the UNEP/SETAC Life Cycle Initiative, a flagship
project is being conducted aiming to provide global guidance and
building consensus on environmental LCIA indicators. In a first
stage, work has been focused on the impacts of climate change,
particulate matter, water use and biodiversity damage due to land
use ( Jolliet et al., 2014 ).
Regarding toxicity impacts, the new version of USEtox
(Rosenbaum et al., 2008 in the current version, Usetox 2.0, 2013 )
provides new subcontinental characterization factors which allow a
more site-speci fic assessment for human and aquatic toxicity. For
some substances, e.g. metals, more detailed geographical factors
would be required. However, terrestrial ecotoxicity characteriza-
tion factors are still missing.
5. Databases in food LCA
In order to effectively implement LCA of any product system, the
inventory data need to be reliable and up-to-date. In the case of
background data used in an LCA study, the speci fic information is
typically extracted from databases. With the increasing interest inthe sustainability of food product systems, databases have evolved
from ones that focussed mostly on industrial processes to ones that
also focus on agri-food systems. Examples of commercial databases
that deal with the food sector are ecoinvent (that also deals with
other non-food systems) and more speci fic ones such as Agrifoot-
print, Food LCA-DK and Agribalyse ( Blonk Consultants, 2014;
Frischknecht et al., 2007; Nielsen et al., 2003; Koch and Salou,
2015 ).
As is often the case in food-related LCA, the datasets in thesedatabases are usually created using data representing speci fic sites
at speci fic times. This means that different databases are not
interchangeable with each other and need to be used with caution
by LCA practitioners. In many cases, the data are presented in a
non-transparent manner that will not allow LCA practitioners to
accurately adapt such data to their speci fic case studies. This can
obviously lead to studies that have ambiguous interpretations and
conclusions that are not comparable to those of other studies. For
non-food inventories, this lack of site-speci ficity is usually not such
a big issue. Different manufacturing sites producing PET are most
likely using the same processes, no matter where the PET is pro-
duced (at least in Europe), and have small variability in the input-
output inventories. In contrast, production of a food item in
different locations of the same region of the same country can make
a huge difference to the inventory data. Therefore, to allow a fair
and meaningful comparison of food production systems, a high
level of geographical speci ficity is needed for agri-food systems.
There is thus a need for speci fic and regionalised databases that are
also well-documented and implemented with flexible data struc-
tures that will allow the user to tailor the data to speci fic case
studies.
6. Role of consumers, governments and of the industry6.1. Role of consumer behaviour and governments towards more
sustainable food
As with any market sector, the consumer can potentially play a
direct role in determining the success of sustainable food products
and of government policies targeted at reducing the environmental
impacts of food production and consumption systems.
In some countries, LCA is becoming a mainstream tool for sup-
porting policy development, such as the case of the EU where LCA is
a fundamental instrument of its Integrated Product Policy ( EC,
2001 ). The implementation of such policies partially addresses
consumers by covering aspects such as the use of Environmental
Product Declarations, Eco-labels, and the development of Green
Public Procurement. However, in general consumers at present are
still not playing an operative role that effectively in fluences the use
of LCA in the food sector. This is due to the average consumer's lack
of knowledge about environmental sustainability issues, and due to
the fact that there are too many labelling systems that in many
cases do not communicate information in a clear and direct
manner. In this context, consumers' associations can have a
fundamental role in promoting the transfer of knowledge to the
consumers and in fluencing their behaviour and food habits, thus
indirectly transferring their feedback back to the supply chain
(Notarnicola et al., 2015 ). It is also essential that the initiatives such
as that of the EU concerning a harmonised and unique LCA based
product footprint (Product Environmental Footprint ePEF; EC,
2013 ) become active in order to effectively and concisely commu-
nicate environmental information about food products to
consumers.
6.2. Role of changes in diet
Many LCA studies (e.g. Mu~noz et al., 2010; Meier and Christen,
2012; Heller et al., 2013 ) have shown that the dietary choices of
the consumer signi ficantly affect the environmental sustainability
of food consumption. Perhaps the most prevalent insight that
emerges is that vegetarian diets seem to generate less environ-
mental burdens compared to animal based ones, and also that the
domestic/use phase is not necessarily negligible in terms of envi-
ronmental impacts ( Foster et al., 2006; Hallstr €om et al., 2015 ).
However, in these assessments the positive aspects of animalB. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 403

production are often not included. These positive aspects, such as
ecosystem services, soil fertility, use of resources otherwise not
available as food, and low levels of pesticide use, are elements
where the LCA approach today is generally weak (see Sections 4
and 7 ). Furthermore, replacing the essential nutrients of animal-
based foods poses nutritional challenges ( Millward and Garnett,
2010 ). As indicated by Heller et al. (2013) and by Smedman et al.
(2010) , environmental assessment of a diet cannot consider only
the daily intake food or its fat, energy or protein content, but must
also comprise other more qualitative aspects of a diet. LCA of di-
etary aspects and health issues must consider more particularised
and inclusive nutrition-based functional units ( Stylianou et al.,
2016 ).
Furthermore, dietary choices and the related consumption
styles of individuals vary greatly from region to region. People in
developed countries, when compared to developing countries, tend
to have diets that are characterised by high consumption of animal-
based unsaturated fats and proteins ( Carlsson-Kanyama et al.,
2003 ). Also, in general, in countries with colder climates, the di-
ets tend to involve high caloric consumption of dairy products and
meat. This highlights the fact that assessment of dietary choicesmust take into account that many different factors arise when food
choices are made including social and cultural ones (see section
4.3).
6.3. Role of the food industry
Thefirst use of LCA in the food industry dates back to 1969,
when the Coca Cola Company used it as a means of evaluating as-
pects concerning the packaging of its products ( Hunt and Franklin,
1996 ). Since then LCA has been widely applied to food packaging,
since packaging has been a subject of public debate and it is an area
where producers can both make and communicate improvements,
such as the Packaging Impact Quick Evaluation Tool (PIQET),
introduced by the Sustainable Packaging Alliance (SPA), and the
Instant LCA Packaging tool ( Intertek, 2015 ). The Tetrapak company
has carried out many LCA studies on their food container products
in order to investigate and improve their environmental sustain-
ability ( Tetrapak, 2016 ). The Nestl /C18e company has included in its
web-based product life cycle management (PLM) software DevEx, a
module called Eco-Design Tool to help the company employees to
assess and develop food products not only during the packaging
phase but during all life cycle stages ( Notarnicola et al., 2012b ).
Other food producers have implementing LCA to guide environ-
mental improvement of the whole life cycle of their food products
as part of their environmental policy e.g. Unilever ( Unilever, 2016 )
and Arla Foods ( Flysj €o and Modin-Edman, 2014 ).
In 2013, the ENVIFOOD Protocol was developed as a food and
drink-speci fic guidance document created by the European Food
Sustainable Consumption and Production Roundtable, a multi
stakeholder initiative co-chaired by the European Commission and
business associations from the food and beverage supply chains.
Speci fically, the Protocol is intended as complementary guidance
for the PEF pilot testing launched by the European Commission(Saouter et al., 2014 ).
The European Feed Manufacturer's Federation (FEFAC) and the
American Feed Industry Association (AFIA) set up a consortium in
2011 with a view to collaborate on environmental footprinting. The
FEFAC and AFIA consortium together with the International Feed
Industry Federation (IFIF), has joined the UN FAO-led Partnership
on benchmarking and monitoring the environmental performance
of livestock supply chains ( FEFAC, 2014 ). Also, the International
Dairy Federation and the UN Food and Agriculture Organization
(FAO, 2010 ) commissioned a study on greenhouse gas emissions
from the dairy sector. A review of other similar initiatives can befound in Notarnicola et al. (2015) . Also, the beverage industry has
developed, via its Beverage Industry Environmental Roundtable
(BIER), protocols for carbon and water footprinting beverage
products (e.g., BIER, 2013 ).
A number of initiatives have also been developed by food re-
tailers regarding the environmental footprint of products on their
shelves. This is the case, for example, for supermarket retailers such
as Casino and Leclerc (France), Migros (Switzerland), and Tesco
(UK) that have evaluated the carbon footprint of their products
(Notarnicola et al., 2012a,b ).
Environmental Product Declarations (e.g. Environdec 2015 ) and
other types of labels entailing the use of LCA, have been extensively
used in industry as a means of communicating transparent and
comparable information about the life-cycle environmental impact
of food products. As already mentioned, such labelling schemes
have not always been successful in communicating environmental
sustainability information in an immediate and transparent
fashion. As we have highlighted before, the development of an EU
LCA based environmental footprint (PEF) is intended as a means to
address these issues and should set an example for the future
development of similar labelling schemes in other regions of theworld.
7. Modelling food waste with LCA
Food waste is a globally critical aspect for sustainable develop-
ment, both from an environmental- and food security perspective
but is also a social issue. About 1.3 billion tons of edible food are
globally wasted along food supply chains, corresponding to one-
third of the food produced for human consumption ( FAO, 2011 ).
This food loss represents a huge ‘avoidable ’environmental burden
and of course a huge concern from a social point of view. However,
the implementation of food waste reduction measures is compli-
cated ( Mourad, 2016; Priefer et al., 2016 ). Each year, nearly 10
million die of hunger and hunger-related diseases ( Nellemann
et al., 2009 ).FAO (2013) has estimated that the environmental
impacts associated with food wastage are about i) 3.3 GtCO
2eq,o f
greenhouse gases (GHG), which makes food wastage the 4th GHG
producer after China, USA and EU, ii) 240 000 m3of irrigation water
wasted, and iii) 1.4 billion hectares cultivated in vain.
Modelling food waste in LCA is common practice because the
reference flow is de fined as a part of the functional unit, and thus
the inventory will include waste generated along the chain (relative
to the reference flow). However, to speci fically assess the impact of
food waste a dedicated effort is needed to highlight this and agreed
modelling guidelines are still missing ( Bernstad Saraiva Schotta and
C/C19anovas, 2015; Corrado et al., 2016 .this issue). Following the stan-
dard procedure as described above will mean that usually the im-
pacts associated with food waste are “hidden ”in the impact
assessment results for different life cycle stages in a supply chain;
for food systems, this is most often the primary production (agri-
culture, fishery). This is nothing that demands new methods but
require a dedicated interpretation of the results, which in turn put
demands on how the LCA model is structured so that the speci fic
impacts of food waste can be extracted. Food waste can be seen as a
symptom of dysfunctional food supply chains, dysfunctional in
both technologically and managerial ways. The latter implies that
there are important non-technological aspects of the solutions to
be developed. When addressing food waste, it is critical that the
measures taken really contribute to real improvements and not just
shift problems around. Thus LCA is an appropriate tool for the
assessment of both the technological and managerial solutions.
However, there is a need to apply LCA in a conscious way. Quan-
tifying the environmental impacts of initiatives to reduce food
waste demands a thorough understanding of the importance ofB. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 404

methodological aspects such as system boundaries, systems
expansion, and the time dimension. It also leads into a discussion of
how to connect LCA to the challenge of food security. Another
strength of LCA is that it can provide a platform for discussion and
mutual learning among supply chain stakeholders since LCA pro-
vides an overarching framework for evaluating initiatives.
Additionally, and in the frame of circular economy, LCA can play
an important role in the evaluation of waste management,
including logistics, by environmentally assessing aspects such as:
– Nutrients recovery used as fertilizers and therefore avoided
fertilizer consumption;
– Water-ef ficiency measures and reuse of treated wastewater; and
– Improved management along the whole production chain in
order to reduce food losses during production and distribution,
in shops, restaurants, catering facilities, and at home.
8. The risk of misusing LCA in the food sectors
8.1. Use of LCA results to support decision-making
The “cradle-to-grave ”perspective and the multi-criteria
approach of LCA makes it a suitable method for supporting deci-
sion making and assessing if alleged eco-innovations are effectively
preferred when considering all life cycle stages and assessing across
different impact categories. However, these multi-phase and multi-
criteria attributes make LCA results very complicated to analyse and
interpret without having a deep understanding of the modelling of
the system studied and the meaning of the impact categories.
There is a clear tendency today to try to include all possible
environmental questions, worries and concerns into LCA with the
aim that this tool will be able to address all environmental issues at
once. In theory, with unlimited time, unlimited resources, with
access to all data instantaneously this could indeed be possible, but
in practice it is not. Today there are huge issues in data availability,
database inter-comparability, modelling approaches, validity and
relevance of some impact assessment methods, in normalisation
and in weighting. There is a real risk that rushing into additional
definition of new impact categories, more exposure compartments,
more complexity, more of everything, will undermine the value and
credibility of LCA. Probably a more ef ficient approach is to apply
appropriate tools in combination with LCA. Examples of such tools
are Environmental impact assessment, Environmental manage-
ment schemes, Human risk assessment, Environmental risk
assessment, Material flow analysis, and Resource ef ficiency
assessment of products. Use of impact categories that most
decision-makers can easily understand and communicate is also
required. Ideally, all impact categories should have the simplicity
and accuracy of the GHG impact category.
The current PEF experience ( EC, 2013 ) has been (and still is)
extremely bene ficial for highlighting, from 27 different sectors who
applied the same method at the same time, what still needs to be
fixed to make LCA a robust decision making tool. There is still a lot
of work ahead, especially in the food sector where LCA was intro-
duced more recently and where the method must be adapted to fit
this particular sector.
In conclusion, to promote LCA as a decision-making tool, both
for industry and for government policy, more robustness, reliability
and representativeness are needed.
8.2. Interpretation of results ehow to avoid misinterpretation
Preventing or minimizing misinterpretation of LCA studies re-
quires, firstly, clarifying the questions the LCA study is addressing.This involves consideration of the system boundaries, the un-
certainties and completeness of the data sources, the allocation
approach, the way impacts are assessed, and the robustness and
completeness of the modelling approaches. However, often LCA
practitioners jump into overly strong conclusions. This can seri-
ously jeopardize the uptake of the, in principle, very sound
approach used in LCA. An area where serious misinterpretation
often occurs is related to the comparison of intensive versus
extensive production, where the impacts associated with each unit
of product are reduced in relative terms (less impacts per unit of
products) while in absolute terms the impacts may be higher (e.g.
more pressure on soil quality, more pressure on aspects not
modelled/included in the evaluation e.g. biodiversity etc.).
Regarding the interpretation of the impact assessment results, this
is another area where often results are misinterpreted. Here we
discuss the case of the toxicity-related impact categories (aquatic
toxicity and human toxicity), discussing the complexities and
related uncertainty within this impact category as an example of
the issue of misinterpretation in LCA.
LCA cannot be used to address questions related to the risk
associated with a product. The current naming of some of the LCAimpact assessment categories suggests exactly the opposite to a
non-expert. ‘Freshwater Ecotoxicity ’and ‘Human toxicity ecancer
enon cancer effect ’reflect the potential impact on ecosystem and
human health, but should not be used to suggest that product A is
actually safer that product B. Toxicity of a chemical must be
assessed by considering concentration in the speci fic receiving
compartment, not just via a quantity as reported in life cycle in-
ventory. No matter how sophisticated the modelling of the impact
assessment is, LCA does not capture essential information such as
the volume of the receiving compartment and time of exposure to
assess the true toxicity of a chemical when released into the
environment. To accommodate the data requirements of LCA (only
calculated as mass), the toxicity impact assessment category is a
‘time-integrated effect per unit mass of chemical release into the
environment ’(Hauschild and Pennigton, 2002 ). This is a pragmatic
and operational approach to deal with toxicity in LCA but it should
never be interpreted as an assessment of the real toxicity impact of
a product. There is just no causality between a mass of a substance
and its toxic effect ( Owens, 1997 ).
Applied to food product LCAs, the aquatic toxicity impact cate-
gories as currently modelled can lead to wrong conclusions when,
for example, comparing conventional and organic farming (see
section 4.4). As discussed earlier, the impact of pesticides on agri-
cultural soil biota, and on plants, butter flies, birds and pollinators
are not included yet, although they are essential for food produc-
tion. Recently, a proposal for integrating new targets species in LCA,
e.g. pollinators, have been made ( Crenna et al. 2016 ).
Overall, to be more meaningful, impact categories in food-
related LCA studies need to re flect the known impacts of agricul-
ture production on the environment (loss of soil fertility, loss of soil
structure, loss of pollinators, etc.). The usual list of impact cate-
gories developed for industrial systems cannot just be reapplied to
food systems without considering the speci ficity of these systems.
A recent review on soil modelling ( Vidal Legaz et al., 2016 this issue )
illustrates that several soil models have been developed focusingspeci fically on agriculture sectors. To the contrary, the challenge for
many models is related to how they can be applied in the LCA
context. In fact, when very comprehensive modelling is proposed
(e.g. the SALCA approach, Oberholzer et al., 2012 ), this is often
associated with a requirement for a huge amount of inventory data
at afield scale, leading to dif ficulties in applying the model to large
systems. Balancing data demand, comprehensiveness and appli-
cability is the way forward.
The LCA concept is easy to understand and very appealing forB. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 405

academic, industry and policy makers because of its holistic nature,
but being able to correctly interpret the results require a high level
of expertise. However, the user-friendliness aspect of LCA tools puts
LCA at the fingertips of almost everyone without the need to un-
derstand all underlying assumptions. As for other disciplines,
running and interpreting an LCA should require a validated diploma
with several years of experience. This is certainly not the case today
although steps have been taken in the right direction through the
development of LCA practitioner certi fication schemes such as that
of the American Life Cycle Association ( ALCA, 2016 ).
9. Conclusions
Ensuring a transition towards more sustainable production and
consumption patterns requires a holistic approach and life cycle
thinking is increasingly seen as a key concept for supporting this
aim ( Sala et al., 2015 ). As food production systems and consump-
tion patterns are among the leading drivers of impacts on the
environment, over time the applications of life cycle thinking and
assessment to food-related supply chains have flourished. Life cycle
assessment has been applied extensively to assessment of agri-cultural systems, and processing and manufacturing activities, and
for comparing alternatives “from field to fork ”and up to food waste
management. However, despite the increasing number of LCA food
studies and a flourishing literature on both methodological aspects
and case studies, several challenges still need to be addressed in
order to ensure that LCA is delivering robust results.
The main challenges highlighted in our analysis are related to
different methodological aspects. Firstly, there is a need to move
beyond the simple rationale that more output per hectare is suf fi-
cient to ensure increasing eco-ef ficiency. In fact, notwithstanding
that increased ef ficiency of land use seems a logical way forward, in
face of the increasing pressure on agricultural land for other pur-
poses such as bioenergy, and pressure from urbanisation and
deserti fication, the current LCA method is incomplete and does not
comprehensively assess some aspects that are critical for long-term
sustainable food production.
The inherent variability of the agricultural system is one
element affecting the assessment at the inventory, impact assess-
ment, and interpretation phases. Compared with other sectors,
food systems are inherently more variable both in the inventory
data and in the possible impacts (e.g. the impact on biodiversity
due to a particular land use may differ dramatically from one
ecoregion to another). How to represent relevant variability in LCA
study results without having to collect such a huge range of data
that these studies become infeasible requires the development of
speci fic methodologies. This also suggests a need for speci fic
guidelines for agricultural inventories, including improving the
quality of the data available in LCA databases. The above mentioned
variability and the geographical speci ficity of food systems calls
into existence the need for clearly structured regionalised datasets
that will allow the LCA practitioner to tailor the data to speci fic case
studies.
Current LCA modelling approaches should be complemented by
other approaches in order to improve the understanding of what is
happening in- field (and potentially subject to speci fic comparisons,
e.g. organic versus non-organic agriculture), and what is off- field
(aka background systems) and which is affected by the reliability
of secondary datasets. For example, for ecotoxicity-related impacts,
it is often observed in LCA studies that the relative share of impacts
associated with the substances applied on field (e.g. pesticides) is
limited compared to the substances used in background systems
that are off- field.
There is a clear need for consensus on more meaningful FUs for
food products, and there is initial work in this field aiming atdeveloping functional units covering the nutritional function of
food ( Sonesson et al., 2016b this issue ). Regarding the comprehen-
siveness of the impacts, there is a need to address the threats to the
environment that are still not properly addressed and modelled in
LCA studies. This implies on the one hand, a need to enlarge and
improve life cycle impact assessment, and on the other hand to find
a way of integrating knowledge coming from other scienti fic do-
mains when modelling within LCA is unfeasible (e.g., integrate
qualitative considerations or warnings related to missing potential
drivers of impacts, for example GMOs). This suggests a need to find
a balance between quantities and qualities as well as exploring
possibilities for implementation of semi-quantitative models in
LCA. The goal should be to have comprehensive and scienti fically
sound measures. This requires simpli fications in order to be
applicable, which in turn puts more pressure on finding ways to
collaborate between disciplines. Even more, this calls for the pro-
vision of clear guidelines for interpretation of results, including
additional guidance by life cycle impact method developers in
clarifying what their models is actually assessing and which are
possible limits and uncertainties in the assessment.
The structure of food systems is very much in fluenced by con-
sumers' choices and behaviours. Understanding this will lead to
better modelling (e.g. the use phase). It will also lead to consider-
ation of the main different aspects that in fluence the choice of a
product, the potential for dietary shifts towards less impacting di-
ets, changes in the perceived environmental quality associated with
different products, the way in which products are consumed and,
even, the amount of wastage associated with food systems. Both
scienti fic and grey literature represent immense sources of
knowledge. However, capitalizing on this knowledge through its
correct use and interpretation are still open challenges within and
beyond LCA. Further research on these challenges will contribute to
making LCA more robust in its role of supporting decision-making
that varies from individual farm decisions up to national and in-
ternational policymaking for more sustainable future food systems.
Acknowledgements
The contribution to this review of the European Commission e
Joint Research Centre was financially supported by the Directorate-
General for the Environment (DG ENV) of the European Commis-
sion in the context of the Administrative Arrangements (No.070201/2015/SI2.705230/SER/ENV.A1): “Indicators and assessment
of the environmental impact of EU consumption ”.
References
Abeliotis, K., Detsis, V., Pappia, C., 2013. Life cycle assessment of bean production in
the Prespa National Park, Greece. J. Clean. Prod. 41, 89 e96.
ACLCA, 2016. Certi fication. Available at: http://www.lcacenter.org/1certi fication.
aspx . accessed 17 03 16.
Adams, R.B., Demmig-Adams, B., 2013. Impact of contrasting food sources on health
versus environment. Nutr. Food Sci. 43 (3), 228 e235.
Ant/C19on, A., Torrellas, M., Nu ~nez, M., Sevign /C19e, E., Amores, M., Mu ~noz, P., Montero, J.,
2014. Improvement of agricultural life cycle assessment studies through spatial
differentiation and new impact categories: case study on greenhouse tomato
production. Environ. Sci. Technol. 48 (16), 9454 e9462 .
ASAE, 2003. ASAE D497.4 FEB03 Agricultural Machinery Management Data.
American Society of Agricultural Engineers .
Audsley, E., 1997. Harmonisation of Environmental Life Cycle Assessment for Agri-
culture. European Commission DG VI Agriculture .
Basset-Mens, C., van der Werf, H.M.G., Durand, P., Leterme, P., 2006. Implications of
uncertainty and variability in the life cycle assessment of pig production sys-
tems. Int. J. LCA 11 (5), 298 e304.
Bernstad Saraiva Schotta, A., C /C19anovas, A., 2015. Current practice, challenges and
potential methodological improvements in environmental evaluations of food
waste prevention ea discussion paper, Resources. Conservation and Recycl. 101,
132e142.
Bessou, C., Basset-Mens, C., Tran, T., Benoist, A., 2013. LCA applied to perennial
cropping systems: a review focused on the farm stage. Int. J. LCA 18, 340 e361.B. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 406

BIER, 2013. Beverage Industry Environmental Roundtable. Beverage Industry Sector
Guidance for Greenhouse Gas Emissions Reporting available from: http://www.
bieroundtable.com/#!energy eclimate/c1y5j . accessed April 2016.
Blonk Consultants, 2014. Agri-footprint Description of Data. V 1.0. Retrieved from:
www.agri-footprint.com/assets/Agri-Footprint-Part2-Descriptionofdata-
Version1.0.pdf .
Boulard, T., Raeppel, C., Brun, R., Lecompte, F., Hayer, F., Carmassi, G., Gaillard, G.,
2011. Environmental impact of greenhouse tomato production in France. Agron.
Sustain Dev. 31 (4), 757 e777.
Bryngelsson, D., Wirsenius, S., Hedenus, F., Sonesson, U., 2016. How can the EU
climate targets be met? A combined analysis of technological and demand-side
changes in food and agriculture. Food Policy 59, 152 e164.
Carlsson-Kanyama, A., Pipping Ekstr €o, M., Shanaha, H., 2003. Food and life cycle
energy in-puts: consequences of diet and ways to increase ef ficiency. Ecol. Econ.
44, 293 e307.
Chobtang, J., Ledgard, S.F., McLaren, S.J., Zonderland-Thomassen, M., Donaghy, D.J.,
2016b. Appraisal of environmental pro files of pasture-based milk production: a
case study of dairy farms in the Waikato region, New Zealand. Int. J. LCA 21 (3),311e325.
Chobtang, J., Ledgard, S.F., McLaren, S.J., Donaghy, D.J., 2016a. Life cycle environ-
mental impacts of high and low intensi fication pasture-based milk production
systems: a case study of the Waikato region, New Zealand. J. Clean. Prod. 140
(part-2), 664 e674.
Ciuffo, B., Sala, S., 2013. Climate-based archetypes for the environmental fate
assessment of chemicals. J. Environ. Manage 129, 435 e443.
Corrado, S., Ardente, F., Sala, S., Saouter, E., 2016. Modelling of food loss within Life
Cycle Assessment: towards a methodological framework. J. Clean. Prod. 140
(part-2), 847 e859. http://dx.doi.org/10.1016/j.jclepro.2016.06.050 .
Crenna, E., Polce, C., Sala, S., Collina, E., 2016. Pollinators in life cycle assessment:
towards a framework for impact assessment. J. Clean. Prod. 140 (part-2),525e536. http://dx.doi.org/10.1016/j.jclepro.2016.02.058 .
da Silva, V.P., van der Werf, H.M., Spies, A., Soares, S.R., 2010. Variability in envi-
ronmental impacts of Brazilian soybean according to crop production and
transport scenarios. J. Environ. Manage 91 (9), 1831 e1839 .
Diamond, J., 2002. Evolution, consequences and future of plant and animal
domestication. Nature 418 (6898), 700 e707.
EC, 2001. Green Paper on Integrated Product Policy COM(2001)68 .
EC, 2013. Commission Recommendation of 9 April 2013 on the use of common
methods to measure and communicate the life cycle environmental perfor-
mance of products and organisations. Off. J. Eur. Communities 56 no. L124, and
related guidance, available at: http://ec.europa.eu/environment/eussd/smgp/
pdf/Guidance_products.pdf (accessed April 2016).
EC-JRC, 2011. International Reference Life Cycle Data System (ILCD) Handbook-
Recommendations for Life Cycle Impact Assessment in the European Context.
Publications Of fice of the European Union .
Environdec, 2015. The International EPD System. http://www.environdec.com/en/
The-International-EPD-System/ .
European Environment Agency (EEA), 2013. EMEP/EEA Air Pollutant Emission In-
ventory Guidebook 2013. Technical Guidance to Prepare National Emission
Inventories. EEA Technical Report No 12/2013. European Environment Agency,
Luxembourg .
FAO, 2006. World Agriculture, towards 2030/2050. FAO, Rome. Available at: http://
www.fao.org/ fileadmin/user_upload/esag/docs/Interim_report_AT2050web.
pdf.
FAO, 2010. Greenhouse Gas Emissions from the Dairy Sector – a Life Cycle Assess-
ment Food and Agriculture Organization of the United Nations. http://www.fao.
org/docrep/012/k7930e/k7930e00.pdf .
FAO, 2011. Global Initiative on Food Loss and Food Waste Reduction. Available at:
http://www.fao.org/save-food/en .
FAO, 2013. Food Wastage Footprint. Impacts on Natural Resources. Summary
Report. Available at: http://www.fao.org/docrep/018/i3347e/i3347e.pdf .
FAO, 2015. Livestock Environmental Assessment and Performance (LEAP) Partner-
ship available at: http://www.fao.org/partnerships/leap/en/ . accessed February
2016.
FAO, IFAD, WFP, 2015. The state of food insecurity in the world 2015. In: Meeting the
2015 International Hunger Targets: Taking Stock of Uneven Progress. Rome,FAO.
FEFAC, 2014. The Feed Chain in Action. Animal Nutrition ethe Key to Animal
Performance, Health and Welfare. http://www.fefac.eu/ files/54738.pdf .
Fenollos, M.L., Ribal, J., Lid /C19on, A., Bautista, I., Juraske, R., Clemente, G., Sanjuan, N.,
2014. In fluence of management practices on economic and environmental
performance of crops. A case study in Spanish Horticulture. Agroecol. and
Sustain. Food Syst. 38 (6), 635 e659.
Feschet, P., Macombe, C., Garrab /C19e, M., Loeillet, D., Saez, A.R., Benhmad, F., 2013.
Social impact assessment in LCA using the Preston pathway. Int. J. LCA 18 (2),
490e503.
Flysj €o, A., Modin-Edman, A.-K., 2014. How to use LCA in a company context ethe
case of a dairy cooperative. In: Proceedings of the 9th International Conference
on Life Cycle Assessment in the Agri-food Sector, San Francisco USA .
Foster, C., Green, K., Bleda, M., Dewick, P., Evans, B., Flynn, A., Mylan, J., 2006.
Environmental Impacts of Food Production and Consumption: a Report to the
Department for Environment, Food and Rural Affairs. Manchester Business
School, Defra, London .
Foster, C., Gu /C19eben, C., Holmes, M., Wiltshire, J., Wynn, S., 2014. The environmental
effects of seasonal food purchase: a raspberry case study. J. Clean. Prod. 73,269e274.
Frischknecht, R., Jungbluth, N., Althaus, H.-J., Doka, G., Dones, R., Heck, T.,
Hellweg, S., Hischier, R., Nemecek, T., Rebitzer, G., Spielmann, M., 2007. Over-
view and Methodology. Ecoinvent Report 1 v2.0. Swiss Centre for Life Cycle
Inventories, Dübendorf, CH retrieved from: www.ecoinvent.org .
Garnett, T., 2011. Where are the best opportunities for reducing greenhouse gas
emissions in the food system (including the food chain)? Food Policy 36,
S23eS32.
Hallstr €om, E., C-Kanyama, A., B €orjesson, P., 2015. Environmental impact of dietary
change: a systematic review. J. Clean. Prod. 91, 1 e11.
Hansson, P., Mattsson, B., 1999. In fluence of derived operation-speci
fic tractor
emission data on results from an LCI on wheat production. Int. J. LCA 4 (4),
202e206.
Hauck, M., Steinmann, Z.J.N., Laurenzi, I.J., Karuppiah, R., Huijbregts, M.A.J., 2014.
How to quantify uncertainty and variability in life cycle assessment: the case of
greenhouse gas emissions of gas power generation in the US. Environ. Res. Lett.
9, 1e7.
Hauschild, M., Pennigton, D., 2002. Indicators for ecotoxicity in life cycle impact
assessment. In: Udo de Haes, H.A., et al (Eds.), Life Cycle Impact Assessment:
Striving towards Best Practice. SETAC Press Proceedings .
Hauschild, M., Goedkoop, M., Guin /C19ee, J., Heijungs, R., Huijbregts, M., Jolliet, O.,
Margni, M., De Schryver, A., Humbert, S., Laurent, A., Sala, S., Pant, R., 2012. Best
existing practice for characterization modelling in life cycle impact assessment.Int. J. LCA 18 (3), 683 e697.
Heller, M.C., Keoleia, G.A., Willett, W.C., 2013. Toward a life cycle-based, diet-level
framework for food environmental impact and nutritional quality assessment: a
critical review. Environ. Sci. Technol. 47 (22), 12632 e12647 .
Hospido, A., Mila i Canals, L., McLaren, S.J., Clift, R., Truninger, M., Edwards-Jones, G.,
2009. The role of seasonality in lettuce production: a case study of environ-
mental and social aspects. Int. J. LCA 14 (5), 381 e391.
Huijbregts, M.A.J., 1998. Application of uncertainty and variability in LCA. Part I: a
general framework for the analysis of uncertainty and variability in Life Cycle
Assessment. Int. J. LCA 3 (5), 273 e280.
Huijbregts, M.A.J., 2001. Uncertainty and Variability in Environmental Life Cycle
Assessment. PhD thesis. University of Amsterdam, Amsterdam .
Hunt, R.G., Franklin, W.E., 1996. LCA How it came about. Int. J. LCA 1 (1), 4 e7.
Intertek, 2015. Instant LCA Packaging. http://www.intertek.com/consumer/
sustainability/instant-life-cycle-assessment-software/ .
IPCC, 2006. Guidelines for National Greenhouse Gas Inventories. In: Agriculture,
Forestry and Other Land Use, Volume 4. IGES, Kanagawa, Japan .
ISO, 2006a. ISO 14040:2006 eEnvironmental Management eLife Cycle Assessment
eprinciples and Framework. International Standards Organization, Geneva .
ISO, 2006b. ISO 14044:2006 -environmental Management dlife Cycle Assessment d
Requirements and Guidelines (Geneva, Switzerland) .
Jolliet, O., Frischknecht, R., Bare, J., Boulay, A., Bulle, C., Fantke, P., Gheewala, S.,
Hauschild, M., Itsubo, N., Margni, M., McKone, T., Mila I Canals, L., Postuma, L.,
Prado-Lopez, V., Ridoutt, B., Sonnemann, G., Rosenbaum, R., Seager, T., Struijs, J.,van Zelm, R., Vigon, B., Weisbrod, A., 2014. Global guidance on environmental
life cycle impact assessment indicators: findings of the scoping phase. Int. J. LCA
19, 962 e967.
Keyes, S., Tyedmers, P., Beazley, K., 2015. Evaluating the environmental impacts of
conventional and organic apple production in Nova Scotia, Canada, through life
cycle assessment. J. Clean. Prod. 104, 40 e51.
Koch, P., Salou, T., 2015. Agrybalise 1.2-Rapport Methodologique. Agence de l ’En-
vironnement et de la Maîtrise de l'Energie, France. http://www.ademe.fr/sites/
default/ files/assets/documents/agribalyse-rapport-methodologique-v1_2.pdf .
Koohafkan, P., Altieri, M.A., 2011. Globally Important Agricultural Heritage Systems
(GIAHS). A Legacy for the Future Food and Agriculture Organization of the
United Nations. www.fao.org/nr/giahs/en
.
Kruse, S.A., 2010. Inclusion of social aspects in Life Cycle Assessment of food. In:
Sonesson, U., Berlin, J., Ziegler, F. (Eds.), Environmental Assessment and Man-
agement in the Food Industry. Woodhead Publishing, pp. 219 e233.
Kruse, S.A., Flysj €o, A., Kasperczyk, N., Scholz, A.J., 2009. Socioeconomic indicators as
a complement to life cycle assessment dan application to salmon production
systems. Int. J. LCA 14 (1), 8 e18.
Kulak, M., Nemecek, T., Frossard, E., Gaillard, G., 2013. How eco-ef ficient are low-
input cropping systems in Western Europe, and what can Be done to improve
their eco-ef ficiency? Sustainability 5 (9), 3722 e3743 .
Lemeilleur, S., Vagneron, I., 2010. Looking for the (missing) indicators of social
sustainability eEvidence from sustainability standards in the coffee sectors. In:
Proceeding of the 4 /C18emes journ /C19ees INRA-SFER-CIRAD de recherchesen sciences
sociales, Rennes (FRA), p. 1 –12. Available at: http://www1.montpellier.inra.fr/
bartoli/moisa/bartoli/download/moisa2010_pdf/WP_8_2010.pdf .
Lenzen, M., Moran, D., Kanemoto, K., Foran, B., Lobefaro, L., Geschke, A., 2012. In-
ternational trade drives biodiversity threats in developing nations. Nature 486
(7401), 109 e112.
McMichael, A.J., Powles, J.W., Butler, C.D., Uauy, R., 2007. Food, livestock production,
energy, climate change, and health. lancet 370 (9594), 1253 e1263 .
Meier, T., Christen, O., 2012. Gender as a factor in an environmental assessment of
the consumption of animal and plant -based foods in Germany. Int. J. LCA 17 (5),
550e564.
Meisterling, K., Samaras, C., Schweizer, V., 2009. Decisions to reduce greenhouse
gases from agriculture and product transport: LCA case study of organic andconventional wheat. J. Clean. Prod. 17 (2), 222 e230.
Meneses, M., Pasqualino, J., Castells, F., 2012. Environmental assessment of the milkB. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 407

life cycle: the effect of packaging selection and the variability of milk produc-
tion data. J. Environ. Manage 107, 76 e83.
Michalský, M., Hooda, P.S., 2015. Greenhouse gas emissions of imported and locally
produced fruit and vegetable commodities: a quantitative assessment. Environ.
Sci. Policy 48, 32 e43.
Mil/C18a i Canals, L., Cowell, S.J., Sim, S., Basson, L., 2007a. Comparing domestic versus
imported apples: a focus on energy use. Environ. Sci. Pollut. R. 14 (5), 338 e344.
Mil/C18a i Canals, L., Romany /C18a, J., Cowell, S.J., 2007b. Method for assessing impacts on
life support functions (LSF) related to the use of ‘fertile land ’in life cycle
assessment (LCA). J. Clean. Prod. 15, 1426 e1440 .
Millward, D.J., Garnett, T., 2010. Plenary Lecture 3: food and the planet: nutritional
dilemmas of greenhouse gas emission reductions through reduced intakes of
meat and dairy foods. Proc. Nutr. Soc. 69 (1), 103 e118.
Mourad, M., 2016. Recycling, recovering and preventing “food waste ”: competing
solutions for food systems sustainability in the United States and France.
J. Clean. Prod. 126, 461 e477.
Mouron, P., Nemecek, T., Scholz, R.W., Weber, O., 2006. Management in fluence on
environmental impacts in an apple production system on Swiss fruit farms:combining life cycle assessment with statistical risk assessment. Agric. Ecosyst.
Environ. 114 (2 e4), 311e322.
Mu~noz, I., Mil /C18a i Canals, L., Fern /C19andez-Alba, A.R., 2010. Life cycle assessment of the
average Spanish diet including human excretion. Int. J. LCA 15, 794 e805.
Nellemann, C., MacDevette, M., Manders, T., Eickhout, B., Svihus, B., Prins, A.G.,
Kaltenborn, B.P. (Eds.), 2009. The Environmental Food Crisis ethe Environ-
ment's Role in Averting Future Food Crises. A UNEP Rapid Response Assess-
ment. United Nations Environment Programme, GRID-arendal www.grida.no .
Nemecek, T., Schnetzer, J., Reinhard, J., 2014. Updated and harmonised greenhouse
gas emissions for crop inventories. Int. J. LCA 1 e18.http://dx.doi.org/10.1007/
s11367 .
Nemecek, T., Jungbluth, N., Mila i Canals, L., Schenck, R., 2016. Environmental im-
pacts of food consumption and nutrition: where are we and what is next? Int. J.
LCA 21, 607 e620.
Nielsen, P.H., Nielsen, A.M., Weidema, B.P., Dalgaard, R., Halberg, N., 2003. LCA Food
Database. www.lcafood.dk .
Notarnicola, B., Tassielli, G., Nicoletti, G., 2003. LCA of wine production. In:
Mattsson, B., Sonesson, U. (Eds.), Environmentally-friendly Food Production.
Woodhead-Publishing, Cambridge .
Notarnicola, B., Curran, M.A., Hayashi, K., Huisingh, D., 2012a. Progress in working
towards a more sustainable agri-food industry. J. Clean. Prod. 28, 1 e8
.
Notarnicola, B., Tassielli, G., Renzulli, P.A., 2012b. Modeling the agri-food industry
with life cycle assessment. In: Curran M.a. Life Cycle Assessment Handbook.
Wiley, , New York, pp. 159 e184.
Notarnicola, B., Tassielli, G., Renzulli, P.A., Lo Giudice, A., 2015. Life Cycle Assessment
in the agri-food sector: an overview of its key aspects, international initiatives,
certification, labelling schemes and methodological issues. In: Notarnicola, B.,
Salomone, R., Petti, L., Renzulli, P.A., Roma, R., Cerutti, A.K. (Eds.), Life Cycle
Assessment in the Agri-food Sector. Springer International Publishing,Switzerland .
Notarnicola, B., Tassielli, G., Renzulli, P.A., Monforti, F., 2016a. Energy Flows and
Greenhouses Gases of EU national breads using an LCA approach. J. Clean. Prod.
140 (part-2), 455 e469. http://dx.doi.org/10.1016/j.jclepro.2016.05.150 .
Notarnicola, B., Tassielli, G., Renzulli, P.A., Castellani, V., Sala, S., 2016b. Environ-
mental impact of food consumption in Europe. J. Clean. Prod. 140 (part-2),
753e765. http://dx.doi.org/10.1016/j.jclepro.2016.06.080 .
O'Brien, D., Shalloo, L., Patton, J., Buckley, F., Grainger, C., Wallace, M., 2012. A life
cycle assessment of seasonal grass-based and con finement dairy farms. Agr
Syst. 107, 33 e46.
Oberholzer, H.-R., Freiermuth Knuchel, R., Weisskopf, P., Gaillard, G., 2012. A novel
method for soil quality in life cycle assessment using several soil indicators.Agron. Sustain. Dev. 32, 639 e649.
Owens, J.W., 1997. Life-cycle assessment in relation to risk assessment: an evolving
perspective. Risk Anal. 17 (3), 359 e365.
Perrin, A., Bausset-Mens, C., Benoît Gabrielle, B., 2014. Life cycle assessment of
vegetable products: a review focusing on cropping systems diversity and the
estimation of field emissions. Int. J. LCA 19, 1247 e1263 .
Pizzirani, S., McLaren, S.J., Seadon, J.K., 2014. Is there a place for culture in life cycle
sustainability assessment? Int. J. LCA 19 (6), 1316 e1330 .
Pizzirani, S., McLaren, S.J., Seadon, J.K., Forster, M.E., Pohatu, P.,
TangiWhaioraPorou, T., Warmenhoven, T., 2016. The distinctive recognition of
culture within LCSA: realising the quadruple bottom line. Int. J. LCA (Under
review) .
Priefer, C., J €orissen, J., Br €autigam, K.R., 2016. Food waste prevention in Europe ea
cause-driven approach to identify the most relevant leverage points for action,
Resources. Conservation and Recycl. 109, 155 e165.
Rivera, X.C.Schmidt, Espinoza Orias, N., Azapagic, A., 2014. Life cycle environmental
impacts of convenience food: comparison of ready and home-made meals.J. Clean. Prod. 73, 294 e309.
R€o€os, E., Karlsson, H., Witth €oft, C., Sundberg, C., 2015. Evaluating the sustainability
of diets dCombining environmental and nutritional aspects. Environ. Sci. Policy
47, 157e166.
Rosenbaum, R.K., Bachmann, T.M., Gold, L.S., Huijbregts, M.A.J., Jolliet, O., Juraske, R.,
K€ohler, A., Larsen, H.F., MacLeod, M., Margni, M., McKone, T.E., Payet, J.,
Schuhmacher, M., van de Meent, D., Hauschild, M.Z., 2008. USEtox
dthe UNEP-
SETAC toxicity model: recommended characterization factors for human
toxicity and freshwater ecotoxicity in life cycle impact assessment. Int. J. LCA 13(7), 532 e546.
Rosenbaum, R.K., Anton, A., Bengoa, X., Bjorn, A., Brain, R., Bulle, C., Cosme, N.,
Djikman, T.J., Fantke, P., Felix, M., Geoghegan, T.S., Gottesbueren, B., Hammer, C.,
Humbert, S., Jolliet, O., Juraske, R., Lewis, F., Maxime, D., Nemecek, T., Payet, J.,
Raesaenen, K., Roux, P., Schau, E.M., Sourisseau, S., van Zelm, R., von Streit, B.,
Wallman, M., 2015. The Glasgow consensus on the delineation between pesti-cide emission inventory and impact assessment for LCA. Int. J. LCA 20 (6),
765e776.
Rothwell, A., Ridoutt, B., Page, G., Bellotti, W., 2016. Environmental performance of
local food: trade-offs and implications for climate resilience in a developed city.
J. Clean. Prod. 114, 420 e430.
Sala, S., Marinov, D., Pennington, D., 2011. Spatial differentiation of chemical
removal rates from air in Life Cycle Impact Assessment. Int. J. LCA 16 (8),748e760.
Sala, S., Farioli, F., Zamagni, A., 2013a. Progress in sustainability science: lessons
learnt from current methodologies for sustainability assessment (Part I). Int. J.
LCA 18, 1653 e1672 .
Sala, S., Farioli, F., Zamagni, A., 2013b. Life cycle sustainability assessment in the
context of sustainability science progress (Part II). Int. J. LCA 18, 1686 e1697 .
Sala, S., Ciuffo, B., Nijkamp, P., 2015. A systemic framework for sustainability
assessment. Ecol. Econ. 119, 314 e325.
Saouter, E., Bauer, C., Blomsma, C., De Camillis, C., Lopez, P., Lundquist, L.,
Papagrigoraki, A., Pennington, D., Martin, N., Schenker, U., Vessia, Ø., 2014.Moving from the ENVIFOOD protocol to harmonized product category Rules
and reference data: current and future challenges of the European food sus-
tainable consumption and production round table. In: 9th International Con-
ference LCA of Food San Francisco. Available at: http://lcafood2014.org/
abstracts/abstract-list.htm .
Schau, E., Fet, A., 2008. Lca studies of food products as background for environ-
mental product Declarations. Int. J. LCA 13 (3), 255 e264.
Scollan, Nigel D., Greenwood, P.L., Newbold, C.J., Y /C19a~nez Ruiz, D.R., Shing field, K.J.,
Wallace, R.J., Hocquette, J.F., 2011. Future research priorities for animal pro-
duction in a changing world. Animal Prod. Sci. 51 (1), 1 e5.
Settanni, E., Tassielli, G., Notarnicola, B., 2010. Combining LCA of food products with
economic tools. In: Sonesson, U., Berlin, J., Ziegler, F. (Eds.), EnvironmentalAssessment and Management in the Food Industry. Woodhead publishing,
pp. 207 e218.
Smedman, A., Lindmark-Månsson, H., Drewnowski, A., Modin Edman, A.K., 2010.
Nutrient density of beverages in relation to climate impact. Food Nutr. Res. 54,
5170 .
Sonesson, U., Lorentzon, K., Andersson, A., Barr, U.-K., Bertilsson, J., Borch, E.,
Brunius, C., Emanuelsson, M., G €oransson, L., Gunnarsson, S., Hamberg, L.,
Hessle, A., Kumm, K.-I., Lundh, Å., Nielsen, T.,€Ostergren, K., Salomon, E.,
Sindh €oj, E., Stenberg, Bo, Stenberg, M., Sundberg, M., Wall, H., 2016a. Paths to a
sustainable food sector: integrated design and LCA of future food supply chains,
the case of pork production in Sweden. Int. J. LCA 21, 664 e676.
Sonesson, U., Davis, J., Flysj €o, A., Gustavsson, J., Witth €oft, C., 2016b. Protein quality as
functional unit d
A methodological framework for inclusion in life cycle
assessment of food. J. Clean. Prod. http://dx.doi.org/10.1016/j.jclepro.2016.06.
115(in this issue).
Soussana, J.F., 2014. Research priorities for sustainable agri-food systems and life
cycle assessment. J. Clean. Prod. 73, 19 e23.
Steinmann, Z.J.N., Hauck, M., Karuppiah, R., Laurenzi, I.J., Huijbregts, M.A.J., 2014.
A methodology for separating uncertainty and variability in the life cyclegreenhouse gas emissions of coal-fueled power generation in the USA. Int. J.
LCA 19 (5), 1146 e1155 .
Stylianou, K.S., Heller, M.C., Fulgoni, V.L., Ernstoff, A.S., Keoleian, G.A., Jolliet, O.,
2016. A life cycle assessment framework combining nutritional and environ-
mental health impacts of diet: a case study on milk. Int. J. LCA 21 (5), 734 e746.
Tetrapak, 2016. Lifecycle Assessment of Packaging. The Tetrapak company. http://
www.tetrapak.com/it/sustainability/managing-our-impact/climate-impact/life-
cycle-assessments .
Thomassen, M.A., van Calker, K.J., Smits, M.C.J., Iepema, G.L., de Boer, I.J.M., 2008.
Life cycle assessment of conventional and organic milk production in the
Netherlands. Agr Syst. 96 (1 e3), 95e107.
Thomassen, M.A., Dolman, M.A., van Calker, K.J., de Boer, I.J.M., 2009. Relating life
cycle assessment indicators to gross value added for Dutch dairy farms. Ecol.
Econ. 68 (8 e9), 2278 e2284 .
Tilman, D., Fargione, J., Wolff, B., D'Antonio, C., Dobson, A., Howarth, R.,
Swackhamer, D., 2001. Forecasting agriculturally driven global environmental
change. Science 292 (5515), 281 e284.
Torrellas, M., An /C19otn, A., Ruijs, M., García Victoria, N., Stanghellini, C., Montero, J.,
2012. Environmental and economic assessmentof protected crops in four Eu-
ropean scenarios. J. Clean. Prod. 28, 45 e55.
Tukker, A., Goldbohm, R.A., De Koning, A., Verheijden, M., Kleijn, R., Wolf, O., ,
et al.Rueda-Cantuche, J.M., 2011. Environmental impacts of changes to healthierdiets in Europe. Ecol. Econ. 70 (10), 1776 e1788 .
Unilever, 2016. Sustainable Living. The Unilever Global Company. https://www.
unilever.com/sustainable-living/ .
USEtox, 2013. Scienti fic Consensus Model. Chemical Toxicity Assessment. Available
at:http://www.usetox.org/model/documentation .
van der Werf, H.M.G., Salou, T., 2015. Economic value as a functional unit for
environmental labelling of food and other consumer products. J. Clean. Prod. 94,394e397.
van der Werf, H.M.G., Kanyarushoki, C., Corson, M.S., 2009. An operational methodB. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 408

for the evaluation of resource use and environmental impacts of dairy farms by
life cycle assessment. J. Environ. Manage 90, 3643 e3652 .
van der Werf, H.M.G., Garnett, T., Corson, M.S., Hayashi, K., Huisingh, D.,
Cedeberg, C., 2014. Towards eco-ef ficient agriculture and food systems: theory,
praxis and future challenges. J. Clean. Prod. 73, 1 e9.
Van Zelm, R., Larrey-Lassalle, P., Roux, P., 2014. Bridging the gap between life cycle
inventory and impact assessment for toxicological assessments of pesticides
used in crop production. Chemosphere 100, 175 e218.
Vidal Legaz, B., De Souza, D.M., Teixeira, R., Anton, A., Putman, B., Sala, S., 2016. Soil
quality, properties, and functions in Life Cycle Assessment: an evaluation of
models. J. Clean. Prod. 140 (part-2), 502 e515.Webb, J., Williams, A.G., Hope, E., Evans, D., Moorhouse, E., 2013. Do foods imported
into the UK have a greater environmental impact than the same foods produced
within the UK? Int J LCA 18 (7), 1325 e1343 .
Weissteiner, C.J., García-Feced, C., Paracchini, M.L., 2016. A new view on EU agri-
cultural landscapes: quantifying patchiness to assess farmland heterogeneity.
Ecol. Ind. 61, 317 e327.
Wiedemann, S., McGahan, E., Murphy, C., Yan, M.-J., Henry, B., Thoma, G.,
Ledgard, S., 2015. Environmental impacts and resource use of Australian beef
and lamb exported to the USA determined using life cycle assessment. J. Clean.
Prod. 94, 67 e75.B. Notarnicola et al. / Journal of Cleaner Production 140 (2017) 399 e409 409